The molecular actions of estrogen in the regulation of vascular health

Abstract

The conflicting implications of the large-scale clinical menopausal hormone therapy trials in humans versus the findings of experimental animal studies underscore the limitations within our understanding of the molecular actions of estrogen. However, recent research has provided improved insight into the actions of estrogen on the endothelium and vascular smooth muscle. This review will outline the actions of estrogen as it contributes to vascular structure, function, and health.

Introduction

Premenopausal women benefit from a reduced incidence of cardiovascular disease relative to similarly aged men (Rosenthal & Oparil, 2000;Tunstall-Pedoe et al., 1994). A primary role for estrogen in this “cardioprotection”, and its role in preventing cardiovascular disease when supplemented following menopause, had long been assumed, but large-scale clinical trials demonstrated that an increased understanding of the actions of estrogen is required. For instance, both the Women’s Health Initiative study and the Heart and Estrogen-Progestin Replacement Study indicated that the postmenopausal use of menopausal hormone therapy is associated with an increase in adverse cardiovascular events (Grady et al., 2002;Hulley et al., 1998;Rossouw et al., 2002;Anderson et al., 2004). The Kronos Early Estrogen Prevention Study, which examined postmenopausal women without cardiovascular risk factors, demonstrated no significant effect of hormone therapy on the progression of atherosclerosis (Kling et al., 2015;Harman et al., 2014), while the Early versus Late Intervention Trial with Estradiol showed beneficial effects of menopausal hormone therapy, but only in women with elevated lipids who were not taking statins (Karim et al., 2005). Taken together, the data suggest that the effects of hormone therapy are dependent on the timing of the initiation, dose, and formulation of the treatment relative to menopause, and the number or degree of pre-existing risk factors. An underlying implication of these data is that the cardiovascular actions of estrogen are complex and incompletely understood.

Estrogen Receptors

The most well-established estrogen receptors (ERs) are ERα and ERβ. The presence of both receptor subtypes has been documented in endothelial and vascular smooth muscle cells (Mendelsohn & Karas, 1999). While the distinct roles of each subtype continue to be elucidated, both ERα and ERβ have been shown to contribute to vascular function (Miller & Duckles, 2008). Receptor expression is modulated by circulating estrogen levels (Ihionkhan et al., 2002), although ERα and ERβ appear to be regulated differently by estrogen concentrations (Miller & Duckles, 2008;Okano et al., 2006), and the effect of estrogen on receptor densities is likely dependent on the tissue type (Haas et al., 2007;Miller & Duckles, 2008).

The classical view of ERα and ERβ is as ligand-activated transcription factors which reside in the cytosol. Within this context, they elicit genomic effects requiring hours to days to become manifest. However, evidence has accumulated over the past 30 years to indicate that ERs and the signaling cascades initiated by ERs are more complex than previously appreciated. For example, cytosolic receptors can also mediate non-genomic responses: ER-mediated responses to estrogen have been observed in the presence of transcriptional inhibitors (Caulin-Glaser et al., 1997). One of estrogen’s most well-established vascular effects, the production of nitric oxide (NO), appears to occur by both genomic and non-genomic mechanisms. On one hand, long-term in vitro administration of estrogen elicits increases in expression of the eNOS mRNA and protein expression (Hishikawa et al., 1995;MacRitchie et al., 1997). On the other hand, activation of endothelial NO synthase (eNOS) occurs rapidly, implicating non-genomic mechanisms as well (Caulin-Glaser et al., 1997;Haynes et al., 2000).

Another advance in estrogen signaling has been the discovery of plasma membrane-bound ERs (Moriarty et al., 2006), the existence of which was debated for decades largely due to the lack of consensus regarding the molecular structure of the receptor (Hisamoto & Bender, 2005). This issue is made more complex by the presence of the various splice isoforms of the ERα, which are expressed under conditions of estrogen deprivation (Li et al., 2003) and preserve estrogen-mediated vascular responses. In exon 2-targeted female ERα knockout animals, the estrogen-mediated protection from vascular injury is preserved (Iafrati et al., 1997;Pare et al., 2002;Karas et al., 1999) due to the retention of splice isoforms of ERα.

In humans, membrane-bound receptors appear to be important for the regulation of vascular function as they have been identified on endothelial cells where the application of membrane-impermeant estrogens results in the activation of eNOS within minutes (Russell et al., 2000). Importantly, cross-talk exists between estrogen-mediated rapid signaling pathways and genomic pathways (Moriarty et al., 2006). Therefore, the vascular actions of estrogen are likely mediated by a complex combination of membrane-associated and cytosolic ERs, as well as ER splice isoforms.

Endothelial Effects of Estrogen

The endothelial layer is an important site of regulation for vascular function and plays a critical role in the determination of vascular health. Endothelial function is associated with estrogen receptors levels, such that male ER knockout mice are associated with reduced basal release of nitric oxide by the endothelium (Rubanyi et al., 1997). The endothelial effects of estrogen are among the most well-described and have been the topic of several reviews (Miller & Mulvagh, 2007;Arnal et al., 2010;Kim & Bender, 2009).

Estrogen may affect endothelial function by increasing sensitivity to vasodilatory factors such as acetylcholine, reducing the concentrations required to evoke similar vasodilatory responses as those observed in estrogen-deprived animals (Miller & Mulvagh, 2007;Gisclard et al., 1988). Data from ovariectomized animals indicate that chronic estrogen supplementation results in an upregulation of eNOS expression, and thereby increasing circulating NO (McNeill et al., 1999;Okano et al., 2006;Stirone et al., 2003). While this upregulation of eNOS is known to occur through genomic mechanisms (McNeill et al., 1999;Stirone et al., 2003), rapid, non-genomic pathways which lead to increases in eNOS function are activated upon estrogen binding at membrane-associated ERs (Russell et al., 2000). The molecular pathways involved in estrogen-mediated increases in endothelial NO include the rapid, estrogen-induced activation of the tyrosine kinase c-Src, followed by sequential activation of phosphatidylinositol-3 kinase, Akt and eNOS (Haynes et al., 2003;Hisamoto et al., 2001;Haynes et al., 2000). In some cellular preparations, the heterotrimeric G proteins Gαi and Gβγ are involved in membrane-initiated responses (Wyckoff et al., 2001). Plasma membrane ERs can be localized to caveolae, specialized lipid rafts abundant in endothelial cells (Chambliss et al., 2000;Li et al., 2003). In addition, ER46 can conform to a type I integral transmembrane protein with a ligand binding ectodomain (Kim et al., 2011). The definition of these various microdomains and components of ER-mediated signaling pathways, resulting in endothelial NO production, provides a variety of therapeutic targets for promotion of vascular homeostasis and health.

The case study of a 31-year old man lacking ERα (Smith et al., 1994) has demonstrated that ERα plays a critical role in the maintenance of endothelial health. Alongside problems such as decreased bone mineral density and incomplete epiphyseal closure, the man had early-onset coronary atherosclerosis (Sudhir et al., 1997a) and lacked the flow-mediated vasodilation response (Sudhir et al., 1997b)}. While these data also indicate that estrogen is an important moderator of endothelial function in both men and women, sex differences in endothelial function have been observed. Whole-body production of nitric oxide, assessed over a 36-hour period, is greater in women in the late follicular phase of the menstrual cycle relative to men (Forte et al., 1998). Additionally, when assessed in either the late follicular or mid-luteal phases of the menstrual cycle, responses to flow-mediated dilation are potentiated in women relative to men (Hashimoto et al., 1995), pointing to an estrogen-based increase in endothelial function in premenopausal women relative to men.

Flow-mediated dilation (FMD) presents a non-invasive means of assessing endothelial function. FMD responses have been shown to be predictive of adverse cardiovascular events (Inaba et al., 2010) and FMD is considered to be a valid clinical test of endothelial function (Thijssen et al., 2011). In support of a direct and functional effect of estrogen on the endothelium, FMD responses are attenuated in the phase of the menstrual cycle when estrogen levels are low (Hashimoto et al., 1995;Williams et al., 2001). Increases in FMD responses when estrogen levels are elevated (Hashimoto et al., 1995), suggest that this effect is estrogen-dependent, and not dependent on progesterone levels. The lack of an FMD response in the man lacking ERα likewise supports a strong role for estrogen in the generation of the FMD response (Sudhir et al., 1997b).

Endothelial function declines markedly at menopause (Celermajer et al., 1994) while estrogen administration in recently postmenopausal women has been shown to increase FMD responses (Lieberman et al., 1994). Similarly, postmenopausal women with vascular dysfunction experience an increase in acetylcholine-induced vasodilation following an acute infusion of estrogen(Gilligan et al., 1994). Improvements in endothelial responsiveness which occur with estrogen administration in postmenopausal women are reduced with increasing age following menopause (Sherwood et al., 2007;Vitale et al., 2008), indicating that the prolonged absence of estrogen elicits deleterious changes within the endothelium which cannot be restored by estrogen treatment (Miller & Duckles, 2008).

Effects of Estrogen on Vascular Smooth Muscle Cells

Primary evidence for the effects of estrogen on vascular smooth muscle cells have been derived from the study of endothelium-denuded arteries. In such in vitro preparations, estrogen has been shown to inhibit vascular smooth muscle cell contraction, which may occur through the inhibition of calcium ion entry into the cell (Crews & Khalil, 1999b;Crews & Khalil, 1999a) and/or through the opening of potassium channels and subsequent cellular hyperpolarization (White et al., 1995;Wellman et al., 1996). In line with these findings, aortic vasoconstriction in response to phenylephrine infusions is reduced in female rats relative to male rats (Stallone et al., 1991;Kanashiro & Khalil, 2001). Similarly, in healthy young humans, the administration of norepinephrine elicits a greater vasoconstriction in men relative to women (Kneale et al., 2000).

Estrogen Effects on Atherosclerotic Factors

One of the most important roles of estrogen in the maintenance of vascular health may be its antiatherogenic properties (Rossouw, 1996). Estrogen has been shown to affect each component of the atherosclerotic cascade (Hisamoto & Bender, 2005), including affecting circulating lipids (1995;Muesing et al., 1996), and the resultant inflammatory responses to the injury triggered by lipids and subsequent matrix deposition and intimal expansion (Beldekas et al., 1981). Estrogen has also been shown to elicit a positive effect on endothelial cell growth (Krasinski et al., 1997), while exerting an inhibitory effect over the growth and proliferation of vascular smooth muscle cells (Kolodgie et al., 1996;Bhalla et al., 1997), both of which contribute to the antiatherosclerotic effects of estrogen (Mendelsohn, 2000). Many of these effects may be NO-mediated.

Conclusions and Implications

The protective actions of estrogen on the vasculature are multi-faceted and profound. It is likely that the direct effects of estrogen on the endothelium and vascular smooth muscle, both through rapid signalling pathways and genomic mechanisms, underlie much of the cardioprotection afforded to premenopausal women. Improved understanding of the molecular actions of estrogen is required to optimize the development of hormonal treatments for postmenopausal women. One implication of such advances has been the development of selective ER modulators (SERMs), which have tissue-specific effects, functioning as ER agonists in some tissues and ER antagonists in others. With greater understanding of the molecular pathways affected by ER activation, a variety of SERMs are currently being engineered with the goal of maximizing cardiovascular and other benefits (e.g. bone, vaginal) without adversely affecting breast or endometrial tissues (Khalil, 2013). While the clinical benefits of current SERMs, primarily raloxifene, appear to be limited (Barrett-Connor et al., 2006;Barrett-Connor et al., 2002), the development of novel SERMs remains a promising area of study for the preservation of cardiovascular health in postmenopausal women (Khalil, 2013), and underlines the importance of furthering our understanding of the molecular actions of estrogen.

New Findings.

1. What is the topic of this review?

This review summarizes the beneficial actions of estrogen on the vasculature, highlighting both molecular mechanisms and functional outcomes.

2. What advances does it highlight?

The net effect of estrogen in women’s vascular health continues to be debated. Recent advances have provided strong evidence for the role of membrane-bound estrogen receptors in the maintenance of normal endothelial function. On a broader scale, functional outcomes of estrogen actions on the vasculature may mediate the reduced risk of cardiovascular disease in premenopausal women.