Short-term exposure to physiological levels of 17b-estradiol enhances 1

Transcription

1 Cardiovascular Research 42 (1999) Short-term exposure to physiological levels of 17b-estradiol enhances 1 endothelium-independent relaxation in porcine coronary artery * Hwee Teoh, Susan W.S. Leung, Ricky Y.K. Man Department of Pharmacology, Faculty of Medicine, The University of Hong Kong, 1/F Li Shu Fan Building, 5Sassoon Road, Hong Kong, PR China Received 10 June 1998; accepted 6 August 1998 Abstract Objectives: While alterations in cholesterol and lipoprotein profiles partly account for menopause being a risk factor for coronary heart disease, recent studies have suggested that 17b-estradiol may have vascular effects. Our aims were to study the short-term effects of 17b-estradiol on vascular function in isolated porcine coronary artery rings. Concomitantly, we sought to determine if physiological concentrations of 17b-estradiol could acutely potentiate relaxation. Results: 17a- and 17b-estradiol at pharmacological (.1 mm) concentrations produced relaxation in U46619-pre-contracted porcine coronary artery rings. Relaxation evoked by 17b-estradiol was not reversed by the estrogen receptor antagonists tamoxifen and ICI Following 20 min exposure to a physiological concentration of 17b-estradiol (1 nm), which on its own had no effect, relaxation elicited by cromakalim, levcromakalim and sodium nitroprusside, but not bradykinin or calcium ionophore A23187, were significantly enhanced. This potentiating action was also insensitive to tamoxifen and ICI Our data provide evidence for an acute indirect relaxant action of 17b-estradiol and suggest that it may be via a tamoxifen- and ICI insensitive estrogen receptor. While this response was only observed at pharmacological concentrations, the potentiation of cromakalim, levcromakalim and sodium nitroprusside relaxation was evident in the presence of a physiological concentration (1 nm) of 17b-estradiol. Conclusions: These results demonstrate that short-term exposure to 17b-estradiol, at concentrations that have no effect on their own, can enhance vasorelaxation. These vascular effects may partly account for some of the acute effects of 17b-estradiol on blood flow Elsevier Science B.V. All rights reserved. Keywords: Estradiol; Vessel; Coronary; Vasodilation; Swine 1. Introduction Increasingly, however, many investigators have studied the vascular effects of 17b-estradiol. Chronic [14 16] and Prior to menopause, women have a lower incidence of acute [17 19] administration of 17b-estradiol at physiocoronary heart disease (CHD) compared to age-matched logical levels in vivo enhances endothelium-dependent men [1 3]. Estrogen replacement therapy reduces CHD vasodilation. Conversely, supraphysiological concentramortality in post-menopausal women [4,5] and diminishes tions of the steroid probably act via the smooth muscle coronary artery atherosclerosis in primates [6]. The protec- cells to elicit relaxation [11,12,20]. Hence, it has been tive effects of estrogen are partially attributable to its proposed that the vascular actions of estrogen may contribfavorable alterations of serum lipoprotein levels [4,7,8], its ute, in part, to the cardioprotective effects in preanti-oxidant properties [9,10] and its actions as a calcium menopausal women [21]. antagonist [11] on voltage operated calcium channels Currently, the mechanism(s) of the acute and chronic [12,13]. vascular effects of estrogen at physiological or conventional pharmacological plasma concentrations remain unclear. * Corresponding author. Tel.: ; fax: The aim of the present study was to determine if physio- address: hteoh@hkusua.hku.hk (H. Teoh) 1 See pages Time for primary review 28 days / 99/ $ see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S (98)00265-X

2 H. Teoh et al. / Cardiovascular Research 42 (1999) logical concentrations of 17b-estradiol could acutely layers after perfusion with Triton X-100 if they exhibited potentiate relaxation in vitro. The rationale for using pigs less than 20% relaxation in response to the same conas an animal model was based on the similarities between centration of bradykinin. After removal of these drugs by the porcine and human hearts. repeated washings, the rings were contracted under the same conditions as described above and the relaxational properties were measured by cumulative additions of the 2. Methods appropriate agents. In some experiments, 17b-estradiol (final concentration of nm) was added to the 2.1. Materials baths 20 min prior to contraction with U46619 and remained present throughout the experiment. Whenever the U46619 (9,11-dideoxy-9a,11a-methanoepoxy prosta- experimental protocol required the use of antagonists, glandin F 2a) was obtained from Biomol (PA, USA) and these agents were added together with indomethacin 20 ICI was a gift from Zeneca (Macclesfield, UK). min before the addition of estradiol. Cromakalim and levcromakalim were generous gifts from SmithKline Beecham, UK. All other chemicals were from 2.3. Data and statistical analyses Sigma (St. Louis, MO, USA). Stocks of 17a-estradiol, 17b-estradiol, ICI , cromakalim, levcromakalim All data shown denote mean6s.e.m. with n indicating and U46619 were made up in absolute ethanol while the number of porcine hearts. Relaxations elicited by the tamoxifen was dissolved in 10% ethanol, A23187 in various vasodilators were expressed as a percentage of DMSO and indomethacin in a buffered sodium carbonate U46619-induced contraction. EC50 values were determined solution. All working solutions were obtained by dilutions using a curve-fitting program (SIGMAPLOT, Jandel Scientific in Krebs Henseleit buffer. Software, CA, USA). The results were analyzed with Student s t-test for paired and unpaired observations Functional studies Analysis of variance (ANOVA) and Dunnett s test were applied where appropriate to determine individual differ- Pigs were processed according to the regulations laid ences between multiple groups of data using a computer down by the Department of Agriculture, Government of statistical package (SIGMASTAT, Jandel Scientific Software). Hong Kong. Porcine hearts of either sex were collected A P value of,0.05 was considered significant. from the local abattoir and rinsed in cold, oxygenated (O 2:CO 2, 95%:5%) Krebs Henseleit solution (KHS; composition in mm: 120 NaCl, 4.76 KCl, 25 NaHCO 3, Results NaH2PO 4?H2O, 1.25 CaCl 2, 1.18 MgSO 4?H2O, 5.5 glucose) before the left anterior descending and right coronary 3.1. Effect of estradiol isomers on pre-contracted artery arteries were isolated. After removal of connective tissue, rings intact coronary artery rings (3-mm wide) were suspended between stainless steel stirrups and stationary support rods U46619 (30 nm) produced g contractions positioned in 5-ml jacketed organ baths filled with oxy- (n575) in porcine coronary artery rings. As illustrated in genated KHS maintained at 378C. To investigate drug Fig. 1, 17b-estradiol ( mm) produced full effects on endothelium-independent relaxation, some por- relaxation of U46619-pre-contracted rings in a concencine coronary arteries were divided into two sections and tration-dependent manner while 100 mm 17a-estradiol perfused for 30 s at 0.5 ml/ s with either 0.5% Triton only elicited relaxations of % (n58). Additions X-100 in KHS or KHS alone. Artery rings were placed of the estradiol vehicle instead of estradiol showed that it under a 2 g tension for 120 min before commencement of did not influence the effects of either stereoisomer sigthe experiment. Bath KHS was changed every 15 min nificantly (data not shown). The action of 17b-estradiol during the whole equilibration period. Isometric tension was not dependent on the presence of intact endothelium was measured by force transducers (FT03, Grass, USA) since pretreatment with Triton X-100 had no significant coupled to an amplifier and a personal computer for data effect (data not shown). This direct relaxant effect of collection (PICO Data Logger, Pico Technology, Cam- 17b-estradiol in porcine coronary artery rings was both bridge, UK). tamoxifen- and ICI insensitive (Fig. 2). The viability of each porcine coronary artery ring was determined by pre-contracting with 30 nm U46619 in the 3.2. Effect of physiological concentrations of 17bpresence of 10 mm indomethacin (a cyclooxygenase estradiol on porcine coronary artery rings inhibitor) and relaxing with 1 mm bradykinin. Only rings that produced $4 g contraction and demonstrated $80% relaxation were used for further studies. Porcine coronary artery rings were considered to have damaged endothelial Bradykinin (0.1 nm 1 mm), A23187 (0.1 nm 1 mm), cromakalim (0.1 nm 100 mm), levcromakalim (0.1 nm 100 mm) and sodium nitroprusside (1 nm 100 mm) all

3 226 H. Teoh et al. / Cardiovascular Research 42 (1999) Fig. 1. Direct relaxing effects of 17a-estradiol (s) and 17b-estradiol (d) on U46619 (30 nm) pre-contracted porcine coronary artery rings. Mean U46619-generated force in the rings treated with 17a- estradiol and 17b-estradiol were g and g, respectively (n58). *P,0.05 vs. corresponding 17b-estradiol effect. relaxed U46619 pre-contracted tissues (Figs. 3 5). Preincubation with 17b-estradiol (1 100 nm) for 20 min did not affect U46619 contraction. The relaxation effects elicited by bradykinin and A23187 were also not significantly altered by 17b-estradiol (Fig. 3). However, 1 nm of 17b-estradiol significantly potentiated cromakalim-, levcromakalim- and sodium nitroprusside-induced endo- Fig. 3. Effects of 17b-estradiol on endothelium-dependent relaxation of porcine coronary artery rings. Tissues were incubated with 17b-estradiol (j, 1 nm; m, 10nM;, 100 nm) or an equal volume of ethanol solvent (s) for 20 min before contracting with 30 nm U Rings were then relaxed with (a) bradykinin or (b) A (n58 for each point). Fig. 2. Effects of 17b-estradiol in the absence (m) and presence of tamoxifen (j, 1 nm; h, 10mM) or ICI (d, 1 nm; s, 10mM) on U46619 (30 nm) pre-contracted porcine coronary artery rings (n57). thelium-independent relaxations (Figs. 4 and 5). Incubation of porcine coronary artery rings with this concentration of 17b-estradiol for 20 min reduced the EC50 values of cromakalim, levcromakalim and sodium nitroprusside from , and mm to , and mm, respectively (n56, 6 and 8, respectively; P,0.05 in all cases). Interestingly, the

4 H. Teoh et al. / Cardiovascular Research 42 (1999) Fig. 5. Effects of 17b-estradiol on sodium nitroprusside-induced relaxation of porcine coronary artery rings. Tissues were incubated with 17b-estradiol (d, 0.1 nm; j, 1 nm; m, 10 nm) or an equal volume of ethanol solvent (s) for 20 min before contracting with 30 nm U Rings were then relaxed with sodium nitroprusside (n58). *P,0.05 vs. corresponding control data. other hand, the enhancement observed with 1 nm 17bestradiol was similar to that induced by 10 nm of the steroid, an effect which was also present in endotheliumfree porcine coronary artery rings (Fig. 7). In contrast to 1 nm 17b-estradiol, the same concentration of 17a-estradiol failed to significantly influence the relaxation elicited by sodium nitroprusside (Fig. 8). In this series of experiments, the EC50 value under control conditions was mm while those after treatment with 17b-estradiol and 17a-estradiol were mm (P,0.05 when compared with control data) and mm (P.0.05 when compared with control data) respectively. 4. Discussion Fig. 4. Effects of 17b-estradiol on endothelium-independent relaxation of porcine coronary artery rings. Tissues were incubated with 1 nm 17b- We report here that acute treatment of porcine coronary estradiol (d) or an equal volume of ethanol solvent (s) for 20 min before contracting with 30 nm U Rings were then relaxed with (a) artery rings with a low concentration (1 nm) of 17b- cromakalim (n56) or (b) levcromakalim (n56). *P,0.05 vs. corresponding estradiol can affect vascular function indirectly through control data. enhancement of vasorelaxation to cromakalim, lev- cromakalim and sodium nitroprusside. In accord with previous findings, we also demonstrate that estradiol can leftward shifts of the cromakalim, levcromakalim and induce direct relaxation in a stereospecific and concensodium nitroprusside concentration response curves fol- tration-dependent manner. In addition, both the direct and lowing exposure to 1 nm 17b-estradiol were also apparent indirect vascular effects of 17b-estradiol appear to be in the presence of 10 mm tamoxifen or 10 mm ICI mediated via a tamoxifen- and ICI insensitive (Fig. 6). As shown in Fig. 5, 0.1 nm 17b-estradiol did not mechanism. More importantly, while the direct effects of significantly affect sodium nitroprusside relaxation. On the this sex hormone only occurred at pharmacological con-

5 228 H. Teoh et al. / Cardiovascular Research 42 (1999) Fig. 7. Effects of 17b-estradiol on sodium nitroprusside-induced relaxation of intact (s, d) and denuded (h, j) porcine coronary artery rings. Tissues were incubated with 1 nm 17b-estradiol (d, j) or an equal volume of ethanol solvent (s, h) for 20 min before contracting with 30 nm U Rings were then relaxed with sodium nitroprusside in a cumulative manner (n58). Mean U46619-generated force in intact and denuded rings were g and g, respectively., *P,0.05 vs. corresponding control data from intact and denuded rings, respectively. Fig. 6. Effects of 1 nm 17b-estradiol in the absence and presence of 10 mm tamoxifen or 10 mm ICI on endothelium-independent relaxation of porcine coronary artery rings. Tissues were incubated with 17b-estradiol (d), 17b-estradiol and tamoxifen (m), 17b-estradiol and ICI ( ) or ethanol solvent (s) for 20 min before contracting with 30 nm U Rings were then relaxed with (a) cromakalim (n57), (b) levcromakalim (n58) or (c) or sodium nitroprusside (n59). centrations, the indirect enhancing effects were apparent at a concentration within the physiological range. Our data showing that 17b-estradiol, but not 17a-estradiol, can significantly produce direct relaxation (Fig. 1) is consistent with those previously reported for agonist- Fig. 8. Stereospecific effect of estradiol on sodium nitroprusside-induced 1 stimulated and high extracellular K -depolarized rat aorta relaxation of porcine coronary artery rings. Tissues were incubated with estradiol vehicle (m), 1 nm 17a-estradiol (s) or 1 nm 17b-estradiol (d) rings and rabbit coronary artery rings [20,22]. The excep- for 20 min before contracting with 30 nm U Rings were then tion to this stereospecificity is the study by Salas et al. [23] relaxed with sodium nitroprusside in a cumulative manner (n55). *P, that demonstrated both isomers were equally potent. Other 0.05 vs. control.

6 H. Teoh et al. / Cardiovascular Research 42 (1999) than possible differences in experimental conditions, the exposure to 1 nm 17b-estradiol enhanced A23187-meexplanation for the discrepancy in this study is not diated endothelium-dependent relaxation in porcine corapparent. onary artery ring segments. Similarly, we found that Estrogen receptors have been detected in human cor- endothelium-dependent relaxation produced by A23187 onary artery and umbilical vein endothelial cell cultures was not affected by short-term treatment with 1 nm 17b- [24] as well as smooth muscle cells of rat aortae [25,26], estradiol (Fig. 3). However, we observed a significant human coronary arteries [27], saphenous veins and mam- alteration in the concentration response curves and EC 50 mary arteries [28,29]. However, immunohistochemical values elicited by cromakalim, levcromakalim and sodium evidence suggest that the classical estrogen receptor is nitroprusside, all of which are endothelium-independent absent in human coronary arteries and that the steroid- relaxing agents. A common feature of our work and that of induced relaxation is independent of this [30]. Indeed, the Bell et al. [42] is the absence of enhancement of endorapid effects observed upon acute application of 17b- thelium-dependent relaxation following acute incubation estradiol are inconsistent with those mediated by the with a physiological concentration of 17b-estradiol. These intrinsically slow gene induction process in classic target findings contradict those obtained from clinical studies tissues of 17b-estradiol [12,17,18,31]. While 17b-estradiol- [17 19,43] and may in part be attributed to the in vitro evoked calcium ion mobilization in porcine coronary systems used by Bell s group and ourselves. A point arteries was insensitive to tamoxifen [12], another estrogen worthwhile noting is that acetylcholine was the only agent receptor antagonist ICI 182,780 decreased the proliferative used in the above mentioned in vivo studies to determine actions of the sex hormone in human endothelial cell the extent of endothelium-dependent vasorelaxation. As cultures [26]. Since estrogen receptors appear to be such, the positive modulatory activities of 17b-estradiol on heterogenous [32 34] and the smooth muscle relaxant endothelium-dependent relaxation reported previously may effect of 17b-estradiol are probably not due to non-specific be peculiar to acetylcholine. We were, however, unable to steroidal effects [22], these acute responses do not appear try and mimic the in vivo situation since acetylcholine to involve the classical estrogen receptor. causes contraction and not relaxation in porcine coronary Bradykinin is believed to cause relaxation via the release arteries. of both NO and endothelium-dependent hyperpolarizing High concentrations of 17b-estradiol are needed to factor [35] while A23187 stimulates NO synthesis by produce direct vasorelaxation actions (Fig. 1). In contrast, raising intracellular calcium concentrations. The absence short-term (20 min) incubation with 1 nm of the active of an estrogen effect on the endothelium-dependent relaxa- estradiol isomer significantly enhanced vasorelaxation to tions induced by both bradykinin and A23187 (Fig. 3) cromakalim, levcromakalim and sodium nitroprusside suggests that the augmentations elicited by 17b-estradiol (Figs. 4 8). This concentration of 17b-estradiol (1 nm) were not due to increased NO release or activity. This is in compares well with the physiological ranges in castrated accord with our observations with the potassium channel and ovariectomized pigs ( 0.2 nm) [44] as well as that in openers, cromakalim and levcromakalim [36]. Interesting- pregnant pigs ( 2 10 nm) [45]. The potentiation of ly, while pregnancy results in an increase in NO synthase relaxation appeared to be of a sharp concentration reactivity in the guinea-pig uterine artery [37], acetylcholine- sponse nature since 0.1 nm 17b-estradiol had no signifimediated relaxation in the same tissue has been found to cant effect on sodium nitroprusside relaxation while those remain unchanged [38]. Hence, it has been suggested that in the presence of 1 nm and 10 nm of 17b-estradiol were pregnancy may induce NO synthase activity in uterine enhanced to a similar level (Fig. 5). Interestingly, bradyvascular smooth muscle and as such not involve NO kinin- and A23187-induced relaxation were unaffected by activity within the endothelium [39] nm of 17b-estradiol (Fig. 3). We did not attempt to Previous studies found that A23187-stimulated relaxa- examine the effects of higher concentrations of 17b-estion in rabbit femoral and canine coronary artery is not tradiol since these would not only directly cause relaxaaltered by long-term treatment with 17b-estradiol [40,41]. tions but would also be physiologically irrelevant. As the In contrast, Bell and colleagues [42] have recently reported porcine hearts were from a local abattoir, we were unable that only porcine coronary artery rings exposed to 1 nm to determine or control the sex distribution of the hearts we 17b-estradiol for h exhibited enhanced response to obtained. While we recognize this as a limitation of our A Hence, to the best of our knowledge, this is the study, the data we present herein were reproducible in first report demonstrating that physiological concentrations every batch of hearts we studied. of 17b-estradiol can acutely (,30 min) modulate relaxa- In conclusion, the acute enhancing effects of such low tion induced by vasodilating agents in vitro. concentrations of 17b-estradiol (1 and 10 nm) in an in Administration of exogenous 17b-estradiol to achieve vitro preparation were in itself intriguing especially since physiological concentrations rapidly (,30 min) improved these concentrations of 17b-estradiol did not, on their own, acetylcholine-induced-endothelium-dependent relaxation in influence the contractile response. Unfortunately, we are female primates [19] and humans [17,18,43]. In contrast, unable at present, to provide a critical explanation as to Bell et al. [42] reported that overnight, but not acute, why these potentiating effects were only evident when

7 230 H. Teoh et al. / Cardiovascular Research 42 (1999) agents eliciting endothelium-independent relaxation were oxidation of low-density lipoproteins in postmenopausal women. Lancet 1994;343: used. Since both the direct and indirect effects of 17b- [9] Shwaery GT, Vita JA, Keaney JF. Antioxidant protection of LDL by estradiol had relatively rapid onsets and were insensitive to physiological concentrations of 17b-estradiol. Requirement for tamoxifen and ICI , it would appear that they might estradiol modification. Circulation 1997;95: be mediated by an estrogen receptor that acts via a nonendothelial [10] Keaney Jr JF, Shwaery GT, Xu A-M, et al. 17b-Estradiol preserves vasodilator function and limits low-density lipoprotein genomic mechanism. However, because the direct action oxidation in hypercholesterolemic swine. Circulation 1994;89:2251 of 17b-estradiol occurred at supraphysiological concen trations and the indirect effect was evident at physiological [11] Jiang C, Poole-Wilson PA, Sarrel PM, Mochizuki S, Collins P, levels, it remains to be determined if these two responses 21 MacLeod KT. Effect of 17b-oestradiol on contraction, Ca current 21 are mediated via the same receptor. The current knowledge and intracellular free Ca in guinea-pig isolated cardiac myocytes. Br J Pharmacol 1992;106: of possible vascular estrogen receptor subtypes, their [12] Han S-Z, Karaki H, Ouchi Y, Akishita M, Orimo H. 17b-Estradiol expression, cellular location(s) and post-receptor activities inhibits Ca influx and Ca release induced by thromboxane A2 in is still in the early phase and the existence of a fast-acting porcine coronary artery. Circulation 1995;91: surface membrane estrogen receptor is still speculative. [13] Sudhir K, Chou TM, Mullen WL, et al. Mechanisms of estrogen- induced vasodilation: in vivo studies in canine coronary conductance Nevertheless, the enhancing vasorelaxant action of 17band resistance arteries. J Am Coll Cardiol 1995;26: estradiol observed in our model may partly account for [14] McCrohon JA, Walters WAW, Robinson JTC, et al. Arterial reactivisome of the acute beneficial effects of 17b-estradiol on ty is enhanced in genetic males taking high dose estrogens. J Am blood flow. Coll Cardiol 1997;29: [15] New G, Timmins KL, Duffy SJ, et al. Long-term estrogen therapy improves vascular function in male to female transsexuals. J Am Coll Cardiol 1997;29: Acknowledgements [16] Herrington DM, Braden GA, Williams JK, Morgan TM. Estrogen modulates coronary vasomotor responses in postmenopausal women with early atherosclerosis. Am J Cardiol 1994;73: The authors would like to thank Godfrey S.K. Man for [17] Gilligan DM, Quyyumi AA, Cannon RO. Effects of physiological excellent technical assistance and Adrian Quan for aiding levels of estrogen on coronary vasomotor function in postmenopausal in the graphics. This study was supported by a Committee women. Circulation 1994;89: [18] Gilligan DM, Badar DM, Panza JA, Quyyumi AA, Cannon RO. on Research and Conference Grant, University of Hong Acute cardiovascular effects of estrogen in postmenopausal women. Kong. HT and SWSL are recipients of a post-doctoral Circulation 1994;90: fellowship award and a postgraduate studentship, respec- [19] Williams JK, Adams MR, Herrington DM, Clarkson TB. Short-term tively, from the University of Hong Kong. RYKM is a administration of estrogen and vascular responses of atherosclerotic member of the Institute of Cardiovascular Science and coronary arteries. J Am Coll Cardiol 1992;20: [20] Jiang C, Sarrel PM, Poole-Wilson PA, Collins P. Acute effect of Medicine, University of Hong Kong. 17b-oestradiol on rabbit coronary artery contractile responses to endothelin-1. Am J Physiol 1992;263:H271 H275. [21] Legato MJ. Coronary artery disease in women. Int J Fertil 1996;41: References [22] Freay AD, Curtis SW, Korach KS, Rubanyi GM. Mechanism of vascular smooth muscle relaxation by estrogen in depolarized rat [1] Barrett-Connor E. Sex differences in coronary heart disease. Why 21 and mouse aorta. Role of nuclear estrogen receptor and Ca uptake. are women so superior? The 1995 Ancel Keys Lecture. Circulation Circ Res 1997;81: ;95: [23] Salas E, Lopez MG, Villarroya M, et al. Endothelium-independent [2] Phillips GB, Pinkernell BH, Jing T-Y. Relationship between serum relaxation by 17a estradiol of pig coronary arteries. Eur J Pharmacol sex hormones and coronary artery disease in postmenopausal 1994;258: women. Arterioscler Thromb Vasc Biol 1997;17: [24] Kim-Schulze S, McGowan KA, Hubchak SC, et al. Expression of an [3] Van der Schouw YT, Van der Graaf Y, Steyerberg EW, Eijkemans estrogen receptor by human coronary artery and umbilical vein MJC, Banga JD. Age at menopause as a risk factor for cardiovascu- endothelial cells. Circulation 1996;94: lar mortality. Lancet 1996;347: [25] Espinosa E, Oemar BS, Luscher TF. 17b-Estradiol and smooth [4] Ettinger N, Friedman GD, Bush T, Quesenberry Jr. CP. Reduced muscle cell proliferation in aortic cells of male and female rats. mortality associated with long-term postmenopausal estrogen Biochem Biophys Res Comm 1996;221:8 14. therapy. Obstet Gynecol 1996;87:6 12. [26] Orimo A, Inoue S, Ikegami A, et al. Vascular smooth muscle cells as [5] Stampfer JM, Colditz FA, Willett WC, et al. Postmenopausal target for estrogen. Biochem Biophys Res Comm 1993;195:730 estrogen therapy and cardiovascular disease. New Engl J Med ;325: [27] Losordo DW, Kearney M, Kim EA, Jekanowski J, Isner JM. Variable [6] Clarkson TB, Anthony MS, Klein KP. Hormone replacement expression of estrogen receptor in normal and atherosclerotic therapy and coronary artery atherosclerosis: the monkey model. Br J coronary arteries of premenopausal women. Circulation Obstet Gynaecol 1996;103: ;89: [7] Dallongeville J, Marecaux N, Isorez D, Zylergberg G, Fruchart J-C, [28] Karas RH, Baur WE, van Eickles M, Mendelsohn ME. Human Amouyel P. Multiple coronary heart disease risk factors are associ- vascular smooth muscle cells express an estrogen receptor isoform. ated with menopause and influenced by substitutive hormonal FEBS Lett 1995;377: therapy in a cohort of French women. Atherosclerosis [29] Karas RH, Patterson BL, Mendelsohn ME. Human vascular smooth 1995;118: muscle cells contain functional estrogen receptor. Circulation [8] Sack MN, Rader DJ, Cannon RO. III. Oestrogen and inhibition of 1994;89:

8 H. Teoh et al. / Cardiovascular Research 42 (1999) [30] Collins P, Shepard M, Beale CM, Dowsett M. The classical estrogen receptor function in the guinea-pig uterine artery is not altered receptor is not found in human coronary arteries (Abstract). Circula- during pregnancy. Eur J Pharmacol 1994;258: tion 1995;92(Suppl 1):I 38. [39] Jovanovic A, Jovanovic S, Grbovic L. Endothelium-dependent [31] Cicenelli E, Ignarro LJ, Lograno M, Matteo G, Falco N, Schonauer relaxation in response to acetylcholine in pregnant guinea-pig LM. Acute effects of transdermal estradiol administration on plasma uterine artery. Hum Reprod 1997;12: levels of nitric oxide in postmenopausal women. Fertil Steril [40] Miller VM, Vanhoutte PM. Progesterone and modulation of endo- 1997;67: thelium-dependent relaxation in estrogen-treated canine coronary [32] Kuiper GGJM, Carlsson B, Grandien K, et al. Comparison of the arteries. Am J Physiol 1991;261:R1022 R1027. ligand binding specificity and transcript tissue distribution of [41] Gisclard V, Miller VM, Vanhoutte PM. Effect of 17b-estradiol on estrogen receptors a and b. Endocrinology 1997;138: endothelium-dependent responses in the rabbit. J Pharmacol Expl [33] Kuiper GGJM, Enmark E, Peltp-Huikko M, Nilsson S, Gustafsson Ther 1988;244: J-A. Cloning of a novel estrogen receptor expressed in rat prostate [42] Bell DR, Rensberger HJ, Koritnik DR, Koshy A. Estrogen pretreatand ovary. Proc Natl Acad Sci 1996;93: ment directly potentiates endothelium-dependent vasorelaxation of [34] Mosselman S, Polman J, Dijkema R. ERb: identification and porcine coronary arteries. Am J Physiol 1995;268:H377 H383. characterization of a novel human estrogen receptor. FEBS Lett [43] Collins P, Rosano GMC, Sarrel PM, et al. 17b-Estradiol attenuates 1996;392: acetylcholine-induced coronary arterial constriction in women but [35] Cowan CL, Cohen RA. Two mechanisms mediate relaxation by not men with coronary heart disease. Circulation 1995;92: bradykinin of pig coronary artery: NO-dependent and -independent [44] Amos MS, Beattie CW. Influence of gonadal steroid hormones upon responses. Am J Physiol 1991;260:H830 H835. the growth of melanoma in Sinclair swine. In: Swine in biomedical [36] Weston AD, Edwards G. Recent progress in potassium channel research. New York: Plenum, 1985: opener pharmacology. Biochem Pharmacol 1992;43: [45] Robertson HA, King GJ. Plasma concentration of progesterone, [37] Weiner CP, Lizasoain I, Baylis SA, et al. Induction of calcium- oestrone, oestradiol-17b and oestrone sulphate in the pig at imdependent nitric oxide synthases by sex hormones. Proc Natl Acad plantation, during pregnancy and at parturition. J Reprod Fertil Sci 1994;91: ;40: [38] Jovanovic A, Grbovic L, Drekic D, Novakovic S. Muscarinic