Antenatal glucocorticoids are used to promote the maturation of fetuses at risk for preterm delivery. While perinatal glucocorticoid exposure has clear immediate benefits to cardiorespiratory function, there is emerging evidence of adverse long-term effects. To determine if antenatal betamethasone alters vascular reactivity, we examined isometric contraction of endothelium-intact coronary and mesenteric arteries isolated from twin fetal sheep at 121-124 days gestation (term being 145 days). One twin received betamethasone (10 μg/h iv) while the second twin received vehicle (0.9% NaCl) for 48 h immediately before the final physiological measurements and tissue harvesting. Fetuses that received betamethasone had higher mean arterial blood pressures than the saline-treated twin controls (53 ± 1 vs. 48 ± 1 mmHg, P < 0.05). Coronary vessels from betamethasone-treated fetuses exhibited enhanced peak responses to ANG II (72 ± 17 vs. 23 ± 6% of the maximal response to 120 mM KCl, P < 0.05). There was no significant difference in response of the coronary arteries to other vasoactive compounds [KCl, U-46619, sodium nitroprusside, 8-bromo-cGMP (8-BrcGMP), isoproterenol, and forskolin]. Contractile responses to ANG II were similar in betamethasone and control mesenteric arteries (48 ± 17 vs. 36 ± 12% of the maximal response to 10-6 M U-46619). Western blot analysis revealed AT1 receptor protein expression was increased by betamethasone in coronary but not in mesenteric arteries. These findings demonstrate that antenatal betamethasone exposure enhances coronary but not mesenteric artery vasoconstriction to ANG II by selectively upregulating coronary artery AT1 receptor protein expression.
- coronary artery
- vascular smooth muscle
consensus statements now widely support the routine administration of antenatal corticosteroids to mothers at risk of preterm delivery between 24 and 34 wk gestation (1, 23a). Since the pioneering work of Liggins and Howie (18) in 1972, a multitude of well-designed studies have shown antenatal corticosteroid administration significantly reduces mortality, respiratory distress syndrome, and intraventricular hemorrhage in preterm infants (7, 18). In the 1990s, multiple courses of corticosteroids were used in an attempt to maximize fetal health (31). However, recent research has identified progressive growth restriction with repetitive doses (12, 16). Potential adverse effects from a single course of therapy during the late stages of development are now emerging (17, 22, 27).
In animal models, acute glucocorticoid exposure elevates the blood pressure of late-gestation fetuses, and ovine small femoral arteries have shown enhanced vasoconstriction to potassium chloride and endothelin-1 immediately after steroid administration (2, 9, 10). Paradoxically, steroid administration has also been associated with enhanced endothelium-dependent vasodilatation to acetylcholine (2). Finally, acute corticosteroid exposure has been associated with increased coronary artery relaxation to bradykinin and nitric oxide, potentially related to increased guanylate cyclase activity (13). The purpose of our study was to clarify whether antenatal glucocorticoid administration alters coronary and mesenteric artery vascular reactivity and to investigate possible mechanisms regulating these effects.
Tissue collection. All procedures were performed within the regulations of the Animal Welfare Act and the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the University of Iowa Animal Care and Use Committee. Time-dated pregnant ewes with twins at 116-119 days gestation (term being 145 days) were obtained from a local source (n = 5). Ewes were anesthetized with 12 mg/kg of thiopental sodium (Abbott Laboratories, Abbott Park, IL), intubated, and ventilated with a mixture of halothane (1%), oxygen (33%), and nitrous oxide (66%). After performing an abdominal flank incision, the uterus was partially externalized and opened over the fetal hindlimbs. Polyethylene catheters were inserted into a femoral artery and vein, and a catheter immersed in amniotic fluid was sutured to the fetal skin. The fetal incisions and uterine openings were closed, and the procedure was repeated for the second twin. All the catheters were exteriorized through a subcutaneous tunnel into a cloth pouch on the ewe's flank. Ampicillin (Sigma, St. Louis, MO) was administered at the completion of surgery (2 g intra-amniotic and 2 g intramuscular to the ewe), followed by intramuscular injections (1 g) to the ewe every 12 h for 3 days. After surgery, the ewe was returned to an individual pen and allowed free access to food and water.
After a 3-day recovery period, one fetus received betamethasone (10 μg/h by continuous intravenous infusion over 48 h; Schering, Kenilworth, NJ) while the control twin received an identical volume of vehicle (0.9% NaCl). The dose and route utilized corresponds to those used by other investigators to ensure consistent corticosteroid exposure (2), while the use of twin controls was chosen to minimize potentially confounding environmental conditions, such as maternal health and diet. The virtual absence of vascular anastomoses between twin placentas in sheep, coupled with the larger volume of distribution seen in the pregnant ewe compared with the exposed fetus, ensured selective administration to a single fetus (20
Isolated vessel contractile responses.2-5% CO2; the composition was as follows (in mM): 130 NaCl, 4.7 KCl, 1.18 KH2PO4, 1.17 MgSO4·7H2O, 14.9 NaHCO3, 1.6 CaCl2·H20, 5.5 dextrose and 0.03 CaN2-EDTA (pH 7.30). The measured osmolality was 293 mosmol/kgH2O.
Endothelium-intact isolated vessels from twin pairs were studied simultaneously. Initially, the contractile response of each vessel to 120 mM KCl was recorded. Subsequent coronary artery contractile responses were then normalized as a percentage of the maximal response. The bath was washed three times with PSS over a 10-min period, and the vessels were allowed to equilibrate again for 20 min before initiation of the dose-response protocols. Cumulative concentration-responses to KCl (5-90 mM), ANG II (10-11 to 10-7 M), and the sympathomimetic phenylephrine (10-9 to 10-4 M) were conducted with addition of increasing concentration of the agent under study at set intervals (10 min for coronary arteries and 5 min for mesenteric arteries). The arteries were then reequilibrated to their baseline with multiple washes of PSS over 1 h before preconstriction of each vessel with 10-6 M U-46619, a thromboxane A2 mimetic. Subsequent mesenteric artery contractile responses were then normalized as a percentage of the maximal response to U-46619, rather than KCl, given the significant difference in the contractile response to KCl between groups. Cumulative concentration responses to acetylcholine (10-10 to 10-5 M), the nitric oxide donor sodium nitroprusside (10-10 to 10-5 M), 8-BrcGMP (10-9 to 10-4 M), the β-agonist isoproterenol (10-10 to 10-6 M), and a direct activator of adenylate cyclase, forskolin (10-11 to 10-6 M), were conducted with addition of increasing concentration of the agent at 8-min intervals. 8-BrcGMP was not utilized with the mesenteric arteries. Microsoft Excel 2000 was used to generate smoothed dose-response curves for each vasoactive agent. All PSS reagents and vasoactive compounds were acquired from Sigma Chemical (St. Louis, MO) with the exception of U-46619, which was supplied by Alexis (San Diego, CA).
Immunoblotting. Western blot analysis for angiotensin type 1 (AT1) and angiotensin type 2 receptor (AT2) was performed as previously described, with extrapolation to endothelial nitric oxide synthase (eNOS) analysis (38). Protein concentrations were determined by the method of Lowry as modified by Peterson (30). Equal protein loading was verified by Ponceau S staining. The AT1 receptor-specific polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was raised in rabbits against an epitope corresponding to amino acids 306-359 of the human AT1 receptor. The AT2 receptor-specific polyclonal antibody (Santa Cruz Biotechnology) was raised in rabbits against an epitope corresponding to amino acids 221-363 mapping the carboxy terminus of the human AT2 receptor. The eNOS-specific monoclonal antibody (BD Biosciences Pharmingen, San Diego, CA) was raised in mice. Nitrocellulose blots (20 μg protein/lane) were incubated with the primary antibody at a 1:1,000 (AT1 and eNOS) or 1:2,000 (AT2) dilution for 2 h at room temperature. Blots were rinsed, washed, and then incubated with either a 1:3,000 dilution of goat anti-rabbit or a 1:2,000 dilution of goat anti-mouse horseradish peroxidase (HRP)-conjugated antibody (Sigma) at room temperature for 1 h. Binding of the secondary antibody was detected using a chemiluminescent system consisting of HRP/hydrogen peroxide oxidation of luminol (Pierce, Rockford, IL). Blots were then exposed to Kodak XAR X-ray film for 1 min. Films were digitized, and the difference between protein signals and background was quantitated using NIH Image (National Institutes of Health, http://rsb.info.nih.gov/nih-image/).
Immunohistochemistry. Isolated vessels were fixed in formalin (10%) and embedded in paraffin, and sections were mounted on glass slides. Sections were deparaffinized in xylenes and hydrated in an ethanol:physiological salt solution (PSS) series. After a 5-min PSS rinse, the sections were incubated in H2O2 (3% in methanol) for 30 min before rinsing with PSS (×2) and blocking with BSA (1% in PSS). Sections were incubated with primary antibody, either rabbit anti-AT1, 1:100 dilution or goat anti-AT2, 1:100 dilution (both from Santa Cruz Biotechnology) at room temperature for 60 min. The sections were then rinsed two times for 10 min per rinse in PSS and stained using a Vectastain Elite kit (Vector Labs, Burlingame, CA). The tissue sections were incubated for 30 min in HRP-conjugated secondary antibody at 1:200 dilution (goat anti-rabbit or donkey anti-goat antibody, as appropriate, both from Santa Cruz Biotechnology), rinsed two times in PSS, and then incubated for 45 min in avidin-peroxidase reagent. After rinsing two times in PSS, the sections were incubated in diaminobenzidine (Sigma Chemical) for 3-5 min. Sections were rinsed in PSS twice for 5 min, rinsed quickly in distilled water, counterstained with 50% hematoxylin for 2 min, dehydrated in an ethanol series, then rinsed with xylenes and mounted with glass coverslips. Incubation with secondary antibody alone was performed for each vessel type to serve as controls. Sections were then viewed under a brightfield microscope (Nikon Optiphot-2) and imaged using a digital camera and Spot software (Diagnostic Instruments, Sterling Heights, MI).
Data analysis. Comparison of physiological parameters was made using two-way ANOVA, factoring for treatment group and timing in relation to infusion. If the overall analysis of variance identified significant differences (P < 0.05), pairwise comparisons were made using Tukey's procedure, with P < 0.05 considered significant. Vascular responses and protein expression were compared using Student's unpaired, two-tailed t-test (with significance at P < 0.05). All analyses were performed using SigmaStat 3.0 for Windows (SPSS, Chicago, IL). All values are presented as means ± SE, and n refers to the number of animals studied.
Physiological parameters. Fetal mean arterial blood pressure was similar between the groups before betamethasone administration (Table 1). The mean arterial blood pressure was higher in the betamethasone-treated than the saline-treated fetuses after the 48-h infusion (53 ± 1 vs. 48 ± 1 mmHg, P < 0.05). There were no significant differences in weight, heart rate, hematocrit, arterial blood gas values, or electrolytes between or within either group.
Vascular reactivity. The maximal vasoconstrictive responses of coronary artery segments to 120 mM KCl and 10-6 M U-46619 were not significantly altered by betamethasone infusion, while the maximal responses of mesenteric artery segments to 120 mM KCl was significantly decreased after betamethasone infusion (Table 2). The dose-response curves for vasoconstriction to increasing concentrations of KCl were similar in betamethasone-infused and saline-infused coronary arteries (Fig. 1A). Although the steroid-exposed coronary arteries achieved significantly more vasodilatation to low concentrations of acetylcholine (P < 0.05 vs. control), the vasoconstriction seen at higher concentrations was not significantly different between groups (Fig. 1B). The maximal contractile response to ANG II was significantly increased in the coronary arteries of the betamethasone-infused vs. the saline-infused animals (72 ± 17 vs. 23 ± 6% of the response to 120 mM KCl, P < 0.05). However, contractile responses to ANG II were similar in betamethasone and control mesenteric arteries (48 ± 17 vs. 36 ± 12% of the maximal response to 10-6 M U-46619). Marked tachyphylaxis to increasing concentrations of ANG II was noted in both coronary and mesenteric arteries (Fig. 1, C and D, respectively).
Fetal mesenteric arteries from the betamethasone-exposed group were significantly less responsive to graded concentrations of KCl than their saline-exposed counterparts (Fig. 2A, P < 0.05). There were no statistically significant intergroup differences in vasoconstriction to phenylephrine between the mesenteric arteries (Fig. 2B). Responses in the betamethasone-exposed group tended to be reduced, but these responses in general were highly variable. Coronary arteries were completely unresponsive to phenylephrine at concentrations up to 10-4 M (data not shown).
No between-group differences were detected in the coronary vasodilatory responses elicited by sodium nitroprusside, 8-BrcGMP, isoproterenol, or forskolin (Fig. 3). The mesenteric artery vasodilatory response to acetylcholine was enhanced after betamethasone administration (Fig. 4A, P < 0.05 vs. control), while the response to sodium nitroprusside was attenuated in the steroid-treated group (Fig. 4B, P < 0.05 vs. control). There were no significant differences in mesenteric artery response to isoproterenol or forskolin (Fig. 4, C and D, respectively).
Immunoblotting. Western blotting consistently demonstrated the presence of AT1 (Fig. 5A) and AT2 (Fig. 6A) receptor protein in both mesenteric and coronary arteries. The immunoblots probed with the AT1 receptor antibody showed the major band at 67 kDa, as previously reported by Marrero et al. (19), while the AT2 receptor antibody showed the major band at ∼68 kDa, as previously reported by Servant et al. (39). AT1 receptor protein expression was upregulated by betamethasone infusion in coronary arteries (P < 0.05 vs. control), but not in mesenteric vessels (Fig. 5B). There was no difference in AT2 receptor protein expression after betamethasone infusion in either the coronary or mesenteric arteries (Fig. 6B). AT1 protein expression was greater in coronary arteries than mesenteric arteries, in contrast to AT2 protein expression, which was greater in the mesenteric arteries (Figs. 5B and 6B).
The immunoblots probed with the eNOS antibody showed the major band at 140 kDa (Fig. 7). The increase in eNOS expression after betamethasone infusion did not reach statistical significance in either the coronary or mesenteric arteries (P = 0.12 and P = 0.14, respectively).
Immunohistochemistry. Immunostaining localized both AT1 and AT2 receptors within the tunica media of each isolated vessel with minimal endothelial staining (Fig. 8).
Emerging information regarding the potential side effects of repeated doses of antenatal corticosteroids has tempered the unbridled use of betamethasone and dexamethasone to promote fetal maturation. In addition to detrimental effects on fetal growth and risk of infection after repetitive administration, the potential exists for cardiovascular maladaptive responses (3, 43). We have shown that direct parenteral administration of betamethasone enhances the coronary, but not mesentery, vasoconstrictor response to ANG II. In addition, betamethasone exposure led to an increase in the expression of ANG II receptors, specifically within the coronary arteries.
The direct effects of ANG II are mediated by two distinct receptors, classified as type 1 (AT1) and type 2 (AT2) based on selective antagonism by peptide and nonpeptidic ligands (5, 41). The tissue distribution and expression of these receptors are developmentally regulated. In general, AT2 receptor expression is high in fetal tissues and decreases with postnatal maturation, while AT1 receptors appear later in fetal life with expression greatest in tissues regulating cardiovascular and fluid and electrolyte homeostasis (4, 6, 33, 40). The functions of the AT2 receptor are unclear, although it may exert proapoptotic and vasodepressor effects (15, 46). The functional role of AT1 receptors in the fetal vasculature has also been incompletely elucidated, although within the cardiovascular system, AT1 receptor expression is responsive to increased glucocorticoid exposure (23, 25, 36). Furthermore, tissue-specific, local renin-angiotensin systems (RAS) are active in the fetus and are important modulators of circulatory function (8, 35), whereas later in life, vascular RAS have been shown to play a role in each step down the pathway of coronary artery disease, from endothelial dysfunction to lipid deposition, inflammation, vascular remodeling, apoptosis, and thrombosis (24, 34).
In the present study, both AT1 and AT2 receptors were present in the coronary and mesenteric arteries. Immunohistochemistry verified the presence of AT1 and AT2 receptor protein in both arteries and localized them predominately within the tunica media. The AT1 receptors identified were functionally active, as evidenced by arterial contractile responses to ANG II. The detection of functional AT1 receptors in the fetal coronary arteries complements our previous studies demonstrating AT1 mRNA and protein are present in isolated fetal ovine renal and mesenteric arteries and that the receptor mediates contraction in response to ANG II (35). The progressive vasoconstriction each vessel displayed to increasing concentrations of ANG II reached a plateau at 10-8 M ANG II. Further increases in ANG II concentration resulted in an exponential decline in vessel tone as strong tachyphylaxis evolved. While displaying the same dose-response pattern, the coronary arteries that were exposed to betamethasone achieved peak responses twice that of their twin controls. This difference may be related to the increased AT1 receptor density as demonstrated by immunoblotting. The exaggerated coronary artery vasoconstriction to ANG II could predispose toward myocardial ischemia during high renin states or episodes of enhanced myocardial oxygen consumption. This finding may be of particular importance if coronary AT1 receptor expression remains permanently elevated after antenatal exposure to glucocorticoids.
The vascular responses to potassium chloride-induced calcium channel activation were similar between treatment groups in the coronary arteries and diminished after betamethasone administration in the mesenteric arteries. Voltage-sensitive calcium influx therefore appears to be unrelated to the association between betamethasone administration and vascular hyperreactivity in the coronary arteries. Likewise, the increase in intracellular calcium release that would have followed phospholipase C activation by U-46619, a thromboxane A2 mimetic, resulted in similar coronary artery contractile responses. These findings supplement the results seen by Anwar et al. (2) in femoral arteries after betamethasone infusion. In their study, betamethasone increased the response of small femoral arterial branches to potassium chloride, but no differences in the vascular response to U-46619 were seen. Given the differential roles played by skeletal muscle resistance vessels and visceral conductance vessels, such as the coronary arteries, it is not surprising that they display differential responses to depolarizing potassium chloride concentrations.
While the betamethasone-exposed coronary arteries initially dilated after acetylcholine activation of eNOS, activation of the muscarinic receptors present within the vascular smooth muscle presumably resulted in overriding vasoconstriction, again involving the activation of phospholipase C. The propensity of coronary arteries to constrict to high concentrations of acetylcholine has been described (45), and the propensity of vessels exposed to betamethasone to constrict more intensely parallels the results seen by Docherty et al. (9) in cerebral arteries. These differences may be a subtle sign of early endothelial dysfunction, resulting in overriding vasoconstriction. The ability of ANG II to effect such dysfunction through activation of plasma membrane-bound NADPH oxidase and subsequent production of reactive oxygen species deserves further consideration (14, 46).
Reminiscent of the coronary artery dilatory response to acetylcholine seen only in the steroid-exposed group, the betamethasone-exposed mesenteric arteries displayed significantly enhanced vasodilation to acetylcholine. These results mirror those seen by both Anwar et al. (2) and Molnar et al. (21) in ovine femoral arteries after steroid exposure. In the face of accentuated endothelium-dependent vasodilation to acetylcholine after betamethasone exposure, the attenuated response of the same mesenteric arteries to a nitric oxide donor, sodium nitroprusside, raises the possibility that eNOS activity was enhanced after steroid administration. Immunoblots subsequently showed a consistent trend toward increased eNOS expression after steroid exposure in both the mesenteric and coronary arteries, although the results did not reach statistical significance. This finding is intriguing, given the well-described protective effect of antenatal corticosteroids on the development of necrotizing enterocolitis, a condition often associated with mesenteric ischemia (7). Highlighting the vessel-specific alterations often seen in vascular research, eNOS expression has been shown to be downregulated in rat aorta and unchanged in ovine femoral arteries after dexamethasone administration (21, 44).
Unlike the results reported by Anwar el al. (2) showing attenuated femoral artery vasodilation to forskolin after exposure to betamethasone, we found no difference in coronary or mesenteric artery responsiveness to adenylate cyclase activation. Isoproterenol, a β-agonist that mediates vasodilatation through G protein-coupled activation of adenylate cyclase, and forskolin, a direct activator of adenylate cyclase, produced similar dose-response relationships. These apparently contradictory results again stress the importance of vessel-specific research and highlight the possibility that the pathway connecting corticosteroids and hypertension may be different from that linking antenatal corticosteroid exposure and coronary artery disease. In fact, our own group and others have shown the postnatal increase in mean arterial blood pressure seen with antenatal glucocorticoid treatment is not abolished by AT1 receptor antagonism, indicating the changes are mediated through mechanisms beyond peripherally accessible AT1 receptors (29, 37).
There is increasing interest regarding the nonpulmonary effects of glucocorticoid exposure to the fetus. Maternal glucocorticoid administration lessens the incidence of complications associated with prematurity. However, a number of studies have indicated prenatal glucocorticoid exposure may lead to permanent effects on cardiovascular homeostasis (11, 28). In the present study, the amount of betamethasone the fetuses received (0.1 mg·kg-1·day-1 based on an average fetal weight of 2.4 kg) approximates the equivalent amount of cortisol required at times of stress (∼5 times the physiological requirements) (42). Furthermore, this betamethasone dose approximates the fetal glucocorticoid exposure after maternal betamethasone administration to promote fetal maturation before preterm delivery, as well as the fetal exposure that results from the normal rise in serum cortisol before parturition at the end of term gestation (26). Thus our findings provide novel information regarding the mechanism of tissue-specific glucocorticoid-induced coronary artery dysfunction after a clinically and physiologically relevant dose of betamethasone. Functionally active ANG II receptors mediating vascular tone are clearly present in the fetal cardiovascular system. If the heightened response of the coronary arteries to ANG II after betamethasone exposure represents a permanently programmed phenotype, rather than simply premature physiological maturation, the exaggerated angiotensin responsiveness may provide a link between antenatal glucocorticoid exposure and cardiovascular morbidity. Continued investigation into the effects of corticosteroids on postnatal cardiovascular health is essential if we are to better understand the long-term consequences of antenatal glucocorticoid treatment.
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 © 2004 the American Physiological Society