Neonatal Exposure to Bisphenol A Alters Rat Uterine Implantation ...

REPRODUCTION-DEVELOPMENT

Neonatal Exposure to Bisphenol A Alters Rat Uterine Implantation-Associated Gene Expression and Reduces the Number of Implantation Sites Jorgelina Varayoud, Jorge G. Ramos, Verońica L. Bosquiazzo, Melina Lower, Mońica Munõz-de-Toro, and Enrique H. Luque Laboratorio de Endocrinología y Tumores Hormonodependientes, School of Biochemistry and Biological Sciences, Universidad Nacional del Litoral, 3000 Santa Fe, Argentina

Endocrine disrupters have been associated with reproductive pathologies such as infertility and gynecological tumors. Using a rat model of early postnatal exposure to bisphenol A (BPA), we evaluated the long-term effects on 1) female reproductive performance, 2) uterine homeobox A10 (Hoxa10) and Hoxa10-target gene expression, and 3) ovarian steroid levels and uterine estrogen receptor ␣ and progesterone (P) receptor expression. Newborn female rats received vehicle, BPA.05 (0.05 mg/kg 䡠 d), BPA20 (20 mg/kg 䡠 d), diethylstilbestrol.2 (0.2 ␮g/kg 䡠 d), or diethylstilbestrol 20 (20 ␮g/kg 䡠 d) on postnatal d 1, 3, 5, and 7. A significant decrease in the number of implantation sites was assessed in the xenoestrogen-exposed females. To address the molecular effects of postnatal xenoestrogen exposure on the pregnant uterus, we evaluated the expression of implantationassociated genes on d 5 of pregnancy (preimplantation uterus). All xenoestrogen-treated rats showed a lower expression of Hoxa10. In the same animals, two Hoxa10-downstream genes were misregulated in the uterus. ␤3 Integrin, which is up-regulated by Hoxa10 in controls, was decreased, whereas empty spiracles homolog 2, which is down-regulated by Hoxa10, was increased. Furthermore a clear down-regulation of estrogen receptor ␣ and P receptor expression was detected without changes in estradiol and P serum levels. The early exposure to BPA produced a lower number of implantation sites in association with a defective uterine environment during the preimplantation period. Alterations in the endocrine-regulated Hoxa10 gene pathways (steroid receptors—Hoxa10—␤3 integrin/empty spiracles homolog 2) could explain, at least in part, the BPA effects on the implantation process. (Endocrinology 152: 1101–1111, 2011)

T

he prototypical endocrine disrupter, bisphenol A (BPA), is an estrogenic chemical produced in large quantities for use primarily in the production of polycarbonate and epoxy resins. In 2004, the estimated production volume of BPA in the United States was approximately 2.3 billion pounds. Of the 1.9 billion pounds of BPA used in 2003, nearly three quarters was used to manufacture various consumer products including polycarbonate containers for storage of foods and beverages (1). To date, numerous BPA effects have been described on endocrine-related organs in different animal models, and

significant levels of BPA have been detected in human samples from different countries (1, 2). We have previously found that exposure to environmentally relevant doses of BPA during the early postnatal period alters the expression of homeobox A10 (Hoxa10) in the rat uterus (3). BPA effect observed during prepubertal period remained in the adulthood, and some Hoxa10-downstream events were fully disrupted (3). Taking into account this and previous results, we propose that an early alteration in Hoxa10 gene expression might affect functional differentiation of the uterus during preg-

ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2011 by The Endocrine Society doi: 10.1210/en.2009-1037 Received September 1, 2009. Accepted December 21, 2010. First Published Online February 1, 2011

Abbreviations: ADI, Acceptable daily intake level; BPA, bisphenol A; CL, corpus lutea; CT, threshold cycle; DES, diethylstilbestrol; E2, estradiol; EMX-2, empty spiracles homolog 2; ER, estrogen receptor; Hoxa10, homeobox A10; IOD, integrated optical density; ITGB3, ␤3 integrin; LOAEL, lowest-observed-adverse-effects-level; NTP, National Toxicology Program; P, progesterone; PR, P receptor; U.S. EPA, United States Environmental Protection Agency.

Endocrinology, March 2011, 152(3):1101–1111

endo.endojournals.org

1101

1102

Varayoud et al.

BPA Affects Uterine Functions

nancy as part of an altered endocrine signal transduction pathway. Hoxa10 is an abdominal B-like homeobox gene that is expressed in the developing genitourinary tract during embryogenesis and in the adult uterus during early pregnancy (4). In Hoxa10-deficient adult female mice, implantation is severely compromised and defective decidualization leads to recurrent pregnancy loss and infertility (5). Furthermore, blocking Hoxa10 expression by administering Hoxa10 antisense oligonucleotides into the uterine lumen results in decreased litter size (6). Interestingly, sex steroid hormones, particularly estradiol (E2) and progesterone (P) and their cognate receptors from the nuclear receptor superfamily, regulate adult Hoxa10 gene expression. Previous results have shown that differential spatiotemporal activation of Hoxa10 by E2 and P in the endometrium may be necessary to achieve distinct stages of functional development to support blastocyst implantation (7). Two Hoxa10-target genes, which are known to be operative downstream in endocrine hormone-regulated Hox gene pathways, are ␤3 integrin (ITGB3) and empty spiracles homolog 2 (EMX-2). Endometrial ITGB3 expression coincides with systemic peak of P, and high endometrial Hoxa10 levels occur in the midsecretory phase of menstrual cycle, around the time of embryo implantation (7). In contrast, at the same time, Hoxa10 directly binds a regulatory element in an enhancer region of the EMX-2 gene and transcriptionally represses its expression (8). In baboons with experimental endometriosis, Hoxa10 expression was lower than that of disease-free animals. This decrease was accompanied by abnormal expression of EMX-2 and ITGB3 (9). In the present study, we hypothesized that an early postnatal exposure to BPA or diethylstilbestrol (DES) could affect the estrogen receptor (ER)/P receptor (PR)-Hoxa-10 pathway producing the failure of endocrine-regulated implantation process. We evaluate the effects of neonatal xenoestrogen exposure on 1) pregnancy rate and the number of embryo implantation sites, 2) implantation-associated uterine gene expression (Hoxa10, EMX-2, and ITGB3) at d 5 of pregnancy (preimplantation period), and 3) ovarian steroid serum levels and their nuclear receptors in uterine tissue at d 5 of pregnancy.

Materials and Methods Animals All procedures used in this study were approved by the Institutional Ethic Committee of the School of Biochemistry and Biological Sciences (Universidad Nacional del Litoral) and were performed in accordance with the principles and procedures out-


lined in the Guide for the Care and Use of Laboratory Animals issued by the United States National Academy of Sciences. Rats of an inbred Wistar-derived strain bred at the Department of Human Physiology (Universidad Nacional del Litoral) were maintained under a controlled environment (22 ⫾ 2 C; lights on from 0600 to 2000 h) and had free access to pellet laboratory chow (Nutricioń Animal, Santa Fe, Argentina) and tap water. The concentration of phytoestrogens in the diet was not evaluated. However, because food intake was equivalent for control and experimental rats (our unpublished observations), we assumed that all animals were exposed to the same levels of phytoestrogens. To minimize additional exposures to endocrine-disrupting chemicals, rats were housed in stainless steel cages with wood bedding, and tap water was supplied ad libitum in glass bottles with rubber stoppers surrounded by a steel ring.

Experimental design Pregnant rats were housed singly, and at delivery, pups were sexed according to anogenital distance and cross-fostered by distributing pups of each litter between different mothers. These actions allowed us to minimize the use of siblings to avoid potential litter effects. Cross-fostered litters were adjusted to 10 pups, with up to 10 female pups per litter if possible. When fewer than 10 females were available, an appropriate number of males were retained. Female pups were assigned to one of five neonatal treatments: 1) controls given corn oil (charcoal stripped) vehicle alone (n ⫽ 17), 2) DES (Sigma, St. Louis, MO) at 0.2 ␮g/kg 䡠 d (DES.2) (n ⫽ 17) or 20 ␮g/kg 䡠 d (DES20) (n ⫽ 10), and 3) BPA (99% purity; Aldrich, Milwaukee, WI) at 0.05 mg/kg 䡠 d (BPA.05) (n ⫽ 17) or 20 mg/kg 䡠 d (BPA20) (n ⫽ 20). Treatments were given on postnatal d 1, 3, 5, and 7 (day of birth, 0) by sc injections in the nape of the neck. Based on an estimate from the United States Environmental Protection Agency (U.S. EPA), the lowest-observed-adverse-effects-level (LOAEL) dose for oral exposure to BPA in rats is 50 mg/kg 䡠 d. Experts from the National Toxicology Program (NTP) suggested a cut-off dose of 5 mg BPA/kg 䡠 d for the classification of low-dose effects, regardless of the administration route, duration of exposure, or the age/life stage at which the exposure occurred (10). Taking into account the above mentioned concepts, the high dose of BPA used in our study is four times that of the low-dose cut-off suggested by NTP’s expert panel and 2.5-fold lower than the U.S. EPA LOAEL. Meanwhile, the low dose is 100 times smaller than the low-dose cut-off suggested by NTP’s expert panel and is similar to the “safe dose” or acceptable daily intake level (ADI) established by the U.S. EPA. The effects of neonatal DES over a wide dose range of exposure have been used to predict potential adverse effects on the reproductive tract (11). According to previous results, low doses of classical estrogens, similar to our doses of DES, are recommended to compare the effects of weaker environmental estrogens (12). No signs of acute or chronic toxicity were observed, and no significant differences in weight gain between xenoestrogen-exposed and control pups were recorded during the experiment. At 21 d of age, animals were weaned, housed four per cage, and held without further treatment. Female rats exposed to xenoestrogens did not exhibit advanced puberty, measured as early vaginal opening compared with controls (data not shown). At the age of 80 d, females rats neonatally exposed to xenoestrogen were assigned to different studies: 1) evaluation of reproductive performance determining the pregnancy rates, the number of corpora lutea (CLs), and implantation and resorption sites at d 18 of pregnancy; and 2) assessment


of uterine gene expressions, ovarian steroid serum levels, and uterine PR and ER␣ during the preimplantation period (d 5).

Evaluation of reproductive performance Control and xenoestrogen-treated female rats (10 –13 rats per group) were housed for two consecutive weeks with sexually mature, untreated males of the same strain and of proven fertility to allow several possible matings. Every morning, vaginal smears were performed to check for the presence of spermatozoa (13). The first detection of a sperm-positive smear marked d 1 of gestation. The pregnancy rates were calculated as the number of pregnant females/number of females housed with a male ⫻ 100. Sperm-positive females were housed separately and killed on d 18 of pregnancy. The ovaries were dissected, and the numbers of profuse irrigated CLs were counted by direct visualization with the aid of a stereomicroscope (Leica Corp., Buffalo, NY). The


1103

two-horned uteri were removed and visually inspected to identify resorption sites and implantation sites. Resorption sites were defined as endometrial sites with an appended amorphous mass without a fetus (Fig. 1F). The number of implantation sites was defined as the result of the total number of placentae with fetuses plus the total number of resorptions sites (14).

Evaluation of uterine gene expression and steroid hormones Groups of neonatal xenoestrogen-exposed rats (seven rats per group) were housed with males and checked by vaginal smears every morning (sperm-positive indicated d 1 of pregnancy). Pregnant females were individually maintained until the morning of d 5 of pregnancy. In our colony, the embryo implantation process occurs on the evening of d 5. Therefore, pregnant rats were autopsied during the late preimplantation period. At this moment,

FIG. 1. Reproductive performance in neonatal xenoestrogen exposed animals. A, The pregnancy rates were calculated by the average of females that were pregnant and the number of females housed with a fertile male. The number of CLs (B) and the number of implantation sites (C) were expressed as the mean ⫾ SEM for each experimental group. The numbers of resorption sites (D) in each individual pregnant rat were plotted, and the horizontal lines are the mean for each experimental group. Asterisks indicate statistical significance compared with the control (*, P ⬍ 0.05; **, P ⬍ 0.01). All results were obtained on d 18 of pregnancy. E and F, Photographs of representative uteri collected on d 18 of pregnancy from a control female (E) vs. a BPA20 neonatally exposed animal (F). Note the lower number of implantation sites in the uterus of a BPA20-treated rat compared with a control. Each black arrow indicates an implantation site, and white arrows indicates resorption sites.

1104

Varayoud et al.


trunk blood was collected, and serum was separated and stored at ⫺20 C until needed for the hormone assay. Uterine tissue was collected and processed for different experimental purposes. For immunohistochemistry, one uterine horn of each rat was fixed in 10% buffered formalin for 24 h at room temperature and embedded in paraffin. For RNA extraction, the other uterine horn of each rat was immediately frozen in liquid nitrogen and stored at ⫺80 C.

Hormone assays Serum levels of E2 and P were determined by RIA after ethyl ether and hexane (Merck, Buenos Aires, Argentina) extraction, respectively (15). The antibodies were provided by G. D. Niswender, and labeled hormones were purchased from PerkinElmer Life and Analytical Sciences (Boston, MA). Assay sensitivities were 1.6 pg/ml and 1.2 ng/ml for E2 and P, respectively. Intra- and interassay coefficients of variation were 3.6 and 11% for E2 and 9 and 14.3% for P, respectively.

Immunohistochemistry Serial sections (5 ␮m in thickness) of paraffin-embedded uterine horns were mounted on 3-aminopropyl triethoxy-silane (Sigma)-coated slides and microwave pretreated for antigen retrieval (16). Primary antibodies were incubated overnight at 4 C. Immunostaining was performed using a rabbit anti-PR (A/B isoforms) antibody (1:500 dilution; Dako Corp., Carpinteria, CA) and a mouse anti-ER␣ (clone 6F-11, 1:200 dilution; Novocastra, Newcastle upon Tyne, UK). Antirabbit/antimouse (biotin conjugate) were purchased from Sigma. Reactions were developed using a streptavidin-biotin peroxidase method using diaminobenzidine (Sigma) as a cromogen substrate. Samples were mounted with permanent mounting medium (PMyR, Buenos Aires, Argentina). Each immunohistochemical run included negative controls replacing the primary antibody with nonimmune mouse serum (Sigma).

Quantification of protein expression The evaluation of PR and ER␣ proteins was performed from the subepithelial stromal cells, because we have previously observed that neonatal BPA exposure affects steroid hormone responsiveness in this compartment in adulthood (3). To measure the integrated optical density (IOD), an image analysis was performed using the Image Pro-Plus 4.1.0.1 system (Media Cybernetics, Silver Spring, MD) as previously described (17). In brief, the images were recorded with a Spot Insight version 3.5 color video camera, attached to an Olympus (Tokyo, Japan) BH2 microscope, using a Dplan ⫻40 objective (at least 10 fields were recorded in each section, and two sections per animal were evaluated). The microscope was set up properly for Koehler illumination. Correction of unequal illumination (shading correction) and the calibration of the measurement system were done with a reference slide. The images of immunostained slides were converted to gray scale, and the subepithelial stromal compartment was delimited (a 300-␮m-wide area adjacent to the epithelium, from the basement membrane toward the outer layers). Using the Auto-Pro macro language (Media Cybernetics), an automated standard sequence operation was created to measure the IOD as a linear combination between the average gray intensity and the relative area occupied by positive cells. Because IOD is a dimensionless parameter, the results were expressed as arbitrary units.


Dual immunofluorescent staining Uterine sections were deparaffinized, rehydrated, and submitted to microwave antigen retrieval. Sections were blocked for 1 h with normal donkey serum (Hoxa10) or goat serum (PR and ER␣) (Sigma). The primary antibodies for ER␣ and PR were described in the immunohistochemistry section. The Hoxa10 detection was performed with a goat anti-Hoxa10 antibody (sc17159, 1/50 dilution; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and antigoat secondary antibodies (biotin conjugate) (Santa Cruz Biotechnology, Inc.). The incubation with primary antibodies was performed overnight at 4 C. The secondary antibodies, Alexa Fluor 488 goat antirabbit (green) (1/500 dilution; Invitrogen Molecular Probes, Eugene, OR), Cy2-conjugated goat antimouse (green) (1/100 dilution) and tetramethylrhodamine isothiocyanate (red)-conjugated streptavidin (1/150 dilution; Jackson ImmunoResearch, West Grove, PA) were incubated for 1 h, and then sections were washed for a total of 45 min in three changes of PBS. Finally, all sections were mounted in Vectashield fluorescent mountant (Vector Laboratories, Inc., Burlingame, CA) with 4⬘,6-diamidino-2-phenylindole dihydrochoride (Fluka; Sigma) and stored in the dark at 4 C. Negative controls included uterine sections incubated with primary antibody buffer solution (3% BSA, 0.1% Tween 20 in PBS) to observe nonspecific staining. All immunostains were evaluated using an Olympus BX-51 microscope equipped for epifluorescence detection with the appropriate filters, and images were recorded using a High-resolution USB 2.0 Digital Color Camera (QImaging Go-3; QImaging, Surrey, British Columbia, Canada).

RT and real-time quantitative PCR analysis RNA extraction and RT Seven uterine horns on d 5 of pregnancy were used for each experimental group. Total RNA was individually extracted using TRIZOL reagent (Life Technologies, Grand Island, NY) according to the manufacturer’s instructions. The concentration and purity of total RNA was determined by measuring the optical density at 260 and 280 nm. All samples were precipitated with ethanol, then dissolved in distilled water, and their quality was verified by gel electrophoresis. Equal quantities (4 ␮g) of total RNA were reverse-transcribed in duplicated reactions using recombinant Avian Myeloblastosis Virus reverse transcriptase (12.5 U; Promega, Madison, WI) and 200 pmol of random primers (Promega); 20 U of ribonuclease inhibitor (RNAout; Invitrogen Argentina, Buenos Aires, Argentina) and 100 nmol of a deoxy NTP mixture were added to each reaction tube in a final volume of 30 ␮l of 1⫻ reverse transcriptase buffer. RT was performed at 42 C for 90 min. Reactions were terminated by heating at 97 C for 5 min and cooling on ice.

Real-time quantitative PCR Each reverse-transcribed product was diluted with ribonuclease-free water to a final volume of 60 ␮l and further amplified in triplicate using the Real-Time DNA Engine Opticon System (Bio-Rad Laboratories, Inc., Waltham, MA) and SYBR Green I dye (Cambrex Corp., East Rutherford, NJ). Primer sequences used for amplification of ER␣, PR, Hoxa10, EMX-2, ITGB3, and ribosomal protein L19, Rpl19 (housekeeping gene), cDNAs are described in Supplemental Table 1, published on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org. For amplifications, 5 ␮l cDNA was combined with a mixture


containing 2.5 U Platinum Taq-DNA polymerase (Invitrogen), 2 mM MgCl2 (Invitrogen), 0.2 mM of each of the four deoxynucleotides (Promega), 5 ␮l of 4⫻ SYBR Green I, and 10 pmol of each primer (Invitrogen) in a final volume of 25 ␮l of 1⫻ PCR buffer. The cycling conditions were as follows: after initial denaturation at 97 C for 3 min, the reaction mixture was subjected to successive cycles of denaturation at 97 C for 15 sec, annealing at 60 C for 15 sec, and extension at 72 C for 15 sec. Product purity was confirmed by dissociation curves and random agarose gel electrophoresis. Controls containing no template DNA were included in all assays, yielding no consistent amplification. The efficiency of PCRs was assessed for each target by amplification of serial dilutions (over five orders of magnitude) of cDNA fragments of the transcripts under analysis. No significant differences in threshold cycle (CT) values were observed for ribosomal protein L19 between the different experimental groups. The CT was determined for each gene of interest in each experimental sample using the Opticon Monitor Analysis Software (MJ Research; Bio-Rad Laboratories) with an automatic fluorescence threshold setting. Relative gene expression data were calculated using the comparative CT method. For each sample, ⌬CT was calculated as the difference in CT between target mRNA and L19 mRNA used as a housekeeping gene. ⌬⌬CT was calculated as the difference between the average ⌬CT for each sample and the average ⌬CT for the control group (baseline). The change in target mRNA, relative to baseline, was calculated as 2⫺(⌬⌬CT). Changes in mRNA expression were expressed as fold change regarding to control group (18).

Statistics All data were calculated as the mean ⫾ SEM. The pregnancy rates were analyzed by the Fisher’s exact test. The differences in the number of implantation sites and CLs between control and DES-treated groups were analyzed using the t test, and comparisons between control and BPA-treated groups were analyzed using one-way ANOVA followed by Dunnett’s test for multiple comparisons (after Bartlett’s test for homogeneity of variance). The number of resorption sites was analyzed using a generalized linear model with a negative binomial response, using the glm.nb function of the statistical software R (The R Foundation for Statistical Computing). Other variables were analyzed by the Kruskal-Wallis test followed by Dunn’s method of multiple comparisons. P ⬍ 0.05 was accepted as significant.

Results The neonatal xenoestrogen-exposed animals showed a lower number of implantation sites Different results were observed between females postnatally exposed to BPA and DES. All female rats exposed to the higher dose of DES (DES20 females) did not show receptive sexual behavior manifested by the absence of spermatozoa in their vaginal smears. Moreover, DES20 females were infertile, showing ovaries lacking mature follicles and functional CLs. Therefore, they were not incorporated into the second part of the study. Unlike DES20, control animals and the rest of the xenoestrogen-exposed females had sperm-positive smears, suggesting that they


1105

showed sexual receptivity (lordosis reflex) in response to stimuli from the male. However, none of the xenoestrogen-treated groups reached the pregnancy rate values of control animals. Pregnancies were not established in 30% of the DES.2, 10% of BPA.05, and 23% of BPA20 animals, whereas controls exhibited a 100% pregnancy rate (Fig. 1A). Even though the number of CLs in pregnant females was similar regardless of the neonatal treatment (CLs/rat, 12–13) (Fig. 1B), the number of implantation sites decreased significantly in DES.2 and BPA20 females compared with controls. A similar trend was observed in the BPA.05 group but with no statistical significance (Fig. 1C). Interestingly, in both BPA groups, not only did the number of rats with more than one resorption site increase, but also the number of resorption sites per rat tended to be higher than in the control group (BPA.05, P ⫽ 0.078; and BPA20, P ⫽ 0.068) (Fig. 1D). Representative d 18 pregnant uteri from control and BPA20-treated females are shown in Fig. 1, E and F, respectively. Because the neonatal exposure to BPA affected the number of implantation sites, next, we studied the expression of uterine markers associated with this event. Ovarian steroid levels We measured serum levels of E2 and P on the morning of d 5 (preimplantation period). No differences in serum concentrations of either ovarian steroid hormone were found between pregnant rats of all experimental groups (E2 serum levels expressed in pg/ml: control, 90.6 ⫾ 3.74; DES.2, 95.0 ⫾ 4.88; BPA20, 101 ⫾ 9.15; BPA.05, 101 ⫾ 12.4; P ⬎ 0.05; and P serum levels expressed in ng/ml: control, 39.7 ⫾ 1.38; DES.2, 42.5 ⫾ 1.72; BPA20, 38.4 ⫾ 2.24; BPA.05, 42.6 ⫾ 1.95; P ⬎ 0.05). ER␣ and PR expression in the preimplantation uterus To address the molecular effect of neonatal xenoestrogen exposure, we examined steroid hormone receptor expression in the preimplantation uterus. The ER␣ and PR mRNA levels on d 5 of pregnancy in samples of rats neonatally exposed to DES.2, BPA.05, and BPA20 were significantly lower compared with those of control animals (Fig. 2, A and B). To determine whether alterations, prompted by postnatal xenoestrogen exposure, extended to ER␣ and PR proteins, we quantified the relative protein expression level in subepithelial stromal cells (Fig. 3). Both isoforms of PR (PRA and PRB) and ER␣ were identified by immunohistochemistry and quantified by image analysis to determine the IOD. The results of quantification are shown in Fig. 3, I and J. A high expression of ER␣ and PR was observed in controls on d 5 of pregnancy (Fig. 3, A and E). In contrast, the uterine stroma of xenoestrogen-ex-

1106

Varayoud et al.



viously reported to be up-regulated and down-regulated by Hoxa10, respectively, were measured by real-time PCR. Notably, the uterine tissue obtained from neonatally BPA20- or DES.2-exposed rats exhibited a lower expression of ITGB3 and a higher expression of EMX-2 compared with control animals (Fig. 4). Thus, lower Hoxa10 expression found in xenoestrogen-treated animals could lead to misexpression of Hoxa10 downstream targets genes.

FIG. 2. Effect of neonatal xenoestrogen exposure on ER␣ and PR mRNA expression levels in the uterine tissue on d 5 of pregnancy. Relative mRNA levels of ER␣ (A) and PR (B) were measured by real-time RT-PCR, and fold expression from control values was calculated for each target by the equation 2⫺⌬⌬CT. Control values were assigned to a reference level of 100, and each column represents the mean ⫾ SEM. Ribosomal protein L19 was used as an internal control (n ⫽ 7 per group). Asterisks denote P ⬍ 0.05, versus the control group.

posed rats showed a lower expression of the ER␣ (Fig. 3, B–D), whereas PR expression was statistically affected only in DES.2- and BPA20-treated rats (Fig. 3, F and H). Fluctuations in the ER␣ and PR protein level between groups were due to variations in both staining intensity and the relative area occupied by positive nuclei in the subepithelial stroma (Fig. 3, A–H). Implantation-associated uterine genes: Hoxa10, EMX-2, and ITGB3 The expression of implantation-related genes, including Hoxa10, EMX-2, and ITGB3, were measured by realtime PCR from the uteri of experimental rats (Fig. 4). A clear decrease in Hoxa10 expression was detected in BPA20, BPA.05, and DES.2 neonatally exposed rats. Immunohistochemical analysis of this homeotic gene revealed that Hoxa10 was localized to the nuclei of many cells in the subepithelial stroma. In addition, glandular epithelial nuclei stained positive with anti-Hoxa10 antibodies (Figs. 4 and 5E). The immunohistochemical results showed that in rats neonatally exposed to xenoestrogens, the Hoxa10 subepithelial uterine expression was lower than in controls (Fig. 4B). To determine whether the decrease in Hoxa10 affected downstream targets, the levels of ITGB3 and EMX-2, pre-

Colocalization of PR, ER␣, and Hoxa10 In the subepithelial stroma, where PR, ER␣, and Hoxa10 are expressed, the question of whether ER␣/PR and PR/Hoxa10 colocalized was addressed using dual immunofluorescence. In the preimplantation uterus (on d 5), ER␣/PR and PR/Hoxa10 colocalized in the subepithelial stromal nuclei (Fig. 5, A and C, and D and F, respectively). No colocalization of Hoxa10/PR and ER␣/PR was observed in the luminal and glandular epithelium. The pattern of coexpression was not modified by the neonatal xenoestrogen treatment.

Discussion Neonatal exposure to endocrine disruptors affects signaling events governed by Hox genes, altering the normal development of the uterus with long-term consequences. We previously demonstrated that BPA exposure during early postnatal life disrupts expression of Hoxa10 and Hoxa11, and we suggested that this effect could reprogram the normal responsiveness of uterine stromal cells to sex steroids during adulthood (3). In the present study, using the same animal model, we showed long lasting effects on the number of implantation sites and on implantation-associated gene expression in female rats neonatally treated with BPA or DES. In agreement with our results, previous work has shown that DES exposure during development permanently reprograms the normal uterine physiological response, and this leads to increased tumor development in genetically susceptible individuals (19). Newbold et al. (20) showed, after neonatal BPA treatment, severe pathologies of the adult uterus, which include benign, premalignant, and neoplastic changes. All results reinforce the idea that early postnatal development is a critical period, highly susceptible to endocrine disrupter effects. In the present work, we showed that animals exposed to a higher dose of DES were infertile. Previously, we studied the hypothalamic effects of females neonatally exposed to DES20, demonstrating that sexual related nuclei were functionally defeminized, showing a significant reduction



1107

FIG. 3. Effects of neonatal xenoestrogen exposure on ER␣ and PR proteins expression on d 5 of pregnancy. Representative photomicrographs of immunohistochemical detection of ER␣ (A–D) and PR (E–H) on uterine sections. The expression of ER␣ and PR in the subepithelial stroma was affected in xenoestrogen-treated rats. LE, Luminal epithelium; ST, subepithelial stroma; GE, glandular epithelium. Scale bar, 50 ␮m. Quantification of ER␣ (I) and PR (J) immunostaining in the subepithelial stroma is expressed as the integrated optical density (IOD), which is a linear combination between the average of immunostaining intensity and the relative area occupied by positive cells. Each column represents the mean ⫾ SEM (n ⫽ 7 per group); *, P ⬍ 0.05 vs. the control group.

in ER␣ expression (21). Here, we observed that the ovaries of DES20 females lack CLs, as previously described in ovaries from adult mice neonatally exposed to E2 (22). The percentage of rats that became pregnant after neonatal exposure to the low dose of DES and both doses of BPA showed a decreased trend, despite the fact that all these females were sexual receptive. Recently, we demonstrated that neonatal exposure to the same doses of BPA affects only proceptive sexual behavior, but females were normally receptive (23). The number of CLs in the ovaries from all the females that became pregnant, regardless of the xenoestrogen treatment, was similar to controls (12–13 CLs/rat). This result suggests that the ovulation rate and CL “activation” were not altered in most of the females after neonatal exposure to xenoestrogen. In contrast, the number of implantation sites was significantly lower in xenoestrogen-treated rats, suggesting an intrinsic uterine defect that preceded the embryo arrival. In addition, BPA-treated groups in particular showed a tendency to have a higher number of resorptions (indicative of postimplantation loss). Other authors have reported that adult exposure to leached components from dental sealants containing BPA

caused toxic effects to reproduction in the mouse (24). These authors observed no effects on the number of implantations; however, they showed an increased number of resorptions (postimplantation loss). Therefore, BPA effects on reproductive performance could be different depending on the animal model used, developmental stage at exposure, and/or the dose administered. Previously published BPA effects include changes at the onset of puberty, in the regularity of the estrus cycle, and in the gross anatomy of the ovaries (25–27). Taking into account the above mentioned results, it cannot be ruled out that there is a disturbance of reproductive/endocrine function with multiple sites of toxicity along the hypothalamic-pituitaryreproductive tract. Further investigation into the effects of BPA on the hypothalamic-gonadal axis, using the animal model reported here, is currently underway in our laboratory. Our next purpose was to investigate whether the effects of xenoestrogens on the number of implantation sites could be associated with disruption of different markers of uterine implantation. To evaluate whether endocrine pathways were affected, we measured serum levels of sex steroids and steroid hormone receptor expression (ER␣

1108

Varayoud et al.



FIG. 4. Effect of neonatal xenoestrogen exposure on implantation-associated gene expression in uterine tissue of d 5 pregnant rats. Hoxa10 mRNA (A) and two downstream genes, EMX-2 (C) and ITGB3 (D), were measured by real-time RT-PCR. Values were normalized with Rpl19 and expressed as 2⫺⌬⌬CT (n ⫽ 7 rats per group). Values are expressed as the mean ⫾ SEM; *, P ⬍ 0.05 vs. control group. B, Immunohistochemical detection of Hoxa10 shows that in rats neonatally exposed to xenoestrogens, the Hoxa10 subepithelial uterine expression was lower than in controls. LE, Luminal epithelium; GE, glandular epithelium.

and PR) on d 5 of pregnancy. The uteri of control animals showed a high expression of ER␣ and PR in subepithelial stromal cells. In neonatal DES.2 and BPA20-treated animals, we detected a lower expression of both receptors specifically in the subepithelial stroma, without changes in serum levels of E2 and P between experimental groups. The BPA selected doses produced different effects on PR expression, suggesting that the uterine response to hormonal milieu at d 5 of pregnancy is differentially affected. BPA.05 diminished the PR mRNA, but not PR protein, whereas BPA20 affected both mRNA and protein expression. In this context, we propose that BPA.05 could affect the ubiquitination of PR protein and consequent degradation. Previous results showed that a reduced ubiquitination of PR contributes to its stabilization and is correlated with

increased response to P (28). The estrogen influence on the ubiquitination system has been shown in many papers (29). ER␣ is commonly associated with ubiquitin ligases (like Smurf) forming protein complexes that regulate the ubiquitination rate of many targets in an estrogen-dependent manner. In particular PR is a substrate for the ubiquitin/ proteasome pathway (30). Ubiquitination of the PR depends on S294 phosphorylation. This was demonstrated using a S294A mutant, which does not undergo ubiquitination and degradation by the proteasome pathway after P binding. These (and others) results provide some empirical data to speculate about the possible influence of xenoestrogens on the control of the ubiquitination process. Hoxa10 expression, an ovarian steroid downstream target gene, was altered in xenoestrogen-treated animals.


FIG. 5. Representative photomicrographs of dual immunofluorescence staining for ER␣/PR (A–C) and PR/Hoxa10 (D–F) in the uterus of a control animal on d 5 of pregnancy. Substantial coexpression of PR and ER␣ in the nuclei of subepithelial stromal cells, evidenced as yellow nuclei in merged images, is shown in C. F, The higher expression of Hoxa10 is coincident with the expression of PR in the subepithelial stroma. ER␣/PR and PR/Hoxa10 were not coexpressed in the luminal or glandular epithelium. GE, Glandular epithelium; LE, luminal epithelium. Scale bar, 100 ␮m.

Thus, the decrease in Hoxa10 expression in the preimplantation uterus was further evidenced in the absence of variation in circulating steroid hormone levels, and it appeared to be due to a decrease in PR and ER␣ expression in the subepithelial stroma. A previous report demonstrated that ER␣ and PR are implicated in the regulation of Hoxa10 expression, using in vitro and in vivo models (reviewed in Ref. 31). The E2 regulation of Hoxa10 was associated with the detection of ER binding of two putative estrogen response element in the 5⬘ regulatory region of Hoxa10. Progestational regulation of Hoxa10 occurred via the PR and therefore is blocked by RU486 (32). In BP.05, the decrease in the mRNA levels of Hoxa10 and the lack of statistical significance in the PR protein expression could also be associated with changes in the expression of nuclear receptor coregulators, altering the transcriptional machinery of steroid-dependent genes (3). Hoxa10 is an essential transcription factor required for endometrial receptivity and embryo implantation, it is regulated in response to E2 and P, and its levels increase dramatically when P levels are high. Therefore, these results could indicate that the normal hormonal control of endometrial receptivity could be affected as a consequence of neonatal exposure to BPA or DES. Two previous studies have evaluated the effects of BPA on the Hoxa10 uterine expression. Smith and Taylor (33) reported a dose-responsive increase in uterine stromal cells Hoxa10 expression in


1109

2- and 6-wk-old mice in utero exposed to BPA, whereas we showed that a postnatal exposure to BPA decreased the uterine Hoxa10 expression (3). Both studies show that Hox genes are targets of endocrine disruption and suggest that exposure during different developmental periods could lead to a different effect. EMX-2 and ITGB3 are two Hoxa10 target genes that are known to be operative downstream in endocrine hormone-regulated Hox gene pathways. EMX-2 is a homeobox gene located outside of the HOX cluster, and it is orthologous to the Drosophila empty spiracles gene (34). Hoxa10 is a direct negative regulator of EMX-2 in reproductive tissue (8). ITGB3 has been proposed to be a bridging molecule between the endometrium and trophoblast, thereby constituting an early link between maternal and fetal tissues (35). ITGB3 expression also coincides with peak systemic P and high endometrial Hoxa10 levels in the midsecretory phase of the menstrual cycle. Moreover, Hoxa10 has been shown to directly regulate the expression of ITGB3 through a consensus Abd-B type HOX binding site located 5⬘ of the ITGB3 gene within its regulatory region (36). Consistently, our results show that those experimental groups exhibiting a lower number of implantation sites and disruption of Hoxa10 additionally showed a lower expression of ITGB3 and a higher expression of EMX-2. Further evidence demonstrating that Hoxa10, ITBG3, and EMX-2 function in a common pathway is seen in diseases with implantation defects, such as endometriosis, where simultaneous misregulation of these three genes has been documented (37, 38). Throughout this study, two issues are interesting to highlight. First, we observed that both doses of BPA showed different effects. Even though this is a common observation in endocrine disrupter studies (21), it is difficult to explain. We could suggest that BPA20 and DES.2 animals showed a more severe alteration of uterine functions, because all components of the Hoxa10 pathway were affected, producing as a consequence a lower number of implantation sites. In contrast, in BPA.05 animals the Hoxa10 pathway is partially affected, without changes in ITGB3 and EMX-2 levels and the number of implantation sites. In accordance with BPA.05 effects, a previous study showed that an experimental murine endometriosis is accompanied by a reduction on Hoxa10 expression without effects on ITGB3 levels (39). In addition, we could not discard that in the BPA.05 animals, different Hoxa10 target genes, such as cyclin-dependent kinase inhibitor 1A (40), genes of the Wnt pathway (41), or FK506 binding protein 4 (42) were affected. Second, the major effects associated with exposure to BPA were found in the uterus of BPA20-treated rats. As detailed in Materials and Methods, the dose of BPA20 is 2.5-fold lower than the LOAEL. The

1110

Varayoud et al.


safe dose or ADI established by both the U.S. EPA and U.S. Food and Drug Administration is calculated based on the LOAEL. We believe that studies showing effects of BPA at levels lower than the current LOAEL dose are highly relevant, because they may have an impact on the dose used to calculate the ADI dose. Several reports indicated that exposure to BPA during critical developmental windows in doses below LOAEL affects reproductive health in adulthood (3, 11, 20, 21, 23, 25, 27). Thus, the concept of LOAEL as currently used is inadequate. In the rat, the uterine development and differentiation is completed during the first postnatal week, representing a critical period to evaluate the effects of xenoestrogens exposure. Furthermore, there is evidence that younger animals metabolize BPA in a less efficient way (43, 44), resulting in higher circulating levels of this compound and implying that oral and nonoral administration of BPA during neonatal life provide the same internal active dose (45). The quantification of exposure levels is difficult to define. Interestingly, a recent report published by the Centers for Disease Control and Prevention (CDC, Atlanta, GA) showed that 93% of the United States population had detectable levels of BPA in their urine, with levels in children being significantly higher than levels in adolescents and adults (46). The combination of data showing reduced conception rates in humans and the common occurrence of female reproductive disease raises concerns that environmental factors may be having a negative impact on female reproductive health. However, the effects of endocrine disruptors are difficult to determine, because the problems that arise do so long after the exposure ended. In addition, it is becoming increasingly clear from epidemiological studies in humans (47), as well as genetic studies in rodents (48), that failed pregnancies occur in large part to faulty uterine function or miscommunication between the embryo and the mother before placentation. Corroborating this paradigm, our results should serve as an alert that a possible relationship exists between reproductive performance and altered uterine function due to neonatal exposure to BPA.

Acknowledgments We Thank Mr. Juan Grant and Mr. Juan C. Villarreal for technical assistance and animal care and Dr. Pablo Beldomenico for advice on statistics. J.V., J.G.R., V.L.B., and E.H.L. are Career Investigators of the Consejo Nacional de Investigaciones Científicas y Tećnicas. Address all correspondence and requests for reprints to: Enrique H. Luque, Laboratorio de Endocrinología y Tumores Hormonodependientes, School of Biochemistry and Biological


Sciences, Casilla de Correo 242, 3000 Santa Fe, Argentina. E-mail: [email protected]. This work was supported by the Universidad Nacional del Litoral and the Argentine National Agency for the Promotion of Science and Technology. Disclosure Summary: The authors have nothing to disclose.

References 1. NTP-CERHR 2008 NTP-CERHR monograph on the potential human reproductive and developmental effects of bisphenol A. NTP CERHR MON 22:i-III1 2. Mahalingaiah S, Meeker JD, Pearson KR, Calafat AM, Ye X, Petrozza J, Hauser R 2008 Temporal variability and predictors of urinary bisphenol A concentrations in men and women. Environ Health Perspect 116:173–178 3. Varayoud J, Ramos JG, Bosquiazzo VL, Munõz-de-Toro M, Luque EH 2008 Developmental exposure to bisphenol A impairs the uterine response to ovarian steroids in the adult. Endocrinology 149: 5848 –5860 4. Krumlauf R 1994 Hox genes in vertebrate development. Cell 78: 191–201 5. Benson GV, Lim H, Paria BC, Satokata I, Dey SK, Maas RL 1996 Mechanisms of reduced fertility in Hoxa-10 mutant mice: uterine homeosis and loss of maternal Hoxa-10 expression. Development 122:2687–2696 6. Bagot CN, Troy PJ, Taylor HS 2000 Alteration of maternal Hoxa10 expression by in vivo gene transfection affects implantation. Gene Ther 7:1378 –1384 7. Taylor HS, Arici A, Olive D, Igarashi P 1998 HOXA10 is expressed in response to sex steroids at the time of implantation in the human endometrium. J Clin Invest 101:1379 –1384 8. Troy PJ, Daftary GS, Bagot CN, Taylor HS 2003 Transcriptional repression of peri-implantation EMX2 expression in mammalian reproduction by HOXA10. Mol Cell Biol 23:1–13 9. Kim JJ, Taylor HS, Lu Z, Ladhani O, Hastings JM, Jackson KS 2007 Altered expression of HOXA10 in endometriosis: potential role in decidualization. Hum Reprod 13:323–332 10. Melnick R, Lucier G, Wolfe M, Hall R, Stancel G, Prins G, Gallo M, Reuhl K, Ho SM, Brown T, Moore J, Leakey J, Haseman J, Kohn M 2002 Summary of the National Toxicology Program’s report of the endocrine disruptors low-dose peer review. Environ Health Perspect 110:427– 431 11. Newbold RR, Jefferson WN, Padilla-Banks E, Haseman J 2004 Developmental exposure to diethylstilbestrol (DES) alters uterine response to estrogens in prepubescent mice: low versus high dose effects. Reprod Toxicol 18:399 – 406 12. Newbold RR 2004 Lessons learned from perinatal exposure to diethylstilbestrol. Toxicol Appl Pharmacol 199:142–150 13. Montes GS, Luque EH 1988 Effects of ovarian steroids on vaginal smears in the rat. Acta Anat 133:192–199 14. Barreto RL, de Jesus Simo˜es M, Amed AM, Soares Juńior JM, Oliveira-Filho RM, Kulay Jr L 2004 Stavudine effects on rat pregnancy outcome. J Obstet Gynaecol Res 30:242–245 15. Bosquiazzo VL, Ramos JG, Varayoud J, Munõz-de-Toro M, Luque EH 2007 Mast cell degranulation in rat uterine cervix during pregnancy correlates with expression of vascular endothelial growth factor mRNA and angiogenesis. Reproduction 133:1045–1055 16. Munõz-de-Toro M, Maffini MV, Giardina RH, Luque EH 1998 Processing fine needle aspirates of prostate carcinomas for standard immunocytochemical studies and in situ apoptosis detection. Pathol Res Pract 194:631– 636 17. Ramos JG, Varayoud J, Bosquiazzo VL, Luque EH, Munõz-deToro M 2002 Cellular turnover in the rat uterine cervix and its


18.

19.

20.

21.

22.

23.

24. 25.

26.

27.

28. 29.

30.

31. 32.

33.

relationship to estrogen and progesterone receptor dynamics. Biol Reprod 67:735–742 Higuchi R, Fockler C, Dollinger G, Watson R 1993 Kinetic PCR analysis: real-time monitoring of DNA amplification reactions. Biotechnology 11:1026 –1030 Cook JD, Davis BJ, Cai SL, Barrett JC, Conti CJ, Walker CL 2005 Interaction between genetic susceptibility and early-life environmental exposure determines tumor-suppressor-gene penetrance. Proc Natl Acad Sci USA 102:8644 – 8649 Newbold RR, Jefferson WN, Padilla-Banks E 2009 Prenatal exposure to bisphenol A at environmentally relevant doses adversely affects the murine female reproductive tract later in life. Environ Health Perspect 117:879 – 885 Monje L, Varayoud J, Luque EH, Ramos JG 2007 Neonatal exposure to bisphenol A modifies the abundance of estrogen receptor ␣ transcripts with alternative 5⬘-untranslated regions in the female rat preoptic area. J Endocrinol 194:201–212 Forsberg JG 1985 Treatment with different antiestrogens in the neonatal period and effects in the cervicovaginal epithelium and ovaries of adult mice: a comparison to estrogen-induced changes. Biol Reprod 32:427– 441 Monje L, Varayoud J, Munõz-de-Toro M, Luque EH, Ramos JG 2009 Neonatal exposure to bisphenol A alters estrogen-dependent mechanisms governing sexual behavior in the adult female rat. Reprod Toxicol 28:435– 442 Darmani H, Al-Hiyasat AS 2006 The effects of BIS-GMA and TEGDMA on female mouse fertility. Dent Mater 22:353–358 Durando M, Kass L, Piva J, Sonnenschein C, Soto AM, Luque EH, Munõz-de-Toro M 2007 Prenatal bisphenol A exposure induces preneoplastic lesions in the mammary gland in Wistar rats. Environ Health Perspect 115:80 – 86 Kato H, Ota T, Furuhashi T, Ohta Y, Iguchi T 2003 Changes in reproductive organs of female rats treated with bisphenol A during the neonatal period. Reprod Toxicol 17:283–288 Maffini MV, Rubin BS, Sonnenschein C, Soto AM 2006 Endocrine disruptors and reproductive health: the case of bisphenol-A. Mol Cell Endocrinol 254 –255:179 –186 Lee EY 2008 Promotion of BRCA1-associated triple-negative breast cancer by ovarian hormones Curr Opin Obstet Gynecol 20:68 –73 Ito I, Hanyu A, Wayama M, Goto N, Katsuno Y, Kawasaki S, Nakajima Y, Kajiro M, Komatsu Y, Fujimura A, Hirota R, Murayama A, Kimura K, Imamura T, Yanagisawa J 2010 Estrogen inhibits transforming growth factor ␤ signaling by promoting Smad2/3 degradation. J Biol Chem 285:14747–14755 Qiu M, Lange CA 2003 MAP kinases couple multiple functions of human progesterone receptors: degradation, transcriptional synergy, and nuclear association. J Steroid Biochem Mol Biol 85:147– 157 Daftary GS, Taylor HS 2006 Endocrine regulation of HOX genes. Endocr Rev 27:331–355 Ma L, Benson GV, Lim H, Dey SK, Maas RL 1998 (AbdB) Hoxa genes: regulation in adult uterus by estrogen and progesterone and repression in mullerian duct by the synthetic estrogen diethylstilbestrol (DES). Dev Biol 197:141–154 Smith CC, Taylor HS 2007 Xenoestrogen exposure imprints ex-


34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

1111

pression of genes (Hoxa10) required for normal uterine development. FASEB J 21:239 –246 Kastury K, Druck T, Huebner K, Barletta C, Acampora D, Simeone A, Faiella A, Boncinelli E 1994 Chromosome locations of human EMX and OTX genes. Genomics 22:41– 45 Sueoka K, Shiokawa S, Miyazaki T, Kuji N, Tanaka M, Yoshimura Y 1997 Integrins and reproductive physiology: expression and modulation in fertilization, embryogenesis, and implantation. Fertil Steril 67:799 – 811 Daftary GS, Troy PJ, Bagot CN, Young SL, Taylor HS 2002 Direct regulation of B3-integrin subunit gene expression by HOXA10 in endometrial cells. Mol Endocrinol 16:571–579 Lessey BA, Castelbaum AJ, Sawin SW, Buck CA, Schinnar R, Bilker W, Strom BL 1994 Aberrant integrin expression in the endometrium of women with endometriosis. J Clin Endocrinol Metab 79:643– 649 Taylor HS, Bagot C, Kardana A, Olive D, Arici A 1999 HOX gene expression is altered in the endometrium of women with endometriosis. Hum Reprod 14:1328 –1331 Lee B, Du H, Taylor HS 2009 Experimental murine endometriosis induces DNA methylation and altered gene expression in eutopic endometrium. Biol Reprod 80:79 – 85 Bromleigh VC, Freedman LP 2000 p21 is a transcriptional target of HOXA10 in differentiating myelomonocytic cells. Genes Dev 14: 2581–2586 Ferrell CM, Dorsam ST, Ohta H, Humphries RK, Derynck MK, Haqq C, Largman C, Lawrence HJ 2005 Activation of stem-cell specific genes by HOXA9 and HOXA10 homeodomain proteins in CD34⫹ human cord blood cells. Stem Cells 23:644 – 655 Daikoku T, Tranguch S, Friedman DB, Das SK, Smith DF, Dey SK 2005 Proteomic analysis identifies immunophilin FK506 binding protein 4 (FKBP52) as a downstream target of Hoxa10 in the periimplantation mouse uterus. Mol Endocrinol 19:683– 697 Domoradzki JY, Thornton CM, Pottenger LH, Hansen SC, Card TL, Markham DA, Dryzga MD, Shiotsuka RN, Waechter Jr JM 2004 Age and dose dependency of the pharmacokinetics and metabolism of bisphenol A in neonatal Sprague–Dawley rats following oral administration. Toxicol Sci 77:230 –242 Matsumoto J, Yokota H, Yuasa A 2002 Developmental increases in rat hepatic microsomal UDP-glucuronosyltransferase activities toward xenoestrogens and decreases during pregnancy. Environ Health Perspect 110:193–196 Taylor JA, Welshons WV, vom Saal FS 2008 No effect of route of exposure (oral; subcutaneous injection) on plasma bisphenol A throughout 24 h after administration in neonatal female mice. Reprod Toxicol 25:169 –176 Calafat AM, Ye X, Wong L-Y, Reidy JA, Needham LL 2008 Exposure of the U.S. population to bisphenol A and 4-tertiary-octylphenol: 2003–2004. Environ Health Perspect 116:39 – 44 Christiansen OB, Nybo Andersen AM, Bosch E, Daya S, Delves PJ, Hviid TV, Kutteh WH, Laird SM, Li TC, van der Ven K 2005 Evidence-based investigations and treatment of recurrent pregnancy loss. Fertil Steril 83:821– 839 Wang H, Dey SK 2006 Roadmap to embryo implantation: clues from mouse models. Nat Rev Genet 7:185–199