Evidence for control of cardiac vagal tone by benzodiazepine receptors

NeuropharmacologyVol. 26, No. 6, pp. 553-559, 1987 Printed in Great Britain. All rights reserved

EVIDENCE

002%3908/87$3.00+ 0.00 Copyright 0 1987 Pergamon Journals Ltd

FOR CONTROL OF CARDIAC VAGAL TONE BY BENZODIAZEPINE RECEPTORS

J. A. DIMICCO Department of Pharmacology and Toxicology, Indiana University School of Medicine, Indianapolis, IN 46223, U.S.A. (Accepted 25 May 1986)

Summary-Previous work has suggested that vagal preganglionic neurons which project to the heart, are tonically inhibited by endogenous y-aminobutyric acid (GABA). This study tested the hypothesis that benzodiazepines, which are thought to act by enhancing GABAergic inhibition, would increase heart rate by suppressing cardiac vagal activity in anesthetized rats, and that pretreatment with Ro 15-1788, a specific benzodiazepine receptor antagonist, would prevent tachycardia induced by benzodiazepines. Midazolam (O.OHmg/kg i.v.), alprazolam (1 mg/kg i.v.) and chlordiazepoxide (10 and 20mg/kg i.v.), all evoked significant increases in heart rate. Pretreatment with atropine methobromide (2 mg/kg i.v.) increased the basal heart rate and prevented tachycardia induced by benzodiazepines. Basal heart rate and blood pressure were unchanged after pretreatment with Ro 15-1788 (10 mg/kg), but subsequent administration of any of the benzodiazepines failed to elicit increases in heart rate in these animals. These findings suggest that benzodiazepines may be potent vagolytics and that this effect should be considered before these agents are administered to patients who have suffered a recent myocardial infarction, in whom vagal tone is thought to be protective against fatal ventricular arrhythmias. Key words: benzodiazepines, heart rate, vagus, parasympathetic, pressure, rats.

Benzodiazepines,

which

comprise

one of the most

widely prescribed classes of drugs, are commonly used for the treatment of anxiety in a variety of clinical settings in which the primary pathology is related to cardiovascular function, particularly in patients who have suffered a recent myocardial infarction (Krogh, McLeen and LaPierre, 1978; Greenblatt, Shader and Abernethy, 1983). Evidence suggests that, in this latter clinical setting, the vagal tone to the heart exerts a significant protective effect against lethal ventricular fibrillation (Epstein, Goldstein and Redwood, 1973). This contention is based in large part on studies in laboratory animals in which, either prior treatment with atropine, or bilateral vagotomy, significantly increased mortality resulting from coronary ligation (Corr and Gillis, 1978; Kent, 1978; Marshal, Muir and Winslow, 1981) or decreased the threshold for electricallyinduced ventricular fibrillation (Verrier and Lown, 1984). Therefore, under conditions of acute myocardial ischemia in which benzodiazepines are often given, administration of any agent which might suppress cardiac vagal tone would deprive the compromised myocardium of an important protective mechanism. Benzodiazepines are now thought to exert their characteristic effects by acting at their own specific receptors, which have been under intense scrutiny in recent years. Benzodiazepine receptors appear to be coupled physically and functionally to receptors for y-aminobutyric acid (GABA), a major inhibitory

autonomic nervous system, arterial

neurotransmitter in the mammalian central nervous system, such that benzodiazepines acting at their receptors, somehow facilitate the interaction between GABA and its receptor or enhance the postsynaptic consequence of this interaction (Haefely, Pole, Schaffner and Laurent, 1983b). Strong evidence supporting this theory derives from studies involving ethyl 8-fluoro-5,6-dihydro-5-methyl-6-oxo-4Himidazo[ 1,5-a] [ 1,4]benzodiazepine-3-carboxylate (Ro 15-1788) a compound which competes with high affinity for these specific benzodiazepine receptors in vitro and blocks the characteristic effects of the administration of benzodiazepine in vivo (Haefely, Bonnetti, Burkard, Cumin, Laurent, Mohler, Pieri, Pole, Richards, Schaffner and Scherschlicht, 1983a). Studies employing agents which act to interfere with or mimic the effects of GABA have provided evidence to suggest that endogenous GABA tonically inhibits cardiac vagal preganglionic neurons, located in the mammalian brainstem (DiMicco, Gale, Hamilton and Gillis, 1979; Williford, Hamilton and Gillis, 1980; Alsip, Simon, Fohl and DiMicco, 1984; Wible and DiMicco, 1986). If so, then stimulation of benzodiazepine receptors, which are coupled to postsynaptic GABA receptors, should enhance endogenous inhibition and thereby further suppress vagal activity. Consistent with this prediction are the results of several previous studies in which the administration of chlordiazepoxide or diazepam was reported to increase the heart rate in laboratory animals (Daniell, 1975; Gerold, Cavero, Riggenbach, Wall

553

J. A. DIMICCO

554

and Haeusler, 1976) and in man (Whitehead, Blackwell and Robinson, 1977) or to suppress activity in the vagal nerve (Sigg and Sigg, 1969) or reflex bradycardia (Hockman and Livingston, 1971; Keim and Sigg, 1973). The availability of a receptor antagonist, such as Ro 15-1788, which could be used to demonstrate the role of specific benzodiazepine receptors in these effects has prompted a systematic examination of this issue. The aim, therefore, was (1) to assess the effects on the heart rate of several representative benzodiazepines in anesthetized rats, and (2) to examine the role of the vagus and of benzodiazepine receptors, by determining the effect of pretreatment with atropine or Ro 15-1788, on these actions. METHODS

Male Sprague-Dawley rats (240-390 g) were anesthetized with urethane 1.35-1.5 g/kg (i.p.) in saline. A femoral artery and vein were cannulated for the measurement of blood pressure and the administration of drugs, respectively. The trachea was cannulated in order to facilitate respiration. Heart rate was determined with a cardiotachometer, triggered by lead II of the electrocardiogram and was recorded continuously, along with mean blood pressure on a Beckman (R5 11A) strip chart recorder. All drugs were administered intravenously in a volume of 1 ml/kg body weight. Agents employed included atropine methobromide (Sigma, St. Louis, Missouri), chlordiazepoxide hydrochloride and midazolam (both generously supplied by Hoffman-La Roche, Nutley, New Jersey), alprazolam (a gift from Upjohn, Kalamazoo, Michigan) and Ro 15-1788 kindly supplied by Dr W. Haefely, F. Hoffman-La Roche and Co. Ltd, Basle, Switzerland. The drug, Ro 15-1788 was administered as a fine suspension prepared by adding one drop each of dimethyl sulphoxide (DMSO) and Tween 80 to 1Omg of drug, mixing this and adding saline to a final volume of 1 ml. Midazolam and alprazolam were dissolved in saline by acidifying with HCl and titrating back to a pH of at least 5.3 with NaOH. All other drugs were dissolved in saline. A similar experimental protocol was employed for the administration of Ro 15-1788 and the benzodiazepines in all experiments. After surgery was completed, at least 30 min was allowed for cardiovascular stabilization before any drug was administered. At this time, the baseline blood pressure and heart rate were noted and either Ro 15-1788 (lOmg/kg) or an equivalent volume of vehicle was administered. Fifteen min later each animal received either a benzodiazepine or vehicle, i.e. acidified saline. Blood pressure and heart rate were then monitored for at least 45 min. In experiments where atropine was the pretreatment, this agent was administered in a dose of 2 mg/kg, 15 min prior to baseline readings for blood pressure and heart rate and the administration

of the vehicle for the benzodiazepine antagonist, and therefore 30min prior to the administration of a benzodiazepine. Statistical analysis consisted of analysis of variance (ANOVA) coupled with Duncan’s multiple range test and Student’s z-test for paired data. The criterion for significance was P < 0.05. RESULTS

All three ~nzodi~epines tested evoked increases in the heart rate in rats treated with vehicle for Ro 15-1788. Increases evoked by midazolam (1 mg/kg) or alprazolam (1 mg/kg) were maximal 15 min after administration with the heart rate gradually returning towards control over the next 30 min (Figs 1 and 2). In neither case was the arterial pressure significantly altered by treatment with drugs. After treatment with Ro 15-1788 (10 mg/kg), which itself failed to influence either cardiovascular parameter, neither alprazolam nor midazolam evoked significant increases in heart rate, suggesting that the tachycardia caused by these agents was a consequence of their interaction at the benzodiazepine receptor. Pretreatment with atropine resulted in higher levels of the baseline heart rate (Table l), and also prevented the tachycardic effect of alprazolam or midazolam administered subsequently. Chlordiazepoxide also caused an increase in the heart rate which was prevented by pretreatment with Ro 15-1788 (Figs 3 and 4). The administration of chlordiazepoxide 10 mg/kg caused a mild tachycardia which failed to reach the level of statistical significance by a narrow margin (0.10 > P > 0.05). The changes caused by infusion of chlor~azepoxide 2Omg/kg appeared to be more complex than were observed after administration of the smaller dose of this agent, but included significant tachycardia 30 and 45 min after administration, which was prevented by pretreatment with Ro 15-1788. In addition, the large dose of the benzodiazepine also evoked trends in both the parameters measured which were resistant to blockade by the benzodiazepine antagonist. Thus, a transient decrease in heart rate was apparent after the administration of chlordiazepoxide 20 mg/kg, in antagonist-pretreated rats at 15 min, a trend which may have offset and therefore obscured the tachycardia mediated by ~nzodi~epine receptors at this time in rats which had not been pretreated with Ro 15-1788. In addition, the benzodiazepine also tended to increase the arterial pressure in a similar manner in both rats treated with the antagonist and in untreated rats, although the change was significantly different from that seen in vehicletreated rats only at 60min after administration of chlordiazepoxide in rats pretreated with the benzodiazepine antagonist. In an expanded series of experiments, the effects of a range of doses of midazolam (0.05-4mg/kg), as weil as a second smaller dose of alprazolam

Benzodiazepines inhibit cardiac vagus

555

__-TT?-Tb Iii1~___----+10

I”

Q

0

__

-10

i

Fig. 1. Effect of midazolam (MZ) (1 mg/kg i.v.) on the heart rate (HR; above) and blood pressure (BP; below), in urethane-anesthetized rats. All values represent mean changes + SE with respect to individual baselines (mean baseline values for each group listed in Table 1). Treatment A consisted of intravenous injection of either Ro 15-1788 (10 mg/kg) or an equivalent volume of vehicle as indicated, while midazolam was administered as treatment B, 15 min later. In one group atropine was administered 15 min prior to the baseline reading at 0 min. Results of ANOVA: *Significantly different from corresponding value in vehicle/vehicle control group; ~Si~ifi~ntly different from ~r~sponding value in animab pretreated with Ro 15-1788 (P < 0.05).

Table I. Changes in blood pressure and heart rate IS min after treatment with midazolam, alprazolam or vehicle: Effect of pretreatment with Ro 15-1788 (10 mg/kg) or atropine (2 mg/kg)

Baseline Treatment Vehicle Vehicle Midazolam 4 mg/kg 4 mg/kg 2 mg/kg 2 mg/kg 2 mgikg l mg/kg

1m&c/kg 1u/kg

Bretreatment

n

Ro 15-1788

5 7

406 i: 22 406+10

101 f8 102?4

-6*7 i-10+6

6 6 9 6 8 6 6 6 6 6 7

362 +6 381 + I3 386 f 7 377 & 10 422 * 7 390 + IO 382 k 15 439 * I I 392 + IO 388 f I3 388 + IO

101*2 98 k 5 105i2 10926 91*5 108k8 lO2?4 93*4 98 +4 91_+5 98&4

+43*4* -8ct6 +31* 128 +7110 -455 +36 +4* o+ I2 +3tt5 +21 *6’ +31 &6* +14+5*

6

368kII 370 f IO 425 i I4 373 + I6

88+4 88 f 5 83k8 88L-4

Ro 15-1788 Ro 15-1788 Atropine Ro IS-1788 Atropine

0.5 mg/kg 0.25 mg/kg 0.05 mg/kg Aipraaolam

1m-/kg 1me/kg f w/kg 0.25mg/kg

Change in: Heart rate Blood pressure (beatsimin) (mm%)

Heart rate (beats/mitt)

Ro 15-1788 Atropine

6 5 5

Blood pressure (mmHg)

-t-37*9* -2k6

+2rt7 +23*9

+2+3 +51t3 +ll$6 O&2 -l&7 -723 OrtS -8t_3 f2ir4 +6+5 +2+4 -I 16 +I*2 +5*4 +5zb3 +81tS +3*4

*Significantly different from baseline by paired t-test (P < 0.05). Atropine (2 mgjkg) was administered I5 min prior to baseline measurements and 30 min prior to administration of midazolam or alprazolam; Ro 15-1788 (IOmgjkg) was administered immediately after baseline measurements and 15 min prior to administration of midazolam, alprazolam, or vehicle.

J. A. DIMICCO

556

+20

+10 P E : a

o-

-10

Fig. 2. Effect of alprazolam (APZ) (1 mg/kg i.v.) on the heart rate (HR; above) and blood pressure below) in utethane-anesthetized rats. See legend Figure 1 for details.

(0.25 mg/kg), on the blood pressure and heart rate were examined (Table 1). Fifteen min after the administration of every dose of midazolam tested, the heart rate was significantly elevated while the blood pressure remained unchanged. The heart rate also tended to increase after the administration of alprazolam 0.25 mg/kg but the change failed to reach the level of statistical significance (0.10 > P > 0.05). Pretreatment with either Ro 15-1788 or atropine consistently prevented tachycardia induced by the benzodiazepines.

DISCUSSION

The results of these studies clearly show that the benzodiazepines tested were similar in eliciting increases in heart rate in anesthetized rats, and that these increases were prevented by pretreatment with a benzodiazepine receptor antagonist. By itself, Ro 15-1788 had no significant effect on either resting blood pressure or heart rate. This is consistent with the idea that (1) at moderate doses, Ro 15-1788 has little or no intrinsic activity at the benzodiazepine receptor, and (2) those benzodiazepine receptors relevant to the effects on heart rate, already noted, are

(BP;

not tonically activated to any significant degree by an endogenous ligand under these experimental conditions. Pretreatment with atropine also blocked the tachycardia induced by benzodiazepines while elevating the basal heart rate, demonstrating both the existence of significant vagal tone in this preparation and the necessity of this tone for the expression of benzodiazepine-induced tachycardia. This finding strongly suggests that the benzodiazepines act to increase the heart rate by a vagolytic action. The site at which benzodiazepines might act to inhibit cardiac vagal activity was not specifically examined in this study but several lines of evidence point to a site of action in central, rather than an interaction with peripheral, nervous mechanisms. Benzodiazepines, including chlordiazepoxide and alprazolam, at doses in excess of those employed here, were shown to have no significant effect on peripheral vagal synaptic function or on ganglionic transmission (Randall, 1960; VonVoigtlander and Straw, 1985). Furthermore, while peripheral benzodiazepine receptors are known to exist, they are pharmacologically distinct from those found in the central. nervous system and only the latter are blocked by Ro 15-1788 (Gee, Yamamura, Roeske and Yamamura, 1984).

Benzodiazepines

inhibit

cardiac

vagus

a

+10-

n

m

a

-lO-

0

1;

3-O TIME

0

n 0

0

10

45

60

(MIN.)

A VEHICLE RO 15-1788

0 CD2 CDZ

10 mglkg 10 mglkg

Fig. 3. Effect of chlordiazepoxide (CDZ) (10 mg/kg i.v.) on the heart rate (HR; above) and blood pressure (BP; below) in urethane-anesthetized rats. All values represent mean changes f SE with respect to individual baselines (baseline values: Vehicle + CDZ: 393 + 15beats/min and 100 f 7 mmHg; Ro 15-1788 + CDZ: 404 f 7 beats/min and 96 k 3 mmHg). The ANOVA revealed that in animals receiving chlordiazepoxide alone changes in the heart rate and blood pressure were not significantly different from corresponding values in animals which had been pretreated with Ro 15-1788 prior to administration of chlordiazepoxide (shown above) or in vehicle/vehicle treated rats (Figs 1 and 2).

Therefore, it seems likely that the vagolytic effects observed were due to the benzodiazepines acting in the brain to depress the efferent activity of the cardiac vagus. Previous studies from this laboratory pointed to the existence of a GABAergic inhibitory mechanism located in the vicinity of cardiac vagal preganglionic neurons in the mammalian brainstem and tonically suppressing their activity. Thus, it is thought that benzodiazepines elevate the heart rate by acting at specific benzodiazepine receptors, coupled to the GABA receptors that mediate this inhibition. As a result of this interaction, cardiac vagal activity is further suppressed and the heart rate increases. Previous studies, which examined the effects of benzodiazepines on autonomic cardiovascular control mechanisms have, with few exceptions, focussed on the sympathetic nervous system (Chai and Wang, 1966; Sigg and Sigg, 1969; Sigg, Keim and Kepner, 1971; Antonaccio and Halley, 1975; Clubley and Elliott, 1977). Where examined, the increased sympathetic discharge, arterial pressure and/or heart rate, evoked by reflex or by central electrical stimulation was found to be markedly depressed by these agents. In addition, however, these studies often suggested

that benzodiazepines might produce mild hypotension and bradycardia, due to depression of basal sympathetic tone. In the present investigation the arterial pressure was unchanged by either midazolam or alprazolam, in spite of the very large doses of the former agent tested. After the administration of atropine, the heart rate was similarly unaffected. These findings suggest that stimulation of benzodiazepine receptors had little or no effect on basal sympathetic tone under these experimental conditions. The present conclusions may differ from those of previous studies due to differences in species or anesthetic employed. However, it is also possible that in some studies the apparent depression of cardiovascular function and/or sympathetic discharge was the result of the effects of the vehicle for which appropriate control experiments were inadequate or entirely lacking (see discussion in Daniel], 1975). Furthermore, the large doses which were often used may have produced effects other than through the benzodiazepine receptor, a possibility that would be difficult to rule out without the use of a specific antagonist, as was employed in the present study. Indeed, the present data demonstrate that chlor-

J. A. D~Mrcco

558

E d

-to -20

_I

-30 ~ +20 I” E mh Q

+to 0

-

-to ~

I

1

I

If

0

30 TIME

0

0

I: 6

A VEHICLE RO 15-1788

I

45

I

60

(MIN)

CD2 CDZ

6 20 mglkg 20 mglkg

I.V.

I.V.

Fig. 4. Effect of chlordiazepoxide (CDZ) (20 mg/kg i.v.) on the heart rate (HR; above) and blood pressure (BP; below) in urethane-an~thetized rats. All values represent mean changes f SE with respect to individual baselines (baseline values: Vehicle + CDZ: 390 f 14 beats/min and 100 + 6 mmHg; Ro 15-1788 + CDZ: 391 & 18 beats/min and 95 rt 5 mmHg). Results of ANOVA: *Significantly different from corresponding value in vehicle/vehicle control group (shown in Figures 1 and 2); tSignificantly

different from corresponding value in animals pretreated with Ro 15-1788(P < 0.05). diazepoxide ZOmg/kg elicited, in addition to the tachycardia mediated by benzodiazepine receptors, and seen also with midazolam and alprazolam, effects which persisted after treatment with Ro 15-1788, including a tendency to a decrease in the heart rate and an increase in arterial pressure. To summarize the present findings are consistent with those studies which have suggested that benzodiazepines may increase the heart rate in man and selectively depress cardiac vagal function in laboratory animals. In patients suffering from angina or a recent myocardial infarction, these drugs are commonly prescribed as anxiolytics. That these same agents may inhibit cardiac vagal efferent activity has important clinical implications in view of evidence suggesting that the vagus may protect the compromised myocardium from ventricular fibrillation. Acknowledgements-This study was supported by USPHS grant NS 19883. The author gratefully acknowledges the technical assistance of Kimberly D. Hankins. REFERENCES Alsip N. L., Simon J. R., Fohl L. D. and DiMicco J. A. (1984) Cardiovascular effects of 3-mercaptopropionic acid and levels of GABA in regions of the brain of guinea pigs. Ne~ropharm~coiogy 23: 349-357.

Antonaccio M. J. and Halley, J. (1975) Inhibition of centrally evoked pressor responses by diazepam: Evidence for an exclusively supramedullary action. Neuropharmacology 14: 649-651.

Chai C. Y. and Wang S. C. (1966) Cardiovascular actions of diazepam in the cat. J. Pharmac. exp. Ther. 154: 271-280.

Ciubley M. and Elliott R. C. (1977) Centrally active drugs and the s~pathetic nervous system of rabbits and cats. Neuropharmacology

16: 609-6 16.

Corr P. B. and Gillis R. A. (1978) Autonomic neural influences on the dysrhythmias resulting from myocardial infarction. Circ. Res. 43: l-9. Daniel1 H. B. (1975) Cardiovascular effects of diazepam and chlordiazepoxide. Eur. J. Pharmac. 32: S&65. DiMicco J. A., Gale K., Hamilton 8. L. and Gillis R. A. (1979) GABA receptor control of parasympathetic outflow to the heart: Characterization and brainstem localization. Science 204: llObl109. Epstein S. E., Goldstein R. E. and Redwood D. R. (1973) The early phase of acute myocardial infarction: Pharmacological aspects of therapy. Ann. Inf. Med. 78: 918-936. Gee K. W., Yamamura S. H., Roeske W, R. and Yamamura H. I. (1984) ~nz~ia~pine receptor heterogeneity: Possible molecular basis and functional significance. Fed. Proc. 43: 2767-2712.

Gerold M., Cavero I., Riggenbach H., Wall M. and Haeusler G. (1976) Analysis of chronotropic responses to diazepam and bromazepam in conscious trained dogs. Eur. J. Pharmac. 35: 361-368. Greenblatt D. J., Shader R. J. and Abernethy D. R. (1983)

Benzodiazepines inhibit cardiac vagus Current status of benzodiazepines. N. Engl. J. Med. 309: 41&416. Haefely W., Bonetti E. P., Burkard W. P., Cumin R., Laurent J-P., Mohler H., Pieri L., Pole P., Richards J. G., Schaffner R. and Scherschlicht R. (1983a) Benzodiazepine antagonists. In: The Benzodiazepines: From Molecblar Biology to Clinical Practice (Costa E., Ed.), pp. 137-146. Raven Press, New York. Haefely W., Pole P., Schaffner R. and Laurent J-P. (1983b) Neuropharmacology of Benzodiazepines: Synaptic mechanisms and neural basis of action. In: The Benzodiazepines: From Molecular Biology to Clinical Practice

(Costa E., Ed.) pp. 21-26. Raven Press, New York. Hockman C. H. and Livingston K. E. (1971) Inhibition of reflex vagal bradycardia by diazepam. Neuropharmacology 10: 3077314. Keim K. K., and Sigg E. B. (1973) Vagally mediated cardiac reflexes and their modulation by diazepam and pentobarbital. Neuropharmacology 1% 319-325. Kent K. M. (1978) Cholinergic innervation of the heart: Electrical stability during congestive heart failure. In: Neural Mechanisms in Cardiac Arrhythmias (Schwartz P. J., Brown A. M., Malliani A. and Zanchetti A., Eds), pp. 99-106. Raven Press, New York. Krogh C., McLean W. M. and LaPierre Y. D. (1978) Minor tranquilizers in somatic disorders. Can. med. Ass. J. 118: 1097-I 108. Marshall R. J., Muir A. W. and Winslow E. (1981) Devel-

559

opment of a severe model of early coronary artery ligation-induced dysrhythmias in the anesthetized rat. Br. Jy Pharmac. 73: 951-959.

Randall L. 0. (1960) Pharmacology of methaminodiazepoxide. Dis. Nerv. Syst. 21, Suppl. 7, 7-10. Sigg E. B. and Sigg T. D. (1979) Hypothalamic stimulation of preganglionic autonomic activity and its modification by chlorpromazine, diazepam and pentobarbital. Inf. J. Neuropharmac. 8: 567-572.

Sigg E. B., Keim K. L. and Kepner K. (1971) Selective effect of diazepam on certain central sympathetic components. Neuropharmacology

10: 621-629.

Verrier R. L. and Lown B. (1984) Behavioral stress and cardiac arrhythmias. Ann. Rev. Physiol. 46: 155-176. VonVoigtlander P. and Straw R. N. (1985) Alprazolam: Review of pharmacological, pharmacokinetic and clinical data. Drug Dev. Res. 6: l-12. Whitehead W. E., Blackwell B. and Robinson A. (1977) Why physicians prescribe benzodiazepines in essential hypertension: A phase IV study. Biol. Psychiat. 12: 597401. Wible J. H. Jr and DiMicco J. A. (1986) Biphasic effects of systemically administered GABA antagonists on cardiac vagal activity. Brain Res. 363: 279-289. Williford D. J., Hamilton B. L., and Gillis R. A. (1980) Evidence that a GABAergic mechanism at nucleus ambiguus influences reflex-induced vagal activity. Brain Res. 182: 19-22.