Jump to
Abstract
Background— Carvedilol is a β-adrenoceptor blocker with a vasodilatory action that is more effective for the treatment of congestive heart failure than other β-blockers. Recently, carvedilol has been reported to reduce oxidative stress, which may consequently reduce the deactivation of adenosine-producing enzymes and increase cardiac adenosine levels. Therefore, carvedilol may also have a protective effect on ischemia and reperfusion injury, because adenosine mediates cardioprotection in ischemic hearts.
Methods and Results— In anesthetized dogs, the left anterior descending coronary artery was occluded for 90 minutes, followed by reperfusion for 6 hours. Carvedilol reduced the infarct size (15.0±2.8% versus 40.9±4.2% in controls), and this effect was completely reversed by the nonselective adenosine receptor antagonist 8-sulfophenyltheophylline (45.2±5.4%) or by an inhibitor of ecto-5′-nucleotidase (44.4±3.6%). There were no differences of either area at risk or collateral flow among the various groups. When the coronary perfusion pressure was reduced in other dogs so that coronary blood flow was decreased to 50% of the nonischemic level, carvedilol increased coronary blood flow (49.4±5.6 to 73.5±7.5 mL · 100 g−1 · min−1; P<0.05) and adenosine release (112.3±22.2 to 240.6±57.1 nmol/L; P<0.05) during coronary hypoperfusion. This increase of coronary blood flow was attenuated by either 8-sulfophenyltheophylline or superoxide dismutase. In human umbilical vein endothelial cells cultured with or without xanthine and xanthine oxidase, carvedilol caused an increase of ecto-5′-nucleotidase activity.
Conclusions— Carvedilol shows a cardioprotective effect against ischemia and/or reperfusion injury via adenosine-dependent mechanisms.
Received July 23, 2003; de novo received December 18, 2003; revision received February 18, 2004; accepted February 25, 2004.
Beta-adrenoceptor antagonists (β-blockers) are used for the treatment of ischemic heart disease because these drugs reduce adrenergic activity.1,2 Carvedilol is a β-blocker that has shown efficacy for chronic heart failure in several large-scale trials.3,4 Carvedilol decreases vascular resistance3 and improves the pathophysiology of chronic heart failure.5 This drug dilates both systemic and coronary vessels,6 which is not a typical characteristic of β-blockers. Although this vasodilatory action may contribute to the beneficial effects of carvedilol in ischemic or nonischemic heart failure,5 it may not be the primary mechanism of cardioprotection, because vasodilators are not always effective at protecting the heart.7 Interestingly, carvedilol can also reduce oxidative stress,8 which causes cellular damage through inactivation of membrane enzymes, pumps, and proteins, such as Na+/K+-ATPase,9 Ca2+ channels,10 and ecto-5′-nucleotidase.11 Ecto-5′-nucleotidase is the enzyme that produces adenosine, and adenosine is believed to ameliorate chronic heart failure or myocardial ischemia.12
To investigate the relationship between the cardioprotective effect of carvedilol and the reduction of oxidative stress on the enhancement of adenosine release, we examined whether carvedilol could reduce infarct size via adenosine- or ecto-5′-nucleotidase–dependent mechanisms in canine hearts. We also investigated whether carvedilol could increase coronary blood flow (CBF) via attenuation of oxidative stress and enhancement of adenosine release in ischemic canine hearts.
Methods
Instrumentation
We have previously reported the details of the instrumentation procedure.13 In brief, hybrid dogs (HBD) mated with the beagle, American fox hound, and Labrador retriever for laboratory use (weighing 15 to 21 kg; Kitayama Labes, Gifu, Japan) were anesthetized by an intravenous injection of sodium pentobarbital (30 mg/kg), intubated, and ventilated with room air mixed with oxygen (100% O2 at flow rate of 1.0 to 1.5 L/min). The arterial blood pH, Po2, and Pco2 before the protocol was begun were 7.38±0.02, 104±3 mm Hg, and 38.7±1.6 mm Hg, respectively. End-diastolic length (EDL) was determined at the R wave on the ECG, and end-systolic length (ESL) was determined at the minimum pressure differential. Then, fractional shortening (FS) was calculated as [(EDL−ESL)/EDL]×100%. Agents were administered into the left anterior descending coronary artery (LAD) via the bypass tube. To constitute the coronary bypass between the carotid artery and the LAD, <30 seconds interruption of the LAD was necessary, but this brief period of ischemia does not provoke either myocardial injury or protection. This study conformed to the Position of the American Heart Association on Research Animal Use adopted by the Association in November 1984.
Experimental Protocols
Protocol 1: Effects of Carvedilol on Adenosine Release and CBF in Nonischemic Myocardium
After hemodynamics became stable, coronary arterial and venous blood samples were obtained for the measurement of adenosine concentrations,11 and the difference between the adenosine levels in coronary arterial and venous blood [VAD(Ado)] was then calculated.
Five HBD dogs were used in protocol 1. Hemodynamic parameters (ie, systolic and diastolic aortic blood pressure and heart rate) were monitored. Carvedilol was infused at 1.5 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 0.09 mg/mL) for 10 minutes, and then coronary perfusion pressure (CPP), CBF, FS, and VAD(Ado) were measured. Carvedilol was dissolved in a small volume of DMSO (final concentration, <0.15%). In a preliminary study, this dose of carvedilol was shown to be the minimum dose that caused maximal coronary vasodilation in ischemic or nonischemic hearts. We also confirmed that this volume of DMSO did not change either coronary hemodynamics or VAD(Ado) in ischemic or nonischemic hearts.
Protocol 2: Effects of Carvedilol or Propranolol on Adenosine Release and CBF in Ischemic Hearts
After hemodynamics became stable, coronary arterial and venous blood samples were obtained for blood gas analysis and for measurement of adenosine11,13 and lactate14 levels. Lactate extraction ratio (LER) was calculated as the coronary arteriovenous difference of the lactate concentration multiplied by 100 and divided by the arterial lactate concentration.
Twenty HBD dogs were used in protocol 2. Hemodynamic parameters were monitored. To examine whether administration of carvedilol caused coronary vasodilation and reduced the severity of myocardial ischemia and whether adenosine-dependent mechanisms are involved in these actions, saline (n=5), α,β-methyleneadenosine diphosphate (AMP-CP) at 80 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 4.8 mg/mL, n=5), or 8-sulfophenyltheophylline (8-SPT) at 30 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 1.8 mg/mL, n=5) was infused into the bypass tube. AMP-CP is an inhibitor of ecto-5′-nucleotidase, whereas 8-SPT is a nonspecific adenosine receptor antagonist. Both agents were dissolved in saline before administration. After confirming that systemic and coronary hemodynamics were unchanged for 5 minutes after each drug infusion, CPP was reduced so that CBF decreased to 50% of the baseline level for 5 minutes. Then, infusion of carvedilol was started at 1.5 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 0.09 mg/mL) and was continued for 10 minutes, while CPP was maintained at the reduced level. A preliminary study showed that the above-mentioned dose of 8-SPT was the minimum dose that prevented coronary vasodilation induced by adenosine at 2 μg · kg−1 · min−1, whereas the dose of carvedilol (1.5 μg · kg−1 · min−1) was the minimum level that caused maximal coronary vasodilation.
In addition, propranolol was infused at 30 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 1.8 mg/mL) to investigate whether it had effects identical to those of carvedilol (n=5). This dose of propranolol corresponds to 15 μg/mL, and the effective dose of propranolol is >10 μg/mL, indicating that the dose of propranolol in the present study is sufficient to antagonize β-receptors of the hearts.
Protocol 3: Influence of the Antioxidant Activity of Carvedilol on CBF
To examine whether carvedilol eliminates oxidative stress and causes adenosine-dependent coronary vasodilation in ischemic hearts, either saline (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 1.5 mg/mL, n=5) or human recombinant superoxide dismutase (SOD) (5340 IU/mg, >99% purity) at 25 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 1.5 mg/mL, n=5) was infused into the bypass tube. CPP was then reduced so that CBF decreased to 50% of the baseline level for 5 minutes. Subsequently, infusion of carvedilol at 1.5 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 0.09 mg/mL) was initiated and continued for 10 minutes, while CPP was maintained at the reduced value. As a marker of oxidative stress, the 8-iso-prostaglandin F2α level was measured in coronary arterial and venous blood, and the arteriovenous difference of 8-iso-prostaglandin F2α [VAD(8-Iso-F2α)] was calculated. We confirmed that this dose of SOD had no effect on either systemic or coronary hemodynamic parameters.11
Protocol 4: Effects of Carvedilol on Infarct Size After 90 Minutes of Ischemia
In HBD dogs, the bypass tube to the LAD was occluded for 90 minutes, followed by reperfusion for 6 hours, together with administration of either saline (n=7, control) or DMSO (0.0167 mL · kg−1 · min−1, n=5) from 10 minutes before occlusion until 1 hour of reperfusion, except at the time of coronary occlusion. Hemodynamic parameters were monitored during myocardial ischemia and after the start of reperfusion. In the carvedilol group (n=5), carvedilol at 1.5 μg · kg−1 · min−1 (an infusion rate of 0.0167 mL · kg−1 · min−1 at a concentration of 0.09 mg/mL) was infused from 10 minutes before coronary occlusion until 60 minutes after the start of reperfusion, except during occlusion. In the carvedilol+8-SPT group (n=6) and the carvedilol+AMP-CP group (n=6), the effect of carvedilol was tested during concomitant administration of either 8-SPT at 30 μg · kg−1 · min−1 or AMP-CP at 80 μg · kg−1 · min−1. In the 8-SPT group (n=6) and the AMP-CP group (n=7), 90 minutes of ischemia and 6 hours of reperfusion were performed during treatments with 8-SPT and AMP-CP, respectively. Either 8-SPT or AMP-CP was infused from 10 minutes before coronary occlusion until 60 minutes after the start of reperfusion, except during occlusion. In all groups, infarct size was assessed after 6 hours of reperfusion.
Protocol 5: Effects of Carvedilol on 5′-Nucleotidase Activity
In human umbilical vein endothelial cells (HUVECs) cultured with or without xanthine (1×10−4 mol/L) and xanthine oxidase (1.6×10−3 U/mL), 5′-nucleotidase activity was measured by an enzyme assay after exposure to carvedilol (0, 1×10−8 to 1×10−5 mol/L) for 15 minutes.15
Analyses
The methods of measuring plasma adenosine levels,11 myocardial ecto-5′-nucleotidase activity,11,16 and plasma lactate levels13 have been reported previously.
Measurement of Infarct Size and Collateral Blood Flow
In protocol 4, the area of myocardial necrosis and the area at risk16 were measured in all of the dogs upon completion of the protocol by an operator who had no knowledge of the treatment given to each animal. Infarct size was expressed as a percentage of the area at risk.
Regional myocardial blood flow was determined as described previously.17 Nonradioactive microspheres (Sekisui Plastic Co) made of inert plastic were labeled with bromine. Microspheres were administered at 80 minutes after the start of coronary occlusion. The radio fluorescence of the stable heavy elements was measured with a wavelength dispersive spectrometer (PW 1480, Phillips Co). Because the level of energy emitted is characteristic of specific elements, it was possible to quantify the radio fluorescence of the heavy element with which the microspheres were labeled. Myocardial blood flow was calculated according to the following formula: time flow=(tissue count)×(reference flow)/(reference count), and was expressed in milliliters per minute per gram wet weight. Endomyocardial blood flow was measured at the inner half of the left ventricular wall.
Exclusion Criteria
To ensure that all of the animals used for analysis of infarct size in protocol 4 were healthy and were exposed to a similar extent of ischemia, the following standards were used for exclusion of unsatisfactory dogs: (1) subendocardial collateral blood flow >15 mL · 100 g−1 · min−1, (2) a heart rate >170 bpm, and (3) >2 consecutive attempts required to terminate ventricular fibrillation using low-energy DC pulses applied directly to the heart.
Statistical Analysis
Statistical analysis was performed by use of ANOVA18,19 to compare data among the groups. When ANOVA indicated a significant difference, paired data were compared by use of the Bonferroni test. Changes of the hemodynamic and metabolic parameters over time were assessed by ANOVA with repeated measures. Results were expressed as the mean±SEM, with a value of P<0.05 being considered significant.
Results
Effects of Carvedilol on VAD(Ado) in Nonischemic Myocardium
Neither systemic hemodynamic parameters (mean blood pressure, 101.0±2.1 versus 98.6±3.2 mm Hg and heart rate, 130.2±3.7 versus 128.0±3.3 bpm) nor FS (20.1±1.0% versus 21.5±1.0%) changed during the infusion of carvedilol. In contrast, CBF was increased (98.4±8.5 versus 112.6±9.6 mL · 100 g−1 · min−1, P<0.05), as was VAD(Ado) (40.9±4.0 versus 68.6±5.5 nmol/L, P<0.05).
Effects of Either Carvedilol or Propranolol on VAD(Ado) During Coronary Hypoperfusion
Administration of either 8-SPT or AMP-CP did not alter the systemic hemodynamics (mean blood pressure, 98.8±6.1 versus 101.8±5.8 mm Hg before and after 8-SPT and 99.0±3.0 versus 102.0±3.2 mm Hg before and after AMP-CP; heart rate, 132.2±6.9 versus 132.4±6.1 min−1 before and after 8-SPT and 131.8±4.6 versus 132.8±3.4 min−1 before and after AMP-CP) or the coronary hemodynamic and metabolic parameters (Figures 1 through 3⇓⇓). Before both CBF and CPP were reduced, there were no significant differences in hemodynamic and metabolic parameters among the 3 groups. In untreated dogs, administration of saline did not affect CPP, LER, or FS. However, addition of carvedilol increased VAD(Ado), CBF, LER, and FS, even in the constant low-CPP state, suggesting that myocardial ischemia was improved by carvedilol. These effects of carvedilol were blunted by administration of either 8-SPT or AMP-CP. Unlike carvedilol, an infusion of propranolol did not alter VAD(Ado), CBF, LER, or FS (Figures 1 through 3⇓⇓).
Figure 1. Effects of carvedilol on coronary hemodynamics in ischemic myocardium. A and B show CPP and CBF, respectively. Statistical analysis was performed by ANOVA followed by Bonferroni’s test.
Figure 2. Changes of difference in adenosine levels between coronary venous and arterial blood [VAD(Ado)] in ischemic myocardium. Carvedilol increased VAD(Ado), which was attenuated by an ecto-5′-nucleotidase inhibitor. Statistical analysis was performed by ANOVA followed by Bonferroni’s test.
Figure 3. Changes of FS (A) and LER (B) in ischemic myocardium. Statistical analysis was performed by ANOVA followed by Bonferroni’s test.
Reduction of Oxidative Stress and Beneficial Effect of Carvedilol in Ischemic Myocardium
In 5 dogs, reduction of CBF caused an increase of VAD(8-Iso-F2α), which was reduced by carvedilol (Figure 4A through 4C). Under these conditions, VAD(Ado) was increased by infusion of carvedilol (Figure 4D). In another 5 dogs, an infusion of SOD did not change either hemodynamic parameters or VAD(8-Iso-F2α) at nonischemic baseline conditions (Figure 4A through 4C). After the reduction of CBF to 50%, VAD(Ado) increased to the level seen in the presence of carvedilol without SOD (Figure 4D), whereas VAD(8-Iso-F2α) did not increase (Figure 4C). Addition of carvedilol did not further attenuate VAD(8-Iso-F2α) or increase VAD(Ado) (Figure 4C and 4D).
Figure 4. Changes of CPP (A), CBF (B), VAD(8-Iso-F2α) (C), and VAD(Ado) (D) in ischemic myocardium. Statistical significance was tested by ANOVA followed by Bonferroni’s test.
Effects of Carvedilol on Infarct Size
Seven of 64 dogs were excluded from analysis because their subendocardial collateral flow was >15 mL · 100 g−1 · min−1, so 57 dogs completed the protocol satisfactorily. Among these 57 dogs, 18 dogs developed ventricular fibrillation at least once, and ventricular fibrillation that matched the exclusion criteria occurred in 15 dogs, so these animals were also excluded from analysis. The numbers of the dogs that met the exclusion criteria of ventricular fibrillation were 2, 2, 0, 2, 3, 3, and 3 in the saline, the DMSO, the carvedilol, the carvedilol+8-SPT, the carvedilol+AMP-CP, the 8-SPT, and the AMP-CP groups, respectively.
Neither aortic blood pressure (≈104 mm Hg) nor heart rate (≈136 min−1) showed any differences among the 7 groups throughout the protocol. The Table shows the area at risk and the endocardial collateral blood flow in the LAD region during myocardial ischemia. There were no significant differences in the area at risk and collateral flow among the 7 groups during myocardial ischemia (Table). Figure 5 shows that carvedilol decreased infarct size compared with the control groups. This protective effect was completely blocked by either 8-SPT or AMP-CP, suggesting that the reduction of infarct size by carvedilol was attributable to an adenosine-dependent mechanism.
Area at Risk and Collateral Blood Flow During Myocardial Ischemia in Each Group
Figure 5. Infarct size as a percentage of area at risk. Infarct size was decreased in carvedilol group compared with control group, and this improvement was blocked by either 8-SPT or AMP-CP. Statistical significance was tested by ANOVA followed by Bonferroni’s test.
Effect of Carvedilol on Ecto-5′-Nucleotidase Activity in HUVECs
In HUVECs, carvedilol increased ecto-5′-nucleotidase activity by 35.4±8.4% (P<0.01) (Figure 6A). Exposure to xanthine and xanthine oxidase decreased ecto-5′-nucleotidase activity, whereas concomitant addition of carvedilol restored ecto-5′-nucleotidase activity to 104.9±8.7% of the baseline levels (P<0.01) (Figure 6B). Neither carvedilol nor xanthine and xanthine oxidase had any effect on cytosolic 5′-nucleotidase.
Figure 6. Ecto-5′-nucleotidase activity of HUVECs in presence or absence of carvedilol or xanthine and xanthine oxidase. Statistical significance was tested by ANOVA followed by Bonferroni’s test.
Discussion
In the present study, we demonstrated that carvedilol increases both adenosine release and CBF in ischemic and nonischemic hearts via reduction of oxidative stress and restoration of ecto-5′-nucleotidase activity. We also showed that carvedilol could limit infarct size and that this effect was attributable to the reduction of oxidative stress and an adenosine- or ecto-5′-nucleotidase–dependent mechanism. These findings suggested that the cardioprotective effect of carvedilol was attributable to an increase of adenosine in ischemic myocardium in addition to its β-blocking action, because propranolol did not mimic this effect.
Influence of Carvedilol on Adenosine Release in Ischemic Hearts
The β-adrenoreceptors in coronary smooth muscle are involved in coronary vasodilation, and their stimulation is thought to increase CBF via the relaxation of vascular smooth muscle and increased myocardial oxygen demand. Therefore, it may seem unusual that a β-blocker like carvedilol would cause coronary vasodilation. There are several possible explanations for the present findings. First, carvedilol itself may cause vasodilation separately from its β-blocking activity. Indeed, although carvedilol does not have a nitroxy moiety, its chemical structure predicts that the drug could also block α1-adrenoceptors,20 which would cause vasodilation. We cannot exclude this possibility, but the role of α1-adrenoceptor blockade in the vasodilatory effect of carvedilol seems likely to be minor, because we have previously reported that blockade of α1-adrenoceptors attenuates adenosine release in ischemic myocardium,21 whereas we found that carvedilol caused an increase of adenosine production. Second, carvedilol may increase vasodilatory substances such as NO or adenosine. We demonstrated that carvedilol could increase cardiac adenosine production independently of its β-blocking effect in the present study, because propranolol did not increase CBF under the same circumstances (Figures 1 through 3⇑⇑). Intriguingly, the carvedilol-induced increases in both adenosine release and coronary vasodilation were greater in ischemic heart than in nonischemic heart. There was a significant difference between the influence of carvedilol on coronary vasodilation under nonischemic and ischemic conditions in the present study, because the percent increases of CBF in nonischemic and ischemic myocardium were 14.4±1.1% and 50.6±10.1% (P<0.05), respectively. One possible explanation is that carvedilol may bind more tightly to β-adrenoreceptors under ischemic conditions than nonischemic conditions, and β-adrenoreceptors are also upregulated in the ischemic heart,22 which may enhance the adenosine-producing effect of carvedilol. Alternatively, even if carvedilol decreases coronary artery tone in nonischemic heart as well as ischemic heart, the activity of other endogenous vasodilators may decrease to maintain coronary autoregulation. Conversely, the effects of other vasodilators may already be maximal in ischemic hearts, so that carvedilol-induced adenosine release becomes a major determinant of coronary artery tone when adenosine-dependent coronary vasodilation is submaximal. A third possibility is that carvedilol may reduce the levels of substances that attenuate adenosine release and are increased in ischemic myocardium. Because carvedilol is reported to decrease oxidative stress and such stress reduces adenosine production, antioxidant activity of carvedilol may be involved in adenosine-dependent coronary vasodilation and cardioprotection. We showed such evidence in the present study.
In this context, several lines of evidence support the concept that adenosine can markedly attenuate ischemia/reperfusion injury,12,23 and we suggest that carvedilol-induced adenosine release is important for cardioprotection.
Mechanism of the Carvedilol-Induced Increase of Cardiac Adenosine
In ischemic hearts, carvedilol caused reduction of oxidative stress and increases in both adenosine release and CBF. Also, in HUVECs under oxidative stress, carvedilol restored ecto-5′-nucleotidase activity to the control level. These findings suggest that carvedilol may eliminate the factors that impaired ecto-5′-nucleotidase activity under ischemic conditions. Oxidative stress is one of these factors. Because oxygen-derived free radicals attenuate the ischemia-induced activation of ecto-5′-nucleotidase, elimination of oxidative stress may increase adenosine release in the ischemic myocardium. We observed that carvedilol could reduce oxidative stress, so this action may explain the present findings. Because ecto-5′-nucleotidase is susceptible to impairment by oxygen-derived free radicals, it is likely that the beneficial effect of carvedilol on myocardial ischemia in the present study was attributable to its antioxidant activity.
Clinical Relevance and Limitations
Carvedilol has been shown to be effective for treating heart failure.5 Its effective clinical dose is about 0.1 to 0.2 μg/mL, and the calculated cardiac concentration of carvedilol in the present study is ≈1 μg/mL. In dogs, carvedilol at 1 and 4 μg/mL decreased blood pressure by 9% and 32%, respectively (data not shown), suggesting that the concentration of 1 μg/mL of carvedilol in canine heats was comparable to a clinical dose of carvedilol. This difference may be also attributable to species differences, the route of administration of carvedilol, or conscious/anesthetic conditions.
The present study hinted that the mechanism by which carvedilol potently ameliorates heart failure, especially ischemic heart failure, may be related to adenosine.12 Carvedilol may have the ability to both antagonize β-adrenoceptors and increase adenosine release.
Tumor necrosis factor-α is inhibited by both carvedilol and adenosine24,25 and has been indicated to have a role in the pathology of congestive heart failure. Because the present study hints that the cardioprotection afforded by carvedilol is adenosine-dependent, it follows that the clinical effects of carvedilol may also be adenosine-dependent. If this hypothesis receives further validation, adenosine and potentiators of adenosine production or adenosine receptor agonists may become candidates for the treatment of heart failure.
Acknowledgments
This study was supported by Grants-in-aid for Scientific Research 12470153 and 12877107 from the Japanese Ministry of Education, Culture, Sports, Science, and Technology; a Health and Labor Sciences Research Grant for Human Genome, Tissue Engineering, and Food Biotechnology (H13-Genome-011); and a Health and Labor Sciences Research Grant for Comprehensive Research on Aging and Health (H13-21seiki(seikatsu)-23, H14Tokushitsu-38) from the Japanese Ministry of Health and Labor and Welfare. The authors gratefully acknowledge the technical assistance of Tomi Fukushima and Junko Yamada during the conduct of the experiments.
References
This Issue
Jump to
Article Tools
Share this Article
- β-Adrenoceptor Blocker Carvedilol Provides Cardioprotection via an Adenosine-Dependent Mechanism in Ischemic Canine HeartsPermalink:
