Dose-Dependent Effect of Endothelin-1 on Blood Flow to Normal and Collateral-Dependent Myocardium
Background Plasma levels of endothelin-1 (ET-1) increase during ischemia and could potentially contribute to impairment of myocardial blood flow (MBF). Because collateral vessels demonstrate enhanced responsiveness to certain vasoconstrictors, blood flow to collateral-dependent myocardium could be particularly sensitive to increases in ET-1 levels.
Methods and Results Studies were performed in 13 dogs in which collateral vessel development was produced by fluoroscopic embolization of the midleft anterior descending coronary artery with a hollow plug 4 to 6 weeks before the study. MBF was measured with radioactive microspheres at baseline and during 30-minute infusions of ET-1 (1, 10, and 100 ng/min) into the left main coronary artery. Because ET-1 stimulates endothelial prostacyclin release, aortic and coronary sinus levels of ET-1 and 6-keto–prostaglandin F1α were measured at the end of each infusion. ET-1 increased MBF from 0.82 mL·min−1·g−1 at baseline to 0.92 mL·min−1·g−1 at 10 ng/min (P<.05), which corresponded to a coronary plasma concentration of 73±16 pg/mL. Blood flow in the collateral zone was less (0.74 mL·min−1·g−1) than in the normal zone (P<.05) and did not increase at an ET-1 dose of 10 ng/min. MBF in the normal and collateral zones significantly decreased when ET-1 was increased to 100 ng/min, corresponding to a coronary sinus concentration of 175±45 pg/mL (P<.05). ET-1 produced dose-related increases in aortic and coronary sinus 6-keto–prostaglandin F1α and the transcoronary difference (P<.05). To assess the importance of prostacyclin in opposing the vasoconstriction produced by ET-1, additional studies were performed after cyclooxygenase blockade with indomethacin. After indomethacin administration, ET-1 (10 ng/min) caused a 120±23% increase in collateral vascular resistance (P<.05) and abolished the vasodilation that this dose produced in the normal zone.
Conclusions Blood flow to normal myocardium is increased at moderate plasma elevations of ET-1, whereas collateral blood flow is unchanged. Only at significantly elevated plasma concentrations of ET-1 is blood flow to normal and collateral-dependent myocardium impaired. Coronary endothelial production of prostacyclin in response to increasing concentrations of ET-1 represents an important means of blunting the vasoconstrictor properties of ET-1 in the canine coronary circulation. Coronary collateral vessels demonstrate a much greater dependence on prostacyclin production in blunting the vasoconstrictor properties of ET-1.
ET-1 is a member of a recently discovered family of peptides with profound vasoconstrictor properties.1 ET-1 has attracted interest clinically because elevated plasma concentrations occur with many vasoconstrictor states such as congestive heart failure2 and pulmonary hypertension.3 Elevations in ET-1 also occur in acute ischemic syndromes such as myocardial infarction4 and coronary vasospasm.5 Currently, it is unknown whether acute elevation of ET-1 in humans can impair myocardial perfusion. In animal experiments, intracoronary infusions of ET-1 have been shown to result in profound coronary vasoconstriction with severe depression of MBF.6 7 8 However, these changes occurred at ET-1 concentrations several orders of magnitude greater than those measured clinically. In addition to interacting with vascular smooth muscle, ET-1 has been shown to stimulate the release of NO and PGI2 from endothelial cells and isolated vessels.9 10 11 12 Thus, ET-1 might be capable of causing coronary vasodilation and vasoconstriction, with the net effect depending on agonist concentration.
After coronary occlusion, residual viable myocardium depends on collateral vessels for blood supply. Coronary collateral vessels demonstrate active vasomotion13 and undergo exaggerated vasoconstrictor responses to certain agonists such as vasopressin.14 It is thus possible that the increased ET-1 levels associated with acute ischemic syndromes could jeopardize blood flow to collateral-dependent myocardium. Currently, no information is available concerning the effect of ET-1 on collateral blood flow in the intact heart, although isolated vessel studies have suggested that collateral vessels may be less sensitive to the vasoconstrictor effect of ET-1 than normal coronary arterial vessels.15
This study was performed to examine the dose-dependent effect of ET-1 on blood flow to normal and collateral-dependent myocardium in a canine model of coronary artery occlusion. Doses of ET-1 were chosen to produce pathophysiologically relevant plasma concentrations. By measuring pressure in the collateral-dependent artery, we could distinguish between vasomotor responses of the collateral vessels and the resistance vessels in the collateral-dependent region in response to a range of ET-1 concentrations.
Induction of Collateral Vessels
Twenty-one mongrel dogs of either sex (20 to 25 kg) were premedicated with morphine sulfate (1 mg/kg IM), anesthetized with sodium thiamyl (20 mg/kg IV), intubated, and ventilated with room air. All dogs received procainamide (6 doses of 300 mg IM every 4 hours), gentamicin (3 mg/kg IV), and aqueous penicillin (20 000 U/kg IM). Under sterile conditions, the right common carotid artery was exposed and cannulated with a 9F sheath. After administration of heparin sulfate (6000 U IV), an 8F Judkins R4 coronary catheter was positioned in the left coronary ostium under fluoroscopic guidance, and a 0.014-in steerable angioplasty guide wire was advanced to the distal LAD. All dogs then received 100 μg IC nitroglycerin through the coronary catheter to cause maximal coronary artery vasodilation. A hollow stainless steel plug (1.1-mm ID, 2.3- to 2.7-mm OD, and 4-mm length) was advanced along the guide wire with a length of flanged PE-90 tubing into the LAD where it was wedged past the first diagonal branch. The guide wire was removed, and the position of the plug was recorded on videotape. The arterial sheath was removed, the right common carotid artery was ligated, and the skin was closed with staples. All dogs received supplemental analgesia. Three dogs suffered sudden death within 2 days of the procedure, probably secondary to ventricular arrhythmia produced by acute plug closure.
The dogs were returned to the laboratory 4 to 6 weeks (mean, 30±4 days) after placement of the intravascular plug, premedicated with morphine sulfate (1 mg/kg IM), anesthetized with α-chloralose (100 mg/kg IV followed by a continuous infusion of 10 mg·kg−1·h−1), intubated, and ventilated with room air supplemented with oxygen to maintain arterial blood gases in the physiological range. All dogs received supplemental morphine throughout the study (1 to 2 mg/h IV). Two 7F NIH catheters were placed in the descending aorta through the femoral arteries for measuring pressure and sampling blood. A similar catheter was introduced into the left common carotid artery and positioned in the left ventricle. A left thoracotomy was performed in the fifth intercostal space, and the heart was suspended in a pericardial cradle. Polyvinyl chloride catheters (3.0-mm OD) were placed in the left atrium through the appendage and in the coronary sinus through its ostium. The coronary arterial plug was located by palpation, and the LAD was mobilized. Heparin sulfate (200 U/kg IV) was administered, and the vessel was occluded proximally. The plug was retrieved through a longitudinal arteriotomy, and the vessel was allowed to bleed freely from the distal end to remove any residual thrombus. The artery was then cannulated with a thin-walled stainless steel cannula (4-mm OD). Pressure at the cannula tip was measured with a 23-gauge tube incorporated into the wall of the cannula. Resistance of the coronary cannula was previously determined to be 0.097 mm Hg·mL−1·min−1 flow. A PE-90 catheter was inserted into the proximal LAD and retrogradely positioned into the left main coronary artery for infusion of ET-1. The position of the catheter tip was determined by palpation and confirmed at autopsy. One dog died during the surgical procedure and was excluded from analysis.
The response to graded intracoronary infusions of ET-1 was studied in 13 dogs. Aortic, left ventricular, and distal coronary pressures at the cannula tip were measured with pressure transducers at the midchest level (Spectramed model TNF-R). Left ventricular pressure was recorded at normal and high gain for measurement of end-diastolic pressure. Data were recorded on an eight-channel direct-writing recorder (Coulbourne Instruments). At baseline and during each infusion of plasma ET-1, blood was collected from the aorta and coronary sinus for plasma measurements of ET-1 and 6-keto-PGF1α. All blood samples were immediately placed on ice and centrifuged, and the plasma was stored at −70°C. ET-1 was infused into the left main coronary artery at increasing doses of 1, 10, and 100 ng/min for 30 minutes at each dose; the infusion rate for all doses of ET-1 was 0.6 cm3/min. Thirty-minute infusions were used to allow plasma levels of ET-1 to rise to steady state levels. Hemodynamic measurements, blood collection, and microsphere determination of MBF were performed during the last 5 minutes of each infusion. MBF was measured at baseline and during ET-1 infusions of 1, 10, and 100 ng/min.
The importance of PGF1α production in modulating the coronary blood flow responses to ET-1 was studied in 5 additional dogs prepared as described above. MBF to the normal and collateral-dependent region was measured under baseline conditions and 30 minutes after cyclooxygenase blockade with 5 mg/kg IV indomethacin. Each dog then received a 30-minute infusion of ET-1 at 10 ng/min IC, and the MBF response was again measured with microspheres. This infusion rate was chosen because in the initial studies in group 1, this dose resulted in significantly increased blood flow to the normal myocardial region. Aortic and coronary sinus plasma levels of 6-keto-PGF1α were measured with each intervention.
Myocardial Blood Flow
MBF was measured with 15-μm-diameter radioactive microspheres obtained as 1.0 mCi in 10 mL of 10% low-molecular-weight dextran. Microspheres were labeled with 141Ce, 51Cr, 85Sr, 95Nb, or 46Sc (NEN Co) and agitated in an ultrasonic bath for at least 10 minutes before injection. For each intervention, 3×106 microspheres were injected into the left atrium while a reference blood specimen was withdrawn from the ascending aortic catheter at a rate of 15 mL/min with a peristaltic pump. Reference sampling was begun at the time of microsphere injection and continued for 90 seconds.
Determination of Collateral-Dependent Myocardium
To allow determination of tissue blood flow in the collateralized region, the shadow technique was used to delineate the collateral-dependent myocardial region.16 For this procedure, microspheres were administered into the left atrium while the coronary cannula was perfused with nonradioactive arterial blood from a pressurized reservoir. Cannula tip pressure was maintained 10 to 15 mm Hg above mean aortic pressure during microsphere injection. In this way, the myocardium distal to the site of occlusion was marked with nonradioactive blood to distinguish it from the remainder of the heart, which was perfused with microsphere-containing blood.
Hearts were excised and fixed in 10% buffered formalin, and the left ventricle was divided into four transverse rings parallel to the mitral valve. The rings were then sectioned radially into 16 segments that were divided into epicardial and endocardial halves. The resultant specimens were weighed on an analytical balance and placed into vials for counting. Myocardial and blood reference samples were counted in a gamma spectrometer with a multichannel analyzer (Packard model 5912) at window settings corresponding to the peak energies of each radionuclide. The activity in each window was corrected for background and overlapping counts between isotopes with a digital computer. Blood flow to each myocardial specimen (Qm) was computed with the formula Qm=Qr×Cm/Cr, where Qr is the reference blood flow rate (milliliters per minute), Cm is counts per minute of the myocardial specimen, and Cr is counts per minute of the reference blood specimen.
Microsphere injection during the shadow technique was used to define the collateral-dependent region. The mean and SD for blood flow to myocardial specimens remote from the collateral region, including the posterior and lateral left ventricular free wall and the posterior interventricular septum, were determined. Tissue specimens in which blood flow during the shadow injection was >3 SD below the mean were considered collateral dependent. Areas of infarction were excluded from analysis.
Preparation and Measurements of ET-1
ET-1, purchased from Peninsula Laboratories, was diluted with normal saline to the appropriate concentrations and mixed with 0.05% BSA. The level of serum ET-1 was measured in 9 dogs and quantified by radioimmunoassay with rabbit antiserum generated against ET-1. The cross-reactivity of the antibody with ET-1 is 100%; with ET-2, 7%; with ET-3, 7%; and with big endothelin, 17%.
Measurement of 6-Keto-PGF1α
6-keto-PGF1α (the stable degradation product of PGI2) was assayed from blood collected in 5-mmol/L EDTA tubes from the aorta and coronary sinus in 9 dogs. The samples were immediately placed on ice and separated by centrifugation at 2500 rpm for 15 minutes at 4°C. Plasma was stored at −70°C until assay. Immunoreactive 6-keto-PGF1α concentrations were determined in reconstituted extracts by enzyme immunoassay (Cayman Chemical Co). The detection limit of the assay is 9 pg/mL. The intra-assay and interassay coefficients of variation are <10%.
Collateral vessel resistance was calculated as the pressure drop from mean aortic pressure to the pressure at the cannula tip distal to the occlusion divided by mean collateral zone blood flow. Small-vessel resistance in the collateral-dependent region was calculated as distal coronary pressure minus left ventricular end-diastolic pressure divided by mean collateral zone blood flow. Total vascular resistance in the normal zone was calculated as aortic pressure minus left ventricular end-diastolic pressure divided by mean normal zone blood flow. Resistance calculations were expressed as coronary resistance units in millimeters of mercury per milliliter per minute per gram. Data were compared with Friedman’s nonparametric ANOVA for repeated measures. Individual comparisons were performed with the Wilcoxon signed-rank test. Data are expressed as mean±SEM.
Table 1⇓ gives hemodynamic variables measured at baseline and during ET-1 infusion. There was no change in aortic pressure, distal coronary pressure, left ventricular end-diastolic pressure, or peak left ventricular systolic pressure in response to increasing doses of ET-1. Heart rate increased during ET-1 infusion and was significantly higher at the 100 ng/min dose compared with baseline (P<.05). There was no change in the transcollateral pressure gradient in response to ET-1.
Myocardial Blood Flow
Left ventricular mass ranged from 85 to 128 g (mean, 105±4 g). The collateral zone weighed 14.5±4 g and comprised 14±1% of the left ventricle. Fig 1⇓ shows the mean MBF in the normal and collateral-dependent regions at baseline and during infusion of ET-1. MBF at baseline in the normal zone was 0.82±0.08 mL·min−1·g−1 and was significantly greater than blood flow in the collateral region, which was 0.74±0.07 mL·min−1·g−1 (P<.05). There was no significant effect of ET-1 at 1 ng/min on MBF to either region. When ET-1 was increased to 10 ng/min, a significant increase in normal-zone blood flow occurred (P<.05), whereas there was no significant change in blood flow to the collateral region. At the highest infusion rate of ET-1 (100 ng/min), mean blood flow decreased to 0.61±0.07 and 0.48±0.06 mL·min−1·g−1 in the normal and collateral zones, respectively (each P<.05).
Table 2⇓ shows blood flow distribution by transmural layers. The endocardial-to-epicardial flow ratio at baseline in the normal zone was 1.11±0.05 and was significantly higher than in the collateral region (0.93±0.07, P<.05). ET-1 did not alter the transmural distribution of blood flow in either the normal or the collateral zone, so the endocardial-to-epicardial flow ratio remained significantly higher in the normal zone at all doses of ET-1 (P<.05). At 10 ng/min, both subendocardial and subepicardial blood flows in the normal region were greater than at baseline; this change was of borderline significance (P<.07). Subendocardial and subepicardial flows were greater in the normal region compared with the collateral zone at each dose of ET-1 (P<.05).
Coronary and Collateral Vascular Resistance
Table 3⇓ gives the values for collateral vascular resistance and coronary vascular resistance in the normal and collateral-dependent regions. Collateral vessel resistance was unchanged during administration of ET-1 at 1 and 10 ng/min but increased significantly at 100 ng/min (P<.01) as MBF to the collateral-dependent region fell significantly compared with baseline. Small-vessel resistance in the collateral region was unchanged during ET-1 compared with baseline at all doses.
Coronary vascular resistance in the normal zone was unchanged compared with baseline at an ET-1 dose of 1 ng/min. At 10 ng/min, normal-zone vascular resistance underwent a small decrease of borderline significance compared with baseline (P<.07). The highest dose of ET-1 caused a significant increase in normal-zone coronary vascular resistance (P<.05) compared with 10 ng/min.
Plasma Levels of ET-1
Table 4⇓ gives the aortic and coronary sinus plasma levels of ET-1 measured at baseline and during infusion of ET-1. Baseline aortic and coronary sinus levels of ET-1 were 4.9±1.7 and 6.7±2.2 pg/mL, respectively, and increased with increasing doses of ET-1 (P<.05). Because ET-1 was infused by the intracoronary route, coronary sinus levels were significantly higher than aortic plasma levels (P<.05).
Plasma Levels of 6-Keto-PGF1α
Fig 2⇓ shows the aortic and coronary sinus plasma levels of 6-keto-PGF1α. Aortic 6-keto-PGF1α increased from 242±39 pg/mL at baseline to 562±94 pg/mL at 100 ng/min (P<.05); coronary venous concentrations increased from 372±43 to 784±153 pg/mL (P<.05). The coronary AV difference of 6-keto-PGF1α also increased with increasing doses of ET-1, indicating that ET-1 stimulated production of PGI2 in the coronary vasculature.
Table 5⇓ lists the hemodynamic data for the 5 dogs that received ET-1 after indomethacin. There was no change in mean aortic, peak left ventricular systolic, or end-diastolic pressure with ET-1. Heart rate decreased from 105±8 to 91±12 bpm after indomethacin administration (P<.07) and to 85±6 bpm after ET-1 infusion (P<.05). Distal coronary pressure was unchanged with indomethacin but decreased to 61±7 mm Hg during ET-1 infusion (P<.07), resulting in a significant increase in the transcollateral pressure gradient.
Coronary Blood Flow Response to ET-1 After Indomethacin Infusion
Table 6⇓ gives MBF data. Mean MBF in the normal region during control conditions was 0.76±0.20 mL ·min−1 · g−1 and decreased to 0.59±0.07 mL ·min−1 · g−1 during infusion of ET-1 after indomethacin. Mean MBF in the collateral region decreased from 0.65±0.16 to 0.46±0.07 mL·min−1·g−1 during ET-1 infusion. This decrease occurred primarily in the subepicardium (P<.05), so the combination of ET-1 and indomethacin caused a significant increase in the endocardial-to-epicardial flow ratio (P<.05).
Administration of ET-1 in the presence of indomethacin caused significant coronary collateral vasoconstriction, resulting in a 120±23% increase in collateral vascular resistance (P<.05). This was in contrast to the dogs in group 1 in which the same dose of ET-1 had no effect on collateral resistance (Fig 3⇓). Coronary vascular resistance in the normal region also increased in response to ET-1 infusion after indomethacin, but this response was much smaller than in the collateral vessels (mean increase, 28±9%). In contrast, the same dose of ET-1 produced a small but significant decrease in normal zone vascular resistance in the dogs in group 1 that did not receive indomethacin (Table 3⇑).
Plasma Levels of 6-Keto-PGF1α
In the dogs that received indomethacin, there was no increase in the production of 6-keto-PGF1α in response to 10 ng/min ET-1. Baseline aortic plasma levels of 411±66 pg/mL decreased to 230±31 pg/mL after indomethacin administration (P<.05). Coronary venous 6-keto-PGF1α levels at baseline were 758±206 pg/mL and decreased to 332±62 pg/mL after indomethacin administration (P<.05). ET-1 infusion failed to increase aortic or coronary venous 6-keto PGF1α levels. In contrast, ET-1 produced a significant increase in aortic and coronary sinus levels of 6-keto PGF1α in the group 1 dogs that did not receive indomethacin.
Several new findings resulted from this study. First, blood flow to normal myocardium increased at moderate plasma elevations of ET-1, whereas collateral blood flow did not change. Second, ET-1 stimulated the endothelial release of PGI2 in a dose-dependent manner, which contributed to the maintenance of MBF in the presence of increased ET-1 levels. Third, inhibition of PGI2 production with indomethacin abolished the increase in normal-zone blood flow that occurred with moderate elevations in plasma levels of ET-1. Fourth, collateral vessels demonstrated a much greater dependence on PGI2 production compared with normal coronary vessels because the administration of indomethacin resulted in severe collateral vessel vasoconstriction at moderate plasma levels of ET-1. The mechanisms and potential implications of these findings will be discussed in detail.
Response of Normal Coronary Vessels to ET-1
Early studies in intact animals documented that ET-1 can cause intense coronary vasoconstriction with reductions of MBF that can result in wall motion abnormalities and even death.6 These MBF reductions were attributed to vasoconstriction of the coronary microvasculature7 and, at higher doses, of the epicardial arteries.6 17 However, the doses of ET-1 used resulted in plasma concentrations that often were several orders of magnitude greater than values reported in pathological states.
Using a smaller dose of ET-1 (0.3-pmol/kg bolus) injected into the left circumflex coronary artery of anesthetized dogs, Tsuchiya et al18 observed a very small (not statistically significant) increase in MBF. At higher doses, coronary blood flow decreased. Donckier et al19 infused ET-1 at 2.5 ng/kg IV for 2 hours in 15 chronically instrumented dogs; a final ET-1 plasma concentration of 32 pg/mL was achieved that caused no change in LAD blood flow or myocardial function. Wang et al20 studied the relation between plasma ET-1 concentration and MBF in awake dogs. After a 20-minute infusion at 1 ng·kg−1·min−1 IV, there was no change in coronary blood flow or large coronary artery diameter; however, the plasma ET-1 level was not significantly different from baseline (4±1 pg/mL). When the infusion rate was increased to 10 ng·kg−1·min−1, the systemic plasma ET-1 level increased to 70 pg/mL and coronary blood flow decreased by 22%. In contrast we observed a significant increase in normal-zone MBF at a similar coronary plasma concentration after a 30-minute infusion of ET-1 at 10 ng/min. This difference may be related to differences between awake and anesthetized animals. In agreement with other studies,6 7 8 we observed a significant decrease in MBF at a higher dose of ET-1 (100 ng/min), which resulted in a coronary sinus concentration of 175±45 pg/mL. These studies indicate that the response to ET-1 is concentration dependent, so either coronary vasodilation or vasoconstriction can occur, depending on the infusion rate.
Transmural Distribution of MBF
When MBF was examined by transmural layers, ET-1 caused no change in the endocardial-to-epicardial flow ratio in the normal region at doses that caused either vasodilation (10 ng/min) or vasoconstriction (100 ng/min). Previous studies6 17 reported that vasoconstrictor doses of ET-1 caused a greater reduction of subepicardial blood flow, but this may be related to the higher doses of ET-1 used in those studies. The endocardial-to-epicardial flow ratio in the collateral region was consistently lower than in the normal zone and was unchanged throughout the ET-1 infusion range.
Mechanism of ET-1–Induced Vasodilation
Our findings indicate that continuous infusions of low doses of ET-1 that result in pathophysiologically relevant plasma concentrations can cause coronary vasodilation. These results support the hypothesis that the net effect of ET-1 on coronary vasomotor tone is dose dependent and that vasoconstriction will occur only at concentrations sufficiently elevated to overcome the endothelial release of PGI2 or NO. Bolus injections of ET-1 have been reported to cause transient coronary vasodilation lasting 30 to 60 seconds, followed by sustained vasoconstriction.20 21 22 In isolated rat hearts21 perfused at a constant flow rate, a 10-pmol bolus of ET-1 resulted in an immediate 14 mm Hg decrease in coronary perfusion pressure followed by sustained vasoconstriction. Studies in rat aortic rings23 showed that after binding ET-1 is internalized with its receptor so that the vasoconstrictor response is diminished with subsequent doses. This may also occur in endothelial cells because the transient vasodilator response to bolus doses of ET-1 is also diminished with repeated administration.24 Thus, short-term administration of ET-1 can produce different results from those obtained during long-term exposure to elevations in this peptide.
The response of coronary blood vessels to ET-1 appears to be dependent on vessel size. Thus, intracoronary infusions of ET-1 (10−8 to 10−7 mol/L) dilated arterioles <130 μm in diameter but did not affect larger arteries in anesthetized dogs.25 After cyclooxygenase blockade with indomethacin, ET-1 caused constriction of coronary arteries, but arterioles <130 μm still dilated, although the mechanism was not studied. In contrast, Homma et al,26 using intravital microscopy, observed prominent vasoconstriction in arterioles <200 μm after intracoronary ET-1 infusion (2.5- to 75-pg bolus doses). However, Ku27 showed that isolated epicardial vessels are less sensitive than intramyocardial vessels to the vasoconstrictor effects of ET-1. It is likely that ET-1 receptor density and subtype differ with vessel size, location, and species28 and that stimulation of endothelium-dependent vasodilators such as PGI2 and NO also vary in efficacy with vessel location and animal species.
To cause vasoconstriction, intraluminal ET-1 must diffuse across the endothelial layer to reach the smooth muscle. In human umbilical vein endothelial cell monolayers, only 6% of ET-1 diffused across the monolayer in the first hour.29 In the absence of endothelial cells, near equilibration of ET-1 occurred over the same time interval. The physiological importance of this finding was demonstrated by Lamping et al,25 who used stroboscopic microscopic visualization of epicardial microvessels in beating hearts of anesthetized dogs. They found that intracoronary administration of ET-1 (10−8 mol/L) resulted in vasodilation of small arterioles and that topical administration produced vasoconstriction. In addition to acting as a diffusion barrier, vascular endothelial cells were recently shown to possess a deamidase enzyme that inactivates ET-1.30 Thus, disease processes that impair the barrier function or vasodilator mechanisms of the vascular endothelium could potentiate the vasoconstrictor properties of ET-1.
Role of PGI2
PGI2 has been demonstrated to exert potent vasodilating effects on the coronary vasculature in normal and collateral-dependent myocardium. Using isolated, fibrillating, blood-perfused canine hearts 4 to 5 weeks after occlusion of the left circumflex coronary artery with an ameroid constrictor, Scholtholt et al31 found that PGI2 produced dose-dependent increases in blood flow in both the normal and collateralized myocardial regions.
Several investigators have demonstrated that PGI2 can blunt the vasoconstrictor properties of ET-1.10 In isolated human aortic endothelial cells, ET-1 (10 μmol/L) was shown to stimulate the release of a wide variety of cyclooxygenase products, including 6-keto-PGF1α.12 In anesthetized dogs, bolus doses of ET-1 (0.03 to 0.3 nmol/kg) elicit large increases in plasma levels of 6-keto-PGF1α.32 In this study, ET-1 produced dose-dependent increases in aortic and coronary sinus plasma levels of 6-keto-PGF1α, the stable degradation product of PGI2. Furthermore, the coronary AV gradient of this product increased with increasing doses of ET-1, indicating that ET-1 stimulated coronary production of PGI2. These results support previous findings in porcine coronary artery strips10 and cultured human aortic endothelial cells12 demonstrating that ET-1 stimulates endothelial release of PGI2. Our results extend these findings by documenting this interaction between ET-1 and PGI2 in the coronary circulation of an intact animal. These opposing influences may explain why coronary blood flow to the normal and collateral regions underwent only modest changes over a wide range of ET-1 plasma concentrations because PGI2 levels increased in parallel with the rise in ET-1 levels. The importance of coronary endothelial production of PGI2 is demonstrated by the finding that normal-zone blood flow failed to increase with ET-1 (10 ng/min) after indomethacin, whereas collateral vasoconstriction to ET-1 was potentiated when PGI2 production was inhibited (Fig 4⇓).
Role of NO
Studies using endothelial cells and isolated vessels have demonstrated that ET-1 can stimulate the production of NO, which promotes vasodilation and opposes vasoconstriction produced by the agonist. In porcine coronary artery strips with intact endothelium, low concentrations of ET-1 (0.1 to 1.0 nmol/L) produced transient relaxation accompanied by a reduction in smooth muscle cytosolic Ca2+ concentration; these responses were abolished by removal of the endothelium or pretreatment with the nitric oxide synthase inhibitor N-nitro-l-arginine.9 ET-1 in higher concentrations (1 to 100 nmol/L) produced vasoconstriction through a direct action on the vascular smooth muscle. Studies in anesthetized dogs have demonstrated that inhibition of NO production with NG-monomethyl-l-arginine potentiated the vasoconstrictor actions of ET-1 (2.5 ng·kg−1·min−1 IV for 120 minutes) on systemic, renal, and pulmonary vasculature.33 However, this did not occur in the coronary circulation because NG-monomethyl-l-arginine alone increased coronary vascular resistance to the same degree as the combination of NG-monomethyl-l-arginine and ET-1, potentially supporting a balanced vasodilator and vasoconstrictor role for ET-1 in the coronary circulation.
Effect of ET-1 on the Collateral Circulation
Collateral vessels demonstrate heightened sensitivity to certain vasoconstrictors such as vasopressin, which can decrease blood flow to the dependent myocardium at doses that do not affect blood flow to normal myocardium.14 Using isolated vessel rings, Parker et al15 found that collateral vessels constricted smaller-than-normal coronary arterial vessels to ET-1. An impaired collateral vasoconstrictor response might be expected to protect collateral blood flow when ET-1 levels are increased. In the present study, MBF in the collateral region did not significantly change even though coronary sinus concentrations of ET-1 exceeded 70 pg/mL. Only at the highest infusion rate, when the coronary blood concentration was 175±45 pg/mL, did collateral blood flow decrease significantly. These results differ from previous reports in isolated collateral vessels15 and indicate that in the intact animal collateral vessels constrict similarly to normal vessels in response to ET-1.
Coronary collateral vessels can synthesize PGI2. Furthermore, in dogs with long-term coronary occlusion, PGI2 appears to cause tonic collateral vessel dilation because cyclooxygenase blockade with indomethacin significantly decreased retrograde blood flow from the cannulated collateral-dependent artery.34 In the present study, ET-1 increased coronary sinus levels of 6-keto-PGF1α. When PGI2 production was inhibited with indomethacin, ET-1 in a dose that caused no change in collateral resistance in the group 1 dogs produced a 120±23% increase in collateral vascular resistance. This finding indicates that PGI2 production represents an important means of blunting the vasoconstrictor effects of ET-1 in the collateral circulation. In contrast, the response of coronary vascular resistance to ET-1 in the normal zone was much less affected by inhibition of PGI2 production. This is in agreement with previous findings that vasodilator prostaglandins are of greater importance in collateral than in normal coronary vessels.34
Well-developed collateral vessels exhibit NO-mediated endothelium-dependent vasodilation in response to agonists such as acetylcholine or bradykinin.35 36 37 Although the present study was carried out relatively early after coronary occlusion, we found that even at this early stage of collateral development the response to endothelium-dependent agonists is intact (unpublished observation). Thus, vasodilation of collateral vessels resulting from ET-1–stimulated production of NO also could potentially improve perfusion of the collateral-dependent region.
ET-1 also could alter blood flow to the collateral-dependent myocardium by influencing vasomotion of the resistance vessels in the collateral zone. Receptor-mediated, endothelium-dependent dilation has been shown to be impaired in microvessels chronically perfused by collateral vessels.38 Possibly for this reason, ET-1 failed to cause vasodilation in the collateral region as it did in the normal zone.
Myocardial ischemia can result in acute elevations of ET-1. The present findings confirm that ET-1 can cause vasoconstriction in both normal and collateral-dependent regions of myocardium, but this occurred only at the highest infusion rate when mean coronary plasma concentrations exceeded clinically measured levels. However, these studies were performed in healthy dogs with intact endothelial function. It is possible that the vasoconstrictor effects of ET-1 would be potentiated in patients with coronary disease because of the coexistence of endothelial dysfunction. This could be especially important in patients with regions of collateral-dependent myocardium because collateral vessels demonstrated a greater dependence on PGI2 production for counteracting the vasoconstrictor effects of ET-1.
Acute elevations in coronary plasma ET-1 concentrations within a pathophysiological range did not impair blood flow to normal or collateral-dependent myocardium. Only at superphysiological plasma concentrations was blood flow to either region decreased. Increased PGI2 production that occurred in response to increasing doses of ET-1 counteracted the vasoconstrictor properties of ET-1 and preserved myocardial blood flow. Collateral vessels demonstrated a greater dependence on PGI2 production to oppose the vasoconstriction caused by ET-1.
Selected Abbreviations and Acronyms
|bpm||=||beats per minute|
|LAD||=||left anterior descending coronary artery|
|MBF||=||myocardial blood flow|
This work was supported by US Public Health Service grants HL-32427, HL-20598, and HL-21872 from the NHLBI and individual National Research Service Award HL-09128 (Dr Traverse). Immunoreactive 6-keto-PGF1α assay kits were generously provided by Solidad Callejas and Upjohn Pharmaceuticals. We gratefully acknowledge the expert technical assistance provided by Todd Pavek, Sara Herrlinger, Melanie Crampton, Paul Lindstrom, Marj Carlson, and Flor Dizon. Secretarial assistance was provided by Sue Quirt.
- Received July 13, 1995.
- Revision received September 21, 1995.
- Accepted September 25, 1995.
- Copyright © 1996 by American Heart Association
Stewart DJ, Cernacek P, Costello KB, Rouleau JL. Elevated endothelin-1 in heart failure and loss of normal response to postural change. Circulation. 1992;85:510-517.
Toyo-oka T, Aizawa T, Suzuki N, Hirata Y, Miyauchi T, Shin WS, Yanagisawa M, Masaki T, Sugimoto T. Increased plasma level of endothelin-1 and coronary spasm induction in patients with vasospastic angina pectoris. Circulation. 1991;83:476-483.
Clozel J-P, Clozel M. Effects of endothelin on the coronary vascular bed in open-chest dogs. Circ Res. 1989;65:1193-1200.
Suzuki S, Kajikuri J, Suzuki A, Itoh T. Effects of endothelin-1 on endothelial cells in the porcine coronary artery. Circ Res. 1991;69:1361-1368.
Lüscher TF, Yang Z, Tschudi M, von Segesser L, Stulz P, Boulanger C, Siebenmann R, Turina M, Bühler FR. Interaction between endothelin-1 and endothelium-derived relaxing factor in human arteries and veins. Circ Res. 1990;66:1088-1094.
Hollenberg SM, Tong W, Shelhamer JH, Lawrence M, Cunnion RE. Eicosanoid production by human aortic endothelial cells in response to endothelin. Am J Physiol. 1994;267:H2290-H2296.
Hautamaa PV, Dai XZ, Homans DC, Bache RJ. Vasomotor activity of moderately well-developed canine coronary collateral circulation. Am J Physiol. 1989;256:H890-H897.
Peters KG, Marcus ML, Harrison DG. Vasopressin and the mature coronary collateral circulation. Circulation. 1989;79:1324-1331.
Parker JL, Myers PR, Mattox ML, Jones AW. Impaired vasoconstrictor responses to endothelin in coronary collateral arteries. Circulation. 1990;82(suppl III):III-336. Abstract.
Patterson RE, Kirk ES. Apparent improvement in canine collateral myocardial blood flow during vasodilation depends on criteria used to identify ischemic myocardium. Circ Res. 1980;47:108-116.
Tsuchiya T, Okumura K, Ishizaka H. Effect of low dose of endothelin-1 on the coronary resistance vessels in dogs: role of endothelium-derived nitric oxide. Circulation. 1991;84(suppl II):II-627. Abstract.
Donckier J, Hanet C, Galanti L, Stoleru L, Van Mechelen H, Hayashida W, Ketelslegers J-M, Pouleur H. Effects of endothelin-1 at pathophysiological concentrations on the normal and post-ischemic myocardium. J Am Coll Cardiol. 1993;21:69A. Abstract.
Wang J, Zeballos GA, Kaley G, Hintze TH. Dilation and constriction of large coronary arteries in conscious dogs by endothelin. Am J Physiol. 1991;261:H1379-H1386.
Asajima H, Sakuma I, Tamura M. Dilatation by endothelin of rat coronary vascular beds through the release of endothelium-derived hyperpolarizing factor which opens Ca2+-activated K+ channel on vascular smooth muscle. Circulation. 1992;86(suppl I):I-485. Abstract.
Hollenberg SM, Shelhamer JH, Cunnion RE. Tachyphylaxis to the vasopressor effects of endothelin in rat aortic rings. Am J Physiol. 1993;264:H352-H356.
LeMonnier de Gouville AC, Mondot S, Lippton H, Hyman A, Cavero I. Hemodynamic and pharmacological evaluation of the vasodilator and vasoconstrictor effects of endothelin-1 in rats. J Pharmacol Exp Ther. 1990;252:300-310.
Lamping KG, Clothier JL, Eastham CL, Marcus ML. Coronary microvascular response to endothelin is dependent on vessel diameter and route of administration. Am J Physiol. 1992;263:H703-H709.
Homma S, Miyauchi T, Goto K, Sugishita Y, Sato M, Ohshima N. Effects of endothelin-1 on coronary microcirculation in isolated beating hearts of rats. J Cardiovasc Pharmacol. 1991;17(suppl 7):S276-S278.
Ku DD. Differential responses between epicardial and intramyocardial coronary arteries to endothelin-induced constriction in dogs. Circulation. 1991;84(suppl II):II-627. Abstract.
Yanagisawa M. The endothelin system: a new target for therapeutic intervention. Circulation. 1994;89:1320-1322.
Wagner WF, Christ G, Wojta J, Vierhapper H, Parzer S, Nowotny PJ, Schneider B, Waldhausl W, Binder BR. Polar secretion of endothelin-1 by cultured endothelial cells. J Biol Chem. 1992;267:16066-16068.
Lerman A, Sandok EK, Hildebrand FL, Burnett JC. Inhibition of endothelium-derived relaxing factors enhances endothelium-mediated vasoconstriction. Circulation. 1992;85:1894-1898.
Altman J, Dulas D, Bache RJ. Effect of cyclooxygenase blockade on coronary collateral blood flow. Circ Res. 1992;70:1091-1098.
Altman J, Dulas D, Pavek T, Laxson DD, Homans DC, Bache RJ. Endothelial function in well-developed canine coronary collateral vessels. Am J Physiol. 1993;264:H567-H572.
Angus JA, Ward JE, Smolich JJ, McPherson GA. Reactivity of canine isolated epicardial collateral coronary arteries: relation to vessel structures. Circ Res. 1991;69:1340-1352.
Flynn NM, Kenny D, Pelc LR, Warltier DC, Bosnjak ZJ, Kampine JP. Endothelium-dependent vasodilation of canine coronary collateral vessels. Am J Physiol. 1991;261:H1797-H1801.
Sellke FW, Quillen JE, Brooks LA, Harrison DG. Endothelial modulation of the coronary vasculature in vessels perfused via mature collaterals. Circulation. 1990;81:1938-1947.