The goal of this study was to determine the role of estrogen receptor subtypes in the development of pressure overload hypertrophy in mice. Epidemiological studies have suggested gender differences in the development of hypertrophy and heart disease, but the mechanism and the role of estrogen receptor subtypes are not established. We performed transverse aortic constriction (TAC) and sham operations in male and female wild-type (WT) mice and mice lacking functional estrogen receptor-α [α-estrogen receptor knockout (α-ERKO)] and mice lacking estrogen receptor-β (β-ERKO). Body, heart, and lung weights were measured 2 wk postsurgery. WT male mice subjected to TAC showed a 64% increase in the heart weight-to-body weight ratio (HW/BW) compared with sham, and WT males have increased lung weight at 2 wk. WT female mice subjected to TAC showed a 31% increase in HW/BW compared with sham, which was significantly less than their male counterparts and with no evidence of heart failure. α-ERKO females developed HW/BW nearly identical to that seen in WT littermate females in response to TAC, indicating that estrogen receptor-α is not essential for the attenuation of hypertrophy observed in WT females. In contrast, β-ERKO females responded to TAC with a significantly greater increase in HW/BW than WT littermate females. β-ERKO females have lower expression of lipoprotein lipase at baseline than WT or α-ERKO females. These data suggest an important role for estrogen receptor-β in attenuating the hypertrophic response to pressure overload in females.
- lipoprotein lipase
epidemiological studies have revealed that, compared with age-matched males, premenopausal females have a decreased incidence of cardiovascular disease, including left ventricular (LV) hypertrophy (LVH), coronary artery disease, and cardiac remodeling after myocardial infarction. Death due to coronary heart disease shows a consistent male-to-female ratio of 2.5–4.5 across populations in countries with different lifestyles and heart disease rates (2). The incidence increases in postmenopausal females (16).
Cardiac hypertrophy occurs when myocytes enlarge in response to pressure or volume overload in an attempt to normalize wall stress. It has been reported that over 50% of patients with LVH secondary to untreated hypertension develop congestive heart failure (17), a clinical condition with a highly unfavorable outcome. It is therefore of great interest that multiple studies have documented gender dimorphism in the development of ventricular hypertrophy (13, 21). Whether females develop less hypertrophy because they have less underlying disease or whether given the same stimulus they have less predisposition to hypertrophy is debated. Gender dimorphism in cardiac remodeling after myocardial infarction in humans has also been observed; in patients with ischemic cardiomyopathy, the heart weight (HW) index was significantly greater in men, and men also exhibited a larger myocyte volume (5). Interestingly, most of the women in this study were postmenopausal. Despite these well-documented gender differences, large randomized clinical trials have failed to show that hormone replacement therapy in postmenopausal women is beneficial in the primary prevention of cardiovascular disease (29), highlighting the need for a better understanding of the molecular mechanisms underlying differences in the development and progression of heart disease with regard to gender and hormone status.
Most animal models of heart failure have reported that females have improved survival, and many report improved contractile function (6, 9, 18, 27). The relationship between sex and hypertrophy is complex and appears to depend on the model/etiology of hypertrophy, age, and the stage of hypertrophy/failure. Dash et al. (6) reported that overexpression of phospholamban leads to increased hypertrophy in males at 15 mo; females develop a similar level of hypertrophy, but it is delayed and does not occur until ∼22 mo of age. As the mice age, there is earlier mortality in the males, such that by 22 mo most of the males have died. Similarly, in mice overexpressing TNF receptor, males exhibit early death and increased hypertrophy (18). In contrast, in a mouse model of hypertrophic cardiomyopathy, associated with an α-myosin heavy chain (MHC) R403Q mutation and a deletion in the actin binding domain of α-MHC, males showed less hypertrophy than females at 10 mo of age (27); at 10 mo, females had normal LV systolic function and concentric hypertrophy, whereas males had a dilated LV and decreased LV systolic pressure. In a model of pressure overload due to banding of the ascending aorta in weanling rats, Douglas et al. (9) found that by 20 mo of age, males exhibited less hypertrophy than females, and males had increased dilation and decreased LV function.
Two known steroid hormone receptors, estrogen receptor (ER)-α and ER-β, mediate the effects of estrogen. Both receptors are expressed in vascular tissue and cardiac myocytes (14, 25). The majority of studies up to this point have concentrated on the vascular effects of estrogen (24, 25). However, it is becoming increasingly clear that estrogen has direct effects on the myocardium independent of vascular hormone action. Van Eickels et al. (32) found that estradiol could attenuate the development of pressure overload hypertrophy in ovariectomized female mice in association with decreased cross-sectional area of cardiac myocytes. This effect was independent of blood pressure because placebo- and estradiol-treated mice showed no differences in developed blood pressure (32).
Given this background, the goal of our study was to establish whether male and female mice would show gender differences in response to pressure overload hypertrophy. We chose to induce hypertrophy using transverse aortic constriction (TAC) because it has been validated as a reproducible model (28) that simulates aortic stenosis and essential hypertension, the most common causes of pathological LVH in humans. We found that males subjected to TAC developed a degree of hypertrophy significantly greater than their age-matched female counterparts. We also examined the role of ER-α and ER-β in mediating attenuation of pressure overload hypertrophy in females after TAC. We hypothesized that if the protection was mediated by either ER-α or ER-β, ablation of that receptor in the corresponding knockout mouse would lead to increased hypertrophy compared with wild-type (WT) females in the setting of pressure overload. We report experimental evidence for the novel observation that ER-β mediates the sexual dimorphic protection in females.
Commercially available C57BL/6 mice were used in the initial study of WT males and females. Mice lacking functional ER-α [α-ER knockout (α-ERKO)] (23) and mice lacking functional ER-β (β-ERKO) (20) as well as WT littermates were used to study the effects of ER deficiency. Notably, both α-ERKO and β-ERKO mice were developed on a C57BL/6 background and backcrossed at least 10 generations (20, 23).
All animal procedures received prior approval from the National Institute of Environmental Health Sciences animal care and use committee. Seven-week-old mice, weighing 18–25 g, were anesthetized with isoflurane. Animals were placed in the supine position under a dissecting microscope, and a 5-mm midline incision was made just above the sternal notch to expose the aorta and carotid arteries. A 7-0 suture was threaded around the aorta between the carotid arteries. Aortic constriction was performed by tying the suture against a 27-gauge needle (28). The needle was then removed, leaving a narrowing of 0.5 mm in diameter. Animals recovered from anesthesia on 100% O2. The mortality rate for animals that recovered from anesthesia was 8%, whereas operative mortality was 18%. Sham animals underwent an identical surgical procedure without placement of the suture.
Animals were killed 2 wk postsurgery. Prestenotic pressure, measured by right carotid artery cannulation performed before death using an Isotec Pressure Transducer (Hugo Sachs Electronics, Harvard Apparatus), was similar in all TAC groups. Prestenotic pressure did not differ between the sham males and females of different genotypes; sham males of all genotypes averaged 112 ± 3 vs. 108 ± 6 cmH2O for females averaged among all genotypes (P > 0.05, not significant). Compared with the sham mice, prestenotic pressure was elevated in all TAC groups. We found no differences in prestenotic pressure between any of the TAC groups (WT male, 152 ± 9.5 cmH2O; α-ERKO male, 145 ± 10 cmH2O; β-ERKO male, 143 ± 13 cmH2O; WT female, 140 ± 20 cmH2O; α-ERKO female, 139 ± 10 cmH2O; and β-ERKO female, 135 ± 9 cmH2O). Total body weight (BW), lung weight (LW), and HW were measured at the time of death. Pieces of each heart and lung were placed in formalin and submitted for histology. Hematoxylin-eosin-stained and Masson-stained sections were evaluated.
Two-dimensional guided M-mode echocardiography (ATL, Philips) was performed on conscious mice as describe previously (11).
A piece of the heart was snap frozen in liquid nitrogen at the time of harvest. RNA was extracted using a Qiagen RNeasy kit. RNA concentration was determined by optical density at 260 nm. Formaldehyde gel electrophoresis was performed to confirm distinct 28S and 18S rRNA. All samples had a 28S band twice the intensity of the 18S band.
Quantitative real-time PCR (RT-PCR) was used to measure gene expression levels. The primer sequences used in RT-PCR are shown in Table 1. First-strand synthesis was conducted for 60 min at 48°C in a 10-μl reaction containing 100 ng total RNA, 5.5 mM MgCl2, 2 μM dNTPs, 4 units RNase inhibitor, and 12.5 units reverse transcriptase. The RT-PCR was carried out in a total volume of 40 μl, which included the first-strand synthesis reaction, to which was added 4 mM MgCl2, 8 mM dNTPs, 1× SYBR green PCR buffer (PE Biosystems), 0.4 μM gene-specific primers, and 2.5 units AmpliTaq Gold DNA polymerase (PE Biosystems). The reaction was analyzed using an Applied Biosystems PRISM 7700 detection system. mRNA expression was normalized to GAPDH and 18S RNA (Amplicon Technologies). Each reaction was performed in triplicate.
Plasma estradiol was assayed in singlicate per animal on 200-ml aliquots using the Ultra-Sensitive Estradiol Double-Antibody RIA kit (Diagnostic Systems Laboratories; Webster, TX) according to the manufacturer's protocol. All final assay samples were quantified using a Packard Multi-Prias 2 Gamma counter. To avoid interassay variation, all assays were carried out in a single setup.
Data are shown as means ± SE. Means were compared by one-way ANOVA, with a post hoc Fisher's test applied for multiple comparisons. Differences were considered significant at P < 0.05.
Pressure overload experiments.
In both male and female WT mice, TAC resulted in a HW-to-BW ratio (HW/BW) that was significantly greater than sham counterparts (Fig. 1A) after 2 wk. Notably, HW/BW in TAC male mice was significantly greater than in TAC female counterparts, whereas HW/BW was virtually identical in shams (Fig. 1A). There was no visible fibrosis in Masson-stained sections in any of the hearts. In this study of WT male and female mice, TAC female mice showed a 31% increase in HW/BW compared with sham; TAC male mice showed a 64% increase in this ratio compared with sham. The LW-to-BW ratio (LW/BW), an indication of LV function, was nearly identical in TAC and sham females (Fig. 1B). However, male TAC mice showed a substantial increase in LW/BW compared with shams (Fig. 1B).
To further evaluate hypertrophy and the hypertrophic stimulus in male hearts compared with female hearts, we used RT-PCR to measure the expression of several genes suggested to be altered by pressure overload hypertrophy. Transcript levels of transforming growth factor-β3 were upregulated 152% in TAC males compared with sham males, whereas TAC females exhibited an upregulation of 30% compared with sham females (Fig. 2A). TNF receptor 12b and matrix metalloproteinase 14 were also upregulated in TAC males versus shams (Fig. 2, B and C). TNF receptor 12b transcript levels showed a 153% increase in TAC males versus sham males, whereas transcript levels in TAC females were elevated only 10% compared with sham females. Matrix metalloproteinase 14 transcript levels were elevated 148% in TAC males compared with sham males; TAC females exhibited a 42% increase transcript levels compared with sham females. As expected, skeletal α-actin and brain natriuretic peptide were all significantly elevated in both male and female TAC hearts compared with sham hearts (Fig. 2, D and E). We also examined estradiol levels in the plasma of male and female sham and TAC mice. As expected, females had higher estradiol levels compared with males, but TAC did not result in a change in estradiol in either male or female mice (Fig. 3).
We were interested in determining the role of ER-α and ER-β in attenuating the development of pressure overload hypertrophy. We therefore performed sham and TAC operations in α-ERKO, β-ERKO, and WT littermate females. We hypothesized that if either ER was important in the attenuation of the hypertrophy observed in females, the hearts lacking that ER would develop a greater degree of hypertrophy than WT females and would be more similar to males. In the α-ERKO experiment, we found no difference in HW/BW between WT and α-ERKO sham females (Fig. 4A). Similarly, TAC also produced no difference in HW/BW between WT and α-ERKO females (Fig. 4A), indicating that ER-α is not essential for the attenuation of the hypertrophic response observed in WT females. In the β-ERKO experiment, we again found no difference in HW/BW between sham WT and β-ERKO females (Fig. 4B). However, TAC resulted in a degree of hypertrophy in β-ERKO females that was significantly greater than their TAC WT female counterparts (Fig. 4B), indicating that ER-β is important in attenuating the development of hypertrophy in response to pressure overload.
Because β-ERKO females showed increased hypertrophy compared with WT females, we performed M-mode echocardiography on sham and TAC mice from these two groups. As shown in Table 2, consistent with the data shown in Fig. 4B, female β-ERKO hearts subjected to TAC had more hypertrophy than WT females. Female β-ERKO hearts showed significantly higher diastolic LV diameter than WT female TAC hearts. Fractional shortening was not significantly different between WT and β-ERKO females.
TAC experiments comparing the degree of hypertrophy in α-ERKO, β-ERKO, and WT littermate males showed increased hypertrophy in TAC males compared with shams (Fig. 5). α-ERKO male hearts subjected to TAC had slightly but significantly less hypertrophy than WT male TAC hearts. Interestingly, the loss of ER-β does not increase hypertrophy in male hearts. These data suggest that, although ER-β may mediate protection in females, in males with lower estrogen levels, the loss of ER-β is not detrimental.
We were interested in gaining insight into how ER-β might result in reduced hypertrophy in females. We therefore used gene profiling to identify altered gene expression among WT, α-ERKO, and β-ERKO female hearts at baseline (data not shown). We found that lipoprotein lipase (LPL), the rate-limiting enzyme in delivery of fatty acids to muscle (10), was significantly downregulated in β-ERKO compared with α-ERKO females. We therefore performed RT-PCR on male and female hearts of the different genotypes. As shown in Fig. 6, at baseline (no sham or TAC surgery), β-ERKO females had significantly less LPL expression than WT or α-ERKO females. α-ERKO males exhibited significantly higher LPL than WT males or β-ERKO males, consistent with the reduced hypertrophy in α-ERKO males and consistent with a role for LPL in the sex-related differences in hypertrophy.
We used RT-PCR to determine the level of RNA for LPL in sham and TAC hearts from β-ERKO, α-ERKO, and WT females. We reasoned that because β-ERKO TAC female hearts resembled TAC male hearts, genes involved in this sexually dimorphic response should be differentially regulated in β-ERKO females compared with WT and α-ERKO females. As shown in Fig. 7A, LPL was downregulated at baseline in sham β-ERKO females compared with sham α-ERKO and WT females, consistent with the decreased expression observed in β-ERKO females without sham surgery (Fig. 6). We found LPL to be downregulated in all females under conditions of pressure overload (Fig. 7A). However, consistent with the increased hypertrophy in β-ERKO females, β-ERKO females exhibited the greatest decline in LPL with TAC, and β-ERKO TAC females had significantly lower LPL than WT TAC females. LPL was also downregulated in both male and female WT TAC hearts; however, LPL was more sharply downregulated in TAC male mice compared with TAC female mice and sham mice (Fig. 7B). Compared with shams, TAC male mice revealed an 83% decrease in transcript levels, whereas TAC females exhibited a 50% decrease in transcript levels. The decrease in LPL in β-ERKO hearts at baseline might enhance susceptibility to hypertrophy; however, LPL is decreased in all hearts after TAC, suggeting that the relationship between LPL, gender, and hypertrophy is complex and requires further study.
ERK 1/2 are involved in the development and progression of cardiac hypertrophy (3). Because estrogen has been reported to have rapid effects on ERK signaling (7), we investigated ERK 1/2 for differences in activation in sham versus TAC females of WT, α-ERKO, and β-ERKO hearts. As shown in Fig. 8, there were no significant differences in phosphorylated ERK 1 or ERK 2 for sham compared with TAC in any of the genotypes. There were no differences in total ERK 1 and ERK 2 between the groups (data not shown).
We found that hypertrophy induced secondary to pressure overload was significantly less in female WT mice compared with age-matched male WT mice at 2 wk postsurgery. We were therefore interested in determining whether this male/female difference was mediated via the ER and, if so, which receptor.
Direct effects of estrogen on the myocardium have been demonstrated (7, 14, 26), but the physiological roles of ER-α and ER-β in the myocardium remain largely unexplored. For the first time, we report the response of ERKO mice to pressure overload hypertrophy. We found that homozygous α-ERKO female mice exhibit a response to TAC that is identical to the response of WT females. This finding indicates that ER-α is not essential for the attenuation of the hypertrophic response. In contrast, we found that homozygous β-ERKO female mice subjected to TAC exhibited an increased degree of hypertrophy compared with WT female mice, comparable to WT males, indicating a role for ER-β in mediating an attenuated response in females to pressure overload. Previous studies have shown that ER-β expression is not changed in the hearts of α-ERKO mice (4). An independent role for ER-α versus ER-β in stress responses in the cardiovascular system has been previously observed. Studies in vascular smooth muscle cells revealed that increases in nitric oxide synthase were mediated by ER-β and antagonized by ER-α (34).
We were interested in identifying ER-β-regulated genes that might be responsible for the reduced hypertrophic response in females. Interestingly, we found that LPL, at baseline, was downregulated in β-ERKO females. LPL was decreased during hypertrophy in all hearts studied. However, the decrease in TAC WT males was greater than in TAC WT females, and, in females subjected to TAC, we observed the most substantial downregulation of LPL in β-ERKO females. Furthermore, consistent with the data in our mouse model, a gender difference in the level of expression of LPL has been observed at baseline by others (12). LPL is the rate-limiting enzyme in the hydrolysis of triglycerides in very-low-density lipoproteins and is therefore an important mediator of fatty acid metabolism (10). There are a number of lines of evidence that suggest a role for LPL in modulating cardiovascular disease and hypertrophy. There is a polymorphism in humans that decreases the activity of LPL and is associated with an increased risk of ischemic heart disease in women but not men (33). Furthermore, mice with cardiac-specific ablation of LPL exhibit decreased fatty acid metabolism and increased carbohydrate metabolism (1), reminiscent of the pathological switch in the myocardial energy substrate from fatty acid to glucose during myocardial hypertrophy (8, 19). Mutations in genes involved in fatty acid metabolism have also been shown to lead to hypertrophy, providing another link. However, LPL is significantly decreased in all hypertrophied hearts, suggesting that the relationship among LPL, gender, and hypertrophy is complex and requires additional study.
We also found that α-ERKO males have increased LPL compared with WT and β-ERKO males, consistent with the reduced hypertrophy observed in α-ERKO males. The female-like expression of LPL in α-ERKO males is consistent with data in the literature showing that disruption of ER-α results in increased numbers of tyrosine hydrolase-immunoreactive neurons in the preoptic region, approaching the numbers seen in WT females (30). Disruption of ER-α has also been shown to feminize the growth hormone Stat 5b pathway and expression of cyp2a4 and cyp2d9 genes in the liver (31). Exactly how disruption of ER-α leads to altered gene expression is unclear, but ER-α and ER-β have been reported to have a “ying-yang” relationship (22). It is therefore possible that lack of ER-α in males, which have limited estrogen, results in enhanced signaling through ER-β and a more feminine expression of LPL. Additional studies will be necessary to define the precise role of ERs in regulating LPL.
Estrogen has been reported to induce ERK 1/2 activation in an ER-dependent manner in rat cardiac myocytes (7), and ERK signaling has been implicated in hypertrophy (3). We therefore examined ERK 1/2 signaling in sham and TAC females of WT, α-ERKO, and β-ERKO hearts. We found no differences in phosphorylated ERK 1 or 2 in females sham versus TAC hearts. These data are consistent with the results of Haq et al. (15), who reported no change in phosphorylation in ERK in hypertrophied human hearts.
In conclusion, our study indicates that males respond to pressure overload with an increased degree of hypertrophy compared with females. For the first time, we established the importance of the direct action of ERs in the myocardial response to pressure overload. We observed that β-ERKO females develop an increased degree of hypertrophy compared with WT females, whereas α-ERKO mice respond in a manner identical to WT females. These findings provide evidence for ER specificity and functionality related to cardiovascular function and have important therapeutic implications for the design of selective ER modulators effective in the prevention and treatment of cardiovascular disease.
M. Skavdahl, J. Clark, P. Myers, T. Demianenko, K. S. Korach, and E. Murphy were supported by the National Institute of Environmental Heath Sciences intramural program; M. Skavdahl was also supported by a grant from Howard Hughes Medical Institute. C. Steenbergen was supported by National Heart, Lung, and Blood Institute (NHLBI) Grant RO1-HL-39752, and H. A. Rockman and L. Mao were supported by NHLBI Grant HL-56687.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 by the American Physiological Society