Perinatal high-fat diet and bisphenol A: Effects on behavior and gene expression in the medial prefrontal cortex.

Abstract

Both high-fat diets and bisphenol A (BPA), an environmental endocrine disruptor, are prevalent in industrialized societies. Previous studies have detected separate effects of BPA and high-fat diets; however, none have assessed possible interactive effects. Here, pregnant dams consumed 0, 40 or 400 μg/kg/day BPA and were fed either a control (CON; 15.8% kcal fat) or high-fat diet (HFD; 45% kcal fat) from gestational day 2 through parturition. The pups were individually dosed with BPA from postnatal days (P) 1–10, while the dams continued to consume one of the two diets. Maternal behavior increased with the HFD while the offspring’s periadolescent social play decreased with BPA, but no interactive effects were observed. Neither HFD nor BPA exposure changed performance on a social recognition task, and only BPA had an effect on the elevated plus maze. BPA increased several cytokines in the medial prefrontal cortex (mPFC) of P10 males, but not females. Expression of several genes related to hormone synthesis and receptors, inflammation, oxidative stress, and apoptosis in the mPFC at P10 and P90 were altered due to BPA and/or HFD exposure with rare interactive effects. BPA resulted in an increase in the gene expression of Esr1 in the mPFC of females at both P10 and P90. Epigenetic analysis at P90 did not show a change in methylation or in the levels of pre-mRNA or microRNA. Thus perinatal BPA and HFD have separate effects but rarely interact.

1. Introduction

The early environment, including maternal care, can have long-lasting effects on the cellular composition [1] and behavior [2] of rodent offspring with broad generalization to other mammalian species [3, 4]. Research in animal models has suggested that endocrine disruptors and dietary intake can alter maternal behavior, offspring behavior and gene expression. Both endocrine disruptors and high-fat diets (HFD) have also been independently found to induce inflammation in the brain and periphery [5, 6, 7]; however, few studies have investigated their possible interactions [8]. This may have implications for humans given the ubiquitous exposure of endocrine-disrupting chemicals in our environment and HFD in Western society.

Endocrine disruptors are compounds that interact with hormone receptors as agonists, antagonists or a mixture of both. Thus endocrine disruptors can inappropriately activate a hormone receptor, even when the hormone is not present, and they can block the actions of hormones. Given that hormones play a role in the development of the nervous system and that hormonal receptors are present during this crucial time, endocrine disruptors can alter the course of development. One such endocrine disruptor is bisphenol A (BPA), a chemical found in polycarbonate plastics, certain dental sealants, and epoxy resins that line canned foods [9]. BPA binds with the highest affinity to estrogen-related receptor γ (ERRγ), but also to estrogen receptor (ER) α, ERβ, and androgen and thyroid receptors [10, 11, 12, 13, 14]. Exposure to BPA has been found to change gene expression, neuroanatomy and several types of behavior. Maternal behaviors during BPA exposure alter the time spent licking and grooming the pups, as well as time spent nursing in an arched-back position [15, 16]. Perinatal exposure to BPA also causes long-term effects on social, sexual, and anxiety behavior in adolescent and adult animals, as well as the number of neurons in a number of neural areas, including the medial prefrontal cortex (mPFC) [17, 18, 19, 20, 21, 22]. On a molecular level, perinatal BPA exposure has been found to affect gene expression of Esr1, Esr2, Errg, as well as dopamine and serotonin receptor-related proteins in the prefrontal cortex of young rats at P21 and mice at P28 [16, 23]. Furthermore, the differences in gene expression of Esr1, Esr2, and Errg have been found to be due to changes in DNA methylation in juvenile mice [16]. Based on the evidence of expression and methylation changes within the cortex and behavior changes in adolescence and adulthood, perinatal BPA exposure has a significant effect on the offspring.

HFD can also affect behavior, inflammation and gene expression. The existing literature presents conflicting evidence of its effect on maternal behavior with both increases and decreases of positive maternal behaviors being reported [24, 25, 26]. In offspring, there are indications of HFD influencing social play, cognitive, and anxiety behaviors [27, 28, 29]. Gestational HFD with 60% fat content has been found to increase pro-inflammatory cytokine expression of IL-1β, microglial activation in the hippocampus of adult male and female offspring [28], and the level of IL-6 in frontal cortex of adult male offspring [30]. A HFD (45%) has also been found to alter gene expression in the hippocampus and performance on the Morris water maze [31], as well as signaling in cultured primary neural stem cells (32). Research on non-human primates also found alterations in the gene expression of serotonin receptors in male and female offspring following maternal HFD [32% fat; 29]. Gestational HFD (60% fat) through P40 is known to affect expression of oxytocin-related genes in the mPFC at P40 [27], and maternal HFD (60% fat) through gestation and lactation also alters opioid and dopamine-related proteins in the mPFC of adult male mice [33]. Few studies have investigated the effect of HFD and BPA on gene expression of hormone receptors in the PFC, despite the interactions of estrogen receptors and inflammation in other brain areas [34].

Although separate exposure to HFD and BPA are known to alter maternal behavior, offspring behavior, gene expression, and inflammation in the brain and periphery [5, 6, 7], the combined effects of both factors have not been investigated. Additionally, previous work in our lab found a higher number of neurons and glia in the mPFC of adult male, but not female, rats following perinatal BPA exposure [22]. The higher number of neurons found in male rats may have a parallel in the higher number of neurons found in a subset of children with autism, a male-biased neurological disorder with impaired social interactions as a core characteristic [35]. To further investigate this similarity, the current study focuses on the gene expression and cytokine levels in the mPFC, as well as social behaviors. We examine the possible interactive effects of BPA exposure and HFD during gestation and the early postnatal period on maternal care, as well as offspring behavior, gene expression, and cytokine levels in the mPFC.

2. Methods

2.1. Breeding

Male and female Long Evans hooded rats (n=58 of each sex) purchased from Harlan Laboratories (now Envigo, Indianapolis, IN, USA) were housed in our vivarium for at least one week before being paired for breeding. They were kept on a 12:12 light/dark cycle and were allowed food and water ad libitum. Precautions were taken to reduce the environmental exposure to endocrine disruptors. All animals were housed in BPA-free polysulfone cages, water was reverse osmosis filtered in glass bottles, and feed was a low phytoestrogen diet (Harlan 2020X; Teklad Diets, Madison, WI, USA) except for the specialized diet (see 2.3) during gestation through P10. All procedures were approved by the University of Illinois Institutional Care and Use Committee, and adhere to the National Institute of Health guidelines on the ethical use of animals.

Breeding pairs were placed in suspended wire bottom cages and checked daily for the presence of sperm plugs. The breeding animals were paired for a maximum of six nights. If no sperm plug was detected after six nights, the male was removed and another male introduced. The day a sperm plug was detected was recorded as gestational day (G) 0 and the dams were singly housed in a polysulfone shoebox cage. From the first day of pregnancy, each female was assigned to one of six groups. The number of litters for the control diet and BPA exposures were: 0 μg BPA/kg = 8; 40 μg BPA/kg = 10; 400 μg BPA/kg = 10. The number of litters for the high-fat diet and BPA exposures were: 0 μg BPA/kg = 10; 40 μg BPA/kg = 10; 400 μg BPA/kg = 10. The experimental design is illustrated in Figure 1.

Figure 1.

Timeline of behavior and tissue collection following perinatal BPA and HFD exposure.

2.2. BPA Dosing

BPA was suspended in tocopherol-stripped corn oil at 0, 0.1 mg BPA/ml, or 1.0 mg BPA/ml in order to administer 0 (control), 40 μg BPA/kg, or 400 μg BPA/kg respectively. Our lab has previously found effects at the 40 and 400 μg BPA/kg doses [22, 36]. Tocopherol-stripped corn oil has no estrogenic activity as evidenced by no increase in uterine weight of pre-pubertal mice [37]. To dose the adult dams, the required amount (0.4 μl/g body weight) was pipetted onto ½ of a cookie (Newman’s Own organic alphabet cookie, vanilla flavor) and given to the animals. The animals readily consumed the cookie making this route of exposure non-stressful and similar to human ingestion of BPA. On G0 and G1, the dams were given ½ of a cookie with 0.4 μl/g tocopherol-stripped corn oil. Starting on G2 through parturition, the dams were given the cookie with the assigned BPA dose. The day of birth was recorded as P0 and the litters were not disturbed. Then daily from P1–10, each pup was individually dosed via pipetting the assigned solution (same as dam) directly into its mouth because lactational transfer of BPA is very low [38]. A plastic pipette tip was gently eased into the pup’s mouth towards the back of the tongue. The oil was dripped onto the pup’s tongue and the experimenter slowly administered small amounts of oil to avoid damaging oral tissue and pulmonary aspiration. The pups easily and quickly consumed the small volume of oil (0.4 ul/g body weight). Although the resulting level of BPA in the brain was not measured in the current study, previous research has shown that oral administration results in BPA being present in brain tissue of both fetal and early postnatal animals [39, 40].

2.3. Diet

Starting on G0 through P10, the dams were fed either a control diet (CON; D10012G, Research Diets Inc., New Jersey, USA) or a high-fat diet (HFD; D12451, Research Diets Inc., New Jersey, USA). The CON diet was comprised of 15.8% kcal fat, 20.3% kcal protein, and 63.9% kcal carbohydrate. The HFD was comprised of 45% kcal fat, 20% kcal protein, and 35% kcal carbohydrate. This diet was chosen because there is evidence that the Western diet has 35% fat on average [41], making the 60% HFD out of range for most people. The animals were allowed to feed ad libitum. Number and sex ratio of the pups and daily body weight for dams and pups were recorded.

2.4. Maternal Behavior Observations

The dams were observed with their litters from P3–10 during the dark cycle using night vision goggles by an observer blind to treatment group. The dams’ behavior was recorded every 3 minutes for 90 minutes (30 observations per night). The behavioral classifications were nursing, licking of pups, pup retrieval, nest building, and being away from the nest. When nursing and licking of pups occurred simultaneously, licking was preferentially recorded. Pup retrieval and nest building were rarely observed and therefore not included in the analyses.

2.5. Periadolescent Social Behavior

Two animals of the same sex and treatment (not cagemates) were paired to assess social behavior. The method of examining play in pairs of same-treatment animals is often used in the literature [42, 43]. Animals between ages P26–40 were isolated in individual cages for one hour before each session and then were paired for social assessments in a neutral cage on four consecutive days one hour before the start of the dark cycle. The animals’ behavior was observed for 20 minutes, and one-minute time point sampling was used to classify the behaviors: sniffing, wrestling, chasing, passive contact (no play, but touching), and solitary (no play and not touching). After data collection, the animals were returned to their home cage. All animals in the study were assessed, and then average from each sex within a litter used in the statistical analysis.

2.6. Elevated Plus Maze

At P90 (adulthood), the one male and one female from each litter was tested for anxiety behavior using the elevated plus maze (EPM). The floor of the EPM apparatus was 50 cm tall with two open arms and two closed arms that were each 50 cm long and 10.5 cm wide. The two closed arms had walls 33.5 cm tall. Each animal was placed on the center and allowed to explore the maze for one 5 minute trial. Time spent in each arm, time in the center and the number of entries into each arm were recorded. The operational definition of “entry” was more than half of the animal’s body had to enter the arm in order to be counted and timed.

2.7. Social Recognition

At P90, another set of littermates not tested in 2.6, were assessed for social memory in the social recognition task using same sex juveniles (P21–30). Before the task began, each juvenile was marked on its back with a different colored non-toxic marker for easy identification. There were three days of habituation, which were identical to testing except the delay period was always 30 minutes. Data was collected during habituation but was not included in the data analysis. Following habituation, the delay periods (15, 45, 90 or 120 minutes) were randomized between animals for the four days of testing. To test social memory, adult animals were introduced to and allowed to investigate a same-sex juvenile for thirty seconds. After the initial introduction, the juvenile animal was removed and the adult was placed back with its cagemate for a delay period.. After the delay period, the same juvenile (familiar) and a novel juvenile were placed in a cage with the adult. The time the adult spent investigating each juvenile was recorded for three minutes. In theory, an animal with good social memory will spend more time investigating the novel juvenile rather than the familiar juvenile. The adult animals never saw the same juveniles on more than one day.

2.8. ELISA assays

On P10, one male and one female from each litter was removed one hour after BPA treatment. The remainder of the litter was allowed to age into adulthood. Each pup was anesthetized using CO₂ then quickly decapitated. The brain was immediately removed, snap frozen in liquid nitrogen. The prefrontal cortex was excised while the brain was frozen and one hemisphere was used for ELISA analysis while the other was used for gene expression analysis (2.9). The hemispheres were stored in a −80°C degree freezer. All tissue was coded to conceal BPA treatment group. A pre-coated ELISA plate (Signosis, Santa Clara, CA) was used to analyze the levels of cytokines. Each plate could bind and measure tumor necrosis factor-α (TNF-α), transforming growth factor-β (TGF-β), MCP-1, interleukin-1α (IL-1α), IL-1β, IL-6, IL-15, and vascular endothelial growth factor (VEGF) simultaneously, allowing for a survey of inflammatory factors. The second hemisphere of the P10 mPFC was lysed following manufacturer’s instructions. Briefly, 1 ml of cell lysis buffer/100 mg of tissue was homogenized on ice. The lysate was centrifuged to separate the tissue from the supernatant, which was then diluted to 100 μg/100 μl per well and used for the protein analysis. The samples were run in duplicate and incubated in the plate overnight followed by three washes. A biotin-labeled antibody (specific to each cytokine) was added and allowed to incubate for 4 hours followed again with three washes. Next, a streptavidin-HRP conjugate was added to each well, incubated for 45 minutes, and followed with three washes. Finally, HRP-substrate was added and allowed to incubate for 30 minutes before the addition of stop solution to each well. The optical density of each well was determined with a microplate reader at 450 nm within 5 minutes.

2.9. RNA Isolation and One-Step Real Time qPCR

Genes related to hormone receptors (Esr1, Esr2, Errg, Ar), steroid synthesis (StAr, Ldlr, Cyp11a1), inflammation (Aif1, C4a, C4b), apoptosis (Casp3), and oxidative stress (Cat, Sod1, Sod2, Glrx) were chosen for investigation. Before genetic analysis, one hemisphere from each animal was randomly selected and total RNA was isolated using TRI reagent (Sigma, St. Louis, MO, USA), followed by Direct-zol™ RNA MiniPrep (Zymo Research, Irvine, CA) according to manufacturer’s instructions. Reverse transcription was performed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Quantitative Real time PCR was performed using the StepOnePlus™ Real-Time PCR System with Power SYBR® Green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA) using the respective forward and reverse primer for each gene (Suppl. Table S1), and were designed by Vector NTI software (Thermo Fisher Scientific, Waltham, MA) and synthesized by Integrated DNA Technologies (Coralville, IA). Standard curves with a slope of −3.30 (SEM 0.30) and R²≥ 0.99 were accepted. A housekeeping gene ribosomal protein L7a (RpL7a) whose expression was not affected by treatment was used to normalize the gene expression data.

2.10. MRI, RNA Isolation and One-Step Real Time qPCR

On P90, one male and one female from each litter was scanned for fat and lean mass in a magnetic resonance imaging (MRI) system. Following the scan, the animals were euthanized via CO2 and the brain was immediately removed. The cortex was excised, snap frozen in liquid nitrogen, and stored in a −80C degree freezer. All tissue was coded to conceal BPA treatment group. The same genes analyzed in the P10 group were analyzed and identical methods were used.

2.11. Genomic DNA isolation and methylation-sensitive PCR analysis of Esr1

Frozen cortex (50 mg) from P90 animals was ground in liquid nitrogen and genomic DNA was then isolated using ZR Genomic DNA™Tissue MiniPrep (Zymo Research, Irvine, California) per manufacturer’s instruction. gDNA was bisulfite converted using the EZ DNA Methylation-Gold Kit (Zymo Research, Irvine, California) following the manufacturer’s instructions. Quantitative PCR used 20 ng of gDNA as the template using Power SYBR Green PCR Master as the reporter. Primers used for each genomic region were designed using the Methprimer website [44], to create four primer sets that targeted the previously described promoter of Rattus norvegicus Esr1 (ENST00000440973.1) [16]. Primer information is in the Suppl. Fig. S1. A serial dilution from a reaction combining 1 sample from each experimental group was used to create an internal standard curve for quantification. Data are presented as the %Methylation (%Methylation= 100/[1+2^((1.33-(Ct_Unmeth-Ct_Meth)/−2.27)]).

2.12. Identification of miRNAs targeting Esr1 3’-UTR

Further exploration of epigenetic mechanisms was undertaken by analyzing the expression levels of three microRNA (miRNA) associated with Esr1. To identify rat miRNAs that might target Esr1 3’UTR, a three-pronged analysis was used. First, a bioinformatic approach was used to map the 3’-UTR of Rattus norvegicus Esr1 gene in miRDB (http://www.mirdb.org/), Targetscan (http://www.targetscan.org/), and PicTar (http://pictar.mdc-berlin.de/) to predict miRNAs that interact with Esr1. Then, a miRNA family that is broadly conserved across vertebrates, targets the Esr1 gene and is bound by BPA was found, to facilitate potential generalizations across species in the future. The final miRNA candidates were selected based on families that overlapped in the three approaches: rno-mir-19b3p (MIMAT0000788), rno-mir-221/222–3p (MIMAT0000890), and rno-mir-22–3p (MIMAT0000791). Quantification of specific miRNA was achieved through individual reverse transcription of previously isolated RNA for enrichment, followed by qPCR with the commercially available TaqMan probes for rno-mir-19b-3p (Assay ID: rno478264_mir [Cat. No. A25576]), rno-mir-221–3p (Assay ID: rno481005_mir [Cat. No. A25576]), and rno-mir-22–3p (Assay ID: rno481004_mir [Cat. No. A25576]), following manufacturer’s instructions. All miRNA content was normalized to a U6 housekeeping snRNA (NCBI Accession: NR_004394; Assay ID: 001973 [Cat. No. 4440887]). Specific localization of each miRNA within the 3’-UTR of the Esr1 rat gene (4037 bp) is illustrated in Suppl. Fig. S2, where only rnomir-19b-3p and rno-mir-22–3p showed at least two individual binding sites along this region. Seed sequences and binding sequences within Esr1 are shown on Suppl. Fig. S2.

2.13. Statistical Analyses

The growth curve for the dams’ gestational body weights was analyzed with a 3 (BPA treatment) × 2 (diet) repeated measures linear mixed model with a first-order ante-dependence covariance structure (determined by small Akaike’s Information Criterion). Litter size, sex ratio, as well as each maternal behavior (nursing, licking, and time away from nest) were analyzed with 3 (BPA treatment) × 2 (diet) ANOVAs. For every measure, only one animal of each sex was used or the animals within the litter were averaged and litter was the unit used for the analysis (e.g., social play). A 3 (BPA treatment) × 2 (diet) × 2 (sex) ANOVA was used to analyze the difference in pup weights at both P1 and P10. The analysis of MRI data was completed on each sex separately given the known sex difference in body weight.

To analyze the periadolescent social behavior, sniffing, chasing and wrestling were combined into one “social play behavior” category. In order to use only one measure per litter, the average counts of each category per litter were calculated and used in the analysis. Each behavior was averaged across days 1–2 and days 3–4. A 3 (BPA treatment) × 2 (diet) × 2 (sex) × 2 (days) repeated measures ANOVA was used to analyze the play behavior data. Significant interactions were further analyzed using Tukey’s LSD post hoc test.

A 3 (BPA treatment) × 2 (diet) × 2 (sex) ANOVA was used to analyze the EPM data. The same analysis was used to assess the percent of time spent investigating the novel juvenile at each delay of the social recognition task and the gene expression data collected from the P10 and P90 cortex. To analyze the ELISA assays, percent change from the control group (CON diet/0 dose BPA) was calculated for each cytokine, and then a 3 (BPA treatment) × 2 (diet) × 2 (sex) ANOVA with plate number as a co-factor was used to analyze the percent change for each cytokine. Tukey’s LSD was used for post hoc tests. When analyzing the effects of BPA treatment, each dose was only compared to the control (0 dose BPA) group. Significant differential expression of miRNA data was defined as a fold change greater than 0.5.

3. Results

3.1. Maternal and Offspring Characteristics

Neither BPA nor HFD had a significant effect on gestational body weights, litter size, or sex ratio (data not shown). There were also no significant differences in offspring body weights at P1; however, BPA treatment differences emerged by P10, F(2, 106) = 5.48, p=.005, with the 40 μg BPA/kg group weighing less than the 0 dose BPA group (p=.002). There were no effects on body weight or body composition in adult rats as assessed by MRI (data not shown).

3.2. Maternal Behavior

The HFD dams spent more time in from the nest compared to CON diet dams; F(1, 52) = 0.25, p=.016 (Fig. 2A). There was also an increase in the amount of time the HFD dams spent nursing the pups that was close to significant, F(1, 52) = 3.84, p=.055 (Fig. 2B). There was no effect of BPA treatment and no interaction on these measures. BPA had a near to significant effect on licking behavior, F(2, 52) = 3.03, p=.057 (Fig. 2C), with the 400 μg BPA/kg exposure group spending less time licking the pups than the 0 dose group (p=.055). Licking was not affected by diet and there was no interaction.

Figure 2.

Maternal behavior assessment (data shown as mean ± SEM in this and all subsequent figures). A) Dams exposed to CON diet spent more time away from the nest compared to the HFD exposed dams, *p<.05. There was no effect of BPA. B) Dams exposed to CON diet spent marginally less time nursing the pups, p=.055. The data are collapsed across BPA treatment because there was no effect of BPA exposure. C) Dams exposed to 400 μg BPA/kg spent marginally less time licking the pups than 0 dose BPA dams, p=.057. The data are collapsed across diet because there was no effect of diet.

3.3. Periadolescent Social Behavior

There was a significant interaction between day and BPA treatment for play, F(2, 160)=6.34; p<.01 (Fig. 3A). No significant differences were found on days 1 and 2 of play behavior; however, BPA treatment differences emerged on days 3 and 4. Play behaviors (sniffing, wrestling and chasing) were significantly higher in 0 dose BPA animals than in the 400 μg BPA/kg animals (p<.05), while the 40 μg BPA/kg animals also appeared to show a decrease in play, the difference with the control animals was not significant. There was also a main effect of sex in the amount of play behavior with males exhibiting more play than females, F(1,160)=6.70, p<.05. There were no main effects of BPA treatment or diet.

Figure 3.

BPA and social behavior separated into days 1–2 and 3–4 of testing sessions collapsed across sex because there were no interactions between sex and BPA. The data are also collapsed across diet. A. 0 dose BPA animals engaged in more play behavior than the animals exposed to 400 μg/kg/day, B. The 40 μg/kg/day BPA exposed animals spent more time alone than the 0 dose animals. C. Animals exposed to 400 μg/kg/day of BPA engaged in more passive contact than did the 0 dose animals.

There was a significant day × treatment interaction for time alone, F(1,160)=4.00, p=.047 (Fig. 3B). Again, days 1–2 was not different, while on days 3–4, 40 μg BPA/kg animals spent more time alone than the 0 dose animals (p<.05). There was a significant main effect of sex in the amount of time spent alone, F(1,160)=14.08, p<.001; with females spending more time alone than males. There were no significant differences in time alone based on maternal diet.

For passive contact, there was also a significant interaction between days and BPA treatment, F(2,160)=5.67, p<.01 (Fig. 3C). Once again, the differences did not occur on days 1–2, but on days 3–4, the 400 μg BPA/kg animals spent significantly more time in passive contact than the 0 dose group (p<.05). There were no significant main effects of BPA or diet.

In each of the social behaviors, there was no significant BPA treatment × sex interaction, which would have suggested an attenuation or introduction of sex differences due to BPA treatment.

3.4. Elevated Plus Maze

There were no significant differences due to BPA or diet in the EPM for any of the measures (time in open arms, time in closed arms, number of crossings into the open arms, number of crossings into the closed arms). There was a significant sex difference in the number of entries into open arms, with females entering the open arms more often, F(1,100)=8.52, p<.01 (Fig. 4). This is a frequently observed sex difference in the EPM [45].

Figure 4.

Perinatal BPA and HFD exposure did not have a significant long-term effect on behavior in the EPM (although see section 3.4 for CON animals alone). A. Time in the open arms. B. Time in the closed arms.

Previous research has shown a reversal of sex differences with BPA exposed males becoming less anxious [46]. In order to compare our results to previous literature, the analysis was restricted to the CON diet animals. There was a significant sex by BPA treatment interaction in the time spent in the closed arms, F(2,62)=3.84, p<.05, and time spent in the open arms, F(2,62)=3.69, p<.05. Post hoc analysis found a significant sex difference between males and females in the time spent in the closed arms (p<.01) and time spent in the open arms (p<.01), and 400 μg BPA/kg/day abolished the sex difference. Males showed a significant dose-dependent increase in the time in open arms with BPA treatment, F(2, 35)=3.70, p<.05, suggesting a anxiolytic effect of BPA. Females did not show significant differences due to BPA treatment.

3.5. Social Recognition

There were no significant differences in social recognition based on either BPA or diet (p>.05; Fig. 5).

Figure 5.

Perinatal BPA and HFD exposure did not result in significant long-term alterations in social recognition in adult animals (collapsed across sex).

3.6. Cytokine Levels in the P10 mPFC

The results of the ELISA assays on the mPFC of P10 pups are in Table 1 as a percent change compared to control animals (CON diet/0 dose BPA). There were main effects of sex in 7 of the 8 cytokines examined. Males had a higher percent change than females in the level of TNFα, TGFβ, MCP-1, IL-1α, IL-15, and VEGF. Females had a higher percent change in the level of IL-6 cytokine than males. The only cytokine that did not show a sex difference was IL-1β. There was a significant main effect of diet on IL-6, F(1,70)=11.25, p<.002. There were no main effects of BPA on the levels of cytokines, but there were significant interactions between sex and BPA exposure (Fig 6): TNFα F(2,68)=4.09, p<.03; MCP-1 F(2,65)=5.93, p<.005, and VEGF F(2,68)=6.81, p<.003. LSD posthoc comparisons found that males, but not females, had increased levels of TNFα, MCP-1 and VEGF following BPA exposure. Males had a significant increase in TNFα following both 40 and 400 μg BPA/kg (p<.03 and p<.001, respectively), and a significant increase in the levels of MCP-1 and VEGF following 400 μg BPA/kg exposure (p<.006 and p<.003, respectively).

Table 1.

Cytokine levels (percent change from control diet/0 BPA) in the mPFC of P10 pups

Diet		Control		High fat			ANOVA
BPA (μg/kg/day)		40	400	0	40	400
TNFα	M	13.4 ± 5.9	10.7 ± 9.8	−2.15 ± 2.0	−4.2 ± 5.7	12.0 ± 8.1	Sex p<.01
	F	5.0 ± 6.7	7.3 ± 5.9	3.8 ± 3.8	−2.6 ± 3.8	4.6 ± 5.9	Sex × BPA p=. 02
TGFβ	M	1.6 ± 5.2	8.0 ± 9.7	9.3 ± 8.4	−9.9 ± 9.1	1.6 ± 7.9	Sex p=.05
	F	0.9 ± 5.9	4.1 ± 6.3	1.9 ± 6.9	−4.5 ± 4.2	2.2 ± 8.7
MCP-1	M	4.6 ± 5.8	3.3 ± 5.5	0.4 ± 3.6	−1.2 ± 4.1	13.2 ± 7.5	Sex p<.001
	F	0.2 ± 5.7	6.9 ± 4.2	8.9 ± 7.6	7.4 ± 4.2	−0.3 ± 3.9	Sex × BPA p<.01
IL-1α	M	11.1 ± 3.9	1.2 ± 6.1	4.8 ± 3.4	−3.9 ± 3.4	14.1 ± 5.3	Sex p<.03
	F	6.3 ± 4.3	2.1 ± 5.3	−.02 ± 7.3	−0.9 ± 3.3	0.5 ± 4.2
IL-1β	M	8.7 ± 8.9	11.8 ± 8.7	−7.9 ± 3.8	−2.5 ± 3.3	13.9 ± 10.0
	F	12.0 ± 10.0	14.3 ± 8.8	5.1 ± 7.9	1.0 ± 4.5	7.7 ± 6.8
IL-6	M	12.4 ± 7.5	20.8 ± 9.1	0.6 ± 7.4	−3.9 ± 2.4	15.7 ± 12.3	Sex p<.001
	F	13.9 ± 8.4	18.6 ± 7.8	5.3 ± 11.5	3.5 ± 5.7	8.0 ± 8.4	Diet p<.001
IL-15	M	10.5 ± 7.2	6.2 ± 8.9	−1.9 ± 3.4	1.2 ± 8.3	15.7 ± 9.4	Sex p<.05
	F	18.0 ± 5.5	5.8 ± 6.1	−7.1 ± 7.7	−0.8 ± 4.1	3.5 ± 4.3
VEGF	M	10.3 ± 4.9	6.2 ± 7.8	−0.6 ± 2.2	−5.7 ± 5.1	16.9 ± 6.5	Sex p<.001
	F	6.9 ± 2.9	−2.3 ± 7.9	−4.6 ± 6.7	−4.7 ± 5.0	−.08 ± 2.9	Sex × BPA p<. 01

Values are expressed as the mean percent change from control (0 μg/kg/day BPA/control diet) ± SEM.

Figure 6.

The levels of pro-inflammatory cytokines that increased in males in the mPFC at P10 following perinatal BPA exposure, but not in females (significant interactions between BPA and sex). The data are collapsed across diet. A. Males exposed to 40 and 400 μg/kg/day BPA had higher levels of TNFα than the 0 dose group. B. Males exposed to 400 μg/kg/day BPA had higher levels of MCP-1 than the 0 dose groups. C. Males exposed to 400 μg/kg/day BPA had higher levels of VEGF than the 0 dose group. *p<.05; p<.006

In order to assess the possible associations between maternal behavior and cytokine levels of the offspring, a two-tailed Spearman’s correlation was conducted comparing maternal licking to each protein. Maternal licking was negatively correlated with levels of IL-1α (r= −.257, p<.02), IL-15 (r= −.243, p<.03) and VEGF (r= −.249, p<.02). The remaining cytokines were not significantly correlated with maternal licking.

3.7. Gene Expression in the P10 mPFC

The results of the qPCR analyses of the P10 mPFC are in Table 2. Females had higher expression than males in: Esr1 (p<.05), Sod1 (p<.04), Aif1 (p<.02), C4a (p<.001), and Casp3 (p<.005). There were main effects of BPA exposure in: Esr1, F(2,89)=7.39, p<.002, and Sod1, F(2,89)=3.75, p<.03. Post hoc comparisons found that animals exposed to 40 μg BPA/kg (p<.02) and 400 μg BPA/kg (p<.001) had higher expression levels of Esr1 in the PFC than 0 dose animals (Fig. 7A); likewise, Sod1 expression was significantly higher following both 40 and 400 μg BPA/kg doses compared to the 0 dose (p<.02). HFD exposure resulted in a significant decrease in gene expression of Sod2 F(1,89)=6.85, p<.02.

Table 2.

Gene expression from the mPFC of P10 male and female rats

Diet		Control			High Fat Diet			ANOVA
BPA		0	40	400	0	40	400
(μg/kg/day)
Hormone receptors and enzymes
Esr1	M	0.74 ± 0.06	0.99 ± 0.02	0.97 ± 0.06	0.85 ± 0.04	1.01 ± 0.06	1.08 ± 0.04	Sex p< .05
	F	0.98 ± 0.04	0.99 ± 0.06	1.05 ± 0.04	0.96 ± 0.06	0.94 ± 0.04	1.13 ± 0.11	BPA p< .002
Esr2	M	0.88 ± 0.09	0.93 ± 0.10	0.98 ± 0.11	1.20 ± 0.10	0.80 ± 0.10	0.79 ± 0.08
	F	0.81 ± 0.09	0.75 ± 0.06	0.75 ± 0.06	0.89 ± 0.12	0.95 ± 0.11	0.88 ± 0.10
Errg	M	0.59 ± 0.09	0.91 ± 0.18	0.91 ± 0.17	1.24 ± 0.24	0.78 ± 0.14	0.79 ± 0.21	Diet*BPA p<.02
	F	0.52 ± 0.06	0.64 ± 0.08	0.75 ± 0.10	1.08 ± 0.15	0.87 ± 0.11	0.82 ± 0.16
								Diet*BPA p<.02
Ar	M	1.02 ± 0.10	0.95 ± 0.07	0.89 ± 0.09	0.84 ± 0.09	0.88 ± 0.09	0.95 ± 0.10
	F	0.85 ± 0.09	0.76 ± 0.05	0.80 ± 0.08	0.85 ± 0.09	0.83 ± 0.06	0.88 ± 0.10
Star	M	1.32 ± 0.13	1.06 ± 0.09	1.13 ± 0.07	1.21 ± 0.09	1.20 ± 0.08	1.15 ± 0.08
	F	1.46 ± 0.09	1.17 ± 0.08	1.20 ± 0.11	1.22 ± 0.11	1.13 ± 0.08	1.19 ± 0.11
Cyp11a1	M	0.91 ± 0.12	0.65 ± 0.10	1.10 ± 0.21	1.14 ± 0.21	1.29 ± 0.23	1.18 ± 0.28
	F	0.79 ± 0.26	1.37 ± 0.20	1.24 ± 0.25	1.24 ± 0.15	1.30 ± 0.22	0.89 ± 0.03
Inflammation or Oxidative Stress Markers
Cat	M	0.96 ± 0.02	1.03 ± 0.02	0.98 ± 0.02	0.96 ± 0.04	0.97 ± 0.04	0.94 ± 0.01
	F	1.03 ± 0.04	1.00 ± 0.02	1.04 ± 0.03	0.98 ± 0.02	0.98 ± 0.01	1.04 ± 0.06
Sod1	M	0.86 ± 0.06	0.99 ± 0.03	0.96 ± 0.02	0.91 ± 0.03	1.01 ± 0.04	1.03 ± 0.03	Sex p<.04
	F	1.00 ± 0.01	1.04 ± 0.04	0.97 ± 0.05	0.96 ± 0.02	1.03 ± 0.03	1.07 ± 0.04	BPA p<.03
Sod2	M	1.15 ± 0.01	1.18 ± 0.09	1.05 ± 0.05	1.02 ± 0.04	1.08 ± 0.08	1.04 ± 0.08	Diet p<.02
	F	1.18 ± 0.08	1.25 ± 0.11	1.04 ± 0.08	0.97 ± 0.08	0.93 ± 0.11	0.94 ± 0.15
Glrx	M	1.06 ± 0.03	1.01 ± 0.03	1.00 ± 0.01	0.95 ± 0.05	1.02 ± 0.04	1.05 ± 0.07	Diet × BPA p<.04
	F	1.18 ± 0.06	0.93 ± 0.16	0.83 ± 0.15	1.06 ± 0.05	1.10 ± 0.04	1.24 ± 0.12
Aif1	M	1.12 ± 0.13	1.08 ± 0.08	1.12 ± 0.06	1.09 ± 0.10	1.24 ± 0.06	1.13 ± 0.08	Sex p<.02
	F	1.26 ± 0.11	1.23 ± 0.09	1.31 ± 0.10	1.39 ± 0.10	1.17 ± 0.14	1.44 ± 0.11
Complement Pathway
C4a	M	0.99 ± 0.11	0.78 ± 0.04	0.83 ± 0.08	0.82 ± 0.05	0.89 ± 0.04	0.93 ± 0.07	Sex p<.001
	F	1.00 ± 0.06	0.93 ± 0.05	1.13 ± 0.14	1.18 ± 0.11	0.97 ± 0.05	1.29 ± 0.09
C4b	M	1.34 ± 0.14	0.79 ± 0.08	0.94 ± 0.15	1.17 ± 0.13	1.13 ± 0.11	1.11 ± 0.26
	F	0.95 ± 0.18	1.15 ± 0.14	1.19 ± 0.12	1.24 ± 0.19	1.08 ± 0.13	1.46 ± 0.24
Apoptosis
Casp3	M	0.94 ± 0.03	0.99 ± 0.03	0.88 ± 0.03	0.90 ± 0.04	0.89 ± 0.07	0.99 ± 0.05	Sex p<.005
	F	1.00 ± 0.05	1.12 ± 0.04	1.03 ± 0.07	0.96 ± 0.05	1.00 ± 0.03	1.14 ± 0.13	Diet × BPA p=.056

Data expressed as mean ± SEM ratio to RpL7a.

Figure 7.

Alterations in gene expression in the mPFC of P10 male and female pups following perinatal BPA and HFD. A. Esr1 expression increased with perinatal exposure to BPA (BPA p<.002). The data are collapsed across diet. B. Errg expression was higher in HFD/0 dose BPA animals than in CON diet/0 dose BPA animals and in HFD/40 μg/kg/day BPA exposure (diet × BPA p<.02). C. Glrx expression was increased in animals exposed to HFD and 400 μg/kg/day BPA compared to HFD animals, 0 dose BPA. No changes occurred in CON diet groups (diet × BPA p<.04) *p<.05; **p<001.

The only significant interactions in gene expression between BPA treatment and diet were found in Errg, F(2,89)=4.47, p<.02 (Fig. 7B) and Glrx, F(2,89)=3.46, p<.04 (Fig. 7C). Post hoc analysis showed that the HFD/0 BPA group had higher Errg levels than both CON diet/0 BPA (p<.001), HFD/40 μg/kg/day (p<.05). Glrx was higher at 400 μg BPA exposure only in the HFD group (p<.05). The remainder of the genes did not show significant differences: Esr2, Ar, Star, Cyp11a1, Cat, and C4b.

In order to assess the possible associations between maternal behavior and resulting gene expression of the offspring, a two-tailed Spearman’s correlation was conducted comparing each maternal behavior to each gene. None of the behaviors or gene expression levels were significantly correlated (data not shown).

3.8. Gene Expression in the P90 Cortex

The results of the gene expression analysis on the cortex at P90 are in Table 3. Females have higher level of gene expression in: Sod2 (p<.001), Aif1 (p<.04), and Casp3 (p=.05), while males had higher levels of gene expression in C4a (p<.001) and C4b (p<.05). HFD resulted in an increase in the expression of genes related to hormone receptors, oxidative stress and apoptosis, including Esr1, F(1,66)=4.63, p<.05; Errg, F(1,66)=4.37, p<.05; Cat, F(1,66)=8.92, p<.005; and Casp3, F(1,65)=4.74, p<.05. There were no main effects of BPA on any of the analyzed genes.

Table 3.

Gene expression from the mPFC of adult male and female rats at P90

Diet		Control			High Fat Diet			ANOVA
BPA		0	40	400	0	40	400
(μg/kg/day)
Hormone receptors and enzymes
Esr1	M	0.92 ± 0.12	1.00 ± 0.06	0.81 ± 0.08	1.10 ± 0.10	0.98 ± 0.08	0.88 ± 0.11	Diet p<.05
	F	0.84 ± 0.06	0.84 ± 0.12	1.01 ± 0.19	0.86 ± 0.14	1.20 ± 0.12	1.32 ± 0.08	BPA × Sex p<.05
Esr2	M	1.10 ± 0.09	1.08 ± 0.15	0.76 ± 0.11	1.03 ± 0.10	0.91 ± 0.16	0.99 ± 0.09	Diet × BPA p<.05
	F	1.05 ± 0.08	0.88 ± 0.10	0.78 ± 0.07	0.88 ± 0.15	1.01 ± 0.08	1.15 ± 0.08
Errg	M	1.19 ± 0.14	1.37 ± 0.20	1.05 ± 0.15	1.36 ± 0.09	1.32 ± 0.11	1.45 ± 0.08	Diet p<.05
	F	1.48 ± 0.15	1.05 ± 0.12	1.23 ±0.11	1.41 ± 0.08	1.50 ± 0.13	1.29 ± 0.05
Ar	M	1.00 ± 0.11	0.93 ± 0.13	0.74 ± 0.09	0.84 ± 0.06	0.88 ± 0.03	0.85 ± 0.06	Diet × BPA p<.03
	F	1.13 ± 0.07	0.81 ± 0.10	1.00 ± 0.10	0.84 ± 0.09	1.07 ± 0.10	1.03 ± 0.05
Star	M	0.91 ± 0.15	0.95 ±0.20	0.69 ± 0.14	0.90 ±0.14	0.89 ±0.07	0.98 ± 0.10
	F	1.20 ± 0.14	0.81 ± 0.10	0.98 ± 0.14	0.83 ± 0.12	1.01 ± 0.09	1.06 ± 0.03
Cyp11a1	M	1.05 ± 0.24	1.26 ± 0.19	1.24 ± 0.54	0.77 ± 0.25	1.16 ± 0.41	1.13 ± 0.49
	F	1.27 ± 0.35	1.37 ± 0.40	1.80 ± 0.44	1.85 ± 0.70	1.17 ± 0.27	1.67 ± 0.38
Inflammation or Oxidative Stress Markers
Cat	M	1.03 ± 0.08	1.00 ± 0.07	0.87 ± 0.07	1.19 ± 0.06	1.05 ± 0.03	1.18 ± 0.06	Diet p<.005
	F	1.07 ± 0.04	0.98 ± 0.12	0.98 ± 0.12	1.16 ± 0.10	1.13 ± 0.06	1.12 ± 0.04
Sod1	M	1.00 ± 0.07	1.00 ± 0.08	0.95 ± 0.04	1.07 ± 0.10	0.97 ± 0.02	0.98 ± 0.05
	F	1.07 ± 0.05	0.91 ± 0.10	1.05 ± 0.08	1.00 ± 0.09	1.10 ± 0.09	1.11 ± 0.04
Sod2	M	0.81 ± 0.08	0.69 ± 0.14	0.78 ± 0.10	0.83 ± 0.12	0.81 ± 0.09	0.61 ± 0.09	Sex p<.001
	F	0.89 ± 0.06	0.97 ± 0.14	0.98 ± 0.12	0.85 ± 0.13	1.26 ± 0.12	1.27 ± 0.16
Glrx	M	1.08 ± 0.12	1.00 ± 0.11	0.90 ± 0.10	0.96 ± 0.08	0.88 ± 0.07	0.87 ± 0.07	Diet × BPA p<.05
	F	1.16 ± 0.06	0.86 ± 0.11	0.88 ± 0.08	0.92 ± 0.08	1.20 ± 0.08	1.30 ± 0.13	Diet × Sex p<.05
Aif1	M	1.04 ± 0.15	1.16 ± 0.18	0.91 ± 0.13	1.00 ± 0.09	0.94 ± 0.11	1.17 ± 0.07	Sex p<.05
	F	1.14 ± 0.13	1.17 ± 0.07	1.10 ± 0.14	1.26 ± 0.15	1.30 ± 0.15	1.28 ± 0.05
Complement Pathway
C4a	M	0.84 ± 0.17	1.16 ± 0.10	0.93 ± 0.13	1.19 ± 0.16	1.00 ± 0.09	1.34 ± 0.19	Diet × BPA × Sex p<.05
								Sex p<.001
	F	0.85 ± 0.04	0.66 ± 0.04	0.64 ± 0.06	0.77 ± 0.08	0.91 ± 0.08	0.76 ± 0.04
C4b	M	0.90 ± 0.08	0.91 ± 0.15	0.93 ± 0.24	0.82 ± 0.10	0.81 ± 0.14	1.06 ± 0.09	Sex p<.05
	F	0.81 ± 0.04	0.59 ± 0.09	0.80 ± 0.11	0.61 ± 0.07	0.91 ± 0.11	0.74 ± 0.10
Apoptosis
Casp3	M	0.80 ± 0.06	0.81 ± 0.04	0.69 ± 0.06	0.85 ± 0.05	0.74 ± 0.04	0.80 ± 0.06	Diet p<.05
	F	0.87 ± 0.05	0.76 ± 0.08	0.74 ± 0.07	0.83 ± 0.04	0.97 ± 0.05	0.97 ± 0.07	Sex p=.05

Data expressed as mean ± SEM ratio to RpL7a.

However, BPA interacted with HFD in several genes: Ar, F(2,66)=3.82; p<.05; Esr2, F(2,66)=3.40; p<.05, Glrx, F(2,64)=3.48; p<.05. Post hoc tests on Ar showed a significant decrease in gene expression following perinatal HFD exposure compared to CON diet only in the 0 BPA group (p<.02). Further, Ar was significantly decreased between 0 and 40 μg/kg/day BPA (p=.05) only in the CON diet animals. Post hoc tests on Esr2 showed that only within the CON diet groups was there a significant reduction in the level of gene expression at 400 μg compared to 0 BPA exposure (p<.02). For Glrx, there was only a trend in post hoc tests between 0 and 400 μg BPA within the CON diet (p=.057). In addition, there was a HFD by sex interaction in Glrx, F(1,64)=4.84; p<.05. Female animals in the HFD/400 μg BPA group had a significant increase in the level of gene expression compared to the 0 BPA group (p<.04), while males did not have significant change.

A BPA by sex interaction was found in expression of Esr1, F(2,66)=3.61, p<.05, and a post hoc test revealed an increase in the level of gene expression in females following 400 μg/kg/day BPA exposure (p<.05), but no differences in males (Fig. 8). There was a three-way interaction of HFD, BPA and sex in C4a, F(2, 65)=3.38, p<.04. However, post hoc tests revealed no significant differences when treated groups were compared to controls (0 dose BPA). The interaction was driven by the different responding of females and males to diet and BPA, but none were significantly different from their own control.

Figure 8.

The increased levels of gene expression of Esr1 in P90 females exposed to 400 μg/kg/day compared to controls; this did not occur in male animals. The data are collapsed across diet. *p<.02

No HFD or BPA effects or interactions were observed in gene expression of Aif1, Star, Cyp11a21, Sod1, C4b, Sod2.

3.9. Esr1 gene expression, premRNA content and promoter DNA methylation

The consistent increase in expression of Esr1 in females due to BPA exposure between P10 and P90 (Fig. 7A and Fig. 8) was further investigated through analysis of the methylation of the Esr1 gene (Esr1) in the P90 mPFC tissue. Comparisons were made to males that did not have a long-term increase. The hypothesis was that the control (0 dose BPA) females had higher levels of methylation than those that had been exposed to BPA especially at the 400 μg/kg/day dose. Despite the changes in gene expression, premRNA (the immature transcript before splicing to mRNA) content of Esr1 in PFC was not altered for males or female animals (Suppl. Fig. S3). Additionally, analysis of CG rich sequences contained in the promoter of the rat gene revealed 4 CpG dense islands and upstream regions. These regions were selected for DNA methylation analysis through Methylation Specific PCR. Regions 1 (+2043 to +2131bp) and 4 (+4755 to +4847bp) showed the highest variation for females and males but no significant effect was detected (Suppl. Fig. S1).

3.10. miRNA analysis of Esr1 in the mPFC

Analysis of the three miRNAs associated with Esr1 revealed that the miRNA levels were not significantly altered by BPA administration as none of the miRNA expression levels were increased by 0.5 or more (Suppl. Fig. S4 and S5).

4. Discussion

The effects of simultaneous perinatal exposure to both BPA and HFD were investigated. BPA influenced adolescent play behavior, the elevated plus maze, and in the mPFC, inflammatory markers and gene expression at P10 and P90,. HFD affected maternal behavior and the expression of a subset of genes. Notably, perinatal BPA and HFD rarely interacted on the measures in this study; the exception being the expression of a small number of genes that did not change in the same direction across ages.

4.1. Maternal behavior

There were separate effects of HFD and BPA exposure on maternal behavior. The dams fed HFD spent less time away from the nest and marginally more time nursing. These effects of HFD on maternal care are supported by previous work showing an increase in nursing and decrease in resting away from nest with HFD ingestion [26, 8, but see also 24]. Thus 45% HFD during the metabolically demanding times of pregnancy and postnatal nursing can facilitate maternal behavior. BPA exposure at the higher dose, on the other hand, resulted in a near significant decrease in the amount of licking behavior exhibited by the dam. This mirrors a pattern in preliminary data from our laboratory in which 400ug/kg/day BPA resulted in less licking than 40ug/kg/day [47]. During the maternal observations, BPA was being administered directly to the pups, not the dam. So the decrease in licking by the dams may be due to BPA-induced changes in the pups, perhaps by alterations in behavior or olfactory cues. Increased maternal licking of pups has been linked to multiple behavioral and cellular alterations in the offspring, including a decrease in the behavioral stress response [1] This is supported in our data with the negative correlation between maternal licking behavior and a reduction in inflammatory cytokines, IL-1α, IL-15 and VEGF.

4.2. Offspring behaviors

Perinatal BPA exposure resulted in several significant long-term alterations on offspring behavior, while perinatal exposure to 45% HFD did not.in the current study. Periadolescent play behavior was decreased in animals exposed to BPA exposure in the latter half of 4 days of observations. Animals in the 0 dose BPA group spent more time engaged in play behaviors than the 400 μg BPA/kg dose, while animals exposed to 40 μg dose spent more time alone than 0 dose, and animals exposed to the 400 μg BPA/kg dose were spent more time in passive contact than 0 dose animals. These results of periadolescent play behavior are consistent with previous studies that found a decrease in play following 40 μg/kg BPA exposure [48]. Ferguson et al. [49] also showed a pattern of decrease in play behavior following prenatal BPA exposure that did not reach significance at the lower doses used (2.5 and 25 μg). Although overall sex differences were found with males spending more time playing and females spending more time alone, there were no sex-specific changes due to BPA in these measures.

Along with the striatum and medial amygdala, play behavior is mediated by the mPFC, with inactivation of either the prelimbic or infralimbic subregions of the mPFC significantly reducing play [50, 51]. Because play behavior was assessed weeks after BPA exposure ended, these data suggest that perinatal BPA exposure may alter the long-term structure of several neural regions including the mPFC.

Evidence of an anxiolytic effect on the EPM due to BPA exposure in males was only uncovered when the CON diet group was analyzed separately, which supports previous literature that does not have a diet component [46, 52; but see also 53, 54]. 0 dose BPA male animals spent more time in the closed arms, but the 400 μg/kg/day BPA exposed males increased their time on the open arms to female levels. We found that perinatal BPA exposure to 400 μg/kg/day abolished the sex difference with an anxiolytic effect in males, which is in agreement with the literature [52].

Anxiety behavior is influenced by the mPFC through cortical-amygdalar circuitry [55]. Both excitotoxic lesions and infusions of muscimol into the mPFC, which inactivate GABAa receptors, reduce anxious behavior in the EPM [56, 57]. Alterations in the anatomy of the mPFC by BPA could also affect anxiety behavior displayed in the EPM, and further research into the neuroanatomical effects of BPA on the mPFC is currently underway.

4.3. Inflammatory cytokines in the mPFC

There was evidence in the present study that male, but not female, offspring had increased levels of pro-inflammatory cytokines, TNFα, MCP-1 and VEGF, in the mPFC at the highest dose of BPA. The sex-specific response to perinatal BPA administration may be due to the sex differences in the immune system that are found early in development [reviewed in 58]. The increased presence of cytokines in the early developing brain may indicate inflammation due to BPA exposure. Inflammation may also shape the behavioral outcomes of the offspring. Previous research has found an increase in anxiety-like behavior and an increase in IL-6 and TNFα following perinatal BPA exposure [6]. Here, the highest dose of BPA produced the increased levels of inflammation in the early postnatal male animals, but the adult male animals showed an anxiolytic effect on of BPA administration, instead of the predicted anxiogenic behavior on the EPM.

Interestingly, perinatal exposure to HFD resulted in a decrease in the protein level of IL-6 in the mPFC, which is contrary to what was reported in the hippocampus and frontal cortex of young animals [28]. However, our study was focused on the mPFC, and the impact of perinatal HFD may be regiondependent. Also, the necessary handling of pups for BPA administration during the first 10 postnatal days could have masked effects since early handling has a potent influence on later behavior [59] and its effects on inflammation are unknown. The current study adds to the evidence that BPA affects cortical inflammation during the early postnatal period and does not appear to interact with a 45% HFD during this exposure period.

4.4. Gene expression in the mPFC at P10 and P90

Of the analyzed genes, only Esr1 was comparably altered at both P10, at the end of exposure, and at P90 in females, which may indicate long-term changes due to BPA. At P10, perinatal BPA exposure significantly increased the gene expression of Esr1 in the mPFC in both sexes. The increase in Esr1 expression seen here is similar but larger than that reported by Kundakovic et al. [16] who found a marginally significant U-shaped curvilinear effect at P28 in mice. Our study included direct BPA administration to the pups from P1–10 in addition to gestational exposure, which may explain the difference in strength of the results. The levels of Esr1 mRNA expression start to decrease in normative studies of P10 rats [60, 61], but in the current study, animals exposed to the high dose of BPA had increased levels of Esr1 expression. Early developmental alterations in the expression of Esr1 could change the sensitivity of the animals to estrogens, particularly if the increase in Esr1 persists, as we observed in the present study. This has also been postulated by Cao et al. [62], who found increases in gene expression of Esr1 in the hypothalamus due to prenatal BPA exposure. At P90, Esr1 continued to be higher in females that had perinatal BPA exposure. Interestingly, Esr1 was the only hormone-related gene in which BPA effects were the same at both P10 and adulthood. The increase in the expression of Era in the mPFC of females persisted well after the final administration of BPA, which could suggest that BPA increases the level of protein transcription.

The consistent increase in Esr1 expression between P10 and P90 in females exposed to BPA was further explored. Several epigenetic mechanisms have been studied in regard to BPA in the brain, such as DNA methylation [16, 63, 64] and histone modifications [65]. In the current study, no clear pattern of methylation was obtained within regions of the gene that defined BPA exposure. In line with this result, the premRNA, or unprocessed transcript, was not significantly different between BPA treatments, concordant with the low impact of DNA methylation on the transcriptional control of the Esr1 gene. DNA methylation mediates polymerase II binding and transcription factor accessibility to regulatory regions; however, the lack of active transcription could indicate that changes in mRNA expression are due to downstream stabilizing or destabilizing epigenetic regulators such as miRNA. However, of the three miRNAs identified, none were significantly increased in 0 dose BPA females, suggesting a different miRNA or mechanism is responsible for the long-term increase in Esr1 expression.

Gene expression of another hormone-related receptor, Errg was also influenced by BPA and HFD at each time point but not in the same direction. At P10, HFD increased the level of Errg expression but in combination with BPA, the levels were not changed. In adulthood there was no effect of BPA on Errg, but HFD resulted in a decrease in expression. ERRγ is an orphan nuclear receptor, which does not bind estrogen [66], but strongly binds BPA [67]. It is likely that ERRγ is, directly or indirectly, a part of the mechanism of action for BPA effects. When activated, ERRγ is able to modulate estrogen receptordependent signaling and even stimulate estrogen response element-mediated transcription without the presence of another ligand [66]. ERRγ has been only recently discovered, and the physiological impact of alterations in its gene expression is unknown. Higher levels of adiposity can increase the level of circulating estrogens [68] and the number of ER binding sites in peripheral tissues [69]. However, this is the first reported evidence of an increase in Errg expression with HFD. The dearth of information on the function of ERRγ leaves the physiological effect of the changes in its expression unclear.

There were several genes that changed their expression in the mPFC due to sex and perinatal exposure to BPA and HFD at either P10 or P90. Expression of genes related to oxidative stress and inflammation were affected by BPA and HFD at P10 and at P90, but rarely at both ages. The levels of Glrx expression showed an interaction between BPA and HFD with similar patterns at P10 and P90; however, what was significantly different from 0 μg/kg/day BPA differed between the ages. In addition, other genes related to inflammation were also altered at P90. Cat and Casp3 were increased following perinatal HFD, while C4a is altered by HFD and BPA in a sex-specific manner. Males had an increase in expression of C4a following perinatal exposure to both BPA and HFD; however, females generally showed a decrease in expression of C4a. Notably, neither HFD nor BPA cause an overall increase in the inflammatory markers, which suggests that not all of the inflammatory pathways are affected by BPA or HFD under these conditions.

4.5. Conclusion

The present study showed that perinatal exposure to BPA has effects on adolescent social behavior that tend to occur at the higher dose. Inflammatory markers and gene expression in the mPFC were also affected. BPA effects were often sex-specific and more reliable at higher doses. Of particular interest is the lasting effect of BPA on the expression of Esr1 in the mPFC. The impact of increased expression of Esr1 in the mPFC is currently unknown, but may indicate increased responding to estrogens. With regard to maternal diet, HFD (45% fat) restricted to the perinatal period had limited effects on the variables assessed in the offspring but did acutely affect the dams’ maternal care. The effects of BPA and HFD rarely interacted in the current paradigm.

Supplementary Material

1

Supplemental Figure S1. CG rich sequences in the promoter of the Esr1 rat gene revealed 4 CpG dense islands and upstream regions within the promoter were selected for DNA methylation analysis through Methylation Specific PCR. Regions 1 (+2043 to +2131bp) and 4 (+4755 to +4847bp) showed the highest variation for females and males but no significant effect was detected.

Supplemental Figure S2. A. Venn diagram to exemplify each search strategy, B. Specific localization of each miRNA within the 3’-UTR of the Esr1 rat gene (4037 bp) is illustrated. Only rno-mir-19b-3p and rno-mir-22–3p showed at least two individual binding sites along this region, C. Specific seed and binding sequence within Esr1.

Supplemental Figure S3. Era premRNA content at P90 was not altered following BPA administration, suggesting that methylation was not the underlying mechanism leading to the long-term increase in Era gene expression.

Supplemental Figure S4. The levels of expression of the housekeeping gene U6 did not show a significant effect following BPA administration.

Supplemental Figure S5. Three miRNAs associated with Esr1 were chosen for gene analysis (A-C). Male animals exposed to 400 μg/kg/day BPA had a higher level of mir-19b-3p expression, which was increased greater than 0.5 fold compared to controls (A), but did not show a higher level of expression of mir-22b-3p or mir-221b-3p (B, C). Females did not show changes in miRNAs following perinatal BPA exposure.

Supplemental Figure 5. Three miRNAs associated with Esr1 were chosen for gene analysis (A-C). Male animals exposed to 400 μg/kg/day BPA had a higher level of mir-19b-3p expression, which was increased greater than 0.5 fold compared to controls (A), but did not show a higher level of expression of mir-22b-3p or mir-221b-3p (B, C). Females did not show changes in miRNAs following perinatal BPA exposure.