Sex differences and hormonal regulation of metabotropic glutamate receptor synaptic plasticity

Abstract

Striking sex differences exist in presentation and incidence of several psychiatric disorders. For example, major depressive disorder is more prevalent in women than men, and women who develop alcohol use disorder progress through drinking milestones more rapidly than men. With regards to psychiatric treatment responses, women respond more favorably to selective serotonin reuptake inhibitors than men, whereas men have better outcomes when prescribed tricyclic antidepressants. Despite such well-documented biases in incidence, presentation, and treatment response, sex as a biological variable has long been neglected in preclinical and clinical research. An emerging family of druggable targets for psychiatric diseases, metabotropic glutamate (mGlu) receptors are G-protein coupled receptors broadly distributed throughout the central nervous system. mGlu receptors confer diverse neuromodulatory actions of glutamate at the levels of synaptic plasticity, neuronal excitability, and gene transcription. In this chapter, we summarize the current preclinical and clinical evidence for sex differences in mGlu receptor function. We first highlight basal sex differences in mGlu receptor expression and function and proceed to describe how gonadal hormones, notably estradiol, regulate mGlu receptor signaling. We then describe sex-specific mechanisms by which mGlu receptors differentially modulate synaptic plasticity and behavior in basal states and models relevant for disease. Finally, we discuss human research findings and highlight areas in need of further research. Taken together, this review emphasizes how mGlu receptor function and expression can differ across sex. Gaining a more complete understanding of how sex differences in mGlu receptor function contribute to psychiatric diseases will be critical in the development of novel therapeutics that are effective in all individuals.

1. Introduction

Sex differences in the incidence, presentation, and treatment responses of several psychiatric and substance use disorders are well-known. Major depressive disorder (MDD), for instance, exhibits sex differences across several levels. Women are approximately twice as likely to have MDD than men, exhibit more severe symptoms, and often experience different symptomatology (Angst & Dobler-Mikola, 1984; Frank, Carpenter, & Kupfer, 1988; Kessler et al., 2005; Kornstein et al., 2000a; Najt, Fusar-Poli, & Brambilla, 2011; Silverstein, 1999; Young et al., 1990). Additionally, evidence suggests that women and men, on average, have superior treatment responses to different antidepressant medications. For instance, women have statistically superior responses to monoamine oxidase inhibitors and selective serotonin reuptake inhibitors than men; conversely, men respond more favorably to tricyclic antidepressants (Hamilton, Grant, & Jensvold, 1996; Kornstein et al., 2000b; Quitkin et al., 2002). In contrast, many reported sex differences in schizophrenia are weighted toward men. For example, men have an earlier age of onset of schizophrenia, with symptoms emerging in the early twenties in men, while symptom emergence is in the late twenties in women (Angermeyer, Goldstein, & Kuehn, 1989; Goldstein, Tsuang, & Faraone, 1989; Hafner et al., 1989; Hambrecht, Maurer, & Hafner, 1992; Sartorius et al., 1986). Men with schizophrenia also exhibit higher incidence of negative symptoms and poorer course compared to women with schizophrenia (Goldstein et al., 1998; Goldstein & Link, 1988; Huber et al., 1980; Salokangas, 1983). Similarly, sex-dependent vulnerabilities to specific disease symptoms have been observed in alcohol use disorders (AUD) and several concerning findings indicate that women are at greater risks for detrimental outcomes. While AUDs are more prevalent in men overall, women are more sensitive to peripheral diseases and cognitive disturbances that can stem from chronic alcohol consumption (Nixon, Tivis, & Parsons, 1995; Tuyns & Pequignot, 1984; Urbano-Marquez et al., 1995). Women who develop AUD also progress through disease milestones more rapidly than men (Diehl et al., 2007; Lewis, Hoffman, & Nixon, 2014; Randall et al., 1999), and women are disproportionately affected by consequences of acute intoxication (Gross & Billingham, 1998). Clearly, many psychiatric diseases display remarkable sex differences in presentation and etiology. With the goal of delivering optimal treatment outcomes to all, understanding the manifestation of these sex differences should be a top priority of biological psychiatry research.

Despite the well-documented sex differences in the epidemiology of psychiatric disorders, it is only relatively recently that researchers have begun to delve into potential biological underpinnings for these observed sex differences. A major driving force behind these recent studies is the National Institutes of Health mandate that sex as a biological variable (SABV) will be factored into research designs, analyses, and reporting in vertebrate animal and human studies. Prior to the mandate, many researchers argued that the circulating gonadal hormones associated with the estrous cycle would introduce too much variability into experimental measures; the result is that many previous basic and preclinical studies excluded female subjects and cells. Thus, much of what is known about biological mechanisms of psychiatric disorders is based on studies in males. The NIMH mandate to include SABV has resulted in an explosion of research documenting sex differences in cell signaling, synaptic plasticity, emotional behaviors, cognition, and social function. In this chapter, we describe and summarize the known sex differences in the functions of metabotropic glutamate (mGlu) receptors. Consistent with the broad and rich neuroendocrinology literature documenting acute effects of gonadal hormones on brain function, significant evidence suggests that sex hormones regulate a variety of mGlu receptor functions through proximal receptor-receptor interactions. However, much less is known about how other factors, including developmental hormone exposure and sex chromosome complement, might influence sex-specific behavioral and neurobiological outcomes. In addition to synthesizing key themes from the existing literature, we hope this chapter will be useful in identifying and highlighting the important gaps in our understanding of sex differences in mGlu receptor function.

2. Sex differences in mGlu receptor function

Metabotropic glutamate (mGlu) receptors are a class of type C G protein-coupled receptors widely distributed in the nervous system. Activation of mGlu receptors via extracellular binding of glutamate triggers intracellular signaling cascades that induce short- and long-term adaptations in synaptic physiology, neuron excitability, and gene expression (Niswender & Conn, 2010). The eight mGlu receptor subtypes are grouped into three families (Groups 1–3) based on sequence homology, cell signaling, and pharmacology (Conn & Pin, 1997).

2.1. Regional distribution and receptor expression

The Group I mGlu receptors, comprised of subtypes mGlu₁ and mGlu₅, are important regulators of excitatory glutamate transmission and synaptic plasticity. mGlu₁ and mGlu₅ receptors are primarily localized to postsynaptic compartments, where they signal through Gq proteins and associated downstream effectors (Hermans & Challiss, 2001; Niswender & Conn, 2010). Our current understanding of the regional distribution and localization of Group 1 receptors in the CNS is largely based on gene-cloning and immunoreactivity studies from preclinical models (Martin et al., 1992; Romano et al., 1995; Van den Pol, 1994; van den Pol, Romano, & Ghosh, 1995). mGlu₁ and mGlu₅ receptors have widespread, generally overlapping distributions throughout most of the brain, with a key exception being the cerebellum, where high levels of mGlu₁ (but not mGlu₅) are expressed in mature Purkinje cells (Baude et al., 1993; Casabona et al., 1997; Fazio et al., 2008). An important caveat to this literature is that most studies investigating the regional distribution of Group 1 receptors were restricted to male rodents or not statistically powered to test for interactions across sexes. As such, the potential for intrinsic sex differences in the regional distribution of mGlu_1/5 receptors remains largely unexplored.

Although major sex differences in the regional distribution of Group 1 receptors have not been reported (Wickens, Bangasser, & Briand, 2018), preclinical studies investigating the potential sex differences in levels of mGlu receptor protein expression have revealed conflicting results. In a Western blot analysis of mGlu receptor subunit protein expression, adult female rats exhibited higher levels of mGlu₅ expression in the prefrontal cortex (PFC) and hippocampus than males, with no sex differences in mGlu₁ receptor protein expression in the PFC or hippocampus (Wang et al., 2015). These findings are in direct contrast to results from a recent study published in 2017, in which female rats had significantly lower levels of mGlu₁ receptor protein expression in the hippocampus than males (Ning et al., 2018). Both of these studies used Sprague-Dawley rats, suggesting that differences in animal strain do not account for the divergent findings. Interestingly, however, these studies employed antibodies raised against different portions of the intracellular C-terminal tail; therefore, it is possible that changes in the proportion of splice variants or post-translational modifications might account for the discrepancies in the literature. Additional studies examining cell type-specific expression of mGlu₁ and mGlu₅, as well as the relative expression of common splice variants and modifications, appear duly warranted.

There are limited published accounts describing sex differences in Group 2 and Group 3 mGlu receptor expression and function. As such, little is known about how these receptor subtypes endogenously differ in male and female brains. Higher levels of mGlu_2/3 receptor expression (assessed with a non-specific antibody) were found in the hippocampus of wild-type female rats in comparison to males, and no sex differences were detected in PFC or amygdala (Wang et al., 2015). In addition, a study in human temporal cortex found higher expression of mGlu₃ receptors in women relative to men (García-Bea et al., 2016). To our knowledge, sex differences in baseline Group 2 receptor expression have not been thoroughly explored in other brain regions. Similarly, to our knowledge, potential sex differences in regional or synaptic localization of Group 3 receptors have not been thoroughly investigated.

2.2. Receptor trafficking

Group 1 receptors are localized to postsynaptic densities and anchored to nearby receptors through interactions with a multitude of scaffolding proteins, most notably including Homer, Shank, and PSD95 (Tu et al., 1999). There are three genes that comprise the Homer family proteins (Homer1–3) (Brakeman et al., 1997; Xiao et al., 1998) each of which has multiple alternative short and long form splice variants involved in the trafficking and signaling of Group 1 receptors (Kammermeier et al., 2000). Long form splice variants of Group 1 Homer proteins (Homer1b and Homer1c) differentially regulate mGlu_1/5 receptor trafficking on the post-synaptic membrane. In cultures from male rodents, Homer1b inhibits mGlu₅ receptor trafficking to the cell surface (Roche et al., 1999), whereas co-expression of Homer1c increases surface expression of mGlu₁ receptors (Ciruela et al., 2000; Xiao, Gustafsson, & Niu, 2006). To our knowledge, these findings have not been replicated in cultures from female animals. Nonetheless, there is little evidence to suggest basal sex differences in Homer protein regulation of Group 1 receptors, and neither Homer1 nor Homer2 knockout mice present with sex-dependent phenotypes in unconditioned behaviors (Szumlinski et al., 2004, 2005). Through their interactions with Homer proteins, the scaffolding proteins Shank and PSD-95 anchor Group 1 mGlu receptors within postsynaptic densities (Tu et al., 1999). Shank proteins are linked to NMDA receptors through PSD-95 associated complexes (Alagarsamy, Sorensen, & Conn, 2001; Xiao et al., 1998). In turn, Shank proteins enable mGlu_1/5 receptors to associate and couple with NMDA receptor complexes (Bertaso et al., 2010). Analysis of the baseline expression of PSD-95 in male and female rats showed no sex differences in protein levels in the hippocampus, amygdala, or PFC (Wang et al., 2015). Taken together, the existing cell signaling literature has not identified major sex differences in trafficking or scaffolding of mGlu receptors.

2.3. Signaling and synaptic plasticity

Throughout the forebrain, mGlu receptors are generally conceptualized as modulatory components that regulate fast neurotransmission through glutamate or GABA receptors. Canonical signaling downstream of Group 1 mGlu receptors proceeds through activation of Gq proteins, phospholipase C (PLC), Protein Kinase C (PKC), and the release of intracellular calcium (Conn & Pin, 1997; De Blasi et al., 2001; Valenti, Conn, & Marino, 2002). In addition to PKC activation, Group 1 receptors can activate a myriad of downstream effectors, including the mitogen-activated protein kinase/extracellular receptor kinase (MAPK/ERK) pathway (Mao & Wang, 2016). mGlu₁ and mGlu₅ receptors can also facilitate the production of endocannabinoids, such as 2-arachadonylglycerol and anandamide (Ohno-Shosaku et al., 2002; Varma et al., 2001), which can attenuate pre-synaptic release probability through CB1 receptors and also regulate post-synaptic plasticity through the transient receptor potential cation channel TRPV1 (Chávez, Chiu, & Castillo, 2010; Grueter, Brasnjo, & Malenka, 2010). A recent study found that, in the rat PFC, 10Hz electrical stimulation evoked mGlu₅ receptor-dependent LTD with distinct latent mechanisms across sexes (Bara et al., 2018). While LTD was mediated by mGlu₅ receptors and endocannabinoid signaling in both sexes, LTD was blocked by the CB1 antagonists in male rats but TRPV1 antagonists blocked LTD in female rats. These findings highlight the tremendous need to revisit basic mechanisms to probe for potential latent or overt differences in mGlu receptor synaptic plasticity.

In addition to their ability to regulate synaptic plasticity, in some nuclei and cell types, mGlu_1/5 receptor activation can elicit excitatory post-synaptic currents through more proximal actions on nonselective cation channels. For example, in several types of neurons within the cerebellum, mGlu_1/5 receptor activation can elicit slow excitatory postsynaptic currents by activating transient receptor potential TRPC3 channels (Hartmann, Henning, & Konnerth, 2011). Interestingly, there are sex differences in cerebellar nuclei firing rates, with cerebellar nuclei cells from female mice having larger mGlu_1/5 receptor currents than males, suggesting sex differences in mGlu_1/5 receptor function (Mercer et al., 2016). However, there were no differences in Group 1 receptor protein levels between male and female mice, but blockade of glutamate transport increased the magnitude of mGlu_1/5-dependent currents in male but not female mice. These data suggest that basal sex differences in cerebellar nuclei mGlu_1/5 currents may not stem from receptor availability but are instead related to the access of glutamate to Group 1 receptors.

Group 2 and 3 mGlu receptors are generally linked to Gi/o proteins, which inhibit adenylyl cyclase to decrease cAMP production. Group 2 and 3 receptors can also initiate a variety of downstream signaling pathways through the liberation of Gβγ subunits, notably including the inhibition of calcium channels and facilitation of potassium channels. Similar to Group 1 receptors, Group 2 and 3 mGlu receptors can also modulate kinase signaling, such as MAPK and phosphatidylinositol 3-kinase (PI3K) pathways. Despite similarities in sequence homology and pharmacology, mGlu₂ and mGlu₃ receptors regulate neurotransmission through distinct mechanisms of action. While both Group 2 receptor subtypes act as presynaptic autoreceptors to reduce glutamate release probability, mGlu₃ receptors are also heavily expressed at postsynaptic sites, where they can induce LTD through a postsynaptic signaling cascade involving mGlu₅ receptors, Homer proteins, PI3K pathway signaling, and the internalization of AMPA receptors. Thus, any of the experience-dependent sex differences in mGlu₅ receptor and/or Homer signaling (described in later sections), might conceivably have ramifications for mGlu₃ receptor function as well.

2.4. Sex hormone regulation

Cellular pools of estrogen and other steroid hormones bind to both nuclear and membrane localized receptors, with varying affinity and response (Fuentes & Silveyra, 2019; Levin, 2009; Watson, Jeng, & Kochukov, 2008). In regions of the hippocampus, PFC, dorsal striatum, and nucleus accumbens, estrogen binds to membrane-localized receptors to activate second messenger systems and affect synaptic function (Milner et al., 2001; Razandi et al., 2004). Notably, several central effects of estradiol (E2) are linked to mGlu_1/5 receptors. E2 alters neuronal excitability and synaptic plasticity through stimulation of Group 1 receptor machinery in both a sex-dependent and independent manner (McEwen et al., 2001; Tonn Eisinger et al., 2018). Membrane-localized estrogen receptors, estrogen receptor α (ERα) and ERβ, are functionally coupled to mGlu_1/5 receptors expressed on the cell surface. Binding of E2 to these ER/mGlu complexes activates G proteins to trigger a glutamate-independent intracellular signaling cascade (Meitzen & Mermelstein, 2011). It is important to note that ER/mGlu complexes are present in both male and female brains, without major sex differences in overall expression. For example, immunocyto-chemistry studies targeting ERs in rat hippocampus (Weiland et al., 1997) and mouse amygdala (Bender et al., 2017) found no sex differences in ERα or ERβ expression levels. A more recent study separating mouse hippocampal ERs into nuclear and membrane populations detected greater extranuclear ERα in tissue from female mice, but no sex differences in ERβ expression (Mitterling et al., 2010). Moreover, ERα expression was highest in the CA1 region of mouse hippocampus during diestrus, while expression of ERβ in CA1 pyramidal cell layers was relatively high across the estrous cycle. We are not aware of any published findings describing Group 1 receptor expression throughout the estrous cycle.

Sex differences in ER/mGlu receptor localization and have been identified in the hippocampus. Co-immunoprecipitation studies indicate that ERα colocalizes with mGlu₁ and mGlu₅ receptors in both males and females (Boulware, Heisler, & Frick, 2013; Tabatadze et al., 2015). Interestingly, E2 treatment selectively increases the co-localization of ERα with mGlu₁ receptors in females, without affecting co-localization with mGlu₁ receptors in males or with mGlu₅ receptors in either sex. Other studies report sex differences in expression of proteins associated with the organization and signaling of ERα/mGlur1 complexes. Transcripts for caveolin 1 and the palmitoylacyltransferase DHHC-7, were expressed at lower levels in the hippocampus of adult female rats compared to males (Meitzen et al., 2019), suggesting that posttranslational modifications may also guide sex differences in ERα/mGlu receptor co-localization. Interestingly, no sex differences in gene expression were detected in hippocampal samples from neonates, suggesting a possible developmental sex-dependent regulation of ERα/mGlu receptor complexes. Taken together, ER signaling promotes the surface expression of mGlu_1/5 receptors to a greater extent in female individuals. Some evidence suggests the converse relationship holds as well (i.e., that mGlu receptors promote the expression of ERα). In ventromedial hypothalamus, genetic deletion of mGlu₅ receptors decreased ERα expression in female mice but increased ERα expression in males (Fagan et al., 2020). In this study, mGlu₅ receptors did not appear to regulate ERβ expression and mGlu₁ receptor function was not addressed. Taken together, the bidirectional relationship between ERα and mGlu_1/5 receptors raises the possibility that sex differences in downstream signaling may govern mGlu_1/5 receptor localization and function.

There are both sex and region differences in the pairing of ER/mGlu receptor subtype coupling and their downstream effects. E2 facilitates IP3 production in hippocampal slices from female rats to a much greater extent than in male rats (Tabatadze et al., 2015). E2-driven IP3 production was abolished by JNJ16259685, indicating an mGlu₁ receptor-dependent mechanism. Importantly, no difference in IP3 production was observed during pharmacological activation of mGlu₁ with the Group 1 agonist DHPG (Tabatadze et al., 2015), suggesting that the sex differences in signaling is manifested upstream of mGlu receptors. Other studies have yielded similar results and identified additional signaling nodes. In hippocampal cultures from female rats, E2 facilitates CREB phosphorylation through a pathway downstream of mGlu₁ receptors, PLC, PKC, IP3 receptors, and MAPK (Boulware et al., 2005). In these studies, an mGlu₅ negative allosteric modulator (NAM) had no effect, again indicating specific coupling between ERα and mGlu₁ receptors. Furthermore, ERα/mGlu₁ receptor-dependent CREB phosphorylation was specific to female hippocampal cultures, whereas similar E2 treatment in male cultures had no effect on CREB phosphorylation (Boulware et al., 2005). Together, these studies indicate that E2 signals through mGlu₁ receptor and canonical G_q signaling pathways in the female hippocampus. By contrast, E2 activates mGlu₅ receptor signaling in striatum. In female striatal cultures, the rapid actions of E2 lead to MAPK-dependent CREB phosphorylation through the activation of ERα/mGlu₅ receptor signaling complexes (Grove-Strawser, Boulware, & Mermelstein, 2010). Similar to the studies in hippocampus cultures, male striatal cultures did not respond to E2 treatment. Thus, while ER-mGlu_1/5 receptor interactions are potentiated in the brains of female individuals, the mGlu receptor subtype involved varies based on region.

Functional studies also support the conclusion that there are sex differences in ERα/mGlu_1/5 receptor regulation of neuronal physiology and synaptic plasticity. In ventromedial hypothalamus, genetic deletion of mGlu₅ receptors decreases current-evoked spike-firing in female but not male mice (Fagan et al., 2020). These actions were linked with E2 signaling, as the decreased excitability in mGlu₅ receptor knockouts was occluded by ovariectomy in control mice and restored by E2. Neurons within cerebellar nuclei also display larger mGlu_1/5 slow excitatory currents in female mice relative to males (Mercer et al., 2016). In addition to basic physiology and excitability, a growing literature now indicates that gonadal hormones regulate mGlu_1/5 receptor-dependent synaptic plasticity. While E2 is synthesized in the hippocampus of both males and females, the hormone exerts sex-specific effects on synaptic plasticity in CA1. In female hippocampal pyramidal cells, E2 suppresses inhibitory synapses through activation of post-synaptic ERα/mGlu receptor complexes, retrograde signaling through anandamide and CB1 receptors, and the attenuation of GABA release probability (Huang & Woolley, 2012). E2 also potentiates excitatory transmission in hippocampus through distinct mechanisms of action in females and males (Oberlander & Woolley, 2016), but it remains unclear whether mGlu receptor signaling is involved in these phenomena.

Naturally occurring fluctuations in apical dendritic spine density of CA1 hippocampal pyramidal cells are mediated, in part, by fluctuating E2 levels across the estrous cycle in female rats (Woolley et al., 1990; Woolley & McEwen, 1993). Studies in nucleus accumbens suggest that these endogenous cycles are regulated by mGlu₁ and mGlu₅. Consistent with the results from functional studies, E2 and other gonadal hormones alter dendritic spine plasticity in medium spiny neurons of the nucleus accumbens through the activation of Group 1 receptors (Gross et al., 2016, 2018). Though mGlu_1/5 receptors are structurally similar and tend to signal through conserved downstream pathways, evidence suggests that mGlu₁ and mGlu₅ receptors regulate spine density through distinct mechanisms. For example, systemic administration of selective mGlu¹ and mGlu₅ positive allosteric modulators increase and decrease spine density, respectively, in the nucleus accumbens of female rats (Gross et al., 2016). In addition, this structural plasticity is linked to gonadal hormones. For instance, E2 decreases spine density in nucleus accumbens core medium spiny neurons, but not dorsal striatum, and this effect was blocked by MPEP (Peterson, Mermelstein, & Meisel, 2015). Similar to E2, the male hormone dihydrotestosterone (DHT) can also influence dendritic plasticity through Group 1 activation. In castrated males, DHT administration decreased dendritic spine density on MSNs in the nucleus accumbens shell, but not core, of castrated male rats and this effect was blocked by pretreatment with MPEP (Gross et al., 2018). Thus, similar to their dissociable roles in mediating gonadal hormone effects on neuronal physiology, mGlu₁, and mGlu₅ receptors differentially regulate hormonal effects on structural plasticity in a region-specific manner.

Group 2 receptor signaling and female sex hormone signaling converge in the regulation of ion channels and activity-dependent gene transcription. Initial studies independently show that Group 2 receptor agonists (Chavis et al., 1994) and E2 each lead to the inhibition of L-type calcium channel currents through either ERα or ERβ (Chaban et al., 2003; Mermelstein, Becker, & Surmeier, 1996). LY341495 (at concentrations selective for mGlu_2/3 receptors) blocks the ability of E2 to attenuate calcium channel currents and CREB phosphorylation (Boulware et al., 2005). Though the mGlu_2/3 agonists suppressed CREB phosphorylation similarly in hippocampal cultures from male rats, E2’s effects were specific to cultures from female rats. Additional studies have begun to tease apart the relative contribution of mGlu₂ vs mGlu₃ receptors. In striatal cultures from female rats, studies using subtype-specific interfering RNA revealed that knockdown of mGlu₃ receptors, but not mGlu₂ receptors, prevented E2 from attenuating CREB phosphorylation (Grove-Strawser et al., 2010). In addition, neither mGlu₂ nor mGlu₃ receptor expression appears to be altered by estrus cycle. Comparable receptor levels were detected in the nucleus accumbens when cocaine/vehicle self-administration rats were killed in the estrus vs. non-estrus phases (Logan et al., 2020). It is important to note, however, that these findings may or may not generalize to other brain areas.

3. Insight gained from behavioral pharmacology, transgenic mice, and disease models

Pharmacological and genetic manipulation of mGlu receptor subtypes in preclinical research have revealed further insights into their biological functions. Over the past two decades, numerous studies have used highly selective pharmacological ligands and transgenic animals to characterize the function of mGlu receptor functions in male preclinical models (each of which have been extensively detailed in (Niswender & Conn, 2010)). In stark comparison, only a handful of studies have addressed mGlu_1/5 receptor function in females, and even fewer studies have made direct comparisons to probe for potential sex differences. Nonetheless, the results of those few studies are consistent with significant sex differences in mGlu receptor functions in basal and altered states.

3.1. Cognition

mGlu₁ (Gil-Sanz et al., 2008) and mGlu₅ (Simonyi, Schachtman, & Christoffersen, 2005; Xu et al., 2009) receptors have been widely implicated in associative learning based on studies performed in male rodents (Luscher & Huber, 2010). However, it remains unclear whether similar contributions of mGlu₁ and mGlu₅ receptors guide cognition in female individuals. In a study using mice with selective deletion of mGlu₅ receptors from parvalbumin-expressing interneurons, there was a sex difference in perseverative behavior on the Barnes maze (Barnes et al., 2015). Male mice lacking mGlu₅ receptors in parvalbumin cells took longer to escape the Barnes maze and displayed more preservative errors than controls. By contrast, while female knockout mice also performed worse than female controls, changes in perseveration were not closely tied to the cognitive disruption. In addition, a genotype by sex interaction was observed with respect to auditory event related potentials, suggesting that mGlu₅ receptor signaling in cortical parvalbumin interneurons may differentially regulate rhythmic activity and behavioral flexibility across sexes. Considering that other studies observed no differences in PFC-dependent cognitive behaviors in models where mGlu₅ receptor function on pyramidal cells was disrupted (Bara et al., 2018), these findings reinforce that mGlu receptor function can regulate behavior in a cell type-specific manner.

Studies investigating sex hormone regulation of learning and memory also provide clear indications that mGlu_1/5 receptor signaling may guide latent sex differences in the molecular mechanisms of cognition. For example, E2 delivery to the dorsal hippocampus of female mice enhances novel object recognition (Fernandez et al., 2008; Zhao et al., 2012; Zhao, Fan, & Frick, 2010), an effect blocked by co-administration of LY367385 or the MEK inhibitor U0126 (Boulware et al., 2013). Considering that the female hippocampus displays enhanced ERK-dependent ER/mGlu₁ receptor signaling (see earlier section), a compelling hypothesis for future research is that mGlu₁ receptors differentially contribute to hippocampal-dependent cognition across sexes. Similarly, some evidence suggests sex-differences in the contribution upregulation of hippocampal mGlu₅ receptors to cognitive processing. In a model of phenylketonuria, male but not female mice, displayed increased mGlu₅ receptor protein levels in the hippocampus that coincided with a deficit in object location recognition (Nardecchia et al., 2018). The increased mGlu₅ receptor expression occurred at the protein, but not transcript, level and was also associated with increased expression of long Homer isoforms. Additionally, administration of MPEP to mutant male mice significantly improved behavioral abnormalities. Taken together, the preclinical literature suggests that potential mGlu₁ and mGlu₅ modulators may display sex differences in pro-cognitive efficacy.

Genetic deletion of mGlu₈ receptors alters hippocampal-dependent learning and memory (Duvoisin et al., 2010). Though mGlu₈ receptor deletion impaired novel object recognition and spatial memory retention in the Morris water maze in both males and females, the effects on memory retention were more pronounced in females. Later studies from the same group found impaired contextual fear learning in female knockout mice relative to wildtypes, whereas males were not affected (Torres et al., 2018). Together, this literature raises the possibility that hippocampal circuits in female individuals are particularly sensitive to manipulations in mGlu₈ receptor function. Transgenic mouse studies have also identified that mGlu₇ receptors are involved in working memory (Hölscher et al., 2004), fear and extinction learning (Goddyn et al., 2008), anxiety-like behaviors (Cryan et al., 2003), and social behaviors (Fisher et al., 2020, 2021). In each of these studies, female mice and male mice were included, similar effects were observed across sexes, and the datasets were collapsed for subsequent analysis and presentation. Thus, there is minimal evidence to suggest notable sex differences in mGlu₇ receptor function regulate cognition or other behaviors.

3.2. Social behaviors

As mentioned previously, Group 1 receptors drive endocannabinoid signaling to regulate synaptic plasticity and behavior (Varma et al., 2001). Previous studies involving prenatal cannabinoid exposure have linked adaptations to PFC mGlu_1/5 receptor function and endocannabinoids to social behavior in a sex-dependent manner. WIN-55,212 administration during gestation causes decreases in adult social exploration and play behaviors in male but not female rats (Bara et al., 2018). There were no differences in the elevated plus maze or in an order recognition task following gestational cannabinoid exposure, suggesting changes were specific to social behaviors in male individuals. In addition, prenatal cannabinoid exposure eliminated endocannabinoid LTD induced by 10-Hz stimulation in layer 5 pyramidal neurons of the PFC in male but not female rats. Previous studies have found this 10-Hz LTD is mediated by mGlu₅ receptors (Luscher & Huber, 2010), and the mGlu₅ PAM CDPPB restored LTD in male rats exposure to prenatal cannabinoids (Bara et al., 2018). Finally, the authors linked the sex-specificity of these phenomena to the latent mechanisms mediating PFC LTD, providing evidence that LTD in male rats is mediated by CB1 receptors while TRPV1 receptors mediate LTD in female rats (Bara et al., 2018). Systemic treatment with either the mGlu₅ PAM or a fatty acid amide hydrolase inhibitor URB597 rescued behavioral deficits in affected male rats, providing a sex-dependent link between mGlu₅ receptor plasticity and social interaction.

Evidence linking mGlu₅ receptor function with social behaviors has also emerged from transgenic animal studies investigating Shank3. Deficits in Shank3, a scaffolding protein that coordinates mGlu₅/Homer intracellular signaling, are associated with the pathogenesis of autism spectrum disorder (Uchino & Waga, 2013; Verpelli et al., 2011; Zoicas & Kornhuber, 2019). Mutant Shank3 mice exhibit significant impairments in social behaviors and hippocampal mGlu₅ receptor signaling, without obvious sex differences (Matas et al., 2021; Peixoto et al., 2016; Wang et al., 2011). While social behaviors related to Shank3 disruption appear comparable in both sexes, Shank3 mutant mice also exhibit changes in repetitive behaviors and motor control, with evidence of sex differences in effects. Male Shank3 mutants displayed greater impairments in several gait parameters than littermate female mutants (Wang et al., 2011), and this phenotype has been linked with lower expression of mGlu₅ receptors in cerebellum in male mice (Matas et al., 2021). Together, these studies suggest that Shank3 regulation of mGlu₅ receptor function exhibits sex differences in specific brain areas (i.e., cerebellum but not hippocampus) and indicate that not all insults to mGlu₅ receptor signaling affect social play in a sex-dependent manner.

3.3. Alcohol use

Genetic and pharmacological manipulations of mGlu_1/5 receptors have revealed a clear association with alcohol seeking behaviors (Joffe et al., 2018). The link between ethanol exposure and mGlu_1/5 receptor function is bidirectional. Acute and chronic applications of ethanol alter mGlu_1/5 receptor function in cultured oocytes and rat cerebellar slices (Belmeguenai et al., 2008; Carta, Mameli, & Valenzuela, 2006; Minami et al., 1998; Netzeband & Gruol, 1995). On the other hand, Group 1 antagonists, such as the prototypical mGlu₅ NAM MPEP, reduce ethanol-seeking and relapse-like behaviors in a broad variety of models and background (Backstrom et al., 2004; Besheer et al., 2008; Hodge et al., 2006; Lominac et al., 2006). While most of these studies were limited to male subjects, a few have found comparable effects in females. In a recent study directly assessing sex differences in binge drinking, MPEP displayed a similar dose-response curve in reducing binge alcohol consumption in male and female mice (Huang, Thompson, & Taylor, 2021). Similarly, MTEP administration decreased binge ethanol intake in male and female mice (Cozzoli et al., 2014). Interestingly, in that study, female but not male mice displayed increased ethanol consumption 24 h after MTEP exposure, suggesting that females may be uniquely sensitive to compensatory or rebound adaptations following mGlu₅ inhibition.

Alterations in mGlu_1/5 receptor signaling in several brain areas have been implicated in binge drinking models. Several studies have implicated Homer2 in the regulation of ethanol sensitivity and ethanol-induced synaptic plasticity (Szumlinski et al., 2005, 2008; Szumlinski, Ary, & Lominac, 2008), including initial studies using mixed-sex cohorts. Homer2 knockout mice of both sexes resist drinking high concentrations of ethanol and do not express ethanol locomotor sensitization or conditioned place preference. By contrast, binge-drinking upregulates mGlu₅ receptors, Homer2, and PI3K signaling in male mice (Cozzoli et al., 2009). Furthermore, both male and female transgenic mice that express a point mutation in mGlu₅ receptors rendering them insensitive to ERK phosphorylation exhibit decreased sensitivity to the aversive properties of high dose ethanol (Campbell et al., 2019). Taken together, this literature indicates that upregulation of Homer2/mGlu₅ receptor expression and activation of ERK-dependent phosphorylation function as adaptive compensatory mechanisms to limit alcohol consumption in both sexes.

Women are more likely than men to drink alcohol to regulate negative emotional state or stress (Peltier et al., 2019). In recent years, the sexually dimorphic bed nucleus of the stria terminalis (BNST) has been identified as a point of convergence for early stress- and alcohol-related disorders in modulating negative affect behaviors (Kasten et al., 2021; Lebow & Chen, 2016). This observation from AUD has been modeled by inducing dependence to ethanol during adolescence in rodents (Crews et al., 2016). In adolescent female mice, acute withdrawal from chronic intermittent ethanol disrupted mGlu₅ receptor LTD in the BNST (Kasten et al., 2020). By contrast, male mice in acute withdrawal exhibited no differences in mGlu₅ receptor LTD, but instead displayed enhanced NMDA receptor functions (Carzoli et al., 2019). Interestingly, these sex-dependent changes in BNST plasticity mapped onto distinct behavioral phenotypes associated with withdrawal. Adolescent female mice showed enhanced anxiety/anhedonia-like behavior in a novelty-induced hypophagia task, whereas male mice in withdrawal displayed elevated freezing during fear conditioning paradigms (Kasten et al., 2020). Two pieces of convergent evidence link the adaptations to BNST mGlu₅ receptor plasticity with female-specific affective behaviors. First, systemic delivery of MTEP to female mice reversed withdrawal-induced increases in latency to each in the novelty-induced hypophagia assay (Kasten et al., 2020). Second, mGlu₅ receptor knockdown in the BNST reproduces anhedonia-like behavior in the novelty-induced hypophagia task in female but not male mice (Kasten et al., 2021). Consistent with the findings by Kasten et al. (2020) in mice, Chandler et al. also observed that adolescent chronic intermittent ethanol exposure enhanced contextual fear learning and spontaneous recovery of freezing behavior in male but not female rats (Chandler, Vaughan, & Gass, 2022). Interestingly, the mGlu₅ PAM CDPPB reversed the withdrawal-induced phenotypes in male rats, although it is not clear from these studies how female individuals would respond to enhanced mGlu₅ receptor function. Together, these findings indicate that mGlu₅ receptors influence behavioral phenotypes relevant to alcohol dependence in adolescence in both males and females, however there is a clear impact of sex on resulting behaviors (Table 1).

Table 1.

Sex differences in mGlu receptor regulation of affective behaviors.

Domain	Family	Subtype	Species/Strain	Sex Diff	Key findings	First author (year)
Anxiety-like and affective behavior	Group I	mGlu1/5	Rat, SD	Y	BLA infusion of the mGlu1/5 agonist DHPG increased punished drinking in VCT in OVX ♀. BLA infusion of the DHPG decreased punished drinking in VCT in ♂ (De Jesús-Burgos, Torres-Llenza, & Pérez-Acevedo, 2012)	De Jesús-Burgos (2016)
		mGlu5	Mouse, C57BL/6J, Grm5(AA/AA)	N	Mice with mGlu5 that cannot be phosphorylated by ERK displayed decreased float time in FST and increased open arm time in the EPM (Cozzoli et al., 2009)	Campbell (2019)
			Mouse, C57BL/6J, Grm5loxp	Y	mGlu5 KD in BNST prolonged latency to eat in NIH in ♀ but not ♂. mGlu5 KD in BNST decreased center time in open field in ♂ but not ♀ (Lebow & Chen, 2016)	Kasten (2021)
			Mouse, C57BL/6J, Grm5loxp, SST-Cre	N	SST-mGlu5-KO mice resisted stress-induced alterations in PFC inhibitory transmission and spontaneous alternation on the Y maze (Di Menna et al., 2018)	Joffe (2022)
			Mouse, C57BL/6J, Grm5loxp, PV-Cre	Y	♂ PV-mGlu5-KO mice displayed more perseverative errors on the Barnes maze. ♀ KO mice did not display more perseverative errors than controls (Luscher & Huber, 2010)	Barnes (2015)
	Group 2	mGlu3	Mouse, C57BL6/J x C57BL6/Ntac, Grm3loxp	N	mGlu3 KD in PFC increased open arm time in EZM and decreased immobility in TST/FST (Maksymetz & Joffe, 2021)	Joffe (2020)
	Group 3	mGlu4	Mouse, C57BL6/J, Grm4−/−	Y	♀ mGlu4 KO mice displayed decreased open arm/center time in EZM and open field. ♂ mGlu4 KO mice displayed increased open arm/center time in EZM and open field (Joffe et al., 2021)	Davis (2012)
		mGlu8	Mouse, C57BL6/J, Grm8−/−	Y	♀ mGlu8 KO mice displayed increased open arm/center time in EZM and open field, and increased startle response. ♂ mGlu8 KO mice displayed decreased open arm/center time in EZM and open field, and increased startle response. (Joffe et al., 2021)	Duvoisin (2010)
Alcohol Use	Group I	mGlu5	Mouse, C57BL/6J	Y	Acute withdrawal from AIE disrupts mGlu5 LTD in the BNST in ♀ but not ♂ (Crews et al., 2016)	Kasten (2020)
			Mouse, C57BL/6NCrl	N	mGlu5 NAM MPEP (30mg/kg) decreases binge drinking (Besheer et al., 2008)	Huang (2021)
			Mouse, C57BL/6J	Y	mGlu5 NAM MTEP (20 mg/kg) decreases binge drinking. One month later, increased drinking was observed in ♀ but not ♂ (Huang et al., 2021)	Cozzoli (2014)
			Mouse, C57BL/6J, Grm5(AA/AA)	N	Transgenic mice with mGlu5 that cannot be phosphorylated by ERK display increased drinking (Cozzoli et al., 2009)	Campbell (2019)

All studies in this table employed mixed-sex cohorts

AIE, adolescent intermittent ethanol vapor; BLA, basolateral amygdala; BNST, bed nucleus of the stria terminalis; EPM, elevated plus-maze; ERK, extracellular signal-regulated kinase; EZM, elevated zero-maze; FST, forced swim test; KD, knock-down; KO, knock-out; LTD, long-term depression; NAM, negative allosteric modulator; OVX, ovariectomized; PFC, prefrontal cortex; PV, parvalbumin; SD, Sprague Dawley; SST, somatostatin; TST, tail suspension test; VCT, Vogel conflict test.

3.4. Drug use

Compared to males, female rodents acquire drug self-administration much more quickly, rapidly progress through milestones of drug use, and exhibit greater levels of drug reinstatement, mediated in part by the estrous cycle (Anker & Carroll, 2011; Carroll et al., 2002; Doncheck et al., 2018; Feltenstein & See, 2007). A significant literature, albeit generally restricted to males, indicates that Group 1 mGlu receptors contribute to drug-seeking and related behaviors. In general, mGlu₁ antagonists and NAMs block some cocaine-seeking behaviors (Xie et al., 2010; Yu et al., 2013), whereas mGlu₁ PAMs may prevent relapse-like behaviors (Loweth et al., 2014). While these studies allude to the involvement mGlu₁ receptors in the adaptations to chronic drug use, there is still limited information on how sex may influence mGlu₁ activity in SUD. In comparison to mGlu₁, we have a better understanding how sex differences in mGlu₅ receptor signaling guide behaviors related to drug use. Some of this evidence comes from studies examining sex hormone contributions to drug-seeking. In female individuals, the subjective effects of smoked cocaine fluctuate over the course of the menstrual cycle (Evans, Haney, & Foltin, 2002). Similar findings have been observed in several rodent models and laboratories: female rodents show increased cocaine-induced hyperlocomotion, enhanced cocaine cue and reward learning, increased motivation to self-administer, increased incubation of craving, and estrous cycle-dependent cocaine reinforcement (Calipari et al., 2017; Jackson, Robinson, & Becker, 2006; Johnson et al., 2019; Roberts, Bennett, & Vickers, 1989). In addition, in female rodent ovariectomy studies, E2 treatment facilitates cocaine self-administration (Hu & Becker, 2003; Lynch et al., 2001), related effects on hyperlocomotion and self-administration were blocked by MPEP administration (Gross et al., 2016; Martinez et al., 2014). Taken together, this literature suggests that changes in drug-seeking and cue attending during female hormonal cycles may be dependent on surges in E2 and potentiated signaling through downstream mGlu₅ receptors (Table 2).

Table 2.

Sex hormone regulation of mGlu receptor synaptic plasticity.

Family	Subtype	Region	Species/Strain	Sexes used	Sex diff	Key findings	First author (Year)
Group I	mGlu1	HPC	Rata	Both	Y	E2 facilitates CREB phosphorylation through ERα/mGlu1, PLC, PKC, and IP3R in ♀ but not ♂ (Boulware et al., 2005)	Boulware (2005)
			Rat, SD	Both (OVX)	Y	E2 suppresses GABA release probability through ERα/mGlu1, anandamide, and CB1R in ♀ but not ♂ (Huang & Woolley, 2012)	Huang (2012)
			Rat, SD	Both (OVX/CAS)	Y	E2 suppression of GABA release probability coincides with greater IP3 production in ♀ than ♂. ERα/mGlu1 complexes in exist in ♂ but are only regulated by E2 in ♀ (Tabatadze et al., 2015)	Tabatadze (2015)
			Rat, SD	♀ (OVX)	n.a.	E2 increases dendritic spine density through activation of mGlu1 (Peterson et al., 2015)	Peterson (2015)
	mGlu5	Striatum	Rata	♀	n.a.	E2 facilitates CREB phosphorylation through ERα/mGlu5 and MAPK (Grove-Strawser et al., 2010)	Grove-Strawser (2010)
		NAC core	Rat, SD	♀ (ovx)	n.a.	E2 decreases dendritic spine density through activation of mGlu5 (Peterson et al., 2015)	Peterson (2015)
		NAC shell	Rat, SD	♂ (CAS)	n.a.	DHT decreases dendritic spine density in the NAC shell, but not core, through mGlu5 (Gross et al., 2018)	Gross (2018)
Group II	mGlu2/3	HPC	Rata	Both	Y	E2 decreases CREB phosphorylation through mGlu2 and/or mGlu3 inhibition of L-type calcium channels in ♀ but not ♂ (Boulware et al., 2005)	Boulware (2005)
	mGlu3	Striatum	Rata	♀	n.a.	E2 decreases CREB phosphorylation through mGlu3 inhibition of L-type calcium channels (Grove-Strawser et al., 2010)	Grove-Strawser (2010)

^aPyramidal neurons cultured from P1-P2 rat pups, experiments performed at 9 d.i.v.; strain not specified

CAS, castrated; CB1R, cannabinoid receptor type 1; CREB, cAMP response element-binding protein; DHT, dihydrotestosterone; GABA, gamma-aminobutyric acid; E2, estradiol; ER, estrogen receptor; HPC, hippocampus; IP3R, inositol triphosphate receptor; MAPK, mitogen-activated protein kinase; n.a., not addressed; NAC, nucleus accumbens; OVX, ovariectomized; PLC, phospholipase C; PKC, protein kinase C; SD, Sprague Dawley

3.5. Anxiety-like and affective behavior

Affective disorders, such as major depressive disorder and anxiety-based disorders, including post-traumatic stress and generalized anxiety disorders, are more prevalent in women than in men (Altemus, Sarvaiya, & Neill Epperson, 2014; Bangasser & Cuarenta, 2021; Kessler et al., 2012). Though monoaminergic systems have historically been the predominant focus for therapeutic targets, growing evidence suggests a functional role for the glutamate system in the treatment of anxiety and affective disorders (Joffe et al., 2019; Mathews, Henter, & Zarate, 2012; Palucha & Pilc, 2007). Indeed, several studies have reported the robust anxiolytic-like effects in preclinical models following pharmacological manipulations of Group 1 receptors (Ballard et al., 2005; Busse et al., 2004; Pietraszek et al., 2005), although this literature is heavily biased toward male subjects.

Stress, a major contributing factor to the development of anxiety and affective disorders, dynamically alters Group 1 receptor function. In the PFC of male rats, exposure to 2-days forced swim stress decreased mGlu₁ and increased mGlu₅ receptor protein expression (Wang et al., 2015). Consistent with a stress-induced dysregulation of PFC mGlu₅ receptor signaling, 20-min of restraint stress in male mice impairs the induction of mGlu₃ LTD that is dependent on mGlu₅ receptor signaling (Di Menna et al., 2018; Joffe et al., 2019). By contrast, female rats exposed to forced swim stress displayed intact mGlu₁ and decreased mGlu₅ receptor expression (Wang et al., 2015). Continued research should account for cell type-specific differences in mGlu₅ receptor signaling, as we recently found that restraint stress potentiates somatostatin interneuron activity through a mGlu₅ receptor-dependent mechanism in both male and female mice (Joffe et al., 2022). Together, these results suggest that stress-induced alterations in mGlu_1/5 receptor signaling may mediate some effects of stress and potentially increase vulnerability to affective behaviors in a sex-specific manner.

In female rodents, E2 can reduce anxiety-like behaviors (Hill, Karacabeyli, & Gorzalka, 2007; Mora, Dussaubat, & Díaz-Véliz, 1996); these effects are mediated by activation of Group 1 mGlu receptors (De Jesús-Burgos et al., 2012) and endocannabinoid mobilization (Hill et al., 2007). Studies from ovariectomized female rats suggest that the ability of mGlu_1/5 receptors to regulate anxiety-like behavior may vary across the estrous cycle. In E2-treated ovariectomized female rats, direct infusion of DHPG into the basolateral amygdala decreased anxiety-related behaviors in the elevated plus maze (De Jesús-Burgos et al., 2012) and Vogel conflict tests (De Jesús-Burgos et al., 2016). By contrast, this manipulation had no effect on non-E2 treated ovariectomized females in either assay or in males in the plus maze. Furthermore, stimulation of mGlu_1/5 receptors within the BLA increased anxiety-like behavior in male rats on the Vogel conflict test (De Jesús-Burgos et al., 2016). It is tempting to speculate that these differences could be related to differential contributions of CB1 receptor signaling to anxiety-like behavior (Bowers & Ressler, 2016), but more studies are needed to fully evaluate that hypothesis. In any case, the results from these studies are consistent with a sex-specific role of amygdalar Group 1 receptors in the attenuation of anxiety and depressive-type behaviors.

Antidepressant-like and anxiolytic-like phenotypes have been observed in mGlu₂ and mGlu₃ receptor knockout mice and following administration of compounds that inhibit mGlu₂ and/or mGlu₃ receptors (Highland et al., 2019; Joffe et al., 2020; Maksymetz & Joffe, 2021). Unfortunately, most of these studies were restricted to male mice. One exception is our recent study reporting a panel of emotional behavioral studies assessing frontal cortex deletion of mGlu₃ receptors (Joffe et al., 2021). We observed that frontal cortex knockdown of mGlu₃ receptors decreased immobility in the forced swim and tail suspension tests and increased open-arm time in the elevated-zero maze. We found no evidence for an interaction between mGlu₃ receptor knockout and sex; however, it is unclear to what extent these findings may generalize to additional behaviors, mGlu₃ receptor function in other brain areas, or acute pharmacological receptor inhibition. Additional studies examining potential sex-dependent behavioral effects of mGlu₂ and mGlu₃ modulators are certainly warranted.

Studies using selective pharmacology and knockout mice implicate Group 3 receptors in anxiety-like behaviors. Systemic treatment with the mGlu₄ PAM ADX88178 increased open-arm entries on the elevated plus maze in both male and female mice (Kalinichev et al., 2014), suggesting that mGlu₄ receptor signaling can reduce anxiety-like behavior in both sexes. Consistent with this hypothesis, male mGlu₄ receptor knockout mice display increased anxiety-like behavior in the open field and elevated plus maze that emerge at age 12 months (Davis et al., 2012). In striking contrast, however, female mGlu₄ receptor knockouts display reduced anxiety-like behaviors in both assays at 6 and 12 months. The authors observed similar sex- and age-dependent effects on sensorimotor function assessed on the accelerating rotarod. Male knockout mice at 12 months, but not 6 months, fell from the rotarod more quickly than wild-type controls, whereas female knockout mice at both ages displayed longer latencies to fall than female wildtypes. Interestingly, in the same studies, mGlu₄ knockout mice displayed altered cued fear learning across all ages and sexes. Thus, the sex- and age-dependencies of mGlu₄ receptor regulation of anxiety-like and sensorimotor behaviors likely occur through actions within one or more distinct neural circuits, rather than at the cellular level of receptor expression or function. Considering that mGlu₄ receptors are highly expressed in corticostriatal and thalamostriatal circuits—which are heavily implicated in anxiety and motor function but less so in fear learning—it is tempting to speculate that these receptor populations contribute to the observed sex- and age-dependent behavioral phenotypes in mGlu₄ knockout mice. Clearly, additional studies will be needed to address this speculative hypothesis.

Numerous studies suggest that mGlu₄ and mGlu₈ may have related and overlapping functions, particularly with respect to anxiety-like behaviors. Similar to the sex-dependent findings in mGlu4 receptor knockout mice, there is a sex difference in anxiety-like behavior in mGlu₈ receptor knockouts. In the elevated plus maze, open-arm time and entries were decreased in male knockouts but increased in female knockouts, respectively suggesting enhanced and reduced anxiety-like behaviors (Duvoisin et al., 2010). Consistent with this finding, the acoustic startle response was enhanced in male knockout mice but attenuated in female knockouts. An important caveat to these findings, however, is that mGlu₈ knockouts of both sexes exhibited less time in the center of an open field relative to matched controls. Therefore, the broad generalizability of the anxiety-like phenotype of female knockout mice should be interpreted with some caution.

3.6. Neurodegeneration

Alzheimer’s disease, a type of dementia, is characterized as a progressive neurological disorder in which atrophy of the brain leads to cell death and subsequent loss of cognitive functions (Johnson et al., 2019). Though the incidence of Alzheimer’s disease does not differ by sex, there are distinct sex-differences in the symptomatology, progression, and treatment of Alzheimer’s disease (Ferretti et al., 2018). Aggregation of β-amyloid proteins in the formation of amyloid plaques is a pathological hallmark of Alzheimer’s disease in both sexes (Murphy & LeVine 3rd, 2010). In fact, accumulation of oligomers induces abnormal clustering of mGlu₅ receptors at the cell membrane and, in turn, alters mGlu₅ glutamate transmission leading to synaptic degradation (Renner et al., 2010). In cortical tissue of wild-type male mice and human men, β-amyloid oligomers display a high-affinity (nanomolar range) interaction with mGlu₅ receptors, an effect not observed in female mice or in women (Abd-Elrahman et al., 2020; Abd-Elrahman & Ferguson, 2022). Furthermore, in primary neuronal cultures from male but not female mice, β-amyloid oligomers activate glycogen synthase kinase 3β signaling in an mGlu₅ receptor dependent manner. Consistent with these actions, systemic treatment with CTEP mGlu₅ inhibition reduced β-amyloid pathology and improved object recognition in male disease model mice but had no effect in females (Abd-Elrahman et al., 2020; Abd-Elrahman & Ferguson, 2022). An mGlu₅ silent allosteric modulator reversed deficits in hippocampal-dependent cognition in male Alzheimer’s disease model mice, but cohorts of female mice were not included in that study (Haas et al., 2017). Taken together, these findings suggest a male-specific contribution of mGlu₅ receptor signaling in β-amyloid-driven pathology in Alzheimer’s disease.

Evidence from Huntington’s disease models similarly suggests that male individuals may be more sensitive to neurodegeneration related to mGlu₅ receptor signaling. MPEP administration in male mouse models of Huntington’s disease inhibits disease progression (Schiefer et al., 2004) and mGlu₅ knockout improves motor behavior and decreases aggregate size of mutant huntingtin (Ribeiro et al., 2014). There is growing preclinical and clinical evidence to suggest that sex may influence the phenotype of Huntington’s disease (Bode et al., 2008; Cao et al., 2019; Zielonka et al., 2013), and preclinical models are beginning to take this variation into account. A recent study examined CTEP administration in male and female Huntington’s disease model mice; although both sexes displayed improvements in HD neuropathology following mGlu₅ blockage, female HD mice required longer CTEP treatment to show motor improvements (Li et al., 2022). Furthermore, mGlu₅ antagonism did not improve cognitive impairments or grip strength in female HD mice. These findings collectively suggest that males may be more sensitive to neurodegenerative processes related to aberrant mGlu₅ receptor signaling. One intriguing possible mechanism could be sex differences in coordination between mGlu₃ and mGlu₅ receptor signaling, as the same group observed that LY379268 exerts sex-dependent effects on Akt, GSK3β, and ERK1/2 phosphorylation in the striatum of Huntington’s disease model mice (Li et al., 2022).

4. Conclusions

Despite preclinical evidence to suggest possible sex-dependent differences in mGlu receptor expression and function, few human studies have been adequately powered to make well-justified conclusions and it has been challenging to make clear comparisons across species. For example, positron emission tomography has enabled the quantification of available mGlu₅ receptors in the intact human brain, and the allosteric radiotracer [¹¹C]-ABP688 has been used to assess sex differences in mGlu₅ receptor availability. While sex differences in mGlu₅ receptor availability were not detected in a small study (DuBois et al., 2016), a larger study found significantly higher rates of [¹¹C]-ABP688 binding potential in men than women across several regions of the PFC, striatum, and hippocampus (Smart et al., 2019). The largest magnitude differences were observed in the orbitofrontal cortex and the dorsolateral PFC, in which binding potentials were 22% and 20% greater in men, respectively. Overall, a whole-brain comparison of mGlu₅ receptor availability found that [¹¹C]-ABP688 binding potential was 17% greater in men. Consistent with these findings, an earlier human PET imaging study using [¹¹C]-ABP688 reported 31% greater global mGlu₅ binding potential in non-smoking men when compared to non-smoking women (Akkus et al., 2013). Taken together, the larger availability of mGlu₅ receptors in women compared to men seems to conflict with the preclinical literature, as major sex differences in mGlu₅ receptor expression have not been consistently reported in rodents (Wickens et al., 2018) and available evidence points toward decreased cortical mGlu₅ receptor expression in male rats compared to females (Wang et al., 2015). Importantly, technical considerations impede direct comparisons between the clinical and preclinical literature. For one, studies using radioligands like [¹¹C]-ABP688 detect mGlu₅ receptors available on the cell surface, while Western blots with targeted antibodies can detect receptor populations in intracellular (Purgert et al., 2014) or nuclear (Jong et al., 2005) membranes. In addition, mGlu receptor antibodies were raised against sections of the intracellular C-terminal tail or the N-terminal extracellular domain, whereas allosteric ligands like [¹¹C]-ABP688 interact with the allosteric binding pocket within the transmembrane domain (Gregory & Conn, 2015). Based on this, sex-dependent changes in splicing, post-translational modifications, or interactions with other proteins (e.g., ER) could disproportionately affect one method of assessing mGlu₅ receptor expression. Moving forward, parallel cross-species studies should be designed to use identical techniques to assess potential sex differences in mGlu receptor expression. Ideal studies would examine expression of transcripts, protein, and available receptors for all mGlu receptor subtypes.

Here, we have summarized key pieces of evidence suggesting sex differences in mGlu receptor functions, highlighting literature showing that the gonadal hormone E2 has acute effects on these mGlu receptor functions. However, much less is known about how other factors, including acute effects of other gonadal hormones (e.g., testosterone, progesterone), developmental gonadal hormone effects (i.e., organizational effects of hormones), and sex chromosome complement might influence sex-specific behavioral and neurobiological outcomes. Indeed, there is a rich literature in neuroendocrinology and the sex differences field showing that these other factors influence behavior and neurobiology. For instance, rodent studies have revealed links between circulating testosterone and anxiety-like behaviors. Mice and rats with the testicular feminization mutation, which renders the androgen receptor insensitive to testosterone, exhibit higher anxiety-like behaviors compared to wildtypes (Chen et al., 2014; Zuloaga et al., 2008, 2011). We have also shown that testosterone exposure in adult mice reduces measures of anxiety–/depressive-like behavior after chronic stress exposure in both males and females (Seney et al., 2013a). Whether mGlu receptor signaling mediates the behavioral effects of androgens remains, by and large, an open question.

Additional sources of sex differences may originate in early developmental events and from sex chromosome genes. Exposure to gonadal hormones during sensitive periods of development (“organizational effects”) is known to cause permanent sex differences in behavior and brain structure (Cooke et al., 1998; Handa et al., 1985; Phoenix et al., 1959). Moreover, sex chromosomes differ between males and females in the presence or absence of a Y and in the dosage of X chromosomes. The role of genetic sex has been difficult to investigate due to mosaicism in females due to X-linked inactivation and the relationship between genetic sex and gonadal sex. We and others have begun to parse genetic sex from gonadal sex by using the Four Core Genotypes (FCG) mice. In the FCG strain, the testes determining gene, Sry, was placed on an autosome after spontaneous deletion from the Y chromosome; thus, both XX mice and XY mice can have either ovaries or testes, dissociating genetic from gonadal sex (Arnold & Chen, 2009; McCarthy & Arnold, 2011). Using the FCG mice, we showed that genetic sex influences anxiety-like behavior (Seney et al., 2013a) and gene expression in mood-relevant brain regions (Barko et al., 2019; Paden et al., 2020; Puralewski, Vasilakis, & Seney, 2016; Seney et al., 2013b). Other studies in FCG mice have revealed sex chromosome influences on sex differences in aggressive, parental, social, cognitive, circadian, and substance use-associated behaviors (Aarde et al., 2021; Barker et al., 2010; Cox & Rissman, 2011; Gatewood et al., 2006; Kuljis et al., 2013; Martini et al., 2020; McPhie-Lalmansingh et al., 2008). Together, it is clear that factors other than circulating E2 influence sex-related outcomes, and that there is much work to be done related to how these factors may impact mGlu function.