(Circulation. 1997;95:709-714.)
© 1997 American Heart Association, Inc.
Articles |
the Perinatal Research Laboratory, Department of Obstetrics and Gynecology, University of Iowa, Iowa City.
Correspondence to Dr Loren P. Thompson, Perinatal Research Laboratory, Department of Obstetrics and Gynecology, University of Maryland School of Medicine, Bressler 11-040, 655 W Baltimore St, Baltimore, MD 21201-1595.
| Abstract |
|---|
|
|
|---|
Methods and Results Female guinea pigs were ovariectomized and allowed to recover for 100 days. Pellets containing 17ß-estradiol (0.25, 0.5, 1.5, and 7.5 mg released over 21 days) were placed subcutaneously for 19 to 20 days. Animals were then anesthetized, and the coronary arteries were excised and cut into ring segments. Rings were placed in small-vessel myographs for measurement of isometric force. Contractile responses of coronary arteries to cumulative addition of U46619 (10-10 to 10-5 mol/L), a thromboxane mimetic, were measured in the presence and absence of nitro-L-arginine (LNA), a selective NO synthase inhibitor, and methylene blue, a guanylate cyclase inhibitor. Low (0.25-mg) but not high (0.5-, 1.5-, or 7.5-mg) doses of estradiol inhibited the maximal contractile responses to U46619 compared with arteries from untreated castrated animals. In addition, both LNA and methylene blue potentiated contractile responses to U46619 of arteries from animals receiving 0.25 and 0.5 mg but not 1.5 and 7.5 mg estradiol. Negative log EC50 values were significantly inhibited at 0.25 and 7.5 mg but unaffected at 0.5 and 1.5 mg estradiol compared with castrated animals.
Conclusions Estradiol at low doses may protect against vasospasm by stimulating endothelium-derived NO release and inhibiting coronary artery contractility.
Key Words: estradiol contractility endothelium thromboxane vasoconstriction
| Introduction |
|---|
|
|
|---|
There are several mechanisms by which estrogen could contribute to a reduction in coronary vascular disease. Estrogen alters the plasma lipoprotein profile by decreasing LDL cholesterol and increasing HDL cholesterol.2 Estrogen may also influence the hemodynamic responses through its action on the blood vessel wall. A bolus injection of 17ß-estradiol (1 µg/kg) decreased sheep uterine and systemic vascular resistance after 120 minutes.14 15 Estrogen, whether administered long-term9 16 17 18 19 20 or short-term to animals or tissues,9 21 inhibits agonist-stimulated constriction and enhances endothelium-dependent dilation of some isolated blood vessels. Several mechanisms of action are proposed. Estrogen acts as a calcium channel antagonist when used short-term at relatively high concentrations22 23 24 25 26 and thus might modulate vascular reactivity via a direct effect on the smooth muscle. Long-term estrogen may also act indirectly through a genomic effect on protein synthesis18 27 28 and enhance the release of endothelium-derived NO.9 16 19 20
NO is an endothelium-derived vasodilator that modulates coronary blood flow.29 Disruption of the endothelium or inhibition of NO synthesis by selective NOS inhibitors contributes to coronary vasoconstriction.29 Because estrogen therapy decreases the risk of coronary artery disease in menopausal women and short-term estradiol supplementation to guinea pigs increases the expression of calcium-dependent NOS mRNA and myocardial NOS activity,27 28 30 31 we hypothesize that estrogen decreases coronary artery reactivity to constrictor agents by enhancing release of endothelium-derived NO. There are no studies that examine the effect of estradiol on coronary artery reactivity over a wide range of doses, despite the paradoxical effects of different doses on cardiovascular disease.1 12 13 Thus, we measured contraction to the cumulative addition of U46619, a thromboxane mimetic, of isolated coronary arteries from castrated guinea pigs receiving long-term 17ß-estradiol replacement therapy of varying doses. We chose U46619 because thromboxane is a potent vasoconstrictor released from activated platelets and is important in initiating in vivo the abnormal vasoconstriction in coronary artery disease.29
| Methods |
|---|
|
|
|---|
Animal and Tissue Preparation
Reproductively mature mixed-breed female guinea pigs (4 to 6 months old) were anesthetized with xylazine (1 mg/kg IM) and ketamine (80 mg/kg IP). Both ovaries were surgically removed through bilateral flank incisions. Castrated animals were allowed to recover for at least 100 days to ensure that the effects of estradiol had disappeared. Animals were then anesthetized and given 21-day continuous-release pellets containing either 0.25, 0.5, 1.5, or 7.5 mg 17ß-estradiol (Innovative Research of America) placed subcutaneously in the abdomen. Control animals were castrated but did not receive pellets. After 20 days, the hearts were removed from anesthetized animals through a thoracotomy and placed into cold physiological buffer solution. A section of the left anterior descending coronary artery on the epicardial surface was dissected free under a microscope, cleaned of adherent tissue, and cut into four rings. Arterial rings were suspended on wires and placed in tissue chambers containing temperature-regulated (37°C) physiological buffer (in mmol/L: NaCl 130, KCl 4.7, CaCl2 1.6, KH2PO4 1.18, NaHCO3 14.9, MgSO4 1.17, EDTA 0.03, and dextrose 11) and aerated with 95% O2/5% CO2. Rings were allowed to equilibrate for 1 hour. The optimal length of each segment was determined by stretching it until the maximal contractile response to 40 mmol/L KCl was achieved. Contractile responses to 120 mmol/L KCl were measured after the optimal length was determined. The vessel was repeatedly rinsed with fresh buffer every 20 minutes for an additional hour before the start of the experimental protocol.
Experimental Protocol
Contractile responses to the cumulative addition of the thromboxane mimetic U46619 (10-10 to 10-5 mol/L) were measured in arteries from untreated castrates and estradiol-treated guinea pigs. To determine the effect of estradiol on the contribution of the NO/cGMP pathway, contractile responses were measured in the presence and absence of either LNA (10-4 mol/L), a selective NOS inhibitor,32 or methylene blue (10-5 mol/L), a guanylate cyclase inhibitor.33 Arteries were treated in separate tissue chambers with either LNA or methylene blue for 30 minutes before the start of the thromboxane concentration-response curve. Contractile responses of treated arteries were compared with responses of untreated arteries from the same animal.
Statistics
Contractile responses to U46619 were normalized to the 120 mmol/L KCl contraction of each ring and reported as their mean±SEM. N indicates the number of animals in each group. Concentration-response curves to U46619 were compared by ANOVA with concentration, estradiol dose, and treatment as independent variables and contractile responses as a dependent variable. If the ANOVA demonstrated a significant difference among dose-response curves, a Newman-Keuls multiple comparison test was performed on individual concentrations. To further characterize the dose-response relationships, -log EC50 values were determined as an index of agonist potency. The -log EC50 value was obtained by linear regression analysis over the range of concentrations producing 20% to 80% maximal contraction. Emax was measured at 10-5 mol/L U46619. Statistical significance among Emax and -log EC50 values was determined with a Student's t test. Responses were considered statistically significant at P<.05.
Drugs
LNA and methylene blue were obtained from Sigma Chemical Co and dissolved in double-deionized water. U46619 was purchased from Cayman Chemical Co. Stock U46619 was dissolved in methyl acetate. An aliquot of stock U46619 was removed for each experiment and dried under argon gas to remove the solvent. U46619 was then reconstituted to the desired concentration in a sodium bicarbonate solution (1 mg/mL).
| Results |
|---|
|
|
|---|
|
|
|
To further illustrate the effect of the estradiol dose on the maximal contractile response to U46619, maximal responses to U46619 were normalized to the maximal response of arteries from the castrated animals (Fig 1B
). Maximal responses to U46619 were compared as a percentage of the mean maximal contractile response of the castrate (% change from castrate). Contractile responses to 10-5 mol/L U46619 of arteries from animals with 0.25 mg estradiol were significantly less than arteries from untreated castrates. The maximal responses of arteries from the other estradiol-treated animals were not significantly different from arteries of untreated castrates, although they appeared to be elevated at the higher doses (eg, 1.5 and 7.5 mg). It should be noted that the interanimal variation among responses rose as the replacement dose increased.
Arteries from animals receiving 7.5 mg estradiol had the largest variability in the contractile responses to U46619. This group of animals had
50% survival rate for the 20-day treatment period compared with almost 100% survival for all other estradiol-supplemented groups. The mechanism underlying the toxicity of the 7.5-mg dose was undetermined.
Effect of NOS and Guanylate Cyclase Inhibition on U46619 Contraction
To determine whether estradiol alters the U46619 contraction by modulating the NO/cGMP pathway, contractile responses were measured in the presence and absence of LNA and methylene blue (Fig 2A through 2E
). LNA had no effect on the -log EC50 but increased Emax values of arteries from untreated castrated guinea pigs and from animals receiving 0.25 and 0.5 mg estradiol (Table
s 1 and 2; Fig 2A, 2B, and 2C![]()
![]()
). LNA had no significant effect on either -log EC50 or Emax values of U46619 of arteries from animals receiving 1.5 mg (Fig 2D
) and 7.5 mg (Fig 2E
) estradiol compared with untreated arteries. Methylene blue also had no effect on -log EC50 values. Methylene blue increased Emax values in arteries from animals receiving 0.25 and 0.5 mg estradiol but had no effect on contractile responses of arteries from animals receiving either 1.5 or 7.5 mg estradiol.
|
| Discussion |
|---|
|
|
|---|
The effect of estradiol on coronary artery reactivity has been studied for several years. Yet, the mechanism by which estradiol inhibits vascular reactivity varies among several preparations and depends on whether the hormone is administered short-term or long-term. Estradiol relaxes arteries directly by stimulating endothelium-derived relaxing factors such as NO9 14 34 and prostacyclin35 and by acting as a calcium channel antagonist.22 25 The direct and immediate inhibitory effects of estradiol on vascular contractility support the hypothesis that estradiol can bind to plasma membrane sites distinct from the classic nuclear receptor.36 However, this does not provide a likely explanation for the effects of estradiol on U46619 contractility reported here. First, estradiol was given long-term, so its effect on contractility is more likely a genomic rather than a short-term membrane effect. Any possible short-term effects from estradiol release on the day the animal was killed are unlikely, because the arteries were placed in tissue baths without estrogen exposure several hours before the experiment was begun. Second, the inhibition of U46619 contractility by estradiol does not occur at the higher doses, as would be expected if the effect were a direct response to estradiol. If estrogen administered long-term inhibits KCl contraction by calcium channel antagonism, we would expect this response to increase with the estradiol dose. Although estradiol appeared to inhibit KCl contraction at two estradiol doses, 0.25 and 7.5 mg, the mechanism is unclear. Since the U46619 responses were normalized to the KCl contraction, the inhibitory effect of estradiol on U46619 contraction would be greater than would be observed if estradiol had a specific effect on KCl contraction. Thus, the differences measured among the KCl contractile responses are random and do not affect the interpretation of our results. Third, if estradiol were acting as a calcium channel antagonist in this study, the potentiation of U46619 contractility by LNA and methylene blue should be similar at all dose levels. Since it was not, we conclude that the most likely mechanism for the estradiol effect on the endothelium is genomic expression of NOS leading to altered NO release. We have previously reported preliminary results demonstrating that NOS activity in the guinea pig myocardium30 and the amount of eNOS-specific mRNA in skeletal muscle28 are increased with low-dose estradiol (0.25 and 0.5 mg) and decreased at doses >0.5 mg estradiol. In addition, we observe a similar phenomenon with brain NOSspecific mRNA in the guinea pig forebrain.31 Unfortunately, we are prevented from accurately quantifying eNOS-specific mRNA in the coronary artery, because we are limited by the small amount of tissue available.
We propose that low-dose estradiol in the guinea pig increases eNOS activity and/or cGMP levels of vascular smooth muscle and decreases contractility. Both inhibition of NOS and guanylate cyclase activation by LNA and methylene blue, respectively, potentiated U46619 contractility in arteries from castrates and from animals receiving 0.25 and 0.5 mg but not those receiving 1.5 mg and 7.5 mg estradiol. This suggests that enzyme activity is stimulated at certain estradiol levels but is inhibited as estradiol levels increase. In contrast, Williams et al37 reported that both low-dose (100 µg/kg) and high-dose (700 µg/kg) estradiol given to monkeys with intact ovaries as weekly intramuscular injections for 6 weeks decreased systemic vascular resistance. The difference from our study may reflect a species difference, a difference in the absolute estradiol levels and time course, or the presence of ovaries. We have previously demonstrated that a 5-day time course of a fairly high-dose estradiol supplementation (500 µg·kg-1·d-1) in guinea pigs with intact ovaries increases myocardial NOS activity.30 Thus, differences in animal species, the presence or absence of ovaries, and the duration of estrogen supplementation or replacement may each contribute to the disparity in the effect of estrogen on vascular responses among different studies.
It is difficult to distinguish in our model whether estradiol exerts a selective effect on either NOS or guanylate cyclase or whether it affects both enzymes. Since the effects of LNA and methylene blue on U46619 contraction parallel each other at all estradiol doses, we suggest that the potentiating effect of methylene blue is due to inhibition of the effect of NO on guanylate cyclase and not via a nonspecific effect of methylene blue on vascular smooth muscle. A limitation of this study is that the effect of estradiol on U46619 contraction of endothelium-denuded rings was not determined because of the technical difficulty in removing the endothelium from these small, thin-walled coronary arteries (ID,
220 µm). However, on the basis of the available evidence that NO is released from coronary artery endothelium, inhibited by L-arginine analogues,38 39 40 and absent from normal vascular smooth muscle,41 it is logical to propose that the effect of LNA is due to inhibition of NO release from the vascular endothelium.
Current evidence demonstrates that long-term estradiol supplementation decreases vascular tone.22 37 42 It also increases agonist-stimulated endothelium-dependent relaxation of isolated arteries16 20 and causes vasodilation of the intact coronary circulation.9 43 Basal NO release is greater in endothelium-intact aortic rings from female rabbits than in male rabbits, suggesting a dependence on the circulating estradiol levels.19 Estradiol increases the expression of eNOS-specific mRNA in skeletal muscle28 and rat aorta.27 Further, we have shown that in pregnancy in guinea pigs, when estradiol levels are increased, uterine artery contractility to norepinephrine is reduced and endothelium-dependent relaxation44 is increased by enhanced NO release. In addition, Ca2+-dependent NOS activity of the guinea pig uterine artery is increased during pregnancy.12 28 Thus, there is ample supportive evidence that estradiol enhances endothelium-derived NO production by vascular tissue. Since NO is a potent vasodilator and acts to decrease vascular responsiveness to vasoconstrictor substances, an increase in endothelium-derived NO release by long-term estrogen stimulation provides a mechanism by which the risk of coronary artery vasospasm may be reduced.
The paradoxical effect of estrogen on U46619 contractility and presumably endothelium-derived NO production may be partly explained by its effect on estrogen receptors. Estrogen receptors are located on vascular smooth muscle cells in the coronary arteries, aorta, and uterine artery.23 45 Low-dose estrogen enhances the expression of its receptor but also downregulates it in vascular tissues by translocation from the cytosol to the nucleus.46 47 This may explain how estradiol could both inhibit and potentiate vasoconstriction, depending on the dose. Estradiol receptor distribution between the cytosol and nuclear fraction differs among tissue types.46 47 It is greater in uterine smooth muscle than in vascular smooth muscle,47 suggesting that estrogen sensitivity may differ among organs and between smooth muscle types. Thus, differences in hormone sensitivity among different organ beds may provide a means for producing selective therapeutic benefit for particular organ beds. Further, Matsuda et al48 have reported that testosterone increases the expression of thromboxane A2 receptors in cultured rat aortic smooth muscle cells and suggest that thromboxane A2 receptors of vascular smooth muscle may be regulated by sex steroids. The effect of estradiol on vascular thromboxane A2 receptor expression has not been reported in the literature.
A few studies support the hypothesis that high-dose estrogen therapy is not beneficial in protecting against cardiovascular disease. In the Veterans Administration trial, men with coronary artery disease treated with high-dose estrogen had higher mortality rates than those treated with placebo.12 In addition, women who use oral contraceptives containing relatively high doses of estrogen have a higher incidence of myocardial infarction than nonusers.13 In patients with prostatic carcinoma receiving daily oral ethinyl estradiol (1 mg for the first 2 weeks followed by 150 mg), there was a pathological increase in the difference between systolic arm and toe blood pressures compared with untreated orchiectomized patients,49 suggesting that high-dose estrogen therapy may reduce arterial blood flow to the lower limbs. Further, in oophorectomized rats, long-term estradiol replacement (1 mg estradiol benzoate per rat once a week for 4 weeks) reduced limb blood flow (measured by 85Sr-microsphere uptake) to the tibia and femur,50 further supporting the hypothesis that high estradiol levels inhibit blood flow by increasing vascular resistance. Thus, the benefit of estrogen therapy on reducing vasoconstriction must be determined on the basis of the dose regimen given for each animal model.
In summary, estradiol at low doses inhibits thromboxane contractility of guinea pig coronary arteries. This is attributed to enhanced endothelium-derived NO release and increased cGMP levels, because both LNA and methylene blue potentiate contraction at the low estradiol doses. At higher estradiol doses, thromboxane contraction is similar to arteries from castrates receiving no estradiol, and LNA and methylene blue do not potentiate the contractile responses. Thus, there is a dose-dependent effect of estradiol that potentiates NO release but is inhibited at higher estradiol doses. These results suggest that estradiol reduces vascular smooth muscle sensitivity to vasoconstrictors by increasing endothelium-derived NO production and thus protects against vasospasm. This may provide the first potential explanation for the protective effect of estradiol against cardiovascular disease at low but not high estradiol levels.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received May 23, 1996; revision received September 6, 1996; accepted September 9, 1996.
| References |
|---|
|
|
|---|
and 7ß estradiol on PGF2a mediated contraction of porcine coronary artery. Adv Prostaglandin Thromboxane Leukotriene Res. 1989;19:277-280.[Medline]
[Order article via Infotrieve]
-nitro-L-arginine: a potent inhibitor of endothelium-derived relaxing factor formation. Eur J Pharmacol. 1990;176:219-223.[Medline]
[Order article via Infotrieve]This article has been cited by other articles:
![]() |
J. M. Muller-Delp, D. B. Lubahn, K. E. Nichol, B. J. Philips, E. M. Price, E. M. Curran, and M. H. Laughlin Regulation of nitric oxide-dependent vasodilation in coronary arteries of estrogen receptor-{alpha}-deficient mice Am J Physiol Heart Circ Physiol, November 1, 2003; 285(5): H2150 - H2157. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. Minshall, D. Pavcnik, P. V. Halushka, and K. Hermsmeyer Progesterone regulation of vascular thromboxane A2 receptors in rhesus monkeys Am J Physiol Heart Circ Physiol, October 1, 2001; 281(4): H1498 - H1507. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. P. Thompson, G. Pinkas, and C. P. Weiner Chronic 17{beta}-Estradiol Replacement Increases Nitric Oxide-Mediated Vasodilation of Guinea Pig Coronary Microcirculation Circulation, July 25, 2000; 102(4): 445 - 451. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Zhang and S. T. Davidge Effect of Estrogen Replacement on Vasoconstrictor Responses in Rat Mesenteric Arteries Hypertension, November 1, 1999; 34(5): 1117 - 1122. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. J. Knot, K. M. Lounsbury, J. E. Brayden, and M. T. Nelson Gender differences in coronary artery diameter reflect changes in both endothelial Ca2+ and ecNOS activity Am J Physiol Heart Circ Physiol, March 1, 1999; 276(3): H961 - H969. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. T. Davidge and Y. Zhang Estrogen Replacement Suppresses a Prostaglandin H Synthase–Dependent Vasoconstrictor in Rat Mesenteric Arteries Circ. Res., August 24, 1998; 83(4): 388 - 395. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |