Author + information
- Received January 2, 2001
- Revision received May 30, 2001
- Accepted June 19, 2001
- Published online October 1, 2001.
- Yuji Oikawa, MDa,
- Kazuhira Maehara, MDa,
- Tomiyoshi Saito, MDa,
- Kazuaki Tamagawa, MDa and
- Yukio Maruyama, MDa,* ()
- ↵*Reprint requests and correspondence:
Dr. Yukio Maruyama, Professor and Chairman, First Department of Internal Medicine, Fukushima Medical University, Hikarigaoka-1, Fukushima 960-1295, Japan
We investigated the changes in coronary vascular resistance caused by angiotensin II, angiotensin-converting enzyme (ACE) inhibition and angiotensin II type 1 or 2 receptor (AT1R and AT2R, respectively) antagonists in chronic heart failure (CHF).
Angiotensin II is an intense vasoconstrictor, and increased angiotensin II in CHF might exert significant vasoconstriction.
Eleven dogs were studied. Before and after three and five weeks of rapid pacing, coronary flow dynamics were evaluated by the coronary pressure–flow relationship (PFR) in long diastole, before and after intracoronary injection of angiotensin II, the ACE inhibitor enalaprilat, the AT1R antagonist L158,809 or the AT2R antagonist PD123319.
Before rapid pacing, angiotensin II reduced the slope of PFR (1.16 ± 0.08 to 0.81 ± 0.07 ml/min/100 g left ventricular mass per mm Hg; p < 0.01) and increased the perfusion pressure at which coronary flow ceased (zero-flow pressure [Pf= 0]), whereas enalaprilat did not change either of them. After rapid pacing, angiotensin II did not change the slope or Pf= 0. In contrast, enalaprilat increased the slope (three weeks: 1.20 ± 0.05 to 1.50 ± 0.03; five weeks: 1.25 ± 0.19 to 1.37 ± 0.08; both p < 0.05) and decreased Pf= 0 after three weeks of pacing, but not after five weeks. Pretreatment with the bradykinin antagonist HOE-140 attenuated the enalaprilat-induced increase in coronary blood flow. L158,809 and PD123319 had no effect both before and after rapid pacing.
This suggests that the coronary vasoconstrictive effect of angiotensin II would disappear and the vasodilatory effect of the ACE inhibitor, partly through bradykinin, would be enhanced in the early stage of CHF.
There is considerable evidence of diminished coronary flow reserve in chronic heart failure (CHF) (1–3), which might play a role in its development and progression. With respect to this, not only would increased ventricular filling pressure and elevated coronary venous pressure impede coronary inflow (4–7), but also various neurohumoral factors activated in CHF, such as sympathetic neuron-derived norepinephrine (NE) (8), angiotensin II (9)and endothelin-1 (10), would affect coronary vascular tone. Recently, the clinical benefits of angiotensin-converting enzyme (ACE) inhibition and angiotensin II type 1 receptor (AT1R) antagonism in the treatment of CHF have been reported (11–13). Angiotensin II has been shown to exert a direct vasoconstrictive effect on coronary arteries in normal subjects (14), and in CHF, it increases not only in circulating blood but also in myocardial tissue. However, it is not known how increased angiotensin II exerts a coronary vasoconstrictive action in CHF. It has been reported recently that AT1Rs are selectively downregulated, and angiotensin II type 2 receptors (AT2Rs) are relatively increased in chronically failing ventricular myocardium (15). If these changes also occur in the coronary arteries, it is possible that they might lead to attenuation of angiotensin II-induced vasoconstriction. In contrast, ACE inhibition increases coronary blood flow in patients with dilated cardiomyopathy (16). However, it is not clear how such vasodilatory action is brought about in CHF after administration of the ACE inhibitor.
The purpose of this study was to investigate whether increased angiotensin II exerts a strong coronary vasoconstrictive action in CHF in a dose-dependent manner, as in the normal heart, and whether the vasodilatory action of the ACE inhibitor is observed, and, if it is, whether it is related to an increase in bradykinin. We investigated the effects of angiotensin II, ACE inhibition and AT1R antagonism on the coronary circulation before and after pacing-induced heart failure in dogs (17–19). In addition, the effect of AT2R antagonist was also examined, because the role of the AT2R-mediated pathway on coronary vascular tone has not been fully defined.
Surgical preparations and instrumentation
Eleven Beagle dogs (NORD Co., Sapporo, Japan) of both genders (weight 10 to 15 kg) were used. General anesthesia was induced by intravenous injection of 20 mg/kg sodium thiopental (Rabonal, Tanabe Co., Ltd., Osaka, Japan), and an endotracheal tube was inserted. Using a piston-type respirator (Compact-15, Kimura Medical Instrument Co., Ltd., Tokyo, Japan) anesthesia was maintained with a mixture of nitrous oxide (30% to 40%), halothane (0.5% to 1.0%) and oxygen (60% to 70%). The heart was exposed under sterile conditions by a left thoracotomy in the fourth intercostal space. An ultrasonic transit time probe (type 3S, Transonic Systems Inc., Ithaca, New York) was implanted around the proximal left anterior descending coronary artery to measure coronary blood flow. Two screw-type unipolar myocardial pacing leads (model 5071, Medtronic Systems Inc., Minneapolis, Minnesota) were placed in the right ventricle. The pacing wires were tunneled subcutaneously to the back. A permanent pacemaker equipped with the DDD mode (Cosmos model 284-05, Intermedics Inc., Freeport, Texas) was implanted under the skin. The atrioventricular node was permanently blocked by injection of 0.1 ml of 37% formaldehyde (Wako Co., Ltd., Osaka, Japan) into the atrioventricular node through the right atrium (20). The heart rate was set at 100 beats/min in the VVI mode. Ampicillin sodium (Amipenix, Asahi Chemical Industry Co., Ltd., Tokyo, Japan) was given intravenously for two days after the operation. The baseline measurements were performed under general anesthesia, as described later. Thereafter, the pacing rate was increased to 240 beats/min by changing the pacing mode from VVI to DDD. Three and five weeks after the rapid right ventricular pacing, the same measurements as in the baseline study were repeated after the pacing rate was returned to 100 beats/min. It was not possible to make the measurements in 6 of the 11 dogs at five weeks.
The study was carried out under the supervision of the Animal Research Committee and in accordance with the Guidelines on Animal Experiments at Fukushima Medical University School of Medicine (approval no. 980010) and the Japanese Government Animal Protection and Management Law (no. 115).
Seven days after the operation, the animals were given intramuscular injections of ketamine hydrochloride (10 mg/kg), and they were then intubated and artificially ventilated as described previously. Eight-French sheath introducers (Termo Co. Ltd., Tokyo, Japan) were inserted into the left carotid artery and external jugular vein for monitoring arterial and central venous pressure, respectively, using pressure transducers and amplifiers (AP-641G, Nihon Koden, Tokyo, Japan). Then, a bolus of 100 IU/kg heparin sodium was injected, followed by 50 IU/kg per h throughout the experiment.
Fifteen milliliters of arterial blood were drawn to measure NE, angiotensin II, atrial natriuretic peptide (ANP) and endothelin-1 concentrations. The blood was collected into tubes containing EDTA and placed on ice immediately. After centrifugation for 15 min at 3,000 rpm at 4°C, the plasma was separated and stored at −20°C. Plasma NE levels were determined by high performance liquid chromatography (21), and ANP, endothelin-1 and angiotensin II concentrations were measured by radioimmunoassay (22–24).
Left ventricular pressure, left ventricular end-diastolic pressure, maximal positive and negative left ventricular pressure derivatives (+dP/dt and −dP/dt, respectively) and aortic pressure were measured with pressure amplifiers using a transducer-tipped pressure monitoring catheter (4F Camino, San Diego, California), which was inserted from the left carotid artery into the left ventricle. Cardiac output and pressures in the right side of the heart were measured using a Swan-Ganz catheter (7F, model 93-121A, American Edwars Laboratories, Santa Ana, California).
Evaluation of the coronary pressure–flow relationship (PFR)
A 5F hand-crafted right Judkins-type catheter was inserted through the sheath introducer into and through the left carotid artery for drug administration into the left coronary artery. The pressure through the sheath introducer inserted into the left carotid artery was regarded as the coronary perfusion pressure. During transient stopping of the pacing, coronary perfusion pressure and coronary blood flow were measured continuously under the following conditions at baseline (n = 11) and after three (n = 11) and five weeks (n = 5) of rapid pacing: 1) before drug administration; 2) 10 s after intracoronary infusion of angiotensin II (0.1, 1.0 and 10 ng/kg) for 30 s; 3) 2 min after bolus injection of the ACE inhibitor enalaprilat (0.2 mg/kg); 4) 2 min after bolus injection of the selective AT1R antagonist L158,809 (0.1 mg/kg) (25); 5) 2 min after bolus injection of the selective AT2R antagonist PD123319 (0.2 mg/kg) (26); and 6) 10 s after intracoronary injection of angiotensin II (10 ng/kg) after blockade of AT1R or AT2R. Then, the effect of enalaprilat was assessed after inhibition of the bradykinin type 2 receptor by intracoronary administration of 0.5 mg of HOE-140 (27)to 4 of the 11 dogs after three weeks of rapid pacing.
Fluoroscopic images were obtained with a Toshiba X-ray system (model KXO-15D). Left ventriculography was performed through a 6F pigtail catheter, which was inserted through the sheath introducer into the left carotid artery. Left ventriculograms were analyzed using a video projector (Ikegami Picture Monitor PM 14-3H, Ikegami Tsushinki, Co., Ltd., Tokyo, Japan). The end-diastolic and end-systolic frames in the end-expiratory phase were selected for volumetric analysis.
Correction of coronary blood flow with respect to myocardial weight
To delineate the perfusion area of the left anterior descending coronary artery, the catheter was advanced just distal to the flow probe, and Evans blue dye (10 mg/ml, Wako Co. Ltd.) was injected through the catheter. After sacrificing the dogs, the wet weight of the myocardium in the perfusion area was measured, and coronary blood flow was normalized to flow per 100 g of left ventricular mass. The corresponding weight of the perfusion area before rapid pacing was estimated by employing the following equation using the left ventriculogram: left ventricular mass = π/6 (L′D′2− LD2) CF3. Thus, left ventricular mass before pacing equals left ventricular mass measured after sacrifice × ([L′D′2− LD2] before pacing/[L′D′2− LD2] after pacing), where L and D are the longest lengths of the long and short axes, respectively, in the ventricular chamber; L′ and D′ are the corresponding lengths of the pericardial silhouette; and CF3is the volume correction factor proposed by Kennedy et al. (28).
Slope and zero-flow pressure of PFR
After stopping the pacing, coronary blood flow was measured at every 5 mm Hg of coronary perfusion pressure after the incisure of the last arterial pressure curve before stopping the pacing until reaching zero flow. The slope of the regression line calculated by using all data points, excluding the point of zero flow, was defined as the slope of PFR. As the correlation coefficient of linear regression was r = 0.99 ± 0.01, the use of linear regression analysis seemed to be reasonable. Then, because the zero-flow pressure (Pf= 0) value was sometimes more difficult to determine accurately in actual tracings, we obtained Pf= 0 by extrapolation of the regression line in each PFR. There were no large differences between extrapolated the Pf= 0 value and Pf= 0 determined from the flowmeter curves.
Data are expressed as the mean value ± SEM. Multiple comparisons were performed using analysis of variance followed by the Fisher post hoc comparisons test. For comparison of paired data, the Student ttest was used. A p value <0.05 was considered statistically significant.
Changes in hemodynamic and neurohumoral variables
Table 1shows the hemodynamic and neurohumoral data obtained with 100 beats/min of right ventricular pacing before and after three and five weeks of rapid pacing. At three weeks of rapid pacing, the mean right atrial pressure and left ventricular end-diastolic pressure increased significantly, whereas the mean aortic pressure, +dP/dt and −dP/dt and cardiac output decreased significantly. Plasma NE, ANP, endothelin-1 and angiotensin II increased significantly after rapid pacing. However, from three to five weeks of rapid pacing, significant changes were not observed.
Coronary PFR before and after three and five weeks of rapid pacing
After rapid pacing, the coronary blood flows with respect to each corresponding perfusion pressure did not change, as compared with the baseline values (i.e., the slope of PFR did not change significantly—baseline: 1.16 ± 0.08; three weeks: 1.20 ± 0.05; five weeks: 1.29 ± 0.08 ml/min/100 g left ventricular mass per mm Hg; p = NS for three and five weeks vs. baseline), whereas Pf= 0 increased slightly at three weeks (baseline: 31.1 ± 1.2; three weeks: 33.9 ± 0.9 [p < 0.05 vs. baseline]; five weeks: 31.7 ± 2.2 mm Hg).
Effect of angiotensin II
Before rapid pacing (n = 11), angiotensin II decreased coronary blood flow in a dose-dependent manner (Fig. 1a). After administration of 10 ng/kg of angiotensin II, the slope of PFR decreased significantly from 1.16 ± 0.08 to 0.81 ± 0.07 ml/min/100 g left ventricular mass per mm Hg (p < 0.01), and Pf= 0 increased significantly from 31.1 ± 1.2 to 33.7 ± 1.3 mm Hg (p < 0.05). Coronary blood flow decreased by 44.2 ± 10.9% at 40 mm Hg, by 50.3 ± 4.9% at 50 mm Hg and by 40.3 ± 3.0% at 60 mm Hg of perfusion pressure. In contrast, angiotensin II changed neither the slope of PFR (three weeks: 1.20 ± 0.05 to 1.18 ± 0.04; five weeks: 1.29 ± 0.08 to 1.27 ± 0.18 ml/min/100 g left ventricular mass per mm Hg; p = NS) nor the Pf= 0 (three weeks: 33.9 ± 0.9 to 33.7 ± 0.9; five weeks: 31.7 ± 2.2 to 31.5 ± 2.2 mm Hg; p = NS) after both three (n = 11) and five weeks (n = 5) of rapid pacing (Fig. 1b, c).
Effect of ACE inhibition
The slope of PFR and the Pf= 0 did not alter after intracoronary injection of enalaprilat before rapid pacing (n = 11) (Fig. 2a). However, the slope of PFR was increased significantly by enalaprilat after three (n = 11) and five weeks (n = 5) of rapid pacing (three weeks: 1.20 ± 0.05 to 1.50 ± 0.03; five weeks: 1.25 ± 0.19 to 1.37 ± 0.08 ml/min/100 g left ventricular mass per mm Hg; p < 0.05 for after vs. before enalaprilat at three and five weeks), and Pf= 0 decreased significantly after three and five weeks of rapid pacing (three weeks: 32.3 ± 1.7 to 27.0 ± 1.3 mm Hg; five weeks: 31.4 ± 0.9 to 28.0 ± 0.9 mm Hg; p < 0.05 vs. before enalaprilat at three and five weeks) (Fig. 2b, c). Thus, the vasodilatory effect of enalaprilat appeared in the failing state.
Effect of bradykinin antagonist on coronary vasodilation induced by ACE inhibition
After infusion of HOE-140, coronary blood flow did not change before and after rapid pacing (data not shown). However, after three weeks of pacing, the enalaprilat-induced increase in the slope of PFR (control state: 1.21 ± 0.04; enalaprilat: 1.45 ± 0.05 [p < 0.05 vs. control state]; pretreatment with HOE-140: 1.28 ± 0.04 ml/min/100 g left ventricular mass per mm Hg), as well as the decrease in Pf= 0 (control state: 32.3 ± 1.7; enalaprilat: 27.0 ± 1.3 [p < 0.05 vs. control state]; pretreatment with HOE-140: 31.8 ± 0.9 mm Hg) disappeared after pretreatment with HOE-140.
Role of AT1and AT2receptors
L158,809 and PD123319 did not alter the slope of PFR and Pf= 0 at baseline (data not shown). However, the administration of angiotensin II after the AT2R antagonist PD123319 decreased the slope of PFR before rapid pacing, indicating that coronary flow reduction after angiotensin II is induced through AT1R (slope before PD123319, after PD123319 and angiotensin II after PD123319: 1.11 ± 0.09, 1.15 ± 0.08 and 0.90 ± 0.04 [p < 0.05 vs. after PD123319], respectively; Pf= 0 values: 30.0 ± 1.1, 30.3 ± 1.4 and 32.1 ± 0.9 mm Hg; p = NS). Moreover, neither L158,809 nor PD123319 changed the slope of PFR or Pf= 0, despite intracoronary administration of angiotensin II (10 ng/kg) after rapid pacing (data not shown).
The major findings of this study may be summarized as follows: first, angiotensin II had a vasoconstrictive effect on the coronary circulation through AT1R in a dose-dependent manner, and endogenous angiotensin II levels would be too low to have this effect in the intact canine heart. Second, the coronary vasoconstrictive effect of angiotensin II through AT1R was greatly attenuated in the failing heart. In other words, the coronary vasculature became insensitive to angiotensin II. Third, a significant vasodilatory effect of ACE inhibition was observed in the failing heart, especially in the relatively early stage of CHF that was partly caused by the accumulation of bradykinin, in addition to the desensitization to an increase in angiotensin II.
To evaluate coronary blood flow dynamics, we used the diastolic PFR during long diastole to reduce the effects of metabolic regulation on coronary vasomotor tone. However, in this procedure, several issues should be taken into account for evaluating the slope of PFR. First, the coronary perfusion pressure after rapid pacing decreased moderately, leading to underestimation of the Pf= 0 value (29). Second, the effects of vascular compliance in coronary vascular beds on diastolic PFR depend on the speed of decrease in perfusion pressure (5,6). Because there was no significant difference in these procedures within the same group or between the baseline and failing heart groups, the effects on coronary hemodynamic data are likely to be similar in this study. Third, as previously mentioned, linear extrapolation of coronary PFRs was used to determine the Pf= 0 value, leading to overestimation of the Pf= 0 value (5). Fourth, increases in right atrial pressure and left ventricular end-diastolic pressure, which have been reported to affect PFR (4–7), were observed in the failing heart. However, in in vivo conditions, PFRs before rapid pacing and three and five weeks after rapid pacing did not differ, probably due to many compensatory factors. Moreover, comparisons of PFRs with and without treatment in the same heart were done using intracoronary injection of drugs; thus, preload changes were negligible. Fifth, we used only single concentrations of enalaprilat, the AT1R antagonist L158,809, the AT2R antagonist PD123319 and HOE-140 for bradykinin type 2 receptor blockade. However, concentrations of enalaprilat (0.2 mg/kg), L158,809 (0.1 mg/kg), PD123319 (0.2 mg/kg) and HOE-140 (0.5 mg/kg) seemed to be sufficient for intracoronary administration (27,30,31).
Attenuation of the vasoconstrictive effect of angiotensin II in CHF
In the nonfailing heart, exogenously administrated angiotensin II had a significant vasoconstrictive effect on the coronary circulation through the AT1R pathway. It should be noted that this vasoconstrictive change occurred despite the availability of compensatory coronary vasomotor mechanisms, which may have minimized the magnitude of the response to the agent. In contrast, the AT1R or AT2R antagonist alone did not change coronary blood flow, suggesting that the AT1R or AT2R pathways had little effect on coronary vasomotion in the normal heart. In contrast, in the failing heart, the vasoconstrictive effect of exogenously administrated angiotensin II was almost gone, suggesting that exogenous, as well as endogenous, angiotensin II may not have a vasoconstrictive effect on the coronary circulation in CHF. With respect to this, Asano et al. (15)reported selective downregulation of AT1Rs in failing human ventricles. Thus, there is a possibility that an altered response to angiotensin II through desensitization of AT1Rs and/or a decrease of AT1Rs might occur in vascular smooth muscle cells in the coronary arterial wall.
Vasodilatory effect of ACE inhibition in CHF
Because AT1R and AT2R antagonists had no effect on the coronary circulation, not only in the baseline state, but also in the heart failure state, probably due to the insensitivity to angiotensin II, the coronary vasodilatory effect of ACE inhibition in the present heart failure model seemed to be partly mediated through endothelium-derived nitric oxide (NO) and prostacyclin, caused by accumulated bradykinin (32–34). In previous studies using a similar pacing-induced heart failure model in dogs, plasma levels of bradykinin were elevated fourfold (34), and plasma levels of bradykinin were 4- to 10-fold increased by enalaprilat (35). The different responses to ACE inhibition of control animals and those with CHF in this study seem to be related to bradykinin levels attained after ACE inhibition, but not to endogenous bradykinin before the administration of the ACE inhibitor, as suggested from the study using HOE-140. In contrast, it has been reported that coronary flow decreased after treatment with HOE-140 in a similar CHF model (34). The reason for the different results is unclear, but methodologic differences must be considered.
In this study, in the late stage of CHF after five weeks of rapid pacing, the difference in coronary flow with and without the ACE inhibitor appeared to be attenuated, as compared with that in the early stage of CHF after three weeks of rapid pacing. According to our previous study using the same model, an increase in coronary blood flow through endothelium-derived NO was preserved until three weeks, but was attenuated after five weeks of rapid ventricular pacing (36). The NO-releasing capacity of the vascular endothelium, presumably through bradykinin release, might also participate in the alteration of the vasodilatory effect of ACE inhibition (37).
Effect of AT2R on coronary flow dynamics
Asano et al. (15)demonstrated that AT1R, but not AT2R, densities are decreased significantly in ventricles with idiopathic dilated cardiomyopathy. However, the localization and function of AT2R in the coronary circulation of the failing heart have not been elucidated. The results of this study suggest that regulation of coronary vascular tone through AT2Rs may not play a role in the coronary circulation. Further study is needed to clarify the functional role of AT2Rs in clinical settings, as well as their localization in the coronary circulation.
In contrast to the action of angiotensin II in the peripheral circulation, the vasoconstrictive effect of the angiotensin II receptor in the coronary circulation would be greatly attenuated in CHF, which might protect against myocardial injury through coronary artery constriction, as expected from significantly increased angiotensin II. Enhanced bradykinin accumulation by ACE inhibition, as well as desensitization to increased angiotensin II, would increase coronary blood flow at least in the short term in CHF. It is necessary to clarify whether the beneficial effect of ACE inhibition on the coronary circulation may also be expected in clinical settings, irrespective of the severity of CHF, and whether long-term treatment with ACE inhibition promotes this vasodilatory action in the coronary circulation.
☆ This study was supported by a Grant-in-Aid for Scientific Research (08670813) from the Ministry of Education, Science, and Culture, Japan.
- angiotensin-converting enzyme
- atrial natriuretic peptide
- angiotensin II type 1 receptor
- angiotensin II type 2 receptor
- chronic heart failure
- nitric oxide
- (coronary) pressure–flow relationship
- Pf= 0
- zero-flow pressure
- Received January 2, 2001.
- Revision received May 30, 2001.
- Accepted June 19, 2001.
- American College of Cardiology
- Canon R.O III.,
- Cunnion R.E,
- Parrillo J.E,
- et al.
- Aversano T,
- Klocke F.J,
- Mates R.E,
- Canty J.M Jr.
- Hofmann J.I,
- Spaan J.A
- Watanabe J,
- Maruyama Y,
- Satoh S,
- Keitoku M,
- Takishima T
- Mitsugi M,
- Saito T,
- Saitoh S,
- Sato M,
- Maruyama Y
- Creager M.A,
- Faxon D.P,
- Cutler S.S,
- Kohlmann O,
- Ryan T.J,
- Gavras H
- Wada A,
- Tsutamoto T,
- Fukai D,
- et al.
- Fowler N.O,
- Holmes J.C
- Asano K,
- Dutcher D.L,
- Port J.D,
- et al.
- Foult J.M,
- Tavolaro O,
- Antony I,
- Nitenberg S
- Armstrong P.W,
- Stopps T.P,
- Ford S.E,
- De Bold A.J
- Sun D,
- Huang A,
- Zhao G,
- et al.
- Steiner C,
- Kovalic A.T.W
- Ohkura Y,
- Nohta H
- Siegel P.K.S,
- Chang R.S.L,
- Mantlo B,
- et al.
- Sudhir K,
- Chou T.M,
- Hutchison S.J,
- Chatterjee K
- Dole W.P,
- Bishop V.S
- Sudhir K,
- Macgregor J.S,
- Gupta M,
- et al.
- Cheng C.P,
- Suzuki M,
- Ohta N,
- Ohno M,
- Wang Z,
- Little W.C
- Kichuk MR, Zhang X, Oz M, et al. Angiotensin-converting enzyme inhibitors promote nitric oxide production in coronary microvessels from failing explanted human hearts (abstr). Am J Cardiol 1997;80:137–42A.
- Zhang X,
- Recchia F.A,
- Bernstein R,
- Xu X,
- Nasjletti A,
- Hintze T.H
- Saito T,
- Tamagawa K,
- Oikawa Y,
- Niitsuma T
- Zanzinger J,
- Zheng X,
- Bassenge E