Cardiac responses elicited by peptides administered to canine intrinsic cardiac neurons

Peptides, Vol. 11,753-761. ©Pergamon Press plc, 1990. Printed in the U.S.A.

0196-9781/90 $3.00 + .00

Cardiac Responses Elicited by Peptides Administered to Canine Intrinsic Cardiac Neurons J. A. A R M O U R , t

BINGXIANG

Y U A N A N D C. K. B U T L E R

Departments of Physiology and Biophysics and Surgery, Faculty of Medicine, Dalhousie University, Halifax, N.S., Canada B3H 4H7 R e c e i v e d 23 M a r c h 1990

ARMOUR, J. A., B. YUAN AND C. K. BUTLER. Cardiac responses elicited by peptides administered to canine intrinsic cardiac neurons. PEPTIDES 11(4) 753-761, 1990.--In order to study the effects of peptides on intrinsic cardiac neurons, substance P, bradykinin, oxytocin, calcitonin gene related peptide, atrial natriuretic peptide and vasoactive intestinal peptide were administered into canine atrial or ventricular ganglionated plexi. When substance P was injected into right atrial or cranial medial ventricular ganglionated plexi heart rate, atrial force and ventricular intramyocardial pressures were augmented. No cardiac changes occurred when similar volumes of saline (i.e., peptide vehicle) were injected into these ganglionated plexi. When bradykinin was injected into atrial or ventricular ganglionated plexi heart rate, atrial force and ventricular force were augmented in --50% and depressor responses were elicited in -50% of these animals. When oxytocin was injected into right atrial ventral ganglionated plexi heart rate and atrial forces were reduced in five of ten dogs studied. No cardiac changes occurred when oxytocin was injected into left atrial or ventricular ganglionated plexi. No responses were elicited when calcitonin gene related peptide, atrial natriuretic peptide or vasoactive intestinal peptide was administered into atrial or ventricular ganglionated plexi. Following acute decentralization of the heart, no significant responses were elicited by repeat administrations of substance P, bradykinin or oxytocin, implying that connectivity with central nervous system neurons was necessary for consistent responses to be elicited. It is concluded that substance P, bradykinin and oxytocin can affect neurons on the heart such that cardiodynamics are modified, these different peptides eliciting different cardiac responses. Atrial natriuretic peptide Neuron Oxytocin

Bradykinin Substance P

Calcitonin gene related peptide Vasoactive intestinal peptide

CARDIAC ganglia have been proposed to contain, in addition to efferent postganglionic parasympathetic neurons (1, 11, 24), afferent neurons (7,18), local circuit neurons (7,18) and efferent postganglionic sympathetic neurons (3, 7, 18). Afferent and local circuit neurons have also been identified in canine stellate and middle cervical ganglia (4,5). Substance P-like and vasoactive intestinal peptide (VIP)-like immunoreactivities have been associated with neuronal somata in canine middle cervical and stellate ganglia (15). When substance P, VIP or oxytocin is administered into an acutely decentralized stellate or middle cervical ganglia, direct or indirect activation of the efferent sympathetic neurons innervating the heart occurs, resulting in augmentation of cardiac chronotropism and inotropism (6,8). Substance P (21, 23, 25), VIP (31) and calcitonin gene related peptide (CGRP) (27) have been associated with somata and axons in, or on, the heart. Primary afferent neurons innervating the heart have been shown to contain calcitonin gene related peptide and substance P (23). Furthermore, bradykinin is known to activate axons of cardiac afferent neurons (17,26).

Cardiac

Chronotropism

Inotropism

Although a number of peptides are known to be involved in cardiovascular regulation (16, 26, 27, 29, 30), it is not known if peptides can modify neurons on the heart such that cardiodynamics are changed. Thus, in the present series of experiments substance P, bradykinin, oxytocin, CGRP and VIP were administered into in situ cardiac ganglionated plexi to determine whether these peptides are capable of modifying neurons in cardiac ganglia and, if that were the case, how such modification might alter cardiodynamics. Since atrial natriuretic peptide has also been associated with cardiac tissues (22), the effects of this peptide on cardiac neurons were also investigated. METHOD

Animal Preparation Mongrel dogs of either sex, weighing 19-26 kg, were tranquilized with sodium thiopental (12-15 mg/kg IV) and anesthetized with alpha chloralose (100 mg/kg IV). Thereafter alpha chloralose (25 mg/kg IV) was administered as a bolus every 2 hours

~Requests for reprints should be addressed to J. A. Armour, Department of Physiology and Biophysics, Faculty of Medicine, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada.

753

754

throughout the experiments or more frequently, as required. Following intubation, positive-pressure ventilation was initiated and maintained using a Bird Mark 7A ventilator. A bilateral thoracotomy was made in the fourth intercostal space and the pericardium incised to expose the heart. Walton Brodie strain gauge arches were sutured to the right and left atria. Miniature solid-state pressure transducers (Konigsberg Instruments, model P190; 5-mm diameter, 1.5-mm thick) were inserted into the midwall regions of the right ventricular conus and sinus, as well as the ventral and lateral walls of the left ventricle, to record regional intramyocardial pressures (9). These sensing devices were employed as intraventricular pressure is not an adequate method of detecting positive inotropic changes induced in ventricles by the efferent sympathetic nervous system (14). Left ventricular chamber pressure was measured using a Bentley Trantec model 800 transducer connected to a Cordis #7 catheter which was inserted into that chamber via a femoral artery. All data, including a lead II electrocardiogram, were recorded on a curvilinear eight-channel Beckman dynograph or an Astro-Med Inc. model MT9500 eightchannel rectilinear recorder.

Administration of Vehicle Saline (0.1 cc of a 0.9% w/v NaC1 solution) was injected via a 1-cc syringe attached to a Teflon-coated 30-gauge needle (Heathco #0317950) into each major cardiac ganglionated plexus studied. The effects, if any, of these injections on cardiodynamics were monitored. The investigated ganglionated plexi were: 1) the ventral right atrial ganglionated plexus located in the fatty tissue on the ventral surface of the right atrium (18); 2) the cranial, intermediate and caudal ganglionated plexi on the ventral surface of the left atrium; 3) the dorsal atrial ganglionated plexus on the dorsal surface of the right and left atrium (1,24); 4) the ganglionated plexi located on the dorsal atrial region adjacent to the inferior vena cava (the inferior vena cava-inferior atrial ganglionated plexus); and 5) the ganglionated plexus on the cranial aspect of the ventricles adjacent to the origins of the aorta and pulmonary arteries--the cranial medial ventricular ganglionated plexus. The ventral component of this ganglionated plexus lies imbedded in fat around the origins of the ventral descending and circumflex coronary arteries (7). Two smaller ganglionated plexi located adjacent to the origins of the left and right marginal arteries, the right and left lateral ventricular ganglionated plexi, were also investigated.

Experimental Protocol for Administration of Peptides in Intact Hearts Substance P induces hypotension when administered intravenously (for instance when testing its systemic effects in the dosage utilized) and thus can compromise a preparation when injected repeatedly. Thus, in order to minimize the numbers of injections of this peptide in any animal, two groups of animals were investigated.

Substance P. Atrial ganglionated plexi. In the first group of animals consisting of 10 dogs, following local injections of saline, 50 txg of substance P (peptide content --85%) in 0.1-cc normal saline was injected via a 1-cc syringe attached to a 30-gauge needle into the fat overlying the right atrial ventral surface. The dose of substance P used has been shown to activate other intrathoracic neurons involved in cardiac regulation (6). In addition, in preliminary experiments this dose was found to be sufficient to modify intrinsic cardiac neurons in such a fashion as to alter cardiodynamics in a detectable fashion. When lesser doses were injected into an active locus lesser responses were elicited, whereas when greater doses were

ARMOUR, YUAN AND BUTLER

injected similar cardiac results were induced and more profound systemic vascular hypotension occurred. The sites of injection were located in the middle of the cranial, intermediate and caudal regions of the fat containing the right atrial ventral ganglionated plexus (18). If no cardiac effects were elicited following an injection, one minute was allowed to elapse before the next injection was made. When cardiac changes were elicited at least five minutes were allowed to elapse before the next injection. All subsequent administrations of peptides were done in this fashion. When an injection of substance P into a locus in a right atrial ganglionated plexus elicited cardiac changes, similar injections were repeated into that locus after --5 minutes (one or two more times) to investigate the effects of repeated administrations of substance P into one site. Thereafter, in these 10 dogs normal saline was administered into the center of the cranial, intermediate and caudal left atrial ventral fat where ganglionated plexi exist. Then 50 fxg of substance P was administered as described above into these same loci. Thereafter, substance P (50 Ixg) was administered into the superior vena cava. In 12 animals of the second group, saline and substance P were administered, as described above, into the middle of the dorsal atrial ganglionated plexus, Then saline and substance P were administered into the inferior vena cava-dorsal atrial ganglionated plexus. Ventricular ganglionatedplexi. In 12 of the animals of the second series of experiments, normal saline was injected into the cranial medial ventricular ganglionated plexi. Thereafter, substance P was administered individually into 6-8 loci in the fat adjacent to the aortic and pulmonary artery roots where the cranial medial ventricular ganglionated plexus is located. The sites chosen were in the ventral, right, dorsal and left components of this plexus as well as the caudal extension of the ventral component adjacent to the ventral descending coronary artery (7). In 6 of these dogs injections of substance P were also made into fat adjacent to the right and left marginal arteries where the right and left lateral ventricular ganglionated plexi, respectively, are located. Bradykinin. In 14 animals, 0.1-cc saline was administered individually into cranial, intermediate and caudal loci in the right atrial ventral ganglionated plexi, as described above. Then 50 Ixg of bradykinin (peptide content - 8 5 % ) was administered into these loci, enough time being allowed to elapse between each injection for the preparation to return to control states. As with substance P, preliminary experiments had determined that administration of lesser doses did not result in maximal changes whereas administration of greater doses resulted in similar changes. Saline and bradykinin were administered into the three ventral left atrial and the single dorsal atrial ganglionated plexi in 10 of these animals and subsequently into the inferior vena cava-dorsal atria ganglionated plexus in 6 of these animals. Then saline and bradykinin were also administered individually into the ventral, dorsal, right and left components of the cranial medial ventricular ganglionated plexi in 15 dogs of this group. Bradykinin was also administered into the superior vena cava.

Oxytocin, CGRP, atriopeptin and vasoactive intestinal peptide. In the first group of 10 animals, oxytocin (1 IU in 0.1 cc of normal saline) was administered into the cranial, intermediate and caudal loci of the fight atrial ventral ganglionated plexus, as described above for administration of substance P, enough time being allowed to elapse between each injection for the preparation to return to a control state. The dose used was the same as that used previously when studying other intrathoracic cardiac neurons (8); preliminary experiments had confirmed that this dose was sufficient to yield maximal results. Then this peptide was administered into the three left atrial ventral ganglionated plexi. Injections of oxytocin preceded the other peptide administrations in two of tile animals; otherwise they occurred after the other injections.

PEPTIDE M O D U L A T I O N OF CARDIAC NEURONS

755

TABLE 1 MAXIMUM CHANGES IN THE HEART RATE (HR), RIGHT ATRIAL SYSTOLIC FORCE (RAF), LEFT ATRIAL SYSTOLIC FORCE (LAF), RIGHT VENTRICULAR CONUS (RVC IMP) AND SINUS (RVS IMP) INTRAMYOCARDIAL SYSTOLIC PRESSURES, LEFT VENTRICULAR VENTRAL WALL INTRAMYOCARDIAL SYSTOLIC PRESSURE (LV IMP) AND LEFT VENTRICULAR CHAMBER SYSTOLIC PRESSURE (IVP) INDUCED BY INJECTIONS OF SUBSTANCE P INTO VENTRAL RIGHT ATRIAL (ABOVE) OR CRANIAL MEDIAL VENTRICULAR (BELOW) GANGLIONATED PLEXI ARE TABULATED

Intervention

HR (beats/rain)

RAF (% change)

LAF (% change)

RVC IMP (mmHg)

RVS IMP (mmHg)

LV IMP (mmHg)

IVP (mm Hg)

129-+-3 129-+3 n.s.

Control Substance P p value

133-+6 172-+12 0.002

Ventral Right Atrial Ganglionated Plexus (n = 10) Intact 100±0 100-+0 22---1 20-+2 98+5 137-+14 123-+13 45±5 45-+5 126±13 0.03 n.s. 0.001 0.005 0.01

Control Substance P p value

129-+6 129___6 n.s.

100_0 100±0 n.s.

Control Substance P p value

149-+6 163-+7 0.02

Cranial Medial Ventricular Ganglionated 100±0 100-0 29-+7 124-+21 118-+21 29-+3 0.006 0.008 n.s.

Acutely Decentralized 100±0 21---1 100-+0 21-+1 n.s. n.s.

18+-2 18-+2 n.s.

87---5 87-+5 n.s.

Plexus (n = 12) 22---3 8 4 +- 10 35-+5 112-+ 18 0.008 0.005

115-+5 115+5 n.s. 109-+9 127-+15 n.s.

The results obtained by substance P administrations into the ventral right atrial ganglionated plexus are given before (intact) and after (acutely decentralized) acute decentralization of the heart and intrathoracic ganglia.

Oxytocin was administered into the dorsal left atrial ganglionated plexus, as well as the cranial medial ventricular ganglionated plexus, in 10 animals of the second group. Oxytocin was also administered into the superior vena cava of 10 animals. In eight of the first group of animals, 200 Ixg of VIP in 0. l-cc saline (peptide content - 7 5 % ) was administered into the middle of the right atrial ventral ganglionated plexus. VIP was also administered to the cranial medial ventricular ganglionated plexus of these dogs. Synthetic human calcitonin gene related peptide (20-50 Ixg; peptide content 78%) was administered into the center of the right atrial ventral ganglionated plexus and, subsequently, into the cranial medial ventricular ganglionated plexus in 6 animals of the second group. Atrial natriuretic peptide (50-100 Ixg of atriopeptin III, peptide content 89%) was administered similarly into the right atrial ventral ganglionated plexus of 4 of these animals, the dorsal atrial ganglionated plexus of 2 of these animals and the cranial medial ventricular ganglionated plexus in 6 of these animals.

sures were measured for five consecutive beats and their means -s.e.m, calculated. Data were averaged immediately prior to and during maximal responses elicited by chemical stimulation. The cardiac responses elicited in each group were then evaluated by comparing data obtained immediately prior to each intervention with maximal changes elicited following each intervention using the two-tailed Student's t-test for paired data. When none of the monitored cardiac parameters were altered following injections, such injections were considered to elicit no cardiac response. RESULTS

Vehicle Administration When 0.1 cc of saline was injected into atrial or ventricular ganglionated plexi, no changes in cardiac rate or force were detected.

Substance P Peptide Administrations Following Acute Decentralization Because cardiac responses were elicited following substance P, bradykinin and oxytocin administration, these agents were also investigated following acute decentralization of the hearts. The right and left cervical vagosympathetic complexes were sectioned and, after waiting - 5 minutes, the three peptides were readministered as described above. Thereafter, all of the connections between the fight and left stellate ganglia with the spinal cord were severed by cutting the vertebral nerves, T1 rami, T2 rami and the thoracic sympathetic chains immediately caudal to the stellate ganglia bilaterally. After waiting - 5 minutes, substance P, bradykinin and oxytocin administrations were repeated.

Data Analyses Heart rate, peak atrial systolic forces, peak systolic intramyocardial pressures and peak systolic left ventricular chamber pres-

When substance P was injected into the right atrial ganglionated plexi of l0 dogs, heart rate and right atrial force, as well as right ventricular sinus and conus and left ventricular ventral and lateral intramyocardial pressures, were augmented (Table 1). Even though left atrial force was not augmented significantly overall, it was augmented in 4 of the 9 animals. Left ventricular chamber systolic pressure was unaffected. Cardiac augmentation occurred --5-15 seconds after the injection, maximum changes occurring --10-20 seconds thereafter (Fig. 1). Cardiac changes were induced in every animal. Loci which elicited cardiac responses when injected with substance P were identified throughout the fight and left atrial or cranial ventricular collections of fat. Active loci were identified in the intermediate region of the fight atrial ventral ganglionated plexus in all 10 dogs, in the cranial region in 6 of the dogs and the caudal region in 5 of the 10 dogs. When substance P was injected into the three ventral left atrial ganglionated plexi, augmentation of left atrial force, as well as

756

ARMOUR, YUAN AND BUTLER

EKG

T L

:}

RAF

LAF 250 rnsec

10 sec ,

5O

RVC IMP

i' i

i I

(mmHg) 0

50

RVS IMP

(rnmHg)

I

0

Lv. j IMP

(mmHg)

0

L LV IMP

ioo

/ L__

(mmHg) 0

LVP (mmHg)

°

¢

FIG. 1. Injection (arrow at bottom) of 50 txg of substance P in 0.1 cc of normal saline into the right dorsal component of the cranial medial ventricular ganglionated plexus resulted in increased heart rate, right atrial force (RAF), left atrial force (LAF), right ventricular conus intramyocardial systolic pressure (RVC IMP), right ventricular sinus intramyocardial systolic pressure (RVS IMP), as well as left ventricular ventral (LVV) and lateral (LVL) wall intramyocardial (IMP) systolic pressures. Left ventricular chamber systolic pressure (LVP) was elevated. EKG = electrocardiogram.

right and left ventricular intramyocardial pressures, occurred in only 2 of the 10 animals. This happened when the intermediate and caudal portions of this ganglionated plexus were injected. When substance P was injected into the dorsal atrial ganglionated plexus cardiac augmentation occurred in 4 of the 10 animals investigated. In contrast, when this peptide was injected into the inferior vena cava-inferior atrial ganglionated plexus no cardiac changes were detected. When substance P was injected into the cranial medial ventricular ganglionated plexi of 12 animals, heart rate, as well as atrial and ventricular forces, was augmented in every animal. With the exception of right ventricular conus intramyocardial pressure, all

monitored parameters were augmented (Table 1). In some animals fight ventricular conus intramyocardial pressure was augmented as well (Fig. 1). When substance P was reinjected into loci in right atrial ventral or cranial medial ventricular ganglionated plexi which previously had elicited cardiac responses when injected with substance P, quantitatively similar responses to those previously induced were elicited. When substance P was injected into loci as close as 2 mm away from an active locus, cardiac responses usually were not elicited. Following acute decentralization of the intrathoracic autonomic nervous system, reinjection of substance P into active loci in atrial (Table 1) or ventricular ganglionated plexi induced no significant

PEPTIDE MODULATION OF CARDIAC NEURONS

757

TABLE 2 MAXIMUM CHANGES IN THE HEART RATE (HR), RIGHT ATRIAL SYSTOLIC FORCE (RAF), LEF1~ ATRIAL SYSTOLIC FORCE (LAF), RIGHT VENTRICULAR CONUS (RVC IMP) AND SINUS (RVS IMP) INTRAMYOCARDIAL SYSTOLIC PRESSURES, LEFF VENTRICULAR VENTRAL (LVV IMP) AND LATERAL (LVL IMP) WALL INTRAMYOCARDIAL SYSTOLIC PRESSURES AND LEFT VENTRICULAR CHAMBER SYSTOLIC PRESSURE (IVP) INDUCED BY INJECTIONS OF BRADYKININ INTO VENTRAL RIGHT ATRIAL (ABOVE) OR CRANIAL MEDIAL VENTRICULAR (BELOW) GANGLIONATED PLEXI ARE TABULATED

Intervention

HR RAF LAF RVC IMP RVS IMP LVV IMP LVL IMP IVP (beats/min) (% change) (% change) (mmHg) (mmHg) (mmHg) (mmHg) (mmHg)

Control Bradykinin p value

169 -+9 183-+12 n.s.

Ventral Right Atrial Ganglionated Plexus (n = 14) Sympathetic Responses (n = 7) 100_+0 22_+3 25_+3 71___9 92_+13 100 -+0 117_+8 29_+4 32_+3 82___14 119-+26 119_+9 n.s. n.s. 0.04 0.03 n.s. n.s.

Control Bradykinin p value

199 + 46 148 -+ 11 n.s.

100 -+0 84 _+7 n.s.

Control Bradykinin p value

162_+10 170_+12 0.04

Cranial Medial Ventricular Ganglionated Plexus (n = 15) Sympathetic Responses (n = 8) 100_+0 100_0 22---3 26_+4 65+10 94_+11 114_+3 105_+3 29_+5 37_+8 78+13 121-+13 0.002 n.s. 0.01 0.04 n.s. n.s.

Control Bradykinin p value

168-+9 155_+8 n.s.

100-0 81_+12 n.s.

101_+12 115_+15 n.s.

Parasympathetic Responses (n = 7) 100_+ 0 21_+2 27___4 84+15 85_+8 20+2 26___4 60_+11 n.s. n.s. n.s. 0.03

122_+13 90_+11 0.02

Parasympathetic Responses (n = 7) 100-+0 24_+4 25+4 79-+ 15 67_+7 24_+3 32+-8 51-+8 0.004 n.s. n.s. 0.01

109-+13 89_+10 0.04

85-+ 10 51_+8 0.01

97+8 93+12 n.s. 88-+8 62_+6 0.001

These results are grouped according to whether tachycardia (sympathetic responses) or bradycardia (parasympathetic responses) was induced.

changes in monitored cardiac indices overall. When the right and left cervical vagosympathetic complexes were severed while the intrathoracic sympathetic ganglia remained connected to the spinal cord, reinjections of substance P induced little or no augmentation of the heart overall. Substance P administered into the right atrial ventral ganglionated plexi in these partially decentralized preparations induced a 14% increase in right atrial force, whereas similar injections in intact preparations had augmented right atrial force by 37%. Similarly, right ventricular conal intramyocardial pressure augmentation was reduced from 210% to 31% and right ventficular sinus intramyocardial pressure from 114% to 30% following sectioning of the vagosympathetic complexes. Although in the intact state substance P initiated a 28% increase in left ventricular intramyocardial pressure, after vagosympathetic trunk sectioning no augmentation was detected in this parameter. When the right and left intrathoracic sympathetic ganglia were decentralized subsequently, thus totally decentralizing the heart, no significant changes were detected overall when substance P was reinjected into atrial or ventficular ganglionated plexus sites which previously had initiated responses (Table 1). When similar doses of substance P were administered intravenously, systemic arterial pressure fell. When systemic vascular pressure was reduced by more than - 1 5 mmHg, left ventricular intramyocardial and chamber pressures were frequently reduced as well; atrial and right ventficular indices were unchanged in these instances.

Bradykinin When bradykinin was injected into the fight atrial ganglionated plexi of 14 dogs, cardiac effects began - 5 - 1 5 seconds after the injection and peaked --10-20 seconds thereafter. Overall, sys-

temic vascular pressure was unaffected. In 7 of the animals tachycardia was induced following injection of bradykinin into the right atrial ganglionated plexus. Augmentation of right ventricular conus and sinus intramyocardial pressures was induced in these animals (Table 2). In some instances other parameters were also augmented. In the other 7 animals minimal bradycardia and atrial force suppression occurred (Table 2). Reduction of left ventricular intramyocardial and chamber systolic pressures also occurred in these animals. When 50 Ixg of bradykinin was administered intravenously, even though no significant change in systemic vascular pressure occurred overall, in a number of instances systemic vascular pressure was reduced. When bradykinin was injected into the ventral left atrial ganglionated plexus, augmentation of left atrial force, as well as fight and left ventficular intramyocardial pressures, occurred in two of the l0 animals tested. When bradykinin was injected into the dorsal atrial ganglionated plexus, augmentation of fight and left atrial forces, as well as ventricular forces, occurred in all animals tested (Fig. 2). When this peptide was injected into the inferior vena cava-inferior atrial ganglionated plexus no cardiac changes were detected. Cardiac changes were induced in all 15 animals tested when bradykinin was injected into the cranial medial ventricular ganglionated plexus. As with right atrial ganglionated plexus injections, cardiac augmentation or depression was induced, depending on the locus studied. In the eight animals in which heart rate was increased, fight atrial force, as well as right ventficular conus and sinus intramyocardial pressures, was increased (Table 2). In the other 7 animals depressor responses were elicited, particularly with respect to left atrial force and left ventricular intramyocardial pressures. Other monitored parameters were also influenced, but

758

ARMOUR, YUAN AND BUTLER

B

A EKG

~

,AF

A__f :i@lUlllml !lWIIWlUlWIIMIIIHIImll!lfl! l!l!!/

J

m

m

m

!

~

f

~

~

m!mm!!mml mmmmm!mmmm!t!.mHl!UmmU .J

5O

RVC IMP

I

(mrnHg)

~#L"'"!~'~!'t'O',,"!,'U"" '''"''''mm"m'm'"""L'''"

0 250 msec

RVS IMP

50

10 sec

--

ilulnm,t,lmlI ,,,,,,,mllt~Ul~lUIllBIIlUlflll~

ELELEL

(rnmHg)

..,

}

.............................................

' IIIIIlllIIIIIIIIIIIIIIIIMIIIIIIIIIIIIIIIIU MIIlilIIMIIIIJIIMn~

o

,oo

LVV IMP (mrnHg)

tp,m,-..,... .............. jwww,....,.......,,,..... .

o

"J AA/ ? .

.

.

.

.

.

.

.

.

100

IMP

..~

(mmHg)

._J

.-"., ~

~

k....../

, u ~ m R , amimi~.m,im.mmm...umlldlil~ "k.-.]k. ~.i,,a,,,,m,, .......... ,,m....................................

o

+

FIG. 2. (A) Injection (arrow at bottom) of 50 ixg of bradykinin in 0.1 cc of normal saline into the dorsal left atrial ganglionated plexus resulted in increased heart rate, right and left atrial forces, right ventricular conus and sinus intramyocardial systolic pressures, as well as left ventricular ventral and lateral wall intramyocardial systolic pressures. (B) Following acute decentralization, administration of bradykinin into the same locus failed to initiate cardiac changes. Abbreviations are the same as in Fig. 1.

with not enough frequency to result in significant changes overall. When bradykinin was reinjected into an active locus in atrial or ventricular ganglionated plexi, quantitatively similar responses to those previously induced were elicited. No cardiovascular responses were elicited when bradykinin was administered into loci in much of the atrial or ventricular fat, indicating the site-specific nature of elicited responses. Reinjection of bradykinin into active loci in atrial or ventricular ganglionated plexi following acute decentralization elicited no significant alterations in the monitored cardiac indices overall (Fig. 2). However, in some animals changes were elicited. When the right and left cervical vagosympathetic complexes were severed first, while leaving the intrathoracic sympathetic ganglia connected to the spinal cord, the majority of the response elicited by bradykinin was reduced.

Oxytocin When oxytocin was injected into the right atrial ganglionated plexus, cardiac responses were elicited in 5 of the 10 dogs investigated. As these injections occurred either before or after the

other peptide injections, the order of peptide administration did not appear to influence the results obtained. In these five animals heart rate was reduced from 139+7 to 120+7 beats per minute (p<0.06); right (100 to 69-+6%, p<0.001) and left (100 to 84 -+ 5%, p<0.001) atrial forces were also reduced. Right and left ventricular wall pressures were unaffected. No cardiac changes were detected when oxytocin was administered into the ventral left atrial, dorsal atrial or cranial ventricular ganglionated plexi in neurally intact preparations or when it was administered intravenously. Following acute decentralization of the heart, reinjection of oxytocin into the right atrial ganglionated plexi elicited no cardiac responses.

Calcitonin Gene Related Peptide, Atrial Natriuretic Peptide and Vasoactive Intestinal Peptide No cardiac changes occurred when CGRP (20-50 t~g), atrial natriuretic peptide (50-100 I~g) or VIP (200 Ixg) was injected into atrial or ventricular ganglionated plexi. DISCUSSION

Different peptides have been associated with neurons and

PEPTIDE MODULATION OF CARDIAC NEURONS

759

EKG

ll~

,11~\ 250msec

1Osec

1oo LVL (rnmHg)

I o

I O0

LVV IMP 0I

(mmHg)

100

LVP I o

(mmHg)

FIG. 3. Injection (arrow at top) of 1 IU of oxytocin in 0.1 cc of normal saline into the cranial region of a ventral right atrial ganglionated plexus resulted in bradycardia and a decrease in right and left atrial systolic force generation. Left ventricular lateral and ventral wall intramyocardial pressures and chamber pressure were affected minimally. Abbreviations are the same as in Fig. 1. axons in the major intrathoracic sympathetic ganglia of cats (10) and dogs (15). Peptides have also been associated with axons (21, 23, 25, 30) and cell bodies (31) on the heart. When peptides such as substance P, VIP or oxytocin are administered into an acutely decentralized canine stellate or middle cervical ganglion, cardiac chronotropism and inotropism are augmented (6,8), indicating that these peptides are capable of directly or indirectly activating intrathoracic sympathetic efferent cardiac neurons. The results of the present experiments demonstrate that peptides can also affect neurons in atrial or ventricular ganglionated plexi such that cardiac chronotropism and/or inotropism are modified. Substance P, or a closely related peptide, has been proposed to act either as a transmitter (20) or modulator (19) mediating synaptic input from primary afferent neurons to central nervous system neurons. Local application of a peptide such as substance P into nervous tissue presumably activates somata and not axons of passage (19). In keeping with this, when substance P is injected into a middle cervical or stellate ganglion, but not a cardiopulmonary nerve or subclavian ansa, cardiac augmentation occurs (6). Thus administration of substance P to cardiac ganglionated plexi presumably activates neurons, not axons, adjacent to the sites of injection. When substance P was administered into atrial or ventricular ganglionated plexi augmentation of heart rate and contractility occurred. Even though substance P reduces systemic vascular resistance when administered intravenously (16), when adminis-

tered into cardiac ganglionated plexi little or no systemic vascular pressure alteration occurred (Table 1). Thus it is unlikely that substance P leaked out of the injection sites in sufficient quantities to alter total peripheral vascular resistance to a significant degree. Rather it appears that the responses elicited were due to direct or indirect activation of sympathetic efferent cardiac neurons. That the population of neurons was located in specific regions of atrial or ventricular fat is indicated by the fact that injections as close as --1 mm away from an active locus rarely elicited cardiac responses. This is consistent with the fact that spontaneous active neurons in atrial (18) or ventricular (7) fat are found in limited numbers of loci rather than being scattered throughout the fat. The population of neurons responsive to substance P differed in the different cardiac ganglionated plexi. Neurons activated by substance P were located consistently in the ventral right atrial and cranial medial ventricular ganglionated plexi and with lesser consistency in dorsal atrial ganglionated plexi. They were rarely identified in the left atrial ventral ganglionated plexi and not at all in the inferior vena cava-inferior atrial ganglionated plexus. Cardiac responses elicited by substance P differed when this agent was administered into right atrial as compared to cranial medial ventricular ganglionated plexi. For instance, substance Psensitive neurons located in right atrial ganglia consistently modified right ventricular conus force, whereas those located in the cranial medial ventricular ganglionated plexus did not. The reverse was true with respect to left atrial force (Table 1). These data

760

ARMOUR, YUAN AND BUTLER

suggest that some degree of regional cardiac control may be exerted by substance P-sensitive neurons, depending upon where they are located on the heart. Furthermore, it appears that neurons which responded to substance P were not distributed throughout atrial or ventricular fat as they were identified in a limited number of loci in atrial or ventricular fat. The locations of these active loci varied between animals, which is consistent with previous anatomical and physiological studies which have demonstrated enormous variability in the locations of ganglia on the atria and ventricles (7,18). Since few neurons in the ventral left atrial ganglionated plexi and none in the inferior vena cava-inferior atrial ganglionated plexus responded to peptides, it appears that few of the neurons in these regions were activated in a manner to alter cardiodynamics by the peptides employed. Overall, insignificant cardiac augmentations were induced by substance P following acute decentralization of the heart. This occurred despite the fact that baseline variables were unaltered following acute decentralization (Table 1). These data imply that direct activation of the efferent sympathetic neurons which exist in cardiac ganglia (13) accounted for a fraction of the augmentor responses elicited as, if only these neurons were activated, their responses would have persisted following acute decentralization. Presumably substance P was not involved to a great extent in facilitation of preganglionic sympathetic input to neurons in cardiac ganglia as such facilitation would have persisted when the cervical vagosympathetic complexes were sectioned while the efferent sympathetic nervous system to the heart remained intact. Since cardiac neurons activated by substance P apparently had, for the most part, to be connected with central nervous system neurons in order to elicit significant cardiac responses, many of the neurons so activated presumably projected their axons centrally. That the majority of such afferent axons coursed in the vagosympathetic complexes was evident from the data obtained in partially decentralized preparations. Furthermore, since repeat injections of substance P in neurally intact preparations usually resulted in cardiac responses similar to those induced previously, it is unlikely that changes induced after acute decentralization could be attributed to tachyphylaxis to a significant degree. Afferent neurons with somata in intrathoracic (2,5) and cardiac (7,18) ganglia have been reported to exist. Perhaps such neurons were activated by substance P, accounting, in part, for the results obtained in the present experiments. Other intrinsic cardiac neurons might have been involved as well, for instance efferent peptidergic neurons innervating the heart. Bradykinin also activated neurons in atrial and ventricular ganglionated plexi such that cardiodynamics were altered. However, in contrast to substance P, bradykinin initiated both augmentor and depressor responses usually depending on the site of injection. Following bradykinin administration into the right atrial or ventricular ganglionated plexi, heart rate was increased in --50% of the animals and reduced in --50% of the animals. Augmentor responses were elicited most consistently in right ventricle force when either atrial or ventricular ganglionated plexi were investigated (Table 2). In contrast, right atrial force changes were consistently induced only when ventricular ganglionated plexi were investigated. These data suggest atrial or ventricular neurons sensitive to bradykinin may not necessarily regulate similar cardiac regions. Bradykinin is a vasodilator agent which becomes rapidly

degraded during passage through the pulmonary vasculature (28). When left ventricular intramyocardial pressure was depressed it might have been due, in part, to systemic vascular hypotension when that occurred. However, hypotension would not likely have induced the reflex bradycardia or negative atrial inotropic changes. Furthermore, when bradykinin was injected into the majority of sites in atrial or ventricular ganglionated plexi no systemic vascular pressure changes occurred. In addition, when the same dose of bradykinin was administered into the superior vena cava little consistent systemic vascular pressure change occurred overall. These data imply that left atrial and ventricular negative inotropic responses elicited following bradykinin administration into cardiac ganglionated plexi were not due to a reduction in systemic vascular resistance alone and presumably were due, in part, to direct or indirect activation of efferent parasympathetic neurons innervating the heart. Since administration of bradykinin into ganglionated plexi on acutely decentralized hearts did not elicit significant cardiac changes, it appears that cardiac responses elicited by bradykinin in intact preparations were to a large part dependent upon activation of efferent cardiac neurons via central nervous system mechanisms. That substance P and bradykinin, two peptides with algesic properties (17,26), can modify such cardiac neurons may have implications with respect to cardiac afferent neuronal mechanisms. However, as responses were still elicited in acutely decentralized preparations, bradykinin apparently can also activate intrinsic cardiac neurons, perhaps efferent peptidergic ones. Oxytocin modified neurons in right atrial ganglionated plexi with less frequency than occurred following administration of substance P or bradykinin. Thus lesser numbers of right atrial neurons appear to be responsive to oxytocin than to substance P or bradykinin. No neurons in left atrial or ventricular ganglionated plexi were modified by oxytocin, implying that the populations of oxytocin-sensitive neurons vary in different cardiac ganglionated plexi. Neurons modified by oxytocin apparently differed from those modified by substance P in another way since they initiated bradycardia and atrial force suppression. That cardiac responses elicited by oxytocin were dependent upon connectivity with the central nervous system is inferred, for reasons given above, from the fact that repeat administration of oxytocin into right atrial ganglionated plexi of acutely decentralized hearts failed to elicit cardiac responses. VIP administered into atrial or ventricular ganglionated plexi failed to induce cardiac changes, despite the fact that VIPimmunoreactive cell bodies and axons have been identified in mammalian cardiac ganglia (31). Thus it appears that few, if any, neurons on the heart are sensitive to this peptide. In contrast, similar doses of VIP administered into stellate or middle cervical ganglia elicit cardiac augmentation (6). Atrial natriuretic peptide and CGRP, two peptides involved in cardiovascular regulation (22, 29, 30), also failed to elicit significant cardiac responses, indicating that if cardiac neurons are activated by these peptides their population is relatively small or they do not result in direct or indirect activation of efferent cardiac neurons. It is concluded that substance P, bradykinin and oxytocin can modify neurons on the heart such that direct or indirect modification of cardiac rate and force occurs. Furthermore, these peptides apparently affect different populations of cardiac neurons since different cardiac responses were elicited by each.

ACKNOWLEDGEMENTS

This work was supported by a Medical Research Council of Canada grant (MA-10122) and the Nova Scotia Heart Foundation. The authors gratefully acknowledge the technical assistance of Richard Livingston.

PEPTIDE MODULATION OF CARDIAC NEURONS

761

REFERENCES

1. Ardell, J. L.; Randall, W. C. Selective vagal innervation of sinoatrial and atrioventricular nodes in canine heart. Am. J. Physiol. 251: H764-H773; 1986. 2. Armour, J. A. Synaptic transmission in chronically decentralized middle cervical and stellate ganglia of the dog. Can. J. Physiol. Pharmacol. 61:1149-1155; 1983. 3. Armour, J. A. Synaptic transmission in the decentralized middle cervical ganglion of the dog. Brain Res. Bull. 10:103-109; 1983. 4. Armour, J. A. Activity of in situ stellate ganglion neurons recorded extracellularly. Can. J. Physiol. Pharmacol. 64:101-111; 1986. 5. Armour, J. A. Neuronal activity recorded extracellularly in chronically decentralized in situ canine middle cervical ganglia. Can. J. Physiol. Pharmacol. 64:1038-1046; 1986. 6. Armour, J. A. Peptidergic modulation of efferent sympathetic neurons in intrathoracic ganglia regulating the canine heart. Proc. Soc. Exp. Biol. Med. 191:60-68; 1989. 7. Armour, J. A.; Hopkins, D. A. Activity of in situ canine ventricular neurons. Am. J. Physiol. 258: H326-H336; 1990. 8. Armour, J. A.; Klassen, G. A. Oxytocin modulation of intrathoracic sympathetic ganglionic neurons regulating the canine heart. Peptides 11:533-537; 1990. 9. Armour, J. A.; Randall, W. C. Canine left ventricular intramyocardial pressure. Am. J. Physiol. 220:1833-1839; 1971. 10. Bachoo, M.; Ciriello, J.; Polosa, C. Effect of preganglionic stimulation on neuropeptide-like immunoreactivity in the stellate ganglion of the cat. Brain Res. 400:377-382; 1987. 11. Blomquist, T. M.; Priola, D. V.; Romero, V. Source of intrinsic innervation of canine ventricles: A functional study. Am. J. Physiol. 252:H638-H644; 1987. 12. Bosnjak, Z.; Kampine, J. P. Cardiac sympathetic afferent cell bodies are located in the peripheral nervous system of the cat. Circ. Res. 64:554-562; 1989. 13. Butler, C.; Watson-Wright, W. M.; Wilkinson, M.; Johnstone, D. E.; Armour, J. A. Cardiac effects produced by long-term stimulation of acutely decentralized thoracic autonomic ganglia or cardiopulmonary nerves: Implications for inter-neuronal interactions within the thoracic autonomic nervous system. Can. J. Physiol. Pharmacol. 66:175-184; 1988. 14. Butler, C. K.; Wong, A. Y. K.; Armour, J. A. Systolic pressure gradients between the wall of the left ventricle, the left ventricular chamber, and the aorta during positive inotropic states: Implications for left ventricular efficiency. Can. J. Physiol. Pharmacol. 66: 873-879; 1988. 15. Darvesh, S.; Nance, D. M.; Armour, J. A.; Hopkins, D. A. Distribution of neuropeptide-like immunoreactivity in intact and chronically decentralized middle cervical and stellate ganglia of dogs. J. Auton. Nerv. Syst. 21:167-180; 1987. 16. Euler, U. S. von; Gaddum, J. H. An unidentified depressor substance

in certain tissue extracts. J. Physiol. (Lond.) 72:74-87; 1931. 17. Foreman, R. D.; Blair, R. W.; Amons, W. S. Neural mechanisms of cardiac pain. In: Cerrero, F.; Morrison, J. F. B., eds. Visceral sensation. Amsterdam: Elsevier Science Publishers; 1986:227-243. (Prog. Brain Res., vol 67.) 18. Gagliardi, M.; Randall, W. C.; Bieger, D.; Wurster, R. D.; Hopkins, D. A.; Armour, J. A. Activity of in vivo canine cardiac plexus neurons. Am. J. Physiol. 255:H789-H800; 1988. 19. Henry, J. L.; Kmjevic, K.; Morris, M. E. Substance P and spinal neurons. Can. J. Physiol. Pharmacol. 53:423-432; 1975. 20. H/Skfelt, T.; Kellerth, J. O.; Nilsson, G.; Pernow, B. Experimental immunohistochemical studies on the localization and distribution of substance P in cat primary sensory neurons. Brain Res. 100:235-252; 1975. 21. Hoover, D. B.; Hancock, J. C. Distribution of substance P binding sites in guinea-pig heart and pharmacological effects of substance P. J. Auton. Nerv. Syst. 23:189-197; 1988. 22. Lattion, A.-L.; Michel, J.-B.; Arnauld, E.; Corvol, P.; Soubrier, F. Myocardial recruitment during ANF mRNA increase with volume overload in the rat. Am. J. Physiol. 251:H890-H896; 1986. 23. Lundberg, J. M.; Franco-Cereceda, A.; Hua, X.; H6kfelt, T.; Fisher, J. A. Co-existence of substance P and catcitonin gene-related peptidelike immunoreactivities in sensory nerves in relation to cardiovascular and bronchoconstrictor effects of capsaicin. Eur. J. Pharmacol. 108:315-319; 1985. 24. Randall, W. C.; Kay, M. P.; Thomas, J. X.; Barber, M. J. Intrapericardial denervation of the heart. J. Surg. Res. 29:101-109; 1980. 25. Reinecke, M.; Weihe, E.; Forssmann, W. G. Substance P-immunoreactive nerve fibers in the heart. Neurosci. Lett. 20:265-269; 1980. 26. Rioux, F.; Bachelard, H.; St-Pierre, S.; Barabr, J. Epicardial application of bradykinin elicits pressor effects and tachycardia in guinea pigs. Possible mechanisms. Peptides 8:863-868; 1987. 27. Sato, A.; Kimura, S.; Goto, K. Calcitonin gene-related peptide as potential neurotransmitter in guinea pig right atrium. Am. J. Physiol. 250:H693-H698; 1986. 28. Steranka, L. R.; Farmer, S. G.; Burch, R. M. Antagonists of B 2 bradykinin receptors. FASEB J. 3:2019-2025; 1989. 29. Verburg, K.; Freeman, R. H.; Villarreal, D.; Brands, M. W. Cardiovascular and renal effects of calcitonin gene-related peptide in hypertensive dogs. Peptides 10:663-669; 1989. 30. Wang, B. C.; Bie, P.; Leadley, R. J., Jr.; Goetz, K. L. Cardiovascular effects of calcitonin gene-related peptide in conscious dogs. Am. J. Physiol. 257:R726-R731; 1989. 31. Weihe, E.; Reinecke, M.; Forssmann, W. G. Distribution of vasoactive intestinal polypeptide-like immunoreactivity in the mammalian heart. Cell Tissue Res. 236:527-540; 1984.