Protective Role of Nerve Growth Factor Against Postischemic Dysfunction of Sympathetic Coronary Innervation
Background Nerve growth factor (NGF) is produced rapidly in myocardium after brief myocardial ischemia. It contributes to the maintenance of neural integrity in several tissues. We examined the effect of exogenous and endogenous NGF on ischemia-induced dysfunction of cardiac sympathetic nerves.
Methods and Results In anesthetized dogs, bilateral stellate stimulation was performed, measuring changes in coronary vascular resistance (%ΔCVR) before and after release of either a 7- or 15-minute occlusion of the left anterior descending coronary artery (LAD). NGF (10 ng·kg−1·min−1, n=5) or vehicle (n=6) was infused into the LAD in dogs during a 15-minute LAD occlusion. In separate experiments, antibody to NGF (anti-NGF, 2 ng·kg−1·min−1, n=5) or vehicle (n=6) was infused into dogs during a 7-minute LAD occlusion. After release of a 15-minute LAD occlusion, attenuation of the coronary constriction to stellate stimulation was seen in the vehicle group (30±3% to 15±1% increase in CVR, P<.05); however, no such reduction was seen in the group receiving NGF. A 7-minute LAD occlusion with reperfusion did not alter %ΔCVR in the vehicle group (36±6% versus 37±7%, P=NS) but attenuated %ΔCVR in the anti-NGF group (39±8% to 17±2%, P<.05).
Conclusions We conclude that exogenously infused and endogenously released NGF protects against postischemic neural stunning of sympathetic cardiac innervation.
A brief period of myocardial ischemia is capable of producing transient dysfunction of the sympathetic and vagal cardiac innervation extending long into the reperfusion period.1 2 3 This “neural stunning” has important clinical implications, because it may produce regional autonomic imbalance, which could contribute to ventricular tachyarrhythmias or impair the favorable transmural distribution of MBF produced by sympathetic neural activation.4 5 Factors responsible for neural stunning have been examined. A study by Miyazaki and Zipes6 suggests that metabolites released from ischemic myocardium, such as adenosine, potassium, and hydrogen ions, affect neurotransmission in the ischemic area. We recently reported that adenosine produced during a brief period of coronary occlusion can attenuate sympathetic coronary constriction through a receptor-mediated mechanism.7 However, it is not known whether there are intrinsic mechanisms that may protect against neural dysfunction after myocardial ischemia.
NGF is a neurotrophic factor that is necessary for differentiation, survival, and regeneration of peripheral sympathetic and sensory nerves.8 Although it was shown recently that NGF is released acutely from the ischemic myocardium,9 its neural protective effect in this situation has not been examined. In a previous study, our laboratory showed that 15 minutes but not 7 minutes of coronary occlusion results in postischemic dysfunction of sympathetic cardiac nerves during reperfusion.1 To address the potential role of NGF in postischemic cardiac neural dysfunction, we examined two hypotheses: (1) Intracoronary infusion of exogenous NGF is capable of protecting against impairment of sympathetic coronary innervation after brief (15 minutes) myocardial ischemia. (2) Neutralizing endogenous NGF with intracoronary anti-NGF antibodies can unmask ischemic sympathetic dysfunction even after 7 minutes of coronary occlusion.
Twenty-two mongrel dogs of either sex weighing 23 to 30 kg were anesthetized with thiopental sodium (25 mg/kg IV). Anesthesia was maintained with α-chloralose (60 mg/kg). Supplemental injections of α-chloralose (20 mg/kg) were given as needed to suppress pressor responses to skin pinch and the corneal reflex. In each dog, muscle tone was flaccid, laryngeal reflexes were suppressed, and eye movements were not present, confirming a surgical level of anesthesia. All dogs were intubated and mechanically ventilated with continuous positive airway pressure at 4 cm H2O. Arterial Po2, Pco2, and pH were maintained within physiological ranges by adjustment of supplemental oxygen, ventilation rate, and tidal volume and by intravenous administration of sodium bicarbonate. Rectal temperature was maintained at 36°C to 37°C with a heating pad.
A polyethylene catheter was inserted into the left carotid artery and connected to a transducer (model P23XL, Viggo-Spectramed) to measure phasic and mean arterial blood pressure. The right femoral vein was cannulated for administration of fluids and pharmacological agents. Proximal and distal aortic catheters were also placed via the femoral arteries for microsphere withdrawal.
Through a ventral neck incision, both cervical vagi were isolated and transected. A left thoracotomy was performed by removal of the second through fifth left ribs. The stellate ganglia were isolated bilaterally, and each was secured within a bipolar microstimulation electrode and covered with gauze soaked in mineral oil. Stimulation parameters were 10-V, 5-ms pulses delivered at 15 Hz (10-second duration).
After a pericardial cradle was formed, a micromanometer-tipped catheter was placed into the left ventricle via the left appendage for the continuous recording of LV dP/dt. This signal was also directed to a cardiotachometer for recording heart rate. A 14-gauge polytetrafluoroethylene (Teflon) catheter was positioned in the left atrium for injection of microspheres. A 26-gauge Teflon catheter was secured retrogradely in each of the LAD and LCx arteries for administration of pharmacological agents. These catheters, which do not blunt coronary reactive hyperemic responses, were flushed with warmed saline and heparin.1
LV Regional Wall Function
%WTh was measured in the perfusion territory of the LAD with a pair of 7-MHz ultrasonic transit-time dimension gauge crystals. Sonomicrometer crystals were implanted in the center of the distal LAD bed where cyanosis was most prominent during a 15-second occlusion. One crystal was sutured on the epicardial surface, and the other was inserted at a 45° angle to a depth of 4 to 7 mm in the subendocardium and directly below the surface crystal. A processing unit (University of Iowa Bioengineering) was used to analyze the sonomicrometer signal. The distance between the crystals was monitored with an oscilloscope and continuously recorded on a chart recorder. At the end of the study, accurate placement of crystals was verified if Evans blue dye injected into the LAD distal to the occlusion site completely stained the excised myocardial segment containing the crystals. This was the case in each animal tested. Wall thickening was measured as the percent change in myocardial thickness from end diastole (measured at the time of onset of the increase in dP/dt) to end systole (measured 50 ms before the time of peak negative dP/dt).
Coronary Flow Velocity
Suction-attached Doppler flow probes designed at the University of Iowa10 11 were attached to the epicardium over both the LAD and the LCx distal to the sites of cannulation for measurements of coronary blood flow velocity. A 20-MHz piezoelectric crystal mounted at a 45° angle in a Silastic cupped housing was held to the epicardial surface by suction (4 mm Hg). This probe minimizes the potential risk of traumatic neural damage that may accompany surgical isolation of the coronary artery, which is necessary for placement of circumferential probes. Reproducible zero-flow velocities are regularly achieved with this probe.10
Radioactive microspheres 15 μm in diameter and labeled with 85Sr, 141Ce, 46Sc, 95Nb, or 113Sn were used to measure regional MBF. The spheres were agitated with a vortex mixer for 5 minutes immediately before use. More than 2×106 microspheres were injected into the left atrium over a 10- to 15-second period, and the catheter was flushed with 5 to 10 mL of warmed saline. Reference samples were obtained from both aortic cannulas with a constant-rate withdrawal pump (5.7 mL/min), beginning 5 seconds before the microsphere injection and continuing for 60 seconds.
At the end of each experiment, the LAD and LCx were ligated just proximal to the catheter tips. Ten milliliters of 2% Evans blue dye was injected via the coronary catheter into the LAD and 10 mL of 2% Bengal rose dye into the LCx to define the perfusion territory. The heart was excised and placed in buffered 10% formalin. Four days later, the free walls of the right ventricle, atria, great vessels, valves, epicardial vessels, and fat were removed. The left ventricle was divided into five to eight parallel rings 0.7 cm thick, and the apical piece was discarded. Each ring of tissue was radially divided into segments of similar sizes, ≈0.5 to 1 g each. Sections were segregated into two groups (LAD completely stained with Evans blue dye and LCx completely stained with Bengal rose dye) and weighed. Tissue samples that were partially stained or contained both stains were discarded. The tissue segments, reference blood samples, and pure isotopic controls were placed into scintillation tubes and analyzed via a sodium iodide gamma detector. MBF to the LAD and LCx regions (expressed as mL·min−1·100 g−1) was calculated, corrected for nuclide overlap with standard computer programs12 by the following equation: MBF=(Cm×Wr/Cr)×100 g, where Cm is tissue activity per gram, Wr is withdrawal rate of the pump, and Cr is total activity of the reference sample. This resulted in at least 500 spheres per sample even in ischemic areas. Adequate mixing of the microspheres was determined by comparison of the activity of the two arterial reference samples for a given nuclide. The activities of the two samples did not differ by >15%.
After instrumentation, propranolol was administered (2 mg/kg IV) and supplemented with doses of 1 mg/kg IV approximately every 2 hours throughout the experiment. This dose is sufficient to block the tachycardia to isoproterenol (3.5 μg IV).1
Effect of Exogenous NGF on Sympathetic Cardiac Innervation After 15 Minutes of LAD Occlusion
The protocol for studies of the effect of exogenous administration of NGF on postischemic dysfunction of cardiac sympathetic nerves is illustrated in Fig 1⇓. Baseline arterial pressure, heart rate, LV dP/dt, wall thickening, and coronary flow velocities in the LAD and LCx beds were determined. Two or three separate stellate stimulations were made, and resultant changes in coronary flow velocity and resistance were averaged. After administration of lidocaine (2%, 2 mL IV) and heparin (2000 U IV), the LAD was occluded just proximal to the catheter tip for 15 minutes by gentle epicardial compression with a cotton-tipped swab.1 Completeness of the occlusion was confirmed when coronary flow velocity measured with the distally placed Doppler flow probe was abolished. Just before release of the LAD occlusion, an additional 1 mL of lidocaine was given.
In 5 dogs, NGF (2.5 S NGF, 20 ng·kg−1·min−1) was infused into the LAD for 15 minutes, and saline was infused into the LCx control bed (NGF group). The infusion was stopped for 1 minute while stellate stimulations were performed and then resumed for an additional 30 minutes, beginning 5 minutes before and ending 10 minutes after a 15-minute LAD occlusion. Fifteen minutes into reperfusion, when hemodynamics had returned to baseline, the response to stellate stimulation was again measured. Measurements were repeated every 15 minutes until 90 minutes after reperfusion. MBF with radiolabeled microspheres was measured before and during the LAD occlusion and 15 minutes after reperfusion. In another 6 dogs, vehicle (BSA 4 mg in 0.5 mL PBS) was administered into the LAD instead of NGF (vehicle group).
Effect of Antagonizing Endogenous NGF on Sympathetic Cardiac Innervation After a 7-Minute LAD Occlusion
In another two groups of dogs, the effect of antagonizing endogenous NGF on neural function after a shorter period (7 minutes) of myocardial ischemia was studied in an experimental model similar to that described for NGF above (Fig 1⇑). Seven minutes of coronary occlusion was chosen because previous experiments show that ischemia of this duration is not associated with sympathetic dysfunction.1 In 5 dogs, anti-NGF (2 ng·kg−1·min−1) was administered into the LAD. This monoclonal antibody (anti-mouse-β [2.5 S] NGF, Boehringer Mannheim) is derived from mouse-mouse hybrid cells and is specific for the β-subunit of both the 2.5 S and 7 S forms of NGF from the mouse, rat, and cow. After a 15-minute infusion, stellate stimulation was tested. Then anti-NGF was resumed for another 22 minutes, beginning 5 minutes before and ending 10 minutes after a 7-minute LAD occlusion (anti-NGF group). The response to stellate stimulation was determined before occlusion and after reperfusion, as described above. In another 6 dogs, saline was administered into the LAD instead of anti-NGF (vehicle group).
Murine NGF and murine anti-NGF were obtained from Boehringer Mannheim. NGF was dissolved in saline with 4 mg BSA and 0.5 mL PBS. Anti-NGF was dissolved in saline. Radiolabeled microspheres were obtained from DuPont Co. Propranolol, chloralose, lidocaine, Evans blue dye, and Bengal rose dye were obtained from Sigma Chemical Co.
Criteria for Acceptable Study
All animals included in the data analysis met the following criteria: mean arterial pressure of >65 mm Hg, arterial pH of 7.35 to 7.45, Pco2 of 30 to 40 mm Hg, Po2 of >70 mm Hg, LAD and LCx flow signals with a signal-to-noise ratio of >20, reactive hyperemia to a 10-second coronary occlusion of >3.0:1, and reproducible hemodynamic and coronary responses to pharmacological agents and stellate stimulations. Three dogs were excluded from analysis because they died of ventricular fibrillation during coronary occlusion or reperfusion (2 dogs in the NGF group and 1 in the anti-NGF group).
Analyses of coronary flow velocity and CVR were made by comparison of changes from baseline in these parameters rather than absolute measurements, because changes in coronary flow velocity correlate excellently with changes in absolute flow in dogs.10 11 %ΔCVRs were calculated by the following formula: %ΔCVR=[(P2×F1)/(P1×F2)−1]×100, where P1 is baseline arterial pressure, P2 is arterial pressure at the time of the maximum decrease in coronary flow, F1 is baseline blood flow velocity, and F2 is blood flow velocity measured simultaneously with P2. Values from two or three stimulations were averaged for analysis.
Differences between baseline hemodynamics and MBF during control conditions and after interventions were detected with Student's two-tailed t test. When multiple measurements were made over time, a repeated-measures ANOVA was used. Post hoc determination of significant differences was tested with the Student-Newman-Keuls multiple comparison test. Statistical significance is defined as P<.05. Data are presented as mean±SEM.
Effect of Exogenous NGF on Sympathetic Cardiac Innervation After a 15-Minute LAD Occlusion
After β-adrenergic blockade, electrical stimulation of the bilateral stellate ganglia produced minimal change in heart rate and arterial pressure and an early and transient decrease in coronary flow velocity accompanied by an increase in CVR in both the LAD and LCx coronary beds (Fig 2⇓). Coronary vasoconstriction to stellate stimulation in the LAD bed was not affected during administration of NGF (peak increase in coronary resistance, 32±3% before versus 32±2% during NGF) or vehicle (30±3% versus 32±6%) (Fig 3⇓). In dogs with vehicle infusion, neurogenic vasoconstriction was markedly reduced in the LAD bed after release of a 15-minute LAD occlusion (15±1% 15 minutes after reperfusion, P<.05 versus before) and did not recover until 75 minutes after reperfusion. In contrast, neurogenic vasoconstriction was not attenuated after release of a 15-minute occlusion (40±4% at 15 minutes after reperfusion, P=NS versus before) in dogs treated with NGF. Neurogenic vasoconstriction was greater in the NGF group than in the vehicle group from 15 minutes to 60 minutes after reperfusion. Throughout the experiment, vasoconstriction to stellate stimulation in the LCx bed was not affected in either group. These data demonstrate that exogenously infused NGF prevents the attenuation of neurogenic coronary vasoconstriction during reperfusion.
Baseline hemodynamics and cardiac contractile function were not different between groups. After release of a 15-minute LAD occlusion, aortic pressure, LV dP/dt, and %WTh decreased significantly and similarly in both NGF and vehicle groups (Table 1⇓). The depression in LV dP/dt and %WTh (myocardial stunning) was maintained throughout the experiment in both groups. MBF in the LAD bed was reduced during the LAD occlusion to a similar degree in both groups and recovered during reperfusion. Thus, during reperfusion, NGF selectively preserved cardiac sympathetic function, whereas myocardial function remained depressed.
Effect of Antagonizing Endogenous NGF With Anti-NGF on Cardiac Innervation After a 7-Minute LAD Occlusion
Coronary vasoconstriction to stellate stimulation in the LAD bed was not altered by intracoronary infusion of anti-NGF (39±8% increase in CVR before versus 37±6% increase during anti-NGF) or vehicle (from 36±6% to 37±6%) (Fig 4⇓). In the vehicle group, 7 minutes of LAD occlusion and reperfusion did not affect neurogenic vasoconstriction in the LAD bed (37±7% increase in CVR at 15 minutes after reperfusion). In contrast, in the anti-NGF group, 7 minutes of LAD occlusion and reperfusion significantly attenuated neurogenic vasoconstriction in the LAD bed (17±2% increase in CVR 15 minutes after reperfusion; P<.05 versus before and versus vehicle group). The attenuation recovered gradually beginning 45 minutes after reperfusion. Vasoconstriction to stellate stimulation in the LCx bed was not altered in either group. These data indicate that neutralizing endogenous NGF with anti-NGF is capable of impairing neurogenic vasoconstriction even during a shorter period (7 minutes) of coronary occlusion and reperfusion.
Baseline heart rate, aortic pressure, LV dP/dt, and %WTh were not different between groups (Table 2⇓). After release of a 7-minute LAD occlusion, aortic pressure, LV dP/dt, and %WTh were reduced significantly but similarly between the groups. MBF decreased markedly during the LAD occlusion in both treated and vehicle groups and recovered after reperfusion (Table 3⇓).
A brief period of myocardial ischemia is capable of impairing cardiac sympathetic innervation.1 13 NGF is a known neurotrophic factor for peripheral sympathetic nerves.8 Although NGF was recently shown to be produced rapidly during brief myocardial ischemia and reperfusion,9 its direct effect on sympathetic cardiac innervation is not known. In this study, we examined the protective effect of NGF on the coronary innervation after brief myocardial ischemia. The major findings are that (1) intracoronary infusion of exogenous NGF protects against postischemic neural stunning of sympathetic cardiac nerves during brief (15 minutes) myocardial ischemia and reperfusion and (2) neutralization of endogenous NGF by anti-NGF is capable of unmasking ischemia-induced attenuation of sympathetic cardiac innervation even after a shorter period (7 minutes) of myocardial ischemia. In the vehicle group as well as in a previous study,1 coronary vasoconstriction to stellate stimulation was not reduced after 7 minutes of coronary occlusion and reperfusion.
NGF and Cardiac Innervation
NGF exists mainly in peripheral sympathetic and sensory nerves and in tissues receiving innervation from those nerves, including the cardiac atrium and ventricle14 and coronary artery.15 It has been shown that NGF is necessary for the survival of sympathetic ganglionic neurons, which may innervate the heart, whereas other neurotrophic factors, such as brain-derived neurotrophic factor, are incapable of maintaining cell survival.16
Basal concentrations of NGF in the cardiac ventricle are relatively low.14 However, a recent preliminary report indicates that NGF is produced rapidly in the ischemic myocardium during brief ischemia and reperfusion.9 In the present study, neurogenic coronary constriction was reduced in the vehicle group but was maintained in the NGF group after 15 minutes of coronary occlusion and reperfusion. Taken together, these data suggest that the amount of endogenous NGF released is not sufficient alone for complete neural protection but when supplemented with exogenous NGF protection against neural dysfunction, is seen after 15 minutes of coronary occlusion and reperfusion. However, after a 7-minute coronary occlusion, endogenous NGF does preserve sympathetic function, because anti-NGF can reduce sympathetic coronary constriction in this situation.
Neural Protective Effect of NGF
The neural protective effect of NGF has been reported in other tissues. NGF improves survival of PC12 cells after ischemia.17 Intraocular treatment with NGF enhances the survival and functional recovery of retinal ganglion cells after 60 minutes of ischemia in cats.18 Protective roles of exogenous NGF on sympathetic nerves have also been reported against oxidant toxicity,19 20 diabetic neuropathy,21 and cisplatin neuropathy.22 Since the causes of neural dysfunction in these conditions are diverse, NGF appears to act at a yet unknown site common to the pathogenic alterations in each of these conditions. In this regard, it is interesting to speculate that the protective effect of NGF may involve the reduction of free-radical formation. In cultured sympathetic neurons, cell death occurs earlier in the absence of NGF and is associated with an increase in reactive oxygen species formation.23 A preliminary study suggests that NGF can enhance endogenous antioxidant enzyme activity in brain tissue.24 These data provide support for an interaction between neurotrophic factors and antioxidant injury during myocardial ischemic, a topic for future investigation.
Mechanism of Cardiac Neural Stunning
Impaired axonal conduction is a likely mechanism of neural stunning during reperfusion. Our previous studies showed that coronary vasoconstriction induced by norepinephrine and that induced by junctional release of norepinephrine by bretylium and tyramine are not altered after 15 minutes of coronary occlusion and reperfusion.1 In an elegant study by Miyazaki and Zipes,6 ischemia or local administration of high concentrations of adenosine alone or with other metabolic products into a diagonal branch of the LAD was capable of reducing the sympathetically induced shortening of the effective refractory period at sites in the myocardium distal to the ischemic or perfused area. The present result that endogenous and exogenous NGF protect against ischemia-induced neural stunning is consistent with the idea of impaired axonal conduction. Exogenous NGF or that released endogenously during coronary occlusion may be taken up through NGF receptors and transported retrogradely to nerve axons to reduce neural stunning.
Possible Mechanism of NGF-Induced Cardiac Neural Protection
Anti-NGF antibody neutralizes the biological activity of NGF very efficiently and selectively both in vitro and in vivo.25 It has been reported that development of the cardiac sympathetic innervation26 in the newborn rat is retarded by anti-NGF. In the present work, we studied the protective role of endogenous NGF in neural stunning of heart. Coronary constriction to stellate stimulation was not changed in the vehicle group after 7 minutes of coronary occlusion and reperfusion, whereas administration of anti-NGF attenuated this response. These data suggest that even after shorter periods (7 minutes) of coronary occlusion and reperfusion, endogenous NGF exerts protective action against neural stunning, because neutralizing endogenous NGF with anti-NGF reduces dysfunction after ischemia.
The effects of NGF and anti-NGF on ischemia-induced neural stunning were not due to changes in the degree of myocardial ischemia. Hemodynamics, myocardial contractile function, and MBF were not different among any of the groups (NGF, anti-NGF, and vehicle). After reperfusion following a 15-minute LAD occlusion, LV dP/dt and %WTh were reduced similarly in both NGF and vehicle groups. Thus, the extent of myocardial stunning was similar. However, the time course of neurogenic coronary vasoconstriction differed between groups. Sympathetic vasoconstriction was attenuated in the vehicle group but not in the NGF group.
NGF increases the number of functional sodium channels through cAMP-dependent protein kinase27 and induces tetrodotoxin-resistant sodium channels.28 These phenomena might contribute to neural protection against ischemia, because voltage-dependent sodium channels play an important role in action potential propagation.29
Activation of protein kinase N may contribute to the beneficial effect of NGF during myocardial ischemia. Protein kinase N, an NGF-activated serine protein kinase, is inhibited by purine analogues, including adenosine.30 We recently reported that adenosine reduces cardiac neurotransmission1 and that endogenously released adenosine is necessary for eliciting neural cardiac stunning.7 This suggests a potential mechanism by which adenosine-induced neural stunning involves NGF. Thus, the neural protection afforded by NGF may be indirectly mediated by the actions of adenosine on serine protein kinase.
Potential Experimental Concerns
Interpretation of our results depends on several methodological considerations. Since stellate stimulations were always performed sequentially over time, it is possible that the attenuation in neurogenic coronary constriction represents a time-related phenomenon rather than an effect of ischemia. This is not likely, for several reasons. First, in the protocols that use anti-NGF, no time-dependent decrease in responsiveness was observed. Second, the coronary vasoconstriction observed in the LCx coronary bed simultaneously with the measured constriction in the LAD bed was consistent throughout the course of the study. Third, in three experiments we studied the effect of NGF without intervening ischemia on the responsiveness in the LAD bed to successive sympathetic stimulations. No effect was observed, suggesting that NGF itself did not act directly to alter sympathetic reactivity in the coronary circulation.
One limitation of this study is the use of stellate stimulation in place of a more physiological stimulus for increasing CVR, such as carotid sinus unloading. The reason for direct stellate stimulation is threefold. First, stellate stimulation elicits a reproducible coronary vasoconstriction, allowing for easier comparison with previous studies that used a similar technique.1 3 31 32 Second, in response to stellate stimulation, the peak coronary constriction occurs within a few seconds of the onset of this stimulus. In contrast, most reflex coronary constrictor stimuli produce peak or steady-state effects much later, when significant changes in arterial pressure and myocardial metabolism would be expected to exert greater influence on coronary resistance measurements. Third, coronary constriction to peripheral sympathetic stimulation is not significantly influenced by central inhibitory projections, vagal modulation of neural impulses, or competition from other reflexes that may be activated during physiological forms of coronary constriction and during myocardial ischemia.
In this study, myocardial ischemia was produced by epicardial coronary occlusion with a cotton-tipped swab. Although this method was chosen because it is relatively atraumatic, pressure from the swab could potentially damage nerve function. We believe that this is unlikely, for several reasons. First, in a previous study1 we demonstrated that using a balloon catheter for intracoronary occlusion produced results similar to that seen with epicardial pressure. Second, if the cotton-tipped applicator had directly damaged the sympathetic nerves, NGF should not have improved neurogenic coronary constriction after 15 minutes of occlusion.
A limitation of this study is the use of an epicardial flow velocity probe without concurrent measurements of epicardial coronary diameter. Changes in epicardial conduit artery diameter due to sympathetic stimulation may alter flow velocity independent of volumetric flow. For two reasons, however, we do not believe that changes in conduit coronary diameter influenced the interpretation of our results. First, the epicardial suction-attached Doppler probe has been validated against electromagnetic flow probes and timed venous collections in anesthetized dogs, demonstrating excellent correlation between changes in flow velocity and absolute changes in flow over a wide range of flow.10 Second, sympathetic activation constricts epicardial coronary arteries.33 If neural stunning had impaired epicardial coronary constriction, this would have led to an augmentation rather than an attenuation of the observed decrease in flow velocity to sympathetic stimulation after reperfusion. Thus, it is most likely that changes in MBF rather than epicardial diameter contribute primarily to the observed changes in coronary flow velocity in this study.
Although our data are consistent with a specific effect of NGF in limiting the degree of neural cardiac stunning, we cannot exclude a nonspecific effect of growth factors on proteins in general. The vehicle we used consisted only of the carrier solvent for NGF and anti-NGF to exclude an effect of constant perfusion through the occluded bed during myocardial ischemia on neural and myocardial function. Because these two proteins had directionally opposite effects on neural dysfunction, simply the presence of protein in the vehicle is not a likely explanation for the beneficial effect of NGF. With only one dose of NGF used, a dose-response relationship cannot be established. However, the dose we used was effective and would be expected to produce similar or lower tissue concentrations compared with those reported in rat heart.34 The dose of 20 ng·kg−1·min−1 IC NGF falls into what is calculated as the high physiological/low pharmacological range based on the assumptions of uniform intravascular distribution of NGF, lack of endothelial uptake or metabolism, and uniform distribution throughout the myocardium. Because of these assumptions, it is possible that the dose used may not apply to physiological conditions.
After 7 minutes of coronary occlusion and reperfusion in the vehicle group, the impaired regional ventricular function was milder than after 15 minutes of coronary occlusion, although neurogenic coronary constriction was not impaired. In contrast, neural stunning was noticed after 7 minutes of coronary occlusion and reperfusion in the group treated with anti-NGF. These findings provide indirect evidence that the mechanisms involved in neural stunning are different from those leading to myocardial stunning. The present data suggest that in the doses used, the effects of NGF and anti-NGF during myocardial ischemia and reperfusion operate specifically on the ischemia-induced dysfunction of the cardiac innervation.
The sonomicrometer crystals were positioned transmurally but did not interrogate function in the most endocardial regions of the heart. However, significant ischemic dysfunction was noted in the outer two thirds of the LV wall. Because ventricular dysfunction is most severe in the subendocardial region, it is possible that we underestimated the magnitude of myocardial stunning and regional dyskinesis.
All animals are treated with propranolol to minimize myocardial metabolic stimulation and the influence of β-adrenergic receptors on the coronary vessels during sympathetic stimulation. However, propranolol has local anesthetic properties that alone cannot account for the observed changes in sympathetic coronary reactivity but if acting in concert with ischemia-induced suppression of sympathetic nerve function may have contributed to some of the observed impairment of sympathetic function.
Significance of Findings
Neural stunning may have important clinical ramifications. Under normal conditions, sympathetic activation produces a favorable redistribution of coronary flow through transmural layers of the myocardium.4 5 35 36 In the presence of a coronary stenosis, however, sympathetic activation can produce myocardial ischemia with attendant lactate production and myocardial dysfunction.37 38 This is a direct effect of α-adrenergic constriction. In humans, reflex sympathetic activation can elicit increases in CVR, with resultant angina pectoris and myocardial ischemia.39 Reflex responses can also produce primary vasoconstriction of the coronary stenosis site in humans with epicardial coronary disease.40 Sympathetic neural dysfunction may confer beneficial effects in the postischemic myocardium by attenuating excess sympathetic drive to the heart, thereby reducing metabolic demand and neurogenic coronary vasoconstriction. Thus, it would appear that in the presence of significant coronary stenoses, neural stunning may be beneficial, preventing further reductions in perfusion to the myocardium at risk and reducing the associated increases in metabolic demand that would put additional stress on the stunned myocardium.
In contrast to these beneficial changes, postischemic neural dysfunction may produce adverse effects. Regional myocardial ischemia may produce a heterogeneity in cardiac autonomic innervation that could predispose to arrhythmias. Furthermore, neural stunning of the cardiac afferent nerves13 may reduce pain perception, contributing to silent myocardial ischemia. Administration of NGF or agents including 4-methylcatechol,38 a stimulator of NGF synthesis, may be useful in preventing postischemic sympathetic cardiac dysfunction. It will be important to test this hypothesis using a more physiological method of eliciting coronary constriction, such as exercise or baroreceptor deactivation.
In conclusion, NGF protects against postischemic neural stunning of sympathetic cardiac nerves. Administration of anti-NGF is capable of uncovering ischemia-induced attenuation of sympathetic cardiac innervation even after a shorter period (7 minutes) of myocardial ischemia. NGF is important for sustained cardiac neural function during myocardial ischemia.
Selected Abbreviations and Acronyms
|anti-NGF||=||murine antibody to β-subunit of NGF|
|CVR||=||coronary vascular resistance|
|%ΔCVR||=||percent change in CVR|
|LAD||=||left anterior descending coronary artery|
|LCx||=||left circumflex artery|
|MBF||=||myocardial blood flow, myocardial perfusion|
|NGF||=||nerve growth factor|
|%WTh||=||percent regional wall thickening|
This study was supported by a Fellowship Grant from the American Heart Association (AHA), Iowa Affiliate, and by a National AHA Grant-in-Aid. Dr Gutterman is the recipient of an AHA Established Investigator Award. We wish to acknowledge the expert technical assistance of Trung Le and the secretarial help provided by Marlene Blakley.
Reprint requests to David D. Gutterman, MD, Associate Professor of Medicine, University of Iowa College of Medicine, Iowa City, IA 52242. E-mail firstname.lastname@example.org.
- Received May 6, 1996.
- Revision received August 8, 1996.
- Accepted August 19, 1996.
- Copyright © 1997 by American Heart Association
Gutterman DD, Morgan DA, Miller FJ. Effect of brief myocardial ischemia on sympathetic coronary vasconstriction. Circ Res. 1992;71:960-969.
Inoue H, Zipes DP. Time course of denervation of efferent sympathetic and vagal nerves after occlusion of the coronary artery in the canine heart. Circ Res. 1988;62:1111-1120.
Ciuffo AA, Ouyang P, Becker LC, Levin L, Weisfeldt ML. Reduction of sympathetic inotropic response after ischemia in dogs. J Clin Invest. 1985;75:1504-1509.
Nathan HJ, Feigl EO. Adrenergic vasoconstriction lessens transmural steal during coronary hypoperfusion. Am J Physiol. 1986;250:H645-H653.
Huang AH, Feigl EO. Adrenergic coronary vasoconstriction helps maintain uniform transmural blood flow distribution during exercise. Circ Res. 1988;62:286-298.
Miyazaki T, Zipes DP. Presynaptic modulation of efferent sympathetic and vagal neurotransmission in the canine heart by hypoxia, high K+, low pH, and adenosine. Circ Res. 1990;66:289-301.
Pettersen MD, Abe T, Morgan DA, Gutterman DD. Role of adenosine in postischemic dysfunction of coronary innervation. Circ Res. 1995;76:95-101.
Thoenen H, Barde YA. Physiology of nerve growth factor. Physiol Rev. 1980;60:1284-1335.
Gu Y, Yong Q, Yoshida K, Kontos MC, Kukreja RC. Rapid increase in NGFI-A mRNA after ischemia/reperfusion and preconditioning in rat heart. Circulation. 1994;90(suppl I):I-249. Abstract.
Marcus ML, Wright CB, Doty DB, Eastham CL, Laughlin DE, Krumm P, Fastenow CF, Brody MJ. Measurements of coronary velocity and reactive hyperemia in the coronary circulation of humans. Circ Res. 1981;49:877-891.
Wangler RD, Peters KG, Laughlin DE, Tomanek RJ, Marcus ML. A method for continuously assessing coronary blood flow velocity in the rat. Am J Physiol. 1981;241:H816-H820.
Pitt B, Shea MJ, Romson JL, Lucchesi BR. Prostaglandins and prostaglandin inhibitors in ischemic heart disease. Ann Intern Med. 1983;99:83-92.
Inoue H, Skale BT, Zipes DP. Effects of ischemia on cardiac afferent sympathetic and vagal reflexes in dog. Am J Physiol. 1988;255:H26-H35.
Korsching S, Thoenen H. Nerve growth factor in sympathetic ganglia and corresponding target organs of the rat: correlation with density of sympathetic innervation. Proc Natl Acad Sci U S A. 1983;80:3513-3516.
Scarisbrick IA, Jones EG, Isackson PJ. Coexpression of mRNAs for NGF, BDNF, and NT-3 in the cardiovascular system of the pre- and postnatal rat. J Neurosci. 1993;13:875-893.
Boniece ER, Wagner JA. Growth factors protect PC12 cells against ischemia by a mechanism that is independent of PKA, PKC, and protein synthesis. J Neurosci. 1993;13:4220-4228.
Siliprandi R, Canella R, Carmignoto G. Nerve growth factor promotes functional recovery of retinal ganglion cells after ischemia. Invest Ophthalmol Vis Sci. 1993;34:3232-3245.
Lovell MA, Markesbery WR, Mattson MP. Neurotrophic factors stimulate antioxidant enzyme activities, and attenuate glutamate-induced free radical accumulation and neurotoxicity. Soc Neurosci. 1994;20:611. Abstract.
Cattaneo A, Rapposelli B, Calissano P. Three distinct types of monoclonal antibodies after long-term immunization of rats with mouse nerve growth factor. J Neurophysiol. 1988;50:1003-1010.
Malfatto G, Rosen TS, Steinberg SF, Ursell PC, Sun LS, Daniel S, Danilo P Jr, Rosen MR. Sympathetic neural modulation of cardiac impulse initiation and repolarization in the newborn rat. Circ Res. 1990;66:427-437.
Kalman D, Wong B, Horvai AE, Cline JJ, O'Lague PH. Nerve growth factor acts through cAMP-dependent protein kinase to increase the number of sodium channels in PC12 cells. Neuron. 1990;2:355-366.
Rudy B, Kirschenbaum B, Rukenstein A, Greene LA. Nerve growth factor increases the number of functional Na channels and induces TTX-resistant Na channels in PC12 pheochromocytoma cells. J Neurosci. 1987;7:1613-1625.
Berliner JA, Navab M, Fogelman AM, Frank JS, Demer LL, Edwards PA, Watson AD, Lusis AJ. Atherosclerosis: basic mechanisms: oxidation, inflammation, and genetics. Circulation. 1995;91:2488-2496.
Volonte C, Rukenstein A, Loeb DM, Greene LA. Differential inhibition of nerve growth factor responses by purine analogues: correlation with inhibition of a nerve growth factor-activated protein kinase. J Cell Biol. 1989;109:2395-2403.
Miyazaki T, Pride HP, Zipes DP. Prostaglandins in the pericardial fluid modulate neural regulation of cardiac electrophysiological properties. Circ Res. 1990;66:163-175.
Zipes DP. Influence of myocardial ischemia and infarction on autonomic innervation of heart. Circulation. 1990;82:1097-1105.
Gerova M, Barta E, Gero J. Sympathetic control of major coronary artery diameter in the dog. Circ Res. 1979;44:459-467.
Kaechi K, Furukawa Y, Ikegami R, Nakamura N, Omae F, Hashimoto Y, Hayashi K, Furukawa S. Pharmacological induction of physiologically active nerve growth factor in rat peripheral nervous system. J Pharmacol Exp Ther. 1992;264:321-326.
Johannsen UJ, Mark AL, Marcus ML. Responsiveness to cardiac sympathetic nerve stimulation during maximal coronary dilation produced by adenosine. Circ Res. 1982;50:510-517.
Chilian WM, Ackell PH. Transmural differences in sympathetic coronary constriction during exercise in the presence of coronary stenosis. Circ Res. 1988;62:216-225.
Brown BG, Lee AB, Bolson EL, Dodge HT. Reflex constriction of significant coronary stenosis as a mechanism contributing to ischemic left ventricular dysfunction during isometric exercise. Circulation. 1984;70:18-24.