Our goal was to assess the effects of BPA on several behaviors in juvenile mice using a very low dose of BPA, within a range comparable to humans. We also limited exposure to the gestational period when most neural development occurs. We did so not only to limit the exposure time but also to remove the complicating effects that BPA may have on maternal behavior 
. Thus by fostering all pups to dams on control diet, we control for the potential contribution of differences in maternal care. Needless to say fostering introduces a new set of issues, which we controlled for here by fostering all the animals used in the experiments. Incorporating BPA in chow we produced tonic levels in mouse dams comparable to humans and we assume these were also achieved in the embryos. BPA readily crosses the placenta 
, exposing fetal tissues to concentrations close to those in maternal placenta 
. Furthermore, in its inactivate conjugated form, BPA, can be unconjugated and re-activated in the fetus 
. Thus, our dosing method is a highly relevant for studying the effects of gestational BPA exposure and subsequent behavioral and genomic profiles. Using a dose well below the LOAEL, given only during gestation, we detected several effects of BPA in diet on juvenile social behaviors as well as embryonic gene expression in whole brain, which may underlie the observed behavioral effects.
Our most striking findings were in females exposed to BPA during a juvenile dyadic social interaction task. This is an “open-ended” task used to identify social, non-social, play-soliciting and investigative behaviors between pairs of mice of the same sex and age 
. BPA exposed females displayed increased social interactions. They engaged in more side-by-side interactions and followed each other more than pairs of control females. They also displayed less self grooming, a non-social behavior. Moreover, in two other measures of partner interactions; nose-to-nose investigations and approaches, BPA treated animals were more gregarious than controls. Together, these results suggest that gestational exposure to BPA in female mice increases social investigations and interactions. BPA exposure, particularly when restricted to critical brain developmental periods, alters the sexually dimorphic brain and in certain behaviors, females appear to be more sensitive to the effects of BPA. For example, BPA exposure decreased the number of tyrosine hydroxylase (TH) positive cells in females, eliminating the sexual dimorphism in the anteroventral periventricular region of the hypothalamus 
. Females exposed to BPA during the period of brain sexual differentiation were less reactive and explored a novel environment less than control females 
. Amphetamine-induced conditioned place preference is also disrupted in female mice exposed to BPA during gestation while males displayed no change in place preference 
. In rats, BPA exposure at various time points during gestation and after birth affects juvenile social behaviors. In juvenile females, BPA increased both non-social and social investigation 
. When tested with males, juvenile females exposed to BPA exhibited reduced play and social grooming. In a study comparing male and female rat juvenile social behavior, BPA reduced social interest in both sexes but a lower dose exposure increased social approach and interactions in females 
. Our data add to these conclusions showing that in juvenile female mice, even at very low doses, BPA in utero
increases their display of social interactions.
Interestingly in the present study, juvenile males were less social than females, exhibiting higher levels of cage exploration and sitting alone than did females. Females on the other hand spent more time in side-by-side sitting than did males. Sex differences have been reported in normal C57BL/6J juvenile mice tested in the same manner reported here 
. In that study, males engaged in more social interactions, and females performed more play soliciting. In CD-1 mice, on several tasks, BPA exposure reduced the sex differences observed in controls 
. Several procedural differences between these studies likely cause the variation in results. In the present studies, all mice were reared on a phytoestrogen-free chow, as opposed to normal rodent chow containing phytoestrogens. Phytoestrogens in chow can enhance sex differences in the brain and in behavior 
. Second, mice in the present study were fostered at birth and we cannot exclude the possibility that fostering may have as yet unidentified effects which alter the display of sex differences. The few studies directly assessing the effects of fostering on maternal rearing behavior have conflicting results 
but raise the possibility that C57BL/6J dams may lick and/or groom the fostered pups more than their biological pups.
In a preference task, juvenile males spent more than 50% of their time in the chamber with an adult male mouse instead of the empty chamber. Exposure to BPA at the current dose did not significantly alter these behaviors. However a 40-fold higher BPA dose decreased male social preference in the identical task 
. Likewise we did not find any effects of sex or BPA on elevated plus maze behavior in the current study. Yet again, exposure to the 40-fold higher BPA dose decreased time in the middle, increased time in the closed arm, and tended to decrease time in the outer most portion of the open arm 
, indicating that BPA at this high dose increased anxiety-related behaviors in the plus maze. These data demonstrate that different juvenile behaviors are affected by different doses of BPA.
There are several potential mechanisms underlying the sex-specific behavioral responses to BPA exposure. Males, but not females, experience a surge of testosterone (and estradiol) during the last few days of gestation 
. Thus endogenous estrogens may out-compete BPA, for example, for estrogen receptor binding sites in males but not in neonatal females that are not experiencing an endogenous hormone surge. Another possibility is that genes on sex chromosomes may interact with BPA 
. Our embryonic gene expression results tend to support the first potential mechanism. We assayed expression levels of estrogen receptor (ER) genes in E18.5 brains to ask if BPA would affect their expression. We selected this embryonic age because it is when brain sexual differentiation occurs under the influence of sex differences in blood levels of testosterone and brain levels of estrogens 
. We hypothesized that if BPA acted via one or more ER it might down regulate this receptor, perhaps in a manner similar to the natural ligand. However, we found no changes in any of the ER or putative ER genes based on BPA exposure. This does not mean that BPA did not act as a direct ligand for one or more of the ERs. In fact it is likely that BPA did influence ERα since the oxytocin receptor is an ERα target gene and we noted changes in Oxtr
. In neonatal cortex, estradiol hypermethylates ERα 
and perhaps the same thing occurs here, but, given that we assayed mRNA in whole brain and ERα expression is not global we may have missed BPA induced changes in expression.
Another mode of BPA action is via direct changes DNA methylation 
. We assayed the three known DNA methyltransferases, enzymes responsible for the deposition of methyl groups onto cytosines when followed by a guanine. Dnmt3a
, a gene responsible for de novo
methylation, was altered by BPA exposure wherein the sex difference in control animals (male>females) was not present in BPA exposed brains because BPA increased Dnmt3a
expression in females. Dnmt3a
has been implicated in rewarding behavior and neuronal plasticity in the adult mouse accumbens 
is the most abundant DNA methyltransferase and is believed to be responsible for the maintenance of DNA methylation. Interestingly, Dnmt1
gene was decreased by BPA in female embryo brains and we speculate that the drop in this gene might be related to the higher levels of Slc1a1
in these same brains.
In male mice, exposed to low doses of BPA that are probably comparable to ours, mRNA for ERβ and several of the NMDA receptors were decreased in hippocampus at day 21 and 56 
. In addition in male rats, LTD and LTP were disrupted by a dose of BPA higher than ours 
. These effects were attributed to functional alterations in dopamine, glutamatergic, and metabotropic glutamate receptors, but in our study none of the genes we assayed, related to the later two pathways, were affected by BPA. Of course, our treatment period was shorter than theirs; in both studies dams were placed on the BPA diet around E7 and pups stayed on the diet through weaning. Moreover, we examined mRNA at a different time point, but none of the three genes (Gria1, Grin2a, Grin2b
) were affected by diet in our study.
Despite the lack of change in NMDA and AMPA receptors, the largest gene expression effect was noted for Slc1a1
, one of the glutamate transporter genes, and interestingly its expression was elevated in brains of females exposed to BPA. Slc1a1
is found throughout the cortex, hippocampus and basal ganglia 
and functions to buffer local glutamate at excitatory synaptic connections 
. Behavioral characterization of the Slc1a1
null mouse is not extensive but with age these animals appear to have impairments in self-grooming and spatial learning 
. In humans single nucleotide polymorphisms (SNPs) within this gene have been associated with repetitive behaviors and anxiety in children with ASD 
, mental retardation 
and obsessive compulsive disorder 
. The last candidate gene we examined, the oxytocin receptor, has long been associated with the display of social behavior in rodents 
. A polymorphic region of the oxytocin receptor (OXTR) in humans has been associated with empathy and stress reactivity 
. Additionally, many have speculated that OXTR is involved with autism spectrum disorders. Two genome wide association studies in autistics patients show linkage associations to chromosomal region 3p25.3, which contains OXTR 
. Together, our findings that in utero
BPA exposure alters expression of Oxtr
suggest a potential mechanism through which early life exposure to BPA can alter normal signaling in the brain and effect adult neurological disorders such as the pathophysiologies associated with ASD, obsessive compulsive disorder and mental retardation.