Sex Hormones, Neurosteroids, and Glutamatergic Neurotransmission: A Review of the Literature Free

Dysregulation of glutamatergic neurotransmission has been implicated in many neurological and psychiatric disorders, including epilepsy [1], chronic pain [2], post-traumatic stress disorder (PTSD) [3, 4], and premenstrual dysphoric disorder (PMDD) [5]. Glutamate is the most common excitatory neurotransmitter in mammalian nervous systems and, as such, glutamatergic neurotransmission can influence wide-ranging brain functions [6]. Emerging research has demonstrated the importance of sex hormones in the regulation of glutamatergic neurotransmission. Thus, it is important to understand how estrogens, progesterone, and testosterone, as well as neurosteroids, may influence normal neurological function, as well as their possible role in neurological conditions. It is widely hypothesized that hormones can affect neurotransmission in a way that is regulatory and protective. For example, a previous review summarized evidence that 17β-estradiol (the major estrogen released by the premenopausal ovary) assists in glutamate reuptake. Estradiol may also interact with metabotropic glutamate receptors in a manner that influences neurological signaling activity, as seen in the figure below (Fig. 1) [7, 8].

As abnormal glutamatergic neurotransmission has been associated with many neurological conditions, it is important to understand how each sex hormone may influence this important aspect of nervous system dysfunction. Thus, this review aims to summarize current knowledge on the influence of estradiol, progesterone, and testosterone on glutamatergic neurotransmission, as well as potential effects from neurosteroids. A few examples of neurological conditions where sex hormone interactions have been implicated will also be reviewed to inform potential future research directions.

Research articles were identified via scholarly databases including PubMed, Google Scholar, and ProQuest. Articles were included if they were original research from peer-reviewed academic journals that dealt with glutamate, estrogen, progesterone, testosterone, neurosteroids, glutamate and sex hormone interactions, or the potential impact of glutamate and sex hormone interactions. Epilepsy, chronic pain, PTSD, and PMDD will be evaluated with importance given to the mechanistic underpinnings for interaction between glutamate and the sex hormones.

Estrogens may be effective in protecting neurons from excitotoxicity caused by glutamate [9, 10]. Estradiol (commonly abbreviated as E2) is the predominant estrogen in terms of activity, and thus, has been the primary focus of research. Estradiol can increase the uptake of glutamate into astrocytes (clearing it from the synaptic cleft), thereby helping to prevent excitotoxicity [11]. It is also thought to activate signaling from metabotropic glutamate receptors through the stimulation of estrogen receptors, showing a potential interaction at the receptor level [12].

Immunocytochemistry has been used to confirm the co-expression of AMPA, GluR1/2/3, and androgen or estrogen receptors, in the septum, amygdala, and hypothalamus of rats. Male and female rats were hormonally manipulated, and glutamate receptor expression was assessed. Estradiol and testosterone both had a stimulatory effect on AMPA receptor expression in the hypothalamus, and estradiol treatment also increased the presence of glutamate receptors in the hypothalamus, a change that was higher in females than males. Both co-localization of glutamate with androgen receptors, and modulation of glutamate receptors in certain brain regions following treatment with estradiol, were identified [13].

Estrogens are also potentially neuroprotective against excitotoxicity induced by ischemia. Long-term administration of estradiol in rats has been shown to protect visual and spatial memory affected by global ischemia, and acute estradiol administration also protected visual memory and neuronal survival after induced global ischemia [14]. Estrogen receptor α (ERα) and β specific agonists were also shown to potentially protect hippocampal CA1 neurons during ischemia [15]. A direct interaction between ERα and metabotropic glutamate receptor 1a has also been observed in rats that underwent hormonal treatment, providing further evidence of estrogen receptor activity in hormonal and glutamatergic interaction [16]. Treatment with estradiol may also increase the internalization of both metabotropic glutamate receptor 1 and ERα [17]. These data are in line with the existing hypothesis of an estrogen receptor/glutamate receptor signaling unit, indicating the involvement of specific receptors in the hormonal/glutamatergic interaction [17].

Excitatory amino acid transporters (EAATs) may also be affected by sex hormones. The expression of GLT-1 and EAAT3, two transporters that assist in glutamate uptake, is upregulated following ischemia (where glutamate levels are known to be elevated), and mRNA and protein levels of these transporters are further increased by progesterone and estrogen as compared to ischemia alone [18]. Increased glutamate transporter expression is protective against excitotoxicity, as it helps remove excess glutamate from the synaptic cleft. Thus, this may be a direct reason for the protective effect of these sex hormones. Estrogens may also be protective against glutamate-induced oxidative stress, as glutamate excitotoxicity is known to cause oxidative stress via the production of reactive oxygen species and the downregulation of glutathione production [19]. Co-treatment with E2, when administering a glutamate injection, effectively reduced oxidative stress when compared to the glutamate injection alone [20].

Several independent laboratories have documented the effects of endogenous estrogen on brain injury, seizure activity, and memory performance. Neural estrogen levels seem to promote seizure activity in rats. More specifically, hippocampal levels of estrogens are higher in rats following kainic acid-evoked seizure activity, and treatment with an aromatase (estrogen synthase) inhibitor decreased kainic acid-evoked seizure activity [21]. Neural estrogen synthesis has also been shown to support memory function in rats and songbirds. Estrogen delivered directly to the hippocampus increases spatial memory performance in rats [22], and inhibition of estrogen synthesis impairs spatial memory function in zebra finches [23, 24]. Taken together, studies on multiple species suggest a strong interaction between estrogen and glutamate-dependent neural functions in the brain.

Progesterone is a primary sex hormone, but as demonstrated in some of the research discussed above, it is often investigated alongside estrogen rather than individually. Because of this, literature solely evaluating the effects of progesterone is lacking. Despite this difference in research quantity, some studies have provided evidence that progesterone should be an item of consideration when examining sex hormones and glutamatergic neurotransmission. For example, one in vitro study using cerebral cortex slices found progesterone to potentially have neuroprotective effects that protected against glutamate-induced toxicity via activation of the neuroprotective mitogen-activated protein kinase and phosphoinositide-3 kinase pathways. Furthermore, progesterone also increased brain-derived neurotrophic factor levels, which have been shown to have neuroprotective properties [25]. Similarly, pretreatment with progesterone reportedly reduced glutamate-induced elevations in intracellular calcium levels in hippocampal slices [26]. Progesterone may also potentially have the ability to enhance the inhibitory response of Purkinje cells to GABA (the main inhibitory neurotransmitter which counters glutamate excitation) while also suppressing excitation from glutamate to reduce excitotoxicity in the brain [27]. Thus, progesterone is thought to have neuroprotective action, though more research using animal models is needed to better understand the effects of progesterone in vivo. It is important to note here that metabolites of progesterone also have potential effects on neurotransmission; these will be discussed in the neurosteroid section below.

In addition to estrogen and progesterone, it is important to also consider the modulating effects of testosterone on glutamatergic neurotransmission, as testosterone may be able to amplify excitotoxic damage to oligodendrocytes [28]. Oligodendrocyte cultures from rat forebrains were prepared and exposed to testosterone for 24 h, and then given a toxic pulse of AMPA with cyclothiazide or kainic acid for 15 min. Cultures were incubated for an additional 24 h and assessed for oligodendrocyte death through lactate dehydrogenase release. AMPA treatments with both cyclothiazide and kainic acid were found to cause oligodendrocyte damage, and testosterone application was shown to enhance AMPA and kainate toxicity, even inducing slight toxicity in cultures that were not exposed to these glutamatergic agents. Testosterone amplified the increase in Ca²⁺ triggered by the activation of AMPA/kainate receptors, thus giving oligodendrocytes a lower threshold for damage [28].

An increased susceptibility to excitotoxicity upon treatment with testosterone is in opposition to more recent animal data showing that testosterone may be protective against neurodegeneration. The protective effects of testosterone have been shown in an animal model of traumatic brain injury [29]. Mice were subjected to cortical impact and were then injected with testosterone or vehicle treatments. While the vehicle group showed Ca²⁺ driven mitochondrial damage, the testosterone-treated group showed significantly reduced damage [29]. It is important to note here that the latter findings may have been due to conversion of testosterone into estradiol via aromatase activity in the brain [30] which may be mediating this observed effect. It is possible that the aforementioned in vitro study may have prevented any conversion of testosterone to estrogen, thereby increasing the potential toxicity in cell culture. Overall, there are few animal studies examining the effects of testosterone, and the existing data are conflicting; thus, additional research is needed to better understand the interaction of testosterone with glutamatergic neurotransmission.

In addition to aromatase conversion of testosterone to estradiol in the brain mentioned above, there are also some neurosteroids which are likely an important part of hormonal interaction with glutamatergic neurotransmission. Neurosteroids, which are created in the brain through reactions with preexisting hormones, are vital to understanding the mechanistic action of progesterone and testosterone on a neuronal level [31]. Progesterone and testosterone are each able to react with 5α-reductase and 3α-hydroxysteroid oxidoreductase to produce either 5α-dihydroprogesterone and allopregnanolone, or 5α-dihydrotestosterone, and androstenediol, respectively, as shown in the figure (Fig. 2). There is significant evidence for an interaction between glutamatergic neurotransmission and several neurosteroidal metabolites of testosterone and progesterone.

Allopregnanolone appears to interact with glutamate release, but data are conflicting, with some research showing an inhibitory effect on release, and other research showing a role for allopregnanolone in the induction of glutamate release [32, 33]. Allopregnanolone action in the peripheral nervous system may also have a role in modifying glutamate uptake [34].

Androstenediol (which can bidirectionally interconvert with testosterone) is documented to allosterically modulate GABAergic activity, which is the main counterbalancing inhibitory neurotransmission from GABA, which is created from glutamate [35]. Dihydrotestosterone can act as an agonist for androgen receptors, causing an uptick in Ca²⁺ as compared to cells untreated with dihydrotestosterone [36].

Thus, neurosteroids may also play an independent role in modulating glutamatergic neurotransmission.

Catamenial epilepsy is characterized by seizure frequency changes at different points in the menstrual cycle, and general epilepsy also has a strong glutamatergic component [1, 37]. Data support an effect of both β-estradiol and glutamate simultaneously on epilepsy, and β-estradiol may be effective in upregulating GLT-1 transporters, downregulating glutamatergic neurotransmission, and increasing positive outcomes overall in rats with induced epilepsy [38]. In a study investigating astrocytes and β-estradiol in the epileptic condition, female rats were ovariectomized and injected with pilocarpine for seizure induction. Results were compared between the ovariectomy only group (control), a lithium-pilocarpine ovariectomy group with no estradiol treatment, a lithium-pilocarpine ovariectomy group with low estradiol dose, a lithium-pilocarpine ovariectomy group with a high dose of estradiol, and a sham group treated with lithium, diazepam, atropine sulfate, and sesame oil. Seizure duration and mortality were lowered with estradiol treatment, alongside other effects indicating potential neuroprotective properties. Notably, glutamate content in the hippocampus actually increased in the low estradiol group, in comparison to all the other groups, but decreased in the high estradiol group. Overall, results from this study suggest a potential protective effect of β-estradiol in epilepsy, alongside the potential ability for β-estradiol to regulate glutamate levels and excitability of neurons in epileptic conditions [38]. In hippocampal slices taken from rats, 4-aminopyridine was able to induce both seizure activity and an increase in levels of excitatory amino acids including glutamate, an effect increased by allopregnanolone [39].

Animal research on pain sensitivity and the influence of sex hormones on glutamatergic neurotransmission has also yielded notable results. The estrous cycle has been shown to influence EAAT and pain sensitivity in the context of early life stress, such that pain sensitivity is greatest in the proestrus phase [2]. The estrous cycle is the reproductive cycle experienced by female rats, a 4-to-5-day cycle that approximates the menstrual cycle in female humans [40]. During the proestrus phase, comparable to the follicular stage in human females, estrogen levels rise and then quickly decline in the estrus phase as follicle-stimulating hormone peaks [40]. Moloney et al. [2] identified glutamatergic neurotransmission and EAATs as important in the body’s pain processing mechanisms, while also being interactive with estrogen receptors. This study examined the effects of maternal separation (early life stress) and the estrous cycle on visceral pain sensitivity and EAATs. A group of female rat pups underwent maternal separation to emulate early life stress, and colorectal distension to induce pain. Estrous-cycle stage was determined, and animals were assessed for pain threshold and behaviors, while postmortem brain tissue EAATs were assessed as well. The early life stress group exhibited a lower pain threshold and increased pain behavior in comparison to the non-stressed group, with both groups demonstrating the lowest pain threshold and highest total pain behaviors in the proestrus phase. There was also a significant effect of estrous-cycle variation on pain behavior and threshold. Furthermore, estrous-cycle phase was shown to have a direct effect on EAAT function, with function being reduced in both the estrus and proestrus phases. The cycle effects in stressed and non-stressed rats were found to be completely opposite: EAAT function in stressed rats was inhibited in low estrogen cycle phases, while in non-stressed rats, EAAT function was inhibited in high estrogen states [2].

When comparing sex hormones, glutamatergic expression and pain, estradiol and testosterone were examined under the hypothesis that estradiol is pronociceptive (pain encouraging) and testosterone is antinociceptive (pain blocking) in a stress-induced rat hypersensitivity model, based on female predominance in pain conditions [41]. Adult rats of both sexes were subjected to a forced swim condition to induce stress and then underwent colorectal distension to induce pain. Groups of male rats were administered estradiol, while groups of female rats were administered testosterone. Results showed that stress-induced hypersensitivity lasted longer in females than it did in males, and ovariectomy blocked visceral hypersensitivity, while orchiectomy facilitated it. Estradiol injections in male rats were found to increase visceral hypersensitivity, while testosterone in female rats attenuated it. Estradiol injected into males was shown to increase excitatory glutamate ionotropic receptor NMDA subunit type 1 expression and decrease inhibitory metabotropic glutamate receptor 2 expression, after the stressful forced swim scenario. Thus, estradiol facilitates stress-induced visceral hypersensitivity, while testosterone attenuates it, potentially through modulation of glutamatergic neuronal activity in the spine [41].

Endometriosis is an example of an extremely painful condition which is strongly tied to hormonal regulation. Enhanced glutamatergic neurotransmission in the anterior insula, along with increased functional connectivity to the medial prefrontal cortex, has been demonstrated in people with endometriosis as compared to age-matched pain-free controls [42]. Other research has speculated a potential neuropathic contribution to endometriosis pain, with 40% of patients reporting pain classified as neuropathic in a recent survey [43]. To date, no research has directly explored the interaction of sex hormones and glutamatergic neurotransmission in endometriosis, so this is an area that may be of interest for future research.

PTSD is marked by an imbalance of excessive excitation (i.e., increased glutamate activity) and deficient inhibition (i.e., reduced GABA activity) [44]. Sex hormones have notable effects on both glutamate and GABA that can influence this imbalance. Estrogen can protect against glutamate excitotoxicity, specifically in the hippocampus and cortex [45], which are areas atrophied by excitotoxicity in PTSD [4, 46]. The progesterone metabolite, allopregnanolone, facilitates GABA activity [47], which is deficient in PTSD [44, 48, 49].

A negative relationship between salivary estradiol levels and the number of reported PTSD symptoms has been reported in a sample of trauma-exposed women, such that more estradiol was associated with less PTSD symptoms [50]. Exogenous estrogen administration in the form of birth control (synthetic estradiol) has been successfully utilized in order to increase the efficacy of extinction training (which has implications for exposure therapies) in women with PTSD [51]. Naturally occurring low estrogen levels seem to impair extinction learning among women with PTSD [52, 53]. This makes sense, as estrogen improves memory in humans and rodent models [54, 55], and thus would help in retaining newly formed extinction memories. Additionally, reduction in glutamatergic neurotransmission in the hippocampus via estrogen has implications for protecting against excitotoxicity-induced hippocampal atrophy that is found in PTSD cases [4]. Therefore, it may be that low estrogen is a liability for PTSD severity and recovery.

Women with PTSD appear to have lower conversion of 5α-dihydroprogesterone to allopregnanolone than controls, implicating reduced activity of the 3α-hydroxysteroid dehydrogenase enzyme [56]. Men with PTSD have reduced activity of 5α-reductase, and thus, make lower levels of allopregnanolone [57, 58]. Unsurprisingly, research has demonstrated reduced allopregnanolone in cerebral spinal fluid of individuals with PTSD of both sexes compared to controls [56‒58]. Given that allopregnanolone is GABAergic, these findings are consistent with the decreased inhibitory tone seen in PTSD. Additionally, GABA inhibits the hypothalamic-pituitary-adrenal axis (HPA-axis) stress response [59], so deficient GABA activity allows for a more reactive stress response, congruent with PTSD symptomatology.

Finally, testosterone appears to have an inhibitory effect on the HPA-axis [60], and individuals with PTSD have lower cerebral spinal fluid testosterone than controls [61].

In summary, neurohormonal patterns are evident in PTSD and these could facilitate excessive excitation and deficient inhibition, along with an overactive HPA-axis stress response.

PMDD is an important condition to consider when examining the interaction between sex hormones and glutamatergic neurotransmission. PMDD directly involves the menstrual cycle and is therefore reactive to hormonal fluctuations. Two studies have evaluated glutamatergic dysregulation in PMDD. One study evaluated glutamate fluctuations across the menstrual cycle of females with PMDD in comparison to control subjects by recruiting 12 PMDD patients, along with 13 healthy controls, with symptoms monitored for 2 cycles [62]. Magnetic resonance spectroscopy was used to measure various metabolites in the prefrontal cortex, including Glx (glutamate and glutamine combined). A phase effect was observed, in which both the PMDD and control groups had lower levels of glutamate/Cr in the luteal menstrual phase than they did in the follicular menstrual phase; however, no significant differences in Glx levels were noted across groups. Similar results were produced when ovulation was taken into account. The study was limited by its small sample size, but it does provide evidence that glutamatergic fluctuations do occur with monthly hormonal fluctuations in females [62]. Females with PMDD may be more reactive to the phase-related glutamatergic alterations, warranting further research.

Additional research has evaluated brain concentrations of glutamate and GABA in participants with PMDD compared to controls [5]. As mentioned above, GABA is the main inhibitory neurotransmitter in the CNS, and as such, plays an important role in balancing out the excitatory effects of glutamate, making it also of interest. Twenty-two PMDD patients and 22 healthy controls were recruited, and GABA and Glx levels were measured with magnetic resonance spectroscopy. In all PMDD subjects, an increase in symptom severity was reported in the luteal menstrual phase, in comparison to the follicular menstrual phase, and scans were performed in the late luteal phase. Results showed that GABA concentrations were lower in the PMDD participants than they were in the controls in the anterior cingulate cortex, medial prefrontal cortex, and left basal ganglia. Results also showed increased Glx levels in the anterior cingulate cortex and medial prefrontal cortex in PMDD females as compared to healthy controls [5]. This suggests GABA and Glx abnormalities in patients with PMDD, which supports the idea that glutamate dysregulation is involved in PMDD (including reduced conversion of excitatory glutamate into the inhibitory neurotransmitter GABA, which could indicate a vitamin B6 deficiency in this disorder) [63]. More work is needed to understand how sex hormones and glutamate/GABA balance may be affected in PMDD.

While this literature review is intended to characterize the effects of sex hormones on glutamatergic neurotransmission, it is important to briefly note a potential bidirectional effect where exposure to monosodium glutamate (MSG) has been shown to alter hormone levels. A study from 2014 divided female Sprague Dawley rats into three groups (no treatment, just MSG, or MSG and diltiazem, a calcium channel blocker used for high blood pressure) with the MSG administered orally [64]. After 2 weeks, the serum levels of estrogen and progesterone were assessed. In the group receiving MSG alone, there was an increase in serum estrogen and progesterone levels when compared to the group given only a laboratory diet. Diltiazem prevented this effect, potentially by preventing calcium overload in the cells [64].

Oral administration of MSG has also been shown to reduce serum estradiol in female rats [65] and reduce testosterone in male rats [66]. It is also worth noting that orally ingested MSG causes swelling of the adrenal glands in mice (sex not specified) [67]. This may have implications for PTSD, particularly for explaining its characteristic phenomenon of hypocortisolism [68‒70], which suggests cortisol release from the adrenals may be affected in the disorder. Given the fact that people with PTSD have been shown to have more glutamate in various brain regions, as well as in serum, as compared to controls [4, 44, 46, 71], the effects of MSG consumption may be of interest in future research.

These studies are notable for their establishment of a potential bidirectional effect present when examining the interaction between sex hormones and glutamate. This is another area which warrants further research.

Glutamatergic excitotoxicity and dysfunction have been implicated in a multitude of different conditions, making the modulation of glutamatergic neurotransmission of particular interest. Emerging research is demonstrating the importance of sex hormones and neurosteroids in modulating this system. Estrogens may be protective against excitotoxicity through modulation of glutamate receptors and transporters in the brain, but opposing effects at the spine may also be present in chronic pain conditions. Better understanding of this hormonal modulation may be vital in the pursuit of optimizing glutamatergic neurotransmission for conditions such as epilepsy, chronic pain, PTSD, and PMDD. More research is needed to expand our understanding of hormonal effects on neurotransmission in neurological and psychiatric disorders where dysregulation of glutamate is apparent.

All papers reviewed herein met the appropriate ethical standards for human subjects or animal research.

The authors have no conflicts of interest to declare.

This work was supported by NSF Grant IOS 2050260 (C. Saldanha).

M.J.G. and K.F.H. conceived of this research, reviewed articles, wrote the manuscript, and approved the final copy. S.L.M. contributed to the PTSD section of the manuscript and contributed to the final review and editing of the manuscript. C.J.S. contributed to the review of animal research, contributed to the writing, and approved the final manuscript.

Since this is a review article, there is no data available for sharing.