Not your mother's hormone therapy: Highly selective estrogen receptor beta agonists as next-generation therapies for menopausal symptom relief.

Although there has been some debate about the identity and cellular localization of ERs, three receptors are well established: ERα, ERß, and G-protein coupled estrogen receptor (GPER or GPER1). ERα and ERß affect cellular function via two general mechanisms: 1) as canonical intracellular sex steroid receptors that bind estrogens in the cytoplasm and then traffic to the nucleus to promote gene transcription ( Hall et al., 2001 ; McKenna et al., 1999 ), and 2) as stimulus-driven receptors that get trafficked to the plasma membrane when activated ( Sheldahl et al., 2008 ) and interact with glutamate receptors to rapidly trigger cell signaling ( Boulware and Mermelstein, 2009 ; Boulware et al., 2005 ). GPER (formerly known as GPR30) is a typical 7 transmembrane domain G-protein located within cell membranes where its primary mode of action is to influence cell signaling ( Barton et al., 2018 ; Prossnitz et al., 2007 ; Prossnitz and Barton, 2023 ; Srivastava and Evans, 2013 ). Because it was not formally identified as an ER until 2008 ( Barton et al., 2018 ; Prossnitz and Barton, 2023 ), clinical drug development targeting this receptor is in its infancy. As a result, we will focus here on ERα and ERß. Both ERα and ERß are expressed throughout the brain, although their relative abundance varies by brain region ( Shughrue et al., 1997 ). In general, ERß is more widely distributed in the brain than ERα ( Kuiper et al., 1998 ), and is the predominant ER in cognitive brain regions including the hippocampus and the frontal and temporal cortices ( Shughrue et al., 1997 ). Although both receptors are expressed in hypothalamic nuclei involved in hot flashes, including the preoptic area and arcuate nucleus, ERα is the principal isoform in these regions ( Kuiper et al., 1998 ; Merchenthaler et al., 2004 ; Shughrue et al., 1997 ). Perimenopause and early menopause have been associated with impaired cognition and “brain fog”, a general term for lack of mental clarity and focus, forgetfulness, attention, memory, and confusion ( Reuben et al., 2021 ; Weber et al., 2013 ; Woods et al., 2000 ). Processing speed and verbal learning and memory are particularly impaired during these stages of the menopausal transition ( Epperson et al., 2013 ; Gervais et al., 2020 ; Karlamangla et al., 2017 ; Kilpi et al., 2020 ; Maki and Jaff, 2022 ; Maki and Weber, 2021 ), but reports also indicate impairments in episodic memories including spatial memory ( Crestol et al., 2023 ; Lissaman et al., 2024 ; Rentz et al., 2017 ), as well as significant alterations to episodic memory neural circuitry particularly focused in the hippocampus ( Jacobs et al., 2016 ). Menopausal women are also at greater risk of dementia, particularly Alzheimer’s disease, later in life ( Chêne et al., 2015 ; Launer et al., 1999 ). Much of what we know about how estrogens and estrogen receptors regulate memory comes from studies of the dorsal hippocampus, a bilateral medial temporal lobe structure that mediates contextual memories involving spatial, object, working, and social information. These types of memories commonly deteriorate in advanced aging due to deterioration and dysfunction in the dorsal hippocampus (rodents)/posterior hippocampus (humans) (e.g., Chawla et al., 2013 ; Langnes et al., 2020 ; Small et al., 2011 ; Winocur, 1992 ). Based largely on >30 years of rodent data showing that cognitive brain regions like the hippocampus are particularly sensitive to estrogens, the detrimental effects of menopause on cognition and hippocampal function have generally been attributed to estrogen loss (for reviews, see Fleischer and Frick, 2023 ; Lacreuse et al., 2015 ; Luine, 2014 ; McEwen, 2001 ; Taxier et al., 2020 ; Torromino et al., 2021 ), despite the fact that levels of multiple sex steroid hormones and gonadotropins are altered by menopause. A detailed accounting of the extensive literature describing effects of exogenous E2 on hippocampal function is beyond the scope of this review but is synthesized in many comprehensive reviews on the subject (e.g., Babayan and Kramár, 2013 ; Bimonte-Nelson et al., 2010 ; Brandt and Rune, 2020 ; Fleischer and Frick, 2023 ; Frick et al., 2015 ; Gibbs, 2010 ; Lacreuse et al., 2015 ; Luine and Frankfurt, 2020 ; Morrison et al., 2006 ; Murakami et al., 2018 ; Rocks and Kundakovic, 2023 ; Sheppard et al., 2019 ; Smith et al., 2009 ; Srivastava et al., 2013 ; Taxier et al., 2020 ). To very briefly summarize relevant work conducted largely in young adult ovariectomized (OVX) rodents, exogenous E2 in the dorsal hippocampus increases within approximately 30–60 minutes CA1 dendritic spine density, presynaptic glutamate release, modes of synaptic plasticity including long-term potentiation (LTP), and the formation of spatial, object recognition, and social memories by rapidly activating numerous cell signaling pathways including protein kinase A (PKA), phosphatidylinositol 3-kinase (PI3K), extracellular signal-regulated kinase (ERK), and mechanistic target of rapamycin (mTOR) that trigger local protein synthesis, epigenetic alterations, and activation of transcription factors like CREB ( Bi et al., 2000 ; Fan et al., 2010 ; Fernandez et al., 2008 ; Fortress et al., 2013 ; Gu et al., 1999 ; Gu and Moss, 1996 ; Hasegawa et al., 2015 ; Hojo et al., 2015 ; Jain et al., 2018 ; Kuroki et al., 2000 ; Oberlander and Woolley, 2016 ; Ogiue-Ikeda et al., 2008 ; Phan et al., 2015 ; Smith et al., 2009 ; Srivastava et al., 2011 ; Srivastava et al., 2008 ; Tuscher et al., 2015 ; Vedder et al., 2013 ; Wu et al., 2011 ; Yokomaku et al., 2003 ; Zhao et al., 2012 ; Zhao et al., 2010 ). E2 also increases neurogenesis in the dentate gyrus region of the hippocampus of female rats within 4 hours ( Mazzucco et al., 2006 ), an effect positively associated with hippocampus-dependent memory ( Duarte-Guterman et al., 2015 ; McClure et al., 2013 ). Many of E2’s effects can be triggered via pharmacological activation of either ERα or ERß ( Kramár et al., 2009 ; Liu et al., 2008 ; Mazzucco et al., 2006 ; Oberlander and Woolley, 2016 ; Waters et al., 2009 ; Zhou et al., 2014 ), suggesting involvement of both receptors in the rapid effects of E2 on dorsal hippocampal structural plasticity and memory. Ultrastructural analyses support this role, as both ERs are abundantly expressed throughout all segments of dorsal hippocampal neurons, including axon terminals, dendrites, and dendritic spines, in which they are positioned near the plasma membrane to interact with neurotransmitter receptors and signaling proteins to rapidly modulate synaptic physiology and morphology ( Almey et al., 2014 ; Boulware et al., 2013 ; Boulware et al., 2007 ; Boulware and Mermelstein, 2009 ; Boulware et al., 2005 ; Milner et al., 2005 ; Milner et al., 2001 ; Mitra et al., 2003 ; Mitterling et al., 2010 ; Sheldahl et al., 2008 ; Waters et al., 2011 ). In support of ERß’s role in regulating hippocampal function, ERß agonists produce myriad changes in the CA1 ( Figure 2 ), including increased post-synaptic strength (fEPSP), enhanced LTP (reduced threshold, increased ceiling), increased GluR1 expression, trafficking, and phosphorylation, increased glutamate neurotransmission, increased synaptic protein expression, and increased CA1 dendritic branching and spine density, as well as increased neurogenesis in the dentate gyrus ( Kramár et al., 2009 ; Liu et al., 2008 ; Mazzucco et al., 2006 ; Oberlander and Woolley, 2016 ; Waters et al., 2009 ; Zhou et al., 2014 ). Accordingly, ERß activation also enhances various forms of hippocampus-dependent memory ( Figure 2 ). For example, systemic injection or dorsal hippocampal infusion of ERß agonists in OVX mice enhance numerous forms of spatial memory ( Liu et al., 2008 ; Paletta et al., 2018 ; Pisani et al., 2016 ; Walf et al., 2008a ; Walf et al., 2006 ), object recognition memory ( Walf et al., 2008a ; Walf et al., 2006 ), social transmission of food preferences ( Clipperton et al., 2008 ; Ervin et al., 2015 ), and contextual fear generalization ( Lynch III et al., 2014 ). Consistent with these effects, a single bilateral infusion of the ERß agonist diarylpropionitrile (DPN) into the dorsal hippocampus of OVX mice enhances spatial and object recognition memory consolidation ( Boulware et al., 2013 ; Pereira et al., 2014 ) in a manner that depends on rapid activation of p42-ERK phosphorylation and interactions with metabotropic glutamate receptor 1a (mGluR1a) at the plasma membrane ( Boulware et al., 2013 ). Although exogenous activation of either ER is sufficient to facilitate hippocampus-dependent memory, infusion of ERα or ERß antagonists into the dorsal hippocampus of OVX mice showed that only ERß is necessary for the consolidation of both spatial and object recognition memories ( Kim and Frick, 2017 ). These data suggest that endogenous dorsal hippocampal neuroestrogens released in response to learning ( Tuscher et al., 2016 ) may preferentially bind ERß to facilitate memory consolidation, which is consistent with the predominance of ERß expression in this brain region ( Shughrue et al., 1997 ). Data from studies of female ERß knockout mice suggest that ERß is also necessary for normal social recognition and discrimination ( Choleris et al., 2003 ; Choleris et al., 2006 ), and spatial memory ( Liu et al., 2008 ; Rissman et al., 2002 ). As such, activation of ERß in the dorsal hippocampus appears to be a critical step in memory formation among OVX females, making this receptor a credible target for maintenance of hippocampus-dependent memories in menopause. Hot flashes (aka, hot flushes, night sweats) are sudden, transient, and intense feelings of heat that produce an initial rise in core body temperature followed by skin vasodilation and sweating to dissipate heat that results in a return of core body temperature to baseline. These sporadic episodes can last from seconds to an hour ( Sturdee et al., 2017 ), and can be extremely embarrassing and disruptive, especially among those who experience frequent hot flashes. Middle-aged women suffer from hot flashes for a median of up to 7–11 years ( Avis et al., 2015 ; Freeman et al., 2014 ), during which time work, activities of daily living, and sleep can be significantly affected ( Hooper et al., 2022 ; Kravitz and Joffe, 2011 ; Tang et al., 2018 ; Tomida et al., 2021 ). In support of their negative impacts, hot flashes are associated with lower life satisfaction, lower self-esteem, increased binge eating, and increased depression ( Hooper et al., 2022 ; Mueller et al., 2024 ; Tomida et al., 2021 ). Although this menopausal symptom may seem unrelated to memory dysfunction, recent research demonstrates clear links. For example, increased frequency and severity of nighttime hot flashes is associated with increased memory dysfunction, slower processing speed, altered brain activity, and plasma markers of Alzheimer’s disease ( Jaff et al., 2020 ; Maki et al., 2020 ; Thurston et al., 2024 ; Triantafyllou et al., 2016 ). This association is likely the result of multiple factors. For example, sleep disruption (particularly reduced REM sleep) is linked to impaired memory functioning ( Boyce et al., 2017 ; Lucey et al., 2021 ; Pérez-Carbonell and Iranzo, 2024 ), and reduced serum E2 in middle-aged women is associated with difficulty falling and staying asleep ( Kravitz and Joffe, 2011 ). In support, the prevalence of chronic insomnia in a sample of women in California aged 35 to 65 years was 36.5% in premenopause, 56.6% in perimenopause, and 50.7% in postmenopause, with increased symptoms of insomnia positively associated with hot flash severity ( Ohayon, 2006 ). Accordingly, longitudinal data from the U.S.-based Study of Women Across the Nation shows that women in their 40s and 50s who experience moderate to severe hot flashes are nearly three times more likely to wake up at night than women who do not report hot flashes ( Kravitz and Joffe, 2011 ). Additionally, cardiovascular disease is associated with increased dementia risk, and middle-aged women with frequent hot flashes exhibit increased aortic calcification and indicators of subclinical atherosclerosis ( El Khoudary et al., 2019 ). Thus, hot flashes may be a symptom of underlying risk for memory loss, and reducing hot flashes, particularly at night, should provide benefits for both comfort and memory functioning. The past decade has seen tremendous progress in understanding the mechanisms underlying hot flashes. The brain’s thermoregulatory circuit consists primarily of hypothalamic nuclei including the arcuate nucleus of the hypothalamus (aka, infundibular nucleus), preoptic area (POA, which includes the median preoptic nucleus (MnPO) and medial preoptic area (MPA)), ventromedial nucleus of the hypothalamus (VMH), and the dorsomedial nucleus of the hypothalamus (DMH) ( Zhang et al., 2021 ). Key for hot flash generation are neurons in the arcuate that co-express the neuropeptides kisspeptin, neurokinin B (NKB, a tachykinin family peptide), and dynorphin ( Burke et al., 2006 ). These neurons, commonly referred to as “KNDy” neurons, are major targets for steroid hormones including estrogens from the ovaries which inhibit the production of the neuropeptides after which the cells are named ( Santoro et al., 2021 ). Relevant for hot flashes are KNDy neurons in the arcuate that project to hypothalamic regions that regulate circulating estrogen levels and heat dissipation. Regarding ovarian estrogens, arcuate KNDy neurons project to cells in the median eminence that produce gonadotropin releasing hormone (GnRH) ( Krajewski et al., 2005 ), thereby regulating ovarian estrogen production by influencing the synthesis and release of luteinizing hormone (LH) and follicle stimulating hormone (FSH) from the pituitary ( Rance et al., 2013 ; Santoro et al., 2021 ). As discussed below, ovarian estrogens feed back to the brain to negatively regulate the activity of KNDy neurons. Importantly, KNDy neurons also project to heat dissipation neurons in the MnPO region of the POA and warm-sensing neurons in the MPA that express neurokinin/tachykinin 3 receptor (NK3R) receptors, the primary receptor for NKB ( Rance et al., 2013 ; Santoro et al., 2021 ; Zhang et al., 2021 ). The importance of KNDy neurons in the biology of hot flashes has been tested primarily in rodents. Although rodents do not sweat, they release excess body heat via their tail, which temporarily increases tail skin temperature (T skin ) and concomitantly reduces core body temperature (T core ). Neurotoxic ablation of arcuate KNDy neurons in female rats significantly reduces T core and T skin , implicating KNDy neurons in mediation of both core temperature and the cutaneous vasodilatation that dissipates body heat ( Mittelman-Smith et al., 2015 ; Mittelman-Smith et al., 2012b ). Supporting the importance of the NK3R-expressing MnPO neurons to which KNDy neurons project, intracranial infusion of the NK3R agonist senktide into the MnPO of female rats decreases T core ( Dacks et al., 2011 ; Mittelman-Smith et al., 2015 ), and systemic injection of senktide in female mice causes a rapid, transient, and simultaneous reduction of T core and increase in T skin ( Fleischer et al., 2021 ; Krajewski-Hall et al., 2018 ; Krull et al., 2017 ). As such, activation of hypothalamic NK3Rs appears critical for thermoregulation and hot flash generation. This conclusion is supported by several lines of evidence in women. For example, as discussed below, NKB, NK3R, and ERα are colocalized in hypertrophied neurons in the arcuate nucleus of postmenopausal women ( Rance and Young, 1991 ; Ruth et al., 2023 ). In addition, genome-wide association (GWAS) data from the WHIMS show that genetic variation in the gene that codes for NK3R ( TACR3 ) is associated with increased odds of hot flashes and night sweats in postmenopausal women aged 50–79 years ( Crandall et al., 2017 ). A similar analysis of data from middle-aged women from the UK Biobank supports TACR3 as the genetic basis of hot flashes ( Ruth et al., 2023 ). Estrogens help maintain optimal T core and prevent excess heat dissipation. For example, OVX in rats increases heat dissipation as indicated by reduced T core and increased T skin , whereas E2 maintains T core , even during environmental heat exposure ( Dacks and Rance, 2010 ; Mittelman-Smith et al., 2012a ; Opas et al., 2006 ). OVX also lowers the threshold at which tail skin vasodilation occurs ( Dacks and Rance, 2010 ; Mittelman-Smith et al., 2012a ). This finding mirrors data from postmenopausal women, as individuals who experience hot flashes have a lower core temperature threshold for sweating than those who do not ( Freedman and Krell, 1999 ). Mechanistically, the effects of E2 on hot flashes are thought to be mediated through the KNDy thermoregulatory circuit ( Figure 3 ). In younger females, estrogens from the ovaries repress the activity of KNDy neurons, thereby suppressing the release of NKB from KNDy neuron projections to the MnPO, as well as MnPO projections to the MPA and MPA projections to sympathetic neurons, thus blocking the heat dissipation response ( Rance et al., 2013 ; Santoro et al., 2021 ; Zhang et al., 2021 ). However, in middle-aged rodents, monkeys, and women, the loss of ovarian estrogens releases KNDy neurons from inhibition, leading to hypertrophy of these neurons, increased expression of kisspeptin and NKB, and increased NKB release onto heat dissipation neurons in the MnPO which then triggers the heat dissipation/hot flash response ( Eghlidi and Urbanski, 2015 ; Hrabovszky et al., 2019 ; Rance, 2009 ; Rance et al., 2013 ; Rance and Young, 1991 ; Rometo et al., 2007 ; Sandoval-Guzmán et al., 2004 ; Santoro et al., 2021 ). Thus, ovarian estrogens play a key role in repressing the neural activity that leads to hot flashes. Accordingly, estrogens are the most effective treatments for reducing hot flashes in menopausal women, with both oral and transdermal formulations of CEE and E2 (with or without a progestin) reducing the frequency and severity of hot flashes by about 75% (see Crandall et al., 2023 for a recent comprehensive review). The KEEPS study illustrates how these treatments benefit both hot flashes and sleep. Women aged 42–58 within three years of their last menstrual period were treated with micronized progesterone plus oral CEE or transdermal E2 treatment for 48 months, with screening at 6, 12, 24, 36, and 48 months. Both estrogen treatments significantly reduced the prevalence of moderate to severe hot flashes, insomnia, and irritability relative to placebo within 6 months, with lasting benefits to insomnia for up to 48 months ( Santoro et al., 2017 ). However, many menopausal individuals and physicians remain wary of estrogen therapy given the well-publicized health risks reported by the WHI. For people whose primary menopausal symptom is hot flashes, relief could be provided by new non-hormonal treatments that specifically target hypothalamic NK3R-expressing neurons. The first of this new class of therapies, fezolinetant, is a small molecule NK3R antagonist developed by Astellas Pharma and approved by the FDA in 2023 under the trade name Veozah ® . A similar drug, elinzanetant, which blocks both NK3R and neurokinin-1 receptors, is currently in development by Bayer, GlaxoSmithKline, and NeRRe Therapeutics. In menopausal women, both drugs reduced hot flash frequency and severity, as well as improved sleep quality ( Abo Elnaga et al., 2024 ; Menegaz de Almeida et al., 2025 ). Thus, neurokinin receptor antagonists could have a tremendous positive impact on the lives of menopausal persons experiencing frequent and severe hot flashes, particularly those with a previous or family history of estrogen receptor positive cancers. However, as with any drug, there are cautionary notes. First, of recent concern for fezolinetant are hepatotoxic effects that increase blood liver enzyme levels and elevate risk of liver injury. These rare but serious side effects caused the FDA to issue a black box warning for Veozah ® on December 16, 2024 that now requires hepatic laboratory tests prior to treatment and frequently during the first 9 months of treatment ( U.S. Food and Drug Administration, 2024 ). Second, the effects of fezolinetant or elinzanetant on cognitive functions, notably memory, have not been tested. This is important because treatment of young rodents with NK3R agonists like senktide enhances object recognition and spatial memories mediated by the hippocampus and frontal cortex ( Zlomuzica et al., 2008 ), suggesting that NK3R antagonists could impair memory. Interestingly, senktide also enhances multiple forms of spatial memory in aged male rats, effects associated with increased acetylcholine levels in the hippocampus and frontal cortex as well as reduced NK3R mRNA expression in the hippocampus ( de Souza Silva et al., 2013 ; Schäble et al., 2011 ). In aged humans with cognitive impairments, single nucleotide polymorphisms in the gene that codes for NK3R are associated with impaired learning and memory and reduced right hippocampal volume ( de Souza Silva et al., 2013 ), suggesting that NK3R activity in the hippocampus is important for cognitive function. Thus, although NK3R antagonists may reduce hot flashes, they could also have unintended detrimental consequences for memory, although this has yet to be tested directly. Third, although NK3R antagonists may be effective for hot flashes, they do not address other menopause symptoms like brain fog, reduced bone density, mood disturbances, or genitourinary changes, which is important because individuals often experience multiple menopausal symptoms concurrently. Finally, although the market potential for hot flash treatments is huge, sales of Veozah ® have been slow due to numerous factors including reluctance of insurers, physicians, and women to try a new type of drug ( Pecci, 2024 ). Despite their perceived downsides, estrogens are well known and address multiple menopausal symptoms. Thus, payers, providers, and patients may be more willing to adopt a novel estrogen-like drug that provides the benefits of estrogens while minimizing the risk of adverse side effects. A key issue for estrogen-based hot flash drug development is to determine which ER or ERs regulate KNDy neuron activity. Both ERs are highly expressed in the MPA and are expressed at lower levels in the arcuate, VMH, and DMH, where ERα expression is substantially higher than that of ERß ( Dellovade and Merchenthaler, 2004 ; Merchenthaler et al., 2004 ; Shughrue et al., 1997 ; Shughrue and Merchenthaler, 2001 ). In the arcuate nucleus of female mice, NKB expression is regulated by ERα, as illustrated by data showing that E2 increased NKB expression in OVX ERßKO, but not ERαKO, mice ( Dellovade and Merchenthaler, 2004 ). More recently, data from ERαKO mice show that NK3R expression is regulated by ERα via both estrogen response element-dependent and -independent mechanisms ( Yang et al., 2016 ). Nevertheless, 25% of kisspeptin-positive neurons in the arcuate nucleus of female rats express ERß ( Esr2 ) mRNA ( Kanaya et al., 2020 ), suggesting functional regulation of KNDy neurons by ERß as well. In support, OVX in either ERαKO or ERßKO mice increased T skin , an effect reversed by E2 in both knockouts and wild type mice ( Opas et al., 2006 ). In addition, agonists that preferentially target ERß have previously been reported to reduce T skin in OVX rats ( McGregor et al., 2014 ; Opas et al., 2009 ; Wilson et al., 2021 ) and decrease hot flash frequency in menopausal women ( Tagliaferri et al., 2012 ; Wang et al., 2020 ). Thus, some evidence supports potential efficacy of selective ERß agonists for menopausal hot flash relief.

As reviewed above, evidence supports key roles for both ERα and ERß in facilitating memory formation and suppressing menopausal hot flashes and heat dissipation in female rodents. Thus, either receptor could theoretically be a drug development target. However, ERα is not a viable candidate for menopausal symptom relief because it promotes tumorigenesis and progression of breast, ovarian, and uterine cancer ( Ali and Coombes, 2000 ; Jia et al., 2015 ; Péqueux et al., 2012 ; Yu et al., 2022 ). In contrast, ERβ inhibits breast, ovarian, and colon cancer cell proliferation ( Chang et al., 2006 ; Cotrim et al., 2013 ; Giroux et al., 2008 ; Jia et al., 2015 ; Treeck et al., 2007 ) and suppresses expression of genes regulated by ERα in cellular models of breast cancer ( Chang et al., 2006 ). ERβ is also neuroprotective, has anxiolytic and antidepressant effects in rodents, and has positive effects on cardiovascular and gastrointestinal diseases ( Chen et al., 2022 ; Miller et al., 2005 ; Vargas et al., 2016 ; Walf and Frye, 2007 ; Walf et al., 2009 ; Walf et al., 2008b ), suggesting additional benefits to targeting this receptor. Moreover, in a mouse model of Alzheimer’s disease, ERα was associated with hippocampal tau hyperphosphorylation and aggregation whereas ERß was associated with hippocampal tau hypophosphorylation ( Xiong et al., 2015 ). Likewise, lentiviral-induced expression of ERβ in a rat model of Alzheimer’s disease reduced hippocampal β-amyloid deposition, attenuated inflammation, and improved learning and memory ( Tian et al., 2013 ). Collectively, the neuroprotective benefits of ERβ, in light of the carcinogenic effects of ERα, strongly support ERβ as the better choice for the development of drugs that provide the benefits of estrogens on memory, hot flashes, and possibly Alzheimer’s-related neuropathology without the health risks of E2 treatment. Moreover, ERβ’s role in mediating mechanisms involved in multiple menopausal symptoms makes ERβ agonists more therapeutically advantageous than NK3R antagonists for polysymptomatic women. Although the idea of using ERß agonists to treat menopausal symptoms is not new, developing potent compounds that are highly selective for just one ER is difficult because ERα and ERß are highly homologous in their ligand-binding pockets. Numerous compounds described as “selective” ERß agonists are barely so, although several have been tested in clinical trials. These compounds fall into two categories, naturally derived products (botanicals, nutraceuticals) and synthetic compounds. Thus far, natural products have produced disappointing results. For example, S-equol (Ausio Pharmaceuticals, LLC), a metabolite of the dietary phytoestrogen daidzein that is 13-fold more selective for ERß ( Muthyala et al., 2004 ) failed to reduce the frequency of hot flashes in menopausal women over age 40 ( NCT00962585 ) or reduce cognitive dysfunction in subjects with Alzheimer’s disease ( NCT04516304 ). A Chinese herbal extract called “menopausal formula 101” (MF101, BioNovo, Inc.) binds ERα and ERß equally well but selectively stimulates ERß-mediated gene transcription ( Cvoro et al., 2007 ). Although one study reported reduced hot flashes in a phase 2 clinical trial ( Grady et al., 2009 ), a 2011 phase 3, double-blind, placebo-controlled, randomized clinical trial to assess safety and efficacy of MF101 for hot flashes and menopausal symptoms in postmenopausal women ( NCT00906308 ) did not report results and is considered of “unknown status” on Clinicaltrials.gov . A newer natural compound called PhytoSERM (NeuTherapeutics, LLC), a mixture of plant-based estrogens that is 83-fold more selective for ERß, has been tested in one clinical trial in women aged 45–60 for hot flashes and memory loss with no results posted ( NCT01723917 ) and is in two additional clinical trials recruiting women aged 45–60 with hot flashes for two other clinical trials related to hot flashes, cognition, and brain metabolism ( NCT06186531 , NCT05664477 ). Thus, the efficacy of this more selective botanical ERß agonist remains unknown. Natural ERß agonists may have thus far fallen short in addressing menopausal symptoms because they were not designed to selectively target only ERß. Synthetic compounds that do so may be more efficacious. The first of these in clinical trials is Eli Lilly’s Erteberel ( LY500307 ), which is 14-fold more selective for ERß over ERα but potently binds to both receptors ( Norman et al., 2006 ). As such, this compound does not show more favorable ERß selectivity than natural substances like S-equol or MF101. A phase 2 clinical trial testing effects of Erteberel on estradiol withdrawal-induced mood symptoms in healthy women aged 45–65 with past perimenopausal depression ( NCT03689543 ) was completed in September 2024 with no results yet posted. However, a recent preprint indicates that LY500307 protects against amyloid pathology and cognitive decline in male and female mice of the App NL-G-F Alzheimer’s model ( Demetriou et al., 2024 ), suggesting possible efficacy for memory. Our group has spent the past 10 years developing ERß agonists that are significantly more potent and selective than existing ERß-targeting compounds and can address multiple menopause symptoms. This work began as an NIH-funded academic collaboration among the Donaldson, Sem, and Frick labs, and the resulting data formed the basis of the company, Estrigenix Therapeutics, Inc., that these investigators founded in 2018. As part of our collaborative efforts, the Donaldson lab synthesized well over a hundred ER compounds, including some up to 1,200-fold more selective for ERß over ERα, and the Sem lab used TR-FRET and cell-based transcriptional assays to assess ER potency, selectivity, and receptor docking ( Sampathi Perera et al., 2018 ; Wetzel et al., 2020 ; Wetzel et al., 2022 ). The Frick lab then conducted in vivo behavioral testing of our most highly selective compounds (ranging in selectivity from 750-fold to 1,200-fold) in mouse models of menopause and Alzheimer’s disease to determine efficacy for memory, hot flashes, and mood-related behaviors. Our published data for one compound, EGX358 (formerly named ISP358–2), are detailed below. Our first report detailed the development of several compounds, of which EGX358 showed the highest ERß selectivity. EGX358 is one of a unique structural class of compounds possessing a cyclohexane-based saturated ring tethered to a phenol ring ( Figure 4E ), making it an A–C estrogen that closely resembles naturally occurring estrogens but without the B and D rings. EGX358 is 750-fold more selective for ERβ over ERα ( Hanson et al., 2018 ), which is >40 times more selective for ERβ than ERβ agonists tested in clinical trials like Erteberel or S-equol. EGX358 is also a potent agonist, with an EC 50 for ERβ of 27.4 nM; although E2 exhibits a more potent EC 50 for ERβ of 0.022 nM it is by nature not selective for ERβ ( Hanson et al., 2018 ). Unlike E2, EGX358 does not cause proliferation of MCF-7 breast cancer cells at concentrations 37x the EC 50 ( Hanson et al., 2018 ). EGX358 also does not bind to the cardiac ion channel hERG at 10 μM or to the major drug metabolizing CYP450 enzymes screened in our panel (3A4, 2D6, 1A2), and exhibited only weak binding (IC 50 of 34 μM) to CYP2C9 ( Hanson et al., 2018 ). Acute EGX358 treatment in mice produced no histological tissue pathology in major organs (heart, liver, kidney) nor hematological abnormalities (complete blood count) ( Hanson et al., 2018 ). Given the combination of high ERβ potency and selectivity plus a favorable preliminary safety profile, we began testing efficacy of EGX358 to facilitate memory consolidation in OVX wild-type mice. For this work, our positive ERβ control has been DPN, a synthetic ERβ agonist with 70-fold ERβ selectivity ( Meyers et al., 2001 ). This compound has been used extensively in basic research to study ERβ-mediated effects on behavioral and biological outcomes, including the Frick lab’s work showing that DPN mimics E2’s ERK-dependent memory consolidation enhancement in OVX female mice ( Boulware et al., 2013 ; Tuscher et al., 2015 ). Although its ERβ selectivity is higher than that of most ERβ agonists in previous or current clinical trials, DPN has not been tested in clinical trials for reasons that are unclear. Because EGX358 is 750-fold more selective for ERβ than ERα ( Hanson et al., 2018 ), and is, therefore, more than 10 times more selective for ERβ than DPN, EGX358 is considerably less likely to bind ERα than DPN. Thus, EGX358 is a safer candidate for clinical trials. Nevertheless, because nearly 25 years of existing data on effects of DPN are available, DPN provides an important pre-clinical reference point against which to judge the effects of EGX358 and other novel compounds. In our initial work, mice were treated with vehicle, DPN, or various doses of EGX358 immediately after training in object recognition (OR) and object placement (OP) tests that assess object recognition and spatial memory, respectively. Memory formation in these tasks requires coordinated activation of the dorsal hippocampus and medial prefrontal cortex ( Tuscher et al., 2018 ), which is important because these brain regions are associated with menopause-related memory dysfunction in humans ( Albert et al., 2017 ; Gervais et al., 2022 ; Maki et al., 2020 ). During the training phase of both tests, mice explore two identical objects in a large square arena ( Figure 4A , B ). Memory is tested 24–48 h later by substituting a novel object for a familiar training object (OR) or by moving one training object to a new location in the arena (OP). Mice who remember the identity and location of the training objects will spend more time than chance with the novel and moved objects, demonstrating intact memory. OVX mice treated with vehicle and tested in the OP and OR tasks 24 or 48 h, respectively, after training, spend chance amounts of time with the two objects, exhibiting impaired memory. For nearly 20 years, the Frick lab has shown repeatedly that systemic injection of 0.2 mg/kg E2 or bilateral dorsal hippocampal infusion of 5 μg E2/hemisphere administered to OVX mice immediately after object training significantly increases the time spent with the novel and moved objects relative to chance and to OVX vehicle-treated mice, demonstrating that exogenous E2 enhances the consolidation and longevity of object recognition and spatial memory in OVX females (e.g., Boulware et al., 2013 ; Fernandez et al., 2008 ; Gresack and Frick, 2006 ; Gross et al., 2021 ; Kim et al., 2016 ; Taxier et al., 2019 ). Similarly, systemic injection of 0.05 mg/kg DPN or bilateral dorsal hippocampal infusion of 10 pg/hemisphere DPN immediately after training also enhances OR and OP memory consolidation in OVX mice ( Boulware et al., 2013 ; Frick et al., 2010 ). As such, E2 and DPN served as important positive controls in our early studies of EGX358. Our first efficacy study tested the extent to which acute post-training treatments of DPN or EGX358 could enhance memory consolation in OVX mice tested in the OR and OP tasks ( Hanson et al., 2018 ). Bilateral dorsal hippocampal infusion of 10 pg DPN/hemisphere, 100 pg EGX358/hemisphere, or 1 ng EGX358/hemisphere enhanced consolidation in both tasks relative to both chance and vehicle-infused mice, with effects of EGX358 comparable to those of DPN ( Hanson et al., 2018 ). We next found that intraperitoneal injection of 0.05 mg/kg DPN or 0.5 mg/kg EGX358 enhanced OR and OP memory consolidation to a similar extent, as did oral gavage of 0.05 mg/kg DPN, 0.5 mg/kg EGX358, or 5 mg/kg EGX358 ( Hanson et al., 2018 ). Thus, acute EGX358 treatment, administered via three routes of administration, enhanced two forms of memory in OVX mice to the same degree as an established commercial ERβ agonist. However, menopausal estrogen therapy is generally a chronic endeavor, so we next assessed the impact of long-term daily oral EGX358 administration on OVX mice. Here, we not only assessed memory but also anxiety-like behaviors (open field, elevated plus maze), depression-like behaviors (tail suspension test, forced swim test), and hot flashes. OVX mice were treated with vehicle, E2 (0.2 mg/kg), DPN (0.05 mg/kg), or EGX358 (0.5 mg/kg) once/day via oral gavage for a total of 63 days ( Fleischer et al., 2021 ). Hot flashes were assessed on treatment day 18, anxiety- and depression-like behaviors on treatment days 22–29, and memory on treatment days 40–58. As in our acute studies described above, all three treatments enhanced memory consolidation in both OR and OP ( Fleischer et al., 2021 ), suggesting that EGX358 was as effective as E2 and DPN in facilitating memory formation ( Figure 4C , D ). To induce a “hot flash”, we injected mice subcutaneously with the NK3R agonist senktide (0.5 mg/kg) after 10 min of baseline thermal imaging of T skin and then recorded T skin every minute after senktide injection for 20 min ( Figure 5F , G ). As in previous studies ( Krajewski-Hall et al., 2018 ; Krull et al., 2017 ), senktide induced a rapid and transient increase in T skin of about 3–4 °C within 5 min ( Fleischer et al., 2021 ). In all groups, this increase peaked about 7 min after injection and returned to baseline within about 15 min ( Fleischer et al., 2021 ). However, the magnitude of the T skin increase was about 25% lower in all treatment groups, as the increase in vehicle-treated mice was ~4 °C compared to ~3 °C in mice treated with E2, DPN, or EGX358 ( Fleischer et al., 2021 ). As with memory, the extent to which EGX358 reduced the senktide-induced increase in T skin was like that of E2 and DPN ( Fleischer et al., 2021 ). Although E2 exhibited minimal anxiolytic effects in the open field, neither DPN nor EGX358 affected measures of anxiety- and depression-like behavior and no treatment affected weekly body weights ( Fleischer et al., 2021 ). Collectively, these data demonstrated that long-term oral treatment with EGX358 could enhance hippocampus- and prefrontal cortex-dependent memory and alleviate a drug-induced hot flash in OVX mice without negative impacts on mood-related measures or body weight. More recently, we extended this work to the EFAD transgenic mouse model of Alzheimer’s disease ( APOE +/+ /5xFAD +/− ), in which late-onset apolipoprotein E ( APOE )-related Alzheimer’s risk is modeled by combining five amyloid precursor protein and presenilin mutations found in early-onset Alzheimer’s families with expression of human APOE gene variants ( Tai et al., 2017 ; Youmans et al., 2012 ). Humans express three APOE gene variants, of which the APOE3 allele is neutral in terms of AD risk, the APOE2 allele is associated with lower risk, and the APOE4 genotype confers significantly higher risk, especially in individuals expressing two copies of APOE4 ( Altmann et al., 2014 ; Bretsky et al., 1999 ; Corder et al., 1993 ; Riedel et al., 2016 ; Roses, 1996 ). By 6 months of age, EFAD mice expressing two copies of APOE4 (E4FAD) mimic the pathology of APOE4/4+ humans, in that female E4FADs exhibit earlier onset of, and more advanced, accumulation of Aβ42, oligomeric Aβ, and amyloid deposits, more hyperphosphorylated tau, and greater synapse degeneration relative to E4FAD males and mice of both sexes that express two copies of APOE3 (E3FAD) ( Cacciottolo et al., 2016 ; Liu et al., 2015 ; Tai et al., 2017 ; Youmans et al., 2012 ). Thus, EFAD mice are an excellent model system in which to test questions related to Alzheimer’s risk factors such as sex, APOE genotype, and estrogen loss. The Frick lab previously showed that acute infusion of E2 into the dorsal hippocampus enhances OR and OP memory consolidation and increases dorsal hippocampal CA1 dendritic spine density in OVX E3FAD and E3/4FAD mice, but not E4FAD mice ( Taxier et al., 2022 ), suggesting that E4FAD females were resistant to the beneficial effects of E2 on memory and CA1 synaptic plasticity. As such, our first study testing effects of long-term oral EGX358 treatment in the EFAD model used OVX E3FAD and E3/4FAD females. At 5 months old, OVX mice were treated orally with vehicle or EGX358 via hydrogel, which served as their sole source of hydration ( Schwabe et al., 2024 ). Between 6 and 7.5 months of age, they were tested in the OR and OP tasks, elevated plus maze, open field, and senktide hot flash test. EGX358 enhanced memory in the OR ( Figure 5A ), but not OP, task in both E3FAD and E3/4FAD females ( Schwabe et al., 2024 ). Although EGX358 treatment benefitted object recognition memory in both genotypes, EGX358 only mitigated the senktide-induced hot flash in E3FAD females ( Figure 5B ) ( Schwabe et al., 2024 ). Uterine weights in both genotypes were increased by treatment but not beyond normal intact uterine weights ( Schwabe et al., 2024 ). As in wild-types, EGX358 did not influence anxiety-like behaviors or body weight ( Schwabe et al., 2024 ). These data suggest that EGX358 could potentially be used to relieve memory and vasomotor symptoms of menopause in some women with established Alzheimer’s pathology without adverse effects on mood, uterine hypertrophy, and body weight. Additional testing of EGX358 in the E3FAD, E3/4FAD, and E4FAD genotypes is underway.

ERβ plays key roles in mediating memory and hot flashes in women and animal models of memory and Alzheimer’s disease. Although the need to alleviate menopausal symptoms is great and estrogen therapy is effective, persistent negative attitudes about estrogen therapy reduce demand and access for traditional hormone treatments. Moreover, those with a personal or family history of certain cancers or cardiovascular disease are not candidates for estrogen therapy. As such, novel next-generation treatments are needed to provide symptom relief, and the development of drugs like ERβ agonists and NK3R antagonists that specifically target mechanisms used by E2 to influence cellular and circuit function will provide numerous options for individuals to personalize their care. The many biological benefits of ERβ activation suggest that highly selective synthetic ERβ agonists like EGX358 could alleviate numerous symptoms for individuals undergoing the perimenopausal and menopausal transitions, thereby providing a therapeutic advantage over NK3R antagonists. Because this work is still in its early stages, more thorough testing of the peripheral effects of compounds like EGX358 must be conducted to determine whether even highly selective ERβ agonists could be contraindicated in women with certain conditions in which ER activation may contribute negatively, like endometriosis and some cancers (e.g., lung, pancreatic, bladder) ( Chen et al., 2022 ). Although our selective synthetic ERβ agonists are related structurally to endogenous E2, they are not steroidal estrogens from which they differ in critical ways that heighten ERβ selectivity and should minimize adverse side effects in women with and without these conditions. We encourage the field to be open to the development of such compounds to provide the symptom relief that menopausal persons urgently need and deserve.

For much of human history, life expectancy at birth was less than 50 years. Thus, relatively few women experienced the massive loss of circulating estrogens that is a defining characteristic of menopause (e.g., Bellantoni and Blackman, 1996 ). In the United States, women’s life expectancy surpassed 50 years of age in 1902 and has increased gradually ever since ( Arias and Xu, 2020 ), reaching 80.2 years of age in 2022 ( Kochanek et al., 2024 ). Thus, within just over a century, the fraction of their lives that menstruating individuals in the U.S. can expect to live beyond childbearing has increased from relatively few years to nearly a third of their lives. As such, many more people are living through the menopausal transition, as well as an extremely extended postmenopausal period. More than half of these individuals will experience at least one of myriad menopausal symptoms, including hot flashes, night sweats, sleep disturbances, mood swings, vaginal dryness, weight gain, fatigue, joint and muscle pain, and the deficits in memory, attention, and concentration referred to as brain fog ( El Khoudary et al., 2019 ; Santoro et al., 2015 ; Williams et al., 2007 ; Woods and Mitchell, 2011 ). Symptoms like hot flashes may last up to a decade and are associated with increased long-term risks of cardiovascular disease and stroke, hypertension, mood disorders, cognitive dysfunction, and Alzheimer’s disease ( Chêne et al., 2015 ; Launer et al., 1999 ; Pinkerton, 2020 ; Santoro et al., 2015 ). Although estrogen therapy can effectively treat many menopausal symptoms ( Faubion et al., 2022 ), the 2002 publication of data from the Women’s Health Initiative (WHI) clinical trial showing increased risks of cancer and stroke in women over age 65 taking estrogen-based hormone therapies ( Rossouw et al., 2002 ) caused estrogen use to plummet ( Ettinger et al., 2012 ; Steinkellner et al., 2012 ), particularly among the symptomatic perimenopausal women who could most benefit from these treatments ( Crawford et al., 2018 ; Weissfeld et al., 2018 ). Not surprisingly, menopausal symptoms can have profoundly negative impacts on health and livelihood; women whose menopausal symptoms go untreated incur deteriorated quality of life, increased anxiety, depression, and stress, greater healthcare costs, more physician visits, and lower work productivity relative to those who use estrogen therapy or are asymptomatic ( Faubion et al., 2023 ; Hooper et al., 2022 ; Mueller et al., 2024 ; Tang et al., 2018 ). In addition to individual impacts, a 2023 study of women aged 45–60 years estimated that the cost to the U.S. of lost workdays due to menopausal symptoms is approximately $1.8 billion ( Faubion et al., 2023 ). Thus, safely and effectively treating menopausal symptoms would reduce personal suffering, improve quality of life, and provide economic benefits for menopausal people and their communities. Unfortunately, the development of therapies to address this need has profoundly lagged demand. Indeed, very few therapeutic advances have been made since estrogens were first isolated from ovarian tissue in the early 1900s ( Allen and Doisy, 1923 ; Allen et al., 1924 ; Doisy et al., 1930 ; Thayer et al., 1931 ; Veler et al., 1930 ). Since then, drug development for menopause in the U.S. has focused largely on various estrogens, as illustrated by currently available treatments. As of 2024, the website for the U.S. Food and Drug Administration (FDA) lists 36 approved medicines for menopause ( FDA Office of Women’s Health, 2017 ); of these, the primary constituent for 33 treatments is some form of estrogen (with or without a progestin), two include a progestin only, and one includes an estrogen and a drug with mixed agonist and antagonist activities at multiple estrogen receptors. The FDA’s website also lists numerous “serious” side effects of menopausal estrogen therapy, including increased risks of blood clots, heart attack, stroke, breast cancer, endometrial cancer (for estrogen-only treatments in those with a uterus), gallbladder and liver problems, high blood pressure, severe allergic reactions, and dementia (for some women over 65 years of age). As such, hormone therapy is not recommended for individuals with a history of vaginal bleeding, bleeding disorders, breast or uterine cancer, blood clots, stroke, heart attack, liver disease, or allergic reactions to hormonal medications. According to the United Nations, the number of women in the world aged 45–54 topped 466 million in 2023 ( UN World Population Prospects, 2024 ). In the U.S. alone, the total number of women aged 45–54 was greater than 20 million in 2023 ( United States Census Bureau, 2023 ). Thus, millions of women in the U.S. and globally are left with few options for symptom relief as they move through the menopausal transition. A more ideal solution would be the development of treatments that provide the benefits of estrogens for symptom relief while minimizing the occurrence of harmful side effects. In this review, we propose that this delicate balance may be achieved with highly selective estrogen receptor (ER) agonists targeting ERbeta (ERß). Of the two canonical ERs, ERalpha (ERα) and ERß, ERα activation promotes the growth and progression of breast, ovarian, and uterine cancer ( Ali and Coombes, 2000 ; Jia et al., 2015 ; Péqueux et al., 2012 ; Yu et al., 2022 ), whereas ERß inhibits breast, endometrial, and ovarian cancer cell proliferation ( Hapangama et al., 2015 ; Hartman et al., 2009 ; Jia et al., 2015 ; Lazennec et al., 2001 ; Yu et al., 2022 ). For this and other reasons to be discussed in this review, ERß would appear the superior target for menopausal therapy. Although so-called “selective estrogen receptor modulators” (SERMs) like tamoxifen and raloxifene are prescribed for conditions such as ER+ breast cancer, these drugs are not truly selective for one ER over the other. Tamoxifen and raloxifene are pharmacologically complicated drugs that bind to both ERs, acting as estrogen antagonists in some tissues (e.g., breast, uterus) and agonists in others (e.g., bone for raloxifene, uterus for tamoxifen) depending on the relative concentrations of ERs in each tissue ( Escande et al., 2006 ; Traboulsi et al., 2017 ; Weatherman et al., 2001 ; Zhao et al., 2005 ). Consequently, they are SERMs by virtue of tissue selectivity rather than receptor selectivity. To date, no truly ER-selective compounds are available to provide menopausal symptom relief. However, as will be discussed in more detail below, our group has developed new ER-specific compounds that are highly potent and selective for ERß over ERα ( Hanson et al., 2018 ; Sampathi Perera et al., 2018 ; Wetzel et al., 2020 ; Wetzel et al., 2022 ), thus providing potential new avenues for symptom relief. This article will begin with a historical perspective on menopausal hormone therapy, with a primary focus on Premarin ® and related conjugated estrogen products like Duavee ® due to their traditional ubiquity as treatments for menopause. This section is followed by overviews of estrogen and ERß effects on memory and vasomotor function and the resulting impacts of menopause on these domains. We then highlight the results of preclinical studies showing that a highly selective ERß agonist can facilitate memory formation and benefit vasomotor function in mouse models of ovarian hormone loss and Alzheimer’s disease. Based on this work, we argue for the development of highly potent and selective ERß agonists as treatments to address menopausal symptoms. Before we begin, however, we should acknowledge several important points. First, because ERs are localized to nearly every tissue in the body, ovarian aging leads to a multitude of heterogeneous symptoms whose incidence varies from person to person. Thus, a one-size-fits-all therapeutic strategy should be eschewed in favor of an arsenal of treatments that give menopausal persons multiple options to treat their specific symptom array. Hence, selective ERß agonism could be one, but not the only, method of symptom relief. Second, because addressing every possible symptom of menopause is beyond the scope of a single review, we will focus here on memory dysfunction and hot flashes, given that these symptoms are not only common in menopause but are also interrelated ( El Khoudary et al., 2019 ; Maki and Jaff, 2022 ; Maki et al., 2020 ). Finally, although we will often use the term “women” to describe the population affected by menopause, especially in reference to data collected in subjects identified as women, we recognize that not all those who undergo menopause identify as a woman and so will use gender neutral terms like “individuals” and “people” or “persons” where possible.