Evidence of Altered Brain Sexual Differentiation in MiceExposed Perinatally to Low, Environmentally RelevantLevels of Bisphenol A

Beverly S. Rubin, Jenny R. Lenkowski, Cheryl M. Schaeberle, Laura N. Vandenberg, Paul M. Ronsheim,and Ana M. Soto

Tufts University, School of Medicine, Boston, Massachusetts 02111

Humans are routinely exposed to bisphenol A (BPA), an es-trogenic chemical present in food and beverage containers,dental composites, and many products in the home and work-place. BPA binds both classical nuclear estrogen receptorsand facilitates membrane-initiated estrogenic effects. Herewe explore the ability of environmentally relevant exposureto BPA to affect anatomical and functional measures of braindevelopment and sexual differentiation. Anatomical evidenceof alterations in brain sexual differentiation were examinedin male and female offspring born to mouse dams exposed to0, 25, or 250 ng BPA/kg body weight per day from the eveningof d 8 of gestation through d 16 of lactation. These studiesexamined the sexually dimorphic population of tyrosine hy-droxylase (TH) neurons in the rostral periventricular preop-

tic area, an important brain region for estrous cyclicity andestrogen-positive feedback. The significant sex differences inTH neuron number observed in control offspring were dimin-ished or obliterated in offspring exposed to BPA primarilybecause of a decline in TH neuron number in BPA-exposedfemales. As a functional endpoint of BPA action on brain sex-ual differentiation, we examined the effects of perinatal BPAexposure on sexually dimorphic behaviors in the open field.Data from these studies revealed significant sex differences inthe vehicle-exposed offspring that were not observed in theBPA-exposed offspring. These data indicate that BPA may becapable of altering important events during critical periods ofbrain development. (Endocrinology 147: 3681–3691, 2006)

BISPHENOL A (BPA) is an environmental estrogenicchemical that has been attracting increased attention

because of its high potential for human exposure. BPA isused in the manufacture of polycarbonate plastics and epoxyresins, and it has been shown to leach from food and bev-erage containers (1, 2) and dental sealants and composites (3)under normal conditions of use (4). BPA is also used in themanufacture of many other products present in the homeand workplace (5) that would be expected to further increaseroutine exposure to this compound. Although assessments ofcurrent exposure levels are limited, several studies have nowconfirmed detectable levels of BPA in human populations. Ina recent study of a reference human population (394 sub-jects), BPA was detected in 95% of the urine samples ana-lyzed (6). Smaller studies have also reported detectable levelsof BPA in all or most individuals examined (7–9). Perhapsmost relevant to the studies to follow, BPA has been mea-sured in amniotic fluid, maternal and fetal plasma, and pla-cental tissue at birth (10, 11) and in breast milk of lactatingmothers (12). These data indicate the very real possibility thatthe developing human fetus and neonate may be exposed toBPA.

Results of animal studies in our laboratory and others

indicate that the developing fetus is more sensitive to BPArelative to the adult (13, 14), a finding that may reflect thelimited capacity of the fetus to metabolize BPA relative to theadult (10, 15–17). The exposure of pregnant female rodentsto low doses of BPA results in pleiotropic effects in theoffspring. Many of the defects that have been described occurlong after BPA exposure has ended. We have reported al-terations in body weight, mammary gland development,reproductive tract tissue, gonadotropin levels, and estrouscyclicity in offspring exposed perinatally to BPA (13, 14,18–20). Other laboratories have reported effects of prenatalexposure to low doses of BPA on the prostate, preputialglands, and epididymis in males (21, 22) and on the time ofvaginal opening and the time from vaginal opening to firstestrus in females (23). In addition, behavioral alterationshave been reported in offspring born to rodents treated withBPA during gestation and/or lactation (24–32). In the studiesto follow, we have examined the offspring of pregnant miceexposed from the evening of d 8 of pregnancy through d 16of lactation to vehicle or to one of two doses of BPA thatshould fall within the range of potential human environ-mental exposure (7).

BPA is an estrogen agonist that can bind to both classicalnuclear estrogen receptors (ERs), ER� and ER� (33, 34). Inaddition, data from recent studies have revealed that lowlevels of BPA can induce rapid, membrane-initiated estro-genic effects (35, 36), suggesting that low levels of BPA ex-posure might interfere with normal estrogenic signaling. Thedeveloping brain is exquisitely sensitive to estrogen andtherefore might be particularly vulnerable to xenoestrogenexposure. Various mechanisms facilitate precise temporal

First Published Online May 4, 2006Abbreviations: Afp, �-Fetoprotein; AVPV, anteroventral periven-

tricular preoptic area; BPA, bisphenol A; BST, bed nucleus of the striaterminalis; BW, body weight; ER, estrogen receptor; ERKO, estrogenreceptor knockout; TH, tyrosine hydroxylase.Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

0013-7227/06/$15.00/0 Endocrinology 147(8):3681–3691Printed in U.S.A. Copyright © 2006 by The Endocrine Society

doi: 10.1210/en.2006-0189

3681

and spatial control of estrogen action during critical periodsof brain development (37). These include the transient ex-pression of both estrogen receptors and aromatizing en-zymes that correspond to transient increases in testosteroneproduction in the developing male. In rodents and otheranimal models, numerous studies have documented thattestosterone secreted by the testes in the developing male isconverted to estradiol in situ by aromatase, which is presentin specific brain regions during critical periods of perinataldevelopment. This conversion of testosterone to estradiolplays an important role in the sexual differentiation of thebrain (for review see Refs. 37 and 38). Unlike the testes, theovaries of the developing female rodent are quiescent peri-natally (39). Recent data have verified the importance of�-fetoprotein (Afp) in protecting the brain of the female fetusfrom defeminization or masculinization by binding circulat-ing estradiol from the mother or neighboring male litter-mates (40). Afp, which is abundant during fetal and earlyneonatal life (41, 42), binds estradiol in rodents with highaffinity. Nonsteroidal estrogens, such as BPA, that exhibit alower affinity for plasma estrogen-binding proteins (43) maybe able to evade this protective mechanism.

In the present studies, we explore the ability of chroniclow-level perinatal exposure to BPA to affect anatomical andfunctional measures of brain development and sexual dif-ferentiation. First, we searched for anatomical evidence ofalterations in brain sexual differentiation. For these studies,we concentrated on a prominent sexually dimorphic regionof the brain in the rostral periventricular preoptic area. Theanteroventral periventricular preoptic area (AVPV) is im-portant for cyclic gonadotropin release and for the LH surgerequired for ovulation (44, 45). Neurons in this region containsexually dimorphic patterns of steroid receptor distribution(46) and peptide expression in rats (47–49). One robust sexdifference that has been observed in the AVPV of both ratsand mice is the sexually dimorphic population of tyrosinehydroxylase (TH) neurons (50–52). TH is the rate-limitingenzyme for dopamine synthesis. The number of TH-positiveneurons in the AVPV is significantly higher in female ratsand mice relative to males, and the sexual dimorphism of thispopulation of neurons appears to be dependent on perinatallevels of gonadal steroids (50, 51). Studies of estrogen re-ceptor knockout (ERKO) mice indicate that the significantdecline in TH neuron number in the male AVPV is depen-dent upon the presence of ER� (52). To determine whetherperinatal BPA exposure can alter a robust anatomical markerof brain sexual differentiation in mice, we compared THneuron number in tissue sections through the rostral periven-tricular preoptic area in male and female littermates born tomothers exposed to BPA or to vehicle only. Herein we reporteffects of perinatal BPA exposure on the expected sex dif-ferences in TH neuron number.

As a second and functional endpoint of BPA action onbrain sexual differentiation, we examined potential effects ofdevelopmental exposure to BPA on sexually dimorphic be-haviors in the open field. Strain-dependent sex differenceshave been reported in open-field behaviors in rats and mice(53, 54), and data from the majority of studies reveal higherlevels of activity in females relative to males (54–56). Al-though circulating hormones in adulthood can affect these

behaviors, differential exposure to gonadal steroids duringthe perinatal period plays an important role in the develop-ment of sexually dimorphic behaviors in the open field (55).More specifically, data from studies of male ERKO micesuggest that masculinization of open-field behavior requiresestrogen action during development (57). The results of thesebehavioral studies further suggest that early exposure to BPAmay alter important events during critical periods of braindevelopment in mice and indicate the need for careful studyof the potential effects of developmental exposure to thiscompound in humans.

Materials and MethodsAnimals

Animals for these studies were maintained in temperature- and light-controlled (14 h light, 10 h dark, lights on at 0400 h) conditions at theTufts–New England Medical Center Animal Facility. The cages andbedding used for housing tested negligible for estrogenicity using theE-SCREEN assay (58). The food (rodent diet 2018; Harlan Teklad, St.Louis, MO) was extracted and assayed as previously described (58), andit also tested negligible for estrogenicity (20 fmol estradiol equivalents/g). Water was supplied by glass bottles only. All experimental proce-dures were approved by the Tufts University-New England MedicalCenter Animal Research Committee in accordance with the Guide forCare and Use of Laboratory Animals.

To generate offspring for these studies, female and male CD-1 micewere purchased (Charles River Laboratories, Wilmington, MA) at 8–10wk of age. Female breeders were housed between cages of breeder malesfor a minimum of 1 wk after arrival. Two days before introduction of amale into the cage, females were exposed to male bedding to stimulateestrous cyclicity and ovulation. Males were placed in the female cages,and the morning on which a vaginal plug was observed was designatedd 1 of pregnancy. On the evening of d 8 of pregnancy, dams wereweighed and implanted sc with Alzet osmotic pumps (Alza Corp., PaloAlto, CA) that were designed to deliver either 50% dimethylsulfoxide(vehicle control) or BPA dissolved in 50% dimethylsulfoxide at the rateof 25 ng or 250 ng BPA/kg body weight (BW) per day (Sigma ChemicalCo., St. Louis, MO) throughout the remainder of the pregnancy andthrough d 16 of lactation. Dams were allowed to deliver naturally, andthe litters were culled to eight pups per mother (four males and fourfemales) 1 d after birth. Litters were weaned on postnatal d 22–24. Onlylitters with normal distributions of males and females were included inthese studies. The average number of pups and mean percentageof female pups in each litter examined in these studies are shown inTable 1.

Vehicle-exposed and BPA-exposed offspring for anatomical studieswere examined before puberty (22–24 d). For behavioral studies, off-spring were initially examined at 6–9 wk of age. However, because ofconcerns regarding potential effects of differences in circulating hor-mone levels in postpubertal animals, additional offspring were exam-ined before puberty (27–29 d of age). It should be noted that vaginalsmear records from our colony indicate that 6- to 9-wk-old group-housed females do not exhibit regular 4- to 5-d estrous cycles (Rubin,B. S., and A. M. Soto, unpublished data). This observation probablyreflects the importance of pheromones to the maintenance of regular

TABLE 1. Composition of the litters used for the studiesdescribed

BPAtreatment

Mean no. of pupsper litter

Proportion of femalepups per litter

Control 12.17 � 0.58 0.591 � 0.02625 ng 11.30 � 0.423 0.524 � 0.052250 ng 11.27 � 0.506 0.524 � 0.034

No significant differences were noted between any experimentalgroup relative to the controls. All litters were culled to eight pups (fourmales and four females) 1 d after birth.

3682 Endocrinology, August 2006, 147(8):3681–3691 Rubin et al. • BPA Alters Brain Sexual Differentiation

estrous cyclicity in mice, which is enhanced by exposure to males or maleurine and can be suppressed by exposure to other females (59–61).

TH neuron number

Tissue preparation. Two male and two female littermates from each litterwere killed at 22–24 d of age. Prepubertal animals were chosen forexamination because of evidence of a potential influence of circulatingsteroid levels on TH expression in adult animals (50). Animals wereinjected with an overdose of pentobarbital and perfused intraventricu-larly with heparinized saline followed by a solution containing 4%paraformaldehyde and 3.0% acrolein. Brains were removed from thecalvarium, placed into phosphate buffer, and stored at 4 C. Brains froma matched set of male and female littermates from seven to eight dif-ferent litters were analyzed for each of the three exposure levels (rangingfrom 0–250 ng BPA/kg BW/d) as detailed above.

Immunocytochemistry protocols

Brains were blocked, and 40-�m sections were cut in the coronal planeon a Vibratome (Technical Products International, St. Louis, MO). Sec-tions were collected beginning rostrally at the diagonal band of Brocaand continuing caudally through the arcuate nucleus. Consecutive brainsections were collected into two numbered tissue boats with each boatreceiving every other section through the region of interest, thus limitingthe number of tissue sections in each boat to facilitate antibody incu-

bations. Both tissue boats for each animal were treated identically so thatevery section through the areas of interest was available for analysis.After pretreatment to remove residual aldehydes, tissues were rinsed,and free-floating sections were incubated for 48 h at 4 C in anti-THmonoclonal antibody (MAB318; Chemicon, Temecula, CA) diluted at1:5000. Secondary antibody was biotin-SP-conjugated donkey anti-mouse (no. 715-066-151; Jackson ImmunoResearch, West Grove, PA).Detection was completed with Vectastain ABC reagent (PK-4000; VectorLaboratories, Burlingame, CA) and diaminobenzidine. A total of sevenchemistries were performed, and each immunocytochemistry run in-cluded tissues from matched male and female littermates from thedifferent treatment groups.

Data analysis

Tissue sections through the AVPV were identified using the MouseBrain Atlas of Paxinos and Franklin (62) as a guide to define the pa-rameters of the region of interest. The region analyzed corresponds tothat identified as the AVPV in Figures 26–30 in the atlas and is indicatedby the shaded areas in Fig. 1. As depicted, this region extends frominteraural 4.42 mm to interaural 3.94 mm, with the consecutive sectionsexamined beginning rostrally at the level of the organum vasculosum ofthe lamina terminalis and extending caudally to the level just before thecrossing of the anterior commissure. TH neurons were identified usinga Zeiss Axioscope (�10 and �40 objectives), and counted by threedifferent observers blind to the sex and treatment groups of the animals.

FIG. 1. Sections through the rostral preoptic area. The drawings depicted were adapted from the Mouse Brain Atlas by Paxinos and Franklin(62), and the AVPV, as defined by the atlas, appears shaded. As shown, these drawings span the region extending from interaural 4.42 mmrostrally through interaural 3.94 mm caudally. Actual tissue sections through most (nine of 11) of the rostral-caudal extent of the AVPV fromone representative female in the study are also shown below the adapted atlas diagrams. The tissue sections depicted here span an areabeginning in the rostral AVPV through the region before the crossing of the anterior commissure. 3V, Third ventricle; aca, anterior commissure;VOLT, organum vasculosum of the lamina terminalis; MPA, medial preoptic area; MnPO, medial preoptic nucleus.

Rubin et al. • BPA Alters Brain Sexual Differentiation Endocrinology, August 2006, 147(8):3681–3691 3683

Because the total number of sections through the AVPV was not identicalfor all animals in the study, TH neuron number was also assessed inseven consecutive sections extending from the caudal AVPV (rostral tothe level of the crossing of the anterior commissure) through the mid-region of the AVPV rostrally in each animal. TH neuron number was alsoassessed in two matched sections through the arcuate nucleus of thehypothalamus.

Behavioral studies

Open-field test at 6–9 wk of age. At 6–9 wk of age, male and femaleoffspring of pregnant dams treated with 0, 25, or 250 ng BPA/kg BW/dwere observed in a novel open field. The open-field apparatus was alarge plastic tub measuring 16 � 24 inches with a wall height of 11 inches.The floor of the tub was divided into squares to facilitate behavioralmeasurements. The apparatus was carefully cleaned with ethanol,rinsed with water, and then dried after each behavioral test. Two iden-tical open fields were available for testing and were rotated during thetesting period. All behavioral tests were conducted in the same room atthe same time of day (1200–1400 h) and were scored by three observers.Before the start of data collection, 3 d of behavioral tests were conductedwith an additional cohort of animals to familiarize the observers with thebehaviors to be recorded and to standardize the scoring of behaviors. Atotal of 94 animals were tested in the open field, including 14–17 malesand females from each of the three treatment groups. Only a single maleand female from each litter were examined in these studies to eliminatepotential litter effects. At the start of each test, the mouse was carefullyplaced into the center of the open field and his/her movements wererecorded over the next 5 min. Measurements included number of rears,time spent in center, time stopped, time grooming, and number of fecalpellets. Initial attempts to score the number of squares entered to assessdistance traveled in a given testing period were abandoned because theanimals moved too quickly to obtain accurate numbers without digitalrecordings.

Open-field tests in prepubertal animals. Additional open-field tests wereperformed in younger prepubertal animals at 27–29 d of age. Theseanimals were examined before vaginal opening and before the estab-lishment of adult gonadal hormone levels. For these studies, controloffspring (n � 10 animals per sex) and offspring born to females treatedwith 250 ng BPA/kg BW/d (n � 12 animals per sex) were examined.Animals were tested in the open field as described above with onechange to the protocol. Because the prepubertal animals were veryactive, we chose to detain each subject in the center of the open fieldcovered by the transfer beaker for a period of 5 sec before the start ofthe test to facilitate accurate scoring of behaviors. This method has beenused by other laboratories. Unfortunately, this procedural change mayhave influenced the time-in-center measurement such that this param-eter was not comparable in the two age groups examined.

Statistics

The data were analyzed using both parametric and nonparametricstatistics. Both showed similar significant differences. The parametricanalyses are presented, and the data are graphed as mean � sem. Theanatomical data and the behavioral data from the 6- to 9-wk animalswere analyzed by two-way ANOVA with sex and BPA as the twoparameters. Planned comparisons of male and female offspring fromeach treatment group were evaluated by t tests, and planned same-sexcomparisons across treatment groups were assessed by ANOVA fol-lowed by Bonferroni post hoc tests. Because the behavioral data for theprepubertal animals contained only a single BPA dose, planned com-parisons of male and female offspring in the control and BPA-exposedgroup and same-sex comparisons across the two treatment groups wereanalyzed by t tests.

ResultsTH neuron number in control and BPA-exposed offspring

Mean number of tissue sections through the AVPV. Comparisonsof the total number of tissue sections through the AVPV(based on the delineation of this region in the Paxinos and

Franklin atlas and independent of the presence or absence ofTH neurons) by two-way ANOVA revealed overall signifi-cance (P � 0.039) and a significant interaction between vari-ables (sex � BPA, P � 0.014). The number of tissue sectionsdiffered significantly in male and female offspring born tovehicle-treated mothers (P � 0.003, t test; see Fig. 2). Tissuesection number did not differ significantly in male and fe-male offspring from either of the BPA-exposed groups. Thesedata are consistent with an increase in the rostral-caudalextent of the AVPV in control females relative to controlmales and a loss of this sex difference in offspring of BPA-treated dams. Comparison of tissue section number by sexacross treatment groups revealed a difference in females (P �0.020, ANOVA), and post hoc analysis revealed significantdifferences between control females and those exposed to 250ng BPA (P � 0.023, Bonferroni).

TH-positive cell number. Analysis (two-way ANOVA) of totalTH cell number revealed an overall significance (P � 0.001)and differences by sex (P � 0.001), BPA (P � 0.020), andBPA � sex (P � 0.014). Subsequent analyses (t test) revealedthat the total number of TH-positive neurons counted insections through the AVPV differ in male and female off-spring born to vehicle-treated mothers (P � 0.001; controlmale TH neuron number � 46% of control females; see Figs.3 and 4) and to mothers exposed to the lowest dose of BPA,25 ng/kg BW/d (P � 0.024; male TH neuron number � 65%of females). No significant sex differences were observed inoffspring exposed to 250 ng BPA/kg BW/d (male TH neuronnumber � 80% of females). Moreover, comparison of femalesacross treatment groups was significant (P � 0.005,ANOVA), and post hoc analysis revealed that females ex-posed to the higher dose of BPA exhibited a significant de-crease in the total number of TH-positive cells relative tocontrol females (P � 0.004, Bonferroni). Females exposed tothe low dose of BPA revealed a 24% decrease in TH neuronnumber and those exposed to the higher dose revealed a 41%

FIG. 2. Number of sections through the AVPV. The average numberof 40-�m tissue sections through the AVPV are shown here for eachtreatment group. Solid bars represent mean (�SEM) female values(n � 7–8 per treatment), and gray bars represent mean (�SEM) malevalues (n � 7–8 per treatment). A significant sex difference in sectionnumber is observed only in control offspring (**, P � 0.003, female vs.male), and the number of sections were decreased in female offspringexposed to the higher dose of BPA relative to controls (*, P � 0.020).The x-axis label refers to the level of BPA exposure of the mothers (perkg BW per day).

3684 Endocrinology, August 2006, 147(8):3681–3691 Rubin et al. • BPA Alters Brain Sexual Differentiation

decline in TH neuron number relative to controls. No sig-nificant differences in TH cell number were noted acrosstreatment groups in the males.

Measurements of total cell counts included assessment ofTH neurons in every section through the AVPV in eachanimal and therefore did not include equal numbers of sec-tions from every brain. When TH-positive cell counts wererestricted to seven consecutive sections through the AVPV ofeach brain (beginning caudally just before the crossing of theanterior commissure and extending rostrally through themidregion of the AVPV), analysis (two-way ANOVA) re-vealed overall significance (P � 0.001) and significant dif-ferences by sex (P � 0.001) and by BPA (P � 0.032). Addi-tional analyses revealed significant sex differences in control

offspring (P � 0.002, t test), but not in the offspring born tomothers treated with either dose of BPA (Fig. 5A). TH neuronnumber differed significantly across treatment groups in fe-males (P � 0.011, ANOVA) and more specifically betweenfemales in the control and 250-ng groups (P � 0.009,Bonferroni).

When the data from these seven sections was further di-vided to assess cell counts in four sections through the medialaspect of the AVPV (Fig. 5B) and three sections through themore caudal aspect of the AVPV (Fig. 5C), it appeared thatBPA exerted a more profound influence on TH-positive neu-ron number in the medial part of the nucleus (overall P �0.004; sex P � 0.026; BPA P � 0.033; and sex � BPA P � 0.040,two-way ANOVA). A robust sex difference in TH neuronnumber was noted in the mid-AVPV in the control group(P � 0.001, t test) but not in either of the BPA-exposed groups.Same-sex comparisons across treatment groups revealed asignificant difference in the females (P � 0.012, ANOVA),and post hoc tests revealed that TH neuron number wassignificantly decreased in females born to mothers treatedwith the 250-ng dose of BPA relative to controls (P � 0.010,Bonferroni). Comparisons of TH-positive neuron number inthe more caudal aspect of the AVPV revealed overall sig-nificance and differences by sex (P � 0.001, two-wayANOVA). Significant sex differences were present in controloffspring (P � 0.001, t test) and offspring exposed to thelower dose of BPA (P � 0.005). The sex difference in animalsexposed to 250 ng BPA/kg BW/d approached significance(P � 0.059).

TH-positive neurons in the arcuate nucleus. No significant sexdifferences in the number of TH-positive cells per sectionwere observed in the arcuate nucleus of control or BPA-exposed offspring. There were also no significant differencesin cell number in BPA-exposed relative to control offspring(see Fig. 6).

Behavioral tests

Open-field tests in 6- to 9-wk-old animals. Male and femaleoffspring born to mothers exposed to 0, 25, or 250 ng BPA/kg

FIG. 3. Photomicrographs of sections through the rostral periven-tricular preoptic area of control and BPA-exposed mice. TH-positiveneurons are shown in representative sections through the mid AVPV(A) and through the caudal AVPV (B) of female and male offspringborn to control dams and born to dams treated with 250 ng BPA/kgBW/d. III V, Third ventricle; OC, optic chiasm. Bar, 100 �m.

FIG. 4. Total number of TH-positive neurons in sections though the AVPV. A, Data are shown for offspring born to control dams and offspringborn to dams treated with two doses of BPA (mean � SEM). Significant sex differences are noted in TH neuron number in control offspring (***,P � 0.001) as well as offspring born to dams treated with the lowest dose of BPA (*, P � 0.024). A significant decline in TH neuron numberis noted in 250-ng females relative to controls (**, P � 0.004). Black bars, females (n � 7–8 per treatment); gray bars, males (n � 7–8 pertreatment). B, The mean female to male ratio of the total TH-positive neuron number was calculated for each pair of littermates examined.As shown, control females have approximately twice the number of TH-positive neurons in the AVPV relative to control males, and the femaleto male ratio is markedly reduced with exposure to the higher dose of BPA. The x-axis label refers to the level of BPA exposure of the mothers(per kg BW per day).

Rubin et al. • BPA Alters Brain Sexual Differentiation Endocrinology, August 2006, 147(8):3681–3691 3685

BW/d were examined in the open field at 6–9 wk of age.Analysis by two-way ANOVA revealed overall significancein three behavioral parameters, and in each case, only themain effect of sex was significant: rears on walls (overall P �0.043; sex, P � 0.009), time stopped (overall P � 0.012; sex P �0.001), and time in center (overall P � 0. 007; sex P � 0.001;Fig. 7). Additional analysis using planned t tests revealedthat male and female offspring born to control dams showeda significant sex difference in all of these parameters (Prears

� 0.023; Pcenter � 0.005; Pstopped � 0.014). Offspring born toBPA-exposed dams showed no sex difference in these pa-rameters with one exception, time stopped in offspring ex-posed to the higher dose of BPA (P � 0.023). Although thevalues for BPA-exposed females drifted toward control malevalues in two parameters (rears and center), no significantdifferences were revealed by ANOVA in comparisons offemales across the three treatment groups.

Behavioral measurements in prepubertal animals. To rule out thepossibility that the behavioral differences observed in controland BPA-exposed offspring at 6–9 wk of age resulted frompotential differences in circulating hormone levels, prepu-bertal offspring born to dams exposed to vehicle or 250 ngBPA/kg BW/d were examined in the open field. Two be-havioral measurements, rearing behavior and time stopped,revealed significant sex differences in the control animals(Prears � 0.010; Pstopped � 0.048, t tests) and not in the BPA-exposed offspring (Fig. 8). BPA-exposed females showed asignificant decline in the number of rears relative to controlfemales (P � 0.028, t test), which is consistent with a mas-culinization of this behavior by BPA. The difference in num-ber of rears in control and BPA-exposed males was not sta-tistically significant. In contrast to the behavioral data fromthe 6- to 9-wk-old animals, sex differences were not observedin the time spent in the center of the open field in control orBPA-exposed prepubertal animals. This measure might havebeen influenced by the procedural modification at the startof the test. For the older animals, the test period began themoment they were placed in the center of the open field. Asmentioned previously, because of the increased activity ofthe younger animals, they were detained in the center of theopen field for a period of 5 sec before the start of the test. Thisprocedural change may have caused some animals to remainin the center of the open field at the start of the test, thuschanging the measure from that in the older animals. Alter-natively, it is possible that the sex difference in center timeemerges later in development. Early assessments of open-field behavior in rats revealed that significant sex differencesin activity levels were apparent in behavioral tests conductedat 52 d of age but not in tests conducted 11 d earlier (55). Also,data from behavioral studies in periadolescent mice (33–43d old) revealed elevated novelty-seeking behaviors in thisage group (63) that might have influenced the time-in-centermeasurement in the younger animals.

DiscussionPerinatal exposure to low levels of BPA alters a marker ofbrain sexual differentiation

The difference in TH neuron number in the AVPV repre-sents one robust anatomical marker of brain sexual differ-entiation that has been documented in rats and mice (50–52).Sexual dimorphism in the population of TH neurons in theAVPV appears to result from the influence of sex steroidsduring the perinatal period of development (50, 64) and isreportedly observed in rat pups by postnatal d 10 (as re-viewed in Ref. 38). Perinatal and/or postnatal exposure totestosterone or estradiol is effective in reducing TH neuronnumber in the AVPV (50, 64). ERKO mice that lack ER� do

FIG. 5. TH neuron number in seven consecutive sections through theAVPV. A, Tissue sections extend caudally from the level just beforethe crossing of the anterior commissure to the midregion of the AVPVrostrally. B and C, The rostral caudal extent of the area is furtherdivided to reveal TH neuron number in four sections through the midAVPV (B) and three sections through the caudal AVPV (C). Blackbars, females (n � 7–8 per treatment); gray bars, males (n � 7–8 pertreatment). Data are depicted as mean � SEM. The x-axis label refersto the level of BPA exposure of the mothers (per kg BW per day). ***,P � 0.0025; **, P � 0.01; *, P � 0.01; a, P � 0.06.

3686 Endocrinology, August 2006, 147(8):3681–3691 Rubin et al. • BPA Alters Brain Sexual Differentiation

not show the sex difference in TH neuron number; however,mice that lack a functioning androgen receptor (Tfm, testic-ular feminization) do maintain the sexual dimorphism in THneuron number (52). Therefore, estrogen action through ER�has been postulated to be important for the significant de-cline in TH neuron number in males relative to females.

As expected, the total number of TH-positive neurons insections through the rostral-caudal extent of the AVPV wassignificantly higher in control females relative to males. Off-spring born to mothers exposed perinatally to the low doseof BPA (25 ng/kg BW/d) also revealed significant sex dif-ferences in total TH neuron number. In contrast, offspringborn to mothers exposed to a 10-fold higher dose of BPA (250ng/kg BW/d) failed to show significant differences in totalTH-positive neuron number. When the analysis of TH neu-ron number was restricted to seven consecutive sectionsextending from the caudal AVPV through the midregion ofthe nucleus in all subjects, only the control animals showeda significant sex difference. Additional analysis of the datasuggested that TH neurons in the midregion of the AVPVmay be particularly vulnerable to perinatal BPA exposure.The loss of sex differences in TH neuron number in BPA-exposed offspring can be attributed primarily to a decreasein TH neuron number in female offspring, which would beconsistent with BPA’s actions as an estrogen.

The total number of tissue sections spanning the rostral-caudal extent of the AVPV was greater in female offspringborn to vehicle-treated mothers relative to their male litter-mates. This finding is consistent with previous reports ofincreased size in the volume of the AVPV in female relative

to male rodents (65). No significant sex differences in therostral-caudal extent of the AVPV were noted in either BPAtreatment group. Moreover, females exposed to the higherdose of BPA exhibited a significant decline in section numberrelative to control females. These data suggest that perinatalBPA exposure may decrease the AVPV volume in the femalebrain, making it more similar to that observed in the male.

Different mechanisms appear to be responsible for thesexual dimorphism in AVPV volume and the sex differencein AVPV TH neuron number, and BPA exposure may in-fluence both. Studies investigating the involvement of apop-totic genes in sexual differentiation of the nervous systemhave revealed that overexpression of the cell survival geneBcl-2 (66) and deletion of the cell death gene BAX (67) de-creased cell death in the AVPV and obliterated sex differ-ences in AVPV volume in mice. However, neither manipu-lation altered the sexual dimorphism in TH neuron numberin this nucleus (66, 67). It is possible that other pathways ofcell death that do not involve the Bcl-2 family of proteins maybe responsible for sex differences in the number of dopamineneurons in the AVPV. Alternatively, developmental expo-sure to gonadal steroids may influence the differentiation ofthe dopaminergic cell phenotype in this nucleus. Such amechanism has been proposed to explain the sexually di-morphic expression of arginine vasopressin in galanin neu-rons of the principal nucleus of the bed nucleus of the striaterminalis (BST) (68). Like the AVPV, the sex differences incell number in the BST can be attributed to differential celldeath (66, 69). In contrast, the marked sex difference in ar-ginine vasopressin expression in this nucleus has been pro-

FIG. 6. TH-positive neurons in the arcuate nucleus. A, Representative sections through the arcuate nucleus of a control female and male showingthe distribution of TH neurons. B, Mean values for TH neuron number per section in two sections through the arcuate nucleus of offspring fromall treatment groups. As shown, no significant sex differences in TH neuron number were revealed in the arcuate nucleus for any treatmentgroup examined. Black bars, females (n � 4 per treatment); gray bars, males (n � 4 per treatment). The x-axis label refers to the level of BPAexposure of the mothers (per kg BW per day).

FIG. 7. Results of behavioral tests conducted in the open field at 6–9 wk of age. Analysis by two-way ANOVA revealed overall significance inthree behavioral measurements in the open field. All three revealed significant sex differences in vehicle-exposed offspring. Data are shownhere for male and female offspring born to dams exposed to vehicle, 25, or 250 ng BPA/kg BW/d. A, Number of rears at the wall; B, time incenter of the open field; C, time stopped. Black bars, females (n � 14–17 per treatment); gray bars, males (n � 14–17 per treatment). The x-axislabel refers to the level of BPA exposure of the mothers (per kg BW per day). *, P � 0.023; **, P � 0.014; ***, P � 0.005.

Rubin et al. • BPA Alters Brain Sexual Differentiation Endocrinology, August 2006, 147(8):3681–3691 3687

posed to result from the ability of galanin cells to alter theirneuronal phenotype in response to perinatal testosteroneexposure (68). Whether this sexually dimorphic expression ofarginine vasopressin by galanin cells is a result of target-dependent mechanisms that may be explained by markeddifferences in the projections of the BST in males and femalesremains to be determined.

Significance of the AVPV

The AVPV is essential for the cyclic pattern of gonado-tropin release and the generation of the preovulatory LHsurge required for ovulation. Lesions of the AVPV abolishedspontaneous LH surges and induced persistent vaginal es-trus in female rats (44, 45), and antiestrogen implants into thisregion blocked the steroid-induced LH surge (70). The dis-ruption of normal development and sexual differentiation ofthis brain region by BPA could contribute to the alterationsin estrous cyclicity in adulthood that has been observed inboth mice and rats exposed perinatally to this environmentalcontaminant (13, 20). A subset of neurons in the AVPVproject to GnRH neurons in the rostral preoptic area that arethought to be involved in the generation of the LH surge(71–73). Moreover, data from electron microscopy studieshave documented synaptic contact between GnRH neuronsin the preoptic area and TH-containing axon terminals (74),and there is evidence that the TH fibers that contact GnRHneurons originate from dopamine neurons in the AVPV (75).Although the role of dopamine in the preovulatory LH surgeis still not understood, these observations suggest a possiblerole for TH neurons in the AVPV in the regulation of go-nadotropin release.

BPA exposure and other anatomical markers of brainsexual differentiation

To date, three studies have reported evidence of alter-ations in other anatomical markers of brain sexual differen-tiation after exposure to BPA during pregnancy and lacta-tion. The expected sex difference in the volume of the locusceruleus was abolished in offspring born to rat dams exposedto 1.5 mg BPA/kg BW/d in their drinking water (30), and thesex difference in the volume of this nucleus was reportedlyinverted in offspring born to dams exposed to 30 or 300 �g

BPA/kg BW/d (31). Sex differences in the number of CRHneurons in subdivisions of the BST were abolished in off-spring born to mothers exposed to 2.5 mg BPA/kg BW/d viadrinking water (76). Only one dose of BPA was examined inthis study, and therefore it is not possible to assess the sen-sitivity of this parameter to perinatal BPA exposure. It shouldbe noted that all of the results discussed above were obtainedwith exposure levels of BPA that were significantly higherthan those assessed in the present study.

Behavioral data reveal a lasting effect of BPA exposure onthe developing brain

Strain-dependent sex differences have been reported inopen-field behaviors in rats and mice (53, 54), and data fromthe majority of studies reveal higher levels of activity infemales relative to males (54–56). Although circulating hor-mones in adulthood can affect these behaviors, differentialexposure to gonadal steroids during the perinatal periodplays an important role in the development of sexually di-morphic behaviors in the open field (55). Results of earlybehavioral studies revealed that neonatal administration oftestosterone to female rats resulted in male-like behaviors inthe open field (55). Data from later studies showed that maleERKO mice exhibited increased rearing and increased centercrossings in the open field relative to wild-type males. Thesedata suggest that ER gene disruption demasculinized be-havior in the open field or that estrogen action during de-velopment is important for masculinization of these behav-iors (57). These findings are supported by results of a studyin which antisense oligonucleotides were used to knockdown ER expression in neonatal females. In this study, fe-males that received the antisense oligonucleotides to ERshowed an elevation of female-associated behaviors in theopen field (77). Data from another test of anxiety and emo-tionality, the elevated plus maze, have revealed that neonatalcastration of male rats results in female-like patterns of be-havior, further suggesting a role for perinatal exposure togonadal steroids in sexually dimorphic behaviors in a novelenvironment (78).

The behavioral data in the open field reported here areconsistent with the idea that perinatal exposure to BPA mayact to alter brain development and brain sexual differentia-tion. As reported, at 27–29 d of age, some behavioral pa-rameters in the open field revealed significant sex differencesin offspring born to control mothers that were not seen inoffspring born to mothers treated with 250 ng BPA/kgBW/d. Measurements of rearing behavior suggest that BPAexposure may have masculinized this behavior in the pre-pubertal female mice. Behavioral measurements in 6- to9-wk-old animals also revealed sex differences in controloffspring that were not observed in offspring born to motherstreated with 25 or 250 ng BPA/kg BW/d, indicating that thebehavioral differences persist beyond weaning. Althoughthe differences were not statistically significant, animals inthe older age group showed a trend consistent with mascu-linization of rearing behavior in the BPA-exposed females.

No attempt was made to monitor estrous cycles in theseyoung females or to assess gonadal steroid levels in eithersex; however, a previous study in our lab found no signif-

FIG. 8. Results of behavioral tests conducted in the open field at 4 wkof age. Two measurements in the open field revealed significant sexdifferences in vehicle-exposed offspring. These significant sex differ-ences were not observed in male and female offspring born to damsexposed to 250 ng BPA/kg BW/d. A, Number of rears at the wall; B,time stopped. Black bars, females (n � 10–12 per treatment); graybars, males (n � 10–12 per treatment). The x-axis label refers to thelevel of BPA exposure of the mothers (per kg BW per day). *, P � 0.05;**, P � 0.025; ***, P � 0.01.

3688 Endocrinology, August 2006, 147(8):3681–3691 Rubin et al. • BPA Alters Brain Sexual Differentiation

icant effects of perinatal BPA exposure on estradiol levels onthe day of the first proestrus (19). As mentioned previously,vaginal smear records of 6- to 9-wk-old females in our colonydo not reveal regular patterns of 4- to 5-d estrous cycles,which is probably attributed, in part, to the absence of malepheromones and the presence of female cagemates (re-viewed in Ref. 59). Although we cannot rule out the possi-bility that differences in circulating hormone levels may haveinfluenced behavioral measures in the older animals, it isunlikely that they could solely account for the loss of sexdifferences in the BPA-exposed offspring; behavioral tests inprepubertal animals also revealed significant sex differencesin control offspring that were not observed in BPA-exposedoffspring. Given the evidence that neonatal estrogen actionthrough ERs may be important for expression of sexuallydimorphic behaviors in the open field, it is possible that theestrogenic actions of BPA are responsible for some of thechanges in the behavioral parameters measured.

To date, behavioral studies have examined offspring ex-posed to significantly higher levels of BPA than those usedhere (greater than 100-fold), and alterations of both sex-dependent and sex-independent behaviors have been re-ported in rats (25, 26, 28, 30, 31). Data from these studiesreveal the ability of perinatal exposure to BPA to exert com-plex, multifaceted developmental effects on behavior inadulthood, and not all of the effects described can be attrib-uted to BPA’s action as an estrogen. It is possible that higherlevels of BPA exposure can exert different or additional ef-fects on behavioral development. Dose-response curves forestrogenic compounds, including BPA, are complex andnonmonotonic for some endpoints (79).

The importance of protecting the developing female brainfrom circulating estrogens

Afp has been identified in many vertebrate species. It isabundant during fetal life and decreases in abundance afterbirth (41, 42). In rodents, Afp binds estradiol with high af-finity and high capacity. Recent data from studies of Afpknockout mice (Afp�/�) lend support to the idea that Afpprotects the developing female mouse brain from masculin-ization and defeminization by binding circulating estradiolthat may reach the female fetus from the mother or fromneighboring male littermates (40). Of particular relevance tothe studies presented here, the number of TH neurons in theAVPV of Afp�/� females was significantly reduced relativeto wild-type females and did not differ from wild-type males.Also relevant was evidence of masculinization and defemi-nization of mating behaviors in the Afp�/� animals. Treat-ment of the pregnant dams with an aromatase inhibitor todecrease estradiol levels during pregnancy prevented themasculinization of TH neuron number and behavior inAfp�/� females. These data indicate the importance of pro-tecting specific regions of the female mouse brain from es-trogen during early development. Relative to estradiol, BPAshows a lower affinity for plasma binding proteins (43),including Afp, and therefore Afp would be expected to pro-vide little protection in shielding areas of the developingbrain from exposure to nonsteroidal estrogens like BPA. Asa result, BPA and other environmental estrogens might be

able to enter regions of the developing female brain that arenormally protected from excessive estrogen exposure, andonce there, they could alter the expected developmental plan.

Levels of BPA exposure in our animals

In the present study, pregnant mice were implanted withosmotic pumps that chronically released BPA at doses of 25or 250 ng BPA/kg BW/d. To our knowledge, the BPA ex-posures used in this study are among the lowest examinedto date, and the lowest doses that have shown significanteffects after perinatal administration. Data from a recentstudy that used the very conservative assumption that levelsof BPA in urine represent the total amount ingested (7) ledto estimates of a maximum daily intake of 0.23 �g BPA/kg/din the individuals studied. This dose is similar to the highBPA dose used in our experiments.

BPA has been shown to cross the placental barrier in miceand rats (15, 80, 81). Although we do not know the precisedose of BPA reaching the developing fetuses in our studies,Zalko et al. (80) reported that 24 h after a single sc injectionof 25 �g tritiated BPA/kg BW to pregnant mice on d 17 ofgestation, approximately 0.4% of the radioactivity adminis-tered to the dams was recovered in the uterus (correspondingto �3.45 ng/g [3H]BPA equivalents), 0.3% in the amnioticfluid (�4.85 ng/ml [3H]BPA equivalents), 0.6% in the pla-centa (�3.14 ng/g [3H]BPA equivalents), and 4.1% waspresent in the entire litter of fetuses (1.23 ng/g [3H]BPAequivalents per fetus). These estimates were obtained after asingle acute injection of a 100- to 1000-fold higher dose ofBPA than the daily doses chronically administered in thepresent study. Therefore it is not possible to extrapolate thelevels of BPA reaching our animals from these measure-ments. However, it is interesting to note that the levels es-timated in the Zalko study (80) appear to fall within the rangeof BPA exposure that has been reported in the human fetal-placental unit: 8.3 ng/ml in amniotic fluid at 15–18 wk ofgestation, 0.3–18.9 ng/ml in maternal plasma, 0.2–9.2 ng/mlin fetal plasma, and 1.0–104 ng/g in the placenta (10, 11, 82).After parturition, neonatal animals in our study continued toreceive exposure to BPA via lactation. Although no mea-surements are yet available in mice, data from a study in ratsshowed that 0.003% of an initial high dose of radiolabeledBPA (100 mg BPA/kg BW) was recovered in the lactatingdam’s milk 1 h after administration (corresponding to �1.0�g equivalent/ml) (81). Recent assessments of BPA in humanbreast milk have revealed mean levels of 0.61 ng BPA/ml(range, 0.28–0.97 ng/ml) in samples from 23 lactating moth-ers (12).

Summary

Exposure of pregnant female mice to low levels of BPAfrom the evening of d 8 of pregnancy through d 16 of lactationresults in lasting effects on the brain of the offspring. The datapresented reveal alterations in sexually dimorphic anatom-ical and behavioral endpoints assessed in offspring of moth-ers exposed to BPA. Although the mechanisms involved inBPA’s actions on the developing brain cannot be delineatedfrom the data presented, we have hypothesized that some ofthe observed effects may be related to the estrogenic activity

Rubin et al. • BPA Alters Brain Sexual Differentiation Endocrinology, August 2006, 147(8):3681–3691 3689

of BPA. Data from studies of ERKO mice have suggested arole for estrogen action during development in the mascu-linization of TH neuron number (52) and open-field behav-iors (57). However, it must be recognized that in addition toits well-documented estrogenicity, BPA may exert other ef-fects on the developing brain. Because of the paucity ofavailable information, it would be premature to speculateabout the potential role of putative nonestrogenic effects ofBPA at this time. The fact that exposure to low environmen-tally relevant levels of BPA results in measurable effectsshould be of considerable concern, particularly if one con-siders that BPA represents only one of many potential en-docrine disruptors in the environment to which humans maybe exposed daily.

Acknowledgments

We acknowledge the expert technical assistance of Dreux Chappelland Leslie McGowan, and we thank Dr. David Damassa for his assis-tance with the statistical analysis.

Received February 14, 2006. Accepted April 19, 2006.Address all correspondence and requests for reprints to: Beverly S.

Rubin, Department of Anatomy and Cellular Biology, 136 HarrisonAvenue, Boston, Massachusetts 02111. E-mail: beverly.rubin@tufts.edu.

This work was supported by Grant ES 08314 from National Instituteof Environmental Health Sciences.

With regard to the disclosure of potential conflicts of interest, all ofthe authors of this manuscript (B.S.R., J.R.L., C.M.S., L.N.V., P.M.R., andA.M.S.) have nothing to declare.

References

1. Brotons JA, Olea-Serrano MF, Villalobos M, Olea N 1994 Xenoestrogensreleased from lacquer coating in food cans. Environ Health Perspect 103:608–612

2. Biles JE, McNeal TP, Begley TH, Hollifield HC 1997 Determination of bis-phenol-A in reusable polycarbonate food-contact plastics and migration tofood simulating liquids. J Agric Food Chem 45:3541–3544

3. Olea N, Pulgar R, Perez P, Olea-Serrano F, Rivas A, Novillo-Fertrell A,Pedraza V, Soto AM, Sonnenschein C 1996 Estrogenicity of resin-based com-posites and sealants used in dentistry. Environ Health Perspect 104:298–305

4. Markey CM, Rubin BS, Soto AM, Sonnenschein C 2003 Endocrine disruptorsfrom Wingspread to environmental developmental biology. J Steroid BiochemMol Biol 83:235–244

5. Markey CM, Michaelson CL, Sonnenschein C, Soto AM 2001 Alkylphenolsand bisphenol A as environmental estrogens. In: Metzler M, ed. The handbookof environmental chemistry. Vol. 3. Part L, Endocrine Disruptors - Part I. Berlinand Heidelberg: Springer Verlag; 129–153

6. Calafat AM, Kuklenyik Z, Reidy JA, Caudill SP, Ekong J, Needham JL 2005Urinary concentrations of bisphenol A and 4-nonylphenol in a human refer-ence population. Environ Health Perspect 113:391–395

7. Arakawa C, Fujimaki K, Yoshinaga J, Imai H, Serizawa S, Shiraishi H 2004Daily urinary excretion of bisphenol A. Environ Health Prevent Med 9:22–26

8. Takeuchi T, Tsutsumi O 2002 Serum bisphenol A concentrations showedgender differences, possibly linked to androgen levels. Biochem Biophys ResCommun 291:76–78

9. Fujimaki K, Arakawa C, Yoshinaga J, Watanabe C, Serizawa S, Imai H,Shiraishi H, Mizumoto Y 2004 Estimation of intake level of bisphenol A inJapanese pregnant women based on measurements of urinary excretion levelof the metabolite. Nippon Eiseigaku Zasshi 59:403–408

10. Schonfelder G, Wittfoht W, Hopp H, Talsness CE, Paul M, Chahoud I 2002Parent bisphenol A accumulation in the human maternal-fetal-placental unit.Environ Health Perspect 110:A703–A707

11. Ikezuki Y, Tsutsumi O, Takai Y, Kamei Y, Taketani Y 2002 Determination ofbisphenol A concentrations in human biological fluids reveals significant earlyprenatal exposure. Hum Reprod 17:2839–2841

12. Sun Y, Irie M, Kishikawa N, Wada M, Kuroda N, Nakashima K 2004 De-termination of bisphenol A in human breast milk by HPLC with column-switching and fluorescence detection. Biomed Chromatogr 18:501–507

13. Rubin BS, Murray MK, Damassa DA, King JC, Soto AM 2001 Perinatalexposure to low doses of bisphenol-A affects body weight, patterns of estrouscyclicity and plasma LH levels. Environ Health Perspect 109:675–680

14. Markey CM, Wadia PR, Rubin BS, Sonnenschein C, Soto AM 2005 Long-

term effects of fetal exposure to low doses of the xenoestrogen bisphenol-A inthe female mouse genital tract. Biol Reprod 72:1344–1351

15. Takahashi O, Oishi S 2000 Disposition of orally administered 2,2-bis(4-hy-droxyphenyl) propane (bisphenol A) in pregnant rats and placental transfer tofetuses. Environ Health Perspect 108:931–935

16. Coughtrie MW, Burchell B, Leakey JE, Hume R 1988 The inadequacy ofperinatal glucuronidation: immunoblot analysis of the developmental expres-sion of individual UDP-glucuronosyltransferase isoenzymes in rat and humanliver microsomes. Mol Pharmacol 34:729–735

17. Matsumoto J, Yokota H, Yuasa 2002 A Developmental increases in rat hepaticmicrosomal UDP-glucuronosyltransferase activities toward xenoestrogensand decreases during pregnancy. Environ Health Perspect 110:193–196

18. Markey CM, Luque EH, Munoz de Toro MM, Sonnenschein C, Soto AM 2001In utero exposure to bisphenol A alters the development and tissue organi-zation of the mouse mammary gland. Biol Reprod 65:1215–1223

19. Munoz de Toro MM, Markey CM, Wadia PR, Luque EH, Rubin BS, Son-nenschein C, Soto AM 2005 Perinatal exposure to Bisphenol A alters peripu-bertal mammary gland development in mice. Endocrinology 146:4138–4147

20. Markey CM, Coombs MA, Sonnenschein C, Soto AM 2003 Mammaliandevelopment in a changing environment: exposure to endocrine disruptorsreveals the developmental plasticity of steroid-hormone target organs. EvolDev 5:1–9

21. Welshons WV, Nagel SC, Thayer KA, Judy BM, vom Saal FS 1999 Low-dosebioactivity of xenoestrogens in animals: fetal exposure to low doses of me-thoxychlor and other xenoestrogens increases adult prostate size in mice.Toxicol Ind Health 15:12–25

22. vom Saal FS, Cooke PS, Buchanan DL, Palanza P, Thayer KA, Nagel SC,Parmigiani S, Welshons WV 1998 A physiologically based approach to thestudy of bisphenol A and other estrogenic chemicals on the size of reproductiveorgans, daily sperm production, and behavior. Toxicol Ind Health 14:239–260

23. Howdeshell KL, Hotchkiss AK, Thayer KA, Vandenbergh JG, vom Saal FS1999 Exposure to bisphenol A advances puberty. Nature 401:763–764

24. Palanza P, Howdeshell KL, Parmigiani S, vom Saal FS 2002 Exposure to a lowdose of bisphenol A during fetal life or in adulthood alters maternal behaviorin mice. Environ Health Perspect 110(Suppl 3):415–422

25. Farabollini F, Porrini S, Della Seta D, Bianchi F, Dessi-Fulgheri F 2002 Effectsof perinatal exposure to bisphenol A on sociosexual behavior of female andmale rats. Environ Health Perspect 110(Suppl 3):409–414

26. Farabollini F, Porrini S, Dessi-Fulgheri F 1999 Perinatal exposure to theestrogenic pollutant bisphenol A affects behavior in male and female rats.Pharmacol Biochem Behav 64:687–694

27. Adriani W, Seta DD, Dessi-Fulgheri F, Farabollini F, Laviola G 2003 Alteredprofiles of spontaneous novelty seeking, impulsive behavior, and response tod-amphetamine in rats perinatally exposed to bisphenol A. Environ HealthPerspect 111:395–401

28. Dessi-Fulgheri F, Porrini S, Farabollini F 2002 Effects of perinatal exposureto bisphenol A on play behavior of female and male juvenile rats. EnvironHealth Perspect 110(Suppl 3):403–407

29. Ishido M, Masuo Y, Kunimoto M, Oka S, Morita M 2004 Bisphenol A causeshyperactivity in the rat concomitantly with impairment of tyrosine hydrox-ylase immunoreactivity. J Neurosci Res 76:423–433

30. Kubo K, Arai O, Ogata R, Omura M, Hori T, Aou S 2001 Exposure tobisphenol A during the fetal and suckling periods disrupts sexual differenti-ation of the locus coeruleus and of behavior in the rat. Neurosci Lett 304:73–76

31. Kubo K, Arai O, Omura M, Watanabe R, Ogata R, Aou S 2003 Low doseeffects of bisphenol A on sexual differentiation of the brain and behavior in rats.Neurosci Res 45:345–356

32. Kawai K, Nozaki T, Nishikata H, Aou S, Takii M, Kubo C 2003 Aggressivebehavior and serum testosterone concentration during the maturation processof male mice: the effects of fetal exposure to bisphenol A. Environ HealthPerspect 111:175–178

33. Kuiper GG, Lemmen JG, Carlsson B, Corton JC, Safe SH, Van Der Saag PT,van der Burg B, Gustafsson J 1998 Interaction of estrogenic chemicals andphytoestrogens with estrogen receptor �. Endocrinology 139:4252–4263

34. Pennie WD, Aldridge TC, Brooks AN 1998 Differential activation by xe-noestrogens of ER � and ER � when linked to different response elements. JEndocrinol 158:R11–R14

35. Wozniak AL, Bulayeva NN, Watson CS 2005 Xenoestrogens at picomolar tonanomolar concentrations trigger membrane estrogen receptor-�-mediatedCa2� fluxes and prolactin release in GH3/B6 pituitary tumor cells. EnvironHealth Perspect 113:431–439

36. Zsarnovszky A, Le HH, Wang HS, Belcher SM 2005 Ontogeny of rapidestrogen-mediated extracellular signal-regulated kinase signaling in the ratcerebellar cortex: potent nongenomic agonist and endocrine disrupting activityof the xenoestrogen bisphenol A. Endocrinology 146:5388–5396

37. McEwen BS, Alves SE 1999 Estrogen actions in the central nervous system.Endocr Rev 20:279–307

38. De Vries GJ, Simerly RB 2002 Sexually dimorphic neural circuits in themammalian brain. In: Pfaff DW, Arnold AP, Etgen AM, Fahrbach SE, RubinRT, eds. Hormones, brain and behavior. New York: Elsevier; 137

39. Greco TL, Payne AH 1994 Ontogeny of expression of the genes for steroido-genic enzymes P450 side-chain cleavage, 3�-hydroxysteroid dehydrogenase,

3690 Endocrinology, August 2006, 147(8):3681–3691 Rubin et al. • BPA Alters Brain Sexual Differentiation

P450 17�-hydroxylase/C17–20 lyase, and P450 aromatase in fetal mouse go-nads. Endocrinology 135:262–268

40. Bakker J, De Mees C, Douhard Q, Balthazart J, Gabant P, Szpirer J, SzpirerC 2006 �-Fetoprotein protects the developing female mouse brain from mas-culinization and defeminization by estrogens. Nat Neurosci 9:220–226

41. Tilghman SM, Belayew 1982 Transcriptional control of the murine albumin/�-fetoprotein locus during development. Proc Natl Acad Sci USA 79:5254–5257

42. Olsson M, Lindahl G, Ruoslahti E 1977 Genetic control of �-fetoproteinsynthesis in the mouse. J Exp Med 145:819–827

43. Nagel SC, vom Saal FS, Welshons WV 1999 Developmental effects of estro-genic chemicals are predicted by an in vitro assay incorporating modificationof cell uptake by serum. J Steroid Biochem Mol Biol 69:343–357

44. Wiegand SJ, Terasawa E 1982 Discrete lesions reveal functional heterogeneityof suprachiasmatic structures in regulation of gonadotropin secretion in thefemale rat. Neuroendocrinology 34:395–404

45. Ronnekleiv OK, Kelly MJ 1986 Luteinizing hormone-releasing hormone neu-ronal system during the estrous cycle of the female rat: effects of surgicallyinduced persistent estrus. Neuroendocrinology 43:564–576

46. Orikasa C, Kondo Y, Hayashi S, McEwen BS, Sakuma Y 2002 Sexuallydimorphic expression of ER � in the anteroventral periventricular nucleus ofthe rat preoptic area: implication in luteinizing hormone surge. Proc Natl AcadSci USA 99:3306–3311

47. Simerly RB 1991 Prodynorphin and proenkephalin gene expression in theanteroventral periventricular nucleus of the rat: sexual differentiation andhormonal regulation. Mol Cell Neurosci 2:473–484

48. Herbison AE 1992 Identification of a sexually dimorphic neural populationimmunoreactive for calcitonin gene-related peptide (CGRP) in the rat medialpreoptic area. Brain Res 591:289–295

49. Okamura H, Yokosuka M, Hayashi S 1994 Induction of substance P-immu-noreactivity by estrogen in neurons containing estrogen receptors in the an-teroventral periventricular nucleus of female but not male rats. J Neuroen-docrinol 6:609–615

50. Simerly RB 1989 Hormonal control of the development and regulation oftyrosine hydroxylase expression within a sexually dimorphic population ofdopaminergic cells in the hypothalamus. Mol Brain Res 6:297–310

51. Simerly RB, Swanson LW, Gorski RA 1985 The distribution of monoamin-ergic cells and fibers in a periventricular preoptic nucleus involved in thecontrol of gonadotropin release: immunohistochemical evidence for a dopa-minergic sexual dimorphism. Brain Res 330:55–64

52. Simerly RB, Zee MC, Pendleton JW, Lubahn DB, Korach KS 1997 Estrogenreceptor-dependant sexual differentiation of dopaminergic neurons in thepreoptic region of the mouse. Proc Natl Acad Sci USA 94:14077–14082

53. Voikar V, Koks S, Vasar E, Rauvala H 2001 Strain and gender differences inthe behavior of mouse lines commonly used in transgenic studies. PhysiolBehav 72:271–281

54. Archer J 1975 Rodent sex differences in emotional and related behavior. BehavBiol 14:451–479

55. Blizard DA, Lippman HR, Chen JJ 1975 Sex differences in open field behaviorin the rat: the inductive and activational role of gonadal hormones. PhysiolBehav 14:601–608

56. Masur J 1972 Sex differences in “emotionality” and behavior of rats in theopen-field. Behav Biol 7:749–754

57. Ogawa S, Lubahn DB, Korach KS, Pfaff DW 1997 Behavioral effects ofestrogen receptor gene disruption in male mice. Proc Natl Acad Sci USA94:1476–1481

58. Soto AM, Lin TM, Justicia H, Silvia RM, Sonnenschein C 1992 An “inculture” bioassay to assess the estrogenicity of xenobiotics. In: Colborn T,Clement C, eds. Chemically induced alterations in sexual development: thewildlife/human connection. Princeton, NJ: Princeton Scientific Publishing;295–309

59. Bronson FH 1971 Rodent pheromones. Biol Reprod 4:344–35760. Clee MD, Humphreys EM, Russell JA 1975 The suppression of ovarian

cyclical activity in groups of mice, and its dependence on ovarian hormones.J Reprod Fertil 45:395–398

61. Champlin AK 1971 Suppression of oestrus in grouped mice: the effects ofvarious densities and the possible nature of the stimulus. J Reprod Fertil27:233–241

62. Franklin GL, Paxinos 2004 The mouse brain atlas. Academic Press: New York

63. Adriani W, Chiarotti F, Laviola G 1998 Elevated novelty seeking and peculiard-amphetamine sensitization in periadolescent mice. Behav Neurosci 112:1152–1166

64. Simerly RB, Swanson LW, Handa RJ, Gorski RA 1985 Influence of perinatalandrogen on the sexually dimorphic distribution of tyrosine hydroxylase-immunoreactive cells and fibers in the anteroventral periventricular nucleusof the rat. Neuroendocrinology 40:501–510

65. Bleier R, Byne W, Siggelkow I 1982 Cytoarchitectonic sexual dimorphisms ofthe medial preoptic and anterior hypothalamic areas in guinea pig, rat, ham-ster, and mouse. J Comp Neurol 212:118–130

66. Forger NG, Rosen GJ, Waters EM, Jacob D, Simerly RB, De Vries GJ 2004Deletion of Bax eliminates sex differences in the mouse forebrain. Proc NatlAcad Sci USA 101:13666–13671

67. Zup SL, Carrier H, Waters EM, Tabor A, Bengston L, Rosen GJ, Simerly RB,Forger NG 2003 Overexpression of Bcl-2 reduces sex differences in neuronnumber in the brain and spinal cord. J Neurosci 23:2357–2362

68. Han TM, De Vries GJ 1999 Neurogenesis of galanin cells in the bed nucleusof the stria terminalis and centromedial amygdala in rats: a model for sexualdifferentiation of neuronal phenotype. J Neurobiol 38:491–498

69. Chung WCJ, Swaab DF, De Vries GJ 2000 Apoptosis during sexual differ-entiation of the bed nucleus of the stria terminalis in the rat brain. J Neurobiol43:234–243

70. Petersen SL, Barraclough CA 1989 Suppression of spontaneous LH surges inestrogen-treated ovariectomized rats by microimplants of antiestrogens intothe preoptic brain. Brain Res 484:279–289

71. Simonian SX, Spratt DP, Herbison AE 1999 Identification and characteriza-tion of estrogen receptor �-containing neurons projecting to the vicinity of thegonadotropin-releasing hormone perikarya in the rostral preoptic area of therat. J Comp Neurol 411:346–358

72. Gu GB, Simerly RB 1997 Projections of the sexually dimorphic anteroventralperiventricular nucleus in the female rat. J Comp Neurol 384:142–164

73. Le WW, Berghorn KA, Rassnick S, Hoffman GE 1999 Periventricular preopticarea neurons coactivated with luteinizing hormone (LH)-releasing hormone(LHRH) neurons at the time of the LH surge are LHRH afferents. Endocri-nology 140:510–519

74. Leranth C, MacLusky NJ, Shanabrough M, Naftolin F 1988 Catecholamin-ergic innervation of luteinizing hormone-releasing hormone and glutamic aciddecarboxylase immunopositive neurons in the rat medial preoptic area. Neu-roendocrinology 48:591–602

75. Horvath TL, Naftolin F, Leranth C 1993 Luteinizing hormone-releasing hor-mone and �-aminobutyric acid neurons in the medial preoptic area are syn-aptic targets of dopamine axons originating in anterior periventricular areas.J Neuroendocrinol 5:71–79

76. Funabashi T, Kawaguchi M, Furuta M, Fukushima A, Kimura F 2004 Ex-posure to bisphenol A during gestation and lactation causes loss of sex dif-ference in corticotropin-releasing hormone-immunoreactive neurons in thebed nucleus of the stria terminalis of rats. Psychoneuroendocrinology 29:475–485

77. McCarthy MM, Schlenker EH, Pfaff DW 1993 Enduring consequences ofneonatal treatment with antisense oligodeoxynucleotides to estrogen receptormessenger ribonucleic acid on sexual differentiation of rat brain. Endocrinol-ogy 133:433–439

78. Lucion AB, Charchat H, Pereira GAM, Rasia-Filho AA 1996 Influence of earlypostnatal gonadal hormones on anxiety in adult male rats. Physiol Behav60:1419–1423

79. vom Saal FS, Timms BG, Montano MM, Palanza P, Thayer KA, Nagel SC,Ganjam VK, Parmigiani S, Welshons WV 1997 Prostate enlargement in micedue to fetal exposure to low doses of estradiol or diethylstilbestrol and oppositeeffects at high doses. Proc Natl Acad Sci USA 94:2056–2061

80. Zalko D, Soto AM, Dolo L, Dorio C, Ratahao E, Debrauwer L, Faure R,Cravedi JP 2003 Biotransformations of bisphenol A in a mammalian model:answers and new questions raised by low-dose metabolic fate studies inpregnant CD1 mice. Environ Health Perspect 111:309–319

81. Snyder RW, Maness SC, Gaido KW, Sumner SCJ, Fennell TR 2000 Metab-olism and disposition of bisphenol A in female rats. Toxicol Appl Pharmacol168:225–234

82. Yamada H, Furuta I, Kato EH, Kataoka S, Usuki Y, Kobashi G, Sata F, KishiR, Fujimoto S 2002 Maternal serum and amniotic fluid bisphenol A concen-trations in the early second trimester. Reprod Toxicol 16:735–739

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

Rubin et al. • BPA Alters Brain Sexual Differentiation Endocrinology, August 2006, 147(8):3681–3691 3691