Effects of Estrogen on the Vascular System

Home , Idoxifene

EstrogenBrazilian andJournal the ofvascular Medical system and Biological Research (2003) 36: 1143-1158 1143 ISSN 0100-879X Review

Effects of estrogen on the vascular system

R.C. Tostes, D. Nigro, Grupo de Pesquisa sobre Hipertensão Arterial, Z.B. Fortes and Departamento de Farmacologia, Instituto de Ciências Biomédicas, M.H.C. Carvalho Universidade de São Paulo, São Paulo, SP, Brasil

Abstract

Correspondence The cardiovascular protective actions of estrogen are partially medi- Key words M.H.C. Carvalho ated by a direct effect on the vessel wall. Estrogen is active both on • Sex hormones Departamento de Farmacologia vascular smooth muscle and endothelial cells where functionally • Estrogen ICB, USP competent estrogen receptors have been identified. Estrogen adminis- • Vascular smooth muscle Av. Lineu Prestes, 1524 tration promotes vasodilation in humans and in experimental animals, • Endothelium, nitric oxide 05508-900 São Paulo, SP • Endothelium-derived Brasil in part by stimulating prostacyclin and nitric oxide synthesis, as well as by decreasing the production of vasoconstrictor agents such as cyclo- hyperpolarizing factor Fax: +55-11-3091-7322 • Angiotensin oxygenase-derived products, reactive oxygen species, angiotensin II, E-mail: [email protected] • Endothelin-1 and endothelin-1. In vitro, estrogen exerts a direct inhibitory effect on • Calcium channels Research supported by PRONEX (No. smooth muscle by activating potassium efflux and by inhibiting • Potassium channels 125/98), FAPESP (Nos. 97/14208-8, calcium influx. In addition, estrogen inhibits vascular smooth muscle 00/04611-4, and 00/12141-8), CAPES cell proliferation. In vivo, 17ß-estradiol prevents neointimal thicken- and CNPq (No. 351365/1997-4). ing after balloon injury and also ameliorates the lesions occurring in atherosclerotic conditions. As is the case for other steroids, the effect of estrogen on the vessel wall has a rapid non-genomic component Received March 14, 2003 involving membrane phenomena, such as alteration of membrane Accepted May 28, 2003 ionic permeability and activation of membrane-bound enzymes, as well as the classical genomic effect involving estrogen receptor activation and gene expression.

Introduction hanced fibrinolysis, and a series of actions on the vasculature, which will be the focus of Extensive epidemiological observations, our review. Since selective estrogen receptor clinical mechanistic studies, and basic labo- modulators (SERMs), such as the synthetic ratory research have suggested that the inci- estrogen-like compounds tamoxifen and dence of cardiovascular disease increases raloxifene, are being considered as an option with menopause and that hormone replace- in hormone replacement therapy due to their ment therapy, or more specifically, estrogen estrogen-agonist effects on some tissues replacement therapy is associated with ben- (liver, bone) and estrogen-antagonistic ef- eficial cardiovascular effects in postmeno- fects on other tissues (breast, uterus), we pausal women (1,2). Estrogen has a multi- included some of the already described vas- tude of biological effects that may account cular effects of SERMs in this review. for its apparent cardiovascular benefits (which remain to be proved in randomized Effects of estrogen clinical trials), including favorable effects on the lipid profile, antioxidant activity, en- The actions of steroid hormones can be

Braz J Med Biol Res 36(9) 2003 1144 R.C. Tostes et al.

divided into two types: those that are de- lasting effects of steroids (genomic effects) layed in onset and prolonged in duration are on cell function. However, the fact that not called “genomic” effects, and those that are all actions of estrogens are delayed in onset rapid in onset and short in duration are called and prolonged in duration implies another “non-genomic” effects. The early effects take mechanism of action (non-genomic effects). place within minutes (e.g., changes in vaso- Some of the estrogen effects are mediated by motor tone) and are mediated by rapid intra- ERs localized on the cell surface since mem- cellular signaling pathways, whereas the de- brane-impermeant forms of estrogen lead to layed effects (e.g., remodeling or lipid alter- the activation of mitogen-activated protein ations) require hours to days to occur and kinase (MAPK) as well as stimulation of require transcriptional events with subse- cGMP production and nitric oxide (NO) re- quent modulation of protein expression. Al- lease. The use of confocal microscopy and though the rapid and delayed effects of ster- flow cytometric analysis has demonstrated oids are clearly distinguishable from each that endothelial cells contain surface recep- other, there are actions that have onset time tors for estrogen, detectable by cell-imper- of minutes and it is not clear as to whether meant ligand binding and anti-ER antibodies. genomic or non-genomic mechanisms apply. Activation of membrane ERs has been linked to novel, rapid signaling pathways induced Estrogen receptors and mechanisms by estrogen. More recently, it was shown of action that several actions of steroid hormones on membranes involve non-nuclear signaling Radioactive estrogens (tritium-labeled pathways, such as coupling to G proteins estrogens) were used to demonstrate for the and generation of second messengers (4). first time intracellular steroid hormone re- Estrogens can induce Ca2+ mobilization, in- ceptors in endocrine target organs through crease levels of cyclic nucleotide second the body. Later, cells expressing the genom- messengers, and activate several kinases, ic steroid receptors were identified and including protein kinase C, phosphatidylin- mapped by binding studies, by immunocy- ositol-3-OH kinase (PI3-kinase), and MAPK tochemistry, and by in situ hybridization. through incompletely defined mechanisms The receptors for estrogen and for all classes (5,6). Adding more complexity, non-genom- of steroidal hormones (androgens, glucocor- ic effects of steroid hormones may be recep- ticoids, mineralocorticoids, progestins, vita- tor-mediated or receptor-independent. Ste- min D), as well as thyroid hormones and roid hormone genomic and non-genomic ef- retinoic acid, belong to the class of intracel- fects may occur simultaneously and may act lular receptors and have been classically at different levels, revealing the high com- defined as nuclear ligand-activated trans- plexity of steroid hormone regulation of cell cription factors. Activation of these recep- function. tors by the corresponding hormones affects The original studies indicated that there gene expression by acting on specific se- was one type of steroid receptors and that the quences in the target genes, known as estro- nature of the biological responses in differ- gen-response elements, and by modulating ent tissues was determined by the network of transcriptional events. Importantly, gene ex- genes available for interaction with the acti- pression may be further regulated through vated receptor. Later, two independent groups positive and negative interactions of estro- cloned the ER complementary DNA (7,8). gen receptors (ER) with transcription factors However, the distribution of this ER could (3). The discovery of intracellular hormone not explain all of the biological effects of receptors led to studies focusing on the long- estrogens, for instance in the ovaries. Later,

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1145 when a second ER was cloned, a useful and are able to reproduce, although they explanation for the dissociation between the show some reduction in litter size (12). ER localization and effects of estrogen was Measurements of mRNA for ER-α and suggested (9). Two types of intracellular ER-ß have shown distributions in the body ERs have been identified: ER-α and ER-ß, that differ quite markedly from each other. and isoforms of ER-α and ER-ß (splice vari- Moderate to high expression of ER-α has ants ER-ß1 and ER-ß2) have been also iden- been identified in the pituitary, kidney, epi- tified. didymis, and adrenal glands, whereas mod- The ligand-binding domain of ER-ß erate to high expression of ER-ß has been shows great homology compared with that found in prostate, lung, bladder, and brain. of ER-α. High homology is also observed in Overlapping high expression of ER-α and the DNA-binding domain or estrogen-re- ER-ß has been identified in ovary, testis and sponse element sites for the ER-α and ER-ß uterus. In the vasculature, ERs have been in the estrogen-sensitive genes. However, identified by different techniques in the en- the ER domains responsible for the interac- dothelium, intima and adventitia and on ad- tion of ER-α and ER-ß with transcription renergic nerve endings of arteries from factors, the ligand-independent N- and C- various territories and several species, in- terminal transcription activation functions-1 cluding humans. and -2, markedly differ between ER-α and A very interesting observation is that the ER-ß. This lack of homology may explain vascular cells are capable of expressing the different effects on an estrogen-response aromatase, the key enzyme in the estrogen element site observed with estrogen agonists synthesis pathway, suggesting that the vas- and antagonists. cular system is capable of local estrogen Furthermore, different ligands cause dif- biosynthesis in vivo, which may lead to acti- ferent conformational changes in ERs, vation of ER and downstream activation of thereby affecting interaction patterns with target genes (13). Locally produced estrogen additional transcription factor co-regulators may therefore act in an endocrine, paracrine and eventually the fine control of gene ex- and autocrine manner on vascular and non- pression. It should also be mentioned that vascular cells. steroid receptor co-regulators react differ- ently with ER-α and ER-ß. For instance, the Effects of estrogen on the arterial two ERs interact with an activator protein-1 wall site in opposite ways in the presence of different ligands (10). Finally, recent work Actions of steroidal hormones on the indicated that ER-ß may counteract or con- arterial wall include alteration or modula- trol ER-α activity in a variety of tissues, tion of ion fluxes and of receptors on smooth including the uterus, and mediate estrogenic muscle cells and modulation of endotheli- effects other than ER-α in some tissues, such um-derived factor production and activity. as the bone or the immune system. In this review we focused on the effects of The discovery and cloning of ER-ß pro- estrogen on components of endothelial and vided a basis for understanding how the smooth muscle cells. However, it is impor- knockout of ER-α could have resulted in a tant to keep in mind that estrogen-induced viable organism and in the continued action effects depend on the vascular bed and on of estrogens in some tissues, although ani- the animal species being considered, indi- mals are sterile and show altered sexual and cating that there is regional and species het- other behaviors (11). On the other hand, erogeneity in the modulatory influence of knockout mice for ER-ß appear quite normal estrogen on vasomotor function. To mention

Braz J Med Biol Res 36(9) 2003 1146 R.C. Tostes et al.

few examples, it has been reported that 17ß- the impaired ex vivo basal release of NO. estradiol induces both endothelium-depend- Effects of estradiol were also described in ent and -independent relaxation in the rat arteries from male animals. Huang et al. (17) aorta but only endothelium-independent re- observed that 17ß-estradiol restores endo- laxation in the rat mesenteric arteries. NO thelial NO release in response to shear stress contributes strongly to the endothelium- in pressurized gracilis muscle arterioles of dependent relaxation induced by 17ß-estra- male spontaneously hypertensive rats (SHR) diol in isolated aortas, whereas in small cere- by up-regulation of endothelial nitric oxide bral arteries both NO and cyclooxygenase synthase (NOS). Conversely, it has been (COX) metabolites contribute to estrogen- reported that endothelium-dependent relax- induced effects. Estrogen treatment increases ation elicited by carbachol and histamine aortic stiffness and potentiates endothelial was attenuated by estradiol in preparations vasodilator function in the hindquarters, but from intact male rats. Moreover, aortic pros- not in the carotid vascular bed. Differences tacyclin release was reduced by about 40% in the mechanisms involved in estrogen ac- after estradiol treatment in tissues from these tions may reflect a differential contribution animals. These results showing that release of mechanisms involved in vascular tone of NO in arteries from male rats is not af- regulation. Furthermore, there is evidence fected by estradiol treatment suggest gender that ER expression may change with patho- specificity for the vascular effects of estro- logical conditions or, inversely, that changes gen. in ER expression may lead to abnormal vas- NO production accounts for most of the cular function (14). endothelium-dependent relaxation activity, and there is extensive evidence showing es- Actions of estrogen on endothelial trogen-induced up-regulation of endothelial cells NO production. Probable mechanisms involved in estradiol-induced increased NO The endothelium plays a major role in production include: 1) transcriptional stimu- vascular tone control by releasing both re- lation of NOS gene expression, 2) inhibition laxing and contractile factors and estrogens of cytokine-induced down-regulation of NOS exert a number of effects on endothelial- gene expression, 3) post-translational modi- derived factors, as summarized below. fication of NOS protein, 4) increased co- Estrogens have been shown to enhance factor or L-arginine availability, 5) non-ge- endothelial-dependent relaxation in arterial nomic activation of second messengers (e.g., rings from different animals and from differ- Ca2+, cAMP) and tyrosine kinase, 6) translo- ent vascular beds, including coronary, mes- cation from the membrane to intracellular enteric, aorta and cerebral arteries. Studies sites, and 7) modulation of NO degrading on humans have demonstrated that estrogen systems (e.g., reactive oxygen radical gen- replacement treatment increases coronary eration and antioxidants). flow and decreases both coronary resistance Induction of constitutive (Ca2+-depend- and peripheral vascular tone. ent) NOS by estrogen has been demonstrated in a variety of tissues, consistent with the Nitric oxide presence of estrogen-response elements in the NOS promoter. In addition to increasing Earlier reports indicated that basal re- NOS production, estrogen induces rapid en- lease of NO is increased in females com- hancement of NOS activity and NO release pared to males (15,16) and that estrogen through nontranscriptional mechanisms and administration to ovariectomized rats restores by reducing its Ca2+ dependence (18). This

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1147 effect seems to be much more intense and NO availability. functionally relevant than the increase in NOS The effects of estrogen on NOS activity expression induced by estrogen and is inhib- are suggested to be important in arterial in- ited by the ER antagonist ICI 182,780, indi- jury. Local delivery of 17ß-estradiol during cating that the effect is mediated by ERs percutaneous transluminal coronary angio- (19). In SHR, estrogen deprivation (induced plasty improved endothelial function, en- by ovariectomy) decreases NOS activity and hanced re-endothelialization and endotheli- expression and NO-derived metabolites (20). al NOS expression and decreased neointima Recent studies indicate that estrogen-in- formation. Recently, Tolbert et al. (24) have duced activation of endothelial NOS is driven shown that the vasoprotective effects of es- by activation of the PI3-kinase/Akt pathway trogen after ligation vascular injury are par- resulting from direct interactions between tially reduced in inducible NOS knockout the ER and the regulatory subunit of PI3- mice, suggesting that estrogen also modu- kinase (6), and requires MAPK activation lates inducible NOS expression and plays a (19). Hisamoto et al. (21) observed that 17ß- role in neointima formation. estradiol, but not 17α-estradiol, caused acute activation of endothelial NOS both in human Endothelium-derived hyperpolarizing factor umbilical vein endothelial cells and in sim- ian virus 40-transformed rat lung vascular Endothelium-derived hyperpolarizing endothelial cells. Activation of endothelial factors (EDHFs) are important mediators of NOS involves the activation of Akt and the vascular relaxation, more specifically in re- phosphorylation of endothelial NOS, which sistance-sized arteries where they regulate is mediated by ER-α via a non-genomic tissue blood flow. The release of EDHFs is mechanism. modulated by a number of influences includ- The effects of estrogen on NOS may also ing agonist stimulation, shear stress, and be associated with its effects on caveolin-1 disease. The chemical identification and func- expression, which inhibits endothelial NOS tional characterization of EDHFs vary de- catalytic activity. Jayachandran et al. (22) pending on vascular size, vascular bed and have shown that endothelial NOS protein species. Three major candidates are the ep- expression and nitrite/nitrate production by oxyeicosatrienoic acids, cytochrome P450 bovine aortic endothelial cells are enhanced metabolites of arachidonic acid, K+ and hy- by 17ß-estradiol, which also stimulates drogen peroxide. Additionally, electrical caveolin-1 transcription and translation coupling through myoendothelial gap junc- through ER-mediated mechanisms. tions serves to conduct electrical changes Similar to estrogen, the SERM raloxifene from the endothelium to the smooth muscle stimulates endothelial NOS mRNA expres- and may mediate or propagate hyperpolar- sion (genomic effects) and also triggers rapid ization. activation of NO synthesis by stimulating Evidence showing that estrogen modu- endothelial NOS (non-genomic effects) via lates EDHF production and release has accu- the PI3-kinase pathway ER signaling (23). In mulated over the past few years. One of the femoral veins, raloxifene induces acute re- first lines of evidence for estrogen-induced laxation both by NO release and by direct hyperpolarization, although not specifically stimulation of vascular smooth muscle cells by EDHF, was provided by Harder and depending on the ovarian hormonal status of Coulson (25). They observed that addition the animal. As we will discuss later, estrogen of diethylstilbestrol to dog coronary arteries also prevents NO degradation due to its anti- hyperpolarized the membrane and reduced oxidant properties, consequently increasing input resistance, which was not related to

Braz J Med Biol Res 36(9) 2003 1148 R.C. Tostes et al.

increased Na+-K+ ATPase pump activity. vein endothelial cells, 17ß-estradiol increases

Later, it was shown that the hyperpolariz- the release of PGI2 via the induction of COX- ing response to acetylcholine in mesenteric 2, but not COX-1 protein (28), and in the arteries isolated from middle-aged rats is cutaneous vasculature of postmenopausal significantly greater in females than in males. women, acute 17ß-estradiol administration Ovariectomy causes a marked reduction in enhanced the response to acetylcholine after acetylcholine-induced hyperpolarization in aspirin, diclofenac, and placebo, but not af- female arteries and this is improved by 17ß- ter celecoxib treatment (29). estradiol replacement therapy (26). Further More recently, an association between evidence for a positive modulation of EDHF estradiol and vasoconstrictor prostaglandins release by estrogen is provided by the obser- has also been established: 1) estrogen re- vation that during pregnancy, a condition placement prevents prostaglandin H synthase with high levels of estrogen, there is an up- (PGHS)-dependent vasoconstriction, that regulation of EDHF-mediated vasodilation. was associated with augmented sensitivity

On the other hand, in pial arteries from fe- of the thromboxane A2 (TXA2)/prostaglan- male rats, ovariectomy increases the NO- din H2 (PGH2) receptor, in resistance-sized and prostacyclin (PGI2)-insensitive dilation mesenteric arteries from ovariectomized (EDHF-mediated dilation) whereas estradiol Sprague-Dawley rats, 2) estrogen improves replacement reduces it, suggesting a nega- vasodilation in isolated mesenteric arteries tive modulation of estrogen on EDHF release. from aged Fisher rats by decreasing PGHS- 2-dependent constriction and PGHS-2 ex- Arachidonic acid metabolites pression, and 3) estrogen decreases the synthesis of contractile COX metabolites. We The actions of estrogen on vascular cells have shown that ovariectomy increases reac- also influence the metabolism of prostaglan- tivity to norepinephrine and reduces sensi- dins and the activity of COX, a key enzyme tivity to acetylcholine in microvessels from in the production of prostaglandins. SHR whereas treatment with estradiol or It has been repeatedly shown that 17ß- estradiol + progesterone similarly restored estradiol stimulates production of vasodila- these altered responses. The COX inhibitors

tor prostaglandins such as PGI2 in a variety indomethacin, diclofenac, as well as ridogrel of preparations such as ovine uterine ar- (a TXA2 receptor antagonist and inhibitor of teries, ovine fetal pulmonary artery endothe- thromboxane synthase), but not dazoxiben lial cells, human umbilical vein endothelial (an inhibitor of thromboxane synthase), also cells, and rat mesenteric and cerebral blood restored norepinephrine and acetylcholine vessels. responses in ovariectomized SHR. The re- α Increased levels of PGI2 upon stimula- lease of PGF2 , but not of TXB2 and 6-keto- tion with estradiol have been associated with PGF1α, was greater in ovariectomized SHR increased expression of the key enzymes than in control SHR microvessels upon stimu-

involved in PGI2 production, phospholipase lation with norepinephrine, suggesting that A2, COX-1, and prostacyclin synthetase estrogen decreases the synthesis of contractile (PGIS), as demonstrated in rat cerebral blood COX metabolites, such as PGH2/PGF2α (30). vessels, where chronic in vivo 17ß-estradiol As mentioned earlier, not all arteries re-

treatment enhances basal PGI2 synthesis by spond in the same way to estradiol. It has increasing COX-1 and PGIS proteins (27). It been reported that estradiol and progeste- seems that the COX-2 pathway also plays a rone treatments differentially alter protein

specific role in estradiol-induced PGI2 syn- expression of the key enzymes involved in thesis and vasodilation. In human umbilical PGI2 production, phospholipase A2, COX-1,

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1149 and PGIS, in uterine and systemic (renal, However, subjects with high estrogen status coronary, mammary, omental) arteries. (premenopausal women) have significantly The modulatory actions of estrogen on lower monocyte chemoattractant protein the COX pathway seem to be mediated (MCP-1) levels than subjects with low estro- through activation of ERs. In pulmonary gen status (postmenopausal women), and artery endothelial cells, PGI2 production is hormone replacement therapy lowers plasma mediated by ER-ß, since it was fully blocked levels of MCP-1 (31). The estrogen-induced by both ER antagonism with ICI 182,780, increase in TNF-α activates endothelial ad- which is not selective for either ER isoform, hesion to leukocytes via increased expres- and the ER-ß-specific antagonist RR- sion of the endothelial adhesion molecules tetrahydrochrysene. On the other hand, the E-selectin, ICAM-1, and VCAM-1 (32). Ki- observation that estrogen increases COX-1 netic studies of mRNA for endothelial adhe- levels in cerebral blood vessels from wild- sion molecules show that, at early time points, type mice, but is ineffective in ER-α knock- estrogen treatment results in increased levels out mice, suggests that ER-α regulation of of mRNA for E-selectin, ICAM-1, and the endothelial COX-1 pathway also appears VCAM-1, whereas at later time points, estro- to contribute to the effects of estrogen. gen-treated cells disclose lower levels of mRNA for these cytokines (32). Adhesion molecules However, at higher or pharmacological doses and after longer periods of exposure, In an inflammatory process, cytokines estrogen decreases cytokine-induced adhe- and other inflammatory mediators elicit re- sion molecule expression by cultured endo- markable phenotypic changes in endothelial thelial cells (33). A prospective, random- cells. Activated endothelial cells release NO ized, placebo-controlled 12-week study on and PGI2, leading to vasodilation, which in healthy, normotensive postmenopausal women turn facilitates leukocyte influx to the sites who received either micronized estradiol of injury. Cytokine-activated endothelial cells alone, or sequentially estradiol combined express adhesion molecules for leukocytes - with a progestagen showed a significant de- E-selectin, intercellular adhesion molecule- crease in the plasma concentrations of soluble 1 (ICAM-1) and vascular cell adhesion mol- ICAM-1, soluble VCAM-1, and thrombo- ecule-1 (VCAM-1) - and are a source of modulin. In the same direction, it has been chemokines that not only have chemotactic shown that Fas ligand expression by the effects on leukocytes, but also are potent vascular endothelium, which inhibits the activators of leukocyte integrins, allowing migration of inflammatory cells into the ves- their interaction with their endothelial counter- sel wall, is improved by estradiol treatment, receptors. resulting in inhibition of leukocyte traffic Many cell types involved in immune and across the endothelium. According to the inflammatory responses are responsive to authors, the maintenance of endothelial Fas estrogen. These include T and B lympho- ligand expression by estradiol may represent cytes, and cells from the myelomonocytic a mechanism of the apparent antiatherogenic lineage such as monocytes, macrophages, effect of estrogen. and mast cells. Actions of estrogen on inflammatory markers seem to depend on lev- Oxidative stress - reactive oxygen species els of estrogen. At physiological concentrations, estrogen increases proinflammatory The vasculoprotective effects of estro- cytokine production - IL-1, tumor necrosis gen have been also partially attributed to a - factor-α (TNF-α), and IL-6 - by monocytes. shift in the NO/superoxide anion (O2 ) bal-

Braz J Med Biol Res 36(9) 2003 1150 R.C. Tostes et al.

ance in the vessel wall, thereby increasing bioavailability in SHR microvessels in vivo. the bioavailability of NO. In human umbili- We observed that oxidative stress in mesen- cal vein cultured endothelial cells, 17ß-es- teric arteries, evaluated by using intravital tradiol decreases expression of the NADPH microscopy and superfusion with hydroethi- oxidase subunit gp91phox and up-regulates dine, was significantly increased in ovariec- endothelial NOS expression, improving the tomized SHR and was attenuated by estra- - NO/O2 balance (34). On the other hand, diol or estradiol + progesterone treatment. estradiol increases endothelium-derived re- Treatment with the superoxide dismutase laxing factor activity in the thoracic aorta of mimetic MnTMPyP, but not with mannitol, ovariectomized female rats in the absence of that decomposes hydroxyl radicals, or L- changes in NOS activity or NOS gene and NAME, an NOS inhibitor, attenuated oxida- protein expression. Since lucigenin-enhanced tive stress in ovariectomized SHR. Treat- - chemiluminescence, an indicator of O2 pro- ment of mesenteries with diphenyleneiodo- duction, of the aorta was decreased in estra- nium, an NAD(P)H oxidase inhibitor, but diol-treated ovariectomized rats compared not with oxypurinol, a xanthine oxidase in- to the placebo group, it was suggested that hibitor, produced a significant reduction of the decreased endothelium-derived genera- oxyradical generation in microvessels from - tion of O2 in response to estrogens accounts ovariectomized SHR (38). for enhanced NO bioactivity and decreased Similar to estrogen, the SERM idoxifene peroxynitrite release (35). Increased oxida- effectively blunts Ang II-induced reactive tive stress and decreased NO production due oxygen species production as evaluated by to estrogen deficiency have been associated confocal laserscanning microscopy using the with higher blood pressure levels. Chronic redox sensitive marker 2',7'-dichlorofluores- estrogen replacement in ovariectomized rats cein and measurement of NAD(P)H oxidase prevents the decrease of plasma levels of activity. This estrogen effect has been ex- nitrites/nitrates as well as the enhancement plained in part by idoxifene-induced down- of blood pressure. regulation of AT1 receptor expression, as Wassmann et al. (36) observed that estro- will be discussed later. gen replacement therapy and angiotensin type The estrogen-mediated induction of the 1 (AT1) receptor antagonism prevent the protein thiol/disulfide oxidoreductases such worsened endothelial dysfunction, enhanced as disulfide isomerase, thioredoxin, thio- vasoconstrictor response to angiotensin II redoxin reductase, and glutaredoxin in vas- - (Ang II) and increased vascular O2 produc- cular endothelial cells may also be involved tion displayed by ovariectomized SHR. Very in the antioxidant properties of estrogen ob- recently, it has been shown that estrogen re- served in the vascular system. Recent studies duces Ang II-induced expression of NAD(P)H have also shown that estrogen inhibits endo- oxidase and peroxynitrite in cultured endothelial cell apoptosis induced by hydrogen thelial cells. Estrogen as well as superoxide peroxide and that phytoestrogens (daidzein, dismutase also inhibited Ang II-induced AT1 genistein and resveratrol) attenuate the oxi- receptor expression and nitrotyrosine stain- dative DNA damage induced by advanced ing, effects that were not inhibited by the ER glycation end-products in vascular smooth antagonist ICI 182,780 (37). muscle cells. The latter action was associ- We have shown that treatment with su- ated with an increase in intracellular total peroxide dismutase also restored changes in glutathione levels. norepinephrine and acetylcholine responses In humans it has been suggested that the in ovariectomized SHR (38). We have also endothelial dysfunction secondary to acute - found evidence that estrogen reduces O2 endogenous estrogen deprivation (surgical

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1151 ovariectomy) is caused by reduced NO avail- ovariectomized rats, whereas diclofenac and ability, which results from COX-dependent verapamil had no effects (38), suggesting production of oxidative stress (39). that, despite the contribution of Ang II to the oxidative stress in SHR, this is not the path- Angiotensin II way that is up-regulated in the absence of estrogen. Estrogen acts on the renin-angiotensin The action of estrogens on angiotensin- system at different points of the cascade: at converting enzyme (ACE) activity have also the formation of Ang II, at the level of Ang II been reported. Brosnihan et al. (42) showed receptors and on Ang II-induced responses. that ACE activity in the circulation and in Estrogen has been shown to increase gene tissues is reduced upon chronic hormone expression and plasma levels of angiotensin- replacement in two animal models of ogen. Estrogen also has effects on the levels postmenopause. Reduced tissue levels (kid- of renin in the circulation: 1) women under ney and aorta) of ACE, paralleled by re- hormone replacement therapy with estrogen duced ACE mRNA concentrations, has also display lower renin levels than those not been described with chronic estrogen re- using such therapy, 2) premenopausal women placement therapy. display lower renin levels than postmeno- These authors also investigated the car- pausal women, and 3) women display lower diovascular responses to administration of renin levels than men. Ang-(1-7) and Ang II in female transgenic Estrogen deficiency has been shown to (mRen2)27 rats receiving estrogen replace- increase the subtype 1 Ang II receptor (AT1 ment. The estrogen-treated rats showed lower receptor) mRNA levels, as well as the effi- blood pressure levels and the magnitude of cacy of Ang II on vasoconstriction (due to the depressor component of the biphasic increased AT1 receptor density), whereas response to Ang-(1-7) was significantly en- estrogen replacement therapy in ovariecto- hanced, whereas the pressor component was mized rats reversed AT1 receptor overex- attenuated. Estrogen replacement also sig- pression. Down-regulation of AT1 receptor nificantly attenuated the pressor response to mRNA and protein expression by estrogen Ang II. In addition, estrogen replacement occurs through activation of ERs and seems therapy was shown to reduce plasma and to be mediated by NO-dependent pathways tissue ACE activity in association with a (40). reduction in circulating levels of Ang II. Estrogen treatment also antagonizes the Because estrogen increased the levels of AT1 receptor-mediated growth-promoting plasma Ang-(1-7) and amplified the vasodi- effects of Ang II in vascular smooth muscle lator actions of Ang-(1-7), while reducing cells and the mechanisms involved in the the formation and vasoconstricting actions action of 17ß-estradiol include attenuation of Ang II, the authors suggested that estro- of AT1 receptor-mediated extracellular sig- gen shifts the vasoconstrictor-vasodilator bal- nal-regulated kinase activation and transac- ance of the renin-angiotensin system (42). tivation of MAPK phosphatase-1 expression The same group later showed that ovari- (41). ectomized monkeys treated with conjugated As mentioned earlier, estrogen and equine estrogen replacement for 30 months SERMs, such as idoxifene, inhibit Ang II- exhibited a reduction of ACE activity in induced reactive oxygen species release. Our association with a significant increase in group has observed that chronic treatment of plasma Ang I and hyperreninemia. Plasma female SHR with losartan caused similar Ang II levels were not increased in monkeys decreases in oxyradicals in both control and treated with estrogen, suggesting that the

Braz J Med Biol Res 36(9) 2003 1152 R.C. Tostes et al.

decrease in ACE curtailed the formation of male pigs, and gender differences in ET-1 the peptide. The Ang II/Ang I ratio, an in receptor density, as well as in the ratio of ET- vivo index of ACE activity, was significantly 1 receptor subtypes, have also been reported. reduced by estrogen treatment, further sup- An increase in the total number of ET-1 porting the biochemical significance of ACE receptors, but not in ET-1 binding Kd, has inhibition by estrogen. been reported in human saphenous vein from Clinical data have shown that postmeno- men compared to women. Furthermore, a

pausal women who received conjugated differential ratio of ETA to ETB receptors, equine estrogen for 3 months exhibited a favoring vasodilator effects in women, has significant reduction in plasma ACE activity also been described (46). that correlated with an increase in NO-medi- Our group showed that arteries from male, ated forearm endothelium-dependent vasodi- but not female, DOCA-salt rats display in- latation (reactive hyperemia) and increased creased sensitivity to ET-1 and to the selec-

serum levels of nitrite/nitrate. However, other tive ETB receptor agonist IRL-1620 both in investigators have observed that acute ad- vitro and in vivo (47,48). Changes in ETB- ministration of estradiol produces no changes mediated vascular responses were associ-

in vascular ACE activity or in the vasocon- ated with increased ET-1 and ETB receptor stricting responses to Ang I or Ang II. gene expression in male, but not in female, Sex-related differences in the effects of animals (49). Furthermore, we observed that blockers of the renin-angiotensin system have ovariectomy increases the vasoconstricting

been documented, with female animals dis- responses to the ETB agonist and that estro- playing a better response. Treatment with gen replacement therapy restores IRL-1620- losartan, an AT1 antagonist, normalizes high induced responses, supporting our suggestion blood pressure in 100% of female SHR that the ovarian hormones modulate ET-1/

against 53% of male SHR (43). Enalapril, an ETB receptor vascular responses/expression ACE inhibitor, produced similar results, nor- in DOCA-salt hypertension (50). Sexual di- malizing blood pressure in 71% of female morphism in vascular reactivity to ET-1 was SHR vs 46% of male SHR (44). also reported in SHR (51). Estradiol, as well as 2-hydroxyestradiol, Endothelin or 2-methoxyestradiol, metabolites of estradiol with little or no affinity for ERs, inhib- Similar to its actions on the renin-angio- ited basal and ET-1-stimulated synthesis in tensin system, the actions of estrogen on the porcine coronary artery endothelial cells. endothelin-1 (ET-1) pathway have been Estradiol-induced inhibition of ET-1 is pos- shown at different points of the cascade: at sibly mediated by inhibition of MAPK activ- its formation, at the level of the receptors ity, and by an ER-independent mechanism and on ET-1-induced responses. (52). On the other hand, the inhibitory ef- It has been reported that 17ß-estradiol fects of ICI 182,780 on the 17ß-estradiol- attenuates ET-1-induced coronary artery con- induced decrease of ET-1 gene expression striction both in vitro and in vivo. In addi- and peptide secretion in cultured bovine cation, increased expression of prepro-ET-1 rotid arterial endothelial cells suggest acti- mRNA has been observed in porcine aortic vation of ER-dependent mechanisms. endothelial cells in the absence of female Studies conducted on healthy postmeno- ovarian hormones. Barber et al. (45) reported pausal women who received continuous hor- that endogenous fluctuations in estrogen in- mone replacement therapy (17ß-estradiol fluence the affinity of ET-1 receptors in combined with norethisterone acetate or coronary arterial smooth muscle from fe- methoxyprogesterone) reported that post-

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1153 menopausal women treated with hormone have confirmed that estrogen stimulates K+ replacement therapy had increased plasma channel gating in coronary smooth muscle. nitrites/nitrates and decreased ET-1 levels Perforated-patch recordings from metaboli- (53). Women receiving the SERM raloxifene cally intact coronary myocytes revealed that had similar changes in plasma nitrites/ni- 17ß-estradiol more than doubles steady-state trates and ET-1 levels as well as in the ratio outward currents in these cells at positive of NO to ET-1 (53). voltages whereas studies of on-cell patches demonstrated a potent stimulatory effect of Actions of estrogen on vascular 17ß-estradiol on the gating of the large-con- smooth muscle cells ductance, Ca2+- and voltage-activated K+ (BKCa) channels. Furthermore, blocking We so far have seen that by acting on BKCa channels in intact arteries inhibited endothelial cells and modulating endotheli- estrogen-induced relaxation. Since the ef- um-derived relaxing and contraction factor fect of 17ß-estradiol on BKCa channels was release/activity, estrogen indirectly affects blocked by inhibiting cGMP-dependent pro- smooth muscle cell tone (Figure 1). Estrogen tein kinase activity and was mimicked by has also direct effects on smooth muscle cells. Here we will try to summarize a series AA of observations showing that 17ß-estradiol EDRFs NOS Nucleus COX may play a role in selectively regulating L-arg ERE ER vascular smooth muscle tone by modulating/ NO Gene transcription PGI2 EDHF modifying ion channel permeability in these EC cells (Figure 1). The voltage-clamp tech- ER K+ nique has helped elucidate the effects of PKC estrogen on ion channels. MAPK + Ca2+ K SMC Hyperpolarization Potassium channels VOC

Ca2+ The actions of estradiol on K+ channels EC have been described both in smooth muscle α PGF2 ET-1 - + Ang II O2 and endothelial cells. ATP-sensitive K TXA2 ACE COX ECE (KATP) channels seem to be involved in di- NAD(P)H EDCFs pET-1 ethylstilbestrol-induced relaxation in isolated Ang I AA rat aorta, since diethylstilbestrol-induced vasodilation is inhibited by glibenclamide Figure 1. Putative mechanisms for the action of estrogen on the vascular system. By acting on endothelial cells, estrogen increases EDRFs (NO, EDHF and PGI2) production/release and tetraethylammonium, but not by paxillin. - and decreases release/actions of EDCFs (PGF2α, TXA2, Ang II, ET-1, O2 ). By acting on Since relaxation in response to forskolin was vascular smooth muscle cells, estrogen activates K+ channels, leading to cell hyperpolariza- antagonized by glibenclamide, it was sug- tion, and blocks activation of Ca2+ channels, decreasing intracellular Ca2+ concentration. Actions of estrogen may be mediated by both direct genomic effects (activation of nuclear gested that diethylstilbestrol-induced relax- receptors and transcriptional events) and non-genomic effects (activation of rapid intracellu- ation in the rat aorta is related to the modula- lar signaling pathways). AA = arachidonic acid; ACE = angiotensin-converting enzyme; Ang I = angiotensin I; Ang II = angiotensin II; Ca2+ channels = calcium channels; COX = tion of KATP channels via cyclic AMP-dependent mechanisms (54). cyclooxygenase; EC = endothelial cells; ECE = endothelin converting enzyme; EDCF = endothelium-derived contracting factor; EDHF = endothelium-derived hyperpolarizing fac- Functional studies also demonstrated that tor; EDRF = endothelium-derived relaxing factor; ER = estrogen receptor; ERE = estrogen- 17ß-estradiol relaxes porcine coronary ar- response element; ET-1 = endothelin-1; K+ channels = potassium channels; L-arg = L- teries by an endothelium-independent mech- arginine; MAPK = mitogen-activated protein kinase; NO = nitric oxide; NOS = nitric oxide synthase; O - = superoxide anion; pET-1 = prepro-endothelin-1; PGF α = prostaglandin F + 2 2 2 anism involving K efflux, and subsequent alpha; PGI2 = prostacyclin; PKC = protein kinase C; SMC = smooth muscle cells; TXA2 = 2+ studies employing the patch-clamp technique thromboxane A2; VOC = voltage operated Ca channel.

Braz J Med Biol Res 36(9) 2003 1154 R.C. Tostes et al.

exogenous cGMP or by stimulating cGMP- striction of Ca2+ entry. Similarly, 17ß-estra- dependent protein kinase activity, the au- diol in isolated rabbit basilar artery induced thors proposed that 17ß-estradiol relaxes cor- relaxation due to inhibition of extracellular onary arteries by opening BKCa channels Ca2+ influx to vascular smooth muscle cells,

via cGMP-dependent phosphorylation (55). since estradiol inhibited CaCl2-induced con- A direct interaction of estradiol with a traction. Furthermore, basal maintained phen- voltage-gated channel subunit was described ylephrine- and KCl-induced [Ca2+]i and con- by Valverde and colleagues (56). They ob- traction of vascular smooth muscle triggered served that estradiol binds to the ß-subunit of by Ca2+ entry from the extracellular space maxi-K channels, which consist of a pore- exhibit differences depending on the pres- forming α-subunit and a regulatory ß-sub- ence or absence of female gonads. Cell con- unit that confers a higher Ca2+ sensitivity on traction and [Ca2+]i due to Ca2+ release from the channel. Binding of estradiol to the ß- the intracellular stores are not affected by subunit activates the maxi-K channels inde- gonadectomy. However, the observation that pendent of the generation of intracellular in the absence of extracellular Ca2+, 17ß- signals and can be triggered by estradiol estradiol still relaxed arteries pre-contracted conjugated to a membrane-impenetrable car- with norepinephrine, suggests that the hor- rier protein (56). mone inhibits intracellular Ca2+ release, an In contrast to the observations on smooth effect that is possibly due to direct interac- muscle cells, neither 17ß-estradiol nor es- tions with the cell membrane or with ion triol affected BKCa channel activity in hu- channel proteins. man umbilical vein endothelial cells. How- Using intact and permeabilized strips and ever, 2-methoxyestradiol, an endogenous me- isolated single cells of smooth muscle from tabolite of 17ß-estradiol, reversibly sup- femoral artery and portal vein, Kitazawa et pressed the amplitude of K+ outward cur- al. (57) observed that estradiol attenuated rents and produced a shift in the activation high KCl-induced force development and curve of BKCa channels to more positive myosin light chain phosphorylation, and pro- potentials. The 2-methoxyestradiol-induced duced rapid and reversible relaxation. Estra- inhibition of BKCa channels is primarily diol also rapidly inhibited voltage-depend- mediated by a decrease in the number of ent L-type Ca2+ channel currents in isolated long-lived openings. smooth muscle cells, suggesting that at pharmacological concentrations estrogen prima- Calcium channels rily reduces Ca2+ influx through inhibition of L-type Ca2+ channels and decreases myosin Functional evidence has suggested that light chain phosphorylation and contraction estrogen modulates Ca2+ entry or intracellu- of smooth muscle. 17ß-estradiol also de- lar Ca2+ release in vascular smooth muscle creased the [Ca2+]i response to ET-1 in cor- cells. In rat aorta and mesenteric artery, 17ß- onary artery smooth muscle cells isolated estradiol significantly reduced the maximum from gonad-intact, sexually mature female pigs, contraction to norepinephrine and KCl with- under conditions in which Ca2+ influx and out affecting potency. In experiments car- sarcoplasmic reticulum Ca2+ reuptake were ried out using Ca2+-free solution in which blocked. The effects of estradiol were blocked Ca2+ stores were depleted, 17ß-estradiol sig- by extracellular lanthanum and by the specific nificantly reduced the contraction to Ca2+ ER antagonist ICI 182,780, indicating that restoration in rat aorta, suggesting that 17ß- estradiol decreases [Ca2+]i in coronary ar- estradiol diminishes the maximum contrac- tery smooth muscle by affecting Ca2+ efflux tile response to norepinephrine due to re- via a receptor-mediated mechanism.

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1155

By using electrophysiological techniques, disease may be associated with an early in- Ogata and colleagues (58) demonstrated that crease of arterial and venous thrombotic 17ß-estradiol as well as the synthetic estro- events: 1) the heart and estrogen/progestin gens, ethynylestradiol and diethylstilbestrol, replacement study showed no differences in inhibited the barium inward current through the rate of primary coronary heart disease the voltage-dependent L-type Ca2+ channel events between active therapy and placebo in cultured rat thoracic aortic smooth muscle after an average of 4.1 years of therapy. The cell lines (A7r5). In the rabbit basilar artery, study also showed an increased risk (50%) estradiol also induces relaxation by inhibit- for nonfatal myocardial infarction and coro- ing voltage-dependent Ca2+ channels. How- nary heart disease death during the first year ever, in this preparation estradiol inhibits of follow-up among women on active thera- both nicardipine-sensitive and -resistant Ca2+ py; 2) in the Nurses’ Health Study, women currents via a pertussis toxin-sensitive GTP- with established heart disease, but not healthy binding protein. women, who had used hormone replacement therapy exhibited an increased risk of myo- Other actions of estrogen cardial infarction recurrence or coronary heart disease death; 3) the Puget Sound Group Other actions of estrogen have been de- Health Cooperative showed that healthy scribed such as activation of the carbon mon- women on hormone replacement therapy for oxide/heme oxygenase/cGMP pathway in short periods of time had double the risk of endothelial cells of human origin (umbilical myocardial infarction than women who had vein and uterine artery), up-regulation of α1- used hormone replacement therapy for longer adrenergic receptors in resistance-sized mes- periods (1-2 years); 4) the Women’s Health enteric arteries and also modulation of pro- Initiative trial also showed, two years ago, a liferation and growth of smooth muscle cells. trend toward early increased cardiovascular Looking at growth responses, it has been risk in healthy women, and very recently this shown that in cultured vascular smooth trial was interrupted due to a recommenda- muscle cells from Sprague-Dawley rats, tion from the Data and Safety Monitoring idoxifene, a SERM, inhibited platelet-de- Board. The reason was that the test statistic rived growth factor-induced DNA synthesis for invasive breast cancer exceeded the stop- and mitogenesis and protected endothelial ping boundary for this adverse effect and the cells from TNF-α-induced apoptosis in vi- global index statistics supported risks ex- tro. Idoxifene also modulates the balloon ceeding benefits among healthy postmeno- denudation-induced vascular injury response pausal women. The results indicated that the by significantly enhancing reendothelializa- hormone replacement regimen in the tion in injured carotid arteries. Women’s Health Initiative trial should not be initiated or continued for primary preven- Cardiovascular versus global actions tion of coronary heart disease (59). of estrogen However, the reasons for this lack of benefit are not yet clear and differ from the As mentioned earlier, extensive epide- beneficial effects predicted in observational miological observations have suggested that studies and in animal models of cardiovas- hormone replacement therapy is associated cular disease. Dose regimen, combination of with beneficial cardiovascular effects in post- estrogen with progestins versus estrogen menopausal women (1,2). However, there is alone, the administration route and duration also evidence that hormone replacement of treatment are some of the factors that may therapy for secondary prevention of heart be involved in the discrepancies. The pres-

Braz J Med Biol Res 36(9) 2003 1156 R.C. Tostes et al.

ence of gene polymorphisms in the setting of Conclusions estrogen therapy has also been implicated as one possible reason for the disappointing Estrogen is active both in vascular smooth results. Polymorphisms in genes related to muscle and endothelial cells and may exert thrombotic events, such as factor V Leiden, its cardiovascular protective actions by a prothrombin, factor VII, fibrinogen and direct effect on the vessel wall. Clinical and plasma activator inhibitor I, could increase animal studies have demonstrated the ben- the risk for cardiovascular disease and throm- eficial effects of estrogen on the vascular botic complications in a subset of women system. However, because estrogen affects and, therefore, would obscure the real ben- so many cellular processes, it is imperative efit of estrogen replacement therapy (60). If to gain a better understanding of the molec- this is the case, estrogen replacement thera- ular mechanisms, both genomic and non- py may be useful to prevent cardiovascular genomic, by which estrogen induces cellular disease in a large number of postmenopausal signals and modulates vascular responses. women, but not in a subset of women who Furthermore, the beneficial clinical effects are at high risk for cardiovascular and throm- of estrogen need to be confirmed in large and botic complications (60). multicenter randomized clinical trials.

References

1. Barret-Connor E & Bush TL (1991). Estrogen and coronary heart 12. Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, disease in women. Journal of the American Medical Association, Sar M, Korach KS, Gustafsson JA & Smithies O (1998). Generation 265: 1861-1867. and reproductive phenotypes of mice lacking estrogen receptor ß. 2. Kanel WB, Hjortland MC, McNamara PM & Gordon T (1976). Meno- Proceedings of the National Academy of Sciences, USA, 95: 15677- pause and risk of cardiovascular disease: the Framingham study. 15682. Annals of Internal Medicine, 85: 447-452. 13. Harada N, Sasano H, Murakami H, Ohkuma T, Nagura H & Takagi Y 3. Beato M & Sanchez-Pacheco A (1996). Interaction of steroid hor- (1999). Localized expression of aromatase in human vascular tis- mone receptors with the transcription initiation complex. Endocrine sues. Circulation Research, 84: 1285-1291. Reviews, 17: 587-609. 14. Zhu Y, Bian Z, Lu P et al. (2002). Abnormal vascular function and 4. Russell KS, Haynes MP, Sinha D, Clerisme E & Bender JR (2000). hypertension in mice deficient in estrogen receptor beta. Science, Human vascular endothelial cells contain membrane binding sites 295: 505-508. for estradiol, which mediate rapid intracellular signaling. Proceed- 15. Hayashi T, Fukuto JM, Ignarro LJ & Chaudhuri G (1992). Basal ings of the National Academy of Sciences, USA, 97: 5930-5935. release of nitric oxide from aortic rings is greater in female rabbits 5. Kelly MJ & Levin ER (2001). Rapid actions of plasma membrane than male rabbits: implications for atherosclerosis. Proceedings of estrogen receptors. Trends in Endocrinology and Metabolism, 12: the National Academy of Sciences, USA, 89: 11259-11263. 152-156. 16. Nigro D, Fortes ZB, Scivoletto R & Carvalho MHC (1990). Simulta- 6. Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW & Liao neous release of endothelium-derived relaxing and contracting fac- JK (2000). Interaction of oestrogen receptor with the regulatory tors induced by noradrenaline in normotensive rats. General Phar- subunit of phosphatidylinositol-OH kinase. Nature, 407: 538-541. macology, 21: 443-446. 7. Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P & 17. Huang A, Sun D, Koller A & Kaley G (2000). 17beta-estradiol re- Chambon P (1986). Human oestrogen receptor cDNA sequence, stores endothelial nitric oxide release to shear stress in arterioles of expression and homology to v-erb-A. Nature, 320: 134-139. male hypertensive rats. Circulation, 101: 94-100. 8. Greene GL, Gilna P, Waterfield M, Baker A, Hoter Y & Shine J 18. Caulin-Glaser T, Garcia-Cardena G, Sarrel P, Sessa WC & Bender JR (1986). Sequence and expression of human estrogen receptor (1997). 17ß-Estradiol regulation of human endothelial cell basal nitric complementary DNA. Science, 231: 1150-1154. oxide release, independent of cytosolic Ca2+ mobilization. Circula- 9. Kuiper GGJM, Enmark E, Pelto-Huikko M, Nilsson S & Gustafsson tion Research, 81: 885-892. JA (1996). Cloning of a novel estrogen receptor expressed in rat 19. Chen Z, Yuhanna IS, Galcheva-Gargova Z, Karas RH, Mendelsohn prostate and ovary. Proceedings of the National Academy of Sci- ME & Shaul PW (1999). Estrogen receptor alpha mediates the ences, USA, 93: 5925-5930. nongenomic activation of endothelial nitric oxide synthase by estro- 10. Paech K, Webb P, Kuiper GG, Nilsson S, Gustafsson J, Kushner PJ gen. Journal of Clinical Investigation, 103: 401-406. & Scanlan TS (1997). Differential ligand activation of estrogen re- 20. Costa SG, Anversa P, Scavone C, Sucupira M & Carvalho MHC ceptors ERα and ERß at AP1 sites. Science, 277: 1508-1510. (1998). Nitric oxide synthase activity in microvessels of SHR and 11. Korach KS (1994). Insights from the study of animals lacking func- normotensive rats: effects of estrogen. Journal of Hypertension, tional estrogen receptor. Science, 266: 1524-1527. 25: S80 (Abstract).

Braz J Med Biol Res 36(9) 2003 Estrogen and the vascular system 1157

21. Hisamoto K, Ohmichi M, Kurachi H et al. (2001). Estrogen induces Albory K, Rosen R, Bohm M & Nickenig G (2001). Endothelial the Akt-dependent activation of endothelial nitric-oxide synthase in dysfunction and oxidative stress during estrogen deficiency in spon- vascular endothelial cells. Journal of Biological Chemistry, 276: taneously hypertensive rats. Circulation, 103: 435-441. 3459-3467. 37. Gragasin FS, Xu Y, Arenas IA, Kainth N & Davidge ST (2003). 22. Jayachandran M, Hayashi T, Sumi D, Iguchi A & Miller VM (2001). Estrogen reduces angiotensin II-induced nitric oxide synthase and Temporal effects of 17beta-estradiol on caveolin-1 mRNA and pro- NAD(P)H oxidase expression in endothelial cells. Arteriosclerosis, tein in bovine aortic endothelial cells. American Journal of Physiolo- Thrombosis, and Vascular Biology, 23: 38-44. gy, 281: H1327-H1333. 38. Dantas AP, Tostes RC, Fortes ZB, Costa SG, Nigro D & Carvalho MH 23. Simoncini T, Genazzani AR & Liao JK (2002). Nongenomic mechan- (2002). In vivo evidence for antioxidant potential of estrogen in isms of endothelial nitric oxide synthase activation by the selective microvessels of female spontaneously hypertensive rats. Hyper- estrogen receptor modulator raloxifene. Circulation, 105: 1368-1373. tension, 39 (Part 2): 405-411. 24. Tolbert T, Thompson JA, Bouchard P & Oparil S (2001). Estrogen- 39. Virdis A, Ghiadoni L, Pinto S, Lombardo M, Petraglia F, Gennazzani induced vasoprotection is independent of inducible nitric oxide syn- A, Buralli S, Taddei S & Salvetti A (2000). Mechanisms responsible thase expression: evidence from the mouse carotid artery ligation for endothelial dysfunction associated with acute estrogen depriva- model. Circulation, 104: 2740-2745. tion in normotensive women. Circulation, 101: 2258-2263. 25. Harder DR & Coulson PB (1979). Estrogen receptors and effects of 40. Nickenig G, Strehlow K, Wassmann S, Baumer AT, Albory K, Sauer estrogen on membrane electrical properties of coronary vascular H & Bohm M (2000). Differential effects of estrogen and progeste- smooth muscle. Journal of Cell Physiology, 100: 375-382. rone on AT(1) receptor gene expression in vascular smooth muscle 26. Sakuma I, Liu MY, Sato A, Hayashi T, Iguchi A, Kitabatake A & cells. Circulation, 102: 1828-1833. Hattori Y (2002). Endothelium-dependent hyperpolarization and re- 41. Takeda-Matsubara Y, Nakagami H, Iwai M, Cui TX, Shiuchi T, laxation in mesenteric arteries of middle-aged rats: influence of Akishita M, Nahmias C, Ito M & Horiuchi M (2002). Estrogen acti- oestrogen. British Journal of Pharmacology, 135: 48-54. vates phosphatases and antagonizes growth-promoting effect of 27. Ospina JA, Krause DN & Duckles SP (2002). 17beta-estradiol in- angiotensin II. Hypertension, 39: 41-45. creases rat cerebrovascular prostacyclin synthesis by elevating 42. Brosnihan KB, Senanayake PS, Li P & Ferrario CM (1999). Bi-direc- cyclooxygenase-1 and prostacyclin synthase. Stroke, 33: 600-605. tional actions of estrogen on the renin-angiotensin system. Brazilian 28. Akarasereenont P, Techatraisak K, Thaworn A & Chotewuttakorn S Journal of Medical and Biological Research, 32: 373-381. (2000). The induction of cyclooxygenase-2 by 17beta-estradiol in 43. Nigro D, Fortes ZB, Scivoletto R, Barbeiro HV & Carvalho MH endothelial cells is mediated through protein kinase C. Inflamma- (1997). Sex-related differences in the response of spontaneously tion Research, 49: 460-465. hypertensive rats to angiotensin-converting enzyme inhibitor. Endo- 29. Calkin AC, Sudhir K, Honisett S, Williams MR, Dawood T & thelium, 5: 63-71. Komesaroff PA (2002). Rapid potentiation of endothelium-depend- 44. Silva-Antonialli MM, Fortes ZB, Carvalho MH, Scivoletto R & Nigro ent vasodilation by estradiol in postmenopausal women is medi- D (2000). Sexual dimorphism in the response of thoracic aorta from ated via cyclooxygenase 2. Journal of Clinical Endocrinology and SHRs to losartan. General Pharmacology, 34: 329-335. Metabolism, 87: 5072-5075. 45. Barber DA, Michener SR, Ziesmer SC & Miller VM (1996). Chronic 30. Dantas AP, Scivoletto R, Fortes ZB, Nigro D & Carvalho MH (1999). increases in blood flow upregulate endothelin-B receptors in arterial Influence of female sex hormones on endothelium-derived vaso- smooth muscle. American Journal of Physiology, 270: H65-H71. constrictor prostanoid generation in microvessels of spontaneously 46. Ergul A, Shoemaker K, Puett D & Tackett RL (1998). Gender differ- hypertensive rats. Hypertension, 34 (Part 2): 914-919. ences in the expression of endothelin receptors in human saphen- 31. Koh KK, Son JW, Ahn JY, Lee SK, Hwang HY, Kim DS, Jin DK, Ahn ous vein in vitro. Journal of Pharmacology and Experimental Thera- TH & Shin EK (2001). Effect of hormone replacement therapy on peutics, 285: 511-517. nitric oxide bioactivity and monocyte chemoattractant protein-1 47. Tostes RCA, David FL, Carvalho MHC, Nigro D, Scivoletto R & levels. International Journal of Cardiology, 81: 43-50. Fortes ZB (2000). Gender differences in vascular reactivity to endo- 32. Cid MC, Kleinman HK, Grant DS, Schnaper HW, Fauci AS & Hoffman thelin-1 in deoxycorticosterone acetate-salt hypertensive rats. Jour- GS (1994). Estradiol enhances leukocyte binding to tumor necrosis nal of Cardiovascular Pharmacology, 35 (Suppl 6): S99-S101. factor (TNF)-stimulated endothelial cells via an increase in TNF- 48. Tostes RCA, David FL, Fortes ZB, Nigro D, Scivoletto R & Carvalho induced adhesion molecules E-selectin, intercellular adhesion mole- MHC (2000). Deoxycorticosterone acetate-salt hypertensive rats cule type 1 and vascular cell adhesion molecule type 1. Journal of display gender-related differences in ETB receptor-mediated vascu- Clinical Investigation, 93: 17-25. lar responses. British Journal of Pharmacology, 130: 1092-1098. 33. Caulin-Glaser T, Watson CA, Pardi R & Bender JR (1996). Effects of 49. David FL, Carvalho MH, Cobra AL, Nigro D, Fortes ZB, Rebouças NA 17ß-estradiol on cytokine-induced endothelial cell adhesion mole- & Tostes RC (2001). Ovarian hormones modulate endothelin-1 vas- cule expression. Journal of Clinical Investigation, 98: 36-42. cular reactivity and mRNA expression in DOCA-salt hypertensive 34. Wagner AH, Schroeter MR & Hecker M (2001). 17beta-estradiol rats. Hypertension, 38 (Part 2): 692-696. inhibition of NADPH oxidase expression in human endothelial cells. 50. David FL, Montezano ACI, Rebouças NA, Nigro D, Fortes ZB, FASEB Journal, 15: 2121-2130. Carvalho MHC & Tostes RC (2002). Gender differences in vascular 35. Barbacanne MA, Rami J, Michel JB, Souchard JP, Philippe M, expression of endothelin and ET A/ET B receptors, but not in cal- Besombes JP, Bayard F & Arnal JF (1999). Estradiol increases rat cium handling mechanisms, in deoxycorticosterone acetate-salt hy- aorta endothelium-derived relaxing factor (EDRF) activity without pertension. Brazilian Journal of Medical and Biological Research, changes in endothelial NO synthase gene expression: possible role 35: 1061-1068. of decreased endothelium-derived superoxide anion production. 51. Fortes ZB, Nigro D, Scivoletto R & Carvalho MH (1991). Influence of Cardiovascular Research, 41: 672-681. sex on the reactivity to endothelin-1 and noradrenaline in spontane- 36. Wassmann S, Baumer AT, Strehlow K, van Eickels M, Grohe C, ously hypertensive rats. Clinical and Experimental Hypertension.

Braz J Med Biol Res 36(9) 2003 1158 R.C. Tostes et al.

Part A, Theory and Practice, 13: 807-816. Mann GE, Vergara C & Latorre R (1999). Acute activation of Maxi-K 52. Dubey RK, Jackson EK, Keller PJ, Imthurn B & Rosselli M (2001). channels (hSlo) by estradiol binding to the beta subunit. Science, Estradiol metabolites inhibit endothelin synthesis by an estrogen 285: 1929-1931. receptor-independent mechanism. Hypertension, 37 (Part 2): 640- 57. Kitazawa T, Hamada E, Kitazawa K & Gaznabi AK (1997). Non- 644. genomic mechanism of 17 beta-oestradiol-induced inhibition of con- 53. Saitta A, Altavilla D, Cucinotta D et al. (2001). Randomized, double- traction in mammalian vascular smooth muscle. Journal of Physiol- blind, placebo-controlled study on effects of raloxifene and hormone ogy, 499 (Part 2): 497-511. replacement therapy on plasma NO concentrations, endothelin-1 58. Ogata R, Inoue Y, Nakano H, Ito Y & Kitamura K (1996). Oestradiol- levels, and endothelium-dependent vasodilation in postmenopausal induced relaxation of rabbit basilar artery by inhibition of voltage- women. Arteriosclerosis, Thrombosis, and Vascular Biology, 21: dependent Ca channels through GTP-binding protein. British Jour- 1512-1519. nal of Pharmacology, 117: 351-359. 54. Martinez C, Sanchez M, Hidalgo A & Garcia de Boto MJ (2001). 59. Writing Group for the Women’s Health Initiative Investigators Involvement of K(ATP) channels in diethylstilbestrol-induced relax- (2002). Risks and benefits of estrogen plus progestin in healthy ation in rat aorta. European Journal of Pharmacology, 413: 109-116. postmenopausal women. Journal of the American Medical Associa- 55. White RE, Darkow DJ & Lang JL (1995). Estrogen relaxes coronary tion, 288: 321-333. arteries by opening BKCa channels through a cGMP-dependent 60. Herrington DM & Klein KP (2001). Pharmacogenetics of estrogen mechanism. Circulation Research, 77: 936-942. replacement therapy. Journal of Applied Physiology, 91: 2776-2784. 56. Valverde MA, Rojas P, Amigo J, Cosmelli D, Orio P, Bahamonde MI,

Braz J Med Biol Res 36(9) 2003