Additional Hypotensive Effect of Endothelin-1 Receptor Antagonism in Hypertensive Dogs Under Angiotensin-Converting Enzyme Inhibition
Background Endothelin-1 (ET-1) may play a role in hypertension. ET-1 receptor antagonism by bosentan lowers blood pressure in hypertension. We evaluated whether the effect of bosentan is still observed under ACE inhibitors (ACEI).
Methods and Results Thirty anesthetized and 18 conscious hypertensive dogs were studied randomly. Anesthetized dogs were divided into 4 groups: group 1 received cumulative doses of bosentan (bolus+30-minute infusion: 0.1 mg/kg+0.23 mg/kg per hour to 3 mg/kg+7 mg/kg per hour); group 2, the same dose-responses after 1 mg/kg enalaprilat; group 3, the vehicle after enalaprilat; and group 4, the dose responses to bosentan followed by enalaprilat. The conscious dogs were divided into 3 groups: group 5 received 2 cumulative doses of bosentan; group 6, the vehicle; and group 7, enalaprilat alone. In groups 1 and 2, bosentan produced dose-related decreases (P=.0001) in left ventricular systolic pressure and mean aortic pressure (AOP). In group 1, bosentan decreased mean AOP by 22%. In group 2, enalaprilat decreased mean AOP by 25% (from 173±26 to 130±25 mm Hg; P<.005); an additional 18% decrease was obtained with bosentan, the mean AOP reaching 98±21 mm Hg (P<.01). In group 3, the effect of enalaprilat alone was a 22% decrease in mean AOP (P<.005). The additive effect of the bosentan-ACEI association was also observed in group 4. In group 5, bosentan reduced mean AOP by 20% (P<.005), whereas mean AOP remained unchanged in group 6. The effect of ACEI alone (group 7) was similar to that of bosentan.
Conclusions Bosentan produces an additional hypotensive effect to that of ACEI, which opens new therapeutic perspectives.
Endothelin-1 (ET-1), a 21–amino acid peptide derived from the vascular endothelium, belongs to a family of peptides endowed with potent vasoconstrictive properties,1 2 which in addition are able to induce cell proliferation and hypertrophy.3 4 In physiological conditions, it is now clear that ET-1 participates in the regulation of vascular tone.5 On the basis of elevated plasma concentrations, ET-1 has also been implicated in various pathological disorders. Recently, the development of endothelin receptor antagonists has shown that ET-1 plays a pathophysiological role because it was demonstrated in ischemic renal failure,6 subarachnoid hemorrhage,6 myocardial infarction,7 and heart failure.8
Given the vasoconstrictive and mitogenic properties of ET-1, a potential role of ET-1 in the pathogenesis of hypertension has raised much interest. However, this role has also been subject to debate because plasma ET-1 concentrations were found to be normal or slightly elevated9 10 in hypertension and the sensitivity to ET-1 reported as increased or decreased.11 12 In the model of the deoxycorticosterone acetate (DOCA)-salt hypertensive rat, authors showed that the combined ETA/ETB ET-1 receptor antagonist bosentan caused a blunting of the rise in blood pressure and abrogated the severe vascular hypertrophy found in this model.13 However, in a model of the spontaneously hypertensive rat (SHR), bosentan was ineffective.14 We demonstrated in the model of renal hypertension (“Page kidney”) in anesthetized dogs that bosentan markedly lowered blood pressure but was less effective on normal blood pressure.15 In this model, angiotensin II was a good candidate for stimulating vascular ET-1 production. We therefore now raise the question of whether this effect of bosentan is still observed in the presence of ACE inhibitors (ACEI). If so, an additional effect of bosentan to ACEI would offer new therapeutic perspectives.
Dose responses to bosentan in the absence or presence of ACEI were performed and compared with ACEI alone in anesthetized hypertensive dogs that had been acutely instrumented for cardiovascular and hormonal assessment. Experiments in which bosentan was first administered and then ACEI was used were also included. Effects of bosentan and ACEI alone were evaluated in two groups of conscious hypertensive dogs.
Forty-eight mongrel hypertensive dogs weighing 17 to 41 kg were studied. Arterial hypertension was produced in all dogs by wrapping the left kidney with silk tissue. This model, a variant of cellophane wrapping initially described by Page,16 is characterized by a persistent hypertension associated with perinephritis. After general anesthesia (induced by 7 mg/kg sodium thiopental IV), intubation, and ventilation, a lumbar incision under sterile conditions was performed. The left kidney was exposed and wrapped tightly with sterile silk tissue. The animals were allowed to recover and were studied after 6 to 8 weeks. At the time of the study, they were afebrile and healthy. Most of the animals (n=30) were studied under anesthesia; a smaller group (n=18) was studied in the conscious state.
The anesthetized dogs comprised 4 groups: 8 dogs received the endothelin receptor antagonist bosentan after a bolus injection of placebo (group 1); 8 dogs received bosentan after a bolus injection of enalaprilat (group 2); 8 dogs received the bosentan vehicle solution after a bolus injection of enalaprilat (group 3); and 6 dogs received bosentan followed by a bolus injection of enalaprilat (group 4). These four groups were studied under anesthesia (20 mg/kg sodium pentobarbital IV), intubated, and ventilated. A fluid-filled catheter was introduced through the left femoral artery into the descending aorta for blood pressure measurement and blood sampling. A flow-directed, balloon-tipped 7F thermodilution catheter (Edwards Swan-Ganz) was advanced from the left femoral vein through the right cardiac chamber and positioned in the pulmonary artery for measurement of right atrial pressure (RAP), drug infusion, and determination of cardiac output. The animals then underwent a left thoracotomy. A catheter was implanted in the left atrial appendage for measurement of left atrial pressure (LAP). A micromanometer (JSI-400, Gifila Scientific Instruments) was inserted into the left ventricle through a stab incision in the apex for measurements of left ventricular systolic pressure (LVSP) and end-diastolic pressure (LVEDP). The left anterior descending coronary artery (LAD) was dissected free near its origin and fitted with a Doppler flow probe (Triton Technology) to measure coronary flow.
The conscious dogs comprised 3 groups: 7 dogs received the bosentan (group 5), 4 dogs received the bosentan vehicle solution (group 6), and 7 dogs received enalaprilat (group 7). Under local anesthesia with lidocaine (2%), a 7F microtip Millar catheter was introduced through the left femoral artery into the left ventricle; a fluid-filled catheter was placed in the left femoral artery for arterial pressure measurement and blood sampling.
Experimental protocol for groups 1, 2, and 3 is shown in Fig 1⇓. After completion of the surgical preparation, the animals were allowed to stabilize for 20 minutes. Bosentan (synthesized at F. Hoffman-La Roche Ltd) was dissolved in water. Four cumulative doses of bosentan (dose 1, bolus 0.1 mg/kg+30-minute infusion at 0.23 mg/kg per hour; dose 2, bolus 0.3 mg/kg+0.7 mg/kg per hour; dose 3, 1 mg/kg+2.33 mg/kg per hour; dose 4, 3 mg/kg+7 mg/kg per hour) were infused 30 minutes after placebo (group 1) or 1 mg/kg enalaprilat injection (ACEI) (group 2). This dose of enalaprilat was chosen to obtain a maximal inhibition of ACE and was based on pilot experiments. It is also known that maximal blockade of angiotensin I pressor response can be achieved by 0.25 mg/kg enalaprilat.17 In group 3, only the bosentan vehicle solution was given after 1 mg/kg enalaprilat injection. Hemodynamic parameters were recorded in the basal state and then every 10 minutes throughout the experiment. Cardiac output was determined in the basal state and at the end of each infusion period. Arterial blood was withdrawn after the 20-minute stabilization period and at the end of each infusion period to determine plasma bosentan concentrations. Blood was also obtained at the beginning and at the end of experiment for hormonal measurements. The experimental protocol in group 4 was performed as above but differed by the inversion of the sequence ACEI-bosentan: the four cumulative doses of bosentan were given first and followed by a bolus injection of 1 mg/kg enalaprilat. Hemodynamic parameters were recorded throughout the experiments and 30 minutes after enalaprilat injection.
Experimental protocol for groups 5 and 6 is shown in Fig 2⇓. After introduction of the catheters, the animals were allowed to stabilize for 20 minutes. Two cumulative doses of bosentan (dose 2, bolus 0.3 mg/kg+30-minute infusion at 0.7 mg/kg per hour; dose 4, bolus 3.0 mg/kg+30-minute infusion at 7 mg/kg per hour) were infused in group 5; group 6 received the bosentan vehicle solution only. Group 7 received a bolus injection of 1 mg/kg enalaprilat alone.
Data Analysis, Hormonal Measurements, and Plasma Bosentan Determinations
Cardiovascular data were measured and analyzed as previously described.18 19 Cardiac output was determined as an average of three measurements by the thermodilution technique with a cardiac output computer (model 9520A Edwards Lab). Total peripheral resistance was reported in peripheral resistance units and calculated as [mean aortic pressure (AOP) (mm Hg)−mean RAP (mm Hg)]/cardiac output (mL/min). Coronary vascular resistance was calculated by the formula [mean AOP (mm Hg)−mean RAP (mm Hg)]/LAD coronary flow (mL/min).
The method for ET-1 measurements has been previously described by our group.2 20 Briefly, ET-1 was measured after plasma extraction on Sep-Pack C18 Cartridges (Waters Associates) by a radioimmunoassay with specific antibodies and synthetic peptides from Peninsula. A possible interference of the infused bosentan solution (at a concentration of 5 mg/mL) in the radioimmunoassay for endothelin was tested and ruled out.
For quantification of bosentan in plasma, a combination of liquid extraction and radioligand competition binding was applied. For extraction, 1 mL methanol was added to 50 μL plasma or cerebrospinal fluid. After rigorous mixing, samples were centrifuged (5 minutes, 3000g) to remove precipitated protein. After evaporation of the methanol phase, samples were redissolved in 50 mmol/L Tris buffer (pH 7.4, 25 mmol/L MnCl2, 1 mmol/L EDTA, 0.5% BSA). Preparation of microsomal membranes and competition binding assays, with the use of [125I] ET-1 and recombinant human ETA receptor expressed in the baculovirus-infected insect cells, were performed as previously described.21 Concentrations of bosentan were computed from calibration curves obtained with spiked plasma. Logit-log of specific binding was plotted against log of bosentan concentrations. The lower sensitivity limit of the assay was 25 ng/mL. Concentrations above the upper limit (2 μg/mL) were measured after appropriate dilutions with Tris buffer. All measurements were performed repeatedly as triplicate determinations.
Data were analyzed by two-way ANOVA for repeated measurements. Differences between treated and control groups (the grouping factor) were assessed with F tests and differences between the levels of the trial factor (the within factor) were assessed with conservative Greenhouse-Geisser tests. A detailed contrast analysis is provided in figures. A probability value of <.05 was considered significant. Computations were performed with SAS statistical software. Data are expressed as mean±SD.
Cardiovascular Effects of Bosentan in Group 1 Compared With Group 2
In groups 1 and 2, administration of bosentan resulted in pronounced effects (P=.0001) on ventricular and aortic pressures, which were clearly dose related. In group 1 (receiving bosentan only), a decrease in left ventricular systolic pressure (LVSP) occurred at the second dose (P<.05) and became maximum at the end of the last infusion dose (P<.005) (Fig 3⇓). The decrease in mean AOP was significant at the third dose (P<.01) and maximum at the last dose (P<.005) (Fig 4⇓). For instance, at that time, mean AOP had decreased from 176±24 to 138±13 mm Hg, which corresponded to a 22% reduction. This was due to a decline in systemic vascular resistance (SVR, from 667±260 to 522±272 mm Hg/mL/min · 10−4; P<.005). Similarly, coronary vascular resistance (CVR) decreased slightly (from 5.8±2.7 to 5.1±2.3 mm Hg/mL per minute, P<.05). Although right and left atrial pressures did not change, a slight 2 mm Hg decrease (P<.01) in left ventricular end-diastolic pressure (LVEDP) was found. Heart rate did not change over time.
In group 2, enalaprilat decreased mean AOP by 25% (from 173±26 to 130±25 mm Hg, P<.005) (Fig 4⇑). An additional 18% decrease was obtained with the maximal dose of bosentan, the mean AOP reaching 98±21 mm Hg (P<.01 versus the level achieved after enalaprilat). The decrease in LVSP (P<.005) was similar with that of mean AOP. At the maximal dose of bosentan, SVR had dropped (P<.05) from 825±329 to 569±342 mm Hg/mL/min·10−4. CVR also declined significantly (from 6.3±2.7 to 3.3±0.8 mm Hg · mL per minute, P<.01). Filling pressures (as reflected by LVEDP, LAP, and RAP) tended to decrease (LVEDP from 6.9±5.5 to 4.4±3.1 mm Hg, NS; LAP from 5.9±3.6 to 4.6±2.7 mm Hg, P<.05; and RAP from 5.8±2.0 to 4.8±0.9 mm Hg, NS). Heart rate increased over time (from 155±2.0 to 167±22 bpm, P<.05).
When groups 1 and 2 were compared, it became clear that the decreases in LVSP and mean AOP were significantly more pronounced in group 2, reflecting an additional effect of bosentan to that of enalaprilat (interaction time×group on LVSP, P=.0008, and on mean AOP, P=.0011).
Cardiovascular Effects of Bosentan in Group 3 Compared With Group 2
In group 3, enalaprilat after 30 minutes produced a 22% decrease in mean AOP (from 163±18 to 128±17 mm Hg; P<.005) and an 18% decrease in LVSP (from 190±15 to 156±13 mm Hg; P<.005) (Figs 3⇑ and 4⇑). No further decrease was observed thereafter. Concomitantly, SVR and CVR declined from 552±170 to 396±166 mm Hg/mL/min·10−4 (P<.005) (Fig 4⇑) and from 5.8±5.9 to 5.1±5.2 mm Hg/mL/min·10−4 (P<.05), respectively. At the end of the experiment, LVEDP tended to decrease (from 7.6±2.4 to 6.0±2.0 mm Hg, NS), whereas LAP and RAP remained at the same level. Heart rate did not change significantly throughout the study. Comparison between groups 2 and 3 revealed a greater decrease in LVSP and mean AOP in group 2 (interaction time×group on LVSP, P=.0004, and on mean AOP, P=.0002).
Cardiovascular Effects of Enalaprilat After Bosentan in Group 4
The dose response to bosentan was characterized by progressive and dose-related decreases in mean AOP that were similar to those observed in group 1. From the beginning to the end of the last dose of bosentan, mean AOP decreased from 151±21 to 112±16 mm Hg (P<.001). LVSP also displayed a decrease from 186±23 to 141±14 mm Hg (P<.005). Thirty minutes after injection of enalaprilat, mean AOP and LVSP reached 74±29 and 101±26 mm Hg, respectively (P<.005 versus last dose of bosentan). Heart rate did not change significantly over time.
Cardiovascular Effects of Bosentan in Group 5 Compared With Group 6
In conscious dogs (group 5), bosentan reduced mean AOP by 20% (from 144±19 to 118±13 mm Hg; P<.005), whereas mean AOP remained unchanged in the time-control group (group 6) (154±17 versus 157±12 mm Hg). Similarly, LVSP decreased in group 5 (from 180±17 to 146±15 mm Hg, P<.001) but not in group 6 (183±13 versus 187±14 mm Hg). The response in group 5 correlated with the dose of bosentan. Systolic and diastolic aortic pressures in groups 5 and 6 are illustrated in Fig 5⇓. No effect on heart rate was observed in either group.
Cardiovascular Effects of Enalaprilat Alone in Group 7
In this group of conscious dogs, enalaprilat administration was associated with a 23% decrease in mean AOP (from 137±19 to 103±16 mm Hg, P<.005), which was present in all dogs and not significantly different from that after bosentan (group 5). Consistently, LVSP also decreased from 174±21 to 141±22 mm Hg (P<.005).
Bosentan Levels and Hormonal Effects
As shown in Fig 6⇓, bosentan levels increased progressively in relation to the dose infused (in groups 1 and 2). Similar plasma concentrations were achieved in both groups.
Fig 7⇓ illustrates plasma ET-1 concentrations in the first 3 groups of anesthetized dogs. Basal plasma ET-1 concentrations were detectable in all dogs. They increased approximately 12-fold in group 1 and 16-fold in group 2 (P<.0001), whereas a moderate increase was seen in group 3 (P<.05). It is noticeable that ET-1 levels achieved in group 2 were significantly greater than those in group 1 (P<.05). In conscious dogs (group 5), plasma ET-1 also rose (P<.0001) with bosentan but remained unchanged in the time-control group (group 6) (Fig 7⇓).
The present study demonstrates that ET-1 receptor antagonism with bosentan lowers aortic and ventricular systolic pressures in perinephritic hypertension. In addition to our previous results,15 a dose response to bosentan has now been established and its hypotensive effects clearly related to vasodilation, as reflected by a decrease in systemic and coronary vascular resistances. These findings are in agreement with a recent study showing that bosentan infused in patients with congestive heart failure decreased blood pressure and peripheral resistances, indicating a reduction in afterload.8 In that study, decreases in right atrial and pulmonary pressures also suggested venous dilation. In our experimental model, as LVEDP decreases there is a possible reduction in left ventricular preload, which may not be demonstrative as much because of the initial low level of these pressures. We previously15 raised the question of a possible influence of anesthesia, which might trigger neurohormonal activation including ET-1 and explain partially the effect of bosentan. This factor does not appear to play a major role because bosentan is also effective in conscious dogs, with the decrease in mean AOP averaging 22% and 20% in anesthetized and conscious dogs, respectively. Interestingly, the effect of bosentan in conscious dogs is similar to that of ACEI.
The major finding of the present study is the demonstration of an additional hypotensive effect of bosentan to that of ACEI in the hypertensive dog. ACEI, however, leads to a shift of the dose-response curve to bosentan because the effect of the latter appears to be significant at the second dose when infused alone but at the last dose in the presence of ACEI. This is not so unexpected inasmuch as larger doses of bosentan are required to obtain a further decrease in blood pressure when the latter was already reduced by ACEI. Again, the hypotensive effect of bosentan can be attributed to a decrease in peripheral resistance, namely in afterload. However, an effect on preload cannot be established because the decrease in LVEDP could be due to ACEI. The effect of bosentan does not reflect a delayed action of ACEI. Indeed, in the group receiving ACEI alone, the maximal decrease in blood pressure was observed after 30 minutes without any further decrease afterward. The infused doses of bosentan were not different between the groups receiving bosentan alone and bosentan after ACEI, as confirmed by similar plasma bosentan concentrations in both groups. It is noteworthy that ACEI still produced an additional effect when bosentan was administered first.
Several mechanisms explaining this additive effect can be postulated. Inhibition of plasma or tissue ACE might be incomplete, so that angiotensin II generation still might stimulate ET-1 production as previously reported22 23 and explain the effect of bosentan. Other enzymatic systems, unaffected by ACEI,24 also may contribute to angiotensin II production. An alternative and tempting hypothesis is that ET-1 synthesis is enhanced secondary to other mechanisms than angiotensin II such as catecholamines, cytokines, tissue hypoxia, or shear stress (caused by remodeling of arteries).25 In any case, the additional effect of bosentan to that of ACEI reinforces the potential role of ET-1 in the pathophysiology of hypertension. This role has remained controversial so far, depending on the hypertensive model. For instance, bosentan is able to prevent the development of hypertension and the remodeling of arteries in DOCA-salt hypertensive rats13 but not in SHR rats.14 It has been hypothesized that endothelin antagonism is effective only in hypertensive models in which vascular overexpression of ET-1 is found (such as DOCA-salt hypertensive rats13 and DOCA-salt SHR26 ). Expression of ET-1 thus certainly deserves to be studied in our model.
The last important finding of our study is the evidence that plasma ET-1 concentrations are raised with bosentan but to a much greater extent if animals are pretreated with ACEI. The increase of plasma ET-1 with bosentan can be accounted for by a displacement of ET-1 from its receptors or a reduced clearance of ET-1 due to ETB receptor antagonism, as previously suggested.8 27 Increased ET-1 synthesis or conversion from big ET-1 does not appear to be involved because plasma big ET-1 was reported to remain unchanged under bosentan.8 To explain a greater increase of plasma ET-1 with bosentan under ACEI, we would favor the hypothesis of baroreflex stimulation in response to a marked hypotension. Indeed, upright tilting has been shown to increase plasma ET-1, and this finding has been attributed to a release of ET-1 from the neurohypophysis, mediated by the baroreceptor reflex.28 In our experimental setup, despite anesthesia, baroreflex stimulation is clear because heart rate increased. ACEI itself does not appear to play a role because plasma ET-1 increased just slightly with ACEI alone, probably as a result of surgery, as we previously observed in an open chest time-control group.2
The present study shows an additional hypotensive effect of ET-1 receptor antagonism to that of ACEI in hypertensive dogs. This effect, which is really additive rather than synergic, suggests different mechanisms of action and raises new hypotheses in the pathophysiology of hypertension. Owing to an effect on afterload, the bosentan-ACEI association also opens new therapeutic perspectives in the treatment of hypertension. Long-term studies are needed to confirm if this effect persists chronically and is able to prevent cardiac hypertrophy and remodeling of arteries.
This work was supported by grant 3.4572.96 from the Fonds National de la Recherche Scientifique and by a grant from the Fonds de Développement Scientifique 1995 (UCL), Brussels, Belgium. We thank M. Durant for her expert secretarial help and F. Vanlinden for his technical assistance in the radioimmunoassay laboratory.
- Received December 18, 1996.
- Revision received February 24, 1997.
- Accepted February 28, 1997.
- Copyright © 1997 by American Heart Association
Donckier JE, Hanet C, Berbinschi A, Galanti L, Robert A, Van Mechelen H, Pouleur H, Ketelslegers J. Cardiovascular and endocrine effects of endothelin-1 at pathophysiological and pharmacological plasma concentrations in conscious dogs. Circulation. 1991;84:2476-2484.
Bobik A, Grooms A, Millar JA, Mitchell A, Grinpukel S. Growth factor activity of endothelin on vascular smooth muscle. Am J Physiol. 1990;258:C408-C415.
Grover G, Dzwonczyk S, Parham CS. The endothelin-1 receptor antagonist BQ-123 reduced infarct size in a canine model of coronary occlusion and reperfusion. Cardiovasc Res. 1993;27:1613-1618.
Schiffrin EL, Deng LY, Larochelle P. Blunted effects of endothelin upon small subcutaneous resistance arteries of mild hypertensive patients. J Hypertens. 1992;10:434-444.
Li JS, Larivière R, Schiffrin E. Effect of a nonselective endothelin antagonist on vascular remodeling in DOCA-salt hypertensive rats. Hypertension. 1994;24:183-188.
Donckier JE, Stoleru L, Hayashida W, Van Mechelen H, Selvais P, Galanti L, Clozel J, Ketelslegers J, Pouleur H. Role of endogenous endothelin-1 in experimental renal hypertension in dogs. Circulation. 1995;92:106-113.
Sansdrap J, Van Mechelen H, Pouleur H, Charlier AA. CROPA: a research-oriented program for automatic processing of cardiovascular signals. In: Computers in Cardiology. Florence, Italy: IEEE Computer Society; 1981:433-436.
Donckier JE, Hanet C, Galanti L, Stoleru L, Van Mechelen H, Robert A, Ketelslegers J, Pouleur H. Low-dose endothelin-1 potentiates volume-induced secretion of atrial natriuretic factor. Am J Physiol. 1992;263:H939-H944.
Berbinschi A, Ketelslegers JM. Endothelin in urine. Lancet. 1989;1:909. Letter.
Kohno M, Yasunari K, Yokokawa K, Murakawa K, Horio T, Takeda T. Inhibition by atrial and brain natriuretic peptides of endothelin-1 secretion after stimulation with angiotensin II and thrombin of cultured hormone endothelial cells. J Clin Invest. 1991;87:1999-2004.
Urata H, Healy B, Stewart RW, Bumpus M, Husain A. Angiotensin II forming pathways in normal and failing human hearts. Circ Res. 1990;66:883-890.
Schiffrin EL, Larivière R, Li J, Sventek P, Touyz R. Deoxycorticosterone acetate plus salt induce overexpression of vascular endothelin-1 and severe vascular hypertrophy in spontaneously hypertensive rats. Hypertension. 1995;25:769-773.