Alterations in the estrogen receptor profile of cardiovascular tissues during aging

Abstract

Estrogen exerts protective effects on the cardiovascular system via three known estrogen receptors: alpha (ERα), beta (ERß), and the G protein-coupled estrogen receptor (GPER). Our laboratory has previously showed the importance of GPER in the beneficial cardiovascular effects of estrogen. Since clinical studies indicate that the protective effects of exogenous estrogen on cardiovascular function are attenuated or reversed 10 years post-menopause, the hypothesis was that GPER expression may be reduced during aging. Vascular reactivity and GPER protein expression were assessed in female mice of varying ages. Physiological parameters, blood pressure, and estrogen receptor transcripts via droplet digital PCR (ddPCR) were assessed in the heart, kidney, and aorta of adult, middle-aged, and aged male and female C57BL/6 mice. Vasodilation to estrogen (E2) and the GPER agonist G-1 were reduced in aging female mice and were accompanied by downregulation of GPER protein. However, ERα and GPER were the predominant receptors in all tissues, whereas ERß was detectable only in the kidney. Female sex was associated with higher mRNA for both ERα and GPER in both the aorta and the heart. Aging impacted receptor transcript in a tissue-dependent manner. ERα transcript decreased in the heart with aging, while GPER expression increased in the heart. These data indicate that aging impacts estrogen receptor expression in the cardiovascular system in a tissue- and sex-specific manner. Understanding the impact of aging on estrogen receptor expression is critical for developing selective hormone therapies that protect from cardiovascular damage.

Introduction

Estrogen exerts protective effects in both the male and female cardiovascular system, such as reducing blood pressure and attenuating the development of vascular inflammation and atherosclerosis, leading to improved cardiac and renal health [1–7]. The estimated 10-year risk for cardiovascular disease is lower in women than in men, despite similar risk factors such as smoking, elevated cholesterol, hypertension, and diabetes [8]. However, this sex-specific protection decreases during menopause and aging [9]. While menopausal hormone therapy was anticipated to be protective, clinical trials produced mixed results [10, 11]. The timing hypothesis has emerged, suggesting that the cardioprotective effects of hormone therapy depend on the delay between onset of menopause and initiation of therapy [12, 13]. While this is most likely one explanation for the altered response, few studies have investigated whether aging impacts estrogen receptor expression in cardiovascular tissues.

Estrogen signals through both genomic and non-genomic pathways. Activation of estrogen receptor alpha (ERα) or estrogen receptor beta (ERβ) induces translocation to the nucleus where the receptors function as transcription factors to alter gene expression [14]. Additionally, estrogen binds a cell surface receptor, G protein-coupled estrogen receptor (GPER), which induces rapid, non-genomic signaling [15]. Differences in the expression of estrogen receptors may vary by tissue, sex, and age, which may provide insight into estrogen’s pleiotropic effects on the cardiovascular system. Although quantitative real-time PCR (qPCR) and reverse transcription qPCR (RT-qPCR) do not allow for direct gene comparison, droplet digital PCR (ddPCR) has emerged as a novel method for determining absolute count of target DNA copies [16]. Therefore, the present study was designed to determine the transcript levels of ERα, ERβ, GPER, and aromatase in three cardiovascular tissues from a rodent model and compare them across age and sex. The overall hypothesis was that transcript levels of all estrogen receptors would be higher in female mice compared with males and that expression levels would be significantly altered by aging.

Materials and methods

Animals

Adult (24 ± 3 weeks) and middle-aged (52 ± 3 weeks) male and female C57BL/6J mice (N = 6–8 per group) were bred at Tulane, and aged male and female C57BL/6 mice (19–20 months) were provided by NIH-NIA. All animals were housed in the Tulane temperature-controlled vivarium with an alternating 12-h, dark and light schedule with free access to standard chow and drinking water. Animal treatment was in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and was approved and monitored by the Tulane University Institutional Animal Care and Use Committee.

Blood pressure

Tail-cuff plethysmography was used to measure blood pressure in awake rodents using an automated tail-cuff volume-pressure recording system (Kent Scientific CODA® system) as previously described [17].

Vascular reactivity

Fresh aortic rings from female mice were mounted on a wire myograph (DMT) and set to a resting tension of 10 mN. After a standard start protocol of exposing the rings to KCl and phenylephrine, rings were washed and preconstricted with 10 μM phenylephrine for 10 min before 5 min exposures to increasing doses of the GPER agonist G-1 (Cayman Chemical) or estradiol (Sigma) as previously described [3, 18].

Immunoblotting

Aortas from female mice of varying ages were snap frozen in liquid nitrogen and lysed in RIPA buffer before separation by electrophoresis on a 10% Bis-Tris Gel. After transferring to a nitrocellulose membrane, blots were blocked using LiCor blocking buffer for 3 h and probed overnight at 4 °C with 1:300 anti-GPER H-300 (Santa Cruz sc-134576, Lot # K2311). The next day, blots were incubated with 1:10,000 secondary anti-rabbit (LiCor) before infrared imaging on the Odyssey system.

Droplet digital PCR

The aorta, heart, and kidneys were harvested from mice under isoflurane and immediately stored in RNAlater^TM solution (Sigma Aldrich, St. Louis, MO) according to the manufacturer’s directions to preserve RNA integrity. The tissues were mechanically homogenized and processed for total RNA isolation using the RNeasy Fibrous Tissue Mini Kit or RNeasy Micro Kit (Qiagen, Germany). The concentration and purity of extracted RNA was measured by NanoDrop spectrophotometry. Samples with nucleic acid concentration > 10.0 ng/ul and an absorbance ratio (260/280 nm) > 2.0 were used for analysis.

Transcript levels were determined using ddPCR as described previously [19, 20] using reagents from the One-Step RT-ddPCR Advanced Kit (Bio-rad, Hercules, CA). Briefly, RNA was combined with One-Step RT-ddPCR Advanced Kit for Probes (Bio-Rad Laboratories, Cat# 1864021, RRID:SCR_008426) and the following PrimePCR Primers with dual-labeled fluorescent probes: GPER (Gper1, Assay ID: dMmuCPE5103031), ERα (Esr1, Assay ID: dMmuCPE5092740), ERß, (Esr2, Assay ID: dMmuCPE5092742), and aromatase (Cyp19a1, Assay ID: dMmuCPE5097863). Droplets were then generated using a Bio-rad QX200 Droplet Generator prior to RT-PCR amplification with the following parameters for 41 cycles: amplification–denaturation at 95 °C for 60 s, annealing at 55 °C for 60 s, and extension at 72 °C for 60 s. All samples were measured using the Bio-Rad QX200 Droplet Reader and analyzed according to the manufacturer’s instructions.

Statistics

All data are presented as mean ± SEM. Data were analyzed with Prism version 8.2.1 (GraphPad Software, San Diego, CA), and significant outliers were identified using the ROUT method. Biological replicates are represented as symbols on most figures, where symbols represent the mean values for each group. Statistical tests and results are listed with each graph, and differences were considered significant when p < 0.05. Two-way ANOVA was used to determine the overall effect of sex and age on ER transcripts in each organ separately, and these results are reported in a table format under each graph. If a significant source of variation was identified, a post-hoc analysis using Sidak’s multiple comparison test was performed, and results due to the impact of aging are presented as asterisks on each graph. Two-way ANOVA was also used to compare ER transcripts by sex and tissue independent of aging. The GPER/ERα ratio was determined for each individual sample, and these values were compared using a two-way ANOVA between tissue and group. A t test was used for immunoblotting results, and two-way ANOVA for vascular reactivity studies.

Availability of data and material

The datasets are available in the Harvard Dataverse repository (link to be added before publication).

Results

To determine the presence and functional significance of GPER in the aorta of females during aging, immunoblotting and vascular reactivity experiments were performed in adult, middle-aged, and aged female C57BL/6J mice. Significant reductions are detected in the vasodilatory response to both the non-selective receptor activation with estradiol (Fig. 1a) as well as the selective GPER agonist G-1 (Fig. 1b). The magnitude of attenuation was amplified by increasing age, particularly in response to estradiol. Aortas from middle-aged and aged female mice also had significantly lower expression of GPER protein in comparison with aortas from adult female mice (Fig. 1c, d).

Fig. 1.

Impact of aging on vasodilation in female aortic rings to (a) estradiol (E2) and (b) the GPER agonist G-1. Two-way ANOVA detected an impact of aging (p = 0.0008), and post-hoc analysis results are denoted on the graph; *p < 0.05 versus adult. c GPER protein expression in aortic lysate from female mice, with age in months denoted for each lane. d The GPER/GAPDH ratio was significantly lower in aging versus adult aortic lysates. Each biological replicate is plotted along with the mean ± SEM. Unpaired t test, *p = 0.02

To determine whether the decreased vascular response to estradiol was due to aging-induced alterations in the expression of estrogen receptors in the cardiovascular system and whether this change was sex-specific, we collected physiological measurements and ER transcript levels from a new cohort of male and female adult, middle-aged, and aged mice. A significant impact of age on SBP is observed in both males and females (Fig. 2a). SBP in females increased from adult to middle age, while a decrease in SBP was observed in both aged groups. Female mice display a significant increase in weight throughout aging, while males decreased in weight after middle age (Fig. 2b). In both sexes, heart weight increases significantly after middle age with aged males having significantly increased heart weight compared with age matched females (Fig. 2c). Differences in kidney weight are observed earlier than in the heart with a significant increase from adult to middle age in male mice (Fig. 2d), but a concomitant decrease in aged males while an increase in aged females. Uterine weight increases in middle-aged female mice and is followed by a significant decrease in aged mice (Fig. 2e).

Fig. 2.

(a) Systolic blood pressure, (b) body weight, (c) heart weight, (d) kidney weight, and (e) uterine weight. Each biological replicate is plotted along with the mean ± SEM. Two-way ANOVA results are provided in the tables, and post-hoc differences due to aging are denoted on the graph; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

In the aorta, GPER mRNA transcript levels remained consistent across male groups (Fig. 3a). However, there was significant interaction between sex and age with a significant increase in GPER mRNA from middle-aged to aged females (p = 0.04). Additionally, there is a significant impact of both sex and aging on ERα mRNA in the aorta, with high expression in adult females compared with middle-aged and aged females (Fig. 3b). Neither ERß nor aromatase mRNA was detected in the aorta (data not shown).

Fig. 3.

Absolute transcript levels of (a) GPER and (b) ERα in the aorta of male and female adult, middle-aged, and aged mice. ERß and aromatase mRNA were not detected. Each biological replicate is plotted along with the mean ± SEM. Two-way ANOVA results are provided in the tables, and results from Sidak’s post-hoc analysis due to aging are denoted on the graph; *p < 0.05

In the heart, both sexes show a significant increase in GPER mRNA from middle-aged to aged groups (Fig. 4a). Additionally, there was a significant sex difference in GPER mRNA levels with both middle-aged and aged male hearts having lower levels of GPER mRNA compared with their age matched female counterparts. There was a significant interaction between sex and age on ERα mRNA levels in the heart. Aging did not alter ERα mRNA levels in the males, but copy number in middle-aged females is significantly higher than in adult mice and is followed by a decrease in aged female mice (Fig. 4b). Moreover, ERα mRNA copies in middle-aged females were significantly greater when compared with middle-aged males. Neither ERß nor aromatase mRNA was detected in the heart (data not shown).

Fig. 4.

Absolute transcript levels of (a) GPER and (b) ERα in the heart of male and female adult, middle-aged, and aged mice. ERß and aromatase mRNA were not detected. Each biological replicate is plotted along with the mean ± SEM. Two-way ANOVA results are provided in the tables, and results from Sidak’s post-hoc analysis are denoted on the graph; *p < 0.05, **p < 0.01, and ***p < 0.001

In the kidney, GPER mRNA levels are significantly greater in middle-aged and aged females compared with age-matched males (Fig. 5a), and GPER levels increased with aging only in females. There is a significant interaction between sex and aging for ERα mRNA in the kidney, with a sex difference found in adult samples and an aging effect only in males (Fig. 5b). ERβ mRNA copy number is low and variable and detectable in the kidneys of adult and middle-aged mice, but not aged mice (Fig. 5c). Aromatase mRNA was below the level of detection in the kidney (data not shown).

Fig. 5.

Absolute transcript levels of (a) GPER, (b) ERα, and (c) ERß in the kidney of male and female adult, middle-aged, and aged mice. Aromatase mRNA was not detected. Each biological replicate is plotted along with the mean ± SEM. Two-way ANOVA results are provided in the tables, and post-hoc differences attributed to aging are denoted on the graph; *p < 0.05

One advantage of ddPCR is that it allows comparisons of copy numbers across tissues and genes. ERα levels were significantly different across all tissues (p < 0.0001), while GPER transcript levels were consistent between the aorta, heart, and kidney (p > 0.99). Significantly higher levels of ERα mRNA were detected in the aorta, exhibiting 2- to 100-fold higher levels in comparison with the heart and kidney. The kidney had significantly higher ERα mRNA compared with both ERß (p < 0.0001) and GPER (p < 0.0001), while GPER transcript in the kidney was 5-fold lower than ERα copy number.

The ratio of GPER to ERα transcript is shown in Fig. 6. The aorta and kidney showed ratios < 1, indicating greater ERα versus GPER mRNA. In contrast, the heart had a near 1:1 ratio of GPER to ERα mRNA in most groups and even greater GPER in adult male and aged males and female hearts.

Fig. 6.

Ratio of GPER to ERα transcript for each biological replicate plotted on log scale. The mean ± SEM was graphed for each group. The dotted line indicates a 1:1 ratio of the two receptors, whereas values < 1 indicate more ERα transcript, and values > 1 indicate greater GPER expression. Two-way ANOVA results are provided in the table, and results from Sidak’s post-hoc tests are denoted on the graph with an asterisk. All male heart values were significantly different from each other as well as all other tissue types

Discussion

The relative ratio between receptors most likely influences the degree of membrane-initiated versus nuclear signaling that is elicited by estrogen and hence the cumulative physiological response. The current study found that absolute transcript levels of ERα, ERβ, and GPER mRNA vary widely among cardiovascular tissues. Female sex was only sometimes associated with higher receptor expression, while the impact of aging was bidirectional and interacted with sex. We also found that aging-induced decreases in GPER protein and function were not associated with changes in transcription. This variability of receptor levels across tissue type and during the aging process suggests that factors other than chronological aging are contributing to the decline in estrogen’s cardioprotective effects that are seen when hormone replacement is initiated many years after menopause.

GPER is an important target in the aorta for estrogen’s rapid and non-genomic signaling [3]. Our previous studies show that exogenous GPER activation in salt-loaded mRen2 hypertensive rats attenuates aortic remodeling [21], and others show that deletion of GPER increases atherosclerosis in the aortic root [22]. Both male and female human coronary arteries express all three estrogen receptors, but a reduced vasodilatory response to estradiol is noted in male arteries [23, 24]. Our previous work found a significant decrease in GPER function and protein expression in the vasculature of middle-aged female rats [25], while the current study showed that this decrease in GPER function and protein was recapitulated in aging female mice. In contrast to this data, aged female mice showed an increase in GPER mRNA in the aorta as well as the heart and kidney. The discrepancy between the protein versus mRNA results in the aorta indicates post-transcriptional modifications of GPER. Epigenetic factors may be at play, since ERα methylation is detected in the aorta and increases with aging and progression of cardiovascular disease [26, 27]. Since others have shown that GPER is susceptible to endocytosis and degradation via the ubiquitin-proteasomal pathway [28], this process may be enhanced during the aging process.

ERα was decreased in the aortas of middle-aged and aged female mice and may contribute to the reduction in vasodilatory response to non-selective estradiol. High levels of ERα mRNA in the aorta coupled with a decrease in response to aging may indicate transcriptional regulation of this receptor in the vasculature during the reproductive phase. ERα mRNA was impacted by aging in all three tissues tested. Interestingly, the impact of aging was tissue-specific and sexually dimorphic with decreased ERα mRNA in middle-aged and aged male kidneys but increased ERα transcript levels in middle-aged female hearts that decreased in aged females. Interestingly, the pattern of cardiac ERα mRNA in females matched uterine weights in these groups, indicating that nuclear receptor expression in the heart may be sensitive to circulating estrogen in a similar manner to the uterotrophic response. Regarding kidney tissue, a reduction in ERα mRNA and protein was similarly shown in the kidneys of aging males using semi-quantitative RT-PCR and Western blotting [29]. Sharma and Thakur also found increased ERα transcript in the kidneys of aging females, whereas similar levels of ERα transcript were found in all female age groups in the current study. The difference between our studies may be due to the mouse strain or PCR technique.

Similar to the aorta, greater levels of ERα mRNA in kidney than in heart are consistent with results found using RT-qPCR [30]. The ratio of ERα to GPER mRNA was 1:1 in the heart, whereas the kidney expressed a greater proportion of ERα mRNA in relation to GPER mRNA. This finding may indicate an important role for ERα in the kidney. Indeed, microarray studies show ERα-mediated regulation of many genes in the kidney [31]. Additionally, mice lacking ERα have significant alterations in kidney size at baseline and in response to unilateral nephrectomy [32]. The equal proportions of these two receptors in the heart were primarily due to ~ 10-fold lower ERα expression in comparison with other tissues rather than increased amounts of GPER. Indeed, this receptor also plays an important role in cardiac tissue and as a potential target for treating diastolic dysfunction in aging women.

Overall a surprising lack of sexual dimorphism was observed when comparing the transcript levels of estrogen receptors in adult mice, similar to the data from our previous study conducted in rats [20]. In the current study, significantly higher ERα mRNA was found in the female aorta and heart, while greater levels of GPER were found in the female heart and kidney. The higher levels of GPER in the female kidney were the only sex difference that was similar to our previous findings in rats. There was no species difference between GPER transcript levels in the aortas of rats and mice [20]. However, the mouse kidney showed lower expression of both ERα mRNA by ~ 5-fold, whereas other tissue transcripts showed relative consistency between the two species. These differences demonstrate the similarities between model organisms in biomedical research and encourage further investigation into gene expression differences between these two mammalian models in comparison with human samples.

Under normal, healthy conditions, undetectable ERβ mRNA in the heart and aorta combined with relatively low levels in the kidney indicates ERα and GPER as the primary receptors mediating estrogen’s effects in the cardiovascular system. Previous studies failed to detect ERβ mRNA in mouse kidneys using RT-qPCR [29], which may reflect the lower threshold for detection by using ddPCR. In contrast to our results in mice, real-time PCR of bovine kidneys detected 3-fold more ERß than ERα [33]. ERß is also detected in human fetal kidneys [34], indicating that species differences may be prevalent in estrogen receptor expression. Consistent with our results in the heart, Pugach et al. (2016) detected ERα, but not ERβ, mRNA in cardiac tissue from rats and mice of both sexes [35]. ERβ protein is detected in the human heart via immunohistochemistry [36], suggesting that species differences may exist. However, reports of ERβ mRNA in human hearts indicate that expression is much lower in healthy controls compared with tissue from patients with coronary disease [37] or aortic stenosis [38]. In combination with animal studies showing a protective effect of ERβ in the heart, these studies may indicate that ERβ is induced by disease [39, 40]. Additionally, aromatase mRNA was not found in any of the analyzed tissues, indicating that estrogen synthesis does not occur within the aorta, heart, or kidney of mice. The presence of aromatase activity and protein in the rat kidney [41] as well as mRNA in human fetal kidney, heart, and brain [42] is inconsistent with our findings.

After years of observational studies indicating that menopausal hormone therapy was cardioprotective [11], the results of the Women’s Health Initiative (WHI), a large, randomized, double-blind trial, directly contradicted this hypothesis [10]. However, data from women who underwent surgical menopause [43], as well as multiple animal models [44], have provided evidence that estrogen protects the cardiovascular system, at least during the reproductive ages. A more recent analyses of the WHI study, that stratified data by 10-year age groups, found that results in younger women (aged 50–59 years) showed positive cardiovascular effects, whereas older women showed negative effects [45, 46]. These age-dependent outcomes underlie the “timing hypothesis,” which postulates that menopausal hormone therapy has a limited window of opportunity to induce beneficial cardiovascular effects [13]. However, mechanistic insight or identification of factors that define this window is lacking. In the current study, the premise that aging-induced alterations in estrogen receptor expression may impact the response to exogenous estrogen was examined. However, estrogen receptor transcripts in the aorta, heart, and kidney of aging mice did not show a consistent pattern and instead were different according to sex and tissue type. Thus, these findings suggest that estrogen receptor mRNA may not change in response to chronological aging and that post-transcriptional changes or other factors may be more important contributors to the decline in the protective cardiovascular effects of estrogen.

In the current study, aged mice were in reproductive senescence [47]; therefore, the impact of chronological aging versus alterations in hormone levels cannot be separated. In addition, this study did not control for estrous cycle, which may influence gene expression [48]. However, our previous study in Lewis rats showed no alterations in GPER expression or function across the estrous cycle [25]. Nevertheless, ddPCR provides increased sensitivity and absolute quantification to allow comparison of data obtained from separate studies. ERβ was detected in mouse kidneys in the current study using ddPCR, while mRNA for this receptor was previously undetected using RT-qPCR [29]. The discrepancies between this study and others on estrogen receptor expression suggest that future research should use ddPCR for its increased accuracy. In conclusion, this study provides a framework on estrogen receptor transcript levels in the cardiovascular system during aging.

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Rakesh Gurrala, Dillion Hutson, Isabella Kilanowski-Doroh, and Sarah Lindsey. The first draft of the manuscript was written by Rakesh Gurrala and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This research was funded by National Institutes Of Health, National Heart, Lung, and Blood Institute, HL133619, National Institutes of General Medical Sciences, P30GM103337 and U54GM104940, National Institute of Diabetes and Digestive and Kidney Diseases, DK107694, National Institute of Neurological Disorders and Stroke and National Institute of General Medical Sciences NS094834, and National Institute on Aging, AG047296.