Epicardial Bradykinin B₂ Receptors Elicit a Sympathoexcitatory Reflex in Rats

Bradykinin is a potent vasodilator of blood vessels in the muscles, kidney, and viscera and also in the coronary circulation.¹² Dilatation of systemic arteries causes a sharp fall in systolic and diastolic blood pressures (BPs).³ The substance can increase capillary permeability and produce edema.¹² In this respect, it is involved in pain reactions² and can evoke cardiovascular reflex responses of so far unknown physiological significance.⁴⁵⁶ In the heart, bradykinin is released in cardiac ischemia and myocardial infarction.⁷

Important afferent reflex pathways stimulated by bradykinin have their origin in the heart.^468910111213 The responses are generally believed to be mediated by cardiac sympathetic afferent fibers. However, there are conflicting reports on decreases and increases in BP, heart rate (HR), and sympathetic nerve activity as well as biphasic responses with respect to these parameters depending on the animal model used.^9101114 Especially in the dog, the responses were difficult to predict.⁹ Vagotomy and sinoaortic denervation were used to avoid putatively confounding reflex responses after cardiac bradykinin application.¹¹ However, vagotomy and sinoaortic denervation may have intrinsic and unpredictable effects on sympathetic outflow.

Bradykinin exerts its effects by stimulation of at least two bradykinin receptors, B₁ and B₂.¹⁵¹⁶ B₂ receptors are the naturally occurring receptors, and B₁ receptors are scarce in normal tissue and expressed in situations of tissue damage.^1516171819 Selective B₂ receptor antagonists such as Hoe 140 had to be available so that the role in rats of B₂ receptors in a putatively sympathoexcitatory circulatory reflex arising from the heart could be addressed.²⁰²¹

We were interested in the putative ability of bradykinin to induce a sympathoexcitatory response characterized by increases in BP and renal sympathetic nerve activity (RSNA) via B₂ receptors in intact rats. To test this hypothesis, we injected bradykinin into the pericardial sac before and after pretreatment with the selective B₂ antagonist Hoe 140 to stimulate cardiac afferent fibers. Adenosine, which is said to induce a sympathoexcitatory response by similar reflex mechanisms,⁴²² was used as a positive control. Phenylbiguanide, a serotonin 5-HT₃ antagonist that induces hypotension, bradycardia, and sympathoinhibition by stimulating cardiac vagal afferent fibers, was used as a further control.²³²⁴ We included cervically vagotomized rats and rats with cardiac autonomic blockade accomplished by instilling 10% procaine into the pericardial sac.

Methods

Male Sprague-Dawley rats (Ivanovas, Kisslegg, FRG) weighing 250 to 300 g were maintained in cages at 24±2°C. They were fed a standard rat diet (No. C-1000, Altromin) containing 0.2% sodium by weight and were allowed free access to tap water.

Preparations

All procedures done in rats were performed in accordance with the guidelines of the American Physiological Society and approved by the local government agency (Regierung von Mittelfranken, Ansbach, FRG). On the day of the experiment and under hexobarbital (30 mg/kg IV, Brevimytal, Eli Lilly) anesthesia, rats were equipped with one femoral arterial and two venous catheters. During the experiments, appropriate anesthesia was achieved with a maintenance infusion of hexobarbital (60 μg/100 g per minute IV) through one of the venous femoral catheters. We used this form of anesthesia to prevent pain reactions during intravenous or intrapericardial administration of bradykinin that might have influenced the response pattern.

Intrapericardial Catheter

As previously described,²⁴ the intrapericardial catheter was constructed from a piece of thin silicone tubing (0.020-inch ID, 0.037-inch OD; Dow Corning) that was formed into an elliptical loop with the longest chord diameter of 2 cm. The loop was fixed with a suture, with the two open ends of the tubing leading out of the tubing being of equal length. One end was used for fluid administration into the pericardial space and the other for fluid withdrawal if more than 30 μL was administered intrapericardially.

Mechanical ventilation was instituted via a tracheal tube, and a high midline thoracotomy was performed. The lobes of the thymus were carefully separated, exposing a small portion of the pericardial sac adherent to the thymus. This part of the pericardial sac was opened, and the loop catheter was inserted into the pericardial sac and advanced toward the left side of the heart. The pericardial sac was closed by apposing the two lobes of the thymus and sealing them together with polyacrylic glue.²⁴²⁵ Finally, the thorax was closed in layers and the open ends of the tubing extensions of the pericardial catheter were exteriorized at the nape. At the end of the respective experiments, 100 μL of concentrated methylene blue in isotonic saline was injected into the pericardial sac via the injection arm; leakage of dye from the pericardial space was assessed visually at autopsy. Only rats without leakage from the pericardial sac were used for final evaluation. The pericardial catheters had a dead space of 4 μL. This amount of saline was used to flush the catheter after injection of substances.

Renal sympathetic nerve recording was performed as described previously.²³²⁴²⁶ Briefly, the left kidney was exposed through a flank incision, and a renal nerve bundle was dissected and placed on a bipolar stainless steel electrode (Cooner Wire Co). The amplified and filtered signal was channeled to an analog/digital oscilloscope (HM 512, Hameg) and a polygraph (model 7DA, Grass Instrument Co) for visual evaluation. An audio amplifier-loudspeaker (Grass model AM8) was used for auditory evaluation, and a rectifying voltage integrator (Grass model 7P10) was used. The integrated voltage signals were displayed on the polygraph. Fig 1 shows an original tracing. The quality of the RSNA signal was assessed by its pulse synchronous rhythmicity and by examination of the magnitude of the decrease in recorded RSNA during ganglionic blockade with the short-acting ganglionic blocker trimethaphan (10 mg/kg IV) with an injection of methoxamine (10 μg IV). When an optimal signal was observed, the recording electrode was fixed to the nerve bundle with silicone adhesive (Sil-Gel 604, Wacker-Chemie). The electrode cable was then secured to the abdominal trunk muscles by a suture.

For the experiments, the arterial catheter was connected to a transducer (Statham P23 Db) that was connected to a Grass polygraph for recording of BP and HR. A 1-hour equilibration and stabilization period was allowed.

Experimental Protocols

Protocol 1

Six anesthetized and instrumented Sprague-Dawley rats were injected in random order with 25 μg bradykinin either intravenously or intrapericardially. The rats were allowed a 20-minute recovery between each injection.

Protocol 2

Six anesthetized and instrumented Sprague-Dawley rats were injected in random order with 32 μg adenosine IV or 50, 200, and 500 μg adenosine IPC. The rats were allowed a 20-minute recovery between each injection.

Protocol 3

Six anesthetized and instrumented Sprague-Dawley rats were injected with 2.5, 12, and 25 μg bradykinin IPC and 0.9, 9, and 90 μg phenylbiguanide IPC in random order before and after intrapericardial pretreatment with 200 μg of the B₂ receptor antagonist Hoe 140. The rats were allowed a 20-minute recovery between each injection. In this and the subsequent protocols, the B₂ receptor antagonist was administered in a volume of 40 μL and the bradykinin and phenylbiguanide boluses in portions of 30 μL each.

Protocol 4

Six anesthetized and instrumented Sprague-Dawley rats were injected with 2.5, 12, and 25 μg bradykinin IPC and 0.9, 9, and 90 μg phenylbiguanide IPC in random order before and after bilateral cervical vagotomy performed by cutting the vagus nerves close to the bifurcation of the common carotid arteries. The rats were allowed a 20-minute recovery between each injection. Rats had to be intubated as their respiration was impaired after bilateral cervical vagotomy because the recurrent nerves were also cut.

Protocol 5

Six anesthetized and instrumented Sprague-Dawley rats were injected with 2.5, 12, and 25 μg bradykinin IPC and 0.9, 9, and 90 μg phenylbiguanide IPC in random order before and after intrapericardial pretreatment with 30 μL of 10% procaine. After dose-response curves were obtained without cardiac autonomic blockade first, a 50-μL bolus of 10 μg methoxamine was injected via the second line in the femoral vein to prove proper neurally mediated bradycardia by stimulation of the arterial baroreceptor reflex. BP, HR, and RSNA changes were assessed. A 10% procaine solution (30 μL) was injected into the pericardial sac, and 5 minutes later, methoxamine was given again. After 5 minutes, the methoxamine injection no longer induced reflex bradycardia in rats with properly implanted intrapericardial catheters. This was interpreted as a sign of proper blockade of efferent and afferent cardiac fibers.^27282930 Three minutes later, either bradykinin or phenylbiguanide was injected as soon as all parameters had returned to baseline values. Thereafter, a further bolus injection of methoxamine was administered to prove ongoing proper cardiac autonomic blockade. Before each dose of bradykinin and phenylbiguanide, additional injections of 30 μL of 10% procaine into the pericardial sac were necessary because the cardiac autonomic blockade lasted only for 15 minutes. Fluid was removed from the pericardial sac and the methoxamine test repeated again as described above. Since the responses to injections of the highest doses of bradykinin and phenylbiguanide never lasted more than 3.5 minutes, the cardiac autonomic blockade achieved with procaine was sufficient.

Protocol 6

Four anesthetized and instrumented Sprague-Dawley rats were injected with six consecutive doses of 25 μg bradykinin IPC. The drug was given every 20 minutes.

Drugs

Adenosine, bradykinin, methoxamine, phenylbiguanide, and procaine were purchased from Sigma Chemical Co. Trimethaphan camsylate (Arfonad, Hoffmann–La Roche) was provided by Roche Laboratories. The B₂ antagonist Hoe 140 was donated courtesy of Dr Schoelkens of Hoechst, Frankfurt, Germany. All drugs except for the previously dissolved trimethaphan were dissolved in saline and prepared anew for every experiment.

Data Analyses

Integrated RSNA was expressed as microvolts integrated over 1-second intervals. The background noise level was recorded as postmortem activity (average of 30 minutes) and subtracted from the measured nerve activity. Because of the limitations of comparing values from multifiber sympathetic nerves between rats, the data are expressed as percent change from control values.

The measured parameters were compared at baseline, maximal response, and recovery before and after drug administration. Furthermore, we tested for possible changes of baseline parameters after administering procaine intrapericardially, cervical vagotomy, or intravenous injection of the B₂ receptor antagonist. The data were statistically analyzed by ANOVA and Newman-Keuls post hoc test³¹ with the CSS statistical software package (StatSoft Inc). Only a priori fixed comparisons were tested. Statistical significance was defined as a value of P<.05. All data are given as mean±SE.

Results

Protocol 1

Protocol 1 evaluated the possibility of distinguishing between intrapericardial and intravenous application of bradykinin by the evoked response pattern in intact rats (Fig 2). Before intrapericardial or intravenous drug administration, baseline BP and HR were 90±6 mm Hg and 400±21 beats per minute (bpm), respectively. ANOVA showed that basal control levels between the injections were not different from these preexperimental values. The two injections were given in random order. Intrapericardial injection of 25 μg bradykinin unequivocally induced marked increases of BP, whereas intravenous application decreased it. The increases in RSNA were nominally not significantly different from each other. HR was not significantly altered after both modes of application. After intrapericardial bradykinin, BP returned to baseline values after 177±9 seconds and RSNA after 184±10 seconds. The time points for intravenous injections of bradykinin were as follows: BP, 175±9 seconds and RSNA, 179±14 seconds.

Protocol 2

Protocol 2 evaluated the effects of intravenous administration of adenosine (32 μg) versus graded intrapericardial application of adenosine (50, 250, 500 μg) on BP, HR, and RSNA responses (Fig 3). Before administration of any drug, baseline BP and HR were 97±6 mm Hg and 388±19 bpm, respectively. ANOVA showed that basal control levels between the various injections were not different from these preexperimental values. Whereas administration of 32 μg adenosine IV led to decreases in BP and HR as well as increases in RSNA, even 500 μg adenosine IPC was not followed by any significant change in the recorded parameters.

Protocol 3

Protocol 3 evaluated the effects of intrapericardial administration of the specific B₂ receptor antagonist Hoe 140 on BP, HR, and RSNA responses to graded doses of bradykinin and phenylbiguanide (Figs 1 and 4). Before administration of any drug, baseline BP and HR were 94±6 mm Hg and 380±21 bpm, respectively. ANOVA showed that basal control levels between the various injections were not different from these preexperimental values. Before the administration of the B₂ receptor antagonist, bradykinin and phenylbiguanide affected BP, HR, and RSNA dose dependently. Bradykinin increased BP and RSNA, and phenylbiguanide decreased BP, HR, and RSNA. Intrapericardial administration of the B₂ receptor antagonist did not affect baseline parameters. After pretreatment with the B₂ receptor antagonist, the responses of BP, HR, and RSNA to intrapericardial bradykinin were abolished and those to phenylbiguanide injection were preserved. After the highest dose of bradykinin, BP returned to baseline values after 177±6 seconds and RSNA after 179±13 seconds during the control period. The respective time points for the highest dose of phenylbiguanide were as follows: BP, 176±9 seconds; HR, 162±9 seconds; and RSNA, 185±12 seconds.

Protocol 4

Protocol 4 evaluated the effects of bilateral vagotomy on BP, HR, and RSNA responses to graded doses of bradykinin and phenylbiguanide (Fig 5). Baseline BP and HR assessed before administration of any drug were 97±8 mm Hg and 405±18 bpm, respectively. Again, ANOVA showed that basal control levels between the various injections were not different from these preexperimental values. Before cervical vagotomy was induced, administration of bradykinin and phenylbiguanide affected BP, HR, and RSNA dose dependently as described above. After bilateral cervical vagotomy, the responses to phenylbiguanide were abolished, whereas the responses to bradykinin were unaffected. After the highest dose of bradykinin, BP returned to baseline values after 178±6 seconds and RSNA after 183±9 seconds. The respective time points for the highest dose of phenylbiguanide during the control period were as follows: BP, 173±9 seconds; HR, 162±7 seconds; and RSNA, 183±7 seconds.

Protocol 5

Protocol 5 evaluated the effects of graded doses of bradykinin and phenylbiguanide on BP, HR, and RSNA responses after cardiac autonomic blockade with 10% procaine (Fig 6). Before administration of any drug, baseline BP and HR were 94±6 mm Hg and 380±21 bpm, respectively. ANOVA showed that basal control levels between the various injections were not different from these preexperimental values. During the control period, bradykinin and phenylbiguanide affected BP, HR, and RSNA dose dependently. After the highest dose of bradykinin, BP returned to baseline values after 173±8 seconds and RSNA after 187±10 seconds. The respective time points for the highest dose of phenylbiguanide were as follows: BP, 170±7 seconds; HR, 165±6 seconds; and RSNA, 180±12 seconds. Bradykinin increased BP and RSNA, whereas phenylbiguanide decreased BP, HR, and RSNA. Intrapericardial administration of procaine only transiently affected baseline parameters. Pretreatment with procaine abolished the responses of BP, HR, and RSNA to intrapericardial bradykinin and phenylbiguanide. After the first injection of 10% procaine into the pericardial sac, mean arterial BP was lowered from 94±6 to 88±4 mm Hg for about 5 minutes. HR exhibited only a transient decrease of −32±9 bpm for about 1 minute. RSNA did not change significantly. After cardiac autonomic blockade with intrapericardial procaine, the responses of the measured parameters to both phenylbiguanide and bradykinin were completely inhibited.

Protocol 6

Protocol 6 evaluated the effects of six consecutive doses of 25 μg bradykinin IPC on BP and RSNA responses (Table). Injections were given every 20 minutes. There were no significant changes in the responses of the measured parameters, nor were there significant alterations in the time course of these responses (time of maximal response, time of recovery).

Discussion

Our results demonstrate that cardiac B₂ receptors are able to induce a marked sympathoexcitatory response most likely via cardiac sympathetic afferent fibers. Bradykinin given intrapericardially elicited a quite different response of mean arterial BP than when injected intravenously at the same dose: In the first case, BP increased, and in the latter, BP decreased. Intravenous bradykinin is a potent vasodilator.¹ Hence, the increases of RSNA after intravenous bradykinin were most likely mediated by the baroreceptor reflex caused by the induced hypotension.³²³³ However, the increases of RSNA and BP in response to intrapericardial bradykinin should have been mediated by neurogenic mechanisms depending on cardiac afferent nerve traffic.¹⁰

It is known that bradykinin stimulates sympathetic afferent fibers from the heart.⁴⁸³⁴³⁵ However, systemic cardiovascular responses have not been uniform and often have been depressant: Excitatory responses have been observed in cats.¹⁴³⁶ In rabbits¹¹ and monkeys,⁹ cardiac bradykinin produced reflex decreases of BP, HR, and sympathetic nerve activity. In dogs, activation of cardiac sympathetic afferents with bradykinin resulted in inhibitory and excitatory as well as biphasic responses.¹³³⁷

The bradykinin receptor mediating the majority of the bradykinin effects is the B₂ receptor.¹⁶ In rats, these receptors have been found in the dorsal root ganglia, the spinal cord, and the area close to the cerebral ventricle.³⁸³⁹⁴⁰ These findings suggest that B₂ receptors are an important part of visceral autonomous pathways and reflexes.

B₁ receptors are synthesized under inflammatory and ischemic conditions.¹⁶¹⁷ The fact that bradykinin is released especially during myocardial ischemia⁷¹⁸ might suggest a role of B₁ receptors under these pathophysiological conditions. The existence of a subgroup of cardiac afferent sympathetic fibers synthesizing B₁ receptors under special conditions is not completely unlikely.^41424344 However, our results in rats demonstrate that cardiac afferent fibers with B₂ receptors alone are capable of eliciting a uniform sympathoexcitatory reflex that could instantly influence the neural control of the circulation without the aid of an additional receptor class being partly dependent on de novo synthesis.

The bradykinin inhibitor Hoe 140 proved to be quite specific and did not influence the responses to intrapericardial phenylbiguanide.²⁰ Phenylbiguanide has been shown to be a serotonin 5-HT₃ receptor agonist that specifically stimulates cardiac vagal afferents if the drug is injected intrapericardially or intravenously.²³²⁴ Both injection routes induced the classic features of the Bezold-Jarisch reflex (hypotension, bradycardia, and sympathoinhibition). On the other hand, intrapericardial bradykinin was characterized by increases of BP and RSNA.

It is assumed that chemostimulation of vagal cardiac afferents induces depressor responses by influencing neural control of the circulation, whereas chemostimulation of cardiac sympathetic afferents may mediate sympathoexcitation and be involved in nociception.⁸ Bradykinin could theoretically stimulate vagal afferents that are involved in cardiovascular control and nociception. However, in our experiments, vagotomy did not alter the responses to intrapericardial bradykinin. This suggests that the afferent information was most likely transmitted to the central nervous system by sympathetic afferent fibers.

We induced cardiac autonomic blockade by repetitive injections of 10% procaine into the pericardial sac. Several authors have used this method in different animal models.^272829304546 Our experiments with intrapericardial procaine demonstrate that the responses to bradykinin and phenylbiguanide injected into the pericardial sac were indeed of cardiac origin.

We used adenosine as a control substance because it occurs endogenously under conditions comparable to those of bradykinin and is also said to stimulate cardiac sympathetic afferent fibers.⁴ We were not able to elicit any responses by intrapericardial application of adenosine. Since we measured only efferent and not afferent sympathetic nerve activity, it is possible that afferent sympathetic nerve traffic was increased, playing a role in, for instance, nociception but not being involved in the neural control of circulation.⁸ Since reflex responses to adenosine were observed during acute cardiac ischemia in dogs, these responses of adenosine might depend on facilitating circumstances such as local increases in H⁺ proton concentration, oxygen free radicals, and/or generation of various mediator substances and related receptors.^81643444748 The assumption that adenosine could depend on unknown confounding conditions is further supported by the fact that direct recordings of cardiac sympathetic C fibers showed conflicting results in the same animal model: Whereas Montano et al²² observed marked increases of cardiac sympathetic C-fiber activity after intracoronary injection of adenosine in the cat, Pan and Longhurst³⁶ could detect no response of cardiac sympathetic C fibers after either intracoronary or intrapericardial administration of this drug in cats with coronary ischemia.

Our results demonstrate that local bradykinin can evoke a powerful excitatory reflex. Notably, the sympathoexcitation occurred when all other cardiovascular reflex mechanisms (arterial baroreceptor reflex, cardiopulmonary reflexes with vagal afferents) were left intact and could have compensated. Hence, cardiac B₂ receptors might play an important role in circulatory control, when bradykinin is released locally, eg, during cardiac ischemia.

Presented in part at the annual meeting of the American Heart Association, Dallas, Tex, November 14-17, 1994.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes