Hormones and Behavior 107 (2019) 26–34

Contents lists available at ScienceDirect

Hormones and Behavior

journal homepage: www.elsevier.com/locate/yhbeh

Effects of adolescent Bisphenol-A exposure on memory and spine density in T

ovariectomized female rats: Adolescence vs adulthood
⁎
Rachel E. Bowmana, , Jennifer Hagedorna, Emma Maddena, Maya Frankfurtb
a
Department of Psychology, Sacred Heart University, Fairfield, CT 06825, United States of America
b
Department of Science Education, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, NY 11549, United States of America

A R T I C LE I N FO A B S T R A C T

Keywords: The endocrine disruptor, Bisphenol-A (BPA), alters many behavioral and neural parameters in rodents. BPA
Adolescence administration to gonadally intact adolescent rats increases anxiety, impairs spatial memory, and decreases
Bisphenol-A dendritic spine density when measured in adulthood. Since BPA's action seems to be mediated through gonadal
Estradiol steroid receptors, the current experiments were done in ovariectomized (OVX) female rats to examine the effects
Hippocampus
on behavior and spine density of adolescent BPA exposure under controlled hormone conditions. OVX (postnatal
Prefrontal cortex
Dentate gyrus
day, PND, 21) female Sprague-Dawley rats (n = 66) received subcutaneous injections of BPA (40 μg/kg/body-
Dendritic spine weight), 17β-Estradiol (E2, 50 μg/kg/bodyweight), or saline during adolescence (PND 38–49). Following the last
Memory injection brains were processed for Golgi impregnation (Exp1), behavioral and spine density in adolescence
(Exp2), or in adulthood (Exp3). In Exp1, E2 increased spine density in CA1 pyramidal cells and BPA decreased
spine density in granule cells of the dentate gyrus (DG). In Exp2, BPA impaired spatial memory on the object
placement (OP) task, E2 increased spine density in CA1, BPA decreased spine density in the DG and the medial
prefrontal cortex (mPFC). When measured in adulthood (Exp3), BPA impaired OP and object recognition (OR)
performance, E2 increased spine density in CA1, and BPA decreased spine density in CA1, the mPFC and the DG.
Results provide novel data on the effects of adolescent BPA in an OVX model and are compared to data in intact
animals and within the context of understanding the importance of the profound neuronal alterations occurring
during adolescent development.

1. Introduction et al., 2015) suggests that the mechanism underlying the behavioral
effects of BPA also involve synaptic plasticity, which is supported in the
Bisphenol-A (BPA), is a synthetic compound commonly found in literature (Elsworth et al., 2015; Hu et al., 2017; Mhaouty-Kodja et al.,
plastics and other materials which has been demonstrated to have be- 2018; Ogiue-Ikeda et al., 2008). Moreover, several studies demonstrate
havioral effects in laboratory animals and humans (for review see that BPA can block gonadal steroid induction of dendritic spines in both
Mhaouty-Kodja et al., 2018). It has been suggested that the endocrine the hippocampus (CA1) and the medial prefrontal cortex (mPFC) in the
disrupter properties of BPA are primarily due to its ability to bind to rat (Inagaki et al., 2012), and inhibit synaptogenesis in CA1 and the
both estrogen and androgen receptors in the CNS (Ahmed et al., 2018; mPFC in both rats (Leranth et al., 2008b; MacLusky et al., 2005) and
Mhaouty-Kodja et al., 2018; Nesan et al., 2018). However it must be primates (Leranth et al., 2008a), two regions of the brain critical to
noted that BPA has also been shown to antagonize thyroid effects at the memory.
receptor level (Moriyama et al., 2002) as well as altering other endo- Most of studies on BPA have been conducted during the perinatal
crine systems (reviewed by Negri-Cesi, 2015). Many studies have de- period. Our studies are designed to assess the potential effects of BPA
monstrated that administration of BPA to pre and postnatal rats results exposure during the adolescent period when the brain is undergoing
in impaired memory (Bowman et al., 2015; Diaz Weinstein et al., 2013; rapid neurodevelopmental changes and may be extremely vulnerable
Eilam-Stock et al., 2012; Goncalves et al., 2010; Inagaki et al., 2012; (Juraska et al., 2013; Schneider, 2013; Shapiro et al., 2017). To date we
Mhaouty-Kodja et al., 2018; Wang et al., 2014). The importance of have assessed the effects of short-term, low-dose levels of BPA exposure
synaptic plasticity in mediating the well-established effects of estrogen during adolescent development, below the current reference safe daily
on memory (Frick et al., 2015; Luine et al., 2018; Luine, 2014; Tuscher limit of 50 μg/kg/day set by the United States Environmental Protection

⁎
Corresponding author at: Department of Psychology, Sacred Heart University, 5151 Park Avenue, Fairfield, CT 06825, United States of America.
E-mail address: [email protected] (R.E. Bowman).

https://doi.org/10.1016/j.yhbeh.2018.11.004
Received 26 July 2018; Received in revised form 9 November 2018; Accepted 14 November 2018
0018-506X/ © 2018 Elsevier Inc. All rights reserved.
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

Fig. 1. Overview of experimental research design timeline.

Agency, (United States Environmental Protectional Agency, 1993) on 2.2. Injections

intact adolescent and adult male and female rats. When administered
and assessed in adolescence, BPA increased anxiety, impaired spatial Subjects were randomly assigned to a control (vehicle control), BPA,
memory (Diaz Weinstein et al., 2013) and decreased dendritic spine or 17β-Estradiol (E2) treatment group. BPA (≥99% purity grade) and
density in the mPFC and CA1 of both male and female rats (Bowman E2 (≥98% purity grade) were obtained from Sigma-Aldrich Corp (St.
et al., 2014). Except for some sex differences, the behavioral and Louis, MO). Each rat received daily subcutaneous (SC) injections of BPA
morphological changes observed in adolescence were maintained when (40 μg/kg/bodyweight), 17β-Estradiol (E2, 50 μg/kg/bodyweight), or
adolescent BPA exposed subjects were assessed in adulthood (Bowman saline during adolescence (PND 38–49). The BPA and E2 were initially
et al., 2015; Bowman et al., 2014). dissolved in ethanol and diluted with saline (BPA) or corn oil (E2) for
Importantly, our previous studies were conducted in gonadally in- the stock solutions (final ethanol concentration < 0.006%, Inagaki
tact rats, which precluded assessing whether previously observed ef- et al., 2012). In the present study SC injections of BPA were used be-
fects were due to BPA or natural fluctuations in gonadal hormones. cause it has been shown that circulating BPA levels in neonatal mice are
Additionally, estrogen manipulation alters many aspects of brain the same following oral or SC exposure (Taylor et al., 2008) and no
structure and function including memory and dendritic spine density in differences were observed between SC and oral administration of BPA
adult and aging rats (Galea et al., 2017; Luine, 2016; Luine and on the reduction in spine synapses in hippocampus and prefrontal
Frankfurt, 2013; Luine, 2014; Scharfman et al., 2007). However, es- cortex in adult rats (Hajszan and Leranth, 2010). Furthermore, the
trogen replacement studies in adolescence in ovariectomized (OVX) rats dosing paradigm used in the current study is the same as all of our
are limited. Thus, we investigated the effects of adolescent BPA ex- previous BPA-related work (Bowman et al., 2015; Bowman et al., 2014;
posure on behavioral measures and dendritic spine density in OVX fe- Diaz Weinstein et al., 2013). The internal dose equivalent associated
male rats. The present series of studies was designed to directly com- with our dosing paradigm is approximately three ng/ml based on the
pare the adolescent effects of BPA or estradiol on both behavior and Taylor et al. (2008) study.
dendritic spine density in adolescent and adult OVX female rats. In
addition, since it is often hard to determine whether behavior alters
spine density or the reverse, rats in the first experiment were sacrificed 2.3. Experimental timeline
immediately after their last injection. Specifically, we examined ado-
lescent BPA or E2 exposure on the following: (1) dendritic spine density Fig. 1 illustrates the overall research design timeline for the three
in the adolescent brain in the absence of behavior, (2) on behavioral experiments. In Experiment 1, a cohort of animals (n = 18) was sacri-
measures and dendritic spine density in adolescence, and (3) behavioral ficed immediately following injections on PND 49. As described below,
measures and dendritic spine density in adulthood. brains were processed for Golgi impregnation. In Experiment 2
(n = 24), injections were immediately followed by behavior testing
during adolescence on PND 49–58. Subjects were sacrificed PND 59 and
2. Materials and methods brains were processed for Golgi impregnation. In Experiment 3, subjects
(n = 24) were aged until young adulthood and received behavioral
2.1. Subjects testing PND 77–86. Subjects were sacrificed PND 87 and brains were
processed for Golgi impregnation.
Experimentally naïve Sprague Dawley female rats (n = 66) were
obtained from Charles River Laboratories (Raleigh, NC, USA). All sub-
jects were bilaterally ovariectomized (OVX) under anesthesia (keta- 2.4. Behavioral measures
mine, 80 mg/kg bodyweight, and xylazine, 10 mg/kg bodyweight) by
the vendor on post-natal day (PND) 21 allowed to recover, and arrived Behavioral measures were conducted in designated behavioral
PND28. Subjects were double housed according to treatment condition testing rooms (21.1 °C, 43 lm/m2) and all behavioral testing occurred
in a common animal colony room, with temperature at 21.1 °C and between 9:00 and 14:00 h. Animals received one behavioral test/day
maintained on a 12/12-h light/dark cycle (lights on 7:00 am), weighed for ten days. The elevated plus maze (EPM) testing was conducted first
weekly, had free access to rat chow (Harlan 2018 Teklad Global) and (behavioral testing day 1), followed by open field (behavioral testing
water (Glass water bottles, Ancare Corporation, Bellmore, NY). All ex- day 2), and object placement (OP, behavioral testing days 3–6) and
perimental procedures were approved by the Sacred Heart University object recognition (OR, behavioral testing day 7–10) trials. To minimize
Institutional Animal Care and Use Committee and in accordance with olfactory cues, all behavioral equipment was wiped down with disin-
the NIH Guide for the Care and Use of Animals. fectant spray and allowed to dry between each individual trial/animal.

27
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

2.5. Elevated plus maze memory (the mPFC). At sacrifice, brains were removed from subjects,
cut into an anterior block (anterior to the optic chiasm) and posterior
Anxiety was measured using the EPM, which is 50.8 cm high and block (between the optic chiasm and the brainstem), and placed in
consists of two open arms, 50.8 × 12.7 cm, and two enclosed arms, solutions provided in the Rapid Golgi Stain Kit (FD NeuroTechnologies,
50.8 × 12.7 × 40.6 cm, with an open top. The two open arms and the Ellicott City, MD). Golgi impregnation was performed as previously
two closed arms are arranged opposite one another around a 10.2 cm described (Frankfurt et al., 2011; Inagaki et al., 2012). Secondary basal
square center. Each rat received one trial on the EPM lasting 5 min dendrites and tertiary apical dendrites were analyzed blindly from
during which the number of entries and duration of time spent in open pyramidal cells from the CA1 region of the dorsal hippocampus and
and closed arms was recorded. An entry occurred when the animal layer II/III of the prelimbic portion of the mPFC and granule cells of the
placed both forelimbs and more than half of their trunk into the arm. suprapyramidal blade of the dentate gyrus (DG). Six cells per region/
The number of visits to and duration of time spent in open arms is used brain were included in the analysis and approximately 6 brains were
as an anxiety index, with fewer open arm visits and/or decreased time quantified per group. Neurons in all areas were chosen for analyses as
spent on open arms being indicative of increased anxious behavior. follows: (1) cell bodies and dendrites were well impregnated; (2) den-
drites were clearly distinguishable from adjacent cells and continuous.
2.6. Open field Spines were counted under oil (100×) using a hand counter and den-
dritic length measured using the Olympus CellSens 1.16 software and
All animals were tested on the OF for measures of locomotor activity an Olympus BX-41 microscope. Spine density was calculated by di-
(peripheral grid visits) and anxiety-related behaviors (central grid viding spine number by the length of the dendrite and data expressed as
visits) (Tovote et al., 2015). Rats were placed one at a time in the OF number of spines/10 μm dendrite.
from a common starting area of a 117 × 70 × 45 cm wooden open top
box, with the floor divided into 23 cm square grids (5 × 3). Activity in 2.9. Data analysis
the field was scored for 6 min. Behaviors recorded included the number
of peripheral grid visits (movements across squares and a measure of Data were analyzed using NCSS software (Kaysville, UT, USA). One-
general locomotor activity), central grid visits (an anxiety-related be- way (treatment), between-subject ANOVAs were used to test for group
havior), and total grid visits. Grid visits were scored when the subjects' differences. Type I error rate was set at 0.05. Effect sizes were calcu-
full body, excluding tail, entered a grid. lated using Eta-squared and Tukey-Kramer Multiple-Comparison Tests
were used for post-hoc analysis, where appropriate.
2.7. Object placement and object recognition trials
3. Results
Spatial memory was assessed using the OP task. All trials were
conducted in an enclosed arena measuring 70 × 70 × 40 cm. The arena 3.1. Experiment 1
was wood with an open top. Three walls were painted grey and one wall
was painted black and white striped which provided an intra-maze In experiment 1, subjects were sacrificed immediately following the
spatial cue. Trials consisted of a sample trial (T1) and a retention trial last injection (PND49) in adolescence.
(T2), separated by an inter-trial delay. In T1, two identical objects were
placed at one end of the open field and amount of time spent exploring 3.1.1. Golgi analysis
the two objects was recorded for 3 min. For T2, one object was moved As shown in Fig. 2, adolescent hormonal manipulations altered
to a novel location within the field. In T2, the time spent exploring the dendritic spine density in a region-specific manner. In the hippocampus
object at the old location and the new location was recorded for 3 min. on pyramidal cells in CA1, spine density was significantly increased by
Thus, the percentage of time spent with the object in the new location E2 treatment compared to CON and BPA groups, on both basal, F
during the total exploration time during T2 is used as an index of OP (2,16) = 7.49, p < 0.006 η2 = 0.52, and apical dendrites, F
performance, (time with new location) / (time with old location + time (2,16) = 6.89, p < 0.008 η2 = 0.50. In the DG, BPA decreased den-
with new location) × 100. This measure is referred to as Ratio. All dritic spine density on granule cells when compared to CON and E2
animals received a 1 min delay acclimation trail (data not shown) and groups, F(2,16) = 9.43, p < 0.003 η2 = 0.57. No significant changes
then were tested for OP memory with 10 min, 30 min, and 1 h inter-trial were observed in the mPFC, p > 0.05.
delays.
Non-spatial, visual memory was assessed using the OR task. The 3.2. Experiment 2
testing field was identical to that used for the OP trials with the ex-
ception that all four walls were painted grey. The OR task is identical to In experiment 2, adolescent exposure to E2, BPA, or CON was fol-
the OP procedure, except that during T2, the retention trial, one of the lowed by behavioral testing PND 49–58. Subjects were sacrificed in
original objects is replaced with a new distinct object. Thus, the per- adolescence, PND 59.
centage of time spent with the novel object during the total exploration
time during T2 is used as an index of OR performance. Subjects received 3.2.1. Behavioral measures
a 1 min delay acclimation trail (data not shown) and were then tested 3.2.1.1. Elevated plus maze. As shown in Table 1, there were no
for object recognition memory with 10 min, 30 min, and 1 h inter-trial differences between the groups for the number of visits to open, F
delays. (2,22) = 0.27, p = 0.77 η2 = 0.03, or closed arms, F
During OP and OR trials, object exploration was defined as subjects (2,22) = 2.91,p = 0.08 η2 = 0.23. There were also no significant
sniffing, whisking, or visually examining the object from no > 2 cm differences in the amount of time spent on open, F(2,22) = 1.11,
away. The objects used for trials included identical pairs of bottles, p = 0.35 η2 = 0.09, or closed, F(2,22) = 0.71, p = 0.50 η2 = 0.07,
cans, ceramic and glass figures. The new location/object was counter- arms.
balanced across treatment.
3.2.1.2. Open field. Also shown in Table 1, is the OF data. There were
2.8. Golgi impregnation no significant effects of adolescent hormone exposure on the number of
peripheral, F (2,23) = 2.22, p = 0.13 η2 = 0.17, central, F
Dendritic spine density was measured in brain regions known to be (2,23) = 1.31, p = 0.29 η2 = 0.11, or total, F(2,23) = 2.16, p = 0.14
involved in spatial memory (the hippocampus) and visual working η2 = 0.17, grid crossings.

28
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

CA1 (data not shown, p > 0.05); however, as shown in 3B, a significant
A.
treatment effect was observed during the retention (T2) trials of the
1.5
CON 10 min, F(2,21) = 4.23, p < 0.03 η2 = 0.31, and 1 h, F(2,20) = 4.47,
* p = 0.03 η2 = 0.33, inter-trial delays. Post-hoc testing revealed that E2
# spines/10 µm dendrite

* E2
decreased T2 exploration times compared to CON (10 min delay) and
BPA
1.0 both CON and BPA (1 h delay). Fig. 3B shows percentage of time spent
with the new object (ratio) across the inter-delay trials. All groups had
intact non-spatial working memory (i.e., ratios above chance
0.5
performance) and there were no significant group differences,
p > 0.05, regardless of adolescent hormone exposure. Thus, E2
treatment altered T2 exploration but did not alter performance of the
OR task.
0.0
Basal Apical
3.2.2. Golgi analysis
Fig. 4A is a photomicrograph illustrating Golgi impregnated den-
drites on granule cells in BPA and CON treated subjects and Fig. 5
DG
B. shows the region-specific effects of adolescent hormone treatment ex-
4 posure on dendritic spine density following adolescent behavior testing.
The same pattern in dendritic spine density observed in experiment 1
# spines/10 µm dendrite

was observed in experiment 2 (Fig. 5). E2 treatment significantly in-

3
* creased spine density in CA1 compared to CON and BPA groups on both
basal, F(2,23) = 5.36, p < 0.02 η2 = 0.34, and apical, F
2 (2,23) = 16.57, p < 0.0001 η2 = 0.61, dendrites. In the DG, BPA de-
creased dendritic spine density compared to CON, but not E2 groups, F
(2,16) = 4.4, p < 0.03 η2 = 0.39. BPA exposure decreased spine den-
1 sity in the mPFC on both basal, F(2,21) = 33.87, p < 0.0001
η2 = 0.78, and apical, F(2,21) = 80.56, p < 0.0001 η2 = 0.90, com-
0
pared to both CON and E2 groups.
CON E2 BPA
3.3. Experiment 3

mPFC
C. In experiment 3, adolescent exposure to E2, BPA, or CON was fol-
20
lowed by behavioral testing in young adulthood PND 77-86 and sub-
jects were sacrificed PND 87.
# spines/10 µm dendrite

15 3.3.1. Elevated plus maze

As shown in Table 1, adolescent hormone treatment exposure did
not impact the number of visits to open F(2,22) = 3.12, p = 0.07
10
η2 = 0.23, or closed arms, F(2,22) = 0.79,p = 0.47 η2 = 0.07, or the
amount of time spent on open arms, F(2,22) = 1.7, p = 0.21 η2 = 0.15
5 and closed arms, F(2,22) = 3.32, p = 0.11 η2 = 0.20.

3.3.2. Open field

0 Also shown in Table 1, is the OF data. There were no significant
Basal Apical
effects of adolescent hormone exposure on the number of peripheral, F
Fig. 2. Effects of adolescent hormone exposure on dendritic spine density in (2,23) = 1.87, p = 0.18 η2 = 0.15, central, F(2,23) = 0.38, p = 0.68
adolescence, in the absence of behavioral testing. Entries are the average # η2 = 0.04, or total, F(2,23) = 1.76, p = 0.20 η2 = 0.14, grid crossings
spines/10 μm ± SEM. All significant effects are p < 0.05 and group differ- when measured in adulthood.
ences are denoted by *. (Panel A) Adolescent E2 exposure led to increased spine
density on both basal and apical dendrites in CA1. (Panel B) BPA decreased 3.3.3. Object placement
dendritic spines on granule cells. There were no significant effects of adolescent
There were no differences in exploration times during the sample
hormone exposure in mPFC (Panel C).
(T1) or retention (T2) trials between the groups for any of the OP inter-
trial delays (data not shown, p > 0.05). Fig. 6A shows percentage of
3.2.1.3. Object placement. There were no differences in exploration time spent with the object in the new location (ratio) across inter-delay
times during the sample (T1) or retention (T2) trials between the trials. BPA exposure significantly impaired spatial memory compared to
groups for any of the OP inter-trial delays (data not shown, p > 0.05). the CON and E2 groups at the 10 min, F(2,23) = 8.35, p = 0.002
Fig. 3A shows percentage of time spent with the object in the new η2 = 0.44, but not the 30 min (p = 0.068) or 1 h (p = 0.065) delays.
location (ratio) across inter-delay trials. While there were no effects of
adolescent hormone exposure on spatial memory at the 10 min inter- 3.3.4. Object recognition
trial delay, BPA exposure significantly impaired spatial memory There were no differences in exploration times during the sample
compared to the CON and E2 groups at the 30 min, F(2,22) = 6.45, (T1) or retention (T2) trials between the groups for any of the OR inter-
p = 0.007 η2 = 0.39, and 1 h, F(2,22) = 9.13, p = 0.002 η2 = 0.48 trial delays (data not shown, p > 0.05). Fig. 6B shows percentage of
delays. time spent with the new object (ratio) across inter-delay trials. There
were no differences between the treatment groups at the 10 min inter-
3.2.1.4. Object recognition. There were no differences in exploration trial delay; however, adolescent BPA exposure lowered non-spatial
times during the sample (T1) across any of the OR inter-trial delays working memory performance at both the 30 min, F(2,23) = 5.05,

29
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

Table 1
Adolescent hormone exposure does not alter EPM and OF behaviors in adolescence (exp 2) and young adulthood (exp 3).
Elevated plus maze Open field

Number of visits Time spent (secs) Number of grid crossings

Open arms Closed arms Open arms Closed arms Peripheral Central Total

Experiment 2:
CON 3.89 ± 0.9 11.0 ± 0.7 36.8 ± 8.8 200.0 ± 8.8 104.5 ± 11.3 5.9 ± 1.2 110.4 ± 12.5
E2 3.63 ± 0.8 7.25 ± 1.3 60.8 ± 14.3 186.9 ± 10.0 63.4 ± 15.0 2.4 ± 0.8 65.8 ± 15.7
BPA 4.57 ± 1.1 8.86 ± 1.3 45.6 ± 11.4 201.9 ± 10.5 78.9 ± 15.2 5.3 ± 2.4 83.9 ± 17.3

Experiment 3:
CON 2.71 ± 0.4 8.9 ± 1.1 39.1 ± 9.7 197.0 ± 12.4 85.6 ± 11.1 3.1 ± 1.2 88.8 ± 11.7
E2 5.13 ± 0.9 10.3 ± 0.8 69.4 ± 12.9 155.6 ± 9.4 97.0 ± 13.5 4.3 ± 2.2 101.3 ± 15.1
BPA 5.25 ± 0.9 10.6 ± 1.2 58.3 ± 11.2 179.5 ± 16.1 65.9 ± 9.7 2.4 ± 0.8 68.3 ± 10.3

Data is the average ± SEM number of entries made and the time spent in the open versus closed arms during a 5 min EPM trial and the number of grid crossings
made during a 6 min OF trial. Adolescent hormone exposure did not alter any EPM or OF measures when measured in adolescence (Experiment 2) or adulthood
(Experiment 3), all p > 0.05.

A. B.
100 100
CON CON
E2 E2
new location, (M + SEM)
Ratio, % time spent with

80 80

Ratio, % time spent with

BPA BPA

new object, M + SEM

60 * * 60

40 40

20 20

0 0
10 min 30 min 1 hr 10 min 30 min 1 hr

Inter-trial delay Inter-trial delay

Fig. 3. OR and OP performance when measured in adolescence. Data are expressed as the percent of total T2 time spent with the object in the new location (OP, Panel
A) and as the percent of total T2 time spent with the new object (OR, Panel B). All significant effects are p < 0.05 and significant differences between groups are
denoted by *. Adolescent BPA exposure decreased object placement (Panel A), but not object recognition (Panel B) performance.

p = 0.02 η2 = 0.32, and 1 h, F(2,23) = 19.31, p = 0.0001 η2 = 0.65,

inter-trial delays.

3.3.5. Golgi analysis

Fig. 7 shows the region-specific effects of adolescent hormone
treatment exposure on dendritic spine density following behavior
testing in adulthood. CA1 basal dendritic spine density was increased
by E2 and decreased by BPA exposure compared to CON, F
(2,15) = 11.9, p < 0.001 η2 = 0.65. CA1 apical dendritic spine density
was decreased by BPA, F(2,15) = 5.59, p < 0.02 η2 = 0.46, compared
to E2 but not CON. In the DG, the same effect in experiment 2 was
observed, BPA decreased dendritic spine density on granule cells
compared to CON, but not E2 groups, F(2,22) = 4.3, p < 0.03
η2 = 0.30. No group differences in spine density where observed on
mPFC basal dendrites (p > 0.05); however, BPA exposure decreased
apical dendritic spine density, F(2,17) = 9.4, p < 0.002 η2 = 0.56,
compared to E2 but not CON.

4. Discussion

In the present study, OVX rats were used to assess the effects of
adolescent BPA or E2 in three experiments. In the first experiment,
Fig. 4. Photomicrograph illustrating Golgi impregnated dendrites on granule spine density was assessed in the absence of behavior. In the next two
cells from experiment 2. Top: BPA, Bottom: Control. Taken under oil at 100×. experiments, the effects of adolescent BPA or E2 on behavior and
Arrows denote spines. dendritic spine density were analyzed at two developmental time points
(adolescence versus adulthood). The overall results demonstrate that
BPA administration in adolescence in the OVX rat has similar effects to
those we have previously described for gonadally intact rats for
memory and dendritic spine density (Bowman et al., 2015; Bowman
et al., 2014; Diaz Weinstein et al., 2013) as well as some differences that

30
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

CA1 from the intact females in that OVX rats show no group differences on
A.
* * any measure of anxiety or activity (EPM and OF) after BPA. This sug-
15
CON gests that the E2 present in intact animals might increase anxiety or that
# spines/10 µm dendrite

there is a complex interaction between E2 and adolescent brain de-

E2
velopment that is not clear. Studies by Juraska and colleagues (re-
BPA
10 viewed in Juraska and Willing, 2017) have shown that female rats
exhibit a larger decrease in the number of neurons in the PFC, as well as
dendritic complexity than male rats during puberty. Thus, in the pre-
sent study, the fact that exogenous administration of E2 to OVX rats
5
also had no effect on anxiety may result from a complex process of
hormonal influences during puberty that cannot be replicated in the
OVX model by simply replacing E2.
0 Unlike the measures of anxiety, spatial memory was impaired fol-
Basal Apical lowing adolescent BPA exposure at both the 30 min and 1 h inter-trial
delays in the OVX rats, which is similar to the BPA effect observed in
intact animals (Diaz Weinstein et al., 2013). Furthermore, the current
DG
B. results show that E2 treated OVX females have spatial memory per-
15 * formance that does not differ from controls. The importance of E2 in
enhancing spatial memory is well documented (Luine et al., 2018;
# spines/10 µm dendrite

Luine, 2014) and the intact performance of spatial memory in E2

* treated OVX subjects is indicative of this effect. In adult rats, OVX
10
without E2 replacement impairs spatial memory (Wallace et al., 2006)
and this differs from the results in the present study in which OVX rats
without E2 replacement did not show any spatial memory impairments.
5 It is interesting to speculate that differences between adolescent and
adult females after OVX might be due to differences between organi-
zational and activational effects of E2 (Koebele and Bimonte-Nelson,
2015), such that in intact rats there is a lasting organizational effect of
0
CON E2 BPA
estrogen on the brain. Alternatively, this may be an example of the
complex interaction between E2 and the adolescent brain that requires
further investigation.
mPFC In contrast to spatial memory, OR performance was intact for all
C.
treatment groups across all trials. This is a novel finding, as OR was not
25
assessed in our previous study on intact female rats (Diaz Weinstein
# spines/10 µm dendrite

et al., 2013) and, to date, no other reports on adolescent BPA or E2

20
treatment effects on adolescent measures of OR have been reported. In
* * general OR has been associated with the prefrontal cortex which un-
15
derdoes plasticity for longer periods than other brain areas (Kolb et al.,
2012) and it seems conceivable that the lack of group differences ob-
10
served in adolescent rats may be due to the ongoing prefrontal cortex
development. It is possible that the extensive neural plasticity during
5 adolescence is potentially masking BPA or E2 effects. This possibility is
supported the results in experiment 3 where both the CON and E2 OVX
0 groups performed even better than in adolescence. The average OR
Basal Apical ratio across trials increased by 4% and 11%, for CON and E2 groups,
Fig. 5. Adolescent hormone exposure effects on dendritic spine density when respectively in adulthood (compared to average performance when
measured following adolescent behavioral testing. Entries are the average # measured in adolescence) and there was a significant BPA induced
spines/10 μm ± SEM. All significant effects are p < 0.05 and group differ- decrease in OR (see below).
ences are denoted by *. (Panel A) E2 exposure increased CA1 basal and apical As previously stated, spatial memory performance in adulthood
dendritic spine density. (Panel B) BPA decreased dendritic spines on granule continued to be impaired by adolescent BPA exposure when examined
cells compared to CON. (Panel C) Basal and apical mPFC dendritic spine density at the 10-min inter-trial delay, and while not significant (p = 0.07) the
was decreased by BPA exposure. same trend was observed at longer delays (similar to experiment 2).
Thus, adolescent BPA in OVX subjects impairs spatial memory in ado-
are discussed below (see Table 2). As described in the Methods, beha- lescence and, while modified, these impairments persist into adulthood.
vioral testing was conducted during the light phase in order to remain The effects of BPA on spatial memory in OVX adults are similar to those
consistent with all previously reported BPA data from our lab (Bowman in adult intact animals in that the impairment of spatial memory is no
et al., 2015; Bowman et al., 2014; Diaz Weinstein et al., 2013) and longer significant in adulthood when measured at longer delays (e.g.,
should be interpreted with this in mind. The present results also suggest 30 min) (Bowman et al., 2015). Overall, it appears that adolescent BPA
that the interaction between E2 and BPA differs in the adolescent and exposure impairs spatial memory in adolescence and these impairments
adult brain. persist only at short inter-trial delays in adulthood, for both intact and
In intact females, we have previously demonstrated that exposure to OVX females. One possible explanation is that BPA's deleterious effects
BPA during adolescence increased anxiety, impaired spatial memory in on spatial memory are declining over time and future studies might
OP, and decreased dendritic spine density in the mPFC and CA1 when address this possibility by aging adolescent exposed subjects longer
measured during adolescence (Bowman et al., 2015; Diaz Weinstein prior to behavioral testing to examine when/if BPA's effect attenuate
et al., 2013). Results in the OVX model (data from experiment 2) differ across time.
Because E2 is known to enhance spatial memory, the finding that

31
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

A. B.
100 100
CON CON

80 E2 80 E2
*
BPA * BPA
new location, (M + SEM)

Ratio, % time spent with

Ratio, % time spent with

new object, M + SEM

60 60

40 40

20 20

0 0
10 min 30 min 1 hr 10 min 30 min 1 hr

Inter-trial delay Inter-trial delay

Fig. 6. OP and OR performance when measured in adulthood. Data are expressed as the percent of total T2 time spent with the object in the new location (OP, Panel
A) and as the percent of total T2 time spent with the new object (OR, Panel B). All significant effects are p < 0.05 and significant differences between groups are
denoted by *. Adolescent BPA exposure decreased object placement at short, but not long delays (Panel A), and decreased the percentage of time spent with the new
object during both the 30 min and 1 h OR trials (Panel B).

OVX controls continue to successfully discriminate between the old and confound our BPA results.
new locations (in the long-term absence of E2) it is of particular in- BPA decreased dendritic spine density in CA1 in intact rats during
terest. Future studies in our lab will consider the role of non-ovarian adolescence (Bowman et al., 2014) but not in OVX rats in adolescence.
sources of E2, such as adipose tissue. Due to the sustained spatial In contrast, BPA decreased spine density in both OVX and intact adults
memory performance of the controls, there is no enhancing effect of E2 (Bowman et al., 2015) suggesting that development may be more im-
and this is the same effect observed when measured in adolescence portant than gonadal state with respect to BPA. In support of this, Chen
(experiment 2). et al. (2018) recently found that extensive dendritic remodeling occurs
Adult OR performance was impaired by adolescent BPA treatment, in CA1 pyramidal cells from day 35 to 55 in intact female rats. In ad-
an effect not seen when measured in adolescence (experiment 2). This dition, although the number of spines was unchanged, there were more
BPA impairment in OVX subjects is different from that in intact adults mature spines later in later adolescence than early adolescence. An-
in which BPA treated males, but not females, were unable to dis- other important consideration is that BPA effects may be mediated by
criminate between the old and new objects at a 30 min inter-trial delay other endocrine systems such as the thyroid axis (Moriyama et al.,
(Bowman et al., 2015). In the context of the present data, the difference 2002) Taken together it appears that development of CA1 during ado-
between OVX female adolescent and adult OR performance supports lescence may not be dependent on E2 and that the effects of BPA in the
the idea that maturation of the prefrontal cortex may be important for adolescent brain are not necessarily E2 dependent in every situation.
non-spatial working memory. The difference between OVX and intact Future studies will address the potential interaction between E2 and
females suggests that in adult intact female rats BPA and E2 interact BPA in OVX adolescent rats.
differently than in adolescence. Results in the mPFC are somewhat different to those in CA1. In the
The neural basis for this differential effect of BPA on memory per- current study E2 had no effect on dendritic spine density in mPFC when
formance in intact vs. OVX and in adolescence vs. adulthood is un- measured immediately in experiment 1 and decreased dendritic spine
known, but changes in dendritic spine density may be a contributing density in both adolescence and adulthood when behavior was tested
factor. In all three experiments in this study, E2 increased dendritic (experiments 2 and 3, respectively). The decreased spine density after
spines on CA1 pyramidal cells. This E2 replacement effect in CA1 has BPA in the PFC in OVX rats is similar to what was seen in adolescent
been demonstrated in adult rats following OVX (Frankfurt and Luine, intact rats (Bowman et al., 2014) but differs from our results in adult
2015; Gould et al., 1990) and therefore it is interesting to note the same intact rats where BPA had no effect (Bowman et al., 2015). In apparent
effect when female rats are OVX at such an early age. contrast to the present BPA induced decrease in spine density in the
BPA had no effect on CA1 dendritic spine density in Experiment 1 or adolescent mPFC, no change in synaptic density was seen in the pre-
Experiment 2 (adolescence) but decreased spine density in CA1 in frontal cortex in juvenile monkeys treated with BPA (Elsworth et al.,
adulthood. The BPA induced decrease in adults OVX rats is consistent 2013). This may reflect differences between assessing spine and synapse
with earlier reports in adult intact male and female rats (Bowman et al., density or may be related to evolutionary differences in prefrontal
2015; Bowman et al., 2014; Eilam-Stock et al., 2012) and OVX adult cortex circuitry.
rats (Inagaki et al., 2012). BPA also induced decreases in synaptic Observed differences between adolescent and adult effects of BPA
density in CA1 in both adult rats (MacLusky et al., 2005) and monkeys on dendritic spines may be due to the ongoing pruning that occurs in
(Elsworth et al., 2013) when administered in the prenatal period. In- the prefrontal cortex (Kolb et al., 2012). Indeed, Shapiro et al. (2017)
terestingly, BPA administration to juvenile monkeys did not alter sy- have recently reviewed the changes that occur in several cytoskeletal
napse density in CA1 or the prefrontal cortex (Elsworth et al., 2013) proteins associated with synaptic plasticity in the prefrontal cortex in
supporting the idea that the interaction between development and go- female mice. During adolescence there were differential changes in
nadal state are critical factors to BPA effects. However, it is important to TRKb receptors, BDNF, PSD -95 and synaptophysin in different pre-
note that there may be other factors confounding the BPA results ob- frontal cortex sub regions. This is of particular interest because many
tained during adolescent exposure after OVX. The process of ovar- cytoskeletal proteins have been shown to be altered by estrogens (Frick,
iectomy at day 21 may have more profound effects on adolescence than 2015) and therefore one may speculate that differences between BPA's
simply removing estrogens. The stress of the surgical procedure, the effects on intact and OVX rats may be due to gonadal steroid interaction
exposure to anesthetic may alter spine density. In addition, the rats in with pruning in the mPFC during adolescence.
the present experiment were fed rat chow that contains phytoestrogens Unlike CA1 and the mPFC, spine density in the DG following BPA
which may counteract the effects of OVX. Thus, the actual effects of has not been well studied. In all three experiments in the present study,
OVX on the adolescent brain require further investigation and may adolescent BPA decreased spine density compared to both CON and E2

32
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

CA1 Table 2
Comparison of BPA effects on intact and OVX females in adolescence and
15 *
* * CON
adulthood.
# spines/10 µm dendrite

E2 Behavior Adolescent Adult Spine Adolescent Adult No behavior

density
BPA
10
OVX
EPM NC ↑ CA1 NC ↓ NC
OF NC NC PFC ↓ ↓ NC
5 OP ↓ (30 and ↓ DG ↓ ↓ ↓
60 min) (10 min)
OR NC ↓ (30 and
60 min)

0 Intact
Basal Apical EPM ↑ anxietya NCc CA1 ↓b ↓c NA
OF ↑ anxietya NCc PFC ↓b NCc NA
OP ↓a NCc DG NAb NA NA
OR NA NCc
DG
B. NC = no change, NA = not assessed.
40 * a
Diaz Weinstein et al., 2013.
b
Bowman et al., 2014.
# spines/10 µm dendrite

c
Bowman et al., 2015.
30

decreases in spine density were only seen after behavior. Although this
20 may be due to regional differences in behavior mediation it may also be
further evidence that the maturation of the mPFC during adolescence
makes it particularly sensitive to stimuli.
10

5. Conclusion
0
Together, adolescent BPA appears to have differential effects on
spatial versus non-spatial working memory and these effects are
maintained in an intact and OVX rats. In addition, BPA exposure has
mPFC differential effects on spatial memory in adolescence and adulthood.
C. While it is difficult to reconcile alterations in spine density with every
15
*
behavioral effect, spine density in key regions is decreased by adoles-
cent BPA but the decreases in spine density seem to be tempered during
# spines/10 µm dendrite

the complex process of neural plasticity that occurs though adolescence.

10
As recent studies show that urinary BPA is detectable in > 95% of
sampled Americans (Lehmler et al., 2018) it seems imperative to better
understand the behavioral and neuronal implications of low-dose ex-
posure during adolescence when the brain is particularly devel-
5 opmentally vulnerable.

Declaration of interests
0
Basal Apical The authors have none.

Fig. 7. Adolescent hormone exposure effects on dendritic spine density when

Acknowledgements
measured following behavioral testing in adulthood. Entries are the average #
spines/10 μm ± SEM. All significant effects are p < 0.05 and group differ-
ences are denoted by *. (Panel A) E2 increased and BPA decreased CA1 basal This work was supported by Sacred Heart University Undergraduate
dendritic spine density compared to CON and BPA decreased CA apical com- Research Initiative grants to J. Hagedorn and E. Madden.
pared to E2. (Panel B) BPA decreased dendritic spines on granule cells com-
pared to CON. (Panel C) Apical dendritic spine density was decreased by BPA References
compared to E2 and there were no effects on basal dendrites.
Ahmed, R.G., Walaa, G.H., Asmaa, F.S., 2018. Suppressive effects of neonatal bisphenol A
on the neuroendocrine system. Toxicol. Ind. Health 36 (6), 397–407.
groups. This decrease in spine density is consistent with the finding that Bowman, R.E., Luine, V., Khandaker, H., Villafane, J.J., Frankfurt, M., 2014. Adolescent
perinatal BPA decreases synaptic density in a dose dependent manner in bisphenol-A exposure decreases dendritic spine density: role of sex and age. Synapse
both intact adult male and female rats (Liu et al., 2016). However, to 68, 498–507.
Bowman, R.E., Luine, V., Diaz Weinstein, S., Khandaker, H., DeWolf, S., Frankfurt, M.,
our knowledge, this is the first report of BPA induced decreases in spine 2015. Bisphenol-A exposure during adolescence leads to enduring alterations in
density in the DG in the adolescent brain. Moreover, the BPA effect on cognition and dendritic spine density in adult male and female rats. Horm. Behav. 69,
the DG was consistent with and without behavior at in all three ex- 89–97.
Chen, Y.W., Akad, A., Aderogba, R., Chowdhury, T.G., Aoki, C., 2018. Dendrites of the
periments. The DG is a region that is known to exhibit active neuro- dorsal and ventral hippocampal CA1 pyramidal neurons of singly housed female rats
genesis which may explain why it is more vulnerable to BPA. exhibit lamina-specific growths and retractions during adolescence that are re-
Lastly the potential effects of behavior on spine density, in- sponsive to pair housing. Synapse 72, e22034.
Diaz Weinstein, S., Villafane, J.J., Juliano, N., Bowman, R.E., 2013. Adolescent exposure
dependent of BPA must be addressed. In both the DG and CA1 spine
to bisphenol-A increases anxiety and sucrose preference but impairs spatial memory
density was the same independent of behavior. Not so in the PFC, where in rats independent of sex. Brain Res. 1529, 56–65.

33
R.E. Bowman et al. Hormones and Behavior 107 (2019) 26–34

Eilam-Stock, T., Serrano, P., Frankfurt, M., Luine, V., 2012. Bisphenol-A impairs memory Liu, Z.H., Ding, J.J., Yang, Q.Q., Song, H.Z., Chen, X.T., Xu, Y., Xiao, G.R., Wang, H.L.,
and reduces dendritic spine density in adult male rats. Behav. Neurosci. 126, 2016. Early developmental bisphenol-A exposure sex-independently impairs spatial
175–185. memory by remodeling hippocampal dendritic architecture and synaptic transmis-
Elsworth, J.D., Jentsch, J.D., Vandevoort, C.A., Roth, R.H., Redmond Jr., D.E., Leranth, sion in rats. Sci. Rep. 6, 32492.
C., 2013. Prenatal exposure to bisphenol A impacts midbrain dopamine neurons and Luine, V.N., 2014. Estradiol and cognitive function: past, present and future. Horm.
hippocampal spine synapses in non-human primates. Neurotoxicology 35, 113–120. Behav. 66, 602–618.
Elsworth, J.D., Jentsch, J.D., Groman, S.M., Roth, R.H., Redmond, E.D., Leranth, C., 2015. Luine, V., 2016. Estradiol: mediator of memories, spine density and cognitive resilience to
Low circulating levels of bisphenol-A induce cognitive deficits and loss of asymmetric stress in female rodents. J. Steroid Biochem. Mol. Biol. 160, 189–195.
spine synapses in dorsolateral prefrontal cortex and hippocampus of adult male Luine, V., Frankfurt, M., 2013. Interactions between estradiol, BDNF and dendritic spines
monkeys. J. Comp. Neurol. 523, 1248–1257. in promoting memory. Neuroscience 239, 34–45.
Frankfurt, M., Luine, V., 2015. The evolving role of dendritic spines and memory: in- Luine, V., Serrano, P., Frankfurt, M., 2018. Rapid effects on memory consolidation and
teraction(s) with estradiol. Horm. Behav. 74, 28–36. spine morphology by estradiol in female and male rodents. Horm. Behav. https://doi.
Frankfurt, M., Salas-Ramirez, K., Friedman, E., Luine, V., 2011. Cocaine alters dendritic org/10.1016/j.yhbeh.2018.04.007. [Epub ahead of print].
spine density in cortical and subcortical brain regions of the postpartum and virgin MacLusky, N.J., Hajszan, T., Leranth, C., 2005. The environmental estrogen bisphenol a
female rat. Synapse 65, 955–961. inhibits estradiol-induced hippocampal synaptogenesis. Environ. Health Perspect.
Frick, K.M., 2015. Molecular mechanisms underlying the memory-enhancing effects of 113, 675–679.
estradiol. Horm. Behav. 74, 4–18. Mhaouty-Kodja, S., Belzunces, L.P., Canivenc, M.C., Schroeder, H., Chevrier, C., Pasquier,
Frick, K.M., Kim, J., Tuscher, J.J., Fortress, A.M., 2015. Sex steroid hormones matter for E., 2018. Impairment of learning and memory performances induced by BPA: evi-
learning and memory: estrogenic regulation of hippocampal function in male and dences from the literature of a MoA mediated through an ED. Mol. Cell. Endocrinol.
female rodents. Learn. Mem. 22, 472–493. 475, 54–73.
Galea, L.A.M., Frick, K.M., Hampson, E., Sohrabji, F., Choleris, E., 2017. Why estrogens Moriyama, K., Tagami, T., Akamizu, T., Usui, T., Saijo, M., Kanamoto, N., Hataya, Y.,
matter for behavior and brain health. Neurosci. Biobehav. Rev. 76, 363–379. Shimatsu, A., Kuzuya, H., Nakao, K., 2002. Thyroid hormone action is disrupted by
Goncalves, C.R., Cunha, R.W., Barros, D.M., Martinez, P.E., 2010. Effects of prenatal and bisphenol A as an antagonist. J. Clin. Endocrinol. Metab. 87, 5185–5190.
postnatal exposure to a low dose of bisphenol A on behavior and memory in rats. Negri-Cesi, P., 2015. Bisphenol A interaction with brain development and functions. Dose
Environ. Toxicol. Pharmacol. 30, 195–201. Response 13 (1559325815590394).
Gould, E., Woolley, C.S., Frankfurt, M., McEwen, B.S., 1990. Gonadal steroids regulate Nesan, D., Sewell, L.C., Kurrasch, D.M., 2018. Opening the black box of endocrine dis-
dendritic spine density in hippocampal pyramidal cells in adulthood. J. Neurosci. 10, ruption of brain development: lessons from the characterization of bisphenol A.
1286–1291. Horm. Behav. 101, 50–58.
Hajszan, T., Leranth, C., 2010. Bisphenol A interferes with synaptic remodeling. Front. Ogiue-Ikeda, M., Tanabe, N., Mukai, H., Hojo, Y., Murakami, G., Tsurugizawa, T., Takata,
Neuroendocrinol. 31, 519–530. N., Kimoto, T., Kawato, S., 2008. Rapid modulation of synaptic plasticity by estrogens
Hu, F., Li, T., Gong, H., Chen, Z., Jin, Y., Xu, G., Wang, M., 2017. Bisphenol A impairs as well as endocrine disrupters in hippocampal neurons. Brain Res. Rev. 57, 363–375.
synaptic plasticity by both pre- and postsynaptic mechanisms. Adv. Sci. (Weinh) 4, Scharfman, H.E., Hintz, T.M., Gomez, J., Stormes, K.A., Barouk, S., Malthankar-Phatak,
1600493. G.H., McCloskey, D.P., Luine, V.N., Maclusky, N.J., 2007. Changes in hippocampal
Inagaki, T., Frankfurt, M., Luine, V., 2012. Estrogen-induced memory enhancements are function of ovariectomized rats after sequential low doses of estradiol to simulate the
blocked by acute bisphenol A in adult female rats: role of dendritic spines. preovulatory estrogen surge. Eur. J. Neurosci. 26, 2595–2612.
Endocrinology 153, 3357–3367. Schneider, M., 2013. Adolescence as a vulnerable period to alter rodent behavior. Cell
Juraska, J.M., Willing, J., 2017. Pubertal onset as a critical transition for neural devel- Tissue Res. 354, 99–106.
opment and cognition. Brain Res. 1654, 87–94. Shapiro, L.P., Parsons, R.G., Koleske, A.J., Gourley, S.L., 2017. Differential expression of
Juraska, J.M., Sisk, C.L., DonCarlos, L.L., 2013. Sexual differentiation of the adolescent cytoskeletal regulatory factors in the adolescent prefrontal cortex: implications for
rodent brain: hormonal influences and developmental mechanisms. Horm. Behav. 64, cortical development. J. Neurosci. Res. 95, 1123–1143.
203–210. Taylor, J.A., Welshons, W.V., Vom Saal, F.S., 2008. No effect of route of exposure (oral;
Koebele, S.V., Bimonte-Nelson, H.A., 2015. Trajectories and phenotypes with estrogen subcutaneous injection) on plasma bisphenol A throughout 24h after administration
exposures across the lifespan: what does Goldilocks have to do with it? Horm. Behav. in neonatal female mice. Reprod. Toxicol. 25, 169–176.
74, 86–104. Tovote, P., Fadok, J.P., Luthi, A., 2015. Neuronal circuits for fear and anxiety. Nat. Rev.
Kolb, B., Mychasiuk, R., Muhammad, A., Li, Y., Frost, D.O., Gibb, R., 2012. Experience Neurosci. 16, 317–331.
and the developing prefrontal cortex. Proc. Natl. Acad. Sci. U. S. A. 109 (Suppl. 2), Tuscher, J.J., Fortress, A.M., Kim, J., Frick, K.M., 2015. Regulation of object recognition
17186–17193. and object placement by ovarian sex steroid hormones. Behav. Brain Res. 285,
Lehmler, H.J., Liu, B., Gadogbe, M., Bao, W., 2018. Exposure to bisphenol A, bisphenol F, 140–157.
and bisphenol S in U.S. adults and children: the National Health and Nutrition United States Environmental Protectional Agency, 1993. Bisphenol A, CASRN 80-05-7.
Examination Survey 2013–2014. ACS Omega 3, 6523–6532. Integrated Risk Information System. U.S. Environmental Protection Agency,
Leranth, C., Hajszan, T., Szigeti-Buck, K., Bober, J., MacLusky, N.J., 2008a. Bisphenol A Washington, DC Available at: http://www.epa.gov/iris/subst/0356.htm.
prevents the synaptogenic response to estradiol in hippocampus and prefrontal cortex Wallace, M., Luine, V., Arellanos, A., Frankfurt, M., 2006. Ovariectomized rats show
of ovariectomized nonhuman primates. Proc. Natl. Acad. Sci. U. S. A. 105, decreased recognition memory and spine density in the hippocampus and prefrontal
14187–14191. cortex. Brain Res. 1126, 176–182.
Leranth, C., Szigeti-Buck, K., Maclusky, N.J., Hajszan, T., 2008b. Bisphenol A prevents the Wang, C., Niu, R., Zhu, Y., Han, H., Luo, G., Zhou, B., Wang, J., 2014. Changes in memory
synaptogenic response to testosterone in the brain of adult male rats. Endocrinology and synaptic plasticity induced in male rats after maternal exposure to bisphenol A.
149, 988–994. Toxicology 322, 51–60.

34