Effects of Estradiol Deprivation on Striatal-Dependent Learning and Memory in Rat Models of Menopause
A Thesis Submitted in Partial Fulfillment of theRequirements of the Renée Crown University Honors Program
and the Distinction in Biology Program atSyracuse University
Rebekah Schwartz
Candidate for Bachelor of Scienceand Renée Crown University Honors
May 2020
Honors Thesis in Biology
Thesis Advisor: _______________________ Donna Korol, PhD
Thesis Reader: _______________________ Susan Parks, PhD
Honors Director: _______________________ Dr. Danielle Smith, Director
© (Rebekah Schwartz May 2020)
ii
Abstract
As women enter the climacteric that leads to menopause, ovarian hormone levels begin to drop, leading to widespread physiological changes in the adult organism. It is thought that declines in estrogens mediate many of these changes. Although estrogens are typically considered reproductive hormones they indeed have far-reaching effects on the body beyond just the reproductive system. Specifically, estrogens have numerous impacts on the brain, influencing essentially all regions studied to date, even those that lack classical estrogen receptors. One of those effects is neural protection against age- and disease-related changes, possibly through facilitation of neural plasticity. Estrogens can protect against AD and PD, perhaps via direct actions on hippocampus, striatum, and related structures. However, estrogens can also produce both enhancements and impairments on learning and memory, depending on the type of problem to be solved and treatment regimen. Specifically, acute treatments of estrogens for two days enhance learning and memory that rely on the hippocampus and impair learning and memory that rely on the striatum (Korol and Kolo, 2002; Korol, 2004). Thus, it appears that estrogens have two types of hormone action on the brain, the activating effects on cognition that may be mixed and depend on physiological state of the individual and trophic or long-term effects on the brain that may be largely protective. The time at which hormone therapy is begun may play a role in how estrogens produce these effects on the brain. The critical window hypothesis states hormone replacement therapy is only effective within a certain window of opportunity after menopause begins (McCarrey and Resnick 2015, Maki et al 2013, Sherwin and Henry 2008). Given that these hormones have such far reaching neurobiological effects, the current study was undertaken to assess how learning and memory is altered when the rat is in a state of hormone deprivation modeled after menopausal women. This study evaluated the changes in cognition resulting from increased hormone deprivation during menopausal transition and whether the brain remains sensitive to the action of estradiol with increased hormone deprivation. An impairment in striatal learning from estradiol after three weeks of hormone deprivation, was not found despite a large body of evidence from previous work. Increasing the duration of hormone deprivation also did not lead to impaired performance as was expected due to the proposed loss of neuroprotective effects of estrogens. Several possible explanations for the lack of estradiol effects are explored.
iii
Executive Summary
Menopause is becoming increasingly salient with the increase in human lifespan. As
lifespan increases, women are spending an increasing proportion of their lives in the
postmenopausal period. As women approach menopause, ovarian hormone levels decline. This
decline, specifically in estrogens, may control the multitude of physiological changes associated
with this period. Estrogens, which we typically consider as one hormone, actually comprise
three different hormones: estrone (E1), estradiol (E2), and estriol (E3) (Cui et al 2013). E2 is
the predominant form of estrogen in the body of reproductively cycling women, and is thus the
focus of this study (Cui et al 2013).
Beyond its role in the reproductive system, E2 mediates a wide array of functions in the
body, most specifically in regards to brain health. E2 plays a protective role in the female brain
against AD and PD, by not only reducing the risk of these neurodegenerative diseases, but also
by ameliorating some of the cognitive deficits associated with them. With the increase in time
spent as a postmenopausal woman, the loss of these protective effects has robust implications
on brain health and progression of neurodegenerative disease.
During menopause, the amount of cycling hormones decreases as a result of the loss of
follicle supply. During the transition into menopause, serum levels of E2 drop by about 85-90%
(Cui et al 2013). This disruption in hormone levels poses implications throughout the body, far
beyond the reproductive system alone. The lack of E2 has implications for learning and
memory, which makes both hormone deprivation and hormone replacement of interest to the
lab.
Previous work in our lab shows that acute E2 given just before maze training impairs
striatum-sensitive learning after three weeks of ovarian hormone deprivation by surgically
induced menopause in rats (Korol and Kolo 2002). Additional findings suggested that the
iv
impairing effects of E2 were even more robust one week after hormone reduction, primarily
because performance in the hormone-deprived rats was still very strong at this time point
especially when compared to the rats that were hormone deprived for three weeks (Korol and
Kolo, 2002). Together, these findings suggest that although increases in E2 in the circulation
impairs striatum sensitive learning, the long-term effects of E2 loss may also impair striatum-
sensitive learning. Thus, estrogens might have two types of hormone action on the brain, the
activating effects on cognition that may be mixed and depend on the type of problem to be
solved and the physiological state of the individual and also the trophic or long-term effects on
the brain that may be largely be neuroprotective.
This study focuses on elucidating effects of hormone deprivation and replacement in a
model of surgical menopause. In this model, hormone deprivation lasts either three or six weeks
to assess how increasing length of hormone deprivation influences learning. This study extends
the deprivation period while also examining effects of different doses of E2 replacement on
learning. The goal of this project was to assess how differing the length of hormone deprivation
after surgically induced menopause in young adult rats impairs learning and memory.
This project’s first aim is to replicate past work finding an impairment on striatal learning
using this same methodology. This study is novel in its extension of this period of hormone
deprivation to six weeks. E2 treatment, regardless of time point, should produce the same
learning results, whereas the rats deprived of ovarian hormones are expected to have impaired
learning with increased time since surgery. For the rats in the six-week time point, this means
that both groups will have an impairment on learning for different reasons, and thus their
learning behavior may appear to be similar.
In contrast to expected results, impaired striatal learning from E2 replacement after three
weeks of hormone deprivation was not shown in this study. Further, an impairment on striatal
learning in vehicle treated rats with a longer, six-week, hormone deprivation was not found.
v
Though this study did not replicate the impairment on striatal learning or identify
significant differences in learning between the two time points and dosages, the work lends itself
to assessing current methodology to make improvements for future studies. The issues
identified in this study provide a basis for a more expansive study looking at the critical window
hypothesis of hormone therapy, as well as new directions investigating the relationships
between stress, E2, and learning.
The cycling of these reproductive hormones, and the confounds they introduce to
experiments has continued to hinder scientific discovery in females. Much of our scientific and
medical knowledge has only been tested in males due to this reason. This has ultimately
resulted in therapies designed for males, as effects on women are difficult to tease out from
inconsistent hormone levels across time. This study works to understand some of the effects of
female reproductive hormones on cognition, working against this bias in the literature towards
male subjects.
vi
Acknowledgements
First and foremost, I would like to thank Donna Korol for accepting me into her lab,
editing multiple drafts of this project and providing continuous support, both research, and non-
research related for the past three years. I would also like to thank Paul Gold, whose expertise
and guidance have helped me to grow throughout my time here at Syracuse. Additionally, I
would like to thank Susan Parks for agreeing to be my reader and editing this project so
efficiently. Special thank you to Robert Gardner for assisting me with statistical analyses. This
project would not have been possible without the mentorship of Stephen Ajayi, Alesia
Prakapenka, and Robert Gardner, and the assistance and support of my fellow lab-mates.
Specifically, I would like to acknowledge Sarah Riddle and Brooke Rhinesmith for helping with
some of the surgeries. I would also like to acknowledge the LAR staff, for working so hard to
take care of all of the rats and ensure their well-being throughout this project. Lastly, I would
like to thank the Reneé Crown Honors Program and the Biology Department for supporting me
in a multitude of ways throughout my undergraduate career and granting me the opportunity to
conduct my research.
vii
Table of Contents
Abstract……………………………………….……………….………………………………………... iii
Executive Summary………………………….……………….………………………………………. iv
Acknowledgements………………………………………………………………………………….. vii
Introduction...…………………………………………………………………………………………… 1
Materials and Methods………………………………………………………………………………... 6
Results…………………………………………………..……………………………………………… 12
Discussion……………………………………………...……………………………………………… 25
Works Cited.…………………………………………………………………………………………… 30
viii
Introduction
Estrogens are typically thought to play a role in reproduction, however, they have far-
reaching effects on other areas of the body, most notably the brain. Estradiol (E2), the most
potent estrogen in reproductive women, is widely known for its protective effects on the brain. A
number of studies show a protective effect of E2 on the brain especially in relation to
neurodegenerative disorders such as Alzheimer’s Disease (AD) and Parkinson’s Disease (PD).
For example, increased risk of PD has been linked to early reduction in estrogens (Ragonese et
al 2006, Villa et al 2006, Benedetti et al 2001). Similarly, postmenopausal treatments with
estrogens is associated with decreased risk and delayed onset of AD, suggesting the loss of
hormones that comes from menopause exposes the brain to degeneration (Gandy 2003, van
Dujin 1999, Tang et al 1996). Interestingly, one brain region shown to degenerate and show
metabolic dysregulation with AD (Korol and Wang 2018) and with normal aging (Gardner et al.,
2020) is the hippocampus.
The hippocampus is positively affected by the action of estrogens, with increased age
and the concomitant loss of estrogens make this brain area more vulnerable to neural insults.
Administration of estrogens is thus thought to be neuroprotective (Daniel 2013). Due to the
beneficial effects of estrogens on the hippocampus, performance is typically enhanced on tasks
utilizing this brain area, and no effects or impairments on other brain areas, of note the striatum
(Daniel 2013). The presence of estrogens can alter the cognitive strategy rats use to solve a
task, thus differentially affecting tasks of learning and memory depending on brain area required
for that process (Korol and Kolo, 2002; Korol, 2004; Daniel and Lee 2004, Daniel et al 1999,
Quinlan et al 2008). Inactivation of the hippocampus with lidocaine causes a shift in rat
preference for striatal-based learning tasks (Packard and McGaugh 1996). Similarly, rats given
ix
lidocaine injection into the striatum show a shift in preference for hippocampal based task
(Packard and McGaugh 1996). Infusion of E2 in the hippocampus enhances learning on place
task with no effect on the response task and infusion into the striatum impairs response learning
with no effect on place learning. This double dissociation between infusion site and effect on
performance (Zurkovsky et al 2007) suggests that E2 acts directly on each structure to
modulate learning in the canonical cognitive task.
Estrogens increase dopamine (DA) function particularly within the basal ganglia, a set of
structures including the striatum important for sensory-motor integration, reward value, and
cued-based behaviors. Within the striatum, estrogens and DA treatment increases firing by
neurons connecting the basal ganglia and the striatum (Arnaud et al 1981, Mizumori et al 2004).
Perhaps this is mediated by estrogen-induced DA release (Mizumori et al 2004). E2 treatments
to ovariectomized (OVXd) mice depleted of DA with a neurotoxin to the nigrostriatal DA pathway
lead to increased DA concentration in the striatum (Dluzen and Horstink 2003). Administration
of dopamine receptor antagonists into OVXd rats receiving 17-estradiol benzoate (EB)
treatment shifts strategy preference between hippocampal and striatal based learning tasks,
suggesting the important role of both EB and DA in strategy selection (Quinlin et al 2008).
Benzoate conjugation of E2 to form EB slows down liver metabolism so it lasts longer in
circulation, better matching what is seen during the estrous cycle. EB could mediate choice of
cognitive strategy both by improving the functions of the hippocampus, but also altering that of
the striatum, which is regulated by DA (Quinlin et al 2008). It is thought that a moderate level of
DA signaling produces optimal striatum-sensitive learning; too little or too much would thus
impair striatum-based cognition.
These robust bidirectional effects of estrogens on brain and cognitive function suggest
that endogenous fluctuations in ovarian hormones across the reproductive cycle will lead to
shifts in learning strategy. Hormone-regulated shifts in cognitive ability might also be found with
x
transitions to physiological states that lose ovarian cyclicity, such as with pregnancy or
menopause.
Menopause occurs when the ovaries become depleted of follicles and subsequently
produce decreased amounts of ovarian hormones. Clinically, menopause is defined as 12
months after the last menstrual cycle and has an average age of onset of 52.5 years of age
(McCarrey and Resnick 2015). The transition into menopause is associated with a multitude of
symptoms, the most common being hot flashes, but often accompanied by night sweats and
sleep problems (von Mühlen et al 1995). The second most common complaint bringing women
into the clinic relates to cognitive deficits. Post-menopausal women claim they experience the
tip-of-the tongue phenomenon, whereby one struggles to recall a specific word, but can recall
those of similar structure and meaning, and also complain of general cognitive fogginess
accompanied by short attention spans and forgetfulness (Brown and McNeill 1966, Mitchell and
Woods 2001).
Female rats exhibit cycling reproductive hormones translatable to the menstrual cycle.
The hypothalamic-pituitary-gonadal axis is responsible for this cycling through stages in females
and the lack thereof in males (Plant 2015). This cycle in rats is referred to as the estrous cycle.
The estrous cycle consists of the stages proestrus, estrus, and diestrus. During proestrus, there
is a spike in levels of estrogens, followed by a spike in progesterone levels. During estrus, rats
are in heat and ovulation occurs. During diestrus, progesterone levels are intermediate.
Female rats also undergo a shift in reproductive hormones occurring at middle age,
slightly differently than what occurs in menopause (Morrison et al 2006). The estrous cycle
begins to become irregular at about 9—12 months of age (Morrison et al 2006). Specifically,
female rats begin to have irregular cycling that ultimately ceases altogether (Morrison et al
2006). The loss of cycling does not coincide with follicular degeneration and decline in estradiol
concentration, which occurs concurrently in menopause (Morrison et al 2006). Acyclic rats,
rather, are in a state of continuous estrus (Morrison et al 2006). Due to these differences, a
xi
surgically induced menopause model via OVX is often used for answering research questions
regarding menopause (Morrison et al 2006).
Hormone therapy (HT) was initially developed to abate hot flashes and is still a common
choice for women experiencing menopausal symptoms (Shanafelt et al 2002). The Women’s
Health Initiative (WHI) Studies, however, have produced a sense of uncertainty about the
effectiveness of HT. The WHI began in 1993 and aimed to provide an overview of the risks and
benefits of hormone therapy, yet had major flaws in methodology. The WHI showed
introduction of estrogen therapy did not improve cognitive function in women who were
postmenopausal for 15 years (Rapp et al 2003) In fact, HT was shown to increase risk of mild
cognitive impairment and dementia in postmenopausal women (Shumaker et al 2003,
Shumaker et al 2004). Despite these flaws, the WHI studies informed the scientific, medical,
and lay communities that dosage, timing, and target tissue (brain, colon, breast, cardio-vascular,
etc.) are important variables for HT.
The results of the WHI led to further investigation of HT that showed effects of HT
depend on timing, formulation, mode of delivery, and regimen (Miller and Harman 2017, Zárate
et al 2017). The WHI ultimately led to many future experiments addressing the flaws within that
study, leading to the development of the “critical window hypothesis”, positing there is a time
period after hormone decline with menopause wherein hormone therapy will remain effective
(McCarrey and Resnick 2015, Maki et al 2013, Sherwin and Henry 2008). HT appears to be
most effective when administered at the time of menopause, becoming less effective as time
passes beyond the onset of menopause. The duration of the critical window depends upon the
age of the individual and the target system or property.
The mechanism underlying the critical window hypothesis is unknown, but could be due
do the need for estrogen-sensitive neurons to adapt to estrogen-depleted environments (Yin et
al 2015). Specifically, these neurons lose their signaling integrity with increased age and
deprivation, thus depriving them of the neuroprotective action of estrogens. One way that
xii
estrogens signal in neurons is through estrogen receptors. Interestingly, after long periods of
hormone deprivation, the beneficial effects of estrogens can be reintroduced by increasing the
expression of the gene encoding for ER, one of the two main types of nuclear receptors for
estrogens (Bean et al. 2015), but not ER, the other nuclear estrogen receptor.
We have shown that E2 has differential effects on different areas of the brain.
Specifically, on a hippocampal-dependent learning task, E2 replacement enhances learning,
whereas on a striatal-dependent task, E2 replacement impairs learning (Korol and Kolo 2002).
Female preference of learning task also changes across the estrous cycle. During proestrus,
when ovarian hormones are high, rats prefer to use a spatial, or hippocampal-dependent,
learning task (Korol et al 2004). Rats in estrus, who have lower ovarian hormone levels prefer
to learn using their striatum (Korol et al 2004). These data alone tell us the critical window
remains open at least for 21 days after ovariectomy (OVX) in female rats.
Our preliminary results investigating the critical window found vehicle control or
“placebo” oil-treated rats showed impaired learning behavior three weeks after OVX as
compared to oil-treated rats trained 6 days after OVX (Korol, unpublished). Our data show
decreased performance in the oil-treated rats, with performance of EB-treated rats remaining
the same across these two time points. Thus, we expect the oil-treated rats to continue to
become worse at performance if deprived from ovarian hormones for an additional three weeks.
Without this knowledge, we may expect EB to still impair learning at 6 weeks post OVX. Our
preliminary data suggest, however, that something is also lost in the impairing effect of EB
because oil rats should continue to worsen while performance with EB treatment remains the
same. This means that with increased deprivation, EB would not have the same impairing effect
with increased hormone derivation because oil rats are also getting worse. In this case, the
impairment caused by the deprivation of E2 may match, or even surpass, the impairment
caused by the presence of E2.
xiii
Perhaps this loss of impairing effect of EB is due to a loss in sensitivity to EB in the
brain. If impairment occurs via disrupted cell-cell signaling, then a lack of impairing effect after
six weeks of hormone deprivation would suggest the impaired performance on the response
task is related to a loss of signaling within the brain, perhaps due to a loss of DA or DA neurons,
or a lack of ERs present for EB to increase DA production.
This experiment aims to determine if sensitivity to EBs changes after six weeks of hormone
deprivation in a rat model of surgically-induced menopause. Secondly, this project aims to
determine how quickly the learning behavior in vehicle-treated rats declines following OVX.
Materials and Methods
Subjects
Young-adult, virgin, female Sprague-Dawley rats (3-mo-old) were obtained from Envigo
(N=33). The rats were individually housed in plastic cages and were on a strict 12:12 light/dark
cycle. Each had free access to food and water until the initiation of food-restriction procedures
to motivate rats to perform the appetitive cognitive task. All rats were OVXd to remove
circulating hormones and trained on a striatum-sensitive response maze learning task either 3
weeks (21-22 days) or 6 weeks (38-44 days) after OVX. Rats were randomly assigned to one
of three hormone treatment groups: vehicle control, low dose of EB, high dose of EB. [Note:
sample sizes for the six-week time point and the 4.5g/kg EB treatment are lower than expected
due to unforeseen health issues that interfered with my ability to train the rats.]
All procedures were conducted according to the National Institutes of Health Guide for
the Care and Use of Laboratory Animals and were approved by the Syracuse University
Institutional Animal Care and Use Committee.
xiv
Ovariectomy
Rats were bilaterally OVXd under isoflurane anesthesia. Rats were put in an induction
chamber filled with 5% isoflurane anesthesia for 10 minutes. The rats’ sides were quickly
shaved and the rat was weighed before transfer to an anesthesia nose cone (1.3-5%) on a
surgery table. Isoflurane was delivered at a rate maintaining one breath per second. Eye
ointment was added to lubricate the eyes. The rats were given an intramuscular injection of
penicillin to prevent infection and a subcutaneous injection of flunixon to prevent pain. The
shaved area was cleaned with 70% ethanol and betadine before making the incision ~150mm
from the lowermost rib. The fascia was cut and the muscle blunt-dissected to locate the fat
pads. The fat pad and ovary were pulled through the hole in the muscle and clamped with soft
tissue forceps on the uterine horn. The ovary was tied off and removed. The remainder of the
fat pad was gently put back into the abdominal cavity. The muscle was sutured together and the
skin was secured with tissue adhesive and closed with 2-3 wound clips. The incision was
cleaned throughout the procedure with 70% ethanol and bacitracin was applied after closing.
The same procedure was followed for the other side. Directly following closure of the wound,
bacitracin was applied to further prevent infection and 10mL of saline (s.c.) was given to replace
lost fluids. Children’s Motrin (2.35mL/500mL water) was given in the drinking water for 24
hours following the surgery with free access to the subjects. Post-operative health and wellness
checks were conducted for the week following surgery to ensure proper recovery.
Estradiol Administration
Rats from each time point were randomly assigned to be in either the EB group or the
vehicle control group. A single injection was given to rats both 48hrs and 24hrs prior to
xv
behavioral training. Injections consisted of either a sesame oil vehicle, a low (4.5µg/kg) dose of
EB in sesame oil, or a high (45 µg/kg) dose of EB in sesame oil. The low dose of EB treatment
was only used for the 3-week group. Final rat numbers for each group were as follows: 3-week
high dose OIL, N = 6; 3-week high dose EB, N = 8; 6-week high dose OIL, N = 4; 6-week high
dose EB, N = 4; 3-week low dose OIL, N = 5; 3-week low dose EB, N = 4. Rats received two
days of a single injection given at the same time as behavioral training would take place. This
regimen controlled for possible interactions in hormone levels with shifting circadian rhythms,
thus ensuring circulating titers are as similar across rats as possible.
Food Restriction
During food restriction, the rats were limited to a certain amount of food each day,
between 6 to 13 grams, supplemented by 3 pieces of Frosted Cheerios®, the food reward used
during behavioral training. Food restriction lasted between 7-10 days until the animal reached
80-85% of their starting weight.
Training
For response learning, rats were trained to make a body turn to the right or left to find
food on a simple T-shaped maze (Figure 1). The training maze was a symmetrical 4-arm maze,
constructed from black Plexiglas®, that had one arm blocked off each trial to make it a T shape,
and was located in a room. The wall cues were curtained to reduce the use of spatial cues
instead of a body turn to guide navigation. The arms of the maze were each 45cm long, 14cm
wide and 7.5cm tall that extended from a 14x14cm central square. At the end of each maze
arm was a cup that contained inaccessible crumbs of the food reward to prevent the rat from
using olfactory cues to guide its navigation.
xvi
The arm that was rewarded had the food reward, half of one frosted Cheerio®, placed
on top of this cup. The direction of the turn was held constant for each rat over the course of
the training session. The correct turn (right or left) for a specific training session was
counterbalanced among the rats. Rats were place in a clean holding cage prior to training to
further prevent them from using olfactory cues. The subjects were put in the training room to
get accustomed to the training environment prior to training. During training, the rat was placed
on the start arm (stem of the “T”) and allowed to enter a single goal arm in attempt to find the
food reward. A choice was defined as entrance of all four paws out of the central square into
one of the choice arms. The rat was removed from the maze approximately 3 seconds following
consumption of the frosted Cheerio®. If the rat did not eat the Cheerio®, the rat was left in the
arm for approximately 10 seconds, or until it just turns around to leave the arm to allow them
time to think.
Time to choose and correct vs incorrect choices were recorded on the data sheet. Any
unusual behaviors were also recorded. There was a 30 second interval between trials, when
the maze was also rotated to prevent using cues from within the maze. If the rat took longer
than two minutes to make a decision, the rat was removed from the maze and placed back into
the cage to start a new trial. This was recorded as no decision on the data collection sheet. All
rats were trained for a total of 75 trials.
Figure 1 Graphic depicting the maze used for response behavioral training.
xvii
Determination of Estrogen Status
Vaginal smears of the rats were taken each day of food-deprivation to track the stage of
the cycle of the rats, to ensure that the OVX was complete and to verify that rats in the EB
groups showed change in vaginal cytology indicating circulating estrogens. Smears were taken
daily at the same time for 7-10 days prior to behavioral testing. Vaginal smears were taken by
soaking a small sterile cotton swab in sterile saline and gently swabbing the vaginal wall. The
slides were fixed with EtOH, and then stained using toluidine blue, coverslipped, and observed
under a light microscope to determine the estrous stage of each rat based on the type of cells
present on each slide. After OVX, rats should stop cycling; thus, smears from oil controls
should resemble those from cycling rats in diestrus (Auletta 1994). The smears taken from the
EB-treated rats after injection should indicate the return of cycling and appear as a proestrous-
like smear.
Sacrifice and Tissue Collection
Immediately following behavioral training, the rat was injected with a lethal overdose of
1mL sodium pentobarbital (50mg/mL, i.p.). Once the rat was no longer responsive to both a toe
pinch and blink test, it was decapitated using a laboratory rodent guillotine. Following
decapitation, trunk blood was harvested for later analysis of serum hormone levels. Given its
high estrogen sensitivity, one uterine horn was dissected out and 1.0 cm weighed to assess
hormone state. The brain was removed from the skull for collection of the striatum, frontal
cortex, hippocampus, cerebellum, and medulla for biochemical analyses used in a different
experiment.
Quantification of Serum Estradiol Levels
xviii
Trunk blood was centrifuged at 3500rpm for 15 minutes (2013Gs) and the serum
supernatant collected and stored at -20C until processing. An ELISA kit (Cayman Chemical)
was used to quantify the E2 levels in the serum, following the manufacturer’s protocol.
Data analysis
Measures
The main cognitive measures were trials to criterion and percent correct choices per trial
block. Learning criterion was set at 9/10 trials with 6 trials in a row correct. Percent correct arm
choices were blocked into groups of 15 trials of training, reflecting accuracy of learning
throughout behavior. Other physiological measures included weight prior to behavioral training,
uterine horn weight, and serum E2 concentration.
Statistical Analysis
T-tests were used to assess differences in uterine horn weight and trials to criterion.
Because of the low sample size, non-parametric analyses using the Mann Whitney U test were
run for trials to criterion to avoid type II errors (false negative). Mann Whitney U tests were not
performed for the time of day analysis due to low sample sizes. Repeated measures ANOVAs
were used to assess differences in treatment condition and learning behavior. 2-way factorial
ANOVAs (Analysis of Variance) were used to compare three-week and six-week data.
Exclusions
Two of the rats in the six-week group were excluded from the data because they failed to
perform on the maze. This was determined by failure to move on the maze for two minutes for
five consecutive trials. One of these rats was treated with oil and the other with 45g/kg EB.
xix
Results
Theoretical Results
The expected behavioral results for the main aim of this experiment are shown in Figure
2 for comparing behavior of rats treated with 45g/kg EB after either 3 weeks or 6 weeks
between OVX and behavioral training. Figure 2A shows an impairment on learning with EB
treatment. Looking between figures 2A and 2B, the oil rats do progressively worse with
increased time between OVX and response training. If sensitivity to estrogens is maintained
despite the loss of function, then replacement would still impair an already deficient response
learning. If the effects of estrogens are lost along with the neural protection, then replacement
of estradiol would produce the same behavior as replacement with oil. Because it appears that
sensitivity to EBs is reduced from 1 to 3 weeks, we predicted that EB would have no effect after
6 weeks. This is why in Figure 2B, EB does not have an impairing effect on response training
because the behavior of oil rats was reduced compared to oil rats 3 weeks after OVX with no
change in EB-treated rats. Essentially, the impairment brought about from increased hormone
deprivation will match or could even exceed the impairment of EB on striatal learning.
xx
1 2 3 4 540
60
80
1003 Week Oil 3 Week EB
% C
orre
ct
A
1 2 3 4 540
60
80
1006 Week Oil6 Week EB
% C
orre
ct
B
Figure 2 Expected performance on the response learning task three weeks (A) and six weeks (B) after OVX and treatment with either 45g/kg EB or sesame oil vehicle. EB impairs striatal learning when administered 3 weeks after OVX (A). Similar performance is expected from both EB and oil treated rats six weeks after OVX (B).NOTE that the oil-treated, hormone deprived rats show significant decline in performance from 3 to 6 week time points.
Task Acquisition
Three-week time point – high dose EB:
All of the rats in this group started around chance performance (50%) and then reached
learning criterion within the 75-trial limit. Evaluation of percent accuracy shows a significant
increase in learning across trial blocks (F(,4,48) = 25.64; p<0.0001). There was no main effect
of treatment for rats on the three-week time point, (F(1,12) = 0.0001; p>0.99). There was no
interaction between performance across trial block and treatment in rats on the three-week time
point (F(4,48) = 0.343; p>0.84) (Figure 3A). There was no significant difference between trials
to reach criterion between these same groups (t(7) = 0.017, p > 0.05; MW U(6,8) = 23.5; p >
0.05) (Figure 3B).
xxi
1 2 3 4 540
60
80
100
Oil (n=6)Estradiol (n=8)
Trial Block
Perc
ent C
orre
ct (
Mea
n ±
S.E.
M)
A
chance
10
20
40
60
80
100
Oil (n=6)Estradiol (n=8)
Tria
ls to
Crit
erio
n
(Mea
n ±
S.E.
M)
B
Figure 3 Peripheral treatment of a high dose of EB (45g/kg) did not show a significant effect on the learning curve (A) nor the trials to reach learning criterion (B) after 3 weeks of hormone deprivation as compared to vehicle-treated rats.
xxii
Time of Day Assessment in 3-week high EB groups:
One possibility for our lack of robust striatum-sensitive learning impairments by EB was
that most rats in this study were trained at earlier times in the day compared to our previous
reports. All of the rats in each of the groups were trained ~ 9:00-10:00AM, except for some of
the rats in the three-week time point treated with the 45g/kg EB were trained at ~12:30PM.
These rats were placed into groups depending on both treatment, oil or EB, and time of training,
9:00AM or 12:30PM (Figure 4; note, splitting up the data is the reason for the smaller sample
size in making the time of day comparison). There were no main effects of treatment or time of
day regardless of trial block (all p’s > 0.5). There is a significant difference in performance
across trial blocks, without regard to treatment or time of day tested in rats on the three-week
time point, excluding the low dose cohort (F(4,40) = 25.357; p<0.0001). There is no interaction
between performance across trial block and time of day tested, regardless of treatment in rats
on the three-week time (F(4,40) = 1.081; p>0.37). There is no interaction between performance
across trial block and treatment, regardless of time of day tested in rats on the three-week time,
excluding the low dose cohort (F(4,40) = 0.345; p>0.84). There is no interaction between
performance across trial block, treatment and time of day tested in rats on the three-week time
point, excluding the low dose cohort (F(4,40) = 0.333; p>0.85). Visual inspection of the data
suggests there is a subtle benefit of being trained in the morning; rats trained in the morning
started out slightly better than those trained in the afternoon (Figure 4A). However, time of day
the rats were trained did not have a significant effect on the trials to reach the learning criterion
in oil-treated rats (t(3) = -0.64, p >0.57) or EB-treated rats (t(6) = 0.56, p>0.59), as shown in
Figure 4B.
xxiii
1 2 3 4 5
20
40
60
80
100
MORNING OIL (n=3)
MORNING EB (n=5)
AFTERNOON OIL (n=3)
AFTERNOON EB (n=3)
Perc
ent C
orre
ct (M
ean
± S.
E.M
)
A
chance
10
20
40
60
80
100MORNING OIL (n=3)
MORNING EB (n=5)
AFTERNOON OIL (n=3)
AFTERNOON EB (n=3)
Tria
ls to
Crit
erio
n
(Mea
n ±
S.E.
M)
B
Figure 4 Comparison of behavior based on treatment of 45g/kg EB or oil vehicle three weeks after OVX and time of day behavioral training took place. The learning curve (A) and the trials to criterion (B) reveal time of day did not play a significant role on behavior.
xxiv
Six-week time point:
Most (8/10) of the rats in this group reached learning criterion within 75 trials without
regard to treatment. Evaluation of percent accuracy shows a significant increase in learning
across trial blocks (F(4,40) = 10.283, p<0.0001). As with the 3-week deprivation, 45 g/kg EB
treatment did not increase trials to criterion when administered 6 weeks after OVX (t(4) = 0.972,
p > 0.05; MW U(4,4) = 4; p > 0.05). There is no main effect of treatment in rats on the six-week
time point (F(1,6) = 0.06; p>0.81). There is no interaction between performance across trial
block and treatment in rats on the six-week time point (F(4,24) = 0.526; p>0.71) (Figure 5A-B).
In comparing the rats in the three-week group to those in the six-week group, the oil rats
did not have higher trials to criterion with increased hormone deprivation on the response task,
as was expected (t(6) = -0.089, p > 0.05; MW U(4,4) = 12, p>0.05). There is no main effect of
testing time on response learning in oil or EB-treated rats from three weeks to six weeks ((F(1,8)
=.204; p>0.66); (F(1,10) = 2.069; p>0.18)). There is no interaction between performance across
trial block and testing time in oil (F(4,32) = 0.968, p>0.43) or EB-treated rats (F(4,40) = 0.216;
p>0.92).
There is a significant difference in performance across trial blocks, without regard to
treatment or time after OVX in oil rats (F(4,72) = 31.411; p<.0001). There is no main effect of
weeks since OVX (F(1,18) = 1.127; p>0.3), or of treatment (F(1,18) = 0.038; p>0.84) on
performance. There is no interaction between performance across trial block and testing time
regardless of treatment (F(4,72) = 0.571, p>0.68), between performance across trial block and
treatment regardless of testing time (F(4,72) = 0.146, p>0.96), or between treatment and testing
time regardless of trial block (F(1,18) = 0.38; p>0.84). There is no interaction between
performance across trial block, treatment and testing time (F(4,72) = 0.787, p>0.53).
xxv
1 2 3 4 540
60
80
100
Oil (n=4)Estradiol (n=4)
Perc
ent C
orre
ct (M
ean
± S.
E.M
)
A
chance
10
20
40
60
80
100
Oil (n=4)EB (n=4)
Tria
ls T
o C
riter
ion
(Mea
n ±
S.E.
M)
Oil (n=4)Estradiol (n=4)
B
Figure 5 45g/kg EB treatment did not impair learning when administered 6 weeks after OVX. The learning curve (A) nor the trials to reach learning criterion (B) differed between EB-treated groups and oil vehicle treated groups.
Three-week time point – lower EB dose:
xxvi
After collecting and analyzing the data for the two time points using the high dose and
noting the lack of impairing effects of EB on learning, a different dose was used to determine if
indeed a lower dose was effective. Although the original findings showing the striatum-sensitive
impairments used the physiologically high dose of EB, recent work from our lab showed that a
lower dose of 4.5g/kg was more effective at impairing response learning and at enhancing
hippocampus sensitive cognition (Pisani et al., 2012). All rats reached the learning criterion
within the 75 trials, with learning curves showing significant improvement across trial blocks
(F(4,28) = 9.429, p < 0.0001). The impairment of EB was still not observed. In fact, the EB-
treated rats were slightly better at solving the maze than the oil-treated rats were, however this
difference was not significant as there was no main effect of treatment on learning curves
(F(1,7) = 0.596; p>0.46) or interaction between performance across trial block and treatment in
rats on the three-week time point with low EB doses (F(4,28) = 0.430; p>0.78) (Figure 6A).
These rats also showed no difference in trials to learning criterion (Figure 6B) when evaluated
parametrically (T(7) = -0.29, p > 0.05) or non-parametrically (MW U(4,5) = 8; p > 0.05).
xxvii
1 2 3 4 520
40
60
80
100
OIL (n=5)EB (n=4)
Perc
ent C
orre
ct (M
ean
± S.
E.M
)
A
Oil (n=5)
Estradiol (n=4)
chance
10
20
40
60
80
100
OIL (n=4)EB (n=4)
Tria
ls to
Crit
erio
n (M
ean
± S.
E.M
)
BOil (n=5)
Estradiol (n=4)
Figure 6 Administration of 4.5g/kg EB did not impair striatal learning when administered 3 weeks following OVX. Learning behavior between the EB-treated rats and the oil vehicle treated rats (A). The trials to reach learning criterion was not significantly different between groups (B).
xxviii
Verification of Circulating Estradiol
Uterine horn weights:
Uterine horn weights (Figure 7) were significantly higher in EB-treated rats than in their
vehicle-treated counterparts, regardless of dose and length of hormone deprivation (Effect of
4.5µg/kg in 3 week time point: (t(3) = 8.60, p = < 0.005; MW U (4,5) = 0, p<0.05); Effect of
45µg/kg in 3 week time point: (t(9)= 3.37, p = 0.001; MW U (6,8) = 5, p<0.05); of Effect 45µg/kg
in 6 week time point: (t(4) = 3.65, p = <0.05; MW U (4,4) = 0, p<0.05)). Rats administered
4.5g/kg EB did not have a lighter uterine horn weight than rats given 45g/kg EB (t(4) = -0.55,
p = 0.61; MW U (4,4) = 7, p>0.05). Uterine horn weights verified ovariectomies were successful
and EB treatments were effective.
3 Week - Low Dose 3 Week - High Dose 6 Week - High Dose0
20
40
60
80
100
5 6 44 8 4
OILEB
Ure
rine
Hor
n W
eigh
t (m
g)
(Mea
n ±
S.E.
M)
*** **
*
OilEstradiol
Figure 7 Assessment of uterine horn weight to verify levels of circulating E2 levels and thus the success of subcutaneous injections prior to training. *p<0.05 **p<0.01 ***p<0.001
xxix
Serum Estradiol Levels
As shown in Figure 8, circulating E2 levels were significantly higher in the 45g/kg EB-
treated rats than the vehicle-treated in the three-week time point (t(8) = -4.20, p<0.01; MW U
(6,8) = 0, p<0.05 and the six-week time point (t(6) = -7.60, p<0.001; MW U (4,4) = 0, p<0.05).
Circulating E2 levels, however, were not significantly different between the 4.5g/kg EB-treated
rats and their vehicle-treated counterparts (t(7) = 1.06, p>0.3; MW U (4,5) = 5, p>0.05). These
data suggest that the higher dose was more effective in circulating through the body. On the
other hand, the uterine horn weight data does suggest a difference between 4.5g/kg EB-
treated rats and their vehicle-treated counterparts. Serum E2 levels were significantly higher in
the 45g/kg EB- treated rats in the three-week group compared to the six-week group treated
with the same dose when evaluated parametrically (t(8) = -2.80, p<0.05) but not non-
parametrically (MW U(4,8) = 6, p>0.05). The serum E2 levels of the rats on the three-week time
point treated with 45g/kg EB were also significantly higher than the rats on the same time point
but treated with 4.5g/kg EB (t(9) = 4.839, p<0.0005; MW U(4,8) = 0, p<0.05). Unexpectedly,
the concentrations of E2 in the serum of oil-treated rats from the six-week time point and the oil
rats used as a control to the three-week rats who received 45g/kg of EB were significantly
different (t(8)=-3.41, p<0.005; MW U(4,6) = 0,p<0.05).
xxx
3 Week - Low Dose 3 Week - High Dose 6 Week - High Dose0
20
40
60
80
100
120
140
160
180
200Oil
Estradiol
[Est
radi
ol] p
g/m
L (M
ean
± S.
E.M
)
**
**
**
***
Figure 8 Serum E2 concentration collected from trunk blood following behavioral training. Each group of rats receiving 45g/kg EB had significanty higher E2 concentrations in the blood than their oil vehicle counterpart. The rats on the six-week time point had significantly higher serum E2 levels than both other groups treated with EB. Interestingly, there was a significant increase in E2 levels in the groups of oil rats that were controlling for groups receiving a 45g/kg EB. *p<0.05 **p<0.01 ***p<0.001
Correlation Plots – Serum Estradiol Level and Behavior
The E2 levels in the trunk blood of the rats, measured by the ELISA, suggested one
reason for the lack of impairment on learning behavior could be due to abnormally high levels of
E2 in the vehicle treated rats. If so, learning scores should covary with serum E2 levels. To
determine if this was observed, the total percent correct arm choices throughout all training
(Figure 9A) or trials to reach criterion (Figure 9B) was plotted against serum E2 concentration.
There was no correlation between serum E2 levels and total percent correct (R2 = 0.033, p >
0.3) or trials to criterion (R2 = 0.014, p > 0.5).
xxxi
0 50 100 150 200 25040
60
80
100
R² = 0.0333595179242643
[Estradiol] (pg/mL) (Mean ± S.E.M)
Tota
l % C
orre
ct
A
0 50 100 150 200 2500
20
40
60
80
100
R² = 0.0142193013278415
[Estradiol] (pg/mL) (Mean ± S.E.M)
Tria
ls to
Crit
erio
n
BFigure 9 Plotting serum E2 concentration against total percent correct arm choices in behavior (A) or trials to reach learning criterion (B). The R2 value indicates that it is not just serum E2 level that is influencing behavior. In this case, there could be something else going on that is controlling behavior.
xxxii
Discussion
The first aim of this study was to validate past work in the lab showing impaired
response learning three weeks after OVX with subsequent EB administration 24 and 48 hours
prior to behavior. Although all rats learned the striatum-sensitive response task quite well, there
was not an impairing effect of EB like was shown previously (Korol and Kolo, 2002; Zurkovsky
et al., 2006; Pisani et al., 2012). The six-week time point was conducted next, to see if there
was any difference between these two time points, specifically looking at whether oil rats
perform worse with increased deprivation. When there was neither a learning impairment in the
six-week time point nor an impairment in performance of vehicle treated rats, a lower dose was
utilized in attempt to replicate our past work. After replicating this experiment with this lower
dosage of EB treatment in exclusively the three-week time point, an impairing effect of EB on
learning was still not found.
Perhaps the reason for the lack of impairment on response learning was due to poor
injections. Circulating levels of E2 validated the efficacy of EB injections. High uterine horn
weight and serum E2 concentration show the injections, however, were effective. The uterine
horn weight and serum E2 concentrations suggest that despite not getting the expected
behavioral results, the injections of EB were effective. However, despite significant differences
in uterine horn weight and serum E2 levels between EB- and oil-treated rats, the vehicle treated
groups nevertheless had unusually high E2 levels. Other work shows vehicle rats should have
serum E2 concentrations <8pg/mL (Pisani et al 2012, Korol and Kolo 2002). These values
greatly differ from the values in this study, where vehicle-treated rats had serum E2
concentrations that ranged from ~27-51pg/mL. Uterine horn weights, even in vehicle-treated
animals were also larger than expected.
xxxiii
A significant difference existed between serum E2 levels between the vehicle treated
rats in the six-week time point and the vehicle treated rats controlling for the high dose treated
rats on the three-week time point. The oil groups corresponding to each EB treatment group
should not differ, as they received no E2 replacement. If anything, perhaps this difference is
related to having more time for E2 to clear from the blood after OVX, but by three weeks, this
should have already occurred. Nonetheless, perhaps this high E2 concentration in vehicle-
treated rats could be the reason for the lack of impairment on behavior in E2 treated rats
compared to their oil counterparts and the lack of impairment in oil-treated rats over time.
Perhaps there was still enough E2 in the body to have an effect on learning and thus the
behavior of those with E2 replacement did not look much different than oil rats. Serum E2
concentration of ~15pg/mL is enough to produce the learning impairment in rats. Thus, oil
animals might have had serum E2 concentrations well above this critical level for impairments,
which could be the reason for the lack of treatment effects (Pisani et al 2012). This would
suggest that despite having circulating E2, there may not have been enough additional E2
above levels seen in vehicle controls to produce impairments in the EB-treated rats. However,
this possibility is less likely due to the high concentration of serum E2. There seems to be a
threshold level of E2 to produce behavioral effects and changes in vaginal cytology (Davidson et
al 1968). The correlations suggest that concentration of E2 did not influence trials to criterion or
total percent correct (Figure 9), perhaps suggesting a threshold effect of E2 in learning behavior
as well.
The oil-treated rats could have higher levels of circulating EB, as shown in uterine horn
weight and serum E2 concentration, due to contamination of the sesame oil used for injection.
Contamination could happen in a number of ways, such as using a pipette that had previously
been used with E2, or putting the sesame oil into E2 glassware. Another study currently being
conducted in the lab has also shown similar high levels of circulating E2 in vehicle treated rats.
Additionally, sesame oil has been shown to be a phytoestrogen, which may in turn influence
xxxiv
results (Wu et al 2006, El Wakf et al 2013). To address this issue within the lab, perhaps we
should begin inclusion of a group with no injection at all. If we included a vehicle injection
group, a no injection group, and the experimental group, we can control for both effects of
injection and whether the vehicle was contaminated.
The XY plots comparing serum EB concentration to both the total percent correct arm
choices (Figure 9A) and trials to criterion (Figure 9B) were created to see if serum EB
concentration was correlated to performance on the maze. If there was a significant correlation
between these variables, it might suggest high EB levels in oil rats prevented an impairment on
behavior, which may indicate a contamination issue. The lack of significance comparing these
relationships in Figure 9, however, decrease the likelihood that sesame oil contamination is the
reason for the lack of E2 impairment, yet the high uterine horn weights suggest it still could play
a role.
The lack of effect of E2 on learning and the lack of change in learning in oil rats with a
longer period of hormone deprivation left many questions unanswered regarding the way the
brain’s sensitivity to E2 over time. One reason underlying the lack of effect of E2 on striatal
learning could related to circadian rhythms and corticosterone (CORT). Corticosterone is the
rodent equivalent of the human hormone cortisol, which plays a role in the stress response
(Feodora et al 2014). CORT, like female reproductive hormones, is also synthesized from
cholesterol, passing through progesterone intermediates during synthesis (Lemaire et al 2005;
Feodora et al 2014). Within the brain, actions of CORT are mediated by glucocorticoid and
mineralocorticoid receptors (Lemaire et al 205). CORT levels are known to fluctuate with
circadian rhythms, specifically, CORT levels are highest midday, at about 2-4 PM and are
highest during proestrus than during the other estrous stages in the F344 rat model (Haim et al
2003).
Male rats who are stressed during behavioral training are known to shift to using
response behavior (Sadowski et al 2009). Therefore, we know stress may also have some
xxxv
implications on learning in female rats. One initial thought explaining the results from this study
was that perhaps rats happened to be less stressed during behavior in the morning, shifting
their performance to slower response learning, and rats trained in the afternoon were more
stressed, potentially masking any hormone effect. If rats trained in the afternoon do learn
quicker because of stress, EB may be less likely to impair learning because stress forces a
cognitive shift that instead favors the striatum. On the other hand, an impairment may be easier
to detect if rats learn quickly because there is more room for impairment, unless the effect of
stress blocks the EB effect. The rats trained in the morning do seem to exhibit lower scores
through the training, yet this difference is small and unlikely to be important, especially
considering the small sample size. Although we did not see a difference in performance when
binning rats based on the time of day, it remains likely that circulating CORT and circadian
rhythms could play a role in learning due to their changes based on estrous cycle stage and
relation to stress. Indicators of stress on the maze include number of fecal boli, time spent
grooming, and time spent rearing vs. hovering. These could have been assayed through video
recording to look at stress as a variable, although this was not done in this study.
Other future work focuses on assessing the effects of differing lengths of ovarian
hormone depletion across ages on both the hippocampus- and striatum-sensitive learning tasks.
This approach will account for any differences in how hormone depletion modulates learning in
multiple brain areas. Within this aim, two tasks per brain area may be used to examine whether
there is a difference in hormone effects based on the task, even if both are associated with the
same brain area. Striatum-sensitive object recognition and/or striatum-sensitive response
learning may impair learning like was previously reported in the literature but not shown in this
study (Kundu et al 2018). Similarly, utilizing hippocampus-sensitive object location and/or
hippocampal place learning may also reveal previously reported effects that were not able to be
reproduced in this study. This approach might lead to the development of batteries of cognitive
xxxvi
tests that tap a number of brain regions, which in turn would be useful for evaluating the effects
of hormone deprivation, interventions, and the critical window hypothesis across the lifespan.
Moving forward, the work of this study has identified multiple future directions and new
interests building upon the findings from this work. The neuroscience field, especially
behavioral neuroscience, has mostly focused on the use of male subjects to avoid any
confounds of cycling hormones in females. As a result, much of what we know about the brain
and behavior is only applicable to males. Because of these hormone fluctuations, we typically
find that results in males cannot be comparable to females. Some therapies, especially those for
neurodegenerative disease, may even be designed with men in mind. Avoiding the confound of
cycling hormones leaves half the population out of the question when addressing behavioral
neuroscience knowledge. The field is beginning to work against this bias with increasing
vigilance, due to the emerging relationships between estrogens, cognition, and
neurodegenerative disease.
xxxvii
Works Cited
Ábrahám, István M., et al. “Action of Estrogen on Survival of Basal Forebrain Cholinergic Neurons: Promoting Amelioration.” Psychoneuroendocrinology, vol. 34, no. SUPPL. 1, Pergamon, Dec. 2009, pp. S104–12, doi:10.1016/j.psyneuen.2009.05.024.
Armstrong, Katrina, et al. “Hormone Replacement Therapy and Life Expectancy After Prophylactic Oophorectomy in Women With BRCA1/2 Mutations: A Decision Analysis.” Journal of Clinical Oncology, vol. 22, no. 6, American Society of Clinical Oncology, Mar. 2004, pp. 1045–54, doi:10.1200/JCO.2004.06.090.
Arnauld, E., et al. “Effects of Estrogens on the Responses of Caudate Neurons to Microiontophoretically Applied Dopamine.” Neuroscience Letters, vol. 21, no. 3, Elsevier, Feb. 1981, pp. 325–31, doi:10.1016/0304-3940(81)90225-1.
Auletta, Carol S. “Vaginal and Rectal Administration.” Journal of the American College of Toxicology, vol. 13, no. 1, 1994, pp. 48–63, doi:10.3109/10915819409140655.
Azcoitia, I., et al. “Estradiol Synthesis within the Human Brain.” Neuroscience, vol. 191, Pergamon, 15 Sept. 2011, pp. 139–47, doi:10.1016/j.neuroscience.2011.02.012.
Bean, Linda A., et al. “Re-Opening the Critical Window for Estrogen Therapy.” The Journal of Neuroscience, vol. 35, no. 49, Dec. 2015, p. 16077 LP-16093, http://www.jneurosci.org/content/35/49/16077.abstract.
Benedetti, Maria D., et al. “Hysterectomy, Menopause, and Estrogen Use Preceding Parkinson’s Disease: An Exploratory Case-Control Study.” Movement Disorders, vol. 16, no. 5, John Wiley & Sons, Ltd, Sept. 2001, pp. 830–37, doi:10.1002/mds.1170.
Brown, Roger, and David McNeill. “The ‘Tip of the Tongue’ Phenomenon.” Journal of Verbal Learning and Verbal Behavior, vol. 5, no. 4, Academic Press, Aug. 1966, pp. 325–37, doi:10.1016/S0022-5371(66)80040-3.
Conrad, Cheryl D., et al. “Acute Stress Impairs Spatial Memory in Male but Not Female Rats: Influence of Estrous Cycle.” Pharmacology Biochemistry and Behavior, vol. 78, no. 3, Elsevier, 2004, pp. 569–79, doi:10.1016/j.pbb.2004.04.025.
Cora, Michelle C., et al. “Vaginal Cytology of the Laboratory Rat and Mouse: Review and Criteria for the Staging of the Estrous Cycle Using Stained Vaginal Smears.” Toxicologic Pathology, vol. 43, no. 6, SAGE PublicationsSage CA: Los Angeles, CA, Aug. 2015, pp. 776–93, doi:10.1177/0192623315570339.
Cui, Jie, et al. “Estrogen Synthesis and Signaling Pathways during Aging: From Periphery to Brain.” Trends in Molecular Medicine, vol. 19, no. 3, Mar. 2013, pp. 197–209, doi:10.1016/j.molmed.2012.12.007.
Daniel, Jill M., and Christopher D. Lee. “Estrogen Replacement in Ovariectomized Rats Affects Strategy Selection in the Morris Water Maze.” Neurobiology of Learning and Memory, vol.
xxxviii
82, no. 2, Academic Press, Sept. 2004, pp. 142–49, doi:10.1016/j.nlm.2004.06.001.
Daniel, Jill M. “Estrogens, Estrogen Receptors, and Female Cognitive Aging: The Impact of Timing.” Hormones and Behavior, vol. 63, no. 2, Academic Press, 1 Feb. 2013, pp. 231–37, doi:10.1016/j.yhbeh.2012.05.003.
Daniel, Jill M., et al. “Effects of Ovarian Hormones and Environment on Radial Maze and Water Maze Performance of Female Rats.” Physiology and Behavior, vol. 66, no. 1, Elsevier Inc., Mar. 1999, pp. 11–20, doi:10.1016/S0031-9384(98)00272-8.
Davidson, Julian M., et al. “Relative Thresholds of Behavioral and Somatic Responses to Estrogen.” Physiology and Behavior, vol. 3, no. 2, Elsevier, Mar. 1968, pp. 227–29, doi:10.1016/0031-9384(68)90090-5.
Dluzen, Dean E., and Martin W. I. M. Horstink. “Estrogen as Neuroprotectant of Nigrostriatal Dopaminergic System Laboratory and Clinical Studies.” Endocrine, vol. 21, no. 1, 2003.
El Wakf, Azza M., et al. “Osteoprotective Effect of Soybean and Sesame Oils in Ovariectomized Rats via Estrogen-like Mechanism.” Cytotechnology, vol. 66, no. 2, Springer, Mar. 2014, pp. 335–43, doi:10.1007/s10616-013-9580-4.
Finch, Amy P. M., et al. Impact of Oophorectomy on Cancer Incidence and Mortality in Women With a BRCA1 or BRCA2 Mutation. 2014, doi:10.1200/JCO.2013.53.2820.
Gandy, Sam. “Estrogen and Neurodegeneration*.” Neurochemical Research, vol. 28, no. 7, 2003.
Gould, Elizabeth, et al. “Gonadal Steroids Regulate Dendritic Spine Density in Hippocampal Pyramidal Cells in Adulthood.” The Journal of Neuroscience, no. 4, 1990, http://www.jneurosci.org/content/jneuro/10/4/1286.full.pdf.
Haim, S., et al. “Serum Levels of Sex Hormones and Corticosterone throughout 4- and 5-Day Estrous Cycles in Fischer 344 Rats and Their Stimulation in Ovariectomized Females.” Journal of Endocrinological Investigation, vol. 26, no. 10, Editrice Kurtis s.r.l., 2003, pp. 1013–22, doi:10.1007/BF03348201.
Hendrix, Susan L. “Bilateral Oophorectomy and Premature Menopause.” American Journal of Medicine, vol. 118, no. 12 SUPPL. 2, Elsevier Inc., 2005, pp. 131–35, doi:10.1016/j.amjmed.2005.09.056.
Kostadinova, Feodora, et al. “Why Does the Gut Synthesize Glucocorticoids?” Annals of Medicine, vol. 46, no. 7, Informa Healthcare, 1 Nov. 2014, pp. 490–97, doi:10.3109/07853890.2014.932920.
Korol, Donna L., and Lacy L. Kolo. “Estrogen-Induced Changes in Place and Response Learning in Young Adult Female Rats.” Behavioral Neuroscience, 2002, doi:10.1037/0735-7044.116.3.411.
xxxix
Korol, Donna L., and Wei Wang. “Using a Memory Systems Lens to View the Effects of Estrogens on Cognition: Implications for Human Health.” Physiology and Behavior, 2018, doi:10.1016/j.physbeh.2017.11.022.
Korol, Donna L., et al. “Shifts in Preferred Learning Strategy across the Estrous Cycle in Female Rats.” Hormones and Behavior, vol. 45, no. 5, Academic Press, May 2004, pp. 330–38, doi:10.1016/j.yhbeh.2004.01.005.
Kundu, Payel, et al. “Licorice Root Components Mimic Estrogens in an Object Location Task but Not an Object Recognition Task.” Hormones and Behavior, vol. 103, Academic Press Inc., July 2018, pp. 97–106, doi:10.1016/j.yhbeh.2018.06.002.
Lemaire, V., et al. “Chapter 3.5 Glucocorticoids and Motivated Behaviour.” Techniques in the Behavioral and Neural Sciences, vol. 15, no. PART 1, Academic Press, 2005, pp. 341–58, doi:10.1016/S0921-0709(05)80020-3.
McCarrey, Anna C., and Susan M. Resnick. “Postmenopausal Hormone Therapy and Cognition.” Hormones and Behavior, vol. 74, Academic Press, Aug. 2015, pp. 167–72, doi:10.1016/J.YHBEH.2015.04.018.
Miller, Virginia M., and S. Mitchell Harman. “An Update on Hormone Therapy in Postmenopausal Women: Mini-Review for the Basic Scientist.” American Journal of Physiology - Heart and Circulatory Physiology, vol. 313, no. 5, American Physiological Society, 6 Nov. 2017, pp. H1013–21, doi:10.1152/ajpheart.00383.2017.
Mitchell, Ellen Sullivan, and Nancy Fugate Woods. Midlife Women’s Attributions about Perceived Memory Changes: Observations from the Seattle Midlife Women’s Health Study. Vol. 10, no. 4, Mary Ann Liebert, Inc, 2001, www.liebertpub.com.
Mizumori, Sheri J. Y., et al. “Parallel Processing across Neural Systems: Implications for a Multiple Memory System Hypothesis.” Neurobiology of Learning and Memory, vol. 82, no. 3, Academic Press, 2004, pp. 278–98, doi:10.1016/j.nlm.2004.07.007.
Packard, Mark G., and James L. Mcgaugh. “Inactivation of the Hippocampus or Caudate Nucleus with Lidocaine Differentially Affects Expression of Place and Response Learning Inactivation of Hippocampus or Caudate Nucleus with Lidocaine Differentially Affects Expression of Place and Response Learning.” Article in Neurobiology of Learning and Memory, vol. 65, 1996, https://www.researchgate.net/publication/312750029.
Pisani, Samantha L., et al. “Acute Genistein Treatment Mimics the Effects of Estradiol by Enhancing Place Learning and Impairing Response Learning in Young Adult Female Rats.” Hormones and Behavior, vol. 62, no. 4, Academic Press, Sept. 2012, pp. 491–99, doi:10.1016/j.yhbeh.2012.08.006.
Plant, Tony M. “The Hypothalamo-Pituitary-Gonadal Axis.” Journal of Endocrinology, vol. 226, no. 2, BioScientifica Ltd., 2015, pp. T41–54, doi:10.1530/JOE-15-0113.
Quinlan, Matthew G., et al. “Use of Cognitive Strategies in Rats: The Role of Estradiol and Its Interaction with Dopamine.” Hormones and Behavior, vol. 53, no. 1, Academic Press, Jan. 2008, pp. 185–91, doi:10.1016/j.yhbeh.2007.09.015.
xl
Ragonese, P., et al. “Implications for Estrogens in Parkinson’s Disease: An Epidemiological Approach.” Annals of the New York Academy of Sciences, vol. 1089, no. 1, Blackwell Publishing Inc., Nov. 2006, pp. 373–82, doi:10.1196/annals.1386.004.
Rapp, Stephen R., et al. “Effect of Estrogen Plus Progestin on Global Cognitive Function in Postmenopausal Women - The Women’s Health Initiative Memory Study: A Randomized Controlled Trial.” Journal of the American Medical Association, vol. 289, no. 20, American Medical Association, May 2003, pp. 2663–72, doi:10.1001/jama.289.20.2663.
Roebuck, Andrew J., et al. “Acute Stress, but Not Corticosterone, Facilitates Acquisition of Paired Associates Learning in Rats Using Touchscreen-Equipped Operant Conditioning Chambers.” Behavioural Brain Research, vol. 348, Elsevier B.V., Aug. 2018, pp. 139–49, doi:10.1016/j.bbr.2018.04.027.
Sadowski, Renee N., et al. “Effects of Stress, Corticosterone, and Epinephrine Administration on Learning in Place and Response Tasks.” Behavioural Brain Research, vol. 205, no. 1, Elsevier, Dec. 2009, pp. 19–25, doi:10.1016/J.BBR.2009.06.027.
Shanafelt, Tait D, et al. “Pathophysiology and Treatment of Hot Flashes.” Mayo Clinic Proceedings, vol. 77, 2002.
Sherwin, Barbara B., and Jessica F. Henry. “Brain Aging Modulates the Neuroprotective Effects of Estrogen on Selective Aspects of Cognition in Women: A Critical Review.” Frontiers in Neuroendocrinology, vol. 29, no. 1, Academic Press, 1 Jan. 2008, pp. 88–113, doi:10.1016/j.yfrne.2007.08.002.
Shumaker, Sally A., et al. “Conjugated Equine Estrogens and Incidence of Probable Dementia and Mild Cognitive Impairment in Postmenopausal Women: Women’s Health Initiative Memory Study.” Journal of the American Medical Association, vol. 291, no. 24, American Medical Association, June 2004, pp. 2947–58, doi:10.1001/jama.291.24.2947.
Shumaker, Sally A., et al. “Estrogen Plus Progestin and the Incidence of Dementia and Mild Cognitive Impairment in Postmenopausal Women - The Women’s Health Initiative Memory Study: A Randomized Controlled Trial.” Journal of the American Medical Association, vol. 289, no. 20, American Medical Association, May 2003, pp. 2651–62, doi:10.1001/jama.289.20.2651.
Tang, Ming Xin, et al. “Effect of Oestrogen during Menopause on Risk and Age at Onset of Alzheimer’s Disease.” Lancet, vol. 348, no. 9025, Lancet Publishing Group, Aug. 1996, pp. 429–32, doi:10.1016/S0140-6736(96)03356-9.
Van Duijn, Cornelia M. “Hormone Replacement Therapy and Alzheimer’s Disease.” Maturitas, vol. 31, no. 3, Elsevier, Mar. 1999, pp. 201–05, doi:10.1016/S0378-5122(99)00005-5.
Villa, Alessandro, et al. “Estrogens, Neuroinflammation, and Neurodegeneration.” Endocrine Reviews, vol. 37, no. 4, Oxford University Press, 1 Aug. 2016, pp. 372–402, doi:10.1210/er.2016-1007.
von Mühlen, Denise Goldani, et al. “A Community-Based Study of Menopause Symptoms and Estrogen Replacement in Older Women.” Maturitas, 1995, doi:10.1016/0378-
xli
5122(95)00917-A.
Wu, Wen-Huey, et al. “Nutrient Physiology, Metabolism, and Nutrient-Nutrient Interactions Sesame Ingestion Affects Sex Hormones, Antioxidant Status, and Blood Lipids in Postmenopausal Women 1.” J. Nutr, vol. 136, 2006, https://academic.oup.com/jn/article-abstract/136/5/1270/4669984.
Yin, Weiling, et al. “Testing the Critical Window Hypothesis of Timing and Duration of Estradiol Treatment on Hypothalamic Gene Networks in Reproductively Mature and Aging Female Rats.” Endocrinology, vol. 156, no. 8, Endocrine Society, Aug. 2015, pp. 2918–33, doi:10.1210/en.2015-1032.
Zárate, Sandra, et al. “Role of Estrogen and Other Sex Hormones in Brain Aging. Neuroprotection and DNA Repair.” Frontiers in Aging Neuroscience, vol. 9, Frontiers Media SA, 2017, doi:10.3389/FNAGI.2017.00430.
Zurkovsky, L., et al. “Estrogen Modulates Learning in Female Rats by Acting Directly at Distinct Memory Systems.” Neuroscience, vol. 144, no. 1, Pergamon, Jan. 2007, pp. 26–37, doi:10.1016/j.neuroscience.2006.09.002.
xlii