Cardiovascular Drugs and Therapy 2000;14:303–307 C Kluwer Academic Publishers. Boston. Printed in U.S.A. °
Long-Term Cardiovascular Effects of High “Osteoprotective” Dose Levels of 17β-Estradiol in Spontaneously Hypertensive Rats Jacques Blacher1 , Hubert Dabire1 , Jean P. Pomies1 , Michel E. Safar1 , and Michael Stimpel 2 1 Department
of Internal Medicine, Broussais Hospital, and INSERM U337, Paris, France; 2 Medical School, University of Cologne, Germany
Summary. The effects of estrogen replacement therapy in menopausal women are more obvious on bones than on the cardiovascular system. The optimal estrogen dosage may differ in these different parts of the body. In hypertensive rats, low doses have been shown to reduce arterial collagen and stiffness, whereas higher dosages are required for osteoprotection. From 4 to 20 weeks of age, female spontaneously hypertensive rats (SHRs) were divided into four groups: without ovariectomy, under placebo or 17βestradiol (10 µg/kg/day), and with ovariectomy under either placebo or 17β-estradiol (same dosage). Serial tail systolic blood pressure measurements were performed, and histomorphometry of the thoracic aorta was determined at the end of the study. Under estrogen, blood pressure was unchanged, whereas the aortic wall–to–lumen ratio was increased, particularly in the presence of ovariectomy. The elastin to collagen ratio was significantly decreased, due both to a decrease in elastin and an increase in collagen density, with no change in media thickness. The latter findings were not observed when ovariectomy was performed. Independent of changes in wall stress, high-dose estrogen increases the aortic extracellular matrix in female SHRs. This increase may be reversed in the presence of ovariectomy, suggesting that estrogen was not the only gonadal factor responsible for altered vascular structure and function. Key Words. female spontaneously hypertensive rats, estrogens, arterial stiffness
In previous studies in ovariectomized WKY rats and SHRs [7,8], we have shown that the dosage of 10 µg/kg/day of 17β-estradiol was able to protect efficiently against rapid bone loss. On the other hand, in order to reduce aortic collagen in rats, low doses (2 µg/kg/day) of 17β-estradiol have been widely used [9–11]. Thus the question arises as to whether the optimal hormonal dosage for bone tissue is identical to that required for vascular structure. This question is potentially important to consider because estrogenic effects on arterial structure might partly account for the previously described increased risk of cardiovascular disease associated with relatively high-dose estrogen administration in men with prostate cancer or a myocardial infarction [12,13]. The purpose of the present study was to determine whether high-dose and long-term administration of 17β-estradiol, in female SHRs significantly modified the aortic wall in terms of vascular thickness and the extracellular matrix composition. Because in previous studies [9–11] the presence of sexual glands was shown to potentially modify the changes in vascular structure, both sham-operated and bilaterally ovariectomized animals were studied.
Methods Natural menopause is associated with an increased incidence of coronary risk and cardiac events . Numerous studies suggest that estrogen replacement therapy protects post-menopausal women from coronary heart disease and could therefore reduce cardiovascular risk [1,2]. However, estrogen replacement therapy acts not only on the cardiovascular system but also on bone metabolism . Clinical randomized trials have shown that the beneficial effect of estrogens on bone and menopausal symptoms is clearly more potent than that obtained for the cardiovascular system [4–6]. Thus animal studies should be developed to differentiate the estrogen effects on bone and arterial vessels.
Study design Forty female SHRs rats (Iffa-Credo, L’Abresle, France) were housed five to seven per cage in our animal room (temperature, 20–22◦ C; humidity, 55–65%; 12-hour light/dark cycles), with free access to tap water.
´ Address for correspondence: Professeur Michel Safar, Medecine ˆ Interne 1, Hopital Broussais, 96 rue Didot, 75674, Paris Cedex 14, France. E-mail: [email protected]
Received 14 May 1999; receipt/review time 2 months; accepted 6 October 1999
Blacher et al.
Animals were sham-operated or bilaterally ovariectomized at the moment of weaning, that is, at the age of 3 weeks. A dorsal approach was used. Treatment of placebo was begun at the age of 4 weeks. Estradiol (10 µg/kg/day) or saline (0.2 mL/100 g) was given each day for 16 consecutive weeks using subcutaneous administration. Four groups were randomly assigned according to the presence of ovariectomy and/or estrogen treatment.
Arterial pressure and body weight in conscious rats Body weight gain was recorded monthly in all animals at the ages of 4, 8, 12, 16, and 20 weeks. Systolic blood pressure (SBP) was measured in awake animals by electrosphygmomanometry of the caudal artery (Pressure Computer LE 5007, Letica°R ). Each measurement was performed at least five times for each animal, and the mean was calculated. Before SBP measurement, rats were made accustom to the measurement device (generally two measurements were necessary for rats to get accustomed). Measurements were performed following 4, 8, 12, and 16 weeks of treatment (i.e., at the ages of 8, 12, 16, and 20 weeks) in all animals at around 22–26 hours after drug administration. Histomorphometric study of the thoracic aorta  After the animals were anesthetized with pentobarbital, a median thoracotomy was performed. Rats were exsanguinated by means of a catheter placed in the right atrium while saline was injected through the femoral catheter. When the liquid exiting the atrium was clear, the circulatory system was perfused with a 4% formaldehyde solution. The animals died within seconds following the onset of formaldehyde infusion. After 1 or 2 minutes, a clamp was positioned on the atrium, and the fixation liquid was allowed to infuse for 3 hours at a pressure equal to the blood pressure of each animal [14–16]. Then the thoracic aorta was dissected out and stored in a 4% formaldehyde solution until the histological study was performed. The different components of the vessel wall were examined in an arterial segment embedded longitudinally in paraffin. Three serial sagittal sections, 5 µm thick, were specifically stained to obtain a monochromatic color associated with the various structures of interest in the aortic media: Sirius red for collagen staining, orcein for elastin, and hematoxylin after periodic acid oxidation for the nucleus. Morphometric analysis was performed with a specialized automated image processor (NS 1500, Nachet-Vision, Paris, France), which was based on morphological mathematical principles and is software controlled. As previously described [14–16], different algorithms were developed to analyze each of the three structures (media thickness, elastin, collagen) revealed by the specific staining in each of the three serial sections.
For image processing, the image is transmitted to the processor via a video camera and can be viewed on the TV monitor. The control of luminosity is automatically adjusted by the software in order to obtain similar contrasts, taking into account the total luminosity transmitted by the video camera. This analogue image is then digitized as follows: each elementary point (pixel) is automatically compared with a threshold; if the gray level of pixel exceeds this threshold, the pixel is given the numeric value 1; otherwise, it is given the numerical value of 0. Threshold parameters are defined by the size of such pixel groups and their local contrast (top-hat transformation). The threshold was determined using the top-hat transformation algorithm so as to minimize variations in nuclear staining and background. Media (M) and lumen (L) cross-sectional areas (CSAs) were measured in arterial samples placed in a gel used for cryosections (medium inclusion Isosystem) and cooled to −20◦ C. When the gel solidified, some transverse sections of the arterial rings were cut from each sample. The sections were examined with a microscope and photographed at a known magnification (×48). When the films were developed, the pictures obtained were projected onto a digitizer in order to measure the area of the vascular lumen and the external area of the vessel. The final magnification used (microscope amplification × projector magnification) was 160. For each sample, the CSAs of the vascular media and lumen were recorded. This latter parameter reflects the degree of arterial hypertrophy better than medial thickness. Different studies have shown that the CSA of the arterial media is the most reliable constant of the vessel wall because it is not influenced by variations of the perfusion pressure [14–16]. Results are presented as mean value ± 1 standard deviation using NCSS 6.0.21 software. A two-way analysis of variance was performed in order to determine an estrogen effect, an ovariectomy effect, and an interaction between those two factors. Posthoc statistical evaluations were performed using the Tukey Kramer test. A P value