Author + information
- Received June 29, 2011
- Revision received August 17, 2011
- Accepted September 13, 2011
- Published online January 24, 2012.
- Georgios Kararigas, PhD⁎,⁎ (, )
- Virginie Bito, PhD§,
- Hanna Tinel, PhD‡,
- Eva Becher, PhD⁎,
- Istvan Baczko, MD, PhD∥,
- Christoph Knosalla, MD†,
- Barbara Albrecht-Küpper, PhD‡,
- Karin R. Sipido, MD, PhD§ and
- Vera Regitz-Zagrosek, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Georgios Kararigas or Dr. Vera Regitz-Zagrosek, Institute of Gender in Medicine and Center for Cardiovascular Research, Charite University Hospital, Hessische Strasse 3-4, 10115 Berlin, Germany
Objectives This study investigated the effects of 17β-estradiol (E2) on gene regulation in human cardiac tissues. We hypothesized that a candidate E2 effect is cardiomyocyte (CM)- and sex-specific, conserved between humans and mice, and that E2 impairs contractile function in male CMs only.
Background Both men and women produce E2 locally from androgenic precursors. E2 regulates cardiovascular function, but specific mechanisms, protective or harmful, are not fully understood.
Methods We performed genome-wide expression profiling of E2-treated cardiac tissues from men and women, and studied gene expression and function in CMs from hearts of male and female E2-treated mice.
Results We found 36 E2-dependent genes regulated in a sex-specific manner. Of these, after E2 exposure, the myosin regulatory light chain interacting protein (MYLIP) gene was induced in tissues of men only. Focusing on Mylip and employing isolated mouse CMs, we confirmed our hypotheses that the E2 effect is CM- and sex-specific and conserved between humans and mice. The E2-treatment led to impaired contractile function in male CMs only, which was characterized by increased Mylip mRNA and protein levels, and decreased myosin regulatory light chain (Mrlc) protein. Our report is the first to our knowledge to show that cardiac Mrlc is an in vivo substrate for Mylip, leading to augmented Mrlc ubiquitination. Of relevance, we found that MYLIP expression levels rise with increasing age in hearts of men.
Conclusions E2 directly influences cardiac gene regulation, and E2 actions may be different between the sexes. Since E2 levels rise in older and/or obese men, pharmacological targeting of MYLIP in men with elevated E2 levels could possibly decrease their risk for the development or progression of cardiovascular disease.
Sex- and menopause-related differences in cardiovascular diseases have drawn considerable attention to the sex steroid hormone 17β-estradiol (E2), which has been largely believed to be beneficial for the heart (1). However, unexpected negative findings from large randomized clinical trials (2,3) and conflicting results from animal studies have led to controversy about the role and actions of E2. Additionally, recent studies have revealed that E2 is also relevant for cardiovascular function in men (4). A better understanding of the effects of E2 on molecular and cellular physiology in the cardiovascular system is necessary.
The primary source of E2 production is the ovary. However, E2 is also produced locally in both sexes through the conversion of testosterone by aromatase (5), with the adipose tissue contributing significantly to its circulating pool (6). E2 modulates a number of biological processes, such as reproduction, development, cell proliferation, differentiation, apoptosis, and metabolism, via gene regulation. E2 directly influences gene expression through the classical cytosolic estrogen receptors (ER) α and β, although rapid nongenomic actions not requiring the ER have also been reported. In the genomic effects, E2 binds to the ER, and the newly formed complex translocates to the nucleus. Subsequently, E2-ER complexes interact with a number of coactivator and corepressor proteins that are recruited to the receptor after ligand binding, to mediate or inhibit transcription (7,8).
We investigated the effect of E2 on the regulation of gene expression in human cardiac tissues. Following genome-wide expression profiling, we selected myosin regulatory light chain interacting protein (Mylip) as our candidate, and we tested the hypothesis that this estrogenic effect occurs in the cardiomyocyte (CM) using isolated CMs from hearts of E2-treated mice. We found that the effect is conserved between humans and mice, and that E2 modulates the expression of Mylip to control cellular contraction. We have therefore identified an effect of E2 on the regulation of gene expression occurring directly in the CM. This effect differs between the sexes, and we report that E2 impairs contractile function in male, but not in female, CMs. Notably, we also found that MYLIP expression is higher in the hearts of men older than 50 years than in the hearts of men younger than 40 years. Considering that E2 levels rise in older and/or obese men and, particularly, that elevated E2 levels in men with heart failure are associated with poor prognosis (4), our findings could be clinically relevant. Further confirmation of these findings at a clinical level could help to develop sex-specific treatments as part of a more individualized medical care, tailored to the specific needs of men and women.
Human cardiac tissue studies
Tissue harvest, culture, and treatment were recently described (9) and are summarized in the Online Methods. Tissues from 6 patients (50% women) undergoing coronary bypass surgery were treated with 10−8 mol/l E2 or 2-hydroxypropyl-β-cyclodextrin (HBC) (vehicle control for E2 treatment) for 24 h. This concentration was chosen to obtain an E2 level of physiological range. The study was approved by the Charite University Hospital Ethics Committee, it complies with the principles outlined in the Declaration of Helsinki, and written consent was obtained.
Nondiseased male human hearts (n = 6 for age <40 years, n = 9 for age >50 years) technically unusable for transplantation were obtained from general organ donors. Experimental protocols were approved by the Scientific and Research Ethical Committee of the Medical Scientific Board at the Hungarian Ministry of Health (ETT-TUKEB) under ethical approval No. 4991-0/2010-1018EKU (339/PI/010).
Hybridization and microarray profiling
Detailed protocols are provided in the Online Methods. Microarray data are deposited in the ArrayExpress database (accession No. E-MEXP-2971).
Microarray data analysis
Middle-aged (11- to 14-month-old) male (n = 20) and female (n = 22) C57BL/6J mice (Charles River Laboratories, Sulzfeld, Germany) were treated in vivo with 0.2 mg/kg E2, a dose reported to lead within 2 h to a 2.5-fold increase in E2 levels up to 129 pg/ml and within the physiological range for mice, returning to baseline after 22 h as E2 gets metabolized (12). Control mice were treated with HBC. The peak serum levels in elderly men are also reported to be up to 129 pg/ml, supporting the choice of E2 dose (13). Five hours after injection, mice were sacrificed and divided into 2 sets. In the first set (n = 4 male HBC, 6 male E2, 5 female HBC, 7 female E2), CMs were isolated for contraction and mRNA measurements. The second set (n = 5/group) was used to extract protein. Unloaded cell shortening was measured with a video-edge detector (Ionoptix, Milton, Massachusetts) at 1, 2, and 4 Hz using field stimulation. Detailed protocols are provided in the Online Methods. The study was approved by the Ethics Committee on Animal Use of the University of Leuven.
Quantitative real-time reverse transcription PCR
Immunoblotting and coimmunoprecipitation
Detailed protocols are provided in the Online Methods.
All data were analyzed statistically using the R version 2.11.0 software. Data are shown as the mean ± SEM. Comparisons between 2 groups were performed with unpaired t test and between multiple groups using 2-way analysis of variance with Tukey's post-hoc test adjusting for multiple comparisons, considering p ≤ 0.05 significant.
E2-dependent regulation of gene expression in human cardiac tissues
To investigate the effects of E2 on global human cardiac gene expression, we treated heart tissues from men and women with E2 or HBC (vehicle), and we performed microarray analysis. After verifying the quality of the microarray data, we fitted a linear model with an empirical Bayesian method (14) to determine differential gene expression. To extract biologically useful information, we examined those probe sets with an unadjusted p value of up to 0.001. This resulted in 77 significant probe sets (Online Table 1). Unsupervised hierarchical clustering of these probe sets revealed that the treatment with E2 was not sufficient to be the sole factor separating the tissues into 2 distinct groups, i.e., E2 and HBC, suggesting that the sex might play an important role in the regulation of gene expression by E2 (Fig. 1). Additionally, we found that there were small fold changes (20% to 40%) in gene expression between the 2 conditions.
Sex-specific regulation of gene expression by E2 in human cardiac tissues
We then asked whether E2 regulates cardiac gene expression in a different manner between male and female tissues. We first calculated the differences in expression between the 2 conditions of E2 and HBC treatment for each tissue. We then fitted a statistical model stratifying the data by the effect of the sex. This approach identified 41 probe sets (p ≤ 0.001) (Online Table 2). Unsupervised hierarchical clustering of these probe sets confirmed the effect of the sex through the separate grouping of E2 effects in male and female divisions (Fig. 2A). Among these candidates, we identified the MYLIP gene, whose expression was induced by E2 in heart tissues of men only (Fig. 2B).
E2 regulates Mylip in a sex-dependent manner in mouse CMs
We selected Mylip as the most interesting candidate because of its potential influence on contraction through regulation of the myosin regulatory light chain (Mrlc) (15–18), and we used the mouse for further analysis. Due to the multicellular nature of the heart tissue, to test the hypothesis that this regulation occurred in the CM, we treated mice in vivo with E2, and we subsequently determined the level of Mylip in isolated CMs. In addition, this approach enabled us to test whether the regulation of Mylip by E2 is conserved between humans and mice. Quantitative real-time reverse transcription PCR confirmed that E2 modulated the expression of Mylip in a sex-specific manner (interaction p < 0.05). In particular, the treatment with E2 induced the expression of Mylip in CMs of male mice (adjusted p ≤ 0.05), whereas the same treatment had no significant effect in female CMs (adjusted p = 0.99) (Fig. 3). Therefore, we have identified a novel regulatory effect of E2 that is CM- and sex-specific and conserved between humans and mice.
Increased Mylip and reduced Mrlc protein levels in CMs of male E2-treated mice only
Next, we investigated whether Mylip protein levels were altered in the same direction as at the transcriptional level. In particular, after E2 exposure, Mylip protein levels were significantly increased in males (adjusted p < 0.01), whereas they did not change in females (adjusted p = 0.76) (Fig. 4A). Mylip mediates ubiquitination and subsequent proteasomal degradation of Mrlc (15,18). On the basis of this, we hypothesized that E2 modulates Mrlc through increased Mylip protein levels. Immunoblotting revealed the sex-dependent abundance of Mrlc protein after treatment with E2. In line with our hypothesis, we found that there was a significant decrease (adjusted p ≤ 0.05) in Mrlc levels in males treated with E2 compared with the male-derived control group, whereas in females, it remained stable (adjusted p = 0.99) (Fig. 4B).
Cardiac Mrlc is an endogenous substrate of Mylip, and increased Mylip abundance enhances in vivo Mrlc ubiquitination
Evidence supporting the notion that Mylip regulates cardiac Mrlc levels has been lacking. Our report offers novel insight into this mechanism. We hypothesized that Mrlc represents an in vivo substrate for Mylip activity in the heart. Coimmunoprecipitation analyses of male left ventricular lysates revealed the binding of Mylip to Mrlc, which was further augmented after E2 exposure (Fig. 5A). Accordingly, the expected basal ubiquitination levels of Mrlc were further increased following E2 treatment (Fig. 5B), supporting our hypothesis.
E2 impairs contractile function in male CMs only
We then postulated that the E2-triggered increase in Mylip levels characterized by the decrease in Mrlc abundance in male CMs leads to impaired contractile function in these cells. We recorded unloaded cell shortening at 1, 2, and 4 Hz. CMs from male mice treated with E2 showed a significantly smaller cell contraction compared with the HBC-treated group (Figs. 6A and 6C). On the other hand, there was an increasing trend in cell contraction in female CMs after E2 compared with controls (Figs. 6B and 6C). Similarly, there was a significant decrease in the rate of contraction of E2-treated male CMs compared with controls, whereas the opposite trend was observed in female CMs (Fig. 6D).
MYLIP expression levels rise with increasing age in male human hearts
Knowing that E2 levels increase in aging men (6,13,19), we sought to support the clinical relevance of our findings. We therefore studied the expression levels of MYLIP in the hearts of younger and older men. Quantitative real-time reverse transcription PCR revealed that there is a higher MYLIP expression in the hearts of men older than 50 years than in the hearts of men younger than 40 years (Fig. 7A). Moreover, we employed 2 public microarray datasets that enabled us to stratify our analysis by age. Using this additional cohort, we further confirmed that in the hearts of men older than 50 years (n = 4), MYLIP expression was higher than in the hearts of men younger than 40 years (n = 4; adjusted p = 0.002) (Fig. 7B).
Our novel findings are that E2 has direct effects in the heart through the regulation of gene expression; these effects can be sex- and CM-specific; E2 regulates Mylip in a conserved fashion in humans and mice; cardiac Mrlc is a substrate for Mylip activity; and E2 impairs cellular contractility in male CMs.
We performed genome-wide profiling of E2-treated human cardiac tissues. The variation in human material was considerable, and the changes in gene expression were not large. Small fold changes due to E2 effects have already been encountered by our group and others (20–22). Small quantitative differences in the amount of RNA might have relevant qualitative effects at a functional level. To this extent, we recently suggested that the heart is a highly regulated environment, where minor changes may have major effects (9). Here, we provide evidence supporting this notion.
Following the generation of the candidate list, the exposure of the tissues to E2 was not sufficient to be the sole factor responsible for the variation in the expression of the identified 77 probe sets between HBC- and E2-treated tissues, as demonstrated by the corresponding clustering. This finding further suggested that the sex might play an important role in the regulation of gene expression by E2. Divergent hormonal effects on gene expression in male and female mice were suggested earlier (23), a finding that we recently confirmed in the human heart for progesterone receptor (9). We now demonstrate that E2 regulates gene expression in a sex-specific manner, particularly in CMs. We have not yet deduced the mechanisms responsible. One possibility could be differences in the expression levels of the ER between the sexes. However, the expression profiling did not reveal any significant changes. On the other hand, the E2-ER complex interacts with other transcriptional cofactors of different function, thereby modulating transcriptional regulation (7,8). We postulate that the identified sex-specific regulation of gene expression by E2 is the result of differences in the recruitment of transcriptional cofactors between the sexes.
A novel observation is that ex vivo treatment of human cardiac tissues with E2 increased the levels of MYLIP in tissues from men, but not from women, compared with control tissues. To verify the idea that this regulation occurs in the CM, we treated mice in vivo with E2, and we subsequently determined the level of Mylip in isolated CMs. As in humans, E2 induced the expression of Mylip in male, but not in female, mouse CMs. We also confirmed that Mylip protein levels were altered in the same direction. Mylip is an E3 ubiquitin ligase that belongs to the ezrin-radixin-moesin (ERM) protein family, and it contains a C-terminal RING finger domain (18). In neuronal cells, Mylip has been shown to bind Mrlc, mediating its ubiquitination and subsequent proteasomal degradation (15,18). Based on this knowledge, we hypothesized that the induction of Mylip by E2 would result in decreased Mrlc levels. Indeed, there was a significant decrease in Mrlc protein levels due to E2 in male CMs, whereas there was no change in the Mrlc content of female CMs. However, the possibility that cardiac Mrlc is a substrate of Mylip had not been explored previously. We now demonstrate the endogenous interaction of Mylip with Mrlc. As expected, this binding was further augmented after E2 exposure. Subsequently, we determined that the in vivo ubiquitination of Mrlc was enhanced in E2-treated CMs.
Modulation of contraction by Mrlc in the heart has been well documented (16,17). Phosphorylation of Mrlc by myosin light chain kinase decreases the distance between myosin and actin, with an increase in Ca2+ sensitivity of the myofilaments and a higher rate of contraction. At baseline, there is a significant level of phosphorylation. Mutant mice for Mrlc phosphorylation show defects in heart contraction (24). Although no data exist regarding changes in the levels of Mrlc expression, we postulated that decreased Mrlc levels would lead to impaired contractile function. Indeed, reduced Mrlc levels would preclude the positive modulatory effect of phosphorylation of Mrlc. Cell shortening measurements demonstrated that with E2 there was a significant impairment of cellular contractility in male CMs only. Our data support the regulatory role of Mrlc and reveal a novel sex-dependent mechanism. The reduced extent and rate of contraction suggest that reduction of Mrlc reproduces the effects of removal of phosphorylation.
In men, E2 is produced in significant quantities by local tissue aromatization of androgenic precursors from the testes and adrenal glands (25). In obese men, there is a marked increase of E2 production (26). In addition, E2 levels in men increase further with advancing age (6,13), and elderly men may have higher concentrations of E2 compared with age-matched women (19). Elevated E2 levels in men are associated with myocardial infarction and coronary artery disease (27) and with an increased risk of stroke (13). Notably, high E2 concentrations are a significant predictor of poor prognosis and higher mortality in men with chronic heart failure and reduced left ventricular ejection fraction (4).
Although elevated E2 levels in older and/or obese men have been associated with an increased risk and incidence of cardiovascular disease, explanations for causal pathways and putative mechanisms for this association have not been identified (4). We suggest that MYLIP could contribute to this association, and we propose that MYLIP could become a pharmacological target in this high-risk group. Relevant to this notion and to support the clinical relevance of our findings, we found that MYLIP expression is higher in the hearts of men older than 50 years than in the hearts of men younger than 40 years. However, no measurements of E2 levels in these same individuals exist so that a direct association could be tested. Furthermore, questions remain about the precise actions of MYLIP that could increase cardiovascular risk. However, it was recently reported that MYLIP is involved in cholesterol regulation through the ubiquitination of the low-density lipoprotein receptor (28). Clearly, avenues for further research in this area are open.
We acknowledge that the sample size for human material was rather small, related to limited fresh material available from closely matched individuals, as well as a low number of publicly available datasets. Yet the internal consistency across the different experiments supports the conclusions. We have also limited ourselves for mouse experiments to 1 dose of E2. The dose was chosen based on literature data that indicate that this dose leads to in vivo E2 levels that at peak are 2.5-fold higher than baseline and within the range of values found in elderly men. Studying the effect of various doses of E2 in vivo would be of interest, but it would require a large number of animals not justified by the current question, namely to confirm a CM-specific effect.
We have shown that E2 exerts direct effects on gene expression in the heart and these actions may be different between the sexes. The present findings may offer insight into possible molecular mechanisms responsible for some of the sex-related differences observed in the development, prognosis, and outcome of cardiovascular disease, and they underscore the potential benefit from sex-specific therapies. Lastly, our work provides evidence and grounds for the notion that E2 is a master regulator through the modulation of gene expression, strongly suggesting an explanatory path for the prominent role of E2 in the development of sexually dimorphic traits in health and disease.
Dr. Kararigas acknowledges A. Brazma, G. Rustici, R. Bourgon, A. Kauffmann, F. Hahne, M. Kapushesky, and J. M. Vaquerizas for their unconditional guidance through the computational and statistical analysis of microarray data and for fruitful discussions. The authors thank V. Riese for technical assistance, and F. C. Luft for critical reading of the manuscript.
For supplemental methods and tables, please see the online version of this paper.
Transcriptome Characterization of Estrogen-Treated Human Myocardium Identifies MYLIP as a Sex-Specific Element Influencing Contractile Function
This work was supported by a Marie Curie Fellowship (to Dr. Kararigas) from the European Union (EU) through the program CARDIOVASC grant No. MEST-CT-2005-020268, the EU-funded EUGeneHeart research project grant No. LSHM-CT-2005-018833 (to Drs. Sipido and Regitz-Zagrosek), a grant from the German Research Foundation (DFG) (FG1054/1 to Dr. Regitz-Zagrosek), and a Travel Fellowship from the Boehringer Ingelheim Fonds (to Dr. Kararigas). Drs. Tinel and Albrecht-Küpper are employees of Bayer HealthCare. Dr. Regitz-Zagrosek has a research agreement with Bayer HealthCare; and has received speaker honorarium from Bayer HealthCare, Berlin Chemie AG, and Dr. Willmar Schwabe GmbH. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Dr. Becher's current address is the German Aerospace Center, Bonn, Germany.
- Abbreviations and Acronyms
- estrogen receptor
- myosin regulatory light chain
- myosin regulatory light chain interacting protein
- Received June 29, 2011.
- Revision received August 17, 2011.
- Accepted September 13, 2011.
- American College of Cardiology Foundation
- Kararigas G.,
- Fliegner D.,
- Gustafsson J.A.,
- Regitz-Zagrosek V.
- Toischer K.,
- Rokita A.G.,
- Unsold B.,
- et al.
- Bornhauser B.C.,
- Olsson P.A.,
- Lindholm D.
- Colson B.A.,
- Locher M.R.,
- Bekyarova T.,
- et al.
- Ding P.,
- Huang J.,
- Battiprolu P.K.,
- Hill J.A.,
- Kamm K.E.,
- Stull J.T.
- Olsson P.A.,
- Korhonen L.,
- Mercer E.A.,
- Lindholm D.
- Mahmoodzadeh S.,
- Dworatzek E.,
- Fritschka S.,
- Pham T.H.,
- Regitz-Zagrosek V.
- Wilson C.L.,
- Sims A.H.,
- Howell A.,
- Miller C.J.,
- Clarke R.B.
- Zelcer N.,
- Hong C.,
- Boyadjian R.,
- Tontonoz P.