Function of human intrinsic cardiac neurons in situ

Am J Physiol Regul Integr Comp Physiol 280:R1736-R1740, 2001. You might find this additional info useful... This article cites 6 articles, 3 of which can be accessed free at: http://ajpregu.physiology.org/content/280/6/R1736.full.html#ref-list-1 This article has been cited by 1 other HighWire hosted articles Heart rate variability Harald M. Stauss Am J Physiol Regul Integr Comp Physiol, November 1, 2003; 285 (5): R927-R931. [Full Text] [PDF] Updated information and services including high resolution figures, can be found at: http://ajpregu.physiology.org/content/280/6/R1736.full.html

Rakesh Christopher Arora, Gregory Matthew Hirsch, Kristine Johnson Hirsch, Camille Hancock Friesen and John Andrew Armour

Downloaded from ajpregu.physiology.org on January 2, 2012

Additional material and information about American Journal of Physiology - Regulatory, Integrative and Comparative Physiology can be found at: http://www.the-aps.org/publications/ajpregu

This infomation is current as of January 2, 2012.

American Journal of Physiology - Regulatory, Integrative and Comparative Physiology publishes original investigations that illuminate normal or abnormal regulation and integration of physiological mechanisms at all levels of biological organization, ranging from molecules to humans, including clinical investigations. It is published 12 times a year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright 2001 by the American Physiological Society. ISSN: 0363-6119, ESSN: 1522-1490. Visit our website at http://www.the-aps.org/.

Am J Physiol Regulatory Integrative Comp Physiol 280: R1736R1740, 2001.

Function of human intrinsic cardiac neurons in situ

RAKESH CHRISTOPHER ARORA,1 GREGORY MATTHEW HIRSCH,1 KRISTINE JOHNSON HIRSCH,2 CAMILLE HANCOCK FRIESEN,1 AND JOHN ANDREW ARMOUR3 Departments of 1Surgery, 2Anesthesiology, and 3Physiology and Biophysics, Dalhousie University, Halifax, Nova Scotia, B3H 4H7, Canada
Received 29 November 2000; accepted in nal form 9 February 2001

Arora, Rakesh Christopher, Gregory Matthew Hirsch, Kristine Johnson Hirsch, Camille Hancock Friesen, and John Andrew Armour. Function of human intrinsic cardiac neurons in situ. Am J Physiol Regulatory Integrative Comp Physiol 280: R1736R1740, 2001.We sought to determine the behavior of intrinsic cardiac neurons in human subjects undergoing cardiac surgery and to correlate their activity with hemodynamics status. A lead II electrocardiogram, pulmonary artery pressure, and systemic arterial pressure were recorded along with extracellular activity generated by right atrial neurons in 10 patients undergoing coronary artery bypass surgery. Identied neurons generated spontaneously activity that was, for the most part, unrelated to the cardiac cycle. Most neurons were activated by gentle mechanical distortion of ventricular epicardial loci. The activity generated by neurons in each patient increased when arterial pressure increased and decreased when arterial pressure fell. Intrinsic cardiac neurons continued to generate activity during cardioplegia and cardiopulmonary bypass, but at reduced levels. Normal neuronal activity was restored postbypass. It is concluded that human intrinsic cardiac neurons generate spontaneous activity and that many receive inputs from ventricular mechanosensory neurites. The latter may account for the fact that their behavior depends, in part, on cardiac dynamics. They are also sensitive to intravenously administered pharmacological agents. These data also indicate that cardiopulmonary bypass and cardioplegia do not induce residual depression of their function. -adrenoceptor agonist; cardiac mechanosensory neurites; cardioplegia; cardiopulmonary bypass

(5, 9). It remains to be established whether neurons in the human heart behave in a manner similar to that identied in animal models. Furthermore, we do not know what effects cardiopulmonary bypass (CPB) and cardioplegia have on the human intrinsic cardiac nervous system. The present experiments were designed to determine whether human intrinsic cardiac neurons generate spontaneous activity and, if so, how they respond to altered cardiovascular status. Additionally, we sought to determine whether cardiac afferent inputs to the human intrinsic cardiac nervous system alter its behavior. Finally, we sought to determine whether CPB and cardioplegia exert deleterious effects on human intrinsic cardiac neuronal function.
MATERIALS AND METHODS

Downloaded from ajpregu.physiology.org on January 2, 2012

MULTIPLE NEURONAL SUBTYPES

have been identied within the mammalian intrinsic cardiac nervous system, both anatomically (4, 6, 8, 10) and functionally (1, 2). It has been proposed that the intrinsic cardiac nervous system is important for the maintenance of adequate cardiac output (11), particularly in disease states such as myocardial ischemia (2). Canine intrinsic cardiac neurons display complex behavioral patterns that rely, to a considerable degree, on their cardiovascular sensory inputs. These inputs, in turn, are dependent on cardiovascular status (1, 2). It is known that the human cardiac efferent nervous system displays functional characteristics similar to those found in animals

General methods. Ethical approval for human experimentation at our institution is reached through a clinical and layperson peer-reviewed process at the Queen Elizabeth II Hospital, Halifax, Nova Scotia, Canada. Once full ethical approval of the study was achieved, patients were approached for entry into the study in accordance with guidelines set forth by the Queen Elizabeth II Hospital Ethics Committee. To qualify for entry into this study, a patient needed to have angiographic evidence of a lesion in at least one major coronary artery that required the coronary artery bypass grafting (CABG) procedure (Table 1). Patients were excluded from this study if they had depressed left ventricular function, required another procedure combined with the CABG procedure, or refused to consent to the recording procedure. Ten patients scheduled for CABG procedures entered this study. Basic demographic data and comorbid variables were collected from each patient upon entry into the study. Left ventricular ejection fractions were determined in each patient by means of transthoracic or transesophageal echocardiography, radionucleotide imaging, or left ventriculography performed at time of selective coronary catheterization. Operative variables relating to pump and aortic crossclamp times, times to extubation postprocedure, as well as intensive care unit and total length of stay were assessed for each patient. Operative procedures. After instigation of general anesthesia, a midline sternotomy was performed to expose the heart, and the appropriate bypass conduit was harvested in the usual
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. http://www.ajpregu.org

Address for reprint requests and other correspondence: J. A. Armour, Dept. of Physiology and Biophysics, Dalhousie Univ., Halifax, Nova Scotia, B3H 4H7, Canada (E-mail: jarmour@is.dal.ca). R1736

0363-6119/01 $5.00 Copyright 2001 the American Physiological Society

HUMAN INTRINSIC CARDIAC NEURONAL FUNCTION

R1737

Table 1. Tabulated data on patients age, sex, diseased vessel territories, and risk factors
Age Sex, male/female Number of disease vessel territories (median IQR) Patients with proximal right coronary artery lesion Patients with cardiac risk factors Current smoker Non-insulin-dependent diabetes mellitus Hypertension Hyperlipemia Medications -blockers Ca2 channel blocker Angiotensin-converting enzyme inhibitor-I Nitrates Ejection fraction, % Pump time, min Cross-clamp time, min Time to extubation, h Length of stay, days (median IQR) 61.7 2.8 (range, 4974) 8/2 3 (23) 6 1 4 2 7 10 4 3 8 61.0 2.9 110.6 12.0 84.2 11.3 7.2 1.3 8 (59)

Values are means SE. Medications provided to patients are listed along with left ventricular ejection fractions, pump time, aortic cross-clamp time, time to extubation, and length of hospital stay. IQR, interquartile range.

fashion for CABG. Once the pericardotomy was completed, baseline systemic and pulmonary artery pressures were recorded along with a two-lead (leads II and V5) electrocardiogram (ECG). Patients were heparinized and underwent aortoright atriocaval cannulation in case CPB compromised hemodynamics, so that CPB could be instigated immediately. Recording neuronal activity. Fatty tissue on the lateral surface of the right atrium that contains the right atrial ganglionated plexus (RAGP) (4, 6, 10) was exposed. A tungsten recording microelectrode was employed to record the extracellular activity generated by right atrial neurons (7). The microelectrode had a shank diameter of 100 m, an exposed tip of 5 m, and an impedance of 911 M at 1,000 Hz. The electrode was held in place by a micromanipulator attached to a variable extension arm (Octopus 1; Medtronic, Minneapolis, MN) to stabilize its motion. The fat on the lateral surface of the right atrium was explored with this microelectrode from its epicardial surface and more deeply to adjacent atrial tissue. An indifferent electrode was attached to the adjacent mediastinum. The RAGP was chosen for investigation because it contains one of the largest collections of intrinsic cardiac neurons without an associated large coronary artery (4). Thus we could explore it with an electrode without potentially injuring a major coronary artery. The midline sternotomy permitted easy access to this plexus and thereby minimized stimulation of mechanosensory nerve endings located in the epicardium that could confound neuronal activity results. The activity generated by right atrial neurons so recorded was amplied differentially by means of two Princeton Applied Research (model 113) ampliers that had band-pass lters set at 300 Hz to 10 kHz and amplication ranges of 100500 , placed in series. The output of these batterydriven ampliers was led to an audio monitor as well as an Astromed MT9500 eight-channel rectilinear chart recorder. Action potentials generated by individual neurons with signal-to-noise ratios greater than 3:1 were studied, with individual neural units being identied by the amplitude and

conguration of their action potentials. Gradually moving the electrode away from an active site reduced the amplitude of recorded action potentials without altering their congurations. Action potentials generated by adjacent axons of passage are of similar magnitude as the background noise. Therefore, with these techniques and criteria, the recording microelectrode identies action potentials generated by neuronal somata (cell bodies) and/or dendrites. Intraoperative procedures. Once an active site was identied, various loci on the exposed epicardial surfaces of the right and left ventricles were touched gently. The epicardial mechanical stimuli so applied were insufcient to distort the heart and thus alter the position of the recording electrode tip. This procedure was performed to determine whether identied RAGP neurons received mechanosensory inputs from such epicardial regions. Then systemic arterial pressure was reduced or increased by 30% after administration of nitroglycerin (50100 g iv boluses) or phenylephrine (50 100 g iv boluses), respectively. Neuronal activity and cardiovascular variables were also monitored during the following interventions: 1) just before instigating total CPB, 2) when the patient was on full CPB before the aortic cross clamp was applied, 3) during cardioplegia as ECG quiescence occurred, and 4) at the end of the initial cardioplegia infusion before starting the rst distal coronary artery anastomosis. A combination of cold blood and crystalloid (4:1 ratio) is employed for cardioplegia at our institution during CPB. Once the coronary revascularization procedures had been completed, neuronal activity and cardiovascular variables were monitored as CPB was discontinued. Lastly, variables were monitored during the subsequent administration of protamine hydrochloride in 0.9 sodium chloride (150250 mg iv). This agent is routinely employed after completion of the CABG procedure to reverse the effects of the previously administered heparin. Protamine was administered over 10- to 15-min periods. Monitored hemodynamic variables were unaffected by this intervention. Data analysis. Cardiac variables were analyzed over 30-s periods of time before and during peak responses elicited by each of the interventions described above. Action potentials with signal-to-noise ratios greater than 3:1 generated within a locus of the RAGP were counted for 30-s periods of time to establish average activity immediately before and during maximal responses elicited by each intervention. Fluctuations in the amplitude of action potentials generated by individual neurons varied by 25 V over several minutes, retaining their same congurations over time. Action potentials recorded in a given locus with the same conguration and amplitude were considered to be generated by the somata and/or dendrites of a single neuron. The means ( SE) of data recorded during control states as well as during each intervention were calculated. ANOVA and paired t-test with Bonferroni correction for multiple tests were employed for statistical analysis where appropriate. A signicance value of P 0.01 was used for these determinations.
RESULTS

Downloaded from ajpregu.physiology.org on January 2, 2012

Patient and operative variables. Patient demographics, comorbid variables, and medications are listed in Table 1. The average age of patients was 62 yr, with 80% of the patients being male. All patients had preserved left ventricular function. All but one patient received at least two grafts that consisted of a left internal mammary artery to the left anterior descending artery and either a radial artery and/or a reversed saphenous vein graft to the remaining diseased terri-

R1738

HUMAN INTRINSIC CARDIAC NEURONAL FUNCTION

Fig. 1. Continuous recording of the activity generated by right atrial neurons and cardiovascular indexes obtained before and after administering a bolus dose of phenylephrine into the circulation (arrow, bottom). Neuronal activity increased soon after this occurred and before any change in recorded cardiovascular variables was detected (note that a paper fold prevented recording partway through the record). ECG, electrocardiogram; AP, aortic pressure; PAP, pulmonary pressure (the same abbreviations are used in the other gures).

tories. Two patients were classied as in-house urgent cases, having been in the coronary care unit awaiting the CABG procedure. There were no complications arising from the intraoperative recording procedure. No in-hospital deaths occurred in this cohort of patients. The majority of patients was extubated within 47 h postoperatively. Both in-house urgent patients had prolonged hospital stays (17 and 18 days). One of these urgent patients developed left lower lobe pneumonia postoperatively. The other patient had low cardiac output postoperatively requiring low doses of dopamine and epinephrine for 24 h. The rest of the patients had uncomplicated recoveries. Neuronal activity. Using the criteria mentioned above, we identied spontaneous activity generated by two to three intrinsic cardiac neurons in each patient Table 2. Alterations in heart rate, pulmonary artery pressure, and aortic pressure, as well as activity generated by right atrial neurons recorded during various interventions
Heart Rate, beats/min Pulmonary Artery Pressure, mmHg Aortic Pressure, mmHg Neuronal Activity, impulses/min

Interventions

Phenylephrine administration, n Before After Before After Before During Before After 71 69 68 68 3 3 24 27 1/12 2/14 1 2 90 126 5/56 7/69

10 2 3 3 2 59 167 144 45 11 18* 27 10* 17 46* 19 46

Nitroglycerin adminstration, n

10

2 26 2/13 1 111 5/66 3 23 2/12 1 95 4/57 Touching epicardium, n 7

72 3 20 2/10 1 100 5/58 2 66 70 3 20 2/10 2 104 6/66 4 202 Cardiopulmonary bypass and cardioplegia, n 9 73 75 3 3 22 23 2/11 3/12 1 1 100 6/60 89 10/69 2 5 83 114

Values are means SE. Neuronal activity increased when aortic pressure increased after phenylephrine administration. Neuronal activity decreased as systemic pressure fell after systemic administration of nitroglycerin. Touching ventral ventricular epicardial loci activated neurons in 7 patients. Neuronal activity was similar before and after cardiopulmonary bypass (pump) and cardioplegia. *P 0.01 control vs. intervention.

before any intervention. Multiple neuronal activity was evidenced by the different amplitudes of their recorded action potentials (Fig. 1; Table 2). The conguration and amplitude of each identied neuronal unit did not vary over time, allowing for comparison of changes in neuronal activity before and after an intervention. The activity so identied was sporadic in nature and, for the most part, not related to a specic phase of the cardiac cycle (Fig. 1). Immediately before the CPB, the activity generated by right atrial neurons in anesthetized patients averaged 59 11 impulses per minute. At that time, these patients systemic arterial pressure was, on average, 90/56 mmHg (Table 2). The activity generated by identied, spontaneously active neurons increased in seven of the 10 patients when limited loci on the exposed anterior epicardium of the right or left ventricles were touched gently (Table 2). No responses were elicited in the other three patients when the exposed surfaces of the two ventricles were touched. In six of these seven patients, additional neurons were recruited when mechanical stimuli were applied to the epicardium, as determined by the varied congurations of the action potentials identied (Fig. 2). When systemic arterial pressure increased after systemic administration of the -adrenoceptor agonist phenylephrine, intrinsic cardiac neuronal activity increased (Fig. 1; Table 2). This included the recruitment of new neuronal units with different amplitudes than identied during baseline conditions (before phenylephrine administration). In some instances, the activity generated by right atrial neurons increased before any detectable changes in monitored cardiovascular variables became evident (Fig. 1). In contrast, when systemic arterial pressure was reduced by the systemic administration of nitroglycerin, intrinsic cardiac neuronal activity decreased (Table 2). The activity generated by intrinsic cardiac neurons remained at control levels (57.4 18.4 impulses per minute) upon instigation of CPB before the application of the aortic cross clamp and infusion of cardioplegia. Intrinsic cardiac neurons continued to generate activity (41.1 28.6 impulses per minute) during infusion of the cardioplegia solution (Fig. 3B). Neuronal activity persisted when the ECG became quiescent, but at a reduced

Downloaded from ajpregu.physiology.org on January 2, 2012

HUMAN INTRINSIC CARDIAC NEURONAL FUNCTION

R1739

Fig. 2. Neuronal responses elicited by a gentle touch of a locus on the left ventricular ventral epicardium (horizontal line below). A burst of activity was generated by right atrial neurons (neural, bottom) during application of this stimulus. Monitored cardiovascular variables remained unchanged.

level. For instance, during a period of prolonged cardiac standstill after the initial cardioplegia period, right atrial neurons generated 21.0 12.1 impulses per minute (P 0.01, compared with control values; Fig. 3C). After completing the CABGs and weaning the patient from CPB, neurons generated activity levels that were similar to those recorded before these interventions were instigated (Table 2). At the end of the procedure, protamine hydrochloride was administered into the systemic circulation to reverse the effects of previously administered heparin. When neuronal activity was monitored in four patients during this intervention, we observed that this peptide enhanced the activity generated by some neurons while activating previously quiescent neurons in other in-

Fig. 4. A burst of activity was generated by right atrial neurons soon after the administration of protamine into the circulation commenced (arrow, bottom). Recorded cardiovascular indexes remained unchanged.

Downloaded from ajpregu.physiology.org on January 2, 2012

stances (Fig. 4). This occurred without any alterations in monitored cardiovascular variables being detected. No untoward reactions were elicited after protamine administration.
DISCUSSION

The results obtained in these experiments demonstrate that populations of human intrinsic cardiac neurons generate spontaneous activity in patients undergoing cardiac surgery after anesthesia is induced and a midline sternotomy is performed. These data support those derived from the canine model (7) in as much as,

Fig. 3. A and B represent a continuous record of the ECG, AP, and right atrial neuronal activity obtained from the beginning of infusing cardioplegia solution until the point of ECG quiescence. Note that neuronal activity persisted in the presence of suppressed cardiac electrical activity (B). C: data obtained after cardioplegia infusion had been instituted for some time, just before the rst distal anastomosis was performed.

R1740

HUMAN INTRINSIC CARDIAC NEURONAL FUNCTION

rst, the activity generated by right atrial neurons was for the most part sporadic in nature and thus unrelated to the cardiac cycle (Fig. 1). Second, as has been found in experimental animals, populations of human intrinsic cardiac neurons receive inputs from ventricular mechanosensory neurites. Third, cardioplegia or CPB procedures do not appear to reduce the capacity of human intrinsic cardiac neurons to generate spontaneous activity after their discontinuance (Table 2). Fourth, the activity generated by human intrinsic cardiac neurons is dependent, in part, on cardiodynamic status. Right atrial neurons were activated in most patients when limited loci on the ventral epicardial surfaces of either ventricle were touched briey (Fig. 2). Presumably, not all investigated neurons responded to this intervention (Table 2) because not all of them received mechanosensory inputs from the epicardial areas investigated with local mechanical stimuli, as occurs in animal models (2, 7). The fact that human intrinsic cardiac neurons receive cardiac mechanosensory inputs may explain, in part, the fact that the activity generated by many identied right atrial neurons changed in concordance with alterations in cardiovascular status (Fig. 1). For instance, when systemic arterial pressure increased as a consequence of the administration of phenylephrine (representing an increase in afterload), neuronal activity increased. Likewise, when systemic arterial pressure decreased after systemic administration of nitroglycerin (representing a reduction in afterload and preload), neuronal activity decreased. These data support those derived from the canine model, in as much as animal intrinsic cardiac neurons are known to receive direct inputs from ventricular mechanosensory neurites and thus are sensitive to changing cardiodynamics (2, 7). In accord with that, most identied neurons became inactive when systemic arterial pressure fell below 60 mmHg. Presumably, that was primarily due to a relative reduction of cardiac mechanosensory inputs to the intrinsic cardiac nervous system. This agrees with the fact that neuronal activity was lowest during cardiac standstill when the ECG was quiescent. Administering the -adrenergic agonist phenylephrine increased right atrial neuronal activity concomitant with increases in systemic arterial pressure (Table 2). These neuronal responses accompanied changes in monitored cardiovascular indexes, presumably reecting increased sensory inputs as well as increasing inputs arising from central reexes as a result of global changes in cardiovascular status. In some patients, activation of right atrial neurons occurred even before any detectable changes in monitored cardiovascular indexes became evident (Fig. 1). Populations of canine intrinsic cardiac neurons are known to be sensitive to exogenous applied -adrenoceptor agonists (3). Thus -adrenoceptor agonists may directly affect some human intrinsic cardiac neurons. Human intrinsic cardiac neurons also proved to be sensitive to protamine (Fig. 4). The fact that some atrial neurons were modied by this peptide is in accord with the fact that canine intrinsic cardiac neurons are sensitive to multiple chemicals, including peptides and amino acids (2).

Because intrinsic cardiac neuronal activity was restored to baseline values by the end of bypass (Table 2), cardioplegia and CPB do not appear to adversely affect the capacity of human intrinsic cardiac neurons to generate spontaneous activity. We have proposed that proper functioning of the nal common regulator of cardiac behavior, the intrinsic cardiac nervous system, may be important in the maintenance of adequate cardiac output (11). This aspect of the human intrinsic cardiac nervous system may be relevant with respect to modifying cardiac function in the perioperative period. Perspectives Human intrinsic cardiac neurons generate spontaneous activity, as is found in animal models. Furthermore, as in the canine model, human intrinsic cardiac neuronal activity is dependent on cardiovascular status. Exogenously administered therapeutic agents can also modify the neurons behavior. These data provide a basis for the development of novel therapy targeting the human intrinsic cardiac nervous system in the perioperative period. The fact that intrinsic cardiac neurons retain their function after CPB implies that the human intrinsic cardiac nervous system can be manipulated to advantage in the postoperative period.
The authors gratefully acknowledge the technical assistance of R. Livingston. This work was supported by the Medical Research Council of Canada (MA-10122), the Nova Scotia Heart and Stroke Foundation, and the Queen Elizabeth II Foundation. REFERENCES 1. Ardell JL. Anatomy and function of mammalian intrinsic cardiac neurons. In: Neurocardiology, edited by Armour JA and Ardell JL. New York: Oxford University Press, 1994, p. 95114. 2. Armour JA. Anatomy and function of the intrathoracic neurons regulating the mammalian heart. In: Reex Control of the Circulation, edited by Zucker IH and Gilmore JP. Boca Raton, FL: CRC, 1991, p. 137. 3. Armour JA. Intrinsic canine cardiac neurons involved in cardiac regulation possess 1-, 2-, 1- and 2-adrenoceptors. Can J Cardiol 13: 277284, 1997. 4. Armour JA, Murphy DA, Yuan BX, MacDonald S, and Hopkins DA. Anatomy of the human intrinsic cardiac nervous system. Anat Rec 297: 289298, 1997. 5. Carlson MD, Geha AS, Hsu J, Martin J, Levy MN, Jacobs G, and Waldo AL. Selective stimulation of parasympathetic nerve bers to the human sinoatrial node. Circulation 85: 1311 1317, 1992. 6. Davies F, Francis ETB, and King TS. Neurological studies on the cardiac ventricles of mammals. J Anat 86: 302309, 1952. 7. Gagliardi M, Randall WC, Bieger D, Wurster RD, Hopkins DA, and Armour JA. Activity of in vivo canine cardiac plexus neurons. Am J Physiol Heart Circ Physiol 255: H789H800, 1988. 8. King TS and Coakley JB. The intrinsic nerve cells of the cardiac atria of mammals and man. J Anat 92: 353376, 1958. 9. Murphy DA, Johnstone DE, and Armour JA. Preliminary observations on the effects of stimulation of cardiac nerves in man. Can J Physiol Pharmacol 63: 649655, 1985. 10. Singh S, Johnson PI, Lee RE, Orfei E, Lonchyna VA, Sullivan HJ, Montoya A, Tran H, Wehrmacher WH, and Wurster RD. Topography of cardiac ganglia in the adult human heart. J Thorac Cardiovasc Surg 112: 943953, 1996. 11. Stevenson RS, Thompson GW, Wilkinson M, Murphy DA, and Armour JA. Neuronal-induced augmentation of cardiac output. Can J Cardiol 15: 13611366, 2000.

Downloaded from ajpregu.physiology.org on January 2, 2012