Abstract
This study compared ex vivo relaxing responses to the naturally occurring human hormone estetrol (E4) vs 17β-estradiol (E2) in eight different vascular beds. Arteries were mounted in a myograph, contracted with either phenylephrine or serotonin, and cumulative concentration-response curves (CRCs) to E4 and E2 (0.1–100 μmol/l) were constructed. In all arteries tested, E4 had lower potency than E2, although the differential effect was less in larger than smaller arteries. In uterine arteries, the nonselective estrogen receptor (ER) blocker ICI 182 780 (1 μmol/l) caused a significant rightward shift in the CRC to both E4 and E2, indicating that the relaxation responses were ER dependent. Pharmacological blockade of nitric oxide (NO) synthases by Nω-nitro-l-arginine methyl ester (l-NAME) blunted E2-mediated but not E4-mediated relaxing responses, while inhibition of prostaglandins and endothelium-dependent hyperpolarization did not alter relaxation to either E4 or E2 in uterine arteries. Combined blockade of NO release and action with l-NAME and the soluble guanylate cyclase (sGC) inhibitor ODQ resulted in greater inhibition of the relaxation response to E4 compared with E2 in uterine arteries. Endothelium denudation inhibited responses to both E4 and E2, while E4 and E2 concentration-dependently blocked smooth muscle cell Ca2+ entry in K+-depolarized and Ca2+-depleted uterine arteries. In conclusion, E4 relaxes precontracted rat arteries in an artery-specific fashion. In uterine arteries, E4-induced relaxations are partially mediated via an endothelium-dependent mechanism involving ERs, sGC, and inhibition of smooth muscle cell Ca2+ entry, but not NO synthases or endothelium-dependent hyperpolarization.
Introduction
The hormone estetrol (estra-1,3,5(10)-trien-3,15α-16α,17β-tetrol or E4) was discovered in 1965 (Hagen et al. 1965). E4 is produced in nature from 17β-estradiol (E2) and estrol (E3) only by the human fetal liver during pregnancy via 15α- and 16α-hydroxylase (Schwers et al. 1965, Gurpide et al. 1966). The concentration of E4 increases during pregnancy in both the fetus and the mother, but 24 h after delivery, the level is undetectable.
The chemical structure of E4 is closely related to that of E2, but the pharmacokinetic, metabolic, and endocrinological profile of E4 is substantially different from that of E2. E2 has high affinity for the estrogen receptor α (ERα) and ERβ (Ki values: 0.2 and 0.05 nmol/l respectively), while E4 has low-to-moderate affinity for ERα and ERβ (Ki values: 4.9 and 19 nmol/l respectively; Coelingh Bennink et al. 2008a). E2 has very low oral availability because over 99% of an oral dose is converted into estrone (E1) and E1 sulfate. Most of the circulating E2 is bound to either albumin or sex hormone binding globulin (SHBG), further decreasing plasma levels of free E2 by a factor of about 30. By contrast, E4 does not undergo phase I metabolism by human HepG2 cells and does not bind to SHBG (Hammond et al. 2008). As a result of these favorable metabolic and protein binding properties, E4 has excellent oral potency despite its low-to-moderate affinity for ERs.
E4 has been studied in many validated models and has been shown to behave as a full ER agonist, similar to E2, in most of these (Coelingh Bennink et al. 2008b, Heegaard et al. 2008, Holinka et al. 2008, Visser & Coelingh Bennink 2009). However, important differences between E4 and E2 were noted. E4 did not induce synthesis of SHBG in vitro as did E2 (Hammond et al. 2008). Recent studies showed that E4 behaved as an ER antagonist in in vitro and in vivo models of estrogen-dependent breast tumors in the presence of E2. E2 behaved as full agonist in these models (Coelingh Bennink et al. 2008c). E4 is now undergoing clinical testing as the estrogenic component of a combined oral contraceptive and for its effects on breast tumors.
The physiological function of E4 is not completely understood, although a role in regulating uterine blood flow has been suggested. Unilateral intrauterine injection of E4 in nonpregnant oophorectomized ewes has been shown to increase uterine blood flow, although with a 15- to 30-fold lower potency than E3 (Levine et al. 1984). Similarly, administration of E2 to ovariectomized ewes results in a rapid uterine vasodilation leading to a rise in uterine blood flow within 30–45 min (Killam et al. 1973). This rise in uterine blood flow is partially mediated via the release of nitric oxide (NO) and resultant increases in cGMP secretion, as shown by local infusion of the NO synthase blocker Nω-nitro-l-arginine methyl ester (l-NAME; Van Buren et al. 1992, Rosenfeld et al. 1996). The dilating effect of E2 on ovine uterine arteries is ER dependent, as shown by local infusion of the nonselective ER blocker ICI 182 780 in nonpregnant ewes (Magness et al. 2005).
To further profile E4, this study was designed to study the in vitro vasorelaxing effects of E4 and to compare them with those of E2. The objectives of this study were 1) to determine the ex vivo relaxing effects of E4 compared to E2 in rat uterine, thoracic aortic, carotid, mesenteric, pulmonary, renal, middle cerebral, and septal coronary arterial segments; 2) to assess the involvement of ERs in the vasorelaxing response to E4 in uterine arteries; and 3) to assess the endothelium-dependent and endothelium-independent vasorelaxing effects of E4 in uterine arteries. The uterine artery was chosen for assessment of the acute vasoactive properties of E4 and for comparison of these to the effects of E2 because the uterine arterial circulation is known to be highly sensitive to the vasodilator effects of E2. The uterine arteries are exposed to high physiological levels of E2 during the follicular and luteal phases of the menstrual cycle and during pregnancy, resulting in major increases in uterine blood flow (Magness 1998).
Materials and Methods
Animals
A total of 40 female nulliparous Sprague Dawley rats (age 12 weeks) were obtained from Charles River Breeding Laboratories, maintained at constant humidity (60±5%), temperature (24±1 °C), and light cycle (0600–1800 h), and fed a standard rat pellet diet (2016 Teklad Global 16% Protein Rodent Diet (Harlan Laboratories, Teklad Diets, Madison, WI, USA) ad libitum. All protocols were approved by the Institutional Animal Care and Use Committee at the University of Alabama at Birmingham and were consistent with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health.
Vessel preparation
Rats were killed by CO2 inhalation and the uterus, thoracic aorta, left common carotid artery, mesentery, lungs, left kidney, brains, and heart were removed and placed in cold Krebs–Ringer buffer (KRB) with the following composition (in mM): 118.5 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgSO4, 1.2 KH2PO4, 25.0 NaHCO3, and 5.5 glucose (all purchased from Sigma–Aldrich). Tissue samples were pinned down onto a 90 mm glass petri dish coated with black Sylgard and soaked in cold KRB. Main uterine arterial segments running along both uterine horns were cleared of adipose and connective tissues. The thoracic aorta and left common carotid artery were cleaned of connective tissue and cut into 2.5–3 mm segments. A fourth-order branch segment of the superior mesenteric artery was dissected from the mesentery; a small (200–300 μm in diameter) pulmonary arterial segment was dissected from the lung; a segment of the left main renal artery was dissected, and a segment of the middle cerebral artery was dissected from the brain. The atria and the left ventricle were removed to expose the interventricular septum of the heart, and a segment of the septal coronary artery was dissected. Aortic and carotid artery segments were mounted between two stainless steel pins, whereas the other segments (all 2 mm long) were mounted between two stainless steel jaws connected with two wires (40 μm in diameter) through the lumen of the segment in a myograph chamber (Danish Myo Technology, Inc., Aarhus, Denmark) filled with 5 ml KRB solution, maintained at 37 °C, and continuously aerated with 95% O2 and 5% CO2. Four arterial segments were analyzed in parallel. The remaining segments were temporarily stored at 4 °C.
Determination of optimal diameters
Thoracic aortae and carotid and pulmonary arteries were passively stretched according to a procedure first described by Mulvany & Halpern (1977). Briefly, the segments were distended stepwise in 100 μm increments measured with a built-in micrometer and the wall tension (N/m) was recorded using data acquisition (Powerlab 8/35, ADInstruments, Colorado Springs, CO, USA) and recording Software (ChartLab7, Colorado Springs, CO, USA). Thoracic aortic segments and carotid arteries were stretched at wall tension corresponding to a pressure of 90 mmHg, whereas pulmonary arteries were stretched at a wall tension corresponding to 40 mmHg. At this passive wall tension, segments were contracted with high K+ KRB (60 mmol/l KCl in KRB solution; replacing equimolar NaCl with KCl), thus generating active wall tension, which was set to a 100% contraction level.
All other arterial segments were progressively and actively stretched to the internal diameter at which the largest contractile response to 10 μmol/l norepinephrine (NE) or 60 mmol/l K+ KRB (middle cerebral arteries) was obtained. This internal diameter was referred to as the optimal diameter and the corresponding active wall tension was set to a 100% contraction level.
Arterial integrity was assessed by contracting arterial segments to either 1 μmol/l phenylephrine (PHE; for uterine, thoracic aorta, carotid, mesenteric, pulmonary, and renal arteries) or 0.1–1 μmol/l serotonin (5-HT; for middle cerebral and septal coronary arteries), followed by endothelium-dependent relaxation with 1 μmol/l acetylcholine (ACh). Arteries that relaxed immediately (>50% of relaxation) were considered to have a functional endothelium.
Experimental protocols
Cumulative concentration-response curves (CRCs) were constructed with PHE (0.16–20 μmol/l) or 5-HT (0.001–10 μmol/l). Arterial segments were then washed with KRB and after 10 min were contracted with a single concentration of the appropriate contractile agent, resulting in a near maximal contraction (80–100% of active wall tension obtained with 10 μmol/l NE or 60 mmol/l K+ depolarization). During a stable contraction, a CRC to E4 (0.1–100 μmol/l) was performed. After a 30-min washout period, segments were again contracted with the appropriate contractile agent and the CRC was repeated with E2 (0.1–100 μmol/l). The order of application of the estrogenic compounds was altered on every experimental day. In a subset of uterine arteries, CRCs to E4 and E2 were run in parallel. Relaxing responses to E2 and E4 were unaltered when run in series compared with parallel application.
We constructed CRCs to the selective ERα agonist propyl-[1H]-pyrazole-1,3,5-triy-triphenol (PPT; 0.1–30 μmol/l; Stauffer et al. 2000) and the selective ERβ agonist 2,3-bis(4-hydroxyphenol)-propionitrile (DPN; 0.1–30 μmol/l; Meyers et al. 2001) and compared their uterine arterial relaxing responses to E4 and E2. To study the involvement of ERs (ERα and ERβ) in the uterine arterial relaxing responses to E4 in comparison to E2, arteries were incubated for 30 min with the nonselective ER antagonist ICI 182 780 (7a,17b-[9-[(4,4,5,5,5-pentafluoropentyl)sulfinyl]nonyl] estra-1,3,5(10)-triene-3,17-diol; 1 μmol/l; Wakeling et al. 1991), after which the arterial segments were contracted with a single concentration of PHE (1–10 μmol/l) followed by a CRC to either E4 or E2.
To assess whether E4 triggered the release of vasodilator and/or contractile prostaglandins in uterine arteries, the cyclooxygenase inhibitor indomethacin (INDO; 10 μmol/l) was used. The involvement of endothelium-derived NO in uterine arterial relaxing responses to E4 was assessed using the nonselective NO synthase blocker l-NAME (100 μmol/l). The combined application of l-NAME and INDO was used to assess the role of endothelium-derived hyperpolarizing factor (EDHF) in E4-mediated relaxing responses in uterine arteries. The involvement of cGMP in E4-mediated relaxation was assessed by incubating uterine arteries with l-NAME, INDO, and the soluble guanylate cyclase (sGC) inhibitor 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ; 10 μmol/l).
The hyperpolarization response, involving both release of EDHF and spread of a hyperpolarizing current, is initiated in the endothelium via activation of small and intermediate calcium-activated K+ channels (SKCa and IKCa respectively; Burnham et al. 2002). Blockade of both KCa channels results in complete blockade of the EDHF response in rat mesenteric arteries (Crane et al. 2003). However, in some arteries, such as skeletal arterioles and coronary arteries, the large-conductance KCa (BKCa) is involved in EDHF-mediated responses (Feher et al. 2010). We did not attempt to study the role of the BKCa channel blocker iberiotoxin in E4- and E2-mediated relaxing responses in uterine arteries. 6,12,19,20,25,26-Hexahydro-5,27:13,18:21,24-trietheno-1 1,7-metheno-7H-dibenzo [b,n] [1,5,12,16]tetraazacyclotricosine-5,13-diium dibromide (UCL 1684; 1 μmol/l; Campos Rosa et al. 2000) was used to block SKCa channels and 1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole (TRAM-34; 1 μmol/l; Wulff et al. 2000) was used to block IKCa channels. All inhibitors were applied 30 min before the addition of PHE. After washing with KRB, arterial segments were again incubated with the appropriate pharmacological blocker(s) and subsequently (after 10 min) contracted with PHE (1–10 μmol/l) followed by a CRC to either E4 or E2.
The role of endothelium in E4-mediated relaxing responses was assessed using endothelium-denuded uterine arteries. The endothelium was mechanically removed by gently rubbing the lumen with a human hair (Osol et al. 1989). Successful denudation was achieved when relaxation to 1 μmol/l ACh was absent. Endothelium-independent relaxing responses to the NO donor sodium nitroprusside (SNP; 0.1–10 000 nmol/l) were performed in PHE (10 μmol/l)-contracted endothelium-intact arteries.
The potential inhibitory effect of E4 on voltage-operated smooth muscle cell Ca2+ entry was assessed by incubating endothelium-intact uterine arteries with three different concentrations (10, 30, and 100 μmol/l) of E4 or vehicle (ethanol) for 10 min in Ca2+–free and high K+ KRB solution, followed by a cumulative addition of CaCl2 (0.01–2.5 mmol/l). To rule out modulating effects of other endothelium-derived relaxing factors, arteries were incubated with l-NAME (100 μmol/l), INDO (10 μmol/l), and ODQ (10 μmol/l). Results were compared with three different concentrations (3, 10, and 30 μmol/l) of E2.
Drugs
E4 was provided by Pantarhei Bioscience B.V. (Zeist, The Netherlands) and dissolved in ethanol (stock solution of 10 mmol/l). ACh, endothelin-1, l-NAME, NE, PHE, serotonin, and SNP were purchased from Sigma–Aldrich and dissolved in distilled H2O. INDO and E2 (Sigma–Aldrich) were dissolved in ethanol (stock solutions of 10 mmol/l). TRAM-34 (Sigma), ODQ (EMD Chemicals, Gibbstown, NJ, USA), PPT, DPN, and UCL 1684 (all from Tocris Bioscience, Ellisville, MO, USA) were dissolved in DMSO.
Data and statistical analysis
Contractile responses were expressed as a percentage of the maximal contractile response to 10 μmol/l NE or 60 mmol/l K+ before the administration of any pharmacological inhibitor. Relaxing responses were expressed as a percentage of the maximal contractile response to PHE or 5-HT. Individual CRCs were fitted to a sigmoid regression curve (GraphPad Prism 5.0). As the sigmoidal curve of E4 could not be fully defined, a constant plateau value (set to 100%) was defined in order for GraphPad Prism to calculate a LOGEC50 (pEC50 value indicating sensitivity). Maximal relaxation to the highest concentration of E4 or E2 (Emax) and pEC50 values are shown as mean±s.e.m. Statistical significance of effects and differences were analyzed using either one-way ANOVA (comparison of pEC50 and Emax) or two-way ANOVA (comparison of CRCs). A Bonferroni post hoc test was used to compare multiple groups. A P value <0.05 was considered statistically significant.
Results
Arterial integrity
Optimal diameters of uterine arteries measured in the wire myograph averaged 317±3 μm and arteries developed an average active wall tension of 4.81±0.10 N/m in response to 10 μmol/l NE. Optimal diameters and active wall tensions for all artery types are summarized in Table 1. Sensitivity (pEC50) and maximal contraction to PHE (uterine, carotid, mesenteric, pulmonary, and renal artery) or 5-HT (middle cerebral and septal coronary artery) and sensitivity and maximal relaxations to 1 μmol/l ACh and the NO donor SNP for all arteries are shown in Table 2.
Optimal diameters and active wall tensions to NE (10 μmol/l) or depolarizing potassium (K+) solution (60 mmol/l KCl in Krebs–Ringer solution) for rat uterine, aortae, carotid, mesenteric, pulmonary, renal, middle cerebral, and septal coronary arteries. Uterine, mesenteric, pulmonary, and renal arteries were contracted with NE. Aortae, carotid, middle cerebral, and septal coronary arteries were contracted with depolarizing K+ solution. When more arteries of the same type were isolated from one rat, the values were averaged per rat. Values are shown as mean±s.e.m.
| Artery type | Uterine | Aorta | Carotid | Mesenteric | Pulmonary | Renal | Middle cerebral | Septal coronary |
|---|---|---|---|---|---|---|---|---|
| Optimal diameter (μm) | 317±3 | 1467±17 | 587±7 | 228±5 | 217±9 | 478±19 | 232±4 | 247±7 |
| Active wall tension (N/m) | 4.81±0.10 | 5.24±16 | 2.07±0.11 | 2.41±0.09 | 1.51±0.16 | 4.92±0.29 | 1.41±0.10 | 0.88±0.07 |
| n | 40 | 6 | 27 | 27 | 11 | 24 | 24 | 19 |
NE, norepinephrine. n denotes the total number of rats used.
Contractile characteristics to phenylephrine or serotonin, responses to a single concentration of ACh (1 μmol/l), and cumulative concentrations of SNP in contracted rat uterine, carotid, mesenteric, middle cerebral, pulmonary, renal, and septal coronary arteries. Sensitivity (pEC50) to contractile agents and to the NO donor SNP is calculated by GraphPad Prism Software as described in the Materials and Methods section and represents the negative logarithmic concentration of the contractile agent that induces a 50% tension level compared to Emax. Maximal contraction to contractile agent (Emax) is calculated at the percentage tension compared to the AWT corresponding to the artery's optimal diameter. The percentage contraction before ACh application is calculated as the percentage tension compared to the AWT corresponding to the artery's optimal diameter. Values are shown as mean±s.e.m.
| Artery type | Uterine | Carotid | Mesenteric | Pulmonary | Renal | Middle cerebral | Septal coronary |
|---|---|---|---|---|---|---|---|
| Contractile agent | Phenylephrine | Phenylephrine | Phenylephrine | Phenylephrine | Phenylephrine | Serotonin | Serotonin |
| n | 10 | 9 | 10 | 6 | 10 | 9 | 7 |
| pEC50 | 5.78±0.03 | 6.57±0.52 | 6.07±0.04 | 5.98±0.36 | 6.11±0.02 | 7.14±0.13 | 6.56±0.10 |
| Emax (% of AWT) | 100±2 | 99±12 | 105±2 | 56±13 | 109±2 | 81±5 | 121±16 |
| n | 21 | 19 | 17 | 6 | 18 | 18 | 15 |
| Percentage of contraction before ACh | 85±3 | 69±4 | 92±3 | 65±8 | 80±5 | 66±4 | 95±8 |
| Percentage of relaxation to ACh | 47±4 | 57±6 | 83±4 | 80±8 | 59±5 | 22±3 | 63±7 |
| n | 10 | 8 | 10 | 6 | 10 | 9 | 6 |
| SNP (pEC50) | 7.30±0.04 | 8.13±0.03* | 6.92±0.07 | 7.17±0.08 | 6.91±0.08* | 6.50±0.10* | 7.62±0.08* |
| Emax (%) | 79±4 | 99±1* | 83±4 | 83±11 | 65±4 | 77±5 | 99±1* |
ACh, acetylcholine; SNP, sodium nitroprusside; NO, nitric oxide; AWT, active wall tension; n, number of experiments. *P<0.05 vs uterine.
Vasorelaxing responses to E4
In uterine arteries, pEC50 for the control estrogenic compound E2 averaged 5.44±0.05 and reached a near maximal relaxation (Emax: 93±1%) in response to the highest concentration tested (100 μmol/l; Tables 3 and 4 and Fig. 1A). pEC50 and Emax for E4 were significantly lower (4.36±0.07 and 74±4% respectively; Tables 3 and 4 and Fig. 1A). Hence, in uterine arteries, E4 was a 14-fold less potent vasodilator compared with E2. Vasorelaxing properties of E4 compared with E2 for the other seven arterial types are shown in Fig. 1B through H and are summarized in Tables 3 and 4. E4 resulted in much lower variability than E2 in pEC50 and Emax values for all artery types tested (Tables 3 and 4). Larger arteries (aorta, carotid, and renal) had lower pEC50 and Emax than smaller arteries (mesenteric, pulmonary, middle cerebral, septal coronary, and uterine).
Vasorelaxing properties (sensitivity and maximal relaxing responses) of E2 compared with E4 in isolated and endothelium-intact rat uterine, aorta, carotid, and mesenteric (fourth-order) arteries. Arteries were contracted with the appropriate contractile agent before application of the estrogenic compound. The contraction before application of E4 or E2 is expressed as percentage of the contraction in response to 10 μmol/l NE. The sensitivity (pEC50) of E4 or E2 is calculated with GraphPad Prism Software as described in the Materials and Methods section and denotes the negative logarithmic concentration of E4 or E2 that induces 50% relaxation compared with the maximal relaxation (Emax). Potency of E2 over E4 is calculated as 10(pEC50(E2)−pEC50(E4)) from paired arteries only. Values are shown as mean±s.e.m.
| Uterine | Aorta | Carotid | Mesenteric | |||||
|---|---|---|---|---|---|---|---|---|
| Artery type | E2 | E4 | E2 | E4 | E2 | E4 | E2 | E4 |
| Control (%) | 100±4 | 95±3 | 95±11 | 96±10 | 101±7 | 97±7 | 89±3 | 92±2 |
| pEC50 | 5.44±0.05 | 4.36±0.07* | 4.61±0.08 | 4.30±0.15 | 4.91±0.05 | 4.39±0.15* | 5.78±0.13 | 4.62±0.08* |
| Emax (%) | 93±1 | 74±4* | 59±7 | 68±8 | 62±4 | 65±6 | 99±1 | 91±3 |
| Potency (fold) | 14±2 | 2±1 | 5±2 | 18±4 | ||||
| n | 16 | 16 | 6 | 6 | 10 | 10 | 8 | 8 |
E2, 17β-estradiol; E4, estetrol; NE, norepinephrine; n, number of experiments. *P<0.05 vs E2.
Vasorelaxing properties (sensitivity and maximal relaxing responses) of E2 compared to E4 in isolated and endothelium-intact rat pulmonary, renal, middle cerebral, and septal coronary arteries. Arteries were contracted with the appropriate contractile agent before application of the estrogenic compound. The contraction before application of E4 or E2 is expressed as percentage of the contraction in response to 10 μmol/l NE. The sensitivity (pEC50) of E4 or E2 is calculated with GraphPad Prism Software as described in the Materials and Methods section and denotes the negative logarithmic concentration of E4 or E2 that induces 50% relaxation compared to the maximal relaxation (Emax). Potency of E2 over E4 is calculated as 10(pEC50(E2)−pEC50(E4)) from paired arteries only. Values are shown as mean±s.e.m.
| Pulmonary | Renal | Middle cerebral | Septal coronary | |||||
|---|---|---|---|---|---|---|---|---|
| Artery type | E2 | E4 | E2 | E4 | E2 | E4 | E2 | E4 |
| Control (%) | 93±14 | 97±6 | 81±2 | 84±3 | 77±6 | 79±4 | 94±5 | 99±12 |
| pEC50 | 5.60±0.09 | 4.35±0.06* | 5.25±0.11 | 4.30±0.05* | 5.49±0.07 | 4.22±0.07* | 5.68±0.10 | 4.74±0.04* |
| Emax (%) | 99±1 | 80±8 | 81±2 | 66±4* | 97±1 | 71±4* | 100±0 | 98±2 |
| Potency (fold) | 20±6 | 11±3 | 22±5 | 10±2 | ||||
| n | 4 | 4 | 8 | 8 | 8 | 8 | 4 | 4 |
E2, 17β-estradiol; E4, estetrol; NE, norepinephrine; n, number of experiments. *P<0.05 vs E2.
Relaxing responses to 17β-estradiol (E2; 0.1–100 μmol/l; closed circles) and estetrol (E4; 0.1–100 μmol/l; open circles) in contracted uterine (A), aorta (B), carotid (C), fourth-order mesenteric (D), pulmonary (E), renal (F), middle cerebral (G), and septal coronary (H) arteries. Values are expressed as mean±s.e.m. *P<0.05 vs E2.
Citation: Journal of Endocrinology 215, 1; 10.1530/JOE-12-0009
Role of ERs in uterine and carotid arterial relaxing responses to E4
To study the contributions of ER subtypes α and β in relaxing responses in uterine arteries, selective agonists for ERα and ERβ were first tested. pEC50 values for the ERα agonist PPT were significantly greater than those for the ERβ agonist DPN (5.43±0.10 vs 4.77±0.25 respectively; Fig. 2A).
Relaxing responses to the ERα agonist PPT (0.1–30 μmol/l; circles) and the ERβ agonist DPN (0.1–30 μmol/l; squares) in PHE-contracted uterine arteries (A). Relaxing responses to 17β-estradiol (E2; 0.1–100 μmol/l; closed symbols) and estetrol (E4; 0.1–100 μmol/l; open symbols) in PHE-contracted uterine arteries in the presence of the nonselective estrogen receptor blocker ICI 182 780 (1 μmol/l) or the absence of ICI 182 780 (VEH; 10 μl DMSO; B). Values are expressed as mean±s.e.m. *P<0.05 vs PPT (A). *P<0.05 vs E4 VEH; **P<0.05 vs E2 VEH (B).
Citation: Journal of Endocrinology 215, 1; 10.1530/JOE-12-0009
The nonselective ER blocker ICI 182 780 (1 μmol/l) resulted in significant rightward shifts in the CRCs to E4 and E2 in uterine (Fig. 2B) arteries. For E4, pEC50 averaged 4.54±0.08 in vehicle and 4.05±0.11 in ICI 182 780-treated uterine arteries (Fig. 2B). For E2, pEC50 averaged 5.56±0.07 in vehicle and 5.12±0.05 in ICI 182 780-treated uterine arteries (Fig. 2B).
Role of endothelium-derived relaxing factors in uterine arterial relaxing responses to E4
The cyclooxygenase inhibitor INDO (10 μmol/l) did not significantly alter sensitivity to E4 or E2 (Fig. 3A). The NO synthase blocker l-NAME (100 μmol/l) caused a significant rightward shift in the CRC to E2, but not to E4 (Fig. 3B). The same trend as with l-NAME alone was observed when uterine arteries were incubated with both l-NAME and INDO (Fig. 3C). We next tested the role of cGMP in mediating the response to E4 and E2 by preventing NO release and action by blocking the sGC with ODQ (10 μmol/l) in combination with l-NAME and INDO. Interestingly, blockade of cGMP release blunted the response to E4 but not to E2 (Fig. 3D). The role of EDHF in mediating relaxing responses was pharmacologically tested by the addition of TRAM-34 (1 μmol/l) and UCL 1684 (1 μmol/l), inhibitors of SKCa and IKCa respectively. Inhibition of SKCa and IKCa channels in the combined presence of l-NAME and INDO did not alter the response to E4 or E2 (Fig. 3E).
Relaxing responses to 17β-estradiol (E2; 0.1–100 μmol/l; closed symbols) and estetrol (E4; 0.1–100 μmol/l; open symbols) in PHE-contracted uterine arteries in the absence of any pharmacological inhibitor (VEH), in the presence of the cyclooxygenase inhibitor INDO (10 μmol/l; A); the NO synthase blocker l-NAME (100 μmol/l; B); the combined presence of l-NAME and INDO (l-N+I; C), the combined presence of l-NAME, INDO, and the selective sGC blocker ODQ (10 μmol/l; l-N+I+O; D); the combined presence of l-NAME, INDO, and the selective IKCa channel blocker TRAM-34 (1 μmol/l) and the selective SKCa channel blocker UCL 1684 (1 μmol/l; E); and in the presence (+ENDO) and absence (−ENDO) of endothelium (F). Values are expressed as mean±s.e.m. *P<0.05 vs E4 VEH; **P<0.05 vs E2 VEH.
Citation: Journal of Endocrinology 215, 1; 10.1530/JOE-12-0009
The contribution of the endothelium to relaxing responses to E4 was assessed by mechanical removal of the endothelium. Successful denudation was confirmed by the absence of relaxation to 1 μM ACh in PHE-contracted arteries (0±1%). A significant rightward shift in the CRCs to E4 and E2 was observed in endothelium-denuded compared with endothelium-intact uterine arteries (Fig. 3F). Table 5 summarizes the effects of pharmacological inhibitors and endothelial denudation on the vasodilator properties of E4 compared with E2 in rat uterine arteries.
Influence of pharmacological inhibitors and endothelial denudation on the vasodilator properties of E2 compared with E4 in rat uterine arteries. The sensitivity (pEC50) of E4 or E2 is calculated with GraphPad Prism Software as described in the Materials and Methods section and denotes the negative logarithmic concentration of E4 or E2 that induces 50% relaxation compared to the maximal relaxation (Emax). See text for inhibitor concentrations used. When more arteries of the same type were isolated from one rat, the values were averaged and regarded as n=1 per rat. Values are shown as mean±s.e.m.
| E2 | E4 | |||||
|---|---|---|---|---|---|---|
| Treatment | pEC50 | Emax (%) | n | pEC50 | Emax (%) | n |
| Control (none) | 5.69±0.10 | 94±1 | 9 | 4.66±0.15† | 75±3† | 9 |
| ICI 182 780 | 5.12±0.05* | 83±2* | 5 | 4.05±0.11*,† | 49±4*,† | 5 |
| Indomethacin (INDO) | 5.75±0.07 | 96±1 | 7 | 4.40±0.06† | 79±3† | 7 |
| Nω-nitro-l-arginine methyl ester (l-NAME) | 5.32±0.08* | 93±1 | 8 | 4.39±0.06† | 76±4† | 8 |
| l-NAME+INDO | 5.26±0.09* | 95±2 | 7 | 4.43±0.10† | 78±5 | 7 |
| l-NAME+INDO+ODQ | 5.34±0.05* | 94±1 | 9 | 3.68±0.08*,† | 35±4*,† | 8 |
| l-NAME+INDO+TRAM-34+UCL 1684 | 5.53±0.11 | 88±4 | 6 | 4.33±0.08† | 70±2† | 6 |
| Endothelial denudation | 5.07±0.06* | 89±1 | 7 | 3.96±0.05*,† | 52±4*,† | 7 |
E2, 17β-estradiol; E4, estetrol; n, number of experiments. *P<0.05 vs control, †P<0.05 vs E2.
Inhibition of calcium entry by E4 in K+-depolarized uterine arteries
From Fig. 3F, it is clear that there is a significant endothelium-independent relaxation response to both E4 and E2. We therefore analyzed whether E4 could inhibit the entry of Ca2+ via voltage-operated Ca2+ channels on smooth muscle cells. Uterine arteries were first depleted of intracellular Ca2+ by washing them three times with 60 mmol/l Ca2+-free K+ KRB, followed by the addition of either vehicle (ethanol), or three different concentrations of E4 or E2. Cumulative addition of CaCl2 resulted in a contraction that was not significantly blocked by the lowest concentrations tested, namely 10 μmol/l E4 (Fig. 4A) and 3 μmol/l E2 (Fig. 4B). Next, we compared the inhibitory effects of higher concentrations of E4 (30 and 100 μmol/l) and E2 (10 and 30 μmol/l). A concentration-dependent inhibitory effect of both E4 and E2 was observed. Overall, the results in Fig. 4 show that E4 has tenfold lower potency than E2 in inhibiting Ca2+ entry in depolarized smooth muscle cells.
Contractile responses to cumulative addition of calcium chloride (CaCl2; 0.01–2.5 mmol/l) during 60 mmol/l K+ depolarization in uterine arteries that were incubated with l-NAME (100 μmol/l) and INDO (10 μmol/l). (A) Effects of 10 μmol/l estetrol (E4; open circles), 30 μmol/l E4 (open squares), and 100 μmol/l E4 (open diamonds) compared to vehicle (VEH; closed circles). (B) Effects of 3 μmol/l 17β-estradiol (E2; closed circles), 10 μmol/l E2 (closed squares), and 30 μmol/l E2 (closed diamonds) compared to vehicle (VEH; open circles). E4 or E2 was applied 10 min before the cumulative application of CaCl2. Values are expressed as mean±s.e.m. *P<0.05 vs VEH; **P<0.05.
Citation: Journal of Endocrinology 215, 1; 10.1530/JOE-12-0009
Discussion
This study assessed the ex vivo relaxing potency of the steroid hormone E4 in comparison to E2 in eight arterial beds, i.e. uterine artery, thoracic aorta, the left common carotid artery, the fourth-order branch of the superior mesenteric artery, pulmonary artery, left main renal artery, middle cerebral artery, and septal coronary artery of the rat. In all arteries tested, E4 had a weaker relaxing potency than E2. Pharmacological blockade experiments revealed that E4 caused relaxation of precontracted rat uterine arteries via both an endothelium-dependent (involving ER) and an ODQ-sensitive mechanism. Furthermore, E4 inhibited smooth muscle cell Ca2+ entry and contraction, albeit with tenfold lower potency than E2.
Estrogens exert their biological effects by binding to specific ERs, primarily ERα and ERβ. Previous in vitro binding studies showed that E4 has moderate affinity for human ERs, with four to five times higher affinity for the ERα compared with the ERβ (Visser et al. 2008). The same study reported that E4 has tenfold lower affinity for the ERα compared with the reference compound diethylstilbestrol (DES) and 100-fold lower affinity for the ERβ compared with DES (Visser et al. 2008). Kitazawa et al. (1997) showed that DES has threefold higher potency than E2 in relaxing rat femoral arteries. Based on these observations, it was postulated that E4 would have weaker vasorelaxant potency than E2. The current ex vivo arterial reactivity study clearly showed that this is indeed the case for a broad spectrum of rat artery types. E4 was a much weaker (∼14-fold) vasorelaxant than E2 in PHE-contracted endothelium-intact uterine arteries in the absence of any pharmacological inhibitor. Thus, substitution of two hydrogen atoms for two hydroxyl groups on the carbon 15 and 16 positions of the E2 molecule appears to result in lower receptor affinity and a reduced functional response.
In smaller resistance arteries, e.g. the middle cerebral artery and fourth-order mesenteric artery, E4 was an even weaker vasorelaxant compared with E2. Surprisingly, in larger elastic arteries, such as the thoracic aorta and common carotid artery, the difference in potency between E4 and E2 was smaller, mainly because of a reduction in sensitivity to E2 compared with its sensitivity in smaller arteries. This reduced sensitivity to E2 in larger arteries is in agreement with findings of others (Lindsey et al. 2011). Furthermore, the differential sensitivities to E2 of the mesenteric, renal, and uterine arteries observed in this study are similar to values reported by others (Naderali et al. 1999, Leung et al. 2005, Scott et al. 2007). Scott et al. observed a pEC50 value of 5.47±0.05 in endothelium-denuded uterine arteries, while Naderali et al. observed a pEC50 value of 6.04±0.06 in the superior mesenteric artery, and Leung et al. (2005) observed a pEC50 value of 5.54±0.22 in the renal artery. Hence, great inter-arterial variability exists in vasomotor responses to estrogenic hormones.
Previous studies have shown that in vivo administration of the nonselective ER antagonist ICI 182 780 into one uterine artery of ovariectomized and E2-treated nonpregnant sheep blunted the E2 effect on uterine blood flow, confirming its ER dependence (Magness et al. 2005). Here, we show that ICI 182 780 resulted in a significant rightward shift in the CRCs to E4 and E2, suggesting the contribution of ER. However, E2-mediated relaxing responses were not inhibited by ICI 182 780 in endothelium-denuded rat uterine arteries (Scott et al. 2007), most likely because the endothelial layer, which expresses ERs, had been mechanically removed. In vitro studies showed that ERα are localized in caveolae in isolated endothelial cells (Chambliss et al. 2000), supporting the conclusion that the lack of inhibitory effect of ICI 182 780 in denuded uterine arteries was due to the mechanical removal of the endothelial layer.
We compared relaxing responses to E4 and E2 with responses to selective agonists for ERα and ERβ in uterine arteries. We showed that the ERα agonist PPT has a more potent relaxing effect on uterine arteries compared with the ERβ agonist DPN. This finding is consistent with the study by Montgomory et al. (2003) in which the relative potency of PPT was greater than DPN in relaxing rat mesenteric arteries ex vivo. The order of potency of the estrogenic compounds and ER agonists in relaxing the rat uterine artery is E2=PPT>DPN>E4.
The current study demonstrated that the relaxing action of E4 in uterine arteries was l-NAME insensitive, suggesting that endothelium-derived NO plays a marginal role in E4-mediated relaxing responses in the rat uterine vasculature. By contrast, relaxing responses to E2 were partially l-NAME sensitive in rat uterine arteries, similar to earlier observations in the in vivo uterine circulation (Van Buren et al. 1992, Rosenfeld et al. 1996). Our data do not provide a clear mechanistic explanation for the l-NAME insensitivity of the uterine arterial relaxing responses to E4. Neither prostacyclin (PGI2) nor EDHF contributed to the relaxing response to either estrogenic hormone. Surprisingly, pharmacological inhibition of sGC resulted in marked blunting of E4-mediated relaxation in uterine arteries, an effort that was not observed in E2-treated vessels.
Both NO synthase and large-conductance Ca2+-activated K+ channels (BKCa) are involved in acute E2-induced vasodilation in oophorectomized nonpregnant ewes (Khan et al. 2000, Rosenfeld et al. 2000). NO activates cGMP-dependent protein kinase G, which subsequently opens BKCa on smooth muscle cells, thus decreasing Ca2+ influx via voltage-gated Ca2+ channels, leading to hyperpolarization and relaxation. A more detailed mechanistic study is needed to determine directly whether E4 can induce the release of cGMP from smooth muscle cells or whether l-NAME-insensitive intracellular NO stores play a role. Endothelial denudation blunted relaxations to E4 and E2 in the uterine artery, probably by disrupting the ER in caveolae on the plasma membrane of endothelial cells.
Both E4 and E2 elicited large endothelium-independent relaxation responses, suggesting that estrogenic compounds may interfere with excitation–contraction coupling, thus inhibiting contraction. Excitatory agonists, such as adrenergic agonists and contractile peptides like angiotensin II, can stimulate smooth muscle contraction through three distinct signaling pathways: Ca2+ influx through membrane Ca2+ channels, IP3-induced Ca2+ release from the sarcoplasmatic reticulum, and Ca2+ sensitizing mechanisms of the contractile machinery (Somlyo & Somlyo 1994, Hilgers & Webb 2005). The contraction-modulating effects of estrogens have been attributed to their Ca2+ antagonistic properties. E2 can inhibit voltage-dependent calcium inward currents (L-type voltage-operated Ca2+ channels or L-VOCC) located on smooth muscle cells, but not on endothelial cells (Shan et al. 1994, Zhang et al. 1994, Nakajima et al. 1995). This leads to reduced intracellular Ca2+ concentration and subsequent lower Ca2+-calmodulin-dependent myosin light-chain phosphorylation and contraction. By contrast, E2 effects have not been ascribed to modulating Ca2+ release from intracellular stores or by IP3 (Kitazawa et al. 1997). High K+ concentrations (>25 mmol/l) can depolarize the smooth muscle cell membrane and stimulate L-VOCC and Ca2+ influx independent of the endothelium. In Ca2+-free conditions, no contraction can occur during high K+ (60 mmol/l) depolarization. Titration of CaCl2 into the organ bath results in a Ca2+-dependent contraction. Using this approach in endothelium-intact uterine arteries, in conditions where NO release and action were blocked, we tested the Ca2+ antagonistic effects of E4 and found that E4 and E2 concentration-dependently blocked L-VOCC-dependent contractions in uterine arteries. E4 was roughly tenfold less potent than E2 in inhibiting these contractions. Interestingly, 3 μmol/l E2 did not block L-VOCC-dependent contractions (Fig. 4A), but resulted in ∼50% relaxation of PHE-contracted uterine arteries (Fig. 1A). The fact that these Ca2+ antagonistic effects of E4 and E2 are seen at higher concentrations might suggest that these Ca2+ antagonistic effects are unrelated to their endothelium and ER-dependent estrogenic effects.
In conclusion, we have demonstrated that E4 is capable of inducing a relaxing response in arteries from a variety of vascular beds. E4 was several fold less potent than E2 in relaxing these arteries, but the differential relaxing potency of E4 vs E2 decreased in larger arteries, mainly because of reduced potency of E2. E4 caused relaxation of contracted uterine arteries via an endothelium-dependent mechanism involving ERs, as well as smooth muscle-dependent inhibition of Ca2+ entry. The mechanism of the discrepancy between the micromolar range estrogenic effects on ex vivo vasorelaxation and the nanomolar range of circulating estrogen levels in vivo is unclear, but the observation does suggest that the effects of estrogens on ex vivo vasorelaxation are pharmacological in nature. The relatively low vasorelaxing potency of E4 might be beneficial in clinical use in that E4 will provide beneficial effects associated with estrogens without altering hemodynamics. Furthermore, previous studies have shown that E4 (compared with E2) has a low first-pass liver metabolism, no SHBG binding, ER antagonistic effects on breast cancer models, and anti-thrombotic properties, suggesting that E4 has promise for clinical use as a component hormone in a combined oral contraceptive and as therapy for breast tumors.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This study was sponsored by Pantarhei Bioscience B.V. (Zeist, The Netherlands). Partial support came from RO1 grant HL087980.
References
Burnham MP, Bychkov R, Feletou M, Richards GR, Vanhoutte PM, Weston AH & Edwards G 2002 Characterization of an apamin-sensitive small-conductance Ca2+-activated K+ channel in porcine coronary artery endothelium: relevance to EDHF. British Journal of Pharmacology 135 1133–1143. (doi:10.1038/sj.bjp.0704551)
Campos Rosa J, Galanakis D, Piergentili A, Bhandari K, Ganellin CR, Dunn PM & Jenkinson DH 2000 Synthesis, molecular modeling, and pharmacological testing of bis-quinolinium cyclophanes: potent, non-peptidic blockers of the apamin-sensitive Ca2+-activated K+ channel. Journal of Medicinal Chemistry 43 420–431. (doi:10.1021/jm9902537)
Chambliss KL, Yuhanna IS, Mineo C, Liu P, German Z, Sherman TS, Mendelsohn ME, Anderson RGW & Shaul PW 2000 Estrogen receptor α and endothelial nitric oxide synthase are organized into a functional signaling module in caveolae. Circulation Research 87 e44–e52. (doi:10.1161/01.RES.87.11.e44)
Coelingh Bennink HJT, Holinka C & Diczfalusy E 2008a Estetrol review: profile and potential clinical applications. Climacteric 11 (Suppl 1) 47–58. (doi:10.1080/13697130802073425)
Coelingh Bennink HJT, Skouby S, Bouchard P & Holinka CF 2008b Ovulation inhibition by estetrol in an in vivo model. Contraception 77 186–190. (doi:10.1016/j.contraception.2007.11.014)
Coelingh Bennink HJT, Singer C, Simoncini T, Genazzani AR & Holinka CF 2008c Estetrol, a pregnancy-specific human steroid, prevents and suppresses mammary tumor growth in a rat model. Climacteric 11 (Suppl 1) 29 doi:10.1080/13697130802040325)
Crane GJ, Gallagher N, Dora KA & Garland CJ 2003 Small- and intermediate-conductance calcium-activated K+ channels provide different facets of endothelium-dependent hyperpolarization in rat mesenteric arteries. Journal of Physiology 553 183–189. (doi:10.1113/jphysiol.2003.051896)
Feher A, Ruthai I, Beleznai T, Ungvari Z, Csiszar A, Edes I & Bagi Z 2010 Caveolin-1 limits the contribution of BK(Ca) channel to EDHF-mediated arteriolar dilation: implications in diet-induced obesity. Cardiovascular Research 87 732–739. (doi:10.1093/cvr/cvq088)
Gurpide E, Schwers J, Welch MT, Vande Wiele RL & Lieberman S 1966 Fetal and maternal metabolism of estradiol during pregnancy. Journal of Clinical Endocrinology and Metabolism 26 1355–1365. (doi:10.1210/jcem-26-12-1355)
Hagen AA, Barr M & Diczfalusy E 1965 Metabolism of 17β-oestradiol-4-14C in early infancy. Acta Endocrinologica 49 207–220.
Hammond GL, Hogeveen KN, Visser M & Coelingh Bennink HJT 2008 Estetrol does not bind sex hormone binding globulin or increase its production by human HepG2 cells. Climacteric 11 (Suppl 1) 41–46. (doi:10.1080/13697130701851814)
Heegaard AM, Holinka CF, Kenemans P & Coelingh Bennink HJT 2008 Estrogenic uterovaginal effects of oral estetrol in modified Allen–Doisy test. Climacteric 11 (Suppl 1) 22–28. (doi:10.1080/13697130701842490)
Hilgers RH & Webb RC 2005 Molecular aspects of arterial smooth muscle contraction: focus on Rho. Experimental Biology and Medicine 230 829–835.
Holinka CF, Brincat M & Coelingh Bennink HJT 2008 Preventive effect of oral estetrol in a menopausal hot flush model. Climacteric 11 (Suppl 1) 15–21. (doi:10.1080/13697130701822807)
Khan LH, Rosenfeld CR, Liu XT & Magness RR 2000 Regulation of the cGMP–cPKG pathway and large-conductance Ca2+-activated K+ channels in uterine arteries during the ovine ovarian cycle. American Journal of Physiology. Endocrinology and Metabolism 298 E222–E228. (doi:10.1152/ajpendo.00375.2009)
Killam AP, Rosenfeld CR, Battaglia FC, Makowski EL & Meschia G 1973 Effect of estrogens on the uterine blood flow of oophorectomized ewes. American Journal of Obstetrics and Gynecology 115 1045–1052.
Kitazawa T, Hamada E, Kitazawa K & Gaznabi AKM 1997 Non-genomic mechanism of 17β-oestradiol-induced inhibition of contraction in mammalian vascular smooth muscle. Journal of Physiology 499 497–511.
Leung FP, Yao X, Lan CW, Ko WW, Lu L & Huang Y 2005 Raloxifene relaxes rat intrarenal arteries by inhibiting Ca2+ influx. American Journal of Physiology. Renal Physiology 289 F137–F144. (doi:10.1152/ajprenal.00353.2004)
Levine MG, Miodovnik M & Clark KE 1984 Uterine vascular effects of estetrol in nonpregnant ewes. American Journal of Obstetrics and Gynecology 148 735–738.
Lindsey SH, Carver KA, Prossnitz ER & Chappell MC 2011 Vasodilation in response to the GPR30 agonist G-1 is not different from estradiol in the mRen2.Lewis female rat. Journal of Cardiovascular Pharmacology 57 598–603. (doi:10.1097/FJC.0b013e3182135f1c)
Magness RR 1998 Maternal cardiovascular and other physiologic responses to the endocrinology of pregnancy. In The Endocrinology of Pregnancy. Ed B Fuller. pp 507–538. Totowa, NJ: Humana Press.
Magness RR, Phernetton TM, Gibson TC & Chen CB 2005 Uterine blood flow responses to ICI 182 780 in ovarioectomized oestradiol-17β-treated, intact follicular and pregnant sheep. Journal of Physiology 565 71–83. (doi:10.1113/jphysiol.2005.086439)
Meyers MJ, Sun J, Carlson KE, Marriner GA, Katzenellenbogen JA & Katzenellenbogen BS 2001 Estrogen receptor-β potency-selective ligands: structure-activity relationship studies of diarylpropionitriles and their acetylene and polar analogues. Journal of Medicinal Chemistry 44 4230–4251. (doi:10.1021/jm010254a)
Montgomory S, Shaw L, Pantelides N, Taggart M & Austin C 2003 Acute effects of oestrogen receptor subtype-specific agonists on vascular reactivity. British Journal of Pharmacology 139 1249–1253. (doi:10.1038/sj.bjp.0705368)
Mulvany W & Halpern MJ 1977 Tension responses to small length changes of vascular smooth muscle cells. Journal of Physiology 265 21P–23P.
Naderali EK, Walker AB, Doyle P & Williams G 1999 Comparable vasorelaxant effects of 17α- and 17β oestradiol on rat mesenteric resistance arteries: an action independent of the oestrogen receptor. Clinical Science 97 649–655. (doi:10.1042/CS19990162)
Nakajima T, Kitazawa T, Hamada E, Hazama H, Omata M & Kurachi Y 1995 17-β-Estradiol inhibits the voltage-dependent L-type Ca2+ currents in aortic smooth muscle cells. European Journal of Pharmacology 294 625–635. (doi:10.1016/0014-2999(95)00602-8)
Osol G, Cipolla M & Knutson S 1989 A new method for mechanically denuding the endothelium of small (50–150 microns) arteries with a human hair. Blood Vessels 26 320–324.
Rosenfeld CR, Cox BE, Roy T & Magness RR 1996 Nitric oxide contributes to estrogen-induced vasodilation of the ovine uterine circulation. Journal of Clinical Investigation 98 2158–2166. (doi:10.1172/JCI119022)
Rosenfeld CR, White RE, Roy T & Cox BE 2000 Calcium-activated potassium channels and nitric oxide coregulate estrogen-induced vasodilation. American Journal of Physiology. Heart and Circulatory Physiology 279 H319–H328.
Schwers J, Govaerts-Videtsky M, Wiqveist N & Diczfalusy E 1965 Metabolism of oestrone sulphate by the previable human foetus. Acta Endocrinologica 50 597–610.
Scott PA, Tremblay A, Brochu M & St-Louis J 2007 Vasorelaxant action of 17β-estradiol in rat uterine arteries: role of nitric oxide synthases and estrogen receptors. American Journal of Physiology. Heart and Circulatory Physiology 293 H3713–H3719. (doi:10.1152/ajpheart.00736.2007)
Shan J, Resnick LM, Liu Q-Y, Wu X-C, Barbagallo M & Pang PKT 1994 Vascular effects of 17β-estradiol in male Sprague-Dawley rats. American Journal of Physiology. Heart and Circulatory Physiology 266 H967–H973.
Somlyo AP & Somlyo AV 1994 Signal transduction and regulation in smooth muscle. Nature 372 231–236. (doi:10.1038/372231a0)
Stauffer SR, Coletta CJ, Tedesco R, Nishiguchi G, Carlson K, Sun J, Katzenellenbogen JA & Katzenellenbogen BS 2000 Pyrazole ligands: structure-affinity/activity relationships and estrogen receptor-α-selective agonists. Journal of Medicinal Chemistry 43 4934–4947. (doi:10.1021/jm000170m)
Van Buren GA, Yang DS & Clark KE 1992 Estrogen-induced uterine vasodilatation is antagonized by l-nitroarginine methyl ester, an inhibitor of nitric oxide synthesis. American Journal of Obstetrics and Gynecology 167 828–833.
Visser M & Coelingh Bennink HJT 2009 Clinical applications for estetrol. Journal of Steroid Biochemistry and Molecular Biology 114 85–89. (doi:10.1016/j.jsbmb.2008.12.013)
Visser M, Foidart J-M & Coelingh Bennink HJT 2008 In vitro effects of estetrol on receptor binding, drug targets and human liver cell metabolism. Climacteric 11 (Suppl 1) 31–40. (doi:10.1080/13697130802056511)
Wakeling AE, Dukes M & Bowler J 1991 A potent specific pure antiestrogen with clinical potential. Cancer Research 51 3867–3873.
Wulff H, Miller MJ, Hansel W, Grissmer S, Cahalan MD & Chandy KG 2000 Design of a potent and selective inhibitor of the intermediate-conductance Ca2+ -activated K+ channel, IKCa1: a potential immunosuppressant. PNAS 97 8151–8156. (doi:10.1073/pnas.97.14.8151)
Zhang F, Ram JL, Standley PR & Sowers JR 1994 17β-Estradiol attenuates voltage-dependent Ca2+ currents in A7r5 vascular smooth muscle cell line. American Journal of Physiology. Cell Physiology 266 C975–C980.
