Girl Power in Glaucoma: The Role of Estrogen in Primary Open Angle Glaucoma

Abstract

Estrogen is essential in maintaining various physiological features in women, and a decline in estrogen levels are known to give rise to numerous unfortunate symptoms associated with menopause. To alleviate these symptoms hormone replacement therapy with estrogen is often used, and has been shown to be fruitful in improving quality of life in women suffering from postmenopausal discomforts. An often forgotten condition associated with menopause is the optic nerve disorder, glaucoma. Thus, estrogen may also have an impact in maintaining the retinal ganglion cells (RGCs), which make up the optic nerve, thereby preventing glaucomatous neurodegeneration. This review aims to provide an overview of possible associations of estrogen and the glaucoma subtype, primary open-angle glaucoma (POAG), by evaluating the current literature through a PubMed-based literature search. Multiple in vitro and in vivo studies of RGC protection, as well as clinical and epidemiological data concerning the well-defined retinal neurodegenerative disorder POAG have been reviewed. Over all, deficiencies in retinal estrogen may potentially instigate RGC loss, visual disability, and eventual blindness. Estrogen replacement therapy may therefore be a beneficial future treatment. However, more studies are needed to confirm the relevance of estrogen in glaucoma prevention.

Introduction

Glaucoma is an ocular progressive neurodegenerative disease, characterized by the death of retinal ganglion cells (RGCs), which constitute the innermost retinal neurons. The gradual loss of RGCs and their axons often results in progressive blindness if the disease is left untreated. Glaucoma is the leading cause of irreversible visual disability and the number of people suffering from glaucoma is estimated to affect approximately 111.8 million by 2040 (Kolko et al. 2015; Tham et al. 2014).

Although all clinical cases with glaucoma display a progressive loss of the RGCs, the disease is clinically divided into various subgroups. The subgrouping is dependent on the intra ocular pressure (IOP), anatomy of the drainage system, and on whether the condition is primary, secondary, or congenital. The most common subtype is primary open-angle glaucoma (POAG) accommodating for approximately 74% of the cases (Quigley and Broman 2006). In this subtype, clinicians distinguish between cases with extensive neurodegeneration combined with increased IOP, classified as high-tension glaucoma (HTG), and cases with extensive neurodegeneration but without any significantly increased IOP, denoted as low-tension glaucoma (LTG) or normal tension glaucoma (NTG). NTG accommodates for approximately 50% of the patients with POAG, thereby supporting the thought of other causal factors for glaucoma besides increased IOP.

Although numerous risk factors have been associated with the development of glaucoma e.g., inheritance, aging (Mallick et al. 2016), and elevated intraocular pressure (IOP), the IOP is the only treatable risk factor (Guedes et al. 2011; Monemi et al. 2005; Stone et al. 1997). In the specific case of NTG, other risk factors have also been described, such as population-based associations with the Japanese population being the most prevalent population with POAG (Iwase et al. 2004). Other risk factors include vascular risk factors predisposing to NTG development such as decreased retrobulbar circulation, increased vascular resistance in ocular vessels (Yamazaki and Drance 1997), decreased blood flow velocity in the central retinal vein (Abegão Pinto et al. 2013), and decreased ocular perfusion pressure (Choi et al. 2007). Speculatively, these changes may be a result of impairment in vascular autoregulation described in patients with NTG (Riccadonna et al. 2003). Further reports connect NTG to other vascular pathologies such as migraine (Phelps and Corbett 1985), Alzheimer’s disease (Chen et al. 2018), tinnitus, Raynaud’s syndrome (Wiermann et al. 2007), as well as nocturnal hypotension (Charlson et al. 2014) and obstructive sleep apnea/hypopnea Syndrome (Lin et al. 2011).

The use of IOP-lowering eye drops, as well as trabeculoplasty and surgery are commonly known to reduce the rate of progression. However, chances of an actual visual rescue are questionable (Broman et al. 2008; Heijl et al. 2002; Caprioli et al. 2016). Thus, recent research has put emphasis on other targets to treat than IOP.

In this context, epidemiological approaches to estimate gender-specific over-representation of POAG have been taking into consideration. Yet, conflicting evidence exists regarding the gender distribution of POAG. Several studies have shown that the prevalence of glaucoma is higher in women (Bengtsson 1981; Coffey et al. 1993; Mitchell et al. 1996; Wilson 2002), whereas other studies have shown that men are more likely to develop the pathology (Dielemans et al. 1994; Khachatryan et al. 2019; Leske et al. 1994; Rudnicka et al. 2006), and other publications report no difference between male and females (Hollows and Graham 1966; Klein et al. 1992; Tielsch et al. 1991; Wensor et al. 1998). Nevertheless, increasing evidence has linked lower estrogen levels with enhanced risk of glaucoma (Feola et al. 2019; Hulsman et al. 2001; Jojua et al. 2005; Lee et al. 2003; Pasquale et al. 2007; Sator et al. 1998; Vajaranant et al. 2014). Hence, it is plausible that estrogen may have protective properties with regards to development of the disease. Systemically, estrogen is well known to maintain bone mass and cardiovascular function in females (Iorga et al. 2017; Maxim et al. 1995). Likewise, estrogen has previously been shown to be neuroprotective in various CNS injuries (Brann et al. 2007; Raghava et al. 2017). Similar protective effects of estrogen have been proposed to occur in the retina, and estrogen has also been mentioned as a future treatment strategy for management of glaucoma (Dewundara et al. 2016; Kolko 2015). In this matter, multiple in vitro studies have provided evidence for an estrogen-mediated neuronal retinal protection (Cao et al. 2003; Nuzzi et al. 2018). More specifically, estrogen has been shown to promote RGC survival by preserving retinal nerve fiber layer (Newman-Casey et al. 2014), increasing RGC proliferation (Revankar et al. 2005), and increasing RGC viability (Giordano et al. 2011; Hayashi et al. 2007; Nakazawa et al. 2006; Prokai-Tatrai et al. 2013), as well as being IOP lowering (Sator et al. 1998; Tint et al. 2010) and protecting outflow channels in the anterior chamber of the eye (Paterson and Miller 1963). Additionally, estrogen has been implicated in improving ocular blood flow (Toker et al. 2003a) which is particularly important in glaucoma patients since vascular dysregulation is a major risk factor for glaucomatous development (Choi and Kook 2015). These and other concepts of estrogen neuroprotection are discussed in detail in the present review.

Taking all the mentioned beneficial characteristics of estrogen into account, sex hormone pathways might be an interesting target to treat in the glaucoma subtype normal tension glaucoma, since NTG patients have an extensive neurodegeneration without a measurable IOP above normal range, thereby making current treatment less desirable (Giorgio et al. 2018). With the onset of menopause and decline of gonadal estrogen production, intracrine regulation mechanisms become a substitution for the estrogen loss (Labrie et al. 2005) making endogenous estrogen-related interactions a point of high interest. Intracrine synthesis of sex hormones takes place in local tissues in different parts of the body. In this way, steroids synthesized from dehydroepiandrosterone (DHEA) produced by adrenals induce their target of action and are degraded locally, which makes it complicated to measure the impact of intracrine hormones on their target tissue (Luu-The and Labrie 2010). As we argue in the present review, the eye and the retina, in particular, are hormone-sensitive entities, and thus, disturbances in intracrine sex hormone metabolism may be of particular interest in POAG and especially the NTG subtype. Yet the specific molecular mechanisms and mode of action are still to be determined. The present review aims to summarize both in vitro and in vivo studies of RGC protection, as well as clinical and epidemiological data concerning the well-defined retinal neurodegenerative disorder glaucoma i.e., POAG.

Method of Literature Search

The literature search was performed in PubMed. The following search words were used: glaucoma, estrogen, neuroprotection, genes, enzymes, intraocular pressure, RGC death, ocular flow. No limitations were made. Reviews were studied to create an overview and guidance for further search. The articles reviewed were those published in English, although no language preference was intended. The abstracts in English were read for articles of relevance in other languages. Figure 1 summarizes the included articles.

Fig. 1.

Flow chart of studies included in the review. Initial search using two different search strategies resulted in 115 studies by which duplicates were removed leading to the selection of 57 original research articles. These were further screened for relevance by two independent researchers resulting in final inclusion of 37 original articles in the review. Additional material was included as supplementary articles (n = 123)

Changes in Estrogen Affects Optic Disc Morphology and Risk of Glaucoma in Humans

Various studies have linked the effects of menstrual cycle with general eye health (Akar et al. 2005a, 2004, 2005b; Hulsman et al. 2001; Jojua et al. 2005; Lee et al. 2003; Nirmalan et al. 2004; Vajaranant et al. 2014). As an example of this, a recent study revealed that the optic disc parameters of normally menstruating patients with non-proliferative diabetic retinopathy (NPDR) change significantly throughout the menstrual cycle (Akar et al. 2005b). Hence, women with severe NPDR had a significant increase in neuroretinal rim area and a significant decrease in cup-shape measure, linear cup/disc ratio, cup/disc area ratio, and cup area. The observed changes took place in the late luteal phase of the cycle, thereby being potentially connected to decreased levels of estradiol and increased levels of progesterone detected in the patients at that time. In addition, NPDR patients also experienced significant retinal sensitivity loss in the luteal phase (Akar et al. 2005a). Similarly, another study revealed that healthy women were also affected by cycle-dependent changes of the optic disc topography (Akar et al. 2004).

Besides being influenced by different stages of the menstrual cycle, ocular health may also be associated with fluctuations in hormonal levels in general.

In this regard, sex hormone concentrations do not only change during the menstrual cycle, yet have also been shown to change in patients with open-angle glaucoma (OAG) in an age-related pattern (Jojua et al. 2005). In this case, a study based on women between 45 and 55 years of age showed lower estradiol levels, followed by a decrease in luteinizing hormone (LH) levels in a cohort of 55–65 individuals, which further substantiates a possible link between OAG and estrogen levels (Jojua et al. 2005).

Additionally, moderately increased risk of developing POAG has been shown to be associated with bilateral oophorectomy, late age at menarche (> 13), increased parity and early age at menopause (< 45) (Hulsman et al. 2001; Lee et al. 2003; Vajaranant et al. 2014). This indicates that a sudden decrease of estrogen, as well as life time estrogen concentrations may have implications for optic nerve health.

Moreover, numerous studies have suggested a neuroprotective property of estrogen in the context of glaucoma. However, these studies show conflicting evidence ranging from postulations that female reproductive factors do not influence OAG development at all (Nirmalan et al. 2004) to the fact that early withdrawal from estrogen might be a significant risk factor (Hulsman et al. 2001; Lee et al. 2003; Vajaranant et al. 2014). Nevertheless, these differences might be present because of different population size, origin, and study conduct. For example, one study was conducted on 5150 patients from rural areas of South India and only roughly 36% of the included subjects were postmenopausal at the time of study (Nirmalan et al. 2004). Additionally, only 20 cases of OAG have been recorded among the postmenopausal cohort which, according to the authors, makes the ability to look at the OAG prevalence limited. Furthermore, increased OAG prevalence in the study in question has indeed been found to be associated with earlier age at natural menopause and shorter duration of exposure to endogenous estrogen, which is also in line with other studies (Hulsman et al. 2001; Lee et al. 2003; Vajaranant et al. 2014). Yet because of the small amount of registered cases, the findings have been deemed to be statistically insignificant and warranted the study non-applicable in the association of OAG prevalence to female reproductive factors.

Regardless, in summary, varying estrogen levels may affect optic disc morphology in a negative manner associated with an epidemiology-based increased risk for development of glaucoma in humans.

Below are examples of female sex hormone related therapies for menopause and contraception and their association with POAG, which further highlights a possible protective effect of female sex hormones.

Hormone Replacement Therapy (HRT)

Increasing interest regarding HRT and its effect on the development of glaucoma has arisen throughout the past decade (Abramov et al. 2005; Battaglia et al. 2004; Harris-Yitzhak et al. 2000; Na et al. 2014; Newman-Casey et al. 2014; Pasquale et al. 2007; Paterson and Miller 1963; Sator et al. 1998; Tint et al. 2010; Toker et al. 2003b; Vajaranant et al. 2014, 2018). Yet it is apparent that no consensus exists concerning its effect on POAG onset and several studies show ambivalent results.

In a study performed on 1044 women whom had undergone bilateral oophorectomy prior to menopause compared with 1070 controls, it was shown that the women who underwent bilateral oophorectomy had increased risk of development of POAG compared with the control group (Vajaranant et al. 2014). However, 11% of women in the study received HRT which did not reduce the association between internal hormonal deficiency and glaucoma development (Vajaranant et al. 2014). Another study by Vajaranant et al., on the other hand, specified that the benefits of HRT could be linked to the specific hormones used for therapy (Vajaranant et al. 2018). Commonly used drug combinations are estradiol combined with levonorgestrel (synthetic progesterone). This particular study identified that a reduction of risk of POAG in African American women was correlated with HRT-containing estrogen and not combined estrogen/progesterone (Vajaranant et al. 2018).

These findings have further been validated by a retrospective cohort analysis verifying that only HRT-containing estrogen had an effect on reducing the risk of POAG compared with combined estrogen/progesterone treatment as well as estrogen/androgen treatment (Newman-Casey et al. 2014). In addition, HRT with estrogen has also been aligned with other modalities to investigate glaucoma, such as evaluation of nerve fiber thickness and IOP measurements. In this regard, it was revealed that HRT with estrogen protected against retinal neuronal fiber damage (Na et al. 2014), while another study revealed IOP-lowering effects to be lower in patients treated with HRT compared to a non-treated cohort (Tint et al. 2010).

The IOP-lowering effect was shown to be unaffected by combined HRT, implying that patients receiving estrogen and progestin (synthetic progesterone) had a decreased risk of glaucoma characterized by IOP above 21 mmHg prior to visual field loss (Pasquale et al. 2007). Additional reductions in IOP from 16–20 mmHg to 13–15 mmHg over the 12-week course of HRT have also been presented in a case study consisting of a 56-year-old woman (Sator et al. 1998). A plausible explanation for these IOP-lowering results might be that female sex hormones may act as protectors towards anterior chamber outflow channels (Paterson and Miller 1963). In this way, the drainage of the fluid remains sufficient which counteracts the elevation of IOP through fluid accumulation.

Results of studies by Abramov et al. (2005) and Toker et al. (2003b) are contradicting the above by showing that there is no immediate connection between either life-time exposure to progesterone and estrogen or estradiol serum levels and IOP. This might, in part, be explained by the fact that the study by Abramov et al. (2005) was initially powered to detect differences in the IOP of 13% and more. Thus, smaller differences in this parameter might not have been detected. Furthermore, the difference in the mean IOP between HRT and non-HRT group of the study (Toker et al. 2003b) has, in fact, been noticed, but was not statistically significant. Interestingly, high levels of postmenopausal testosterone, however, seem to have an association with IOP elevation (Toker et al. 2003b), indicating that deviations in hormonal balances are not only correlated with female sex hormones, but also male sex hormones.

Among other instigating parameters for glaucoma, HRT has been sought to affect plasma viscosity and ocular flow beneficially in patients with chronic POAG and NTG (Battaglia et al. 2004). HRT has been shown to reduce vascular resistance distal to the ophthalmic artery in postmenopausal women matching the index to young women (Harris-Yitzhak et al. 2000), indicating that HRT appears to be fruitful in terms of improving ocular blood flow. These findings are also consistent with the fact that HRT with estrogen causes vasodilation and increases blood flow in major peripheral arteries (Harris-Yitzhak et al. 2000).

Taken together, HRT with estrogen has a positive impact in protecting against glaucoma. Thus, HRT may be worth recommending to postmenopausal women with a higher risk of glaucoma despite the lack of obvious systemic postmenopausal symptoms.

Oral Contraceptives (OC)

The use of OC has also been connected to the development of POAG (Abramov et al. 2005; Pasquale and Kang 2011; Wang et al. 2016). In line with this, women taking OC for more than three years have been revealed to have an increased risk of self-reported POAG evaluated through health care questionnaires in the USA (Wang et al. 2016).

Likewise, an OC treatment for more than five years has also shown a 25% increased risk of POAG (Pasquale and Kang 2011). However, the mechanism behind the increased risk of glaucoma in response to OC seems to be independent of influence on IOP levels (Abramov et al. 2005). Although speculative, these findings may be explained by the previous presented studies in the “Hormon replacement therapy (HRT)” paragraph in which numerous studies showed that the addition of progesterone to treatment counteracted possible protective effects of estrogen.

To conclude, evidence regarding the female sex hormones relation to POAG development seems controversial. Albeit estrogen has been shown to be predominantly protective, other hormones used in replacement therapy and oral contraceptives may deemingly increase the risk of glaucoma. Thus, more randomized control trials are needed to elucidate such relationship.

Molecular and Genetic Factors Connected to the Role of Estrogen in POAG

The Genetic Association of Estrogen and POAG

Accumulating studies have proven that genetics is one of the major contributing factors in glaucoma development (Bailey et al. 2018; Kosior-Jarecka et al. 2019; Mabuchi et al. 2010; Pasquale et al. 2013).

In this context, studies have reviewed estrogen and testosterone single-nucleotide polymorphisms (SNPs) and their relationship to POAG (Bailey et al. 2018; Pasquale et al. 2013). Pasquale et al. determined a genomic association from two large genetic glaucoma studies—GLAUGEN and NEIGHBOR (Pasquale et al. 2013). Furthermore, estrogen pathway-related SNPs were associated with a high-tension glaucoma (HTG) subtype of POAG among women, whereas no association was found between estrogen pathway SNPs and NTG (Pasquale et al. 2013).

A SNP set of Catechol-O-methyltransferase (COMT) is a possible driving factor of this association, as it showed particularly strong association with HTG in women.

In estrogen metabolism, COMT acts as a negative modulator of estrogen activity. The enzyme is involved in O-methylation of estrone and estradiol cathecols, which blocks the activity of estrogen. Furthermore, COMT blocks the following oxidation of the methylated estrogen E1/E2 cathecols to their quinone metabolites which reduces the formation of mutagenic DNA adducts and oxidative damage inflicted through redox cycling (Cavalieri and Rogan 2016; Lavigne et al. 2001; Wu et al. 2019; Yager 2015). Steroid sulfatase (STS), on the other hand, converts inactive estrogen sulfates into active estrogens, thus being a positive modulator of estrogenic activity (Garbacz et al. 2017). Interestingly, STS SNP in the estrogen pathway is the only one apart from COMT, associated with both NTG and HTG in women (Pasquale et al. 2013). These findings may imply the pro- and anti-estrogenic balance as one of the determining factors for POAG development in women.

It has been postulated that estrogen receptor (ESR) SNPs may determine the clinical phenotype of POAG patients, meaning POAG subtype distribution. Hence, ESR2 polymorphisms have been seen to correlate with elevated IOP in female patients with HTG (Mabuchi et al. 2010). Other evidence suggests that ESR2 polymorphisms are also related to higher prevalence of open-angle glaucoma (OAG) in men and the NTG subtype in women which generally is not characterized by elevated IOP (de Voogd et al. 2008; Pasquale et al. 2013). Contradictorily, a study by Kosior-Jarecka et al. did not establish any significant differences in frequency of ESR SNPs either between POAG compared to control groups or HTG and NTG (Kosior-Jarecka et al. 2019). Despite that, IOP levels were linked to SNPs in ESR1 and ESR2 in patients with HTG and NTG, respectively. Hence, HTG patients with ESR1 SNP had the lowest maximal IOP and NTG patients with ESR2 SNP experienced the lowest initial IOP.

It is, however, unclear, whether the SNPs in the mentioned stages of estrogen signaling correlate with POAG development in women due to immediate connection to their influence on estrogenic activity or through other RGC survival/death-specific mechanisms.

Overall, the relationship between genetics and glaucoma is seen on different levels of estrogen signaling and presents a clear sex-specific tendency. Further studies elucidating the genetic nature of the disease as well as the precise mechanisms in which the genes are influencing glaucoma pathophysiology are needed to complete the understanding.

The Metabolism of Estrogen Mediated by CYB1B1 May Influence the Development of POAG

The role of estrogen in POAG development may be observed in different stages of estrogen metabolism. CYP1B1, an enzyme from the cytochrome P450 family, is partially responsible for estrogen breakdown (Jansson et al. 2001), and linking CYP1B1 with the first ever identified POAG gene, the MYOC gene is of particular interest (Vincent et al. 2002).

Earlier studies have shown that GLC1A (an alias name for MYOC) is expressed in the human ciliary body, retina, and trabecular meshwork (Fingert et al. 1998; Kubota et al. 1997). The MYOC gene codes a mutant recombinant protein myocilin which has been shown to block perfusion in the anterior segment of the eye and, specifically, in the trabecular meshwork, thus leading to IOP elevation (Fautsch et al. 2000).

In 2002, mutations in CYP1B1 together with mutations of MYOC gene were shown to be implicated in digenic inheritance of Early-Onset Glaucoma (Vincent et al. 2002). This study presented a family with autosomal dominant glaucoma with segregation of both MYOC and CYP1B1 mutations with the disease and the carriers of both MYOC and CYP1B1 mutations had the earliest age of onset—around 27 years (Vincent et al. 2002). Hence, it has been proposed that CYP1B1 could be a modifier for the MYOC gene expression by means of an unknown common pathway (Vincent et al. 2002).

CYP1B1 has also been implicated in the development of Primary Congenital Glaucoma (Stoilov et al. 1997) and POAG (Acharya et al. 2006; Melki et al. 2004). The CYP1B1 is abundantly expressed in the human eye, particularly in the cornea, ciliary body, iris, and retina (Kaur et al. 2011). Yet no CYP1B1 expression was found in the trabecular meshwork (Kaur et al. 2011), indicating that CYB1B1 may alter IOP by other pathways and mechanisms.

In line with this, estrogen was established to be a mediator in the MYOC/CYP1B1 interaction (Mookherjee et al. 2012). Hence, both E2 and its biosynthesis pathway enzymes are present in the human trabecular meshwork and retinal pigment epithelium cells. Additionally, functional activity of estrogen responsive elements (EREs) has been identified in the MYOC promoter. In connection to this, human trabecular meshwork cells were treated with E2 for 24 h which led to a subsequent increase in endogenous myocilin showing that increased concentrations of E2 can upregulate MYOC in the human trabecular meshwork cells (Mookherjee et al. 2012). Thus, increased concentrations of estrogen in trabecular meshwork might actually lead to IOP elevation due to increased outflow resistance caused by accumulation of myocilin.

These findings describe a probable non-beneficial role of excessive amount of estrogen on anterior structures of the eye, which potentially may deviate from what is seen in the retina, where estrogen contradictorily has been shown to lower IOP by mechanisms independent of myocilin-related effects (Sator et al. 1998; Tint et al. 2010), as stated in the paragraph “Experimental Research in Estrogen Neuroprotection”. Additionally, it has been concluded that the pathologic state of the carrier of CYP1B1 mutations is determined by allelic states of the gene (Banerjee et al. 2016), which may imply fluctuations of individual risk of POAG development in response to allele variation.

17-Betahydroxysteroid Dehydrogenase Expression and Activity Regulate Aqueous Humor Flow

17-betahydroxysteroid dehydrogenases (17HSDs) is another enzyme class involved in biosynthesis and inactivation of sex hormones. The enzymes catalyze the hybrid transfer between 17β-hydroxy and 17-ketosteroid pairs in a positional and stereospecific manner. Since both estrogen and androgens possess their highest metabolic activity in their 17β-hydroxy form, 17HSD-enzymes are directly involved in regulating steroid hormone action (Peltoketo et al. 1998). In connection to this, a study (Coca-Prados et al. 2003) determined the presence of both α- and β-ER in cell lines from normal and glaucomatous cadaver eyes. Furthermore, the study established the presence of three members of 17HDS family (two, five and seven) in the human ciliary epithelium cell lines. The enzymes were highly responsive to both androgens and E2, and 17HDS7 acted as a regulator of intra/paracrine estrogenic environment (Coca-Prados et al. 2003).

The analyzed cells in the study by Coca-Prados et al. (2003) originated from the ciliary epithelium, a part of the ciliary body responsible for the synthesis of the aqueous humor (Delamere 2005). The relationship between the synthesis of the aqueous humor and its subsequent drainage through the anterior chamber structures of the eye is responsible for the IOP (Delamere 2005). Previous reports have connected IOP lowering with HRT in nonglaucomatous eyes (Tint et al. 2010) and in a single patient with POAG (Sator et al. 1998). As the ciliary body was previously shown to release factors of autocrine regulation that can control aqueous humor outflow (Coca-Prados et al. 1999), it is tempting to suggest that the highly specific expression of 17HDS subtypes in the ciliary epithelium may account for the decrease in IOP observed after taking HRT. In this regard, further research is required to validate the presence of this phenomenon in POAG (Fig. 2).

Fig. 2.

Intracrine hormonal synthesis in the retina. Retinal sex steroids are synthesized from dehydroepiandrosterone (DHEA) originating from the adrenal cortex. The present diagram illustrates enzymes relevant for intraocular hormonal synthesis. Androstenedione synthesized from DHEA acts as a progenitor to retinal androgens. The 17-ßHDS5 enzyme converts androstenedione into testosterone and is potentially implicated in intracrine hormonal synthesis in the retina and regulation of IOP via influence of aqueous humor outflow. Furthermore, mutations in the 17-ßHDS enzyme family have been associated with a higher prevalence of POAG in both sexes. Testosterone is converted into the end product of androgen transformation: DHT. Retinal androgens are the primary source of retinal estrogen. The molecular transformations from androgens into estrogens occur with the help of cytochrome system (CYP19A1). Both the cytochrome and 17-ßHDS systems are implicated in the retinal estrogen turnover. CYP1B1, taking part in E2 conversion into hydroxyestrogens, is particularly implicated in digenic inheritance of POAG in both sexes. SNPs of STS, an enzyme involved in bidirectional conversion of active and inactive estrogens, have been connected to both HTG and NTG subtypes of POAG in women. Additionally, SNPs of COMT, an enzyme involved in deactivation of hydroxyestrogens via their methylation, were associated with inhibition of estrogenecity. Therefore, the retina is an organ that exhibits diverse internal sex hormone turnover which potentially has to be taken into consideration, should local hormonal treatment be considered for the therapy of POAG in the future. ADHEA dehydroepiandrosterone, IOP intraocular presdsure, POAG primary open-angle glaucoma, CYP cytochrome. MYOC MYOC gene coding protein myocilin. COMT Catechol-O-methyltransferase. STS steroid sulfatase. SNP single-nucleotide polymorphism

Molecular Pathways Associated with Estrogen and POAG

Classical estrogen receptors (ERs) are abundant in the eye (Gupta et al. 2005; Wickham et al. 2000) at several cellular levels including RGCs (Jiang et al. 2019). It has been speculated that estrogen-mediated signaling leads to RGC protection by several different mechanisms (Jiang et al. 2019; Nakazawa et al. 2006; Revankar et al. 2005; Yu et al. 2004).

For example, activation of protein-kinase pathways such as GPR30 has been shown to be localized on endoplasmic reticulum and the binding of estrogen E2 to it results in mobilization of intracellular calcium and the PI3K/Akt pathway that sends a proliferative impulse (Revankar et al. 2005). In the eye, estrogen has recently been shown to prevent apoptosis via the same receptor and signaling pathway in a model of mouse RGC degeneration (Jiang et al. 2019). It has also been shown that estrogen (E2) possesses neuroprotective properties in axotomized RGCs increasing their viability through ERK-c-Fos signal transduction pathway (Nakazawa et al. 2006).

In addition, estrogen also counteracts direct neurotoxic stimuli through further biochemical pathways, independent of classical estrogen receptor signaling (Yu et al. 2004). A study by Yu et al. evaluated the neuroprotective effects of E2 in the retina against two promoters of cellular degeneration: hydrogen peroxide, H₂O₂, and light. Neuroprotective benefits of systemic administration of E2 were tested in vivo via electroretinogram of ovariectomized female rats prior to 24 h of constant light exposure followed by quantitative histology and TUNEL assay of their eyes. Animals pretreated with E2 showed a significant rescue of cells in outer nuclear layer, significantly reduced photoreceptor apoptosis verified by lesser amount of apoptotic DNA via TUNEL assay, and protected retinal function seen by ERG traces. Furthermore, it has been shown that photoreceptor cell lines pretreated for 30 min with E2 prior to H₂O₂ exposure demonstrated significantly lesser quantity of apoptotic DNA (16% TUNEL-positive cells compared to 53% in the group without estrogen). According to the study, this effect is independent of estrogen receptor signaling and involves the insulin/PI3K/Akt cascade where activation of the retinal insulin receptors β is involved in mediating the neuroprotective effects of estrogen (Yu et al. 2004).

An interesting point of view on estrogen-mediated neuroprotection originates from research in phytoestrogens. Phytoestrogens are plant-derived phenolic compounds chemically resembling endogenous estrogen, which are commonly found in botanical species such as soy, red clover, kudzu, hops, licorice, rhubarb, yam, chasteberry, dong quai, and black cohosh (Hajirahimkhan et al. 2013). In this regard, a study has both confirmed the E2 neuroprotective properties and investigated those of plant phenols (Ondricek et al. 2012). The study used the RGC-5 model with iodoacetic acid (IAA) as a cytotoxic stimulus that insults glycolytic metabolism and, thus, subsequent metabolism in the mitochondria contributing to an increase in the production of reactive oxygen species (ROS). Not only in the presence of E2, but also in the presence of phytoestrogen, RGC viability increased and ROS production decreased.

Furthermore, it has been shown that studied flavonoids are effective in preventing IAA-induced cell damage and decreasing ROS production. Some of them also had different effects on phosphorylation of kinases involved in cellular signaling and apoptosis (Akt, MAPK, RSK). An additional study has also established that the short wavelength cone-mediated response was improved in postmenopausal women consuming phytoestrogens compared to those not consuming them (Eisner and Demirel 2011). Speculatively, this might also be the case concerning RGCs. This suggests that women who more frequently consume phytoestrogen-rich foods (e.g., in Asian cultures) might be subject to different physiological outcomes depending on the flavonoid spectrum of the food consumed. This might also be correlated with the fact that POAG prevalence in Asia is one of the lowest in the world (especially in East-Asia region) (Kapetanakis et al. 2016).

Overall, the discussed findings suggest that estrogen might be neuroprotective in RGCs by acting through different pathways primarily reducing RGC apoptosis and enforcing proliferative signaling. This is also relevant for glaucoma, since apoptosis is a mechanism of cellular death in POAG (Kerrigan et al. 1997).

Experimental Research in Estrogen-Mediated Neuroprotection

The Effect of Estrogen in Experimental Models of Glaucoma

Approaches of basic in vivo and in vitro research may also shed light on the link between glaucoma and estrogen. In this matter, exacerbation of visual dysfunction has been noticed in an animal model of experimental glaucoma and menopause (Feola et al. 2019). In the study, rats were subjected to ovariectomy followed by induction of unilateral ocular hypertension. After comparing visual acuity and retinal structure of menopausal rats to control, it was seen that an estrogen deficiency may intensify visual defects caused by ocular hypertension.

Additionally, daily treatment with E2 of mice with surgically elevated IOP resulted in significant neuroprotective benefits and prevented loss of visual function (Prokai-Tatrai et al. 2013). Neuroprotection in this case was measured as a decreased number of apoptotic RGCs compared to the control group. The study also reported systemic adverse events caused by high circulating estrogen levels characterized by significant uterothropic effect in the treatment group which may induce endometrial cancer. This might have to be taken into consideration in the future, should estrogen-based therapies for glaucoma be developed.

Estrogen and Excitotoxicity

Retinal excitotoxicity is a common cause for retinal neurodegeneration, and it is postulated that estrogen may be beneficial in preventing excitotoxic damage, thereby promoting retinal neuroprotection in different kinds of retinal neurons. Hence, a study confirmed a beneficial effect of non-feminizing estrogens (incl. E2) on photoreceptors (Nixon and Simpkins 2012). Here, the tested compounds were applied to a cone photoreceptor cell line insulted with 5 mM glutamate and E2 rescued the studied cell population. Even though photoreceptors are not implicated in glaucoma pathogenesis, various retinal neurons could share similar responses in estrogen neuroprotection. Hence, similar research was conducted concerning effects of non-feminizing estrogen analogs and E2 on RGCs against glutamate-induced excitotoxicity in rats (Kumar et al. 2005), in which E2 along with the analogs protected cells in RGC-5 cell death model against excitotoxicity (Kumar et al. 2005). It has also been suggested that estrogen under these circumstances affects neuroprotection independently of the classical estrogen receptor pathway. Hayashi et al. sheds light on one of the mechanisms behind E2 neuroprotection in NMDA injury by proving that the protective effect is exerted through p-ERK signaling pathway (Hayashi et al. 2007). This evidence in animals is consistent with results in human studies showing that ERK-c-Fos pathway might be involved (Nakazawa et al. 2006).

Single intravitreal injection of NMDA in rats is often used as model for experimental glaucoma (Yamashita et al. 2011). Full-field electroretinograms (ERGs) and analyses of transverse sections of the retina have been used to assess the influence of E2 on retinal function in these models (Yamashita et al. 2011). Here, the study showed that rats subjected to combined NMDA and E2 had less aggravated amplitudes of the ERGs compared to NMDA group. Therefore, the study established estrogen as a neuroprotective against glutamate excitotoxicity. There was, however, no morphological differences between the groups, indicating that the protective effects of E2 occurred in a functional improvement rather than prohibiting structural damage. These findings are supported by Russo et al., which showed that systemic administration of E2 30 min before an ischemic insult minimized elevation of glutamate levels in rat (Russo et al. 2008). The study also showed significant reduction in RGC loss in rats with elevated IOP.

Another mechanism of E2 neuroprotection against NMDA insult may be reduction of 14-3-3 zeta protein phosphorylation (Koseki et al. 2012). 14-3-3 is a family of proteins involved in mediation of essential anti-apoptotic signaling, and high levels of the zeta 14-3-3 isoform have been previously found in the retina (Ivanov et al. 2006; Masters and Fu 2001). In the study by Koseki et al. phosphorylation of 14-3-3 zeta in rat RGCs were significantly depleted after NMDA insult, but subsequent E2 implantation significantly prevented the decrease in the protein phosphorylation (Koseki et al. 2012). Thus, levels of 14-3-3 zeta may be of importance in retinal neurotoxicity.

Taken together, it is tempting to suggest that estrogen may have an essential role in protecting against glutamate excitotoxicity in the retina, which may prevent the development of retinal neurodegeneration linked to glaucoma.

Estrogen and Retinal Energy Metabolism

Among other tissues in the human body, the retina has one of the highest energy requirements (Wong-Riley 2010). RGCs are almost completely dependent on the energy yielded by oxidative metabolism of glucose, and more controversially, lactate, in mitochondria (Hertz et al. 2007; Vohra et al. 2019). In line with this, various factors may threaten mitochondrial function and, thus, compromise retinal energy metabolism leading to retinal pathology.

In glaucoma, a disease with a predominantly late-onset (an age of 40+) mitochondrial dysfunction is a major risk factor, since it is known to increase with aging (Chistiakov et al. 2014; Quigley 2011). Among others, mitochondrial health is closely connected to oxidative stress triggered by excessive ROS formation in malfunctioning mitochondria (Zorov et al. 2014). In the state of oxidative stress, cellular membranes are damaged by ROS, thus being exposed to a direct cyctotoxic stimulus, which is counteracted by the presence of antioxidants.

Antioxidants are chemical entities that can alleviate oxidative stress. Estrogen has been established to exert antioxidant activity by several means (Behl et al. 1997; Giordano et al. 2011; Manthey and Behl 2006; Simpkins et al. 2008; Wang et al. 2003; Yang et al. 2004).

Firstly, estrogen’s activity to protect neurons against oxidative stress has been correlated with its molecular structure—the presence of a phenolic hydroxyle group in the C3 position of the A steroid ring (Behl et al. 1997). In phenolic antioxidants, the hydroxyle reduces the rates of oxidation by transferring the hydrogen atom to free radicals and this mechanism in estrogen is not dependent on ER activation (Behl et al. 1997; Foti 2007; Manthey and Behl 2006).

Secondly, estrogen’s antioxidant properties have also been connected to its direct effect via binding to estrogen receptors alpha and beta. A variety of genes implicated in antioxidant defense; neuronal survival and synaptic plasticity have been upregulated in a human neuroblastoma cell line transfected with ER-α (Manthey and Behl 2006).

Furthermore, it has been shown that estrogen binding sites including ER-β are also present in mitochondria and estrogen treatment of human lens epithelial cells resulted in a dose-dependent protective effect on ATP production and cell viability (Simpkins et al. 2008; Wang et al. 2003; Yang et al. 2004). Additionally, a study has evaluated E2 effect on RGC mitochondria from patients with Leber’s hereditary optic neuropathy (LHON) (Giordano et al. 2011). In the study, E2 was shown to decrease levels of ROS in the affected cells, induce activity of superoxide dismutase 2 antioxidant enzyme, increase RGC viability, stabilize mitochondrial membrane potential, induce mitochondrial biogenesis, and enhance energy competence of the cells (Giordano et al. 2011). In comparison, an earlier study by Inagaki et al. established the presence of rare LHON-related in POAG patient population (Inagaki et al. 2006) linking mitochondrial dysfunction in LHON with POAG. Additionally, changes in mitochondrial DNA and nuclear DNA genes encoding mitochondrial proteins may influence mitochondrial function in POAG and, thus, contribute to its development (Lascaratos et al. 2012). Therefore, although speculative, the findings of Giordano et al. may be partially applicable to the pathophysiological processes taking place in glaucomatous RGCs (Giordano et al. 2011).

Overall, estrogen seems to be protective in terms of preventing mitochondrial damage and sustaining energy production. Considering the high energy requirements in retinal cells, it is tempting to hypothesize that estrogen has a critical role in preserving RGCs and preventing glaucomatous damage.

Estrogen and Ocular Blood Flow

Vascular dysfunction is a known risk factor for development of glaucoma (Choi and Kook 2015) and various studies have connected inadequate systemic and ocular circulation to the development of both POAG (Feke and Pasquale 2008; Gherghel et al. 2007; Pemp et al. 2009; Su et al. 2008) and NTG (Gasser and Flammer 1991; Oettli et al. 2011; Su et al. 2008). The specific relationship of ocular blood flow and POAG has been thoroughly reviewed in detail (Chan et al. 2017; Nakazawa 2016; Patel et al. 2018), thus this paragraph aims to summarize key points of ocular blood flow in POAG linking it to estrogen.

In POAG and NTG, increased variability in nocturnal systemic blood pressure, increased systemic arterial thickness and intima media thickness have been determined compared to control groups (Mroczkowska et al. 2013). Additionally, POAG has been associated with coronary artery bypass and vascular surgery (Topouzis et al. 2011), which implies its association with coronary artery disease. In this regard, systemic estrogen has been shown to influence acetylcholine coronary response in atherosclerotic coronary arteries and be protective against coronary artery disease in women, but not in men (Collins et al. 1995).

Similarly, increased variability in systemic blood pressure has been associated with glaucomatous neurodegeneration. While some studies report no significant differences in systemic blood pressure monitored throughout a 24-h period (Gherghel et al. 2010; Kim et al. 2010; Riccadonna et al. 2003), other studies emphasize nocturnal hypotensive episodes as potentially contributing to the disease progression (Graham et al. 1995; Hayreh et al. 1994; Kaiser et al. 1993; Tokunaga et al. 2004). A possible connection between glaucoma progression and nocturnal blood pressure dipping may be hypothesized due to its relationship with ocular perfusion pressure (OPP). It has been shown that nocturnal hypotension was correlated with mean OPP fluctuations in NTG patients, which in itself is a risk factor for NTG development (Choi et al. 2006). Furthermore, a study has shown different associations between OPP and retinal circulations in males and females (Tobe et al. 2012). In this particular study, females were shown to have a positive linear association between retinal microcirculation and OPP, whereas males exhibited a negative correlation, implying gender-dependent vascular autoregulation in POAG (Tobe et al. 2012). In the context of blood pressure regulation, E2 was also shown to be a potent vasodilator (Caulin-Glaser et al. 1997). Yet, despite its vasodilating activity, a large prospective analysis of 619 postmenopausal women revealed a correlation between hypertension and increased endogenous E2 levels which was eliminated after adjustment for BMI (Wang et al. 2012). In this regard, metabolic status of POAG patients is worth taking into consideration, as the incidence of POAG was previously correlated with obesity (Jung et al. 2020) and Metabolic Syndrome (Kim et al. 2016).

In the case of ophthalmic disorders, it is not only evident that there is a difference in how gonadal hormones influence their course in males and females, but also that this difference may be attributed to the ocular blood flow (Eisner 2015; Schmidl et al. 2015). Evidently, sex-related differences in ocular blood flow can be observed even in healthy subjects. A study evaluated sex-related differences in optic nerve head and choroidal circulation via Laser Speckle Flowgraphy and found out that the mean blur rate (MBR) in the optic nerve head in females was significantly higher than in males (Yanagida et al. 2015). The MBR is a measure of blood flow velocity in the method (Luft et al. 2016), therefore the findings of Yanagida et al. (2015) indicate higher blood flow velocities in the optic discs in healthy females. Another point of interest is that aging affects the retrobulbar ocular blood flow both in male and female in a manner comparable to glaucoma or age-related macular degeneration suggesting that vascular dysregulation gains even a higher significance as a risk factor for POAG with age (Harris et al. 2000).

Additionally, discrepancies between ocular blood flow between pre- and postmenopausal women have also been reported. One study examined flow velocities and resistive index in central retinal and ophthalmic arteries and concluded that not only higher blood concentrations of estrogen correlate positively with these parameters, but also that testosterone acts as an antagonist to this improvement (Toker et al. 2003a). However, a study by Siesky et al. (2008) showed that there was no significant difference between pre- and postmenopausal women in terms of general ocular hemodynamics.

Nonetheless, estrogen has been shown to improve ocular blood flow by having a vasodilatory effect (Faria et al. 2011) and improving blood flow in the inferotemporal retinal artery consequently improving optic nerve fiber layer parameters in the inferotemporal region of the optic nerve head (Deschênes et al. 2010). The possible explanation for the vasodilating properties of estrogen might lie in its influence on NO production. NO is a crucial vasodilator for maintaining retinal autoregulation. Hence, in one in vitro study E2 induced a rise in NO levels in cerebral and peripheral vascular endothelium via eNOS activation and E2 receptor mediated signaling (Nevzati et al. 2015).

In summary, estrogen affects blood flow in a positive manner through improved perfusion through mainly vasodilation, which is predominantly seen in women. Enhanced ocular blood flow may ultimately lead to better retinal health ultimately protecting against glaucomatous neurodegeneration (Fig. 3; Table 1).

Fig. 3.

Summary of neuroprotective properties of estrogen. Estrogen acts as a neuroprotector in the retina by several means. It increases retinal ganglion cell survival by acting on membrane estrogen receptors and insulin receptor-β while at the same time activating several reaction cascades that regulate apoptosis and cell proliferation. Estrogen also protects visual function in conditions of excitotoxicity by activating ERK-c-Fos pathway and decreasing 14-3-3 zeta phosphorylation levels in response to glutamate insult. Furthermore, estrogen offers mitochondrial protection via its binding to classical estrogen receptors and by exhibiting antioxidant activity which improves energy metabolism

Table 1.

A summary of possible neuroprotective mechanisms of estrogen in glaucoma

Mode of action/mechanism	In vivo	In vitro	References
RGC survival
Protects against retinal neuronal fiber damage	+		Newman-Casey et al. (2014)
Increases RGC proliferation		+	Revankar et al. (2005)
Increases RGC viability		+	Giordano et al. (2011), Hayashi et al. (2007), Nakazawa et al. (2006), Ondricek et al. (2012), Prokai-Tatrai et al. (2013), and Wang et al. (2003)
IOP regulation
Lowers IOP	+		Sator et al. (1998) and Tint et al. (2010)
Protects anterior chamber outflow channels	+		Paterson and Miller (1963)
Excitotoxicity
Protects against glutamate excitotoxicity	+	+	Hayashi et al. (2007), Koseki et al. (2012), Russo et al. (2008), and Yamashita et al. (2011)
Visual function
Prevents loss of visual function	+		Prokai-Tatrai et al. (2013)
Improves retinal function	+		Yamashita et al. (2011)
Mitochondrial-related protection
Preserves RGCs from cytotoxic stimuli	+		Yu et al. (2004)
Decreases ROS production		+	Giordano et al. (2011) and Ondricek et al. (2012)
Upregulates genes implicated in antioxidant defense, neuronal survival and synaptic plasticity		+	Manthey and Behl (2006)
Protects ATP production and mitochondria		+	Giordano et al. (2011), Simpkins et al. (2008), and Wang et al. (2003)

Conclusion

The present review has introduced novel protective features of estrogen in maintaining retinal homeostasis, which is a crucial element in preventing glaucoma. On a molecular level, estrogen protects retinal neurons from glutamate excitotoxicity and regulates cellular metabolism by improving mitochondrial function through upregulation of the antioxidative defense and ensuring ATP production. Morever, estrogen provides a positive impact by lowering IOP and protecting the anterior chamber outflow channels, as well as maintaining ocular blood flow.

Reduced life span exposure to estrogen is correlated with increasing risk of POAG development. Thus, deficiencies in retinal estrogen may, therefore, potentially instigate retinal neurodegeneration, visual disability, and eventual blindness. Ophthalmologists, therefore, need to be aware that changing levels of sex hormones, especially estrogen, in different phases of a woman’s life may influence clinical outcomes, such as IOP, visual function, and optic nerve health. HRT treatment may perhaps be beneficial in future treatment and/or prevention strategies of retinal neurodegeneration associated with glaucoma. However, future studies are required to further investigate estrogen as a possible neuroprotective agent to treat glaucoma.

Author Contributions

Conceptualization: RV; Methodology: KF and RV; Validation: BSVT and MK; Investigation: KF, BSVT, MK, and RV; Writing—original draft preparation: KF, BSVT, and RV; Writing—review and editing: MK and RV.

Funding

The work is carried out as a part of the BRIDGE – Translational Excellence Programme (bridge.ku.dk) at the Faculty of Health and Medical Sciences, University of Copenhagen, funded by the Novo Nordisk Foundation (Grant Agreement No. NNF18SA0034956).

Compliance with Ethical Standards

Conflict of interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the review or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.