The Enteric Nervous System

The Enteric Nervous System

John Barton Furness PhD, FAA Department of Anatomy and Cell Biology, University of Melbourne, Victoria, Australia

The Enteric Nervous System

The Enteric Nervous System

John Barton Furness PhD, FAA Department of Anatomy and Cell Biology, University of Melbourne, Victoria, Australia

© 2006 John B. Furness Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First published 2006 Library of Congress Cataloging-in-Publication Data Furness, John Barton. The enteric nervous system / John B. Furness. p. ; cm. Includes bibliographical references and index. ISBN-13: 978-1-4051-3376-0 ISBN-10: 1-4051-3376-7 1. Gastrointestinal system--Diseases. I. Title. [DNLM: 1. Enteric Nervous System--physiology. 2. Neurons --physiology. WL 600 F988e 2006] RC817.F87 2006 616.3--dc22 2005024527 ISBN-13: 978-1-4051-3376-0 ISBN-10: 1-4051-3376-7 A catalogue record for this title is available from the British Library Set in 10/13½ Sabon by Sparks, Oxford – www.sparks.co.uk Printed and bound by Narayana Press, Odder, Denmark Commissioning Editor: Alison Brown Editorial Assistant: Saskia Van der Linden Development Editor: Rob Blundell Production Controller: Kate Charman For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

Contents

Preface, ix Abbreviations, xi

1: Structure of the enteric nervous system, 1 The enteric plexuses, 3 Interconnections between the plexuses, 14 Extent of the ganglionated plexuses, 15 Intramural extensions of extrinsic nerves, 17 Electron microscope studies, 17 Enteric glia, 20 The structural similarities and functional differences between regions may have an evolutionary basis, 21 Development of the enteric nervous system, 23 Maturation of enteric neurons and development of function, 26 Changes in enteric neurons with aging, 27 Summary and conclusions, 28

2: Constituent neurons of the enteric nervous system, 29 Shapes of enteric neurons, 31 Cell physiological classifications of enteric neurons, 43 Functionally defined enteric neurons, 53 Neurons in human intestine with equivalence to those investigated in laboratory animals, 76 Summary and conclusions, 78

v

vi

CONTE NTS

3: Reflex circuitry of the enteric nervous system, 80 Evolution of ideas about enteric circuitry, 80 Motility controlling circuits of the small and large intestine, 81 Intrinsic secretomotor and vasomotor circuits, 88 Assemblies of neurons, 93 Circuits in the esophagus and stomach, 96 Co-ordination of motility, secretomotor, and vasomotor reflexes, 98 Circuits connecting the intestine, biliary system, and pancreas, 98 Sympathetic innervation of the gastrointestinal tract, 99 Summary and conclusions, 101

4: Pharmacology of transmission and sites of drug action in the enteric nervous system, 103 Chemical coding and multiple transmitters, 103 Transmitters of motor neurons that innervate the smooth muscle of the gut, 104 Transmitters at neuro-neuronal synapses, 111 Sites within the reflex circuitry where specific pharmacologies of transmission can be deduced to occur, 120 Transmission from entero-endocrine cells to IPANs, 126 Roles of interstitial cells of Cajal in neuromuscular transmission, 127 Transmitters of secretomotor and vasodilator neurons, 128 Synapses in secretomotor and vasodilator pathways, 130 Transmitters of motor neurons innervating gastrin cells, 130 Summary and conclusions, 130

5: Neural control of motility, 132 Rhythmic activity of gastrointestinal muscle, 132 Structure and properties of interstitial cells of Cajal, 134 Relationship between slow wave activity and neural control, 138 Gastric motility, 140 Patterns of small intestine motility and their intrinsic neural control, 147 Motility of the colon, 157 Neural control of the esophagus, 159 Gall bladder motility, 160 Sphincters, 161 Muscle of the mucosa, 165

C ONT ENT S

v ii

Mechanism of sympathetic inhibition of motility in non-sphincter regions, 166 Sympathetic innervation of the sphincters, 169 Physiological effects of noradrenergic neurons on motility in undisturbed animals, 170 Reflex activities of sympathetic neurons that affect motility, 171 Summary and conclusions, 178

6: Enteric neurons and the physiological control of fluid secretion and vasodilation, 180 Water and electrolyte secretion in the small and large intestines, 180 Reflex control of water and electrolyte secretion, 182 Secretion of gastric acid, 189 Pepsinogen secretion, 194 Gastric secretion of bicarbonate, 195 Secretion into the gall bladder, 195 Pancreatic exocrine secretion, 196 Summary and conclusions, 198

7: Disorders of motility and secretion and therapeutic targets in the enteric nervous system, 200 Therapeutic endpoints for motility disorders, 201 Therapies for secretory diarrheas, 205 Enteric neuropathies involving neuronal loss or phenotypic changes, 206 Mitochondriopathies with intestinal manifestations, 207 Irritable bowel syndrome and plasticity of enteric neurons, 208 Summary and conclusions, 210 Epilogue: the future of enteric neurobiology, 211 References, 214 Index, 267

Preface

The enteric nervous system is of special interest because it is the only substantial grouping of neurons outside the central nervous system that form circuits capable of autonomous reflex activity. In humans it contains around 500 million neurons that fall into about 20 functional classes. Because of its size, complexity, and certain structural similarities, it has been likened to a second brain. Although the enteric nervous system was discovered almost 150 years ago, and several remarkably insightful hypotheses about its functions were made in the 19th century, a long period ensued in which progress was meagre in comparison to the effort made, because methods available were not adequate to determine the intrinsic circuitry of the enteric nervous system and the properties of its constituent neurons. In the last 20–30 years, new techniques, and excellent application of such techniques, have provided a wealth of information on the structural complexity, neuron types, and connectivity of the enteric nervous system and on the transmitters and cell physiology of enteric neurons. Beginning at an earlier time, and proceeding in parallel, have been investigations of the patterns of movement and secretory functions of the digestive tract, and their control. This book aims to integrate the detailed cellular knowledge of the enteric nervous system with the more macroscopic information that is provided by physiological studies of organs, especially in the living animal or human. In doing so, I have tried to deal with the emergence of knowledge in historical perspective, where possible by drawing on early information to acknowledge the contributions made by pioneers of enteric neurobiology, and in places to reproduce original illustrations from early publications. I hope that the reader will enjoy this approach. I have also created many new illustrations, especially of the organization of enteric nerve circuits, which I hope will provide an understanding of the enteric nervous system that the written word cannot easily convey. The first four chapters lay the groundwork, by dealing with the structure of the enteric nervous system, the defining cell physiological, morphological, ix

x

PRE FACE

and neurochemical properties that allow its neurons to be functionally classified, the enteric neurotransmitters and the intrinsic nerve circuits within the alimentary tract. This is followed by two chapters on gastrointestinal physiology, first on the contractile activity of the muscular walls of the digestive tract and the second on secretory function. In these two chapters I try to develop an understanding of the roles of enteric neurons and how they perform these roles. I have also sought to relate control through enteric circuits to control exerted by the vagus and the sympathetic innervation of the digestive organs, and to a lesser extent through the pelvic nerves. The involvement of altered structure and function of the enteric nervous system in some disease states is well recognized. Nevertheless, how to use the new-found knowledge of the enteric nervous system to understand the relations between changes in the neurons and clinical manifestations of disease is a challenge. Moreover, how the neurons might be manipulated by therapeutic compounds to ameliorate disorders of the digestive system is elusive, in many cases. The problems of understanding and treating digestive diseases that involve the enteric nervous system, or functions controlled by the enteric nervous system, are touched on throughout the book, and are specifically discussed in Chapter 7. In writing this book I have relied on the assistance and advice of many colleagues who have generously read and commented on parts of book, in some cases through several drafts. My special thanks go to Dr Paul Andrews, Dr Joel Bornstein, Dr Axel Brehmer, who also helped me with the interpretation of some of the older literature published in German, Dr Nadine Clerc, Dr Helen Cox, Dr Roberto de Giorgio, Dr Giorgio Gabella, Dr Peter Holzer, Dr Terumasa Komuro, Dr Alan Lomax, Dr Kulmira Nurgali, Dr Michael Schemann, Dr Keith Sharkey, Dr Henrik Sjövall, Dr Werner Stach, who provided previously unpublished micrographs, Dr Jean-Pierre Timmermans, Dr Marcello Tonini and Dr Heather Young. For assistance in the preparation of the illustrations I am very grateful to Melanie Clarke, Anderson Hind, and Trung Nguyen, and for editorial help and assistance with the references, to Emma James. I would also like to thank the many colleagues who gave permission for illustrations to be included in the book. I hope that this book succeeds in linking the extensive knowledge of the structure and cell physiology of the enteric nervous system to an understanding of digestive physiology, and that in so doing it helps provide a rational basis for therapeutic intervention, and even reasons why some interventions may fail. I enjoyed writing the book, although at times it was a hard task. I hope that in reading the book you encounter only the enjoyment. John B Furness Melbourne, May 2005

Abbreviations

AC, adenylyl cyclase ACh, acetylcholine AChE, acetylcholine esterase ADP, after-depolarizing potential AH, designation of neurons having slow after-hyperpolarizing potentials AHP, after-hyperpolarizing potential AMP, adenosine monophosphate AMPA, alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid AP, action potential ATP, adenosine triphosphate BK, large-conductance potassium channel BMP, bone morphogenic protein BN, bombesin (the mammalian form also referred to as GRP, below) cAMP, cyclic adenosine monophosphate CCK, cholecystokinin CFTR, cystic fibrosis transmembrane conductance regulator CGRP, calcitonin gene-related peptide ChAT, choline acetyltransferase CM, circular muscle CNS, central nervous system DAG, diacyl glycerol DMP, deep muscular plexus DMPP, dimethyl phenyl piperazinium DYN, dynorphin ECL, enterochromaffin-like (cell) EJP, excitatory junction potential ENK, enkephalin EPSP, excitatory post-synaptic potential GABA, γ-aminobutyric acid GAL, galanin gCav, voltage-sensitive calcium conductance gK , Ca2+-dependent K+ conductance Ca gNav, voltage-dependent Na+ conductance GRP, gastrin-releasing peptide (also known as mammalian bombesin) xi

xii

AB B RE V I ATI O N S

Gs, stimulating G-protein 5-HT, 5-hydroxytryptamine (serotonin) HCN, hyperpolarization activated non-specific cation conductance HVA, high-voltage activated calcium current IAHP , AHP current IBS, irritable bowel syndrome ICav, voltage-sensitive calcium current ICC, interstitial cell(s) of Cajal Ih, hyperpolarization-activated cation current IK, intermediate-conductance potassium channel IKATP, ATP-dependent potassium current IPAN, intrinsic primary afferent neuron IPSP, inhibitory post-synaptic potential LM, longitudinal muscle MAP2, microtubule associated protein 2 MELAS, multisystem mitochondriopathy MMC, migrating myoelectric complex MNGIE, mitochondrial neurogastrointestinal encephalomyopathy MP, membrane potential Muc, mucosa L-NAME, L-nitro-arginine methyl ester nAChRs, nicotinic acetylcholine receptors NANC, non-adrenergic, non-cholinergic NFP, neurofilament protein Nic, nicotinic NK, neurokinin NO, nitric oxide NOS, nitric oxide synthase NPY, neuropeptide tyrosine, usually known as neuropeptide Y P2X, purine receptor 2X P2Y, purine receptor 2Y PACAP, pituitary adenylyl cyclase activating peptide PCR, polymerase chain reaction PDBu, phorbol dibutyrate PHI, peptide histidine isoleucine PHM, peptide histidine methionine PKA, protein kinase A PKC, protein kinase C PLC, phospholipase C PPADS, pyridoxal-phosphate-6-azophenyl-2΄,4΄-disulfonic acid PVG, prevertebral ganglion PYY, peptide tyrosine tyrosine Rin, input resistance RT, room temperature SAC, stretch activated channel SGLT1, Na+/glucose co-transporter 1 SK, small-conductance potassium channel

A BBR EV IAT IONS

xiii

SM, submucosa SOM, somatostatin SSPE, sustained slow post-synaptic potential STC, slow-transit constipation TEA, tetraethylammonium TK, tachykinin TRH, thyrotropin-releasing hormone TTX, tetrodotoxin TTX-R INaV, TTX-resistant sodium current VIP, vasoactive intestinal peptide VPAC, vasoactive intestinal peptide; pituitary adenylyl cyclase activating peptide

1: Structure of the enteric nervous system

A vast amount of neural tissue, which constitutes the enteric nervous system, is embedded in the wall of the gastrointestinal tract. Within the enteric nervous system, nerve cells and supporting (glial) cells are grouped in small clusters, the enteric ganglia, which are interconnected by nerve fiber bundles (Fig. 1.1). The individual ganglia are small, but are so numerous that the system as a whole contains millions of nerve cells. The processes of these nerve cells connect with other neurons and innervate the muscle, secretory epithelium, and blood vessels of the digestive tract, biliary system, and pancreas. Processes of nerve cells from outside the digestive tract also connect with enteric neurons, and intermingle with processes of enteric neurons. A remarkable aspect of the enteric nervous system is that its reflex circuits are capable of directing the functions of the digestive system without relying on commands from the brain or spinal cord. This independence is modulated by the rich interchange of signals between the enteric and central nervous systems. The first clear descriptions of ganglionated plexuses within the wall of the digestive tract were those of Meissner (1857), Billroth (1858), and Auerbach (1862a,b, 1864). Remak (1840, 1852) had earlier noted the presence of microscopic ganglia in the walls of the pharynx and stomach, but his descriptions do not suggest that he recognized a ganglionated plexus. Following their discovery, the enteric ganglia and plexuses attracted considerable attention and numerous descriptions of their organization were published, including those of Henle (1871), Drasch (1881), Dogiel (1895b, 1899), Cajal (1911), Kuntz (1913, 1922), Hill (1927), Schabadasch (1930a,b), Stöhr (1930), and Irwin (1931). These studies, and the contemporary literature they cite, provide detailed information on the sizes, arrangements and interconnections of the ganglia. The descriptions that Meissner, Billroth, and Auerbach provided of the general organization of the ganglionated plexuses, based on quite primitive techniques to reveal the nerve tissue, were not superseded by work in the subsequent 100 years and the descriptions of the arrangements of the enteric plexuses that 1

2

CHAPTER 1

are set out in the following pages were essentially established by the time of the reviews of Schabadasch (1930a,b) and Stöhr (1930). An English translation of Auerbach’s 1864 description has been published (Furness & Costa 1987). The enteric nervous system of the tubular digestive tract (the esophagus, stomach, and intestines) is formed of a number of interconnected networks, or plexuses, of neurons, their axons, and enteric glial cells (Fig. 1.1). In the

Fig. 1.1 The enteric plexuses as they are seen (A) in wholemounts and (B) in transverse section.

The drawings depict the small intestine. There are two ganglionated plexuses, the myenteric and the submucosal plexuses, in addition to nerve fibers that innervate the muscle layers, the mucosa and intramural arterioles. Nerve fibers enter the intestine with mesenteric blood vessels in paravascular nerves (B). Adapted from Furness and Costa (1980).

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

3

small intestine and colon, most nerve cells are found in two sets of ganglia, the ganglia of the myenteric (Auerbach’s) plexus and of the submucosal plexus (often referred to as Meissner’s plexus, but see below). The axons of these nerve cells innervate other ganglia and the tissues of the digestive organs, such as the muscle layers and the mucosa. The enteric plexuses Myenteric plexus The myenteric plexus is a network of nerve strands and small ganglia that lie between the outer longitudinal and inner circular muscle layers of the external muscle coat of the intestine (Fig. 1.2). The network is continuous around the circumference and along the gastrointestinal tract (Fig. 1.3). The myenteric ganglia vary in size, shape, and orientation between animal species and from

Fig. 1.2 Drawing of a whole-

mount of the myenteric plexus of the human small intestine, prepared by Auerbach and published in Henle’s Textbook of Histology in 1871. Myenteric ganglia, internodal strands, and small nerve trunks of the secondary component of the myenteric plexus (arrows) can be seen. The secondary nerve strands supply fibers to the circular muscle and deeper layers. Calibration: 1 mm.

4

CH AP TE R 1

Fig. 1.3 Schabadash’s (1930a) depiction of the myenteric plexus of the pyloric canal in the cat. The large dark areas are the ganglia, which are connected by nerve fiber bundles of various calibers. The continuity of the plexus around and along the gut wall can be clearly seen. Calibration: 2 mm.

one part of the intestine to another (Fig. 1.4), but the shape of the meshwork is usually characteristic and readily identified in any major region from a particular species (Irwin 1931, Gabella 1981a). Although the pattern is easily recognized, considerable variation in the size of ganglia is encountered. In the ileum of the guinea-pig, ganglia range in size from 5 to over 200 nerve cell bodies. Single nerve cell bodies are occasionally encountered outside the main meshwork of the plexus, usually adjacent to a nerve strand. The ganglia are sometimes referred to as the nodes of the plexus because they lie at the junctions of nerve strands, which in turn are called internodal or interganglionic strands, or sometimes interganglionic connectives. The ganglia are deformable and are distorted by the movements of the muscle. Thus measurement of such features as their shape and spatial density must take into account the state of contraction of the gut wall (Gabella & Trigg 1984). Three components of the myenteric plexus are described (Fig. 1.5): a primary plexus, a secondary plexus, and a tertiary plexus (Auerbach 1864, Schabadasch 1930a,b, Stöhr 1930, Li 1940, 1952). Together, the ganglia and internodal strands make up the primary meshwork of the myenteric plexus. Many of the nerve fibers in an internodal strand do not enter the ganglion with which a strand connects, but pass over the ganglion, usually between the ganglion and the longitudinal muscle, and continue in another internodal strand. Finer nerve fiber bundles, constituting the secondary component of the plexus, branch from the primary internodal strands or arise from ganglia, but do not usually link adjacent ganglia. The secondary strands run parallel to the circular muscle bundles and often cross internodal strands. They run on the inner aspect of the primary plexus and ganglia, between the primary plexus and the circular muscle (Schabadasch 1930b, Stöhr 1952). Auerbach (1864) traced nerve processes from the secondary strands to the circular muscle, a connection that has been confirmed (Wilson et al. 1987). The secondary

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

Fig. 1.4 Drawings of the myenteric plexus in different regions of the gastrointestinal tract of the guinea-pig: A, esophagus; B, pylorus; C, duodenum; D, ileum; E, colon; F, rectum. Note that patterns of the ganglia differ between regions. They also differ between species. All at the same magnification, except B. The calibration lines are 1 mm apart. Reproduced from Irwin (1931).

5

6

CH AP TE R 1

1 1

2

3 2 3

Fig. 1.5 The three components of the myenteric plexus found in small animals are shown in a drawing of a wholemount from the guinea-pig small intestine. Common to all species is the primary component of the plexus (1), consisting of the ganglia and internodal strands (interganglionic connectives), and the secondary component (2), consisting of nerve strands lying parallel to the circular muscle (across the page). The tertiary plexus (3) is found only where the longitudinal muscle is thin; in such regions, few nerve fibers are found within the longitudinal layer. Neuron cell bodies are depicted as white ovals in the ganglia. Redrawn from Furness and Costa (1987). Calibration: 100 μm.

strands can be seen in Auerbach’s drawing (Fig. 1.2). The tertiary meshwork (tertiary plexus) is made up of fine nerve bundles that meander in the spaces between the meshwork formed by the primary plexus (Richardson 1958, Llewellyn Smith et al. 1993, Furness et al. 2000) (Fig. 1.5). Nerve bundles of the tertiary plexus can be traced from primary internodal strands, ganglia and secondary strands. The definition of the tertiary plexus given here accords with that of Stöhr (1930), which is different from that given by Schabadasch and Li. The definitions of the latter authors combine the secondary and tertiary plexuses under the name secondary plexus and they call the tertiary plexus those fine fibers that run parallel to and extend into the circular muscle and which join the deep muscular plexus. I refer to these fibers as the circular muscle plexus, or simply as the circular muscle innervation. Submucosal plexus A submucosal ganglionated plexus is found in the small and large intestine (Figs 1.6, 1.7), and was first described in the mid-19th century by Meissner (1857) and Billroth (1858). Although scattered ganglia are found in the submucosal layer in the esophagus and stomach, these do not form a ganglionated plexus comparable to that of the intestines. In general, the interconnecting strands of the submucosal plexus are finer and the ganglia are smaller than those of the myenteric plexus (Henle 1871,

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

7

Fig. 1.6 Distribution of enteric ganglia in the tubular digestive tract. The gastrointestinal tract is represented schematically in longitudinal section to reveal the myenteric ganglia, which form a continuous plexus from the upper esophagus to the internal anal sphincter, and the submucosal plexus, which is prominent in the small and large intestines. Isolated ganglia occur in the gastric and esophageal submucosa and in the mucosa throughout the digestive tract. From Furness et al. (1991).

8

CH AP TE R 1

Fig. 1.7 Drawing of the submucosal plexus of the small intestine of a 6-day-old child, published by Billroth (1858). It accurately depicts the ganglia, with nerve cells drawn as small circles, and the connecting strands. Note that Billroth depicts ganglia and nerve strands at two levels. Calibration (approx): 250 μm.

Goniaew 1875, Timmermans et al. 2001). The plexus is continuous around the circumference and along the length of the small and large intestines. The arrangements of ganglia in the submucosal plexus, and the functional types of neurons in these ganglia, differ between species (Scheuermann et al. 1987b,c, Hoyle & Burnstock 1989, Timmermans et al. 1990). In large animals, good examples being the pig and human, submucosal ganglia form distinct, but interconnected, plexuses that lie at different levels, as first clearly described by Schabadasch (1930b). Two or sometimes three layers of ganglia have been distinguished (Schabadasch 1930b, Gunn 1968, Hoyle & Burnstock 1989, Timmermans et al. 2001). Ganglia at different depths contain different populations of neurons, these variations being apparent in the shapes and chemical natures of the constituent nerve cells. The inner ganglionated plexus (closer to the gut lumen) has been likened to the plexus described by Meissner (1857) and the outer has been identified with that described by Henle (1871) and Schabadasch (1930b). Because it is not completely clear who should be credited with the discovery of individual components of the submucosal plexus, it seems sensible to refer to the most obvious groupings as the inner and outer submucosal plexuses (Timmermans et al. 2001), the inner being that closer to the intestinal lumen, and the outer that closest to the circular muscle layer. Among the neurons of the outer plexus are some that supply innervation to the circular and even to the longitudinal muscle (Sanders & Smith 1986, Furness et al. 1990a, Timmermans et al. 1994, 1997,

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

9

Porter et al. 1999). The outer submucosal plexus also supplies innervation to the mucosa. The inner submucosal plexus has few neurons that supply the muscle, but many that innervate the mucosa (Porter et al. 1999, Timmermans et al. 2001). In small mammals, typified by the guinea-pig, there is generally a single layer of submucosal ganglia, and these ganglia contain secretomotor neurons, but not motor neurons that supply the external muscle; in fact, in the guinea-pig there are four main types of neurons in the submucosal ganglia of the small intestine (Furness et al. 1984, 2003a). The ganglia of the submucosal plexus in small mammals most closely resemble those of the inner submucosal plexus of larger species. Paucity of ganglia in the submucosa of the esophagus and stomach Extensive submucosal ganglionated plexuses, such as those found in the small and large intestines, do not occur in the esophagus (Harting 1934, Schofield 1960, Rash & Thomas 1962, Christensen & Rick 1985a, Izumi et al. 2002). Small groups of nerve cell bodies are occasionally found adjacent to submucosal mucus-secreting glands that are scattered along the esophagus, although some investigators have reported that there are no nerve cell bodies at all in the submucosa of the esophagus (Christensen & Rick 1985a). Submucosal ganglia are absent or extremely rare in the stomach of small animals (guinea-pig and rat) and are sparse, but clearly present, in larger mammals, such as dog, human, and cat (Schabadasch 1930a, Kyösola et al. 1975, Stach et al. 1975, Radke et al. 1978, Christensen & Rick 1985a, Furness et al. 1991, Schemann et al. 2001, Colpaert et al. 2002) and denervation and tracing experiments show that the intrinsic innervation of the gastric mucosa is derived almost entirely from the myenteric ganglia (Furness et al. 1991, Pfannkuche et al. 1998). Submucosal nerve cells that do occur in the stomach are more common in the antrum. Some of the myenteric ganglia extend into the clefts (septa) between the large blocks of circular muscle in the stomach and can be mistaken for submucosal ganglia. Ganglia in the mucosa Small groups of nerve cell bodies occur in the lamina propria of the mucosa in the small and large intestine, and, rarely, in the stomach (Drasch 1881, Vau 1932, Stöhr 1934, Ohkubo 1936, Isisawa 1939, 1949, Lassmann 1975, Newson et al. 1979, Fang et al. 1993, Balemba et al. 1998). Stöhr (1934) has suggested that these are displaced (ectopic) submucosal ganglia. These nerve cells are almost always close to the muscularis mucosae, that is, they are close to the inner submucosal plexus.

10

CH AP TE R 1

Subserosal plexus This is a plexus of fine nerve bundles that is found in the connective tissue layer at the surfaces of digestive organs, for example between the serosal lining of the peritoneal cavity and the external muscle of the intestine (Schabadasch 1930b). These nerve bundles connect extrinsic nerves and nerves of the deeper layers of the gut wall, as was recognized and described by Auerbach (1864). Small ganglia sometimes occur in the subserosal plexus, particularly in the esophagus and stomach and near the mesenteric attachment of the intestine and on the surface of the rectum. Some subserosal ganglia lie within or adjacent to the branches of the vagus nerves as they enter the walls of the stomach and esophagus. Longitudinal muscle innervation and the tertiary plexus The longitudinal muscle is innervated by a longitudinal muscle plexus, which consists of fine bundles of nerve fibers that run parallel to and within the muscle, or by the tertiary component of the myenteric plexus, which consists of axons in bundles that lie against the inner surface of the muscle (Richardson 1958). How the muscle is innervated seems to be simply determined by its thickness. In large animals, and in small animals where this muscle layer is thickened, for example in the teniae which occur in the large intestines of some species, a longitudinal muscle plexus is observed. Where the muscle layer is less than about 10 muscle cells thick, it is innervated exclusively by fine nerve bundles of the tertiary component of the myenteric plexus. These bundles are frequently found in small grooves at the inner surface of the muscle (Llewellyn Smith et al. 1993). The tertiary plexus is described in more detail above (see Fig. 1.5). Circular muscle innervation Fine nerve bundles that run parallel to the length of the muscle cells are found throughout the thickness of the circular muscle (Fig. 1.8). These bundles connect with the primary and secondary components of the myenteric plexus and with the deep muscular plexus in the small intestine. The nerve fiber bundles of the circular muscle plexus form a continuous meshwork both around the circumference of the intestine and, through oblique interconnecting nerve strands, along its length. In small mammals, most of the axons within the circular muscle plexus derive from motor neurons with cell bodies in the myenteric ganglia, but there are some fibers that come from nerve cells in the outer ganglia of the submucosal plexus. Fibers that originate from submucosal ganglia are more numerous in larger species (see above and Chapter 2).

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

11

Fig. 1.8 Nerve fibers in the cir-

cular muscle. This micrograph is of a wholemount preparation of the circular muscle of the guinea-pig small intestine, stained with the Champy-Maillet zinc iodide and osmium technique. The level of focus corresponds to the deep muscular plexus. Major nerve fiber bundles run approximately parallel to the long axes of the muscle cells and there are many connections between these bundles. Calibration: 50 μm.

Deep muscular plexus and submuscular plexus An aggregation of nerve fiber bundles is found near the inner part of the circular muscle layer of the small intestine (Li 1937, 1940, Taxi 1965, Gabella 1972b, 1974) (Fig. 1.9) and also of the large intestine (Stach 1972, FaussonePellegrini & Cortesini 1984, Faussone-Pellegrini 1985, Christensen & Rick 1987b). These concentrations of innervation were described by Cajal (1895,

Fig. 1.9 Diagram to illustrate the nerve supply to the mucosa of the small intestine, as seen in histological section. The nerve fibers are in small bundles that form a continuous network in the connective tissue of the mucosa, the lamina propria (lp). The mucosal nerve network can be divided into interconnecting subglandular, periglandular, and villous components. Muscularis mucosae, mm; gland, gl.

12

CH AP TE R 1

1911) who provided two descriptive names, plexus musculaire profond (deep muscular plexus) and plexus sous-musculeux (submuscular plexus) (Cajal 1911). A concentration of fibers near the inner surface of the circular muscle coat is not observed in the canine stomach (Furness et al. 1990a), although there is some degree of concentration of innervation in the human stomach (Faussone-Pellegrini et al. 1989). It is of interest that Cajal provided two names, because there is a subtle difference that these two names accommodate. In the small intestine, the plexus separates a thin layer of muscle cells from the bulk of the circular muscle, and here it has been generally referred to as the deep muscular plexus (Li 1937, Gabella 1974). In some regions, good examples being the dog and pig colon, there is no layer of muscle internal to the plexus, which lies at the extreme inner surface of the circular muscle, adjacent to the submucosa (Stach 1972, Christensen & Rick 1987b, Furness et al. 1990a). In this position, it can be called the submuscular plexus. The nerve bundles of the deep muscular and submuscular plexuses form continuous meshworks around the circumference and along the intestine. Their predominant orientation is parallel to the direction of the circular muscle, with frequent oblique connections between adjacent bundles (Fig. 1.8). The reason why part of the innervation of the circular muscle is concentrated close to its inner surface is not known, but the axons have the same spectrum of neurochemical types as axons in the rest of the circular muscle. It may simply be that the circular muscle is innervated asymmetrically, just as the innervation of arteries is asymmetric (in arteries, the axons are primarily at the outer border of the muscle coat) and the innervation of the longitudinal muscle through the tertiary plexus is asymmetric. Interstitial cells of Cajal (ICC) lie in close proximity to nerve fibers of the deep muscular plexus and they have a critical role as intermediates in transmission between the axons of motor neurons and the smooth muscle cells (Chapters 4, 5). Innervation of the muscularis mucosae The layers of smooth muscle at the surface of the mucosa, adjacent to the submucosa, are known as the muscularis mucosae. In general, this consists of inner bundles of circularly disposed smooth muscle cells and outer longitudinally oriented smooth muscle, but in some places it is thin and has smooth muscle bundles at various orientations. In the esophagus, the muscle bundles are arranged primarily in a longitudinal direction. Fine nerve fiber bundles that run parallel to the long axes of the muscle cells make up the innervation of the muscularis mucosae. In the small intestine, bundles of smooth muscle

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

13

that are considered part of the muscularis mucosae make finger-like intrusions into the cores of the villi and in the stomach similar slivers of muscle are found between gastric glands. These intra-mucosal muscle bundles are also accompanied by nerve fibers. Mucosal innervation The structure of the mucosa varies more from one part of the gastrointestinal tract to another than do the structures of other layers. In the small intestine, it consists of the muscularis mucosae, the connective tissue (lamina propria), into which simple tubular glands protrude (the intestinal crypts, or glands of Lieberkuhn), and the villi. The lining of the glands and the surface of the villi are a single layer columnar epithelium. A dense network of fine interconnecting nerve bundles that is found throughout the connective tissue (lamina propria) of the mucosa (Fig. 1.9) makes up the mucosal innervation and was described in the 19th century (Billroth 1858, Drasch 1881, Müller 1892, Berkley & Baltimore 1893, Cajal 1895, 1911). The mucosal innervation in the small intestine can be divided into different components: a subglandular plexus, a periglandular plexus, a villous subepithelial plexus, and a plexus of the villous core. There is some specificity in the nerve fibers that contribute to the different components. For example, in guinea-pig small intestine, calretinin immunoreactive secretomotor neurons selectively supply the subglandular and periglandular components (Brookes et al. 1991a, Clerc et al. 1998b). In the stomach and colon, nerve fibers are found throughout the depth of the mucosa, and there is also a dense mucosal innervation in the gall bladder. A sparser plexus of nerve fibers occurs adjacent to the mucosal epithelium in the esophagus, which is a protective stratified epithelium that is devoid of secretory elements. The nerve fibers that innervate the mucosa lie in the connective tissue of the lamina propria, they do not penetrate the epithelium, which is a single layer in the stomach, small intestine, and colon. Some nerve fibers, which are believed to be sensory, penetrate the inner layers of the stratified epithelium that lines the esophagus (Rodrigo et al. 1975, Clerc & Condamin 1987). Nerve fibers in the mucosa inevitably come close to entero-endocrine cells of the gastric and intestinal epithelium. It is difficult to define any special relationship with the entero-endocrine cells, even using electron microscopy. Nevertheless, there are functional interactions between nerve fibers of the mucosal plexus and the endocrine cells (Chapter 2). Nerve fibers in the mucosa also come close to cells of the immune system, e.g. lymphocytes, and to the lymph nodules (Peyer’s patches) that occur in the mucosa (Chapter 2).

14

CH AP TE R 1

Vascular innervation Two types of nerve fiber bundles are associated with arteries and arterioles of the alimentary tract (and of other organs). The first of these are paravascular nerve fibers which follow the arteries, using the tract taken by the arteries as a conduit. Paravascular fibers carry axons that are destined to supply other structures, as well as axons that innervate blood vessels. In the alimentary tract these other targets include the enteric ganglia, intestinal smooth muscle, and the mucosa. In addition, the arteries and arterioles are surrounded by a continuous meshwork of fine anastomosing nerve fiber bundles, the perivascular plexus, which contains both motor fibers to the blood vessels and vascular primary afferent (sensory) nerve fibers. Veins of the mesentery also have a perivascular plexus, but very few nerves have a perivascular relation with veins within the gut wall. Nerve fibers associated with intestinal lymphatic vessels are quite rare (Furness 1971). Nerve fibers associated with Brunner’s glands and glands in the esophagus The numerous small mucus-secreting glands in the submucosa of the duodenum, the glands of Brunner, are innervated by fine nerve fibers (Drasch 1881, Stach & Hung 1978, Ferri et al. 1984), as are the mucus-secreting glands in the esophagus. Functional studies suggest that the innervation of Brunner’s glands comes from the vagus, not from neurons of the enteric nervous system (Moore et al. 2000). The observation that vagal nerve stimulation promotes secretion of mucus, presumed to derive from Brunner’s glands, is an old one (Wright et al. 1940), but it is only this recent study (Moore et al. 2000) that suggests that the effect is not via the enteric nervous system. In the esophagus I have observed small ganglia in the submucosa adjacent to the glands, confirming earlier reports (Kadanoff & Spassowa 1959). Interconnections between the plexuses Although the plexuses have been described above as if they are separate entities, they are in fact joined by numerous nerve fiber bundles. Auerbach (1864) observed connections between extrinsic (vagal and mesenteric) nerves and the myenteric plexus via the subserosal plexus, and he also observed connections between the myenteric and submucosal plexuses. Drasch (1881) confirmed the connection of the myenteric plexus and the submucosal plexus and recognized that fibers from the submucosal plexus innervate the mucosa. Various authors detected fibers running from myenteric neurons into the circular muscle (Dogiel 1899, Stöhr 1930). Connections within the myenteric plexus, with one neuron sending a process to end on another myenteric neuron of

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

15

the same or an adjacent ganglion were also described by several investigators (Dogiel 1895b, 1899, Kuntz 1922, Waddell 1929). Kuntz also found similar connections in the submucosal plexus. The fiber bundles that connect the myenteric and submucosal plexuses run almost perpendicular to the circular muscle layer. These have been called “penetrating fiber bundles” or “vertical fibers” (Furness et al. 1990a, Brehmer et al. 1998). Extent of the ganglionated plexuses Ganglia are found in the wall of the pharynx near its junction with the esophagus (Remak 1840, 1852, Shimazaki 1998). There are also small ganglia within the tongue (Remak 1852, Sbarbati 2002). I am hesitant to identify oral and pharyngeal ganglia as part of the enteric nervous system, because there is no evidence that they provide the type of local control that is exerted by the enteric nervous system. On the other hand, there is a well-developed myenteric plexus from the most oral part of the esophagus to the stomach in all species (Sabussow 1913, Abe 1959, Mann & Shorter 1964, Christensen & Robison 1982, Wu et al. 2003b). The myenteric plexus is then present continuously along the digestive tract until the internal anal sphincter (Schofield 1968). The ganglionated submucosal plexus is prominent and continuous from the first part of the duodenum to the level of the internal anal sphincter (Fig. 1.6). The numbers of nerve cells contained in the myenteric and submucosal plexuses have been estimated by counts of nerve cell bodies per unit area of gut surface (Furness & Costa 1987, Gabella 1987). Myenteric nerve cells are numerous throughout the digestive tube, varying in density from about 1000 to about 15 000 cell bodies per square centimeter. The lower values are found in the esophagus and proximal stomach, high values in the distal stomach and large intestine, and intermediate values in the small intestine. In total, the enteric nervous system of an individual mammal contains 2–1000 million nerve cells, depending on the size of the mature animal (Furness & Costa 1987, Gabella 1987, Karaosmanoglu et al. 1996). The total number of nerve cells in the myenteric plexus of the sheep small intestine has been estimated to be 31.5 million, in the guinea-pig, 2.75 million, and in the mouse, 403 000 (Gabella 1987). Total numbers in the submucosa of the small intestine were: sheep, 50 million; guinea-pig, 950 000; and mouse, 330 000. It has been estimated that the number of nerve cells in the enteric ganglia is about the same as the number in the spinal cord (Furness & Costa 1980). Although no direct counts have been made, the total number of enteric neurons in the human gastrointestinal tract, estimated by extrapolating data from other mammals, is probably in the range of about 200–600 million.

16

CH AP TE R 1

Ganglia of the biliary system The biliary tract and the gall bladder develop from a diverticulum of the small intestine, and the ganglia that have been described in the walls of the gall bladder, cystic duct, hepatic duct, and common bile duct (Gerlach 1873, Dogiel 1895a, Harting 1930, Ottaviani 1941, Hermann 1951, Sutherland 1966, 1967, Cai & Gabella 1983, Mawe & Gershon 1989b) are regarded as part of the enteric nervous system. A ganglionated plexus lies external to the muscle coat in the gall bladder. Ganglia are found between the muscle bundles and internal to the muscle in the lamina propria of the mucosa (there is no muscularis mucosae in the gall bladder). In some species, for example the guinea-pig, nerve cells in the lamina propria are very rare (Sutherland 1966, Cai & Gabella 1983). The ganglionated plexus of the cystic duct, hepatic duct, and upper part of the common bile duct is also external to the muscle. In the part of the common bile duct closer to the duodenum, ganglia lie between the longitudinal and circular muscle layers and are also found internal to the musculature. There is continuity between the ganglionated plexuses of the biliary system and those of the duodenum (Cai & Gabella 1983, Mawe & Gershon 1989a, Padbury et al. 1993). Pancreatic ganglia Ganglia in the pancreas have been described by many authors (Cajal 1891, Müller 1892, De Castro 1923, Coupland 1958, Watari 1968). These ganglia are small and are connected to each other by nerve fiber bundles. They are probably best regarded as enteric ganglia because the pancreas develops as an outgrowth of the midgut, and because of their direct connections to duodenal ganglia (Kirchgessner & Gershon 1990). On the other hand, they can be looked upon as parasympathetic ganglia, analogous to the ganglia of the salivary glands, which are also secretory glands of the digestive tract. Ganglia in the trachea and bronchi The lungs develop as outgrowths of the foregut, and, like the digestive system, they have intrinsic ganglia, in this case in the walls of the trachea and bronchi (Fisher 1964). The ganglia form a well-developed plexus and appear to be a principal source of innervation of the airways of the lung. Nevertheless, the innervation of the trachea retains a peculiar relationship with the enteric nervous system into adult life: neurons with the phenotype of inhibitory neurons to gut muscle (immunoreactivity for VIP and NOS) project from the esophagus, where their cell bodies reside, to the trachea (Fischer et al. 1998; Moffatt et al. 1998). These neurons were identified by injecting a retrogradely

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

17

transported dye into the trachea, which was subsequently detected in nerve cells in the esophagus. Intramural extensions of extrinsic nerves Branches of the vagus and pelvic nerves continue to be distinct and recognizable for considerable distances after they enter the walls of the esophagus, stomach, and rectum. The intramural pelvic nerves run along the large intestine, between the longitudinal and circular muscle layers, and can be traced at least as far as the transverse colon (Iljina & Lawrentjew 1932, Fan 1955, Lee 1956a,b, Christensen et al. 1984, Fukai & Fukuda 1984, McRorie et al. 1991). A good description is provided by Stach (1971). Myelinated fibers in the intramural nerves can be shown by degenerative section or tracing to arise from the pelvic nerves (Christensen & Rick 1987a). Branches of the vagus nerve enter the stomach at its junction with the esophagus and fan out to supply prominent intramural vagal branches that ramify at the level of the myenteric plexus (Mitchell 1940, Christensen & Rick 1985b, Wang & Powley 2000). Mesenteric nerve fiber bundles, containing sympathetic efferent axons and afferent nerve fibers, follow the small arteries into the gut wall and branch with them. They soon become very fine and cannot be recognized as distinct entities within the enteric plexuses. Electron microscope studies Ultrastructure analysis shows that enteric ganglia are remarkably compact, consisting of the cell bodies of neurons, supporting cells (enteric glia), and nerve cell processes (Richardson 1958, Hager & Tafuri 1959, Taxi 1959, 1965, Ono 1967, Baumgarten et al. 1970, Gabella 1972a, Cook & Burnstock 1976a, Yamamoto 1977, Wilson et al. 1981, Komuro et al. 1982b). A lowpower view typically shows nerve cell bodies and areas consisting of numerous axons with thin processes of enteric glial cells (Fig. 1.10). The enteric ganglia, unlike other autonomic ganglia, do not contain blood vessels, connective tissue cells or collagen fibrils. The absence of connective tissue and the close packing of neurons and glia give enteric ganglia an appearance similar to the central nervous system. The ganglia are not encapsulated, but lie in the connective tissue framework between the muscle layers or in the submucosa. Within the ganglia, individual nerve cells are only partly surrounded by glial cells. The surface elements of the ganglia have a thin layer of basal lamina and beyond this are fibroblasts and collagen fibrils which, although in parts aligned with the surface, form a discontinuous covering (Fig. 1.10). Thus the enteric ganglia receive their nutrient by diffusion from the blood vessels

18

CH AP TE R 1

Fig. 1.10 The ultrastructure of enteric ganglia. A: At a low power, enteric nerve cells (N = neuronal nucleus) can be seen to present bare surfaces to the extracellular space (arrows). Glial cells are found in the ganglia, but only partly surround the nerve cells and axon bundles (G = glial cell nucleus). From Komoro et al. (1982). B: An area of neuropil adjacent to a nerve cell (N = neuronal nucleus) and a glial cell (G = glial cell nucleus) in a myenteric ganglion. The neuropil contains many profiles of axons (and perhaps dendrites), some of which contain transmitter vesicles, and some of which contain neurofilamants and neurotubules. Arrow points to a synapse. All axons are unmyelinated. From Gabella (1981).

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

19

through the connective tissue that surrounds them. Tracers, such as peroxidase, injected into the blood stream diffuse to the ganglia and penetrate the clefts between their neuronal and glial elements (Jacobs 1977). Thus there is no effective blood–ganglion barrier, and drugs that do not enter the central nervous system have effects at enteric ganglia. It is common for enteric neurons to have large parts of their surfaces bare at the ganglion surface (Gabella 1972a, Komuro et al. 1982b). Sections through myenteric nerve cell bodies reveal profiles that vary considerably in size and in fine structural detail. Cook and Burnstock (1976a) distinguished eight types of neuron within the myenteric ganglia of the guinea-pig small intestine. However, how these relate to functional types is unknown. The only neurons in which ultrastructure and cell type have been clearly related are Dogiel type II neurons, which contain large numbers of mitochondria (Chapter 2). However, neurons in submucosal ganglia of the guinea-pig small intestine do not vary much in their ultrastructural features and, in conventional electron microscopy, appear to belong to one class (Wilson et al. 1981), although of course this is not true in a functional sense. It seems that all nerve cells in the myenteric and submucosal plexuses receive ultrastructurally recognizable synaptic inputs (Cook & Burnstock 1976a, Wilson et al. 1981, Komuro et al. 1982a, Pompolo & Furness 1988). The synapses typically show pre- and post-synaptic densities and presynaptic accumulations of transmitter storage vesicles (Fig. 1.11), similar to synapses elsewhere. There are also frequent close appositions, without pre- and post-synaptic densities,

Fig. 1.11 Enteric axon varicosities containing transmitter vesicles. A: An axon (Ax) forming a synapse on a myenteric nerve cell (N). There are both pre- and post-synaptic densities at the synapse (arrows). B: An axon (Ax) at the surface of a nerve fiber bundle within the muscle coat. The axon contains transmitter vesicles and is bare at the surface that faces the muscle (m). Axons within the nerve bundle are only partly separated by the processes of glial cells (G). Both micrographs from Gabella (1972).

20

CH AP TE R 1

between synaptic vesicle-containing axonal profiles and nerve cells. Bare surfaces of axons are encountered where nerve bundles come close to other effectors, such as smooth muscle cells, the interstitial cells of Cajal, or the mucosal epithelium (Fig. 1.11) (Gabella 1972a). In general, presynaptic specialization is not recognized at these close approaches, although rare instances of presynaptic densities have been described (Klemm 1995). Enteric glia The enteric ganglion cells and nerve fiber tracts are supported by numerous glial cells. Nucleated satellite cells (glial cells) around nerve cell bodies of the enteric ganglia were described by Dogiel (1899), and many subsequent light microscopists pointed out these cells, both in the ganglia and nerve strands of the gastrointestinal plexuses, and in general referred to them as Schwann cells (see Stöhr 1952). More recent studies, which have used immunohistochemical markers to locate enteric glia, reveal that they are common in the ganglia and nerve fiber bundles. Enteric glia express glial fibrillary acidic protein (Jessen & Mirsky 1980, 1983) and the S-100 Ca2+ binding protein (Ferri et al. 1982b), both of which are typical of central nervous system astrocytes. Glial cells in other autonomic ganglia do not contain glial fibrillary acidic protein (Jessen & Mirsky 1983). Electron microscope studies also show that enteric neuronal satellite cells resemble glial cells or astrocytes of the central nervous system more than Schwann cells of other peripheral ganglia or nerve trunks (Gabella 1971, Gabella 1972a, Cook & Burnstock 1976b, Gabella 1981b, Komuro et al. 1982b). The glial cells partly surround nerve cell bodies and axons in the ganglia, leaving bare large areas of neuronal membrane at the surfaces of ganglia. There is a marked contrast in relationships of glial cells to axons in small mammals (for example, guinea-pig or rat) and large mammals (such as cat or human). In small mammals, glial cell processes fail to penetrate all the interstices between nerve cell bodies and between axons in the neuropil (Gabella 1972a, Komuro et al. 1982b). In fact, many nerve processes are in direct membrane-to-membrane contact with each other; the glial cells only separate them into groups and rarely form a sheath around an individual axon. In contrast, in enteric ganglia of human and monkey, axons are separated from one another by intervening glial cell processes (Baumgarten et al. 1970). It is noteworthy that Auerbach (1864) had recognized the differing relations of neurites and supporting cells between humans and some other mammals. Glial cells in nerve strands of the myenteric plexus of small mammals give rise to radiating lamellae which divide the axons into large bundles, and up to 600 neurites may be associated with one glial cell (Gabella 1981b). The ratio of numbers of glial cells to nerve cells increases with species size, in myenteric

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

21

ganglia from 1.1 in mice to 4.5 in sheep and in submucosal ganglia from 0.6 to 1.5 (Gabella & Trigg 1984). Glial cell nuclei can readily be distinguished from those of neurons (Fig. 1.10). The glial nuclei have conspicuous clumps of chromatin, particularly adjacent to the nuclear envelope, and the nuclear surface often displays deep invaginations. The most prominent distinguishing feature of the majority of enteric glial cells is the numerous 10 nm gliofilaments. The cytoplasm also contains smooth and rough endoplasmic reticulum, numerous free ribosomes, mitochondria, lysosomes, and microtubules. Groups of gliofilaments criss-cross the cell bodies and run parallel to the long axes of the cell processes. The gliofilaments appear to be anchored to dense aggregations of material adjacent to the surface membrane. There are some points of apposition between glial cells and between glial cells and neurons where there are small areas of cytoplasmic density inside each plasma membrane. Gap junctions between enteric glia are not often seen, but dye filling indicates that they are coupled to each other (Hanani et al. 1989). The structural similarities and functional differences between regions may have an evolutionary basis A nervous system embedded in or surrounding the gut tube is found throughout the animal kingdom, and is seen for example in simple animals such as Hydra (Hansen et al. 2002, Shimizu et al. 2004). It is thought that the ganglia that form the primitive brains of helminths, and eventually the brains of higher animals, as well as chains of ganglia that provide body segments, were derived from the nervous system around the gut tube (Bullock & Horridge 1965). However, at the same time that primitive brains in the form of cephalic ganglia evolve, there is development of a recognizable enteric nervous system (Fig. 1.12), for example in insects (Ganfornina et al. 1996), leeches (Ábrahám & Minker 1958), octopus (Alexandrowicz 1928), and snails (Campbell & Burnstock 1968, Röszer et al. 2004). Perhaps the most primitive enteric nervous system is that of the marine polyp Hydra (Shimizu et al. 2004). In this species the nodes of the plexus contain only single nerve cells. Nevertheless, when the enteric nervous system in Hydra is ablated, peristalsis, mixing movements and defecation are severely compromised (Shimizu et al. 2004). Although there is a primitive gut tube with little regional specialization in typical invertebrates, the mammalian gut exhibits distinct regional specializations. In turn, the enteric nervous system has evolved to control the functions of the different regions of the mammalian digestive system. So, although the general arrangement of the myenteric ganglia looks similar in all regions of the tubular digestive tract, reflecting a common origin, its functions show important specializations. In the striated muscle part of the esophagus, the

22

CH AP TE R 1

A

Fig. 1.12 The enteric nervous system in the octopus. A: A ganglion of the superficial plexus. This plexus occurs in the connective tissue at the surface of the intestine, a similar location to the subserous plexus in mammals. B: Cross-sectional drawing of the enteric plexuses of cephalopods. Ganglionated plexuses are found at the serosal surface of the intestine and in the external muscle. Cm: circular muscle; Imp: intramuscular plexus; Lm: longitudinal muscle; lp: lamina propria; Epi: epithelium lining the gut tube. From Alexandrowicz (1928).

enteric nervous system is extensive, but it has a limited control over organ function compared to the enteric nervous system of the small intestine. In fact, the motor functions of the striated muscle esophagus are largely controlled through motor pattern generators in the brainstem (Chapters 3, 5), although enteric neurons may have modulatory roles (Chapter 2). By contrast, the enteric nervous system of the small intestine has substantial roles in the intrinsic control of its movements, fluid exchange with the lumen, and local blood flow. It is as if the enteric nervous system of the striated muscle part of the esophagus has retained a structural complexity that is excessive to the needs of the organ. In the stomach, the enteric nervous system has evolved to influence gastric acid secretion and movement, while considerable influence

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

23

over these functions is also exerted through the vagus. In fact, the sophisticated intrinsic reflex control of organ function that occurs in the small and large intestines is less well developed in the stomach (see Chapter 5). Development of the enteric nervous system The enteric nerve cells are not generated in the gut, they arise from precursor cells that migrate from the neural crest. One of the first investigations of the origin of enteric ganglia was that of Kuntz (1910) who examined the precursor and developing ganglia of the sympathetic chains, sympathetic pre-vertebral ganglia, cardiac plexuses, and the gastrointestinal tract in pig embryos of different ages. He concluded that the enteric neuron precursors reach the developing gut by migration along the vagus. This conclusion is essentially correct (Gershon et al. 1993, Young & Newgreen 2001, Gershon 2002). Experimental confirmation of an origin from the vagal neural crest was obtained by Jones (1942) and Yntema and Hammond (1954) who removed parts of the developing neural crest from early chick embryos. Yntema and Hammond (1954) found that removal of rostral regions, but not removal of caudal regions, prevented the development of enteric ganglia throughout the gut, and concluded that vagal level neural crest cells were the source of enteric ganglia. There was initial disagreement about a sacral source in birds: one investigator reported that removal of sacral level neural crest diminished the formation of ganglia in the hindgut (Jones 1942), and it was soon after argued that this observation was mistaken (Yntema & Hammond 1954). Ablation of part of the developing chick is a dramatic intervention that might interrupt signaling in the embryo even if the enteric neurons had a non-vagal origin. An elegant solution was devised by Le Douarin who transplanted quail neural crest into chickens from which the equivalent region was removed (Le Douarin & Teillet 1973, 1974). Quail cells have a nuclear marker that is absent in chick, and these experiments showed that vagal neural precursors migrate to the gut, to form enteric neurons and glial cells throughout its length, and that the precursors from sacral sources contribute to neural populations in the hindgut. In mammals there is also a small contribution of sacral crest cells to the colonization of the rectum (Kapur 2000). In both species, the entry of cells of sacral origin does not occur until vagal cells have arrived. Moreover, if the vagal source is eliminated, the progeny of sacral cells that enter the gut of the chick remain in the hindgut, the rest of the gastrointestinal tract being uncolonized (Burns et al. 2000). In mammals, most recent work has been conducted in mouse, using excellent markers of the early partly differentiated neurons. A wave of migrating cells flows down the gut, from the foregut to the end of the hindgut (Fig. 1.13) in the period between embryonic days 9.5 and 14.5 (Kapur et al. 1992, Young et al. 1998). As the wave advances, maturing

24

CH AP TE R 1

Fig. 1.13 The developing enteric nervous system. The intestine of an embryonic day 10 mouse is shown, with neurons revealed by staining for the transcription factor Phox2b. At this stage, myenteric neurons extend throughout the foregut and the furthest migrated nerve cell is in the midgut (arrow), but no cells have yet appeared in the hindgut. From Young et al. (1998).

cells are halted in their progress and subsequently form ganglia. Thus behind the wave front there is a maturing enteric nervous system. Substantial advances have been made in identifying the signaling mechanism involved in determining the migration and maturation of enteric neurons and glial cells, primarily through the investigation of transgenic animals, but also through the investigation of the genetic basis of disorders of enteric nervous system development (Gershon 2002, Newgreen & Young 2002a,b, Young et al. 2004). One of the first clues came from genetic analysis of Hirschsprung’s disease, a disorder in which enteric nerve cells fail to colonize the gut completely. Most commonly, enteric ganglia are missing from the rectum and sigmoid colon, but the absence of enteric ganglia is sometimes more extensive (Newgreen & Young 2002a, Swenson 2002). About half of the cases are attributable to abnormalities of a cell recognition and signaling complex that consists of the membrane inserted receptor-tyrosine kinase, Ret, and its partner molecules, glial-derived neurotrophic factor (GDNF) and GFR 1 (Gershon 2002, Newgreen & Young 2002a). Nerve cells only occur in the first part of the foregut (that will become the esophagus) of mice in which the c-ret gene that encodes Ret or the gene for GDNF is knocked out (Schuchardt et al. 1994, Moore et al. 1996, Pichel et al. 1996). The reciprocal defect, a failure of the development of esophageal ganglia, occurs when the transcription factor Mash-1 is inactivated (Guillemot & Lo 1993). Mash-1 knockout also prevents the development of a subset of neurons in the gastrointestinal tract beyond the esophagus, so although this region is colonized, it is not normal (Gershon 2002). Another regulatory system that controls the development of the enteric nervous system is the endothelins (ETs) which bind to the cell surface receptors

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

25

ETA and ETB (Newgreen & Young 2002a). In the enteric nervous system, the important ET form appears to be ET3, and the receptor is ETB. Current evidence is consistent with the theory that ET3 is produced by the developing mesenchyme, and that the ETB receptor is on enteric neurons (Kapur et al. 2004). Mutations in the ET and ET receptor genes cause about 5% of Hirschsprung’s disease and distal bowel aganglionosis in the piebald and lethal spotting mouse and the spotting lethal rat. Thus defects in endothelin or its receptor results in incomplete colonization of the bowel. It has been suggested that the defects cause premature differentiation of the precursors (Hearn et al. 1998, Wu et al. 1999), thus limiting the availability of proliferating and migrating neuron and glial cell precursors to colonize more distal gut. However, the complexity of the interactions of neurons and their environment during the development of the enteric nervous system suggests a number of other possible mechanisms (Gershon 2002). Transcription factors that initiate enteric neuron differentiation have been identified. If the genes for either Phox2b or Sox10 are knocked out, no neurons enter the gut (Newgreen & Young 2002a). Other genes that influence the development of the enteric nervous system have been identified, and, although their roles are not yet accurately defined (Gershon 2002, Newgreen & Young 2002a, Young et al. 2004), it is likely that they soon will be in this rapidly advancing field. The precursors of enteric neurons must supply the 15 to 20 individual types that are present in the enteric nervous system. Therefore signaling mechanisms must exist that influence the final differentiation, the density of neurons, the connections that they make, the placement of myenteric neurons between the external muscle layers, and the positioning of other neurons in the submucosa. The molecules that have so far been identified in directing the proliferation, migration, and differentiation are those whose deficiency or manipulation cause substantial changes in enteric neuron number or migration; the control of the final differentiation, positioning, and connectivity of enteric neurons has yet to be unravelled. Development of enteric neurons of the submucosal plexus, pancreas, and gall bladder The biliary system and pancreas develop as outgrowths from the gut, and the similarities in the structure of the neurons within these organs would suggest that they derive from enteric neural precursor cells. In mice, the submucosal plexus is colonized 2–3 days after the myenteric plexus at the same level of the intestine, and might therefore be assumed to derive from the myenteric plexus. In human, the delay in formation of the submucosal plexus is greater, lagging behind the myenteric plexus by 2–3 weeks (Wallace & Burns 2005). These assumptions have been tested experimentally, leading to the

26

CH AP TE R 1

observation that myenteric neural precursor cells migrate from the myenteric plexus to the submucosa and pancreas under the influence of netrins and the netrin receptor, DCC (Jiang et al. 2003). Maturation of enteric neurons and development of function Development of function in the enteric nervous system begins in fetal life and continues for some time post-partum. This period varies considerably between species, amongst other things depending on the development program of the species, and the maturity at birth. In human, there is informative data from studies on aborted fetuses, fetuses in utero and from investigation of pre-term infants (Milla 1993). Neural crest cells enter the human foregut at about week 4, are observed in the developing stomach at 6 weeks, and the myenteric plexus is identifiable in the small intestine by about 8–9 weeks, the submucosal plexus is present 2–3 weeks later, and by around 12–14 weeks the gut has developed its adult form and pattern of enteric ganglia (Milla 1993, Wallace & Burns 2005). Cyclic regular electrical activity and propulsive activity occurs in the small intestine in human, sheep, and dog at about 70–80% of the full term of gestation (Buéno & Ruckebusch 1979, Milla 1993), which suggests that the enteric nervous system in the small intestine is functional at this time. In the lamb, mature migrating myoelectric complexes (MMC; Chapter 5) were seen at 10 days before birth, whereas in the dog they commenced 15 days postnatal. For premature human infants born at 34 weeks or less, there is a lack of co-ordinated activity of swallowing and esophageal motility, but within about a further week the infant sucks, and the motility of the stomach and small intestine shows features of the mature pattern, including the presence of the MMC and postprandial patterns of movement that are qualitatively similar to the adult (Berseth 1989, Milla 1993). The MMC has a shorter cycle time in the fetus and for the first few weeks after birth. In mice, which are considerably less mature at birth, movement of fluid in an anal direction is independent of enteric neurons to the end of gestation, but if the neurons are lacking the mice die soon after they are born (Anderson et al. 2004). Effective neuromuscular transmission from excitatory and inhibitory motor neurons to the muscle of the small intestine is seen for rabbit at E17 (embryonic day 17; gestation 29–35 days, usually 31 days) and in the mouse at E17 (Gershon & Thompson 1973). Consistent with the maturation of the function of the enteric nervous system in utero, evidence of neuronal differentiation, such as the expression of neurotransmitter systems, occurs early. In mice at embryonic day 10 (E10), enzymes that reduce NADPH were seen in neurons adjacent to the stomach (Branchek & Gershon 1989). The NADPH reductase activity is probably due

STR U C TU R E OF T HE ENT ER IC NERV OU S S Y S T EM

27

to the presence of nitric oxide synthase (NOS), the synthesizing enzyme for the enteric transmitter nitric oxide (NO). At E12, the enteric neuropeptide NPY was observed in enteric neurons. Other neural markers that appear during fetal life include choline acetyltransferase, 5-HT, CGRP, enkephalin and VIP (Pham et al. 1991). The early appearance of NADPH reductase/NOS is consistent with the observation that some neurons immediately behind the advancing front of migrating neurons express NOS (Young et al. 2002). By electron microscopy, identifiable synapses between enteric neurons are first seen in the small intestine at E38 (term is 55–60 days) in the guinea-pig (Gershon et al. 1981), indicating that nerve circuits were beginning to be assembled 60–70% through gestation. In mouse intestine (gestation 21 days), synapses with a mature appearance were first detected at E16.5 (Vannucchi & Faussone-Pellegrini 2000), although primitive synapse-like structures were transiently present between 12.5 and 16.5 days. Structural studies confirm the functional data that maturation of the enteric nervous system continues beyond birth (Matini et al. 1997). Changes in enteric neurons with aging Considerable loss of enteric neurons may occur in animals and human towards the ends of their lifespan. Between 6 and 24 months of age in the rat, the numbers of myenteric neurons stained by the NADH diaphorase method declined by 40% in the jejunum and 63% in the colon (Santer & Baker 1988). This method is general, and was thought at the time to stain all neurons. It is a histochemical method that detects NAD:NADH oxidoreductase of the respiratory chain and should detect metabolically active neurons. The guinea-pig small intestine harbors 2.75 million NADH diaphorase positive neurons in the young adult (3–4 months old), which falls to 1.1 to 1.6 million in guinea-pigs aged 26–30 months (Gabella 1989). Moreover, the nerve cell bodies in guinea-pigs became smaller and their surfaces were more angular. In contrast, in rats maintained on a restricted diet, no decline in myenteric neuron numbers was observed between 4 and 24 months (Johnson et al. 1998b). Given these contrasting results, a comparison of rats on a free-feeding diet and a food-restricted diet were compared (Cowen et al. 2000). Rats on an unrestricted diet ate an average of 47 g of rat food per day and lost about 50% of myenteric neurons, whereas rats restricted to 25 g per day had the same numbers of neurons at 4 and 24 months. The loss was selective for cholinergic neurons (neurons that stained for choline acetyltransferase) over NOS neurons. There is also evidence for enteric neuron loss with age in human (Wade & Cowen 2004), although it is not known whether this is diet related.

28

CH AP TE R 1

Summary and conclusions The enteric nervous system is composed of thousands of small ganglia (most being contained in the myenteric and submucosal nerve plexuses), the nerve fibers that connect the ganglia, and nerve fibers that supply the muscle of the gut wall, the mucosal epithelium, arterioles, and other effector tissues. The ganglia contain nerve and glial cells, but not connective tissue elements, and in many respects are similar in structure to the central nervous system. Nerve fiber bundles within the enteric nervous system consist of the axons of enteric neurons, axons of extrinsic neurons that project to the gut wall, and glial cells. The myenteric plexus forms a continuous network, extending from the upper esophagus to the internal anal sphincter. The ganglionated submucosal plexus is present in the small and large intestines, but is absent from the esophagus and contains only very few ganglia in the stomach. Large numbers of neurons are contained in the enteric nervous system, about 200–600 million in human. This is far more neurons than occurs in any other peripheral organ and is similar to the number of neurons in the spinal cord. Neuronal plexuses occur around the gut tube from the simplest animals, such as Hydra, to the most advanced mammals. In mammals, the general structures of the enteric ganglia are comparable throughout the gut, although their functional roles have diverged. The enteric nervous system originates from neural crest cells that colonize the gut during intra-uterine life. It becomes functional in the last third of gestation in human, and continues to develop following birth. There is some loss of enteric neurons in mammals of advanced age.

2: Constituent neurons of the enteric nervous system

Enteric neurons have been classified by their shapes, their physiological properties, specific histochemical, and immunohistochemical staining, and other distinguishing features, including the structures that they innervate, the transmitters they utilize, and the connections that they receive. One of the major advances of the last 15–20 years has been success in correlating most of these features, so that characterizing profiles of all functionally identified neurons in the guinea-pig ileum have now been established (Table 2.1; Fig. 2.1). Enteric neurons can be grouped by their functions as intrinsic primary afferent neurons, interneurons, and motor neurons. Intrinsic primary afferent neurons (IPANs) are sometimes referred to as “intrinsic sensory” neurons, a terminology that is discussed below. Although extensive data on each neuron type has only been obtained for the guinea-pig, the enteric nervous system performs the same functions in all mammalian species, and therefore all species have essentially the same functionally defined neurons. For example, all mammals have excitatory and inhibitory motor neurons that innervate the muscle of the gastrointestinal tract. There are sufficient data to identify in other species the equivalent neurons (orthologs) that have been identified by function in the guinea-pig. One clue is shape. It is generally true that neurons that serve the same functions in different species have the same shape (Peters et al. 1991), a principle that appears to apply to the enteric nervous system (Brehmer et al. 1999a, Furness et al. 2004b). Examples of neurons that have the same shapes and serve the same functions in different species are motoneurons in the ventral horns of the spinal cord, cerebellar Purkinje cells, pyramidal cells in the cerebral cortex, retinal ganglion cells, and dorsal root ganglion cells. It is thus justified to use morphology as a first indicator of the equivalence of neurons in different species. Other clues are chemical properties related to the transmitters that neurons utilize, and projections to targets. In relation to shape, it should be noted that there is a tendency for equivalent neurons to have more elaborate dendritic trees in larger animals, and less elaborate dendrites in smaller mammals (Purves & Lichtman 1985, Purves et al. 1986, Tabatabai et al. 1986). 29

30

CH AP TE R 2

Fig. 2.1 A: The types of neurons in the small intestine of the guinea-pig, all of which have been defined by their functions, cell body morphologies, chemistries, key transmitters, and projections to targets. The numbers adjacent to the neurons correspond to the numbers in Table 2.1, which lists each of the neuron types by their functions and provides data on their chemistries and the percentages of their cell bodies in the myenteric or submucosal ganglia. LM: longitudinal muscle; MP: myenteric plexus; CM: circular muscle; SM: submucosal plexus; Muc: mucosa. Note that the muscularis mucosae (MM) is very thin in the guinea-pig small intestine and the neurons that innervate this muscle have not been identified in this species. In other species they are in the submucosal ganglia. B: Types of neurons in the small intestine of the pig. In many cases the equivalent neurons have been identified in the pig and guinea-pig, and they have been given the same numbers. A, adapted from Furness et al. (2004a); B, modified from Timmermans et al. (2001).

C O N STITU EN T N EU R O N S OF T HE ENT ER IC NERV OU S S Y S T EM

31

For example, in the enteric nervous system, Dogiel type I neurons (Fig. 2.1) have considerably less elaborate dendrites in mice (Nurgali et al. 2004) than in larger mammals, such as guinea-pigs, pigs, dogs, and humans (Dogiel 1899, Stach 1980, Furness & Costa 1987). In general, enteric neurons are smaller, simpler in form, and easier to classify, in small mammals (e.g. in guinea-pigs) and become more elaborate, and more difficult to classify, in larger species of animal, including humans (Brehmer et al. 2004a). Table 2.1 includes data on the chemical coding of the different classes of enteric neurons in the guinea-pig small intestine. Most of these data were gathered over a period of around 20 years, from about 1975 to 1995, and have been summarized and reviewed in a number of papers (Furness et al. 1984, 1995b, Costa et al. 1996, 2000). Subsequently, CART (cocaine and amphetamine-regulated transcript peptide) has been identified as a constituent of a minority of myenteric neurons, and of the VIP-containing secretomotor/vasodilator neurons (Ellis & Mawe 2003). These neurons also contain corticotrophin-releasing factor (CRF) (Liu et al. 2005). Another peptide that, like CART, is involved in regulation of feeding, orexin, is found in enteric neurons (Kirchgessner & Liu 1999a). Orexin occurred in IPANs in the myenteric and submucosal plexusus, and in about 25% of VIP-immunoreactive submucosal neurons. Two further chemical markers for IPANs have also been identified recently, these being the neuronal nuclear protein, NeuN, which occurs in both the nucleus and cytoplasm of IPANs (Chiocchetti et al. 2003) and binding sites for isolectin B4 from Bandiera simplificifolia (Hind et al. 2005), which binds specifically to α-D-galactose end-groups of glycoproteins (Laitinen 1987). Both cytoplasmic NeuN and IB4 binding appear to be specific to IPANs, and to label the entire population of these neurons in the guinea-pig ileum. Shapes of enteric neurons The first and most enduring classification of enteric neurons by their shapes was that of Dogiel (Figs 2.2, 2.3, 2.7). He provided a comprehensive description of neuron morphologies in the myenteric and submucosal plexuses of the intestine from human, guinea-pig, rabbit, rat, dog, and cat (Dogiel 1895b, 1899). The majority of his illustrations are of guinea-pig and human samples. He described three neuron types, now generally referred to as Dogiel types I, II, and III. Between the publication of his first paper and the more comprehensive publication of 1899, La Villa (1898) published observations that revealed cell types in the rabbit intestine (Fig. 2.4) that corresponded closely to those described and depicted by Dogiel.