Estrogenic regulation of memory: The first 50 years

Abstract

This review highlights fifty years of progress in research on estradiol’s role in regulating behavior(s). It was initially thought that estradiol was only involved in regulating estrus/menstrual cycles and concomitant sexual behavior, but it is now clear that estradiol also influences the higher order neural function of cognition. We provide a brief overview of estradiol’s regulation of memory and some mechanisms which underlie its effects. Given systemically or directly into the hippocampus, to ovariectomized female rodents, estradiol or specific agonists, enhance learning and/or memory in a variety of rodent cognitive tasks. Acute (within minutes) or chronic (days) treatments enhance cognitive functions. Under the same treatment conditions, dendritic spine density on pyramidal neurons in the CA1 area of the hippocampus and medial prefrontal cortex increase which suggests that these changes are an important component of estrogen’s ability to impact memory processes. Noradrenergic, dopaminergic and serotoninergic activity are also altered in these areas following estrogen treatments. Memory enhancements and increased spine density by estrogens are not limited to females but are also present in castrate males. In the next fifty years, neuroscientists need to determine how currently described neural changes mediate improved memory, how interactions among areas important for memory promote memory and the potential significance of neurally derived estrogens in normal cognitive processing. Answering these questions may provide significant advances for treatment of dementias as well as age and neuro-degenerative disease related memory loss.

1. Introduction

In the early years of Hormones and Behavior, approximately fifty years ago, the effects of estrogen on behavior were quite limited and consisted of regulating female sexual behavior (Hart, 1970) and perhaps influencing ingestive behavior (ter Haar, 1972). While behavior was believed to be regulated in the CNS, many scientists thought that estrogens did not act directly in the brain but, rather, altered other hormones or chemicals which in turn acted in the brain to regulate behavior. After all, receptors for estradiol were only beginning to be described in the uterus in the late 60s (Jensen et al., 1968), and it would take several more years for the demonstration of estrogen receptors in the brain (Pfaff, 1968; Stumpf, 1972). Fast forward to 2020; it is now recognized that multiple receptors for estrogens are present throughout the brain and are located in cell nuclei, axons, cell membranes of neurons, dendritic spines, and pre- and post-synaptic terminals (Almey et al., 2015). Moreover, estrogens have been shown to impact a wide variety of behaviors, besides sexual behaviors, and include mood and affective disorders and the higher order neural function, cognition. As we discuss in this review, a body of research also demonstrates that cognition is regulated by estrogens in males as well as females.

The recognition that gonadal steroids can alter cognition function was a slowly evolving process based on research primarily in rodents, but also sub-human primates and humans. However, this work is still met with some skepticism in the neuroscience community. While estrogen is recognized to impact cognition, some question whether this interaction is consequential. Recognition of estrogenic influences on neural function is also intertwined with issues of its impact (or lack thereof) during the menopausal transition and post menopause (Koebele and Bimonte-Nelson, 2015) and with the general resistance of many in the scientific community to consider sex as a biological variable in research (Cahill and Aswad, 2015). Yet, in the last 50 years, substantial research has accumulated showing that estrogens affect cognitive function, and it is important to acknowledge that Hormones & Behavior has been supportive of this work by publishing many research articles on this topic and two special issues devoted to estradiol and cognition. Volume 34 (1998) “Estrogen effects on cognition across the lifespan” provided, what seemed at the time, a very bold compendium of an emerging research field. However, the field successfully matured and 17 years later, an even larger volume, 74 (2015) “Estradiol and cognition: molecules to mind” was published which confirmed and expanded upon the original research presented in 1988. In this article, we provide an historical perspective on estradiol and cognition derived mainly from our own research.

2. Effects of estradiol on memory tasks

The demonstration that estradiol binds to receptors in the hippocampus and causes estrogen-dependent alterations in the morphology of dendritic trees, specifically the density of spines in this structure, spurred investigations of its effects on spatial memory. In addition, we and other groups had shown that estrogens affected cholinergic neurotransmission in basal forebrain nuclei and their targets, particularly the hippocampus (Luine, 1985). Cholinergic systems had long been tied to cognition and death of these neurons is a major contributor to memory loss in aging and Alzheimer’s disease (Bartus et al., 1982). Moreover, monoamines had been shown to strongly influence cognition, and gonadal hormones also acted through these systems to regulate gonadotropin release (Rance et al., 1981). Thus, it was not surprising when various labs demonstrated that estrogens enhanced learning and memory (See Luine, 2014 and Frick, 2015 for reviews) and that cholinergic and monoaminergic containing neurons were site(s) for estrogen actions in enhancing enhance memory (see section 5).

In Figure 1, behavioral results from a variety of studies are shown. Early studies utilized hippocampal dependent tasks like the Morris water maze (Fig. 1 A) and 8-arm radial maze (Fig 1B). In the water maze, ovariectomized (OVX) mice showed longer latencies and swim paths to find a submerged platform than intact mice (42 vs 28 sec) and treatment with estradiol to OVX subjects for 5 five weeks decreased them (Xu and Zhang, 2006). On the radial arm maze, OVX rats had Silastic capsules containing 25% estradiol or 100% cholesterol for 30 days prior to and continuing through 24 days of maze training (Fig. 1B). Estradiol treated subjects performed better showing a greater number of correct arm choices than OVX + cholesterol subjects in the first eight visits averaged over 24 days of training. For a comprehensive review of early studies on estradiol effects on learning and memory, see Dohanich, 2002.

Figure 1:

Estradiol enhances learning and/or memory in a variety of rodent tasks

A. Water maze – Effects of 5-week treatment with estradiol benzoate (EB, 20,100, 200 ug/kg, S.C.) on (A) escape latency of finding platform and (B) distance traveled reach the platform of OVX mice in water maze. The x-axis denotes each day’s average of four training session. Value differs from the Sham at ## p < 0.001. Value differs from the vehicle-treated OVX group at * P < 0.05 and ** P < 0.01. Reprinted from Zu et al, 2006 by permission.

B. Radial arm maze -Effects of estradiol administer prior to and during training on a radial arm maze. OVX rats were implanted with 5 mm Silastics containing 25% estradiol or 100% cholesterol for 30 days prior to and continuing through 24 days of maze training. (A) Mean number of correct arm choices in the first eight visits averaged over 24 days of training (*P < 0.01; ANOVA). (B) Mean number of correct arm choices in the first eight visits presented as 4-day blocks over 24 days of training. Reprinted from Daniel et al, 1997 by permission.

C. Object placement memory - Effects of acute E₂ and testosterone (T) treatment on place memory in males. Hormones were administered to castrate males immediately following the training trial and were tested two hours later. Time spent exploring objects at old and new location is shown for vehicle- and T-treated, 750 ug/kg (left panel) and vehicle- and E-treated, 20 ug/kg, castrated subjects (right panel). * P < 0.05, by paired t test. Reprinted by permission from Jacome et al, 2016.

D. Recognition memory - Effects of chronic estradiol treatment on object and place recognition memory. OVX rats received 2 days of S.C. EB, 50 ug/kg, and object and place memory was tested separately 48 h after the last dose. Entries are ratios (new/old + new) of time spent exploring each object and objects in each location for vehicle and EB-treated subjects. Dotted line at 0.5 indicates spending the same amount of time exploring new and old objects or locations. *** p < 0.001. Data redrawn from Jacome et al (2010).

E. Inhibitory Avoidance - Effects of acute E₂ on crossover latencies in an inhibitory avoidance test. Crossover latencies are plotted for OVX rats given vehicle (open bars) or estradiol (10 ug) immediately post-training, 0, 1, 2, or 3 h post training and tested 24 h later. * p < 0.05. Redrawn from Rhodes and Frye, 2004.

More recently, recognition memory tasks have been utilized to investigate memory processes and to determine possible effects of drugs and hormones (Ennaceur and Meliani, 1992). These tasks utilize the natural novelty seeking, exploratory nature of rodents and, thereby, mitigate possible confounding influences of task requirements, experience, reinforcements (positive or negative), or psychological performance variables. Subjects receive a training trial where they explore two identical objects, and after an inter-trial delay of an hour to several hours, they are returned to arena for a test trial. In the test or retention trial, subjects view the same objects but one of the objects is moved to a new location (object placement, OP). Replacement of one object with a new/novel object tests object recognition memory (OR) in the retention trial. Both object and place memory rely on the hippocampus and the prefrontal cortex, but memory for objects relies less on the hippocampus and more on the prefrontal cortex than place memory while the opposite relationship is true for place memory (Luine, 2015).

Fig. 1D shows results of recognition memory testing, object and place, in OVX females treated with estradiol benzoate for two days. No differences in exploration of objects are found in the training trial (data not shown), but in the test trial conducted 4 h later, estradiol treated females had exploration ratios greater 0.5 suggesting remembrance of old objects and old locations while OVX + vehicle rats had exploration ratios of 0.5 showing chance performance. Thus, enhancements in recognition memory are similar to results with other spatial and non-spatial memory tasks following chronic estrogen regimens. These estrogen-dependent effects occur mainly through genomic mediation by nuclear estrogen receptors, are pervasive over the lifespan, and contribute to well-known sex differences in cognition, changes in cognition over the menstrual and estrous cycles, and to declines in cognition during aging (See Luine 2014 for details).

Recently, estrogens have been shown to enhance some memory tasks very rapidly, within minutes to hours, following administration to OVX rodents (Fig 1C; Table 1). Possible acute effects of estrogens were investigated when it was demonstrated that estrogen receptors, in addition to being localized intracellularly, were also present in cell membranes of neurons, dendritic spines, and in pre- and post-synaptic terminals (Almey et al., 2015). Membrane receptors coupled to cell signaling cascades are able to rapidly orchestrate hormonal or other agonistic effects, and estrogens have now been shown to rapidly initiate female rodent sexual behavior, avian male sexual displays, nutrient ingestion, and social behaviors within minutes (Micevych et al., 2015). Thus, experiments in the Luine and MacLusky laboratories were the first to report that estradiol, given 30 min before a training trial or immediately after the training trial, enhanced OR and OP 2 - 4 h later in female OVX rats (Luine et al., 2003). As shown in Fig 1C, estradiol or testosterone given to castrate rats immediately following the training trial, enhances object and place memory 2 h later. Thus, memory in males is also rapidly enhanced by gonadal hormones. A variety of estrogens and selective SERMs enhance memory in these tests (Inagaki et al., 2012; Inagaki et al., 2010; Jacome et al., 2016; Luine et al., 2003; Phan et al., 2012; Phan et al., 2015) and enhancements are rapid, occurring within 40 min post estradiol injection (Phan et al., 2015). In addition, intra-hippocampal injection of estradiol is sufficient to rapidly enhance object or place memory (Fernandez et al., 2008). Also shown in Fig 1E is that OVX rats are impaired in inhibitory avoidance as compared to intact rats and that estradiol given immediately after the training trial, but not 1-3 h later, returns cross-over latencies to levels present in intact females (Rhodes and Frye, 2004). The time dependency of the effect post training shows that estradiol is acting on mnemonic processes to enhance memory consolidation and this relationship is also present in OR and OP testing (Inagaki et al., 2010; Luine et al., 2003).

Table 1:

Effects of acute and chronic Estradiol treatments and changes in physiological states on recognition memory and spine density.

Treatment/Physiological state	Memory Effect	Spine Change		Reference
		Hipp	mPFC
S.C. INJECTION
E2 (2 Days)	↑ OR, ↑ OP	?	?	Jacome et al, 2010
E2 (2 Days)	↑ OP	↑	=	Luine & Frankfurt, 2012
E2 (2 Days)	↑ OR, ↑ OP	?	?	Scharfman et al, 2007
E2 (3 Days)	?	↑	?	Gould et al, 1990
E2 (5 Days)*	↑ OP	↑	?	Li et al, 2004
E2 (7 Days)	?	?	↑	Khan et al, 2013
E2(2–4 h)	↑ OR, ↑ OP	↑	↑	Luine et al, 2003 Inagaki et al, 2010; 2012 Jacome et al, 2010; Frye et al, 2007
DES (4 h)	↑ OR, ↑ OP	?	?	Luine et al, 2003
E2 (40 min)*	↑ OR, = OP	↑	?	Phan et al, 2012; 2015
DPN (40 min)*	= OR, = OP1 = OR, ↑ OP2	↓ =	? ?	Phan et al, 20111 Phan et al, 20152
DPN (4 h)	↑ OR, ↑ OP = OP	?	?	Jacome et al, 2010 Frye et al, 2007
PPT (40 min)*	↑ OR, ↑ OP	↑	?	Phan et al, 2011; 2015
PPT (4 h)	= OR, = OP ↑ OP	?	?	Jacome et al, 2010 Frye et al, 2007
HIPPOCAMPAL INJECTION
E2 (48 h)*	↑ OR	?	?	Fernandez et al, 2008; Fortress et al, 2013; 2014
E2 (48 h)*	↑ OR, ↑ OP	?	?	Boulware et al, 2013
E2 (48 h)*	?	↑	↑	Tuscher et al, 2016
DPN (48 h)*	↑ OR, ↑ OP	?	?	Boulware et al, 2013
PPT (48 h)*	↑ OR, ↑ OP	?	?	Boulware et al, 2013
PHYSIOLOGICAL STATE
Ovariectomy	↓ OR, ↓ OP	↓	↓	Wallace et al, 2006
Castration	↓ OP	?	?	Luine, 2015
Proestrus	? ↑ OR* ↑ OP	↑ ? ?	? ? ?	Woolley et al, 1990; Kinsley et al, 2006; Gonzalez-Burgos et al, 2005 Walf et al, 2009 Frye et al, 2007
Aged females	↓ OR, ↓ OP	↓	↓	Wallace et al, 2007;Luine et al, 2011
Cas + E (2 h)	↑ OP	↑	?	Jacome et al, 2016
Cas + E (2 h)	?	↑	?	Avila et al, 2017

Estradiol treatments were given to OVX subjects; All experiments conducted in rats except * which were in mice. Updated from Luine and Frankfurt, 2020.

The functional relevance of these rapid mnemonic changes is unknown, but it is possible that they may play a role in responding to stressors. Stress causes peripheral release and central synthesis of estradiol (Liu et al., 2011) which might, for example, promote beneficial memories such as the location or details of an attack for animals or the locations of traffic jams or dangerous locales for people. Since phytoestrogens also enhance memory by acting through estrogen receptors, our ancestor hunter-gatherers, or grazing animals today, may have benefited from the presence of phytoestrogens in clover and soy plants because they would enhance place memory for these high protein containing plants by acting through estrogen receptors (Luine et al., 2006). Finally, because estradiol is synthesized de novo in the hippocampus and other brain areas (Kato et al., 2013), estrogens may contribute to normal memory formation (see Frick, 2015 for further discussion). See Luine (2015) for further discussion of the possible functional relevance of rapid effects of estrogens on memory.

In sum, a large body of research shows that estradiol, selective estrogen receptor modulators (SERMs), androgens, and various environmental estrogens/androgens modulate cognitive function in both females and males (more evidence exists for females). Generally, effects are positive in nature, resulting in enhancements of learning, memory or both, but in some cases, mainly striatal dependent memory tasks, estrogens impair cognition (Korol and Pisani, 2015). The literature is not always consistent; see Luine, 2014 for a discussion of the many factors which influence outcomes of estrogen treatments on cognition. In the next sections, we consider mechanisms which may underlie the effects of estrogens on cognitive processes and suggest future directions for the research.

3. Estrogens increase dendritic spine density

Studies by Gould, Woolley and Frankfurt, working in the McEwen laboratory, which showed fluctuations of CA1 hippocampal dendritic spines over the estrous cycle and restoration of spine density in OVX females by estradiol (Gould et al., 1990), was seminal in stimulating research on possible influences of estrogens in cognitive functions. The hippocampus and, later, the medial prefrontal cortex (mPFC) had been identified as important in learning and memory (Churchwell and Kesner, 2011; Churchwell et al., 2010), and dendritic spines had been postulated to play an important role in the process (Jedlicka et al., 2008; Leuner et al., 2003). Thus, many studies investigated possible relationship(s) between estrogen dependent enhancements in learning and memory and increases in hippocampal and mPFC spine density (Frankfurt and Luine, 2015; Luine, 2016, 2020; Luine and Frankfurt, 2013; Luine and Frankfurt, 2012). Some of these studies are summarized in Table 1 which shows experiments where either recognition memory or spine density, or both, was measured following estradiol treatment or in physiological states where gonadal hormone levels are different.

That estradiol or specific estrogen receptor agonists influence spine density in the mPFC and hippocampus has been repeatedly demonstrated (Frankfurt and Luine, 2015). In gonadally-intact female rats, dendritic spine density is greatest when estrogen levels peak during the 4-5-day estrous cycle (Gonzalez-Burgos et al., 2005; Kinsley et al., 2006; Woolley et al., 1990). Ovariectomy decreases dendritic spine density in CA1 and mPFC compared to gonadally-intact rats (Wallace et al., 2006). Spine stability also appears to depend on estradiol synthesized locally within the hippocampus as well as from the ovaries. Inhibition of aromatase, the enzyme which converts testosterone estrogens, also decreases hippocampal spine density (Zhou et al., 2010). The relative importance of central vs peripheral estrogen synthesis to spine maintenance needs further investigation. Estradiol administration to OVX rats restores the levels of dendritic spines in CA1 (Gould et al., 1990; Luine and Frankfurt, 2013), as well as spine synapses in the hippocampi of OVX monkeys (Leranth et al., 2002) and rats (Woolley and McEwen, 1992). Indeed, estradiol has been shown to increase both dendritic spine and spine synapse density in CA1 and the mPFC of non-human primates (Hajszan and Leranth, 2010; Leranth et al., 2008; Tang et al., 2004). Other examples of chronic effects of estrogens on spine density include those seen during and after pregnancy (Kingsley et al, 2006) and following ingestion of various plant estrogens (Luine et al., 2006). Lastly, Bisphenol A (BPA), an endocrine disruptor acting through estrogen receptors, decreases dendritic spine density in the mPFC and CA1 following chronic administration to adolescent and adult rats of both sexes (Bowman et al., 2015; Bowman et al., 2014).

As reviewed in Luine et al. (2018), acute administration of estrogens or agonists has been shown to increase dendritic spine density in PFC and CA1 within 30 – 40 min (Inagaki et al., 2012), Table 1, and within 4 h in other estrogen sensitive brain areas like the ventromedial nucleus (Frankfurt and Luine, 2015) and arcuate nucleus (Christensen et al., 2011). Although some differences in effects have been reported, agonists for both ERα and ERβ alter spine density (Murakami et al., 2006; Phan et al., 2011). In the PFC, agonists of the G-protein-linked estrogen receptor (GPER), but not ERα/β receptors, increased spine density, and the opposite selectivity was found in CA1 (Ye et al., 2019). Overall, it is accepted that estradiol promotes increased dendritic spine plasticity in brain areas implicated in memory; however, which receptor(s) mediates the changes and whether differences in receptor mediation exist among brain areas remains to be resolved. In addition, it is not clear whether estrogens cause maturation of existing immature, filopodia spines (Avila et al., 2017) or the addition of new, immature spines, or both (see Frankfurt and Luine, 2015 for further discussion).

4. Is increased spine density by estradiol necessary and sufficient for enhancing recognition memory?

Whether estrogen-dependent increases in spine density in the hippocampus and mPFC are causal for estrogen-dependent enhancements in memory consolidation or simply reflect different aspects of estrogenic regulation of neural function has not been definitively established. Table 1 summarizes experiments where enhancements in memory and/or increased spine density in mPFC and hippocampus, or both parameters, were found following estradiol treatment. The pattern of changes suggests causality between spines and memory function. However, the timing of measurements is often not the same. For example, increased CA1 spines are present 30 min after systemic estradiol injection, whereas increased OP was assessed 2 h (Jacome et al., 2016) or 4 h later (Inagaki et al., 2012). However, Phan et al (Phan et al., 2011; Phan et al., 2015) found simultaneous enhancement of OP and increases in CA1 spines 40 min after injection of estradiol or an ERα agonist, providing evidence that estradiol-induced effects on spines and cognition are temporally aligned and thus, increased spinogenesis may contribute to increased memory consolidation.

Neurochemical investigations show that memory consolidation during recognition memory tasks requires activation of signaling pathways such as such as the extracellular signal-regulated kinase/mitogen activated protein kinase (ERK/MAPK) pathway, cyclic AMP response element binding protein (CREB), and activation of the mammalian target of rapamycin (mTOR) protein synthesis pathway (Fortress and Frick, 2014; Fortress et al., 2014; Fortress et al., 2013). Because inhibition of these pathways following bilateral infusion of estradiol into the hippocampus impairs recognition memory consolidation (Fortress and Frick, 2014; Fortress et al., 2014; Fortress et al., 2013), we tested whether such treatments would also block increases in dendritic spines (Tuscher et al., 2016). Within 30 min of estradiol infusion into the hippocampus, dendritic spine density on pyramidal cells in CA1 increased and was maintained at 2 h. Hippocampal injection of mTOR or ERK inhibitors immediately prior to estradiol infusion blocked increases in CA1 dendritic spine density. This result, coupled with previous behavioral studies, is evidence that estradiol-dependent increases in spines are necessary for the memory-enhancing effects of estradiol.

5. Do estrogenic effects on afferent projections from monoaminergic systems alter spine density and cognition?

Pyramidal neurons in CA1 and layer II/III of the mPFC, which respond to estrogens with increased spine density, are glutamatergic, and estrogens cause changes in expression of several markers in these neurons including glutamate receptor expression (Khan et al., 2013; Phan et al., 2015), as well as increases in glutamatergic synapses which are abundant and known to contribute to memory (Khan et al., 2013). However, it is unclear whether estrogens act directly on ERs present in pyramidal neurons or on afferent endings that synapse on these neurons. We investigated a possible role for monoaminergic afferents in contributing to estrogen-dependent changes since both areas receive extensive monaminergic innervation from nuclei in the brainstem, and there is also evidence for monoaminergic dysfunction in Alzheimer’s Disease (Simic et al., 2017). Table 2 summarizes changes in monoaminergic activities in these areas following acute (30 min), sub-chronic (2 days) and chronic (28 days) estrogenic treatments which enhance memory in OVX subjects. In the PFC, activity of norepinephrine (NE) and dopamine (DA) increased after all treatments. Acute effects (30 min post estradiol), were greatest, approximately 85% increases, while longer treatments, 2 or 28 days, were associated with smaller increases, 26 to 75%. In contrast, serotonin (5-HT) activity decreased by 27% 30 min following estradiol, but sub-chronic or chronic treatments increased activity by approximately 30%. Thus, changes in monoaminergic terminals on mPFC pyramidal neurons may underlie estrogen actions in this area. A different pattern was found in CA1 following estrogenic treatments. Neither DA or 5-FIT activity was altered following estrogenic treatments, but NE activity was decreased by approximately 30%, following acute and sub-chronic, but not chronic treatment. Thus, it is possible that actions of estradiol on neural plasticity required for cognitive enhancement may also involve the monoaminergic innervation of pyramidal cells in both CA1 and the mPFC. In support of this idea, extensive interactions between the hippocampus, prefrontal cortex and ascending monoaminergic pathways in mediating cognitive function have been recently reviewed (Bueno-Junior and Leite, 2018).

Table 2:

Effects of estradiol and an ERβ agonist on monoaminergic activity

OVX rats received 20 μg/kg of estradiol immediately following a recognition memory training trial, were sacrificed 30 min later and monoamines (Inagaki et al, 2010) measured. OVX rats received 2 days treatment with the ERβ agonist DPN and sacrificed 48 h following the last injection; this treatment and time enhances recognition memory (Jacome et al, 2010). Twenty-eight days following estradiol Silastic capsule implantation and one day after radial arm maze testing, subjects were sacrificed (Luine et al, 1998). Changes in activity are % change from vehicle and represent either ratio of metabolite to monoamine or metabolite level. ND, non-detectable.

6. Concluding Remarks

In the last fifty years, knowledge of estradiol’s role in regulating behaviors has evolved from simply controlling estrus/menstrual cycles and concomitant sexual behavior to exerting wide ranging influences on higher order neural functions including mood, anxiety, neurological disorders and learning/memory. Our focus has been on estrogen’s promotion of cognitive function, and we provided information and validation for this view. While behavioral neuroendocrinologists have provided fundamental information on this topic, future research needs to address precisely how spine plasticity and signaling pathways mediate improved memory. Interactions among areas important for memory also need to be further investigated since we recently found that infusion of estradiol into the hippocampus increased spine density not only in CA1 after 30 min, but also in the mPFC after 2 h (Tuscher et al., 2016). Increases in mPFC spine density are consistent with White and McDonald’s concept of parallel information processing (White and McDonald, 2002), and suggest that estrogens may orchestrate such effects and contribute to cognition by turning on specific neural systems which input to and are activated by the hippocampus. In addition, possible functions for acute changes in cognition by estrogens need to be determined. Overall, this area of research has seen notable advances and constant evolution over the last fifty years.

Highlights.

This review highlights fifty years of progress in research on estradiol’s role in regulating behavior(s). We provide a brief overview of estradiol’s regulation of memory and some mechanisms which underlie its effects. We also identify questions that need answering in the next 50 years.