INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CLINICAL STUDIES HAVE DEMONSTRATED that an increased risk of vasovagal syncope in elderly men and women is associated with reduced baroreflex sensitivity (12). Reported studies have shown that baroreflex sensitivity is reduced in postmenopausal women and restored to a normal level after estrogen replacement therapy (13, 32). These correlative clinical findings have been confirmed in experimental studies in which estrogen depletion and repletion were associated with reduced and enhanced baroreflex sensitivity, respectively (5, 19, 32). Nonetheless, most, if not all, of these studies dealt with the short-term impacts of estrogen depletion (2-3 wk) and estrogen repletion (minutes to hours) on baroreflex sensitivity in ovariectomized (Ovx) rats (5, 20). The mechanism by which estrogen enhances baroreflex sensitivity is not known. Furthermore, some suggested mechanisms, based on reported studies including our own, may pertain only to short-term (minutes to hours) effects on baroreflex pathways (5, 20, 25, 26), which may or may not apply to the long-term effects of estrogen on baroreflex sensitivity. To our knowledge, none of the reported studies has attempted to link the favorable effect of estrogen on baroreflex sensitivity to its cellular effects at the cardiac myocyte level. The vascular wall and heart tissue have been shown to contain specific high-affinity receptors for estrogen both in humans (16) and in animals (2). Several studies have documented estrogen ability to stimulate endothelial nitric oxide (NO) synthase (eNOS), which may explain, at least in part, the favorable effects of estrogen on vascular biology (10, 24).

eNOS, originally identified in large vessel endothelium, is also expressed in cardiac myocytes. In cardiac myocytes, eNOS is quantitatively associated with caveolin (6), the structural protein of caveolae, which serves to inhibit eNOS. Cell stimulation with Ca²⁺-mobilizing agonists promotes calmodulin binding to eNOS and caveolin dissociation from the enzyme, rendering the enzyme active (7). The caveolin isoform (caveolin-3) expressed in myocytes (6, 7, 30), therefore, modulates the catalytic activity of cardiac eNOS and hence regulates NO production and its biological effects. In this study, we tested the hypothesis that 17-estradiol (E₂) enhancement of baroreflex-mediated bradycardia, observed in humans and in experimental animals (5, 13, 19, 20), involves estrogen modulation of cardiac eNOS. We hypothesized that E₂ enhances the catalytic activity of cardiac eNOS by differentially modulating eNOS association with the inhibitory protein caveolin-3 and/or the stimulatory protein calmodulin. To test this hypothesis, we used sham-operated (control) rats, Ovx rats, and Ovx rats treated with E₂ (Ovx + E₂). E₂ or vehicle was provided over 12 wk to Ovx rats. This estrogen treatment resulted in physiological levels of serum E₂. At the end of the treatment period, baroreflex-mediated bradycardia was evaluated and the cardiac tissues from all groups were used for the measurements of NOS activity and eNOS expression as well as its association with caveolin-3 and calmodulin.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Twenty-four female Sprague-Dawley rats were used in the present study. The rats were obtained from Charles River (Raleigh, NC) at 9-10 wk of age (180-200 g). The rats were randomly separated into three groups: sham operated (n = 8), Ovx (n = 8), and Ovx + E₂ (n = 8). Sham operation or ovariectomy was performed in all rats. Experiments were performed in strict accordance with institutional animal care and use guidelines.

Ovariectomy. Bilateral ovariectomy was performed as described in our previous studies (5, 20). Briefly, the rat was anesthetized with methohexital sodium (50 mg/kg ip). The lower part of the back was shaved, and a single 2- to 3-cm incision was made in the skin to expose the back muscles. A small 1- to 2-cm incision was made in the muscles overlying the ovaries on both sides, and the ovaries were isolated, tied off with sterile suture, and removed. The muscles and the skin were sutured separately, and the rats were allowed to recover. Sham operation was performed by exposing the ovaries without isolation. After ovariectomy or sham operation, each rat received a subcutaneous injection of buprenorphine hydrochloride (Buprenex, 30 µg/kg) to control pain and an intramuscular injection of 50,000 U/kg of penicillin G and penicillin G procaine in an aqueous suspension (Durapen). Rats were housed in separate cages. Two weeks later, E₂ replacement (Ovx + E₂) or vehicle (Ovx) was administrated subcutaneously via osmotic minipumps for 12 wk (45 µg per rat per 30 days in 180 µl saline). The minipumps were replaced every 4 wk.

Intravascular cannulation. Two days before the experiment, the rats were prepared for measurement of blood pressure (BP) according to our previous studies (4-6). Briefly, the rats were anesthetized with methohexital sodium (50 mg/kg ip). Polyethylene-50 catheters filled with heparinized saline (100 U/ml) were introduced via the left femoral artery and vein to the abdominal aorta and inferior vena cava, respectively, for measurements of BP and intravenous injection of drugs, respectively. The catheters were tunneled subcutaneously and exteriorized at the back of the neck between the scapulae. Incisions were closed with surgical clips and swabbed with povidone-iodine solution. Each rat received a subcutaneous injection of buprenorphine hydrochloride (Buprenex, 30 µg/kg) to control pain and an intramuscular injection of 50,000 U/kg of penicillin G and penicillin G procaine in an aqueous suspension (Durapen). Rats were housed in separate cages.

Blood E₂ concentration measurements. Two weeks after sham operation or Ovx, blood samples were collected from the rats of each group (week 0). The blood samples were obtained from conscious rats held in restrainers. A volume of 0.5 ml of blood was taken from a tail vein. After E₂ replacement, blood samples were collected every week over the course of the experiment. The estrogen concentration in the samples was measured using an Estradiol DSL-4400 Radioimmunoassay kit (Diagnostic Systems Laboratories; Webster, TX) according to manufacturer's instructions, as in our previous studies (5, 20).

Baroreflex sensitivity measurements. On the day of the experiment, the arterial catheter was connected to a Gould-Statham pressure transducer, and BP was displayed on a Grass polygraph. Heart rate (HR) was computed from BP waveforms by a Grass tachograph and displayed on another channel of the polygraph. A period of 30 min was allowed at the beginning of the experiment for stabilization of BP and HR. A baroreflex curve was constructed by intravenous injection of phenylephrine (PE; 1-16 µg/kg) at 5-min intervals in all rats. PE was dissolved in saline, and the injection volume was kept constant at 0.05 ml/100 g body wt with a flush volume of 0.1 ml saline. The mean arterial pressure (MAP) and HR values before and after PE administration were used for construction of the baroreflex curves and calculation of baroreflex sensitivity as in our previous studies (5, 20).

NOS activity assays. A NOS Assay kit (Calbiochem; La Jolla, CA) was used to determine NOS activity. Briefly, ventricular tissue was homogenized in a homogenization buffer [25 mM Tris · HCl (pH 7.4), 1 mM EDTA, and 1 mM EGTA]. The tissue homogenate was centrifuged at 100,000 g for 60 min at 4°C. The pellet containing membrane-associated NOS was resuspended in homogenization buffer. The tissue extracts were incubated with a reaction buffer [50 mM Tris · HCl (pH 7.4), 6 µM tetrahydrobiopterin, 2 µM flavin adenine dinucleotide, and 2 µM flavin mononucleotide] containing 1.25 mM NADPH, 0.75 mM CaCl₂, and 1 µCi [³H]arginine at 37°C. To determine NOS activity, duplicate samples were incubated for 30 min in the presence of N-nitro-L-arginine methyl ester (L-NAME; 1 mM) or vehicle. The reaction was stopped by the addition of 400 µl of cold stop buffer [50 mM HEPES (pH 5.5) and 5 mM EDTA]. Equilibrated resin (100 µl) was added into each reaction samples. After centrifugation, the radioactivity of elute was quantified. Radiolabeled counts per minute of L-citrulline generation were measured and used to determine L-NAME-inhibited NOS activity.

Immunoblotting. Ventricular tissue was homogenized in a homogenization buffer [50 mM Tris (pH 7.5), 0.1 mM EGTA, 0.1 mM EDTA, 2 µm leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1% (vol/vol) Nonidet P-40, 0.1% SDS, and 0.1% deoxycholate]. After centrifugation (12,000 g for 5 min), protein was quantified in the supernatant using a modified Lowry assay procedure (the Bio-Rad Protein assay system). For Western blot analysis of eNOS, caveolin-3, and calmodulin, various amounts of protein (80, 40, and 40 µg, respectively) were loaded onto SDS-PAGE gels with the appropriate concentration (7.5%, 12%, and 12%, respectively) for best resolution. Proteins were transferred to nitrocellulose membranes after electrophoresis. After being blocked with 5% nonfat dry milk in Tris-buffered saline (TBS), the blots were incubated with the specified primary antibody [anti-eNOS dilution at 1:1,000, anti-caveolin-3 dilution at 1:5,000 (Transduction Laboratories; San Diego, CA), and anti-calmodulin dilution at 1:1,000 (Santa Cruz Biotechnology; Santa Cruz, CA)] in TBS buffer containing 5% nonfat dry milk overnight at 4°C. After four washes, the blots were incubated with second antibody dilution in TBS buffer containing 5% nonfat dry milk. After three additional washes in TBS buffer with 0.1% (vol/vol) Tween 20, the blots were detected by an ECL system and exposed to X-ray film (29).

Immunoprecipitation. Ventricular tissue lysates were prepared using the same method as that described for immunoblotting. Lysates containing 5 mg of detergent soluble protein were precleared by incubating with Protein A-Sepharose for 1 h at 4°C and then transferred to a fresh tube. Fourteen microliters of anti-eNOS polyclonal antibody (250 µg/ml, Transduction Laboratories) were added and gently mixed for 1 h at 4°C. Fifty microliters of 50% Protein A-Sepharose slurry were added to the mixture. After an overnight incubation rotating at 4°C, immune complexes were collected by centrifugation, washed four times with 1 ml of immunoprecipitation buffer lacking protease inhibitors, and disrupted by boiling in immunoblotting sample buffer. The supernatant was then analyzed by SDS-PAGE, followed by protein immunoblotting for detection (29).

Immunohistochemical analysis. Ventricular tissues were fixed in 4% paraformaldehyde-PBS solution for 4 h at 4°C. The tissues were then transferred to 20% sucrose-PBS solution and incubated for 48 h at 4°C. The sucrose-infiltrated tissues were frozen and sectioned. Cryostat sections (10 µm in thickness) were immunostained with anti-eNOS polyclonal antibody (Transduction Laboratories) using a modification of the avidin-biotin complex method (ABC). An ABC kit (Vector Laboratories; Burlingame, CA) was used to perform immunohistochemistry. Briefly, sections were incubated serially with the following solutions: 1) 0.1% H₂O₂ for 30 min to block endogenous peroxidase activity, 2) normal animal serum solution containing 0.4% Triton X-100 for 1 h to reduce nonspecific binding, 3) primary antibody against eNOS (1:100) solution for 48 h at 4°C, 4) diluted biotinylated secondary antibody solution for 1 h at 4°C, 5) ABC reagent solution for 30 min at room temperature, 6) peroxidase substrate solution until the desired stain intensity developed, and 7) clear and mount. PBS (pH 7.4) was used to dilute each solution and to wash the sections for three times after each step (9).

Statistical analysis. Data are presented as means ± SE. The relationship between increases in the MAP evoked by PE and the associated decreased in HR was assessed by regression analysis for individual animals as described in our previous studies (5, 20). The regression coefficient (slope of the regression line), expressed as beats per minute per millimeter of mercury, was taken as an index of baroreceptor reflex sensitivity. ANOVA was used in the analysis of data with the level of significance set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Blood E₂ concentration in sham-operated, Ovx, and Ovx + E₂ rats. The baseline values of the blood E₂ concentration in rats of the three groups were similar (data not show). Two weeks after the Ovx operation, the blood E₂ concentration was markedly decreased compared with that in sham-operated rats at week 0 (Fig. 1). The blood E₂ level in the Ovx group remained at a low level throughout the experiment, with an average of 4.59 ± 0.72 pg/ml. E₂ replacement resulted in an increased blood E₂ concentration. During the first 6 wk after E₂ replacement, the blood E₂ concentration in Ovx + E₂ rats was higher than that in sham-operated rats; it then declined to control levels by weeks 10-12. The average blood E₂ level in Ovx + E₂ group rats was 20.9 ± 3.05 pg/ml. In the sham-operated rats, there was greater variability in the E₂ level compared with other groups, perhaps because samples were obtained without regard for estrous stage (average, 14.55 ± 3.42 pg/ml in weeks 0-12). In general, the E₂ level in the sham-operated rats was higher than that in the Ovx rats but lower than that in Ovx + E₂ rats. Nonetheless, the E₂ levels in Ovx + E₂ rats were within normal range.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   The blood estradiol concentrations in sham-operated rats or ovariectomized rats treated with vehicle (Ovx) or 17-estradiol (Ovx + E2) measured 2 wk after sham or Ovx operation (week 0). * P < 0.05 compared with sham-operated values; #P < 0.05 compared with Ovx + E2 values.

Effect of Ovx and estrogen replacement on baroreflex sensitivity. The baseline values of MAP (104.9 ± 2.3, 99.7 ± 4.7, and 100.9 ± 3.3 mmHg) and HR (430 ± 5, 437 ± 22, and 410 ± 14 beats/min) of sham-operated, Ovx, and Ovx + E₂ rats, respectively, were similar. The increase in MAP and decrease in HR evoked by intravenous bolus administration of PE (1-16 µg/kg) in sham-operated, Ovx, and Ovx + E₂ rats are shown in Fig. 2, A and B. PE elicited dose-dependent similar pressor responses in all groups of rats. In contrast, the bradycardic responses associated with comparable PE-evoked presser responses were significantly attenuated in Ovx compared with sham rats (P < 0.05). E₂ replacement enhanced the bradycardic responses of Ovx rats to similar values as those observed in sham-operated rats (Fig. 2, B and C). On the other hand, no significant difference existed when the baroreflex-mediated tachycardic responses to reductions in BP elicited by nitroprusside were compared in all groups (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Phenylephrine-evoked increments in mean arterial pressure (MAP; A) and the associated bradycardic responses [heart rate (HR); B] obtained in sham-operated, Ovx, and Ovx + E2 rats. Baroreflex sensitivity (BRS; C), measured by phenylephrine, was significantly reduced by ovariectomy and restored to sham-operated values after E2 replacement (P < 0.05). Values are means ± SE. * P < 0.05 vs. sham-operated values; #P < 0.05 vs. Ovx + E2 values.

Effect of estrogen on myocardial NOS activity. Twelve weeks after E₂ replacement, ventricular NOS activity was significantly reduced by 32.3 ± 9.5% in Ovx rats compared with sham-operated control rats (Fig. 3). E₂ replacement restored myocardial NOS activity to that of the control (sham operated) level (Fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Effect of estrogen depletion and replacement on nitric oxide synthase (NOS) activity measured by L-[3H]citrulline generation in homogenized ventricular tissue. Membrane-associated NOS activity was significantly (* P < 0.05) reduced in Ovx rats compared with sham-operated rats (n = 6 rats/group) and increased in Ovx + E2 rats (n = 6) compared with Ovx rats (#P < 0.001).

eNOS protein expression in the myocardium of sham-operated, Ovx, and Ovx + E₂ rats. As shown in a representative blot (Fig. 4A) and by densitometry analysis (n = 6 rats/group; Fig. 4B), a significant reduction in the ventricular eNOS protein level was detected in Ovx rats compared with sham-operated rats. Treatment of Ovx rats with E₂ increased the eNOS protein expression to a level similar to that observed in sham-operated rats (Fig. 4). Comparable results were obtained from an immunohistochemical study using the same antibody. Myocytes in the sections of sham-operated rat ventricles showed a diffused background with granular positive focal staining. Strong staining intensity for the endothelium of microvessels could also be observed (Fig. 5A). In the sections from Ovx rats, ventricular myocytes displayed weaker staining for eNOS compared with sham-operated rats, and the endothelium of microvessels showed moderate staining intensity (Fig. 5B). Treatment with E₂ restored ventricular as well as endothelial eNOS to control (sham operated) levels (Fig. 5C).

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Effect of estrogen on ventricular endothelial NOS (eNOS) protein expression. A: protein lysates were prepared from ventricular tissues from sham-operated, Ovx, and Ovx + E2 rats. Equal amounts of protein (80 µg) were subjected to 7.5% SDS-PAGE and analyzed by immunoblotting as detailed in MATERIALS AND METHODS. eNOS expression was reduced in Ovx rats compared with sham-operated rats, whereas E2 replacement restored eNOS expression to control (sham operated) levels. B: densitometric analysis of blots (n = 6). Values are means ± SE. * P < 0.05 vs. sham-operated values; #P < 0.05 vs. Ovx + E2 values.

View larger version (86K): [in this window] [in a new window]   Fig. 5.   Immunohistochemical detection of eNOS in the ventricular myocytes of sham-operated, Ovx, and Ovx +E2 rats. A: myocytes showing a diffused background with granular positive focal staining in the sections of sham-operated rat ventricles. Strong staining for the endothelium of microvessels could also be observed. B: in the sections of Ovx rat ventricles, myocytes displayed weaker staining for eNOS compared with that of sham-operated rats; the endothelium of microvessels showed moderate staining. C: treatment of Ovx rats with E2 restored eNOS staining to a staining similar to that in sham-operated rat ventricles. The solid line in B represents 100 µm.

Coassociation of caveolin-3 and calmodulin with eNOS in hearts of sham-operated, Ovx, and Ovx + E₂ rats. As shown in Fig. 6, there was an enhanced binding of caveolin-3 with eNOS in tissue lysates of hearts of Ovx rats compared with sham-operated rats. Treatment with E₂ decreased the binding of caveolin-3 with eNOS in the hearts of Ovx + E₂ rats compared with Ovx rats. On the other hand, the calmodulin signal detected in primary eNOS immunopricipitates showed a looser association in the cardiac tissue lysates from Ovx rats compared with control (sham operated) values. E₂ replacement increased the binding of calmodulin with eNOS in the cardiac tissue lysates of Ovx + E₂ rats.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Analysis of eNOS association with caveolin-3 and calmodulin in the hearts of sham-operated, Ovx, and Ovx + E2 rats measured by immunoprecipitation (IP). Protein lysates were prepared from sham-operated, Ovx, and Ovx + E2 rat ventricular tissues. Tissue lysates containing 5 mg detergent-soluble protein were subjected to immunoprecipitation with 14 µl anti-eNOS polyclonal antibody (250 µg/ml). Proteins were separated and analyzed by immunoblotting as detailed under MATERIALS AND METHODS. The association of eNOS with caveolin-3 was increased in Ovx rats compared with sham-operated rats, whereas it was decreased in Ovx + E2 rats compared with Ovx rats. The association of eNOS with calmodulin was decreased in Ovx rats compared with sham-operated rats, whereas it was increased in Ovx + E2 rats. Results are representative of 3 independent experiments.

Caveolin-3 and calmodulin protein level in heart tissues from sham-operated, Ovx, and Ovx + E₂ rats. As shown in Fig. 7, caveolin-3 expression was similar in tissue lysates obtained from the hearts of the three groups of rats. Similarly, there were no significant differences in calmodulin expression in heart tissues of the three groups of rats.

View larger version (40K): [in this window] [in a new window]   Fig. 7.   Effect of estrogen on cardiac expression of caveolin-3 and calmodulin. Protein lysates were prepared from the ventricular tissues of sham-operated, Ovx, and Ovx + E2 rats. Equal amounts of protein (40 µg for caveolin-3 and 40 µg for calmodulin) were loaded onto 12% SDS-PAGE gels and analyzed by immunoblotting as detailed under MATERIALS AND METHODS. Caveolin-3 and calmodulin expressions were unaltered in the hearts of Ovx or Ovx + E2 rats compared with those in sham-operated rats. Results are representative of 3 independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates the importance of estrogen in the modulation of eNOS protein expression and activity in the rat heart. The study was extended to investigate the involvement of two counterbalancing allosteric modulators, caveolin-3 and calmodulin, in the regulation of eNOS activity by estrogen. Our data show 1) ovariectomy reduced the enzymatic activity as well as the protein expression of eNOS in cardiac tissue and 2) eNOS activity is regulated partly by direct protein interactions between the enzyme and its two regulatory proteins, caveolin-3 and calmodulin. These results provide the first analysis of the molecular mechanisms underlying the estrogen-dependent regulation of eNOS expression/activity in the rat heart and its possible role in estrogen enhancement of baroreflex-evoked bradycardia.

We reasoned that alterations in eNOS-derived NO in cardiac myocytes may contribute to the changes in baroreflex sensitivity observed with chronic alterations in estrogen status because 1) cardiac myocyte eNOS-derived NO contributes to the parasympathetic mediated bradycardic response via a mechanism that involves the activation of the myocyte muscarinic receptor (7) and 2) estrogen enhances the baroreflex-mediated bradycardic, but not the tachycardic, response (20), which suggests an interaction between estrogen and the parasympathetic limb of the baroreflex arc because the latter plays the major role in this response (1, 4). Therefore, we investigated the effects of chronic estrogen depletion and repletion on the regulation of cardiac NOS activity and on baroreflex sensitivity in female rats. Initially, we studied whether chronic hormone changes after Ovx and E₂ replacement can influence myocardial NOS activity. We found that the myocardial NOS activity was decreased in Ovx rats compared with sham-operated control rats, whereas chronic E₂ treatment of Ovx rats was accompanied by a restoration of normal activity of myocardial NOS. The alterations in myocyte NOS-derived NO production were associated with similar changes in baroreflex sensitivity in the same rats. The present finding that the reduced baroreflex sensitivity in Ovx rats, which agrees with reported findings (5, 19, 20), was associated with decreased activity of eNOS in the myocardium seems to implicate eNOS-derived NO in the mediation of the baroreflex sensitivity response. In support of this notion, reported findings have shown that inhibition of NOS activity by chronic treatment with NOS inhibitors, such as L-NAME, decreases baroreflex sensitivity in conscious rats (23, 28). There is, however, some controversy with regard to the role of NOS-derived NO in the regulation of baroreflex sensitivity. Liu et al. (15) found that NOS inhibition increased baroreflex sensitivity in conscious rabbits. It is possible that the animal species, the duration of NOS inhibition, and the method of baroreflex sensitivity measurement may account for the differences in the findings of the latter group and the reported findings (23, 28).

The results of the present study also show that estrogen treatment of Ovx rats restores baroreflex sensitivity to control levels, which highlights a role for chronic (monthly) estrogen replacement in the modulation of the baroreflex control of HR. Previous studies including our own have dealt with the impact of short-term (minutes to hours) estrogen treatment on baroreflex sensitivity (5, 19, 20, 25, 26). The clinical relevance of the present finding is apparent because it reflects similar findings in women (13). It is also important to note that we have demonstrated, for the first time, that the restoration of baroreflex sensitivity in Ovx-E2 treated rats was associated with a restoration of eNOS-derived NO production in cardiac myocytes of these rats.

Therefore, the results of the present study provide correlative evidence consistent with the hypothesis that alterations in myocyte eNOS may explain the changes in baroreflex sensitivity caused by chronic alterations in estrogen levels. Taken together, it seems that NOS-derived NO acts to enhance baroreflex-mediated bradycardia in conscious rats. Nonetheless, it is imperative to point out that other mechanisms may also contribute to estrogen effects on baroreflex sensitivity. For example, a previous study from our laboratory has shown that after short-term (2 day) exposure, estrogen enhancement of baroreflex sensitivity involves the central nervous system (19). More recently, estrogen has been shown to modulate neuronal NOS activity in the brain stem, which contains neuronal pools implicated in baroreflex control (31).

It was important to investigate the mechanism by which E₂ enhances NO production in cardiac tissue. The latter is regulated not only by the amount of NOS protein in cardiac tissue but also by its catalytic activity. We demonstrated that the expression of cardiac eNOS proteins was downregulated by chronic estrogen depletion and normalized upon E₂ replacement. Therefore, it is reasonable to conclude that the reduction in NO production in Ovx rats resulted, at least partly, from reduced cardiac eNOS protein expression. In agreement with our findings, others have shown that estrogen facilitates the expression of eNOS level and activity in endothelial cells (11, 14, 17, 18). These effects are estrogen specific because they are not mediated by progesterone (14) and are inhibited by estrogen receptor blockade (17). Nonetheless, the present results are the first to demonstrate the enhancement of eNOS expression by estrogen in the rat heart.

The precise molecular mechanism of estrogen regulation of eNOS activity still remains unknown. Notably, eNOS activity is under a dynamic regulation, and several molecules have been implicated in modulating eNOS activity, such as caveolin, calmodulin, and heat shock protein 90 (8, 27). For example, eNOS/caveolin association is very important in the physiological function of eNOS. NO production may be regulated through a reciprocal and competitive interaction of calmodulin and caveolin with eNOS, with NOS activation and inhibition promoted through binding with calmodulin and caveolin, respectively. Here, we demonstrated an enhanced association of eNOS with caveolin-3 and a diminished association of eNOS with calmodulin in the hearts of Ovx rats. Furthermore, chronic E₂ administration to Ovx rats resulted in normalization of the interaction of eNOS with caveolin-3 and calmodulin in the heart and restored vagally mediated bradycardia. These results suggest that the E₂ interacton with cardiac eNOS, which is directly correlated with the favorable effects on the cardiac baroreflex response, involves not only an upregulation of eNOS protein expression but also the modulation of eNOS association with its regulatory proteins, calmodulin and caveolin-3.

Because an enhanced association of eNOS with caveolin-3 and a diminished association of eNOS with calmodulin in the Ovx rat heart have been found, it was important to determine whether chronic treatment with estrogen could influence the expression of caveolin and calmodulin proteins in cardiac myocytes. Notably, in the hepatic tissue from cirrhotic animals, a reduced activity of eNOS was associated with a sevenfold increase in the binding of eNOS with caveolin-1, and the protein expression of caveolin-1 was markedly increased (29). Furthermore, it has recently been reported that caveolin-1 and vascular eNOS expressions in cerebral blood vessels are altered, in opposite directions, with chronic changes in estrogen status in female rats (33). In contrast, the present results show that chronic changes in estrogen status had no impact on caveolin-3 or calmodulin expression in the rat cardiac myocyte. Therefore, it seems that estrogen influences the activity of cardiac eNOS by regulating the association of eNOS with its regulatory proteins rather than affecting the expressions of caveolin-3 or calmodulin proteins. Finally, the increased cardiac NOS activity represents the sum of the activity of all three NOS isoforms. However, because the basal expression of neuronal NOS and inducible NOS in heart tissues is negligible (21, 34), their contribution to the results obtained in the present study was expected to be minimal. Furthermore, because the method employed in the present study allowed measurement of the membrane-associated NOS activity (3, 22), the results suggest that the activity of the membrane-associated eNOS in cardiac myocytes is regulated by estrogen.

In summary, we report that chronic alterations in circulating estrogen levels have a profound influence on eNOS activity and its protein expression in cardiac myocytes. Our findings demonstrate that the effects of estrogen on cardiac eNOS functional activity are not limited to an increased eNOS protein expression. We report, for the first time, that E₂ reduces the association of cardiac eNOS with the inhibitory protein caveolin-3 and increases the association with the facilitator protein calmodulin. Our findings may help to elucidate the molecular mechanism underlying the effects of estrogen on cardiac eNOS activity and perhaps explain the favorable effect of estrogen on baroreflex sensitivity. Nonetheless, the possibility cannot be ignored that estrogen may enhance baroreflex sensitivity also by acting on pathways within the baroreflex arc other than the cardiac myocyte.