The role of the estrogen receptor (ER) subtypes in the modulation of vascular function is poorly understood. The aim of this study was to characterize ex vivo the functional properties of small arteries and their response to estrogens in the mesenteric circulation of female and male ER-β knockout mice (β-ERKO) and their wild-type (WT) littermates. Responses to changes in intraluminal flow and pressure were obtained before and after incubation with 17β-estradiol or ER-α agonist propyl-pyrazole-triol (3 h; 10 nM). Cumulative concentration-response curves to acetylcholine, norepinephrine, and passive distensibility were compared with respect to sex and genotype. The collagen and elastin content within the vascular wall and ER expression were also determined. Endothelial morphology was visualized by scanning electron microscopy. 17β-Estradiol and propyl-pyrazole-triol-treated arteries from female β-ERKO and WT mice showed enhanced flow-mediated dilation, but this was not evident in males. Distensibility was decreased in arteries from β-ERKO females. Sex differences in myogenic tone were observed in 17β-estradiol-treated arteries, but were similar between β-ERKO and WT mice. Acetylcholine- and norepinephrine-induced responses were similar between groups and sexes. ER-α was similarly expressed in the endothelium and media of arteries from all groups studied, as well as ER-β in WT animals. Endothelial morphology was similar in arteries from animals of both sexes and genotype; however, arterial elastin content was decreased, and collagen content was increased in β-ERKO male compared with WT male and with β-ERKO female. We suggest that ERs play a sex-specific role in estrogen-mediated flow responses and distensibility, and that deletion of ER-β affects artery structure but only in male animals. Further studies in β-ERKO mice with established hypertension and in α-ERKO mice are warranted.
the male predominance in the risk for cardiovascular disease (CVD) has generally been attributed to protection afforded to women by the female sex steroid hormones. Human and animal studies (2, 8) support the classically held hypothesis that ovarian steroids exert important cardiovascular protective actions in females and that their decline after menopause increases vulnerability to CVD. However, the explanations seem to be more complex than first thought, and the underlying mechanisms require further exploration.
Males are also susceptible to the beneficial effects of exogenous estrogens, as shown, for example, by restoration of endothelial nitric oxide (NO) release to flow-induced shear stress in arterioles of male hypertensive rats (22), enhancement of endothelium-dependent dilation in male-to-female transsexuals (38), and by the lowering blood pressure in hypogonadal men (25). Moreover, estrogen receptor (ER) polymorphisms in men, as well as women, have been linked to susceptibility and development of CVD, including myocardial infarction, angina pectoris, coronary insufficiency, hypertension, and atherosclerosis (11, 44).
ERs have been identified in vascular endothelium and smooth muscle of a wide range of blood vessels and species (33, 48), including an observation from one of our laboratories in the mouse small femoral artery (30). There is, however, relatively little information about the role of the different ERs (ER-α, ER-β) in vascular function and how the two subtypes interact in response to estrogens, partly because of a lack of subtype-specific antagonists.
The development of the ER knockout (ERKO) mice has, therefore, been useful in determination of the contribution of the receptor subtypes in determination of vascular function.
To date, the majority of in vitro studies, investigating cardiovascular function of ERKO mice, have focused on the aorta (43, 52), but studies of the smaller arteries contributing to the control of peripheral resistance and blood pressure are lacking. Indeed, it is not known whether the hypertension, which develops with age in β-ERKO mice of both sexes (52), is associated with small-artery dysfunction.
We hypothesized that the disruption of ER-β would result in functional abnormalities of small arteries, thus contributing to the elevation in blood pressure characteristic of this genotype. Functional characteristics of small mesenteric arteries of β-ERKO mice were assessed, and responses to exogenous estrogens were investigated and compared with wild-type (WT) mice. Small-artery function was evaluated by investigating responses to flow and changes in intraluminal pressure (myogenic tone and distensibility). Dilator and constrictor function were assessed by determining responses to acetylcholine (ACh) and norepinephrine (NE). The expression and localization of ER-α and ER-β in the vascular wall, endothelial morphology, and the collagen and elastin content were also assessed.
Experiments were performed on age-matched (14–22 wk old) male and female WT mice (ER-β +/+, C57BL/6J) and homozygous mutant mice lacking the gene for ER-β (ER-β −/−); (26). Mice lacking ER-β developed normally and were indistinguishable as young adults from WT littermates. Mice were housed in climate-controlled rooms, and standard rodent chow and water were available ad libitum. Animals were killed by a rising concentration of CO2, and the mesentery was removed and placed in ice-cold physiological salt solution (PSS) for dissection of resistance-sized arteries. Guidelines were followed for the care and use of experimental animals, as issued by Stockholm's Södra Djurförsöksetiska Nämnd and the Institute for Laboratory Animal Research Guide for Care and Use of Laboratory Animals. The study was also approved by the ethical committee of animal experiments at the Karolinska Institutet, Stockholm, Sweden.
Assessment of Vascular Function
Second-order mesenteric arteries (∼200 μm internal diameter) were dissected free of connective tissue and mounted on a pressure myograph (Living Systems Instrumentation), as described previously (16, 27). Vessels were equilibrated (45 min at 60 mmHg) while continuously superfused with PSS (95% oxygen in 5% CO2, 37°C). Viability was assessed by contraction to an extraluminal application of a high-potassium buffer (125 mmol/l) and NE (1 μmol/l) and by dilatation to ACh (1 μmol/l).
For the flow protocol, arteries were preconstricted with NE (1 μmol/l, 15 min) to achieve an ∼50% reduction of the initial internal diameter and until a stable constrictor response had developed. Flow was then initiated and increased incrementally at 5-min intervals (0–103.7 μl/min). Responses to flow were measured before and after incubation with either 17β-estradiol (ER-α and ER-β agonist) or propyl-pyrazole-triol (PPT) (ER-α agonist) 0.01 μmol/l for 3 h, with a 30-min washout period, as our laboratory has previously reported in human and rodents studies (9, 13, 27).
As a time control, responses to flow were carried out after 3-h incubation in PSS (sham incubation).
Cumulative concentration-response curves to NE (0.01–10 μmol/l) and ACh (0.001–3 μmol/l) were also performed separately while arteries were pressurized at 60 mmHg.
To enable the calculation of arterial myogenic tone and distensibility, responses to pressure were measured over a range of 20–100 mmHg (20-mmHg increments starting at 20 mmHg), with vessels maintained at each pressure step for 3 min. At the end of the experiment, the vessel was perfused with calcium (Ca2+)-free PSS, and the pressure steps were repeated. Myogenic tone and distensibility were then calculated.
Arterial Morphology and Structure
Second-order mesenteric arteries from each animal (male and female WT and β-ERKO) were dissected and at least one immersed in 2.5% glutaraldehyde solution in a sodium cacodylate buffer (0.15 M, pH 7.3, 24 h) and postfixed in 1% osmium tetraoxide in sodium cacodylate buffer containing 75 mM sucrose. Following dehydration in acetone and drying in a critical point dryer with CO2, samples were mounted, coated with gold palladium, and examined under a JEOL 820 (JEOL USA, Peabody, MA) scanning electron microscope. Digital images (×1,000 magnification) were collected randomly from endothelium of longitudinal arterial sections.
The collagen and elastin content were examined in 4-μm transverse paraffin sections of the small mesenteric arteries using Miller's elastic staining protocol. Color images were captured with a Nikon Eclipse E 1000 microscope (×40 objective) using a digital imaging system (Nikon ACT 1) under identical conditions of color intensity, brightness, and contrast. Images with intact structure were included in the study after careful exclusion of spoiled sections.
Immunohistochemical Analysis of ER-α and ER-β in the Mesenteric Artery
Paraffin sections (4 μm thick) of WT and β-ERKO mouse mesenteric arteries were immunostained with polyclonal rabbit antibodies against mouse ER-α and ER-β (1:100 dilution; ER-β sc-8974 and ER-α sc-542, Santa Cruz Biotechnology). Sections were incubated overnight at 4°C with the different antibodies diluted in phosphate-buffered solution-Tween. The specificity of these antibodies has been described elsewhere (7). For the negative control, the primary antibody was omitted. Human endometrial tissue section served as a positive control for ER-α (32, 48) and ovarian tissue for ER-β (48).
Chemicals and Solutions
Composition of PSS (in mmol/l) is as follows: 119 NaCl, 4.7 KCl, 2.5 CaCl2, 1.17 MgSO2, 25 NaHCO3, 1.18 KH2PO4, 0.026 EDTA, 5.5 D(+) glucose, pH 7.4. Ca2+-free PSS contained EGTA (0.001 mmol/l) and the phosphodiesterase inhibitor and Ca2+ channel blocker papaverine (0.001 mmol/l) and lacked CaCl2. Papaverine was dissolved in distilled water. 17β-Estradiol and PPT were dissolved in 95% ethanol. All chemicals were obtained from Sigma-Aldrich (Stockholm, Sweden) with the exception of PPT, which was obtained from Tocris Cookson (Bristol, UK).
All statistical analysis was performed using STATISTICA (version 6, StatSoft).
Relaxation to flow was calculated as percent change of NE-induced preconstriction. Myogenic tone and distensibility (× lied between 20 and 100 mmHg in 20-mmHg incremental steps) was calculated as: where WSS is wall shear stress; p is viscosity in poise (dyn/cm2) at 37°C; Q̇ is flow rate (μl/s); and r is artery radius (μm). Viscosity of PSS is 0.7 cP. The factor 10−9 in the equation is to correct for the use of both μl/s for flow and μl/s and μm for arterial radius (1 μl = 10−9 μm3).
Statistical significance between response curves to flow, ACh, NE, and pressure in WT and β-ERKO mice was assessed using repeated-measures ANOVA. The log of the concentration (M) of a drug required to produce 50% of the maximum response (EC50) was calculated after data were fitted to a sigmoidal curve (GraphPad Prism Software). Values of maximum responses and EC50 values were assessed using two- or one-way ANOVA, whichever was the most appropriate. Results are expressed as the means ± SE, unless indicated otherwise, and n denotes the number of animals used in each experiment. The evaluation of ER expression was performed using semiquantitative analysis. The staining was scored blindly by three observers, + if >5–25% staining, ++ if 25–50%, +++ if >50%, and the mean of the score in arteries from each animal was used for statistical analysis.
The distribution of elastin and collagen was evaluated in high-resolution printed color digital images by point counting in a 1-cm2 lattice to determine the elastin and collagen/total area volume density, as previously described (17, 49). The area occupied by elastin and collagen was expressed as a percentage of the total artery area. The point counting was carried out by two observers and blinded, and the mean value for arteries from each animal is reported. At least five views were captured and analyzed for each artery.
Mann-Whitney statistics were applied for analysis, and P < 0.05 in all comparisons was considered to be statistically significant.
Body Weights and Artery Diameters
There was no difference in body weight between WT females (WTF) and β-ERKO females (BF) or between WT males (WTM) and β-ERKO males (BM) (see Table 1). WTM and BM were significantly heavier than WTF and BF (P < 0.05). There was no difference in basal internal diameter nor in diameter after preconstriction with NE before the flow response in any of the experimental groups studied (see Table 1).
Dilatation to flow was observed in all groups and was characterized by an immediate response followed by a sustained dilatation. No difference in the response to flow over the entire range of flow increments was observed between WT and β-ERKO mice (n = 12–13 in each group, P > 0.05 ANOVA, Fig. 1). The range of WSS values calculated before change in flow steps was also similar between WT and β-ERKO mice (WSS in dyn/cm2: WTF from 6 ± 1 to 55 ± 9, WTM 10 ± 5 to 73 ± 16, BF 7 ± 1 to 69 ± 10, BM 11 ± 3 to 62 ± 17 under range of flow rate from 9 to 103.7 μl/min) and, therefore, within the physiological range (10).
After incubation with 17β-estradiol, flow-mediated dilation increased to a similar extent in both groups of females: WT (%relaxation at maximum flow: 5 ± 3 before vs. 42 ± 13 after incubation; P <0.05, n = 11, Fig. 2A) and β-ERKO (%maximal response: −3 ± 5 before vs. 39 ± 13 after incubation; P <0.05, n = 12, Fig. 2B). Enhanced dilation to flow was also observed after incubation with the selective ER-α agonist PPT in mesenteric arteries from WTF (%maximal response: 9 ± 6 before vs. 42 ± 16 after incubation, P < 0.05, n = 7, Fig. 2C).
17β-Estradiol also increased the flow response in arteries from WTM (%maximal response: 16 ± 7 before vs. 43 ± 14 after incubation; P < 0.05, n = 13, Fig. 3A), but, in contrast to the BF mice, 17β-estradiol did not affect the flow response in arteries from BM animals (%maximal response: 8 ± 4 before vs. 14 ± 2 after incubation, n = 9, Fig. 3B), nor in arteries from WTM mice pretreated with PPT (WT: %maximal response: 5 ± 3 before vs. 7 ± 3, n = 5, Fig. 3C). Similarly, PPT had no effect on the flow response in arteries from BM mice (%maximal response: 2 ± 3 before vs. 8 ± 6, n = 6, Fig. 3D).
ACh-induced dilatation and NE-induced constrictor responses were similar between groups (P > 0.05). Deletion of ER-β had no influence on these dilatory and contractile responses (n = 7–10, Table 1).
There was no difference in myogenic tone between any of the groups studied. Mesenteric arteries demonstrated myogenic tone of 10 ± 3, 9 ± 2, 7 ± 3, and 9 ± 2% (e.g., at intraluminal pressure of 60 mmHg) in WTF and BF and WTM and BM, respectively. After incubation with 17β-estradiol, a significant decrease in myogenic tone was observed in arteries from both WTF and BF (P < 0.05), but not from WTM or BM (n = 11–13, Fig. 4, A and B, respectively).
Increases in intraluminal pressure from 20 to 100 mmHg led to a stepwise increase in diameter of the mesenteric vessels from WT and β-ERKO male and female mice. Vascular distensibility (passive increase in diameter as a function of pressure) was not significantly different between males and females in either group, nor was it different between WTM and BM. However, mesenteric arteries from WTF were more distensible (P < 0.05) than those of BF (n = 11–13, Fig. 5, A and B, respectively).
ER-α and -β were localized within both endothelium and media. No staining for ER-β was observed in arteries from BF or BM (Fig. 6). Semiquantitative analysis of ERs revealed that ER-α was similarly expressed in the endothelium and media in arteries from all groups studied, as well as ER-β in WTM and WTF animals (P > 0.05, n = 3–6, Table 1).
Endothelial Morphology, Collagen, and Elastin
The entire surface of the arterial lumen, obtained from mesenteric arteries with internal diameter comparable with that of arteries involved in functional studies, was examined by scanning electron microscopy. In arteries from WT and β-ERKO mice, a continuous sheath of elongated and tightly connected endothelial cells covered the luminal surface of the intima, with no changes evident in respect to genotype or sex (Fig. 7, n = 3–4).
Collagen and elastin distribution was similar in arteries from WTF and BF (Fig. 8, A and C, and Table 1). Arterial elastin area was lower in BM compared with WTM and to BF (P < 0.05). The area of the arterial wall occupied by collagen was increased in BM compared with WTM and BF (P < 0.05, n = 3–6, Fig. 8, B–D, and Table 1).
The principal finding of this study was that disruption of ER-β results in sex-specific effects on estrogen-enhanced flow-mediated dilation, arterial distensibility, and structural composition of the small artery wall.
Estrogens have previously been observed to improve in vivo flow-mediated dilation in both men and women (18, 28) and in vitro to enhance flow response in isolated arteries from women with preeclampsia (47). 17β-Estradiol also improves flow-mediated dilatation in arteries from rodents, including isolated small arteries from prepubertal female (9) and hypertensive male (22) and female rats (21). Receptor subtype dependence of this potentially important functional response has not been considered previously, but this study suggests sex and receptor subtype dependence of the responses, in which ER-β appears to play a role in male mice, but in females involvement of ER-α is implicated. The central role for ER-α in the female concurs with recent observations from our laboratories in isolated, small arteries from healthy postmenopausal women, in which PPT (an ER-α agonist) but not genistein (a natural ligand for ER-β) mimicked the 17β-estradiol-induced dilation in response to flow (27).
Arterial dilation to flow arises from physical tangential force-shear stress at the endothelial surface (10). We have not investigated the mechanisms behind the estrogen-enhanced dilation to flow; however, it may occur through NO via endothelial NO synthase (9, 21, 27, 47), endothelium-derived hyperpolarizing factor (23), and/or second-messenger signaling pathways (3, 20, 45).
Both ERs are expressed in the vasculature of humans and rodents (30, 31, 48). We demonstrate here the presence of ER-α and ER-β in the mesenteric artery wall, with no difference in expression between sexes in the WT group. ER expression has been shown to be tissue and vascular bed specific, as well as hormone status and sex dependent (6, 19, 39). Our data suggest sex differences in the subtype of ER utilized to mediate the physiological response to flow. Other lines of evidence support sex-linked properties of the ERs, e.g., estrogen-induced vascular protection has been attributed to ER-α in female mice (4, 40), whereas the same function has been attributed to ER-β in males (1, 29).
Since myogenic tone was similar between WT and β-ERKO mice of both sexes, ER-β appears not to modulate myogenic tone. The lack of sex differences in either WT or β-ERKO mice agrees with previous studies in mouse mesenteric (15) and rat cerebral arteries (46). A reduction in myogenic tone has been shown to be due to estrogen-dependent increase in NO synthesis in female arteries (12, 50), and the observation that basal production of NO is reduced in the aorta of the α-ERKO male mouse (43) would suggest that the estrogen-induced reduction in myogenic tone is ER-α dependent. As in our study, 17β-estradiol preincubation led to a similar decrease in myogenic tone in arteries from female animals (both WT and β-ERKO mice); it may be presumed that ER-β plays no role in this response and that ER-α predominates. Our laboratory has recently reported reduced myogenic tone in response to 17β-estradiol and PPT, but not genistein, in small arteries from postmenopausal women (27), which would also lend support to ER-α playing a functional role in basal tone in the female vasculature. Myogenic tone in small arteries from WTM and BM mice was, however, insensitive to incubation with 17β-estradiol, thus differing from one report in male rat femoral arteries, in which prolonged (>12 h) incubation with 17β-estradiol reduced basal tone through a NO-dependent pathway (22). This discrepancy between studies may be related to vascular bed specificity or incubation time.
The observation of decreased arterial distensibility in mesenteric arteries from BF mice compared with WT animals accords with others studies suggesting that estrogens affect arterial distensibility. Estrogen supplementation to aging rats has been shown to increase mesenteric artery distensibility (51), and in postmenopausal women hormonal replacement therapy increases compliance in the carotid artery (5, 35) and systemic arterial circulation (41). The observation that distensibility was similar in mesenteric arteries from BM and WTM mice is consistent with studies of healthy men, in which decreased concentrations of endogenous estrogens did not alter arterial compliance (28) and of reports in male-to-female transsexuals, in which estrogen therapy did not alter arterial compliance (37).
Disruption of ER-β resulted in alteration of arterial wall structure in the mesenteric artery, as assessed by evaluation of collagen and elastin content, but only in males. This might argue against the preserved distensibility in arteries from BM, but reduced in BF. The mechanism for this remains unclear. The abnormal organization of extracellular matrix (ECM) proteins, altered elastin and collagen content, and their nonenzymatic glycation-distorted spatial organization of elastic fibers in the resistance vasculature has been associated with development of hypertension in humans and experimental animals (14, 24, 42), and this may be a contributing mechanism here. Abnormal ECM composition characterized by accumulation of collagen has been demonstrated in lungs from BM mice (34), further supporting the role of ER-β in modulation of the ECM. Future studies of small-artery structure and mechanisms of remodeling in prehypertensive and hypertensive β-ERKO mice could provide added insight into the mechanisms responsible (36).
In conclusion, the absence of the ER-β is associated with sex-specific effects on estrogen-enhanced, flow-mediated dilation, distensibility, and vascular structure in the small mesenteric arteries of the mouse. Estrogens, through the action of ER-β, may increase distensibility and hence vascular compliance in the vasculature of the female mouse. In the male mouse, the same receptor may have an important influence on the control of estrogen-enhanced, flow-mediated dilation, whereas, in the female, ER-α appears central to this response. In the male animals, deletion of ER-β significantly altered the content of elastin and collagen within the vascular wall, while myogenic tone was unaffected by the loss of ER-β in either male or female mice. The sex-specific differences in myogenic tone after incubation with 17β-estradiol implicate ER-α in the modulation of basal tone of the female vasculature. Further studies in β-ERKO mice with established hypertension and studies in α-ERKO mice are warranted.
We acknowledge the support of the Swedish Heart and Lung foundation, the Centre of Gender Related Medicine at the Karolinska Institutet, and Tommy's the Baby Charity, UK.
J. Gustafsson has financial interests (consultancy, stock ownership, and pending corporate grants) with KaroBio AB, Sweden.
We gratefully thank Kjell Hultenby for help with studies on small-artery morphology and structure.
↵* G. Douglas and M. N. Cruz contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society