INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE MECHANICAL PROPERTIES of large elastic arteries are important determinants of circulatory physiology (24). Decrease in arterial compliance, which reflects the ability of an artery to expand and recoil with cardiac pulsation and relaxation (1), has been identified as an independent risk factor for cardiovascular disease (2, 28). Premenopausal women are at lower risk for developing cardiovascular disease than men of similar age (4). Data on estrogen effects on arterial distensibility are contradictory. Pregnancy, a physiological condition characterized by a wide variety of morphological and hormonal changes, including increased estrogen formation, is associated with increased aortic distensibility, together with decreased venous distensibility and viscoelasticity in the lower, but not the upper, limb (8). Menopause is associated with impaired left ventricular (LV) systolic performance (29) due, at least in part, to a decrease in arterial distensibility (15). Furthermore, a protective effect of long-term estrogen therapy on age-related changes in systemic arterial compliance has been described in postmenopausal women (18, 26, 33). In contrast, gender has no significant effects on carotid midwall strain in normal adults (5), and administration of estrogens increases femoral and brachial artery stiffness in men (10).

Little is known about the putative modulatory influence of estrogen on arterial stiffness. A recent study in spontaneously hypertensive rats (SHR) reported that estrogen decreases aortic lumen, medial cross-sectional area, and elastin-to-collagen ratio (6). Inasmuch as arterial structural composition influences arterial elasticity, our first goal was to study the effect of chronic estrogen therapy on arterial stiffness in the rat. Thus thoracoabdominal pressure pulse wave velocity (PWV), which is inversely related to arterial wall distensibility, was measured in anesthetized rats.

The endothelium and its derivatives are known to influence systemic arterial compliance (13, 14). Furthermore, many reports have described evidence that estrogen modulates vascular endothelium-dependent responses (9, 17). In animals, the vascular action of estrogen has been mostly investigated in vitro, and little information is available regarding the effects of estrogen on endothelial function in vivo. Therefore, our second goal was to explore estrogen effects on regional hemodynamics in rats. We monitored changes in central arterial blood pressure and carotid resistance (CR) and hindquarters vascular resistance (HQR) after administration of vasoactive agents in rats.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

All experiments and protocols were performed in accordance with the regulations of the Canadian Council on Animal Care and were approved by the Animal Care Committee at Laval University. Studies were carried out on female Sprague-Dawley (SD) rats (Charles River, St. Constant, QC, Canada) weighing 200-250 g and kept under standard conditions for 1 wk before experiments.

Estrogen treatment. Animals of the appropriate groups were ovariectomized by bilateral flank incision under isoflurane anesthesia. Experimental treatments began on the day after ovariectomy. Rats were randomly assigned to 3 groups of 15 animals each: sham-operated SD rats receiving vehicle (Sham), ovariectomized SD rats receiving vehicle (OVX), and ovariectomized SD rats receiving 17-estradiol (OVX-E₂). 17-Estradiol (E₂, 5 mg in Silastic tubing) was inserted subcutaneously in the dorsal area of the animal.

Measurement of hemodynamic parameters. After 20 days of treatment, PWV and hemodynamic responses to bradykinin (BK), sodium nitroprusside (SNP), and angiotensin II (ANG II) were measured in isoflurane-anesthetized fasted rats.

Aortic blood pressure, PWV, and pulse amplification in anesthetized rats. After completion of surgery, rats were allowed to equilibrate for 1 h before measurements of baseline central and peripheral aortic blood pressures and PWV as described earlier (31). PWV (cm/s) was calculated as the distance between the two central and peripheral cannula tips divided by transit time. The distance between the two central and peripheral cannula tips was measured in situ after postmortem fixation by sticking a damp cotton thread onto the aorta (8.4 ± 0.2, 8.5 ± 0.2, and 8.3 ± 0.3 cm in Sham, OVX, and OVX-E₂ rats, respectively). Transit times (ms) between the two central and peripheral pressure signals were measured on-line for each 5-s period by peak detectors of the AcqKnowledge software that systematically shifted in time the peripheral pressure waveform with respect to the central pressure waveform and determined the value of the time.

Hemodynamic responses to BK, SNP, and ANG II in anesthetized rats. After PWV measurement, baseline cardiovascular parameters (central arterial blood pressure, heart rate, HQBF, CBF, HQR, and CR) were recorded. Then hemodynamic responses to BK (0.03-3 µg/kg), SNP (1 and 10 µg/kg), and ANG II (10 and 100 ng/kg) were assessed. PWV and BK hemodynamic studies were measured in one set of rats (n = 10/group) and SNP and ANG II responses in another set (n = 5/group).

1

1

1

1

1

1

1

1

Aortic histomorphometry, wall stress, and elastic modulus. At the end of hemodynamic studies, rats were perfused for 30 min with 10% formaldehyde-containing phosphate-buffered saline. Because fixation pressure can affect aortic dimensions, rats were perfused at their in vivo mean carotid blood pressure level. The histomorphometric technique has been described in detail previously (31). A 1-cm sample of the thoracic descending aorta was excised, immersed in 10% formaldehyde solution, dehydrated in graded ethanol solutions, and embedded in paraffin. Aortic sections (20 µm thick) were stained with hematoxylin-eosin for determination of aortic internal diameter and medial thickness as described earlier (31).

Steroid measurements. Determination of steroids was performed using high-performance gas chromatography and negative chemical ionization mass spectrometry. On each day of analysis, six calibration curve standards ranging from 5 to 200 pg/ml were prepared using male gonadectomized rat serum. For E₂ extraction from rat serum, 400 µl of a 0.5 M sodium acetate solution and a methanolic solution (50 µl) containing internal standards were added to each tube. Aliquots (400 µl) of study samples were added to the appropriate tubes, and all tubes were vortexed for 1 min. A solution of 1-chlorobutane (3 ml) was added to each tube and mixed. Tubes were centrifuged, and the organic extracts were purified on LC-Si SPE columns. Columns and the adsorbed material were washed with 6 ml of 1:9 (vol/vol) ethyl acetate-hexanes. The analytes of interest were then eluted using 6 ml of 50:50 (vol/vol) ethyl acetate-hexanes. The eluates were collected and evaporated at 35°C. The dried residue was reconstituted in ethyl acetate (0.4 ml) and vortexed for 15 s. An aliquot of 200 ml was transferred to a glass tube for E₂ assay. The extracts were then completely evaporated (Speed-Vac Evaporator, Savant Instruments, Farmingdale, NY). Pentafluorobenzoylchloride in ethyl acetate (50 µl, 1:10, wt/vol) and pyridine in ethyl acetate (500 µl, 1:99, vol/vol) were added to the dried residue of E₂, and the samples were incubated for 30 min at 60°C. After evaporation of the reagent mixture, a solution of 0.5 M NaHCO₃ (1 ml) was added to the tubes, which were then left for 15 min at room temperature. Hexanes (2 ml) were added to the samples, which were mixed for 2 min and left at room temperature for 10 min. The organic phase was evaporated at 50°C, and the final extract was reconstituted in 40 µl of isooctane and then transferred into a conical vial for injection into the gas chromatography-mass spectroscopy instrument.

Data analysis. Values are means ± SE; n refers to the number of rats in each group. Experiments were performed according to a randomized block design. Data were analyzed by repeated-measures analysis of variance followed by Student's modified t-test with Bonferroni's post hoc test for multiple comparisons between means. Differences between means were considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Physiological parameters. OVX rats exhibited a significant (P < 0.05) body weight gain and uterus atrophy that were completely abolished by E₂ treatment (Table 1). As expected, E₂ levels were below the detection limit (10 pg/ml) in OVX rats. Estrogen replacement (OVX-E₂ rats) restored E₂ levels to values measured in Sham rats (Table 1).

Aortic blood pressures, indexes of aortic wall stiffness, LV mass, and thoracic aorta and cardiac histomorphometry. Central and peripheral aortic systolic, diastolic, and mean blood pressures were similar in all groups (Table 2). Estrogen treatment also increased LV mass (Fig. 1). Central and peripheral pulse pressures were increased in OVX-E₂ compared with OVX control rats (Table 2). Pulse pressure increased by 25% between central and peripheral aortic recording sites in all groups; aortic pressure wave amplification, wall stress, and elastic modulus were not modified (Table 2). Although medial thickness was not affected, internal diameter and lumen cross-sectional area were significantly decreased in OVX-E₂ rats (Table 2). Estrogen significantly increased the medial thickness-to-internal diameter ratio. Ovariectomy significantly reduced PWV, which was normalized by estrogen replacement (Fig. 1).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Thoracoabdominal pulse wave velocity (PWV) and left ventricular (LV) mass in sham-operated rats treated with vehicle (Sham), ovariectomized rats treated with 17-estradiol (OVX-E2), and ovariectomized rats treated with vehicle (OVX). Values are means ± SE. *P < 0.05 compared with OVX. ¶P < 0.05 compared with Sham.

Baseline cardiovascular parameters and hemodynamic responses to BK, SNP, and ANG II. Baseline values of cardiovascular variables in anesthetized animals were similar in all groups (Table 3). The resting hemodynamic variables before injection of each dose of vasoactive agents were similar in all groups (Table 4). Figure 2 depicts the hemodynamic responses to the endothelium-dependent vasodilator BK. Bolus injection of BK produced dose-dependent falls in MABP, CR, and HQR. Ovariectomy tended to reduce BK-induced hypotension and vasodilatory responses, but it failed to reach the level of statistical significance. However, estrogen replacement increased BK-induced hypotension (P < 0.05 at 0.3 µg/kg), and this was associated with a marked potentiation of the vasodilatory responses in the hindquarters but not in the carotid vascular bed.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Dose-dependent bradykinin (BK)-induced changes in mean arterial blood pressure (MABP), carotid resistance (CR), and hindquarters resistance (HQR) in Sham, OVX, and OVX-E2 rats. Values are means ± SE. *P < 0.05 compared with OVX.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Sodium nitroprusside (SNP)- and ANG II-induced changes in MABP, CR, and HQR in Sham, OVX, and OVX-E2 rats. Values are means ± SE.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Mechanical and structural properties of arteries and their interactions have been studied extensively with respect to aging, hypertension, and antihypertensive treatments. We designed this study to investigate the effect of estrogen replacement on arterial distensibility and endothelial function. We determined in vivo aortic wall properties by measuring thoracoabdominal PWV. This parameter was used for the calculation of two other indexes of wall stiffness: elastic modulus and wall stress. The measurement of PWV has been widely used for estimating elasticity and compliance of large vessels in humans (2, 8), rats (3, 31), and mice (35).

The results of this study demonstrate for the first time that ovariectomy significantly decreases PWV, suggesting an increase in arterial distensibility. To our knowledge, this finding has not been reported earlier. Interestingly, there were no significant changes in the calculated elastic modulus and wall stress in the rat. This could indicate that arterial wall dimensions/composition play a minor role in the estrogen-related changes in the measured PWV.

Estrogen-treated rats were associated with increased PWV, LV hypertrophy, and increased pulse pressure. The LV hypertrophy observed in OVX-E₂ rats is consistent with the increased LV mass and the ratio of mass to cross-sectional area of right ventricular papillary muscles documented in estrogen-treated rabbits (24). Furthermore, it is well established that LV hypertrophy is linked to increased aortic stiffness and pulse pressure (7, 31). Histomorphometric analysis of the descending aorta showed that estrogen produced an increase in medial thickness-to-internal diameter ratio and mostly a decrease in medial thickness, internal diameter, and lumen cross-sectional area. Our histomorphometric findings are in line with previous observations in SHR (6). The mechanisms by which ovariectomy and E₂ affect cardiac and aortic geometry are unclear. However, one cannot exclude the possible impact of body size and growth rate on aortic dimensions. For example, chronic ANG I-converting enzyme inhibition with perindopril, which decreases body weight and slows body weight gain, is associated with reduced aortic and carotid dimensions (19). Indeed, body weight was significantly lower in Sham and OVX-E₂ rats than in OVX rats.

Arterial stiffness can be influenced by several factors such as heart rate (16, 21), blood pressure (33), atherosclerosis plaque (3, 34), age (19, 33), and body size and composition and height (11, 15, 27, 30). Indeed, the heart rate-dependent reduction in arterial distensibility and compliance is primarily due to the viscous nature and inertial behavior of the vessel wall, which implies that a shortening of the time available for recoil results in vessel stiffening (16, 21). It is unlikely that the difference in PWV observed in the present study could result from an increase in heart rate or blood pressure, inasmuch as both parameters did not differ between the experimental groups. This does not preclude the possible subtle impact of the marginal increases in heart rate or blood pressure observed in OVX-E₂ rats. Thus the combination of small but nonsignificant increases in heart rate and blood pressure could have contributed, at least in part, to the increased PWV in estrogen-treated rats. A recent study demonstrated that estrogen significantly reduces the elastin-to-collagen ratio as a result of a reduction of elastin and an increase in collagen in the aorta of ovariectomized SHR (6). Although we did not measure aorta wall composition in the present study, it is plausible to speculate that the decreased PWV observed in OVX rats could result, at least in part, from changes in the aorta wall structure and composition. Another explanation for the decreased arterial stiffness in OVX rats resides in the larger body size. As stated earlier, body size, composition, and height are major determinants in systemic arterial compliance. Short stature is associated with disproportionally high systolic and pulse pressures, faster heart rate, and shortened return times for reflected waves (22, 30). In contrast to ovariectomy, estrogen replacement significantly decreased body weight; however, the aorta length was similar in all groups. The decreased arterial distensibility after estrogen treatment observed in this study is in line with the observation of Giltay and colleagues (10), who showed that estrogen treatment for 4 but not 12 mo decreases femoral and brachial artery distensibility in young men. However, this contrasts with other studies where postmenopausal estrogen replacement therapy increases systemic arterial compliance (18, 26, 33). The reasons for these conflicting findings are unknown, but factors such as dose and duration of hormone therapy and experimental approaches could lead to contradictory results.

Hemodynamics and endothelial factors play a major role in blood vessel structure and function (12). In the present study, we evaluated endothelial function by measuring hemodynamic responses to vasoactive agents. Ovariectomy did not affect significantly the hemodynamic responses to endothelium-dependent and -independent vasoactive agents. On the other hand, the BK-induced hindquarters, but not carotid, vasodilatory hemodynamic responses were potentiated by estrogen supplementation. This could suggest a complex tissue/vasoactive agent-dependent heterogeneity in estrogen-modulatory effects on vascular function. The effect of estrogen on BK responses was more obvious on hindquarters peripheral resistance than on arterial blood pressure. This indicates that measurement of systemic arterial blood pressure alone in rats does not allow reliable evaluation of cardiovascular changes. Furthermore, the estrogen effect on BK-induced hypotension was significantly potentiated at 0.3 but not 3 µg/kg. Why estrogen potentiates BK responses (hypotension) at a low but not a high dose is unclear. However, some explanations can be suggested: 1) estrogen differentially modulates BK receptor-coupled mechanisms at low and high doses, and 2) higher doses of BK may induce vasoconstrictor responses (via BK type 1 receptors on smooth muscle cells), which may have masked a possible enhancement of endothelial vasodilator factor synthesis and/or release (mediated by BK type 2 receptors).

BK triggers the release of endothelial factors, including vasodilator prostaglandins, endothelium-derived hyperpolarizing factor, and nitric oxide, which in turn relax the underlying smooth muscle. Because estrogen did not affect SNP and ANG II hemodynamic responses, it is tempting to suggest that estrogen restores endothelial vasodilator function in OVX rats. The putative mechanisms responsible for the beneficial effects of estrogen on endothelial function in the present study remain to be identified. As suggested by others (9, 18), it is possible that chronic estrogen treatment enhances endothelial function via a nitric oxide-dependent pathway through nongenomic and/or transcriptional mechanisms. However, there are also studies reporting that estrogenic treatment does not affect (20) or attenuates (23) acetylcholine-mediated vasodilatory responses. The reasons for these conflicting findings are unknown, but factors such as animal strain, type and dose of vasoactive agents, dose and duration of hormone therapy, and experimental approaches appear to influence the observations.

Our findings showed that estrogen potentiates endothelium (BK)-dependent vasodilatory responses in the hindquarters and yet reduces arterial distensibility in OVX rats. Experimental studies have demonstrated that the endothelium plays an important and active role in vascular remodeling and arterial distensibility (8, 30). Levy and co-workers (13, 14) reported that removal of the endothelium increases carotid compliance and diameter in Wistar-Kyoto rats and SHR, suggesting that vasoconstrictive compounds of endothelial origin are involved in the control of the viscous and elastic properties of the arterial wall. We have not tested the influence of estrogen on the vasopressor/vasoconstrictor responses in the present study. Inasmuch as BK-induced vasodilatory responses were potentiated by estrogen treatment, it is unlikely that the decrease in arterial elasticity observed in estrogen-treated rats is the result of a defective endothelial vasorelaxant function. However, this does not preclude a possible upregulation of endothelial vasoconstrictive factors. Further studies are needed to test this hypothesis.

In conclusion, the present study demonstrates for the first time that estrogen deficiency increases aortic distensibility as measured by PWV in normotensive rats. Estrogen replacement-induced increase in arterial stiffness was associated with LV hypertrophy and amplified pulse pressure. The deleterious effect of estrogen on aortic distensibility is somewhat surprising, given increasing evidence of the cardioprotective properties of estrogen. Although any extrapolations from the rat model to the human require cautions, the estrogen-induced decrease in arterial distensibility observed in the present study could represent one of the multiple negative side effects of estrogen replacement therapy. We also showed that estrogen potentiates endothelial vasodilator function in the hindquarters but not in the carotid bed, suggesting a regional heterogeneity in the modulatory effects of estrogen on vasomotor function.