17β-Estradiol Prevents Dysfunction of Canine Coronary Endothelium and Myocardium and Reperfusion Arrhythmias After Brief Ischemia/Reperfusion
Background Brief myocardial ischemia is associated with myocardial and coronary endothelial dysfunction caused by oxygen free radicals released during reperfusion. Estrogen, known to have antioxidant activity, may prevent these complications.
Methods and Results We assessed the effect of 2 weeks of treatment with 17β-estradiol (E, 100 μg·kg−1·d−1, n=12) or placebo (P, n=15) on myocardial and coronary endothelial function during the first 2 hours of reperfusion in dogs subjected to 15 minutes of ischemia induced by occlusion of the left anterior descending coronary artery (LAD). Our results show that the incidence of ventricular arrhythmias significantly decreased in E (3 of 12) compared with P (11 of 15). Systolic shortening, significantly depressed in P during early reperfusion, was maintained at preischemic levels in E. During reperfusion, the increase in LAD flow to acetylcholine, attenuated in P (60±6%), was preserved in E (151±28%) and was associated with increased serum nitrite/nitrate concentration. n-Pentane in exhaled gas in vivo, an index of lipid peroxidation, increased significantly during early reperfusion in P (from 9.1±1.9 to 41.6±13.0 ppb, P<.05) but not in E (23.0±6.9 ppb). In vitro, arterial segments from E generated significantly less superoxide anion after hypoxia/reoxygenation than those from P. Ischemic/reperfused LAD segments from E also revealed a better preservation of endothelium-dependent relaxation in vitro (maximum relaxation, 42±4% versus 24±4% in P; P<.05).
Conclusions Estrogen protects against endothelial and myocardial dysfunction resulting from brief ischemia/reperfusion. This protection may relate to an antioxidant effect of estrogen.
Brief myocardial ischemia is a prevalent clinical event in patients with established CAD.1 Although the event itself does not induce myocardial infarction, it presents serious clinical challenges such as reperfusion arrhythmias and myocardial as well as coronary endothelial dysfunction (stunned myocardium and endothelium).1 2 3 4 Recent evidence indicates that the massive generation of oxygen free radicals and/or related calcium overload at the time of reperfusion is most likely responsible for the complications related to ischemia/reperfusion. This view is supported by studies in which a variety of interventions that exert antioxidant activity prevented or attenuated injury resulting from brief ischemia/reperfusion.5 6 7
On the other hand, the sex differences in CAD and related morbidity and mortality are well documented and have been attributed, in large part, to differences in endogenous sex hormones.8 Studies have found ≈50% reduction in CAD-related coronary events in postmenopausal women on estrogen replacement therapy.9 10 The beneficial effect of estrogen replacement therapy on lipoprotein metabolism11 accounts for only 25% to 50% of the observed risk reduction, suggesting that other factors are involved.8 12
Recent studies also show that estrogen directly modifies the function of the endothelium and may attenuate vasomotor dysfunction, a possible trigger of acute coronary occlusion.13 14 Although the exact mechanisms for this protection are not yet clear, several acute and chronic studies suggest a direct effect of estrogen on the activity of the l-arginine/NO pathway in vascular endothelial cells. In vivo, direct intracoronary infusion of 17β-estradiol enhances coronary vasodilation in response to acetylcholine within 20 minutes after injection.15 Similarly, chronic estrogen treatment was shown to increase serum nitrate and nitrite levels in postmenopausal women.16 Estrogen upregulates the expression of the cNOS in female guinea pigs.17
Estrogen also has an antioxidant potential that may contribute to its protective effect on the endothelium. The antioxidant properties of the estrogen molecule may relate to the presence of a phenol ring that scavenges oxygen free radicals, resulting in hydroxylated products,18 19 or may prevent lipid peroxidation.20 Chronic administration of 17β-estradiol in hypercholesterolemic swine is reported to limit oxidation of LDL and preserve endothelial vasodilator function.13 Oxidation of LDL is associated with abnormal endothelial function; thus, the prevention of LDL oxidation by estrogen may be responsible for the restoration of vasodilator responses to ACh and, to a certain extent, for the antiatherosclerotic effects of the hormone.13 21
On the basis of these observations, we hypothesized that estrogen treatment may prevent or reduce the incidence of ventricular arrhythmias and myocardial and coronary endothelial dysfunction resulting from free radical generation during brief ischemia/reperfusion. To test this hypothesis, we assessed the effects of chronic administration of 17β-estradiol on the incidence of ventricular arrhythmias and changes in coronary endothelial and myocardial systolic function during the first 2 hours of coronary reperfusion after 15 minutes of ischemia. We also conducted ex vivo studies to quantify coronary vascular reactivity, using epicardial coronary artery segments harvested from dogs subjected to 15 minutes of coronary arterial occlusion and 20 minutes of reperfusion. To determine whether the beneficial effects of estrogen on ischemia/reperfusion injury are related to its action on oxygen free radicals and/or NO generation, we measured n-pentane levels in exhaled gases as well as nitrite/nitrate concentration in coronary venous blood during ischemia/reperfusion periods. The generation of oxygen free radicals will be determined by measurement of n-pentane from exhaled gas. n-Pentane is a product of lipid peroxidation resulting from interaction of oxygen free radicals with membrane lipids.22 23 24 We also measured superoxide generation during hypoxia/reoxygenation challenge in vitro using vascular tissue obtained from in vivo studies.
Animals and General Preparation
Male mongrel dogs (n=39) weighing 25 to 32 kg and in good general health were used for this study. Dogs were treated for 2 weeks with a daily subcutaneous injection of 17β-estradiol (100 μg·kg−1·d−1) (n=19) or an equal volume of the vehicle (n=20). Dogs were allowed free access to the regular diet and water. Blood samples were obtained for measurement of 17β-estradiol by a standard radioimmunoassay.25 All protocols and procedures of this investigation were approved by the Animal Care and Use Committee of Georgetown University and conformed to the guiding principles of the American Physiological Society. Animals were divided into three groups: In group 1, 6 estrogen-treated and 7 placebo-treated dogs were used for in vivo ischemia/reperfusion studies. One placebo animal died of arrhythmia on reperfusion, leaving 6 placebo animals available for the completion of the study (n=13). In group 2, 6 estrogen- and 8 placebo-treated dogs were used for the ex vivo vascular reactivity studies. In the placebo group, 2 animals died on reperfusion, leaving 6 available for completion of ex vivo experiments (n=14). In group 3, 7 estrogen- and 5 vehicle-treated dogs were used to study nitrite/nitrate formation in vivo and superoxide production in vitro. Two dogs from the estrogen-treated group, however, died during surgery, leaving only 5 available for measurement of NO metabolites (n=12).
Dogs were fasted overnight before surgery, with water provided ad libitum. After induction of anesthesia with sodium thiopental (25 mg/kg), the trachea of each dog was intubated and the lungs were mechanically ventilated. Anesthesia was then maintained with halothane (1.5%), and a thoracotomy was performed through the left fifth intercostal space. The following instrumentation was then established.
1. A polyvinyl chloride catheter was placed into the right femoral artery. A catheter-tipped micromanometer (Millar Inc) was passed through this catheter into the aortic arch for measurement of aortic pressure. This catheter was also used to withdraw blood for measurement of RMBF.
2. A silicon rubber catheter was implanted into the left atrium to be used for microsphere injections for RMBF measurements and for passing a micromanometer into the LV cavity for measurement of LV pressure.
3. A pair of ultrasonic dimension transducers (Vernitron) was implanted 10 mm apart in the subendocardium of a region of the anterior free wall of the LV supplied by the LAD. This was used to measure SS as described previously.26
4. An ultrasonic flow probe (Transonic Systems Inc) was positioned around the LAD distal to the second major diagonal branch to measure LAD flow.
5. A 24-gauge Teflon catheter (Deseret Medical Inc) was inserted via the first diagonal branch into the LAD with its tip positioned at the level of the first diagonal branch. This catheter was used to study coronary vasodilatory responses.
6. A rubber band was looped around the LAD proximal to the first diagonal branch, and free ends of the band were passed through a narrow Silastic tube; this snare was used for occlusion of the LAD.
7. A gas-sampling catheter (20 gauge×15 cm) was inserted into the trachea via a side port on the endotracheal tube, with the tip of catheter located near the carina. This catheter was used for sampling end-tidal gases for the measurement of pentane concentration.
8. Five-lead ECG electrodes were established to monitor ECG.
Surgical preparations were identical to those described in in vivo studies, except for the measurement of RMBF and end-tidal pentane concentrations. Thus, the left atrial and gas-sampling catheters were not included.
Animals were prepared as described in group 1 except for steps 2, 5, and 7. Instead, a 24-gauge Teflon catheter was inserted into the great coronary vein for blood sample collection for measurement of NO metabolites.
In Vivo Studies
Routine arterial blood gases, electrolytes, and hematocrit were measured in each dog the day before surgery, and abnormal values (mainly pH and potassium ion levels) were corrected. End-tidal halothane, carbon dioxide, and arterial oxygen saturation were continuously monitored with a POET agent analyzer equipped with a pulse oxymeter (Criticare Systems). End-tidal carbon dioxide levels were maintained between 33 and 37 mm Hg, and the end-tidal concentration of halothane was maintained at 1.5% throughout the study. Body temperature was maintained in the normothermic range with a warming blanket. ECG, aortic pressure, LV pressure, LAD flow, and SS were monitored continuously. All signals were filtered and recorded on a polygraph. Analog signals were digitized at an eight-channel sweep rate of 200 Hz with an analog-to-digital converter (ADAC model 1012) and PDP 11/73 microcomputer (Digital Equipment Corp) programmed for analysis of hemodynamic data.
Coronary occlusion and reperfusion
All dogs in this part of the study were subjected to 15 minutes of ischemia, followed by 120 minutes of reperfusion. Cessation of LAD flow, wall motion changes, cyanosis of the anterior wall, and typical ECG changes immediately after LAD occlusion were used as signs of successful induction of anterior wall ischemia. To ensure that no substantial collateral blood flow to the ischemic regions occurred during occlusion and that reperfusion after release of LAD was successful, microspheres were injected for RMBF measurement before LAD occlusion and 10 minutes into ischemia in 12 dogs (6 dogs per group).
Coronary arterial vasodilatory responses
Coronary vasoreactivity was tested by infusion of drugs over 60-second periods into the LAD via the indwelling catheter. Each test was performed before LAD occlusion and at 15, 30, 60, 90, and 120 minutes into reperfusion, allowing enough time for recovery between tests. Endothelium-dependent vasodilation was tested with ACh (Sigma Chemical Co). The dose of ACh varied between 0.5 and 2.0 μg, based on changes in coronary blood flow between individual animals. The same dose of ACh as used for the baseline study was also used in the reperfusion study for each animal. Endothelium-independent vasodilatory response was tested with 10 μg SNP (Sigma). These doses of drugs injected into the LAD over 60 seconds resulted in reproducible LAD flow increases without causing significant systemic hemodynamic changes.3
End-tidal gas samples were collected for pentane measurements before LAD occlusion and at 5, 15, 30, 60, 90, and 120 minutes into reperfusion. A breath sample (15 mL) was collected at the end of expiration with a 20-mL syringe attached to the gas sampling catheter. An aliquot of 3 mL of sample gas was aspirated from the collecting syringe into an airtight glass syringe (Hamilton Co) and injected into the injection port of the gas chromatograph. The gas chromatographic analysis was performed with a model 10S Plus gas chromatographic system with a photo ionization detector (Photovac International Inc). The separation of pentane from isoprene was accomplished with a 9-m CP WAX 52 analytical capillary column (composition, Carbowax PEG 20M) and a 39-cm SE-30 packed precolumn (composition, 100% dimethylpolysiloxane). The oven temperature was maintained at 45°C, and the carrier gas (flow rate, 20 mL/min) was ultrapure nitrogen (Roberts Oxygen). All calibration gases were obtained from Chem Service. For calibration purposes, different concentrations of standards (10 ppb to 100 ppm) were made by the technique described earlier.23 Our preliminary studies confirmed that this modified system was able to separate n-pentane (retention time, 188 seconds) from isoprene (retention time, 982 seconds). The detection limit of the assay was between 5 and 10 ppb for both gases. The within-run reproducibility computed from six measurements of three different concentrations (10 ppb, 1 ppm, and 100 ppm) had a coefficient of 5% to 12%, and the standard curves of the calibration gases were linear up to 100 ppm (r=.99). Anesthetic gases did not interfere with pentane measurements.
Analysis of ventricular arrhythmias
Reperfusion ventricular arrhythmias were analyzed in all 27 dogs used in groups 1 and 2, based on two-channel ECG recordings (leads II and V) at a speed of 25 mm/s made during the first 5 minutes of both ischemia and reperfusion. Isolated premature ventricular beats, ventricular bigeminy, ventricular salvos, VT, and VF were identified in accordance with the Lambeth Convention guidelines.27 All dogs remained in sinus rhythm during reperfusion.
SS was calculated as described earlier26 before ischemia, 10 minutes into ischemia, and at 15, 30, 60, 90, and 120 minutes of reperfusion.
Polystyrene-divinylbenzene microspheres (15±0.2 μm in diameter; Triton) suspended in saline with 0.2% Tween 80 were used to measure RMBF, as described previously.28 At the conclusion of the study, the dog was killed with a pentobarbital overdose, and the ischemic region of the anterior wall of the LV was harvested. This ischemic area was identified by techniques described previously.29 Five tissue samples (cubic samples of the whole myocardial thickness weighing ≈2 g) were obtained near the center of the ischemic area. The average value of five samples was used to calculate RMBF for each time point for each dog.
Ex Vivo Vascular Reactivity Studies
After 15 minutes of ischemia and 20 minutes of reperfusion, the heart was arrested by electrical stimulation, excised, and placed immediately into chilled Krebs-Ringer bicarbonate buffer. The epicardial portion of the LAD (outside diameter, 1.5 to 2 mm distal to ligation site) and an epicardial portion of the LCx of similar size were isolated and removed. The arteries were cut into 2-mm rings. The rings were then placed into tissue chambers containing Krebs-Ringer bicarbonate buffer at 37°C and aerated with 95% O2/5% CO2, mounted on stainless steel hooks, and suspended from force-displacement transducers (Harvard Apparatus) that measure isometric tension developed parallel to the circular smooth muscle fibers. The tension developed was continuously recorded on a multichannel ink recorder (Harvard). Resting tension was maintained at 2 g. The tissues were allowed ≈60 minutes of equilibration time after suspension in the chambers before testing of vasodilatory responses.
Vascular reactivity assay
Rings from the LAD served as the vessel segments preexposed to ischemia/reperfusion, whereas rings from the LCx served as control segments. Sequences for relaxation tests were randomly assigned. Sufficient time for recovery (20 to 30 minutes) was allowed between tests. Relaxation responses were tested on rings precontracted with 32 mmol/L KCl. In our previous study, 32 mmol/L KCl produced 50% to 70% of maximal contraction in our system.3 Incremental concentrations of SNP (10−9 to 10−4 mol/L) or ACh (10−9 to 10−4 mol/L) were injected into the tissue baths to produce cumulative concentration-response curves.
Measurement of Serum Nitrite and Nitrate
NO production was measured in placebo- and estradiol-treated animals subjected to 15 minutes of ischemia followed by 60 minutes of reperfusion. Blood was collected before ischemia and at 5, 15, and 60 minutes into reperfusion via the intravenous catheter. Serum concentrations of nitrite and nitrate were determined by the technique described by Takashi et al.30
In Vitro Measurement of Superoxide Anion
The in vitro effect of hypoxia and reoxygenation on superoxide production31 was studied in LCx harvested from placebo- or 17β-estradiol–treated dogs. LCx segments were incubated in a sealed test tube in potassium phosphate buffer at 37°C and aerated with a mixture of 95% nitrogen/5% CO2 for 15 minutes. The average partial oxygen pressure of the hypoxic incubation buffer was 31±2 mm Hg. The segments were then reoxygenated with 95% O2/5% CO2 for 20 minutes. Aerobic control baseline experiments were performed similarly, with LCx segments exposed to normoxic conditions. Superoxide production in response to hypoxia/reoxygenation was measured by the cytochrome c reduction technique,32 and the data were expressed as percent increase in superoxide over baseline.
For in vivo studies, vasodilatory responses to ACh were calculated as percent changes from baseline values. SS, end-tidal pentane, and nitrite/nitrate concentration were also reported as absolute values. The prevalences of dogs demonstrating any ventricular arrhythmia were compared between the two groups.
For ex vivo studies, vasodilatory responses to ACh and SNP were quantified on the basis of the cumulative dose-response curves. Responses were expressed as percent relaxation from the precontracted state. If more than one ring from a dog was tested, average values were used to represent one dog. In general, multivariate analysis was performed by ANOVA, with Scheffe´'s test to obtain levels of statistical significance for multiple comparisons. For analysis of in vivo results, repeated-measures ANOVA was used, with Dunnett's test to adjust for the multiple comparisons made with baseline values. The incidence of arrhythmias during reperfusion between groups was compared by Fisher's exact test. For the superoxide generation studies, the unpaired t test was used. Values are expressed as mean±SEM. Statistical significance was achieved at a value of P<.05.
Plasma levels of estrogen in control dogs were near the detection limits of the assay (20±8 pg/mL), whereas animals that received estrogen had substantially higher levels (732±42 pg/mL).
Hemodynamic Profiles and RMBF
Brief ischemia followed by 120 minutes of reperfusion had no significant effect on mean arterial pressure, heart rate, LVEDP, dP/dt, and RPP. These parameters were also comparable in both placebo- and estrogen-treated animals throughout the whole study period (Table 1⇓). The observation that RPP, dP/dt, and LVEDP during coronary occlusion and reperfusion were not different suggests that myocardial oxygen consumption was comparable between groups. It is also worth noting that RMBF decreased equally in the region of ischemia in both groups of animals during LAD occlusion, suggesting that collateral flow to the ischemic region was not altered by estrogen treatment.
The incidence of arrhythmias observed during ischemia was not statistically different between placebo-treated (6 of 15) and estrogen-treated (4 of 12) groups. However, the incidences of reperfusion arrhythmias of all classifications were significantly different between groups (P<.01). In placebo dogs, 11 of 15 developed various types of ventricular arrhythmias: ventricular bigeminy progressed to VT in 2 placebo dogs and were counted as VT; in 3 other dogs, VT progressed to VF and counted as VF, and all died (leaving 12 placebo dogs for completion of groups 1 and 2 studies). Conversely, reperfusion arrhythmias were effectively prevented in estrogen-treated dogs except for 3 dogs (of 12) that developed premature ventricular beats (P<.01) (Table 2⇓).
Anterior Wall SS
SS was comparable between placebo- and estrogen-treated animals both before and during ischemia. During reperfusion, SS was severely depressed (≈30%) compared with preischemic values and did not recover up to 120 minutes into reperfusion. In estrogen-treated dogs, SS was maintained at preischemic values during the whole reperfusion period (Fig 1⇓).
Vasodilatory Responses to ACh and SNP In Vivo
Baseline LAD flow remained steady throughout the whole duration of the study. During preischemia, both LAD flow and percent increase in LAD flow over baseline in response to ACh were comparable between placebo- and estrogen-treated animals (Table 3⇓). In placebo dogs, vasodilation to ACh was significantly depressed during reperfusion. The percent increase in LAD flow in response to ACh dropped from 136±37% during preischemia to 60±6% at 15 minutes (P<.05) and remained significantly depressed up to 60 minutes into reperfusion (Fig 2⇓). In the estrogen-treated group, the vasodilatory responses to ACh were preserved during 120 minutes of reperfusion. The response to ACh in estradiol-treated dogs was 151±29% at 15 minutes of reperfusion. The responses to SNP were comparable in both treatment groups and were not significantly affected by ischemia or reperfusion (Table 3⇓).
End-Tidal Concentration of n-Pentane
Brief ischemia (15 minutes) and 5 minutes of reperfusion increased end-tidal n-pentane concentration in placebo animals from 9.1±1.9 to 41.6±13.0 ppb (P<.05). n-Pentane levels remained significantly elevated during early reperfusion and reached a peak at 30 minutes (51.6±17.9 ppb) (Fig 3⇓). There were no significant increases in n-pentane concentration in the estrogen-treated group during reperfusion.
In Vitro Vascular Reactivity Studies
The response of LAD segments to a single concentration of KCl (32 mmol/L) was not affected by ischemia and/or reperfusion and was not different between estrogen- and vehicle-treated groups. The mean tension developed was 2.4±0.3 and 2.9±0.3 g in segments from estradiol and placebo-treated animals, respectively. These responses were comparable to those of control segments obtained from nonischemic LCx (2.5±0.3 g in estradiol- and 2.4±0.3 g in placebo-treated animals).
In contrast, ischemia/reperfusion resulted in a significant attenuation of ACh-induced relaxation in LAD segments from placebo animals compared with LCx segments from the same animals (Fig 4⇓). The maximum percent relaxations in response to ACh were 80.4±15.8% and 24.0±4.0% (P<.05) in nonischemic and ischemic rings, respectively. Estrogen treatment had no effect on control rings but significantly improved endothelium-dependent relaxation to ACh (percent maximum relaxation, 42.1±4.1%) in ischemic segments. The responses to ACh in estradiol-treated ischemic segments, however, remained significantly (P<.05) attenuated compared with their nonischemic counterparts. The responses to SNP were not different between all groups studied.
Measurement of Nitrite and Nitrate Concentration
Estradiol-treated dogs had a baseline nitrite/nitrate value of 32.3±4.5 μmol/L, compared with 15.8±3.6 μmol/mL in placebo animals (P<.05). Reperfusion had no significant effect on nitrite/nitrate levels in placebo animals, which remained at preischemic levels at all time points studied. In the estradiol group, however, the nitrite/nitrate concentration was significantly increased above preischemic levels at 5 minutes into reperfusion (47.5±4.9 versus 32.3±4.5 μmol/L; P<.05) but returned to baseline levels at 15 and 60 minutes (Fig 5⇓).
Measurement of In Vitro Superoxide Generation
LCx segments from placebo dogs generated a significantly higher concentration of superoxide anion than those from estrogen-treated animals in response to hypoxia/reoxygenation. The mean increase in superoxide production above baseline value was 45.0±5.6% in placebo, compared with 20.4±7.1% in the estrogen group (P<.05) (Fig 6⇓).
Our present results clearly support the hypothesis that estrogen treatment prevents or attenuates the incidence of ventricular arrhythmias, endothelial dysfunction, and myocardial systolic dysfunction occurring during the early phase of reperfusion after brief ischemia (within 15 minutes). Such beneficial effects appear to be mediated, at least in part, through the antioxidant activity of estrogen.
Myocardial ischemia/reperfusion induces ventricular arrhythmias1 2 and is associated with myocardial stunning.1 Clinically, reperfusion arrhythmias and myocardial stunning are likely to occur in settings of myocardial infarction with early revascularization, rest and effort angina, coronary artery bypass surgery, and cardiac transplantation.1 33 Recent studies have demonstrated that a variety of agents with antioxidant properties protect the heart against reperfusion-induced VF and myocardial stunning.34 35 Oxygen free radicals generated during the early phase of reperfusion may damage membrane lipids or proteins, leading to arrhythmogenic conditions such as ionic disturbances in potassium or calcium.2 33 35 36 Similarly, free radical generation and/or disturbance in calcium homeostasis may also be involved in the pathogenesis of myocardial stunning.6 33 35
Brief ischemia/reperfusion can also induce temporary endothelial vasodilatory dysfunction.3 4 5 We have previously shown that a 10- to 20-minute period of ischemia/reperfusion causes “endothelial stunning” associated with impairment of endothelium-dependent vasodilation without evidence of endothelial damage.3 Similar results were reported in canine coronary artery segments subjected to longer periods of ischemia showing depressed endothelium-dependent but not endothelium-independent vasodilatory responses.37 A number of observations support a role of the oxygen free radicals generated during reperfusion in inducing endothelial dysfunction.5 6 7 In addition, drugs that scavenge and/or inhibit free radical generation effectively preserve endothelial function from reperfusion-related dysfunction.5 33 38 Free radicals can directly injure endothelial cell membrane and/or indirectly alter the production or biological activity of endothelium-derived vasoactive substances such as prostacyclin, adenosine, endothelium-derived relaxing factors, and endothelin.5 39
The present data clearly show that estrogen treatment significantly attenuates ischemia/reperfusion-induced ventricular arrhythmias and myocardial systolic dysfunction. In a similar study, attenuation of reperfusion arrhythmias was reported with the acute administration of the equine estrogen Premarin.36 Acute 17β-estradiol administration was also shown to reduce infarct size in rabbit myocardium subjected to 30 minutes of ischemia followed by 4 hours of reperfusion. This effect was specific, not observed with 17α-estradiol, and was not associated with an increase in myocardial blood flow or alteration in hemodynamics.40 Our data also demonstrate that estrogen pretreatment preserves endothelium-dependent vasodilation after ischemia/reperfusion both in vivo and ex vivo. An improvement of endothelium-dependent relaxation with estrogen treatment has also been demonstrated in atherosclerotic coronary arteries of ovariectomized monkeys41 42 and postmenopausal women.14 In the present study, protection of endothelium-dependent relaxation was similarly observed in the isolated arterial segment preparation. Ischemic LAD segments from estrogen-treated dogs relaxed significantly more in response to ACh than rings from the placebo group. This protection, however, was not complete, since the response of ischemic segments from estrogen-treated dogs remained significantly depressed compared with nonischemic rings. The discrepancy between in vivo and ex vivo results is interesting and may be physiologically significant. The isolated arterial segment preparation provides a model to evaluate the direct effect of estrogen on vascular reactivity independent of the other possible neuronal and/or humoral effects of the hormone. In the present model, the ex vivo data support a direct effect of the hormone on coronary vascular reactivity. However, the segments used ex vivo are from epicardial arteries and may not relate to the status of the whole coronary circulation. On the other hand, the in vivo measurement of LAD flow changes reflects the status of the microvasculature, which is responsible for a major portion of the total coronary vascular resistance, and may be more relevant to the response of the whole coronary vascular bed.
The exact mechanisms involved in the protective effect of estrogen are not yet clear. Recent studies have demonstrated a direct effect of estrogen on the metabolism of the l-arginine/NO pathway.17 Our observation of an elevated coronary venous nitrite/nitrate concentration in estrogen-treated animals is in agreement with these studies. Estrogen replacement therapy has similarly been shown to increase serum nitrate and nitrite levels in postmenopausal women,16 and estradiol treatment in guinea pigs increased the activity and mRNA expression of the cNOS in heart, kidney, and skeletal muscle.17 However, this effect of estrogen on cNOS, which requires a few days to be expressed, cannot account for the acute improvement in endothelium-dependent vasodilation observed as early as 15 minutes after exposure to estrogen in some studies.14 41 All these observations suggest the possible involvement of other mechanisms in mediating the effect of estradiol on coronary vascular reactivity.
The antioxidant properties of estrogen may also contribute to its protective effect against endothelial dysfunction associated with ischemia/reperfusion.13 43 Our data show that estrogen treatment decreases in vivo n-pentane output resulting from ischemia/reperfusion as well as superoxide anion production induced by hypoxia/reoxygenation in vitro. These results suggest an inhibitory effect of estrogen on oxygen free radical generation and lipid peroxidation. Oxygen free radicals can directly inactivate NO in the vascular wall and/or indirectly damage the endothelium by oxidizing LDL cholesterol.21 Estrogen has been shown to inhibit LDL oxidation in hypercholesterolemic swine13 and postmenopausal women.44 Estrogen may exert its antioxidant action directly by scavenging circulating free radicals18 19 or indirectly through activation of antioxidant enzymes45 and/or alteration of synthesis and release of endothelium-derived vasoactive mediators, such as prostacyclin and NO.14 17 46 In vitro studies have shown that various estrogens in micromolar concentrations scavenge hydroxyl radicals with a potency directly related to their chemical structure.18 19 It is unlikely that this mechanism is solely responsible for the protective effect of the hormone in the present model, in view of the plasma levels (picomolar) of estrogen achieved with our treatment protocol. More studies are needed to elucidate the mechanisms involved in mediating the antioxidant effect of estrogen in the present model.
Our conclusion that the beneficial effects of estrogen are mediated by its antioxidant activity is largely based on the premises that, in our model, we have produced a comparable degree of regional ischemia and that n-pentane measurements in exhaled gas accurately estimate lipid peroxidation resulting from oxygen free radicals generated during ischemia/reperfusion. Although the duration and anatomic site of coronary occlusion as well as appearance of cyanosis and ECG changes were identical between groups, it is extremely difficult to ensure a similar degree of ischemia among various groups, because dogs have substantial collaterals. However, in our study, the major parameters determining ischemia during coronary occlusion were carefully monitored. Collateral blood flow was estimated by measurement of RMBF, and RPP and LVEDP were also monitored to estimate oxygen consumption. All these parameters were comparable in all groups during ischemia and the early phase of reperfusion.
The in vivo detection of oxygen free radical generation during ischemia/reperfusion is extremely difficult, because of their highly reactive and ultra–short-lived nature.47 The generation of free radicals has been demonstrated by electron paramagnetic resonance spectroscopy.38 However, sample processing for this method is tedious and time-consuming, and the procedure is considered unsuitable for most types of in vivo studies. For this reason, a common approach to estimate oxygen free radical generation is to measure in vivo the end products of lipid peroxidation or to determine free radical production in vitro. Oxygen free radicals interact with membrane lipids, resulting in the generation of aldehydes, such as malonyldialdehyde, and volatile hydrocarbons, such as ethane and pentane.22 23 24 47 Gas chromatography has been used to monitor ethane and pentane in exhaled gases. A recent report, however, has shown that normal exhaled air contains a relatively high concentration of isoprene, which cannot be distinguished from pentane by gas chromatographic techniques.24 In the present study, we used a modified and highly sensitive (detection limit of 3 to 5 ppb) technique that allows the separation of pentane from isoprene in exhaled air. Pentane concentration was measured in the end-tidal gas, which contains the highest concentration of pentane. These values, however, were not corrected for body weight or minute ventilation and thus cannot be used directly to compare with absolute values obtained from other studies measuring pentane concentration in mixed expired breath. Our technique does not quantify total output of pentane. However, since tidal volume and minute ventilation were maintained at steady levels during the entire study period, the end-tidal pentane concentration should accurately reflect changes in the extent of lipid peroxidation.
In summary, we have shown that pretreatment with 17β-estradiol attenuates reperfusion arrhythmias and temporary dysfunction of coronary endothelium and myocardium associated with brief ischemia/reperfusion. Since these cardiovascular changes are related to a surge of free radicals, our data suggest that the protective effect of estrogen may be mediated by its antioxidant properties. The cardiovascular events related to ischemia/reperfusion are prevalent in patients with established CAD. In the postischemic heart, endothelial dysfunction may promote coronary spasm, leading to another bout of ischemia. Endothelial dysfunction could also induce adherence of blood cells such as platelets and neutrophils, resulting in thrombosis and vascular obstruction, and further release of harmful substances such as free radicals, vasoconstrictors, and inflammatory mediators, which may ultimately lead to myocardial infarction.5 Thus, the antioxidant effect of estrogen may account for a significant reduction in cardiovascular morbidity and mortality in CAD patients on estrogen replacement therapy. It is worth noting, however, that the plasma levels of estrogen in our study are supraphysiological, since they are higher than those achieved in postmenopausal women with estrogen replacement therapy.48 More studies will be needed to test whether the same degree of protection could be achieved in postmenopausal women with lower doses of estrogen.
Selected Abbreviations and Acronyms
|CAD||=||coronary artery disease|
|cNOS||=||constitutive NO synthase|
|LAD||=||left anterior descending coronary artery|
|LCx||=||left circumflex artery|
|LVEDP||=||LV end-diastolic filling pressure|
|RMBF||=||regional myocardial blood flow|
- Received April 26, 1996.
- Revision received July 1, 1996.
- Accepted July 5, 1996.
- Copyright © 1996 by American Heart Association
Hearse DJ. Myocardial injury during ischemia and reperfusion: concept and controversies. In: Yellon DM, Jennings RB, eds. Myocardial Protection: The Physiology of Reperfusion and Reperfusion Injury. New York, NY: Raven Press; 1992:13-33.
Kim YD, Fomsgaard JS, Heim KF, Ramwell PW, Thomas G, Kagan E, Moore SP, Coughlin SS, Kuwahara M, Analouei A. Brief ischemia-reperfusion induces stunning of endothelium in the canine coronary artery. Circulation. 1992;85:1473-1482.
Bolli R, Triana JF, Jeroudi MO. Prolonged impairment of coronary vasodilation after reversible ischemia: evidence for microvascular ‘stunning.’ Circ Res. 1990;67:332-343.
Mullane KM, Young M. The contribution of neutrophil activation and changes in endothelial function to myocardial ischemia-reperfusion injury. In: Yellon DM, Jennings RB, eds. Myocardial Protection: The Pathophysiology of Reperfusion and Reperfusion Injury. New York, NY: Raven Press; 1992:59-83.
Grady D, Rubin SM, Petitti DB, Fox CS, Black D, Ettinger B, Ernster VL, Cummings SR. Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann Intern Med. 1992;117:1016-1037.
Bush TL, Fried LP, Barrett-Connor E. Cholesterol, lipoproteins, and coronary heart disease in women. Clin Chem. 1988;34(8B):B60-B70.
Bush TL, Barrett-Connor E, Cowan LD, Criqui MH, Wallace RB, Suchindran CM, Tyroler HA, Rifkind BM. Cardiovascular mortality and noncontraceptive use of estrogen in women: results from the Lipid Research Clinics Program Follow-up Study. Circulation. 1987;75:1102-1109.
Keaney JF Jr, Shwaery GT, Xu A, Nicolosi RJ, Loscalzo J, Foxall TL, Vita JA. 17β-Estradiol preserves endothelial vasodilator function and limits low-density lipoprotein oxidation in hypercholesterolemic swine. Circulation. 1994;89:2251-2259.
Gilligan DM, Quyyumi AA, Cannon RO. Effects of physiological levels of estrogen on coronary vasomotor function on postmenopausal women. Circulation. 1994;89:2545-2551.
Collins P, Rosano GMC, Sarrel PM, Ulrich L, Adamopoulos S, Beale CM, McNeill JG, Poole-Wilson PA. 17β-Estradiol attenuates acetylcholine-induced coronary arterial constriction in women but not in men with coronary artery disease. Circulation. 1995;92:24-30.
Rosselli M, Imthurn B, Keller PJ, Jackson EK, Dubey RK. Circulating nitric oxide (nitrite/nitrate) levels in postmenopausal women substituted with 17β-estradiol and norethisterone acetate: a two year follow-up study. Hypertension. 1995;25:848-853.
Weiner CP, Lizasoain I, Baylis SA, Knowles RG, Charles IG, Moncada S. Induction of calcium-dependent nitric oxide synthases by sex hormones. Proc Natl Acad Sci U S A. 1994;94:5212-5216.
Simon BC, Cunningham LD, Cohen RA. Oxidized low-density lipoproteins cause contraction and inhibit endothelium-dependent relaxation in the pig coronary artery. J Clin Invest. 1990;86:75-79.
Heim KF, Makila UM, Leveson P, Ledley GS, Thomas G, Rackley C, Ramwell PW. Detection of pentane as a measurement of lipid peroxidation in humans using gas chromatography with a photoionization detector. In: Paubert-Braguet M, ed. Lipid Mediators in the Immunology of Shock. New York, NY: Plenum Publishing Co; 1987:103-108.
Theraux P, Franklin D, Ross J Jr, Kemper WS. Regional myocardial function during acute coronary artery occlusion and its modification by pharmacological agents in the dog. Circ Res. 1974;35:896-908.
Walker MJ, Curtis MJ, Hearse DJ, Campbell WF, Janse MJ, Yellon DM, Cobbe SM, Coker SJ, Harness JB, Harron DWG, Higgins AJ, Julian DG, Lab MJ, Manning AS, Northover BJ, Parratt JR, Reimersma RA, Riva E, Russell DC, Sheridan DJ, Winslow E, Woodward B. The Lambeth Conventions: guidelines for the study of arrhythmias in ischemia, infarction and reperfusion. Cardiovasc Res. 1988;22:447-455.
Kowallik P, Schulz R, Guth BD, Schade A, Paffhausen W, Gross R, Heusch G. Measurement of regional myocardial bloodflow with multiple colored microspheres. Circulation. 1991;83:974-982.
Bolli R, Zhu WX, Thornby JI, O'Neill PG, Roberts R. Time-course and determinants of recovery of function after reversible ischemia in conscious dogs. Am J Physiol. 1988;254:H102-H114.
Takashi H, Nakanishi T, Nishimura M, Tanaka H, Yoshimura M. Measurement of serum levels of nitrate ions in men and women: implications of endothelium-derived relaxing factor in blood pressure regulation of atherosclerosis. J Cardiovasc Pharmacol. 1992;20:S214-S216.
Fox KA. Reperfusion injury: a clinical perspective. In: Yellon DM, Jennings RB, eds. Myocardial Protection: The Pathophysiology of Reperfusion and Reperfusion Injury. New York, NY: Raven Press; 1992:151-163.
McHugh NA, Cooks M, Schairer JL, Bidgoli MM, Merrill GF. Ischemia and reperfusion-induced ventricular arrhythmias in dogs: effects of estrogen. Am J Physiol. 1995;268:H2569-H2573.
VanBenthuysen KM, McMurtey IF, Horwitz LD. Reperfusion after acute coronary occlusion in dogs impairs endothelium-dependent relaxation to acetylcholine and augments contractile reactivity in vitro. J Clin Invest. 1987;79:265-274.
Bolli R, Patel BS, Jeroudi MO, Lai EK, McCay PB. Demonstration of free radical generation in ‘stunned’ myocardium of intact dogs with the use of the spin trap alpha-phenyl N-tertbutyl nitrone. J Clin Invest. 1988;82:476-485.
Hale SL, Birnbaum Y, Kloner RA. 17β-Estradiol protects ischemic myocardium after occlusion and reperfusion in the rabbit. Presented at the American Heart Association Scientific Conference on Hormonal, Metabolic, and Cellular Influences on Cardiovascular Disease in Women; October 19-21, 1995; San Diego, Calif.
Williams JK, Adams MR, Klopfenstein HB. Estrogen modulates responses to atherosclerotic coronary arteries. Circulation. 1995;81:1680-1687.
Gerland M, Ganz P. How do we explain the clinical benefits of estrogen? From bedside to bench. Circulation. 1995;92:5-8.
O'Connell MB. Pharmacokinetics and pharmacologic variation between different estrogen products. J Clin Pharmacol. 1995;35:185-245.