MORPHOLOGICAL AND IMMUNOHISTOCHEMICAL PATTERNS OF THE INTRINSIC GANGLIONATED NERVE PLEXUS IN THE MOUSE HEART

(1)

LITHUANIAN UNIVERSITY OF HEALTH SCIENCES MEDICAL ACADEMY

Kristina Rysevaitė

MORPHOLOGICAL AND

IMMUNOHISTOCHEMICAL PATTERNS

OF THE INTRINSIC GANGLIONATED

NERVE PLEXUS IN THE MOUSE HEART

Doctoral Dissertation Biomedical Sciences,

Biology (01 B)

(2)

The dissertation has been prepared during the period of 2007–

2011 at the Institute of Anatomy, Medical Academy, Lithuanian

University of Health Sciences

Scientific Supervisor:

Prof. Dr. Neringa Paužienė (Medical Academy, Lithuanian

University of Health Sciences, Biomedical Sciences, Biology –

01 B)

(3)

LIST OF FREQUENTLY USED ABBREVIATIONS

ACh – acetylcholine

AChE – acetylcholinesterase AV – atrioventricular AVN – atrioventricular node

cChAT – conventional form of ChAT CGRP – calcitonin gene related peptide ChAT – choline acetyltransferase CHT – high-affinity choline transporter DBH – dopamine beta hydrohylase DRA – dorsal right arial subplexus

DVM – dorsal motor nucleus of the vagus ENP – epicardiac neural plexus

GNPHH – ganglionated nerve plexus of the heart hilum

HCN4 – hyperpolarization activated cyclic nucleotide-gated potassium channel 4

HH – hilum of the heart ICG – intrinsic cardiac ganglia

ICNS – intrinsic cardiac nervous system ICNs – intrinsic cardiac neurons

INP – intrinsic neural plexus IR – immunoreactive

IVC – inferior vena cava LA – left atria

LAu – left auricle

LC – left coronary subplexus LD – left dorsal subplexus

LOM – ligament of Marshall or neural fold of the left atrium LPV – left pulmonary vein

MD – middle dorsal subplexus MPV – middle pulmonary vein NA – nucleus ambiguus

NOS – nitric oxide synthase NE – norepinephrine

NPY – neuropeptide Y

pChAT – choline acetyltransferase of a peripheral type PGP 9.5 – protein gene product 9.5

PVs – pulmonary veins RA – right atria

(6)

RAu – right auricle

RPV – right pulmonary vein RV – right ventral subplexus SA – sinoatrial

SAN – sinoatrial node

SIF – small intensively fluorescent cell SP – substance P

TH – tyrosine hydroxylase

VAChT – vesicular acetylcholine transporter VC – vena cava

VLA – ventral left atrial subplexus VRA – ventral right arial subplexus

(7)

INTRODUCTION

Actuality of the Study

The intrinsic cardiac nervous system plays a crucial role in the regu-lation of heart rate, atrioventricular nodal conduction, and inotropism of atria and ventricles (Baumgart and Heusch, 1995; Cifelli et al., 2008; Feigl, 1998; Gorman et al., 2000; Randall et al., 2003; Tsuboi et al., 2000). Intrinsic ganglionated cardiac plexus integrates input from multiple sources including vagal efferent and afferent neurons, extrinsic sympathetic and spinal sensory neurons. The balance between the stimulatory sympathetic and inhibitory parasympathetic inputs are important for control of cardiac function (Ardell et al., 1991; Pauza et al., 1997b; Smith, 1999). It is widely recognized that autonomic nervous system modulates the cardiac electro-physiology of the heart and influences the genesis of cardiac arrhythmias or sudden cardiac death (Racker and Kadish, 2000).

The recent elucidation of the complete mouse genome in combination with transgenesis and gene targeting in embryonic stem cells have opened excellent opportunities to breed numerous genetically modified mouse lines for experimental modeling and investigation of molecular mechanisms of the role of sympathetic-parasympathetic imbalance in cardiac arrhythmia predisposition (Kanazawa et al., 2010; Koentgen et al., 2010). The mouse, as an animal model, is widely used in cardiovascular researches. It have been developed transgenic mouse models, in which specific genes involved in cardiac development, are modified (Feintuch et al., 2007). It is reported that neuronal populations within the intrinsic cardiac neurons of adult mice and human are competently comparable as they exhibit a similar neuroche-mical phenotype manifested predominantly by choline acetyl-transferase and tyrosine hydroxylase (Mabe et al., 2006). Morphologically, the intrinsic cardiac nervous system corresponds to the neural ganglionated plexus, which is frequently subdivided according to layers of heart wall into epi-cardial, myocardial and endocardial (Marron et al., 1995). Cardiac neuro-anatomical investigations have demonstrated that intrinsic cardiac neural plexus may be considered as a complex of distinct ganglionated subplexuses (Pauza et al., 2000). Intrinsic ganglia related to particular subplexuses are distributed at specific atrial or ventricular regions around the sinuatrial node, the roots of caval and pulmonary veins, and near the atrioventricular node (Arora et al., 2003; Batulevicius et al., 2008; Pauza et al., 2002b; Pauza et al., 2000). In contrast to human and very few species of laboratory animal hearts, neuroanatomy of the mouse heart was rather poorly examined for a

(8)

long while. In spite of very few investigations to this date (Ai et al., 2007; Hoard et al., 2008; Hoard et al., 2007; Mabe et al., 2006; Maifrino et al., 2006), both the distribution of sympathetic, vagal and sensory nerve fibers within the mouse intrinsic cardiac nerve plexus was almost unknown.

Aim and Objectives

The aim of the present study was to determine the structural organiza-tion of the intrinsic neural plexus in the total, non–secorganiza-tioned mouse heart, identifying the immunohistochemistry of nerve fibers and neurons located within this intrinsic neural plexus.

The objectives of the study:

1. To seek out the neural sources and neural pathways, by which mediastinal nerves supply the mouse heart.

2. To ascertain the structural organization of the mouse cardiac neural plexus.

3. To assess the ganglion size and the number of ganglionic cells in the mouse hearts in order to compare this animal model with others.

4. To identify the distribution of cholinergic, adrenergic and peptidergic neural structures in the whole-mount mouse heart preparations using double immunohistochemical labeling.

Originality and Implications

This is the first detailed anatomical investigation of intrinsic gangliona-ted nerve plexus in the total (i.e., non-parceled and non-sectioned) mouse heart. This study demonstrates for the first the distribution of cholinergic, adrenergic and peptidergic nerve fibers and neurons in the whole-mount preparation of the mouse heart. The technique of the whole-mount prepa-ration allows precise identifying and mapping of the all intrinsic cardiac ganglia. We also identified their immunohistochemical properties and inter-connections of intrinsic cardiac neurons within the atria, interatrial and interventricular septa. This study mapped in detail the distribution of the mouse intrinsic cardiac nerves and ganglia and this may help in attempts to stimulate and/or to ablate selectively the functionally distinct intrinsic neural pathways for investigations of arrhythmic heart. The neuroanatomy of the mouse heart demonstrated in this study facilitates further investigations with this animal model and, thereby, it should increase our knowledge of physiologic roles of distinct intrinsic nerves and individual intrinsic ganglia.

(9)

1. REVIEW OF LITERATURE 1.1. Innervation of the Heart

Parasympathetic innervation of the heart is carried by the Xth cranial (vagus) nerve, which originates in the medulla oblongata. In the human, it is the superior, inferior, and thoracic branches of the vagus nerve that innervate the heart (Kawashima, 2005). The nerves supplying the heart join in the cardiac plexus, an accumulation of mixed neurons located cranial and dorsal to the heart. In the human heart, left and right-sided cardiac plexuses surround the brachiocephalic trunk and the aortic arch, respectively, and form part of a larger cardiac plexus that lies between the aorta and pulmo-nary trunk (Kawashima, 2005; Pauza et al., 1997a; Pauza et al., 2000). Nuclei in the ventral lateral medulla project via the vagus nerve to postganglionic neurons in the cardiac nerve plexus. Physiological, viral-tracing, and degeneration studies showed that neurons of the dorsal motor nucleus of the vagus (DVM) and of the nucleus ambiguus (NA) innervate the heart (Standish et al., 1994). Furthermore, it has been shown that the ganglion cluster located at the left atrium adjacent to the inferior vena cava receives input from both the DVM and NA, (Massari et al., 1995; Standish et al., 1995) whereas, the ganglion cluster located at the right pulmonary vein-left atrial junction receives fibers from the NA only (Massari et al., 1994).

A submacroscopic anatomical investigations of the human extrinsic cardiac nervous system showed that (1) the superior, and middle cervical, and the cervicothoracic (stellate) ganglia, composed of the inferior cervical and 1st thoracic ganglia, were mostly consistent sources of symphatetic imputs to the heart; (2) the superior, middle, and inferior cardiac nerves innervated the heart by simple following the descent great arteries; (3) the thoracic cardiac nerve in the posterior mediastinum followed a complex course because of the long distance to the middle mediastinum; (4) the cranial cardiac nerve and branch tended to distribute into the heart medially, and the caudal cardiac nerve and branch tended to distribute into the heart laterally; (5) the mixing positions (cardiac plexus) of the sympathetic cardiac nerve and the vagal cardiac branch, as well as the definitive morphology of brachial arteries with the recurrent laryngeal nerves, tended to differ on both sides (Kawashima, 2005). The dorsal right atrial, right ventral and middle dorsal subplexuses of the sheep intrinsic cardiac nervous system (ICNS) receive the main extrinsic neural input from the right cervi-cothoracic and right thoracic sympathetic T2 and T3 ganglia as well as from the right vagal nerve. The left dorsal subplexus is supplied by sizeable

(10)

extrinsic nerves from the left thoracic T4–T6 sympathetic ganglia and the left vagal nerve (Saburkina et al., 2010). Convergence of inputs from extrinsic cardiac (vagal and cardiopulmonary (CPN)) nerves on intrinsic cardiac neurons was investigated in the pig (Smith, 1999).

There is a degree of asymmetry in the distribution of preganglionic symphatetic neurons in the cat medulla (Massari et al., 1995; Massari et al., 1994). Injection of a retrograde tracer into the atrioventricular (AV) ganglion of the cat results in the labeling of twice as many cells on the left side of the medulla as on the right side (Massari et al., 1995), whereas injection of a retrograde tracer into the sinoatrial (SA) ganglion showed asymmetrical distribution of labeled preganglionic neurons (Massari et al., 1994).

Preganglionic neurons form synapses on postganglionic neurons in autonomic ganglia, and there is a substantial degree of "convergence" in the parasympathetic nervous system. However, there is considerable diversity in the degree of convergence and divergence among species, also among different autonomic ganglia and individuals of a single species (Wang et al., 1995). The cardiac ganglion of the cat appears to be a degree of divergence rather than convergence: one preganglionic neuron projects on average to 13 or 32 postganglionic neurons (Wang et al., 1995). Similarly, studies by Massan and colleagues (Massari et al., 1995) indicate that a small number of neurons in the medulla are capable of controlling significant numbers of postganglionic cardiac neurons.

The ultrastructural development of the cardiac ganglia in the chicken can be divided into three phases: (1) migration and aggregation of neuroblasts on days 3.5-5; (2) differentiating ganglia, days 5-10; (3) maturing ganglia, days 11 to hatching. The development of cholinergic me-chanisms precedes that of adrenergic meme-chanisms. As a consequence the parasympathetic-cholinergic control becomes functional and plays a role in cardiac function earlier than the sympathetic-adrenal neural control (Baptista and Kirby, 1997).

In the mouse, nerve cells are first seen in the dorsal mesocardium at 10.5 days after fertilization. Well developed nerve tracts that can be identified using the neurofilament marker NF160D, extend through this region and reach the heart by 12.5 days after fertilization (Hildreth et al., 2008). Innervation of the outflow tract also occurs later in the mouse, with the first neural elements first seen between the separating aorta and pulmonary trunk. Staining against tyrosine hydroxylase (TH) and the parasympathetic neuron marker vasoactive cholinesterase transporter protein (VAChT) reveals the presence of sympathetic and parasympathetic neurons within the main nerves innervating the arterial and venous poles of the heart at 11.5

(11)

days after fertilization and 12.5 days after fertilization (Hildreth et al., 2008). Sympathetic innervation has been shown to occur at later stages compared with parasympathetic innervation in both avian and mammalian species (Kirby et al., 1980; Shoba and Tay, 2000). In the mouse, autonomic efferent innervation precedes sensory innervation, as shown by the later appearance of calcitonin gene related peptide (CGRP) positive nerve fibers in comparison to neuropeptide Y (NPY) immunoactivity in the rat (Shoba and Tay, 2000).

It has been proposed that cardiac parasympathetic neurons from the DVM and the NA project their axons to the intrinsic cardiac neurons and that the neurons from the DVM regulate cardiac inotropism, while those in the NA are related to heart rate control (Armour, 2008; Gatti et al., 1995). Some studies have shown the importance of the intrinsic cardiac ganglia in modulating relay between extrinsic autonomic nerves and the nodal tissues of the cardiac conduction system. Specifically, autonomic inputs from the left and right vagal branches are regulated to varying degrees by different cardiac ganglia, revealing complex interlinking pathways in the control of heart rate. These pathways have been revealed in studies of atrial fibrillation, which can be induced through electrical stimulation of auto-nomic nerves or cardiac ganglia (Patterson et al., 2005; Scherlag et al., 2005a; Scherlag et al., 2005b). A study in the dog (Hou et al., 2007) revealed that the right and left vagal trunks exert sympathetic control over heart rate, but that both inputs are modulated through specific, interlinking pathways, with these pathways differing between each node. There is, however, a degree of variation in these pathways between different individuals. Failure completely to attenuate the autonomic responses following ablation of these ganglia suggested that the modulation involves other ganglia within the cardiac network (Hou et al., 2007).

1.2. Neurotransmiters in Cardiac Innervation

Regulation of cardiac function by the autonomic nervous system plays crucial roles in the response of the organism to external stimulation. The sympathetic nervous system increases heart rate and the force of contraction via effects on the function of the sarcoplasmic reticulum within the cardio-myocytes and on ion channel activity. Preganglionic sympathetic axons from neurons in the T1-T5 spinal segments project to secondary sympa-thetic neurons that are located in the sympasympa-thetic chain, as well as the me-diastinal and intrinsic cardiac ganglia (Armour, 2008; Horackova et al., 1999; Kawashima, 2005; Richardson et al., 2006). Most of these sympa-thetic neurons contain TH, which is required for the synthesis of

(12)

norepi-nephrine (Maslyukov et al., 2006). The parasympathetic system, in contrast, counteracts the action of the sympathetic nerves, slowing the heart rate. Intrinsic parasympathetic neurons are primarily located within the atrial epicardial ganglia and these neurons contain choline acetyltransferase (Mabe et al., 2006), which is obligatory for the synthesis of acetylcholine (ACh). Postganglionic cholinergic axons control cardiac function by releasing ACh, which has a direct inhibitory influence on cardiomyocytes and a prejunctional inhibitory effect on sympathetic axons (Weihe et al., 2005). Under normal conditions, the balance between these two opposing systems is well maintained. The local environment plays an important role in determining the specific neurotransmitter used by presumptive autonomic postganglionic neurons.

Most sympathetic neurons express norepinephrine (NE) as a primary neurotransmitter, along with a range of peptides that modulate signaling. Norepinephrine binds to a range of adrenergic receptors, broadly classified as a and b. Under normal circumstances, a high affinity b-adrenoceptor binds to mul-tiple G proteins, activating adenylyl cyclase to produce cAMP. This then improves excitation-contraction coupling within the cardiac muscle, increasing heart rate and force. Norepinephrine is produced from its precursor, hydroxylase, dopamine b-hydroxylase, and phenylethanolamine N-methyl transferase, all found within the cell body of sympathetic neurons. Once produced, the NE is transported along the axon to nerve terminals where it is stored. In all species studied to date, maturation of the sympathetic innervation and onset of function occurs late in the fetal, or early in the neonatal, heart. Despite this, response to catecholamines, in the form of tachycardia, occurs well before birth, suggesting that b-adrenergic recaptors must be present during the fetal period, perhaps as early as 9.5 days after fertilization in the mouse (Liu et al., 1999). Indeed, although few TH positive neurons are found in the heart until late in gestation, myocardial cells expressing TH, dopamine beta-hydroxylase, and phenylethanolamine N-methyl transferase are found interspersed throughout the myocardium in the mouse early in its gestation. By mid-gestation, they have localized to the regions of formation of the sinus and atrioventricular nodes (Ebert and Thompson, 2001). These data are supported by the finding that, following sympathectomy, TH activity can still be detected in the chick heart (Stewart and Kirby, 1985). Although the roles of these myocardially produced cate-cholamines remain unclear, the fact that the heart can respond to these fac-tors from early in gestation suggests that they may be playing an important role. A wide range of cotransmitters are found within the sympathetic ner-vous system, functioning alongside norepinephrine. Although the mecha-nisms underlying the specification of these cotransmitters remain unclear,

(13)

cell lineage plays an important role. Of these cotransmitters, the adeno-sinergic system, using adenosine, is the earliest functionally responsive sys-tem in the rodent heart, with activation of the adenosine receptor slowing the heart rate from 8 days after fertilization in the rat embryo (Porter and Rivkees, 2001). Adenosine inhibits NE release from sympathetic nerve endings and is important both as a vasodilator and for its anti-arrhythmic properties. Similar to catecholamines, cardiac adenosine receptors are functional well before sympathetic innervation occurs. It has been suggested that fetal plasma may be the source early in gestation (Sawa et al., 1991).

Neuropeptide Y (NPY) is the most abundant peptide in the heart, being found in postganglionic sympathetic neurons synapsing on postganglionic parasympathetic neurons in the cardiac ganglia, endocardium, and myocardium (Palmiter et al., 1998). It has been shown to inhibit release of catecholamines and regulate the vasoconstrictor action of norepinephrine. Neuropeptide Y has also been shown to produce vasoconstriction of coronary vessels in many species, including human (Michel et al., 1989). Neuropeptide Y is first seen late in development in the region of the sinus and atrioventricular nodes and in the intrinsic cardiac ganglia. Levels increase in the immediate postnatal period, and then remain steady thereafter. Some studies have suggested that the peptide can produce long-term inhibition of the responses to vagal stimulation in the sinus and atrioventricular nodes (Potter, 1987). Neuropeptide Y has been suggested to play a crucial role as a neuromodulator in the complex interactions that take place between different branches of the sensory and autonomic innervation of the heart (Horst, 2000).

Acetylcholine (ACh) is the main neurotransmitter released by preganglionic and postgangionic vagal nerve terminals. Two types of choli-nergic receptor have been described, namely muscarinic and nicotinic. Mus-carinic receptors are the main type found on cardiac effector cells, whereas nicotinic receptors are found on intrinsic parasympathetic neurons. Acetyl-choline is mainly synthesized at the nerve endings where it is stored until release (Loffelholz and Pappano, 1985). Acetycholine can generally be detected in the heart at an earlier stage than can norepinephrine, reflecting the earlier onset of parasympathetic innervation than sympathetic ingrowth (Pappano, 1977).

Vasoactive intestinal peptide (VIP) is a peptidergic cotransmitter in cholinergic parasympathetic neurons. Neurons expressing this peptide are localized primarily in relation to blood vessels in microvascular beds (Della et al., 1983), and to the sinus and atrioventricular nodes and coronary vessels (Weihe et al., 1984). The peptide has direct effects on heart rate. In the dog, it has been reported to be twice as potent as norepinephrine in

(14)

increasing heart rate (Rigel, 1988). It has also been suggested to modulate the activity of acetylcholine and norepinephrine (Ferron et al., 1985).

Somatostatin immunoreactivity is associated with intramural para-sympathetic neurons, localizing to nerve fibers in the myocardium, endo-cardium, and the conduction system. The actions of somatostatin, slowing the heart rate and decreasing cardiac output, closely resemble those caused by vagal stimulation. It exerts its suppressing cardiac effects by acting on calcium channels excited by b-adrenoceptor activity (Diez et al., 1985) and/or by inducing the release of acetylcholine from intracardiac para-sympathetic neurons (Wiley et al., 1989), but may also modulate sym-pathetic neurotransmission by acting at a postjunctional site.

The neurotransmitters substance P (SP) and calcitonin gene-related peptide (CGRP) act on the heart via the sensory-motor pathway. Both transmitters are stored in granules within the nerve terminal, are coreleased on stimulation, and interact with parasympathetic nerves (Armour et al., 1993). SP-containing nerve fibers are found surrounding blood vessels within the myocardium, and within the sinus and atrioventricular nodes (Crick et al., 1994). Substance P, a potent vasodilator, although it has no effect on heart rate, may control parasympathetic activity (Corr, 1992). Calcitonin gene-related peptide is scarce in the human heart (Crick et al., 1994). It is a potent vasodilator of peripheral blood vessels, including the coronary arteries. In the mouse, it is expressed in sensory nerves within the heart, appearing around the time of birth. These, with other data, suggest that the sensory innervation of the heart occurs later than autonomic innervation (Gordon et al., 1993).

The myocardium also secretes endocrine factors that regulate the cardiac nervous system. The two most abundantly expressed proteins are the atrial and brain natriuretic peptides, which exert an inhibitory effect on sym-pathetic input to the heart. These factors are mainly expressed in the atria, although they are also expressed in other regions of the heart. They have additional roles, such as diuretic and natriuretic homeostatic function in the kidney, and inhibitory effects on the renin-angiotensin-aldosterone system (McGrath and de Bold, 2009; McGrath et al., 2005). A number of other nonadrenergic noncholinergic transmitters have been localized within the heart of various species, including opioids, neurotensin, and peptide histidine isoleucine, albeit that their actions remain poorly defined.

1.3. Physiology of Intrinsic Cardiac Neurons

It has been shown that neurons with distinct electrophysiological behavior differ in their number of dendritic processes and patterns of

(15)

synaptic inputs (Edwards et al., 1995; Klemm et al., 1997). Many cardiac neurons are synaptically coupled to each other and/or excited by stimulation of extrinsic nerves. However, approximately 10% of intrinsic cardiac neurons appear to lack a synaptic input as determined by staining whole-mount preparations of guinea pig cardiac plexus with antibodies to synaptophysin (Klemm et al., 1997).

Histological studies have shown that a complex network of neurons exists within the mammalian cardiac ganglion. Intrinsic parasympathetic neurons are not the only neurons present and there is evidence suggesting that sensory neurons, interneurons, and efferent neurons are found in intracardiac ganglia. Furthermore, the presence of sympathetic fibers (Smith, 1999) and afferent nerve fibers (Hardwick et al., 1995) suggests that there is potential for interaction between these elements. Sympathetic nerve terminals are often found in close proximity to parasympathetic nerve terminals (Randall et al., 1965). This anatomical arrangement of sym-pathetic and parasymsym-pathetic nerve terminals makes it possible for transmitters released from the nerve terminals of one division to diffuse readily to terminals of the other division, as well as to cardiac muscle cells. Small intensely fluorescent (SIF) cells, which contain catecholamines, have also been shown to be present within mammalian intracardiac ganglia (Hassall and Burnstock, 1986; Jacobowitz et al., 1967; Seabrook et al., 1990). Therefore, both the soma and axon terminal of parasympathetic neurons may be under the physiological influence of catecholamines.

Sympathetic and parasympathetic (vagal) nervous systems exert antagonistic effects on the heart and interaction between the two systems is well established (Levy, 1971). In addition to parasympathetic efferent and sensory afferent neurons, there also exists a population of interneurons within the cardiac ganglia. The axons of these cells may reside within its particular ganglion or travel out into another ganglion cluster. This arrangement has been found in most species, including the dog (Xi et al., 1991), guinea pig (Steele et al., 1996), and rabbit (Papka, 1976). Inter-neurons within cardiac ganglia mediate lateral interactions between various ganglion neurons and allow a convergence of different inputs. Electrical stimulation of the stellate ganglion or the vagosympathetic trunk produces responses in ganglion cells, with variable latencies indicating polysynaptic connections (Gagliardi et al., 1988).

Visceral cardiac afferent neurons also exist in cardiac ganglia, which are identified by the presence of the neuropeptides SP and CGRP (Franco-Cereceda, 1988).The presence of these afferents in close association with parasympathetic neurons has led to postulation of the existence of a local reflex circuit within the heart. Spontaneous activity has been demonstrated

(16)

in canine cardiac neurons (Gagliardi et al., 1988). The spontaneous firing is entrained to events in the cardiac and respiratory cycles even with disconnection from the central nervous system, although the amount of activity is substantially less. These observations strongly suggest that cardiac afferent fibers transmit cardiopulmonary information to ganglion cells in a local reflex arc as well as through higher centers.

The sensory neurons associated with cardiac function have been identified inside the nodose, C1-T4 dorsal root, mediastinal and intrinsic cardiac ganglia (Armour, 2008; Foreman, 2007; Hopkins and Armour, 1989; Horackova et al., 1999). Recent findings suggest that intrinsic sensory cardiac neurons are also involved in local neural circuits via their axonal projections to efferent neurons distributed within the same or neighboring intrinsic ganglia (Armour, 2008). Possibly, this diversity of neurons composes an integrative neuronal network, which modulates extrinsic autonomic projections to the heart and mediates local cardiac reflexes (Armour, 2008; Armour and Ardell, 1994). In addition, the integrative function of the intrinsic cardiac neurons is under the tonic influence of neurons from the insular cortex, brainstem and spinal cord (Armour, 2008; Armour and Ardell, 1994). To define sensory neuronal subpopulations, immunohistochemistry for CGRP and SP are commonly used in recent anatomical investigations. These both neuropeptides have been employed to identify peptidergic class of nociceptors, although CGRP is clearly expressed in some non-nociceptive neurons as well (Lawson et al., 2002; Lawson et al., 1996; Lawson et al., 1993). Although SP is considered as a pain transmitter, receptors for this neuropeptide are expressed not only on neurons, but also on the surface of cardiomyocytes, endothelial cells and immunocytes, such as lymphocytes and macrophages (Church et al., 1996; Cook et al., 1994; Goode et al., 1998; Ho et al., 1997). Furthermore, it was demonstrated that SP contributes to a dilated cardiomyopathy and is essential for the pathogenesis of encephalomyocarditis viral myocarditis (D'Souza et al., 2007; Robinson et al., 2009). Both the CGRP and SP play a role of counter regulator in hypertension and coronary flow (Robinson et al., 2009).

1.4. Intrinsic Cardiac Nerve Plexus

The extensive network of neuronal cell bodies receiving parasym-pathetic vagal input and comprising the intrinsic cardiac ganglia of the mammalian heart has long been known (Meiklejohn, 1914; Woollard, 1926). These ganglia could well represent the final common pathway through which the diverse, extrinsic neural signals to the heart are modified

(17)

before being transmitted to the effector tissues, yet the precise function of the network and the way in which it mediates vagal input are largely unknown. A detailed description of the location, distribution, and pro-jections of the intracardiac ganglia has been provided for the heart in numerous ma-mmalian species (Armour et al., 1997; Arora et al., 2003; Baptista and Kirby, 1997; Batulevicius et al., 2003; Batulevicius et al., 2004; Batulevicius et al., 2005; Batulevicius et al., 2008; Ellison and Hibbs, 1976; Hoover et al., 2004; Hoover et al., 2009; Horackova et al., 1999; Horackova et al., 2000; Yuan et al., 1994; King and Coakley, 1958; Leger et al., 1999; Moravec and Moravec, 1984; Pardini et al., 1987; Parsons et al., 1987; Pauza et al., 1997a; Pauza et al., 2002b; Pauza et al., 1999; Pauza et al., 2000; Pauza et al., 1997b; Pauziene et al., 2000a; Pauziene et al., 2000b; Saburkina and Pauza, 2006; Saburkina et al., 2009; Saburkina et al., 2010). Cardiac ganglia are located in epicardium and within the myocardium or en-docardium (Batulevicius et al., 2003; Batulevicius et al., 2005; Pauza et al., 2002a; Pauza et al., 2000; Saburkina et al., 2010). Whereas most of the gan-glia are epicardial, the septal gangan-glia are located on the inner surface of the atria. Cardiac ganglionic cells located in the right atrium are associated with control of the sinoatrial node and neurons in the region of the inferior vena cava modulate AV conduction (Armour and Ardell, 1994).The exact anato-mical distribution of the intracardiac ganglia varies among species (Pauza et al., 2002a).

Cardiac ganglia are distributed in different regions of the atria in a number of mammalian species, surrounding the SA node, around the roots of the vena cava and pulmonary veins, interatrial septum, and in the proximity of the AV node. A typical cardiac ganglion consists of neurons, satellite cells and SIF cells. The mammalian cardiac ganglia contain unipolar, bipolar and multipolar neurons with differing dimensions and shapes. The neurons and satellite cells of the cardiac ganglia originate from neural crest cells that migrate to the heart. Upon arriving in the outflow tract the cells segregate into parasympathetic neurons and supporting cells to form the cardiac ganglia (Baptista and Kirby, 1997).

According to Armour et al. (1997) the human intrinsic cardiac nerve plexus is distributed extensively, most of its ganglia being located on the posterior surfaces of the atria and superior aspect of the ventricles. Atrial ganglia in the human heart were identified on 1) the superior surface of the right atrium, 2) the superior surface of the left atrium, 3) the posterior surface of the right atrium, 4) the posterior medial surface of the left atrium (the latter two fuse medially where they extend anteriorly into the interatrial septum), and 5) the inferior and lateral aspect of the posterior left atrium. Ventricular ganglionated plexuses were located in fat 1) surrounding the

(18)

aortic root, 2) at the origins of the right and left coronary arteries (the latter extending to the origins of the left anterior descending and circumflex coronary arteries), 3) at the origin of the posterior descending coronary artery, 4) adjacent to the origin of the right acute marginal coronary artery, and 5) at the origin of the left obtuse marginal coronary artery (Armour et al., 1997). Pauza et al. (2000) concluded that the human heart is innervated by seven subplexuses: the right atrium was innervated by two subplexuses, the left atrium by three, the right ventricle by one, and the left ventricle by three subplexuses. The highest density of epicardiac ganglia was identified near the heart hilum, especially on the dorsal and dorsolateral surfaces of the left atrium, where up to 50% of all cardiac ganglia were located (Pauza et al., 2000). The study of epicardiac ganglia in the human fetuses showed that the topography and structural organization of epicardiac neural plexus were typical for hearts of adult human. The largest ganglion number comprising 77% of all counted ganglia was identified on the dorsal atrial surface. Fetal epicardiac plexus in gestation period of 15-40 weeks contained 929±62 ganglia (Saburkina and Pauza, 2006).

A three-dimensional description of the distribution and organization of the canine intrinsic cardiac nervous system was developed in order to characterize its full extent physiologically. Collections of ganglia associated with nerves, i.e., ganglionated plexuses, were identified in specific locations in epicardial fat and cardiac tissue. Distinct epicardial ganglionated plexuses were consistently observed in four atrial and three ventricular regions, with occasional neurons being located throughout atrial and ventricular tissues. One ganglionated plexus extended from the ventral to dorsal surfaces of the right atrium. Another ganglionated plexus, with three components, was identified in a fat on the left atrial ventral surface. A ganglionated plexus was located on the mid-dorsal surface of the two atria, extending ventrally in the interatrial septum. A fourth atrial ganglionated plexus was located at the origin of the inferior vena cava extending to the dorsal caudal surface of the two atria. On the cranial surface of the ventricles a ganglionated plexus that surrounded the aortic root was identified. This plexus extended to the right and left main coronary arteries and origins of the ventral descending and circumflex coronary arteries. Two other ventricular ganglionated plexuses were identified adjacent to the origins of the right and left marginal coronary arteries (Yuan et al., 1994). Later on, it was performed study on whole canine hearts to highlight the differences of intrinsic neural plexus (INP) in dog and human. Pauza et al. (2002b) identified 13 locations between the canine ascending aorta and pulmonary trunk, around the pul-monary veins, and on every side of the superior vena cava, through which mediastinal cardiac nerves accessed the canine heart. Intrinsic nerves from

(19)

these locations extended within the canine epicardium by seven neuronal subplexuses. Intrinsic nerves and ganglia were found to be widely distributed in topographically consistent atrial and ventricular regions. The canine right atrium, including the sinoatrial node, was innervated by two subplexuses, the wall of the left atrium by three, and the right and left ventricles by two subplexuses. Depending on the age of the animal, the number of intrinsic ganglia per one canine heart might range from 400 up to 1,500. By taking into account the ganglion size and potential approximate number of neurons residing inside a ganglion of a certain size, it was estimated that on average about 80,000 intrinsic neurons are associated with the canine heart. A comparative analysis of the morphological patterns of the canine and human intrinsic cardiac neural plexuses showed that the topography of these plexuses may be considered as quite similar, but the structural and quantitative differences of the intrinsic cardiac neural subplexuses between dogs and humans are significant (Pauza et al., 2002a; Pauza et al., 2002b; Pauza et al., 1999; Pauza et al., 2000).

Sheep is routinely used in experimental cardiac electrophysiology and surgery, however it is noted a possible distinct neural control of the ventricles in the human and sheep hearts (Saburkina et al., 2010). Intrinsic cardiac nerves extend from the venous part of the ovine heart hilum along the roots of the cranial (superior) caval and the left azygos veins to both atria and ventricles via five epicardial routes: the dorsal right atrial, middle dorsal, left dorsal, right ventral, and ventral left atrial nerve subplexuses. Intrinsic nerves proceeding from the arterial part of the heart hilum along the roots of the aorta and pulmonary trunk extend exclusively into the ventricles as the right and left coronary subplexuses. Sheep hearts contained an average of 769±52 epicardial ganglia. Cumulative areas of epicardial ganglia on the root of the cranial vena cava and on the wall of the coronary sinus were the largest of all regions (Saburkina et al., 2010). Despite substantial interindividual variability in the morphology of ovine epicardiac neural plexus (ENP) right-sided epicardial neural subplexuses supplying the sinuatrial and atrioventricular nodes are mostly concentrated at a fat pad between the right pulmonary veins and the cranial vena cava. This finding is in sharp contrast with a solely left lateral neural input to the human atrioventricular node, which extends mainly from the left dorsal and middle dorsal subplexuses (Pauza et al., 2000; Saburkina et al., 2010).

The nerves entered the porcine epicardium at five sites: (1) ventro-medially to the origin of the superior vena cava, (2) dorsally to the origin of the superior vena cava, (3) among the pulmonary veins, (4) dorso-medially to the origin of the left azygos vein, and (5) ventrally to the left pulmonary vein (Batulevicius et al., 2008). Numerous ganglia and interconnecting

(20)

nerves were found to be concentrated in an epicardial fat in five atrial and six ventricular regions of the porcine heart (Arora et al., 2003). The five atrial ganglionic fields were identified (1) the ventral right atrial, (2) the right vena cava-right atrial, (3) the dorsal atrial, (4) the interatrial septal, and (5) the left superior vena cava-left atrial ones. Six ventricular ganglionic fields were identified in close proximity to the (1) roots of the aorta and pulmonary artery (craniomedial), extending along the left main coronary artery to the (2) ventral interventricular and (3) circumflex coronary arteries. (4) ganglionic fields were also identified around the origin of the dorsal interventricular coronary artery as well as the (5) right main and (6) right marginal coronary arteries. Isolated neurons were identified scattered throughout the cranial interventricular septum (Arora et al., 2003). Other investigators divided the porcine ENP according to the neural pathways (subplexuses): the left atrium received nerves by four subplexuses, left ventricle by three subplexuses, right atrium and right ventricle each by two subplexuses (Batulevicius et al., 2008). The estimated total number of the intrinsic ganglia per porcine heart was 362±52. About 55% of ganglia per porcine heart were accumulated on the left atrium while 36% on the right atrium. The percentage of ganglia within the porcine ventricular and para-aortic regions was 7.6% and 1.6%, respectively (Batulevicius et al., 2008).

The nerves entering the guinea pig heart were found both in the arterial and venous part of the heart hilum (Batulevicius et al., 2005). The nerves from the arterial part of the heart hilum proceeded into the ventricles, but the nerves from the venous part of the hilum formed a nerve plexus of the cardiac hilum located on the heart base. Within the guinea pig epicardium, intrinsic nerves divided into six routes and proceeded to separate atrial, ventricular and septal regions (Batulevicius et al., 2005). The intracardiac neurons from adult guinea pigs were amassed within 329±15 ganglia. The hearts of young guinea pig contained significantly fewer ganglia, only 211±27 (Batulevicius et al., 2005). Of all identified neurons in the guinea pig heart, 85-90% were located in ganglia (ganglionic neurons), the rest being isolated (individual neurons) (Horackova et al., 1999). The remarkable similarity was found in the architecture of the intracardiac nerve plexuses between guinea pig and rat (Batulevicius et al., 2003; Batulevicius et al., 2004; Batulevicius et al., 2005). Extrinsic nerves entered the rat heart in both the arterial and venous parts of the cardiac hilum. Extracardiac nerves entering the rat heart were found amid aorta and pulmonary trunk as well as along both right and left cranial veins. The nerves from the arterial part of the heart hilum extended directly to the ventricles but the nerves from the venous part of the hilum interconnected among themselves forming a nerve plexus of the cardiac hilum on the heart base. Within the rat

(21)

epicardium, intrinsic nerves clustered into six routes by which they selectively projected to different rat heart regions. Ventral wall of the ventricles was supplied by three neural subplexuses, dorsal ventricular wall by one subplexus; each atrium received nerves from two distinct subplexuses (Batulevicius et al., 2004). Also, the distributions of cardiac ganglia and vagal efferent projections to cardiac ganglia in mice and rats were quite similar both qualitatively and quantitatively (Ai et al., 2007). Using the tracer into the left NA, cardiac ganglia of different shapes and sizes were marked in the sinoatrial (SA) node, atrioventricular (AV) node, and lower pulmonary vein (LPV) regions on the dorsal surface of the atria. In each region, several ganglia formed a ganglionated plexus. The plexuses at different locations were interconnected by nerves. Vagal efferent fibers ramified within cardiac ganglia, formed a complex network of axons, and innervated cardiac ganglia with very dense basket endings around individual cardiac principal neurons (Ai et al., 2007).

The human intrinsic cardiac ganglia (ICG) range in size from those containing a few neurons to large ganglia measuring up to 0.5 x 1 mm (Armour et al., 1997). The human heart was estimated to contain more than 14,000 neurons. Neuronal somata varied in size and shape. Many axon terminals in intrinsic cardiac ganglia contained large numbers of small, clear, round vesicles that formed asymmetrical axodendritic synapses, whereas a few axons contained large, dense-cored vesicles (Armour et al., 1997). The canine intrinsic cardiac nervous system contains a variety of neurons interconnected via plexuses of nerves. Intrinsic cardiac ganglia range in size from ones comprising one or a few neurons along the course of a nerve to ones as large as 1 x 3 mm estimated to contain a few hundred neurons. Intrinsic cardiac neuronal somata vary in size and shape, up to 36% containing multiple nucleoli. Electron microscopic examination demonstrated typical autonomic neurons and satellite cells in intrinsic cardiac ganglia (Yuan et al., 1994). Many of their axon profiles contained large numbers of clear, round, and dense-core vesicles. Asymmetrical axodendritic synapses were common (Yuan et al., 1994).

Approximately 3,000 neuronal somata were estimated to compose intrinsic cardiac nervous system (ICNS) of the pig (Arora et al., 2003). Some ganglia contained more than 100 neurons. Neuronal somata had dimensions of roughly 33.1 (short axis) by 46.3 (long axis) microms (Arora et al., 2003). Other authors calculated, that on average, the porcine heart contained about 12,000 intrinsic neurons (Batulevicius et al., 2008).

An estimate number of neuronal somata in guinea pig individual gan-glia were 100-300 (Horackova et al., 1999). The total number of the intra-cardiac neurons estimated per atria was 1,510±251 (Leger et al., 1999), in

(22)

the whole heart were counted 2,321±215 neurons, and this number did not differ significantly between young and adult animals. In adult guinea pigs approximately 60% of the intracardiac neurons were distributed within gan-glia of not more than 20 neurons, but the gangan-glia of such size accumulated only 45% of the neurons in young animals (Batulevicius et al., 2005). Ganglia in the guinea pig heart contained three sub-populations of neurons: approximately 80% of ganglionic neurons were large – 15-40 microm dia-meters, approximately 20% had smaller diameters - less than 15 microm and 5% of ganglionic neurons were small – less than 20 microms (Horackova et al., 1999).

The total number of intrinsic cardiac neurons in old rats was 6,576±317 (Batulevicius et al., 2003). The juvenile animals contained significantly fewer such neurons, only 5,009±332. Approximately 70% of all intracardiac neurons were amassed within the heart hilum, while 30% of the neurons were distributed epicardially. Within the interatrial septum, only 11±11 neurons were identified in the juvenile and 6±4 neurons in old rats (Batulevicius et al., 2003).

The morphological organization and structure-function correlation of mammalian intracardiac ganglion cells using conventional intracellular microelectrode techniques were applied to the tissue whole mount prepa-ration of the canine intracardiac ganglia (Xi et al., 1991). Somata were elon-gated (mean 62 x 40 microns) and had 2-12 primary dendrites restricted within the ganglion. Almost half of the neurons had either a short axon that was traced only within the ganglion or no axon distinguishable. Authors suggest that these neurons may have been intraganglionically active neurons. The other cells had a long axon that either coursed out of the ganglion to peripheral cardiac tissue or exited the ganglion via intergan-glionic nerve to innervate more remote cardiac tissue or cells in other intra-cardiac ganglia. Interaction between neurons was suggested by the close proximity of processes from different neurons. Previously defined electro-physiological cell types (R-, S-, and N-cells), which were significantly different in their passive and active membrane properties, had different mor-phological features of the somata but not the axonal or dendritic processes. Intraganglionic or long axon neurons were not associated with a particular electrophysiological cell type (Xi et al., 1991). Most neuronal somata of porcine intrinsic cardiac nervous system were multipolar, a small population of unipolar neurons being identified in atrial and ventricular tissues (Arora et al., 2003). A morphological study of neurons in the nerve plexus that lies beneath the pulmonary arteries on the myocardium of the left atrium of rats and guinea pigs revealed at least two major types of neuron: unipolar (61.2%) and multipolar (38.8%). The neuron somata exhibit no significant

(23)

difference in their length or width (Pauza et al., 1997b).Classification of nerve cells in the terminal plexus of the rat AV junction according to their three-dimensional (3-D) morphology confirmed that they could be divided into three categories: (1) large unipolar neurons with axonal projections directed toward the interventricular junction, (2) large unipolar or bipolar neurons, and (3) small multipolar interneurons (Moravec and Moravec, 1998). It is demonstrated that there are diverse populations of cardiac ganglion cells in the guinea pig and that some of these neurons may act as interneurons within the intrinsic cardiac plexuses (Steele et al., 1994). Neurons from adult mouse heart, maintained in culture, were primarily unipolar, and 89% had prominent neurite outgrowth after 3 days. Many neurites formed close appositions with other neurons and nonneuronal cells. Neurite outgrowth was drastically reduced when neurons were kept in culture with a majority of nonneural cells eliminated (Hoard et al., 2007).

1.5. Immunohistochemical Characterization of Intrinsic Cardiac Neurons

The studies of autonomic ganglia have shown that specific combinations of neuropeptides and other potential neurotransmitters distinguish different functional types of neurons (Hoover et al., 2009). Therefore it is highly likely that vagal transmission in the heart is modified by sympathetic, sensory and intrinsic neurons and those cardiac ganglia are complex integrators of convergent neuronal activity rather than simple relays. The findings demonstrate that the human intrinsic cardiac neuronal somata has a complex neurochemical anatomy, which includes the presence of a dual cholinergic/nitrergic phenotype for most of its neurons, the presence of noradrenergic markers in a subpopulation of neurons, and innervation by a host of neurochemically distinct nerves (Hoover et al., 2009). Most human intrinsic cardiac neuronal somata displayed immunoreactivity for the cholinergic marker ChAT and the nitrergic marker neuronal nitric oxide synthase (NOS). A subpopulation of intrinsic cardiac neurons also stained for noradrenergic markers. While the most intrinsic cardiac neurons received cholinergic innervation evident as punctate immunostaining for the high affinity choline transporter (CHT), some lacked cholinergic inputs (Hoover et al., 2009). AChE and TH-immunoreactive (IR) cell bodies were observed in the atrial ganglionated plexuses in the pig heart (Crick et al., 1999a). Experiments demonstrated that guinea pig cardiac ganglia contain prominent pericellular baskets of varicose nerve terminals of sympathetic and sensory origin (Steele et al., 1994). The results indicate that all postganglionic neurons in guinea pig cardiac ganglia are likely to utilize

(24)

acetylcholine as a neurotransmitter, regardless of their functional role in circuitry of cardiac innervation, and each of these neurons is likely to receive cholinergic input (Mawe et al., 1996).

The patterns of coexistence of neurochemicals in guinea pig cardiac ganglion cells were examined using a multiple-labelling immunohisto-chemistry. Many neurons were found to contain the somatostatin immuno-reactivity with various combinations of immunoimmuno-reactivity for SP, and NOS (Mawe et al., 1996; Steele et al., 1994), also neurons were protein gene product 9.5 (PGP 9.5) immunoreactive, exhibiting ChAT, TH, NPY, VIP immunoreactivity (Horackova et al., 1999; Parsons et al., 2006). The study of distribution of cholinergic and adrenergic neurons in whole-mount prepa-rations of the guinea pig atria showed that all neuronal somata expressing PGP 9.5-IR also expressed ChAT-IR, suggesting that these neurons were cholinergic. No neuronal somata expressed TH-IR or contained detectable amines but these elements were expressed by somata of small cells throughout the atria, primarily associated with ganglia (Leger et al., 1999). Whereas, previously it was reported that immunoreactivity for ChAT was also observed in a large proportion of the small TH-IR neurons that exist in guinea pig cardiac ganglia (Mawe et al., 1996). The cell bodies of the rat intramural ganglion cells localized between the right and left branches of the bundle of His were TH-IR and dopamine beta hydrohylase immunoreactive (DBH-IR) (Moravec et al., 1990).

The neurochemistry of intracardiac neurons of the rat intrinsic ganglia was investigated in whole-mount preparations. This technique allowes to study the morphology of ganglionated nerve plexus found within the atria as well as of individual neurons. The results of this study indicated a moderate level of chemical diversity within the intracardiac neurons of the rat. Such chemical diversity may reflect functional specialisation of neurons in the intracardiac ganglia (Richardson et al., 2003). The VIP, CGRP, pituitary adenylate cyclase-activated peptide (PACAP), SP and TH immunoreactivity was observed in nerve fibres within the ganglia, but never in neuronal somata of the rat cardiac ganglia (Richardson et al., 2003). Catecholamine-handling intrinsic ganglion neurons were observed as SIF cells and large-diameter neurons (Slavikova et al., 2003). The SIF cells are most probably dopaminergic and serotonergic neurons, whereas large-diameter intrinsic cells seem to represent a subpopulation of norepinephrine – and/or epineph-rine-secreting neurons (Slavikova et al., 2003). Moravec and colleagues investigations show that ganglion cells of sulcus terminalis as well as the epicardial ganglia enclosed between the superior vena cava and ascending aorta are VIP-IR and TH-negative, but NPY-IR and DBH-IR (Moravec et al., 1990).

(25)

To defined the basic phenotypic properties of the dissociated neurons, cells from adult mouse heart were maintained in culture (Hoard et al., 2007). All neurons in coculture showed immunoreactivity for a full complement of cholinergic markers, but about 21% also stained for tyrosine hydroxylase (Hoard et al., 2007). Disruption of cholinergic function in diabetic mice cannot be attributed to a loss of cardiac cholinergic neurons and nerve fibers or altered cholinergic sensitivity of the atria. Instead, the decreased res-ponses to vagal stimulation might be caused by a defect of preganglionic cholinergic neurons and/or ganglionic neurotransmission. The increased density of cholinergic nerves observed at the sinoatrial node of diabetic mice might be a compensatory response (Mabe and Hoover, 2011).

1.6. Immunohistochemistry of Nerves and Nerve Fibers in Intrinsic Cardiac Nervous System

It is reported that the neonatal human heart possesses a rich supply of autonomic nerves (Chow et al., 1995). The atria possess at least two popu-lations of nerves, presumably sympathetic and vagal, whereas the walls of the ventricles are innervated principally by presumptive sympathetic nerves (Chow et al., 1995). Numerous PGP-IR nerves were found in the atrial myocardium, forming perivascular plexuses and lying in close apposition to myocardial cells. Fewer PGP-IR nerves were found amongst the myo-cardium of the ventricles. Both DBH-IR and TH-IR nerves demonstrated a similar pattern of distribution as that of PGP-IR nerves; in the atria, however, they were less numerous, while in the ventricles, their density approximated to that of PGP-IR nerves. The density of AChE positive nerves in the walls of the atria was less than that of PGP-IR nerves although their distribution patterns were similar. In the ventricles, AChE positive nerves were rarely observed (Chow et al., 1995). Later, it was reported that there were more AChE-IR nerves in the subendocardial area than in the subepicardial area of the myocardium. In the atrium, AChE-positive nerves were more numerous than IR nerves. By contrast, there were more TH-positive nerves than AChE-TH-positive nerves in the ventricle. Predominancy of the distribution density at the anterior to the posterior wall of the ventricle was observed for TH-positive nerves (Kawano et al., 2003). Moreover, peptidergic, nitrergic, and noradrenergic nerves provided substantial innervation of intrinsic cardiac ganglia (Hoover et al., 2009).

The predominant neural subpopulation displayed AChE activity, throughout the endocardium, myocardium and epicardium in the pig heart. A large proportion of nerves in the ganglionated plexus of the atrial epicardial tissues displayed AChE activity, together with their cell bodies.

(26)

Tyrosine hydroxylase immunoreactive nerves were the next most prominent subpopulation throughout the heart. Endocardial TH- and NPY-immunoreactive nerves also displayed a right to left gradient in density, whereas in the epicardial tissues they showed a ventricular to atrial gradient (Crick et al., 1999a).

Varicose nerve fibers that were immunoreactive for ChAT were abundant in guinea pig cardiac ganglia, with every cardiac neuron lying in close apposition to one or more labelled varicosities. ChAT-IR nerve fibers were also observed in large vagosympathetic fiber bundles, in intergan-glionic fiber bundles, and passing individually within the myocardium (Mawe et al., 1996). Amine- and TH- containing varicosities were also present in guinea pig ganglia, representing potential sites for adrenergic modulation of ganglionic neurotransmission (Leger et al., 1999). High-affinity choline transporter (CHT) and AChE immunoreactive nerve fibers and nerves were very abundant in the sinus and AV nodes, bundle of His, and bundle branches. Both markers also delineated a distinct nerve tract in the posterior wall of the right atrium. AChE-positive nerve fibers were more abundant than CHT-IR nerves in working atrial and ventricular myo-cardium. CHT-IR nerves were rarely observed in left ventricular free wall. Both markers were associated with numerous parasympathetic ganglia that were distributed along the posterior atrial walls and within the interatrial septum, including the region of the AV node (Hoover et al., 2004). The distribution of CHT-IR nerve fibers and parasympathetic ganglia in the guinea pig heart suggests that heart rate, conduction velocity, and auto-maticity are precisely regulated by cholinergic innervation. In contrast, the paucity of CHT-IR nerve fibers in the left ventricular myocardium implies that vagal efferent input has little or no direct influence on ventricular contractile function in the guinea pig (Hoover et al., 2004).

Cholinergic innervation of the rat heart was studied using various cholinergic markers including AChE, VAChT, choline acetyltransferase of a peripheral type (pChAT) and the conventional form of ChAT (cChAT). The density of pChAT-positive fibers was very high in the conducting system, high in both atria, the right atrium in particular, and low in the ventricular walls. pChAT and VAChT immunoreactivities were closely associated in some fibers and fiber bundles in the ventricular walls (Yasuhara et al., 2007). Yasuhara and colleagues indicate that intrinsic cardiac neurons homogeneously express both pChAT and cChAT and innervation of the ventricular walls by pChAT- and VAChT-positive fibers provides morpho-logical evidence for a significant role of cholinergic mechanisms in ventri-cular functions (Yasuhara et al., 2007). Nerve fibres showing DBH-IR and TH-IR were present in the ganglia; some of these fibres being closely

(27)

associated with the ganglion cells or with the cells showing TH-IR (Forsgren et al., 1990).Cholinergic nerve fibers in the mouse heart were abundant in the sinus and atrioventricular nodes, ventricular conducting system, interatrial septum, and much of the right atrium, but less abundant in the left atrium. The right ventricular myocardium contained a low density of cholinergic nerves, which were sparse in other regions of the working ventricular myocardium. Some cholinergic nerves were also associated with coronary vessels (Mabe et al., 2006).

Calcitonin gene-related peptide immunoreactive nerves were the most abundant peptide-containing subpopulation after those possessing NPY immunoreactivity. They were most abundant in the epicardial tissues of the ventricles of the pig heart (Crick et al., 1999a). Populations of axons con-taining SP were observed in guinea pig cardiac ganglia. Intrinsic axons containing SP-IR were very rare. The regions of the heart with the densest innervation by axons of intrinsic neurons were the cardiac valves, the atrio-ventricular node and the sino-atrial node (Steele et al., 1996). Nerve fibres showing SP-IR, CGRP-IR were present in the rat ganglia; some of these fibres being closely associated with the ganglion cells or with the cells showing TH-IR (Forsgren et al., 1990).

1.7. Autonomic Control and Innervation of Mammalian Cardiac Conduction System

The so-called specialized tissues within the heart are the sinus node, the atrioventricular conduction system, and the Purkinje network (Anderson et al., 2009). Atrioventricular (AV) nodal conduction time is known to be modulated by the autonomic nervous system. Parasympathetic stimulation slows conduction through the AVN via the hyperpolarizing effects of the neurotransmitter acetylcholine (Mazgalev et al., 1986), and sympathetic stimulation accelerates AV nodal conduction via the effects of norepi-nephrine. Many studies have shown that autonomic modulation of the sinoatrial node (SAN) causes the pacemaker site to shift in many species. Sympathetic stimulation typically induces the SAN pacemaker to shift toward the superior vena cava, whereas parasympathetic stimulation shifts the leading pacemaker toward the inferior vena cava (Boineau et al., 1989; Fedorov et al., 2006; Shibata et al., 2001). Clinically, the AV junctional rhythm is evident in patients with sick sinus syndrome, in those receiving isoproterenol infusion, and in patients subsequent to slow pathway ablation (Matsushita et al., 2001). In normal heart, the AVN is located near the apex of the triangle of Koch, and has at least two inputs, each with unique electrophysiologic properties. The atrioventricular node has electrical

(28)

connection between the atria and ventricles, serving as the gateway to the His-Purkinje system (Moe et al., 1956). The specialized tissues of the AVN, its transitional cells, and inputs are collectively termed the atrioventricular (AV) junction (Billette, 2002). Proximal atrioventricular bundle, atrio-ventricular node, and distal atrioatrio-ventricular bundle are distinct anatomic structures with unique histological characteristics and innervation (Racker and Kadish, 2000). It has long been known that the AV junction can act as a pacemaker (Watanabe and Dreifus, 1968). The pacemaking activity of the AV junction now has been demonstrated to reside predominantly in the slow pathway (Dobrzynski et al., 2003), which is an AVN input located within the isthmus between the tricuspid annulus and the coronary sinus ostium in both rabbits and humans (Inoue and Becker, 1998; Medkour et al., 1998). Autonomic control of the AV junction has been the subject of numerous studies. Structurally, the distribution of sympathetic and parasympathetic neurons, as well as the distribution of β-receptors, has been documented in rat AV junction (Petrecca and Shrier, 1998). Functionally, autonomic modu-lation of AV junctional conduction properties is well documented. Rate control during atrial fibrillation has been attempted with vagal stimulation of the AV junction using subthreshold stimulation, which excites intra-cardiac neurons without causing myocyte contraction (Mazgalev et al., 1999). The feasibility of the subthreshold stimulation has been tested with regard to energy requirements for chronic neurostimulation in canines (Soos et al., 2005) and catheter placement to achieve parasympathetic stimulation in humans (Schauerte et al., 2001). Despite many structural and functional studies, few have examined the subthreshold stimulation technique and quantitative immunohistochemistry to investigate autonomic control and innervation of the AV junctional pacemaker in the same preparation (Dobrzynski et al., 2003; Hucker et al., 2007).

The morphology of the human atrioventricular node, atrioventricular bundle and bundle branches is described by serial sections of tissue bounded by the ostium of the coronary sinus, the pars membranacea, the septal leaflet of the tricuspid valve and the atrial and ventricular septa (Roberts and Castleman, 1979). Later on serial semithin and thin sections of the interatrial septum and atrioventricular junction of adult rat were examined in light and electron microscopes (Moravec and Moravec, 1984). It was observed rings of specialized tissue mainly in hearts from rats, mice, and guinea pigs, negative for connexin 43 (Cx43) but positive for hyperpolarization activated cyclic nucleotide-gated potassium channel 4 (HCN4). Each ring takes its origin from an inferior extension of the atrioventricular node. The rightward ring runs around the vestibule of the tricuspid valve, whereas the leftward ring encircles the mitral valve. On returning toward the atrial septum, the

MORPHOLOGICAL AND IMMUNOHISTOCHEMICAL PATTERNS OF THE INTRINSIC GANGLIONATED NERVE PLEXUS IN THE MOUSE HEART

Kristina Rysevaitė