- Email: [email protected]
Distribution of vagal afferent fibers of the guinea pig heart labeled by anterograde transport of conjugated horseradish peroxidase
Journal of the Autonomic Nervous System, 36 (1991) 13-24
13
© 1991 Elsevier Science Publishers B.V. All rights reserved 0165-1838/91/$03.50 JANS 01200
Distribution of vagal afferent fibers of the guinea pig heart labeled by anterograde transport of conjugated horseradish peroxidase Mark Quigg Department of Anatomy, Karolinska Institutet, Stockholm, Sweden, and Department of Neurology, University of Virginia, Charlottesville, Virginia, US.A. (Received 16 August 1990) (Revision received 15 April 1991) (Accepted 17 May 199l)
Key words: Heart; Vagal sensory fiber; Nodose ganglia; Horseradish peroxidase conjugate; Anterograde transport; Guinea pig Abstract To determine the distribution of vagal afferent fibers in the heart, wheat germ agglutinin conjugated horseradish peroxidase (WGA-HRP) or choleragenoid conjugated horseradish peroxidase (B-HRP) was injected into nodose ganglia of guinea pigs. Anterogradely labeled fibers and beaded, terminal-like arborizations were observed in the ascending aorta and aortic arch, the pulmonary trunk and arteries, posterior atrial walls, atrioventricular valves, and ventricles. Control experiments with injection of B - H R P into the cervical vagus nerve indicated that labeled fibers observed in the heart originated from sensory neurons in the nodose ganglia. Neither the density nor distribution of labeling differed between W G A - H R P and B-HRP. Injection of tracer into the left or right nodose ganglion shows that these regions of the heart are bilaterally innervated, although labeling in the left or right posterior atrium was denser after injection into the ipsilateral ganglion. Comparison with a previous study on the distribution of sympathetic afferent fibers in the guinea pig heart suggests that the two afferent systems maintain a complementary, but not mutually exclusive, distribution within the ventricles. Whereas afferents with their source in the spinal ganglia are mainly distributed with coronary arteries on the anterior-superior surface of the ventricles, afferent fibers with their source in the nodose ganglia are concentrated within peri-arterial regions of the posterior-inferior ventricular epicardium and the posterior septal ventricular endocardium. These differences in distributions of afferent systems could play a role in post-infarction autonomic dysfunction and in the symptoms that accompany angina pectoris.
Introduction
The heart receives its afferent innervation from both vagal and sympathetic nerves [26,30]. In the
Correspondence: M. Quigg, c / o H~kan Aldskogius, Department of Anatomy, Karolinska Institutet, Box 60400, S-10401 Stockholm, Sweden.
case of the vagal afferent system, sensory cell bodies located within the nodose ganglia send central fibers to the nucleus tractus solitarius and the dorsal motor vagal nucleus [19]. Their peripheral axons descend the vagus to enter the cardiac plexus at the base of the heart. The peripheral distribution of vagal afferent fibers and their sensory endings within the heart is still unclear because of the difficulty in distinguishing vagal from
14
sympathetic and afferent from efferent innervation. Clinical observations of autonomic symptoms after acute myocardial infarction suggest that the distribution of these two afferent systems are not homogeneous [23]. Anatomical confirmation, however, is difficult to find among earlier studies [6,12,14,16,18,20] of cardiac afferent innervation that have depended upon unspecific nerve labeling methods. Conversely, a subset of cardiac afferent fibers that contain neuropeptides [8] can be specifically labeled by immunohistochemical methods. Accordingly, fibers containing substance P [4,7,9,24,28,33] and calcitonin generelated peptide [25,34] have been identified in various regions of the heart and coronary arteries. These studies, however, do not exclude contributions from the dual innervation of the heart.
In the present study, anterogradely transported conjugates of horseradish peroxidase, wheat germ agglutinin conjugate (WGA-HRP) [1] and choleragenoid /3 subunit conjugate (BHRP) [15], are used to visualize selectively the distribution of vagal afferent fibers and endings within the heart. These results are compared to those of an earlier study in which cardiac sympathetic afferent fibers were studied With the same technique [27].
Materials and Methods
Male and female guinea pigs (200-400 g, n = 6) were anesthetized with xylazine (2.5 mg/kg i.m., Bayer) and ketamine (50 mg/kg i.m., Parke-
.A
I
Es
I
1 cm Fig. 1. Cardiac innervation map of vagal afferent fibers with origins in the left nodose ganglion labeled by anterograde transport of B - H R P after 31 h post-injection survival time. Black squares mark locations of labeled fibers or fiber groups in this cranial to caudal (A-I) series of transverse sections of the guinea pig heart. T: trachea; Es: esophagus; Ao: aorta; PA: pulmonary artery; RA, RAur, RV: right atrium, auricle, ventricle; LA, LAur, LV: left atrium, auricle, ventricle; Br: bronchi.
15 Davis), and either the left or right nodose ganglion was surgically exposed. 1 /zl of 2% W G A HRP (Sigma) or 1% B - H R P (List Biol. Labs Campbell, CA) in distilled water was pressure-injected with a glass micropipette over a period of 5-10 min to eliminate leakage. Afterwards, the pipette was withdrawn slowly to discourage backflow of tracer. After survival times ranging from 12 to 52 h, the animals were re-anesthetized and cannulated in the abdominal aorta. The animals were retrogradely perfused with a prewash of body temperature 0.9% saline and then fixed with 2 1 of 2% paraformaldehyde and 1.5% glutaraldehyde in phosphate buffer (ice-cold, pH 7.4). Nodose ganglia, vagus nerve, mediastinum, and brainstem were excised, postfixed by 15-30 min immersion in fixative, and then rinsed and stored overnight in 30% sucrose/phosphate buffer solution (ice-cold, pH 7.4). Brainstems and vagal nerves were sectioned transversely and longitudinally respectively at 50 /xm on a freezing microtome and incubated with tetramethyl benzidine (TMB, Sigma) [17] to check for successful nodose injections. If the distal vagal nerves showed anterogradely-transported tracer, and if brainstem perikarya, labeled by uptake and retrograde transport of tracer by efferent axons passing through the nodose ganglion [15,29], were found in the ambiguous and dorsal motor nuclei in the brainstem, the corresponding mediastinum was then processed. Before sectioning, pieces of the inferior vena cavae (IVC) and parietal pericardia were stripped from the mediastina and saved for incubation with mediastinal sections. Mediastina were sectioned transversely at 90 /xm with a freezing microtome; prior to freezing, the heart chambers were filled with a 2:1 mixture of 30% sucrose/ phosphate buffer and Tissue-Tek (Miles Laboratories, Naperville, IL). This reinforcement and minimum handling kept the sections intact and untangled during TMB incubation and mounting. The mediastinal sections were traced at a magnification of 5.5 x with the aid of a light projector; labeled fibers and sensory endings identified by the presence of a fine, dark blue HRP reaction product were scanned under a light microscope at a total magnification of 250 x and mapped onto
the traced sections as shown in Fig. 1. From each specimen every other section of all animals was mapped. Mapped sightings were counted by anatomical region to determine a regional frequency distribution, thereby allowing comparisons in labeling between different parts of the heart after left versus right nodose ganglia injection. To confirm that the labeling observed within the heart was not in part due to tracer uptake from passing vagal efferent fibers, in 1 animal 1 /zl of B - H R P was injected into the left cervical vagus nerve 1 cm distal to the nodose ganglion. After a survival time of 32 h, the animal was processed as above.
Results
Nodose ganglia injections No differences between labeling produced by the anterograde transport of W G A - H R P versus that of B - H R P were found. Although W G A HRP, compared to B-HRP, often formed smaller grains of reaction product, this did not affect the distribution or the frequency of labeled fibers in the heart. Both tracers showed a similar timecourse of labeling intensity. Initial experiments established that a survival time of 28-32 h led to the best overall labeling regardless of tracer. Brainstem sections, used as labeling controls before proceeding to process heart sections, showed extensive fiber labeling in the dorsal motor nucleus of the vagus and in the nucleus ambiguous [15,29]. Anterograde transport of tracer demonstrated a consistent distribution of labeled fibers within transverse sections of the heart (Fig. 1). Labeled anatomical regions include the ascending aorta and aortic arch, the pulmonary trunk and arteries, the left and right atrial walls and interatrial septum, the tricuspid and mitral valves, the posterior interventricular septum, and the posterior ventricular subepicardium in close association with coronary arteries. Additional fibers noted in the mediastinal cross sections were associated with the lower trachea and mainstem bronchi.
16 Labeling was absent from mounted sections of the parietal pericardium as well as from intracardiac ganglia. All regions of the heart were bilaterally innervated. Although no regions displayed lateralization, the atrium ipsilateral to the injected ganglion appeared to show denser labeling than the contralateral atrium. The greatest density of labeling was found within the walls of the ascending aorta and aortic arch, the pulmonary trunk and arteries, and the interjunctional space between the two vessels (Figs. 2A-D). The aortic wall, particularly in the region of the arch adjacent to the pulmonary arteries, was densely innervated. Fiber bundles associated with small arteries entered this interjunctional space (Figs. 2A, B). These heavily labeled bundles ramified along their course and spread upon the aortic adventitia and penetrated into the media. Smaller fiber bundles penetrated into the aortic media and formed complicated, intramural arborizations that often displayed beaded globules resembling terminals along their branches and ends (Figs. 2A-C). These arborizations were not diffusely distributed upon the aorta but were grouped together into discrete patches. Similar arborizations, somewhat rarer and less extensive, were seen within the pulmonary arteries (Fig. 2D) and other great vessels. At the light microscope level, it was difficult to determine whether these arborizations penetrated into the media or were surrounded by invaginations of adventitia [2,13]. The left and right atria were also richly labeled (Figs. 3A-H). Atrial labeling was concentrated at, though not limited to, the posterior atrial walls at their junctions with the pulmonary veins (Figs. 3A, B), and the venae cavae (Figs. 3C, D). In these junctional regions, fiber bundles ran along the posterior atrial epicardium where they often ramltlcd and penetrated into adjacent myocardium. Terminal-like structures were observed within the myocardium or within the subendocardial layer. Within the atrial walls two types of terminallike structures were observed. Complicated, beaded arborizations similar to those described above (Figs. 2A-D, 3E, G) were commonly found.
Less frequent were single, discrete globules distinguished from the surrounding artifact by a regular, spindle shape and a short section of fiber leading away from the structure (Figs. 3A, C). Beaded fibers and arborizations also were found within the interatrial septum (Figs. 3E, F), usually toward the posterior wall. Long, nonbranching fibers without beading, presumably passing fibers, ran within the connective tissue between the atria and the roots of the aorta and pulmonary artery (Fig. 1). Labeled fibers were not detected in the auricles. The ventricles, although consistently labeled among the experiments, were more sparsely innervated by vagal afferent fibers than the regions described above. (Figs. 4A-D, 5D) Labeled fibers within the ventricles were usually found close to branches of the coronary arteries. Several labeling patterns were observed in the interventricular septum (Figs. 4A-D). The first consisted of beaded fibers closely adherent to the posterior septal arteries (Fig. 4A). These fibers ran in the connective tissue between the vessel and the ventricular endocardium. A second group were arborizations, often within the myocardium (Fig. 4B) located in the posterior septum not far from the insertion of the aortic root. A third group of arborizations terminated in the subendocardial layer in the posterior septum (Fig. 4C). This last group of arborizations was found on both sides of the interventricular septum. Thin bundles of fibers continued along the subepicardium adjacent to posterior arteries of the ventricles. Occasionally they would branch and follow arterioles supplying the ventricular myocardium (Fig. 5D). Although labeled fibers were found associated with other branches of the coronary system--for example, the left anterior descending artery--labeled fibers, except those near posterior-inferior vessels, disappeared as the sections continued caudally (Fig. 1). In addition to the above regions, the tricuspid and mitral valves were found to contain HRPlabeled fibers. Labeling consisted of simply branched, beaded fibers that were scattered upon the valve surface (Figs. 5A-C). In the case of the mitral valve, these fibers appeared to extend toward the papillary muscles (Fig. 5C).
17
B V
Ao
Fig. 2. A, B: Photomicrograph and accompanying camera lucida drawing of a transverse section of ascending aorta after injection of B - H R P into left nodose ganglion. C: Arborization labeled by B - H R P within interjunctional space between aortic arch and pulmonary artery. D: Intramural arborization within pulmonary artery after W G A - H R P injection into left nodose ganglion, bv: blood vessel; *: lumen. Bars = 100 p,m.
t8
B
RA
D
,k~
SVC
' ,tl' /
'
I I
k
, <
E /I"
RA
H
LA Fig. 3. A, B: Close-up and overview of subendocardial, spindle-shaped termination and associated fibers of the posterior wall of the right atrium after injection of W G A - H R P into the right nodose ganglion. C, D: termination within myocardium between the superior vena cava (SVC) and the right atrium. B-HRP, left nodose ganglion injection. E, F: Subendocardial ending and nearby fiber bundles within the interatrial septum near the posterior wall. WGA-HRP, right nodose ganglion injection. G, H: myocardial and subendocardial arborizations within posterior wall of left atrium. B-HRP, left nodose ganglion injection. RA, LA: right and left atrium, SVC: superior vena cavae. A, C, E, G: bars = 50/zm. B, D, F, H: bars = 100/zm.
19
D
NV • J
//"
/~
CA
/// /
/
j
jt
J /
/
LV
Fig. 4. Photomicrographs (A-C) and overview camera lucida (l)) of a section of posterior interventricular septum after injection of W G A - H R P into the right nodose ganglion. A: Extensively branching fibers between a coronary artery and the endocardium of the right ventricle. B" Myocardial arborization. C: Fiber with prominent varicosities within the subepicardium of the left ventricle. RV, LV: right and left ventricle; ca: coronary artery. *: lumen. Bars = 100/xm.
20
C LA
--_
pap
Fig. 5. A: Fibers upon the tricuspid valve. WGA-HRP, right nodose ganglion injection. B-C: close-up and overview of fibers upon the mitral valve descending toward the papillary muscle. D: Weakly labeled fibers within the subepicardium of the posterior-inferior ventricle. B-HRP, left nodose ganglion injection. E: Fan-shaped array of fibers upon the IVC. B-HRP, right nodose ganglion injection. F: Fibers running between left and right mainstem bronchi. RA, RV: right atrium and ventricle; LA, LV: left atrium and ventricle; RB, LB: right and left bronchi. Bars = 100/zm.
21 Finally, labeled fibers were seen to fan along and descend the IVC (Fig. 5E) and were observed entwined about the lower trachea and mainstem bronchi (Fig. 5F). Because the respiratory system was not the focus of this study, these fibers were omitted from the mapped transverse sections of the heart (Fig. 1).
Cervical vagus injections Injection of B - H R P into the vagus nerve distal to the nodose ganglion resulted in sparse labeling of perikarya in the ambiguous and dorsal motor nuclei of the ipsilateral brainstem. Reaction product within the vagus nerve was restricted to several mm distal to the injection site. No labeled fibers or arborizations were observed in transverse sections of the mediastinum.
Discussion
An important question raised by the technique used in the current study was whether the observed labeling was purely a result of anterogradely labeled primary afferent nerve fibers or in some degree augmented by efferent vagal fibers. The presence of labeled neurons within the ambiguous and dorsal motor nuclei of the brainstem, via uptake en passant of tracer and its retrograde transport along efferent axons [15,29], shows that conjugated HRP enters passing axons. The possibility of anterograde transport by efferent fibers en passant can therefore not be excluded. Earlier tracing studies, however, have demonstrated the absence of anterograde transport by efferent fibers. In studies of visceral innervation, injections of W G A - H R P into the nodose ganglion after supranodose vagotomy and into the cervical vagus have shown the absence of significant anterograde efferent transport due to en passant uptake of tracer [21,22]. The lack of anterograde transport by efferent fibers has been shown in the central nervous system [3] as well. Likewise, in the current study, labeled fibers were absent from sections of the mediastinum after the injection of conjugated tracer into the cervical vagus. Our findings, therefore, in accord with previous tracing studies indicate that anterograde
transport of conjugated HRP is a specific method of tracing vagal afferent fibers originating from the nodose ganglia. Although this discussion focuses upon the afferent innervation of the heart, the impressive labeling of afferent fibers within the aorta and pulmonary trunk created by this technique deserves some comment. To our knowledge the current study has been the first to label receptors of the aorta and other great vessels by the anterograde transport of conjugated HRP. The endings labeled by conjugated HRP are identical to those visualized through methylene blue [13] and silver techniques [2]. These receptors are classically considered vagal [26], although sympathetic [16] and other afferent pathways [31] have been histologically or physiologically demonstrated. It is instructive to note that aortic fibers of vagal origin differ remarkably from those of sympathetic origin when labeled by the anterograde transport of conjugated HRP. In contrast to the thick fibers and complex arborizations of the former, the latter appeared as thin, beaded strands often coupled with the vasa vasorum within the walls of the aorta and pulmonary arteries [27]. Regarding the innervation of the heart itself, the distribution of selective labeling by conjugated HRP shows some important differences from previous histological studies. Earlier histological investigations using methylene blue or silver techniques in conjunction with nerve lesions [12,20] have resulted in conflicting and equivocal data. More recent histological investigations of sensory innervation have centered on a subset of cardiac afferents that contain neuropeptides such as substance P (SP) and calcitonin gene related peptide (CGRP) [4,7,9,24,25,28,33,34]. Attempts have been made to isolate vagal versus sympathetic afferent innervation through chemical or surgical nerve lesions [4,24]. To place the current study in the context of previous studies, the majority of HRP-labeled fibers, originating from the nodose ganglia and terminating in the heart, were found in the posterior atria in regions corresponding to atrio-venous junctions. Both compact, globular and diffuse arborizations, corresponding to earlier descriptions of compact and diffuse unencapsulated endings
22
[6,12,18,20] were represented. These endings were found to be supplied by both vagi, although in accord with earlier studies [12] labeling of the left or right atrium was favored by tracer injection into the ipsilateral nodose ganglion. The AV valves and coronary arteries, in concordance with earlier visualization techniques [6,12,18,20], were also innervated with HRP-labeled fibers. Furthermore, HRP-labeled fibers of vagal origin corresponded to the denser and wider distribution of fibers containing neuropeptides. Like SP-immunoreactive (IR) fibers, HRP-labeled fibers were more abundant in the atria than in the ventricles. Similarly, within the ventricles, where the number of SP-IR fibers were greatest at the base and diminished toward the apex [9], the number of HRP-labeled fibers diminished as the transverse sections continued caudally, toward apex and diaphragm. The HRP fibers within the posterior interventricular septum, a region containing the AV bundle, suggest that a network of HRP fibers, in part, could correspond to neuropeptide-containing fibers found within the distal conduction system [7,25,33]. Afferents of vagal origin and those of sympathetic origin, when similarly traced with conjugated HRP [27], failed to appear in or about intracardiac ganglia. The lack of HRP labeling, or its relative sparsity, underscores previous findings that neurons containing SP and other neuropeptides can be of extrinsic or intrinsic origin [4]. The relative sparsity of HRP fibers in comparison to those visualized by immunohistochemical methods could also indicate limitations of the tracing technique or could reflect that fibers perpendicular to the section plane are difficult to see in transverse sections [1] (most studies cited dealt with flat mount preparations). The selectivity of antea'ograde tracing facilitates the comparison of the vagal versus sympathetic contributions to the afferent innervation of the heart. Clinical observations of the autonomic symptoms after acute myocardial infarction suggest that the distributions of these two afferent systems are not identical [23]. Parasympathetic overactivity (sinus bradycardia or atrioventricular block) was related to posterior infarction, sympathetic overactivity (transient hypertension or si-
nus tachycardia) to anterior infarction. The distribution of afferents could also contribute to the intensity and presentation of angina pectoris, specifically the predominance of vagally-mediated symptoms--bronchospasm, nausea, and vomiting for example--that often accompany posterior myocardial infarction [11]. A greater density of vagal afferent innervation in the posterior/ inferior arteries could play a role in post-infarction cardiac responses. Physiological studies have clearly demonstrated site-specific vagal depressor and sympathetic pressor responses through pharmacological stimulation of the Bezold-Jarish reflex [5]. In this manner, Walker et al. [32] showed that intracoronary injections of nicotine or veratridine into the circumflex artery caused a greater depressor response than injections into the left anterior descending artery. Moreover, Inoue and Zipes [10] produced similar results with epicardial application of nicotine or bradykinin, reporting that drug application onto the epicardium of the midposterior left ventricle gave the greatest hypotensive response. In agreement with physiological and clinical observations are the results of the current study and those of an earlier investigation that examined the distribution of cardiac sympathetic afferent innervation [27]. Anterogradely transported W G A - H R P was used to selectively label afferent fibers with origins in spinal ganglia of C~,, Cs, and TI_3. The frequency of sympathetic afferent fibers decreased as the transverse sections continued caudally, so that labeled fibers were rarely found along ventricular arteries as they approached the diaphragmatic surface of the heart. Although vagal afferent fibers also demonstrate a caudal decrease in density, unlike sympathetic afferent fibers they remain present when attending posterior-inferior coronary arteries. Labeled fibers were seen with posterior vessels even at the most caudal transverse sections. This pattern, in summary, gives vagal afferent fibers predominance in the posterior-inferior regions of the arterial supply to the ventricles; sympathetic afferent fibers, the anterior regions. This inhomogeneous distribution of afferent systems correlates with clinical and physiological findings that suggest ischemia
23
may invoke different magnitudes of activation of sympathetic or vagal afferent nerves according to the site of tissue injury.
Acknowledgements This project and other related work would not have been possible without the advice, knowledge, and support of Drs. Lars-G6sta Elfvin and H~ikan Aldskogius of the Department of Anatomy of Karolinska Institutet. Thanks also to AnnaStina H6ijer and Britt Meijer for their technical skills and Nila Saliba of U.Va. Medical Publications for her points on style. Dr. Bj6rn Lindh provided excellent instruction in operative technique. This study was supported by grants from the Swedish Medical Research Council (project nos. 5189 and 5420) and from Karolinska Institutet.
References 1 Aldskogius, H., Elfvin, L.-G. and Forsman, C.A., Primary sensory afferents in the inferior mesenteric ganglion and related nerves of the guinea pig. An experimental study with anterogradely transported wheat germ agglutinin horseradish peroxidase conjugate, J. Auton. Nerv. Syst., 15 (1986) 179-190. 2 Aumonier, F.J., Histological observations on the distribution of baroceptors in the carotid and aortic regions of the rabbit, cat, and dog, Acta Ana., 82 (1972) 1-16. 3 Brodal, P., Dietrichs, E,, Bjaalie, J.G., Nordby, T. and Walberg, F., Is lectin-coupled horseradish peroxidase taken up and transported by undamaged as well as damaged fibers in the central nervous system?, Brain Res., 278 (1983) 1-9. 4 Dalsgaard, C.-J., Franco-Cereceda, A., Saria, A., Lundberg, J.M., Theodorsson-Norheim, E. and H6kfelt, T., Distribution and origin of substance P- and neuropeptide Y-immunoreactive nerves in the guinea pig heart, Cell Tiss. Res., 234 (1986) 477-485. 5 Ferguson, D.W. and Abboud, F.M., Pathophysiology of hypotension and shock. In J.W. Hurst et al. (Eds.), The Heart, McGraw-Hill, New York, 1990, pp. 442-459. 6 Floyd, K., Light microscopy of nerve endings in the atrial endocardium. In A. Hainsworth, C. Kidd, and R.J. Linden (Eds.), Cardiac Receptors, Cambridge University Press, London, 1979, 3-26. 7 Forsgren, S., The distribution of nerve fibers showing substance P-like immunoreactivity in the conduction sys-
tem of the bovine heart: dense innervation in the atrioventricular bundle, Anal. Embryol., 179 (1989) 485-490. 8 Franco-Cereceda, A. Calcitonin gene-related peptide and tachykinins in relation to local sensory control of cardiac contractility and coronary vascular tone, Acta Physiol. Scand., 133 (1988) 1-63. 9 Hougland, M.W. and Hoover, D.B., Detection of substance P-like immunoreactivity in nerve fibers in the heart of guinea pigs but not rats, J. Auton. Nerv. Syst., 8 (1983) 295-301. 10 Inoue, H. and Zipes, D.P., Increased afferent vagal responses produced by epicardial application of nicotine on the canine posterior left ventricle, Am. Heart J., 114 (1987) 757-764. 11 James, T.N., Rossi, L. and Hageman, G.R., On the pathogenesis of angina pectoris and it's silence, Trans. Am. Clin. Climat. Assoc., 100 (1988) 81-99. 12 Khabarova, A.Y., The afferent innervation of the heart, Consultants Bureau, New York, 1963, 1-475. 13 Krauhs, J.M., Structure of rat aortic baroreceptors and their relationship to connective tissue, J. Neurocytol., 8 (1979) 401-404. 14 Linden, R.J. and Kappagoda, C.T., Histology of sensory nerve endings in the heart. In R.J. Linden and C.T. Kappagoda (Eds.), Atrial Receptors, Cambridge University Press, London, 1982, pp. 6-30. 15 Lindh, B., Aldskogius, H. and H6kfelt, T., Simultaneous immunohistochemical demonstration of intra-axonally transported markers and neuropeptides in the peripheral nervous system of the guinea pig, Histochemistry, 92 (1989) 367-376. 16 Malliani, A., Cardiovascular Sympathetic Afferent Fibers. In R.H. Adrian et al. (Eds.), Reviews of Physiology, Biochemistry, and Pharmacology 94, Springer-Verlag, Berlin, 1982, pp. 11-74. 17 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem., 26 (1978) 106-117. 18 Miller, M.R., Kasahara, K., Studies on the nerve endings of the heart, Am. J. Anat., 115 (1964) 217-233. 19 Natelson, B.H., Neurocardiology, an interdisciplinary area for the 80's, Arch. Neurol., 42 (1986) 178-184. 20 Nettleship, W.A., Experimental studies on the afferent innervation of the cat's heart, J. Comp. Neurol., 64 (1963) 115-131. 21 Neuhuber, W.L., Sensory vagal innervation of the rat esophagus and cardia: a light and electron microscopic tracing study, J. Auton. Nerv. Syst., 20 (1987) 109-255. 22 Neuhuber, W.L., Vagal afferent fibers almost exclusively innervate islets in the rat pancreas as demonstrated by anterograde tracing, J, Auton. Nerv. Syst. 29 (1989) 13-18. 23 Pantridge, J.F., Autonomic disturbance at the onset of acute myocardia infarction. In P.J. Schwartz et al. (Eds.), Neural Mechanisms in cardiac arrhythmias. Raven Press, New York, 1978, pp. 7-17.
24 24 Papka, R.E., Urban, I., Distribution, origin, and sensitivity to capsaicin of primary afferent substance P-immunoreactive nerves in the heart, Acta Physiol. Hungarica, 69 (1987) 459-468. 25 Parsons, R.L. and Neel, D.S., Distribution of calcitonin gene-related peptide immunoreactive nerve fibers in the mudpuppy cardiac septum, J. Auton. Nerv. Syst. 21 (1987) 135-143. 26 Pick, J., The Autonomic Nervous System, Lippincot, Philadelphia, 1970. 27 Quigg, M., Elfvin, L.-G. and Aldskogius, H., Distribution of cardiac sympathetic afferent fibers in the guinea pig heart labeled by anterograde transport of wheat germ agglutinin-horseradish peroxidase, J. Auton. Nerv. Syst. 25 (1988) 107-118. 28 Reinecke, M., Weihe, E. and Forssman, F.G., Substance P-immunoreactive nerve fibers in the heart, Neurosci. Lett., 20 (1980) 265-269. 29 Reference omitted. 30 Schlant, R.C., Silverman, M.E. and Roberts, W.C., In J.W.
31
32
33
34
Hurst et al. (Eds.), The Heart, McGraw-Hill, New York, 1990, pp. 20-35. Thames, M.D., Johannsen, U.J. and Mark, A.L., In search of a cardiogenic hypertensive chemoreflex, Circulation, 75 (1987) 643-650. Walker, J.L., Abboud, F.M. and Mark, A.L., Preferential distribution of inhibitory cardiac receptors in the inferior wall of the left ventricle. In P.J. Schwartz et al. (Eds.), Neural mechanisms in cardiac arrhythmias. Raven Press, New York, 1978, 271-274. Wharton, J., Polak, J.M., McGregor, G.P., Bishop, A.E. and Bloom, S.R., The distribution of substance P-like immunoreactive nerves in the guinea pig heart, Neuroscience 6 (1981) 2193-2204. Wharton, J., Gulbenkian, S., Mulderry, P.K., Ghatei, M.A., McGregor, G.P., Bloom, S.R. and Polak, J.M., Capsaicin induces a depletion of calcitonin gene-related peptide (CGRP)-immunoreactive nerves in the cardiovascular system of the guinea pig and rat, J. Auton. Nerv. Syst., 16 (1986) 289-309.
13
© 1991 Elsevier Science Publishers B.V. All rights reserved 0165-1838/91/$03.50 JANS 01200
Distribution of vagal afferent fibers of the guinea pig heart labeled by anterograde transport of conjugated horseradish peroxidase Mark Quigg Department of Anatomy, Karolinska Institutet, Stockholm, Sweden, and Department of Neurology, University of Virginia, Charlottesville, Virginia, US.A. (Received 16 August 1990) (Revision received 15 April 1991) (Accepted 17 May 199l)
Key words: Heart; Vagal sensory fiber; Nodose ganglia; Horseradish peroxidase conjugate; Anterograde transport; Guinea pig Abstract To determine the distribution of vagal afferent fibers in the heart, wheat germ agglutinin conjugated horseradish peroxidase (WGA-HRP) or choleragenoid conjugated horseradish peroxidase (B-HRP) was injected into nodose ganglia of guinea pigs. Anterogradely labeled fibers and beaded, terminal-like arborizations were observed in the ascending aorta and aortic arch, the pulmonary trunk and arteries, posterior atrial walls, atrioventricular valves, and ventricles. Control experiments with injection of B - H R P into the cervical vagus nerve indicated that labeled fibers observed in the heart originated from sensory neurons in the nodose ganglia. Neither the density nor distribution of labeling differed between W G A - H R P and B-HRP. Injection of tracer into the left or right nodose ganglion shows that these regions of the heart are bilaterally innervated, although labeling in the left or right posterior atrium was denser after injection into the ipsilateral ganglion. Comparison with a previous study on the distribution of sympathetic afferent fibers in the guinea pig heart suggests that the two afferent systems maintain a complementary, but not mutually exclusive, distribution within the ventricles. Whereas afferents with their source in the spinal ganglia are mainly distributed with coronary arteries on the anterior-superior surface of the ventricles, afferent fibers with their source in the nodose ganglia are concentrated within peri-arterial regions of the posterior-inferior ventricular epicardium and the posterior septal ventricular endocardium. These differences in distributions of afferent systems could play a role in post-infarction autonomic dysfunction and in the symptoms that accompany angina pectoris.
Introduction
The heart receives its afferent innervation from both vagal and sympathetic nerves [26,30]. In the
Correspondence: M. Quigg, c / o H~kan Aldskogius, Department of Anatomy, Karolinska Institutet, Box 60400, S-10401 Stockholm, Sweden.
case of the vagal afferent system, sensory cell bodies located within the nodose ganglia send central fibers to the nucleus tractus solitarius and the dorsal motor vagal nucleus [19]. Their peripheral axons descend the vagus to enter the cardiac plexus at the base of the heart. The peripheral distribution of vagal afferent fibers and their sensory endings within the heart is still unclear because of the difficulty in distinguishing vagal from
14
sympathetic and afferent from efferent innervation. Clinical observations of autonomic symptoms after acute myocardial infarction suggest that the distribution of these two afferent systems are not homogeneous [23]. Anatomical confirmation, however, is difficult to find among earlier studies [6,12,14,16,18,20] of cardiac afferent innervation that have depended upon unspecific nerve labeling methods. Conversely, a subset of cardiac afferent fibers that contain neuropeptides [8] can be specifically labeled by immunohistochemical methods. Accordingly, fibers containing substance P [4,7,9,24,28,33] and calcitonin generelated peptide [25,34] have been identified in various regions of the heart and coronary arteries. These studies, however, do not exclude contributions from the dual innervation of the heart.
In the present study, anterogradely transported conjugates of horseradish peroxidase, wheat germ agglutinin conjugate (WGA-HRP) [1] and choleragenoid /3 subunit conjugate (BHRP) [15], are used to visualize selectively the distribution of vagal afferent fibers and endings within the heart. These results are compared to those of an earlier study in which cardiac sympathetic afferent fibers were studied With the same technique [27].
Materials and Methods
Male and female guinea pigs (200-400 g, n = 6) were anesthetized with xylazine (2.5 mg/kg i.m., Bayer) and ketamine (50 mg/kg i.m., Parke-
.A
I
Es
I
1 cm Fig. 1. Cardiac innervation map of vagal afferent fibers with origins in the left nodose ganglion labeled by anterograde transport of B - H R P after 31 h post-injection survival time. Black squares mark locations of labeled fibers or fiber groups in this cranial to caudal (A-I) series of transverse sections of the guinea pig heart. T: trachea; Es: esophagus; Ao: aorta; PA: pulmonary artery; RA, RAur, RV: right atrium, auricle, ventricle; LA, LAur, LV: left atrium, auricle, ventricle; Br: bronchi.
15 Davis), and either the left or right nodose ganglion was surgically exposed. 1 /zl of 2% W G A HRP (Sigma) or 1% B - H R P (List Biol. Labs Campbell, CA) in distilled water was pressure-injected with a glass micropipette over a period of 5-10 min to eliminate leakage. Afterwards, the pipette was withdrawn slowly to discourage backflow of tracer. After survival times ranging from 12 to 52 h, the animals were re-anesthetized and cannulated in the abdominal aorta. The animals were retrogradely perfused with a prewash of body temperature 0.9% saline and then fixed with 2 1 of 2% paraformaldehyde and 1.5% glutaraldehyde in phosphate buffer (ice-cold, pH 7.4). Nodose ganglia, vagus nerve, mediastinum, and brainstem were excised, postfixed by 15-30 min immersion in fixative, and then rinsed and stored overnight in 30% sucrose/phosphate buffer solution (ice-cold, pH 7.4). Brainstems and vagal nerves were sectioned transversely and longitudinally respectively at 50 /xm on a freezing microtome and incubated with tetramethyl benzidine (TMB, Sigma) [17] to check for successful nodose injections. If the distal vagal nerves showed anterogradely-transported tracer, and if brainstem perikarya, labeled by uptake and retrograde transport of tracer by efferent axons passing through the nodose ganglion [15,29], were found in the ambiguous and dorsal motor nuclei in the brainstem, the corresponding mediastinum was then processed. Before sectioning, pieces of the inferior vena cavae (IVC) and parietal pericardia were stripped from the mediastina and saved for incubation with mediastinal sections. Mediastina were sectioned transversely at 90 /xm with a freezing microtome; prior to freezing, the heart chambers were filled with a 2:1 mixture of 30% sucrose/ phosphate buffer and Tissue-Tek (Miles Laboratories, Naperville, IL). This reinforcement and minimum handling kept the sections intact and untangled during TMB incubation and mounting. The mediastinal sections were traced at a magnification of 5.5 x with the aid of a light projector; labeled fibers and sensory endings identified by the presence of a fine, dark blue HRP reaction product were scanned under a light microscope at a total magnification of 250 x and mapped onto
the traced sections as shown in Fig. 1. From each specimen every other section of all animals was mapped. Mapped sightings were counted by anatomical region to determine a regional frequency distribution, thereby allowing comparisons in labeling between different parts of the heart after left versus right nodose ganglia injection. To confirm that the labeling observed within the heart was not in part due to tracer uptake from passing vagal efferent fibers, in 1 animal 1 /zl of B - H R P was injected into the left cervical vagus nerve 1 cm distal to the nodose ganglion. After a survival time of 32 h, the animal was processed as above.
Results
Nodose ganglia injections No differences between labeling produced by the anterograde transport of W G A - H R P versus that of B - H R P were found. Although W G A HRP, compared to B-HRP, often formed smaller grains of reaction product, this did not affect the distribution or the frequency of labeled fibers in the heart. Both tracers showed a similar timecourse of labeling intensity. Initial experiments established that a survival time of 28-32 h led to the best overall labeling regardless of tracer. Brainstem sections, used as labeling controls before proceeding to process heart sections, showed extensive fiber labeling in the dorsal motor nucleus of the vagus and in the nucleus ambiguous [15,29]. Anterograde transport of tracer demonstrated a consistent distribution of labeled fibers within transverse sections of the heart (Fig. 1). Labeled anatomical regions include the ascending aorta and aortic arch, the pulmonary trunk and arteries, the left and right atrial walls and interatrial septum, the tricuspid and mitral valves, the posterior interventricular septum, and the posterior ventricular subepicardium in close association with coronary arteries. Additional fibers noted in the mediastinal cross sections were associated with the lower trachea and mainstem bronchi.
16 Labeling was absent from mounted sections of the parietal pericardium as well as from intracardiac ganglia. All regions of the heart were bilaterally innervated. Although no regions displayed lateralization, the atrium ipsilateral to the injected ganglion appeared to show denser labeling than the contralateral atrium. The greatest density of labeling was found within the walls of the ascending aorta and aortic arch, the pulmonary trunk and arteries, and the interjunctional space between the two vessels (Figs. 2A-D). The aortic wall, particularly in the region of the arch adjacent to the pulmonary arteries, was densely innervated. Fiber bundles associated with small arteries entered this interjunctional space (Figs. 2A, B). These heavily labeled bundles ramified along their course and spread upon the aortic adventitia and penetrated into the media. Smaller fiber bundles penetrated into the aortic media and formed complicated, intramural arborizations that often displayed beaded globules resembling terminals along their branches and ends (Figs. 2A-C). These arborizations were not diffusely distributed upon the aorta but were grouped together into discrete patches. Similar arborizations, somewhat rarer and less extensive, were seen within the pulmonary arteries (Fig. 2D) and other great vessels. At the light microscope level, it was difficult to determine whether these arborizations penetrated into the media or were surrounded by invaginations of adventitia [2,13]. The left and right atria were also richly labeled (Figs. 3A-H). Atrial labeling was concentrated at, though not limited to, the posterior atrial walls at their junctions with the pulmonary veins (Figs. 3A, B), and the venae cavae (Figs. 3C, D). In these junctional regions, fiber bundles ran along the posterior atrial epicardium where they often ramltlcd and penetrated into adjacent myocardium. Terminal-like structures were observed within the myocardium or within the subendocardial layer. Within the atrial walls two types of terminallike structures were observed. Complicated, beaded arborizations similar to those described above (Figs. 2A-D, 3E, G) were commonly found.
Less frequent were single, discrete globules distinguished from the surrounding artifact by a regular, spindle shape and a short section of fiber leading away from the structure (Figs. 3A, C). Beaded fibers and arborizations also were found within the interatrial septum (Figs. 3E, F), usually toward the posterior wall. Long, nonbranching fibers without beading, presumably passing fibers, ran within the connective tissue between the atria and the roots of the aorta and pulmonary artery (Fig. 1). Labeled fibers were not detected in the auricles. The ventricles, although consistently labeled among the experiments, were more sparsely innervated by vagal afferent fibers than the regions described above. (Figs. 4A-D, 5D) Labeled fibers within the ventricles were usually found close to branches of the coronary arteries. Several labeling patterns were observed in the interventricular septum (Figs. 4A-D). The first consisted of beaded fibers closely adherent to the posterior septal arteries (Fig. 4A). These fibers ran in the connective tissue between the vessel and the ventricular endocardium. A second group were arborizations, often within the myocardium (Fig. 4B) located in the posterior septum not far from the insertion of the aortic root. A third group of arborizations terminated in the subendocardial layer in the posterior septum (Fig. 4C). This last group of arborizations was found on both sides of the interventricular septum. Thin bundles of fibers continued along the subepicardium adjacent to posterior arteries of the ventricles. Occasionally they would branch and follow arterioles supplying the ventricular myocardium (Fig. 5D). Although labeled fibers were found associated with other branches of the coronary system--for example, the left anterior descending artery--labeled fibers, except those near posterior-inferior vessels, disappeared as the sections continued caudally (Fig. 1). In addition to the above regions, the tricuspid and mitral valves were found to contain HRPlabeled fibers. Labeling consisted of simply branched, beaded fibers that were scattered upon the valve surface (Figs. 5A-C). In the case of the mitral valve, these fibers appeared to extend toward the papillary muscles (Fig. 5C).
17
B V
Ao
Fig. 2. A, B: Photomicrograph and accompanying camera lucida drawing of a transverse section of ascending aorta after injection of B - H R P into left nodose ganglion. C: Arborization labeled by B - H R P within interjunctional space between aortic arch and pulmonary artery. D: Intramural arborization within pulmonary artery after W G A - H R P injection into left nodose ganglion, bv: blood vessel; *: lumen. Bars = 100 p,m.
t8
B
RA
D
,k~
SVC
' ,tl' /
'
I I
k
, <
E /I"
RA
H
LA Fig. 3. A, B: Close-up and overview of subendocardial, spindle-shaped termination and associated fibers of the posterior wall of the right atrium after injection of W G A - H R P into the right nodose ganglion. C, D: termination within myocardium between the superior vena cava (SVC) and the right atrium. B-HRP, left nodose ganglion injection. E, F: Subendocardial ending and nearby fiber bundles within the interatrial septum near the posterior wall. WGA-HRP, right nodose ganglion injection. G, H: myocardial and subendocardial arborizations within posterior wall of left atrium. B-HRP, left nodose ganglion injection. RA, LA: right and left atrium, SVC: superior vena cavae. A, C, E, G: bars = 50/zm. B, D, F, H: bars = 100/zm.
19
D
NV • J
//"
/~
CA
/// /
/
j
jt
J /
/
LV
Fig. 4. Photomicrographs (A-C) and overview camera lucida (l)) of a section of posterior interventricular septum after injection of W G A - H R P into the right nodose ganglion. A: Extensively branching fibers between a coronary artery and the endocardium of the right ventricle. B" Myocardial arborization. C: Fiber with prominent varicosities within the subepicardium of the left ventricle. RV, LV: right and left ventricle; ca: coronary artery. *: lumen. Bars = 100/xm.
20
C LA
--_
pap
Fig. 5. A: Fibers upon the tricuspid valve. WGA-HRP, right nodose ganglion injection. B-C: close-up and overview of fibers upon the mitral valve descending toward the papillary muscle. D: Weakly labeled fibers within the subepicardium of the posterior-inferior ventricle. B-HRP, left nodose ganglion injection. E: Fan-shaped array of fibers upon the IVC. B-HRP, right nodose ganglion injection. F: Fibers running between left and right mainstem bronchi. RA, RV: right atrium and ventricle; LA, LV: left atrium and ventricle; RB, LB: right and left bronchi. Bars = 100/zm.
21 Finally, labeled fibers were seen to fan along and descend the IVC (Fig. 5E) and were observed entwined about the lower trachea and mainstem bronchi (Fig. 5F). Because the respiratory system was not the focus of this study, these fibers were omitted from the mapped transverse sections of the heart (Fig. 1).
Cervical vagus injections Injection of B - H R P into the vagus nerve distal to the nodose ganglion resulted in sparse labeling of perikarya in the ambiguous and dorsal motor nuclei of the ipsilateral brainstem. Reaction product within the vagus nerve was restricted to several mm distal to the injection site. No labeled fibers or arborizations were observed in transverse sections of the mediastinum.
Discussion
An important question raised by the technique used in the current study was whether the observed labeling was purely a result of anterogradely labeled primary afferent nerve fibers or in some degree augmented by efferent vagal fibers. The presence of labeled neurons within the ambiguous and dorsal motor nuclei of the brainstem, via uptake en passant of tracer and its retrograde transport along efferent axons [15,29], shows that conjugated HRP enters passing axons. The possibility of anterograde transport by efferent fibers en passant can therefore not be excluded. Earlier tracing studies, however, have demonstrated the absence of anterograde transport by efferent fibers. In studies of visceral innervation, injections of W G A - H R P into the nodose ganglion after supranodose vagotomy and into the cervical vagus have shown the absence of significant anterograde efferent transport due to en passant uptake of tracer [21,22]. The lack of anterograde transport by efferent fibers has been shown in the central nervous system [3] as well. Likewise, in the current study, labeled fibers were absent from sections of the mediastinum after the injection of conjugated tracer into the cervical vagus. Our findings, therefore, in accord with previous tracing studies indicate that anterograde
transport of conjugated HRP is a specific method of tracing vagal afferent fibers originating from the nodose ganglia. Although this discussion focuses upon the afferent innervation of the heart, the impressive labeling of afferent fibers within the aorta and pulmonary trunk created by this technique deserves some comment. To our knowledge the current study has been the first to label receptors of the aorta and other great vessels by the anterograde transport of conjugated HRP. The endings labeled by conjugated HRP are identical to those visualized through methylene blue [13] and silver techniques [2]. These receptors are classically considered vagal [26], although sympathetic [16] and other afferent pathways [31] have been histologically or physiologically demonstrated. It is instructive to note that aortic fibers of vagal origin differ remarkably from those of sympathetic origin when labeled by the anterograde transport of conjugated HRP. In contrast to the thick fibers and complex arborizations of the former, the latter appeared as thin, beaded strands often coupled with the vasa vasorum within the walls of the aorta and pulmonary arteries [27]. Regarding the innervation of the heart itself, the distribution of selective labeling by conjugated HRP shows some important differences from previous histological studies. Earlier histological investigations using methylene blue or silver techniques in conjunction with nerve lesions [12,20] have resulted in conflicting and equivocal data. More recent histological investigations of sensory innervation have centered on a subset of cardiac afferents that contain neuropeptides such as substance P (SP) and calcitonin gene related peptide (CGRP) [4,7,9,24,25,28,33,34]. Attempts have been made to isolate vagal versus sympathetic afferent innervation through chemical or surgical nerve lesions [4,24]. To place the current study in the context of previous studies, the majority of HRP-labeled fibers, originating from the nodose ganglia and terminating in the heart, were found in the posterior atria in regions corresponding to atrio-venous junctions. Both compact, globular and diffuse arborizations, corresponding to earlier descriptions of compact and diffuse unencapsulated endings
22
[6,12,18,20] were represented. These endings were found to be supplied by both vagi, although in accord with earlier studies [12] labeling of the left or right atrium was favored by tracer injection into the ipsilateral nodose ganglion. The AV valves and coronary arteries, in concordance with earlier visualization techniques [6,12,18,20], were also innervated with HRP-labeled fibers. Furthermore, HRP-labeled fibers of vagal origin corresponded to the denser and wider distribution of fibers containing neuropeptides. Like SP-immunoreactive (IR) fibers, HRP-labeled fibers were more abundant in the atria than in the ventricles. Similarly, within the ventricles, where the number of SP-IR fibers were greatest at the base and diminished toward the apex [9], the number of HRP-labeled fibers diminished as the transverse sections continued caudally, toward apex and diaphragm. The HRP fibers within the posterior interventricular septum, a region containing the AV bundle, suggest that a network of HRP fibers, in part, could correspond to neuropeptide-containing fibers found within the distal conduction system [7,25,33]. Afferents of vagal origin and those of sympathetic origin, when similarly traced with conjugated HRP [27], failed to appear in or about intracardiac ganglia. The lack of HRP labeling, or its relative sparsity, underscores previous findings that neurons containing SP and other neuropeptides can be of extrinsic or intrinsic origin [4]. The relative sparsity of HRP fibers in comparison to those visualized by immunohistochemical methods could also indicate limitations of the tracing technique or could reflect that fibers perpendicular to the section plane are difficult to see in transverse sections [1] (most studies cited dealt with flat mount preparations). The selectivity of antea'ograde tracing facilitates the comparison of the vagal versus sympathetic contributions to the afferent innervation of the heart. Clinical observations of the autonomic symptoms after acute myocardial infarction suggest that the distributions of these two afferent systems are not identical [23]. Parasympathetic overactivity (sinus bradycardia or atrioventricular block) was related to posterior infarction, sympathetic overactivity (transient hypertension or si-
nus tachycardia) to anterior infarction. The distribution of afferents could also contribute to the intensity and presentation of angina pectoris, specifically the predominance of vagally-mediated symptoms--bronchospasm, nausea, and vomiting for example--that often accompany posterior myocardial infarction [11]. A greater density of vagal afferent innervation in the posterior/ inferior arteries could play a role in post-infarction cardiac responses. Physiological studies have clearly demonstrated site-specific vagal depressor and sympathetic pressor responses through pharmacological stimulation of the Bezold-Jarish reflex [5]. In this manner, Walker et al. [32] showed that intracoronary injections of nicotine or veratridine into the circumflex artery caused a greater depressor response than injections into the left anterior descending artery. Moreover, Inoue and Zipes [10] produced similar results with epicardial application of nicotine or bradykinin, reporting that drug application onto the epicardium of the midposterior left ventricle gave the greatest hypotensive response. In agreement with physiological and clinical observations are the results of the current study and those of an earlier investigation that examined the distribution of cardiac sympathetic afferent innervation [27]. Anterogradely transported W G A - H R P was used to selectively label afferent fibers with origins in spinal ganglia of C~,, Cs, and TI_3. The frequency of sympathetic afferent fibers decreased as the transverse sections continued caudally, so that labeled fibers were rarely found along ventricular arteries as they approached the diaphragmatic surface of the heart. Although vagal afferent fibers also demonstrate a caudal decrease in density, unlike sympathetic afferent fibers they remain present when attending posterior-inferior coronary arteries. Labeled fibers were seen with posterior vessels even at the most caudal transverse sections. This pattern, in summary, gives vagal afferent fibers predominance in the posterior-inferior regions of the arterial supply to the ventricles; sympathetic afferent fibers, the anterior regions. This inhomogeneous distribution of afferent systems correlates with clinical and physiological findings that suggest ischemia
23
may invoke different magnitudes of activation of sympathetic or vagal afferent nerves according to the site of tissue injury.
Acknowledgements This project and other related work would not have been possible without the advice, knowledge, and support of Drs. Lars-G6sta Elfvin and H~ikan Aldskogius of the Department of Anatomy of Karolinska Institutet. Thanks also to AnnaStina H6ijer and Britt Meijer for their technical skills and Nila Saliba of U.Va. Medical Publications for her points on style. Dr. Bj6rn Lindh provided excellent instruction in operative technique. This study was supported by grants from the Swedish Medical Research Council (project nos. 5189 and 5420) and from Karolinska Institutet.
References 1 Aldskogius, H., Elfvin, L.-G. and Forsman, C.A., Primary sensory afferents in the inferior mesenteric ganglion and related nerves of the guinea pig. An experimental study with anterogradely transported wheat germ agglutinin horseradish peroxidase conjugate, J. Auton. Nerv. Syst., 15 (1986) 179-190. 2 Aumonier, F.J., Histological observations on the distribution of baroceptors in the carotid and aortic regions of the rabbit, cat, and dog, Acta Ana., 82 (1972) 1-16. 3 Brodal, P., Dietrichs, E,, Bjaalie, J.G., Nordby, T. and Walberg, F., Is lectin-coupled horseradish peroxidase taken up and transported by undamaged as well as damaged fibers in the central nervous system?, Brain Res., 278 (1983) 1-9. 4 Dalsgaard, C.-J., Franco-Cereceda, A., Saria, A., Lundberg, J.M., Theodorsson-Norheim, E. and H6kfelt, T., Distribution and origin of substance P- and neuropeptide Y-immunoreactive nerves in the guinea pig heart, Cell Tiss. Res., 234 (1986) 477-485. 5 Ferguson, D.W. and Abboud, F.M., Pathophysiology of hypotension and shock. In J.W. Hurst et al. (Eds.), The Heart, McGraw-Hill, New York, 1990, pp. 442-459. 6 Floyd, K., Light microscopy of nerve endings in the atrial endocardium. In A. Hainsworth, C. Kidd, and R.J. Linden (Eds.), Cardiac Receptors, Cambridge University Press, London, 1979, 3-26. 7 Forsgren, S., The distribution of nerve fibers showing substance P-like immunoreactivity in the conduction sys-
tem of the bovine heart: dense innervation in the atrioventricular bundle, Anal. Embryol., 179 (1989) 485-490. 8 Franco-Cereceda, A. Calcitonin gene-related peptide and tachykinins in relation to local sensory control of cardiac contractility and coronary vascular tone, Acta Physiol. Scand., 133 (1988) 1-63. 9 Hougland, M.W. and Hoover, D.B., Detection of substance P-like immunoreactivity in nerve fibers in the heart of guinea pigs but not rats, J. Auton. Nerv. Syst., 8 (1983) 295-301. 10 Inoue, H. and Zipes, D.P., Increased afferent vagal responses produced by epicardial application of nicotine on the canine posterior left ventricle, Am. Heart J., 114 (1987) 757-764. 11 James, T.N., Rossi, L. and Hageman, G.R., On the pathogenesis of angina pectoris and it's silence, Trans. Am. Clin. Climat. Assoc., 100 (1988) 81-99. 12 Khabarova, A.Y., The afferent innervation of the heart, Consultants Bureau, New York, 1963, 1-475. 13 Krauhs, J.M., Structure of rat aortic baroreceptors and their relationship to connective tissue, J. Neurocytol., 8 (1979) 401-404. 14 Linden, R.J. and Kappagoda, C.T., Histology of sensory nerve endings in the heart. In R.J. Linden and C.T. Kappagoda (Eds.), Atrial Receptors, Cambridge University Press, London, 1982, pp. 6-30. 15 Lindh, B., Aldskogius, H. and H6kfelt, T., Simultaneous immunohistochemical demonstration of intra-axonally transported markers and neuropeptides in the peripheral nervous system of the guinea pig, Histochemistry, 92 (1989) 367-376. 16 Malliani, A., Cardiovascular Sympathetic Afferent Fibers. In R.H. Adrian et al. (Eds.), Reviews of Physiology, Biochemistry, and Pharmacology 94, Springer-Verlag, Berlin, 1982, pp. 11-74. 17 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem., 26 (1978) 106-117. 18 Miller, M.R., Kasahara, K., Studies on the nerve endings of the heart, Am. J. Anat., 115 (1964) 217-233. 19 Natelson, B.H., Neurocardiology, an interdisciplinary area for the 80's, Arch. Neurol., 42 (1986) 178-184. 20 Nettleship, W.A., Experimental studies on the afferent innervation of the cat's heart, J. Comp. Neurol., 64 (1963) 115-131. 21 Neuhuber, W.L., Sensory vagal innervation of the rat esophagus and cardia: a light and electron microscopic tracing study, J. Auton. Nerv. Syst., 20 (1987) 109-255. 22 Neuhuber, W.L., Vagal afferent fibers almost exclusively innervate islets in the rat pancreas as demonstrated by anterograde tracing, J, Auton. Nerv. Syst. 29 (1989) 13-18. 23 Pantridge, J.F., Autonomic disturbance at the onset of acute myocardia infarction. In P.J. Schwartz et al. (Eds.), Neural Mechanisms in cardiac arrhythmias. Raven Press, New York, 1978, pp. 7-17.
24 24 Papka, R.E., Urban, I., Distribution, origin, and sensitivity to capsaicin of primary afferent substance P-immunoreactive nerves in the heart, Acta Physiol. Hungarica, 69 (1987) 459-468. 25 Parsons, R.L. and Neel, D.S., Distribution of calcitonin gene-related peptide immunoreactive nerve fibers in the mudpuppy cardiac septum, J. Auton. Nerv. Syst. 21 (1987) 135-143. 26 Pick, J., The Autonomic Nervous System, Lippincot, Philadelphia, 1970. 27 Quigg, M., Elfvin, L.-G. and Aldskogius, H., Distribution of cardiac sympathetic afferent fibers in the guinea pig heart labeled by anterograde transport of wheat germ agglutinin-horseradish peroxidase, J. Auton. Nerv. Syst. 25 (1988) 107-118. 28 Reinecke, M., Weihe, E. and Forssman, F.G., Substance P-immunoreactive nerve fibers in the heart, Neurosci. Lett., 20 (1980) 265-269. 29 Reference omitted. 30 Schlant, R.C., Silverman, M.E. and Roberts, W.C., In J.W.
31
32
33
34
Hurst et al. (Eds.), The Heart, McGraw-Hill, New York, 1990, pp. 20-35. Thames, M.D., Johannsen, U.J. and Mark, A.L., In search of a cardiogenic hypertensive chemoreflex, Circulation, 75 (1987) 643-650. Walker, J.L., Abboud, F.M. and Mark, A.L., Preferential distribution of inhibitory cardiac receptors in the inferior wall of the left ventricle. In P.J. Schwartz et al. (Eds.), Neural mechanisms in cardiac arrhythmias. Raven Press, New York, 1978, 271-274. Wharton, J., Polak, J.M., McGregor, G.P., Bishop, A.E. and Bloom, S.R., The distribution of substance P-like immunoreactive nerves in the guinea pig heart, Neuroscience 6 (1981) 2193-2204. Wharton, J., Gulbenkian, S., Mulderry, P.K., Ghatei, M.A., McGregor, G.P., Bloom, S.R. and Polak, J.M., Capsaicin induces a depletion of calcitonin gene-related peptide (CGRP)-immunoreactive nerves in the cardiovascular system of the guinea pig and rat, J. Auton. Nerv. Syst., 16 (1986) 289-309.