Airway Chemoreceptors in Vertebrates

Airway Chemoreceptors in the Vertebrates Structure, Evolution and Function

Cover Illustrations Left-hand side figure Immunocytochemical double staining for acetylcholinesterase (AchE) and neuronal nitric oxide synthase (nNOS) of two neuroepithelial cells (NECs) in the lung of the ray-finned fish, bichir Polypterus bichir bichir. Nitrergic nerve terminals (arrowed) are seen running between these cells (green). Micrograph courtesy of G. Zaccone. Right-hand side figure Immunocytochemical triple staining for vesicular glutamate transporter 2 (VGLUT2; red fluorescence), calbindin D28k (CB; blue fluorescence) and myelin basic protein (MBP; green fluorescence) of a neuroepithelial body (NEB) in a rat intrapulmonary airway. The CB-immunoreactive (ir) NEB is contacted by a CB- and VGLUT2-ir vagal nodose sensory nerve fiber, which is wrapped in an MBP-ir myelin sheath that ends in the immediate neighborhood of the NEB. VGLUT2 expression is seen in extensively branching nerve terminals between the NEB cells. The image shows a combination of the three color channels of a maximum value projection of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). Also see Brouns et al., Chapter 11, Figure 2. Reproduced by kind permission of Dirk Adriaensen.

Airway Chemoreceptors in the Vertebrates Structure, Evolution and Function

Editors

Giacomo Zaccone

Deparment of Animal Biology and Marine Ecology Messina University, Messina, Italy

Ernest Cutz

Department of Paediatric Laboratory Medicine University of Toronto, Toronto, Canada

Dirk Adriaensen

Department of Veterinary Sciences University of Antwerp, Antwerp, Belgium

Colin A. Nurse

Department of Biology Mc Master University, Hamilton, Canada

Angela Mauceri

Department of Animal Biology and Marine Ecology Messina University, Messina, Italy

Science Publishers

Enfield (NH)   Jersey   Plymouth

iv  |  Airway Chemoreceptors in the Vertebrates

Science Publishers

www.scipub.net

234 May Street Post Office Box 699 Enfield, New Hampshire 03748 United States of America General enquiries : [email protected] Editorial enquiries : [email protected] Sales enquiries : [email protected] Published by Science Publishers, Enfield, NH, USA An imprint of Edenbridge Ltd., British Channel Islands Printed in India © 2009 reserved ISBN 978-1-57808-614-6 (hardcover) Library of Congress Cataloging-in-Publication Data

Airway chemoreceptors in the vertebrates : structure, evolution and function / editors, Giacomo Zaccone ... [et al.].    p. cm.   Includes bibliographical references and index.   ISBN 978-1-57808-614-6 (hardcover)   1.  Airway (Medicine). 2. Chemoreceptors I. Zaccone, Giacomo. QP123.A34 2008

612.2--dc22

2009006218

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 or otherwise, without the prior permission of the publishers, in writing. The exception to this is when a reasonable part of the text is quoted for purpose of book review, abstracting etc. This book is sold subject to the condition that it shall not, by way of trade or otherwise be lent, re-sold, hired out, or otherwise circulated without the publisher’s prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser.

“What is the Genius if not that productive force generating things that are worthy showing themselves in the presence of God and the Nature, and that therefore have followed in the time? All the Mozart’s works are like this. There is a creative force that continues to act from generation to a generation, and that never it would have to be get exhausted” — Wolfgang Goethe to Eckermann

Preface It is both an honor and a privilege for me to write this Preface to “Structure, evolution and function of the airway chemoreceptors in the vertebrates”. The full credit for realizing this book is due to Professor Zaccone who had initiated this undertaking and solicited contributions from experts from diverse fields of vertebrate biology with shared interest in the investigation of the structure, function and evolution of airway chemoreceptors. The end result is a one of a kind publication reflecting a truly global effort from five continents encompassing expertise in marine biology, zoology, animal physiology, cellular and molecular biology as well as pathobiology. The book provides a comprehensive up to date account of the information available on the morphological, physiological and evolutionary aspects of specialized cells distributed within the epithelia lining the air conducting structures of air breathing species or in the gills of fishes. Where available, recent advances in cellular and molecular aspects of oxygen and carbon dioxide sensing, relevant to the function of the chemoreceptor cells are highlighted. These cells, by virtue of their role as sensors that detect and signal changes in the external and internal environments, likely played an important role in the survival of various species. Consequently, these cells were conserved throughout evolution as evidenced by their occurrence in ancient fish to higher vertebrates including humans. These cells are commonly referred to as neuroendocrine/neurosecretory (NE) or pulmonary neuroendocrine cells (PNECs) in mammals, owing to the expression of a variety of neural and endocrine cell markers as well as prominent innervation that carries the signals to the central nervous system (CNS). While in primitive species solitary NE cells predominate, innervated clusters of NE cells, referred to as neuroepithelial bodies (NEBs), are observed only in the lungs of amphibia and higher vertebrates, perhaps reflecting adaptation from an aqueous to a terrestrial environment.

vi  |  Airway Chemoreceptors in the Vertebrates The book includes 16 contributions divided into seven chapters based on the evolution from primitive to more complex species. Accordingly, the first chapter covers NE cells in the airways and carotid labyrinth of aquatic vertebrates. Recent studies on oxygen-sensitive NE cells in the gills of fish and larval amphibians are reviewed. Of interest are electrophysiological studies using the patch-clamp technique demonstrating remarkable similarities with the O2 sensing mechanism in mammals that depends on reversible inhibition of an O2-sensitive K+ current. Further similarity with mammals is the expression of serotonin (5-HT) in these cells, which may be acting as a neurotransmitter of the hypoxia stimulus as well as a mediator of vascular responses. Because of the relatively low solubility of O2 in water as compared to ambient air, the physiological consequences of reduced pO2 have more profound effects in aquatic vertebrates as opposed to air breathing animals. In addition, in fish and larval amphibians there is rapid elimination of CO2, implicating that in the aquatic environment it is hypoxia rather than hypercapnia that plays a dominant role in driving cardiorespiratory responses. The carotid labyrinth (CL) in fish and amphibians is analogous to the carotid body (CB) of higher vertebrates. It is a paired structure consisting of a dense capillarylike plexus located at the site of bifurcation of the common carotid artery. As in CB, glomus cells of CL are distributed within the intervascular stroma as single cells or in clusters and are innervated by an extensive network of afferent, efferent and reciprocal synapses. The glomus cells and related neural elements express a range of amines and neuropeptides likely acting as neurotransmitters or neuromodulators. The function of CL is multimodal and includes chemoreception, regulation of vascular tone to maintain blood supply to the eyes and brain, baroreception, and maintenance of brain and intraocular temperature. The chapter on NE cells in the lung of amphibians and in the respiratory organs of air-breathing fish reviews the morphology and postulated functions of these cells. An interesting observation from the phylogenetic perspective is the pattern of NE cell evolution in the respiratory organs. In the simpler species, such as air breathing fish (Polypterus and Amia), solitary non-innervated NE cells of the open type (reaching the airway lumen) and closed type (without luminal contact) are found distributed within the ciliated airway epithelium. A more complex structural arrangement with innervated solitary NE cells of closed type are observed in Triturus, whereas NE cells of open type are found in lungfish Protopterus. Interestingly, clusters of innervated NE cells resembling mammalian NEB are observed only in amphibians and higher vertebrates. Another striking similarity between NE cells in amphibians and mammals is the expression of 5-HT and neuropeptides, including bombesin, gastrin-releasing peptide (GRP), calcitonin and others. Although at present there are no direct physiological studies on amphibian NE cells or NEBs, by analogy with mammalian counterparts they are presumed to function as airway O2 sensors involved in the control of breathing or alternatively in local paracrine regulation of airway and/or vascular responses.

Preface  |  vii

Of particular interest are studies on “bimodal breathers” that include various amphibian vertebrates that use different anatomical structures (i.e., skin, gills, lungs) to achieve gas exchange. Many amphibious vertebrates, at some stage of their development use a combination of skin and gills plus lungs to breath both water (skin and/or gills) and air (skin and/or lungs), thus actually representing “trimodal” breathers. The respiratory control of amphibious vertebrates, using multiple modes of gas exchange, is complex and therefore requires sophisticated sensory systems for regulation, including various chemoreceptor structures. For example, pulmonary stretch receptors in amphibians not only respond to pulmonary volume changes but also to increasing intrapulmonary CO2 levels by decreasing their firing rate. Additional receptor systems include olfactory CO2 receptors, arterial chemoreceptors represented by the CL, arterial O2-sensitive chemoreceptors located in the aortic arch, and chemoreceptors within the pulmonary vasculature and the airways represented by NEBs. Air breathing fish and amphibians illustrate a fascinating functional transition stage during evolution of terrestrial tetrapods from their aquatic fish-like ancestors. It has been pointed out that animals that are more aquatic in nature possess sophisticated receptors and respond primarily to changes in pO2 levels in the internal environment. Transition to air-breathing and terrestrial life necessitated CO2/pH sensitive receptors that are involved in respiratory control. Finally it is noted that during evolution, CO2 chemoreception translocates centrally into the brain stem, while O2 chemoreception remains peripherally located. The NE cell system in the reptilian respiratory tract includes solitary cells that are present in both extra and intrapulmonary airways and NEBs that are found mostly in an intrapulmonary location. As in other species, NE cells in the airway mucosa of tortoises, lizards and snakes express 5-HT and peptides such as calcitonin, calcitonin gene-related peptide (CGRP), and in some species leu-enkephalin. Peptidergic innervation, expressing VIP, SP, NPY and PYY, has also been described. Although the precise function of NE cells in reptilian lung remains unknown, based on morphological observations the possible role as mechanoreceptors has been suggested. The latter, however, remains speculative as there are no direct functional studies on NE cells in the reptilian respiratory system. In contrast, more physiological data is available on airway receptors in birds. The avian respiratory system differs from other species in that it consists of non-expandable lungs and series of air sacs providing a unique unidirectional system of airflow. In the avian respiratory system, both chemoreceptors and mechanoreceptors appear to play a role in feedback mechanisms of respiratory control. In birds, as in mammals, respiratory structures are innervated by the vagus nerve. Based on physiological studies, two main types of receptors have been identified in birds: an inspiratory-inhibitory CO2-sensitive receptor located in the lung, corresponding to vagally innervated NEBs described in several avian species, and a slowly adapting mechanoreceptor presumed to be located in the walls of the air sacs or subpleural membrane since birds lack a diaphragm and have non-expanding lungs. The intrapulmonary chemoreceptors (IPC) show sensitivity to

viii  |  Airway Chemoreceptors in the Vertebrates pCO2 but not pO2. During the last 40 years a great deal of progress has been made in our understanding of the CO2 signal transduction mechanism in avian lungs. Recent evidence indicates that IPC sense changes in intracellular pH (pHi), resulting from the activity of carbonic anhydrase that catalyses hydration/dehydration of CO2 rather than from direct CO2 sensing. Other recent data shows that pH sensing in IPC also involves additional mechanisms including membrane based acid–base transporters, modulation by Ca2+ influx, and the expression of a pH-sensitive TREK tandem pore domain K+ leak channel. Due to the relative ease with which the air sac cavities and membranes of the avian respiratory system can be accessed for experimental purposes, they provide a useful model for further studies of both the peripheral and central components of respiratory control. The Chapter on PNECs in mammalian lungs reviews recent findings on the structure, molecular markers, ontogeny and postulated functions. The general morphologic features including the ultrastructure and immunohistochemical/molecular markers of NE cells and NEBs in mammalian lung have been previously well defined. However the complexity of NEB innervation became apparent only recently. Thus far, only postnatal rat had been studied extensively in terms of the origin of NEB cell innervation and its neurochemical coding, which was extended to mice in a present contribution. Due to known species variations in lung innervation, the findings in the rat and mouse lungs would need to be verified in other species including human to better define their functional significance. Experimental denervation and neural tracing studies combined with multilabel immunohistochemistry and confocal imaging have revealed a complex neural network in rodent NEBs, consisting of vagus nerve derived afferent nerve fibers originating in nodose ganglion neurons, a CGRP expressing component originating in spinal ganglia, and a nitrergic component derived from intrinsic airway ganglia. Thus a complex picture is emerging where different potential signals that may include changes in airway pO2 and possibly pCO2 as well as mechanical forces could be transduced by NEB cells and modulated by their complex innervation. Based on observations of a large number of myelinated vagal nodose afferents in rodent intrapulmonary airways that selectively innervate NEBs, it has been proposed that discharges from NEBrelated myelinated vagal afferent fibers may be part of the already characterized vagal myelinated receptors in lower airways. Therefore, NEBs in mammals can be viewed as multimodal airway receptors that may serve different functions during ontogeny, during transition from the euoxic (relative fetal hypoxia) aqueous intrauterine environment to an extrauterine life dependent on air breathing, and subsequently in the postnatal and adult lung. Recent electrophysiological studies, using the patch-clamp technique combined with molecular approaches, have provided evidence indicating that NEB cells are transducers of hypoxic stimuli via a membrane bound molecular complex (“oxygen sensor”), characteristic of specialized cells that monitor and signal hypoxia in different parts of the body to maintain homeostasis. The O2 sensor in NEB cells of mammalian

Preface  |  ix

lungs and of the related human small cell lung carcinoma cell line (H146) has been partially characterized and consists of a hydrogen peroxide (H2O2) generating, multicomponent NADPH oxidase coupled to O2 sensitive K+ channels. Under normoxia, the oxidase tonically generates H2O2 that, used as a second messenger, modulates the O2-sensitive K+ channel activity via a redox mechanism involving cystein residues in critical components of the K+ channel gating mechanism. Hypoxia leads to decreased generation of H2O2 resulting in K+ channel closure that triggers downstream events leading to neurotransmitter release. Studies using oxidase (gp91phox/Nox2) deficient neonatal mice confirmed the critical role of gp91phox/ Nox2 protein in NEB O2 sensing, as these mice failed to respond to hypoxia in both in vitro and in vivo studies. Thus the predominant O2 sensor in NEBs appears to be a complex of Nox2/Kv3.3 (and possibly combinations with other Kv alpha subunits). Future studies will determine the role of recently described Nox homologues and their potential O2 sensitive K+ channel partners (i.e., acid sensitive two pore domain TASK channels). The membrane based O2 sensor used by pulmonary NEBs appears to be unique since other O2 sensing cells (i.e., CB glomus cells, adrenal medullary cells and pulmonary artery smooth muscle cells) use mitochondria as the primary source of H2O2 that modulates K+ channel activity, suggesting cell and organ specific O2 sensing mechanisms rather than a monolithic system. However in spite of progress in the molecular characterization of O2 sensing, the precise role(s) for NEBs in the lung remain(s) undefined. Postulated functions include that of airway O2 sensors involved in respiratory control, particularly during the perinatal period. There is accumulating evidence that NEB cells could also be sensing pCO2/pH changes in the airways and thus may represent bi-modal receptors similar to the CB. Clearly many more studies, using physiological approaches are required to establish the role and function of NEBs in normal and diseased lungs. The observation of large relative numbers of PNECs/NEBs in fetal lung compared to the adult focused attention on a possible role of these cells during lung development. Since PNEC are the first cell type to differentiate within the primitive airway epithelium, prior to the emergence of other airway epithelial cell types, a role in lung growth and epithelial differentiation has been suggested. The question of the origin, ontogeny and molecular regulation of PNEC differentiation has been a subject of recent investigations. Although the precise origin of PNECs is at present unclear, current evidence suggests derivation from the foregut endoderm that gives rise to pluripotent epithelial progenitors, analogous to endocrine cells in the gastrointestinal tract and pancreas. Putative PNEC progenitors have been identified using antibodies against early neuronal developmental markers, such as FORSE-1 that recognizes the “forebrain surface embryonic” antigen expressed in the developing CNS. During early stages of fetal lung development, FORSE-1 antibody labeled all the cells of the primitive airway epithelium and later became restricted to 5-HT positive PNEC/NEB. At later stages of lung development there was gradual loss of the FORSE-1 epitope with

x  |  Airway Chemoreceptors in the Vertebrates further PNEC differentiation. The presence of a few FORSE-1/5-HT expressing cells in postnatal lung is consistent with the retention of progenitors in mature lungs. At the molecular level, it is now evident that genes involved in neuronal fate determination belonging to the family of basic-loop-helix (bHLH) transcription factors, such as achaete-scute-homolog-1 (Mash 1 in rodents and hASH 1 in humans) and hairy-enhancer of split (Hes1), play related but opposite roles in governing PNEC lineage fate in the lung. Mash1 k/o mice lack PNECs and die of respiratory failure soon after birth. It is of interest that the lungs of heterozygote (Mash+/-) mice show 50% reduction in the number of PNEC/NEB accompanied by an abnormal breathing pattern, supporting the role of these cells as airway sensors. There is also accumulating evidence that NEBs may serve as a pulmonary stem cell niche, providing a cell microenvironment for toxin-resistant stem cells as demonstrated in the naphthalene mouse model of acute airway injury. Based on these recent studies, an overall picture is emerging where PNEC/NEB can be viewed as a multifunctional cell system with roles that are developmentally regulated. Thus during the early stages of lung development, PNECs could be modulators of fetal airway growth and differentiation, at the time of birth as airway O2 sensors involved in neonatal adaptation, and postnatally and beyond as providers of a lung stem cell niche that is important for airway epithelial regeneration and in lung carcinogenesis. An additional role for NEBs may involve modulation of immune responses in the airways via production of a variety of peptide mediators such as CGRP that, in addition to vasodilatory effects, exhibits chemotactic properties, and bombesin/GRP that appears to be involved in mast cell recruitment and activation. In an experimental mouse model of asthma, increased exocytosis of dense-cored granules (the storage site for amine and peptide mediators) from NEB cells was observed after allergen challenge, suggesting a potential role for these cells in the pathophysiology of asthma. When comparing morphologic, physiological and developmental aspects of pulmonary NEBs, the presumed airway O2 sensors, with that of CBs, well established and extensively studied arterial chemoreceptors, a number of similarities and fundamental differences emerge. At the anatomical level, CBs form discrete paired organs composed of densely vasularized clusters of glomus cells that produce and secrete amine and peptide neurotransmitters and are innervated by the glossopharyngeal nerve. These anatomical features have facilitated physiological and other experimental studies on CBs that defined their role as principal arterial chemoreceptors monitoring arterial pO2, pCO2 and pH. In contrast NEBs monitor pO2 in inhaled air, are composed of small cell clusters widely dispersed within the epithelium of intrapulmonary airways, and are innervated mainly by vagal afferents. The relatively small number of NEB, representing <1% of all airway epithelial cells, together with a difficult access, complicates direct experimental studies. This obstacle has been partially overcome by developing methods to study NEBs in vitro and by the use of NEB cell related tumor cell lines to investigate their biochemical and molecular aspects. Using the patch-clamp

Preface  |  xi

and carbon fiber amperometric methods, it has been shown that NEB cells, as their CB glomus cell counterparts, exhibit features of neurosecretory cells with membrane properties of excitable cells. In both cell types the response to hypoxia is modulated by O2 sensitive K+ channels that trigger downstream events resulting in neurotransmitter release. Possible interactions and complementary activity between the two chemoreceptor systems, i.e., the airway based NEB and arterial CB is evidenced by an observation of differences in the time course of structural and functional maturation. During the early perinatal period, NEBs appear to mature in advance of the CB, suggesting a critical function in neonatal adaptation. Theoretically, detection of changes in pO2 directly in the airway facilitates a more rapid response to hypoxia that may be of importance to the neonate. The chapter on solitary chemosensory cells (SCCs) in the airways reviews data on the distribution, immunocytochemistry, fine structure, ontogeny and function of these cells in different species from aquatic vertebrates to mammals. Studies in primary aquatic vertebrates have identified SCCs in the epidermis, oropharynx and gills. These diffusely distributed chemoreceptor cells are related to but distinct from the taste buds. SCCs constitute a third chemosensory modality in addition to the better known senses of taste and smell. The elementary unit of SCCs consists of a single bipolar cell contacted by an afferent nerve fiber. The overall shape of the cell is determined by its location within a particular epithelium, usually with their apical surface contacting the lumen to receive sensory stimuli. Their presumed function is to monitor the incoming water or air stream for possible irritants or toxins, and for predator avoidance. Phylogenetically, SCCs appear prior to the development of taste buds. In mammals SCCs have been described at specific sites in the digestive and respiratory systems. For example, in the rat SCCs were found in the nasal cavity, in the vallate papillae of the tongue, in the larynx, trachea and bronchi. In humans, they have so far been described in the nasal cavity. The so-called “brush cells”, which may be related to SCCs, are characterized by distinct apical microvilli and pleiomorphic cytoplasmic granules. These cells have been described in the epithelia of trachea, bronchi and even in the alveolar region of the rat and human respiratory system. Immunohistochemical studies have identified in SCCs expression of alpha-gustducine and phospholipase C beta 2, both markers of gustatory sensory receptor cells. The role of these enzymes in the function of SCCs is at present unknown but could include local antimicrobial defense via G-protein mediated secretory responses. The tracheal epithelium of mammalian the respiratory system harbors solitary PNECs that express 5-HT as well as CGRP and bombesin (in human).These cells, as their NEB counterparts in the intrapulmonary airways are extensively innervated as shown by recent studies using confocal microscopy. Tracheal PNECs also express O2 sensing properties and may effect bronchomotor tone as demonstrated in experiments using isolated tracheal preparations with intact and denuded epithelium. It has been

xii  |  Airway Chemoreceptors in the Vertebrates proposed that, at least in the guinea pig, solitary PNECs may control spontaneous airway contraction, and may potentially be involved in the pathophysiology of asthma. The last chapter discusses the evolutionary trends in CO2/H+ chemoreception. In most vertebrates, CO2/pH-sensitive chemoreceptors of the airway passages play an important role in regulating ventilation. In strictly water-breathing fish, cardiorespiratory reflexes are responsive primarily to changes in O2 concentration, but a modulatory role is played by water-sensing, CO2-sensitive, branchial chemoreceptors that are activated by hypercarbia. The transition to air breathing has resulted in the emergence of central CO2/pH chemosensitivity, while the peripheral airway CO2/pH chemoreceptors play a modulatory role in ventilatory control. In bimodal breathers, current evidence supports the existence of both water-sensing CO2/pH chemoreceptors located in the gills or orobranchial cavity, and CO2-sensitive pulmonary stretch receptors. In air-breathing vertebrates, several types of CO2/pH-sensitive receptor groups have been identified, particularly in the nasal epithelium. These olfactory CO2-sensitive chemoreceptors are inhibitory, but their precise role is unclear. In birds and reptiles, intrapulmonary CO2/pH-sensitive chemoreceptors are linked to pulmonary stretch receptors in which the pattern of response to the primary stimuli of changes in lung volume, pressure or wall tension is modulated by CO2. The function of these receptors is to reduce dead space ventilation and enhance the efficiency of CO2 elimination during hypercarbia. Thus, the CO2/pH-sensitive chemoreceptors of the respiratory tract share a common purpose of modulating breathing pattern. It is hoped that the readers of this book will be inspired by the enthusiasm and fascination of the authors with the complexity and diversity of airway chemoreceptors, which at a structural and functional level share many fundamental similarities across the various species. Opportunities abound for further investigations of airway sensory systems and their roles in normal physiology and in disease. As the future of the planet earth is being threatened by global warming and climate change, a better understanding of the sensory systems that monitor and respond to changes in the oxygen and carbon dioxide concentrations in the environment may be of vital importance for the survival of all species.

Ernest Cutz, Md, Frcpc Professor of Pathology The Hospital for Sick Children and University of Toronto Toronto, Ontario, Canada

Contents Preface

v Neurosecretory Epithelial Cells (NEC’s) in the Airways and Carotid Labyrinth of Aquatic Vertebrates: Morphology, Distribution, Innervation and Function

1. Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates Michael G. Jonz and Colin A. Nurse

1

2. Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells in Indian Catfishes A. Gopesh

31

3. Serotonergic Neuroepithelial Cells in Fish Gills: Cytology and Innervation Yannick J.R. Bailly

61

Neurosecretory Cells (NEC’s) in the Lung of Amphibians and Accessory Respiratory Organs of the Air-breathing Fishes and in Amphibian Carotid Labyrinth: Structural Morphology and Function 4. Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes Lucyna Goniakowska-Witalńska, Anna Pecio and Dagmara Podkowa

99

5. The Amphibian Carotid Labyrinth Tatsumi Kusakabe

125

6. Neuroendocrine System of the Amphibian Extrapulmonary Airways Luis Miguel Pastor García and Esther Beltrán-Frutos

141

7. Chemoreceptive Control of Ventilation in Amphibians and Air-Breathing Fishes Warren Burggren and Tien-Chien Pan

151

xiv  |  Airway Chemoreceptors in the Vertebrates Neuroepithelial Bodies(NEB’s) in the Lung of Reptiles: Structural Morphology, Immunohistochemistry and Function   8. Neuroendocrine System of the Reptilian Respiratory Tract Luis Miguel Pastor García, Giacomo Zaccone and Esther Beltrán-Frutos

185

  9. Airway Receptors in Birds M. Fabiana Kubke, Roderick A. Suthers and J. Martin Wild

199

10. Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors Steven C. Hempleman and Jason Q. Pilarski

213

Pulmonary Neuroepithelial Cells in Mammals: Structure, Molecular Markers, Ontogeny and Functions 11. Diverse and Complex Airway Receptors in Rodent Lungs  Inge Brouns, Isabel Pintelon, Ian De Proost, Jean-Pierre Timmermans and Dirk Adriaensen

235

12. Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies E. Cutz, W.X. Fu, H. Yeger, J. Pan and C.A. Nurse

269

13. Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System in the Developing Mammalian Lung H. Yeger, J. Pan and E. Cutz

291

14. Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators: Some New Findings and a Review of the Literature  Alfons T.L. Van Loomel, Tania Bollé and Peter W. Hellings

311

15. Neuroepithelial Bodies and Carotid Bodies: A Comparative Discussion of Pulmonary and Arterial Chemoreceptors Alfons T.L. Van Lommel

331

Solitary Chemosensory Cells in the Airways of Mammals: Distribution, Immunocytochemistry, Fine Structure and Function 16. Solitary Chemosensory Cells in the Airways of Mammals  A. Sbarbati, M.P. Cecchini, C. Crescimanno, F. Merigo, D. Benati, M. Tizzano and F. Osculati

359

17. Solitary Chemosensory Cells: Phylogeny and Ontogeny Anne Hansen and Thomas E. Finger

377

18. Functional Importance of Pulmonary Neuroendocrine Cells Staffan Skogvall

389

19. CO2/H+ Chemoreceptors in the Respiratory Passages of Vertebrates K.M. Gilmour and W.K. Milsom

403

Index

427

Color Plate Section

List of Contributors Adriaensen, D. Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171 BE-2020, Antwerp, Belgium Bailly, Y.J.R. Cytologie et Cytopathologie Neuronales, Département Neurotransmission et Sécrétion Neuroendocrine, Institut de Neurosciences Cellulaires et Intégratives CNRS UMR 7168, 5, rue Blaise Pascal, 67084 Strasbourg, France Beltrán-Frutos, E. Department of Cellular Biology and Histology, Medical School, University of Murcia, Campus de Espinardo, 30100, Murcia, Spain Benati, D. Department of Morphological-Biomedical Sciences, Anatomy and Histology Section, University of Verona, Medical Faculty, Italy Bollé, T. Leuven Catholic University, Medical Faculty, Department of Morphology and Molecular Pathology, Minderbroedersstraat 12 Brouns, I. Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171 BE-2020, Antwerp, Belgium Burggren, W. Department of Biological Sciences, University of North Texas,, Denton, TX 76205, USA Cecchini, M.P. Department of Morphological-Biomedical Sciences, Anatomy and Histology Section, University of Verona, Medical Faculty, Italy

xvi  |  Airway Chemoreceptors in the Vertebrates Crescimanno, C. Faculty of Exercise and Sport Science, Kore University, Enna, Italy Cutz, E. CIHR Group on Lung Development Division of Pathology and Department of Paediatric Laboratory Medicine, The Hospital for Sick Children, and Department of Laboratory Medicine & Pathobiology, University ofToronto, Canada De Proost, I. Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171 BE-2020, Antwerp, Belgium Finger, T.E. Department of Cell and Developmental Biology, University of Colorado, School of Medicine, Mail Stop 8108, PO Box 6511 Aurora, CO 80045 Fu, W.X. Division of Neuroscience, Oregon Health Science University, Beaverton, Oregon, USA Gilmour, K.M. Department of Biology, University of Ottawa, Ottawa, ON, Canada Goniakowska-Witalińska, L. Jagiellonian University, Department of Comparative Anatomy, Institute of Zoology, Krakow, Poland Gopesh, A. Department of Zoology, University of Allahabad, Allahabad 211002, India Hansen, A. Department of Cell and Developmental Biology, University of Colorado, School of Medicine, Mail Stop 8108, PO Box 6511 Aurora, CO 80045 Hempleman, S.C. Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640, USA Jonz, M.G. Department of Biology, University of Ottawa, Ottawa, ON, Canada Kubke, M.F. Department of Anatomy with Radiology, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand Kusakabe, T. Laboratory for Anatomy and Physiology, Department of Sport and Medical Science, Kokushikan University, Tokyo, Japan

List of Contributors  |  xvii

Merigo, F. Department of Morphological-Biomedical Sciences, Anatomy and Histology Section, University of Verona, Medical Faculty, Italy Milsom, W.K. Department of Zoology, University of British Columbia, Vancouver, BC, Canada. Nurse, C.A. Department of Biology, McMaster University, Hamilton, ON, Canada Osculati, F. IRCCS Centro Neurolesi “Bonino-Pulejo” Messina, Italy Pan, J. CIHR Group on Lung Development, Division of Pathology and Department of Paediatric Laboratory Medicine, The Hospital for Sick Children, and Department of Laboratory Medicine & Pathobiology, University of Toronto, Canada Pan, T.-C. Department of Biological Sciences, University of North Texas, Denton, TX 76205, USA Pastor, L.M. Department of Cellular Biology and Histology, Medical School, University of Murcia, Campus de Espinardo, 30100, Murcia, Spain Pecio, A. Jagiellonian University, Institute of Zoology, Department of Comparative Anatomy, R. Ingardena 6, 30-060 Krakow, Poland Pilarski, J.Q. Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640, USA Pintelon, I. Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171, BE-2020 Antwerp, Belgium Podkowa, D. Jagiellonian University, Institute of Zoology, Department of Comparative Anatomy, R. Ingardena 6, 30-060 Krakow, Poland Sbarbati, A. Department of Morphological-Biomedical Sciences, Anatomy and Histology Section, University of Verona, Medical Faculty, Italy

xviii  |  Airway Chemoreceptors in the Vertebrates Skogvall, S. PharmaLundensis AB, Bio Medical Center D10, 221 84 Lund, Sweden Suthers, R.A. Medical Sciences, Indiana University, Bloomington IN 47405, USA Timmermans, J.-P. Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171, BE-2020 Antwerp, Belgium Tizzano, M. Rocky Mountain Taste and Smell Center, Department of Cell and Developmental Biology, University of Colorado, Denver, Aurora, USA Van Lommel, A.T.L. Leuven Catholic University, Medical Faculty, Department of Morphology and Molecular Pathology, Minderbroedersstraat 12, 3000 Leuven, Belgium Wild, J.M. Department of Anatomy with Radiology, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand Yeger, H. Division of Pathology, Department of Paediatric Laboratory Medicine, The Research Institute, CIHR Lung Development Group, The Hospital for Sick Children, Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada Zaccone, G. Department of Animal Biology and Marine Ecology, Faculty of Science, University of Messina, Salita Sperone 31, Messina, S. Agata I-98 166, Italy

Neurosecretory Epithelial Cells (NEC’s) in the Airways and Carotid Labyrinth of Aquatic Vertebrates: Morphology, Distribution, Innervation and Function 1. Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates Michael G. Jonz and Colin A. Nurse

1-30

2. Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells in Indian Catfishes A. Gopesh

31-60

3. Serotonergic Neuroepithelial Cells in Fish Gills: Cytology and Innervation Yannick J.R. Bailly

61-98

1 Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates Michael G. Jonz1* and Colin A. Nurse2­

Abstract

All vertebrates respond to acute hypoxic stress with reflex cardiorespiratory adjustments initiated by peripheral O2 chemoreceptors. In mammals, these specialized cells are stimulated by hypoxia and lead to activation of sensory nerves, allowing central integration and resulting physiological responses, such as hyperventilation. While the role of O2 chemoreceptors has been well characterized in mammals, relatively little is known about O2 chemoreception in aquatic vertebrates. This chapter reviews the morphological, physiological and developmental studies that have examined the role of O2-sensitive neuroepithelial cells (NECs), which display characteristics similar to mammalian chemoreceptors, and potential chemosensory pathways of the gills in fish and larval amphibians. The gill is a multifunctional organ that receives sensory innervation from the glossopharyngeal or vagus cranial nerves. Extracellular recordings from these nerves provide strong evidence for the gill as a site of O2 chemoreception. NECs of the gill epithelium, most of which contain the neurotransmitter serotonin, are neurosecretory and associated with nerve fibres. In agreement with the ‘membrane hypothesis’, patch-clamp recordings have indicated that NECs isolated from the zebrafish gill respond to acute hypoxia with decreased K+ channel activity and membrane depolarization. Thus, NECs are potential O2 sensors of the gill that may initiate cardiorespiratory reflexes in aquatic vertebrates. The few available studies that have examined O2 sensing in fish have indicated that a functional system of O2 chemoreception begins to develop before complete formation of the gills, but another extrabranchial site must regulate hypoxic responses during earlier stages. Aquatic vertebrates

Department of Biology, University of Ottawa, 30 Marie Curie, PO Box 450, Station A, Ottawa, ON, K1N 6N5, Canada. 2 Department of Biology, McMaster University, Hamilton, ON, Canada. *Author for correspondence: Email: [email protected], Phone: +1 613 562 5800 ext. 6051, Fax: +1 613 562 5486 1

2  |  Airway Chemoreceptors in the Vertebrates represent attractive models for pursuing studies of O2 sensing, from cellular mechanisms to behavioural responses. These studies will provide important information about O2 sensing and respiratory regulation in aquatic species, and how this system has evolved in vertebrates.

Introduction Vertebrates have evolved elaborate strategies for maintaining adequate levels of arterial O2 during hypoxic stress, and these are fundamental for adaptation to new or changing environments. Similar to mammals, aquatic vertebrates (e.g. fish and larval amphibians) display physiological changes when confronted with O2 lack. These responses include hyperventilation, decreased heart rate (bradycardia), and changes in vascular resistance (Malvin, 1989; Burggren and Pinder, 1991; Burggren and Doyle, 1986; Burleson et al., 1992; Jia and Burggren, 1997a; Burleson and Milsom, 2003; Milsom and Burleson, 2007). Specialized chemoreceptors detect changes in respiratory gases (O2 and CO2) as well as pH, and can initiate these adaptive cardiorespiratory reflexes. In mammals, central chemoreceptors sensitive to CO2 and pH are located in the brainstem, while peripheral chemoreceptors sensitive to CO2, O2, and pH are located in the carotid bodies (González et al., 1994; López-Barneo et al., 2001; Feldman et al., 2003; Putnam et al., 2004). Additionally, O2-sensitive neuroepithelial bodies (NEBs) of the pulmonary epithelium have been described (Youngson et al., 1993; Cutz and Jackson, 1999; Fu et al., 2002). Adult amphibians also display mechanisms of central CO2 sensing (Milsom, 2002) and possess a peripheral O2-sensing organ similar to the mammalian carotid body, called the carotid labyrinth (Kusakabe, 2002; see also Kusakabe’s chapter, this volume). Notably, there is presently no convincing evidence of central O2 chemoreception involved in respiration in vertebrates (Milsom and Burleson, 2007). In fish and larval amphibians, peripheral respiratory chemoreceptors are located primarily in the gills and have recently been identified as NECs (Malvin, 1989; Jia and Burggren, 1997a; Burleson and Milsom, 2003; Jonz and Nurse, 2003; Jonz et al., 2004). Adaptive responses to hypoxia are exquisitely exemplified in these animals because they live in an aqueous environment where the solubility of O2 is relatively low compared to that of air (Burleson et al., 1992; Milsom, 2002). Thus, naturally-occurring fluctuations in the partial pressure of O2 (Po2) have profound physiological consequences in aquatic vertebrates, compared with air-breathing or terrestrial species. In addition, because of the high flow rates over the gills during water-breathing, and the high capacitance for CO2 in water, there is rapid CO2 excretion and minimal amounts of arterial CO2. Therefore, in aquatic environments, hypoxia, rather than hypercapnia, appears to play a dominant role in driving cardiorespiratory responses in fish and larval amphibians (see Gilmour, 2001; Milsom, 2002; and Gilmour and Milsom, this volume). This chapter will review the morphological and physiological studies that have focused on peripheral O2 chemoreception in the gills of aquatic vertebrates, since these

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  3

animals have recently emerged as useful models for studying the evolution of O2 sensing at the cellular level. Where possible, evidence will be compared with mammalian studies to offer an evolutionary perspective on the homologous relationship of O2 sensing structures among different vertebrate taxa. Beginning with a brief description of the structure and phylogenetic significance of the gills, the organization and morphological features of O2 chemoreceptors and their innervation will be presented. In addition, more recent electrophysiological evidence from gill O2 chemoreceptors will help unravel the mechanisms of O2 sensing at the cellular level, and the use of aquatic species, particularly the zebrafish, as models for studying the development of O2 sensing in vertebrates will be discussed.

Morphology Gill Organization in Aquatic Vertebrates The gills of aquatic vertebrates are designed to maximize respiratory surface area and gas exchange in an aqueous environment with a low O2 solubility. While the number and general anatomy of the gills vary within the anamniotes, the organization of the internal gills in all aquatic vertebrates reflects a common phylogenetic or ancestral origin. A region of the embryonic endoderm in vertebrates, called the pharynx, gives rise to respiratory organs, such as the lungs of terrestrial vertebrates and the internal gills (i.e. the pharyngeal arches), of fish and larval amphibians (Weichert, 1967; Gilbert, 1988). In all embryonic vertebrates, a series of bilateral aortic arches branches from the ventral aorta to the paired dorsal aortae to form the anterior arterial system (Figure 1A). In fish and aquatic stages of amphibians, these aortic arches course through the pharyngeal arches and provide blood to the gills once they develop (Weichert, 1967). Although most vertebrates have six pairs of aortic arches during embryogenesis, this number is usually reduced by the time the adult stage is reached. In teleost fish (e.g. goldfish, trout, zebrafish), the first two arches, the mandibular and the hyoid (I and II), degenerate and leave aortic arches III, IV, V and VI to develop with the gills (i.e. gill arches 1, 2, 3 and 4; Figure 1B). This arrangement is also found in aquatic or larval stages of amphibians. By contrast, in mammals and adult amphibians (i.e. after metamorphosis), aortic arch V is lost, in addition to I and II (Figure 1C, 1D). This leaves aortic arches III, IV and VI that go on to form the internal carotid artery, aortic arch proper, and pulmonary artery, respectively, such as in mammals (Weichert, 1967; Gilbert, 1988; Figure 1D). As illustrated in Figure 1, the organization of the aortic arches in vertebrates has remained relatively conserved throughout phylogenesis. For example, the first gill arch (aortic arch III) in fish and larval amphibians is homologous to the internal carotid artery, a site just distal to the bifurcation of the common carotid artery, in adult anurans and mammals. As will be discussed in the following sections, these regions are important because they represent sites of O2 sensing in vertebrates.

4  |  Airway Chemoreceptors in the Vertebrates

Figure 1: Illustration of the changes in aortic arch organization throughout development and phylogenesis in vertebrates. Numerals I through VI indicate aortic arches from anterior to posterior. Grey areas represent arterial tissue that has degenerated. All diagrams are in ventral view. A. Typical condition of aortic arches in vertebrate embryos. Six pairs of arches connect ventral and paired dorsal aortae. B. Organization of aortic arches in adult teleost fish. Arches I and II degenerate, leaving aortic arches III-VI (i.e. gill arches 1-4). Dashed lines indicate connection of afferent and efferent branchial arteries via capillaries. C. Modification of aortic arches as found in adult anuran amphibians. As in fish, arches I and II degenerate. In addition, arch V also degenerates. Note the ventral aorta is modified to become the common carotid artery, which bifurcates into the internal and external carotids (i.e. gill arch 1 and ventral aorta in fish). D. Modification of aortic arches as observed in adult mammals. As in amphibians, only arches III, IV and VI remain. In addition, the right dorsal aorta disappears, leaving the systemic aorta (aorta proper) on the left side. As shown in this illustration, the site of the common carotid bifurcation and the internal carotid artery in mammals (D) is homologous with the first gill arch in fish (B). aba, afferent branchial artery; cc, common carotid artery; da, dorsal aorta; eba, efferent branchial artery; ec, external carotid artery; ic, internal carotid artery; lsa, left subclavian artery; pca, pulmocutaneous artery; pulm. aorta, pulmonary aorta; rsa, right subclavian artery; sa, systemic arch; sys. aorta, systemic aorta; va, ventral aorta. Modified from Weichert (1967).

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  5

Teleost fish have four pairs of gill arches, which contain numerous gill filaments that are divided into two parallel rows, or hemibranchs. Each gill filament gives rise to many respiratory lamellae, where gas exchange occurs (Figure 2A). Blood flow through the gill filaments and lamellae occurs by way of the filament arteries (afferent to efferent) and vascular sinus, respectively, and generally opposes the flow of water over the gills to increase gas exchange during ventilation (see also Figures 3C, 3D). The gill filaments and lamellae are covered by a thin epithelium composed of several cell types that provides a boundary between the external environment and the extracellular fluids. These topics, including the fine structure of the gill epithelium, have been reviewed previously by several authors (Laurent and Dunel, 1980; Laurent, 1984; Hughes, 1984; Wilson and Laurent, 2002; Olson, 2002; Evans et al., 2005).

Figure 2: Representation of the general structures of the internal gills of fish and amphibians. A. Gill filament (F) and respiratory lamellae (L) of a teleost fish (zebrafish, Danio rerio). B. The internal gills of Xenopus laevis larvae are composed of filament-like regions (F) and many respiratory terminal branches (TB) analogous to gill lamellae in fish. Reprinted from Saltys et al. (2006) with permission.

The gills of larval amphibians differ from those of fish in their general morphology. In anurans (e.g. frogs, toads), internal gills develop in premetamorphic larvae on the same four pharyngeal arches as in fish, and give rise to a series of irregular gill ‘tufts’ that divide several times (Figure 2B; see also Figures 5A, 5B). These tufts ultimately form respiratory terminal branches, the sites of gas exchange, and are homologues of respiratory lamellae in the fish gill (Malvin, 1989; Minnich et al., 2002; Saltys et al., 2006). During metamorphosis, the internal gills of anurans are resorbed, leaving the lungs and skin for gas exchange. In urodeles, such as salamanders, external gills are present throughout the larval stage and may persist in neotenic forms (i.e. aquatic adults).

O2 Chemoreceptors of the Gill The gill is a multifunctional organ that performs a variety of physiological processes, such as respiration, ionoregulation, acid-base balance, and nitrogen excretion, and is

6  |  Airway Chemoreceptors in the Vertebrates composed of a variety of cell types (Wilson and Laurent, 2002; Evans et al., 2005). Ultrastructural investigations are integral in implicating the gill as the primary site of O2 chemoreception in fish. More recently, such observations have been confirmed in fish, and amphibian larvae, using confocal and immunofluorescence techniques. This section discusses the available morphological evidence that implicates the gills of aquatic vertebrates in O2 chemoreception, and describes the chemoreceptors themselves. The morphology of peripheral O2 chemoreceptors has been well described in terrestrial vertebrates. Type I (glomus) cells of the mammalian carotid body and NEBs of the pulmonary epithelium are sensitive to changes in arterial or airway Po2, respectively, and exhibit several morphological similarities. For example, both cell types are neurosecretory, as evidenced by the retention of cytoplasmic dense-cored vesicles and chemical neurotransmitters, and type I cells and NEBs receive extensive afferent innervation primarily from the glossopharyngeal (IX) and vagus (X) cranial nerves, respectively, which allows reflex physiological changes in response to hypoxia (González et al., 1994; Cutz and Jackson, 1999). NECs of the fish gill resemble O2 chemoreceptors of terrestrial vertebrates and were first described in teleost fish by Dunel-Erb et al. (1982). NECs were shown to reside in the primary filament epithelium of the gills, where they may be exposed to the incident flow of water during respiration and the arterial blood supply, and to possess characteristics typical of cells involved in secretion and neurotransmission. These characteristics include storage of the neurotransmitter serotonin (5-hydroxytryptamine, 5-HT) and, at the ultrastructural level, retention of cytoplasmic dense-cored vesicles (DCVs) and innervation by adjacent nerve fibres (Dunel-Erb et al., 1982; Bailly et al., 1992; see also Bailly’s chapter, this volume). Unlike mammalian type I cells and NEBs, gill NECs are not typically organized into clusters but are usually found as solitary cells. In addition, NECs with similar morphology and distribution have been described in the gill filaments of all fish species studied to date (Dunel-Erb et al., 1982; Donald, 1987; Bailly et al., 1992; Zaccone et al., 1994, 1997; Jonz and Nurse, 2003; Saltys et al., 2006), indicating that these cells are highly conserved among these animals. Therefore, for these reasons, and because the site of the mammalian carotid body and the first gill arch in fish share common evolutionary origins (see previous section), NECs are strongly believed to bear a phylogenetic relationship with mammalian O2 chemoreceptors and allow for O2 chemosensory responses in the gill. Recent experiments using the combined techniques of confocal microscopy and immunofluorescence have allowed visualization of entire populations of NECs in whole-mount preparations of the gills of several teleost species, including zebrafish, goldfish and trout ( Jonz and Nurse, 2003; Saltys et al., 2006). These studies revealed that NECs were invariably dispersed throughout the gill filaments and lamellae of all gill arches. Trout, however, do not have NECs of the lamellae (Saltys et al., 2006). A similar account of NECs on all gill arches has been reported in other species (Burleson and Smatresk, 1990; Sundin et al., 2000). The generally diffused organization of NECs throughout the gill arches in fish and amphibians may reveal a more primitive

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  7

distribution of O2 chemoreceptors in these animals that arose relatively early in vertebrate evolution. A reduction from multiple O2-sensing sites on aortic arches III to VI (gill arches 1-4) in aquatic vertebrates to a single O2-sensing organ (i.e. the carotid body) on the third aortic arch in mammals (recall Figure 1), may have been associated with the transition from aquatic to aerial respiration, and suggests a phylogenetic link between NECs and carotid body type I cells (see also Burleson and Milsom, 2003; Milsom and Burleson, 2007). It is interesting that this proposed evolutionary phenomenon is exemplified by the changes observed during ontogenesis in amphibians, such as Xenopus. As larvae pass through metamorphosis and gill respiration is replaced by aerial respiration, O2 chemoreceptors are lost along with the gill arches and are replaced by a single O2-sensing organ, the carotid labyrinth, in the adult stage. The size of NECs varies from species to species, but the general morphology of gill NECs appears to be highly conserved (Saltys et al., 2006). Confocal immunofluorescence studies on whole mounts ( Jonz and Nurse, 2003; Saltys et al., 2006) have verified previous conclusions from tissue section experiments that 5-HT-containing NECs are concentrated at the distal regions of the gill filaments (Bailly et al., 1992), and further showed that a second population of smaller NECs are present in the secondary epithelium of proximal respiratory lamellae (i.e. lamellae nearest to the gill arch) where they were most numerous (Figures 3A, 3B; 4A-4C). All NECs are confined to the efferent aspects of the filament and lamellar epithelium. The ‘efferent’ designation refers to the efferent flow of oxygenated arterial blood away from gas exchange surfaces as it returns to the gill arches and systemic circulation (Figures 3C, 3D; 4D). In these regions, NECs may be subjected directly to the flow of water that is incident upon the gills during ventilation. The location of gill NECs specifically to the distal filaments and proximal lamellae may provide further advantages for detection of changes in water Po2. From our observations in zebrafish, the confinement of NECs to the distal regions of the filaments appears to be related to the general shape of the filaments themselves, which ‘flare’ outward (see Figure 3A) in each hemibranch directly into the flow of water over the gills, thereby maximizing potential exposure to changes in water Po2. Furthermore, because the proximal lamellae receive a continuous flow of inspired water during resting and hypoxic conditions, and are sites in the gill where ventilation and perfusion are favoured over distal lamellae (Hughes, 1972; Booth, 1978), NECs of the proximal lamellae are also well placed. Indeed, lamellar NECs have been shown to be exposed directly to the external environment in zebrafish (Figure 3B). On the other hand, because NECs of the filament and lamellae are located near the efferent arterial vasculature and vascular sinus, respectively, that carries oxygenated blood, they may also detect changes in arterial Po2, such as during periods of hypoxaemia (Figures 3C, 3D; 4D). The orientation of NECs within the gill epithelium is unclear. While the above confocal studies have not confirmed if NECs of the gill filaments are exposed directly to the external or arterial environments (or both), previous studies have classified 5-HTpositive NECs of the filaments as ‘closed’ to the external environment (i.e. internally

8  |  Airway Chemoreceptors in the Vertebrates

Figure 3: Distribution of neuroepithelial cells (NECs) in the gills of zebrafish (Danio rerio). A, B. 5-HT immunohistochemistry of whole-mount gill tissue. A. NECs of the gill filaments (F) are concentrated in the distal filament regions (arrowheads), while NECs of the lamellae (L) tend to be distributed within the proximal gill regions (arrows). Scale bar = 50 µm. B. NECs of the respiratory lamellae (arrow) directly contact the external environment (EE). Scale bar = 10 µm. Images reprinted from Jonz and Nurse (2003) with permission. C, D. General orientation of gill structures and the location of filament NECs in longitudinal (C) and transverse (D) planes. NECs are located within the efferent aspect of the filament epithelium between the incident flow of water during ventilation and the arterial blood flow. aFA, afferent filament artery; eFA, efferent filament artery; eNB, extrinsic nerve bundle; F, filament; FE, filament epithelium; iNB, intrinsic nerve bundle; L, lamellae; NEC, neuroepithelial cell; NP, nerve plexus. Large arrows indicate the flow of water during ventilation, small arrows indicate blood flow, and dashed arrows indicate possible pathways of O2 diffusion to NECs. C and D are reprinted from Jonz et al. (2004) with permission.

oriented), and those that are 5-HT-negative and immunoreactive for enkephalins as ‘open’ and potentially exposed to the exterior (reviewed by Zaccone et al., 1994). However, this method of classification would not account for the 5-HT-positive NECs of the lamellae in zebrafish that are clearly of the open type ( Jonz and Nurse, 2003;

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  9

Figure 4: NECs of the gill display morphological characteristics of O2 chemoreceptors. Confocal images from whole-mount immunohistochemistry of zebrafish using antibodies against 5-HT (green) and the neuronal marker zn-12 (red). A. NECs of the filament (F, arrows) and lamellae (L, arrowheads) are associated with nerve fibres. Scale bar = 50 µm. B. Higher magnification of a gill filament showing NECs and associated nerve bundles (NB) and nerve fibres (NF). Note the large nuclei in NECs and points of contact between NECs and nerve fibres (arrowheads). Scale bar = 5 µm. C. Image similar to B and rotated 60º around its longitudinal axis demonstrates the presence of synaptic vesicles (SV), labelled with antibodies against SV2, in the basal cytoplasm of NECs facing a nerve bundle (NB). A nerve fibre (NF) also makes contact with a NEC. Same scale as in B. D. Higher magnification of a gill filament similar to that shown in A that was tilted back 90º to reveal a cross-section through the filament. Note the location of NECs (arrows) in the filament epithelium (F) with nearby nerve bundle (NB), a NEC (arrowhead) and nerve fibres of the lamellae (L), a ‘chain neuron’ (ChN), and efferent filament artery (eFA). Scale bar = 10 µm. E. High magnification image showing the close association between NECs of the lamellae and nerve fibres (arrowheads). Additional labelling of cytoplasmic synaptic vesicles with SV2 antibodies produced a yellow signal. Scale bar = 5 µm. A-E are reproduced from Jonz and Nurse (2003) with permission.

and see Figure 3B). While little is known of the orientation of O2 chemoreceptors in amphibians (see Jia and Burggren, 1997b), it is interesting to note that the distribution of gill NECs in larval Xenopus appears to correspond to the general organization of

10  |  Airway Chemoreceptors in the Vertebrates vascular branching in the gill tufts of anurans (see Malvin, 1989; Minnich et al., 2002). This close relationship between the distribution of NECs and arterial branching may suggest that NECs in Xenopus are capable of detecting changes in arterial O2, as well as environmental changes. In this regard, Perry and Gilmour (2002) suggested that cells residing in a multilayered epithelium, such as that of the gill, need not be exposed directly to water or arterial blood and may be capable of sensing changes in both exterior and arterial environments. As will be discussed later in Section 3.1, the orientation of NECs within the gill has important implications in determining the initiation of adaptive physiological effects by hypoxia versus hypoxaemia. As shown in Figure 4, NECs of the zebrafish gill filaments and lamellae label strongly for antibodies against 5-HT and the synaptic vesicle protein SV2 ( Jonz and Nurse, 2003; Saltys et al., 2006). Although 5-HT labelling appears to be diffused throughout the cell, at high magnification a large nucleus is observed that is circumscribed by an intensely fluorescent 5-HT-positive cytoplasm, where the neurotransmitter appears to be stored (Fig. 4B). Consistent with previous ultrastructural reports of DCVs in gill NECs facing nerve profiles (Dunel-Erb et al., 1982; Bailly et al., 1992), confocal images have confirmed that NECs of the gill filaments and lamellae in fish are rich in SV2-positive synaptic vesicles within the basal cytoplasm that may allow for neurotransmitter release onto adjacent nerve fibres during periods of hypoxia (Figures 4C, 4E; Jonz and Nurse, 2003; see also Saltys et al., 2006). Interestingly, studies have suggested that DCV degranulation (suggestive of exocytosis) occurred in NECs after exposure to acute hypoxia, and that 5-HT depletion may have occurred in zebrafish NECs following chronic hypoxia (Dunel-Erb et al., 1982; Jonz et al., 2004). Since hypoxia is also known to induce 5-HT release in O2-sensitive NEBs of the pulmonary epithelium in mammals (Cutz et al., 1993; Fu et al., 2002), these studies suggest an important neurochemical role for 5-HT in O2 sensing in the fish gill. Observations from confocal studies have indicated that the general morphology and distribution of gill NECs in aquatic stages of amphibians are similar to those of fish. Saltys et al. (2006) showed that in larvae of Xenopus laevis, 5-HT-positive NECs were identified in the gill tufts and respiratory terminal branches (Figures 5A, 5B) and resembled the distribution of NECs of the gill filaments and lamellae in fish. In addition, NECs of Xenopus gill filaments were also SV2-positive, indicating a role for these cells in the storage and secretion of neurotransmitters (Figure 6C). NECs have also been described in the external gills of the neotenic (i.e. aquatic adult) tiger salamander (Ambystoma tigrinum) using SV2 immunolabelling techniques (Figure 5C) and at the ultrastructural level (Figure 5D), and indicate the neurosecretory nature of NECs in salamander ( Jonz and Nurse, 2006; Goniakowska-Witaliñska et al., 1993). However, 5-HT immunoreactivity of NECs of the salamander gill has not been described. Although most studies characterizing the morphology and distribution of gill NECs have exploited the expression of 5-HT in these cells for their identification, it is important to note that not all gill NECs contain 5-HT. A relatively small proportion

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  11

Figure 5: NECs in amphibian gills. A. Confocal imaging of NECs and nerve fibres of the filaments (F) and respiratory terminal branches (TB) of the gill labelled with antibodies against 5-HT (green) and zn-12 (red) in larval Xenopus laevis. Scale bar = 50 µm. B. A region of the image in A shown at higher magnification reveals the close association of NECs and nerve fibres, as evidenced by colocalization of 5-HT and zn-12 labelling (arrowheads, and in inset). Reproduced from Saltys et al. (2006) with permission. Scale bar = 10 µm in B (inset 5 µm). C. Cryosection of a gill lamella (L) of the tiger salamander, Ambystoma tigrinum, labelled with antibodies against SV2 reveals the presence of NECs (arrows). Reprinted from Jonz and Nurse (2006) with permission. Scale bar = 10 µm. D. Transmission electron micrograph of a single NEC of the tiger salamander gill. Shown are the nucleus of a NEC (N), a neighbouring basal cell (BC), Golgi complexes (asterisk denotes general region), cytoplasmic dense-cored vesicles (arrows), and numerous mitochondria. Modified from Goniakowska-Witaliñska et al. (1993) with permission.

of NECs have been described in the gill filaments and respiratory lamellae of zebrafish, goldfish, trout, and Xenopus that label positive for antibodies against SV2, but are immunonegative for 5-HT (Figure 6A-6C; Jonz and Nurse, 2003; Saltys et al., 2006).

12  |  Airway Chemoreceptors in the Vertebrates

Figure 6: Some gill NECs in aquatic vertebrates do not contain 5-HT. Confocal images of the gills of zebrafish (A), goldfish (B), and Xenopus laevis (C) labelled with antibodies against 5-HT, SV2, and zn-12 (in A). NECs of the filaments (F) and lamellae (L) positive for both 5-HT and SV2 are labelled green/yellow (arrows), while those that are SV2-positive but 5-HT-negative are labelled in red (arrowheads). Scale bars = 10 µm in A, 25 µm in B and C. Reproduced from Saltys et al. (2006) with permission.

In addition, 5-HT-negative NECs have been reported in the gills of other fish species (Zaccone et al., 1994). It has been proposed that 5-HT-negative NECs may represent a proliferative population of NECs that has not yet synthesized, or taken up, 5-HT ( Jonz and Nurse, 2003; Jonz et al., 2004). Alternatively, these may be neurosecretory NECs for which the neurotransmitter is yet to be identified. This would suggest that if 5-HT-negative NECs are involved in O2 sensing, other chemicals may play a role in afferent neurotransmission between O2 chemoreceptor and sensory nerve fibre, or in paracrine pathways, within the gill. Several studies have indicated that, in addition to 5-HT, potential neurochemical candidates may include nitric oxide or catecholamines, since there is evidence that their respective synthetic enzymes, neuronal nitric oxide synthase and tyrosine hydroxylase, have been localized to NECs (Zaccone et al., 2006; Burleson et al., 2006), and a variety of neuropeptides may also be involved (for reviews see Zaccone et al., 1994, 1997). By comparison, several neurochemicals have been described in the carotid body and may play important excitatory, inhibitory or modulatory roles in O2 sensing. These vary from species to species but may include acetylcholine, ATP, catecholamines (dopamine, norepinephrine), 5-HT, GABA, and neuropeptides (González et al., 1994; Nurse, 2005; Prabhakar, 2006; Lahiri et al., 2006). Although not yet supported by physiological data, histochemical evidence suggests that the neurochemical basis of O2 chemoreception in the gill may involve multiple populations of NECs (i.e. 5-HT-positive and negative NECs of the filament and lamellae), multiple neurotransmitters or neuropeptides, and perhaps a diversity of excitatory, inhibitory and modulatory mechanisms. Interestingly, a variety of

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  13

neurochemicals, such as ACh, 5-HT and dopamine applied exogenously to the gill have been shown to have stimulatory effects on sensory nerve fibres and cardiorespiratory reflexes (Burleson and Milsom, 1995a, 1995b).

Innervation of Gill O2 Chemoreceptors According to the embryonic vertebrate plan, four cranial nerves originally supply sensory and/or motor fibres to the pharyngeal region. The trigeminal nerve (V) provides innervation to the mandibular arch; the facial nerve (VII) supplies the hyoid arch; the glossopharyngeal nerve (IX) innervates the third arch (i.e. first gill arch); and the vagus nerve (X) associates with the remaining posterior arches (Weichert, 1967; Sadler, 2004). However, as development proceeds, and as the pharyngeal arches become specialized to support gills (see Figure 1), branches of the glossopharyngeal and vagus nerves become the primary sources of innervation to the gill arches in both fish and larval amphibians (Rugh, 1951; Nilsson, 1984; Sundin and Nilsson, 2002). In teleost fish, the gill arches are innervated primarily by pretrematic and posttrematic branches of the glossopharyngeal and vagus nerves, which carry parasympathetic and motor (efferent) fibres to the gill vasculature and skeletal muscles, and visceral sensory (afferent) fibres from chemoreceptors of the gill filaments (Nilsson, 1984; de Graaf, 1990; Sundin and Nilsson, 2002). All nerve fibres that enter the gill filaments and lamellae are carried by the branchial nerve, a large nerve trunk that extends the length of the gill arch and sends projections into each filament at its base. Several studies have previously shown the presence of neurons and nerve fibres in the gills (Bailly and Dunel-Erb, 1986; Bailly et al., 1989; de Graaf, 1990; Donald, 1984, 1987; Dunel-Erb and Bailly, 1986; Dunel-Erb et al., 1989), and ultrastructural evidence has indicated that 5-HT-positive NECs of the gill filaments in teleost fish receive innervation (Dunel-Erb et al., 1982; Bailly et al., 1992; see also Bailly’s chapter in the present volume). Recent confocal immunofluorescence studies have confirmed these earlier findings and have shown that in whole-mount preparations, NECs of the gill filaments and respiratory structures (i.e. lamellae or terminal branches) in a variety of teleost and amphibian species are intimately associated with nerve fibres immunolabelled with a zebrafish-derived neuronal marker, zn-12 (Figures 4A-4E; 5A, 5B; 6A; Jonz and Nurse, 2003; Saltys et al., 2006). Furthermore, experiments designed to characterize the source of this innervation revealed that NECs of the gill filaments receive innervation both from nerve fibres with corresponding cell bodies that are ‘extrinsic’ to the gill, as well as from nerve fibres with cell bodies that are ‘intrinsic’ and located within the gill filaments ( Jonz and Nurse, 2003). NECs of the lamellae, on the other hand, receive only extrinsic innervation. These experiments involved the removal (and consequent denervation) and maintenance of gill tissue in short-term explant culture. After 48 h, immunolabelling with zn-12 indicated that only nerve fibres of extrinsic origin (i.e. those that were separated from the cell body after explantation) had

14  |  Airway Chemoreceptors in the Vertebrates

Figure 7: Denervation of the zebrafish gill revealed that NECs are innervated by two separate populations of nerve fibres. Confocal images show NECs (green) and nerve fibres (red) labelled with antibodies against 5-HT and zn-12, respectively. A. In control tissue prepared for immunohistochemistry immediately after removal, NECs of the filament (F, arrows) were associated with fibres of an intrinsic nerve bundle (NB). Extrinsic nerve fibres were also seen associated with NECs in the lamellae (L, arrowhead), and an intrinsic ‘chain neuron’ (ChN) is also visible. B. In gill tissue removed from zebrafish and maintained in explant culture for 2 d, nerve fibres of the intrinsic nerve bundle remained and were associated with NECs of the filament. However, nerve fibres of the lamellae had degenerated. A chain neuron also survived denervation. Scale bar = 10 mm.

degenerated, while intrinsic neurons had remained intact throughout the culture period (Figure 7). Extrinsic nerve fibres originate from large nerve bundles that run the length of the gill filaments (Figure 4A) and form an extensive nerve plexus that wraps around the efferent filament artery, where contact with NECs is made (Figures 4B, 4D). The remaining innervation to NECs of the filaments in zebrafish is supplied by 5-HTpositive intrinsic neurons, such as the proximal neurons at the base of the filaments. Proximal neurons are multipolar and also send projections that terminate locally at the base of the filament artery (Figures 8A, 8B). In teleosts, intrinsic innervation of the filament artery originates from serotonergic and cholinergic neurons and is believed to control blood flow in the gill by way of modulating vascular resistance at the site of a contractile segment of the efferent filament artery (Bailly et al., 1989; Dunel-Erb et al., 1989). Extrinsic and intrinsic innervation patterns in the zebrafish gill are summarized in a schematic in Figure 8C. While the innervation patterns of NECs in the fish gill have been well defined, less well known is the innervation of NECs in the gills of amphibian larvae. Previous studies have shown the presence of nerve fibres in the gills of a larval amphibian (Rana

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  15

Figure 8: NECs in zebrafish (Danio rerio) receive innervation from intrinsic and extrinsic sources. A, B. Immunohistochemical labelling of proximal filament neurons (which innervated NECs) terminating (arrowheads) at the gill vasculature shown in the longitudinal plane (A) and rotated about the vertical axis by 70º (B). Scale bar = 20 µm in A (applies to B). C. Schematics summarizing the intrinsic (left) and extrinsic (right) innervation in of NECs in the zebrafish gill. Small arrows indicate the direction of nerve fibres that issue from the nerve plexus of the filament and innervate NECs of the lamellae. BN, branchial nerve; ChN, chain neuron; DPN, deep proximal neuron; eBA, efferent branchial artery; eFA, efferent filament artery; NEC, neuroepithelial cell; SPN, superficial proximal neuron. Reprinted from Jonz and Nurse (2003) with permission.

catesbeiana), but these were not attributed to any chemosensory or respiratory role (Kusakabe and Kawakami, 1992; Kusakabe et al., 1993). The organization of nerve fibres in the gills of Xenopus resembles that of the fish gill. For example, nerve fibres course through the gill tufts and into the respiratory terminal branches where they associate with NECs in both regions (Figures 5A, 5B; Saltys et al., 2006). However, while only neurons intrinsic to the fish gill contain 5-HT in some species, at least some extrinsic nerve fibres in the Xenopus gill that contact NECs appear to contain 5-HT ( Jonz and Nurse, 2003; Saltys et al., 2006). In addition, neurons intrinsic to the Xenopus gill filaments have also been reported (Saltys et al., 2006), as in the fish gill. Notably, studies at both the confocal and ultrastructural levels have shown that NECs of the external gills of the neotenic tiger salamander appear not to receive s (Figures 5C,D; Goniakowska-Witaliñska et al., 1993; Jonz and Nurse, 2006), and no innervation was reported within respiratory tissues of the gill (Malvin and Dail, 1986). Thus, a potential role for NECs of the external gills in amphibians as O2 chemoreceptors involved in respiratory regulation is not clear. Similar to the innervation of O2-sensitive type I cells of the carotid body, innervation of gill NECs in aquatic vertebrates includes a sensory component (Milsom and Brill, 1986; Burleson and Milsom, 1993; Straus et al., 2001. However,

16  |  Airway Chemoreceptors in the Vertebrates the presence of nitrergic and serotonergic nerve fibres in the gills that contact NECs ( Jonz and Nurse, 2003; Zaccone et al., 2006) suggests that the response of O2 chemoreceptors to hypoxia may potentially be influenced by efferent neurotransmission. Efferent innervation of O2 chemoreceptors has been observed in mammals. For example, nitrergic neurons located within the glossopharyngeal nerve in rat innervate the carotid body and provide efferent inhibition to type I cells, thereby modulating O2-chemosensory responses to hypoxia (Campanucci and Nurse, 2007). Therefore, nitrergic nerve fibres that innervate gill NECs in fish (Zaccone et al., 2006) may represent a feedback pathway homologous to that of the rat carotid body. In addition, the presence of intrinsic neurons in the gill that also contact NECs ( Jonz and Nurse, 2003) suggest that these cells, though not well characterized, may also be candidates for a feedback response pathway.

Effects of Chronic Hypoxia and Hyperoxia Chronic exposure to hypoxia in fish induces lasting physiological changes that may underlie long-term adaptation to hypoxic environments. For example, hyperventilation is induced by hypoxic stimulation of O2 chemoreceptors of the gill (Burleson and Milsom, 2003), and this response appears to be modified by chronic hypoxia. In a variety of species, previous acclimation to low Po2 for a period of 1-4 wk resulted in an increased ventilatory response (i.e. frequency, amplitude, and opercular pressure) during subsequent exposure to acute hypoxia (Figure 9A), compared to the acute hypoxic response of control fish maintained at normoxic levels (Kerstens et al., 1979; Johnston et al., 1983; Burleson et al., 2002). Similar ventilatory changes following exposure to chronic hypoxia also occur in mammals (Powell, 2007). Consistent with these findings, acclimation of zebrafish to hyperoxia for 4 wk produced a blunted hyperventilatory response to acute hypoxic exposure (Figure 9B); however, previous acclimation of zebrafish to chronic hypoxia had no effect (Vulesevic et al., 2006). Catecholamine secretion from chromaffin cells appears also to be stimulated by O2 chemoreceptors of the gill and leads to adaptive physiological changes (Randall and Perry, 1992; Reid and Perry, 2003). Following cholinergic stimulation, catecholamine secretion by chromaffin cells was higher in trout previously exposed to chronic hypoxia than in normoxic controls (Montpetit and Perry, 1998). These studies, therefore, suggest that O2-sensitive NECs of the gills, which may initiate changes in ventilation and catecholamine release, undergo morphological or physiological changes following chronic exposure to hypoxia or hyperoxia,­ and such plasticity may underlie adaptation to such environments. Indeed, confocal studies have shown that in zebrafish maintained in chronic hypoxia, gill NECs increased in size and extended neuron-like processes toward adjacent nerve fibres, compared to normoxic controls (Figures 9C-9F; Jonz et al., 2004). Furthermore, the density of 5-HT-containing NECs was shown to decrease within the gill epithelium in zebrafish chronically

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  17

Figure 9: O2 chemoreceptor plasticity in the fish gill. A. Changes in opercular pressure during ventilation in response to acute hypoxia in the channel catfish (Ictalurus punctatus) acclimated to normoxia (150 mmHg, control) or chronic hypoxia (75 mmHg). Note the increased hypoxic sensitivity in chronically-exposed fish. Modified from Burleson et al. (2002). B. Changes in ventilation frequency in response to acute hypoxia in zebrafish (Danio rerio) acclimated to normoxia (control) or chronic hyperoxia (350-450 mmHg). Note the blunted hypoxic sensitivity in chronicallyexposed fish. Modified from Vulesevic et al. (2006). In the zebrafish gill, NECs labelled with SV2 antibodies increase in number (C, D, arrows), and NECs labelled with 5-HT antibodies extend long neuron-like processes (E, F, arrows), following exposure to chronic hypoxia (35 mmHg) compared to control. F, filament; iNB, intrinsic nerve bundle. Scale bar = 20 µm in C and D, 10 µm in E and F. Reprinted from Jonz et al. (2004) with permission.

exposed to hyperoxia (Vulesevic et al., 2006), consistent with blunting of the hypoxic hyperventilatory response under these conditions, while there were no changes in the density of 5-HT-containing NECs following chronic hypoxia ( Jonz et al., 2004). In the latter study, however, a population of 5-HT-negative NECs did appear to proliferate in

18  |  Airway Chemoreceptors in the Vertebrates response to chronic hypoxia, but O2-sensitivity in these cells has not been investigated. In mammals, changes in the morphology and number of O2-sensitive type I cells of the carotid body were associated with exposure to chronic hypoxia. Chemoreceptor hypertrophy and proliferation occurred after in vitro and in vivo exposure (Stea et al., 1992; Mills and Nurse, 1993; Nurse and Vollmer, 1997; Wang and Bisgard, 2002). Furthermore, chronic exposure to hypoxia has also been shown to induce physiological changes at the cellular level, such as ion channel expression and neurotransmitter regulation in type I cells, and altered sensitivity of the carotid body to hypoxia (Stea et al., 1992; Wyatt et al., 1995; Jackson and Nurse, 1997; Bisgard, 2000). While such physiological responses have not yet been investigated in aquatic vertebrates, it appears that both morphological and physiological plasticity in peripheral O2 chemoreceptors may be important for maintaining ventilatory drive, and other adaptive processes, during acclimatization to hypoxia in all vertebrates.

Physiological Evidence of Gill O2 Chemoreceptors Branchial Nerve Recordings Evidence from earlier studies investigating the role of the gill in initiating cardiorespiratory reflexes in fish in response to water hypoxia and hypoxaemia was instrumental in implicating the gill as an important site of O2 chemoreception (reviewed by Burleson et al., 1992; and see Reid’s chapter, this volume). However, it was not until branchial nerve recordings from isolated perfused gill preparations were performed that direct physiological evidence of gill O2 chemoreceptors was available (Milsom and Brill, 1986; Burleson and Milsom, 1993). These studies reported an increase in discharge frequency from sensory fibres of the branchial nerves associated with decreased Po2 during extracellular electrophysiological recording. This response resembled the increase in carotid sinus nerve discharge in mammals during exposure of carotid body chemoreceptors to a decrease in arterial Po2 (González et al., 1994). Furthermore, these experiments indicated that chemosensory responses appeared to be produced by a decrease in both water and arterial Po2, and suggested the presence of internally and externally oriented chemoreceptors (Milsom and Brill, 1986; Burleson and Milsom, 1993). Therefore, these results are consistent with morphological evidence suggestive of sensory innervation of O2-sensitive NECs of the gill ( Jonz and Nurse, 2003, 2005). As in fish, the gills of larval amphibians have been shown to be important sites of O2 chemoreception involved in the initiation of cardiorespiratory reflexes in response to external hypoxia ( Jia and Burggren, 1997a, 1997b). Furthermore, nerve recordings from isolated gill preparations of amphibian larvae (R. catesbeiana) have demonstrated an increase in discharge frequency of branchial nerve fibres associated with a decrease in external Po2 (Straus et al., 2001). While there is some evidence for a group of internally oriented O2 chemoreceptors in Rana ( Jia and Burggren, 1997b),

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  19

a correlation between decreased internal or arterial Po2 and chemosensory responses has not been investigated in amphibians.

Patch-clamp Recordings Single cell patch-clamp recordings have been successful in revealing important steps in the O2-sensing response in respiratory chemoreceptors at the cellular level. While the precise mechanisms of O2 chemoreception remain controversial (see below), the patch-clamp technique has permitted demonstration of the role of membranebound ion channels that are sensitive to changes in O2 tension. O2-sensitive ion channels, such as those that are permeable to K+, Ca2+, and Na2+, have been described in a variety of mammalian cell types, including central and peripheral neurons, vascular smooth muscle cells, adrenal medullary chromaffin cells, and pulmonary NEBs (Cutz and Jackson, 1999; López-Barneo et al., 2001; Plant et al., 2002; Campanucci et al., 2003; Neubauer and Sunderram, 2004). Of particular interest to the present discussion, are those studies that have explored the roles of O2-sensitive ion channels in mammalian respiratory chemoreceptors, such as type I cells of the carotid body, which initiate adaptive ventilatory responses to hypoxia. According to the ‘membrane hypothesis’ , it is proposed that hypoxia depolarizes type I cells by way of K+ channel inhibition, and this leads to a series of events, including Ca2+ influx through voltagegated Ca2+ channels, release of neurotransmitters, and subsequent activation of afferent nerve fibres (Peers, 1997; López-Barneo et al., 2001; Weir et al., 2005). A class of K+ channels, called two-pore (2P) domain or ‘leak’ channels, have been implicated in the chemoreceptor response to hypoxia. Leak K+ channels function independently of voltage change and conduct K+ ions at resting membrane potential, when other K+ channels may be closed. Thus, these channels make a significant contribution to setting membrane potential and input resistance (i.e. excitability) of the cell (for reviews see Lesage and Lazdunski, 2000; Goldstein et al., 2001; Lesage, 2003). Interestingly, leak K+ channels in O2 chemoreceptors of mammals and fish are sensitive to Po2, and inhibition of these channels appears to mediate hypoxia-induced membrane depolarization. In type I cells isolated from rat carotid body, decreased Po2 reduced membrane current that was carried by K+ and resistant to the K+ channel blockers tetraethylammonium (TEA) and 4-aminopyridine (4-AP) (Buckler, 1997, 2007). The O2-sensitive K+ current was carried through leak channels that were closely related to the TASK subfamily of tandem-P-domain K+ channels, and exhibited complete inhibition in the presence of quinidine (Buckler et al., 2000). Figures 10A and 10B show membrane currents from rat type I cells during application and removal of hypoxia, and isolation of the O2-sensitive leak K+ current, respectively. In NECs isolated from the gills of adult zebrafish, similar O2-sensitive currents were found ( Jonz et al., 2004). As shown in Figures 10C and 10D, the application of hypoxia reversibly reduced membrane currents carried by K+ in NECs at resting membrane potential (−53 mV) and across a range of

20  |  Airway Chemoreceptors in the Vertebrates voltages in a manner similar to that shown in type I cells. These O2-sensitive currents in NECs were resistant to TEA and 4-AP but were completely inhibited by quinidine ( Jonz et al., 2004). A similar resistance of O2-sensitive K+ channels to TEA and 4-AP, and sensitivity to quinidine, was reported in other mammalian chemoreceptors, such as type I cells, peripheral neurons, and NEB-derived H-146 cells (Buckler, 1997; Buckler et al., 2000; Campanucci et al., 2003; O’Kelly et al., 1999). Therefore, biophysical and

Figure 10: Patch-clamp recording of isolated O2 chemoreceptors from the rat carotid body (A, B) and the zebrafish gill (C, D). A. Current-voltage (I-V) relationship from voltage-clamp recordings showing inhibition of membrane current in response to hypoxia (10 mmHg) in carotid body type I cells of rat. B. Difference current from A reveals a hypoxia-sensitive background K+ current. Reprinted from Buckler (1997) with permission. C. I-V relationship from voltage-clamp recordings in zebrafish NECs showing reversible inhibition of membrane current after hypoxic solution (25 mmHg, Hox) was added to the recording chamber. Nox, normoxia; Rec, recovery. D. Isolation of the difference current from C revealed that, as in rat type I cells, the hypoxia-sensitive current appeared to be due to background K+ channels. C and D reprinted from Jonz et al. (2004) with permission. Dashed lines in B and D represent the best fit of the data with the Goldman-Hodgkin-Katz current equation, and indicate voltage independence of the O2-sensitive current.

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  21

pharmacological properties suggest that the O2-sensitive current in NECs is carried by leak K+ channels, though the particular subtype remains to be determined. In addition, current-clamp recordings from zebrafish NECs also demonstrated that hypoxiainduced depolarization was resistant to TEA and 4-AP, but was completely abolished in the presence of quinidine ( Jonz et al., 2004). In another study, exposure of adult and larval zebrafish to quinidine in the surrounding water stimulated an increase in ventilation frequency ( Jonz and Nurse, 2005). Thus, these data strongly indicate a role for leak K+ channels in hypoxic depolarization at the cellular level, and in initiating the hypoxic response, in these animals. Cellular mechanisms of O2 sensing have not been described in detail in other aquatic species, but a putative O2-sensitive K+ current was reported in NECs of catfish (Burleson et al., 2006). Regulation of leak K+ channels by hypoxia, therefore, appears to be a fundamental mechanism that has been relatively conserved and may have appeared early in vertebrate evolution.

Development of O2 Chemoreception in Aquatic Vertebrates The development of O2 sensing by chemoreceptors of the mammalian carotid body has been well described (e.g. Donnelly, 2005; Bairam and Carroll, 2005). The development of cardiorespiratory regulation in fish and amphibians, and the effects of hypoxia during early life, have been investigated at the systemic level (for reviews see Rombough, 1988; Burggren and Pinder, 1991), but only recently have studies focused on the development of O2 chemoreception in these animals (reviewed by Jonz and Nurse, 2006). This section will discuss these aspects and will place particular focus on the zebrafish, since this animal represents a useful developmental model and is the only non-mammalian species for which O2 chemoreceptors have been characterized in larvae and adults. Respiratory regulation and sites of gas exchange in aquatic vertebrates change dramatically throughout development. This is due primarily to organogenesis and differentiation of specialized tissues for gas exchange and transport, as well as increases in total body size. During embryonic and larval stages, when fish and amphibians are small, respiration occurs primarily via cutaneous gas exchange. The entire skin surface, including the yolk sac and fins, is permeable to O2 (Rombough, 1988; Burggren and Pinder, 1991; Wells and Pinder, 1996). Therefore, diffusion of O2, rather than convective transport, is sufficient to meet metabolic demands during early life. In zebrafish, for example, larvae between 3 and 14 d postfertilization (dpf ) survive by cutaneous respiration alone (Rombough, 2002), and convective delivery of O2 to tissues is not a requirement for survival until 14 dpf (Pelster and Burggren, 1996; Jacob et al., 2002). Similarly, cutaneous diffusion plays a more significant role than O2 transport in developing Xenopus compared to adults (Territo and Burggren, 1998).

22  |  Airway Chemoreceptors in the Vertebrates As development proceeds in fish and simple diffusion can no longer sustain increasing O2 and metabolic requirements, reliance on cutaneous respiration decreases and the developing gills become the primary site of gas exchange. In zebrafish, for example, complete formation of the gills occurs relatively late in development, and larvae become completely reliant on the gills for respiration at about 21 dpf (Rombough, 2002; Jonz and Nurse, 2006). In light of the changes in sites of gas exchange that occur in aquatic vertebrates during development, it is perhaps not surprising that identification of O2-sensing sites during these early stages is not straightforward. Zebrafish embryos exposed to anoxia at or before 1 dpf survive by entering a state of suspended animation, where even cell division is arrested (Padilla and Roth, 2001). However, at 2 dpf anoxia induces mortality (Padilla and Roth, 2001) and zebrafish embryos begin to respond to decreased O2­with rapid pectoral fin movements to increase cutaneous gas exchange ( Jonz and Nurse, 2005, 2006). At 3 dpf, hypoxia induces an increase in the frequency of opercular and gill movements and the adult-like hyperventilatory response begins to develop (Figure 11A). However, confocal studies have shown that O2 chemoreceptors of the developing gills, which begin to appear at 3 dpf, are not innervated until about 7 dpf, and innervation coincides with a significant rise in the hyperventilatory response to hypoxia (Figure 11A; Jonz and Nurse, 2005). Therefore, these data indicate that a population of O2 chemoreceptors must exist outside the respiratory regions of the gill (i.e. extrabranchial) and be responsible for inducing hypoxic reflexes before O2sensing pathways of the gill develop. Though such extrabranchial O2 chemoreceptors of developing or adult aquatic vertebrates have not been described, potential sites may include the oropharyngeal cavity (Milsom et al., 2002; Burleson and Milsom, 2003), the non-respiratory regions of the gill arches, and possibly the skin of embryos and larvae ( Jonz and Nurse, 2006). The development of central pathways that receive sensory afferents from peripheral O2 chemoreceptors, and regulate appropriate cardiorespiratory responses, is not well defined in aquatic vertebrates. In zebrafish, regular ventilatory or buccal movements do not develop until after 8 dpf, despite formation of functional O2-sensing mechanisms in the gill during the first week ( Jonz and Nurse, 2005). This suggests that adult-like central pathways that control cardiorespiratory reflexes are not functional during these early stages. Interestingly, Turesson et al. (2006) showed that the hyperventilatory response to acute hypoxia was completely inhibited in larvae by exogenous application of MK801, a blocker of ionotropic N-methyl-d-aspartate (NMDA) receptors, beginning at 13 dpf. While some sensitivity to MK801 developed after 8 dpf, the drug was without effect during earlier stages (Figure 11B). These results suggest that a central, glutamatergic NMDA-receptor-dependent mechanism forms relatively late in zebrafish development and may play a role in the central integration of peripheral O2-chemosensory responses (Turesson et al., 2006). In adult catfish, regions of the medulla homologous to the mammalian nucleus tractus solitarius (NTS) receive central projections of the branchial nerves (Kanwal and Caprio, 1987). In addition, injection

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  23

Figure 11: Development of the hypoxic response in zebrafish (Danio rerio). A. Ventilation frequency in anaesthetized zebrafish larvae exposed to acute hypoxia (25 mmHg, Hox) compared to control (Cont). The hyperventilatory response to hypoxia is present at 3 days postfertilization (dpf ) and increases significantly at 7 dpf. Modified from Jonz and Nurse (2005). B. Inhibition of the hyperventilatory response to acute hypoxia (30 mmHg) in anaesthetized zebrafish following application of the ionotropic glutamate receptor (NMDA) blocker, MK801. Sensitivity of the hypoxic hyperventilatory response to MK801 developed between 8 and 13 dpf, suggesting development of central glutamatergic neurotransmission during this time. Modified from Turesson et al. (2006).

of glutamate into these areas induces cardiorespiratory reflexes that mimic hypoxic stimulation (Sundin et al., 2003a), and pharmacological and immunohistochemical studies have demonstrated the role of central ionotropic NMDA receptors and a potential glutamatergic pathway involved in cardiorespiratory control (Sundin et al., 2003b; Turesson and Sundin, 2003).

Summary of Potential O2 Sensing Pathways of the Gill From the available morphological and physiological evidence, it appears as though NECs of the gills in aquatic vertebrates act as peripheral O2 sensors that initiate cardiorespiratory reflexes during exposure to acute hypoxia. Decreases in either arterial or water Po2 may decrease O2 tension in the surrounding epithelial tissue of the gill, and lead to inhibition of O2-sensitive leak K+ channels and subsequent membrane depolarization. This cellular response is expected to lead to Ca2+-dependent release of neurotransmitters contained within synaptic vesicles, such as 5-HT, and activation of postsynaptic sensory nerve fibres. As summarized in Figure 8C, NECs of the zebrafish gill filaments are innervated by two populations of nerve fibres: those with cell bodies intrinsic or extrinsic to the gill. Extrinsic nerve fibres may originate from cranial nerve

24  |  Airway Chemoreceptors in the Vertebrates ganglia (i.e. petrosal or nodose ganglia) and relay sensory information centrally to respiratory regions of the medulla, thus controlling ventilatory changes in response to hypoxia. The intrinsic innervation of NECs is more likely to modulate local vascular changes in the gill during periods of hypoxic stress, since nerve fibres of intrinsic neurons also terminate at proximal regions of the efferent filament artery ( Jonz and Nurse, 2003). The diffuse organization of NECs in the gills of fish and amphibians suggests that, in addition to stimulating cardiorespiratory responses to hypoxia, these cells may detect small gradients in water or arterial Po2 to induce local changes in vascular resistance that may result in the diversion of blood flow to better ventilated areas of the gill ( Jonz and Nurse, 2003). Such a system may act along with the process of lamellar recruitment, which increases perfusion of distal lamellae during periods of hypoxia to increase respiratory surface area (Booth, 1978, 1979; Sundin and Nilsson, 1997), thereby maintaining balance of the ventilation-perfusion ratio. The remaining steps in the cellular response to hypoxia, including the molecular identification of the O2 sensor itself, have not been investigated in aquatic vertebrates, and are at present controversial in mammalian systems. While a universal O2 sensor has not been described in vertebrates, current models have suggested that a plasma membrane or intracellular heme protein may be involved in O2 sensing and regulate ion channel activity in the carotid body (López-Barneo et al., 2001). It was proposed that heme oxygenase-2 (HO-2) may be the O2 sensor that couples to Ca2+-dependent K+ channels (Williams et al., 2004), however recent studies report that O2 sensitivity is preserved in carotid body type I cells and adrenal medullary chromaffin cells in HO-2 knockout mice (Ortega-Sáenz et al., 2006). Alternatively, the redox hypothesis proposes that reactive oxygen species produced by NADPH oxidase, or the mitochondrial electron transport chain, regulate ion channel activity during hypoxia (López-Barneo et al., 2001). Plasma membrane NADPH oxidase appears to act as an O2 sensor in pulmonary NEBs (Fu et al., 2000), but not in myocytes, type I cells or chromaffin cells (Archer et al., 1999; He et al., 2002; Thompson et al., 2002). More recently, AMPactivated protein kinase (AMPK) has been proposed as the key mediator of hypoxic chemotransduction in the carotid body, where it appears to couple inhibition of mitochondrial phosphorylation to that of O2-sensitive K+ channels (Wyatt and Evans, 2007). Thus, there appears to be a diversity of mechanisms in vertebrates by which O2 sensitivity may be conferred in peripheral chemoreceptors. Aquatic vertebrates, especially the zebrafish, represent attractive models for pursuing studies of many aspects of O2 sensing, from cellular mechanisms to behavioural responses. Further electrophysiological characterization of gill NECs, and the use of zebrafish mutants with perturbations in O2 chemoreception, may provide important clues about the specific roles of ion channels in O2 sensing, and perhaps an ubiquitous O2 sensor in vertebrates. In addition, further confocal studies may lead to identification of the neurochemical basis of chemosensory signalling in the gill, and localization of O2 chemoreceptors during early embryonic stages. Generally, such studies will provide

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  25

important information about O2 sensing and respiratory regulation in aquatic species, and how this system has evolved in vertebrates.

References Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat Rec. 270A:25-40. Archer, S.L., Reeve, H.L., Michelakis, E., Puttagunta, L., Waite, R., Nelson, D.P., Dinauer, M.C., and Weir, E.K. 1999. O2 sensing is preserved in mice lacking the gp91 phox subunit of NADPH oxidase. Proc Natl Acad Sci USA. 96:7944-7949. Bailly, Y., and Dunel-Erb, S. 1986. The sphincter of the efferent filament artery in teleost gills: I. Structure and parasympathetic innervation. J Morphol. 187:219-237. Bailly, Y., Dunel-Erb, S., Geffard, M., and Laurent, P. 1989. The vascular and epithelial serotonergic innervation of the actinopterygian gill filament with special reference to the trout, Salmo gairderi. Cell Tissue Res. 258:349-363. Bailly, Y., Dunel-Erb, S., and Laurent, P. 1992. The neuroepithelial cells of the fish gill filament: indolamine-immunocytochemistry and innervation. Anat Rec. 233:143-161. Bairam, A., and Carroll, J.L. 2005. Neurotransmitters in carotid body development. Respir Physiol Neurobiol. 149:217-232. Bisgard, G.E. 2000. Carotid body mechanisms in acclimatization to hypoxia. Resp Physiol. 121:237-246. Booth, J.H. 1978. The distribution of blood flow in the gills of fish: application of a new technique to rainbow trout (Salmo gairdneri). J Exp Biol. 73:119-129. Booth, JH. 1979. The effects of oxygen supply, epinephrine, and acetylcholine on the distribution of blood flow in trout gills. J Exp Biol 83:31-39. Buckler, K.J. 1997. A novel oxygen-sensitive potassium current in rat carotid body type I cells. J Physiol. 498:649-662. Buckler, K.J. 2007. TASK-like potassium channels and oxygen sensing in the carotid body. Resp Physiol Neurobiol. 157:55-64. Buckler, K.J., Williams, B.A., and Honoré, E. 2000. An oxygen-, acid- and anaesthetic-sensitive TASK-like background potassium channel in rat arterial chemoreceptor cells. J Physiol. 525:135-142. Burggren, W.W., and Doyle, M. 1986. Ontogeny of regulation of gill and lung ventilation in the bullfrog, Rana catasbeiana. Respir Physiol. 66:279-291. Burggren, W.W., and Pinder, A.W. 1991. Ontogeny of cardiovascular and respiratory physiology in lower vertebrates. Annu Rev Physiol. 53:107-135. Burleson, M.L. and Smatresk, N.J. 1990. Effects of sectioning cranial nerves IX and X on cardiovascular and ventilatory reflex responses to hypoxia and NaCN in channel catfish. J Exp Biol. 154:407-420 Burleson, M.L., and Milsom, W.K. 1993. Sensory receptors in the first gill arch of rainbow trout. Resp Physiol 93:97-110. Burleson, M.L., and Milsom, W.K. 1995a. Cardio-ventilatory control in rainbow trout: I. Pharmacology of branchial, oxygen-sensitive chemoreceptors. Resp Physiol. 100:231-238. Burleson, M.L., and Milsom, W.K. 1995b. Cardio-ventilatory control in rainbow trout: II. Reflex effects of exogenous neurochemicals. Resp Physiol. 101:289-299.

26  |  Airway Chemoreceptors in the Vertebrates Burleson, M.L., and Milsom, W.K. 2003. Comparative aspects of O2 chemoreception: anatomy, physiology, and environmental adaptations. In: Oxygen Sensing: Responses and Adaptation to Hypoxia, S. Lahiri, G.L. Semenza, and N.R. Prabhakar (Eds.). Marcel Dekker, New York. pp. 685-707. Burleson, M.L., Smatresk, N.J., and Milsom, W.K. 1992. Afferent inputs associated with cardioventilatory control in fish. In: Fish Physiology, Vol. XIIB, W.S. Hoar, D.J. Randall, and A.P. Farrell (Eds.). Academic Press, San Diego. pp. 389-426. Burleson, M.L., Carlton, A.L., and Silva, P.E. 2002. Cardioventilatory effects of acclimatization to aquatic hypoxia in channel catfish. Resp Physiol Neurobiol. 131:223-232. Burleson, M.L., Mercer, S.E., and Wilk-Blaszczak, M.A. 2006. Isolation and characterization of putative O2 chemoreceptor cells from the gills of channel catfish (Ictalurus punctatus). Brain Res. 1092:100-107. Campanucci, V.A., and Nurse, C.A. 2007. Autonomic innervation of the carotid body: role of efferent inhibition. Resp Physiol Neurobiol. 157:83-92. Campanucci, V.A., Fearon, I.M., and Nurse, C.A. 2003. A novel O2-sensing mechanism in rat glossopharyngeal neurones mediated by a halothane-inhibitable background K+ conductance. J Physiol. 548:731-743. Cutz, E., and Jackson, A. 1999. Neuroepithelial bodies as airway oxygen sensors. Resp Physiol. 115:201-214. Cutz, E., Speirs, V., Yeger, H., Newman, C., Wang, D., and Perrin, D.G. 1993. Cell biology of pulmonary neuroepithelial bodies—validation of an in vitro model. I. Effects of hypoxia and Ca2+ ionophore on serotonin content and exocytosis of dense core vesicles. Anat Rec. 236:41-52. de Graaf, P.J.F. 1990. Innervation pattern of the gill arches and gills of the carp (Cyprinus carpio). J Morphol. 206:71-78. Donald, J.A. 1984. Adrenergic innervation of the gills of brown and rainbow trout, Salmo trutta and S. gaidneri. J Morphol. 182:307-316. Donald, J.A. 1987. Comparative study of the adrenergic innervation of the teleost gill. J Morphol. 193:63-73. Donnelly, D.F. 2005. Development of carotid body/petrosal ganglion response to hypoxia. Respir Physiol Neurobiol. 149:191-199. Dunel-Erb, S., and Bailly, Y. 1986. The sphincter of the efferent filament artery in teleost gills: II. Sympathetic innervation. J Morphol. 187:239-246. Dunel-Erb, S., Bailly, Y., and Laurent, P. 1982. Neuroepithelial cells in fish gill primary lamellae. J Appl Physiol Resp Environ Exercise Physiol. 53:1342-1353. Dunel-Erb, S., Bailly, Y., and Laurent, P. 1989. Neurons controlling the gill vasculature in five species of teleosts. Cell Tissue Res. 255:567-573. Evans, D.H., Piermarini P.M., and Choe, K.P. 2005. The multifunctional fish gill: dominant site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous waste. Physiol Rev. 85:97-177. Feldman, J.L., Mitchell, G.S., and Nattie, E.E. 2003. Breathing: rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci. 26:239-266. Fu, X.W., Wang, D., Nurse, C.A., Dinauer, M.C., and Cutz, E. 2000. NADPH oxidase is an O2 sensor in airway chemoreceptors: evidence from K+ current modulation in wild-type and oxidase-deficient mice. Proc Natl Acad Sci USA. 97:4374-4379. Fu, X.W., Nurse, C.A., Wong, V., and Cutz, E. 2002. Hypoxia-induced secretion of serotonin from intact pulmonary neuroepithelial bodies in neonatal rabbit. J Physiol. 539: 503-510. Gilbert, S.F. 1988. Developmental Biology. 2nd Ed. Sinauer Associates Inc., Massachusetts.

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  27 Gilmour, K.M. 2001. The CO2/pH ventilatory drive in fish. Comp Biochem Physiol Mol Integr Physiol. 130:A219-A240. Goldstein, S.A.N., Backenhauer, D., O’Kelly, I., and Zilberberg, N. 2001. Potassium leak channels and the KCNK family of two-P-domain subunits. Nat Rev Neurosci. 2:1-11. Goniakowska-Witaliñska, L., Zaccone, G., and Fasulo, S. 1993. Immunocytochemistry and ultrastructure of the solitary neuroepithelial cells in the gills of the neotenic tiger salamander Ambystoma tigrinum (Urodela, Amphibia). Eur Arch Biol. 104:45-50. González, C., Almaraz, L., Obeso, A., and Rigual, R. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol Rev. 74:829-898. He, L., Chen, J., Dinger, B., Sanders, K., Sundar, K., Hoidal, J., and Fidone, S. 2002. Characteristics of carotid body chemosensitivity in NADPH oxidase-deficient mice. Am J Physiol Cell Physiol. 282:C27-C33. Hughes, G.M. 1972. Morphometrics of fish gills. Resp Physiol. 14:1-25. Hughes, G.M. 1984. General anatomy of the gills. In: Fish Physiology, Vol. XA, W.S. Hoar, and D.J. Randall (Eds.).Academic Press, San Diego. pp. 1-72. Jackson, A., and Nurse, C.A. 1997. Dopaminergic properties of cultured rat carotid body chemoreceptors grown in normoxic and hypoxic environments. J Neurochem. 69:645-654. Jacob, E., Drexel, M., Schwerte, T., and Pelster, B. 2002. Influence of hypoxia and of hypoxemia on the development of cardiac activity in zebrafish larvae. Am J Physiol Regul Integr Comp Physiol. 283:R911-R917. Jia, X.X., and Burggren, W.W. 1997a. Developmental changes in chemoreceptive control of gill ventilation in larval bullfrogs (Rana catesbeiana): I. reflex ventilatory responses to ambient hyperoxia, hypoxia and NaCN. J Exp Biol. 200:2229-2236. Jia, X.X., and Burggren, W.W. 1997b. Developmental changes in chemoreceptive control of gill ventilation in larval bullfrogs (Rana catesbeiana): II. sites of O2-sensitive chemoreceptors. J Exp Biol. 200:2237-2248. Johnston, I.A., Bernard, L.M., and Maloiy, G.M. 1983. Aquatic and aerial respiration rates, muscle capillary supply and mitochondrial volume density in the air-breathing catfish Clarias mossambicus acclimated to either aerated or hypoxic water. J Exp Biol. 105:317-338. Jonz, M.G., and Nurse, C.A. 2003. Neuroepithelial cells and associated innervation of the zebrafish gill: a confocal immunofluorescence study. J Comp Neurol. 461:1-17. Jonz, M.G., and Nurse, C.A. 2005. Development of oxygen sensing in the gills of zebrafish. J Exp Biol. 208:1537-1549. Jonz, M.G., and Nurse, C.A. 2006. Ontogenesis of oxygen chemoreception in aquatic vertebrates. Resp Physiol Neurobiol. 154:139-152. Jonz, M.G., Fearon, I.M., and Nurse, C.A. 2004. Neuroepithelial oxygen chemoreceptors of the zebrafish gill. J Physiol. 560:737-752. Kanwal, J.S., and Caprio, J. 1987. Central projections of the glossopharyngeal and vagal nerves in the channel catfish, Ictalurus punctatus: clues to differential processing of visceral inputs. J Comp Neurol 264:216-230. Kerstens, A., Lomholt, J.P., and Johansen, K. 1979. The ventilation, extraction and uptake of oxygen in undisturbed flounders, Platichthys flesus: responses to hypoxia acclimation. J Exp Biol. 83:169-179. Kusakabe, T. 2002. Carotid labyrinth of Amphibians. Microsc Res Tech. 59:207-226. Kusakabe, T., and Kawakami, T. 1992. Distribution of CGRP, substance P, VIP and somatostatin immunoreactive nerve fibres in the internal gills of larvae of the bullfrog, Rana catesbeiana. Arch Histol Cytol. 55:243-249.

28  |  Airway Chemoreceptors in the Vertebrates Kusakabe, T., Kawakami, T., Hori, H., Bandou, Y., and Takenaka, T. 1993. Immunohistochemical coexistence of calcitonin gene-related peptide and substance P in the nerve fibers of the internal gills of bullfrog (Rana catesbeiana). Neurosci Lett. 158:59-62. Lahiri, S., Roy, A., Baby, S.M., Hoshi, T., Semenza, G.L., and Prabhakar, N.R. 2006. Oxygen sensing in the body. Prog Biophys Mol Biol. 91:249-286. Laurent, P. 1984. Gill internal morphology. In: Fish Physiology, Vol. XA. W.S. Hoar, and D.J. Randall (Eds.). Academic Press, San Diego. pp 73-183. Laurent, P., and Dunel, S. 1980. Morphology of gill epithelial in fish. Am J Physiol Reg Integr Comp Physiol. 238:R147-R159. Lesage, F. 2003. Pharmacology of neuronal background potassium channels. Neuropharmacology. 44:1-7. Lesage, F., and Lazdunski, M. 2000. Molecular and functional properties of two-pore-domain potassium channels. Am J Physiol Renal Physiol. 279:F793-F801. López-Barneo, J., Pardal, R., and Ortega-Sáenz, P. 2001. Cellular mechanisms of oxygen sensing. Annu Rev Physiol. 63:259-287. Malvin, G.M. 1989. Gill structure and function: amphibian larvae. In: ComparativePulmonary Physiology: Current Concepts, S.C. Wood (Ed.). Marcel Dekker, New York. pp. 121-151. Malvin, G.M., and Dail, W.G. 1986. Adrenergic innervation of the gills, pulmonary arterial plexus, and dorsal aorta in the neotenic salamander, Ambystoma tigrinum. J Morphol. 189:67-70. Mills, L., and Nurse, C.A. 1993. Chronic hypoxia in vitro increases volume of dissociated carotid body chemoreceptors. NeuroReport. 4:619-622. Milsom, W.K. 2002. Phylogeny of CO2/H+ chemoreceptors in vertebrates. Resp Physiol Neurobiol. 131:29-41. Milsom, W.K., and Brill, R.W. 1986. Oxygen sensitive afferent information arising from the first gill arch of yellowfin tuna. Resp Physiol. 66:193-203. Milsom, W.K., and Burleson, M.L. 2007. Peripheral arterial chemoreceptors and the evolution of the carotid body. Resp Physiol Neurobiol. 157:4-11. Milsom, W.K., Reid, S.G., Rantin, F.T., and Sundin, L. 2002. Extrabranchial chemoreceptors involved in respiratory reflexes in the neotropical fish Colossoma macropomum (the tambaqui). J Exp Biol. 205:1765-1774. Minnich, B., Bartel, H., and Lametschwandtner, A. 2002. How a highly complex threedimensional network of blood vessels regresses: the gill blood vascular system of tadpoles of Xenopus during metamorphosis. A SEM study on microvascular corrosion casts. Microvasc Res. 64:425-437. Montpetit C.J., and Perry, S.F. 1998. The effects of chronic hypoxia on the acute adrenergic stress response in the rainbow trout (Oncorhynchus mykiss). Physiol Zool. 71:377-386. Neubauer, J.A., and Sunderram, J. 2004. Oxygen-sensing neurons in the central nervous system. J Appl Physiol. 96:367-374. Nilsson, S. 1984. Innervation and pharmacology of the gills. In: Fish Physiology, Vol. XA, W.S. Hoar, and D.J. Randall (Eds.). Academic Press, San Diego. pp. 185-227. Nurse, C.A.. 2005. Neurotransmission and neuromodulation in the chemosensory carotid body. Autonom Neurosci: Basic and Clinical. 120:1-9. Nurse, C.A. and Vollmer, C. 1997. Role of basic FGF and oxygen in control of proliferation, survival, and neuronal differentiation in carotid body chromaffin cells. Dev Biol. 184:197-206. O’Kelly, I., Stephens, R.H., Peers, C., and Kemp, P.J. 1999. Potential identification of the O2-sensitive K+ current in a human neuroepithelial body-derived cell line. Am J Physiol Lung Cell Mol Physiol. 276: L96-L104.

Oxygen-sensitive Neuroepithelial Cells in the Gills of Aquatic Vertebrates  |  29 Olson, K.R. 2002. Vascular anatomy of the fish gill. J Exp Zool. 293:214-231. Ortega-Sáenz, P., Pascual, A., Gómez-Díaz, R., and López-Barneo, J. 2006. Acute oxygen sensing in heme oxygenase-2 null mice. J Gen Physiol. 128:405-411. Padilla, P.A., and Roth, M.B. 2001. Oxygen deprivation causes suspended animation in the zebrafish embryo. Proc Natl Acad Sci USA. 98:7331-7335. Peers, C. 1997. Oxygen-sensitive ion channels. Trends Pharmacol Sci. 18:405-408. Pelster, B., and Burggren, W.W. 1996. Disruption of hemoglobin oxygen transport does not impact oxygen-dependent physiological processes in developing embryos of zebra fish (Danio rerio). Circ Res. 79:358-362. Perry, S.F., and Gilmour, K.M. 2002. Sensing and transfer of respiratory gases at the fish gill. J Exp Zool. 293:249-263. Plant, L.D., Kemp, P.J., Peers, C., Henderson, Z., and Pearson, H.A. 2002. Hypoxic depolarization of cerebellar granule neurons by specific inhibition of TASK-1. Stroke. 33:2324-2328. Powell, F.L. 2007. The influence of chronic hypoxia upon chemorcception. Resp Physiol Neurobiol. 157:154-161. Prabhakar, N.R. 2006. O2 sensing at the mammalian carotid body: why multiple O2 sensors and multiple transmitters? Exp Physiol. 91:17-23. Putnam, R.W., Filosa, J.A., and Ritucci N.A. 2004. Cellular mechanisms involved in CO2 and acid signalling in chemosensitive neurons. Am J Physiol Cell Physiol. 287:C1493-C1526. Randall, D.J., and Perry, S.F. 1992. Catecholamines. In: Fish Physiology, Vol. XIIB. W.S. Hoar, D.J. Randall, and A.P. Farrell (Eds.). Harcourt Brace Jovanovich, San Diego. pp. 255-300. Reid, S.G., and Perry, S.F. 2003. Peripheral O2 chemoreceptors mediate humoral catecholamine secretions from fish chromaffin cells. Am J Physiol Regul Integr Comp Physiol. 284:R990-R999. Rombough, P.J. 1988. Respiratory gas exchange, aerobic metabolism, and effects of hypoxia during early life. In: Fish Physiology, Vol. XIA. W.S. Hoar, and D.J. Randall (Eds.). Academic Press, New York. pp. 59-161. Rombough, P.J. 2002. Gills are needed for ionoregulation before they are needed for O2 uptake in developing zebrafish, Danio rerio. J Exp Biol. 205:1787-1794. Rugh, R., 1951. The Frog: its Reproduction and Development. McGraw-Hill, New York. Sadler, T.W. 2004. Langman’s Medical Embryology. 9th Ed. Lippincott, Williams and Wilkins, Philadelphia. Saltys, H.A., Jonz, M.G., and Nurse, C.A. 2006. Comparative study of gill neuroepithelial cells and their innervation in teleosts and Xenopus tadpoles. Cell Tissue Res. 323:1-10. Stea, A., Jackson, A., and Nurse, C.A. 1992. Hypoxia and N6,O2’-dibutyryladenosine 3’,5’-cyclic monophosphate, but not nerve growth factor, induce Na+ channels and hypertrophy in chromaffin-like arterial chemoreceptors. Proc Natl Acad Sci USA. 89:9469-9473. Straus, C., Wilson, R.J., and Remmers, J.E. 2001. Oxygen sensitive chemoreceptors in the first gill arch of the tadpole, Rana catesbeiana. Can J Physiol Pharmacol. 79:959-962. Sundin, L., and Nilsson, G.E. 1997. Neurochemical mechanisms behind gill microcirculatory responses to hypoxia in trout: in vivo microscopy study. Am J Physiol Reg Integr Comp Physiol. 272:R576-R585. Sundin, L., and Nilsson, S. 2002. Branchial innervation. J Exp Zool. 293:232-248. Sundin, L., Reid, S.G., Rantin, F.T., and Milsom, W.K. 2000. Branchial receptors and cardiorespiratory reflexes in a neotropical fish, the tambaqui (Colossoma macropomum). J Exp Biol. 203:1225-1239. Sundin, L., Turesson, J., and Taylor, E.W. 2003a. Evidence for glutamatergic mechanisms in the vagal sensory pathway initiating cardiorespiratory reflexes in the shorthorn sculpin Myoxocephalus scorpius. J Exp Biol. 206:867-876.

30  |  Airway Chemoreceptors in the Vertebrates Sundin, L., Turesson, J., and Burleson, M. 2003b. Identification of central mechanisms vital for breathing in the channel catfish, Ictalurus punctatus. Resp Physiol Neurobiol. 138:77-86. Territo, P.R., and Burggren, W.W. 1998. Cardio-respiratory ontogeny during chronic carbon monoxide exposure in the clawed frog Xenopus laevis. J Exp Biol. 201:1461-1472. Thompson, R.J., Farragher, S.M., Cutz, E., and Nurse, C.A. 2002. Developmental regulation of O2 sensing in neonatal adrenal chromaffin cells from wild-type and NADPH-oxidasedeficient mice. Pflügers Arch 444:539-548. Turesson, J., and Sundin, L. 2003. N-methyl-D-aspartate receptors mediate chemoreflexes in the shorthorn sculpin Myoxocephalus scorpius. J Exp Biol. 206:1251-1259. Turesson, J., Schwerte, T., and Sundin, L. 2006. Late onset of NMDA receptor-mediated ventilatory control during early development in zebrafish (Danio rerio). Comp Biochem Physiol A. 143:332-339. Vulesevic, B., McNeil, B., and Perry, S.F. 2006. Chemoreceptor plasticity and respiratory acclimation in the zebrafish Danio rerio. J Exp Biol. 209:1261-1273. Wang, Z.Y., and Bisgard, G.E. 2002. Chronic hypoxia-induced morphological and neurochemical changes in the carotid body. Microsc Res Tech. 59:168-177. Weichert, C.K. 1967. Elements of Chordate Anatomy. McGraw-Hill, New York. Weir, K.E., López-Barneo, J., Buckler, K.J., and Archer, S.L. 2007. Acute oxygen-sensing mechanisms. N Engl J Med. 353:2042-2055. Wells, P.R., and Pinder, A.R. 1996. The respiratory development of Atlantic salmon. II. Partitioning of oxygen uptake among gills, yolk sac and body surfaces. J Exp Biol. 199:2737-2744. Williams, S.E., Wootton, P., Mason, H.S., Bould, J., Iles, D.E., Riccardi, D., Peers, C., and Kemp, P.J. 2004. Hemoxygenase-2 is an oxygen sensor for a calcium-sensitive potassium channel. Science. 306:2093-2097. Wilson, J.M., and Laurent, P. 2002. Fish gill morphology: inside out. J Exp Zool. 293:192-213. Wyatt, C.N., and Evans, A.M. 2007. AMP-activated protein kinase and chemotransduction in the carotid body. Resp Physiol Neurobiol. 157:22-29. Wyatt, C.N., Wright, C., Bee, D., and Peers, C. 1995. O2-sensitive K+ currents in carotid body chemoreceptor cells from normoxic and chronically hypoxic rats and their roles in hypoxic chemotransduction. Proc Natl Acad Sci USA. 92:295-299. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature 365:153-155. Zaccone, G., Fasulo, S., and Ainis, L. 1994. Distribution patterns of the paraneuronal endocrine cells in the skin, gills and the airways of fishes as determined by immunohistochemical and histological methods. Histochem J. 26:609-629. Zaccone, G., Fasulo, S., Ainis, L., and Licata, A. 1997. Paraneurons in the gills and airways of fishes. Microsc Res Tech. 37:4-12. Zaccone, G., Mauceri, A., and Fasulo, S. 2006. Neuropeptides and nitric oxide synthase in the gill and the air-breathing organs of fishes. J Exp Zool. 305A:428-439.

2 Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells in Indian Catfishes A. Gopesh

Abstract Carotid labyrinth of land vertebrates is known for the last two centuries but that of fishes for the last few decades only. It has been found to occur in catfishes (Siluroidei) alone and there is much in common between carotid labyrinth of fish and amphibian in points of basic structure, topography and innervation. The structure, homology and functions are discussed in detail. The presence of peculiar paraneuronal cells in close vicinity of carotid labyrinth is a significant feature. The fish carotid labyrinth and associated pseudobranchial neurosecretory system, comprising of paraneuronal cells have been described to throw light on the possible function of these structures in the biology of these fishes. In its structure the carotid labyrinth seems to be more elaborate then that of higher vertebrates and its association with neurosecretory cells in these groups of fishes suggest a possible chemosensory function for the former and a paracrine role for the latter. Although an in depth investigation is needed to reach any conclusion, these cells seem to play a role in conditions of hypoxia, signalling the surfacing behaviour of these fishes, which is a common behavioural feature of fishes of this group. Nothing can be said with certainty at present, but the presence of pseudobranchial neurosecretory cells in close proximity of carotid labyrinth and efferent branchial vessel suggests that respiration in these fishes is a much more complicated process than was thought earlier.   More in depth investigations on the carotid labyrinth and associated pseudobranchial neurosecretory are likely to reveal more facts about the biology of catfishes that will help in understanding the nature of its derivatives in higher vertebrates, including carotid body of mammals.

Department of Zoology, University of Allahabad, Allahabad - 211002, India. Email: anita_gopesh@ yahoo.co.in

32  |  Airway Chemoreceptors in the Vertebrates Keywods: Carotid labyrinth, pseudobranch, paraneurons, neurosecretory cells, microcirculation, vascular channels, catfishes, hypoxia, serotonin

Introduction Carotid labyrinth—a characteristic swelling of the common carotid, at the junction of the external and internal carotid artery in amphibians was highlighted for the first time by Swammerdam (1738). Initially it was believed to be an emergent structure of land vertebrates which made its first appearance in Amphibia (Kingsley, 1926). In subsequent years, the carotid labyrinth and similar homologous structure were studied in great detail in amphibians and higher tetrapods (de Kock, 1954; Carman, 1955; Adams, 1958; Abraham, 1969; Biscoe, 1971; Fujita et al., 1980; Kusakabe, 2002). These studies established that in higher vertebrates the carotid labyrinth has undergone great evolutionary modification in morphology and is recognized as a carotid body (Adams, 1958; Biscoe, 1971). Meanwhile, sporadic reports regarding the presence of carotid labyrinth in a few catfishes also appeared (Singh, 1959, 1960; Vashisht and Kapoor, 1965). Later on it was established that a well developed carotid labyrinth is ubiquitously present in all Table 1:

Catfish species in which the carotid labyrinth is reported (Modified from Prakash, 1993). Fishes

Authors

1. Clarias batrachus

Srivastava and Singh, 1980; Olson et al., 1981; Singh, 1982; Vashisht and Kapoor, 1965; Srivastava et al., 1993; Prakash, 1993.

2. Heteropneustes fossilis

Srivastava and Singh, 1980; Olson et al., 1990; Singh, 1982; Munshi and Hughes, 1987; Srivastava et al., 1993.

3. Mystus seenghala

Singh, 1959; Srivastava and Singh, 1980; Singh, 1982; Prakash, 1993.

4. Rita rita

Singh, 1959; Vashisht and Kapoor, 1965; Srivastava and Singh, 1980; Singh, 1982; Srivastava et al., 1993; Prakash, 1993.

5. Clupisoma garua

Vashisht and Kapoor, 1965; Srivastava and Pandey, 1984b; Pandey, 1985; Prakash, 1993.

6. Wallago attu

Singh, 1959, 1960; Vashisht and Kapoor, 1965; Srivastava and Pandey, 1984b; Srivastava et al., 1988; Srivastava et al., 1993.

7. Eutropiichthys vacha

Vashisht and Kapoor, 1965; Srivastava and Pandey, 1984b; Pandey, 1985; Srivastava et al., 1993. (Table 1 contd.)

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  33

(Table 1 contd.) 8. Silonia silondia

Srivastava and Pandey, 1984b; Pandey, 1985.

9. Mystus aor

Singh, 1959, 1960; Tripathi, 1985.

10. Mystus menoda

Singh, 1959.

11. Ompak pabda

Tripathi, 1985; Srivastava et al., 1987; Srivastava et al., 1993.

12. Tachysurus thalassinus

Srivastava and Pandey, 1984a; Pandey, 1985; Srivastava et al., 1993.

13. Tachysurus maculatus

Srivastava and Pandey, 1984a; Pandey, 1985; Srivastava et al., 1993.

14. Arius caelatus

Pandey, 1985.

15. Arius subrostratus

Pandey, 1985.

16. Glyptothorax pectinopterus

Pandey, 1985.

17. Pseudoecheneis sulcatus Pandey, 1985. 18. Ictalurus punctatus

Olson et al., 1981.

19. Ictalurus melas

Olson et al., 1981.

20. Plotosus canius

Gopesh et al., 2002.

21. Osteogeniosus militaris

Gopesh et al., 2002.

the catfishes among teleosts (see, Srivastava, 2005; Table 1). The discovery of carotid labyrinth in fishes offers an opportunity to understand its origin and evolution in lower vertebrates as it has survived a long history of vertebrate evolution.

Structure The carotid labyrinth in fishes, like amphibians, is a paired spongy enlargement situated at the junction of first efferent branchial artery and lateral dorsal aorta and fed by these (Olson et. al., 1981; Srivastava and Singh, 1981; Singh, 1982; Tripathi, 1985). The external carotid, the internal carotid and ophthalmic arteries arise from this swelling (Figure 1). Basically carotid labyrinth in fishes comprises a mass of capillary like plexuses in which capillary like channels branch and anastomose (Figures 2, 3 & 4). A central chamber is found in the middle of the carotid labyrinth which represents the meeting point of the first efferent branchial artery and the lateral dorsal aorta (Figures 2d). These two arteries form lateral dorsal aorta axis within the carotid labyrinth. The external carotid artery and the internal artery open into the central chamber from opposite sides (Figure 2a) Due to the plastic nature of the external carotid artery and its branches, great variation is present regarding the topographical relationship between head arteries

34  |  Airway Chemoreceptors in the Vertebrates

Figure 1: Schematic drawings, showing the position of the carotid labyrinth (stippled zone) with respect to the first efferent branchial artery (1), the lateral dorsal aorta (l.d.a.) the external carotid artery (e.c.) the internal carotid artery (i.c.) and the ophthalmic artery (o.a.) in Clarias batrachus (a), Heteropneustes fossilis (b), Mystus seenghala (c) and Rita rita (d). (2, 3, 4, 2nd, 3rd and 4th efferent branchial arteries) (Srivastava et al., 1981).

and carotid labyrinth, among catfishes (Srivastava and Pandey, 1984b). Internally the carotid labyrinth shows a highly vascular architecture (Figures 2,3 & 4). The interior of carotid labyrinth comprises of main blood vessels, their bases and blood filled channels of small diameter (Figures 3 & 4). The spongy nature of the carotid labyrinth is due to the presence of a large number of capillaries (Singh, 1959, 1960; Srivastava and Singh, 1980; Singh, 1982; Pandey, 1985; Tripathi, 1985). There is an organized plexus of capillaries and arterioles within carotid labyrinth which are inter connected and are invariably associated with the principal head arteries, namely the external carotid, the internal carotid and the ophthalmic arteries (Figure 5). There are basically two sets of channels; one is confluent with the base of the external carotid artery and the other with that of the base of internal carotid artery. The same basic organization of fish carotid labyrinth is met within freshwater catfishes of the plains, as well as hill-streams and in marine catfishes, with minor differences (Srivastava and Pandey, 1984a; Pandey, 1985; Gopesh et al., 2002). The corrosion replica cast studies on Ictalurus melas, Ictalurus punctatus and Clarias batrachus (Olson et al., 1981) and Heteropneustes fossilis (Munshi and Hughes, 1987; Olson et al., 1990), confirm the histological findings that many small capillaries developing from the head arteries anastomose with each other to form the bulk of the carotid labyrinth.

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  35

Figure 2: Transverse section of carotid labyrinth of (a) Clarias batrachus, showing the common chamber, the first efferent branchial artery, the lateral aorta, the internal carotid artery and the ophthalmic artery along with the capillaries associated with them. x 170 (b) Plotosus canius showing central chamber and plexus of arterioles and vascular channels. x 170 (c) Rita rita showing the lateral aorta and the internal carotid artery along with the capillaries associated with them. x 94 (d) Rita rita showing the first efferent branchial artery and lateral dorsal aorta along with the capillaries associated with them. x 94 (Singh, 1982).

Microcirculation The vascular channels are peculiar insofar as they appear to be neither capillaries, nor sinusoids, but form a system of capillary-like microcirculation approaching arterioles (Figure 6). They seem to arise at several points from the lateral dorsal aorta axis within the body of carotid labyrinth. The vascular plexus is enormous and fills the carotid labyrinth from end to end (Figure 5). Intervascular spaces are generally present and are filled with some cellular elements in varying degrees in different species (Figures 4, 6 & 7). The bases of the internal carotid artery, the external carotid artery, the ophthalmic artery and in some cases the orbital artery, are joined through this plexus to the main dorsal aorta axis within carotid labyrinth (Figure 5). The external and the internal carotid arteries open directly into the lateral dorsal aorta axis within carotid labyrinth, apart from their connections via the vascular channel plexus.

36  |  Airway Chemoreceptors in the Vertebrates

Figure 3: Transverse sections of carotid labyrinth in (a) Rita rita, showing layout of microvascular channels, in relation to the main artery (a) x 352 (b) Magnified view of a portion of Figure (a) showing microvascular channels opening into the artery. Note the thick wall of artery showing tunica media and tunica adventia x 706 (c) Clarias batrachus, showing plexus of microvascular channels extending out to connect one artery with the other x 352 (d) A portion of figure (c) magnified to show immediate repeated branching of the microvascular channels after their origin from the lumen of the main artery. [Note shunt vessels (sh.v.) and secondary vessels (s.v.) in the plexus of carotid labyrinth x 704] (Prakash, 1993).

The wall of the capillaries consists of an endothelial lining and an outer thin layer of fibro-elastic tissue (Figures 6 & 7). The fibro-elastic coat is relatively thin as compared to the thick wall of the main arteries (Prakash, 1993). This coat is common with the main coat of the associated head artery (Figure 5). The branching occurs frequently and each branch carries a small share of the common coat around its endothelial lining (Figures 6 & 7). Such channels, by definition seem to be arterioles (Shepro, 1980). They are distensible insofar as smooth muscles have been located in addition to tortuous fibres associated with them. Sphincter muscles are also found to surround the arterioles occasionally (Prakash, 1993). The outer vascular stroma consists of connective tissue fibre, filling the intervascular spaces (Figure 7). These fibres are continuous with

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  37

Figure 4: Transverse sections of carotid labyrinth of Clupisoma garua (a) Showing the plexus of the blood filled vascular channels that connect the lumen of the lateral dorsal aorta, with that of the internal carotid artery x 100 (b) Magnified view of a portion enclosed by the rectangle in figure (a) x 310 (c) Showing the plexus of vascular channels and outer vascular stroma. Arrow indicates the opening of one of the vascular channels into the lateral dorsal aorta x 100 (d) Magnified view of a portion, showing fibro-elastic coat of the intervascular stroma adjacent to the endothelial lining x 310 (Pandey, 1985).

the fibro-elastic coat of the vascular channels. There is no trace of pseudobranchial epithelial cells, either covering the vascular channels or scattered in the stroma except in Wallago attu (Figure 7b: Srivastava and Pandey, 1984a&b; Pandey, 1985).

38  |  Airway Chemoreceptors in the Vertebrates

Figure 5: Longitudinal section of the carotid labyrinth of Eutropiichthys vacha, showing large plexus of vascular channels across the length of the carotid labyrinth, from end to end. (Note the origin of first efferent branchial artery, lateral dorsal aorta at the proximal end and the ophthalmic artery and the orbital artery at its distal end). x 110 (Pandey, 1985).

Fish and Amphibian Carotid Labyrinth Fish carotid labyrinth is basically similar to that of amphibian carotid labyrinth, in both the basic structure and topography with respect to the aortic arches found in catfishes and amphibians (Pandey, 1985; Singh and Srivastava, 1979). In its internal architecture the piscine carotid labyrinth displays all the three basic components of the amphibian carotid labyrinth, namely the ‘main chamber’ , the ‘external carotid retia’ and the ‘internal carotid retia’ (Carman, 1964, 1967a, 1967b). In addition, there is a

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  39

Figure 6: Cross section of carotid labyrinth of Mystus seenghala (a) Showing the sphincter muscle around the meta-arteriole forming a complete ring profile x 704 (b) Showing complete ring profile of sphincter muscle around the meta-arteriole [Note: A constricted sphincter muscle (arrow) x 1760 (under oil)] (Prakash, 1993).

striking similarity of plan of the aortic arches and head arteries between the catfishes and the amphibians (Figures 8e, 8f ). In both, the internal and external carotid arteries arise from the dorsal part of the third aortic arch. The topographical relationship of the amphibian carotid labyrinth with the third aortic arch on one hand, and the internal and external carotid arteries, on the other, are also shown in case of catfishes (Srivastava and Singh, 1980). Besides the presence of carotid labyrinth, catfishes (Figure 8e) resemble the amphibians (Figure 8f ) in the plan of aortic arches and head arteries and differ from that of the closely-related cyprinids (Figure 8a). The catfishes, both air-breathing and non air-breathing, are, therefore, singular among fishes, not only in having a carotid labyrinth, but in having the amphibian pattern of head arteries too, particularly the

40  |  Airway Chemoreceptors in the Vertebrates

Figure 7: T.S. of the carotid labyrinth of Wallago attu (a) Showing the intervascular stroma, fibre coat, endothelial lining and peculiar cellular tissue x 185 (b) Showing the isolated cell masses, surrounded by the connective tissue fibres of the intervascular stroma x 350

external and the internal carotid arteries—a feature not shown even by the lung fishes (Goodrich, 1930). Thus, it appears that in catfishes (Figure 8e), like in amphibians (Figure 8f ), the plan of head arteries is so designed as to ensure supply of most oxygenated blood to the head region, since the oxygen demand for the brain and eye is greatest among the body organs (Hoffert and Fromm, 1972). It is interesting to note that amphibians with bimodal respiration present the simplest condition among the tetrapods (Smith, 1960). The presence of the amphibian pattern of head arteries and the ubiquitous presence of carotid labyrinth among the catfish group, both in air-breathing catfishes like Clarias batrachus, Heteropneustes fossilis, Rita rita and non air-breathing catfishes like Mystus seenghala, Wallago attu, Clupisoma garua, Eutropiichthys vacha, Plotosus canius and Osteogeniosus. militaris raises an important question. Was the air-breathing habit lost secondarily in majority of modern non air-breathing catfishes or did carotid labyrinth develop irrespective of the air-breathing habit in the catfishes in the course of evolution?

Innervation As in amphibians (Ishii and Oasaki, 1966), fish carotid labyrinth is also innervated by the glossopharyngeal nerve (Singh, 1982). However, terminations of nerve endings have not been located so far in fishes, which may provide an insight into the sensory function of carotid labyrinth.

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  41

Figure 8: Schematic drawings showing the position of the pseudobranch (a-d) or the carotid labyrinth (e-f ) (hatched) with respect to branchial and head arteries in fishes and amphibians. Numbers 1-4 denote correspondence to aortic arches first to four. (Note the shift in the position owing to the supported telescoping of a segment.) (Srivastava et al., 1994).

Homology There are varied opinions regarding the origin and evolution of carotid labyrinth in fishes (see Srivastava, 2005). The pseudobranch of the non-catfish but amphibious teleosts seems to be the most plausible candidate for the homology between the pseudobranch and carotid labyrinth. In catfishes carotid labyrinth is invariably present while the pseudobranch is altogether absent. The pseudobranch in carp is typically situated on the wall of the gill cavity (Figure 9c) but those of amphibious teleosts are situated far away from the gill cavity (Figure 9d), a feature which is very similar to carotid labyrinth. The glandular pseudobranch of non-catfish amphibious teleosts (Figure 9d) is different from the lamellae of covered pseudobranch of Labeo rohita (Figure 9c: Roy et al., 1997). Interestingly, histological investigations of pseudobranch in non-catfish amphibious teleosts viz. Anabas testudineus, Channa punctatus, Notopterus chitala and Notopterus notopterus reveal characteristic intermediate stages of modification into the carotid labyrinth (Munshi and Hughes, 1981; Tripathi, 1985; Srivastava et al., 1988).

42  |  Airway Chemoreceptors in the Vertebrates

Figure 9: Schematic drawings showing the supposed evolutionary course of transformation of the hyoidean hemibranch into teleost pseudobranch and later in turn into the carotid labyrinth: (a) Showing normal position of hyoidean pseudobranch. (b) Hypothetical primitive stage for carotid labyrinth as represented by embryonic condition of pseudobranch found in Labeo rohita (c). (d) As seen in non catfish teleosts, A. testudineus, C. punctatus and N. notopterus. (e) As seen in adult condition of Clarias batrachus and Heteropneustes fossilis (Note the vestigial pseudobranchial cavity ps.c.). (f ) As seen in adult condition of Mystus seenghala and Rita rita (Note the absence of the pseudobranchial cavity). I. Mandibular arch II. Hyoidean arch III. Branchial arch (Tripathi, 1985).

The carotid labyrinths of catfishes share the same topographical relationship with principal head arteries as the pseudobranch in other teleosts (Singh and Srivastava, 1979; Srivastava and Singh, 1980; Singh, 1982; Pandey, 1985). The topographical relationship of the burried glandular pseudobranchs of Anabas testudineus, Channa punctatus, Notopterus chitala and Notopterus notopterus with main head arteries (Figures 8b, 8c, 8d and 9d) shows a definite trend of modification in the direction of condition found in the carotid labyrinth of catfishes (Figures 8e, 9e and 9f ). These stages

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  43

have been referred to as ‘sub-carotid labyrinth’ grade, representing the intermediate transitional stages from pseudobranch to carotid labyrinth (Srivastava et al., 1988). Besides the pattern of aortic arches and head arteries of catfishes also stand out as a link between the pattern of head arteries of generalized bonyfish and the carotid labyrinth possessing amphibians (Singh and Srivastava, 1979). Similarly the microvascular architecture of carotid labyrinth also shares several basic features with the pseudobranch of teleosts (Prakash, 1993). The above mentioned similarities point towards the origin of carotid labyrinth from the pseudobranch. However the absence of pilaster/pillar cells and pseudobranchial epithelial cells in the pseudobranch of Notopterus notopterus, Notopterus chitala and Channa punctatus and its presence in Anabas testudineus (Munshi and Hughes, 1981) and Macropodus opercularis (Ladich, 1985, 1987) does not seem to support the hypothesis of transformation of pseudobranch into carotid labyrinth. According to another view the origin of carotid labyrinth has been attributed to the precursor of pseudobranch, the mandibular gill. The presence of smooth muscle cells and sphincters in the vascular channels of carotid labyrinth (Figure 6), has given way to this hypothesis (Prakash, 1993). The catfish carotid labyrinth is a plexus of small diameter channels, lined with endothelium (Figures 6 & 7) considered only arterioles (Hughes, 1984; Munshi and Hughes, 1987; Olson et al., 1990) or only capillaries (Singh, 1982; Pandey, 1985), or both (Prakash, 1993). These were either held to represent the secondary vessels of Vogel (see Olson et al., 1990) or shunt vessels (Figures 3c&d: see Srivastava et al., 1994). The loss of sheet ‘flow’ type of secondary lamellae and their replacement by arteriolelike capillary-plexus (Figure 10) in the vascular architecture of the carotid labyrinth and that of the gills of dipnoans represent parallel evolution in the same direction. This type of microcirculation in both may have originated from shunt vessels of the basic gill organization, though quite independently of each other. These additional evidences give support to the hypothesis of the evolutionary transformation of pseudobranch into carotid labyrinth, dismissing the arguments against such a view based on the absence of pilaster cells. Thus, the microvascular architecture of the carotid labyrinth was proposed to have evolved by elaboration of the corpus cavernosum (see Laurent, 1974, 1984) and/or marginal channels (Hughes and Munshi, 1979) of the arterio-arterial pathway, and the pseudobranch by retention of the ‘pillar cell sheet flow system’ . However, a prominent corpus cavernosum is found in chondrostean gills, but they also possess a pseudobranch, like holosteans and teleosteans. The corpus cavernosum is also found in extant chondrichthyans, indicating its presence in early gnathostomes. The dipnoan gill (Laurent, 1984) probably shows a different though parallel morphological elaboration of the corpus cavernosum into a plexus of arterioles forming arborescent gills. Evolutionary replacement of secondary gill lamellae into capillaries of rete mirabile seems to be yet another possibility, if the choroid gland of teleost eye is considered to be

44  |  Airway Chemoreceptors in the Vertebrates

Figure 10: Schematic drawing showing a comparison between the basic organization of glandular pseudobranch of the three amphibians teleosts A. testudineus (a) C. punctatus and N. notopterus (b) and that of carotid labyrinth of catfishes (c & d). Histology of secondary lamellae of the pseudobranch as seen in T.S. of pseudobranch in Anabas testudineus (a) C. punctatus and N. notopterus (b) Histology of vascular channels of carotid labyrinth in Wallago attu (c) and in other catfishes (d) (b and c represent intermediate condition) (Srivastava et al., 1988).

a transformed gill as speculated by Goodrich (1930) and Munshi and Hughes (1987). It is likely that the piscine and amphibian carotid labyrinth may have been derived from the same structure though independently. This hypothesis brings catfishes closer to primitive osteichthyans. Regarding the homology of carotid labyrinth Srivastava (2005) has opined that “the matter may be viewed in the light of the fact that the point is of homology between pseudobranch and carotid labyrinth and not between their constituents. When evolutionary transformation of organs takes place, not only function but structure also undergoes changes comprising loss, alteration and modification of existing constituents or replacement of existing structural components with ‘new’ components. This forms the basis of acquisition of new functions. Supposed transformation of the gill into pseudobranch and pseudobranch into carotid labyrinth are good examples of tremendous plasticity of gill of vertebrates. On the matter of homology, ontogenetical investigations of carotid labyrinth (Singh, 1982) has shown similarity in the origin of the structure with that of the pseudobranch (Ladich, 1985). In both, the first appearance is in the form of ‘small capillary loops’’. In addition, to these facts, presence of peculiar neurosecretory cells located in the gill region of certain groups of Indian teleosts (Table 2) including catfishes, is significant

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  45 Table 2: List of teleostean species in which the pseudobranchial neurosecretory system is known to occur. S.No. 1.

Order Siluriformes

Species

Author

Clarias batrachus

Srivastava et al., 1981

2.

Heteropneustes fossilis

Singh, 1982; Gopesh, 1983

3.

Mystus seenghala

4.

Ompak pabda

Srivastava et al., 1981; Gopesh, 1983

5.

Ompak bimaculatus

Gopesh, 1983

6.

Mystus aor

Gopesh, 1983

7.

Mystus cavassius

Gopesh, 1983

8.

Wallago attu

Gopesh, 1983

9.

Clupisoma garua

Gopesh, 1983

10.

Ailia coila

Gopesh, 1983

11.

Eutropiichtys vacha

Gopesh, 1983

12.

Mystus vittatus

Gopesh, 1983

13.

Rita rita

Devi, 1987

14.

Tachysurus thalassinus*

Gopesh, 1983

15.

Tachysurus maculatus*

Gopesh, 1983

16.

Photossus canius*

Gopesh et al., 2002

17.

Osteogeniosis militaris*

Gopesh et al., 2002

18.

Perciformes

Glossogobius giuris

Srivastava et al., 1981; Gopesh, 1983

19.

Clupeiformes

Notopterus notopterus

Srivastava et al., 1981; Gopesh, 1983

Notopterus chitala

Srivastava et al., 1981; Gopesh, 1983

20. 21.

Channiformes

Channa punctatus

Devi, 1987

22.

Atheriformes

Xenentodon cancila

Devi, 1987

* Marine species.

evidence in favour of the proposed evolutionary transformation of the pseudobranch into the carotid labyrinth (Srivastava et al., 1981; Gopesh, 1983). To conclude enough evidence is available to support the view of evolutionary transformation of the pseudobranch into the carotid labyrinth. The major sequence of events envisaged during the transformation are summarized in Figure 10 and Table 3.

Functions The mammalian and avian carotid bodies are well known for their peripheral chemoreceptor function (Abraham, 1969; Biscoe, 1971). Similarity between carotid body and labyrinth suggests that the latter also serves as a chemo and/or baroreceptor. Although

46  |  Airway Chemoreceptors in the Vertebrates Table 3:

1.

Hypothetical sequence of events in the evolutionary transformation of carotid labyrinth from pseudobranch in fishes, as suggested by the comparative anatomy of these structures in amphibious teleosts (Srivastava et al., 1988).

FREE PSEUDOBRANCH STRUCTURES 1. Secondary lamellae with pseudobranchial epithelial layers. 2. Interlamellar space. 3. “Sheet-flow” type of vascular channels 4. Pilaster cells present

2.

COVERED PSEUDOBRANCH (LABEO) & ENCLOSED PSEUDOBRANCH (ANABAS) STRUCTURES RETAINED

STRUCTURES LOST

1. Secondary lamellae with pseudobranchial epithelial layers

1. Interlamellar space

2. “Sheet-flow” type of vascular channels 3. Pilaster cells present 3.

4.

5.

SUB-CAROTID STAGE/ENCLOSED PSEUDOBRANCH (NOTOPTERUS & CHANNA) STRUCTURES RETAINED

STRUCTURES LOST

1. Secondary lamellae with pseudo-branchial space

1. Interlamellar space

2. “Sheet-flow” type of vascular channels

2. Pilaster cells

CAROTID LABYRINTH (WALLAGO) STRUCTURES RETAINED

STRUCTURES LOST

1. “Sheet-flow” type of vascular channels replaced by plexus of vascular channels

1. Interlamellar space

2. Isolated pseudobranchial cell masses

2. Pilaster cells

CAROTID LABYRINTH (OMPAK, TACHYSURUS) & OTHER CATFISHES STRUCTURES RETAINED

STRUCTURES LOST

1. “Sheet-flow” type of vascular channels replaced by plexus of vascular channels

1. Interlamellar space 2. Pilaster cells 3. Pseudobranchial epithelia

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  47

anatomically carotid labyrinth seems ideally suited to regulate respiratory gas exchange or systematic arterial pressure, it is presumed that it serves another as yet unidentified function, in conjunction with or in lieu of chemo-baroreception in catfishes (Olson et al., 1981). Based on the investigations carried on the carotid labyrinth, so far, the likely functions of carotid labyrinth can be enumerated as: 1. Regulation of the improvement in quality and pressure of blood, supplied to the eye and brain (Tripathi, 1985; Munshi and Hughes, 1987; Srivastava et al., 1994).   The arrangement and organization of blood capillaries allows the blood flowing through the carotid labyrinth to pass from the common chamber either directly or through the capillary plexus or through both to any of the head arteries. It is well known that the flow of blood through capillaries enhances metabolic exchange activity between the blood and the surrounding tissue. The carotid labyrinth in these fishes seems to be engaged in enhancing the rate of the metabolic activity, on one hand and in altering the nature of blood going to the retina and brain on the other. 2. Plasma skimming for improving the haematocrit value of blood going to the brain (Munshi and Hughes, 1987; Prakash, 1993). 3. Enzymatic activity (Tripathi, 1985).   Carbonic anhydrase activity similar to that suspected for the pseudobranch [transport of oxygen (Copeland, 1951), pH regulation (Maetz, 1956) hydration of CO2 into plasma bicarbonate, for better oxygenation of blood (Hoffert, 1966), and for the gill acid-base regulation (Randall and Dexboeck, 1984)] is a likely function of carotid labyrinth. 4. Maintenance of brain and intraocular temperatures above the ambient with the help of counter current between arterial blood from the carotid and opercular arteries (Linthicum and Carey, 1972).   A carotid rete mirabile has been observed in Tuna fish, which has a counter current vasculature. This structure acts as a heat exchanger with venous blood. In catfishes, however, carotid labyrinth has not been observed to serve a similar function, as arterioles and veins, in these group of fishes are not closely associated and intraocular. Carotid labyrinth and brain temperatures are not found to be significantly above ambient temperature (Olson et al., 1981). 5. Neuroendocrine mediated sensory function—baro and/or chemoreception (Gopesh, 1983; Srivastava and Gopesh, 1987; Olson et al., 1990).

Pseudobranchial Neurosecretory System The similarities between the carotid body of higher vertebrates and carotid labyrinth of fish suggest that the latter also serves as a chemo or baroreceptor (Olson et al., 1981).

48  |  Airway Chemoreceptors in the Vertebrates Information about any extrabranchial peripheral chemoreceptors was lacking in fish with the exception of that on the pseudobranch (Laurent and Rouzeau, 1972a, 1972b). Revelation of peculiar neurosecretory cells in the gill region in certain Indian teleost (Srivastava et al., 1981; Gopesh, 1983) is an interesting finding in this regard. These cells belonging to a peripheral neurosecretory system—the pseudobranchial neurosecretory system, may play an important role in such a function. Investigation on these cells have resulted into the establishment of a third neuroendocrine system for fish—the pseudobranchial neurosecretory system (Srivastava and Gopesh 1987; see Srivastava, 2005), besides the two earlier known, the hypothalamo—hypophyseal system (pituitary) and the caudal neurosecretory system. This system belongs to the category of diffuse neuroendocrine system. Many examples of a similar system have emerged during last three decades (Lauweryns et al., 1972; Cutz et al., 1975; Rogers and Haller, 1978; Wasano and Yamamoto, 1978, 1979; Goniakowska-Witalinska, 1981; Dunel-Erb, 1994; Dunel-Erb et al., 1982; Laurent, 1984; Zaccone et al., 1995; Bailly et al., 1989, 1992) which have been included under

Figure 11: Schematic sketch of dissected palate of the catfish Clarias batrachus, showing position of pseudobranchial neurosecretory cell masses (nsm) with relation to the gill cavity, muscle fascicles (m) aortic arches (I, II, III) and carotid labyrinth (c.l.) (Gopesh, 1983).

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  49

Figure 12: Cross section of a portion of the gill region falling between the first two gill arches in C. batrachus (a) Showing presence of lobes containing neurosecretory cell perikarya (pk) and bundle of neurosecretory cell processes (b.n.c.p.) coursing along the surface of a dorsal branchial muscle, first and second efferent branchial arteries and dorsal aorta joining the two. x 100. (b) Showing presence of lobes containing neurosecretory cell perikarya in close proximity of efferent branchial vessel and carotid labyrinth. x 62 (Gopesh, 1983).

Figure 13: Profiles of contact between neurosecretory cell processes and carotid labyrinth in C. batrachus (a) Showing a small bundle of neurosecretory cell processes (arrow) entering the capillary plexus of the carotid labyrinth x 80. (b) A magnified view of the area enclosed in the square in Figure (a) x 400. (c) Showing plexus of microvascular channels surrounding a principal head artery and a bundle of the pseudobranchial neurosecretory cell processes embedded in the plexus of channels. x 704. (d) Higher magnification of a portion of Figure (c) x 1025 (Gopesh, 1983; Tripathi, 1985).

50  |  Airway Chemoreceptors in the Vertebrates the broader category of APUD (Amine precursor uptake decarboxylation) system by Pearse (1969). These cells produce biologically active substances and act as regulators of some function towards homeostasis acting via neurocrine, endocrine and/or paracrine mechanisms (Kvetnoy et al., 2000). Pseudobranchial neurosecretory cells are present as jelly like masses, containing neurosecretory cell perikarya in close proximity of the first two efferent branchial vessels, lateral dorsal aorta connecting them and the pseudobranch (Glossogobius giuris) or the carotid gland (Notopterus notopterus, Notopterus chitala) or the carotid labyrinth [(in catfishes) see Table 2]. The diffuse mass is organized in lobes interspersed between fascicles of the dorsal branchial muscles (lavotores arcuum branchialis) interlacing with muscles and the anterior aortic arch vessels (Figures 11 & 12). At times, axon processes, taking long, tortuous courses are observed to end in carotid labyrinth (Figures 13a-13d) or in close contact with capillaries of any of the blood vessels (Figure 16). These neurosecretory cells are comparable with the concept of ‘Paraneurons’ (Fujita, 1976, 1977; Fujita and

Figure 14: Cross section of a lobe head showing different profiles of cell perikarya in (a) C. batrachus x 570; (b) Notopterus notopterus x 500; (c) Hetoropneustes fossilis x 600; (d) higher magnification of figure (c) showing neurosecretory cells and glial cells x 1025; (e) Ompak pabda x 400 (f ) Higher magnification of Figure (e) x 850; (g) O militaris x 400; (h) higher magnification of (g) x 850

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  51

Kobayashi, 1975, 1979; Fujita et al., 1980, 1988; Gopesh and Srivastava, 1997; Gopesh et al., 2002; Kanno, 1977; Zaccone et al., 1977, 1999). A number of perikarya of the pseudobranchial neurosecretory cells and small pericytes (glial cells) are present in each lobe (Figures 14 & 15). The cells are round or pear shaped, with prominent nuclei, one or two nucleoli and granular cytoplasm (Figure 15d). Cells in different stages of the secretory cycle viz., young, mature and spent, coexist in one mass (Figures 15b, 15c). Cells are monopolar and their axons run in a long, tortuous course as bundles. These axon bundles lie in close vicinity of the main blood vessels - the first efferent branchial vessel and/or the lateral dorsal aorta (Figure 12). A definite neurohaemal organ is lacking. The axon endings are intermingled with fine capillary net work of these vessels (Figure 16). The presence of the same type of neurohaemal contact in all the groups of fishes in which this system has been investigated (Gopesh, 1983), suggests a paracrine role for the system. The cells show positive reaction with neurosecretion specific stains like Acid Violet, Aldehyde fuchsin and Iron haematoxyline, which shows that secretory

Figure 15: Pseudobranchial neurosecretory cells, as seen in a cross section of a lobe head in a clupeid, N. chitala (a) Single lobe head containing perikarya x 80. (b) Magnified view of Figure (a), showing arrangement of the perikarya into discrete groups x 140. (c) A magnified view of area enclosed in the rectangle in Figure (b) Note different cell profile x 400. (d) Magnified view of a single neurosecretory perikarya showing nucleus, nucleolus, hyaline cytoplasm and a ring like vascular space around the perikarya. x 950 (Gopesh, 1983).

52  |  Airway Chemoreceptors in the Vertebrates

Figure 16: View of a complete unit of pseudobranchial neurosecretory system in C. batrachus, as seen in a lucky plane, in a series of sections of the gill region, showing lobe-head, containing neurosecretory cell perikarya (pk) the bundle of axon processes (c.pr) axon endings (a.ed) in close contact of blood vessel (bv) and its capillaries (cp). x 126 (Gopesh, 1983).

material is proteinaceous in nature. The cells and their processes demonstrated positive immunoreactivity with serotonin, neurophysin and neuropeptide-Y in Clarias batrachus and Heteropneustes fossilis (Gopesh et al., 2003) indicating production and active transportation of the secretory products to the neurohaemal contact sites. The colocalization of several peptides and monoamines in the gill neuro-endocrine cell system has also been observed (Mauceri et al., 1999; Zaccone et al., 1989a, 1995, 1996, 1997, 1999, 2003a, 2006). Positive immunoreactivity with neurophysin, serotonin and neuropeptide-Y in the cells of same mass indicates the presence of more than one bioactive substance in the cells of a single mass. Multiple products produced by individual cells and stored in the same secretory vesicle is well documented (Tischler, 1990; Kusakabe et al., 1991, 1993, 1996) which may be true for the pseudobranchial

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  53

neurosecretory cells also. However, this needs to be investigated by double staining and other methods. Pseudobranchial neurosecretory system may have more than one role in catfishes. The catfishes are in general bottom dwelling fishes, perpetually faced with hypoxia. Hypoxic conditions of the surrounding water influences the activity of pseudobranchial neurosecretory cells to signal surfacing activity (Gopesh, 1983). Positive immunoreactivity with serotonin (a biogenic amine) corroborates a role of these cells in respiratory physiology, especially in conditions of hypoxia—a role attributed to similar cells in fishes. This is similar to the neuroendocrine cells which have been observed to respond to hypoxia especially the gill epithelial cells of fish (Dunel-Erb, 1994; Dunel-Erb et al., 1982; Bailly et al., 1989; Jonz et al., 2004) and the glomus cells of rat (Kusakabe et al., 1993). Photoperiod dependent circadian rhythm initiates a secretory cycle in the pseudobranchial neurosecretory cells (Pandey, 1987; Singh, 1992) which may be responsible for regulation of circadian rhythm in the diel pattern of locomotory activity of these fishes. However, the presence of as many as three bioactive substances viz. serotonin, neurophysin and neuropeptide-Y in the mass of pseudobranchial neurosecretory cells, suggests a multiple functional role of these cells in the biology of these fishes (Gopesh et al., 2003). It is possible that these cells may be either of a neurogenic origin like cells of the APUD series (Pearse, 1969) or non-neurogenic like cells of skin and gills of fishes (Zaccone et al., 1977, 1999). Origin of peptide producing cells from a common neuroendocrine programme has been postulated (Pearse, 1986; Day and Salzet, 2002) and the final expression of these cells are presumably determined by microenvironmental circumstances of an individual candidate. Activation of specific genetic switches can lead to the expression of a partial or full neuroendocrine phenotype in a variety of cell types (Day and Salzet, 2002). The speculated hypothesis of origin of these cells is illustrated in Figure 17.

Discussion The organization of the pseudobranchial neurosecretory system is in conformity with the current concept of neuroendocrine system (Toni, 2004). Accordingly it is a set of cells organized in single organ and diffuse elements, which share co-production of amine hormone/transmitters, peptide hormones/transmitters and specific markers of determination. All the anatomical structures of the system may be part of a wide functional circuit, based on ‘internal secretions’ (Toni, 2004). Perhaps this represents an informational supersystem of the hypothalamo-pituitary axis, autonomous nervous system, APUD (Amine precursor uptake decarboxylation) system, immune and any

54  |  Airway Chemoreceptors in the Vertebrates

Figure 17: A hypothetical view of comparative differentiation of neurons, neurosecretory cells and paraneurons during development (Gopesh, 1983).

other body system controlling the homeostatic balance by performing autocrine, paracrine and endocrine regulations. Presence of amine and peptide containing neuroepithelial cells with chemosensory and paracrine function in the epithelial lining of fish respiratory surfaces is well established (Zaccone et al., 1995, 1997, 1999, 2003; Burleson and Milson, 2003; Jonz and Nurse 2003; Jonz et al, 2004). Similar neuroepithelial cells are also widely distributed within airways mucosa of terrestrial vertebrate (Lauweryns et al., 1972; Sorokin and Hoyt, 1989; Cutz 1997). The production of serotonin by the neuroepithelial cells (Bailly et al., 1992; Zaccone et al., 1989a&b, 1992, 2006) and the pseudobranchial neurosecretory cells (Gopesh et al., 2003) may be suggestive of a homology between the two cell types as the production of common peptides has already been explained on the basis of the common embryonic origin of these cells (Pearse, 1986). Thus, it appears that the amine and peptide containing neuroepithelial cells and the pseudobranchial neurosecretory

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  55

cells have evolved from the same common ancestral cell type. These cells might have derived at different periods of time from the embryonic neuroectodermal cells—or ecto-endodermal cells (Figure 17) but share a common neuroendocrine programme which is manifested by possession of the acronymous amino—handling characteristics of the series. The amphibians living in conditions similar to those of amphibious catfishes have shown the presence of a carotid labyrinth (Kusakabe, 1991, 2002), structurally and functionally similar to the carotid labyrinth of catfishes. This clearly reflects on the role of their micro-environment in determining the final expression. In addition the pseudobranchial neurosecretory cells, present in close association of carotid labyrinth in catfishes seem to be similar to those of oxygen sensing NE cells of gill and airways of fishes (Dunel-Erb, 1994; Dunel-Erb et al., 1982; Sundin et al., 1998b) and pulmonary NE cells (Mercer et al., 2000) and glomus cells (Kusakabe, 1991, 2002) in mammals. It is imperative to investigate further the pseudobranchial neurosecretory cells, along with carotid labyrinth in catfishes for better understanding of the evolution of chemosensory system in vertebrates as the gill perfusion and ionic transport in fishes has been found to be a complex mechanism involving a variety of neurocrine, endocrine, paracrine and autocrine signals through autonomic innervation of gill (Olson, 2002; Evans, 2002). This will go a long way in elucidating the general evolution of chemosensory in vertebrates, in general and in fishes, in particular.

References Abraham, A. 1969. Microscopic innervation of the heart and blood vessels in vertebrates, including man. Pergamon Press, London. Adams, W.E. 1958. Comparative morphology of the carotid body and carotid sinus. Charles C. Thomas, Springfield III. Bailly, Y., Dunel-Erb, S., Geffard, M., and Laurent, P. 1989. The vascular and epithelial serotonergie innervations of the actinopterygian gill filaments, with special reference to the trout. Salmo gairdneri. Cell. Tissues Res. 258: 349-363. Bailly, Y., Dunel-Erb, S., and Laurent, P. 1992. The neuro epithelial cells of the fish gill filament: Indolamine - immunocytochemistry and innerretion. Anat. Rec. 233: 143-161. Biscoe, T.J. 1971. Carotid body structure and function. Physiol. Rev. 51: 427-496. Burleson, M.L., and Milsom, W.K. 2003. Comparative aspects of O2 Chemoreception, anatomy, physiology and environmental adaptation. In : Oxygen Sensing, Responses and Adaptation to Hypoxia, S. Labin, G. Semanga, and N. Prabhakar (Eds.). Marcel Dekker, New York, Marcel Dekkar, pp. 685-708. Carman, J.B. 1955.The carotid labyrinth in Hyla aurea, with a note on that in Leiopelma hochstetteri. J. Anat. 83: 503-525. Carman, J.B. 1964. Carotid labyrinth in Anura. J. Anat. 98: 490. Carman, J.B. 1967a. The carotid labyrinth in the anuran Breviceps mossambicus. Trans. Roy. Soc., New Zealand, Zool. 10: 1-15.

56  |  Airway Chemoreceptors in the Vertebrates Carman, J.B. 1967b. The morphology of the carotid labyrinth in Bufo bufo and Leiopelma hochstetteri. Trans. Roy. Soc., New Zeal. Zool. 10: 71-76. Copeland, D.E. 1951. Function of glandular pseudobranch in teleosts,. Amer. J. Physiol. 167: 725. Cutz, E. 1997. Structure, molecular markers, ontogeny and distribution of pulmonary neuroepithelial bodies. In: Cellular and Molecular Biology of Airway Chemoreceptors., Cutz E. (Ed.). Landes Bioscience, Austin, T.X. : 1-33. Cutz, E., Chan, W., Wong, V., and Comen, P.E. 1975. Ultrastructure and flourescence histochemistry of endocrine (APUD type) cells in tracheal mucosa of human and various animal species. Cell Tissue Res. 158: 425-437. Day, R., and Salzet, M. 2002. The neuroendocrine phenotype, cellural plasticity, and the search for genetic switches: redefining the diffuse neuroendocrine system. Neuro. Letter. 23 (no5/6): 447-451. Devi, V 1987. Studies on the pseudobranchial neurosecretory system of certain teleosts, with particular reference to its function. D.Phil Thesis, University of Allahabad, Allahabad, India. Dunel-Erb, S. 1994. The neuroepithelial cells of the fish gill filament. In: Advance in Fish Research., B.R. Singh (Ed.). Narendra Publishing House, Delhi, pp. 29-42. Dunel-Erb, S., Bailly, Y., and Laurent, P. 1982. Neuroepithelial cells in fish gill primary lamellae. J. Appl. Physiol. 53: 1342-1353. Evans, D.H. 2002. Cell signaling and ion transport across the fish epithelium. J Exp Zool 293: 336-347. Fujita, T. 1976. The gastro-enteric endocrine cell and its paraneuronic nature. In: Chomaffin, Enterochromaffin and Related Cells., R.E. Coupland and T. Fujita (Eds.). Elsevier Amsterdam, pp. 191-208. Fujita, T. 1977. Concept of paraneurons. Arch. Histol. Jap. 40 (Suppl.): 1-12. Fujita, T., and Kobayashi, S. 1975. Parareuron—the new brothers of neurons. Igakuno Ayumi (Tokyo) 94: 638-640. Fujita, T., and Kobayashi, S. 1979. Current view on the paraneuronic concept. TINS 2: 27-30. Fujita, T., Kobayashi, S., Yui, R., and Iwanaga, T. 1980. Evolution of neurons and paraneurons. In: Hormones, Adaptation and Evalution, S. Ishii, T. Hirano, and M. Wade (Eds.). Jap. Sci. Soc. Press. Tokyo, Springer Verlag, Berlin. Fujita, T., Kanno, T., and Kobayashi, S. 1988. The paraneurons. In: Olfaction and Taste XI. N. Kurihara, N. Singh, and S.N. Ogava, (Eds.). Springer, Berlin, Heidelberg, New York. Fujita, T., Kobayashi, S., Yui, R., and Iwanaga, T. 1980. Evolution of neurons and paraneurons. In : Hormones, Adaptation and Evalution. S., Ishii, T., Hirano, and M., Wade, (Eds.). Jap. Sci. Soc. Press. Tokyo, Springer Verlag, Berlin. Goniakowska-Witalinska, L. 1981. Neuroepithelial bodies in the lungs of tree frog, Hyla arbora L. A scanning and transmission electron microscope study. Cell Tissue Res. 217: 435-441. Goodrich, E.S. 1930. Studies on the structure and development of Vertebrates. Macmillan, London, Reprinted Dover Publications, New York (1958). Gopesh, A. 1983. Studies on peculiar pseudobranchial neurosecretory cells in certain teleostean fishes. D.Phil. Thesis, University of Allahabad, Allahabad, India. Gopesh, A., and Srivastava, C.B.L. 1997. Paraneuronic nature of fish pseudobranchial neurosecretory cells. Jour. Freshwater Biol. 9(2): 78-81.

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  57 Gopesh, A., Prakash, S., and Jaiswal, U. 2002. Carotid labyrinth and associated pseudobranchial neurosecretory system in two marine catfish, Plotosus canius and Osteogeneiosus militaris. Science Letters, Nat. Acad. Sci. (India), 25(a-b): 298-302. Gopesh, A., Jaiswal, U., and Prakash, S. 2003. Immunocytochemical localization of neurophysin, neuropeptide Y and serotonin in the pseudobranchial neurosecretory cells of the catfish species, Clarias batrachus and Heteropneustes fossilis. Proc. Nat. Acad. (India). 73(B): 275-281. Hoffert, J.R. 1966. Observations on ocular fluid dynamics and carbonic anhydrase in tissues of the lake trout (Salvelinus namacycush) Comparative Biochem. Physiol. 17: 107-114. Hoffert J.R., and Fromm, P.O. 1972. Teleost retinal metabolism as affected by acetazolamide. Proc. Soc. Exp. Biol. Med. 139: 1060-1064. Hughes, G.M. 1984. General Anatomy of the gills. In: Fish Physiology Vol. X, Part A, W.S. Hoar, and D.J. Randall (Eds.) “Fish Physiology” Vol. X, Part A: Academic Press Inc., London, pp. 1-72. Hughes, G.M., and Munshi, J.S.D. 1979. Fine structure of the gills of some Indian air-breathing fishes. J. Morphol., 160: 169-194. Ishii, K., and Oasaki, T. 1966. Electron microscopy of the chemoreceptor cells of carotid labyrinth of the toad, Nature, 212: 1499-1500. Jonz, M.G., and Nurse, C.A. 2003. Neuroepithelial cells and associated innervation of the Zebra-fish gill: a confocal immunoflouresence study. J. Comp. Neurol, 461: 1-17. Jonz, M.G., Fearon, I.M., and Nurse, C.A. 2004. Neuroepithelial oxygen chemoreceptors of the Zebra-fish gill. J. Physiol. 560(3): 737-752. Kanno, T. 1977. Physiology of paraneurons Arch. Histol. ( Jap.) 40 (Suppl.): 13-29. Kingsley, J.S. 1926. Vertebrates. In: Outlines of Comparative Anatomy. Publisher of Vertebrates Publisher John Murray, London, p. 276. (Printed in India by G.W. Lawrie & Co., Lucknow, 1957.) de Kock, L.L. 1954. The intraglomerular tissues of the carotid body. Acta. Anat. 21: 108-116. Kusakabe, T. 2002 Carotid labyrinth of amphibians. Microsc. Res.Tech. 59:207-226. Kusakabe, T., Anglade, P., and Tsuji, S. 1991. Localization of substance P, CGRP, VIP, neuropeptide Y, and somatostatin immunoreative nerve fibres in the carotid labyrinths of the some amphibian species. Histochemistry. 96(3): 255-260. Kusakabe, T., Powell, F.L., and Ellisman, M.H. 1993. Ultrastructure of the glomus cells in the carotid body of chronically hypoxic rats: With special reference to the similarity of amphibian glomus cells. Anat Rec. 237: 220-227 Kusakabe, T., Kawakami, T., and Takenaka, T. 1996. Coexistence of neuropeptide in the amphibian carotid labyrinth. An application of double immunolabelling in combination with a multiple dye filter. Adv Exp Med Biol. 410: 325-328. Kvetnoy, I.M., Reiter, R.J., and Khavinson, V.H. 2000. Claude Bernard was right : hormones may be produced by “non-endocrine” cells. Neuro endocrinal Lett. 21: 173-174. Ladich, V.F. 1985. Development of the pseudobranch of the paradise fish Macropodus operculari (Anatantoidea, Teleostes), Zool. J. Anat. 113: 119-132. Ladich, V.F. 1987. Gross and fine structure of the pseduobranch of the paradise fish Macropodus opercularis, Zool. J. Anat. 114: 489-505. Laurent, P. 1974. Pseudobranch receptors in teleosts. In: Handbook of Sensory Physiology, Autrum et al. (Eds.). Springer Verlag Berlin. Laurent, P. 1984. Gill Internal Morphology. In: Int. Physiology Vol. XA, W.S. Hoar, and D.J. Randall (Eds.) Int. Physiology Vol. XA Academic Press, New York.

58  |  Airway Chemoreceptors in the Vertebrates Laurent, P., and Rouzeau, J.D. 1972a. Afferent neural activity from pseudobranch of teleosts. Effects of PO2, pH, osmotic pressure and Na+ ions. Resp. Physiol. 14: 307-331. Laurent, P., and Rouzeau, J.D. 1972b. Arterial chemoreceptors activity from in situ teleost pseudobranch during change in PO2 of inhaled water. Laurent, P., and Dunel-Erb, S. 1984. The pseudobranch: Morphology and function In: Fish Physiology Vol. XB, W.S. Hoar, and D.J. Randall (Eds.). Academic Press, New York. Lauweryns, J.M., Cokelaere, M., and Theunyck, P. 1972. Neuroepithelial bodies in the respiratory mucosa of various mammals. Z. Zellforsch. Mikrosk, Anat. 135: 569-592. Linthicum, D.S., and Carey, F.G. 1972. Regulation of brain and eye temperature by the blue tuna. Comp.Biochem.Physiol. 43A: 425-433. Maetz, J. 1956. La role biologique de l’anhydrase carbonigue chez-quelaues Teleosteans. Bull. Biol. Belg. Suppl. 40: 1-29. Mauceri, A., Fasulo, S., Ainis, L., Licata, A., Lauriano, E.R., Martinez, A., Mayor, B., and Zaccone, G. 1999. Neuronal Nitric oxide synthase (nNOS) expression in the epithelial neuroendocrine cell system and nerve fibres in the gill of the catfish, Heteropneustes fossilis. Acta. Histochem. 101: 437-448. Mercer, S.E., Wilk-Blaszcrak, M.A., and Burleson, M. 2000. A comparative model for the investigation of the electrophysiological basis of oxygen sensing. FASEB J 407: 2. Munshi, J.S.D., and Hughes, G.M. 1981. Gross and fine structure of the pseudobranch of the climbing perch Anabas testudineus (Block) J. Fish Biol., 19: 427-438. Munshi, J.S.D., and Hughes, G.M., 1987. Microvascular organization of the pseudobranch and carotid labyrinth of two species of air-breathing teleostean fishes. J. Fish. Biol. 31: 705-714. Olson, K.R. 2002. Vascular anatomy of the fish gill. J EXP Zool, 293: 214-231. Olson, K.R., Flint, K.B., and Budde Jr., R.B. 1981. Vascular corrosion replicas of chemobaroreceptors in fish: The carotid labyrinth in Ictaluridae and Clariidae. Cell. Tissue Res. 219: 535-541. Olson, K.R., Munshi, J.S.D., Ghosh, T.K., and Ojha, J. 1990. Vascular organization of the head and respiratory organs of the air breathing catfish, Heteropneustes fossilis. Jour. Morph, 203: 165-179. Pandey, A. 1987. Studies on surfacing activity of certain catfishes in the particular reference to the role of pseudobranchial neurosecretory system, D.Phil. Thesis, University of Allahabad, India. Pandey, K.N. 1985. Studies on the structure of carotid labyrinth in certain catfishes. D.Phil. Thesis. University of Allahabad, Allahabad, India. Pearse, A.G.E. 1969. The cytochemistry and ultrastructure of polypeptide hormone producing cells of the APUD series and embryonic physiologic and pathologic implications of the concept. J. Histochem. Cytochem. 17: 303-313. Pearse, A.G.E. 1986. The diffuse neuroendocrine system : peptides amines placodes and the APUD theory. Progr. Brain Res. 68: 25-31. Prakash, S. 1993. Studies on carotid labyrinth of some fishes D.Phil. Thesis. University of Allahabad, Allahabad, India. Ramaswami, L.S. 1945. The occurrence of a pseudobranch in the family ophiocephalidae (Teleostomi) Current Sci. Bangalore 14: 105-106. Randall, D.J., and Dexboeck, C. 1984. Oxygen and carbon diaoxide transfer across fish gills. In: Fish Physiology, W.S. Hoar, and D.J. Randall (Eds.), Academic Press, London. Rogers, D.C., and Haller, C.J. 1978. Innervation and cytochemistry of the neuroepithelial bodies in the ciliated epithelial of the toad lung (Bufo marinus). Cell Tissue Res. 195: 395-410.

Carotid Labyrinth and Associated Pseudobranchial Neurosecretory Cells  |  59 Roy, P.K., Ghosh, T.K., and Munshi, J.S.D. 1997. The pseudobranch : Their structure, function in freshwater teleosts. In: Advances in Fish Research (Vol. 2), B.R. Singh (Ed.) Narendra Publishing House, Delhi, India, pp. 81-86. Shepro, D. 1980. The microvascular system. In: The Cell Biology of iInflammation., G. Weissmann (Ed.). Elsener/North Holland Biomedical Press, Amsterdam, New York, Oxford. Singh, G.P. 1959. A note on the carotid labyrinth in fresh water teleosts. Sci. and Cult. 24: 475476. Singh, G.P. 1960. The blood vascular system of some freshwater teleosts. Ind. J. Zool. (1): 2758. Singh, K.P. 1992. Studies on surfacing behaviour of certain fishes. D.Phil. Thesis, University of Allahabad, Allahabad, India. Singh, M. 1982. Studies on the structure and development of the pseudobranch in certain fishes, with particular reference to their innervation. D.Phil. Thesis, Allahabad University, Allahabad, India. Singh, M., and Srivastava, C.B.l. 1979. Head arteries of some teleost fishes, with particular reference to the air-breathing catfishes. Nat. Acad. Sci. Letters, Vol. 2(6): 240-242. Smith, H.M. 1960. Evolution of Chordate Structure. Holt Rinehart and Winston, Inc. Sorokin, S.P., and Hoyt, R.F. 1989. Neuroepithelial bodies and small granule cells. In: Lung Cell, D. Massaro (Ed.). Lung Cell, Marcel Dekker, New York, pp. 199-344. Srivastava, C.B.L. 2005. Catfish Soft Anatomy. Kitab Mahal, Allahabad, India. Srivastava, C.B.L., and Gopesh, A., 1987. A third system of neurosecretion in fish the— psuedobranchial system. In: Advances in cytology and genetics V.S. Bhatnagar (Ed.). Allahabad University Press, Allahabad, India. Srivastava, C.B.L., and Pandey, K.N. 1984a. Carotid labyrinth of a marine catfish Tachysurus thalassinus (Rupp.) Proc. Nat. Acad. Sci. India (B) 54(I): 6-10. Srivastava, C.B.L., and Pandey, K.N. 1984b. Carotid labyrinth of catfishes. Proc. Nat Acad. Sci. India 54(11): 204-212. Srivastava, C.B.L., and Singh, M. 1980. Occurrence of carotid labyrinth in the catfish group of teleost fishes. Experientia (Switzerland), 36: 651-653. Srivastava, C.B.L., and Singh, M. 1981. Structure of the carotid labyrinth of the air breathing catfish Clarias batrachus (Linn) Proc. Nat. Acad. Sci. India 51(B) IV: 329-333. Srivastava, C.B.L., Singh, M., and Gopesh, A. 1981. A new neurosecretory system in fish, located in the gill region. Experientia, 37 : 850-851. Srivastava, C.B.L., Tripathi, S., Pandey, K.N., and Singh, M. 1988. Fish Carotid labyrith—a transformed pseudobranch. Arch. Biol. (Bruxelles) 99: 33-50. Sundin, L. Holmgren, S. and Nilsson, S., 1998b. The oxygen receptor of the teleost gill? Acta. Zool (Stockholm) 79:207-214. Srivastava, C.B.L., Tripathi, S., and Pandey, K.N. 1994. Fish carotid labyrinth. In: Advances in Fish Biology, H.R. Singh (Ed.). Hindustan Publication Corporation, New Delhi. Swammerdam. 1738. Bybel de Nature, Leyden: Severinus 2. In: The Comparative Morphology of the Carotid Body and Carotid Sinus., W.E. Adams (Ed.). Spring field, III: Charles C. Tischler, A.S. 1990. The dispersed neuroendocrine cells: The structure function, regulation and effects of xenobiotics on this system. Semin. Surg. Oncol. 9(5): 387-393. Toni, R. 2004. The neuroendocrine system: organization and homeostatic role. J. Endocrinol. Invest. 27 Suppl. (6): 35-47.

60  |  Airway Chemoreceptors in the Vertebrates Tripathi, S. 1985. Studies on the carotid labyrinth in teleosts. D.Phil. Thesis, University of Allahabad, Allahabad, India. Vashisht, H.S., and Kapoor, N.N. 1965. Relation of the efferent branchial vessels to the circulus cephalicus and other arteries of the branchiocephalic region in some fresh water teleosts. Res. Bull. Punjab Univ. Sci. N.S. 16: 129-145. Wasano, K., and Yamamoto, T. 1978. Honoamines containing granulated cells in frog being. Cell Tissue Res. 193: 201-209. Wasano, K., and Yamamoto, T. 1979. APUD-type recepto-secretory cells in the chicken lung. Cell Tissue Res. 201: 197-205. Wittenberg, J.B., and Haedrich, R.L. 1974. The choroid rete mirabile of the fish eye II. Distribution and relation to the pseudobranch and to the swim bladder rete mirabile. Biol. Bull., 149 (1): 137-156. Zaccone, G., Fasulo, S., and Ainis, L. 1995. The neuroendocrine epithelial cell system in the respiratory organs of air-breathing and teleost fishes. Int. Rev. Cytol. 157: 277-314. Zaccone, G., Mauceri, A., Ainis, L., Lo. Cascio, P., and Ricca, M.B. 1996. Location of immuno reactive endothelin in the neuroendocrine cells of fish gill. Neuropeptides. 30: 50-57. Zaccone, G., Fasulo, S., Ainis, L., and Licata, A. 1997. Paraneurons in the gills and air ways of fishes. Microsc Res Tech. 37: 4-12. Zaccone, G., Mauceri, A., Ainis, L., Fasulo, S., and Licata, A. 1999. Paraneurons in the skin and gills of fishes. In: Ichthyology, Recent Research Advances, D.N. Saksena (Ed.). Science Publishers, Inc., Enfield N.H.: 417-447. Zaccone, G., Goniakowska-Witalinska, L., Lauweryns, J.M., Fasulo, S., and Tagliafierro, G. 1989a. Fine structure and serotonin immunohistochemistry of the neuroendocrine cells in the lungs of the bichirs, Polypterus delhezi and P. ornatipinnis. Basic Appl Histochem. 33: 277-287. Zaccone, G., Ainis, L., Mauceri, A., Lo Cascio, P., Giudice, F.L., and Fasulo, S. 2003a. NANC nerves in the respiratory air-sac and branchial vasculature of the Indian catfish. Heteropneustes fossilis. Acta Histochem. 105: 151-163. Zaccone, G., Mauceri, A., and Fasulo, S. 2006. Neuropeptides and Nnitric oxide synthase in the gill and the air breathing organs of fishes. J. Exp. Zool. 305A: 428-439.

3 Serotonergic Neuroepithelial Cells in Fish Gills: Cytology and Innervation Yannick J.R. Bailly

Abstract Positioned strategically at the interface between respiratory water and arterial blood flows, the secretory neuroepithelial cells NECs present in the gill filaments of fishes and exhibit morphofunctional features of O2 chemoreceptors present in the lungs of air-breathing vertebrates.   Our anatomical and physiological knowledge about the indolamine secretory properties, hypoxia sensitivity and complex innervations of the fish gill NECs are reviewed in a comprehensive description of the fish gill nervous system. Acting as receptosecretory paraneurons, the fish gill NECs are probably involved in local and central control of branchial functions through the paracrine production of serotonin and their synaptic relationships with the sympathetic and intrinsic branchial nervous systems. In response to hypoxia, NECs could locally adapt the respiratory surface through direct serotonin-mediated effects on the branchial vasculature. Vasomotor responses to hypoxia may also be mediated by a central reflex triggered by the NECs through their afferent synapses to the CNS which in turn would act on the filament vasculature via branchial autonomic motor innervation and intrinsic neurons.

Keywords: Fish gill, neuroepithelial cell, serotonin, oxygen chemoreceptor, electron microscopy

Introduction Fishes exhibit pronounced responses to hypoxia such as hyperventilation, bradycardia, and variations in gill vascular resistance. These responses have been shown to arise Cytologie et Cytopathologie Neuronales, Département Neurotransmission et Sécrétion Neuroendocrine, Institut de Neurosciences Cellulaires et Intégratives CNRS UMR 7168, 5, rue Blaise Pascal, 67084 Strasbourg, France, Email : [email protected]

62  |  Airway Chemoreceptors in the Vertebrates principally from peripheral O2 chemoreceptors located in the gills (Milsom and Brill, 1986; Burleson et al., 1992). Such O2 sensors are believed to be the secretory neuroepithelial cells (NECs) that have been identified in the gill filaments of all species examined so far including both teleost (Dunel-Erb et al., 1982; Bailly et al., 1989, 1992; Zaccone et al., 1997; Jonz and Nurse, 2003; Saltys et al., 2006) and non-teleost species (Laurent, 1984; Bailly et al., 1992; Zaccone et al., 1992, 1997; GoniakowskaWitalinska et al., 1995). This is suggested by several morphofunctional features that fish gill NECs share with other peripheral O2 chemoreceptor cells such as the carotid body cells of mammals (Gonzalez et al., 1994) and birds (King et al., 1975) and the NECs present in the lungs of air-breathing vertebrates (Cook and King, 1969; Wasano and Yamamoto, 1978, 1979; Lauweryns et al., 1973, 1982, 1983, 1985, 1986; Rogers and Haller, 1978; Goniakowska-Witalinska, 1981; Lauweryns and Van Lommel, 1982, 1983, 1987; Youngson et al., 1993) including lungfishes (Zaccone et al., 1989; Adriaensen et al., 1990; Adriaensen and Scheuermann, 1993). In fish gills, most NECs are located in the distal part of the gill filaments in the deepest layer of the filament epithelium, very near the efferent filament (eFA) and lamellar (efLA) arteries, and facing the respiratory water flow (Dunel-Erb et al., 1982). This is a strategic position for potentially monitoring changes in arterial or ambient PO2 suggesting a chemosensory role for these cells. In addition, gill NECs are able to secrete serotonin, which is stored in dense-cored cytoplasmic vesicles (DCVs), and display a complex innervation pattern (Dunel-Erb et al., 1982; Bailly et al., 1989, 1992; Jonz and Nurse, 2003; Saltys et al., 2006). However, the physiological and morphological responses of these presumptive receptors to hypoxia, reminiscent of anoxic degranulation displayed by the NECs in the mammalian lungs (Lauweryns and Cokelaere, 1973; Lauweryns et al., 1978, 1983), have been poorly documented at the cellular level in fishes (Dunel-Erb et al., 1982) until the recent demonstration of the pronounced morphological changes that gill NECs undergo in response to chronic hypoxia ( Jonz et al., 2004). This article reviews the morphofunctional approaches that have been used in our laboratory to investigate the 5-HT content and the innervations of NECs in fish gills (Dunel-Erb et al., 1982; Bailly et al., 1989, 1992; Bailly, 1991).

Material and Methods Fish Species Fish gill NECs were studied in eight teleostean species: the rainbow trout, Onchorynchus mykiss (Richardson) (n=23); the perch, Perca fluviatilis (Linné 1736) (n=10); the pike-perch, Sander lucioperca (n=2); the sea perch, Dicentrarchus labrax (n=2); the black bass, Micropterus dolomieui (Lacépède) (n=3); the catfish, Ictalurus melas (Rafinesque) (n=3); the tilapia, Oreochromis mossambica (n=3); and the eel, Anguilla anguilla (n=4). In addition, a chondrostean fish, the sturgeon Acipenser

Serotonergic Neuroepithelial Cells in Fish Gills  |  63

baeri (Linné 1736) (n=2), and a chondrichthyan fish, the dogfish Scyliorhinus canicula (Linné 1736) (n=3), were examined.

Anaesthesia All fishes were anaesthetized with 0.01% MS-222 (Sandoz) before experimentation.

Biogenic Amines Histochemistry Biogenic amine histofluorescence was performed according to the method of Falck and Owman (1965) on small pieces of gill arches from perch, pike-perch, trout, catfish, black bass and eel. In some experiments on catfish, the serotonin-depleting agent parachlorophenylalanine (PCPA, Koe and Weissman, 1966; Jequier et al., 1967; Sanders-Bush and Massari, 1977) was injected intraperitoneally (i.p., 50 mg in 2 ml saline per 200 g fish) twice at 24 h intervals before sampling. Gill samples were frozen at –80°C before freeze-drying at –40°C for 5 d under vacuum. The samples were then warmed to 40°C and exposed for 1, 2 or 3 h to 70% humid vapours of formaldehyde at 80°C before being embedded in paraffin. Sections (10 µm thick) were examined with an epifluorescence microscope equipped with ultraviolet illumination and appropriate filters for paraformaldehyde-induced fluorescence of biogenic amines (Dunel-Erb et al., 1982; Dunel-Erb and Bailly, 1986).

Immunocytochemistry Fixation of the Gills

The gills of the anaesthetized fishes were perfused via the ventral aorta with 25 ml of ice-cold buffered (pH 7.5) paraformaldehyde (4%) or glutaraldehyde (5%) as described previously (Bailly et al., 1992). Small pieces of gills were then immersed for 2 h in the perfusion fixative before being rinsed overnight in the fixative buffer containing 10% sucrose (2°C) and frozen in isopentane at –80°C.

Antibodies

Rabbit polyclonal antibodies against a bovine serum albumin (BSA)-glutaraldehyde5-HT immunogen (Immunotek, Marseille, France; Geffard et al., 1985; Bailly et al., 1992) were used (1/500-1/5000) for both horseradish peroxidase (HRP) immunohistochemical (trout, perch, black bass, sea perch, tilapia and sturgeon) and immunocytochemical (trout) assays on glutaraldehyde-fixed gills. In addition, rabbit polyclonal antibodies against a BSA-paraformaldehyde-5-HT immunogen (INC, Stillwater, MN; Bailly et al., 1992) were used (1/500) in immunohistofluorescence assays on paraformaldehyde-fixed gills of trout, perch, black bass and catfish. Finally,

64  |  Airway Chemoreceptors in the Vertebrates rabbit polyclonal antibodies against a BSA-glutaraldehyde-5-methoxytryptamine (5-MT) immunogen (Geffard et al., 1985) were used for HRP-immunohistochemistry on glutaraldehyde-fixed gills of all species.

Indirect Immunohistofluorescence and HRP-Immunohistochemistry

Cryostat sections (10 µm thick) of paraformaldehyde- and glutaraldehyde-fixed gills were processed following, respectively, classical indirect immunohistofluorescence (Hökfelt et al., 1975; Bailly et al., 1992) and immunoperoxidase (Sternberger, 1979; Geffard et al., 1985; Bailly et al., 1992) procedures using fluorescein isothiocyanate (FITC)-labelled (1/20-150) and HRP-labelled (1/50-1200) sheep anti-rabbit immunoglobulins (Biosys, Compiègne, France).

Pre-embedding Indirect HRP-Immunocytochemistry

Immunoperoxidase-treated vibratome sections (150 µm) of glutaraldehyde-fixed trout gills were postfixed in 1% OsO4 in water before dehydration and flat embedding in Araldite 502. Ultrathin sections were counterstained with uranyl acetate (50% in ethanol) before examination with a Siemens Elmiskop 101 transmission electron microscope (TEM) at 60 kV.

Immunocytochemical Controls

Negative controls were conducted by omitting the 5-HT or 5-MT antibodies and/ or the labelled anti-rabbit immunoglobulins in the incubation media. In addition, alternate sections were incubated with the free anti-5-HT or 5-MT antibodies or with the anti-5-HT or 5-MT antibodies pre-adsorbed with their specific immunogen as previously described (Bailly et al., 1989, 1992). No specific labelling was observed in any of these negative controls.

Ultrastructural Study The gills (trout, catfish, eel, perch, sturgeon and dogfish) were perfused rapidly via the ventral aorta with 5% glutaraldehyde in 0.05 M cacodylate buffer (pH 7.4). The gill arches and filaments were then dissected and further fixed for 1 h at 4°C before washing in the fixative buffer and post-fixation in 1% osmium tetroxide for 1 h. After dehydration and embedding in Araldite, semi-thin sections of the gill filaments were cut and stained with toluidine blue. Ultra-thin sections were counterstained with uranyl acetate and lead citrate before TEM examination. In TEM study of untreated and 5-OHDA-treated trout, a special fixation was used for ultrastructural preservation of the biogenic amines (Tranzer and Richards, 1976).

Serotonergic Neuroepithelial Cells in Fish Gills  |  65

Surgical Denervation Surgical denervation was carried out in trout and perch. The post trematic glossopharyngeal (IX) nerve and/or the pretrematic first vagus (X) nerve were cut unilaterally proximal to the gill to induce nerve degeneration in the first gill arch. The post trematic first vagus nerve and/or the pretrematic second vagus nerve to the second gill arch were cut (Figure 1). Unilaterally proximal to the gill to induce nerve degeneration in the second gill arch. Sectioning of the sympathetic nerve was performed by cutting the post trematic glossopharyngeal nerve to the first gill arch and the post trematic vagus nerve to the second gill arch. The first and second gill arches were fixed 8, 12 or 21 d after surgery, dissected bilaterally and processed either for amine histofluorescence or histological immunoperoxidase detection of 5-HT. TEM examination was carried out according to the procedures described above, using the non-sectioned side as a control.

Chemical Degeneration Chemical degeneration was carried out in the trout and the perch according to the schedules indicated in Table 1 (0.5-1 ml i.p., 0.1 ml i.c.). Sympathectomy was achieved using the neurotoxins 5- and 6-hydroxydopamine (5- and 6-OHDA, Sigma). Chemical degeneration of indolaminergic nerves was induced using the specific neurotoxins 5,6di-hydroxytryptamine (5,6-DHT) and 5,7-di-hydroxytryptamine (5,7-DHT). In the later case, 30 min before administration of 5,7-DHT, the trouts were given 50 mg/kg methysergide bimaleate i.p. and 25 mg/kg desmethylimipramine i.p. respectively to protect from 5,7-DHT mimetic effects and prevent its uptake by catecholaminergic terminals (Gershon and Sherman, 1982). Control trouts and perches (n=2 in each experiment) were sham-injected with 0.06% saline only. The gills were then prepared for amine histofluorescence, immunohistochemistry and TEM analysis.

Table 1:  Catecholamine- and indolamine-specific neurotoxin injection schedules. Fish species

Neurotoxin

Dose

Route

Post-inoculation time of gill fixation

Rainbow trout n = 3 6-OHDA

10, 15, 60 mg/kg i.p.

7, 6 and 1 d

Perch n = 3

6-OHDA

10, 15, 60 mg/kg i.p.

7, 6 and 1 d

Perch n = 3

5-OHDA

50 mg/kg

i.c.

30 min

Rainbow trout n = 3 5-OHDA

2 x 75 mg/kg

i.p.

10 h and 4 h

Rainbow trout n = 3 5,6-DHT

20 mg/kg

i.p.

1d

Rainbow trout n = 3 5,6-DHT

15 mg/kg

i.p.

3h

Rainbow trout n = 3 5,7-DHT

110 mg/kg

i.p.

3h

66  |  Airway Chemoreceptors in the Vertebrates

Serotonergic Neuroepithelial Cells in Fish Gills  |  67

Hypoxia Cytological analysis of the organelle and DCV content of the NECs was carried out at the ultrastructural level in the gills of trouts (n=3) submitted to acute hypoxia by transferring them from normoxic to hypoxic water (PO2 < 10 Torr) for 30 min before processing for TEM examination.

Results Histology of Fish Gill NECs In fish, the gill filament or primary lamella supports parallel respiratory lamellae. The gill filament itself is surrounded by a multilayered filament epithelium distinct from the thin, bilayered respiratory epithelium that covers the respiratory lamellae (Figure 2). The filament epithelium consists of several cell types, including pavement cells, mast cells, chloride cells, and incompletely differentiated cells, in addition to Figure 1: Innervation of the gill arches by the branchial nerves in Perca fluviatilis. A. Drawing of lateral view of the left side of the body. The entry of the nerves to the gill arches (1, 2, 3, and 4) is shown more precisely by removing the dorsal part of the gill filaments and muscles (m). The glossopharyngeal nerve (IX) has two main branches: the pretrematic (pr) supplies the pseudobranch (PSB, not shown), and the post trematic (pt) reaches the first gill arch. The post trematic branch divides in the gill arch (*) to supply the rakers (R), efBA, efferent branchial artery. The four distinct ganglia of the vagus (X1, X2, X3, X4) give off four nerve trunks, each dividing principally into pretrematic and post trematic branches. The first branch innervates the first (pr1) and second (pt1) gill arches; the second branch innervates the second (pr2) and third (pt2) gill arches; the third branch innervates the third (pr3) and fourth (pt3) gill arches; the fourth branch innervates the fourth (pr4) gill arch and gives rise to visceralis (v) and cardiac (c) branches. The sympathetic chain (sy) runs under the cranial nerves, and bears ganglia that connect with them. Red and blue double arrows respectively show where the IX post trematic branch and the X pretrematic and post trematic branches were sectioned to induce degeneration in the first and second gill arches. The head kidney (HK) lies beneath the gills; cg, celiac ganglion; l, and d, lateral and dorsal branches of the vagus. B. Connection (arrowhead) of the sympathetic chain (sy) with the glossopharyngeal ganglion (IX) in Perca fluviatilis. Osmium tetroxide staining. gg, sympathetic ganglion. Scale bar = 1 mm. C. Distribution and composition of the glossopharyngeal (IX) and vagus (X) branchial nerves in the teleosts. The pretrematic and post trematic branches of both IX and X nerves contain intrinsic branchial neurons (yellow), proprioceptive somatic and general visceral sensory components (black) as well as cranial autonomic motor (orange) components. Motor fibres of the special visceral system (red) as well as the spinal autonomic system (green) enter the gill arches by the post trematic branches (pt) only. psb, pseudobranch; sgc, sympathetic ganglionic chain; 1, 2, 3, 4, first, second, third and fourth gill arches.

68  |  Airway Chemoreceptors in the Vertebrates

Figure 2: NECs resting on the basal lamina of the filament epithelium. Onchorynchus mykiss. A. Toluidin blue staining. Two NECs with clear cytoplasm (arrows) are visible in the filament epithelium (FE) surrounding efferent lamellar arteries (efLA) and subepithelial processes of the central venous sinus (stars). IFE, interlamellar filament epithelium; L, lamellae. Bar = 100 µm. B. Electron micrograph showing two NECs in the bottom layer of the filament epithelium. The left NEC sends a basal process (arrow) close to the basal lamina (bl), whereas the right NEC is largely apposed on it. Subepithelial smooth muscle cells (m) and nerve bundles (n) surround the wall of the central venous sinus (CVS). s, Schwann cell. Bar = 1 µm.

Serotonergic Neuroepithelial Cells in Fish Gills  |  69

mucous cells that are open to the external milieu on the mucosal side of the filament epithelium and NECs present on the basal lamina that lines the serosal side of the filament epithelium on its “efferent” side facing the respiratory water flow (Figure 2). We found 5-HT NECs in the gill filaments of all teleosts (Dunel-Erb et al., 1982) and non-teleosts (Laurent, 1984; Bailly et al., 1989, 1992) studied (Figures 2, 3).

Serotonin in Fish Gill NECs The monoamine content of the NECs was for the first time demonstrated in our laboratory (Dunel-Erb et al., 1982) using the histofluorescence technique of Falck and Owman (1965). This study indicated that serotonin (5-hydroxytryptamine, 5-HT) is the major monoamine contained in these cells, because the yellow fluorescence of the NECs revealed by this approach reached a maximum brightness after 2 h of exposure to formaldehyde vapour and completely disappeared after 2 min of ultraviolet exposure (Figure 4A). Furthermore, an intraperitoneal injection of the serotonin-depleting agent PCPA completely eliminated the fluorescence of the NECs in catfish gill filaments (Dunel-Erb et al., 1982). Finally, both FITC and HRP indirect immunohistochemistry demonstrated the presence of 5-HT and 5-MT in the fish gill NECs (Figure 4B-4D). The highest intensity of formaldehyde-induced fluorescence and indolamine immunoreactivity often occurred in the basal soma of the NECs and especially in the short processes projecting towards the basal lamina of the filament epithelium (Figures 2B, 4B-4D, 5A). The largest number of indolamine-containing NECs were isolated or clustered cells in the distal half of the filaments, whereas single NECs were found in the proximal half of the filaments (Figures 3, 4). Almost all the 5-HT NECs observed in the fish species studied were encased in the deeper layer of the filament epithelium and never reached the external surface located about 50 µm from the NEC apex (see below). On the opposite side, the efferent lamellar arteries that collect arterial blood from the respiratory lamellae were located within 25-30 µm of the NECs (Figures 2A, 4C, 4D). Nevertheless, the NECs were separated from the arterial blood circulating in the subepithelial processes of the central venous sinus (Figures 2A, 4D) by only 15-20 µm of connective tissue and the delicate endothelial wall of the sinus (100 nm thick).

Cytology of Fish Gill NECs At the ultrastructural level, the NECs were mostly characterized by numerous DCVs and a large number of organelles indicating a high metabolic rate, presumably associated with the production of the DCVs. These vesicles (80-100 nm) appeared to be completely filled with dense material when a procedure for specific preservation of biogenic amines was used (Tranzer and Richards, 1976). In many cases, the accumulation of indolamine-related fluorescence and immunostaining as well as of vesicles in a part of the cell close to the basal lamina suggested that these cells were

70  |  Airway Chemoreceptors in the Vertebrates

Figure 3: Three-dimensional diagram showing the topography of the 5-HT NECs within the second proximal and fourth distal quarters of the trout gill filament. The frequency of the NECs (red cells) increases in the distal part of the filament. The proximal part of the filament contains isolated 5-HT NECs and neurons (N). Both NECs and neurons are absent on the afferent side of the filament. Only the main parts of the filament vasculature are represented. Two filament nerve bundles (n) run parallel to the efferent filament artery (efFA). Arrowheads indicate the direction of blood flow in the respiratory arterial vasculature, and arrows indicate the direction of water (w) over the efferent edge of the filament and between the respiratory lamellae (L). afLA, afferent lamellar artery; afFA afferent filament artery; C, skeletal cartilage rod of the filament; CVS, central venous sinus; efLA, efferent lamellar artery.

Serotonergic Neuroepithelial Cells in Fish Gills  |  71

Figure 4: Serotonin NECs in the distal half of the filament. A, B, Ictalurus melas; C, D, Onchorynchus mykiss. A. Formaldehyde-induced indolamine fluorescence. B. Serotonin-immunofluorescence. C, D. 5-HT-immunoperoxidase. Most NECs rest or send processes (arrow in D) close to the basal lamina (arrowheads) of the filament epithelium (FE). Some NECs send processes towards neighbouring NECs (arrows in C). White asterisk, efferent filament artery; black asterisk, efferent lamellar arteries; star, central venous sinus. Bar = 50 µm in A, B, C and 10 µm in D.

72  |  Airway Chemoreceptors in the Vertebrates polarized (Figure 5A). The presence of secreted cores outside the cell membrane close to the basal lamina of the filament epithelium indicated that exocytosis of DCVs occurred at the NEC plasmalemma (Figure 5B, 5C).   Pre-embedding immunoperoxidase detection of serotonin showed that the NEC population in the distal part of the trout gill filament was made up of both intensely and weakly 5-HT immunoreactive cells (Figure 6A, 6C). Non-serotonergic NECs were not found. Much of the 5-HT immunolabelling covered the dense core of the cytoplasmic DCVs. In the intensely labelled NECs, most DCVs were seen in the basal pole of the cells and in fine processes that ran tortuously between the epithelial cells (Figure 6B, 6C). These processes were in direct contact with the basal lamina or a neighbouring NEC. In the dogfish and the sturgeon, NECs were seen either in close vicinity or in direct contact with extracellular spaces in the depth of the filament epithelium (Figures 7, 8A). In the trout, 5-HT immunoreactive processes were found touching the external medium in areas of the filament that surrounded the base of the respiratory lamellae (Figure 8B, 8C).

Innervation of Fish Gill NECs Throughout the filaments, a dense extensive subepithelial nerve plexus was inserted between the basal lamina of the filament epithelium and the vascular system. On the efferent side of the filament, nerve fibres emanating from this plexus unequivocally innervated three types of structures: the vascular walls of the efferent filament and lamellar arteries (Bailly and Dunel-Erb, 1986; Dunel-Erb and Bailly, 1986; Bailly et al., 1989; Dunel-Erb et al., 1989) and the central venous sinus (Bailly, 1991) as well, the subepithelial smooth muscle cells intermingled with the nerve plexus (Figures 9A-9C, 12, 14B, 14C) and the NECs (Figures 5B, 5C, 7, 8A, 9C, 10, 11, 12, 13, 14B) (DunelErb et al., 1982; Bailly, 1991; Bailly et al., 1992). The nerve fibres of the subepithelial plexus formed varicose profiles. The first type displayed swollen shapes with a clear neuroplasm containing scattered neurotubules and a few DCVs (Figure 9A). Smaller nerve profiles were considered preterminal and contained several mitochondria, small clear vesicles, and small DCVs. In the area of the NECs, nerve fibres were separated from the subepithelial plexus and ran beneath the basal lamina opposite those cells. Most of them displayed a second type of ultrastructural aspect with numerous clear vesicles and a few DCVs and mitochondria in a dense neuroplasm (Figures 5B, 5C, 9C, 10A, 10B, 11B, 11C, 12). Nerve profiles were seen passing across the basal lamina, whereas NEC processes only occasionally passed through the basal lamina down to the subepithelial nerve profiles. After crossing the basal lamina, nerve fibres finally came into contact with NECs (Figures 10B, 10C, 11A, 11D). The apposition of more than one nerve profile to the same NEC indicated that either several nerves converged on a single cell or the same nerve fibre had several contacts “en passant”. Terminal profiles in contact with NECs were of two types. The first type was large and contained few organelles,

Serotonergic Neuroepithelial Cells in Fish Gills  |  73

Figure 5: Basal indolamine secretion of the NECs. A. Formaldehyde-induced fl uorescence of indolamines in two NECs in the gill fi lament of the eel Anguilla anguilla. Note enhanced fl uorescence in the basal pole of NECs and fl uorescent nerve profi les in the subepithelial area beneath the NECs. Bar = 100 μm. B, C. Trout, Onchorynchus mykiss. Electron micrograph of the basal pole of two NECs showing enrichment with DCVs (arrowheads) opening on the subjacent basal lamina (bl). Arrows point to dense cores presumably released out of the NECs. Subepithelial nerve varicosities (2) display dense neuroplasm containing many clear vesicles with a few DCVs (arrowheads) and approach the NECs very closely. m, subepithelial smooth muscle cell; 1, nerve varicosity with clear neuroplasm and few vesicles. Bar = 1 μm.

74  |  Airway Chemoreceptors in the Vertebrates

Figure 6: Ultrastructure of 5-HT NECs in the distal region of the gill filament of Onchorynchus mykiss. 5-HT-immunoperoxidase staining. A. Two adjacent 5-HT NECs are cut in a plane tangential to the surface of FE. The upper NEC is intensely labelled and contains many DCVs; the lower displays low immunoreactivity and few DCVs. No labelling can be seen in the nuclei. B. Short basal processes (arrows) of this 5-HT immunolabelled NEC contact the basal lamina (bl) of FE. White arrowheads show DCVs. C. Of the two NECs visible on the electronmicrograph, the left one displays weak immunostaining while the flattened one contains numerous 5-HT DCVs (white arrowheads) and sends processes (arrows) between basement cells. These processes display intense immunostaining and numerous granular vesicles. A-C, bar = 1 µm.

Serotonergic Neuroepithelial Cells in Fish Gills  |  75

Figure 7: A. Cluster of NECs in the filament epithelium of the dogfish Scyliorhinus canicula. Thin processes emanating from neighbouring epithelial cells delineate a cluster of two NECs. The NECs are in very close vicinity of enlarged extracellular spaces (asterisks) and appose their basal pole on the FE basal lamina (bl), which is thinner beneath the NECs. Nerve fibres emanating from small nerve bundles form varicosities (n) very close to the basal lamina beneath the NECs. B. High magnification of the basal pole of the right NEC seen in Figure A. Arrowheads show DCVs. Bar = 1 µm.

76  |  Airway Chemoreceptors in the Vertebrates

Figure 8: NECs and surrounding media. A. NEC in the bottom of the filament epithelium of the sturgeon Acipenser baeri. Thin extensions of neighbouring cells separate a NEC from the basal lamina (bl) of the filament epithelium (FE) and from enlarged intercellular space (asterisk). Subjacent nerve varicosities (n) are very close to the basal lamina. White arrowheads show DCVs in the basal cytoplasm of the NEC. B, C. Tangential section to the surface of the filament epithelium (FE) of the trout Onchorynchus mykiss. Immunolabelled processes of NECs filled with 5-HT vesicles weave their way through the mucous cells (mc) and microvilli of epithelial cells to reach (arrow in C) the external medium (asterisks). Bar = 1 µm.

Serotonergic Neuroepithelial Cells in Fish Gills  |  77

Figure 9: Innervation of NECs and subepithelial smooth muscle cells in trout gill filament Onchorynchus mykiss. A. Bundled nerve fibres with clear neuroplasm containing a few clear vesicles and DCVs (1) run beneath a subepthelial smooth muscle cell (m). Free nerve profiles (n) approach the basal lamina of the filament epithelium at the level of a NEC. B. Two types of nerve varicosities innervate the subepithelial smooth muscle cells (m): one is clear with a few vesicles (1), the other is filled with numerous clear vesicles and DCVs and mitochondria (2). C. Type 2 nerve varicosities (2) are closely apposed to the basal lamina in front of a NEC. Below, a type 1 nerve varicosity (1) innervates a subepithelial smooth muscle cell (m). Arrowheads show DCVs in the NEC cytoplasm. Bars = 1µm.

78  |  Airway Chemoreceptors in the Vertebrates few DCVs and large mitochondria; the second type was smaller and contained a dense population of clear vesicles and a few DCVs. In both cases, this accumulation of DCVs at the NEC membrane facing the nerve profiles suggests efferent synaptic relationships with two distinct types of nerves (Figure 10B, 10C). Indeed, biogenic amine-specific histofluorescence and immunohistochemistry as well as electron microscopy combined with degeneration methods confirmed the presence of at least two types of nerves in the subepithelial plexus. The first type displayed a green formaldehyde-induced fluorescence, presumably involving a secondary amine in view of the extended time necessary for cyclization (3h). These nerves were located beneath the filament epithelium and extended all around the whole filament in all species studied (Dunel-Erb et al., 1987; Bailly, 1991). They disappeared after either surgical (preand post-trematic branchial nerves) or chemical (5-, 6-OHDA) sympathectomy. At the ultrastructural level, intact nerve varicosities as well as nerve endings degenerated either by post-trematic section (Figure 12) or by the 5- and 6-OHDA treatments (Figure 11) were seen in the subepithelial plexus and close to the NECs within the epithelium (Figure 11). Typically, osmiophilic granules accumulated in these terminals after 5-OHDA treatments (Figure 14A). In the distal half of the filament, the type 1 nerve profiles located close to the NECs were not reactive with indolaminespecific antibodies and displayed electron-lucent luminal content after chemical sympathectomy. Nerve terminals close to the subepithelial smooth musculature of the filament also displayed similar degenerating features (Figsure 11A, 11B, 11C, 12). Under the NECs, subepithelial intact nerve varicosities displayed type 2 content with clear and granular vesicles. No qualitative changes were detected within the NECs after the sympatholytic treatments (Figures 11A, 11B, 11C, 12). The second type of innervation concerns neurons and nerve fibres that innervate the efferent arteries and contribute to the subepithelial plexus, mostly in the proximal half of the filament. This innervation displayed a yellow fast-fading formaldehyde-induced fluorescence after less than 2 h cyclization time typical of a primary amine such as serotonin (Dunel-Erb et al., 1989). In all species studied except the tilapia and the sturgeon, both FITC and HRP indirect immunohistochemistry demonstrated the presence of 5-HT and 5-MT in the filament neurons and their nerve processes that innervate the efferent filament and lamellar arteries and run in the subepithelial plexus (Bailly et al., 1992; Figure 13A, 13B). The indolamine neurons send short, thick and densely ramified processes to the arterial walls and beneath the interlamellar filament epithelium at the efferent side of the lamellae. They also innervate the subepithelial parenchyma surrounding the processes of the central venous sinus by long and varicose serotonergic fibres that also approach the filament epithelium close to the NECs or connect the neurons to the filament nerve (Figure 13A, 13B, 13C). Similarly to sympatholytic treatments, severance of the branchial nerves (pre- and post-trematic branches) did modify either the yellow fluorescence or the indolamine immunoreactivity of this innervation. At the ultrastructural level, the filament neurons contain both clear and dense-cored

Serotonergic Neuroepithelial Cells in Fish Gills  |  79

Figure 10: Innervation of NECs in the trout Onchorynchus mykiss and catfish Ictalurus melas gill filament. A. Trout. Type 2 nerve varicosities (2) are closely apposed to the basal lamina (bl) in front of the basal pole of a NEC. B, C. Catfish. Nerve fibres have crossed the basal lamina (bl) of the filament epithelium and form type 1 and type 2 varicosities innervating NECs. DCVs clearly accumulate in the NEC cytoplasm close to the nerve varicosities suggesting possible afferent transmission at these synapses. Note the electron opaque thickening of synaptic membranes at the sites of vesicle accumulation (white arrows). Bars = 1 µm.

80  |  Airway Chemoreceptors in the Vertebrates

Figure 11: Chemical degeneration of the NEC sympathetic innervation. Onchorynchus mykiss. A. 5-OHDA (150mg/kg i.p.). Intact nerve varicosities (2) approach the basal lamina (bl) of the filament epithelium (FE) beneath a NEC, whereas presumptive catecholaminergic nerve fibres of the sympathetic system display empty, degenerated varicosities (1) in subepithelial nerve bundles and close to the apical pole of the NEC in FE. B, C. High magnification of the subepithelial nerves beneath the NEC viewed in 11A. D. 6-OHDA (85 mg/kg). Degenerated catecholaminergic nerve endings (1) in contact with a NEC in the filament epithelium (FE). Bars = 1 µm.

Serotonergic Neuroepithelial Cells in Fish Gills  |  81

Figure 12: Innervation of a NEC 15 days after section of the post-trematic nerve. Onchorynchus mykiss. Degenerating nerve varicosities with clear neuroplasm and collapsed organelles (1) are seen close to a subepithelial smooth muscle cell (m) and a NEC in the filament epithelium (FE). Intact nerve varicosities (2) approach the basal lamina (bl), which is thinned beneath the NEC. See the free vesicles (white arrow) between NEC and bl. Bar = 1 µm.

vesicles. The neurons exhibited a strong electron-dense HRP-labelling indicating that large amounts of serotonin are concentrated in the core of the DCVs (Bailly, 1991; Bailly et al., 1992). Nerve fibres displaying 5-HT immunostaining ran between the filament epithelium and the central venous sinus. In association with unlabelled fibres they formed numerous small bundles surrounded by a Schwann cell sheath. These bundles emitted 5-HT nerve varicosities that closely approached the wall of the central venous sinus and the basal lamina of the filament epithelium (Bailly, 1991). With the electron microscope, these varicosities were observed to contain clear and

82  |  Airway Chemoreceptors in the Vertebrates

Figure 13: Serotonin-containing neurons and NECs in the gill filament of Onchorynchus mykiss. A-C. Immunoperoxidase detection of serotonin in the neurons and NECs along the efferent edge of the filament. A. A subepithelial 5-HT-containing nerve fibre (white arrowheads) forms a varicosity “en passant” (arrow) close to the basal pole of a 5-HT-immunolabelled NEC (arrowheads). 5-HT filament neurons (stars) closely innervate efferent lamellar arteries (asterisks) and give off subepithelial fibres. FE, filament epithelium. B. High magnification of the innervated 5-HT NEC described in Figure 13A. C. This tangential section through the basal layer of the filament epithelium shows 5-MT-immunolabelled nerve fibres (white arrowheads) innervating (arrows) two 5-MT NECs (arrowheads). D. Lysing subepithelial nerve endings (2) close to the basal lamina (bl) beneath a NEC 24 h after 5,6-DHT treatment (20 mg/kg ip.). Arrow points to a DCV fusing with the basal membrane of the NEC. Bar = 40 µm in A, 50 µm in B and C, 1 µm in D.

Serotonergic Neuroepithelial Cells in Fish Gills  |  83

granular vesicles and remained intact after sectioning the branchial nerves as well as after the sympatholytic treatments. The indolamine-specific neurotoxins 5, 6- and 5, 7-DHT induced degeneration of these nerves beneath the NECs. These toxins also degenerated the indolaminergic nerve varicosities located in the thin bundles running in the subepithelial parenchyma, close to the serotonergic neurons in the proximal half of the filament and innervating the smooth muscle cells in the walls of efferent arteries. In these nerves, as in the NECs, the DCVs displayed enlarged electron-opaque core, and vacuoles contained wispy material (Figure 14). Twenty-four hours after 5, 6-DHT injection, while the NECs recovered a normal cytological aspect, damaged subepithelial varicosities were lysing or displayed damaged mitochondria, multivesicular and dense bodies, autophagic vacuoles, and abnormal cored vesicles (Figures 13C, 14B, 14C).

The Fish Gill NECs Are O2 Sensitive Hypoxia is known to cause a decrease in fluorescence intensity from NECs and to increase secretory exocytosis of dense cores from DCVs (Keith et al., 1981; Lauweryns et al., 1973, 1978). A similar phenomenon was provoked by 30 min hypoxia in the NECs of the trout gill filament epithelium. They displayed disorganized mitochondria with partial destruction of cristae. Moreover, they exhibited a relatively small number of DCVs indicating degranulation (Figure 15).

Discussion NECs as Paracrine Serotonergic Oxygen Sensors of the Fish Gill Filament Tissue The morphological studies reviewed in this report show that the gill filaments of all of the actinopterygian fish species studied contain innervated NECs that contain indolamines, as demonstrated by several approaches including pharmacologically controlled formaldehyde-induced histofluorescence and indolamine-specific immunocytochemistry. Furthermore, the NECs take up the neurotoxic serotonin analog 5,6-DHT (Bailly et al., 1992). At the ultrastructural level, the indolamines (mostly serotonin) are concentrated in DCVs that undergo paracrine exocytosis at the basal pole and processes of the NECs. These features indicate that NECs are non-neuronal sites of indolamine synthesis in the gills and that in the actinopterygian gill filament, indolamines are reliable NEC markers. This remains to be shown in the chondrichthyan NECs, which display a high basal concentration of DCVs at the ultrastructural level. These serotonergic properties of the NECs together with their strategic position at the efferent edge of the filament that cleaves the flow of the inhaled water before the respiratory lamellae suggest that they could act as oxygen sensors to the composition of the environment by enhancing local paracrine activities, similar to the pulmonary NECs of air-breathing

84  |  Airway Chemoreceptors in the Vertebrates

Figure 14: Chemical degeneration of the indolamine innervation of the subepithelial smooth muscle cells and NECs in the gill filament of Onchorynchus mykiss. A. Indolaminecontaining nerve fibres (2) display opaque DCVs (arrows) in a subepithelial nerve bundle 3 h after 5,6-DHT (15 mg/kg i.p.) treatment. Presumptive catecholaminergic 5,6-DHT–insensitive nerve fibres (1) display normal DCVs (arrowheads). B. Twenty four hours after 5,6-DHT treatment (20 mg/kg i.p.) degenerating nerve varicosities (2) have lost most organelles and display swollen vacuoles beneath a NEC. A neighbouring intact nerve varicosity (1) displays a catecholaminergic fibrelike content. Dust-like material is visible in vacuoles within the NEC (asterisks) and type 2 nerve varicosities. bl, basal lamina of the filament epithelium; m, subepithelial smooth muscle cell. C. Intact catecholaminergic-like (1) and degenerating nerve varicosities (2) innervate the subepithelial smooth muscle cells (m) beneath the NEC viewed in Figure 14B. Bars = 1µm.

vertebrates (Lauweryns and Cokelaere, 1973). Indeed, the NECs are located between inhaled oxygenated water and the arterial blood circulating in the subepithelial vasculature close to the basal pole of the NECs (i.e., lamellar arteries and processes of the central venous sinus). At this strategic place, the NECs could monitor the oxygenation of the gill tissue. In the gill filament of the dogfish, such functional relationships with the surrounding tissue are well illustrated by a large system of baso-lateral interdigi-

Serotonergic Neuroepithelial Cells in Fish Gills  |  85

Figure 15: NEC ultrastructure after 30 min hypoxia. Onchorynchus mykiss. Compare with Figures 5A and C. Note the pronounced disorganization of mitochondria and the relatively small number of DCVs. bl, basal lamina. Bar = 1 µm.

tations between the NECs and neighbouring cells and by quasi-direct contacts with the intercellular medium (Laurent, 1984). In the actinopterygian fishes, the gill NECs have fewer baso-lateral processes. In teleosts, such processes even contact the inhaled water. The degranulation of NECs that occurs in fishes (Dunel-Erb et al., 1982) and in mammals (Lauweryns and Van Lommel, 1982) during environmental hypoxia provides further evidence for a role for NECs as oxygen sensors. The immuno-electron microscopic studies reported above show that the serotonin contained in DCVs is secreted by exocytosis towards the basal cells and lamina of the filament epithelium. Thus, the serotonin released from the NECs is likely to have

86  |  Airway Chemoreceptors in the Vertebrates

Serotonergic Neuroepithelial Cells in Fish Gills  |  87

both epithelial and subepithelial targets, such as epithelial ionocytes, subepithelial vasculature and smooth musculature, endothelial cells, fibrocytes and the nerve plexus (Dunel-Erb et al., 1982, 1989; Laurent, 1984; Bailly et al., 1989, 1992; Bailly, 1991).

Epithelial and Subepithelial Targets of Indolamine Release from NECs In the teleost gills, the distal halves of the filament contain the bulk of NECs that may be responsible for the vasomotor action of 5-HT on the respiratory efferent vasculature (Östlund and Fänge, 1962; Reite, 1969; Katchen et al., 1976; Thomas et al., 1979). In this part of the filament, the efferent arteries are poorly innervated and thus are a likely primary target of the 5-HT released from the NECs. Indeed, serotonin has been shown to cause a redistribution of blood flow within the gill from the distal to the proximal part of the filaments. This is consistent with a reduced functional respiratory surface area (Sundin et al., 1995; Sundin and Nilsson, 2000). The NECs in the distal half of

Figure 16: Diagram of branchial nerves and neurons and their targets in the gill of the trout Onchorynchus mykiss. abm, filament abductor skeletal muscle; adm, filament adductor skeletal muscle; AE, gill arch epithelium; AF, apical part of the filament; afBA, afferent branchial artery; afFA, afferent filament artery; afLA, afferent lamellar artery; AH, anterior hemibranch; AVAef, efferent arterio-venous anastomosis; B, gill arch bone; BF, basal part of the filament; BV, branchial vein; CVS, central venous sinus; C, skeletal cartilage rod of the filament; efBA, efferent branchial artery; efFA, efferent filament artery; efLA, efferent lamellar artery; FE, filament epithelium; FN, filament nerve; GA, gill arch; IFE, interlamellar filament epithelium; L, lamella; NEC, neuroepithelial cell in red; NV, nutrient vasculature; PH, posterior hemibranch; Pr, pretrematic nerve; Pr.e, external branch of Pr; Pr.i, internal branch of Pr; Pt, post-trematic nerve; Pt.e, external branch of the Pt; Pt.e.a, anterior branch of Pt.e; Pt.e.c, central branch of Pt.e; Pt.i, internal branch of Pt; R, gill raker; S, septum; sp, sphincter of efFA; TB, taste bud; *, dorsal edge of the gill arch; w, water flux.   1, acetylcholinesterase-positive neurons located in the subepithelial ramifications of Pr.i and Pt.i.. They innervate R, AE and TB. 2, acetylcholinesterase-positive neurons located in Pr.e and Pt.e. They innervate sp and the proximal segment of efFA. 3, acetylcholinesterase- and 5-HT-positive neurons associated with FN emanating from Pr.e in PH and from Pt.e.a in AH. They innervate sp, efFA and proximal efLA. 4, filament 5-HT neurons. They innervate efFA, efLA, CVS and IFE. 5, subepithelial filament nerve plexus made of varicose nerve fibres emanating from the 5-HT filament neurons and of varicose catecholaminergic nerve fibres of the sympathetic system. They innervate the subepithelial filament smooth musculature of the efferent edge (not shown) and the NECs. 6, special visceral motor innervation of abm and adm, emanating from Pt.e.c. The spinal autonomic nervous system (catecholaminergic sympathetic) is not detailed. It innervates afFA, afLAs, efLAs efFA, NV, CVS and NECs.

88  |  Airway Chemoreceptors in the Vertebrates the filament may be responsible for this effect. In the proximal halves of the filament, which contain few NECs, the perfusion of these arteries may be modulated by 5-HT from the intrinsic indolaminergic neurons, which provide a rich 5-HT innervation for efferent arteries (Bailly et al., 1989). This would provide an opposing mechanism for increasing functional respiratory surface area by favouring lamellae recruitment. The complexity of the vasodilator actions of serotonin in the branchial vasomotor control is further increased as serotonin in veins is involved in both endothelium-dependent and -independent inhibitory mechanisms (Leff et al., 1987; Martin et al., 1987). Indeed, subepithelial processes of the central venous sinus come very close (15-20 µm) to the basal pole of the NECs in the distal part of the gill filaments of the trout, whereas serotonergic neurons in the proximal part provide innervation to the thin (100 nm) endothelial wall of the sinus (Bailly, 1991). Although the branchial vasomotor effects of water acidification have been shown to include a serotonergic component (Sundin and Nilsson, 2000), the involvement of 5-HT mechanisms in the branchial osmoregulatory function has not been extensively investigated. However, in many cases serotonin has potent effects on the transepithelial movements of water and electrolytes across gill epithelia (Kisloff and Moore, 1976; Berridge and Schlue, 1978; Reinach and Candia, 1978; Donowitz et al., 1980; Klyce et al., 1982; Zimmerman and Binder, 1984; Ahlman et al., 1984; Byrne and Dietz, 1997). These effects have been shown to involve adenylate cyclase activity–dependent mechanisms in the ionocytes of the lamellibranchs (Scheide and Dietz, 1986) and the trout Oncorhynchus mykiss (Guibbolini and Lahlou, 1987). In molluscs, gill cyclase activity is stimulated by neuronal 5-HT (Scheide and Dietz, 1983; Weiss and Drummond, 1985). In the fish gill, it remains unclear whether adenylate cyclase– dependent hydromineral exchanges are regulated by 5-HT from neuronal and/or neuroepithelial origin. A subepithelial smooth musculature is attached to the basal lamina of the filament epithelium, which is innervated by catecholaminergic nerve fibres of the sympathetic nervous system as other skeletal smooth muscles in the teleost gill (Dunel-Erb and Bailly, 1987) and to a lesser extent by serotonergic nerve fibres (Bailly et al., 1989, Bailly, 1991). Such a muscle layer may modulate the surface area and orientation of the filament epithelium, a mechanism that could also involve serotonin from the NECs in the distal part of the filament where serotonergic innervation is lacking.

NEC Relationships with the Filament Nervous System At the optical level using indolamine-specific histofluorescence and immunocytochemistry, we were able to demonstrate the relationships between NECs and the filament serotonergic neurons in several species including the trout and the perch. Isolated NECs were innervated by fibres sent by these neurons in the proximal part of the filament, whereas similar nerve fibres were less frequent beneath the numerous NECs

Serotonergic Neuroepithelial Cells in Fish Gills  |  89

present in the distal part of the filament. Thin, varicose catecholamine-specific nerve fibres have been revealed by the paraformaldehyde-induced fluorescence technique throughout the subepithelial parenchyma beneath the NECs (Dunel-Erb et al., 1982, 1989; Bailly, 1983, 1991; Donald, 1984; Dunel-Erb and Bailly, 1986) but were difficult to distinguish accurately from the indolamine-containing nerves using this approach. However, variations in their ultrastructural aspects confirmed that at least two types of nerve profiles innervated the NECs. Both types of nerves crossed the basal lamina of the filament epithelium and formed postsynaptic varicosities facing accumulation of DCVs in the NECs (Dunel-Erb et al., 1982; Bailly, 1983, 1991; Bailly et al., 1992). The first type was most often closely associated with NECs in the epithelium and established contacts with the subepithelial smooth muscle cells. Similar nerve endings established synapses with the serotonergic neurons in the distal part of the filament. Close to the distal NECs, these nerve endings remained unlabelled by the indolamine-specific antibodies. However, they degenerated after treatment with the catecholamine-specific neurotoxins 5- and 6-hydroxydopamines but were insensitive to the indolaminespecific neurotoxins 5, 6- and 5, 7-DHT. Furthermore, in the filament of the surgically denervated gills (sectioning of either both pre- and post-trematic nerves or just posttrematic nerve), the first type of nerve endings, presumably catecholaminergic, degenerated, whereas the serotonergic innervation remained intact throughout the filament (Dunel-Erb et al., 1989; Bailly et al., 1989, 1992). These data show that many NECs, particularly in the distal half of the filament, establish synapses with catecholaminergic sympathetic nerve fibres. This suggests that the activity of the NECs may be modulated by the sympathetic nervous system, which has been shown to modify 5-HT release from many other types of indolaminergic paraneurons (see references in Bailly et al., 1992), including the pulmonary NECs (Lauweryns et al., 1982, 1983; Redick and Hung, 1984; Lauweryns and Van Lommel, 1987). Numerous synaptic-like contacts between the NECs and nerve endings of the first type display the ultrastructural features of afferent synapses (Dunel-Erb et al., 1982; Bailly, 1991; Bailly et al., 1992), suggesting that NEC activity may modulate the functioning of the sympathetic nerves under the influence of substances such as serotonin (Mylecharane and Phillips, 1989; Wallis, 1989). The second type of nerve profiles close to the NECs was most frequently observed beneath the basal lamina of the filament epithelium just under the basal pole of the NECs, although it was also seen within the epithelium making synapses with the NECs in the proximal part of the filament. At these contacts, the nerve profiles are filled with clear vesicles and DCVs, whereas DCVs accumulate at the active zone in the NEC. This suggests that the NECs may be influenced by these nerves and reciprocally regulate the activity of the nerve fibres. This second type of nerve ending likely belongs to the 5-HT filament neurons since they remained intact after either surgical or 6- and 5-OHDA-induced chemical sympathectomy but degenerated after treatment with the indolamine-specific neurotoxins 5, 6- and 5, 7-DHT (Bailly, 1991; Bailly et al., 1992). Most of these nerve endings were seen in the subepithelial parenchyma beneath the

90  |  Airway Chemoreceptors in the Vertebrates NECs, suggesting that in addition to synaptic contacts, there may be a loose association between NECs and filament neurons in the subepithelial tissue, as described for the enterochromaffin cells in the gut mucosa (Pettersson, 1979; Ahlman et al., 1984). Some of the filament neurons are likely candidates for relaying the hypoxic signal from the NECs. They thus have afferent sensory functions, in addition to a local motor control of the efferent respiratory vasculature in response to hypoxia (Bailly et al., 1989, 1992; Bailly, 1991). In the distal half of the filament where these neurons are lacking, a direct release of diffusing neuroactive agents such as serotonin from the NECs may account for some of these effects. In the trout gill filament, twinned or isolated NECs occurred with intense or weak 5-HT immunoreactivity according to the amount of DCVs present in the cells. These differences may reflect some renewal process of the NECs, most of which are located in the developing area of the apical filament epithelium at the base of the lamellae (Laurent, 1984; Zenker et al., 1987). Based on morphological and ultrastructural criteria, the NECs probably differentiate from precursor cells of neural crest origin (Dunel-Erb et al., 1982; Bailly et al., 1992), which migrate very early into the gill arches when filaments are just beginning to form ( Jonz and Nurse, 2005, 2006). Such progenitor cells would give rise to proliferating NECs, which migrate from the arches into the filaments during development. In developing zebrafish, filament 5-HT-positive NECs become innervated a few days after NECs arrive in the gill arches, suggesting that they resemble adult O2 chemoreceptors and may already be functional. This is temporally correlated with the sudden increase in hypoxia-induced gill ventilation, indicating that a chemosensory pathway is activated in the gill at this developmental stage ( Jonz and Nurse, 2005, 2006). In the adult fish, difference in 5-HT content of NECs present in the developing apical part of the filaments may be related to the degree of NEC differentiation. This population of NECs displays quantitative seasonal variations, indicating that these cells are locally renewed (Dunel-Erb, unpublished) in a manner similar to the pulmonary NECs and enterochromaffin cells, which also differentiate from epithelial precursors (Inokuchi et al., 1983; Cutz et al., 1985a; Kruger et al., 2002). Serotonin can stimulate cell division and promote growth and differentiation of the aminergic nerve cells including the serotonergic neurons (De Vitry et al., 1986; McCobb et al., 1988; Whitaker-Azmitia and Azmitia, 1989; Hawley, 1989; Hamon and Emerit, 1989; Lauder, 1990; Gaspar, 2004). The fetal lung of mammals contains numerous well-differentiated 5-HT NECs, which proliferate further towards term, suggesting a role for these cells during embryonic life and neonatal adaptation (Cutz et al., 1985b; Chua and Perks, 1999). Whether the indolamines released by the NECs in the distal halves of the filaments are involved in the continuous development of the filament and its innervation requires further investigation. In addition, seasonal variations indicate a negative correlation between the number of NECs and the population of serotonergic neurons (Dunel-Erb, unpublished) and underline the importance of identifying the factors regulating the NEC population and more generally gill physiology.

Serotonergic Neuroepithelial Cells in Fish Gills  |  91

Figure 17: Diagram of possible interactions between the NECs and their environment and innervation in the proximal (A) and distal areas of the teleost gill filament. A. Chemoreceptive function and innervation of an isolated NEC in the proximal part of the filament. B. Paracrine secretory function, targets and innervation of the NECs in the apical area of the filament. (1) Environmental factors act directly on the NECs or through FE (O2, CO2, ions, pH, temperature, intercellular messengers). (2) Chemosensory pathway: afferent synapses on the 5-HT neurons (N) connect to the central nervous system through the filament nerve (FN). (3) Afferent synapses on catecholaminergic sympathetic nerves (functioning). (4) Substances (i.e., indolamines, peptides) diffuse from the NECs and act on the subepithelial nerve plexus (development, functioning). (5) The catecholaminergic sympathetic system modulates the activity of the NECs via efferent synaptoid contacts. (6) Active substances (hormones, ions) diffuse from the arterio-venous blood of the central venous sinus (CVS) up to the NECs. (7) Local control of respiration through an intrinsic sensorimotor way: 5-HT vasomotor control of efferent lamellar arteries by the NECs. (8) 5-HT control of CVS by the NECs (permeability and extent).

92  |  Airway Chemoreceptors in the Vertebrates The histological and ultrastructural results reviewed above indicate that fish gill NECs contain and release serotonin in their environment. Also, they establish synapses with at least two types of innervation: the sympathetic nervous system and the intrinsic filament neurons. Our knowledge of the NECs present in the respiratory organs of vertebrates suggests that they could be receptosecretory paraneurons, probably activated by both environmental and nervous stimuli. In the fish gill, NECs are probably involved in local and central modulation of branchial functions through their interactions with the branchial nervous system and paracrine secretion of substances such as serotonin (Figure 17). In these ways, the adaptative response to hypoxia could involve filament NECs in two different, but not mutually exclusive ways. Firstly, hypoxia-induced exocytotic secretion of serotonin by the NECs could have local effects directly on the efferent arterial vasculature of the filament or via stimulation of the intramural filament neurons. This would provide a local modulation of the respiratory surface. Secondly, oxygen chemoreception may depend on a centrally mediated reflex triggered by hypoxia and mediated by NECs through the afferent-like synapses, which could then send information to the central nervous system. This would in turn trigger reflex vasomotor responses via autonomic motor innervation and filament neurons. Important questions remain to be addressed in order to understand the physiological roles played by the filament NECs. One of them concerns the expression and location of the serotonin receptor subtypes, which remains completely unknown in this tissue. This issue has been partially addressed using the 5-HT1/5-HT2 receptor antagonist methysergide, which eliminated the serotonininduced vasoconstriction of gill vessels (Fritsche et al., 1992; Sundin et al., 1995) but had no effect on the arterio-venous vasodilatory effect of serotonin (Sundin, 1995), suggesting that at least two different 5-HT receptor subtypes mediate the vasomotor effects of serotonin in fish gill filaments.

References Adriaensen, D., and Scheuermann, D. 1993. Neuroendocrine cells and nerves of the lung. Anat. Rec., 236: 70-85. Adriaensen, D., Scheuermann, D., Timmermans, J., and De-Groodt-Lasseel, M. 1990. Neuroepithelial endocrine cells in the lung of the lungfish Protopterus aethiopicus: An electron and fluorescence-microscopical investigation. Acta Anat. (Basel), 139: 70-77. Ahlman, H., Grönstad, K., Nilsson, O., and Dahlström, A. 1984. Biochemical and morphological studies on the secretion of 5-HT into the gut lumen of the rat. Biogenic Amines, 1: 63-73. Bailly, Y. 1983. Recherche sur quelques innervations de la branchie des Poissons. Thesis, Cytology. Université P. et M. Curie, Paris. Bailly, Y. 1991. Organisation et morphologie des innervations branchiales chez les Téléostéens. Thesis Neurosciences. Université P. et M. Curie, Paris. Bailly, Y., and Dunel-Erb, S. 1986. The sphincter of the efferent filament artery in the teleost gill. I. Structure and parasympathetic innervation. J. Morphol., 187: 219-237.

Serotonergic Neuroepithelial Cells in Fish Gills  |  93 Bailly, Y., Dunel-Erb, S., Geffard, M., and Laurent, P. 1989. The vascular and epithelial serotonergic innervation of the actinopterygian gill filament with special reference to the trout, Salmo gairdneri. Cell Tissue Res., 258: 349-363. Bailly, Y., Dunel-Erb, S., and Laurent, P. 1992. The neuroepithelial cells of the fish gill filament: Indolamine_immunocytochemistry and innervation. Anat. Rec., 233: 143-161. Berridge, M., and Schlue, W. 1978. Ion-selective electrode studies on the effect of 5-hydroxytryptamine on the intracellular level of potassium in an insect gland. J. Exp. Biol., 78: 203-216. Burleson, M., Smatresk, N., and Milsom, W. 1992. Afferent inputs associated with cardioventilatory control in fish. In: Fish Physiology, Vol. XIIB, W. Hoar, D. Randall, and A. Farrel (Eds.). Academic Press, San Diego, pp. 389-426. Byrne, R., and Dietz, T. 1997. Ion transport and acid-base balance in freshwater bivalves. J. Exp. Biol., 200: 457-465. Chua, B., and Perks, A. 1999. The pulmonary neuroendocrine system and drainage of the fetal lung: effects of serotonin. Gen. Comp. Endocrinol., 113: 374-387. Cook, R., and King, A. 1969. A neurite receptor complex in the avian lung: electron microscopical observations. Experientia, 25: 1162-1164. Cutz, E., Yeger, H., Wong, V., Bienkowsky, E., and Chan, W. 1985a. In vitro characteristics of pulmonary neuroendocrine cells isolated from rabbit fetal lung. I. Effects of culture media and nerve growth factor. Lab. Invest., 53: 672-683. Cutz, E., Gilan, J., and Bryan, A. 1985b. Neuroendocrine cells in the developing human lung: morphologic and functional considerations. Pediatr. Pulmonol., 1: 21-29. De Vitry, F., Hamon, M., Catelon, J., Dubois, M., and Thibault, J. 1986. Serotonin initiates and autoamplifies its own synthesis during mouse central nervous system development. Proc. Natl. Acad. Sci. USA, 63: 8629-8633. Donald, J. 1984. Adrenergic innervation of the gills of brown and rainbow trout, Salmo trutta and S. gairdneri. J. Morphol., 182: 307-316. Donowitz, M., Charney, A., and Asarkof, N. 1980. Effect of serotonin on active electrolyte transport in rabbit ileum, gall bladder and colon. Am. J. Physiol., 239: 463-472. Dunel-Erb, S., and Bailly, Y. 1986. The sphincter of the efferent filament artery in teleost gills. II. Sympathetic innervation. J. Morphol., 187: 239-246. Dunel-Erb, S., and Bailly, Y., and Laurent, P. 1982. Neuroepithelial cells in fish gill primary lamellae. J. Appl. Physiol: Respirat. Environ. Exercise Physiol., 53: 1342-1353. Dunel-Erb, S., and Bailly, Y. 1987. Smooth muscles in relation to the gill skeleton of Perca fluviatilis: organization and innervation. Cell Tissue Res., 247: 339-350. Dunel-Erb, S., Bailly, Y., and Laurent, P. 1989. Neurons controlling the gill vasculature in five species of teleosts. Cell Tissue Res., 255: 567-573. Falck, B., and Owman, A. 1965. A detailed methodological description of the fluorescence method for the demonstration of biogenic amines. Acta Univ. Lund Sect., 27: 1-23. Fritsche, R., Thomas, S., and Perry, S. 1992. Effects of serotonin on circulation and respiration in the rainbow trout Oncorhynchus mykiss. J. Exp. Biol., 173: 59-73. Gaspard, P. 2004. Genetic model to understand how serotonin acts during development. J. Soc. Biol., 198: 18-21. Geffard, M., Dulluc, J., and Rock, A.-M. 1985. Antisera against the indolalkylamines: Tryptophan, 5-hydroxytryptamine, 5-methoxytryptophan and 5-methoxytryptamine tested by an enzyme-linked immunosorbent assay method. J. Neurochem., 44: 1221-1228.

94  |  Airway Chemoreceptors in the Vertebrates Gershon, M., and Sherman, D. 1982. Selective demonstration of serotonergic neurons and terminals in electron micrographs: loading with 5,7-dihydroxytryptamine and fixation with NaMnO4. J. Histochem. Cytochem., 30: 769-773. Goniakowska-Witalinska, L. 1981. Neuroepithelial bodies in the lung of the tree frog. A scanning and transmission electron microscopic study. Cell Tissue Res., 217: 435-441. Goniakowska-Witalinska, L., Zaccone, G., Fasulo, S., Mauceri, A., Licata, A., and Youson, J. 1995. Neuroendocrine cells in the gills of the bowfin Amia calva. An ultrastructural and immunocytochemical study. Folia Histochem. Cytobiol., 33: 171-177. Gonzalez, C., Almaraz, L., Obeso, A., and Rigual, R. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol. Rev., 74: 829-898. Guibbolini, M., and Lahlou, B. 1987. Basal and stimulated adenylate cyclise activity in the gill epithelium of the rainbow trout. Gen. Comp. Endocrinol., 67: 178-188. Hamon, M., and Emerit, M. 1989. Facteurs chimiques impliqués dans la differenciation neuronale. In: Médecine Périnatale (Fœtus et Nouveau-né). Flamarion, Paris, pp. 194-200. Hawley, M. 1989. Mitogenic neurotransmitters. Nature, 340: 97. Hökfelt, T., Fuxe, K., and Golstein, M. 1975. Application of immunohistochemistry to studies on mono amine cell system with special references to nervous tissue. Ann. NY Acad. Sci., 254: 407-432. Inokuchi, H., Fujimoto, S., and Kawai, K. 1983. Cellular kinetics of gastrointestinal mucosa with special references to the gut endocrine cells. Arch. Histol. Jpn., 46: 137. Jequier, E., Lovenberg, W., and Sjoerdsma, A. 1967. Tryptophan hydroxylase inhibition, the mechanism by which p-chlorophenylalanine depletes rat brain serotonin. Mol. Pharmacol., 3: 274-278. Jonz, M., and Nurse, C. 2003. Neuroepithelial cells and associated innervation of the zebrafish gill: a confocal immunofluorescence study. J. Comp. Neurol., 461: 1-17. Jonz, M., and Nurse, C. 2005. Development of oxygen sensing in the gills of zebrafish. J. Exp. Biol., 208: 1537-1549. Jonz, M., and Nurse, C. 2006. Ontogenesis of oxygen chemoreception in aquatic vertebrates. Resp. Physiol. Neurobiol., 154: 139-152. Jonz, M., Fearon, I., and Nurse, C. 2004. Neuroepithelial oxygen chemoreceptors of the zebrafish gill. J. Physiol., 560: 737-752. Katchen, M., Olson, K., and Wayne, C. 1976. Effect of histamine and serotonin on isolated perfused trout gills of rainbow trout (Salmo gairdneri). Fed. Proc., 35: 528. Keith, I., Wiley, L., and Will, J. 1981. Pulmonary neuroendocrine cells: decreased serotonin fluorescence and stable argyrophil cell numbers in acute hypoxia. Cell Tissue Res., 214: 201-205. King, A., King, D., Hodges, R., and Henry, J. 1975. Synaptic morphology of the carotid body of the domestic fowl. Cell Tissue Res., 162: 459-473. Kisloff, B., and Moore, E. 1976. Effect of serotonin on water and electrolyte transport in the in vivo rabbit small intestine. Gastroenterology, 71: 1033-1038. Klyce, S., Palkama, K., Harkonen, M., Marshall, W., Huhtanii, S., Mann, K., and Neufeld, A. 1982. Neural serotonin stimulates chloride transport in the rabbit corneal epithelium. Invest. Ophtalmol. Vis. Sci., 23: 181-182. Koe, B., and Weissman, A. 1966. p-Chlorophenylalanine: a specific depletor of brain serotonin. J. Pharmacol. Exp. Ther., 154: 499-516. Kruger, G., Mosher, J., Bixby, S., Joseph, N., Iwashita, T., and Morrison, S. 2002. Neural crest stem cells persist in the adult gut but undergo changes in self-renewal, neuronal subtype potential, and factor responsiveness. Neuron, 35: 657-669.

Serotonergic Neuroepithelial Cells in Fish Gills  |  95 Lauder, J. 1990. Ontogeny of the serotonergic system in the rat: Serotonin as a developmental signal. In: The Neuropharmacology of Serotonin. IV. Serotonin in Development and Aging. P. Whitaker-Azmitia and S. Peroutka (Eds.). Ann. NY Acad. Sci., 600: 297-314. Laurent, P. 1984. Gill internal morphology. In: Fish Physiology, Vol. X. A.W. Hoar and D. Randall (Eds.). Academic Press, New York, pp. 73-183. Lauweryns, J., and Cokelaere, M. 1973. Hypoxia-sensitive neuroepithelial bodies: Intrapulmonary secretory neuroreceptors modulated by the CNS. Z. Zellforsch. Mikrosk. Anat., 145: 521-540. Lauweryns, J., and Van Lommel, A. 1982. Morphometric analysis of hypoxia-induced synaptic activity in intrapulmonary neuroepithelial bodies. Cell Tissue Res., 226: 201-214. Lauweryns, J., and Van Lommel, A. 1983. The intrapulmonary neuroepithelial bodies after vagotomy: Demonstration of their sensory neuroreceptor-like innervation. Experientia, 39: 1123-1124. Lauweryns, J., and Van Lommel, A. 1987. Ultrastructure of nerve endings and synaptic junctions in rabbit intrapulmonary neuroepithelial bodies: A single and serial section analysis. J. Anat., 151: 65-83. Lauweryns, J., Cokelaere, M., and Theunynck, P. 1973. Serotonin producing neuroepithelial bodies in rabbit respiratory mucosa. Science, 180: 410-413. Lauweryns, J., Cokelaere, M., Lerut, T., and Theunynck, P. 1978. Cross-circulation studies on the influence of hypoxia and hypoxaemia on neuroepithelial bodies in young rabbits. Cell Tissue Res., 193: 319-339. Lauweryns, J., de Bock, V., Verhofstad, A., and Steinbusch, H. 1982. Immunohistochemical localization of serotonin in intrapulmonary neuroepithelial bodies. Cell Tissue Res., 226: 215-223. Lauweryns, J., de Bock, V., Guelinckx, P., and Decramer, M. 1983. Effects of unilateral hypoxia on neuroepithelial bodies in rabbit lungs. J. Appl. Physiol. Respirat. Environ. Exercise Physiol., 55: 1665-1668. Lauweryns, J., Van Lommel, A., and Dom, R. 1985. Innervation of rabbit intrapulmonary neuroepithelial bodies: Quantitative and qualitative ultrastructural study after vagotomy. J. Neurol. Sci., 67: 81-92. Lauweryns, J., Van Ranst, L., and Verhofstad, A. 1986. Ultrastructural localization of serotonin in the intrapulmonary neuroepithelial bodies of neonatal rabbits by use of immuno-electron microscopy. Cell Tissue Res., 243: 455-459. Leff, P., Martin, G., and Morse, J. 1987. Differential classification of vascular smooth muscle and endothelial cell 5-HT receptors by use of tryptamine analogs. Br. J. Pharmac., 91: 321-331. Martin, G., Leff, P., Cambridge, D., and Barret, V. 1987. Comparative analysis of two types of 5-hydroxytryptamine receptor mediating vasorelaxation: Different classification using tryptamines. 336: 365-373. McCobb, D., Cohan, C., Connor, J., and Kater, S. 1988. Interactive effects of serotonin and acetylcholine on neurite elongation. Neuron, 1: 377-385. Milsom, W., and Brill, R. 1986. Oxygen sensitive afferent information arising from the first gill arch of the yellow fin tuna. Respirat. Physiol., 66: 193-203. Mylecharane, E., and Phillips, C. 1989. Mechanisms of 5-hydroxytryptamine-induced vasodilatation. In: The Peripheral Actions of 5-hydroxytryptamine. J. Fozard (Ed.). Oxford University Press, Oxford, pp. 147-181. Ostlund, E., and Fänge, R. 1962. Vasodilation by adrenaline and noradrenaline, and the effects of some other substances on perfused fish gills. Comp. Biochem. Physiol., 5: 307-309.

96  |  Airway Chemoreceptors in the Vertebrates Pettersson, G. 1979. The neural control of the serotonin content in mammalian enterochromaffin cell. Acta Physiol. Scand., Suppl. 470. Redick, M., and Hung, K.S. 1984. Quantification of rabbit pulmonary neuroepithelial bodies in pre- and postnatal rabbits. Cell Tissue Res., 238: 583-587. Reinach, P., and Candia, O. 1978. Effect of tryptamine on active Na and chloride transport in the isolated bullfrog cornea. Biochim. Biophys. Acta, 510: 327-338. Reite, O. 1969. The evolution of vascular smooth muscle responses to histamine and 5-hydroxytryptamine. Acta Physiol. Scand., 75: 221-239. Rogers, D., and Haller, C. 1978. Innervation and cytochemistry of the neuroepithelial bodies in the ciliated epithelium of the toad lung (Bufo marinus). Cell Tissue Res., 195: 395-410. Saltys, H., Jonz, M., and Nurse, C. 2006. Comparative study of gill neuroepithelial cells and their innervation in teleosts and Xenopus tadpoles. Cell Tissue Res., 323: 1-10. Sanders-Bush, E., and Massari, J. 1977. Actions of drugs that deplete serotonin. Fed. Proc., 36: 2149-2153. Scheide, J., and Dietz, T. 1986. Serotonin stimulated adenylate cyclase in the gill of a fresh water mussel and its relationships to sodium transport. Physiol. Zool., 56: 585-596. Scheide, J., and Dietz, T. 1986. Serotonin regulation of gill cAMP production, Na, and water uptake in freshwater mussels. J. Exp. Zool., 240: 309-314. Sternberger, L. 1979. Immunocytochemistry. In: Basic and Clinical Immunology, 2nd ed. S. Mc Clubey (Ed.), Wiley and Sons, New York. Sundin, L. 1995. Serotonergic vasomotor control in fish gills. Braz. J. Med. Biol. Res., 28: 12171221. Sundin, L., and Nilsson, G. 2000. Branchial and circulatory responses to serotonin and rapid ambient water acidification in rainbow trout. J. Exp. Zool., 287: 113-119. Sundin, L., Nilsson, G., Block, M., and Löfman, C. 1995. Control of gill filament blood flow by serotonin in the rainbow trout, Oncorhynchus mykiss. Am. J. Physiol., 37: 1224 -1229. Thomas, S., Belaud, A., and Peyraud, C. 1979. Arguments for serotoninergic adjustments in gill blood circulation in fish. IRCS Med. Sci., 7: 543. Tranzer, J., and Richards, J. 1976. Ultrastructural cytochemistry of biogenic amines in nervous tissue: methodologic improvements. J. Histochem. Cytochem., 24: 1178-1193. Wallis, D. 1989. Interaction of 5-hydroxytryptamine with autonomic and sensory neurons. In: The Peripheral Actions of 5-hydroxytryptamine. J. Fozard (Ed.). Oxford University Press, Oxford, pp. 220-246. Wasano, K., and Yamamoto, T. 1978. Monoamine-containing granulated cells in the frog lung. Cell Tissue Res., 193: 201-209. Wasano, K., and Yamamoto, T. 1979. APUD-type receptor-secretory cells in the frog lung. Cell Tissue Res., 201: 197-205. Weiss, C., and Drummond, G. 1985. Biochemical properties of adenylate cyclise in the gill of Aplysia californica. Comp. Biochem. Physiol., 80: 251-255. Whitaker-Azmitia, P., and Azmitia, E. 1989. Stimulation of astroglial serotonin receptors produces culture media which regulates growth of serotonergic neurons. Brain Res., 497: 80-85. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature, 365: 153-155. Zaccone, G., Goniakowska-Witalinska, L., Lauweryns, J., Fasulo, S., and Tagliaferro, G. 1989. Fine structure and serotonin immunohistochemistry of the neuroendocrine cells in the lungs of Polypterus delhezi and P. ornatipinnis. Basic Appl. Histochem., 33: 277-287.

Serotonergic Neuroepithelial Cells in Fish Gills  |  97 Zaccone, G., Lauweryns, J., Fasulo, S., Tagliaferro, G., Ainis, L., and Licata, A. 1992. Immunocytochemical localization of serotonin and neuropeptides in the neuroendocrine paraneurons of teleost and lungfish gills. Acta Zool., 73: 177-183. Zaccone, G., Fasulo, S., Ainis, L., and Licata, A. 1997. Paraneurons in the gills and airways of fishes. Micr. Res. Tech., 37: 4-12. Zenker, W., Fergusson, H., Barker, I., and Woodward, B. 1987. Epithelial and pillar cell replacement in gills of juvenile trout, Salmo gairdneri Richarson. Comp. Biochem. Physiol., 86: 310-317. Zimmermann, T., and Binder, H. 1984. Serotonin-induced alteration in colonic transport in the rat. Gastroenterology, 86: 310-317.

Neurosecretory Cells (NEC’s) in the Lung of Amphibians and Accessory Respiratory Organs of the Air-breathing Fishes and in Amphibian Carotid Labyrinth: Structural Morphology and Function 4. Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes Lucyna Goniakowska-Witaliñska, Anna Pecio and Dagmara Podkowa 5. The Amphibian Carotid Labyrinth Tatsumi Kusakabe

99-124

125-140

6. Neuroendocrine System of the Amphibian Extrapulmonary Airways Luis Miguel Pastor Garcia and Esther Beltran-Frutos

141-150

7. Chemoreceptive Control of Ventilation in Amphibians and Air-Breathing Fishes Warren Burggren and Tien-Chien Pan

151-184

4 Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes Lucyna Goniakowska-Witali´nska*, Anna Pecio, and Dagmara Podkowa

Abstract In the lungs of 13 species of Amphibia investigated so far, solitary neuroendocrine (NE) cells, as well as a group of these cells called “neuroepithelial bodies” (NEB), have been observed. They are located mainly in the ciliated epithelium covering the apical part of the Ist and IInd order septa deeply protruding into the air space of the lung and may monitor gas composition. In two anuran species, Hymenochirus and Xenopus, the ciliated epithelium is absent and the NE cells do not occur, as in apodan amphibians. Substantial diversity in the structure of the NEBs has been observed. In anurans the NEBs are composed of 20-100 small NE cells, while in tailed amphibians they comprise 3-6 large cells. The NEBs are mostly of the “closed type” and the NE cells are covered by a thin layer of surrounding cells that are ciliated, goblet or pneumocytes. Only in two species, Bufo marinus and Ambystoma tigrinum, the NEBs are of “open type” and communicate with the air space via single cell equipped with microvilli and one atypical cilium with an 8+1 microtubule arrangement. NE cells possess characteristic densecored vesicles (DCVs) of various diameter, in which the serotonin and several neuropeptides have been demonstrated by immunohistochemical methods. The basal part of NEBs and some NE cells is innervated by sensory nerve terminals morphologically of afferent and efferent type. In ontogeny the NE cells in anuran amphibians appears after metamorphosis. In respiratory organs such as lungs and air bladders of the air-breathing fishes, solitary NE cells with or without innervation are found. As in other vertebrates, in the lungs of amphibians and airbreathing fishes the NE cells form an epithelial endocrine system that acts as endocrine or paracrine receptors.

Jagiellonian University, Institute of Zoology, Department of Comparative Anatomy, R. Ingardena 6, 30-060 Krakow, Poland. *Author for correspondence: Email: [email protected]

100  |  Airway Chemoreceptors in the Vertebrates Keywords: neuroendocrine cells, immunocytochemistry, amphibia, air-breathing fishes

Introduction The solitary neuroendocrine (NE) cells have been observed in the respiratory epithelium of airways and lungs of numerous tetrapod species. A group of these cells connected by nerve endings was termed “neuroepithelial bodies” (NEBs) by Lauweryns and Peuskens (1972) . The NE cells have been investigated mostly in mammals (see review in Scheuermann, 1987, 1997; Sorokin and Hoyt, 1989; Van Lommel et al., 1999). These cells were also observed in other vertebrates such as birds (Cook and King, 1969; Wasano and Yamamoto, 1979) and reptiles (Pastor et al., 1989; Scheuermann et al., 1983; Wasano and Yamamoto, 1976) as well as in the respiratory organs of air-breathing fishes (Adriaensen et al., 1990; Adriaensen and Scheuermann, 1993; Zaccone et al., 1989a, 1989b, 1994, 2006). In the lungs of tailed and anuran amphibians, NE cells are dispersed mainly in the ciliated epithelium as solitary cells or form NEBs (Adriaensen et al., 1994; Gomi et al., 1994; Goniakowska-Witalińska, 1980a, 1981, 1982, 1997; Goniakowska-Witalińska and Cutz, 1990; Goniakowska-Witalińska et al., 1990, 1992; Matsumura, 1985; Rogers and Haller, 1978, 1980; Scheuermann et al., 1989; Wasano and Yamamoto, 1979). In apodan amphibians, the ciliated epithelium is reduced and NE cells are absent (own unpublished data). Ultrastructural investigations have shown that the cytoplasm of NE cells contains characteristic dense-cored vesicles that constitute the storage sites of amine and biogenic peptides. These neuroendocrine markers have been visualized by fluorescence and immunohistochemical methods (Adriaensen et al., 1994; Bodegas et al., 1993; Cutz et al., 1986; Gomi et al., 1994; Goniakowska-Witalińska et al., 1990, 1992; Rogers and Haller, 1978; Scheuermann et al., 1989). The current review contains published observations and some new data and may explain the evolution of these endocrine systems from solitary NE cells, as in the respiratory organs of air-breathing fishes and tailed amphibians, to NEB invested with nerve endings as in the lungs of anurans and higher vertebrates.

The Structure of Amphibian Lungs The lungs of anuran Amphibia are thin-walled sacs that have internal air space divided by numerous septa of the Ist, IInd and IIIrd order into many chambers of different size (Figure 1). The apical part of septa are head-like and filled with smooth muscle cells and vein, or in the IIIrd order septa exclusively with smooth myocytes (Figure

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  101

Figure 1: Diagram representation of anuran lung with several septa dividing the internal space into several chambers of different dimensions.

2). The ciliated epithelium with goblet cells is situated exclusively on the apical part of Ist and IInd order septa. The internal lung wall, the sides of all septa and apical parts of IIIrd order septa are covered with one type of pneumocytes with an underlying network of capillaries (Dierichs, 1975; Goniakowska-Witalińska, 1981, 1986, 1994, 2001; Hermida et al., 2002; Meban, 1973).

Figure 2: Diagrams of cross-sections through the anuran lung with septa of the first (I), second (II) and third order (III) protruding deeply into the air space (A) with their head-like apical part covered by ciliated epithelium (CE). AR, artery; EE, external epithelium; LW, lung wall; SM, smooth muscle cells; V, vein.

102  |  Airway Chemoreceptors in the Vertebrates In some anuran species such as Hymenochirus and Xenopus from the aquatic family Pipidae, the ciliated epithelium is absent and the lung wall in the proximal part of the lungs is reinforced by a cartilage plate as in Hymenochirus (Figure 3) (own unpublished data) or cartilage bulbous protrusions on apical parts of Ist and IInd order septa as observed in Xenopus (Goniakowska-Witalińska, 2001). In tailed amphibians such as newts (Triturus, Notophthalmus) the lungs have a smooth internal surface (Figure 4) and a ciliated epithelium with goblet cells is located exclusively along the pulmonary artery and vein (Goniakowska-Witalińska, 1980a, 1980b, 1994; Hightower et al., 1975). The lungs of salamanders, e.g., Salamandra,

Figure 3: Semithin epon section through the lung of Hymenochirus sp. The wall lung is strengthened by cartilage (CR). The smooth internal surface is lined by capillaries covered by one type of pneumocyte. A, air space. Bar 200 μm.

Figure 4: Diagram of the cross-section of newt lungs with smooth internal surface. The ciliated epithelium (CE) with goblet cells is located along the pulmonary artery (AR) and vein (V). The remaining surface is lined by respiratory epithelium (RE) covering capillaries. A, air space; EE, external epithelium.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  103

Megalobatrachus, Hynobius, and Ambystoma, have corrugated internal surfaces due to low septa dividing the internal space into several shallow chambers of different size. The septa are lined by ciliated epithelium with goblet cells (Goniakowska-Witalińska, 1978, 1982; Goniakowska-Witalińska et al., 1992; Hashimoto et al., 1983; Matsumura, 1985; Meban, 1979). The internal surface of the lungs of apodan amphibians, e.g., Typhlonectes, is lined only by pneumocytes. A ciliated epithelium is absent (Figure 5). The apical part of low septa and thicker parts of the lung wall are strengthened by a bar of cartilage (own unpublished data). A similar situation persists in three other investigated species (Welsch, 1981). Toews and MacIntyre (1978) reported that in Typhlonectes compressicauda three lung sacs exist.

Figure 5: Semithin epon section through the lung of apodan Typhlonectes sp. A cartilage bar (CR) is present in the lung wall. A, air space; V, vein. Bar 100 μm.

NE Cells in Developing Lungs The lung sacs in amphibian larvae are at first smooth internally and without septa, with few capillaries deeply located in the lung wall. The lungs mainly have a hydrostatic role. About halfway through development, the lungs of anuran larvae become folded by low septa as in Pelobates fuscus. A few ciliated and goblet cells can be observed along the pulmonary vein and artery and the rest of the internal surface is covered by pneumocytes and a rich underlying network of capillaries (Goniakowska-Witalińska, 2001), but the NE cells are absent. The larval lungs of Bombina variegata are thin walled with numerous capillaries (Figure 6). Later, their internal surfaces become slightly folded, but the NE cells are lacking. Numerous septa may be seen just after metamorphosis, their apical parts augmented by smooth muscle cells and vein (Figure 7). The first solitary NE cells (Figure 8) in the ciliated epithelium are observed (own unpublished data) .

104  |  Airway Chemoreceptors in the Vertebrates

Figure 6: Semithin epon section of larval lung of Bombina variegata. The lung is thin walled and lined by respiratory epithelium without ciliated and goblet cells. A, air space. Bar 100 μm.

In larval lungs with a smooth internal surface (Figure 4) such as those of the newt Triturus cristatus carnifex (Goniakowska-Witalińska, 1980b), the NE cells are absent. The larval lungs of Salamandra salamandra just before metamorphosis (stage III) possess a folded internal surface consisting of low septa that divide the internal air space into several chambers. The external gills are reduced, a thin air-blood barrier is observed, and active respiration via lungs begins. Ciliated epithelium with goblet cells covers the apical part of the main septa where two types of large, solitary, noninnervated NE cells are located on the basal membrane (Table 1). Some of them extend closely along the surface epithelium but do not contact the air space. Type I NE cells

Figure 7: Semithin epon section of part of the lung of Bombina variegata just after metamorphosis with long septum enlarged in their apical part by vein and smooth muscle cells (SM) and covered by ciliated epithelium The respiratory epithelium (RE) is observed on both sides of septum and lung wall. A, air space; V, vein. Bar 100 μm.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  105

Figure 8: Semithin epon section. Larger magnification of the apical part of septum covered by ciliated epithelium containing two neuroendocrine cells with underlying vein (V). SM, smooth muscle cells; A, air space. Bar 10 μm.

possess typical small dense-cored vesicles (DCVs) 90-150 nm in diameter and larger ones ranging from 150 to 270 nm, and type II NE cells with larger DCVs 150-450 nm in diameter (Goniakowska-Witalińska, 1982).

Localization and Structure of NEB The greatest concentration of both solitary NE cells and NEBs occurs in the proximal region of the lungs in association with ciliated epithelium located on the apical part of main septa protruding into the air space (Figure 2). In anurans, e.g., in the frog Rana nigromaculata, the NEB are situated deeply in the ciliated epithelium (Wasano and Yamamoto, 1978). In species such as Rana lessonae and Bufo bufo, round and flat NEB contain a single layer of 20-25 NE cells covered with a sheet of ciliated cells (Cutz et al, 1986; Goniakowska-Witalińska and Cutz, 1990). In Rana temporaria, NEB on the main septa slightly protruded into the air space and are covered by a ciliated epithelium (Figure 9) (own unpublished data; Bodegas et al., 1993). In the toad Melanophryniscus stelzneri, NEBs are composed of 4-6 cells and are covered by pneumocytes (Hermida et al., 2003). In Bombina orientalis (Figures 10, 11) and B. variegata, NEBs protruding into the lung lumen consist of 25-100 NE cells that form three or four layers. NEBs are covered mainly with ciliated cells with the exception of those enveloped by pneumocytes, i.e., NEBs on IIIrd order septa in B. orientalis (Goniakowska-Witalińska et al., 1990; Goniakowska-Witalińska and Cutz, 1990). The Ist and IInd order septa in the lungs of the tree frog Hyla arborea are lined mainly with pneumocytes with few patches of ciliated cells in the vicinity of

106  |  Airway Chemoreceptors in the Vertebrates Table 1:

The occurrence of solitary NE cells (S), cells grouped in NEB and ranges of diameter (in nm) of dense core vesicles (DCVs) in the NE cells of Amphibia. Species

Triturus alpestris Salamandra salamandra

Numerous, small DCVs

Scarce, large DCVs

69-189

-

S

1

I 90-150 and 150-270

-

S

2

II 150-450

References

S

Hynobius nebulosus

I 46-69

-

NEB+S

3

Ambystoma tigrinum

I 70-140

300-600

NEB+S

4

II 140-260 and 320-700 Ambystoma mexicanum

I 58

NEB -

NEB+S

5

60-100

-

NEB

6

30-140 and 170-420

210-450

NEB+S

7

II 73 Bufo marinus Bufo bufo Bufo viridis Rana nigromaculata

60-110

-

NEB

8

Hyla arborea

54-110

290-862

NEB+S

9

Bombina variegata

52-130

300-700

NEB+S

7

I 60-100

-

NEB+S

10

Bombina orientalis

II 120-200 Melanophryniscus 40-100 NEB 11 stelzneri References: 1. Goniakowska-Witalińska, 1980a. 2. Goniakowska-Witalińska, 1982. 3. Matsumura, 1985. 4. Goniakowska-Witalińska et al., 1992. 5. Scheuermann et al., 1989. 6. Rogers and Haller, 1978. 7. Goniakowska-Witalińska and Cutz, 1990. 8. Wasano and Yamamoto, 1978. 9. Goniakowska-Witalińska, 1981. 10. Goniakowska-Witalińska et al., 1990. 11. Hermida et al., 2003.

dome-shaped NEBs covered exclusively by pneumocytes (Goniakowska-Witalińska, 1981). In two species of toads, Bufo marinus (Rogers and Haller, 1980) and B. viridis (Goniakowska-Witalińska and Cutz, 1990) on the main septa are special spherical protrusions into the air space, 200-300 μm in diameter and 150-180 μm in height, which are filled with connective tissue or smooth muscle cells. NEBs covered by ciliated epithelium are located in apical parts of these protrusions. In the lungs of the fully aquatic Xenopus and Hymenochirus, the ciliated epithelium is reduced and NE cells are absent (own unpublished data). In tailed amphibians such as Triturus alpestris, Salamandra salamandra and Ambystoma mexicanum, solitary NE cells have been observed in the ciliated epithelium (Goniakowska-Witalińska, 1980a, 1982; Scheuerman et al., 1989). Ambystoma tigrinum

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  107

Figure 9: Scanning electron microscope. The first and second order septa in the lung of Rana temporaria. The NEBs are located under the patches of ciliated epithelium slightly protruding into the air space. The rest of the internal surface is lined by pneumocytes covering the network of capillaries. Bar 100 μm.

Figure 10: Scanning electron microscope. The first order septum in the lung of Bombina orientalis is covered by ciliated epithelium and pneumocytes. In the central part of septum the dome-shaped NEB covered by pneumocytes with longer microvilli and few ciliated cells is seen. Bar 10 μm.

possesses solitary NE cells in between pneumocytes and NEBs consists of 3 to 6 NE cells (Figure 12) covered by goblet cells in the vicinity of ciliated cells (GoniakowskaWitalińska et al., 1992). In Hynobius nebulosus and H.n. tokyoensis, solitary NE cells and NEBs are located under a thick layer of mucous cells (Matsumura, 1985) or ciliated epithelium (Gomi et al., 1994). Observations of the apodan Typhlonectes sp. (Figure 5) revealed that the ciliated epithelium as well as NE cells are absent (own unpublished data).

108  |  Airway Chemoreceptors in the Vertebrates

Ultrastructure and Innervations The solitary NE cells and NEBs observed in amphibian lungs are of the “closed” type, i.e., they are separated from the lumen of the lung by the enveloping sheet of ciliated cells, pneumocytes or goblet cells of varying thickness (Figure 11). In some cases these sheets are very thin, from 0.1 to 4 μm, as in Ambystoma tigrinum (Figures 12, 16) or Bombina variegata (Figure 13). In some investigated species, both solitary NE cells and NEBs remain in direct contact with the basal lamina (Figures 13, 14, 15) and in other species they may be separated from it by a thin cytoplasmic layer of the surrounding cells (Figure 16). From among all investigated species of Amphibia only two, Bufo marinus and Ambystoma tigrinum, possess NEBs of the “open type”, i.e., they are in contact with air space via one cell: the apical cell in Bufo (Rogers and Haller, 1980) and NE cell type II in Ambystoma (Goniakowska-Witalińska et al., 1992). These cells on the free surface are equipped with several microvilli and single modified cilium with an 8+1 microtubule arrangement. The same situation was observed in the turtle Pseudemis scripta elegans (Pastor et al., 1987; Scheuermann et al., 1983). The NE cells are round or oval. Smaller cells are found in anuran species (ca 10 μm in diameter) in which the NEB are composed of 20 to 100 cells. Tailed amphibians possesses larger cells (to 20 μm in length) and the NEB are formed with 3-6 NE cells.

Figure 11: Diagram of NEB of “closed type” in Bombina orientalis. The numerous small NE cells (NE) with characteristic nuclei and DCVs are surrounded by ciliated cells (CC) and covered by a thin cytoplasmic layer interconnected by desmosomes. In the basal part of NEB and connective tissue (CT), numerous afferent endings (arrow) and efferent nerve endings (arrowhead) form a “nerve plexus”. The nerve endings are also located in between the cells. Some of the NE cells possess lamellar bodies. A, air space.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  109

Figure 12: Diagram of NEB of the “open type” in Ambystoma tigrinum. The 3-6 NE cells of type I (NE I) with small DCV are surrounded and covered by goblet cells (G). In some NEB, larger NE cells of type II (NE II) that accompanied the NEB are in contact with air space (A) through the narrow apical space with microvilli and one atypical cilium. This cell possesses larger DCV. The nerve endings (N) of afferent (arrow) and efferent (arrowhead) nerve endings are observed in the basal part of NEB and between NE cells. In some nerve endings the connections (asterisk) with afferent and efferent nerve ending can be seen. CC, ciliated cell; CT, connective tissue.

Figure 13: Transmission electron microscopy (TEM). Solitary NE cell (NE) in the lungs of Bombina variegata just after metamorphosis in between ciliated cell (CC) and pneumocyte (PN). The NE cell is covered by a thin cytoplasmic sheet of ciliated cells interconnected by desmosomes. The cytoplasm of NE cell contains a large nucleus with invaginations of the nuclear membrane, centrosome, Golgi complexes, mitochondria and small DCVs. A, air space; CT, connective tissue. Bar 1 μm.

110  |  Airway Chemoreceptors in the Vertebrates

Figure 14: TEM. NE cell in Bombina variegata. (a) Basal part of NE cell with numerous small DCVs. CT, connective tissue. Bar 1 μm. (b) Higher magnifications of DCVs. Bar 0.5 μm.

The NE cells in NEB adhere tightly and are connected to each other and neighbouring cells by desmosomes. In Bombina variegata, NE cells are separated from each other by thin cytoplasmic protrusions of ciliated cells that cover them. Capillary blood vessels in the connective tissue under NEB often occur. The cytoplasm of the NE cells is lighter than that of the surrounding cells. The characteristic large nuclei have numerous deep invaginations of nuclear membrane and patches of condensed chromatin (Figures 11, 12, 13, 16). The cytoplasm contains Golgi complexes, numerous long actin filaments, centrosome, few lysosomes, multivesicular bodies, numerous elongated mitochondria, as well as rough endoplasmic reticulum. In the NE cells of Bombina orientalis, several lamellar bodies also occur and constitute the storage sites of surface active material characteristic for pneumocytes (Figure 15). The function of these bodies in NE cells is unknown. In some species a single cilium between NE cells was observed. In Hyla arborea and Bombina orientalis, some NE cells in their basal part possess thin cytoplasmic protrusions (1.5 μm long) that penetrate to the connective tissue (Figure 15). Throughout the cytoplasm of NE cells but prevalently in the basal part of cells, characteristic DCVs occur (Figures 13, 14, 15, 17, 18). They are polymorphic and have an electron-dense interior with a less dense periphery between the limiting membrane and the dense centre. DCVs in different secretory phases and coated vesicles are often observed in the vicinity of the basal cell membrane. The size of DCVs varies in the investigated species (Table 1) but in most species one type of NE cells with DCV of 30-200 nm diameter occurs. Only in two species of tailed amphibians, Salamandra salamandra and Ambystoma tigrinum, was a second type of NE cells with larger DCV (140-450 nm) observed (Figures 17, 18). In the NE cells of Hyla arborea, Bufo bufo, Bombina variegata and Ambystoma tigrinum, a few large DCVs ranging from 210 to 852 nm in diameter are seen.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  111

Figure 15: TEM. Basal part of NEB of Bombina orientalis with numerous nerve endings (N). The NE cells (NE) contain elongated mitochondria, numerous small DCVs and lamellar bodies (LB) characteristic of pneumocytes. Note protrusion of NE cell (arrowhead) to the connective tissue (CT). Bar 1 μm.

The basal part of NEB is enriched by several intraepithelial sensory nerves that form “nerve plexus”. In the connective tissue near the NEB several unmyelinated profiles are also observed as well as in between NE cells (Figures 11, 12, 14, 15, 17, 18). In the axoplasm neurotubule, glycogen particles, neurofilaments, mitochondria and various types of vesicles occur. The nerve fibres are morphologically of afferent and efferent types (Figures 15, 17, 18). The afferent nerve fibres and terminals contain glycogen particles, a few small clear vesicles 40-100 nm in diameter and abundant mitochondria with long cristae. The efferent nerve fibres and terminals are characterized by various numbers of small lucent vesicles (40-100 nm in diameter) and a few DCVs 90-110 nm in diameter. In Bufo viridis, such nerve endings also possess concentric lamellated bodies (Goniakowska-Witalińska and Cutz, 1990). In most of the investigated amphibian species, afferent and efferent nerve endings were observed in different compositions. The solitary NE cells in investigated amphibian lungs are not innervated, with the exception of NE cells in Triturus alpestris, in which the ciliated epithelium contains exclusively solitary NE cells with efferent nerve endings located in the basal or apical part of cells (Goniakowska-Witalińska, 1980a). In Bombina variegata, B. orientalis and Ambystoma tigrinum, nerve endings are found in the basal part of NEB, in between the NE cells and also in the apical parts near the air space (Figures 11, 12, 15, 17, 18). In Hyla arborea, Bufo bufo, B. viridis and Bombina variegata, several sections show up to 10 nerve endings in the basal part of NEB. In Hyla the nerve profiles are predominantly of afferent type, while in three other species they are of the efferent type. In several species, synaptic junctions between nerve endings and NE cells have been observed. The synaptic zone ranges from 0.3 to 0.5 μm, and the extracellular space between synaptic membranes is about 20 nm wide (Goniakowska-Witalińska,

112  |  Airway Chemoreceptors in the Vertebrates

Figure 16: TEM. Ambystoma tigrinum. Neuroepithelial body consisting of 3 NE cells (NE) with characteristic nuclei with patches of condensed chromatin, surrounded and covered by goblet (G) cells in the vicinity of a ciliated cell (CC). Note the long protrusion of the goblet cell in between NE cells and connective tissue (CT) and thin layer of goblet cells between the air space (A) and NEB. Bar 10 μm.

Figure 17: TEM. Ambystoma tigrinum. Two types of NE cells (NE I and NE II) with DCVs and nerve endings (arrow). NE I with small DCVs and lysosome (L), and NE II with larger DCVs of different diameter. CT, connective tissue. Bar 1 μm.

1981, 1997; Goniakowska-Witalińska and Cutz, 1990; Goniakowska-Witalińska et al., 1990). Afferent and efferent nerve endings were found in the basal part of NEB and in between NE cells in the lung of the toad Melanophryniscus stelzneri (Hermida et al., 2003). In Bufo marinus (Rogers and Haller, 1978, 1980), 60% of NEB innervations consist of adrenergic axons, 20% of cholinergic axons, and the remaining 20% of both types of nerve fibres. The reciprocal synapses are also seen.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  113

Figure 18: TEM. Ambystoma tigrinum. Two NE cells of type I (NE I) with small DCVs and with nerve endings between cells of afferent type on right side and efferent type (arrow). G, goblet cell. Bar 1 μm.

In tailed amphibians such as Hynobius nebulosus, NEB cholinergic and adrenergic nerve types have been observed (Matsumura, 1985), while in Ambystoma mexicanum exclusively efferent cholinergic nerves and reciprocal synapses of these nerves were found (Scheuermann et al. 1989). Serial sections of nerve endings in the NEB of two species, Ambystoma tigrinum and Bombina orientalis, have revealed that both afferent and efferent types of terminals can be formed by the same nerve fibre that branch at the some distance from the endings (Goniakowska-Witalińska et al., 1990; Goniakowska-Witalińska, 1997). This situation was previously demonstrated in the lungs of rabbit and cat by Lauweryns and Van Lommel (1987) and Van Lommel and Lauweryns (1993).

Immunohistochemistry Histochemical investigations (Table 2) were performed on several species of Amphibia in both anurans and urodeles by fluorescence (Rogers and Haller, 1978; Scheuermann et al., 1989) or immunocytochemical methods (Adriaensen et al., 1994; Bodegas et al., 1993; Cutz et al., 1986; Gomi et al., 1994; Goniakowska-Witalińska et al., 1990, 1992). Biogenic amine serotonin is involved in the synthesis, storage and release of coexisting regulatory peptides and acts as a vaso- and bronchioconstrictor (Owman et al., 1973; Will et al., 1984). Serotonin is the principal amine present in NE cells of all investigated amphibian species with both techniques as well as in NE cells of mammals and other vertebrates (Adriaensen and Scheuerman, 1993; Sorokin and Hoyt, 1989; Van Lommel et al., 1999; Zaccone et al., 2006).

Immunohistochemical investigations in the lungs of Amphibia. Enkephalin 7B2

Ser

NSE

Bom

GRP

Calc

Leu

Met

Som

CCK

SP

VIP

Ref.

C

x

+

+/-

+

+

-

-

x

-

N

x

-

x

x

x

-

-

x

-

-

-

-

2

-

x

-

C

x

+

x

x

x

x

x

x

x

x

x

x

N

x

x

+

x

x

x

x

x

x

x

x

x

x

+

x

x

x

x

+

+

x

x

x

x

5

x

+

+/-

+

+

+

x

x

+

x

x

x

7

x

+

-

+

x

+

x

x

+

x

x

x

8

Triturus alpestris

Ambystoma mexicanum 4

Ambysoma tigrinum C Hynobius nebulosus tokyoensis C Cynops pyrrhogaster C

114  |  Airway Chemoreceptors in the Vertebrates

Table 2:

marinus C

x

+

x

x

x

x

x

x

x

x

x

x

1

C

x

+

+

-

-

-

-

x

-

-

-

-

2

N

x

-

+

+/-

+/-

-

-

x

-

-

+

-

C

x

+

+

-

-

-

-

x

-

-

-

-

N

x

-

+

+

+

-

-

x

-

-

+

-

+

+

x

x

x

x

x

x

x

x

x

x

6

x

+

x

x

x

x

x

x

x

x

x

x

3

C

x

+

+

-

-

-

-

x

-

-

-

-

2

N

x

-

+

+/-

+/-

-

-

x

-

-

+

-

Bufo viridis

Rana lessonae 2

Rana temporaria C Bombina orientalis C Bombina bombina

(Table 2 contd.)

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  115

Bufo

Bombina variegata C

x

+

+

-

-

-

+

x

+

-

-

-

N

x

-

+

+/-

+/-

-

-

x

-

-

+

-

C

x

+

+/-

-

-

-

-

x

-

+

-

-

N

x

-

+

+/-

+/-

-

-

x

-

-

+

-

2

Hyla arborea 2

C, NE cells; N, nerves; +, positive; +/-, weak positive; -, negative; x, not observed. 7B2, protein from human and porcine pituitary gland; Ser, serotonin; NSE, neuron-specific enolase; Bom, bombesin; GRP, gastrin-releasing peptide; Calc, calcitonin; Som, somatostatin; CCK, cholecystokinin; SP, substance P; VIP, vasoactive intestinal peptide. References: 1. Rogers and Haller, 1978. 2. Cutz et al., 1986. 3. Goniakowska-Witalińska et al., 1990. 4. Scheuermann et al., 1989. 5. GoniakowskaWitalińska et al., 1992. 6. Bodegas et al., 1993. 7. Gomi et al., 1994. 8. Adriaensen et al., 1994.

116  |  Airway Chemoreceptors in the Vertebrates

(Table 2 contd.)

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  117

Serial sections of NEB of Rana temporaria have shown the co-localization of both serotonin and 7B2, a protein of 23 kDa isolated from human and porcine pituitary gland (Bodegas et al., 1993). Gastrin releasing peptide (GRP) is an airway constrictor (Palmer et al., 1987) and is closely related to bombesin. In Amphibia the GRP was found in solitary NE cells of the tailed amphibians Triturus and Hynobius. The coexistence of GRP and bombesin was observed in NE cells of Triturus alpestris and also in the submucosal nerves of pulmonary septa in four species of anurans. Bombesin is a predominant peptide in the human lung (Cutz et al., 1984) and stimulates the release of other regulatory peptides and acetylcholine. It acts as a bronchioconstrictor (Belvisi et al., 1991) and as a growth factor for epithelial cells in the respiratory tract (Sunday et al., 1990). In amphibian species, it was found exclusively in the urodeles Triturus, Cynops and Hynobius. Somatostatin inhibits bronchioconstriction and secretion of regulatory peptides (Bloom, 1978) and was found in fetal and neonatal mammals. Somatostatin was found in Bombina variegata and two species of tailed amphibians. Neuron-specific enolase (NSE) is a universal marker of the diffuse neuroendocrine system (Wharton et al., 1981) and was demonstrated in various mammals (Adriaensen and Scheuermann, 1993). In amphibians, NSE was observed in solitary NE cells, NEB and associated nerves of five anuran species. The coexistence of NSE and serotonin was found in two urodele species and five anuran species. Calcitonin is a peptide often observed in mammals and lizards (Adriaensen and Scheuermann, 1993) but in amphibians it was found only in solitary NE cells of two species of urodele. Substance P (SP) was found only in the nerve fibres of five anuran species. Nerves containing SP represent a sensory type of innervation (Polak and Bloom, 1984). Met-enkephalin demonstrated in mammals (Cutz et al., 1981; Lauweryns and Van Ranst, 1987) was also observed in Bombina variegata and Ambystoma tigrinum together with leukephalin and serotonin. Cholecystokinin (CCK) was found exclusively in the NE cells of Hyla arborea, while vasoactive intestinal peptide (VIP) was not observed in the NE cells of the investigated amphibians. The presented results show a large variability of the bioactive substances in the NE cells of the investigated amphibians. The same substances were detected in mammals, suggesting a recepto-sensory function of NE cells in Amphibia.

NE Cells in Air-breathing Fishes In the respiratory organs of fish (lungs and air bladder), the NE cells are associated with the ciliated epithelium, as in Amphibia. NEBs were not observed, only solitary NE cells can be found. The NE cells are situated on a basal membrane and reach the air space by a narrow surface with microvilli.

118  |  Airway Chemoreceptors in the Vertebrates The solitary, innervated, serotonin-positive NE cells in the lungfish Protopterus aethiopicus are located in a pneumatic duct lined with ciliated epithelium with goblet cells. In the cytoplasm of NE cells, DCV ranging from 75 to 150 nm in diameter were observed (Adriaensen et al., 1990; Zaccone et al., 1989b). In the air bladder of two species of bichir Polypterus (P. ornatipinnis and P. delhezi), solitary NE cells without innervation were found in small dispersed islets of ciliated epithelium with goblet cells. Numerous DCV were found, 80-165 nm in diameter, in different stages of maturation and secretion to basement membranes (Zaccone et al., 1989a). These cells were serotonin positive by a fluorescence procedure (Scheuermann and De Groodt-Lasseel, 1982) and immunohistochemical method (Zaccone et al., 1989a) and also somatostatin positive (Adriaensen and Scheuermann, 1993). In the air bladder of the bowfin, Amia calva, three types of solitary, non-innervated NE cells were found in small patches of ciliated epithelium and in between pneumocytes: type I with small DCV 70-190 nm in diameter, type II with larger DCV 120-260 nm in diameter and contacting the air space via one cilium, and type III with the largest DCV 447-800 nm in diameter and a characteristic wide halo between dense core and limiting membrane (Goniakowska-Witalińska, 1997). The teleost catfish Pangasius hypophthalmus respires through the gills and air bladder. The internal surface of the air bladder is lined exclusively with one type of epithelial cell covering a rich network of capillaries. The epithelial cells contain lamellar bodies characteristic of pneumocytes. NE cells were not observed in the entire air bladder as well as the pneumatic duct (Podkowa and Goniakowska-Witalińska, 1998), which is probably connected with a lack of ciliated epithelium, as in Xenopus and Hymenochirus. Another teleost species, Corydoras aeneus, uses a modified posterior part of its intestine for respiration. Two types of non-innervated solitary NE cells located in the vicinity of goblet cells were found. The DCVs were dispersed in cytoplasm and range from 69 to 230 nm in diameter (Podkowa and Goniakowska-Witalińska, 2002).

Discussion Physiological experiments on mammals suggest that the NEB are hypoxia-sensitive and oxygen-sensing chemoreceptors (Lauweryns and Cokelaere, 1973; Lauweryns and Van Lommel, 1986; Youngson et al., 1993). However, no physiological studies have been performed on amphibian NEB. The complex structure and innervation of afferent and efferent nerve types in some species, such as Bufo marinus, B. bufo, B. viridis, Bombina variegata, B. orientalis and Ambystoma tigrinum, suggests chemoreceptor function for the NEB. On the other hand, the immunohistochemical investigations show the occurrence of serotonin and several regulatory peptides in NE cells of amphibians, reptiles, birds and mammals (Cutz et al., 1986; Adriaensen and Scheuermann, 1993;

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  119

Sorokin and Hoyt, 1989; Van Lommel et al., 1999) as well as in air-breathing organs in fishes (Zaccone et al., 1994, 2006). This review of amphibian and air-breathing fish reveals the substantial diversity of the endocrine system in their structure and innervation of both of NE cells and NEB. The occurrence of neuroendocrine markers suggests the possible release of active substances to the afferent nerve terminals (receptor functions) and to the basement membrane and intracellular space (paracrine function) evolving systemic reflex or local changes. Capillaries or veins were observed under the NEB in several species so the bioactive substances may also be transported by the blood. The localization of single non-innervated NE cells in the ciliated epithelium as well as between pneumocytes in air-breathing fishes and amphibians suggests a paracrine role for NE cells, i.e., local stimulation of ciliary and mucous activity or smooth muscle cells. From the phylogenetic perspective the following pattern of evolution of endocrine system in respiratory organs may be proposed: (1) the solitary non-innervated NE cells of closed type in the ciliated epithelium (as in the larval lung of Salamandra) or solitary, non- innervated cells in contact with the air space (as in air-breathing fishes Polypterus and Amia); (2) the solitary innervated NE cells of “closed type” (as in Triturus) or of “open type” (as in lungfish Protopterus); and (3) the NEB of closed type (as in most amphibians and other vertebrates) evolving to complicated structure and innervation, and NEB of “open type “ as in some amphibians and higher vertebrates. In conclusion, the solitary NE cells (in both amphibians and air-breathing fishes) as well as NEB in amphibian lungs are similar in ultrastructure, innervations and immunohistochemistry to the NEB of other terrestrial vertebrates including mammals. Their localization on strategic positions at the apical part of septa deeply protruding into the air space suggests their control of the air flow and in consequence their role as oxygen-sensing chemoreceptors.

Acknowledgements This paper was supported by DS/BiNoZ/IZ/769/2007.

References Adriaensen, D., and Scheuermann, D.W. 1993. Neuroendocrine cells and nerves of the lung. Anat. Rec., 236: 70-85. Adriaensen, D., Scheuermann, D.W., Timmermans, J.-R., and De Groodt-Laseel, M.H.A. 1990. Neuroepithelial endocrine cells in the lung of the lungfish Protopterus aethiopicus. Acta Anat. (Basel), 139: 70-77. Adriaensen, D., Scheuermann, D.W., Timmermans, J.-P., Gomi, T., Mayer, B., and De GroodtLassel, M.H.A. 1994. The neuroepithelial endocrine and nervous system in the respiratory tract of Cynops pyrrhogaster with special reference to the distribution of nitric oxide synthase and serotonin. Microsc. Res. Tech., 29: 79-89.

120  |  Airway Chemoreceptors in the Vertebrates Belvisi, M.G., Stretton, C.D., and Barnes, P.J. 1991. Bombesin-induced bronchoconstriction in the guinea pig: Mode of action. J. Pharmacol. Exp. Ther., 258: 36-41. Bloom, S.R. 1978. Somatostatin and the gut. Gastroenterology, 75: 145-147. Bodegas, M.E., Montuenga, L.M., Polak, J.M., and Sesma, P. 1993. Immunohistochemical colocalization of 7B2 and 5HT in the neuroepi­thelial bodies of the lung of Rana temporaria. Cell Tissue Res., 273: 137-140. Cook, R.D., and King, A.S. 1969. A neurite-receptor complex in the avian lung: Electron microscopical observations. Experientia, 25: 1162-1164. Cutz, E., Chan, W., and Track, N.S. 1981. Bombesin calcitonin and leuenkephalin immunoreactivity in endocrine cells of human lung. Experientia, 37: 765-767. Cutz, E., Gillan, J.E., and Track, N.S. 1984. Pulmonary endocrine cells in the developing human lung and during neonatal adaptation. In: The Endocrine Lung in Health and Disease. K.L. Becker and A.F. Gazdar (Eds.). W.B. Saunders, Philadelphia, p. 210. Cutz, E., Goniakowska-Witalińska, L., and Chan, W. 1986. An immunohistochemical study of regulatory peptides in lungs of amphibians. Cell Tissue Res., 244: 227-233. Dierichs, R. 1975. Electron microscopic studies of the lung of the frog. II. Topography of the inner surface by scanning and transmission electron microscopy. Cell Tissue Res., 160: 399-410. Gomi, T., Kikuchi, Y., Adriaensen, D., Timmermans, J.-P., De Groodt-Lassel, M.H.A., Kimura, A., Naruse, H., Ishikawa, Y., Kishi, K., and Scheuermann, D.W. 1994. Immunocytochemical survey of the neuroepithelial endocrine system in the respiratory tract of the Tokyo salamander, Hynobius nebulosus tokyoensis Tago. Histochemistry, 102: 425-431. Goniakowska-Witalińska, L. 1978. Ultrastructural and morphometric changes in the lung of european salamander, Salamandra salamandra L. Cell Tissue Res., 191: 343-356. Goniakowska-Witalińska, L. 1980a. Endocrine-like cells in the lungs of the newt Triturus alpestris Laur. Cell Tissue Res., 210: 521-524. Goniakowska-Witalińska, L. 1980b. Ultrastructural and morphometric changes in the lung of the newt, Triturus cristatus carnifex Laur. during ontogeny. Cell Tissue Res., 130: 571-583. Goniakowska-Witalińska, L. 1981. Neuroepithelial bodies in the lung of the tree frog Hyla arborea L. Cell Tissue Res., 217: 435-441. Goniakowska-Witalińska, L. 1982. Development of the larval lung of Salamandra salamandra L. A scanning and transmission electron microscopic study. Anat. Embryol., 164: 113-137. Goniakowska-Witalińska, L. 1986. Lung of the tree frog, Hyla arborea L. A scanning and transmission electron microscopic study. Anat. Embryol., 174: 379-389. Goniakowska-Witalińska, L. 1994. The lungs of Amphibia. In: The Respiratory System of Non-Mammalian Vertebrates, Vol. 2. L.M. Pastor (Ed.). Murcia University Press, Murcia, pp. 74-112. Goniakowska-Witalińska, L. 1997. Neuroepithelial bodies and solitary neuroendocrine cells in the lungs of Amphibia. Microsc. Res. Tech., 37: 13-30. Goniakowska-Witalińska, L. 2001. Development and ultrastructure of the amphibian lungs— Scanning and transmission electron microscopy study. In: Vertebrate Functional Morphology. H.M. Dutta and J.S. Datta Munshi (Eds.). Science Publishers, Inc., Enfield, Plymouth, pp. 241-265. Goniakowska-Witalińska, L., and Cutz, E. 1990. Ultrastructure of neuroendocrine cells in the lungs of three anuran species. J. Morphol., 203: 1-9. Goniakowska-Witalińska, L., Lauweryns, J.M., and Van Ranst L. 1990. Neuroepithelial bodies in the lungs of Bombina orientalis (Boul.). In: Chemoreceptors and Chemoreceptor Reflexes. H. Acker (Ed.). Plenum Press, New York, pp. 111-117.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  121 Goniakowska-Witalińska, L., Lauweryns, J.M., Zaccone, G., Fasulo, S., and Tagliafierro, G. 1992. Ultrastructure and immunocytochemistry of the neuroepithelial bodies in the lung of the tiger sala­mander, Ambystoma tigrinum (Urodela, Amphibia). Anat. Rec., 234: 419-431. Hashimoto, T., Gomi, T., Kimura, A., and Tsuchiya, H. 1983. Light and electron microscopic study of the lung of giant salamander. J. Med. Soc. Toho, Japan, 29: 724-733. Hermida, G.N., Farias, A., and Fiorito, L.E. 2002. Neuroepithelial bodies in the lung of Melanophryniscus stelzneri stelzneri (Weyenberg, 1875) (Anura, Bufonidae). Biocell, 26: 347355. Hermida, G.N., Farias, A., and Fiorito, L.E. 2003. Neuroepithelial bodies in the lung of Melanophryniscus stelzneri stelzneri (Anura, Bufonidae). Anat. Embryol., 207: 373-379. Hightower, J.A., Burke, J.D., and Haar, J.L. 1975. A light and electron microscopic study of the respiratory epithelium of the adult newt Notophthalmus viridescens. Can. J. Zool., 53: 465472. Lauweryns, J.M., and Cokelaere, M. 1973. Hypoxia-sensitive neuro-epithelial bodies: intrapulmonary secretory neuroreceptors modulated by the CNS. Z. Zellforsch. Mikrosk. Anat., 145: 521-540. Lauweryns, J.M., and Peuskens, J.C. 1972. Neuroepithelial bodies (neuroreceptor or secretory organs?) in human infant bronchial and bronchiolar epithelium. Anat. Rec., 172: 471-482. Lauweryns, J.M., and Van Lommel, A. 1986. Effect of various vagotomy procedures in the reaction to hypoxia of rabbit neuroepithe­lial bodies: Modulation by intrapulmonary axon reflexes? Exp. Lung Res., 11: 319-339. Lauweryns, J.M., and Van Lommel, A. 1987. Ultrastructure of nerve endings and synaptic junctions in rabbit intrapulmonary neuroepi­thelial bodies: A single and serial section analysis. J. Anat., 151: 65- 83. Lauweryns, J.M., and Van Ranst, L. 1987. Leu-7 immunoreactivity in human, monkey and pig bronchopulmonary neuroepithelial bodies and neuroendocrine cells. J. Histochem. Cytochem., 35: 687-690. Matsumura, H. 1985. Electron microscopic studies of the lung of the salamander, Hynobius nebulosus. III. A scanning and electron microscopic observations on neuroepithelial bodies. Okajimas Folia Anat. Jpn., 62: 187-204. Meban, C. 1973. The pneumocytes in the lung of Xenopus laevis. J. Anat., 114: 235-244. Meban, C. 1979. An electron microscope study of the respiratory epithelium in the lungs of the fire salamander (Salamandra sala­mandra). J. Anat., 128: 215-224. Owman, C., Hakanson, R., and Sundler, F. 1973. Occurrence and function of amines in endocrine cells producing polypeptide hormones. Fed. Proc., 32: 1785-1791. Palmer, J.B.D., Cuss, F.M.C., Mulderry, P.K., Ghatei, M.A., Springall, D.R., Cadieux, A., Bloom, S.R., Polak, J.M., and Barnes, P.J. 1987. Calcitonin gene-related peptide is localized to human airway nerves and potently constricts human airway smooth muscle. Br. J. Pharmacol., 91: 95101. Pastor, L.M., Ballesta, J., Hernandez, F., Perez-Tomas, R., Zuasti, A., and Ferrer, C. 1987. A microscopic study of the tracheal epithelium of Testudo graeca and Pseudemis scripta elegans. J. Anat., 153: 171-183. Pastor, L.M., Ballesta, J., Castells, M.T, Perez-Tomas, R., Marin, J.A., and Madrid, J.F. 1989. A microscopic study of the lung of Testudo graeca (Chelonia). J. Anat., 164: 19-39. Podkowa, D., and Goniakowska-Witalińska, L. 1998. The structure of the airbladder of the catfish Pangasius hypophthalmus Roberts and Vidthayanon 1991, (previously P. sutchi Fowler 1937). Fol. Biol. (Kraków), 46: 189-197.

122  |  Airway Chemoreceptors in the Vertebrates Podkowa, D., and Goniakowska-Witalińska, L. 2002. Adaptations to the air breathing in the catfish (Corydoras aeneus, Callichthydae). A histological and ultrastructural study. Fol. Biol. (Kraków), 50: 63-83. Polak, J.M., and Bloom, S.R. 1984. Regulatory peptides. Localization and measurement. In: The Endocrine Lung in the Health and Disease. K.L. Becker and A.F. Gazdar (Eds.). W.B. Saunders, Philadelphia, pp. 300-327. Rogers, D.C., and Haller, C.J. 1978. Innervation and cytochemistry of the neuroepithelial bodies in the ciliated epithelium of the toad lung (Bufo marinus). Cell Tissue Res., 195: 395-410. Rogers, D.C., and Haller, C.J. 1980. The Ultrastructural characteristics of the apical cell in the neuroepithelial bodies of the toad lung (Bufo marinus). Cell Tissue Res., 209: 485-498. Scheuermann, D.W. 1987. Morphology and cytochemistry of the endocrine epithelial system in the lung. Int. Rev. Cytol., 106: 35-88. Scheuermann, D.W., and De Groodt-Lasseel, M.H.A. 1982. Monoamine-containing granulated cells in the Polypterus lung. Verh. Anat. Ges., 76: 301-302. Scheuermann, D.W., De Groodt-Lasseel, M.H.A., Stilman, C., and Meisters, M.L. 1983. A correlative light- and electron microscopic study of neuroepithelial bodies in the lung of the red eared turtle, Pseudemis scripta elegans. Cell Tissue Res., 234: 249-269. Scheuermann, D.W., Adriaensen, D., Timmermans, J.P., and De Groodt-Lasseel, M.H.A. 1989. Neuroepithelial endocrine cells in the lung of Ambystoma mexicanum. Anat. Rec., 225: 139149. Sorokin, S.P, and Hoyt, R.F. 1989. Neuroepithelial bodies and solitary small-granule cells. In: Cell Biology. D. Massaro (Ed.). M. Dekker Inc., New York, Basel, pp. 191-344. Sunday, M.E., Hua, J., Dai, H.B., Nusrat, A., and Torday, J.S. 1990. Bombesin increases fetal lung growth and maturation in utero and in organ culture. Am. J. Respir. Cell Mol. Biol., 3: 199-205. Toews, D., and MacIntyre, D. 1978. Respiration and circulation in an apodan amphibian. Can. J. Zool., 56: 998-1005. Van Lommel, A.V., and Lauweryns, J.M. 1993. Ultrastructure and innervation of neuroepithelial bodies in the lungs of newborn cats. Anat. Rec., 236: 181-190. Van Lommel, A., Bolle, T., Fannes, W., and Lauweryns, J.M. 1999. The pulmonary neuroendocrine system: the past decade. Arch. Histol. Cytol., 62: 1-16. Wasano, K., and Yamamoto, T. 1976. Granule-containing cells in the snake respiratory mucosa. Acta Anat. Nippon, 51: 299. Wasano, K., and Yamamoto, T. 1978. Monoamine-containing granu­lated cells in the frog lung. Cell Tissue Res., 193: 201-209. Wasano, K., and Yamamoto, T. 1979. APUD-type recepto-secretory cells in the chicken lung. Cell Tissue Res., 201: 197-205. Welsch, U. 1981. Fine structure and enzyme histochemical observations on the respiratory epithelium of the caecilian lungs and gills. A contribution to the understanding of the evolution of the vertebrate respiratory epithelium. Arch. Histol. Jap., 44: 117-125. Wharton, J., Polak, J.M., Cole, G.A., Marangos, P.J., and Pearse, A.G.E. 1981. Neuron specific enolase as an immunocytochemical marker for the diffuse neuroendocrine system in human fetal lung. J. Histochem. Cytochem., 29: 1359-1364. Will, J.A., Keith, I.M., Buckner, C.K., Chacko, J., Olson, E.B., and Weir, E.K. 1984. Serotonin and the pulmonary circulation. In: The Endocrine Lung in Health and Disease. K.L. Becker and A.F. Gazdar (Eds.). W.B. Saunders, Philadelphia, p. 137.

Neuroendocrine Cells in the Lungs of Amphibians and Air-Breathing Fishes  |  123 Youngson, Ch., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature, 365: 153-155. Zaccone, G., Goniakowska-Witalińska, L., Lauweryns, J.M., Fasulo, S., and Tagliafierro, G. 1989a. Fine structure and serotonin immunohistochemistry of the neuroendocrine cells in the lungs of the bichirs Polypterus delhezi and P. ornatipinnis. Basic Appl. Histo­chem., 33: 277-287. Zaccone, G., Tagliafierro, G., Goniakowska-Witalińska, L., Fasulo, S., Ainis, L., and Mauceri, A. 1989b. Serotonin-like immunoreactive cells in the pulmonary epithelium of ancient fish species. Histochemistry, 92: 61-63. Zaccone, G., Fasulo, S., and Ainis, L. 1994. Distribution patterns of the paraneuronal endocrine cells in the skin, gills and the airways of fishes as determined by immunohistochemical and histological methods. Histochem. J., 26: 609-629. Zaccone, G., Mauceri, A., and Fasulo, S. 2006. Neuropeptides and nitric oxide synthase in the gill and the air-breathing organs of fishes. J. Exp. Zool., 305A: 428-439.

5 The Amphibian Carotid Labyrinth Tatsumi Kusakabe

Abstract The amphibian carotid labyrinth is a small maze-like vascular expansion where the common carotid artery bifurcates into the internal and external carotid arteries and has been electrophysiologically confirmed to have arterial chemo- and baroreceptor functions analogous to those of the mammalian carotid body and carotid sinus. The carotid labyrinths of anurans are spherical and those of urodeles are oblong. In the intervascular stroma of both anuran and urodelan carotid labyrinths, the glomus cells (type I cells, chief cells) are distributed singly or in clusters between connective tissue cells and smooth muscle cells and emit intense fluorescence for biogenic monoamines. In fine structure, the glomus cells are characterized by many densecored vesicles in their cytoplasm. The glomus cells have long thin cytoplasmic processes, some of which are closely associated with smooth muscle cells (g-s connection), endothelial cells (g-e connection), and pericytes (g-p connection). Afferent, efferent, and reciprocal synapses are found on the glomus cells. The morphogenesis of the carotid labyrinth starts in the larvae at the point where the carotid arch descends to the internal gills. Through the early stages of larval development, the slightly expanded region of the external carotid artery becomes closely connected with the carotid arch. Thereafter, the expanded region becomes globular, and at the final stage of metamorphosis, the carotid labyrinth is close to its adult form. In fine structure, the glomus cells appear as early as the initial stage of larval development. At the middle stages of development, the number of dense-cored vesicles increases remarkably. Distinct afferent synapses are found in juveniles, although efferent synapses can be seen during metamorphosis. The carotid labyrinth is innervated by nerve fibers containing several kinds of regulatory neuropeptides such as SP, CGRP, VIP, NPY, and others. Double-immunolabeling in combination with a multiple dye filter system demonstrates the coexistence of two different neuropeptides. The three-dimensional fine structure of vascular corrosion casts suggests that the amphibian carotid labyrinth has the appropriate architecture for controlling vascular tone, and the findings throughout metamorphosis reveal that the vascular regulatory function begins at Laboratory for Anatomy and Physiology, Department of Sport and Medical Science, Kokushikan University, 7-3-1 Nagayama, Tama, Tokyo 206-8515, Japan, Fax: F81-42-339-7238, Email: [email protected]

126  |  Airway Chemoreceptors in the Vertebrates an early stage of metamorphosis. The ultrastructural characteristics of the glomus cells during and after metamorphosis suggest that the glomus cells contribute to chemoreception after metamorphosis. In addition, immunohistochemical studies suggest that vascular regulation in the carotid labyrinth is under peptidergic innervation. Thus, the multiple functions of the carotid labyrinth underline the importance of this relatively small organ for maintenance of homeostasis and appropriate blood supply to the cephalic region.

Keywords: Carotid labyrinth; carotid body; chemoreceptor; ontogeny, peptidergic innervation; ultrastructure, immunohistochemistry; amphibia

Introduction The amphibians have a pair of small characteristic vascular expansions at the bifurcation of the common carotid artery into the internal and external carotid arteries (Adams, 1958). These vascular expansions have been called the carotid labyrinths because their appearance is that of a maze-like vasculature (Ishida, 1954; Carman, 1955; Kobayashi and Murakami, 1975; Toews et al., 1982; Kusakabe, 1990a, 2002; Kusakabe et al., 1995b). The amphibian carotid labyrinths function as a peripheral arterial chemo- and baroreceptor sensitive to changes in arterial O2 and CO2 tension, in hydrogen ion concentration, and in blood pressure (Ishii et al., 1966). Thus, the amphibian carotid labyrinths are considered to correspond to the mammalian carotid body and carotid sinus. Accordingly, the carotid labyrinths play an important role in the regulation of respiratory and cardiovascular systems. The carotid labyrinths also function in controlling the blood flow into the internal carotid artery through the intervention of the closely positioned chemoreceptor and smooth muscle cells (Ishii and Kusakabe, 1982; Kusakabe et al., 1987). Thus, the multiple functions of the carotid labyrinths underline the importance of this relatively small organ for maintenance of homeostasis and appropriate blood pressure and blood supply to cephalic regions. For a detailed history of the determination of some possible functions of the carotid labyrinths, see my earlier review (Kusakabe, 2002). In the present review, I introduce the general morphology and ontogeny of the amphibian carotid labyrinth and its innervation. In addition, I discuss the histological similarity between the amphibian carotid labyrinth and the carotid body of chronically hypoxic mammals.

Comparative Morphology of the Carotid Labyrinths in anurans and urodeles The three-dimensional structure of the carotid labyrinth in a number of species of amphibians was first studied using serial sections and reconstruction methods (Ishida,

The Amphibian Carotid Labyrinth  |  127

1954; Carman, 1955, 1967a, 1967b). Thereafter, Kobayashi and Murakami (1975) applied corrosion casting and scanning electron microscopy to observe the threedimensional fine structure of this organ in the bullfrog. Noguchi and Kobayashi (1977), Toews et al. (1982), and Kusakabe (1990a) demonstrated the three-dimensional fine structure of the carotid labyrinth of some species of urodeles. Anuran carotid labyrinths are nearly spherical in shape but differ in size (0.4-1.0 mm in diameter) from species to species (Figure 1). The carotid labyrinths are classified into two groups according to the origin of the external and internal carotid arteries (Figure 2). One group, which includes Rana, Hyla, Bufo, and others, is characterized by the presence of a vascular

Figure 1: Diagrams representing the location of the bullfrog (A) and the newt (B) carotid labyrinths, and scanning electron micrograph of a vascular cast of the carotid labyrinth. cca, common carotid artery; cl, carotid labyrinth; eca, external carotid artery; ica, internal carotid artery; mz, vascular maze (Kusakabe, 1990a).

128  |  Airway Chemoreceptors in the Vertebrates

Figure 2: Diagram showing the characteristics of the carotid labyrinths in some species of anurans (A-a, b, B), and urodeles (C). cca, common carotid artery; cch, central chamber; eca, external carotid artery; ica, internal carotid artery; ri, vascular ring. Scale bar = 100 µm (Kusakabe, 1990a).

ring at the proximal end and some vascular routes at the distal end of the labyrinth. The external and internal carotid arteries originate from these structures. The other group, which includes Xenopus, is characterized by the external carotid artery opening directly from the central chamber or the common carotid artery, and by the internal carotid artery originating from within the vascular maze. The vascular maze is more complicated in Xenopus, less so in Rana and Bufo, and simpler in Hyla. The carotid labyrinths in urodeles are oblong and have a distal tapered end (Figure 1). They are 0.6-0.8 mm long and 0.3-0.4 mm wide, and their fundamental organization is similar to that in anurans. Thus, most amphibian carotid labyrinths have the appropriate architecture for controlling vascular tone. In sections stained with hematoxylin and eosin, the amphibian carotid labyrinths are composed of a sinusoidal plexus and intervascular stroma to make a complicated maze-like structure (Figure 3). In both anurans and urodeles, the intervascular stroma contains several types of cells: glomus cells (type I cells, chief cells), smooth muscle cells, fibroblasts, mast cells, melanophores, pericytes, endothelial cells, and others (Chowdhary, 1951; Rogers, 1963; Ishii and Oosaki, 1969, Kobayashi, 1971a; PoulletKrieger, 1973; Ishii and Kusakabe, 1982; Kusakabe, 1990b, 1991a). The glomus cells are distributed singly or in clusters of 2-4 cells between connective tissue cells and smooth muscle cells, and they show a positive chromaffin reaction. In fluorescence histochemistry, the glomus cells emit intense fluorescence for biogenic monoamines (Banister et al., 1967; Kobayashi, 1971a; Böck and Gorgas, 1976; Kusakabe, 1990b). In fine structure, the glomus cells of various species of amphibians are enveloped by thin processes of type II sustentacular cells (Rogers, 1963; Ishii and Oosaki, 1969; Kobayashi, 1971b; Poullet-Krieger, 1973; Ishii and Kusakabe, 1982; Kusakabe, 1990b, 1992a, 1992b). In the pipid frog Xenopus laevis, the newt Cynops pyrrhogaster, and juveniles of the bullfrog Rana catesbeiana, a part of the glomus cell surface is directly

The Amphibian Carotid Labyrinth  |  129

Figure 3: Toluidine blue stained semithin section of the carotid labyrinth of Rana catesbeiana. The carotid labyrinth has a maze-like structure. e, endothelial cells; gc, glomus cell; l, lumen of sinusoid; m, melanophore; sm, smooth muscle cell. Scale bar = 50 µm (Kusakabe et al., 1995b).

in contact with the surrounding connective tissue, losing its covering of type II cells (Ishii and Kusakabe, 1982; Kusakabe, 1992a). In the newt glomus cells, a group of electron-dense nuclear inclusion bodies is sometimes found (Kusakabe, 1990b). The cytoplasm of glomus cells is characterized by numerous dense-cored vesicles 60-120 nm in diameter. In Xenopus laevis and Cynops pyrrhogaster, coated pits indicating the final stage of exocytosis are often found on the surface of the cell body and its processes (Figure 4) (Ishii and Kusakabe, 1982; Kusakabe, 1990b). This may indicate a secretory function in the glomus cells. In addition, some parts of the Xenopus glomus cells make close contact with intervascular smooth muscle cells without a covering of type II cell processes (Figure 5). This close contact was named the g-s connection by Ishii and Kusakabe (1982). Coated pits are also observed at the g-s connection of juvenile bullfrogs (Kusakabe, 1992a). In Xenopus, stimulation of the peripheral cut end of the glossopharyngeal nerve causes a decrease in number of dense-cored vesicles (Yates et al., 1970; Ishii and Kusakabe, 1982). Many dense-cored vesicles aggregate at the g-s connection (Figure 5), and the membrane of some vesicles fuses with the plasma

130  |  Airway Chemoreceptors in the Vertebrates

Figure 4: Some membrane invaginations (arrows) of the cytoplasmic processes (Cp). Dcv, dense-cored vesicles. Scale bar = 0.2 µm (Kusakabe, 1990b).

Figure 5: Close apposition (g-s connection) between a glomus cell (GC) and a smooth muscle cell (Sm) observed in a Xenopus carotid labyrinth. Scale bar = 1.0 µm (Ishii and Kusakabe, 1982).

membrane of the glomus cell, suggesting exocytosis. These findings indicate that the secretory product may influence smooth muscle cells under nervous control. Kusakabe et al. (1987) confirmed the role of the g-s connection in controlling vascular tone of the carotid labyrinth. Losing their covering of supporting cells, long cytoplasmic processes of the glomus cells extend toward the endothelium (Figure 6). In Cynops pyrrhogaster, the cytoplasmic processes are closely associated with the endothelial cells or the pericytes without intervening substances. These contacts have also been named the g-e connection and the g-p connection, respectively (Kusakabe, 1990b). If contentreleasing occurs in the g-e connection, catecholamines in dense-cored vesicles might be released into the blood vessels, because it is commonly accepted that the plasmalemmal vesicles in the endothelial cells are involved in intracellular transport (Palade, 1953; Simionescu et al., 1975). However, these might also be structures facilitating blood gas

The Amphibian Carotid Labyrinth  |  131

Figure 6: Cytoplasmic processes in a newt glomus cell extend into the subendothelial stroma. The luminal cytoplasm makes a g-e connection (arrow). E, endothelial cell; L, lumen of sinusoid; S, supporting cell. Scale bar, 1.0 µm (Kusakabe, 1990b).

uptake by glomus cells for chemoreception (Kusakabe, 1990b). Between two adjacent glomus cells, desmosome-like membrane thickenings are often observed in the toad Bufo japonicus, and an aggregation of dense-cored vesicles was visible in one of the glomus cells (Ishii and Oosaki, 1969). This indicates a communication between the clustered glomus cells. In Xenopus laevis, close apposition of glomus cell membranes interpreted as gap junctions is frequently observed between adjacent glomus cells (Ishii and Kusakabe, 1982). Since the glomus cells are usually arranged in clusters, excitation of one glomus cell may electrotonically spread to other cells in the cluster. Consequently, a cluster may behave as a functional unit. Enclosed by supporting cells, nerve endings lie close to the glomus cells to make synapses (Figure 7). Ishii and Oosaki (1966, 1969) described, for the first time, two types of synapses, afferent and efferent, in Bufo japonicus. Thereafter, the same synapses were reported in other species of anurans and urodeles (Kobayashi, 1971b; Kusakabe, 1990b). The first type of nerve ending is morphologically efferent and is characterized by the accumulation of clear synaptic vesicles up to 50 nm in diameter. The second type of ending is morphologically afferent and membrane thickenings are conspicuous on the glomus cell membrane where dense-cored vesicles are aggregated. In the salamanders, another type of afferent synapse is observed, in which an aggregation of clear vesicles is found in the glomus cell cytoplasm just beneath juxtaposition to a nerve ending (Kobayashi, 1971b). In addition to these two types of synapses, reciprocal synapses consisting of afferent and efferent type synapses on the same nerve ending are found in toads (Yamauchi, 1977). Serial sections have suggested that most nerve endings may be reciprocal in the newt carotid labyrinth (Kusakabe, 1990b).

Ontogeny of the Carotid Labyrinth Corrosion casting and scanning electron microscopy have also been applied for a precise analysis of the ontogenesis of the carotid labyrinth in the bullfrog during

132  |  Airway Chemoreceptors in the Vertebrates

Figure 7: A: Synaptic junctions between a glomus cell and a nerve ending in the newt carotid labyrinth. The arrow and arrowheads indicate efferent (Ne·1) and afferent (Ne·2) types of synapses, respectively. Scale bar, 0.2 µm (Kusakabe, 1990b).

larval development and metamorphosis (Kusakabe, 1991b). In fact, this method has provided several new findings that could not be detected in early observations using serial sections and reconstruction methods (Mishima, 1944a, 1944b). The morphogenesis of the carotid labyrinth starts at the point where the carotid arch descends to the internal gills. The transformation of the appearance of the carotid labyrinth can be summarized in the following six phases (Figure 8). (1) Through the early stages of larval development (stage I-V, Taylor and Kollros, 1946), the slightly expanded region of the external carotid artery becomes closely connected with the carotid arch. (2) By the last of the foot stages (stages XVII), the expanded region becomes globular. (3) At the middle of the metamorphic stages (stage XXII), a number of protuberances appear on the surface of the globular expansion. (4) At stage XXIII, these form a rudimentary vascular maze. (5) At stage XXIV, this globular expansion is completely surrounded by a primitive maze-like structure. (6) At the final stage of metamorphosis (stage XXV), the carotid labyrinth is nearly completed and is close to its adult form. From just before the completion of metamorphosis, the forepart of the carotid arch disappears. The carotid arch and the forepart of the external carotid artery are thereafter called the common carotid artery and the internal carotid artery, respectively. Ultrastructural studies in the expanded region of the external carotid artery in bullfrog larvae have shown the appearance of the glomus cells even as early as the initial stages of development (stages I-X) (Kusakabe, 1992b). The glomus cells are located singly in the septum between the carotid arch and vascular expansion. Their cytoplasm contains fewer granules, 50-60 nm in diameter. At stage III, although a few nerve fibers are close to the glomus cells, there are no ultrastructural characteristics of nerve endings. It therefore appears that the glomus cells are non-functional at the

The Amphibian Carotid Labyrinth  |  133

Figure 8: Diagram showing the characteristics of the morphogenesis of the carotid labyrinth in Rana catesbeiana at stage III (a), stage XVII (b), stage XXII (c), stage XXIII (d), stage XXIV (e), and stage XXV (f). ca, carotid arch; cca, common carotid artery; eca, external carotid artery; exp, vascular expansion; mz, vascular maze; pt, protuberance; rem, remnant vessel; ri, vascular ring; rt, vascular route (Kusakabe, 1991b).

initial stage of development. At the middle stages of development (stages XV-XXI), some glomus cells show a tendency to form small clusters. Between clustered glomus cells, gap junctions are often found. The number of dense-cored vesicles, 60-80 nm in diameter, increases remarkably, and coated pits indicating exocytosis are sometimes detected. At stage XV, efferent synaptic junctions with typical membrane thickening and accumulation of a number of clear synaptic vesicles are first observed. In addition, there are some glomus cells that make close contact with the intervascular smooth muscle cells, the g-s connection. At the metamorphic climax (stages XXII-XXV), the close apposition of the glomus cells and the neighboring cells, such as smooth muscle cells (g-s connection), endothelial cells (g-e connection), and pericytes (g-p connection), is frequently observed (Kusakabe, 1992a). The g-s connections are more frequently found in juveniles than in larvae. Distinct afferent synapses can be found in juveniles, but they cannot be identified in any larval stages (Kusakabe, 1992b). These suggest that the vascular regulatory function through the g-s connection may start at an early stage of the metamorphic climax, and that the chemosensory function may start immediately after metamorphosis.

Innervation in the Carotid Labyrinth and its Histochemical Nature The amphibian carotid labyrinth is innervated by the carotid/sinus nerve, a branch of the glossopharyngeal nerve (Figure 9) (Rogers, 1963; Ishii et al., 1966; Ishii and Oosaki, 1969). In the toad, the greater part of the carotid nerve is derived from the sympathetic nerve and partly from the vagal nerve (Ishii and Ishii, 1973). The sympathetic fibers are vasoconstrictors of the vasculature of the carotid labyrinth. In the vagal nerve, slowly conducting fibers are chemosensory and rapidly conducting fibers are barosensory.

134  |  Airway Chemoreceptors in the Vertebrates

Figure 9: Schematic representation of the innervation of the carotid labyrinth in Bufo vulgaris. a., aorta; a.t., aortic trunk; c.c.a., common carotid artery; c.l., carotid labyrinth; c.n., carotid nerve; e.c.a., external carotid artery; i.c.a., internal carotid artery; l.n., laryngeal nerve; p.c.a., pulmocutaneous artery; v.br., vagal branch (Ishii et al., 1966).

Several kinds of regulatory neuropeptides have been demonstrated in various amphibian organs with complicated vasculature (Kusakabe et al., 1994b, 1995a, 1996b, 1996c). In the carotid labyrinth, occurrence and distribution of some neuropeptides have been compared in some species of anurans and urodeles (Kusakabe et al., 1991, 1993a), and their ontogeny and coexistence have also been examined in bullfrogs (Kusakabe, 1992c; Kusakabe et al., 1993c, 1994a, 1995c). Details are given in our previous review (Kusakabe et al., 1995b). In brief, specific immunoreactivity of substance P (SP), calcitonin gene-related peptide (CGRP), vasoactive intestinal polypeptide (VIP), neuropeptide Y (NPY), somatostatin (SOM), galanin (GAL), and FMRFamide (FMRF) is recognized in the nerve fibers distributed in the intervascular stroma of the carotid labyrinth. There are some differences in the distribution and abundance of immunoreactive fibers and among the species (Kusakabe et al., 1991). Generally SP, CGRP, VIP, and NPY immunoreactive fibers are more numerous than SOM, GAL, and FMRF fibers. These fibers form complicated networks at the divergence of the intervascular stroma. Because SP, CGRP, VIP, and NPY are vasoactive in nature, the regulation of vascular tone of the labyrinth is controlled by the direct regulation of the peptidergic terminals in addition to the direct regulation of the sympathetic nerves and the indirect regulation through the g-s connection. The ontogeny of some peptidergic fibers has been demonstrated in the bullfrog (Kusakabe, 1992c). At an early stage of larval development (stage III), CGRP immunoreactive nerve fibers first appear in the wall of the external carotid artery and the carotid arch, and in a slender septum between these two vessels, but SP and VIP fibers are not yet found. At stage V, SP immunoreactive fibers first appear in the

The Amphibian Carotid Labyrinth  |  135

arterial wall and in the septum, and VIP fibers are found at an early metamorphic stage (stage XXII). From 1 to 5 weeks after metamorphosis, these peptidergic fibers increase in number to varying degrees. By 8 weeks after metamorphosis, the distribution and abundance of these fibers closely resemble those of the adults. SP, CGRP, and VIP fibers during larval development and metamorphosis may be non-functional and start to take part in the function of the carotid labyrinth only after metamorphosis. The immunohistochemical coexistence of two neuropeptides, SP and CGRP, in the carotid labyrinth was first speculated in two serial sections (Kusakabe et al., 1991). Then this speculation was confirmed using double immunohistochemical staining with an individual filter system (Kusakabe et al., 1993c). Double immunohistochemical staining revealed the coexistence of SP and CGRP in most nerve fibers in the bullfrog carotid labyrinth. In addition, the combination of the double immunofluorescence technique and alternate consecutive sections further demonstrated the possible coexistence of multiple neuropeptides (Kusakabe et al., 1994a). Later, a multiple dye filter system was applied to this field and demonstrated the precise coexistence of these neuropeptides (Kusakabe et al., 1996c, 1996d). Most fibers show the coexistence of SP and CGRP, and a few fibers show SP immunoreactivity but not CGRP immunoreactivity.

Histological Similarity between the Amphibian Carotid Labyrinth and the Carotid Body of Chronically Hypoxic Mammals The mammalian carotid bodies are enlarged after chronic hypoxic exposure (Blessing and Wolff, 1973; Heath et al., 1973; Laidler and Kay, 1975a, 1975b, 1978; Barer et al., 1976; Pequignot and Hellström, 1983; Dhillon et al., 1984; McGregor et al., 1984). In sections stained with hematoxylin eosin, the chronically hypocapnic hypoxic carotid body is found to be enlarged several fold and is very similar to the amphibian carotid labyrinth (Figure 10) (Kusakabe et al., 1993b). In addition, the following ultrastructural features of mammalian glomus cells agree with the ultrastructural characteristics of amphibian glomus cells: (1) incomplete covering of glomus cells by supporting cells; (2) long thin cytoplasmic processes in the subendothelial stroma; and (3) close apposition of the glomus cells and endothelial cells, pericytes, plasma cells, and others (Kusakabe et al., 1993b). Immunohistochemical affinity for neuropeptides in the hypoxic carotid body is also similar to that in the amphibian carotid labyrinth. In the intervascular stroma of the enlarged hypoxic carotid body, VIP, NPY, SP, and CGRP immunoreactive fibers, which are commonly found in the amphibian carotid labyrinth, are observed (Kusakabe et al., 1998). When rats are exposed to chronically hypocapnic hypoxia, the arterial O2 tension (PaO2) decreases from 94.0±2.3 to 36.2±0.6 mmHg (Hirakawa et al., 1997). This value is similar to the PaO2 in undisturbed conscious toads Bufo marinus, although the PaO2 in the toads varies widely (West et al., 1987).

136  |  Airway Chemoreceptors in the Vertebrates

Figure 10: Toluidine blue stained semithin section through the center of a chronically hypoxic rat carotid body (CB). The enlarged carotid body with great enlargement of vasculature (*) is similar to the amphibian carotid labyrinth (Figure 3). Scale bar = 100 µm (Kusakabe et al., 1993b).

The morphological similarity between the amphibian carotid labyrinth and the carotid body of chronically hypoxic rats may, in part, depend on the PaO2.

Acknowledgments I am grateful to Prof. G. Zaccone, Editor of this issue, for inviting me to submit this review article. Thanks are also due to Prof. R.C. Goris of Yokohama City University School of Medicine for his help in editing this manuscript.

References Adams, W.E. 1958. The Comparative Morphology of the Carotid Body and Carotid Sinus. Charles V Thomas, Springfield, Illinois, pp. 202-214. Banister, R.J., Portig, P.J., and Vogt, M. 1967. The content and localization of catecholamines in the carotid labyrinths and aortic arches of Rana temporaria. J. Physiol., 192: 529-535. Barer, G.R., Edwards, C., and Jplly, A.I. 1976. Changes in the carotid body and ventilatory response to hypoxia in chronically hypoxic rats. Clinic. Sci., 50: 311-313. Blessing, M.H., and Wolff, H. 1973. Befunde an Glomus Caroticum der Ratte nach Aufenhalt im einer Simulierten Höhe von 7500m. Virchows Arch. Pathol. Anat. Physiol., 360: 78-97.

The Amphibian Carotid Labyrinth  |  137 Böck, P., and Gorgas, K. 1976. Catecholamines and granule content of the carotid body type-I cells. In: Chromaffin, Enterochromaffin and Related Cells. R.E. Coupland and T. Fujita (Eds.). Elsevier, Amsterdam, pp. 355-374. Carman, J.B. 1955. The carotid labyrinth in Hyla aurea, with a note on that in Leiopelma hochstettri. J. Anat., 89: 503-525. Carman, J.B. 1967a. The carotid labyrinth in the anuran Breviceps mossambicys. Trans Roy. Soc. New Zeal. Zool., 10: 1-15. Carman, J.B. 1967b. The morphology on the carotid labyrinth in Bufo bufo and Leiopelma hochstetteri. Trans Roy. Soc. New Zeal. Zool., 10: 71-76. Chowdhary, D.S. 1951. The carotid labyrinth of Rana tigrina. Nature (Lond.), 167: 1074. Dhillon, D.P., Barer, R.R., and Walsh, M. 1984. The enlarged carotid body of the chronically hypoxic and chronically hypoxic and hypercapnic rat: A morphometric analysis. Quart. J. Exp. Physiol., 69: 301-317. Heath, D., Edwards, C., Winson, M., and Smith, P. 1973. Effects on the right ventricle, pulmonary vasculature, and carotid bodies of the rat exposed to, and recovery from, simulated high altitude. Thorax, 28: 24-28. Hirakawa, H., Nakamura, T., and Hayashida, Y. 1997. Effect of carbon dioxide on autonomic cardiovascular responses to systemic hypoxia in conscious rats. Am. J. Physiol., 273: R747-R754. Ishida, S. 1954. So-called carotic body of the amphibia. Igaku Kenkyu, 24: 1024-1050. Ishii, K., and Ishii, K. 1973. Fiber composition and derivation of afferent and efferent nerve fibers in the carotid nerve innervating the carotid labyrinth of the toad. Tohoku J. Exp. Med., 109: 323-337. Ishii, K., and Kusakabe, T. 1982. The glomus cell of the carotid labyrinth of Xenopus laevis. Cell Tissue Res., 224: 459-463. Ishii, K., and Oosaki, T. 1966. Electron microscopy of the chemoreceptor cells of the carotid labyrinth of the toad. Nature (Lond.), 212: 1499-1500. Ishii, K., and Oosaki, T. 1969. Fine structure of the chemoreceptor cell in the amphibian carotid labyrinth. J. Anat., 104: 263-280. Ishii, K., Honda, K., and Ishii, K. 1966. The function of the carotid labyrinth in the toad. Tohoku J. Exp. Med., 88: 103-116. Kobayashi, S. 1971a. Comparative cytological studies of the carotid body. 1. Demonstration of monoamine-storing cells by correlated chromaffin reaction and fluorescence histochemistry. Arch. Histol. Jpn., 33: 319-339. Kobayashi, S. 1971b. Comparative cytological studies of the carotid body. 2. Ultrastructure of the synapses on the chief cell. Arch. Histol. Jpn., 33: 387-420. Kobayashi, S., and Murakami, T. 1975. Scanning electron microscopic observation of the fine threedimensional distribution of the blood vessels in the frog carotid labyrinth. In: The Peripheral Chemoreceptors. M.J. Purxes (Ed.). Cambridge Univ. Press, Cambridge, pp. 301-313. Kusakabe, T. 1990a. Comparative studies of the vascular organization of carotid labyrinths of anurans and caudates. J. Morphol., 204: 47-55. Kusakabe, T. 1990b. Ultrastructural studies of the carotid labyrinth in the newt Cynops pyrrhogaster. Zool. Sci., 7: 201-208. Kusakabe, T. 1991a. The occurrence of melanosomes in the newt glomus cell. Arch. Histol. Cytol., 54: 81-87. Kusakabe, T. 1991b. Morphogenesis of the carotid labyrinth in the bullfrog, Rana catesbeiana, during larval development and metamorphosis. Anat. Embryol., 184: 133-139.

138  |  Airway Chemoreceptors in the Vertebrates Kusakabe, T. 1992a. Intimate apposition of the glomus and smooth muscle cells (g-s connection) in the carotid labyrinth of juvenile bullfrogs. Anat. Embryol., 185: 39-44. Kusakabe, T. 1992b. Ultrastructural characteristics of glomus cells in the external carotid artery during larval development and metamorphosis in the bullfrogs, Rana catesbeiana. Anat. Rec., 233: 461-466. Kusakabe, T. 1992c. Ontogeny of substance P-, CGRP-, and VIP-containing nerve fibers in the amphibian carotid labyrinth of the bullfrog, Rana catesbeiana: an immunohistochemical study. Cell Tissue Res., 269: 79-85. Kusakabe, T. 2002. Carotid labyrinth of amphibians. Microsc. Res. Tech., 56: 207-226. Kusakabe, T., Ishii, K., and Ishii, K. 1987. A possible role of the glomus cell in the controlling vascular tone of the carotid labyrinth of Xenopus laevis. Tohoku J. Exp. Med., 151: 395-408. Kusakabe, T., Anglade, P., and Tsuji, S. 1991. Localization of substance P, CGRP, VIP, neuropeptide Y, and somatostatin immunoreactive nerve fibers in the carotid labyrinths of some amphibian species. Histochemistry, 96: 255-260. Kusakabe, T., Kawakami, T., and Takenaka, T. 1993a. Distribution of FMRFamide-immunoreactive nerve fibers in the carotid labyrinth of the bullfrog, Rana catesbeiana in corresponding differential interference-contrast (Nomarski) images. Arch. Histol. Cytol., 56: 277-282. Kusakabe, T., Powell, F.L., and Ellisman, M.H. 1993b. Ultrastructure of the glomus cells in the carotid body of chronically hypoxic rats: with a special reference of the similarity of the amphibian glomus cells. Anat. Rec., 237: 220-227. Kusakabe, T., Kawakami, T., Tanabe, Y., Fujii, S., Bandou, Y., and Takenaka, T. 1993c. Coexistence of substance P and calcitonin gene-related peptide in the nerve fibers of the carotid labyrinth of the bullfrog, Rana catesbeiana. Brain Res., 603: 153-156. Kusakabe, T., Kawakami, T., and Takenaka, T. 1994a. Coexistence of substance P, neuropeptide Y, VIP, and CGRP in the nerve fibers of the carotid labyrinth of the bullfrog, Rana catesbeiana: a double labelling immunofluorescence study in combination with alternate consecutive sections. Cell Tissue Res., 276: 91-97. Kusakabe, T., Kawakami, T., and Takenaka, T. 1994b. Localization of immunoreactive neuropeptides in the kidney of the bullfrog, Rana catesbeiana, by immunofluorescence. Cell Tissue Res., 278: 41-44. Kusakabe, T., Kawakami, T., and Takenaka, T. 1995a. Calcitonin gene-related peptide and substance P in the pharynx and lung of the bullfrog, Rana catesbeiana. Cell Tissue Res., 279: 115-121. Kusakabe, T., Kawakami, T., and Takenaka, T. 1995b. Peptidergic innervation in the amphibian carotid labyrinth. Histol. Histopathol., 10: 185-202. Kusakabe, T., Kawakami, T., Ono, M., Hori, H., and Takenaka, T. 1995c. Localization of galanin-immunoreactive nerve fibers in the carotid labyrinth of the bullfrog, Rana catesbeiana: comparison with substance P-immunoreactive fibers. Cell Tissue Res., 281: 63-67. Kusakabe, T., Kawakami, T., and Takenaka, T. 1996a. Coexistence of neuropeptides in the amphibian carotid labyrinth: An application of double immunolabelling in combination with a multiple dye filter. Adv. Exp. Med. Biol., 410: 325-328. Kusakabe, T., Kawakami, T., Ono, M., and Takenaka, T. 1996b. Distribution of peptidergic nerve fibers in the bullfrog lingual papillae demonstrated by a combination of double immunofluorescence labelling and a multiple dye filter. Histochem. J., 28: 289-297. Kusakabe, T., Matsuda, H., Kawakami, T., Syoui, M., Kurihara, K., and Takenaka, T. 1996c. Ontogeny of neuropeptide-containing fibers in the bullfrog taste organ. Dev. Brain Res., 96: 285-289.

The Amphibian Carotid Labyrinth  |  139 Kusakabe, T., Kawakami, T., Ono, M., Syoui, N., Kurihara, K., and Takenaka, T. 1996d. Precise coexistence of regulatory peptides in the nerve fibers of the amphibian carotid labyrinth demonstrated by a combination of double immunofluorescence labelling and a multiple dye filter. Brain Res. 735: 307-310. Kusakabe, T., Hayashida, Y., Matsuda, H., Gono, Y., Kawakami, T., Powell, F.L., Ellisman, M.H., and Takenaka, T. 1998. Hypoxic adaptation of the peptidergic innervation in the rat carotid body. Brain Res., 806: 165-174. Laidler, P., and Kay, J.M. 1975a. The effect of chronic hypoxia on the number and nuclear diameter of type I cells in the carotid bodies of the rats. J. Pathol., 79: 311-320. Laidler, P., and Kay, J.M. 1975b. A quantitative morphological study of the carotid bodies of the rats living at a simulated altitude of 4300 meters. J. Pathol., 117: 183-191. Laidler, P., and Kay, J.M. 1978. A quantitative study of some ultrastructural features of the type I cells in carotid bodies of rats living at a simulated altitude of 4300 m. J. Neurocytol., 7: 183-192. McGregor, K.H., Gil, J., and Lahiri, S. 1984. A morphometric study of the carotid body in chronically hypoxic rats. J. Appl. Physiol., 57: 1430-1438. Mishima, D. 1944a. On the development of the carotid gland of anurans. I. In Rhacophprous arborea schlegelli. Kaibougaku Zasshi, 22: 399-408. Mishima, D. 1944b. On the development of the carotid gland of anurans. II. In Bufo formosus. Kaibougaku Zasshi, 22: 399-408. Noguchi, R., and Kobayashi, S. 1977. On the vascular architecture of the carotid labyrinth in Cynops pyrrhogaster and Onychodactylus japonicus. Arch. Histol. Jpn. 40: 347-360. Palade, G.E. 1953. Fine structure of blood capillaries. J. Appl. Physics, 24: 1424. Palme, F. 1934. Die Paraganglein über dem Herzen und im Endingungsgebiet des Nervus Depressor. Z. Mikrosk. Anat. Forsch., 36: 391-420. Pequignot, J.M., and Hellström, S. 1983. Intact and sympathectomized study of the carotid bodies of long-term hypoxic rats: A morphometric light microscopical study. Virchows Arch. Pathol. Anat., 400: 235-243. Poullet-Krieger, M. 1973. Innervation du labyrinthe carotidien du crapaud Bufo bufo: etude ultrastructurale et histochimique. J. Microsc. (Paris), 18: 55-64. Rogers, D.C. 1963. Distinct cell types in the carotid labyrinth. Nature (Lond.), 200: 492-493. Rogers, D.C. 1966. A histological and histochemical study of the carotid labyrinth in the anuran amphibians, Bufo marinus, Hyla aurea and Neobatrachus pictus. Acta Anat., 63: 249-280. Simionescu, N., Simionescu, M., and Palade, G.E. 1975. Permeability of muscle capillaries to small heme-peptides. Evidence for the existence of patent transendothelial channels. J. Cell Biol., 64: 586-607. Taylor, A.C., and Kollros, J.J. 1946. Stages in the normal development of Rana pipiens larvae. Anat. Rec., 94: 7-23. Toews, D., Shelton, G., and Boutilier, R. 1982. The amphibian carotid labyrinth: some anatomical and physiological relationships. Can. J. Zool., 60: 1153-1160. West, N.H., Topor, Z.L., and Van Vliet, B.N. 1987. Hypoxemic threshold for lung ventilation in the toad. Respir. Physiol., 70: 377-390. Wilson, D.A., O’Neill, J.T., Saido, S.I., and Traystman, R.J. 1981. Vasoactive intestinal polypeptide and canine cerebral circulation. Circ. Res., 48: 138-148. Yamauchi, A. 1977. On the recepto-endocrine property of granule-containing (GC) cells in the autonomic nervous system. Arch. Histol. Jap. 40 (Suppl.): 147-161. Yates, R.D., Chen, I., and Duncan, D. 1970. Effects of sinus nerve stimulation on the carotid body glomus cells. J. Cell Biol., 46: 544-552.

6 Neuroendocrine System of the Amphibian Extrapulmonary Airways Luis Miguel Pastor García* and Esther Beltrán-Frutos Abstract The histology of the epithelium of extrapulmonary airways of amphibian is revised in this article, with special reference to the neuroendocrine system. Cell types of the epithelium differ, especially in anurans. Several regulatory peptides can be identified in the laryngotracheal epithelium. Cells positive to bombesin are found only in amphibian and urodele species, while serotonin cells are found only in the urodele. The ultrastructure of the neuroendocrine cells is similar to that described for the lung neuroepithelial bodies of the urodele or amphibian species. Intraepithelial nerves are only found in the anuran laryngeal epithelium. More studies are required to understand the neoroendocrine system in amphibians.

Keywords: Larynx, trachea, epithelium, neuroendocrine cells, nerves, amphibian

Introduction: Extrapulmonary Airways Epithelium of Amphibian1 Before showing the few data of neuroendocrine cells in amphibian extrapulmonary airways is important to know the histology of this area. In Amphibia, little attention has been paid to the epithelium of the extrapulmonary airways, which have only been studied only in the urodele Hinobius nebulosus (light and electron microscopy) Department of Cellular Biology and Histology, Medical School, University of Murcia, 30100, Espinardo, Murcia, Spain, Telephone: (34)-968-363949, Fax: (34)-968-364150, Email: bioetica@ um.es *Author for Correspondence. 1 Part of this text is extracted, with editorial permission, from Histology, Ultrastructure and Immunohistochemistry of the Respiratory Organs in Non-Mammalian Vertebrates, L.M. Pastor (Ed.), Ediciones Secretariado de Publicaciones de la Universidad de Murcia, 1995.

142  |  Airway Chemoreceptors in the Vertebrates (Matsumura et al., 1986), in the anuran Rana perezi (light and lectin histochemistry) (Castells et al., 1990) (transmission and scanning electron microscopy) (Pastor et al., 1998) and Apoda Siphonops annulatus (light microscopy) (Kuehne and Junqueira, 2000). Also some histological dates of this epithelium are shows in other articles (Urodela: Naruse et al., 2005 Cynops pyrrhogaster; Anuran: Zaccone et al., 2004 Rana esculenta and Discoglossus pictus; Bodegas et al., 1995a, Bodegas et al., 1995b Rana temporaria). The trachea of Apoda Siphonops annulatus is mainly made up by incomplete cartilage rings lined by a respiratory epithelium (ciliated and mucous cells) of variable morphology, according to the region observed. In this area a rich vascularization is observed (Kuehne and Junqueira, 2000). The urodele laryngotracheal epithelium is constituted by basal, goblet and ciliated cells (Matsumura et al., 1986; Gomi et al., 1994). In the arytenoid region of Hynobius nebulosus there is a ciliated columnar epithelium, with goblet and ciliated mucous cells, which present some folds in the ventral and dorsal parts. In the cricotracheal region, the epithelium is formed by flat or cuboidal mucous cells, and ciliated cells are only sporadically observed. Both the goblet and mucous cells of the arytenoid and cricotracheal regions show abundant acidic sulphate mucopolysaccharides (Matsumura et al., 1986). Also in Rana esculenta and Discoglossus pictus this epithelium is formed by a pseudostratified ciliated mucous epithelium (Zaccone et al., 2004). The laryngotrachea in Cynops pyrrhogaster (Naruse et al., 2005) is a tubular structure that opens ventromedially into the esophagus via an elongated opening, the aditus laryngis. Five regions can be differentiates. In crosssection, the lateral cartilages initially appear round, becoming triangular and sickleshaped towards the distal regions. In the middle portion of the tubular laryngotrachea, the laryngeal sphincter, consisting of striated muscle, surrounds the lateral cartilage plates. The lining of the laryngotrachea is the pseudostratified ciliomucous epithelium, where mucous cells are abundant among ciliated cells. A fibromuscular layer was observed between the epithelium and lateral cartilages and beneath both the dorsal and ventral surfaces of the epithelium. Blood vessels were also observed in this layer. The laryngeal cavity in Rana perezi is covered by a pseudostratified epithelium, non-ciliated over most of its surface, which is constituted by cuboidal secretory cells, which are not of goblet type, and basal cells. Also, some patches of pseudostratified epithelium composed of ciliated, basal and goblet cells are located next to the aditus laryngis or in areas next to the insertion of the vocal cords into the laryngeal cavity. In the posterior chamber simple tubuloalveolar glands are observed (Pastor et al., 1998; Castells et al., 1990). The short bronchi of Rana perezi are covered by a ciliated pseudostratified epithelium with goblet cells. Histochemically, the secretory cells are weakly PAS-positive and show reactivity to the lectins DBA and WGA, while the goblet cells also react with PNA and Con-A (Castells et al., 1990). The only urodele species studied ultrastructurally has been Hynobius nebulosus (Matsumura et al., 1986). Scanning electron microscopy distinguishes an arytenoid

Neuroendocrine System of the Amphibian Extrapulmonary Airways  |  143

and a cricotracheal portion in the larynx of this species, part of which communicates with the air sacs. Abundant ciliated and non-ciliated cells can be observed. The latter are classified as mucous cells that are individually located among the ciliated cells, and that present a slightly elevated surface with short microvilli. In the cricotracheal region the ciliated cells are numerous in the lateral portions of the larynx, and the mucous cells show more microvilli. Alongside these mucous cells, other cells can be found that have microfolds on their apical surface forming a complicated network. These cells are abundant in the most caudal part of this area (Matsumura et al., 1986). With transmission electron microscopy the laryngotracheal epithelium is seen to be pseudostratified and to mainly consist of ciliated, goblet and basal cells. In the arytenoid region, the mucous cells present a morphology typical of goblet cells, and the ciliated and basal cells are similar to those observed in the epithelium of mammalian airways. In the cricotracheal portion, the cells of the epithelium decrease in height, becoming cuboidal or squamous. The goblet cells present a different morphology; the mucous granules are small and rounded or elliptical and are situated in the apical cytoplasm of the cells. This same cell type, more squamous and with few granules, appears in the first portions of the epithelium where respiratory exchange takes place and has some laminar bodies in its cytoplasm. These cells would have two roles in the larynx: producing mucus and phospholipids and intervening in gaseous exchange. In this same cricotracheal region, other mucous cells are present that show the characteristic of possessing numerous supranuclearly distributed mitochondria. For this Matsumura et al. (1986), these are cells which present microfolds under the scanning electron microscope. Together with these cells in both portions of the larynx, endocrine cells and macrophages are located in the laryngeal epithelium. Scanning electron microscopy of the surface of the frog larynx, both in the anterior and posterior chambers, only showed cells with a polygonal morphology, marked cellular boundaries and apical microvilli. At higher magnification, the structure of the apical microvilli, as well as the presence of small hollows in the apical membrane that possibly derive from a secretory activity of this cell, were observed in greater detail. In the posterior chamber of the larynx, circular openings were also observed among these cells. These were probably glandular apertures with secretory material. The bronchi showed ciliated and secretory cells that lacked glandular apertures (Pastor and GómezPascual 1995; Pastor et al., 1998). Transmission electron microscopy of the epithelia of the chambers and of the glands of the posterior chamber showed three cellular types: secretory, basal, and endocrine. Intraepithelial nerves, leukocytes and plasma cells were also identified in the epithelium. The secretory cells were found in the posterior chamber where they formed simple tubular glands. These cells were prismatic, while those of the lining of the chambers were usually cuboid and presented abundant short apical microvilli, abundant endoplasmic reticulum, dense mitochondria, multivesicular bodies and a cytoplasm with numerous secretory granules. These granules were more abundant

144  |  Airway Chemoreceptors in the Vertebrates in the glandular cells, where they sometimes filled the supranuclear cytoplasm. These secretory granules can be classified into five types: Type I, strongly electron dense, with a variable size and delimited by a membrane; Type II, with electron-lucent peripheral halo and a dense core; Type III, electron-lucent; Type IV, electron-lucent, with myelinized portions; Type V, with a myelinic structure similar to that of the lamellar bodies of the Type II pneumocytes. Type V granules were the most abundant in the glands, occupying almost all the cytoplasm of the cell. Frequently, these granules coalesce, forming tall granules. The glandular lumen may be filled with granules that are not straightened or with filaments of varying size. Tubular myelin was not found. The basal cells were scarcely differentiated, and they appeared in both the luminal and glandular epithelia (Pastor et al., 1998).

Neuroendocrine System of Amphibian Extrapulmonary Airways Immunocytochemical studies of neuroendocrine (NE) cells in the extrapulmonary airways of Amphibian are scarce. In the urodele Cynops pyrrhogaster, 5HT (serotonin), somatostatine-, calcitonin- and bombesin-like NE cells have been identified in the ciliomucous epithelium of the trachea (Adriaensen et al., 1994). In Hynobius nebulosus, 5HT-immunopositive, solitary NE cells have been labelled in the larynx, whereas 5HT, somatostatin, calcitonin, CGRP (calcitonin gene-related peptide), bombesin, and NSE (neuron-specific enolase) immunopositive cells have been identified in the trachea (Gomi et al., 1994). The density was high in the area with the laryngeal sphincter and decreased caudally in the laryngotracheal tube. The distribution of 5HTimmunoreactive cells appears to be related with the organization of smooth muscles in the fibromuscular layer and airway portion. The cells may be involved in regular shrinkage of the laryngotracheal cavity. No neuroepithelial bodies (NEBs) are observed (Kikuchi, 1995). Recently the distribution of 5HT-positive cells in serial sections of the respiratory tract of Cynops pyrrhogaster were stained using immunohistochemical methods for serotonin, a periodic acid-Schiff method and hematoxylin-eosin staining. 5HT-positive solitary NE cells were distributed widely throughout the ciliomucous epithelium of the laryngotrachea. The NEBs were rarely found in the caudal portion of the laryngotrachea. The density of NE cells was high in the middle portion of the laryngotrachea, particularly in the region surrounded by a striated laryngeal sphincter, and the densities were low cranially towards the aditus laryngis and caudally in the pulmonary sacs. The epithelium consisted of both ciliated and mucous epithelial cells. The morphology and distribution of 5HT-positive NE cells were similar to those described in mammals. The close localization of NE cells with ciliary and mucous cells and the distribution of NE cells in the laryngotrachea suggest that NE cells play important roles in the defense and repair of the ciliomucous epithelium,

Neuroendocrine System of the Amphibian Extrapulmonary Airways  |  145

such as increasing mucous secretion, ciliary beat frequency and the proliferation of epithelial cells (Naruse et al., 2005). Only an ultrastructural description of NE cells in Urodela extrapulmonary airways has been carried out. In the trachea of Cynops pyrrhogaster, single or small groups of NE cells could be found in the basal region of tracheal epithelium. The cells showed typical small dense-cored vesicles (150-250 nm) (Adriaensen et al., 1994). In urodela nervous plexus, histochemical observations show the coexistence of VIP (vasoactive intestinal peptide) and NOS (nitric oxide synthase), which seemingly form a non-adrenergic, non-cholinergic inhibitory neurotransmission (Adriaensen et al., 1994). There are numerous studies dealing with the ultrastructure of the lung epithelium and the immunohistochemistry of the lung NE cells and NEB in anurans (Goniakowska-Witalinska, 1995). These works have also paid attention to the innervation of the lungs and pharynx of anurans (Goniakowska-Witalinska, 1995; Kusakabe et al., 1995). In Anuran the first description of the presence of NEC and peptidergic innervation was in the larynx of Rana temporaria (Bodegas et al., 1995a). These authors showed the presence of 7b2, bombesin, SP (substance P), ET-1 (endothelin-1) and chromogranin in NE cells. More recently, in the anuran Rana esculenta and Discoglossus pictus, NE cells positive for Leu-enkephalin, VIP, CCK 8 (cholecystokinin octapeptide), and endothelin-big were described (Zaccone et al., 2004). In the anuran pharynx, nerve plexus containing NOS (Rana temporaria), CGRP, and SP (Rana catesbeiana) have also been described (Kusakabe et al., 1995; Bodegas et al., 1995b). In the larynx, only CGRP and SP-immunopositive nerve have been identified in the frog Rana temporaria (Bodegas et al., 1995a). In the buccal cavity of this species, the immunorreactivity was more varied, as serotonin, PGP 9.5 (Protein Gene Product 9.5), PHI (peptide histidine isoleucine amide), and CCK-immunopositive nerves were encountered (Bodegas et al., 1995a). Positive nerve fibers for Leu-enkephalin were found along the fibromuscular layer between arytenoid cartilage and laryngeal epithelium, and also in the laryngeal muscles (Zaccone et al., 2004). Met-enkephalinimmunostained fibers were located in the submucosal layer and terminated in the smooth muscle fibers of the sphincters. In the arytenoid region of Rana esculenta, a high density of CCK-8-positive and tyrosine hydroxylase-positive nerve fibers was found in laryngeal sphincters and the fibromuscular layer attached to the lateral cartilage. High numbers of VIP-positive nerve fibers were found in the core of connective tissue of the laryngeal mucosa. They formed perivascular plexuses in the walls of large vessels. NSEimmunopositive nerve fibers were also present in submucosal vasculature (Zaccone et al., 2004). In Rana perezi, immunocytochemistry also identified scattered epithelial NEC and nerve fibers in the laryngeal mucosa (Pastor et al., 1998). Immunosera directed to mammalian regulatory factors revealed epithelial cells apparently belonging to different cell types. Bombesin (Figure 1a) and 7B2 peptide (Figure 1b) immunoreactive cells

146  |  Airway Chemoreceptors in the Vertebrates

1a 1a

1b 1b

Figure 1: (a) Bombesin immunopositive cell lying among secretory cells (arrow). Their morphology seems to be of closed type. X 400; (b) 7B2 immunoreactive cell in the laryngeal epithelium (arrow). The cell extends an apical process toward the epithelial surface. X 400.

were observed. These epithelial endocrine cells were in close contact with the basal lamina. The morphology of these cells varied; some cells appeared elongated with the apex reaching the lumen (Figure 1b), and others were round or oval (Figure 1a). The nerve elements found in the larynx consisted mainly of numerous CGRPimmunoreactive subepithelial nerve fibers with prominent varicosities. Some nerve fibers immunoreactive for PHI were also detected. No immunoreactivity was found after the application of antisera against chromogranin, PGP 9.5, 5HT, SP, CCK, or ET-1 (Pastor et al., 1998). By transmission electron microscopy the endocrine cells were seen to be more electron-lucent in the larynx epithelium and showed small secretory granules (125 nm in diameter) in their cytoplasm with a clear peripheral halo and a dense core (Figure 2). Among the epithelial cells and in the glands, intraepithelial nerves could be found (Figure 3). Some nerves carried internal, clear vesicles, and other nerves had a clear peripheral halo and a dense core. The lamina propria of the larynx

Neuroendocrine System of the Amphibian Extrapulmonary Airways  |  147

2 Figure 2: Secretory cell of the surface epithelium showing varied types of granules. The basal NEC (asterisk) show a clear cytoplasm, in contrast with that of the secretory cells. Neuroendocrine granules (arrows). X 18000.

3 Figure 3: Intraepithelial, electron-lucent nerve endings contain myelinic-like inclusions and clear vesicles (arrow). X 10500.

148  |  Airway Chemoreceptors in the Vertebrates presented fibrocytes, numerous amyelinic and myelinic nerves, as well as melanocytes. The amyelinic nerves were usually close to the basal lamina, and occasionally they ran parallel to the former (Figure 4) (Pastor et al., 1998).

N

4 Figure 4: Amyelinic nerves (N) are seen in subepithelial region. X 12500.

Discussion Studies of the neuroendocrine system in extrapulmonary airways of amphibian are limited to a few species and only in one Urodela (Cynops pyrrogaster) and one Anuran (Rana perezi) have ultrastructural studies been made (Adriaensen et al., 1994; Pastor et al., 1998). In the four frogs that have been studied by light microscopy, the solitary NE cells are found only in this region, and 5HT-immunopositive cells are not observed Bodegas et al., 1995a; Pastor et al., 1998; Zaccone et al., 2004). In Urodela, too, solitary NE cells are found but 5HT-immunopositive cells are observed in the two species studied (Adriaensen et al., 1994; Gomi et al., 1994; Kikuchi, 1995; Naruse et al., 2005). The peptides that are found in the frogs are bombesin, SP, Leu-enkephalin, VIP, CCK8, endothelin-big, and ET-1. In Rana perezi, 7B2-immunopositive cells are also found, while in Rana temporaria they have been localized only in the lung and glottis. In Rana temporaria, positive cells of the neuroendocrine marker chromogranin are also observed (Bodegas et al., 1995a; Pastor et al., 1998; Zaccone et al., 2004). In Urodela, somatostatine-, calcitonin-, and bombesin-positive cells are found in the two species

Neuroendocrine System of the Amphibian Extrapulmonary Airways  |  149

studied and CGRP-positive cells only in Hynobius nebulosus (Adriaensen et al., 1994; Gomi et al., 1994). These results suggest heterogeneity in the immunocytochemical coding of respiratory laryngeal NE cells, both among closely related species in Anuran and between the Urodela and the Anuran order. In Urodela the species studied showed similar NE cells. Only the bombesin-containing cells have been identified in the airways in all the amphibians studied. The ultrastructure of the NE cell is very similar to of that of lung NEB in Anura and Urodela. The size of the granules in NE cells is similar in the species studied (Goniakowska-Witalinska, 1995). The results of nerves in Rana perezi (Pastor et al., 1998) are similar to the findings of CGRP in Rana temporaria, but are unique in identifying PHI-immunopositive nerves, which, in frogs, have been observed previously only in the buccal cavity and lung (Bodegas et al., 1995a). Intraepithelial nerves have been described in the airways of reptiles, birds, and mammals (Walsh and Mclelland, 1974; Laitinen, 1985; Pastor et al., 1988). To date, these nerves have been described in the anuran larynx of Rana perezi. They have a sensory function in mammals, though their possible implication in phonation cannot be ignored because they are abundant in the mammalian laryngeal mucosa and have been detected both immunohistochemically and ultrastructurally (Lima-Rodrigues et al., 2004). In conclusion, the neuroendocrine system of extrapulmonary amphibians is little known but there seems to be great variability between the species studied.

References Adriaensen, D., Scheuermann, D.W., Timermans, J.P., Gomi, T., Mayer, B., and De GroodtLasseel, M.H.A. 1994. Neuroepithelial endocrine and nervous system in the respiratory tract of Cynops pyrrhogaster with special reference to the distribution of nitric oxide synthase and serotonin. Microsc. Res. Tech., 29: 79-89. Bodegas M.E., Montuenga, L.M., and Sesma, P. 1995a. Neuroendocrine diffuse system of the respiratory tract of Rana temporaria: An immunocytochemical study. Gen. Comp. Endocrinol., 100: 145-161. Bodegas, M.E., Villaro, A.C., Montuenga, L.M., Moncada, S., Riveros-Moreno, V., and Sesma, P. 1995b. Neuronal nitric oxide synthase immunoreactivity in the respiratory tract of the frog, Rana temporaria. Histochem. J., 27: 812-818. Castells, M.T., Ballesta, J., Pastor, L.M., Madrid, J.F., and Marín, J.A. 1990. Histochemical characterization of glycoconjugates in the epithelium of the extrapulmonary airways of several vertebrates. Histochem. J., 22: 24-35. Gomi, T., Kikuchi, Y., Adriaensen, D., Timermans, J.P., De Groodt-Lasseel, M.H.A., Kimura, A., Naruse, H., Ishikawa Y., Kishi, K., and Scheuermann, D.W. 1994. Immunocytochemical survey of the neuroepithelial endocrine system in the respiratory tract of the Tokyo salamander, Hynobius nebulosus tokyoensis Tago. Histochemistry, 102: 425-431. Goniakowska-Witalinska, L. 1995. The histology and ultrastructure of the amphibian lung. In: Histology, Ultrastructure and Immunohistochemistry of the Respiratory Organs in NonMammalian Vertebrates. L.M. Pastor (Ed.). Secretariado Publicaciones Universidad Murcia, Murcia, pp. 71-112.

150  |  Airway Chemoreceptors in the Vertebrates Kikuchi, Y. 1995. Structure of the respiratory system of the Urodela Hynobius nebulosus tokyoensis Tago, with special reference to the distribution of serotonin-immunoreactive cells in its respiratory tract epithelium. Kaibogaku Zasshi, 70: 541-553. Kuehne, B., and Junqueira, L.C.U. 2000. Histology of the trachea and lung of Siphonops annulatus (Amphibia, Gymnophiona). Rev. Brasil. Biol., 60: 167-172. Kusakabe, T., Kawakami, T., and Takenaka, T. 1995. Calcitonin gene-related peptide and substance P in the pharynx and lung of the bullfrog, Rana catesbeiana. Cell. Tissue Res., 279: 115-121. Laitinen, A. 1985. Ultrastructural organization of intraepithelial nerves in the human airway tract. Thorax, 40: 488-492. Lima-Rodrigues, M., Nunes, R., and Almeida, A. 2004. Intraepithelial nerve fibers project into the lumen of the larynx. Laryngoscope, 114: 1074-1077. Matsumura, H., and Setoguti, T. 1986. Correlated light microscopic and transmission and scanning electron microscopic studies of the laryngotracheal epithelium in Salamanders, Hynobius nebulosus. J. Electron Microscop., 35: 66-76. Naruse, H., Gomi, T., Kimura, A., Adriaensen, D., and Timmermans, J.P. 2005. Structure of the respiratory tract of the red-bellied newt Cynops pyrrhogaster with reference to serotoninpositive neuroepithelial endocrine cells. Anat. Sci. Int., 80: 97-104. Pastor, L.M., and Gómez-Pascual, A. 1995. The extrapulmonary airways in amphibians. In: Histology, Ultrastructure and Immunohistochemistry of the Respiratory Organs in NonMammalian Vertebrates. L.M. Pastor (Ed.). Secretariado Publicaciones Universidad Murcia, Murcia, pp. 53-69. Pastor, L.M., Ballesta, J., Castells, M.T., Pérez-Tomás, R., Madrid, J.F., and Marín, J. A. 1988. A light and electron microscopic study of the epithelium of the extrapulmonary airways of Mauremis caspica and Lacerta lepida (Reptilia). J. Submicrosc. Cytol. Pathol., 20: 25- 36 Pastor, L.M., Bodegas, M.E., Gallego-Huidobro, J., and Pallares J. 1998. The ultrastructure and immunohistochemistry of the laryngeal mucosa of Rana perezi (Amphibia:Anura). J. Submicrosc. Cytol. Pathol., 30: 55-63. Walsh, C., and Mclelland, J. 1974. Intraepithelial axons in the avian trachea. Z. Zellforsch. Mikrosk. Anat., 147: 209-217. Zaccone, V., Mauceri, A., Lo Cascio, P., Minniti, F., Parrino, V., and Fasulo, S. 2004. Immunohistochemical study of the innervation of pulmonary vessels and smooth muscles in the respiratory tract of two frog species. Acta. Histochem., 106: 179-193.

7 Chemoreceptive Control of Ventilation in Amphibians and Air-Breathing Fishes Warren Burggren* and Tien-Chien Pan

Abstract Ventilation is a critically important process in providing O2 to the respiratory surfaces and removing CO2 from them. When either environmental gas composition or tissue demands change, then adjusting ventilation through rate and amplitude modifications is the most direct response to ensure respiratory gas exchange. In amphibious vertebrates breathing both air and water and using a suite of respirator structures (which can include skin, external gills, internal gills, lungs, gas bladders and intestines), the process of ventilatory adjustment can be complex indeed. The present review examines the morphology, physiology and evolutionary biology of ventilatory responses to altered O2 and CO2 levels in amphibians and air-breathing fishes. Additionally, the vital role in modulating ventilatory responses of both centrally and peripherally located chemoreceptors and mechanoreceptors, as investigated by in vitro and in vivo methods, is examined. Finally, this analysis concludes by posing an extensive list of areas in lower vertebrate respiratory control deserving future investigation.

Introduction Many vertebrates exploit some combination of aquatic and aerial gas exchange to provide O2 uptake and CO2 elimination. In fishes, exploitation of aerial gas exchange has evolved independently many times, involving a variety of air breathing organs (for general reviews see Johansen, 1970; Randall et al., 1981; Little, 1983; Graham, 1997; Maina, 2002). Indeed, air-breathing occurs in at least 49 known families of fish (Graham, 1997). In the Amphibia, a large proportion of the more than 6000 Department of Biological Sciences, University of North Texas, Denton, TX 76205, USA. *Auhtor for Correspondence.

152  |  Airway Chemoreceptors in the Vertebrates amphibian species dwell in water (e.g. the anuran amphibians and especially frogs), using their lungs for aerial gas exchange and their skin for aquatic gas exchange. From a developmental perspective, almost all air-breathing fishes and amphibians exhibit early embryonic/larval stages that are strictly aquatic and use solely water for gas exchange, but subsequently undergo a fascinating and complex developmental transition that includes the capacity for air breathing. The term “bimodal breather” has been used extensively in describing various amphibious vertebrates, but some confusion as to the meaning of this term still persists. “Mode” is typically defined as “..a way of doing something...", hence “biomodal” refers to two ways of doing something. It follows, then, that “bimodal gas exchange” refers to the two ways in which gas exchange is achieved, not the two respiratory media (water, air) that are used (Figure 1). Thus, here we use the term “biomodal” to mean that two different respiratory structures are used.1 This may at first seem like a trivial semantic diversion. Consider, however, that many amphibious vertebrates, at some stage in their development, are actually trimodal breathers that us various combinations of skin plus gills plus lungs to breath both water (skin and/or gills) and air (skin and/or lungs). In many respects, trimodal breathing represents a much more complex respiratory Respiratory Mode

Respiratory Medium

SKIN (non-specialized body surfaces)

GILLS

WATER

(external and internal)

AIR

LUNG-LIKE ORGANS (non- pulmonary origin)

LUNGS (derived from ventral gut)

Major interaction Minor interaction

Figure 1: Interrelationships between modes of breathing with various respiratory organs and the two respiratory media—water and air. Many amphibious vertebrates use combinations of respiratory modes during their developmental life cycle, as well as concurrently as adults. The skin is the only respiratory organ that can serve equally well when either water- or air-exposed. Note that in some species of air- breathing fishes the gills do not entirely collapse when air exposed, and can still participate in some degree of gas exchange. Similarly, some non-pulmonary air-breathing organs in fishes can continue to exchange gas at slow rates even when water-filled (e.g. the labyrinth organ of labyrinthodontid fishes). 1

It could be argued that “diffusion” and “convection” are indeed also “modes” of gas exchange, but here we shall confine the use of mode to structure rather than process.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  153

situation compared with the far simpler respiratory circumstances of animals that use almost exclusively either gills or lungs. Indeed, the non-linear developmental transition of many amphibians from breathing only with skin (early larvae) → skin + gill breathing → skin + gill + lung breathing → skin + lung breathing (adults) has been used a physiological model for complexity change and its analysis (Burggren and Monticino, 2005; Burggren, 2006). Tremendous variety is to be found in the combinations of various modes of breathing in amphibious vertebrates. A review of this material is beyond the scope of this chapter. However, evident from Figure 1 is that in any vertebrate using multiple respiratory modes and multiple respiratory media, the control of ventilation process is potentially quite complicated. In this chapter we will discuss the chemoreceptive control of air breathing in amphibious vertebrates using multiple modes of gas exchange. To begin this process, let us first briefly consider from a ventilatory control point of view both the physico-chemical characteristics of the respiratory media as well as the nature respiratory modes (structures).

The Respiratory Media Water and air differ enormously in attributes important to the process of gas exchange: density, viscosity and oxygen capacitance. These three factors interact to make breathing water a very different process from breathing air. Because water is heavier, more viscous and has a much lower O2 capacity than air, animals actively breathing water—that is, using muscle power to generate a flow of water over their respiratory surfaces—will have to pump a 30-40 times greater volume of water than an air breather would have to pump of air. Thus, the cost of breathing water in aquatic fishes, while apparently quite variable, is certainly much more expensive than in vertebrates that breathe air (see Randall, 1970; Steffenson and Lomholt, 1983; Maina, 2002). Consequently, there is an additional energetic burden on aquatic vertebrates to ensure that ventilation is carefully monitored and regulated. Also, because aquatic vertebrates like fishes necessarily ventilate their gills with a high volume of water—and because CO2 has a much greater capacitance coefficient for CO2 than O2, metabolically produced CO2 is quickly washed out of the blood. Typically, fishes have a venous blood PCO2 of less than 1 kPa, compared with much higher values typically in the range of 5-8 kPa for terrestrial, air-breathing animals. For equivalent molar quantities of CO2 entering a given volume of water or air, the very high solubility of CO2 in water means that the increase in measurable PCO2 in air will be much higher than the PCO2 increase in water. In other words, elimination of a large amount of CO2 into water will produce only very small increases in PCO2 in the exhalant water stream. The addition of CO2 into water will result in a variable degree of fall in the pH of exhalent water, because the actual change in water pH for a given molar quantity of CO2 eliminated from the blood is dependent upon the

154  |  Airway Chemoreceptors in the Vertebrates exhalent water’s buffer capacity. While some aquatic species are tolerant of only a very narrow suite of water characteristics, others can range more freely and experience a significant range of water quality, including buffer capacity. Thus, monitoring of either PCO2 or pH in exhalent water for the purposes of regulation of aquatic ventilation will be unreliable for an aquatic animal. As a consequence of these physico-chemical differences in air and water, aquatic vertebrates have evolved ventilatory control systems that predominantly affect the uptake of O2 (Smatresk, 1990; Taylor et al., 1999; Florindo et al., 2004; Vulesevic and Perry, 2006). Ventilatory changes for CO2 elimination are rarely necessary, though fishes do have some ventilatory responses to aquatic hypercarbia (e.g. Perry and McKendry, 2001). Body fluid pH in strictly water-breathing vertebrates is maintained in large part by the controlled elimination of H+ and HCO3- ions, since the high solubility of CO2 in water makes untenable retention of CO2 in the blood to be “blown off ” in a regulated fashion. In contrast, in terrestrial air-breathing animals air is relatively inexpensive to metabolically pump through lungs or lung-like organs, and O2 is in abundance. Minute-to-minute ventilatory control thus tends to center around elimination of CO2 to maintain of appropriate body fluid pH levels, though internal or, more rarely, environmental hypoxia can nonetheless profoundly stimulate ventilation. Gas exchange and ventilatory control complexity reaches a zenith in amphibious, bimodal breathers that have to face concurrently both the advantages and disadvantages of air and water as a respiratory medium. We will return to this topic after considering the respiratory organs, themselves.

Modes of Gas Exchange: The Respiratory Structures of Vertebrates Collectively, amphibians and air-breathing fishes show examples using all four major categories of respiratory structure of vertebrates: skin, gills, non-pulmonary airbreathing organs, and lungs. Many species, either as larvae or adults, show combinations of exchangers as either bimodal or trimodal breathers.

Skin

All animals have some capacity for gas exchange via their generalized body surface (skin). Even in heavily scaled fishes or furry mammals there is measurable O2 uptake and CO2 elimination via the skin. In some lightly scaled or scale-less aquatic fishes and in primarily aquatic amphibians, cutaneous gas exchange can account for up to 40% of O2 uptake and, in amphibians, even larger proportions of CO2 elimination (Feder and Burggren, 1985). The rest of the gas exchange in these bimodal or tri-modal breathers occurs by gills, ABOs or lungs. In terrestrial vertebrates with thin, relatively moist skin (e.g. toads), there is a reduced role of the skin in O2 uptake, which occurs primarily via pulmonary routes, but the skin retains importance in CO2 elimination.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  155

There is an historical, dominant view in the literature that cutaneous gas exchange “just happens”—that is, it cannot be regulated per se and instead, gas exchange across the generalized body surface merely reflects the partial pressure gradients for gases between blood and surrounding air or water. While both the relative role of the skin in gas exchange declines as metabolic rate increases and the transcutaneous partial pressure gradients are certainly of pre-eminent importance (e.g. Pinder et al., 1991), recent experiments on terrestrial toads have shown that, overall, cutaneous blood flow, as well as regionalized capillary recruitment and derecruitment, can actively regulate cutaneous gas exchange (Burggren and Vitalis, 2004) to an extent not previously appreciated. Skin of aquatic vertebrates can also be actively ventilated via behavioral mechanisms. One of the biggest obstacles to effective cutaneous gas exchange is the build-up of a boundary layer of stagnant water adjacent to the skin, essentially increasing the diffusion distance for respiratory gases. By positioning the body in a current of water or by actively swimming or creating other body movements, fishes and aquatic amphibians can disrupt these boundary layers and increase the efficiency of cutaneous gas exchange. Whether there is localized sensory monitoring of changes in tissue O2 and CO2 that reflexly alter either cutaneous perfusion or activities that “ventilate” the skin is currently unknown.

Gills

The structure and respiratory function of gills has been extensively documented in fishes (for recent reviews and entry into the extensive literature see Maina, 2002; Olson, 2002; Wilson and Laurent, 2002; Evans et al., 2005). Briefly, in all but the most primitive fishes, the branchial arches (typically 4 to 6 pairs depending on genus) are enclosed in paired internal branchial chambers. Buccal pumping drives large volumes of water across the gills in a direction counter to that of the blood flow within the individual gill filaments. Gill ventilation is carefully tuned to oxygen demands via sensory feedback involving receptors located within the branchial chambers, on the gill surfaces, or internally in excurrent (arterialized) blood (see Chapter 1, this volume). The gills of larval and neotenous amphibians have been much less examined compared to fish gills (Malvin, 1989; Pinder and Burggren, 1986; Maina, 2002). External gills contribute to gas exchange in early developmental stages in both fishes and amphibians, and persisting in adults in a few neotonous amphibian species. Since these gills are not enclosed in a ventilated, internal chamber, they are faced with the issues of boundary layer build-up as skin. However, once again behavioral activities enhancing gas exchange involve orientation of the body in currents or, in the case of some amphibians with external gills, doing “pushups” to wave the gills and break up boundary layers in hypoxic water. Generally, gills are solely instruments of aquatic gas exchange, collapsing to a fraction of their original surface area when removed from the buoying effect of water and exposed to air. However, the gills of a few amphibious

156  |  Airway Chemoreceptors in the Vertebrates fishes that venture on to land (e.g. Periopthalmus, Boleophthalmus) have mechanical spacers that hold apart the individual filaments and allow some continuing aerial gas exchange (see Graham, 1997).

Non-pulmonary Air-Breathing Organs

Non-pulmonary air-breathing organs (ABOs) are found in the air-breathing fishes excluding the lungfishes, which have true lungs (see below). ABOs are found in many shapes and forms in air-breathing fishes (Randall et al., 1981; Graham, 1997). They have evolved as both de novo structures (e.g. labyrinth organs of the gourami, Trichogaster trichopterus and the Siamese fighting fish, Betta splendens), and through partial modification of organs used for other purposes (e.g. the hindgut of the African weather loach, Misgurnus anguillacaudatus or the swim bladder of the arapaima, Airpaima gigas). Lung-like ABOs, which can be quite elaborate with an alveolar-like structure, always retain a residual gas volume are never exposed to water. However, labyrinth organs in epibranchial chambers alternate between being water and air filled, and the gut breathers must accommodate both air and regular gastrointestinal contents. ABOs tend to be regularly ventilated, with the ventilation rate increasing with higher metabolic demand, increasing temperature (which heightens the metabolic rate), and with decreasing environmental O2 levels. The reflex mechanisms by which air ventilation in ABOs of air-breathing fishes are regulated are not nearly as well categorized as for amphibians or aquatic fishes, as will be discussed below.

Lungs

True lungs are found only in the lungfishes (Lepidosiren, Neoceratodus, Protopterus) and in tetrapod vertebrates including, of course, the amphibians. These structures are ventrally derived outgrowths of the esophagus, a definition that differentiates them from swim bladders that might otherwise occupy the same region of the body cavity and have a similar structure to primitive lungs. Lungs are also perfused by arteries derived from the branchial arch VI, comprising a pulmocutaneous artery in amphibians and a pulmonary artery in lungfishes. Having made this embryological/anatomical distinction, the mechanisms of ventilation of the lungs of lower vertebrates are quite similar to those of ABOs derived from swim bladders. Rather than a diaphragmatic mechanism as in mammals, for example, the lungs are ventilated by positive pressure produced by buccal gas compression in both lungfishes (McMahon, 1969; DeLaney and Fishman, 1977) and amphibians (Shoemaker et al., 1992; Jorgensen, 2000; Vasilakos et al., 2006), though some dispute still exists over the precise mechanics and patterns of gas flow (see Fernandes et al., 2005). The interior structure of the lungs of amphibians and lungfishes is quite variable, but generally they are more secular than the more highly alveolarized lungs of reptiles and mammals.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  157

Having reviewed the respiratory structures of bimodally breathing air-breathing fishes and amphibians, now let us turn to how the specific role of chemoreceptors in the regulation of their ventilation.

Chemoreceptors and Ventilatory Control in Amphibians Sensory Systems for Ventilation Regulation The role of the respiratory system of amphibians is to maintain appropriate respiratory gas composition within the circulating body fluids. To achieve this task, information on respiratory gases, acid-base status, and ventilatory performance transduced by various O2- and CO2-sensitive chemoreceptors and mechanoreceptors must reach the brain, especially the medulla. It is in this central nervous system structure where the respiratory rhythm is both formed and modulated, generating appropriate respiratory behaviors to meet the tissues’ gas exchange requirements. Sensing both the external and internal environment is the key to effective regulation of ventilation. Thus amphibians, not surprisingly, show a variety of sensory receptors that monitor both peripheral and central changes in respiratory gases and pH. The general subject of chemoreceptors in amphibians has previously been reviewed in depth (Smatresk, 1990; West and Van Vliet, 1992; Kusakabe, 2002; Reid, 2006; Gargaglioni and Milsom, 2007), and our intent here is to provide a general overview.

Respiratory Tract Chemoreceptors

Pulmonary Stretch Receptors. Pulmonary stretch receptors deliver dynamic information on the extent of lung deflation and inflation to the brain stem via the vagus nerve. Generally, amphibian pulmonary stretch receptors are stimulated by a dynamic increase in lung volume or pulmonary wall tension, which in turn increases expiration and inhibits inspiration (West and Van Vliet, 1992; Wang et al., 1999; Reid et al., 2000; Sanders and Milsom, 2001; Reid, 2006; Gargaglioni and Milsom, 2007). These receptors are divided into three groups. The first group responds to the degree of lung inflation, and can be viewed as a pulmonary volume receptor. The second group of phasic pulmonary stretch receptors is stimulated by the rate of inflation, increasing their firing frequency when the rate of stretch increases. Individual receptors in the last, distinct, group actually respond to both rate and extent of stretch (Milsom and Jones, 1977; Kinkead and Milsom, 1996; Reid, 2006). Reid and West (2004) investigated the role of phasic pulmonary stretch receptor (rate-sensitive) in ventilation in the cane toad, Bufo marinus, using tidally ventilation instead of the more commonly used unidirectional ventilation method. Efferent neural recording of trigeminal nerve activity showed that stimulation of the phasic pulmonary stretch receptor increased overall breathing frequency.

158  |  Airway Chemoreceptors in the Vertebrates While pulmonary stretch receptors in amphibians are primarily responsive to their distortion when pulmonary volume changes, these receptors also respond to increasing intrapulmonary CO2 levels by decreasing their firing rate (Milsom and Jones, 1977; Reid et al., 2000; Reid and West, 2004; Reid, 2006; Gargaglioni and Milsom, 2007). The interaction between these two kinds of stimuli is responsible for the overall respiratory input from the lungs to the brainstem in the bullfrog (Sanders and Milsom, 2001; Reid, 2006). Indeed, the CO2-sensitive stretch receptor in amphibians may represent the archetype for specialized CO2 receptors found in higher vertebrates (Milsom and Jones, 1977; Milsom, present volume).   Olfactory CO2 Chemoreceptors. Olfactory receptors of amphibians are also CO2 sensitive, and they respond to elevated CO2 levels by sending inhibitory afferent signals that ultimately inhibit breathing, which is likely to be a defensive mechanism (Getchell and Shepherd, 1978; Sakakibara, 1978; Coates and Ballam, 1990). The information is conveyed via the olfactory nerve, since transection of that nerve eliminates the CO2 response (Coates, 2001). The population of CO2-sensitive olfactory receptors is relatively rare. In the salamander, only 1 to 2% of olfactory receptors responded to 5% CO2 while the remainder were stimulated by odorants (Getchell and Shepherd, 1978). The response of these receptors showed dose-dependent increases for CO2 levels from 0.5 to 10% in the bullfrog (Coates and Ballam, 1990). Carbonic anhydrase (CA), a family of enzyme catalyzing the hydration of CO2, was found to participate in the CO2 sensing mechanism in amphibian olfactory epithelium. Coates et al. (1998) reported that CA immunoactivity was localized mainly in the dorsal and ventral regions, where 23 out of 1222 sites examined responding to 5% CO2. Inhibition of the enzyme (CA) by acetazolamide attenuated the response by 65%. These findings support the evidence of the rare presence of CO2-sensitive olfactory receptors found in salamanders and indicate the role of CA in CO2 detection in olfactory epithelium (Coates et al. 1998).   Other Receptors in the Respiratory Tract. In addition to pulmonary stretch receptors and olfactory CO2 receptors, narial mechanoreceptors can be identified that are sensitive to water. They prevent water from entering the respiratory tract by inhibiting ventilation upon submergence. The feedback from this type of receptor also contributes to the overall output of breathing frequency (West and Van Vliet, 1992). Taste cells are also sensitive to water; however, no evidence has shown its relationship to the ventilation of animals. A population of water-sensitive receptors is also located in the glottis and pharynx, and inhibits lung ventilation during swallowing and water entry (West and Van Vliet, 1992).

Arterial Chemoreceptors

Hypoxic stimulation of lung ventilation in adult anuran amphibians is mediated primarily by peripheral O2-sensitive receptors that monitor arterial blood. At least two locations have been identified for these chemoreceptors. The first is the carotid labyrinth, which is a highly vascular plexus located in the bifurcation of the common

Ventilation Control in Amphibians and Air-Breathing Fishes  |  159

carotid artery forming the internal and external carotid arteries (Kusakabe, 2002). Though similar in many respects to the mammalian carotid body, these structures are not homologous. Both chemo- and baroreceptor functions have been confirmed for the carotid labyrinth through electrophysiological recording (Van Vliet and West, 1992) and nerve ablation ( Jones and Chu, 1988). Arterial O2-sensitive chemoreceptors are also located in the aortic arch. Injection of sodium cyanide and perfusion with hypoxic or hypoxic-hypercapnic solutions result in discharge of the receptors within the aortic arch, indicating the presence of O2 chemoreceptors (Van Vliet and West, 1992). However, the aortic chemoreceptor has received lesser attention as compared to the carotid labyrinth in amphibian. Receptors within the pulmonary vasculature also participate in chemoreception for ventilation. Injection of cyanide into the pulmonary arterial circulation causes fictive hyperventilation, suggesting the presence of pulmonary arterial O2-sensitive receptors (Wang et al., 2004). Denervation of the recurrent laryngeal nerves innervating the baroreceptor within the pulmocutaneous arteries caused a threefold increase in pulmonary blood flow and increased net transcapillary fluid flux, suggesting that pulmocutaneous baroreceptors protect the anuran lung by regulating pulmonary blood flow (Smits et al., 1986). Neuroepithelial bodies are also plentiful in amphibian lungs (Goniakowska-Witalińska, 1997). These structures, which are located mainly in the ciliated epithelium of the apical part of the septa, may also play a role in chemoreception involving intrapulmonary gas composition.

Afferent and Efferent Innervation

Receptors providing environmental cues for ventilatory regulation are distributed throughout the lungs, as well as some locations in the central arterial circulation. Afferent information bound for the medulla (the site of respiratory rhythm generation) occurs via a variety of nerves carrying sensory fibers from these receptors including cranial nerves I, V, IX and X (see Table 1). Modulation of chemoreceptor performance occurs via efferent (“motor”) innervation. The carotid labyrinth of anurans is innervated by neurons containing regulatory neuropeptides thought to modulate chemoreceptor sensitivity and vascular tone (Kusakabe et al., 1995; Kusakabe, 2002). The neuroepithelial bodies in the lungs of amphibians also receive efferent innervation, but the physiological significance of this neural innervation has not been identified (Goniakowska-Witalińska, 1997).

Central Nervous System and Ventilatory Chemoreception

1 Hypoxia. To the best of our knowledge there is no evidence for the existence of an O2 receptor in the central nervous system that directly monitors changes in brain blood PO2 and induces physiological responses, such as hyperventilation. However, some cell groups are thought to participate in the pathways responding to hypoxia.

160  |  Airway Chemoreceptors in the Vertebrates Table 1:

Afferent innervation of structures bearing chemo- and mechanoreceptors regulating gas exchange in amphibians.

Anatomical Structure(s)

Cranial Nerve Carrying Afferent Fibers

Reference

• Nares • Olfactory epithelium

Cranial nerve I Cranial nerve V

Sakakibara, 1978 West and Van Vliet, 1992 Coates, 2001

• Tongue

Cranial nerve IX

Inoue, 1978 West and Van Vliet, 1992

• Pharynx • Glottis • Lungs • Carotid labyrinth

Cranial nerve IX Cranial verve X

West and Burggren, 1983 Van Vliet and West , 1986 West and Van Vliet, 1992 Kusakabe, 2002

The nucleus isthmi (NI), a mesencephalic structure located between midbrain and the cerebellum, inhibits the hypoxic ventilatory response in toads, Bufo paracnemis, by inhibiting the increase in tidal volume that would normally accompany hypoxia. It shows the regulatory role of structures in the CNS in the hypoxic hyperventilatory response. Glutamate and nitric oxide (NO) may be two of the possible candidates that mediate this inhibitory effect (Gargaglioni and Branco, 2004).   CO2. Despite multiple locations for CO2 chemorecptors in amphibians, the CO2sensitive receptors present in the ventral medulla of the central nervous system, which arise in late larval development, are considered the dominant sensory site for CO2 chemoreception in amphibians as well as other tetrapods (Smatresk and Smits, 1991; West and Van Vliet, 1992; Torgerson et al., 1997; Taylor et al., 2003). Stimulation of these receptors by high PCO2 and low pH caused both an increase in ventilation frequency and tidal volume (West and Van Vliet, 1992; Wang et al., 1999). Central chemosensitivity to CO2 and pH is enhanced by a 9-day-exposure to hypercapnia (3.5% CO2) as investigated by both in vivo monitoring of breathing frequency and in vitro neural recording from brainstem-spinal cord preparations in an adult anuran, Bufo marinus (Gheshmy et al., 2006) In mammals, several sites within the central nervous system exhibit CO2 chemoreception, including the nucleus tractus solitarius, the locus coeruleus, the midline medullary raphe, the ventral respiratory group, the fastigal nucleus, and the retrotrapezoid nucleus (Feldman et al., 2003). Among these many structures, only the locus coeruleus (LC) has been described in amphibians (Noronha-de-Souza, et al., 2006). In the adult toad, Bufo schneideri, lesions in the LC diminish the hyperventilatory response to hypercarbia, and injection of acidic solution into the LC induces hyperventilation (Noronha-de-Souza et al., 2006). Increased immunoreactivity of c-fos after exposure to 5% CO2 indicates that the nucleus was activated by hypercarbia. In addition to the inhibitory effect on hypoxic hyperventilation, the nucleus isthmi (NI) has a similar inhibitory effect to hypercapnia-induced hyperventilation. The

Ventilation Control in Amphibians and Air-Breathing Fishes  |  161

NI differentiates during metamorphosis when the transition of branchial ventilation to pulmonary ventilation occurs. Chemical lesion of the NI enhanced hypercarbic hyperventilation, demonstrating the inhibitory role of the NI when respiratory stimulus is high (Gargaglioni and Branco, 2004). As mentioned earlier, the NI does not function as a direct sensor for CO2 or pH in the CNS, because lesions in the NI do not affect resting breathing frequency (Gargaglioni and Branco, 2004).

Brain Respiratory Centers The two different ventilatory acts of frogs—the more frequent and rhythmic buccal ventilation and the more irregular and stronger lung ventilation—appear to be generated by two distinct coupled central pattern generators (CPGs) (Wilson et al., 2002; Vasilakos et al., 2006). Pulmonary respiratory rhythms in amphibians originate from central pattern generators located in the medulla (see McLean et al., 1995; Perry et al., 1995; Milsom et al., 1999). Unlike CPGs in mammals and birds, the CPGs in adult amphibians do not provide for a constant pulmonary ventilation rhythm, but rather generate motor output for frequent but irregularly spaced breaths. Input from sites in the dorsal brainstem caudal to the optic chiasma clusters breaths into small groupings. Segmental generators in the medulla produce the primary rhythm, and are subsequently entrained to create the typical intermittent pattern of pulmonary ventilation (Reid et al., 2000). Nitric oxide provides excitatory input to the bullfrog’s CPGs in the brainstem (Hedrick et al., 1998; Hedrick and Morales, 1999; Harris et al., 2002). Buccal respiratory rhythms in adult amphibians are modulated by activity in both caudal and rostral levels of the brainstem (Wilson et al., 2002). The buccal oscillator is coupled to the pulmonary oscillator via chloride-mediated, opiod-sensitive mechanisms (Vasilakos et al., 2006). Modulation of the normal ventilation pattern in adult amphibians is based on sensory input that occurs via several brain structures. Elevation of the respiratory drive results in tegmental and medullary inputs that modify the burst pattern of motor output to respiratory muscles (Reid et al., 2000). The nucleus isthmi, a mesencephalic structure situated between the cerebellum and the roof of the midbrain, is thought to modify the hypoxic and hypercapnic drives (Kinkead et al., 1997; Gargaglioni and Branco, 2004), a process involving both glutamate, nitric oxide , substance P, etc. (Perry et al., 1995; Gargaglioni and Branco, 2004). The developmental changes in the location, anatomy, neurochemistry and function of amphibian central pattern generators have been investigated because of both the usefulness of the model in understanding lower vertebrate ventilation patternsits used as a model for considering the evolution of CPGs. This interesting subject is beyond the scope of this review, but the reader is referred to recent reviews by (Gdovin et al. 1999; Straus, 2000; Hedrick et al., 2005)

162  |  Airway Chemoreceptors in the Vertebrates

Regulation of Ventilation by Chemoreceptors Ventilatory responses in amphibians represent a complex integration of input from pulmonary stretch receptors, olfactory chemoreceptors and intrapulmonary and arterial chemoreceptors (Figure 2). Not surprisingly, then, there is no “standard” hypoxic or hyepercapnic response. Thus, observation of the nature of a hypoxic or hypercanpic drive based on data from the popular brainstem model is perhaps more useful in teasing apart the “internal wiring” of the brainstem than it may be in describing the actual responses of the whole animals. The ventilatory response of fictive lung breathing in brainstem preparations was briefly discussed above, and it is not our intention to review these CNS responses (for reviews see Milsom et al., 1999; Reid, 2006). Here we will focus on in vivo, whole animal responses.

Ventilatory Responses to Lung Inflation and Hypoxia

Intact, conscious amphibians typically exhibit a strong hypoxic drive for both buccal and pulmonary ventilation, a finding long recognized for anurans (e.g. Babak, 1911; Smyth, 1939; reviewed by West and van Vliet, 1992; see also Branco and Glass, 1995; Hou and Huang, 1999). Not only does inspiration of hypoxic gas stimulate ventilation, but hyperoxia actually inhibits ventilation. In the toad Bufo marinus, hyperoxia inhibits ventilation even though hypercapnia and respiratory acidosis ensues (Toews and Kirby, 1985; West et al, 1987), indicating that the hypoxic drive can dominate in controlling ventilation in this toad. A typical finding in studies showing hypoxic stimulation of ventilation is that not only is pulmonary minute ventilation increased, but the pattern of ventilation changes as a consequence of inspiration of hypoxic gas (Pinder and Burggren, 1986; West and Van Vliet, 1992; Kinkead and Milsom, 1994; Gardner et al., 2000; Gargaglioni and Branco, 2000; Gargaglioni et al., 2002). Does inspiration of hypoxic gas stimulate lung ventilation through reduction of arterial PO2 or reduction of arterial blood oxygen concentration? Ventilatory responses to hypoxia persist independently from changes in blood O2 carrying capacity in Bufo paracnemis (Wang et al., 1994; Andersen et al., 2003), indicating that blood-facing receptors are monitoring PO2. Hypoxic ventilatory responses appear to have a seasonal component in some anurans. In Bufo paracnemis, toads that respond vigorosly to hypoxia at 25oC during summer show no hypoxic response at 25oC in winter, despite the fact that blood gases showed no seasonal effect (Bicego-Nahas et al., 2001), suggesting that seasonal effects are affecting some aspect of the chemoreceptors or the integration of the information they provide to the CNS. Rana catesbeiana shows enhancement of temperaturedependent hypoxic ventilatory responses in winter, and reduction in summer, with intermediate responses in spring and autumn (Rocha and Branco, 1998) The neotenous axolotl, Ambystoma mexicanum, provides an interesting perspective in an “adult” amphibian (or at least one that is no longer developing) that ventilates

Ventilation Control in Amphibians and Air-Breathing Fishes  |  163

Figure 2: Electroneurograms representing fictive breathing recorded from the laryngeal branch of the vagus nerve (XI) and the mandibular branch of the trigeminal nerve (Vm) in unidirectionally ventilated, decerebrate bullfrogs, Rana catesbeiana. In A, frogs were ventilated with air, while in B animals are ventilated with 3% CO2 in air. “Low pressure” corresponds to 1 cmH2O, while “high pressure” corresponds to 5 cmH2O. Note that stimulation of stretch receptors by increased ventilation pressure in the lungs suppressed Vm burst amplitude such that fictive lung ventilations (taller spikes in Vm recordings) became indistinguishable from fictive buccal oxcillations (shorter spickes). CO2 stimulated absolute fictive lung ventilation, primarily by reducing apnea length rather than breathing depth. These experiments show the complex nature of the interactions between mecho- and chemoreceptors in modulating the central rhythm generators in anuran amphibians (from Sanders and Milsom, 2001).

with both gills and lungs. Hypoxia stimulated ventilation rate of both the gills and the lungs, as did infusion of NaCN into the ventilatory stream or the arterial bloodstream (McKenzie and Taylor, 1966). Interestingly, norepinephrine stimulated gill ventilation but not lung ventilation rates. The axolotl thus shows similar ventilatory responses to larval amphibians.

164  |  Airway Chemoreceptors in the Vertebrates The caecilian Typhlonectes natans shows an interesting suite of ventilatory responses to hypoxia, differing somewhat from other amphibians, in that aquatic hypoxia affects neither breathing frequency nor mechanics (Gardner et al., 2000). Yet, aerial hypoxia increases ventilation frequency as in other amphibians, The salamander Desmognathus fuscus responds to hypoxic exposure with an increase in buccal pumping, even though as adults they lack lungs (Sheafor et al., 2000), similarly to how lunged salamanders would respond, though the role of this buccal hyperventilation in the observed maintenance of oxygen uptake in milder hypoxia is unknown.

Ventilatory Responses to Hypercapnia

Interpretation of hypercapnic responses in amphibians is confounded by the considerable capacity for cutaneous CO2 elimination. With the potential for CO2 loss across the skin, arterial PCO2 values will be lower for a given inspired PCO2 than in reptiles, birds or mammals, for example. Short of concurrently measuring blood PCO2 and acid-base parameters along with ventilation, quantitative determination of the sensitivity of the pulmonary hypercapnic response—and certainly any comparison with similar exposure in reptiles, for example—is problematic. Anuran amphibians typically respond to elevations in aerial CO2 with increased lung ventilation (see reviews by West and Van Vliet, 1992; Reid, 2006) resembling terrestrial tetrapod vertebrates (see other chapters, this volume). Most urodeles, however, show little or no ventilatory responsive to hypercapnia, and lung ventilation frequency is not correlated with arterial PCO2 (see West and Van Vliet, 1992). The predominantly skin-breathing salamander Cryptobranchus alleganiensis responds to aquatic hypercapnia with an increase in pulmonary ventilation (Boutilier and Toews, 1981), more like anurans. In caecilians, where independent and combined exposure to aerial and aquatic hypoxia has been determined, aquatic rather than aerial hypercapnia is the more potent ventilatory stimulant (Gardner et al., 2000). Similar to the hypoxic ventilatory response in anuran amphibians, there is seasonal variation in the extent of the hyperventilation stimulated by hypercapnia, with winter bullfrogs (Rana catesbeiana) showing a temperature-independent muting of the ventilatory response to 3-5% inspired CO2 (Bicego-Nahas and Branco, 1999).

Chemoreceptors and Intermittent Ventilation Amphibians are typically intermittent lung breathers (see Boutilier, 1984; Smatresk, 1990; Feder and Burggren, 1992; Taylor et al., 1999; Reid and West, 2004). Apneic periods (essentially, diving in aquatic species) range from a few seconds to literally hours, depending upon species, metabolic rate, and temperature. Understanding the dynamics of control of intermittent ventilation in non-endothermic vertebrates has vexed researchers for decades (Gottlieb and Jackson, 1976; Burggren and Shelton,

Ventilation Control in Amphibians and Air-Breathing Fishes  |  165

1979; Boutilier and Shelton, 1986; West et al., 1989; Milsom, 1991; Kinkead and Milsom, 1996), as they have tried to understand how the lung ventilation is reflexly stimulated. Compounding the analysis is the fact that amphibians are bi- or trimodal breathers, which provides a whole additional layer of complexity of chemoreceptive control. The simplest hypothesis for what triggers the initiation of air breathing is that here exist regulatory set-point(s) for “acceptable” blood PO2, PCO2 or [H+]. When a threshold level is crossed (increased PCO2 or decreased PO2 or pH), then lung ventilation is triggered. This “threshold hypothesis” has much appeal, fitting in with the steady-state, homeostatic view of ventilatory control in mammals and birds, which typically are constant breathers that experience relatively little variation in blood gases and acid-base status. Unfortunately, analyses of blood gases and pH during bouts of intermittent breathing in amphibians reveal only a moderate correlation between a short-term specific threshold (e.g. PO2 15 kPa or pH 7.45) and the onset of lung ventilation following an apneic period in anuran amphibians (e.g. Coelho and Smatresk, 2003; Boutilier and Shelton, 1986). Feedback from intrapulmonary chemoreceptors may be more influential in terminating or initiating apneic episodes, but Kinkead and Milsom (1996) report an indirect modulatory effect rather than a direct control of the intermittent breathing pattern by such receptors. One of the confounding factors in understanding how the internal respiratory environment influences intermittent breathing in amphibians may lie in the large contribution of cutaneous gas exchange to total gas exchange in most amphibians. In the adult bullfrog, for example, the skin accounts for approximately 10-25% of total O2 uptake and up to 80% of total CO2 elimination (Gottlieb and Jackson, 1976; Burggren and West, 1982). Intracardiac admixture of systemic venous blood draining the skin with systemic venous blood from non-cutaneous systemic vascular beds will elevate arterial PO2 and reduce arterial PCO2 an effect that may grow as the apneic period progresses. This diminishes the signal for arterial blood-monitoring chemoreceptors (e.g. aortic arch, carotid labyrinth) that would normally occur during interruption of pulmonary ventilation in a strictly lung breather. Clearly, the additional study of factors—both peripheral and CNS - terminating apnea in intermittent breather is highly warranted.

Development of Chemoreceptive Ventilatory Control in Amphibians Almost all amphibians begin life with embryonic/larval stages that are almost entirely aquatic. Later in their life cycle, they develop from bimodal (skin, gills) into trimodal breathers with the addition of pulmonary ventilation. The changing importance of respiratory organs during development, evident from gas exchange partitioning studies, has been investigated in many anuran species (see Burggren and West; 1982; Burggren

166  |  Airway Chemoreceptors in the Vertebrates and Just, 1992; de Souza and Kuribara, 2006). Given the considerable ontogenetic restructuring of gas exchange organs and their perfusion that also happens to occur, not surprisingly the sites of chemoreceptors and mechanoreceptors provide sensory feedback involved in the regulation of ventilation also change following larval development.

Ventilatory Responses to Lung Inflation and Hypoxia

Most studies on the development of the respiratory regulatory system have focused on anuran larvae (“tadpoles”), which have become popular models for probing vertebrate respiratory development (Reid and Milsom, 1998; Gdovin et al., 1999; Straus, 2000; Wassersug and Yamashita, 2000; Straus et al., 2001). In early larval stages, anurans respond to aquatic hypoxia by increasing buccal pumping frequency, which in turn increase irrigation of the internal gills (see Burggren and Just, 1992). This response is evident as early as the Taylor-Kollros stage I, even before the appearance of internal gills (Burggren and Doyle, 1986). As lungs develop, pulmonary gas exchange is also increased by hypoxic stimulation. The sensory system involved in the hypoxic stimulation of gill ventilation appears to involve receptors at two locations. The larvae of bullfrogs (Rana catesbeiana) as early as stage V through stage XIX show rapid response (from 1.3 to 3.3 sec depending on stages) to inhalation of hypoxic or hyperoxic water or water laced with sodium cyanide (NaCN), a stimulant of O2-sensitive receptors ( Jia and Burggren, 1997a; Straus et al., 2001). This response is subsequently abolished by the removal of the first gill arch (Figure 3, Jia and Burggren, 1997b). Neurophysiological recordings have subsequently been made from O2-sensitive neurons on the first gill arch of bullfrog tadpoles (Strauss et al., 2001). A second, slower hypoxic response to inhalation of hypoxic water (varying from 7.7 to 19 sec) persists after the first gill arch has been removed, indicating another population of more centrally located O2sensitive receptors in larval anurans (West and Van Vliet, 1992; Jia and Burggren, 1997b). The specific location and structure of these “non-branchial” receptors has not been identified. The carotid labyrinth is an important site of chemoreception in adult amphibians, as discussed above, but in the anurans Rana catesbeiana and Xenopus laevis, and the urodele Ambystoma tiginnum, this structure is not fully developed until the completion of metamorphosis (Malvin, 1985, 1989; Kusakabe, 2002). The adrenergic cells of the branchial shunt vessels in larval Ambystoma tigrinum may also be the site of arterial blood chemoreceptors (Malvin and Dail, 1986). In vitro characterization of the respiratory neural output from the brain stem of premetamorphic bullfrog (stages VIII to XVI) shows that neither the gill nor lung fictive ventilation frequency is affected by severe hypoxia (Winmill et al., 2005). This indicates the absence of central O2-sensitive receptors for stimulating ventilation during late larval development and supports the notion of arterial O2 receptors in anuran larvae. However, direct evidence indicating the existence of peripheral O2-sensitive receptors reflexly affecting both gill and lung ventilation is still lacking.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  167

The increase in gill ventilation frequency in response to aquatic hypoxia in larval anuran amphibians quickly diminishes as lung ventilation begins and becomes progressively more important to O2 consumption. Eventually hypoxic branchial ventilatory responses and branchial ventilation itself then disappears with subsequent development, and is replaced by the typical ventilatory responses of the adults (West and Burggren, 1982; Burggren and Doyle, 1986).

Pulmonary Stretch (Mechano-)Receptors

In addition to O2- and CO2-sensitive receptors, pulmonary mechanoreceptors are also involved in regulation of gill ventilation in anuran larvae. Gill ventilation frequency

Figure 3: Effect on branchial ventilation of the injection of water containing 0.5% NaCN into the inhalent water stream in an unanesthetized stage VI larva of the bullfrog (Rana catesbeiana). (A) Control animal with intact gill arch 1. (B) Larva following surgical removal of gill arch 1. Note that the rapid (within 2 sec) onset of the response to NaCN is completely eliminated with removal of gill arch 1 (from Jia and Burggren, 1997b).

168  |  Airway Chemoreceptors in the Vertebrates typically decreases following a single air breath in the larvae of Rana catesbeina (Figure 4). In bullfrog larvae at stage XVII-XIX, artificial inflation of lungs with nitrogen, air or oxygen temporarily reduces gill ventilation frequency (West and Burggren, 1983). This finding is supported by the study on decerebrate larvae at the same developmental stage (Gdovin et al., 1998). Larvae at stage XVI-XIX showed reduced gill ventilation frequency following lung inflation by cranial nerve VII recording and electromyogram of the buccal levator muscle (Gdovin et al., 1998). After the initial decrease in gill ventilation frequency, lung inflation with nitrogen subsequently increased gill ventilation; on the other hand, initial oxygen inflation was subsequently followed by a reduction in gill ventilation (West and Burggren, 1983). These experiments suggest that input from the pulmonary stretch receptor initially causes a reflex reduction in gill ventilation frequency, while the longer term changes resulting from nitrogen or oxygen inflation are mediated by input from O2-sensitive chemoreceptors in the lungs or pulmonary vessels. These spatio-temperal interactions of chemo- and mechanoreceptors in larval stages likely ensure optimal O2 acquisition from both respiratory media after the pulmonary system has developed, but before the gills undergo developmentally associated apoptosis. These interactions between chemo- and mechanoreceptors can also help minimize the loss of O2 from blood through the gills into surrounding water when environmental aquatic hypoxia reverses the PO2 gradient across the branchial membranes.

CO2-Sensitive Chemoreceptors

The location of central CO2-sensitive chemoreceptors in larval anuran amphibians has been demonstrated in vitro. Hypercapnia stimulates fictive gill ventilation in stage X to XIX bullfrog larvae. After stage XX, perfusion of the brain stem with hypercapnic solution increasingly stimulates fictive lung ventilation (Torgerson et al., 1997). The locations of CO2-sensitive receptors are within the ventral medulla: chemical and protease lesions at specific sites localized these chemoreceptors to be adjacent to the origin of cranial nerves V and X (Taylor et al., 2003). There is as yet no evidence for the presence of peripheral CO2-receptors on the internal gills in larval amphibian during early development, in contrast to the presence of these in air-breathing fishes (see below). In summary, amphibian anuran larvae respond to both hypoxia and hypercapnia by increasing gill ventilation frequency in the early stages and then show a developmental transition to predominant adjustments in lung ventilation. The location of receptors sensing ambient O2 level is on the first branchial arch in early development, along with some other likely sites, such as the aorta or brain stem. The peripheral O2-sensitive chemoreceptors migrate during development from the gill arch(es) to the carotid labyrinth following the completion of metamorphosis. CO2-sensitive chemoreceptors are found within ventral medulla. The investigation of the development of ventilatory control in amphibians has been heavily focused on anuran larvae. While the cardiovascular anatomical and physiological

Ventilation Control in Amphibians and Air-Breathing Fishes  |  169

development of urodeles (salamanders) has been characterized (see Malvin, 1985, 1989), we know relatively little about the extent to which the development of respiratory regulation in anurans maps onto salamanders and newts.

Buccal Pumping

Reflex inhibition of gill ventilation

1 cm H2 0

Air breath

Time (10 sec) Figure 4: A single air breath results in a reflex inhibition of gill ventilation in an unrestrained larva (St TK XIX) of the bullfrog, Rana catesbeiana (after West and Burggren, 1983).

Chemoreceptive Control in Air-Breathing Fishes In addition to internal gills ventilated by a conventional piscine buccal pump, air-breathing fishes typically possess various methods for exploiting air-breathing, including nonpulmonary air-breathing organs (ABOs) or, in the case of the lungfishes, true lungs (see above)(Figure 1). Like their gills, the ABOs/lungs of air-breathing fishes have sensory innervation, allowing transmission of chemo- and mechanoreceptor information from these organs to the CNS, and allowing motor control over ventilation of the gills (and perhaps even neuromodulation of the sensors themselves), as we will now consider.

Chemoreception in Air-Breathing Fishes Although a systematic examination of chemoreceptors in air-breathing fishes is lacking, several studies of both an in vivo and in vitro nature reveal their existence in both central and peripheral locations.

Central Chemoreception

The appearance of central CO2/pH chemoreception is often linked to terrestriality related to air breathing and the associated elevated venous blood PCO2, as previously mentioned. However, air-breathing fishes also possess central CO2/pH chemoreception. In in vitro brainstem preparations, fictive air-breathing frequency increased following hypercarbia in superfusing solution in the long nose gar, Lepisosteus osseus (Wilson et al., 2000; Remmers et al., 2001). In the South American lungfish, Lepidosiren

170  |  Airway Chemoreceptors in the Vertebrates paradoxa, the reduction of pH in the solution perfusing the isolated fourth cerebral ventricle increased lung ventilation and breathing frequency (Sanchez et al., 2001). These data suggest the presence of central acid-base and CO2 receptors in a few species of air-breathing fishes. However, there is insufficient evidence and too few species examined to conclude that central CO2/pH chemoreception evolves concurrent with the evolution of air-breathing in fishes.

Peripheral Chemoreception

As in water-breathing fishes and larval amphibians, the branchial O2 chemoreceptors of air-breathing fishes monitor gas composition near the gills and control the net level of ventilation—i.e. ventilation of gills and ABOs (Smatresk, 1990). On the other hand, some species, such as lungfish, rely on internal arterial receptors for regulating respiratory and cardiovascular behavior in response to hypoxia or hypercapnia (Perry et al., 2005). We now consider peripheral chemoreception of air-breathing fishes. Cranial nerve denervation has been a direct method to test peripheral chemoreceptive control, despite confounding side effects such as stress, metabolic depression and decreased arterial PO2 (Graham, 1997; McKenzie et al., 1991). Denervation of cranial nerves IX and X had no effect on air-breathing responses to aquatic hypoxic conditions in the bowfin, Amia calva. Thus, the O2-sensitive chemoreceptor responsible for increasing air-breathing frequency does not appear to reside on the gills of Amia calva (Hedrick and Jones, 1999). However, pseudobranch ablation in the same species abolished the air-breathing responses to aquatic hypoxia, indicating that the O2-sensitive chemoreceptors may be located on the pseudobranch instead of gills (McKenzie et al., 1991). Mechanical movement and compression of the gas bladder of Amia calva stimulates a ventilatory response, indicating the likely presence of a stretch receptor (Hedrick and Jones, 1999). The African lungfish (Protopterus dolloi) is another species that relies only on internal O2 chemoreceptor, because only aerial hypoxia induced the secretion of catecholamines and cardiorespiratory responses (Perry et al., 2005). The long nose gar (Lepisosteus osseus) also possesses internal chemoreceptors for O2. A decrease in arterial PO2 and injection of NaCN into the ventral aorta both stimulated air-breathing frequency; however, air-breathing frequency also increased as O2 level in air bladder fell, suggesting peripheral ventilatory control mechanism also exists in this species (Smatresk et al., 1986). In addition to O2-sensitive chemoreceptors, lungfish may also have peripheral CO2/pH-sensitive receptors. Lung ventilation increased by 20% in hypercarbia (6.5 KPa in both water and air) when the cerebral ventricular system was superfused with normocarbic solution in the South American lungfish, Lepidosiren paradoxa (AminNaves et al., 2007). Innervation of the ABOs has also been studied in the Indian catfish, Heteropneustes fossilis, the Asian catfish, Pangasius bypophthalmus, and the Nile bichir, Polypterus bichir

Ventilation Control in Amphibians and Air-Breathing Fishes  |  171

bichir. Expression of various neuropeptides was found in the air-breathing organs of these species (Mauceri et al., 2005; Zaccone et al., 2007). Neuroendocrine cells and their innervation have been located in the lungs of Protopterus aethiopicus, Amia calva, Polypterus delhezi, Polypterus ornatipinnis and Polypterus bichir bichir (Zaccone et al., 1989, 1995, 2007). Immunoreactivity of several neuropeptides was found in these cells. The role of these transmitters may be autonomic control of circulation and respiration. However, the relative importance and significance of these signals to the respiratory responses of air-breathing fishes is still enigmatic, and additional studies are needed to link the morphology, function and innervation of the neuroendocrine cells.

Ventilatory Responses in Air-Breathing Fishes Precise regulation of the ventilation of both gills and ABOs has been studied in numerous species of air-breathing fishes (see for example Randall et al., 1981; Graham, 1997; Brauner et al., 2004). In contrast to ventilatory chemoreception in amphibians, where there is considerable conformity in regulatory patterns, amongst the air- breathing fishes there appear to be three major groupings: aquatic hypoxia driven, aerial hypoxic driven, and a hybrid pattern of responses evident in the lungfishes. This information is grouped and summarized here according to the diverse ventilatory responses to hypoxia or hypercapnia in air and water.

Ventilation Driven Primarily by Aquatic Hypoxia

The first group of air-breathing fishes, primarily responsive to aquatic hypoxia, comprises both facultative and obligatory air-breathers. As an example, aquatic hypoxia below 6.5 KPa stimulates both gill ventilation and air-breathing frequency in the South American tamoatá, Hoplosternum littorale (Affonso and Rantin, 2005) and the jeju, Hoplerythrinus unitaeniatus (Oliveira et al., 2004). Also in this first category is the gourami, Trichogaster trichopterus, an obligate air-breather from South-East Asia that has a labyrinth organ contained within a suprabranchial chamber. This species also responds to both aquatic and aerial hypoxia (PO2 ~7 KPa) and aquatic hypercapnia (PCO2 ~3 KPa) by increasing air-breathing frequency. Hypoxia also increases O2 uptake by the labyrinth (Burggren, 1979). In the bowfin, Amia calva, air breaths are categorized into two types. The first includes exhalation followed by inhalation, and it is stimulated by both aerial and aquatic hypoxia. The second is characterized by inhalation only, which may be used for regulating gas bladder volume and buoyancy (Hedrick and Jones, 1999). Low aquatic O2 partial pressure is the main stimulant driving these respiratory responses, and the branchial chemoreceptor is the predominant sensor eliciting these responses. This group of responses for air-breathing fishes resemble those observed in water-breathing teleost fishes (see Jonz and Nurse, Chapter 1) and larval anuran amphibians (e.g. Burggren and Just, 1992; Straus, 2000).

172  |  Airway Chemoreceptors in the Vertebrates

Ventilation Driven Primarily by Aerial Hypoxia

The second group of air-breathing species includes fishes such as the mudskipper, Periophthalmodon schlosseri and the Australian desert goby, Chlamydogobius eremius. Unlike the first group of air-breathing fishes that respond primarily to aquatic hypoxia, the species in this latter group respond only modestly, if at all, to PO2 changes in water. Rather, they increase gill and ABO ventilation frequency markedly in response to aerial hypoxia. The mudskipper, for example, responds to aerial hypoxia by increasing both air-breathing frequency and tidal volume of the vascularized buccopharyngeal cavity (Aguilar et al., 2000). The desert goby decreases its opercular movements in aerial hypoxia as severe as a PO2 of ~2 KPa. In a (futile) attempt to cope with experimental severe aerial hypoxia, this species relies more on the bubbles in buccal cavity for O2 acquisition, increasing the percentage of total O2 consumption via buccal bubbles during aerial hypoxic exposure (Thompson and Withers, 2002).

Ventilation Driven by both Aquatic and Aerial Hypoxia—the Lungfishes

The last functionally-categorized group of air-breathing fishes is represented by the lungfish. The Sarcopterygii are characterized by true lungs resembling those of amphibians. They have reduced gills (especially the anterior-most arches) and generally share similar ventilatory control mechanism and responses with tetrapod vertebrates. Lung ventilation is stimulated by both aerial and arterial hypoxia (below 7 KPa) in the South American lungfish, Lepidosiren paradoxa and the African lungfish, Protopterus dolloi (Sanchez et al., 2001; Perry et al., 2005). Aquatic hypoxia has little or no effect on pulmonary ventilatory rate of these species. However, the Australian lungfish, Neoceratodus forsteri, responds to aquatic hypoxia (3 kPa) with increased branchial ventilation and air-breathing frequency (Fritsche et al., 1993). The difference in response among lungfish species may be due to the relative importance of air breathing. The Australian lungfish is a facultative air breather that begins using lung ventilation in aquatic hypoxia. The other two species are obligate airbreathers with reduced gill surface area ( Johansen, 1970; Graham, 1997; Fritsche et al., 1993; Sanchez et al., 2001). The respiratory regulatory system of Neoceratodus may rely more on the signals from water, it being more critical as a respiratory medium for maintaining normal aerobic metabolism in this more aquatic species. However, the existence of branchial O2-sensitive chemoreceptors in the American and African lungfishes cannot be excluded until such time as more direct loss-of-function experiments on gills—e.g. denervation of cranial nerves IX and X—are completed. The major differences in hypoxic ventilatory responses discussed above likely reflect differences in habitat rather than fall into any sort of strict taxonomic pattern. Airbreathing fishes of the Amazon Basin and many South-East Asian habitats experience both hypoxia and hypercapnia on a daily and seasonal basis due to alternating cycles of photosynthesis, respiration, decay of vegetation, flooding, etc. In contrast, the desert goby and lungfish live in temporary ponds and may be completely out of water during

Ventilation Control in Amphibians and Air-Breathing Fishes  |  173

the dry season. The mudskipper resides in mud burrows filled with extremely hypoxic water during low tide, where they survive in the environment by solely using aerial gas exchange (Aguilar et al., 2000; Sanchez et al., 2001; Thompson and Withers, 2002). Heavy utilization of aerial gas exchange may have contributed to the loss of aquatic gas sensing in these species. Such adaptation to their habitat makes the chemoreceptive control of ventilation in this group more similar to adult anurans, despite the fact that no evidence has pointed out the actual location of their O2-sensitive chemoreceptors. Also, little work has been done regarding the existence of fish-like (peripheral) or amphibian-like (central) CO2 chemoreception in these species.

Conclusions General Trends Air-breathing fishes and amphibians occupy a fascinating functional transition point in the evolution of terrestrial tetrapods from their aquatic fish-like ancestors. Not surprisingly, considerable attention has been paid to the chemoreceptors that regulate ventilation, as well as the ventilatory responses themselves. Perhaps reflecting the extreme diversity of the air-breathing habit in air-breathing fishes and amphibians, there are relatively few general lessons that can be derived with certainty from both interpretation of existing studies and planning of future ones. However, a few key principles do emerge: 1. the more aquatic in nature the animal, the greater is the tendency to have sophisticated receptors for, and to respond primarily to, changes in O2 levels in the interior milieu; 2. as a transition to air-breathing and terrestrial life develops, the greater is the likelihood of having CO2-/pH-sensitive receptors that participate in regulation of ventilation; 3. central chemosensitivity is a highly conserved trait, evident in all semiamphibious and amphibious animals.; 4. despite the difference in exact location of O2-sensitive receptors, O2 chemoreception remains peripherally located while CO2 chemoreception turns centrally during evolution.

Unanswered Questions/Future Experiments When our understanding of a subject like chemoreception in amphibious vertebrates is so incomplete, not surprisingly several areas ripe for future experimentation emerge, including: 1. the role of daily and seasonal influences on chemoreception and ventilatory control, particularly in those animals that live in environments with large changes in temperature, pH, CO2 and O2 levels;

174  |  Airway Chemoreceptors in the Vertebrates 2. the specific location and morphological/neurophysiological characterization of chemoreceptors—both central and peripheral—in air-breathing fishes and amphibians; 3. explanation of the lack of hypercapnic ventilatory responses in animals whose isolated brainstems prove to be exquisitely sensitive to CO2 and pH; 4. the role, if any, of efferent innervation of chemoreceptors, and the associated extent of neuromodulation that might occur; 5. better understanding of the developmental changes in chemoreception, brought into an explicit “evo-devo” context; 6. the interaction of chemo- and mechano-receptors in the regulation of both aquatic and aerial gas exchange; 7. the effect of chronic hypoxia and hypercarbia (hypercapnia) on ventilatory behavior in bimodal breathers, especially during development when physiological plasticity may be at its greatest. 8. whether in animals heavily exploiting the cutaneous gas exchange the general body surface has respiratory chemo-receptors involved in facilitating behaviors or processes. Since a strong interest exists in the evolution of chemoreception, as reflected in the many chapters considering this subject in the present volume, a final perspective is that investigators be encouraged to take a truly comparative, multi-species, systematic approach. Currently, we are typically faced with attempting to fit into a patchy mosaic of emerging information the results of an in vivo study of branchial denervation here, and an in vitro investigation of brainstem responses there, most likely carried out on distantly related species. The most rapid progress will come when, instead, a systematically robust and taxonomically relevant suite of species is concurrently investigated by the same investigators under the same experimental conditions with the same techniques. The rewards of such an approach, though demanding, will be manifold.

Acknowledgements The authors acknowledge the National Science Foundation (operating grant #IOB0614815 to WB) for support during the preparation of this chapter.

Notes 1. It could be argued that “diffusion” and “convection” are indeed also “modes” of gas exchange, but here we shall confine the use of mode to structure rather than process.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  175

References Affonso, E. G., and Rantin, F.T. 2005. Respiratory responses of the air-breathing fish Hoplosternum littorale to hypoxia and hydrogen sulfide. Comp. Biochem. Physiol. C. 141:275-280. Aguilar, N.M., Ishimatsu, A., Ogawa, K., and Huat, K.K. 2000. Aerial ventilatory responses of the mudskipper, Periophthalmodon schlosseri, to altered aerial and aquatic respiratory gas concentrations. Comp. Biochem. Physiol. A. 127:285-292. Amin-Naves, J., Giusti, H., Hoffmann, A., and Glass, M.L. 2007. Components to the acidbase related ventilatory drives in the South American lungfish Lepidosiren paradoxa. Respir. Physiol. Neurobiol. 155:35-40. Andersen, J.B., Hedrick, M.S., and Wang, T. 2003. Cardiovascular responses to hypoxia and anaemia in the toad Bufo marinus. J. Exp. Biol. 206:857-865. Babak, E. 1911. Folia Neurobiol., Lpz. 539. Bicego-Nahas, K.C., and Branco, L.G.S. 1999. Seasonal changes in the cardiorespiratory responses to hypercarbia and temperature in the bullfrog, Rana catesbeiana. Comp Biochem Physiol. A Mol. Integr. Physiol. 124:221-229. Bicego-Nahas, K.C., Gargaglioni, L.H., and Branco, L.G.S. 2001. Seasonal changes in the preferred body temperature, cardiovascular, and respiratory responses to hypoxia in the toad, Bufo paracnemis. J. Exp. Zool. 289:359-365. Boutilier, R.G. 1984. Characterization of the intermittent breathing pattern in Xenopus laevis. J. Exp. Biol. 110:291-309. Boutilier, R.G., and Toews, D.P. 1981. Respiratory, circulatory and acid-base adjustments to hypercapnia in a strictly aquatic and predominantly skin-breathing urodele, Cryptobranchus alleganiensis. Respir. Physiol. 46:177-192. Boutilier, R.G., and Shelton, G. 1986. Gas exchange, storage and transport in voluntarily diving Xenopus laevis. J. Exp. Biol. 126:133-155. Branco, L.G.S., and Glass, M.L. 1995. Ventilatory responses to carboxyhaemoglobinaemia and hypoxic hypoxia in Bufo paracnemis. J. Exp. Biol. 198:1417-1421. Brauner, C.J., Matey, V., Wilson, J.M., Bernier, N.J., and Val, A.L. 2004. Transition in organ function during the evolution of air-breathing; insights from Arapaima gigas, an obligate airbreathing teleost from the Amazon. J. Exp. Biol. 207(9):1433-8143. Burggren, W.W. 1979. Bimodal gas exchange during variation in environmental oxygen and carbon dioxide in the air breathing fish Trichogaster trichopterus. J. Exp. Biol. 82:197-213. Burggren, W.W. 2006. Complexity change during physiological development.  In: Comparative Developmental Physiology, S. Warburton, W. W. Burggren, B. Pelster, C. Reiber, and J. Spicer (Eds.). Oxford University Press, New York, pp. 174-190. Burggren, W.W., and Shelton, G. 1979. Gas exchange and transport during intermittent breathing in chelonian reptiles. J. Exp. Biol. 82: 75-92. Burggren, W.W., and West, N.H. 1982. Changing respiratory importance of gills, lungs, and skin during metamorphosis in the bullfrog Rana catesbeiana. Respir. Physiol. 47:151–164. Burggren, W.W., and Doyle, M. 1986. Ontogeny of regulation of gill and lung ventilation in the bullfrog, Rana catesbeiana. Respir. Physiol. 66:279-291. Burggren, W.W. and Just, J.J. 1992. Developmental changes in amphibian physiological systems. In: Environmental Physiology of the Amphibia, M.E. Feder and W.W. Burggren (Eds.). University of Chicago Press, Chicago, pp. 467-530.

176  |  Airway Chemoreceptors in the Vertebrates Burggren, W.W., and Vitalis, T.Z. 2004. The interplay of cutaneous water loss, gas exchange and blood flow in the toad, Bufo woodhousei: Adaptations in a terrestrially-adapted amphibian. J. Exp. Biol. 208:105-112. Burggren, W.W., and Monticino, M.G. 2005. Assessing Physiological Complexity. J. Exp. Biol. 208:3221-3232. Coates, E.L. 2001. Olfactory CO2 chemoreceptors. Respir. Physiol. 129:219–229. Coates, E.L., and Ballam, G.O. 1990. Olfactory receptor response to CO2 in bullfrogs. Am. J. Physiol. 258:R1207–R1212. Coates, E.L., Wells, C.M., and Smith, R.P. 1998. Identification of carbonic anhydrase activity in bullfrog olfactory receptor neurons: histochemical localization and role in CO2 chemoreception. J. Comp. Physiol. A. 182:163–174. Coelho, F.C., and Smatresk, N.J. 2003. Resting respiratory behavior in minimally instrumented toads-effects of very long apneas on blood gases and pH. Braz. J. Biol. 63:35-45. DeLaney, R.G., and Fishman, A.P. 1977. Analysis of lung ventilation in the aestivating lungfish Protopterus aethiopicus. Am J Physiol. 233:R181-R187. de Souza, S.C., and Kuribara, C.M. 2006. Metabolic scaling associated with unusual size changes during larval development of the frog, Pseudis paradoxus. J. Exp. Biol. 209:1651-1661. Evans, D.H., Piermarini, P.M., and Choe, K.P. 2005. The multifunctional fish gill: dominant site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous waste. Physiol. Rev. 85:97-177. Feder, M.E., and Burggren, W.W. 1985. The regulation of cutaneous gas exchange in vertebrates. In: Current Topics and Trends: Comparative Physiology and Biochemistry. Vol. A: Respiration, Circulation, Metabolism. R. Gilles (Ed.). Springer-Verlag, Berlin, pp. 101-113. Feder, M.E., and Burggren, W.W. 1992. Environmental Physiology of the Amphibia. University of Chicago Press, Chicago. Feldman, J.L., Mitchell, G.S., and Nattie, E.E. 2003. Breathing: rhythmicity, plasticity, chemosensitivity. Annu. Rev. Neurosci. 26:239-266. Fernandes, M.S., Giusti, and Glass, M.L. 2005. An assessment of dead space in pulmonary ventilation of the toad Bufo schneideri. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 142:446-450. Florindo, L.H., Reid, S.G. Kalinin, A.L. Milsom, W.K., and Rantin, F.T. 2004. Cardiorespiratory reflexes and aquatic surface respiration in the neotropical fish tambaqui (Colossoma macropomum): acute responses to hypercarbia. J. Comp. Physiol. B. 174:319-28. Fritsche, R., Axelsson, A., Franklin, C.E., Grigg, G.G., Holmgren, S., and Nilsson, S. 1993. Respiratory and cardiovascular responses to hypoxia in the Australian lungfish. Respir Physiol. 94:173-187. Gardner, M.N., Smits, A.W., and Smatresk, N.J. 2000. The ventilatory responses of the caecilian Typhlonectes natans to hypoxia and hypercapnia. Physiol. Biochem. Zool. 73:23-29. Gargaglioni, L.H., and Branco, L.G.S. 2000. Role of nucleus isthmi in the ventilatory response to hypoxia of Bufo paracnemis. Respir. Physiol. 119:31-39. Gargaglioni, L.H., and Branco, L.G.S. 2004. Nucleus isthmi and control of breathing in amphibians. Respir. Physiol. Neurobiol. 143:177-186. Gargaglioni, L.H., and Milsom, W.K. 2007. Control of breathing in anuran amphibian. Comp. Biochem. Physiol. A Mol, Integr. Physiol. 147:665-684.. Gargaglioni, L.H., Coimbra, N.C., and Branco, L.G.S. 2002. Chemical lesions of the nucleus isthmi increase the hypoxic and hypercarbic drive to breathing of toads. Respir. Physiol. Neurobiol. 132: 289-299.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  177 Gdovin, M.J., Torgerson, C.S., and Remmers, J.E. 1998. Neurorespiratory pattern of gill and lung ventilation in the decerebrate spontaneously breathing tadpole. Respir. Physiol. 113:135–146. Gdovin, M.J., Torgerson, C.S., and Remmers, J.E. 1999. The fictively breathing tadpole brainstem preparation as a model for the development of respiratory pattern generation and central chemoreception. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 124:275-286. Getchell, T.V., and Shepherd, G.M. 1978. Responses of olfactory receptor cells to step pulses of odour at different concentrations in the salamander. J. Physiol. 282:521-540. Gheshmy, A., Vukelich, R., Noronha, A., and Reid, S.G. 2006. Chronic hypercapnia modulates respiratory-related central pH/CO2 chemoreception in an amphibian, Bufo marinus. J. Exp. Biol. 209:1135-1146. Gilmour, K.M., Milsom, W.K., Rantin, F.T., Reid, S.G., and Perry, S.F. 2005. Cardiorespiratory responses to hypercarbia in tambaqui Colossoma macropomum: chemoreceptor orientation and specificity. J. Exp. Biol. 208:1095-1107. Goniakowska-Witalińska, L.1997. Neuroepithelial bodies and solitary neuroendocrine cells in the lungs of Amphibia. Microsc. Res. Tech. 37:13-30. Gottlieb, G., and Jackson, D.C. 1976. Importance of pulmonary ventilation in respiratory control in the bullfrog. Am. J. Physiol. 230:608-613. Graham, J.B. 1997. Air-Breathing Fishes. Academic Press, New York. Harris, M.B., Wilson, R.J. Vasilakos, K., Taylor, B.E., and Remmers, J.E.. 2002. Central respiratory activity of the tadpole in vitro brain stem is modulated diversely by nitric oxide. Am. J. Physiol. Regul. Integr. Comp. Physiol. 283:R417-428. Hedrick, M.S., and Jones, D.R. 1999. Control of gill ventilation and air-breathing in the bowfin Amia calva. J. Exp. Biol. 202:87-94. Hedrick, M.S., and Morales, R.D. 1999. Nitric oxide as a modulator of central respiratory rhythm in the isolated brainstem of the bullfrog (Rana catesbeiana). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 124:243-251. Hedrick, M.S., Morales, R.D. Parker, J.M., and Pacheco, J.L. 1998. Nitric oxide modulates respiratory-related neural activity in the isolated brainstem of the bullfrog. Neurosci. Lett. 251:81-84. Hedrick, M.S., Chen, A.K., and Jessop, K.L. 2005. Nitric oxide changes its role as a modulator of respiratory motor activity during development in the bullfrog (Rana catesbeiana). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 142:231-240. Hou, P.C., and Huang, S.P. 1999. Metabolic and ventilatory responses to hypoxia in two altitudinal populations of the toad, Bufo bankorensis. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 124:413-421. Inoue, I. 1978. Reflex discharges in the hypoglossal nerve elicited by stimulating parts of the frogs’ tongue. J. Kyushu Dent. Sci. 32:119-131. Jia, X.X., and Burggren, W.W. 1997a. Developmental changes in chemoreceptive control of gill ventilation in larval bullfrogs (Rana catesbeiana). I. Reflex ventilatory responses to ambient hyperoxia, hypoxia and NaCN. J. Exp. Biol. 200:2229-2236. Jia, X.X., and Burggren, W.W. 1997b. Developmental changes in chemoreceptive control of gill ventilation in larval bullfrogs (Rana catesbeiana). II. Sites of O2-sensitive chemoreceptors. J. Exp. Biol. 200:2237–2248. Johansen, K. 1970. Air breathing in fishes. Fish Physiology Vol. IV. W.S. Hoar and D.J. Randall (Eds.). Academic Press, New York.

178  |  Airway Chemoreceptors in the Vertebrates Jones, D.R., and Chu, C. 1988. Effect of denervation of carotid labyrinths on breathing in unrestrained Xenopus laevis. Respir. Physiol. 73:243-255. Jonz, M., and Nurse, C. 2009. Oxygen-sensitive neuroepithelial cells in the gill of aquatic vertebrates. In: Airway chemoreceptors in the vertebrates. Structure, evolution and function. G. Zaccone, E. Cutz, D. Adriaensen and C.A. Nurse (Eds.). Science Publisher. Enfield, NH. pp. 1-30. Jorgensen, C. B. 2000. Amphibian respiration and olfaction and their relationships: from Robert Townson (1794) to the present. Biol. Rev. Camb. Philos. Soc. 75:297-345. Kinkead, R., and Milsom, W.K. 1994. Chemoreceptors and control of episodic breathing in the bullfrog (Rana catesbeiana). Respir. Physiol. 95:81-98. Kinkead, R., and Milsom, W.K. 1996. CO2-sensitive olfactory and pulmonary receptor modulation of episodic breathing in bullfrogs. Am. J. Physiol. 270(1 Pt 2):R134-R144. Kinkead, R., Harris, M.B., and Milsom, W.K. 1997. The role of the nucleus isthmi in respiratory pattern formation in bullfrogs. J. Exp. Biol. 200:1781-1793. Kusakabe, T. 2002. Carotid labyrinth of amphibians. Microsc. Res. Tech. 59:207–226. Kusakabe, T., Kawakami, T., and Takenaka, T. 1995. Peptidergic innervation in the amphibian carotid labyrinth. Histol. Histopathol. 10:185-202. Little, C. 1983. The Colonisation of Land. Cambridge University Press, Cambridge. Maina, J.N. 2002. Fundamental structural aspects and features in the bioengineering of the gas exchangers: comparative perspectives. Adv. Anat. Embryol. Cell. Biol. 163:III-XII, 1-108. Malvin, G.M. 1985. Cardiovascular shunting during amphibian metamorphosis. In: Cardiovascular Shunts: Ontogenetic, Phylogenetic and Clinical Aspects. Alfred Benzon Symposium 21., K. Johansen and W.W. Burggren (Eds.). Munksgaard, Copenhagen. pp. 163-178. Malvin, G.M. 1989. Gill structure and function: Amphibian larvae. In: Comparative Pulmonary Physiology: Current Concepts. Vol. 39, Lung Biology in Health and Disease, S.C. Wood (Ed.). Marcel Dekker, New York, pp. 121-151. Malvin, G.M., and Dail, W.G. 1986. Adrenergic innervation of the gills, pulmonary arterial plexus, and dorsal aorta in the neotenic salamander, Ambystoma tigrinum. J. Morphol. 189:67-70. Mauceri, A., Fasulo, S., Minniti, F., Cascio, P.L., Maisano, M., and Zaccone, G. 2005. Neurochemical features of the innervation of respiratory organs in some air-breathing fishes. Ital. J. Zool. 72:175-181. McKenzie, D.J., and Taylor, E.W. 1996. Cardioventilatory responses to hypoxia and NaCN in the neotenous axolotl. Respir. Physiol. 106:255-262. McKenzie, D.J., Burleson, M.L., and Randall, D.J. 1991. The effects of branchial denervation and pseudobranch ablation on cardioventilatory control in an air-breathing fish. J. Exp. Biol. 161:347-365. McLean, H.A., Kimura, N., Kogo, N., Perry, S.F., and Remmers, J.E. 1995. Fictive respiratory rhythm in the isolated brainstem of frogs. J. Comp. Physiol. A. 176:703-713. McMahon, B.R. 1969. A functional analysis of the aquatic and aerial respiratory movements of an African lungfish, Protopterus aethiopicus, with reference to the evolution of the lungventilation mechanism in vertebrates. J. Exp. Biol. 51:407-430. Milsom, W.K. 1991. Intermittent breathing in vertebrates. Annu. Rev. Physiol. 53:87-105. Milsom, W.K., and Jones, D.R. 1977. Carbon dioxide sensitivity of pulmonary receptors in the frog. Experientia. 33:1167-1168. Milsom, W.K., Reid, S.G., Meier, J.T., and Kinkead, R. 1999. Central respiratory pattern generation in the bullfrog, Rana catesbeiana. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 124:253-264.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  179 Noronha-de-Souza, C.R., Bicego, K.C., Michel, G., Glass, M.L., Branco, L.G., and Gargaglioni, L.H. 2006. Locus coeruleus is a central chemoreceptive site in toads. Am. J. Physiol. Regul. Integr. Comp. Physiol. 291:R997-R1006. Oliveira, R.D., Lopes, J.M., Sanches, J.R., Kalinin, A.L., Glass, M.L., and Rantin, F.T. 2004. Cardiorespiratory responses of the facultative air-breathing fish jeju, Hoplerythrinus unitaeniatus (Teleostei, Erythrinidae), exposed to graded ambient hypoxia. Comp. Biochem. Physiol. A. 139:479-485. Olson, K.R. 2002. Vascular anatomy of the fish gill. J. Exp. Zool. 293:214-231. Perry, S.F., and McKendry, J.E. 2001. The relative roles of external and internal CO2 versus H+ in eliciting the cardiorespiratory responses of Salmo salar and Squalus acanthias to hypercarbia. J. Exp. Biol. 204:3963-3971. Perry, S.F., McLean, H.A., Kogo, N., Kimura, N., Kawasaki, H., Sakurai, M., Kabotyanski, E.A., and Remmers, J.E. 1995. The frog brainstem preparation as a model for studying the central control of breathing in tetrapods. Braz. J. Med. Biol. Res. 28:1339-1346. Perry, S.F., Gilmour, K.M., Vulesevic, B., McNeill, B., Chew, S.F., and Ip, Y.K. 2005. Circulating catecholamines and cardiorespiratory responses in hypoxic lungfish (Protopterus dolloi): a comparison of aquatic and aerial hypoxia. Physiol. Biochem. Zool. 78:325-334. Pinder, A.W., and Burggren, W.W. 1986. Ventilation and partitioning of oxygen uptake in the frog Rana pipiens: effects of hypoxia and activity. J. Exp. Biol. 126:453-468. Pinder, A.W., Clemens, D., and Feder, M.E. 1991. Gas exchange in isolated perfused frog skin as a function of perfusion rate. Respir. Physiol. 85:1-14. Randall, D.J. 1970. Gas exchange in fishes. In: Fish Physiology. Vol. IV, W.S. Hoar and D.J. Randall (Eds.). Academic Press, New York. pp. 253-292. Randall, D.J., Burggren,W.W., Haswell, M.S., and Farrell, A.P. 1981. The Evolution of Air Breathing in Vertebrates. Cambridge University Press, Cambridge. Reid, S.G. 2006. Chemoreceptor and pulmonary stretch receptor interactions within amphibian respiratory control systems. Respir. Physiol. Neurobiol. 154:153-164. Reid, S.G., and Milsom, W.K. 1998. Respiratory pattern formation in the isolated bullfrog (Rana catesbeiana) brainstem-spinal cord. Respir. Physiol. 114:239-255. Reid, S.G., and West, N.H. 2004. Modulation of breathing by phasic pulmonary stretch receptor feedback in an amphibian, Bufo marinus. Respir. Physiol. Neurobiol. 142:165–183. Reid, S.G, Meier, J.T., and Milsom, W.K. 2000. The influence of descending inputs on breathing pattern formation in the isolated bullfrog brainstem-spinal cord. Respir. Physiol. 120:197-211. Remmers, J.E., Torgerson, C., Harris, M., Perry, S.F. , Vasilakos, K., and Wilson, R.J.A. 2001. Evolution of central respiratory chemoreception: a new twist on an old story. Respir. Physiol. 129: 211-217. Rocha, P.L., and Branco, L.G. 1998. Seasonal changes in the cardiovascular, respiratory and metabolic responses to temperature and hypoxia in the bullfrog Rana catesbeiana. J. Exp. Biol. 201:761-768. Sakakibara, Y. 1978. Localization of CO2 sensor related to the inhibition of the bullfrog respiration. Jpn. J. Physiol. 28:721-735. Sanchez, A.P., Hoffmann, A., Rantin, F.T., and Glass, M.L. 2001 Relationship between cerebrospinal fluid pH and pulmonary ventilation of the South American lungfish, Lepidosiren paradoxa (Fitz.). J. Exp. Zool. 290:421-425. Sanchez, A.P., Soncini, R., Wang, T., Koldkjaer, P., Taylor, E.W., and Glass, M.L. 2001. The differential cardio-respiratory responses to ambient hypoxia and systemic hypoxaemia in the South American lungfish, Lepidosiren paradoxa. Comp. Biochem. Physiol. A. 130:677-687.

180  |  Airway Chemoreceptors in the Vertebrates Sanders, C.E., and Milsom, W.K. 2001. The effects of tonic lung inflation on ventilation in the American bullfrog Rana catesbeiana Shaw. J. Exp. Biol. 204:2647–2656. Sheafor, E.A, Wood, S.C., and Tattersall, G.J. 2000. The effect of graded hypoxia on the metabolic rate and buccal activity of a lungless salamander (Desmognathus fuscus). J. Exp. Biol. 203:3785-3793. Shoemaker, V.H., Hillman, S.S., Hillyard, S.D., Jackson, D.C., McClanahan, L.L., Withers, P.C., and Wygoda, M.L.1992. Exchange of water, ions and respiratory gases in terrestrial amphibians. In: Environmental Physiology of Amphibians, M.E. Feder and W. W. Burggren (Eds.). University of Chicago Press, Chicago. pp. 125-150. Smatresk, N.J. 1990. Chemoreceptor modulation of endogenous respiratory rhythms in vertebrates. Am. J Physiol. 259:R887-R897. Smatresk, N.J., and Smits, W. 1991. Effects of central and peripheral chemoreceptor stimulation on ventilation in the marine toad, Bufo marinus. Respir. Physiol. 83:223–238. Smatresk, N.J., Burleson, M.L., and Azizi, S.Q. 1986. Chemoreflexive responses to hypoxia and NaCN in longnose gar: evidence for two chemoreceptor loci. Am. J. Physiol. 251: R116-R125. Smits, A.W., West, N.H., and Burggren, W.W. 1986. Pulmonary fluid balance following pulmocutaneous baroreceptor denervation in the toad. J. Appl. Physiol. 61:331-337 Smyth, D.H. 1939. The central and reflex control of respiration in the frog. J. Physiol. (London) 95:305-327. Steffensen, J.F., and Lomholt, J.P. 1983. Energetic cost of active branchial ventilation in the sharksucker, Echeneis naucrates. J. Exp. Biol. 103:185-192. Straus, C. 2000. Ontogeny of respiratory muscle control. Evidence from the amphibian model. Rev. Mal. Respir. 17:585-590. Straus, C., Wilson, R.J., and Remmers, J.E. 2001. Oxygen sensitive chemoreceptors in the first gill arch of the tadpole, Rana catesbeiana. Can. J. Physiol. Pharmacol. 79:959-962. Taylor, B.E., Harris, M.B., Leiter, J.C., and Gdovin, M.J. 2003. Ontogeny of central CO2 chemoreception: chemosensitivity in the ventral medulla of developing bullfrogs. Am. J. Physiol. 285:1461–1472. Taylor, E.W., Jordan, D., and Coote, J.H. 1999. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol. Rev. 79:855-916. Toews, D.P., and Kirby, S. 1985. The ventilatory and acid-base physiology of the toad, Bufo marinus during exposure to environmental hyperoxia. Respir. Physiol. 59: 225-230. Thompson, G.G., and Withers, P.C. 2002. Aerial and aquatic respiration of the Australian desert goby, Chlamydogobius eremius. Comp. Biochem. Physiol. A. 131:871-879. Torgerson, C.S., Gdovin M.J., and Remmers, J.E. 1997. Ontogeny of central chemoreception during fictive gill and lung ventilation in an in vitro brainstem preparation of Rana catesbeiana. J. Exp. Biol. 200:2063-2072. Van Vliet, B.N., and West, N.H. 1986. Cardiovascular responses to electrical stimulation of the recurrent laryngeal nerve in conscious toads (Bufo marinus). J. Comp. Physiol. B. 156:363-375. Van Vliet, B.N., and West, N.H. 1992. Functional characteristics of arterial chemoreceptors in an amphibian (Bufo marinus). Respir. Physiol. 88:113-127. Vasilakos, K., Kimura, N. Wilson, R.J., and Remmers, J.E. 2006 Lung and buccal ventilation in the frog: uncoupling coupled oscillators. Physiol. Biochem. Zool. 79:1010-1018. Vulesevic, B., and Perry, S.F. 2006. Developmental plasticity of ventilatory control in zebrafish, Danio rerio. Respir. Physiol. Neurobiol. 154:396-405. Wang, T., Branco, L.G.S., and Glass, M.L. 1994. Ventilatory responses to hypoxia in the toad Bufo paracnemis before and after a decrease in haemoglobin oxygen-carrying capacity. J. Exp. Biol. 186:1-8.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  181 Noronha-de-Souza, C.R., Bicego, K.C., Michel, G., Glass, M.L., Branco, L.G., and Gargaglioni, L.H. 2006. Locus coeruleus is a central chemoreceptive site in toads. Am. J. Physiol. Regul. Integr. Comp. Physiol. 291:R997-R1006. Oliveira, R.D., Lopes, J.M., Sanches, J.R., Kalinin, A.L., Glass, M.L., and Rantin, F.T. 2004. Cardiorespiratory responses of the facultative air-breathing fish jeju, Hoplerythrinus unitaeniatus (Teleostei, Erythrinidae), exposed to graded ambient hypoxia. Comp. Biochem. Physiol. A. 139:479-485. Olson, K.R. 2002. Vascular anatomy of the fish gill. J. Exp. Zool. 293:214-231. Perry, S.F., and McKendry, J.E. 2001. The relative roles of external and internal CO2 versus H+ in eliciting the cardiorespiratory responses of Salmo salar and Squalus acanthias to hypercarbia. J. Exp. Biol. 204:3963-3971. Perry, S.F., McLean, H.A., Kogo, N., Kimura, N., Kawasaki, H., Sakurai, M., Kabotyanski, E.A., and Remmers, J.E. 1995. The frog brainstem preparation as a model for studying the central control of breathing in tetrapods. Braz. J. Med. Biol. Res. 28:1339-1346. Perry, S.F., Gilmour, K.M., Vulesevic, B., McNeill, B., Chew, S.F., and Ip, Y.K. 2005. Circulating catecholamines and cardiorespiratory responses in hypoxic lungfish (Protopterus dolloi): a comparison of aquatic and aerial hypoxia. Physiol. Biochem. Zool. 78:325-334. Pinder, A.W., and Burggren, W.W. 1986. Ventilation and partitioning of oxygen uptake in the frog Rana pipiens: effects of hypoxia and activity. J. Exp. Biol. 126:453-468. Pinder, A.W., Clemens, D., and Feder, M.E. 1991. Gas exchange in isolated perfused frog skin as a function of perfusion rate. Respir. Physiol. 85:1-14. Randall, D.J. 1970. Gas exchange in fishes. In: Fish Physiology. Vol. IV, W.S. Hoar and D.J. Randall (Eds.). Academic Press, New York. pp. 253-292. Randall, D.J., Burggren,W.W., Haswell, M.S., and Farrell, A.P. 1981. The Evolution of Air Breathing in Vertebrates. Cambridge University Press, Cambridge. Reid, S.G. 2006. Chemoreceptor and pulmonary stretch receptor interactions within amphibian respiratory control systems. Respir. Physiol. Neurobiol. 154:153-164. Reid, S.G., and Milsom, W.K. 1998. Respiratory pattern formation in the isolated bullfrog (Rana catesbeiana) brainstem-spinal cord. Respir. Physiol. 114:239-255. Reid, S.G., and West, N.H. 2004. Modulation of breathing by phasic pulmonary stretch receptor feedback in an amphibian, Bufo marinus. Respir. Physiol. Neurobiol. 142:165–183. Reid, S.G, Meier, J.T., and Milsom, W.K. 2000. The influence of descending inputs on breathing pattern formation in the isolated bullfrog brainstem-spinal cord. Respir. Physiol. 120:197-211. Remmers, J.E., Torgerson, C., Harris, M., Perry, S.F. , Vasilakos, K., and Wilson, R.J.A. 2001. Evolution of central respiratory chemoreception: a new twist on an old story. Respir. Physiol. 129: 211-217. Rocha, P.L., and Branco, L.G. 1998. Seasonal changes in the cardiovascular, respiratory and metabolic responses to temperature and hypoxia in the bullfrog Rana catesbeiana. J. Exp. Biol. 201:761-768. Sakakibara, Y. 1978. Localization of CO2 sensor related to the inhibition of the bullfrog respiration. Jpn. J. Physiol. 28:721-735. Sanchez, A.P., Hoffmann, A., Rantin, F.T., and Glass, M.L. 2001 Relationship between cerebrospinal fluid pH and pulmonary ventilation of the South American lungfish, Lepidosiren paradoxa (Fitz.). J. Exp. Zool. 290:421-425. Sanchez, A.P., Soncini, R., Wang, T., Koldkjaer, P., Taylor, E.W., and Glass, M.L. 2001. The differential cardio-respiratory responses to ambient hypoxia and systemic hypoxaemia in the South American lungfish, Lepidosiren paradoxa. Comp. Biochem. Physiol. A. 130:677-687.

182  |  Airway Chemoreceptors in the Vertebrates Sanders, C.E., and Milsom, W.K. 2001. The effects of tonic lung inflation on ventilation in the American bullfrog Rana catesbeiana Shaw. J. Exp. Biol. 204:2647–2656. Sheafor, E.A, Wood, S.C., and Tattersall, G.J. 2000. The effect of graded hypoxia on the metabolic rate and buccal activity of a lungless salamander (Desmognathus fuscus). J. Exp. Biol. 203:3785-3793. Shoemaker, V.H., Hillman, S.S., Hillyard, S.D., Jackson, D.C., McClanahan, L.L., Withers, P.C., and Wygoda, M.L.1992. Exchange of water, ions and respiratory gases in terrestrial amphibians. In: Environmental Physiology of Amphibians, M.E. Feder and W. W. Burggren (Eds.). University of Chicago Press, Chicago. pp. 125-150. Smatresk, N.J. 1990. Chemoreceptor modulation of endogenous respiratory rhythms in vertebrates. Am. J Physiol. 259:R887-R897. Smatresk, N.J., and Smits, W. 1991. Effects of central and peripheral chemoreceptor stimulation on ventilation in the marine toad, Bufo marinus. Respir. Physiol. 83:223–238. Smatresk, N.J., Burleson, M.L., and Azizi, S.Q. 1986. Chemoreflexive responses to hypoxia and NaCN in longnose gar: evidence for two chemoreceptor loci. Am. J. Physiol. 251: R116-R125. Smits, A.W., West, N.H., and Burggren, W.W. 1986. Pulmonary fluid balance following pulmocutaneous baroreceptor denervation in the toad. J. Appl. Physiol. 61:331-337 Smyth, D.H. 1939. The central and reflex control of respiration in the frog. J. Physiol. (London) 95:305-327. Steffensen, J.F., and Lomholt, J.P. 1983. Energetic cost of active branchial ventilation in the sharksucker, Echeneis naucrates. J. Exp. Biol. 103:185-192. Straus, C. 2000. Ontogeny of respiratory muscle control. Evidence from the amphibian model. Rev. Mal. Respir. 17:585-590. Straus, C., Wilson, R.J., and Remmers, J.E. 2001. Oxygen sensitive chemoreceptors in the first gill arch of the tadpole, Rana catesbeiana. Can. J. Physiol. Pharmacol. 79:959-962. Taylor, B.E., Harris, M.B., Leiter, J.C., and Gdovin, M.J. 2003. Ontogeny of central CO2 chemoreception: chemosensitivity in the ventral medulla of developing bullfrogs. Am. J. Physiol. 285:1461–1472. Taylor, E.W., Jordan, D., and Coote, J.H. 1999. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol. Rev. 79:855-916. Toews, D.P., and Kirby, S. 1985. The ventilatory and acid-base physiology of the toad, Bufo marinus during exposure to environmental hyperoxia. Respir. Physiol. 59: 225-230. Thompson, G.G., and Withers, P.C. 2002. Aerial and aquatic respiration of the Australian desert goby, Chlamydogobius eremius. Comp. Biochem. Physiol. A. 131:871-879. Torgerson, C.S., Gdovin M.J., and Remmers, J.E. 1997. Ontogeny of central chemoreception during fictive gill and lung ventilation in an in vitro brainstem preparation of Rana catesbeiana. J. Exp. Biol. 200:2063-2072. Van Vliet, B.N., and West, N.H. 1986. Cardiovascular responses to electrical stimulation of the recurrent laryngeal nerve in conscious toads (Bufo marinus). J. Comp. Physiol. B. 156:363-375. Van Vliet, B.N., and West, N.H. 1992. Functional characteristics of arterial chemoreceptors in an amphibian (Bufo marinus). Respir. Physiol. 88:113-127. Vasilakos, K., Kimura, N. Wilson, R.J., and Remmers, J.E. 2006 Lung and buccal ventilation in the frog: uncoupling coupled oscillators. Physiol. Biochem. Zool. 79:1010-1018. Vulesevic, B., and Perry, S.F. 2006. Developmental plasticity of ventilatory control in zebrafish, Danio rerio. Respir. Physiol. Neurobiol. 154:396-405. Wang, T., Branco, L.G.S., and Glass, M.L. 1994. Ventilatory responses to hypoxia in the toad Bufo paracnemis before and after a decrease in haemoglobin oxygen-carrying capacity. J. Exp. Biol. 186:1-8.

Ventilation Control in Amphibians and Air-Breathing Fishes  |  183 Wang, T., Taylor, E.W., Reid, S.G., and Milsom, W.K. 1999. Lung deflation stimulates fictive ventilation in decerebrate, paralyzed and unidirectionally ventilated toads (Bufo marinus). Respir. Physiol. 118:181–191. Wang, T., Taylor, E.W., Reid, S.G., and Milsom, W.K. 2004. Interactive effects of mechano- and chemo-receptor inputs on cardio-respiratory outputs in the toad. Respir. Physiol. Neurobiol. 140:63–76. Wassersug, R.J., and Yamashita, M. 2000. The mechanics of air-breathing in anuran larvae: implications to the development of amphibians in microgravity. Adv. Space. Res. 25:20072013. West, N.H., and Burggren, W.W. 1982. Gill and lung ventilation responses to steady-state aquatic hypoxia and hyperoxia in the bullfrog tadpole. Respir Physiol. 47:165-176. West, N.H., and Burggren, W.W. 1983. Reflex interactions between aerial and aquatic gas exchange organs in larval bullfrogs. Am. J. Physiol. 244:R770-R777. West, N.H., and Van Vliet, B.N. 1992. Sensory mechanisms regulating the cardiovascular respiratory systems. In: Environmental Physiology of the Amphibans, M.E. Feder and W.W. Burggren (Eds.). Academic Press: New York, pp. 151-205. West, N.H., Topor, Z.L., and Van Vliet, B.N.. 1987. Hypoxemic threshold for lung ventilation in the toad. Respir. Physiol. 70:377-390. West, N.H., Smits, A.W., and Burggren, W.W. 1989. Factors terminating nonventilatory periods in the turtle, Chelydra serpentina. Respir. Physiol. 77:337-350. Wilson, J.M. and Laurent, P. 2002. Fish gill morphology: inside out. J. Exp. Zool. 293:192-213. Wilson, R.J.A., Harris, M.B., Remmers, J.E., and Perry, S.F.. 2000. Evolution of air-breathing and central CO2/H+ respiratory chemosensitivity: new insights from an old fish? J. Exp. Biol. 203:3505-3512. Wilson, R.J., Vasilakos, K., Harris, M.B., Straus, C., and Remmers, J.E. 2002. Evidence that ventilatory rhythmogenesis in the frog involves two distinct neuronal oscillators. J. Physiol. 540:557-570. Winmill, R.E., Chen, A.K., and Hedrick, M.S. 2005. Development of the respiratory response to hypoxia in the isolated brainstem of the bullfrog Rana catesbeiana. J. Exp. Biol. 208:213– 222. Zaccone, G., Goniakowska-Witalinska, L. , Lauweryns,J.M., Fasulo, S., and Tagliafierro, G. 1989. Fine structure and serotonin immunohistochemistry of the neuroendocrine cells in the lungs of the bichirs Polypterus delhezi and P. ornatipinnis. Basic Appl. Histochem. 33:277-287 Zaccone, G., Fasulo, S., and Ainis, L. 1995. Neuroendocrine epithelial cell system in respiratory organs of air-breathing and teleost fishes. Int. Rev. Cytol. 157:277-314. Zaccone, G., Mauceri, A., Maisano, M., Giannetto, A., Parrino, V., and Fasulo, S. 2007. Innervation and neurotransmitter localization in the lung of the Nile bichir Polypterus bichir bichir. The Anatomical Record. 290: 1166-1177.

Neuroepithelial Bodies(NEB’s) in the Lung of Reptiles: Structural Morphology, Immunohistochemistry and Function 8. Neuroendocrine System of the Reptilian Respiratory Tract Luis Miguel Pastor Garcia, Giacomo Zaccone and Esther Beltrán-Frutos 9. Airway Receptors in Birds M. Fabiana Kubke, Roderick A. Suthers and J. Martin Wild 10. Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors Steven C. Hempleman and Jason Q. Pilarski

185-198

199-212

213-234

8 Neuroendocrine System of the Reptilian Respiratory Tract Luis Miguel Pastor García1*, Giacomo Zaccone2 and Esther Beltrán-Frutos1

Abstract The tracheal epithelium of lizards and snakes differs with respect to the equivalent epithelium of tortoises. While the first two orders have a prismatic pseudostratified epithelium with ciliated, basal and secretory cells or flat pseudostratified epithelium made up of basal and secretory cells, the epithelium in tortoises is the same in both the cartilaginous and membranous tracheal areas, with a uniform distribution of ciliated cells and goblet cells. The morphology of the cells that compose the bronchial-type epithelium in Chelonia, as well as in Ophidia and Squamata, is very similar to that of the trachea. In the extrapulmonary airways of lizards and snakes, secretory cells do not show the same morphology as the mucous cells of turtles and are classified as a serous type. In the trachea, solitary neuroendocrine cells can be observed by silver and immunohistochemical techniques and electron microscopy in tortoises, lizards and snakes. They have also been observed in Chelonia, where they present serotonin. Under the electron microscope, neuroendocrine cells of Chelonia, Lacertilia and Ophidia show granules of small diameter (90-120 nm approx.). Neuroendocrine cells, alone or forming groups, the latter denominated neuroepithelial bodies, are found in the lung trabeculae. In most species these cells show immunoreactivity to serotonin and calcitonin. Other regulatory peptides including calcitonin gene-related peptide, Leu-enkephalin, endothelin-1 and -2 are found depending on the species studied. Ultrastructurally both solitary neuroendocrine cells and those that form

Department of Cellular Biology and Histology, Medical School, University of Murcia, Campus de Espinardo, 30100, Murcia, Spain. 2 Department of Animal Biology and Marine Ecology, Faculty of Science, University of Messina, Salita Sperone 31, Messina, S. Agata I-98 166, Italy. *Author for correspondence: E-mail: [email protected], Phone: (34)-968-363949, FAX: (34)-968364150 1

186  |  Airway Chemoreceptors in the Vertebrates neuroepithelial bodies show low electron density and present dense-core granules in the basal region. The diameter of the granules, which have been described ultrastructurally in a variety of reptiles, varies between 120 and 140 nm. The neuroepithelial bodies are preferentially located in the primary trabeculae where clusters, formed by a varying number of granule-containing cells, are surrounded by flattened secretory or ciliated cells. Abundant peptidergic innervation is observed in the lung, along with vascularization and ganglionic cells. In conclusion, the neuroendocrine system in the respiratory system of reptiles shows a high degree of morphological organization, although more studies in more species are needed in order to demonstrate the differences that exist between the reptile families.

Keywords: trachea, lung, epithelium, neuroendocrine cells, nerves, reptiles.

1. Introduction: Epithelium of Extrapulmonary Airways1 The neuroendocrine cells in the respiratory tract of reptiles are found in extrapulmonary and intrapulmonary airways. The studies of these cells have a morphological character. Before we consider these a summary of histological data from reptiles airways is shown below. Although few species have been studied (Tesík, 1984; Pastor, 1987a; Pastor et al., 1988; Pastor, 1990; Beorlegui and Sesma, 1993; Pastor et al., 1997), it is probably safe to say that the larynx, trachea, and primary bronchi of all reptiles are made up of a pseudostratified epithelium, a lamina propria and an adventitia. The lamina propria usually shows few cells, with mainly collagen fibres, capillaries and nerves. In Lacertilia, Ophidia and Amphisabaena the epithelium differs depending on whether it is coating the cartilaginous portion or the membranous portion (between cartilages) of the trachea. This second area shows a prismatic pseudostratified epithelium with ciliated, basal and secretory cells, whereas in the cartilaginous portion the epithelium is flat pseudostratified and made up of basal and secretory cells (Pastor et al., 1988; Pastor, 1997; Pastor, 1990; Beorlegui and Sesma, 1993; Castells et al., 1990). In Chelonia, the epithelium is the same in both the cartilaginous and membranous areas, with a uniform distribution of ciliated cells and secretory cells (goblet cells) (Pastor et al., 1987a 1988). From a histochemical point of view, the mucous cells of the Chelonia show sialosulphomucins, whereas neutral mucosubstances predominate in Lacertilia (Castells et al. 1990a). By using lectins, differences can also be observed between the secretory cells of Chelonia and those of Lacertilia and Ophidia, since the first are labelled with

1

Part of this text is extracted with editorial permission from Histology, Ultrastructure and Immunohistochemistry of the Respiratory Organs in Non-Mammalian Vertebrates, L.M. Pastor (Ed.), Ediciones Secretariado de Publicaciones de la Universidad de Murcia, Murcia, 1995.

Neuroendocrine System of the Reptilian Respiratory Tract  |  187

SBA and DBA WGA, and the second with Con-A, WGA and LTA. After digestion with neuraminidase, PNA showed an increased affinity for goblet and non-goblet cells in all reptiles (Castells et al., 1990a). Scanning electron microscopy shows two clearly differentiated areas. The cartilagenous portions show flattened non-ciliated cells with marked cell margins and microvilli. In the membranous portion, patches of ciliated cells are observed together with nonciliated cells that show microvilli and occasionally rounded apical surfaces (Pastor et al., 1988, Pastor, 1990). The secretory cells of Lacertilia, Ophidia (Pastor et al., 1988; Pastor, 1990) and Amphisbaena (Pastor et al., 1997) have a similar morphology to that of the serous cells observed in mammals in the subepithelial glands of the respiratory apparatus, as well as the trachea of some species. These cells show a well-developed endoplasmic reticulum and secretory granules with a clearly defined limiting membrane and have no tendency to coalesce. The granules are usually electron dense or have a large electron-dense centre, although on occasion they have been seen to be electron-lucent with a granular-looking matrix. Some of them may be of myelinic appearance or empty. The granules with myelinic appearance are very similar to the lamellar bodies of type II pneumocytes of the lung, which suggests that there is a high concentration of phospholipids in these granules. The majority of the granules are situated apically in the cell, making it protrude into the lumen of the airway (Pastor et al., 1988; Pastor, 1990; Beorlegui and Sesma, 1993). The secretory cells of Chelonia are similar to the goblet cells of mammals. Their granules are variably electron dense, electron-lucent ones being abundant, and they tend to fuse (Pastor et al., 1987a, 1988). Along with the previously described cells, other cells, related to the immune system of these species, such as lymphocytes, macrophages, plasma cells and leukocytes, also appear in the epithelium.

Neuroendocrine System of Extrapulmonary Airways Together with these cell types, the epithelium also presents neuroendocrine cells that can be identified by silver and immunocytochemical techniques, having also been observed in Chelonia, presenting 5-HT (serotonin) (Pastor et al., 1987a, 1988). Neuroendocrine cells immunoreactive to PHI (peptide histidine isoleucine amide), PYY (peptide YY) and L-ENK (Leu-enkephalin) (Beorlegui et al., 1994a) are also found in Lacertilia (Podarcis hispanica). Under electron microscope, the neuroendocrine cells show granules of small diameter (90-120 nm approx.) in Chelonia, Lacertilia and Ophidia (Tesík, 1984, Pastor et al., 1987a; Pastor, 1990) (Figures 1, 2, 3). In the latter, groups of neuroendocrine cells have been described, similar to the neuroepithelial bodies (NEBs) found in the lung, but without innervation (Pastor, 1990) (Figure 2). Both in Lacertilia and in Chelonia, intraepithelial nerves have been described that show cholinergic-like agranular vesicles and peptidergic-like large-diameter vesicles (Pastor et al., 1988).

188  |  Airway Chemoreceptors in the Vertebrates

B

Figure 1: Tracheal epithelium of Pseudemys scripta elegans. Basal (B) and neuroendocrine cells (asterisk) are present. X 3,400.

S

Figure 2: Neuroendocrine cells (asterisk) in the tracheal epithelium of Natrix maura showing an electron-lucent cytoplasm. Ciliated and secretory columnar cells are observed. The secretory cells show diverse types of secretory granules. X 3,500.

Intrapulmonary Airways Epithelium Two surface epithelia can be clearly differentiated in reptile lungs. On the one hand, that which usually covers the primary trabeculae, bronchi and intrapulmonary trachea

Neuroendocrine System of the Reptilian Respiratory Tract  |  189

Figure 3: Neuroendocrine cell in the trachea of Pseudemys scripta elegans. The endocrine granules show a thin halo surrounding an electron-dense core (arrows). X 18,000.

if they exist (bronchial-type epithelium), and on the other, that which is related to gasexchange and situated on the faveoli or ediculae (gas-exchange epithelium). In all species studied, the first usually has a similar structure to that of the trachea. In chelonians, it is formed by basal, ciliated, mucous and neuroendocrine cells, the latter being solitary or forming groups (NEB). In Chelonia midas, mucous intraepithelial glands are also observed (Solomon and Purton, 1984). In Amphisbaenidae, and in some Lacertidae, the primary trabeculae principally present a gas-exchange epithelium on their surface, with only a few areas of ciliated cells, neuroendocrine cells (solitary or forming NEB) and secretory cells. Some groups of isolated ciliated cells have also been found in the edicular wall of Rhacodactylus leachianus (Perry et al., 1989). Other Lacertidae mainly show an epithelium with ciliated and secretory cells in their primary trabeculae. In the Ophidiae studied, the intrapulmonary trachea shows a similar epithelium to that of the trachea. The primary trabeculae have an epithelium with secretory, ciliated, neuroendocrine and basal cells. Finally, there are some Ophidiae species that present a simple squamous epithelium in the saccular portion of the lung (Luchtel and Kardong, 1981). The ciliated cells present a structure similar to that of mammals, and the mucous cells secrete sialosulphomucins and show an affinity for the lectins Con-A, WGA, DBA, PNA and SBA in Testudo graeca, and for DBA, SBA and WGA in Mauremys caspica (Pastor et al., 1989; Castells et al., 1990b). The secretory cells of the lizard Lacerta lepida have few acid groups and are reactive to Con-A, DBA and WGA. The snakes Elaphe scalaris and Natrix maura show secretory cells that contain sialosulphomucins and label with Con-A and WGA. The distribution of carbohydrates in the respiratory tract of the turtle Caretta caretta during embryogenesis and post-natal development has also been studied by fluorescein-labelled lectins (Sharma and Schumacher, 1992). In stage 30 embryos there is a strong positivity to the lectin WGA on the mucous

190  |  Airway Chemoreceptors in the Vertebrates surface, and in stage 25 and 26 embryos the surface mucus of the bronchi is reactive with SBA and VVA. The epithelium of intrapulmonary airways has been ultrastructurally studied in some ophidians, lizards and turtles. In the latter, this epithelium is made up of mucous cells, basal cells, ciliated cells and solitary neuroendocrine cells or neuroendocrine cells forming NEB (Pastor et al., 1989; Solomon and Purton, 1984). In Ophidia and Lacertilia, mainly secretory and ciliated cells have been described (Klemm et al., 1979; Perry et al., 1989; Luchtel and Kardong, 1981; Maina, 1989; Pastor, 1995), although, in the latter, neuroendocrine cells forming NEBs have also been demonstrated (Van Den Steen et al., 1994). In the snake Natrix maura, solitary and NEB neuroendocrine cells, ciliated and basal cells are found in this epithelium (Pastor, 1995). The morphology of the cells that make up the bronchial-type epithelium in Chelonia, as well as in Ophidia and Squamata, is very similar to that of the trachea. As occurs in the extrapulmonary airways of lizards and snakes, the secretory cell is not the same in morphology as the mucous cells of turtles and is classified as a serous type (Klemm et al., 1979; Perry et al., 1989; Luchtel and Kardong, 1981). Ultrastructurally, the latter present granules with an irregular membrane and a variably electron-dense content, whereas the former present granules with a more regular membrane and a content that is usually more electron dense, occasionally having lighter cores or empty granules. The basal cells show abundant filaments in the cytoplasm and are scarce in secondary or tertiary trabeculae (Pastor, 1995).

Neuroendocrine System of Intrapulmonary Airways The neuroendocrine cells are solitary or form NEBs and present a clear cytoplasm with toluidine-blue stain. The first have been described in the primary and secondary trabeculae of Testudo graeca, Mauremys caspica, and Pseudemys scripta elegans (Figure 4, 5) (Pastor et al., 1987b) and Lacerta muralis (Ravazzola et al., 1981), although in Testudo

Figure 4: Argyrophilia cells (arrows) are found of the primary bronchus of the turtle Mauremys caspica. X 260.

Neuroendocrine System of the Reptilian Respiratory Tract  |  191

graeca they can be observed in the gas-exchange epithelium, and in Pseudemys scripta elegans they are also seen in the tertiary trabeculae. These cells show immunoreactivity to 5-HT (Pastor et al., 1987b), except in Lacerta muralis, where only calcitonin has been observed (Ravazzola et al., 1981). Neuroepithelial bodies, mainly located in the respiratory epithelium covering the primary septa, are found in Testudo graeca, Pseudemys scripta elegans, Mauremys caspica and Lacerta lepida. They are immunoreactive to 5-HT and also present argyrophilia (Pastor et al., 1987b) (Figure 6). In Lacerta muralis, NEBs immunoreactive to calcitonin have also been described (Ravazzola et al., 1981). The neuroendocrine cells

Figure 5: Serotonin-inmunoreactive cell (arrow) in the primary septa of the lung of Testudo graeca. X 400.

Figure 6: Serotonin-inmunoreactive cells (arrow) grouped as a NEB in a primary septum of the lung of Testudo graeca. X 400.

192  |  Airway Chemoreceptors in the Vertebrates of four species have been studied immunocytochemically in more detail. In the first, the Cheloniae Pseudemys scripta elegans, immunoreactivity to 5-HT and CGRP (calcitonin gene-related peptide) has been identified in NEB, and in isolated neuroendocrine cells located in the ciliated epithelium of the intrapulmonary bronchus, the primary, secondary and tertiary trabeculae (Adriaensen et al., 1991). Moreover, enkephalin-like (ENK-like) peptide is found in NEBs between the ciliated cells of the trabeculae and in consecutive sections, the reactivity to CGRP is colocalized with that of 5-HT in the NEB. That is to say that all the NEBs positive to ENK-like peptide also contain CGRPlike peptide, whereas only part of the second have ENK-like peptide (Adriaensen et al., 1991). The second species is Basiliscus vittatus, of the Iguanidae family and the Squamata order. Here, the NEBs are immunoreactive to 5-HT and calcitonin, but not to CGRP, PGP, 9.5 (Protein Gene Product 9.5) and Leucine-7 (Van Den Steen et al., 1994). The third species is Podarcis hispanica, a lizard. Immunoreactivity to serotonin, calcitonin, CGRP, PHI, and ENK-L is found in pulmonary neuroendocrine cells (Beorlegui et al., 1994a). Immunoreactivity to serotonin, calcitonin and ENK-L is found in NEB and calcitonin and ENK-L in solitary cells of the gas-exchange epithelium of Blanus cinereus lung (Pastor et al., 1997) (Figure 7). In Podarcis hispanica, the coexistence of immunoreactivities to 5HT, calcitonin and CGRP is found in some cells, while in other cells only 5HT/calcitonin or CGRP immunoreactivities are found (Beorlegui et al., 1994b). Ultrastructurally the solitary neuroendocrine cells (Figure 8), or those that form NEB, show low electron density and present dense-core granules in the basal region.

Figure 7: A Nomarski’s image at higher magnification of a NEB (asterisk) positive for leu-5-enkephalin. Blanus cinereus. X 1,150.

Neuroendocrine System of the Reptilian Respiratory Tract  |  193

S

E Figure 8: One neuroendocrine cell (E) within epithelium of a primary trabecle in the lung of Natrix maura. Secretory cells (S) are observed overlying the neuroendocrine cell. X 6,000.

Figure 9: Scanning electron micrograph of a primary trabecle of the lung of Blanus cinereus. Ciliated cells in patches are found (arrow) and a presumed NEB is covered by secretory cells (asterisk). The rest (arrowhead) represents gas-exchange epithelium. X 1,100.

194  |  Airway Chemoreceptors in the Vertebrates The diameter of the granules varies between 120 and 140 nm. The NEBs have been described ultrastructurally in turtles Pseudemys scripta elegans and Testudo graeca, in lizard Basiliscus vittatus and in the worm lizard, Blanus cinereus (Pastor et al., 1989; Van Den Steen et al., 1994; Pastor 1995). The NEBs are preferentially localized in the primary trabeculae (Figure 9), on occasion at the connection between two septa, and they present an ovoid or round morphology in oblique or tangential sections. Clusters formed by a varying number of granule-containing cells are surrounded by flattened pneumocytes or secretory cells (Pastor, 1995; Van Den Steen et al., 1994; Scheuermann et al., 1983) (Figure 10) or ciliated and mucous cells (Pastor et al., 1989). The morphology of the neuroendocrine cells of the NEB is similar to that described for solitary neuroendocrine cells. Nerve endings in the NEB of reptiles are observed in Pseudemys scripta elegans and Testudo graeca between neuroendocrine cells. These nerves are probably cholinergic terminals (Scheuermann et al., 1983). Adrenergic terminals have been described by Scheuermann et al. (1983) in Pseudemys scripta elegans. In Basiliscus vittatus, unmyelinated axons have also been described in close vicinity to the NEB. From these, epithelial cell extensions pass through the basal membrane surrounding the nerves and forming basket-like structures (Van Den Steen et al., 1994). On occasions some intraepithelial nerves are observed in the basal portion

NEB

10 Figure 10: NEB in primary trabecle of the lung of Blanus cinereus. Secretory cells (arrow) cover the NEB. X 12,500.

Neuroendocrine System of the Reptilian Respiratory Tract  |  195

of NEB (Pastor et al, 1989). These nerves are mainly of a cholinergic nature, although some adrenergic or peptidergic ones also exist. Smooth muscle bands, extrinsic pulmonary arteries and arterioles are sparsely innervated by adrenergic fibres. A dense distribution of vasoactive intestinal polypeptide immunoreactive axons is formed on the pulmonary arteries and in the large veins of the snake Acrochordus granulatus (Donald and Lillywhite, 1989). Somatostatin immunoreactive axons are not observed. The perivascular plexus of VIP (vasoactive intestinal peptide) may represent part or all of the vagal postganglionic innervation of the pulmonary vasculature. Intramural ganglia are also demonstrated in the intraparenchymal connective tissue of the interstitium, being more abundant in the posterior portion of the lung in the species Chelodina longicollis (Smith and MacIntyre, 1987). This ganglion is composed of granule-containing cells that show argent affinity, and a bright blue-white formaldehyde-induced fluorescence, a sign of dopamine. With immunocytochemistry the presence of 5-HT is occasionally detected. The ganglion cells present synaptic contacts with cholinergic nerves. VIP-positive ganglion cells have also been observed in snakes (Donald and Lillywhite, 1989). The evidence suggests that granule-containing cells are neuroendocrine cells with a similar structure and function to the chief cells of the carotid body (Donald and Lillywhite, 1989). Numerous nerve fibres positive to NSE (neuron-specific enolase), PGP9.5, chromogranin, tyrosine hydroxylase, CGRP, calcitonin, bombesin, NPY (neuropeptide Y), substance P, VIP, and PYY, have been found in the lung interstitium of Podarcis hispanica. In perivascular innervation from this species, immunoreactivities in blood vessels are found. Also, in Podarcis hispanica, neurons positive to NSE and PGP9.5 are found in the lung interstitium (Beorlegui et al., 1994a).

Discussion In summary, the extrapulmonary airway epithelium of reptiles shows two clear structures, one similar to that of mammals, as is the case of Chelonia, although without the presence of a gland or special cell types, and a second specific to Lacertilia, Ophidia and Amphisbaena, in which the secretory cells are probably of a serous nature with phospholipid production. These cells are similar to the secretory cells observed in the frog and it is possible that they elaborate a substance whose composition is between a serous-type protein secretion and one rich in phospholipids similar to the lung surfactant. As in mammals the neuroendocrine system shows solitary neuroendocrine cells in extrapulmonary airways and solitary cells and NEBs in lungs. In the mammalian trachea, the presence of neuroendocrine cells is well recognized (Ericson et al., 1972; Cutz et al., 1975), being diffusely distributed throughout the trachea and showing morphological similarities to the neuroendocrine cells of pulmonary NEBs (Sonstegard et al., 1976). These cells apparently contain 5-HT (Dey et al., 1981) and secretory

196  |  Airway Chemoreceptors in the Vertebrates granules of different size and shape, depending on the species (Cutz et al., 1975). They are similar to those found in the NEB of the lung of Pseudemys scripta elegans (Sheuermann et al., 1983; Pastor et al., 1989). The presence of 5-HT suggests that the tracheal neuroendocrine cells of reptiles are similar to those of higher vertebrates (Dey et al., 1981). No clusters have been described in the extrapulmonary airways of the adult vertebrate. The result observed in Natrix maura is unexpected and contradicts the idea that the NEBs are located only in intrapulmonary airways (DiAgustine and Sonstegard, 1984). The presence of 5-HT in neuroendocrine cells of reptile lung is commonly seen; also, most of the species studied showed the presence of calcitonin. These substances are found in the Urodela (Adriaensen et al., 1994) respiratory system but are not observed in Anura. The peptide bombesine located in neuroendocrine cells of amphibians (Adriaensen et al., 1994) is not found in reptiles. The ultrastructure of reptile NEBs is similar to that of amphibians. In the latter, the NEBs are frequently covered by secretory cells or ciliated cells and the NEB does not contact the lumen (Rogers and Haller, 1978; Wasano and Yamamoto, 1978; Goniakowska-Witalinska et al., 1992). In reptiles also, the NEBs are lined with both mucous and ciliated cells, and only occasionally contact with the lumen has been observed (Sheuermann et al., 1983). Cells in NEBs are always covered by at least a thin cytoplasmic extension of a neighbouring cell, indicating that luminal contact is not required. Stronger still, it appears that contact to the airspace is avoided in some lower vertebrates (Van Den Steen et al., 1994). Ultrastructural evidence also exists for a basketlike innervation of NEB in some reptiles. This way of innervation possibly represents an evolutionarily different concept for interaction between NEB corpuscular cells and nerve fibres. This organization possibly represents a mechanoreceptor function (Van Den Steen et al., 1994). In conclusion, the neuroendocrine system of reptiles shows a complex organization, but further studies are necessary to determine similarities and differences between the large number of families and species.

References Adriaensen, D., Scheuermann, D.W., Timmermans, J.P., and De Groodt-Lasseel, M.H.A. 1991. Calcitonin gene-related peptide, enkephalin and serotonin coexist in neuroepithelial bodies of the respiratory tract of the red-eared turtle, Pseudemys scripta elegans. An immunocytochemical study. Histochemistry, 95: 567-572. Adriaensen, D., Scheuermann, D.W., Timmermans, J.P., Gomi, T., Mayer, B., and De GroodtLasseel, M.H.A. 1994. Neuroepithelial endocrine and nervous system in the respiratory tract of Cynops pyrrhogaster with special reference to the distribution of nitric oxide synthase and serotonin. Microsc. Res. Tech., 29: 79-89. Beorlegui, C., and Sesma, P. 1993. Estudio histológico del epitelio traqueal de la lagartija Podarcis hispanica (Steindacher, 1870). In: Actas VIII Congreso Nacional de Histología. M.A. Peinado and J.A. Pedrosa (Eds). Gráficas Catena, Jaén, pp. 189-190.

Neuroendocrine System of the Reptilian Respiratory Tract  |  197 Beorlegui, C., Sesma, P., and Martínez, A. 1994a. An immunocytochemical study of the respiratory system of Podarcis hispanica (Reptilia). Gen. Comp. Endocrinol., 96: 327-338. Beorlegui, C., Sesma, P., and Martínez, A. 1994b. Colocalization of immuroneactivities to serotonin, calcitonin and CGRP in endocrine pulmonary cells of the iberian lizard Podarcis hispanica (Reptilia). Tissue Cell., 26: 817-825. Castells, M.T., Ballesta, J., Pastor, L.M., Madrid, J.F., and Marin, J.A. 1990a. Histochemical characterization of glycoconjugates in the epithelium of the extrapulmonary airways of several vertebrates. Histochemical J., 22: 24-35. Castells, M.T., Ballesta, J., Madrid, J.F., Hernandez, F., Martinez-Menarguez, J.A., and Aviles, M. 1990b. Histochemistry of glycoconjugates in the lung of several vertebrates. Zool. Jb. Anat., 120: 331-346. Cutz, E., Chan, W., Wong, V., and Conen, P.E. 1975. Ultrastructure and fluorescence histochemistry of endocrine (APUD-type) cells in tracheal mucosa of human and various animal species. Cell Tissue Res., 158(4): 425-437. Daniels, C.B., Barr, H.A., Power, J.H.T., and Nicholas, T.E. 1990. Body temperature alters the lipid composition of pulmonary surfactant in the lizard Ctenophorus nuchalis. Exp. Lung Res.,16: 435-449. Dey, R.D., Echt, R., and Dinerstein, R.J. 1981. Morphology, histochemistry and distribution of serotonin-containing cells in tracheal epithelium of adult rabbit. Anat. Rec., 199(1): 23-31. DiAugustine, R.P., and Sonstegard, K.S. 1984. Neuroendocrine like (small granule) epithelial cells of the lung. Environ. Health Perspect., 55: 271-295. Donald, J.A., and Lillywhite, H.V. 1989. Vasoactive intestinal polypeptide-immunoreactive nerves in the pulmonary vasculature of the aquatic file snake Acrochordus granulatus. Cell Tissue Res., 255: 585-588. Ericson, L.E., Hakanson, R., Larson, B., Owman, C., and Sundler, F. 1972. Fluorescence and electron microscopy of amine-storing enterochromaffin-like cells in tracheal epithelium of mouse. Z. Zellforsch. Mikrosk. Anat., 124(4): 532-545. Goniakowska-Witalinska, L., Lauweryns, J.M., Zaccone, G., Fasulo, S., and Tagliafierro, G. 1992. Ultrastructure and immunocytochemistry of the neuroepithelial bodies in the lung of the tiger salamander, Ambystoma tigrinum (Urodela, Amphibia). Anat. Rec., 234(3): 419431. Klemm, R.D., Gatz, R.N., Westfall, J.A., and Fedde, M.R. 1979. Microanatomy of the lung parenchyma of a tegu lizard Tupinambis nigropunctatus. J. Morphol., 161: 257-280. Luchtel, D.L., and Kardong, K.V. 1981. Ultrastructure of the lung of the Rattlesnake, Crotalus viridis oreganus. J. Morphol., 169: 29-47. Maina, J.N. 1989. The morphology of the lung of the black mamba Dendroaspis polylepis (Reptilia: Ophidia: Elapidae). A scanning and transmission electron microscopic study. J. Anat., 167: 31-46. Pastor, L.M. 1990. A morphological study of the tracheal epithelium of the snake Natrix maura. J. Anat., 172: 47-57. Pastor, L.M. 1995. The histology of reptilian lung. In: Histology, Ultrastructure and Immunohistochemistry of the Respiratory Organs in non-Mammalian Vertebrates. L.M. Pastor (Ed.). Secretariado de Publicaciones de la Universidad de Murcia, pp. 129-153. Pastor, L.M., Ballesta, J., Hernandez, F., Perez-Tomas, R., Zuasti, A., and Ferrer, C.A 1987a. Microscopic study of the tracheal epithelium of Testudo graeca and Pseudemys scripta elegans. J. Anat., 153: 171-183.

198  |  Airway Chemoreceptors in the Vertebrates Pastor, L.M., Ballesta, J., Perez-Tomas, R., Marin, J.A., Hernandez, F., and Madrid, J.F. 1987b. Immunocytochemical localization of serotonin in the reptilian lung. Cell Tissue Res., 248: 713-715. Pastor, L.M., Ballesta, J., Castells, M.T., Perez-Tomas, R., Madrid, J.F., and Marin, J.A. 1988. A light and electron microscopic study of the epithelium of the extrapulmonary airways of Mauremys caspica and Lacerta lepida (Reptilia). J. Submicrosc. Cytol. Pathol., 20: 25-36. Pastor, L.M., Ballesta, J., Castells, M.T., Perez-Tomas, R., Marin, J.A., and Madrid, J.F. 1989. A microscopic study of the lung of Testudo graeca. J. Anat., 164: 19-39. Pastor, L.M., Pallares, J., Horn, R., and Zaccone, G. 1997. Morfología Microscópica de las Vías Extrapulmonares y Pulmón de Blanus Cinereus (Reptilia). In: IX Congreso Nacional de la SEH. Universidad de Navarra, Navarra, p. 150. Perry, S.F., and Duncker, H.R. 1978. Lung architecture, volume and static mechanics in five species of lizards. Respir. Physiol., 34: 61-81. Perry, S.F., Bauer, A.M., Russell, A.P., Alston, J.T., and Maloney, J.E. 1989. Lungs of the Gecko Rhacodactylus leachianus (Reptilia: Gekkonidae): A correlative gross anatomical and light and electron microscopic study. J. Morphol., 199: 23-40. Ravazzola, M., Orci, L., Girgis, F., Galan Galan, F., and McIntyre, I. 1981. The lung is the major organ source of calcitonin in the lizard. Cell. Biol. Int. Rep., 5: 937-944. Rogers, D.C., and Haller, C.J. 1978. Innervation and cytochemistry of the neuroepithelial bodies in the ciliated epithelium of the toad lung (Bufo marinus). Cell Tissue Res., 195(3): 395-410. Scheuermann, D.W., De Groodt-Lassel, M.H.A., Stilman, C., and Meisters, M.L. 1983. A correlative light, fluorescence- and electron-microscopic study of the neuroepithelial bodies in the lung of the red-eared turtle Pseudemys scripta elegans. Cell Tissue Res., 234: 249-269. Sharma, R., and Schumacher, U. 1992. Histochemical characterization of carbohydrate residues during the morphogenesis of grastrointestinal and respiratory systems of Caretta caretta. Acta Histochem., 93: 411-432. Smith, D.G., and MacIntyre, D.H. 1987. Innervation of the lung of the Australian snake-necked tortoise Chelodina longicollis. Comp. Biochem. Physiol., 87C: 439-444. Solomon, S.E., and Purton, M. 1984. The respiratory epithelium of the lung in the green turtle (Chelonia mydas). J. Anat., 139: 353-370. Sonstegard, K.S., Cutz, E., and Wong, V. 1976. Dissociation of epithelial cells from rabbit trachea and small intestine with demonstration of APUD endocrine cells. Am. J. Anat., 147(3): 357-373. Tesík, I. 1984. The ultrastructure of the tracheal epithelium in European common lizard (Lacerta agilis L.) and in sand lizard (Lacerta vipira Jacq.). Anat. Anz., 155: 329-340. Van Den Steen, P., Van Lommel, A., and Lauweryns, J.M. 1994. Neuroepithelial bodies in the lung of Basiliscus vittatus (Reptilia, Iguanidae). Anat. Rec., 239: 158-169. Wasano, K., and Yamamoto, T. 1978. Monoamine-containing granulated cells in the frog lung. Cell Tissue Res., 193(2): 201-209.

9 Airway Receptors in Birds M. Fabiana Kubke1*, Roderick A. Suthers2 and J. Martin Wild1

Structure of the Airways in Birds The respiratory system of birds consists of the non-expandable lungs and a series of air sacs providing a unidirectional system of air flow (Maina, 2002). The avian lung is formed by a series of tubular structures, the parabronchi, through which air flows unidirectionally, and a series of air capillaries where gas exchange with the blood takes place (Dunker, 1971; Duncker, 1972; Duncker, 1974; Scheid, 1979). The air sacs of birds constitute thin walled chambers located largely in the thoracic and abdominal cavities and that communicate with the avian lung via the ostia (Duncker, 1972; Duncker, 1974; McLelland, 1989). The anterior group of air sacs consists of the anterior, cervical, interclavicular, and anterior thoracic air sacs, which are connected to the ventrobronchi. The posterior group of air sacs consists of the posterior thoracic and the abdominal air sacs, connecting to the bronchial system directly via the posterior primary bronchus or via the laterobronchus. Powered by respiratory muscles, air sacs act as bellows in the ventilatory mechanism. The wall of the air sac consists essentially of a poorly vascularized simple squamous luminal epithelium covering a thin layer of connective tissue underlain by peritoneal epithelium (Dunker, 1971; Cook et al., 1978; Bezuidenhout, 2005). Scattered in the epithelium are clusters of cuboidal or columnar ciliated cells as well as associated basal and goblet cells and columnar epithelial cells filled with secretory vesicles (Cook et al.,

Department of Anatomy with Radiology, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand. 2 Medical Sciences, Indiana University, Bloomington IN 47405, USA. *Author for correspondence: E-mail: [email protected], Phone: (+64) (9) 373-7599 Ext 86002, Fax: (+64) (9) 373-7484. 1

200  |  Airway Chemoreceptors in the Vertebrates 1986). The epithelium is described as columnar and ciliated at the ostia where abundant smooth muscle cells are also found (Cook et al., 1978). A small number of muscle cells have been observed in the air sac membranes of some species, probably extending from the smooth muscle layer present around the bronchi (Bennett and Malmorfs, 1970; Dunker, 1971; Groth, 1972). The control of ventilation in birds is still not completely understood. Both chemoreceptors and mechanoreceptors appear to play a role in feedback mechanisms to the respiratory system. A precise control of ventilatory mechanisms must exist, at least in some birds, since it is crucial for the precise temporal patterns of ventilation that underlie the production of complex vocal output (Hartley, 1990).

Innervation of the Lungs and Air Sacs The air sacs themselves are variously said to be either “devoid of nerves” (Fedde et al., 1963) or to be richly innervated (Groth, 1972; Cook et al., 1978). Such apparent contradictions may reflect differences in density of innervation in different regions of the air sacs, because those areas closer to the ostia are more densely innervated than more distal areas. Indeed, a study of the vagal innervation of the zebra finch air sac revealed two regions within the air sac, one with a high density nerve plexus and a second region that appeared almost devoid of innervation (Kubke et al., 2004). The innervation of the air sacs has been studied by histochemistry, immunocytochemistry and light microscopy. These studies revealed a cholinergic plexus, and an adrenergic plexus associated with blood vessels and smooth muscle cells (Bennett and Malmorfs, 1970). Synaptic and/or neuroendocrine vesicles are also seen as dense puncta along the dense nerve plexus in the air sacs and around blood vessels (Kubke et al., 2004). In the caudal thoracic sac of the chicken, for example, axons found in small bundles have the ultrastructural appearance of cholinergic, adrenergic or peptidergic axons. Putative afferent nerve endings have also been described in the air sac and bronchoperitoneal membranes (Bennett and Malmorfs, 1970; Groth, 1972), and vagal innervation has been demonstrated in the air sacs of zebra finches (Kubke et al., 2004). The smooth muscle cells of the ostia are also richly innervated by cholinergic fibres and fibres containing dense core vesicles (Cook et al., 1986). Adrenergic terminals have been described associated with the smooth muscle cells of the primary bronchi and the ostia and on the pulmonary aponerosis, but not in the muscular walls of the primary, secondary or tertiary bronchi (Bennett and Malmorfs, 1970). In birds, as in mammals, respiratory structures are innervated by the vagus nerve, and the activity of single vagal afferent fibres has been correlated with changes in respiratory activity (Fortin et al., 1994; reviews in Gleeson, 1987; Gleeson and Molony, 1989). Changes in the volume of the air sacs, without a change in PCO2, have been shown to affect ventilatory timing, and this effect is dependent on an intact vagus

Airway Receptors in Birds  |  201

(Ballam et al., 1982; Ballam et al., 1985; Gleeson and Molony, 1989). Vagal terminals are also found in the air sacs of songbirds (Kubke et al., 2004) where they innervate structures resembling the neuroepithelial bodies (NEBs) described in the airways and lungs of other vertebrates (Figure 1). The distribution of vagal terminals in the air sacs in songbirds is characterized by a non-homogeneous, dense network of fibres in the air sac membrane. In most cases, fibre bundles run along the air sac membrane, giving out branches along the way.

Figure 1: Structure of vagal terminations in the air sac of the zebra finch following injections of cholera toxin B into the vagus nerve.

Air Sac and Lung Receptors Airway receptors are classified on the basis of their sensitivities to a variety of stimuli and their axonal conduction velocity (Widdicombe, 2001). Two main types of receptors have been revealed in birds by physiological studies: an inspiratory-inhibitory CO2sensitive receptor located in the lungs (Peterson and Fedde, 1968; Molony, 1974), and a less well studied, slowly adapting mechanoreceptor having a vagal axon, a regular respiratory modulation, and an end-inspiratory peak discharge only when tidal volume is elevated above eupneic resting values (Molony, 1974; Gleeson and Molony, 1989). The latter type of receptor is unlikely to be located in the lung, if only because the lung, in the absence of a diaphragm, is largely non-expanding, especially during inspiration. It has been suggested, therefore, that these receptors are instead located in the walls of the air sacs or the membranes associated with them (Gleeson and Molony, 1989), such as the richly innervated saccopleural membrane between the wall of the thoracic air sac and the parietal pleura (Cook et al., 1978).

202  |  Airway Chemoreceptors in the Vertebrates

Chemoreceptors Sensitivity to PCO2 is mediated by intrapulmonary chemoreceptors (IPC), which are a homogeneous group of CO2 sensitive receptors located in the parabronchial gas exchange area of the avian lung (Scheid et al., 1974; Burger et al., 1976b; Nye and Burger, 1978; Nye et al., 1982). Sensitivity is also mediated by extrapulmonary afferents (EPA) consisting of central chemoreceptors and peripheral arterial chemoreceptors carried by the middle cardiac nerve and carotid body. In birds, EPAs are sensitive to changes in arterial PO2 and PCO2 (Bouverot and Leitner, 1972; Barnas et al., 1978; Nye and Powell, 1984) and EPAs interact with IPCs to produce changes in ventilation (Adamson et al., 1994). Although the carotid body is the primary organ that mediates the response to hypoxia in mammals, in birds both the central and peripheral arterial chemoreceptors, as well as the IPCs, are involved in this response (Menna and Mortola, 2003). IPCs are sensitive to changes in PCO2 but not PO2; they mediate a reflex response that inhibits ventilation upon a decrease of PCO2 (Burger et al., 1976a; Nye and Burger, 1978; Clanton et al., 1982; Fedde et al., 1982; Hempleman and Burger, 1985). The contribution of IPCs to the regulation of ventilation in birds can be demonstrated even in the absence of any influence from the carotid body (Adamson and Solomon, 1993) e.g., increased PCO2 results in increased ventilation in the duck (Powell et al., 1978; Scheid et al., 1978) even when arterial PCO2 remains unchanged (Scheid et al., 1978). In the vagotomized bird, sensitivity to PCO2 is not in the carotid receptors (Fedde et al., 1963) and temporally occluding the lung—which prevents chemical signals reaching extrapulmonary receptors—has the same effect on ventilation, suggesting that the IPCs are sufficient to mediate the compensatory ventilatory response (Peterson and Fedde, 1968). These receptors have been studied neurophysiologically and their discharge patterns are proportional to –logPCO2 at high PCO2 but are rectilinear at low PCO2 (Fedde and Peterson, 1970; Fedde et al., 1974; Osborne and Burger, 1974; Tschorn and Fedde, 1974; Burger et al., 1976a; Barnas et al., 1978; Bouverot, 1978; Nye and Burger, 1978; Powell et al., 1978; Barker et al., 1981). A reflex reduction in amplitude of ventilation occurs when firing increases in IPC afferents (Fedde and Peterson, 1970; Scheid et al., 1978) and stimulating vagal axons results in apnoea (Peterson and Nightingale, 1976).

Mechanoreceptors Several lines of evidence suggest, but do not prove, that the air sacs of birds contain mechanoreceptors which convey sensory information to the central nervous system via the vagus nerve. Changes in the volume of the air sacs have been shown to elicit a respiratory reflex via the vagus nerve. Ballam et al. (1985) suggested on physiological grounds that at least some of this vagal innervation might be supplied by slowly adapting mechanoreceptors such as those that Gleeson and Molony identified in the

Airway Receptors in Birds  |  203

vagus nerve as being sensitive to air sac expansion (Molony, 1974; Gleeson, 1987; Gleeson and Molony, 1989). Indeed, the demonstration of a compensatory reflexive decrease in the amplitude of expiratory muscle EMG as a result of a small injection of air into the air sacs during singing in northern cardinals could suggest that an air sac receptor effects a reflex modulation of the respiratory motor pattern associated with song (Suthers et al., 2002). The somatosensory receptors mediating this expiratory muscle response to changes in respiratory pressure have not specifically been identified in birds, although the presence of mechanoreceptors have been demonstrated in the air sac membrane (Ballam et al., 1985; Kubke et al., 2004). However, the receptors sensing air sac expansion need not be located within the air sacs themselves, but could instead be located in other tissues, such as the respiratory muscles themselves—despite the paucity of spindles in these muscles in non-songbirds (DeWet et al., 1971)—or in the membranes associated with the air sacs.

Neuroepithelial Bodies in the Air Sacs and Lungs Vagally innervated structures have been described in the air sacs of chickens, swans and zebra finch (Groth, 1972; Kubke et al., 2004). In particular, Groth (1972) described several types of what he interpreted as sensory endings in the air sacs, including small, knob-like structures (Endkolben) in chickens and clew-like and web-like nerve fibre complexes in swans. Although structures resembling those that Groth described in swans were not found in the air sacs of zebra finches, a small number of structures similar to the Endkolben that Groth found in chickens were identified in the region of the air sac that lies closer to the ostia, where the air sac is characterized by a dense reticular nerve mesh (Figure 1) (Kubke et al., 2004). These structures could best be described as a characteristic vagal terminal structure of rather varying morphology forming a complex reticular arrangement of delicate fibres delimiting and innervating a cluster of cells with a racemose appearance of variable size. The compact aggregate of cells in the racemes contain synaptic/neuroendocrine vesicles and 5HT, indicating that they are neurosecretory in nature (Cook et al., 1978; Kubke et al., 2004). In his silver stained material, Groth (1972) could not identify a close relation of the Endkolben to any kind of “helper-cell” and claimed all the sensory endings were of a non-encapsulated nature. However, the study by Kubke et al. (2004) shows (a) that the racemes (Groth’s Endkolben) are innervated by the vagus nerve, (b) that cells are characteristically associated with the vagal nerve terminals, (c) that these cells form a compact aggregate and contain neurosecretory vesicles, and (d) that the cells within the raceme are also positive for serotonin, a feature that is characteristic of neuroendocrine cells in other parts of the respiratory tract of birds (Salvi and Renda, 1992; Adriaensen et al., 1994). These results, and the fact that some of the cellular aggregates are found clearly in close association with blood vessels, together conspire to strongly suggest that the cells of the racemes are neuroendocrine in nature (Buckley and Kelly, 1985; Hou and Dahlstrom,

204  |  Airway Chemoreceptors in the Vertebrates 1996; Langley and Grant, 1999; Portela-Gomes et al., 2000). Indeed, the presence of synaptic vesicles within the vagus nerve terminal itself, suggests that at least some of these nerve terminals could be presynaptic in nature (Adriaensen and Scheuermann, 1993; Dememes et al., 2000; Widdicombe, 2001). The structures described by Kubke et al. (2004) are reminiscent of the NEB found in the airways of birds and other vertebrates (e.g., Lauweryns et al., 1972; Bower et al., 1978; Cook et al., 1986; McLelland and Macfarlane, 1986; McLelland, 1989), and the immunocytochemical staining patterns are consistent with this hypothesis. Apart from the punctuate staining associated with the nerve plexus and blood vessels, a synaptic vesicle antibody labelled only the racemes in the air sac (Kubke et al., 2004). Thus, the cells within the racemes may represent the vesicle-containing cells already described by Cook (1986). Neuroepithelial cells (NEC) rich in secretory granules have been described in the airways of all vertebrates studied, including birds, and can be seen as individual cells (NECs) or as aggregations forming NEBs (Cook et al., 1986; McLelland and Macfarlane, 1986; McLelland, 1989). NEBs are characterized by the aggregations of serotonin-containing neuroendocrine cells filled with neurosecretory vesicles and receiving vagal and spinal afferent innervation, as well as efferent inputs (Widdicombe, 2001; Adriaensen et al., 2003). Although the structures described in the air sac of the zebra finch appear to conform to the general structure of previously described NEBs, they cannot categorically be classified as actual NEBs, since some of the racemes in the air sacs appear to be larger than those previously described, and because it is possible that not all cells within the raceme make contact with the underlying epithelium (Kubke et al., 2004). However, the general morphology, the innervation and the staining patterns argue in favour of them representing NEB-like structures. NEBs and NECs have been shown to occur in different sizes and arrangements, and the structures described in the air sac of zebra finches may represent additional morphological variation of this ubiquitously distributed structure. The ostia of the abdominal air sac of the domestic fowl are also said to contain NECs (Cook et al., 1986). Cook et al. classified these cells into three types based on morphology and location. Type I NECs were found in the ganglia in the submucosal layer of the ostium, contained small granular vesicles and were innervated by nerve terminals containing small agranular vesicles as well as large dense-cored vesicles. Type II NECs, filled with large dense-cored vesicles, were found in clusters and associated with the capillary system receiving innervation of terminals containing primarily small agranular vesicles, although occasional terminals with large dense-core vesicles were also observed. Type III neuroendocrine cells consisted of columnar cells that were not apparently associated with the epithelium and did not appear to be innervated, and a second subpopulation with a more basal position which did receive innervation by terminals containing small clear vesicles. The ostia of the domestic fowl are found to contain NEBs (Cook et al., 1986)

Airway Receptors in Birds  |  205

The lung is also known to have a neuroepithelial endocrine system consisting of either NECs or NEBs (Cook and King, 1969; King et al., 1974; Walsh and McLelland, 1974a&b; Wasano and Yamamoto, 1979; Lopez et al., 1983; Cook et al., 1986; McLelland and Macfarlane, 1986; Adriaensen et al., 1994). In the quail, several subtypes could be found that expressed different types of chemicals (somatostatin, bombesin and serotonin) (Salvi and Renda, 1992; Adriaensen et al., 1994). These cells were found to be isolated or formed clusters of cells characterized by the expression of 5HT, by the presence of dense core vesicles, the receipt of afferent and efferent innervation on their basal pole (Cook and King, 1969; King et al., 1974; Wasano and Yamamoto, 1979; Lopez et al., 1983) and their location in the ciliated epithelium of primary and secondary bronchi. Intraepithelial axons innervating the NECs and NEBs have clear vesicles and asymmetric densities, suggesting efferent as well as afferent innervation (Adriaensen et al., 1994). NEBs have been proposed to perform dual afferent and efferent functions including a role as O2 receptors (Patel and Honore, 2001; Peers and Kemp, 2001; Widdicombe, 2001; Kemp et al., 2002; Kemp et al., 2003), and it is therefore possible that they have a similar function in the air sacs of birds, where they would be in an ideal location to sense the level of O2 in inspired air. NECs also are reminiscent of Merkel cells and, in particular, Grandry corpuscles found in the epidermis of birds. In this case, these cells, which are also of neural crest origin, have been proposed to play a mechanoreceptive role (Gottschaldt, 1985; Halata et al., 2003). NEBs in particular, are also believed to exhibit a mechanoreceptive function (van den Steen et al., 1994). Whether the structures described in the air sac of zebra finches share chemoreceptive and/or mechanoreceptive functions remains to be established. Regardless of the receptive modality of the cells, the possible dual pre- and postsynaptic nature of their innervation may implicate these structures in regulating feedback mechanisms involved with the control of ventilation.

Innervation of Neuroepithelial Bodies Injections of 3H-leucine into the nodose ganglion revealed vagal terminations on NEBlike structures in the ostia of chickens, demonstrating their afferent innervation (Bower et al., 1978). However, many afferent fibres have been shown to also play a presynaptic role (Adriaensen and Scheuermann, 1993; Dememes et al., 2000; Widdicombe, 2001). In general, the innervation of NEBs has been proposed to perform a dual preand postsynaptic function (McLelland, 1989). Indeed, the presence in the NEBs of zebra finches of synaptic vesicle immunoreactivity in the presynaptic terminal, as well as in the neurosecretory cells themselves, is consistent with such a proposal (Kubke et al., 2004). However, these NEB-like structures could receive a separate efferent innervation. The origin of these efferents has thus far remained unclear (Adriaensen et al., 2003) but one candidate might be the retrogradely labelled neurons identified in the ventrolateral medulla of songbirds following injection of cholera toxin B-chain

206  |  Airway Chemoreceptors in the Vertebrates (CTB) into the vagus nerve or into the air sacs. It is pointed out, however, that the specific target(s) of these particular vagal efferent neurons have not been identified by anterograde tracing methods. Regardless of the possible vagal efferent innervation of the air sacs, the functional nature of the afferent vagal innervation remains enigmatic. The physiological evidence of a reduction in the amplitude of the EMG from abdominal expiratory muscles following injection of air into the air sacs during singing (Suthers et al., 2002) would seem to suggest the presence of a mechanoreceptor, possibly located in the air sacs themselves, that is capable of mediating changes in volume or pressure. However, the NEB-like structure of the receptors described in the air sacs of zebra finches is consistent with their having a chemoreceptor/neuroendocrine, as well as, or instead of, a mechanoreceptor role (Kubke et al., 2004). The presynaptic component could mediate local neuroendocrine functions - as suggested by its close association with blood vessels—while the postsynaptic component, via its input to the nucleus tractus solitarius (nTS), could initiate centrally mediated reflexes. The efferent component of such a reflex might well involve vagal neurons in the ventrolateral medulla (see below). Indeed, Groth (1972) recognized that the air sacs, by constituting the major component of the ventilatory system, might be the origin of breathing reflexes mediated by mechanoreceptors, while Cook et al. (1986) suggested that such feedback mechanisms could mediate a sphincter-like response in the smooth muscles of the ostia.

Relationship with CNS Structures Respiration in birds may be the result of a central pattern generator, which is likely to be modulated by vagal input, as it is in mammals (Fortin et al., 1994). Vagal branches of the nodose ganglion carry sensory information about changes in PO2 from chemoreceptors around the common carotid artery that may contribute to resting ventilatory drive, while other vagal branches innervate IPCs that supply the CNS with information about PCO2. Application of CTB to the vagus nerve or injected into the air sacs of zebra finches and cowbirds reveal several populations of labelled sensory and motor structures in the medulla (Wild, 2004). Nucleus tractus solitarius (nTS), which receives the great majority of vagal afferents, is consistently labelled by both these procedures, particularly densely around the solitary tract where respiratory inputs have been shown to terminate in other species (Katz and Karten, 1983; Fortin et al., 1994). Two populations of vagal efferent neurons are also labelled, one located dorsomedially, which is the parasympathetic dorsal motor nucleus of the vagus (DMNX), and another that forms an arc around the ventrolateral periphery of nucleus retroambigualis (RAm) in the caudal medulla. This nucleus projects upon both respiratory (expiratory) and vocal motoneurons (Wild, 1993a; Sturdy et al., 2003; Kubke et al., 2005) and it also

Airway Receptors in Birds  |  207

receives major descending inputs from the song control nucleus robustus arcopallialis (RA) in the telencephalon and from the dorsomedial intercollicular nucleus (DM) in the midbrain (Wild, 1993a; Wild, 1993b; Reinke and Wild, 1997; Wild et al., 1997; Reinke and Wild, 1998; Wild et al., 2000; Sturdy et al., 2003). As in mammals, RAm is, therefore, a true nexus in the respiratory-vocal control system (Holstege, 1989; Wild, 2004). In mammals injections of CTB into the airway lumen (Perez Fontan and Velloff, 2001) result in the labelling of neurons near and within nucleus ambiguus and the authors suggested that some of these vagal motoneurons may provide direct innervation of epithelial or vascular effector organs in the airway mucosa. In addition, in the rat, cells in nTS, which receive vagal afferent input from slowly and rapidly adapting pulmonary mechanoreceptors, project to neuronal somata in the vicinity of nucleus ambiguus, and it has been proposed that these neurons may in fact be located in nucleus retroambigualis (Otake et al., 2001; Ezure et al., 2002). Similar connections between nTS and RAm in birds (Wild et al., 2009) could constitute reflex circuitry ensuring that vagally mediated feedback from respiratory structures reaches vagal efferent neurons. The specific function of these neurons is not known, but it is conceivable that they are somehow involved in effecting motor corrections in response to air sac expansion.

Conclusions Although the regulation of ventilation may rely on similar receptors in birds and mammals, the lack of an expanding lung and presence of air sacs could suggest that some of the mechanoreceptive and perhaps chemoreceptive functions have been shifted from the lung to the air sacs in birds. The similarities in the airway receptors and the CNS structures with which they are connected speak of the conservative nature of the respiratory system. Although very limited work has been done on airway receptors in birds as compared to mammals, new findings on the similarities in the two classes might show that birds are a more tractable model in which to examine the function of these receptors, due to the relative ease with which the air sac cavities and membranes can be accessed for experimental purposes. In addition, the respiratory centres that receive descending vocal control input in a way comparable to that in mammals (Wild, 2004) are amenable to both in vitro and in vivo physiological studies in adult animals (e.g. Sturdy et al;., 2003; Kubke et al., 2005), where respiratory mechanisms may be probed from a more integrative point of view.

Acknowledgements This work was supported by NIH R01 NS29467-07 to RA Suthers (PI) and JM Wild.

208  |  Airway Chemoreceptors in the Vertebrates

References Adamson, T.P., and Solomon, I.C. 1993. Effects of low intrapulmonary PCO2 on ventilatory sensitivity to PaCO2 in chickens. Respir Physiol 94:163-171. Adamson, T.P., Myers, B.K., and Solomon, I.C. 1994. Chicken intrapulmonary and cardiac nerve afferents interact in ventilatory control. Respir Physiol 98:283-293. Adriaensen, D., and Scheuermann, D.W. 1993. Neuroendocrine cells and nerves of the lung. Anat Rec 236:70-85; discussion 85-76. Adriaensen, D, Scheuermann, D.W., Gomi, T., Kimura, A., Timmermans, J.P., and De GroodtLasseel, M.H. 1994. The pulmonary neuroepithelial endocrine system in the quail, Coturnix coturnix. Light- and electron-microscopical immunocytochemistry and morphology. Anat Rec 239:65-74. Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: Extremely complex airway receptors. Anat Rec 270:25-40. Ballam, G., Clanton, T., Kaminski, R., and Kunz, A. 1985. Effect of sinusoidal forcing of ventilatory volume on avian breathing frequency. J Appl Physiol 59:991-1000. Ballam, G.O., Clanton, T.L., and Kunz, A.L. 1982. Ventilatory phase duration in the chicken: role of mechanical and CO2 feedback. J Appl Physiol 53:1378-1385. Barker, M.R., Burger, R.E., and Nye, P.C. 1981. Respiratory inhibition from chicken intrapulmonary chemoreceptors reduced by increasing rate of repetitive PCO2 changes. Q J Exp Physiol 66:367-376. Barnas, G.M., Mather, F.B, and Fedde, M.R. 1978. Are avian intrapulmonary CO2 receptors chemically modulated mechanoreceptors or chemoreceptors? Respir Physiol 35:237-243. Bennett, T., and Malmorfs, T. 1970. The adrenergic nervous system of the domestic fowl (Gallus domesticus (L.)). Z Zellforsch 106:22-50. Bezuidenhout, A.J. 2005. Light and electron microscopic study of the thoracic respiratory air sacs of the fowl. Anat Histol Embryol 34:185-191. Bouverot, P. 1978. Control of breathing in birds compared with mammals. Physiol Rev 58:604655. Bouverot, P., and Leitner, L.M. 1972. Arterial chemoreceptors in the domestic fowl. Respir Physiol 15:310-320. Bower, A., Parker, S., and Molony, V. 1978. An autoradiographic study of the afferent innervation of the trachea, syrinx and extrapulmonary primary bronchus of Gallus gallus domesticus. J Anat 126:169-180. Buckley, K., and Kelly, R.B. 1985. Identification of a transmembrane glycoprotein specific for secretory vesicles of neuronal and endocrine cells. J Cell Biol 100:1284-1294. Burger, R.E., Coleridge, J.C., Coleridge, H.M., Nye, P.C., Powell, F.L., Ehlers, C., and Banzett, R.B. 1976a. Chemoreceptors in the paleopulmonic lung of the emu: discharge patterns during cyclic ventilation. Respir Physiol 28:249-259. Burger, R.E., Nye, P.C., Powell, F.L., Ehlers, C., Barker, M., and Fedde, M.R. 1976b. Response to CO2 of intrapulmonary chemoreceptors in the emu. Respir Physiol 28:315-324. Clanton, T.L., Ballam, G.O., Moore, R.K., and Kunz, A.L. 1982. Rapid ventilatory responses to changes in insufflated CO2 in awake roosters. J Appl Physiol 53:1371-1377. Cook, R., Valcart, C., and King, A. 1986. The abdominal air sac ostium of the domestic fowl:a sphimoter regulated by neuro-epithelial cells? J. Anat. 149:101-111.

Airway Receptors in Birds  |  209 Cook, R.D., and King, A.S. 1969. Nerves of the avian lung: electron microscopy. J Anat 105:202. Cook, R., Vailland, C., and King, A. 1978. The structure and innervation of the saccopleural membrane of the domestic fowl, Gallus gallus: an ultrastructural and immunohistochemical study. J Anat 150:1-9. Cook, R., Vaillant, C., and King, A. 1986. The abdominal air sac ostium of the domestic fowl: a sphincter regulated by neuro-epithelial cells? J Anat 149:101-111. Dememes, D., Seoane, A, Venteo, S., and Desmadryl, G. 2000. Efferent function of vestibular afferent endings? Similar localization of N-type calcium channels, synaptic vesicle and synaptic membrane-associated proteins. Neuroscience 98:377-384. DeWet, P.D., Farrell, P.R., and Fedde, M.R. 1971. Number and morphology of muscle spindles in the transversus abdominis muscle of the chicken. Poult Sci 50:1349-1357. Dunker, H.R. 1971. The lung air sac system of birds. A contribution to the functional anatomy of the respiratory apparatus. Ergebnisse der Anatomie. Bd. 45:1-171. Duncker, H.R. 1972. Structure of avian lungs. Respir Physiol 14:44-63. Duncker, H.R. 1974. Structure of the avian respiratory tract. Respir Physiol 22:1-19. Ezure, K., Tanaka, I., Saito, Y., and Otake, K. 2002. Axonal projections of pulmonary slowly adapting repceptor relay neurons in the rat. J Comp Neurol 446:81-94. Fedde, M.R., and Peterson, D.F. 1970. Intrapulmonary receptor response to changes in airwaygas composition in Gallus domesticus. J Physiol 209:609-625. Fedde, M., Burger, R., and Kitchell, R. 1963. Localization of vagal afferents involved in the maintenance of normal avian respiration. Poult Sci 42:1224-1236. Fedde, M.R., Gatz, R.N., Slama, H. and Scheid, P. 1974. Intrapulmonary CO2 receptors in the duck: I. Stimulus specificity. Respir Physiol 22:99-114. Fedde, M.R., Kiley, J.P., Powell, F.L., and Scheid P. 1982. Intrapulmonary CO2 receptors and control of breathing in ducks: effects of prolonged circulation time to carotid bodies and brain. Respir Physiol 47:121-140. Fortin, G., Foutz, A.S., and Champagnat, J. 1994. Respiratory rhythm generation in chick hindbrain: effects of MK-801 and vagotomy. Neuroreport 5:1137-1140. Gleeson, M. 1987. The role of vagal afferents in the peripheral control of respiration in birds. In: The Neurobiology of the Cardiorespiratory System, E. Taylor (Ed.) Manchester University Press, Manchester, pp. 51-79. Gleeson, M., and Molony, V. 1989. Control of breathing. In: Form and Function in Birds, A.S. King and J. McLelland (Eds.) Academic Press, London, pp. 439-484. Gottschaldt, K.M. 1985. Structure and function of avian somatosensory receptors. In: Form and Function in Birds, A. King and J. McLelland (Eds.) Academic Press, London. Groth, H.P. 1972. Licht und fluoreszenzmikroskopische Untersuchungen zur Innervation des Luftsacksystems der Vogel. Z Zellforsch 127:87-115. Halata, Z., Grim, M., and Bauman, K.I. 2003. Friedrich Sigmund Merkel and his “Merkel cell”, morphology, development, and physiology: review and new results. Anat Rec 271A:225-239. Hartley, R.S. 1990. Expiratory muscle activity during song production in the canary. Respir Physiol 81:177-187. Hempleman, S.C., and Burger, R.E. 1985. Comparison of intrapulmonary chemoreceptor response to PCO2 in the duck and chicken. Respir Physiol 61:179-184. Holstege, G. 1989. Anatomical study of the final common pathway for vocalization in the cat. J Comp Neurol. 284:242-252. Hou, X.E., and Dahlstrom, A. 1996. Synaptic vesicle proteins in cells of the sympathoadrenal lineage. J Auton Nerv Syst 61:301-312.

210  |  Airway Chemoreceptors in the Vertebrates Katz, D.M., and Karten, H.J. 1983. Visceral representation within the nucleus of the tractus solitarius in the pigeon, Columba livia. J Comp Neurol 218:42-73. Kemp, P.J., Lewis, A., Hartness, M.E, Searle, G.J., Miller, P., O’Kelly, I., and Peers, C. 2002. Airway chemotransduction: from oxygen sensor to cellular effector. Am J Respir Crit Care Med 166:S17-S24. Kemp, P.J., Searle, G.J., Hartness, M.E., Lewis, A., Miller, P., Williams, S., Wootton, P., Adriaensen, D., and Peers, C. 2003. Acute oxygen sensing in cellular models: Relevance to the physiology of pulmonary neuroepithelial and carotid bodies. Anat Rec 270:41-50. King, A.S., McLelland, J., Cook, R.D., King, D.Z., and Walsh, C. 1974. The ultrastructure of afferent nerve endings in the avian lung. Respir Physiol 22:21-40. Kubke, M.F., Ross, J.M., and Wild, J.M. 2004. Vagal innervation of the air sacs in a songbird, Taenopygia guttata. J Anat 204:283-292. Kubke, M.F., Yazaki-Sugiyama, Y., Mooney, R., and Wild, J.M. 2005. Physiology of neuronal subtypes in the respiratory-vocal integration nucleus retroamigualis of the male zebra finch. J Neurophysiol 94:2379-2390. Langley, K., and Grant, N.J. 1999. Molecular markers of sympathoadrenal cells. Cell Tissue Res 298:185-206. Lauweryns, J.M., Cokelaere, M., and Theunynck, P. 1972. Neuro-epithelial bodies in the respiratory mucosa of various mammals. A light optical, histochemical and ultrastructural investigation. Z Zellforsch Mikrosk Anat 135:569-592. Lopez, J., Diaz de Rada, O., Sesma, P., and Vazquez, J.J. 1983. Silver methods applied to semithin sections to identify peptide-producing endocrine cells. Anat Rec 205:465-470. Maina, J.N. 2002. Some recent advances on the study and understanding of the functional design of the avian lung: morphological and morphometric perspectives. Biol Rev Camb Philos Soc 77:97-152. McLelland, J. 1989. Anatomy of the lungs and air sacs. In: Form and Function in Birds, A.S. King and J. McLelland (Eds.) Academic Press, London, pp. 221-279. McLelland, J., and Macfarlane, C. 1986. Solitary granular endocrine cells and neuroepithelial bodies in the lungs of the Ringed Turtle Dove (Streptopelia risoria). J Anat 147:83-93. Menna, T.M., and Mortola, J.P. 2003. Ventilatory chemosensitivity in the chick embryo. Respir Physiol Neurobiol 137:69-79. Molony, V. 1974. Classification of vagal afferents firing in phase with breathing in Gallus domesticus. Respir Physiol 22:57-76. Nye, P.C., and Burger, R.E. 1978. Chicken intrapulmonary chemoreceptors: discharge at static levels of intrapulmonary carbon dioxide and their location. Respir Physiol 33:299-322. Nye, P.C., and Powell, F.L. 1984. Steady-state discharge and bursting of arterial chemoreceptors in the duck. Respir Physiol 56:369-384. Nye, P.C., Barker, M.R., and Burger, R.E. 1982. Chicken intrapulmonary chemoreceptor discharge frequency reduced by increasing rate of repetitive PCO2 changes. Q J Exp Physiol 67:607-615. Osborne, J.L., and Burger, R.E. 1974. Intrapulmonary chemoreceptors in Gallus domesticus. Respir Physiol 22:77-85. Otake, K., Nakamura, Y., Tanaka, I., and Ezure, K. 2001. Morphology of pulmonary rapidly adapting receptor relay neurons in the rat. J Comp Neurol 430:458-470. Patel, A.J., and Honore, E. 2001. Molecular physiology of oxygen-sensitive potassium channels. Eur Respir J 18:221-227. Peers, C., and Kemp, P.J. 2001. Acute oxygen sensing: diverse but convergent mechanisms in airway and arterial chemoreceptors. Respir Res 2:145-149.

Airway Receptors in Birds  |  211 Perez Fontan, J.J., and Velloff, C.R. 2001. Labeling of vagal motoneurons and central afferents after injection of cholera toxin B into the airway lumen. Am J Physiol Lung Cell Mol Physiol 280:L152-L164. Peterson, D.F., and Fedde, M.R. 1968. Receptors sensitive to carbon dioxide in lungs of chicken. Science 162:1499-1501. Peterson, D.F., and Nightingale, T.E. 1976. Functional significance of thoracic vagal branches in the chicken. Respir Physiol 27:267-275. Portela-Gomes, G.M., Lukinius, A., and Grimelius, L. 2000. Synaptic vesicle protein 2, A new neuroendocrine cell marker. Am J Pathol 157:1299-1309. Powell, F.L., Fedde, M.R., Gratz, R.K., and Scheid, P. 1978. Ventilatory response to CO2 in birds. I. Measurements in the unanesthetized duck. Respir Physiol 35:349-359. Reinke, H., and Wild, J. 1997. Distribution and connections of inspiratory premotor neurons in the brainstem of the pigeon. J Comp Neurol 379:347-362. Reinke, H., and Wild, J. 1998. Identification and connections of inspiratory premotor neurons in songbirds and budgerigar. J Comp Neurol 391:147-163. Salvi, E., and Renda, T. 1992. An immunohistochemical study on neurons and paraneurons of the pre- and post-natal chicken lung. Arch Histol Cytol 55:125-135. Scheid, P. 1979. Respiration and control of breathing in birds. Physiologist 22:60-64. Scheid, P., Gratz, R.K., Powell, F.L., and Fedde, M.R. 1978. Ventilatory response to CO2 in birds. II. Contribution by intrapulmonary CO2 receptors. Respir Physiol 35:361-372. Scheid, P., Slama, H., Gatz, R.N., and Fedde, M.R. 1974. Intrapulmonary CO2 receptors in the duck: III. Functional localization. Respir Physiol 22:123-136. Sturdy, C.B., Wild, J.M., and Mooney, R. 2003. Respiratory and telencephalic modulation of vocal motor neurons in the zebra finch. J Neurosci 23:1072-1086. Suthers, R.A., Goller, F., and Wild, J.M. 2002. Somatosensory feedback modulates the respiratory motor program of crystallized birdsong. Proc Natl Acad Sci U S A 99:5680-5685. Tschorn, R.R., and Fedde, M.R. 1974. Effects of carbon monoxide on avian intrapulmonary carbon dioxide-sensitive receptors. Respir Physiol 20:313-324. van den Steen, P., van Lommel, A., and Lauweryns, J.M. 1994. Neuroepithelial bodies in the lung of Basiliscus vittatus (Reptilia, Iguanidae). Anat Rec 239(2):158-169. Walsh, C., and McLelland, J. 1974a. Granular “endocrine” cells in avian respiratory epithelia. Cell Tissue Res 153:269-276. Walsh, C., and McLelland, J. 1974b. Intraepithelial axons in the avian trachea. Z Zellforsch Mikrosk Anat 147:209-217. Wasano, K., and Yamamoto, T. 1979. APUD-type recepto-secretory cells in the chicken lung. Cell Tissue Res 201:197-205. Widdicombe, J. 2001. Airway receptors. Respir Physiol 125:3-15. Wild, J.M. 1993a. The avian nucleus retroambigualis: a nucleus for breathing, singing and calling. Brain Res 606:319-324. Wild, J.M. 1993b. Descending projections of the songbird nucleus robustus archistriatalis. J Comp Neurol 338:225-241. Wild, J.M. 2004. Functional neuroanatomy of the sensorimotor control of singing. Ann N Y Acad Sci 1016:438-462. Wild, J.M., Kubke, M.F., and Mooney, R. 2009. Avian nucleus retroambigualis: cell types and projections to other respiratory-vocal nuclei in the brain of the zebra finch (Taeniopygia guttata). J. Comp. Neurol. 512: 768-783.

212  |  Airway Chemoreceptors in the Vertebrates Wild, J., Li, D., and Eagleton, C. 1997. Projections of the dorsomedial nucleus of the intercollicular complex (DM) in relation to respiratory-vocal nuclei in the brainstem of pigeon (Columba livia) and zebra finch (Taeniopygia guttata). J Comp Neurol 377:392-413. Wild, J.M., Williams, M.N., and Suthers, R.A. 2000. Neural pathways for bilateral vocal control in songbirds. J Comp Neurol 423:413-426.

10 Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors Steven C. Hempleman* and Jason Q. Pilarski

Introduction Avian intrapulmonary chemoreceptors (IPC) are CO2 sensitive vagal afferents that help control breathing in birds. IPC sensory endings have been localized by reflex and electrophysiological methods to the gas exchange region of the lungs, their cell bodies lie in the nodose ganglia, and their afferent axons project to the brainstem. IPC provide important phasic and tonic sensory feedback that helps control breathing depth and rate and CO2 homeostasis in birds. In this manner, intrapulmonary chemoreceptors in avian lungs have functional similarity to pulmonary stretch receptors in mammalian lungs, although the stimulus modalities detected (chemical vs. mechanical) are different. Important advances in the mechanistic understanding of CO2 chemotransduction in IPC have occurred in the 40 years since their discovery (Hempleman and Posner, 2004), and is the subject of this chapter. The reflex role of IPC in avian respiration has been reviewed recently (Milsom, 2002; Milsom et al., 2004; and in this volume), and will be mentioned only briefly as it pertains to chemotransduction. Unfortunately, the histology of IPC sensory processes in the avian lung remains unknown, and discussion of the anatomical correlates to IPC function must await future study. Existence of avian IPC was revealed in 1968 by Peterson and Fedde and by Burger with their demonstration of a rapid, vagal reflex that produced immediate hypoventilation or apnea when avian lungs were abruptly insufflated with fresh air. Further studies, using unidirectional ventilation to keep respiratory volume Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640, USA. *Author to Correspondence.

214  |  Airway Chemoreceptors in the Vertebrates constant while changing CO2, showed that this profound inhibitory reflex arose from intrapulmonary chemoreceptor neurons that increase their discharge with decreasing CO2 tension, not from mechanoreceptors (Scheid and Piiper, 1986; Powell, 2000). IPC action potential discharge is distinctly phasic in tidally breathing or ventilated birds. Most IPC reach peak discharge rate during mid-late inspiration, then taper to near 0 Hz discharge by the end of expiration (Fedde and Scheid, 1976; Berger et al., 1980). Most IPC are also highly responsive to the rate of CO2 change (Berger et al., 1980; Tallman and Grodins, 1982; Hempleman and Bebout,1994; Tallman et al., 1996), therefore intrabreath variation in IPC discharge rate is even larger than would be expected from already large intrabreath oscillations of lung PCO2 (Powell et al., 1981). IPC are unusually dynamic and responsive respiratory chemoreceptors, seemingly adapted by evolution for monitoring respiration in the rigid avian lung. This chapter begins with a review of the avian respiratory system in the context of IPC function. Then many of the experiments that have revealed the cellular mechanisms behind IPC activity as well as the current conceptual model of IPC signal transduction are described, which incorporates many of these experiments. Finally, recent studies investigating the development and maturation of IPC activity are discussed, and also how IPC activity may be influenced by changes in body size.

The Avian Respiratory System Avian parabronchial lungs evolved differently from mammalian alveolar lungs (Duncker, 1971; Scheid and Piiper, 1986; Powell, 2000; Duncker, 2004). Avian lungs are non-expandable, they have a very small functional residual capacity, and they are flow-through ventilated by body wall movements driving tidal volume changes of anatomically distal airsacs. There are only three main orders of bronchial branching in avian lungs: primary, secondary and tertiary. The fundamental units of gas-exchange in avian lungs are tube-like, rigid, tertiary bronchi or “parabronchi”, which are open at both ends to either secondary bronchi or airsac ostia. Parabronchi are not blind-ended compliant sacs like mammalian alveoli. There are about 200 parallel parabronchi in a duck lung, each about 1-2 mm in diameter and several cm long. In Figure 1 (modified from Duncker [1971]) an avian lung is shown with the parabronchi cut away, leaving dark stubs. Had the parabronchi not been cut away, they would occupy the spaces between the secondary bronchi. Gas flows through the parabronchi on the way to and from the airsacs. Each parabronchus is surrounded by a rigid parenchymal mantle consisting of air capillaries that arise from parabronchi, and blood capillaries that arise from surrounding pulmonary arterioles and venules. Pulmonary gas exchange occurs by diffusion in and between the air and blood capillaries. The convective and diffusive arrangements in the parabronchi form a highly efficient cross-current gas exchanger (Scheid and Piiper, 1972).

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  215

Figure 1: Schematic view of avian lung modified from Duncker (1971). In this picture, the tertiary bronchi (parabronchi) have been cut away from the secondary bronchi, leaving numerous small black stubs (see text).

Because the functional residual capacity (FRC) of the avian parabronchi is only about 10% that of mammalian alveoli (just a few cm3 in 1 kg. ducks, Powell, 2000), during each tidal breath the parabronchial gas volume is completely flushed with inspired gas, and the O2 and CO2 concentrations in the parabronchial mantle fluctuate between near-venous levels (5-6% CO2), and near-inspired levels (0-1% CO2) (Powell et al., 1981). Since IPC sensory endings are located in the parabronchial mantle (Hempleman and Burger, 1984), IPC experience large CO2 oscillations with each breath. This produces large tidal variations in IPC action potential frequency and provides sensory feedback about the magnitude and timing of within-breath CO2 elimination (Fedde and Peterson, 1970; Fedde and Scheid, 1976; Berger et al., 1980; Tallman and Grodins, 1982; Furilla and Bernstein, 1995; Powell, 2000). In contrast, mammalian lungs have a large resident gas volume (FRC) that minimizes tidal CO2 fluctuations within the alveolar gas-exchange zone, and breath-by-breath feedback is provided by stretch-sensitive mechanoreceptors (PSR).

Inverse CO2 Sensitivity in IPC IPC have some unusual properties compared to many other commonly studied respiratory chemoreceptors,most notably their“backwards”response to CO2 (hypocapnia excites, hypercapnia inhibits) and extremely rapid response to CO2 (Figure 2). Action potential discharge rate is inversely proportional to PCO2 in IPC, but it is directly proportional to PCO2 in carotid bodies (Fitzgerald et al., 1990; Iturriaga et al., 1991; Hempleman et al., 1992; Gonzalez et al., 1994; Iturriaga et al., 1998), most mammalian

216  |  Airway Chemoreceptors in the Vertebrates

Figure 2: Spike train discharge from a single unit intrapulmonary chemoreceptor in a Japanese quail (Coturnix japonica), modified from Hempleman et al., 2005. Note the rapid onset of action potentials when CO2 is abruptly reduced from 6% to 0% (at arrow), followed by spike frequency adaptation of discharge rate. The heavy gray line is a running average of IPC discharge rate calculated over successive 0.25 sec intervals.

medullary chemoreceptors (Dean et al., 1989; Richerson, 1995; Erlichman and Leiter, 1997; Wellner-Kienitz, 1998a) and snail pneumostome ganglia (Erlichman and Leiter, 1997). However, reptilian IPC (Scheid et al., 1977), some mammalian laryngeal mechanoreceptors (Coates et al., 1996), and many mammalian medullary raphe CO2 chemoreceptors (Richerson, 1995) have an inverse CO2 sensitivity like that of avian IPC. Therefore inverse CO2 sensitivity may be more common than previously thought, and other CO2 chemoreceptors might share some aspects of signal transduction with IPC.

Intracellular pH (pHi) is an important Stimulus for IPC IPC increase their discharge when airway PCO2 decreases (Fedde and Peterson, 1970). Although most IPC are extremely sensitive to both steady and rapidly changing intrapulmonary PCO2, they are much less sensitive to acute extracellular infusion of fixed acids and bases like HCl (Burger et al., 1974) or HCO3- (Adamson and Burger, 1986). This suggests that IPC respond to either molecular CO2 (which is freely diffusible across cell membranes) or intracellular pH alterations produced by CO2, but not to extracellular pH. Figure 2 shows action potentials from an IPC responding to a CO2 step stimulus— note inverse CO2 sensitivity with rapid response and partial adaptation.

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  217

Several lines of evidence suggest that IPC in fact sense intracellular pH rather than CO2 directly. These include experiments with carbonic anhydrase inhibitors, SO2, the Na/H+ exchange inhibitor dimethyl amiloride, HCO3-/Cl- exchange inhibitor DIDS, the imidazole blocker diethypyrocarbonate, and IPC responses to chronic respiratory acidosis and metabolic acidosis.

Carbonic Anhydrase Inhibitors IPC CO2 transduction is blocked by acetazolamide (Figure 3), a cell permeable carbonic anhydrase (CA) inhibitor that slows the CA catalyzed reaction between CO2 + H2O ↔ H2CO3 ↔ H+ + HCO3- (Maren, 1977; Powell et al., 1978; Scheid et al., 1978; Neubauer, 1991; Hempleman et al., 2000). Interestingly, IPC treated with acetazolamide increase their discharge frequency and become almost completely insensitive to the normally inhibiting effects of increased CO2. IPC treated with acetazolamide act as if they were continually sensing very low CO2 levels, independent of the actual CO2 level. In contrast, benzolamide, a more potent CA inhibitor that has only limited membrane solubility, has significantly less effect on IPC response compared to acetazolamide (Hempleman et al., 2000). Taken together, these experiments indicate that CO2 hydration-dehydration must occur rapidly by CA catalysis inside

Figure 3: The cell-membrane permeable carbonic anhydrase inhibitor acetazolamide raises mean IPC discharge rate and nearly abolishes IPC response to CO2. Left panel shows normal IPC response to step changes in CO2 stimulus. Right panel shows response after acetazolamide administration. Figure modified from Hempleman et al., 2000. IPC discharge rate (fIPC is shown as μ± SEM, n=7).

218  |  Airway Chemoreceptors in the Vertebrates IPC endings, that H+ or HCO3- must be altered intracellularly to affect IPC discharge, and that the products of CO2 hydration (H+ or HCO3-), rather than CO2, are the actual stimuli detected by IPC.

Contrasting the Effects of SO2 and NH3 with CO2 Acute inspiration of sulfur dioxide (SO2) inhibits spontaneous IPC discharge (Chiang et al., 1978). Since SO2 forms acid when hydrated (SO2 + H2O ↔ H2SO3 ↔ H+ + HSO3-), its effect on IPC is thought to be due to intracellular acidification. Conversely, acute inhalation of ammonia excites IPC (Vivoni et al., 2001; Kumaraswamy et al., 2004). Since NH3 acts as a base (NH3 + H2O ↔ NH4OH ↔ NH4+ + OH‑), its effect on IPC is likely due to intracellular alkalinization. Interestingly, neither SO2 nor NH3 is as rapid or potent as CO2 in modifying IPC discharge, but this may be partly explained by CO2 having intracellular carbonic anhydrase to catalyze its hydration, while SO2 and NH3 hydration is most likely uncatalyzed.

Transmembrane Acid-Base Exchangers in IPC CO2 Transduction Dimethyl amiloride (DMA, Figure 4, top panel), a blocker of the acid-base regulating Na+/H+ antiport, significantly inhibits IPC response to PCO2 and decreases IPC discharge (Hempleman et al., 2003). This suggests that H+ accumulates inside IPC endings when the Na+/H+ antiport is blocked, thereby reducing intracellular pH (pHi) and inhibiting IPC discharge. Conversely, 4, 4’-diisothiocyanatostilbene-2, 2’disulfonic acid (DIDS; Figure 4, bottom panel), a blocker of the acid-base regulating HCO3-/Cl- antiport, increases IPC discharge (Shoemaker and Hempleman, 2001). This suggests that HCO3- accumulates inside IPC endings when the HCO3-/Clantiport is blocked, thereby increasing pHi and exciting IPC discharge. In our current model of IPC CO2 transduction, it was proposed that these transmembrane acid-base exchangers help kinetically balance the H+ and HCO3- produced by CA catalyzed CO2 hydration.

Imidazole is a likely Intracellur Target for H+ Modulation of IPC Activity Relatively little is known about acid/base-sensitive ion channels expressed in IPC, but it seems likely that such channels, along with carbonic anhydrase, acid-base transporters and protein buffers, underlie IPC signal transduction. In other respiratory CO2

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  219

Figure 4: Response of IPC to steady levels of inspired (intrapulmonary) PCO2 between 5 torr and 45 torr. Top panel shows IPC discharge rate (μ± SEM, n=7 at 1 μmol/kg, n=8 at 3 μmol/kg, n=5 at 8 μmol/kg) with and without DMA (dimethyl amiloride), an inhibitor of the membrane Na+/H+ exchanger. DMA inhibits IPC discharge at all PCO2. Bottom panel shows IPC discharge rate (μ± SEM, n=12) with and without DIDS (4, 4’-diisothiocyanatostilbene-2, 2’-disulfonic acid), an inhibitor the membrane HCO3-/Cl- exchanger. DIDS excites IPC discharge at all PCO2. Modified from Shoemaker and Hempleman, 2001 and Hempleman et al., 2003.

220  |  Airway Chemoreceptors in the Vertebrates chemoreceptors, acid-sensitive ion channels have been identified (Putnam et al., 2004) and include background K+ leak channels (Buckler et al., 2000; Bayliss et al., 2001), Ca2+-activated K+ channels (Wellner-Kienitz et al., 1998b; Peers and Green, 1991), and pH-sensitive Cl- leak channels (Petheo et al., 2001). H+ gating of ion channels is assumed to occur through conformational changes in the channel protein. According to Reeves’ (1972) “alphastat regulation” theory, many proteins that respond to physiological pH changes near 7.0 contain histidine-imidazole groups. Because imidazole pKa is close to the pKw of water, imidazole ionization ratio ([Im‑]/ [ImH], or “alphaIm”) and the [OH‑]/[H+] ratio of body fluids change proportionately with changes in physiological pH. The degree of alphaIm ionization is thought to provide a steric mechanism linking changes in acid-base balance with changes in protein conformation and function. Nattie (1986a, 1986b, 1988) tested for alphastat regulation in mammalian medullary CO2 chemoreceptors using diethyl pyrocarbonate (DEPC), a chemical that specifically binds and inhibits potential imidazole buffering of H+ at normal physiological pH. DEPC attenuated or eliminated mammalian central CO2 sensitivity (Nattie 1986a, 1986b, 1988; Krause et al, 1998), and DEPC had the same inhibitory effect on pulmonate snail CO2 sensitivity (Lu et al., 1998). Recent tests on avian IPC in vivo demonstrated a profound inhibitory effect of DEPC on CO2 chemosensitivity and discharge rate (Figure 5; Pilarski and Hempleman, 2006), suggesting that imidazole groups and alphastat regulation are also critical in IPC CO2 chemotransduction. Further studies are needed to determine the target proteins for DEPC in IPC, which may include acid-sensitive ion channels, acidbase transporters, pH modulated enzymes, and protein H+ buffers. However, DEPC does not seem to interfere with carbonic anhydrase function, at least in avian erythrocytes (Pilarski and Hempleman, 2006).

Chronic Hypercapnia Resets IPC CO2 Sensitivity, but not IPC pH Sensitivity Acclimatization to elevated inspired CO2 recalibrates the physiological relationship between PCO2 and pH, and provides another way to test whether IPC respond to PCO2 directly, or to pH. Sustained hypercapnia (7% inspired CO2) produces respiratory acidosis which is compensated over several weeks by metabolic alkalosis (Bebout and Hempleman, 1999). This compensatory metabolic alkalosis elevates HCO3- in body tissues and alters CO2 buffering, causing intracellular and extracellular pH to be higher for any given PCO2 level after acclimatization (Bebout and Hempleman, 1999). Acclimatized birds have a normal arterial pH at the elevated, acclimatized hypercapnic CO2 level, but they have an alkalotic pH when they are acutely exposed to normal CO2. Interestingly, acclimatized birds show increased IPC discharge at all PCO2 stimulus levels compared to normal birds; however, IPC discharge is nearly

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  221

Figure 5: Cycle triggered stimulus histograms of dynamic IPC discharge (μ± SEM, n=10) as inspired CO2 was stepped between 0% and 7% (11 sec stimulus cycle, 0.25 sec bin width, two cycles shown): Symbols: (closed circle) = Mean control IPC discharge, (open circle) = Mean IPC response after 15-150 mM DEPC. Control response shows characteristic adaptation and increased discharge rate during CO2 down-step, and low discharge rate at high CO2 (significant IPC entrainment to CO2 step, P <0.05). After i.v. infusion of DEPC, IPC entrainment to CO2 step was eliminated (P >0.5). Modified from Pilarski and Hempleman, 2006.

identical in acclimatized and normal birds if IPC discharge rate is referenced to the pH produced by various PCO2 stimulus levels. This is consistent with the hypothesis that IPC responds to pH, not to CO2 directly (Bebout and Hempleman, 1999).

The Role of Calcium in IPC Response to CO2 Most IPC are sensitive to both tonic and phasic CO2 stimuli. When IPC experience an abrupt increase or decrease in CO2 stimuli, they often show an exaggerated discharge response immediately following the stimulus change, followed by partial spike frequency adaptation (SFA). In many neurons, SFA is caused by opening of apamin-sensitive small conductance calcium-activated potassium channels (sKCa) in response to intracellular calcium influx (Hille, 1992a). However, in IPC, block of sKCa channels by apamin has no significant effect on SFA (Hempleman et al., 2006), so other mechanisms must be responsible for SFA. Even so, Ca++ influx through voltagegated Ca++ channels does significantly modulate IPC response to CO2. Inorganic Ca++ channel blockers Cd++ or Co++ tonically stimulate IPC discharge rate (Hempleman et

222  |  Airway Chemoreceptors in the Vertebrates al., 2006). Likewise, the L-type Ca++ channel blocker nifedipine tonically stimulates IPC discharge, while the L-type channel opener Bay-K 8644 tonically depresses IPC discharge rate (Hempleman et al., 2006). Transmembrane influx of Ca++ into IPC via L-type voltage-gated Ca++ channels thus inhibits overall excitability, and this modulates IPC response to CO2. Large conductance calcium-activated potassium channels (BKCa) contribute to this negative feedback regulation of IPC excitability. Administering charybdotoxin, a specific BKCa channel blocker, relieves about 40% of the Ca++-related feedback inhibition, and increases IPC excitability.

Evidence for pH-Sensitive Trek Tandem Pore Domain Potassium Leak Channels in IPC CO2 Transduction In IPC, as with other respiratory chemoreceptors, some of the most pressing signal transduction questions involve the mechanisms of primary stimulus sensing: how does alkaline pHi /decreasing PCO2 increase IPC membrane excitability and spike generation, and how does acid pHi /increasing PCO2 decrease IPC excitability? New evidence indicates that IPC express TREK-like potassium leakage channels that may help explain this linkage (Bina and Hempleman, 2007). TREK channels belong to the tandem pore domain family of potassium leak channels that includes TASK, TREK, and TRAAK. Because these potassium leak channels help set the resting membrane potential, physiochemical agents that change TREK open-state probability can have a direct effect on cell excitability (Patel et al., 1999; Buckler et al, 2000; Bayliss et al., 2001; Patel and Honore, 2001; Jiang et al., 2005; Lopes et al., 2005). While all tandem pore domain channels are potassium leak channels, different members of the tandem pore domain channel family are modulated by different physiochemical agents. For example TREK channels are opened by membrane stretch, intracellular acidosis, arachidonic acid, certain polyunsaturated fatty acids, heat, and certain gas anesthetics (relative anesthetic potency is ether > chloroform > halothane > isoflurane). In contrast, TASK channels are closed by extracellular acidosis, and opened by halothane and isoflurane but not chloroform or ether. Experiments show that IPC excitability is depressed by ether, chloroform, halothane, and isoflurane in the expected order of potency for TREK-type channels (Bina and Hempleman, 2007). Since IPC appear to express TREK-like channels that should be opened by intracellular acidosis and closed by alkalosis, this may explain the fundamental inverse CO2/pH sensitivity of IPC. More work is needed to test TREK channel contribution to IPC CO2 sensitivity. Considerable evidence suggests that TASK channels may underlie chemoreception in carotid body glomus cells and pulmonary neuroepithelial bodies of mammals (Buckler et al., 2000; Kemp et al., 2002) as well as central CO2 chemoreceptors (Bayliss

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  223

et al., 2001; Putnam et al., 2004). TREK channels in IPC may play an analogous role to TASK channels in central CO2 chemoreceptors, except that presence of TREK in IPC confers their characteristic inverse CO2 sensitivity (alkaline pHi excites, acid pHi inhibits).

Current Model of IPC CO2 Transduction IPC appear to sense intracellular H+ as do other CO2 chemoreceptors (Putnam et al., 2004), but they are stimulated by low CO2 and [H+] and inhibited by high CO2 and [H+]. IPC also have unusual responses to the inhibition of carbonic anhydrase, Na+/H+ antiports, and HCO3–/Cl– antiports. Intracellular carbonic anhydrase inhibition does not merely slow the IPC response to CO2 as it does in other respiratory chemoreceptors—it eliminates it. IPC are also maximally stimulated rather than inhibited by intracellular CA inhibition, which suggests that their intracellular [H+] becomes very alkalotic. To explain this, it was proposed that intracellular [H+] in IPC is controlled as a dynamic steady state, and that it is critically dependent on the rates of several counterbalancing reactions (Figure 6). In our model, when CO2 levels are high the catalyzed hydration of CO2 to H+ and HCO3– occurs very rapidly, and is balanced by rapid transmembrane extrusion of H+ by Na+/H+ antiports, HCO3- by HCO3–/Cl- antiports, and intracellular H+ buffering. Conversely, when CO2 levels are low, the catalyzed dehydration of H+ and HCO3– to CO2 predominates and the antiport extrusion rates probably decrease. Under normal conditions, it is proposed that the dynamic balance of these reactions is uniquely determined by the CO2 stimulus level, thereby producing a characteristic pHi and IPC discharge rate. For example, with CA inhibition in this model, the CO2 hydration rate is considerably slowed but H+ and HCO3- extrusion continues. With its dynamic balance upset, the CA-inhibited IPC becomes alkaline and discharges maximally. Likewise, if Na+/H+ exchange is blocked but catalyzed CO2 hydration continues, H+ accumulates in IPC endings and discharge decreases. If HCO3–/Cl- exchange is blocked, HCO3– accumulates intracellularly and discharge increases. Action potential discharge allows calcium ion influx through charybdotoxin-sensitive L-type Ca++ channels, and this results in negative feedback on IPC excitability exerted through large conductance calcium activated potassium channels (BKCa). Tandem pore domain TREK ion channels may be the primary transducer that couples changes in pHi to changes in transmembrane potential (alkalinity closes TREK channels, and could underlie depolarization by low CO2; acidity opens TREK channels and this could underlie inhibition by high CO2). Critique of Model. Although based on empirical data, the current model seems rather complicated to explain the very rapid and precise CO2 responses of IPC (Figure 2). If carbonic anhydrase, protein/imidazole H+ buffers, Na+/H+ exchangers,

224  |  Airway Chemoreceptors in the Vertebrates

Figure 6: Current model of IPC chemotransduction (see text for discussion).

Cl-/HCO3– exchangers, various ion channels, and H+ sensitive TREK leak channels were all located randomly in the IPC sensory ending, how could simple diffusion deliver CO2-derived H+ and HCO3– to all signal transduction components rapidly enough to produce responses like in Figure 2? A mechanism that could explain this is subcellular organization of critical CO2 transduction elements into functional units (Hempleman and Posner, 2004). While no experimental evidence for such organization currently exists in IPC, other than the extraordinary rapidity of IPC CO2 response, there is considerable evidence for subcellular organization enhancing signal transduction in other cells. This includes carbonic anydrase-HCO3–/Cl– transport metabolons (Sterling et al., 2001), metabolic substrate channeling (Ovadi and Saks, 2004), organization of inner membrane cytochromes for electron transfer and oxidative phosphorylation in mitochondria, and organization of local synaptic circuits in dendritic spines (Kennedy et al., 2005). Further work is needed to test the idea of metabolons or metabolic channeling in IPC chemotransduction, but it may have to await development of histochemical techniques for visualizing IPC receptive endings.

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  225

Ontogeny: Development and Maturation of Avian IPC Since avian embryos develop in cleidoic eggs outside the mother’s body, they offer a classic, accessible system for exploring vertebrate development and maturation. Nevertheless, only one study so far has investigated avian IPC development (Pilarski and Hempleman, 2007), even though a lot of previous work has been done on the prenatal, paranatal and postnatal changes in avian physiology occurring during the conversion from chorioallantoic (CAM) to pulmonary respiration (Abramovici, 1967; Dawes and Simkiss, 1969; Freeman and Misson, 1970; Erasmus et al., 1970/71; Tazawa et al., 1971; Tazawa, 1981, 1982, 1986, 1987). This work highlights several critical developmental periods during avian hatching, which are delimited by the events of internal pipping (chick breaks through the inner shell membrane and starts rebreathing gas from the internal air cell of the egg), external pipping (chick breaks through the external calcite shell and starts breathing fresh atmospheric air), and hatch-out (chick separates completely from chorioallantoic membrane and relies entirely on pulmonary respiration). The physiological changes associated with these events present major homeostatic challenges for the chick’s respiratory exchange and control systems (Vince and Tolhurst, 1975; Visschedijk, 1968a, 1968b). As discussed here, the “prenatal” period comprises 6 h immediately prior to internal pipping, the “paranatal” period begins at internal pipping and ends with external pipping, and the “postnatal” period begins with external pipping and extends for 4 d post-hatching. Recordings of IPC discharge from prenatal, paranatal, and postnatal ducklings indicate that IPC undergo development and maturation, likely triggered by the conversion from CAM to pulmonary respiration (Pilarski and Hempleman, 2007). IPC were not present (or more likely inactive—that is not discharging) in prenatal chicks. IPC were first observed in paranatal chicks, but the paranatal IPC response to CO2 was almost entirely phasic with little or no tonic CO2 sensitivity (Figure 7). During the postnatal period IPC retained their phasic CO2 sensitivity, but showed a gradual increase in tonic CO2 sensitivity toward adult levels. Therefore, evidence indicates that IPC are silent while the lungs are filled with amniotic fluid (prenatal period), and start discharging once the lungs become filled with air (paranatal period). Immature IPC exhibit increased peak discharge frequencies and greater spike frequency adaptation compared to adult ducks, and also exhibit blunted tonic sensitivity in response to a maintained CO2 stimulus (Figure 7). Since phasic IPC CO2 sensitivity is fully developed when air-breathing commences, yet tonic IPC CO2 sensitivity exhibits postnatal maturation, the cellular mechanisms that underlie phasic and tonic IPC action potential discharge may be independent processes with different developmental trajectories.

226  |  Airway Chemoreceptors in the Vertebrates

Figure 7: Cycle-triggered stimulus histograms (μ± SEM) of IPC action potential discharge as inspired CO2 was stepped between 0% and 7% (11 sec stimulus cycle, 0.250 sec bin width, two cycles shown; n = # of IPC single units). Ducklings had higher peak discharge frequencies and increased magnitudes of adaptation than adult ducks. Although IPC both developmental periods displayed significant entrainment to the CO2 stimulus, each period exhibited a unique IPC discharge pattern (P < 0.05). Modified from Pilarski and Hempleman, 2007.

CO2 Signal Transduction in IPC is Affected by Animal Body Size Allometry is the study of the scaling of biological traits (Y) with body mass (M) according to the power function Y=aMb (Schmidt-Nielsen, 1984). It is well known that biological rates (Y) in small animals are usually higher than those in large animals, and that they scale in proportion to Y=aM­­­-¼. This can be observed in the scaling of breathing rate, heart rate, gait frequency, or mass-normalized metabolic rates for animals over a wide range of body mass (Schmidt-Nielsen, 1984). Our laboratory

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  227

recently tested whether action potential spike rates in IPC scale allometrically with body size (Hempleman et al., 2005). Neurons like IPC generate spike trains that provide sensory information about the timing and magnitude of breathing. IPC monitor breath-by-breath intrapulmonary CO2 oscillations, which are entrained to breathing frequencies that vary from about 10 min-1 in a 100 kg resting ostrich to more than 175 min-1 in a 4g resting hummingbird (Frappell et al., 2001). Since breathing frequency scales approximately to M-¼ (higher in small birds, lower in large birds), it was reasoned that IPC discharge frequencies may scale in the same way to maintain a constant amount of spike information transmitted within each breath. Phasic action potential discharge pattern, as quantified by the peak IPC discharge rate and the magnitude of IPC spike frequency adaptation, did in fact scale between aM‑0.22 and aM‑0.26, i.e., higher in smaller birds and lower in larger birds, just like breathing rate (Figure 8). Mass-dependent scaling of neural coding in IPC spike trains may be necessary for preserving information transmission with decreasing body size. However, since peak action potential frequency is limited to about 300Hz in most neurons due to highly conserved genetics and kinetics of voltage-gated ion channels (Hille, 1992b), very small birds may hit an upper limit for modulating peak discharge rates, and rely more heavily on increased spike frequency adaptation to encode rapid changes in lung CO2. This needs to be tested further with birds like ostriches and hummingbirds at the extremes of body mass.

Comparison of Ipc to Pulmonary Neuroepithelial Bodies Aves, like all air- breathing vertebrates studied, possess neuroepithelial/neuroendocrine cells (NECs) in their respiratory tracts, often in aggregations, resembling mammalian neuroepithelial bodies or NEBs (Cook et al., 1986; McLelland and Macfarlane, 1986, Kubke et al., 2004). Avian NEB and NEC receive vagal innervation, and are rich in secretory granules. In birds, NEC and NEB are known by their histology and histochemistry, but their function is not understood. In mammals, NEB histology and histochemistry are very well documented, and cellular electrophysiological studies indicate that mammalian NEBs have functional properties consistent with chemoreceptors involved in airway O2 sensing (Kemp et al., 2002) and perhaps mechanoreception (Adriaensen et al., 2006). At the present time, experimental evidence suggests that avian IPC (known exclusively by their functional properties, not by their structure) and avian NEBs/ NECs (known exclusively by their structural properties, not by their function) seem to be distinctly different receptors. IPC are highly responsive airway CO2 chemoreceptors with no O2 sensitivity, while NEB, at least in mammals, are O2 chemoreceptors or perhaps mechanosensors, with negligible CO2 sensitivity.

228  |  Airway Chemoreceptors in the Vertebrates

Figure 8: Compared to IPC in large birds, smaller birds have higher peak IPC discharge rates (μ± SEM) in response to an inspired CO2 stimulus steps. Shown here are averaged IPC discharge responses to CO2 steps from lovebird (Agapornis roseicollis), quail (Coturnix japonica), pigeon (Columbia livia), duck (Anas platyrhynchos) and goose (Anser anser). Peak IPC discharge rates scale allometrically with a mass exponent of ‑¼. Breathing rates in birds also scales allometrically with a mass exponent of -¼. In the text we discuss the importance of scaling IPC information transmission rate with breathing rate in birds of all body sizes. Modified from Hempleman et al. (2005).

Summary Important gains in the mechanistic understanding of CO2 signal transduction have occurred in the 40 years since IPCs were discovered. Many lines of evidence now indicate that IPC sense changes in intracellular pH (pHi) resulting from the catalyzed hydration/dehydration of CO2, i.e., CO2 + H2O ↔ H2CO3 ↔ H+ + HCO3-, and not CO2 directly. Highly responsive IPC provide breath-by-breath sensory feedback about tidal breathing that may have evolved in concert with the unique structural organization of the avian respiratory system. Because avian lungs are rigid with a small functional residual capacity, the parabronchi do not change volume, and the parabronchi are completely flushed with inspired gas by each tidal breath, producing the large CO2 oscillations detected by IPC. In contrast, mammalian lungs have compliant lungs and a large functional residual capacity; this minimizes tidal CO2

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  229

fluctuations within the alveoli, and breath-by-breath feedback is provided by stretchsensitive mechanoreceptors (PSR). The most recent work on IPC reveals that pH “sensing” involves several cellular components besides carbonic anhydrase and membrane acid-base transporters previously described. These include: (1) protein H+ buffering by imidazole groups and alphastat regulation that are critical to IPC signal transduction; (2) modulation of CO2 transduction by Ca++ influx that has a depressive effect on IPC discharge; and (3) pH-sensitive TREK tandem pore domain potassium leak channels that may underlie IPC chemoreception. Finally, two new areas of IPC research were reviewed. The first involves the development and maturation of IPC chemotransduction, which are shown to be a useful comparative model for studying cellular mechanisms underlying the normal (and possibly abnormal) development of respiratory control by CO2. A second area involves the effect of body size on IPC neural spike coding. This research shows how IPC sensory neurons, which monitor both the timing and magnitude of tidal CO2 stimuli, maintain information transmission as body size and breathing rates change. Initial experiments indicate that IPC discharge frequency scales allometrically between aM-0.22 and aM-0.26, like breathing rates. Despite all that has been learned about IPC chemotransduction mechanisms there are still many important questions that remain unanswered. Most of these questions involve the histology, electrophysiology, and cell biology of the relatively inaccessible IPC endings within the lung tissue. For example, what type(s) of ion channels are involved in IPC primary stimulus sensing? How are the functional elements of IPC signal transduction organized intracellularly? And, what do avian IPC receptor endings look like? These questions and more await future investigations as well as the applications of modern neurobiological techniques.

Acknowledgements This work was supported by grants from the U.S. National Science Foundation, IBN0217815, and from the U.S. National Institutes of Health, HL087269-01.

References Abramovici, A., 1967. L’evolution du pH du plasma et des liquides extra embryonnaires de l’embryon de poulet au cours du development normal. C. R. Acad. Sci. (Paris) 265:336-339. Adamson, T.P., and Burger, R.E. 1986. Sodium bicarbonate infusion increases discharge frequency of intrapulmonary chemoreceptors only at high CO2. Respir. Physiol. 66(1):83‑93. Adriaensen, D., Brouns, I., Pintelon, I., De Proost, I., and Timmermans, J.-P. 2006. Evidence for a role of neuroepithelial bodies as complex airway sensors: comparison with smooth muscleassociated airway receptors. J. Appl. Physiol. 101:960-970.

230  |  Airway Chemoreceptors in the Vertebrates Bayliss, D.A., Talley, E.M., Sirois, J.E., and Lei, Q. 2001. Task-1 is a highly modulated pHsensitive ‘leak’ K channel expressed in brainstem respiratory neurons. Respir. Physiol. 129:159-174. Bebout, D.E., and Hempleman, S.C. 1999. Chronic hypercapnia resets CO2 sensitivity of avian intrapulmonary chemoreceptors. Am. J. Physiol. 276 (Reg. Int. Comp. Physiol. 45): R317‑R322. Berger, P.J., Tallman Jr., R.D., and Kunz, A.L.. 1980. Discharge of intrapulmonary chemoreceptors and its modulation by rapid FICO2 changes in decerebrate ducks. Respir. Physiol. 42:123130. Bina, R.W., and Hempleman, S.C. 2007. Evidence for TREK-like tandem-pore domain channels in intrapulmonary chemoreceptor chemotransduction. Evidence for TREK-like tandem pore domain leak channels in intrapulmonary CO2 chemoreception. Respir. Physiol. Neurobiol. 14;156(2):120-31. Buckler, K.J., Williams, B.A., and Honore, E. 2000. An oxygen-, acid- and anesthetic sensitive TASK-like background potassium channel in rate arterial chemoreceptor cells. J. Physiol. (London). 525:135-142. Burger, R.E. 1968. Pulmonary chemosensitivity in the domestic fowl. Fed. Proc. 27:328. Burger, R.E., Osborne, J.L., and Banzett, R.B. 1974. Intrapulmonary chemoreceptors in Gallus domesticus: adequate stimulus and functional localization. Respir. Physiol. 22:87‑97. Chiang, M.J., Berger, P.J., and Kunz, A.L. 1978. A study of the effect of SO2 on pacing and intrapulmonary chemoreceptor discharge in the domestic fowl. Respir. Physiol. 33(2):229‑239. Coates, E.L., Knuth, S.L., and Bartlett Jr., D. 1996. Laryngeal CO2 receptors: influence of systemic PCO2 and carbonic anhydrase inhibition. Respir. Physiol. 104(1):53‑61. Cook, R., Vailland, C., and King, A. 1986. The abdominal air sac ostium of the domestic fowl: a sphincter regulated by neuroepithelial cells? J. Anat. 149:101-111. Dawes, C., and Simkiss, K. 1969. The acid-base status of the blood of the developing chick embryo. J. Exp. Biol. 50:70-86 Dean, J.B., Lawing,W.L., and Millhorn, D.E. 1989. CO2 decreases membrane conductance and depolarizes neurons in the nucleus tractus solitarii. Exp. Brain Res. 76(3):656‑661. Duncker, H.-R. 1971. The lung air sac system of birds. Adv. Anat. Embryol. Cell Biol. 45:7-171. Duncker, H.-R. 2004.Vertebrate lungs: structure, topography and mechanics. A comparative perspective of the progressive integration of respiratory system, locomotor apparatus and ontogenetic development. Respir. Physiol. Neurobiol. 144(2-3):111-124. Erasmus, Bde. W., Howell, B.J., and Rahn, H. 1970/71. Ontogeny of acid-base balance in the bullfrog and chicken. Respir. Physiol. 11:46-53. Erlichman, J.S., and Leiter, J.C. 1997. Comparative aspects of central CO2 chemoreception. Respir. Physiol. 110:177‑185. Fedde, M.R., and Peterson, D.F. 1970. Intrapulmonary receptor response to changes in airway‑gas composition in Gallus domesticus. J. Physiol. (London). 209(3):609‑625. Fedde, M.R., and Scheid, P. 1976. Intrapulmonary CO2 receptors in the duck: IV. Discharge pattern of the population during a respiratory cycle. Respir. Physiol. 26:223‑227. Fitzgerald, R.S., Shirihata, M., and Lahiri, S. 1990. Amiloride and carotid body chemoreception of hypercapnia and hypoxia. Respir. Physiol. 81(3):337‑347. Frappell, P.B., Hinds, D.S., and Boggs, D.F. 2001. Scaling of respiratory variables and the breathing pattern in birds: an allometric and phylogenetic approach. Physiol. Biochem. Zool. 74(1):75-89.

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  231 Freeman, B.M., and Misson, B.H. 1970. pH, Po2, and Pco2 of blood from the foetus and neonate of Gallus domesticus. Comp. Biochem. Physiol. 33:763-772. Furilla, R.A., and Bernstein, M.H. 1995. Intrapulmonary CO2 rise time and ventilation in ducks. J. Appl. Physiol. 79(5):1397‑1404. Gonzalez, C., Almaraz, L. Obeso, A., and Rigual, R. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol. Rev. 74:829‑898. Hempleman, S.C., and Burger, R.E. 1984. Receptive fields of intrapulmonary chemoreceptors in the pekin duck. Respir. Physiol. 57:317‑330. Hempleman, S.C., and Bebout, D.E. 1994. Increased venous PCO2 enhances dynamic responses of avian intrapulmonary chemoreceptors. Am. J. Physiol. (Reg. Int. Comp. Physiol.): 266: R15‑R19. Hempleman, S.C., and Posner, R.G. 2004. CO2 transduction mechanisms in avian intrapulmonary chemoreceptors: experiments and models. Respir. Physiol. Neurobiol. 144(23):203-214. Hempleman, S.C., Powell, F.L., and Prisk, G.K. 1992. Avian arterial chemoreceptor responses to steps of CO2 and O2. Respir. Physiol. 90: 325‑340. Hempleman, S.C., Rodriguez, T.A., Bhagat, Y., and Begay, R.S. 2000. Benzolamide, acetazolamide and signal transduction in avian intrapulmonary chemoreceptors. Am. J. Physiol. (Reg. Int. Comp. Physiol.): 279:R1988-R1995. Hempleman, S.C., Adamson, T.P., Begay, R.S., and Solomon, I.C. 2003. CO2 transduction in avian intrapulmonary chemoreceptors is critically dependent on transmembrane Na+/H+ exchange. Am. J. Physiol. (Reg. Int. Comp. Physiol.) 284(6):R1551-R1559. Hempleman, S.C., Kilgore, Jr., D.L., Colby, C. Bavis, R.W, and Powell, F.L. 2005. Spike firing allometry in avian intrapulmonary chemoreceptors: matching neural code to body size. J. Exp. Biol. 208(Pt 16):3065-73. Hempleman, S.C., Egan, S.X., Pilarski, J.P., Adamson, T.P., and Solomon, I.C. 2006. Calcium and avian intrapulmonary chemoreceptor response to CO2. J. Appl. Physiol. 101(6):1565-1575. Hille, B. 1992a. In: Ionic Channels of Excitable Membranes, 2nd ed. Sinauer Associates, Sunderland, MA. pp. 79-81. Hille, B. 1992b. In: Ionic Channels of Excitable Membranes, 2nd ed. Sinauer Associates, Sunderland, MA. pp. 525-544. Iturriaga, R., Mokashi, A., and Lahiri, S. 1991. Carbonic anhydrase and chemoreception in the cat carotid body. Am. J. Physiol. (Cell Physiol.) 261(30):C565‑C573. Iturriaga, R., Mokashi, A., and Lahiri, S. 1998. Anion exchanger and chloride channel in cat carotid body chemotransduction. J. Auton. Nerv. Syst. 70(1‑2):23‑31. Jiang, C., Rojas, A., Wang, R., and Wang, X. 2005. CO2 central chemosensitivity: why are there so many sensing molecules? Respir. Physiol. Neurobiol. 145(2-3):115-126. Kemp, P.J., Lewis, A., Hartness, M., E., Searle, G.J., Miller, P., O’Kelly, I., and Peers, C. 2002. Airway chemotransduction: from oxygen sensor to cellular effector. Am. J. Respir. Crit. Care Med. 166(12 Pt 2):S17-S24. Kennedy, M.B., Beale, H.C., Carlisle, H.J., and Washburn, L.R. 2005. Integration of biochemical signaling in spines. Nature Rev. Neurosci. 6(6):423-434. Krause, W.L., Kazemi, H., and Burton, M.D. 1998. Potential role for imidazole in the rhythmic respiratory activity of the in vitro neonatal rat brainstem. Neurosci. Lett. 251(3):153-160. Kubke, M.F., Ross, J.M., and Wild, J.M. 2004. Vagal innervation of the airsacs in a songbird, Taenopygia guttata. J. Anat. 204:283-292.

232  |  Airway Chemoreceptors in the Vertebrates Kumaraswamy, M., Solomon, I.C., and Hempleman, S.C. 2004. Ammonia, pH, and signal transduction in avian intrapulmonary chemoreceptors.  FASEB J. 18(5): A1060-A1061. Lopes, C.M., Rohacs, T., Czirjak, G., Balla, T., Enyedi, P., and Logothetis, D.E. 2005. PIP2 hydrolysis underlies agonist-induced inhibition and regulates voltage gating of two-pore domain K+ channels. J. Physiol. 564(Pt 1):117-29. Lu, D.C., Erlichman, J.S., and Leitner, J.C. 1998. Diethyl pyrocarbonate (DEPC) inhibits CO2 chemosensitivity in Helix aspersa. Respir. Physiol. 111(1):65‑78. Maren, T.H. 1977. Use of inhibitors in physiological studies of carbonic anhydrase. Am. J. Physiol. 232(4):F291‑F297. McLelland, J., and Macfarlane, C. 1986. Solitary granular endocrine cells and neuroepithelial bodies in the lungs of Ringed Turtle Dove (Streptopelia risoria). J. Anat. 147:83-93. Milsom, W.K. 2002. Phylogeny of CO2/H+ chemoreception in vertebrates. Respir. Physiol. Neurobiol. 131(1-2): 29-41. Milsom, W.K., Abe, A.S., Andrade, D.V., and Tattersall, G.J. 2004. Evolutionary trends in airway CO2/H+ chemoreception. Respir. Physiol. Neurobiol. 144(2-3):191-202. Nattie, E.E. 1986a. Diethyl pyrocarbonate (an imidazole binding substance) inhibits rostral VLM CO2 sensitivity. J. Appl. Physiol. 61:843-850. Nattie, E.E. 1986b. Intracisternal diethylpyrocarbonate inhibits central chemoreception in conscious rabbits. Respir. Physiol. 64:161-176. Nattie, E.E. 1988. Diethyl pyrocarbonate inhibits rostral ventrolateral medullary H+ sensitivity. J. Appl. Physiol. 64:1600-1609. Neubauer, J.A. 1991. Carbonic anhydrase and sensory function in the central nervous system. In: The Carbonic Anhydrases. S. Dodgson, R. Tashian, G. Gros, and N. Carter (Eds.), Plenum Press, New York. pp. 49‑70. Ovadi, J., and Saks, V. 2004. On the origin of intracellular compartmentation and organized metabolic systems. Mol. Cell. Biochem. 256/257:5-12. Patel, A.J., and Honore, E. 2001. Anesthetic-sensitive 2P domain K+ channels. Anesthes. 95(4):1013-1021. Patel, A.J., Honore, E., Lesage, F., Fink, M., Romey, G., and Lazdunski, M. 1999. Inhalational anesthetics activate two-pore-domain background K+ channels. Nature Neurosci. 2(5):422-426. Peers, C., and Green, F.K. 1991. Inhibition of Ca-activated K currents by intracellular acidosis in isolated type I cells of neonatal rat carotid body. J. Physiol. (London). 437:589-602. Peterson, D.F., and Fedde, M.R. 1968. Receptors sensitive to carbon dioxide in lungs of chicken. Science. 162:1499‑1501. Petheo, G.L., Molnar, Z., Roka, A., Makara, J.K., and Spat, A. 2001. A pH-sensitive chloride current in the chemoreceptor cell of rat carotid body. J. Physiol. 535(Pt 1):95-106. Pilarski, J.Q., and Hempleman, S.C. 2006. Imidazole binding reagent diethyl pyrocarbonate (DEPC) inhibits avian intrapulmonary chemoreceptor discharge in vivo. Respir. Physiol. Neurobiol. 150(2-3):144-154. Pilarski, J.Q., and Hempleman, S.C. 2007. Development of intrapulmonary chemoreceptors. Respir. Physiol. Neurobiol. 157(2-3):393-402. Powell, F.L. 2000. Respiration. In: Sturkie’s Avian Physiology, 5th Ed, G.C. Whittow (Ed.), Academic Press, San Diego, CA, pp. 233-264. Powell, F.L., Fedde, M.R., Gratz, G.K., and Scheid, P. 1978. Role of avian intrapulmonary chemoreceptors in the ventilatory response to inhaled CO2: effects of acetazolamide in the duck. In: Respiratory Function in Birds, Adult and Embryonic, J. Piiper (Ed.), Berlin, Springer‑Verlag, pp. 164-167.

Mechanisms of CO2 Sensing in Avian Intrapulmonary Chemoreceptors  |  233 Powell, F.L., Geiser, J., Gratz, R.K., and Scheid, P. 1981. Airflow in the avian respiratory tract: variations of O2 and CO2 concentrations in the bronchi of the duck. Respir. Physiol. 44:195‑213. Putnam, R.W., Filosa, J.A., and Ritucci, N.A. 2004. Cellular mechanisms involved in CO2 and acid signaling in chemosensitive neurons. Am. J. Physiol. Cell Physiol. 287(6):C1493-C1526. Reeves, R.B. 1972. An imidazole alphastat hypothesis for vertebrate acid-base regulation: tissue carbon dioxide content and body temperature in bullfrogs. Respir. Physiol. 14:219236. Richerson, G.B. 1995. Response to CO2 of neurons in the rostral ventral medulla in vitro. J. Neurophysiol. 73(3):933‑944. Scheid, P., and Piiper, J. 1972. Cross-current gas exchange in avian lungs: effects of reversed parabronchial air flow in ducks. Respir. Physiol. 16(3):304-312. Scheid, P., and Piiper, J. 1986. Control of breathing in birds. In: Handbook of Physiology, Sec. 3: The Respiratory System, Vol II, Part 2, A.P. Fishman, N.S. Cherniack, J.G. Widdicombe, and S.R. Geiger (Eds.), American Physiological Society, Bethesda, pp. 815‑832. Scheid, P., Kuhlmann, W.D., and Fedde, M.R. 1977. Intrapulmonary receptors in the tegu lizard: II. Functional characteristics and localization. Respir. Physiol. 29:49‑62. Scheid, P., Gratz, R.K., Powell, F.L., and Fedde, M.R. 1978. Ventilatory response to CO2 in birds. II. Contribution by intrapulmonary chemoreceptors. Respir. Physiol. 35:361‑372. Schmidt-Nielsen, K. 1984. In: Scaling: Why is Animal Size so Important? Cambridge University Press, New York. Shoemaker, J.S., and Hempleman, S.C. 2001. Avian intrapulmonary chemoreceptor discharge rate is increased by anion exchange blocker “DIDS”. Respir. Physiol. 128:195-204. Sterling, D., Reithmeier, R.A.F., and Casey, J.R. 2001. A transport metabolon. Functional interation of carbonic anhydrase II and chloride/bicarbonate exchangers. J. Biol. Chem. 276(51):47886-47894. Tallman, R.D., Jr., and Grodins, F.S. 1982. Intrapulmonary CO2 receptor discharge at different levels of venous PCO2. J. Appl. Physiol. 53(6):1386‑1391. Tallman, R.D., Jr., Ghazanshahi, S.D., and Khoo, M.C.K. 1996. Transduction dynamics of intrapulmonary CO2 chemoreceptors. Respir. Physiol. 106(2):115‑125. Tazawa, H. 1981. Compensation of diffusive respiratory disturbances of the acid-base balance in the chick embryo. Comp. Biochem. Physiol. A 69:333-336. Tazawa, H. 1982. Regulatory process of metabolic and respiratory acid-base disturbances in embryos. J. Appl. Physiol. 53:1449-1454. Tazawa, H. 1986. Acid-base equilibrium in birds and eggs. In: Acid-Base Regulation in Animals, N. Heisler (Ed.), Elsevier, Amsterdam, pp. 203-233. Tazawa, H. 1987. Embryonic respiration. In: Bird Respiration Vol. II, T.J. Seller (Ed.), CRC Press, Boca Raton, FL, pp. 3-41. Tazawa, H., Mikami, T., and Yoshimoto, C. 1971. Respiratory properties of chicken embryonic blood during development. Respir. Physiol. 13:160-170. Vince, M.A., and Tolhurst, B.E. 1975. The establishment of lung ventilation in the avian embryo: the rate at which lungs become aerated. Comp. Biochem. Physiol. A 52(2):331-337. Visschedijk, A.H.J. 1968a. The air space and embryonic respiration. 1. The pattern of gaseous exchange in the fertile egg during the closing stages of incubation. Brit. Poul. Sci. 9:173-184. Visschedijk, A.H.J. 1968b. The air space and embryonic respiration. 3. The balance between oxygen and carbon dioxide in the air space of the incubating chicken egg and its role in stimulating pipping. Brit. Poul. Sci. 9:197-210.

234  |  Airway Chemoreceptors in the Vertebrates Vivoni, M., Castro, M.I., Lojo, L., and Furilla, R.A. 2001. Airway ammonia negates the normal ventilatory response to airway CO2 in garter snakes. Respir. Physiol. 127(1):13-22. Wellner-Kienitz, M.-C., and Shams, H. 1998a. Hyperpolarization-activated inward currents contribute to spontaneous electrical activity and CO2/H+ sensitivity of cultivated neurons of fetal rat medulla. Neuroscience. 87(1):109-121. Wellner‑Kienitz, M.‑C., Shams, H., and Scheid, P. 1998b. Contribution of Ca++‑activated K+ channels to central chemosensitivity in cultivated neurons of fetal rat medulla. J. Neurophysiol.70:2885‑ 2894.

Pulmonary Neuroepithelial Cells in Mammals: Structure, Molecular Markers, Ontogeny and Functions 11. Diverse and Complex Airway Receptors in Rodent Lungs  235-268 Inge Brouns, Isabel Pintelon, Ian de Proost, Jean-Pierre Timmermans and Dirk Adriaensen 12. Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies E. Cutz, W.X. Fu, H. Yeger and C.A. Nurse

268-290

13. Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System in the Developing Mammalian Lung 291-310 H. Yeger, J. Pan and E. Cutz 14. Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators: Some New Findings and a Review of the Literature Alfons T.L. Van Loomel, Tania Bollé and Peter W. Hellings 15. Neuroepithelial Bodies and Carotid Bodies: A Comparative Discussion of Pulmonary and Arterial Chemoreceptors Alfons T.L. Van Lommel

311-330

331-358

11 Diverse and Complex Airway Receptors in Rodent Lungs Inge Brouns, Isabel Pintelon, Ian De Proost, Jean-Pierre Timmermans and Dirk Adriaensen*

Abstract Nowadays, it is clear that pulmonary neuroepithelial bodies (NEBs) are organized as integrated receptor complexes. The groups of pulmonary neuroendocrine cells that are able to store and release transmitters upon appropriate stimulation are, apart from their thin apical processes, crowned by specialized Clara-like cells in most mammalian species. The profuse contacts of many different populations of sensory and motor nerve fibers with NEB cells strongly suggests that NEBs are able to transduce sensory information and conduct it to the CNS, while the activity of NEB cells can be modulated via autocrine (NEB cells), paracrine (Clara-like cells) and neurocrine (innervation) pathways.   This chapter summarizes our current knowledge of the origin, neurochemical coding and morphology of the selective innervation of pulmonary NEBs in rodents (rats/mice) against a background of ultrastructural information, thereby providing supporting evidence for NEBs being diverse and complex airway receptors with the capacity of chemo- and/or mechanoreceptors.

Introduction Recent functional morphological data about the innervation of pulmonary neuroepithelial bodies (NEBs) in rodents suggest that NEBs represent an extensive

Laboratory of Cell Biology and Histology, Department of Veterinary Sciences, University of Antwerp, Groenenborgerlaan 171, BE-2020 Antwerp, Belgium *Author for Correspondence.

236  |  Airway Chemoreceptors in the Vertebrates population of very complex intraepithelial receptors. Although the physiological significance of the complex innervation pattern of NEBs is still an enigma, their connections with sensory nerve terminals, and therefore the sensory nature of NEBs, is now beyond dispute. A few years ago, pulmonary NEBs were added to the list of presumed electrophysiologically identified afferent receptors in the lower airways, which until then included slowly adapting stretch receptors (SARs), rapidly adapting stretch receptors (RARs) and C-fiber receptors only (Widdicombe, 2001). As early as 1949 (Fröhlich), a (chemo)receptor function was suggested for NEBs because of their close association with nerve terminals. At the end of the 20th century, a possible dual role for NEBs in healthy lungs was proposed (Sorokin and Hoyt, 1990, 1993): (1) during early stages of lung organogenesis, pulmonary neuroendocrine cells (PNECs) acting via their amine and peptide transmitters would function as local modulators of lung growth and differentiation; (2) later in fetal life and post-natally, PNECs and, particularly, innervated NEBs would play a role as airway chemoreceptors. Recently, some of the functional facets of the pulmonary neuroendocrine system have been reviewed extensively (Adriaensen et al., 2003, 2006; Linnoila, 2006), including a potential novel role of PNECs/NEBs as guardians of lung stem cell niches (Linnoila, 2006). Because of the presumed receptor function of NEBs, the current chapter will summarize our present knowledge on the nervous connections of pulmonary NEBs in rats and mice. The chemical coding, exact location and origin of the nerve terminals in contact with pulmonary NEBs will be outlined. A short overview of literature data on the ultrastructural characteristics of pulmonary NEBs in rodents is intended to help to put the data in their correct perspective. Finally, the functional implications of our recent findings, and the findings of others, on the possible working mechanisms of pulmonary NEBs in rodents will be evaluated.

General Morphological Features and Ultrastructural Characteristics of Pulmonary Neuroepithelial Bodies in Rats and Mice More than 30 years ago, the presence of pulmonary NEBs (Lauweryns et al., 1972) in rats (Cutz et al., 1974) and mice (Hung et al., 1973) was reported. The rather recent discovery of NEBs in the pulmonary epithelium (in comparison to other specific pulmonary cell types) may be explained by the fact that they are highly scattered and, in many species, are not easily detected without specific staining. Early observations of NEBs in rodents were based on the ultrastructural characteristics of NEB cells (Hung et al., 1973; Cutz et al., 1974; Jeffery and Reid, 1975), or on silver staining (Wasano, 1977; Hung, 1984) and formaldehyde-induced fluorescence (Hage, 1976). Fortunately, advances in the methods for immunolabeling

Diverse and Complex Airway Receptors in Rodent Lungs  |  237

substances in tissue sections in the early 1980s resulted in powerful tools for those interested in the pulmonary diffuse neuroendocrine system (Lauweryns et al., 1982) and have greatly increased our knowledge of the distribution and “chemical coding” of PNECs/NEBs. As in most other species, NEBs in mice and rats are found in bronchi, bronchioles and respiratory areas. Whereas NEBs in mice were reported to be preferentially located at intrapulmonary airway bifurcations (Wasano, 1977; Wasano and Yamamoto, 1981), such a favourite position is less apparent for NEBs in (newborn) rats (Carabba et al., 1985; Gomez-Pascual et al., 1990; De Proost et al., 2007). The total number of pulmonary NEBs in different rat strains has been estimated to vary between 3000 and 4000 (Van Genechten et al., 2004). Ultrastructurally, pulmonary NEBs in rodent lungs occur as small, well-defined clusters of PNECs in which all constituent cells rest on the basement membrane. The apical poles of PNECs are covered by a unicellular layer of flattened non-ciliated epithelial cells, the so-called Clara-like cells. Only a few narrow pores between these cells allow direct communication between the NEB cells and the airway lumen, via slender apical processes (Hung and Loosli, 1974; Van Lommel and Lauweryns, 1993). The most important ultrastructural feature for the indisputable identification of PNECs is the demonstration of dense-cored vesicles (DCVs), which are secretory granules that consist of an electron-dense core, surrounded by a limiting membrane that is separated from the core by a clear space. Transmission electron microscopical studies confirmed the presence of DCVs in the subnuclear cytoplasm of PNECs in rats (Van Lommel and Lauweryns, 1993) and mice (Hung and Loosli, 1974) and were even able to provide indirect evidence for the release of the content of DCVs by exocytosis (Van Lommel and Lauweryns, 1993). The released transmitters may bind to structures in very close proximity to NEBs, e.g., smooth muscle cells and NEB-associated nerve terminals, or may be taken up in the bloodstream by fenestrated capillaries that are sometimes found less than 1 µm from the base of NEBs (Van Lommel and Lauweryns, 1993).

Secretory Products and Cytoplasmic Contents of Pulmonary Neuroendocrine Cells in Rodent Lungs After routine fixation and light microscopical staining, only very few NEBs can be recognized in the brightfield microscope, and none of them unambiguously. Numerous methods have, however, been developed to selectively identify PNECs and NEB cells. For a historical overview of the most relevant approaches to examine PNECs/NEBs, we would like to refer to other reviews (Scheuermann, 1987; Sorokin and Hoyt, 1989; Adriaensen and Scheuermann, 1993; Adriaensen et al., 2003). Nowadays, immunohistochemical methods using fluorescent labels, visualized by fluorescence microscopy or confocal microscopy, are widely used to detect pulmonary NEBs. While immunohistochemistry can determine which transmitters are stored in NEB cells, at

238  |  Airway Chemoreceptors in the Vertebrates least some of the detected proteins can also serve to specifically “mark” NEBs when multiple immunohistochemical stainings are performed. One of the routinely used NEB marker molecules is protein gene-product 9.5 (PGP9.5), a pan-neuronal and neuroendocrine marker of the carboxyl-terminal ubiquitin hydrolase family (Thompson et al., 1983; Wilkinson et al., 1989). Pulmonary NEBs in different rat (Lauweryns and Van Ranst, 1988b; Van Genechten et al., 2004) and mouse strains (Lauweryns and Van Ranst, 1988b) could selectively be detected (Figures 3a, 4, 6a). Also, neuron-specific enolase (NSE), a glycolytic enzyme localized primarily to the neuronal cytoplasm, can be used to detect NEB cells in rodents (Sheppard et al., 1982; Cole et al., 1982). Aromatic L-amino acid decarboxylase, an enzyme that catalyzes the decarboxylation of all L-aromatic amino acids, has been reported in NEBs of mice and rats (Lauweryns and Van Ranst, 1988a). Recently, synaptic vesicle protein 2 (SV2) was demonstrated to be a pan-neural marker for NEB cells in mice (Pan et al., 2006). Antibodies against the calcium-binding protein calbindin D28k (CB) can be used to label all cells in rat NEBs (Figure 2a) (Brouns et al., 2000), whereas in mice only a minor part of the NEB cells can be labeled with this marker (Brouns et al., 2009). Positive staining with antibodies against vesicular acetylcholine transporter (VAChT), a protein present in the membrane of cholinergic secretory vesicles, suggested that NEB cells in rodents may have the ability to store and release acetylcholine (Brouns et al., 2002b). In rabbits, the presence of serotonin (5-hydroxytryptamine; 5-HT) and the release of 5-HT after stimulation of NEB cells has been described using different experimental set-ups. Regardless of some sparse reports (Luts et al., 1991; Verástegui et al., 1997a), it appears to be rather difficult to detect 5-HT in rodent NEB cells due to their low 5-HT concentrations (own unpublished observations; Cutz et al., 1974; Wasano, 1977; Gomez-Pascual et al., 1990). In contrast to humans and primates, where gastrin-releasing peptide (GRP) is a major neuropeptide produced by PNECs and NEBs (Linnoila, 2006), rodent PNECs and NEBs store and release calcitonin gene-related peptide (CGRP) as the most prominent transmitter (Uddman et al., 1985; Cadieux et al., 1986; Lauweryns and Van Ranst, 1987; Luts et al., 1991; Shimosegawa and Said, 1991a; Verástegui et al., 1997a). In rats, calcitonin has also been observed in almost all NEB cells (Gosney and Sissons, 1985; Luts et al., 1991), sometimes even completely colocalized with CGRP (Shimosegawa and Said, 1991a), whereas in mice calcitonin was only observed in some NEB cells (Luts et al., 1991). Helodermin, a member of the vasoactive intestinal polypeptide (VIP) family, is found in some cells in rodent NEBs (Luts et al., 1991, 1994). While even recent reviews do not take the purine transmitters of NEBs into account, quinacrine histochemistry (Olson et al., 1976; Crowe and Burnstock, 1981) has unambiguously revealed the accumulation of quinacrine in rodent pulmonary

Diverse and Complex Airway Receptors in Rodent Lungs  |  239

NEBs, strongly suggesting that ATP is accumulated in secretory vesicles of NEBs (Brouns et al., 2000, 2009).

Selective Innervation of Neuroepithelial Bodies Long before the existence of pulmonary NEBs was known, several authors described intraepithelial varicose nerve terminals that were concentrated in groups, irregularly distributed along the airways of different animal species (Berkley, 1894; Larsell, 1921; Larsell and Dow, 1933; Elftman, 1943). Since that time, many researchers have observed an unquestionable innervation of NEBs in both light and electron microscopical investigations. For visualization of nerve fibers that contact pulmonary NEBs, an unambiguous and simultaneous identification of PNECs and nerves is, however, essential (for methodological overview, see Adriaensen et al., 2003).

Observations in Rodents at the Electron Microscopical Level Most of the earlier information about the innervation of NEBs has been obtained using transmission electron microscopy (TEM). Different types of morphologically characterized nerve terminals have been described in contact with NEBs in mice (Hung et al., 1973; Hung and Loosli, 1974; Wasano, 1977; Wasano and Yamamoto, 1981) and rats ( Jeffery and Reid, 1973; Cutz et al., 1974; Hung, 1984; Carabba et al., 1985; Lauweryns and Van Ranst, 1987; Van Lommel and Lauweryns, 1993). The most often reported type of innervation of NEBs in rodents consists of nerve fibers that are observed to penetrate the basement membrane as unmyelinated processes that widen and enter into the intercellular space between the NEB cells. The terminals of these nerve fibers are packed with numerous mitochondria, suggesting, on morphological grounds, that these are afferent nerve endings. Locally, the nerve endings are seen to form synapses with NEB cells (Wasano, 1977; Wasano and Yamamoto, 1981; Van Lommel and Lauweryns, 1993). At the level of these asymmetric synaptic contacts, DCVs accumulate near electron-dense cone-shaped thickenings of the surface membrane of the NEB cells. This type of synapse is indicative for signals passing from NEB cell to nerve ending, implying afferent signaling to the central nervous system (CNS). Frequently, these nerve endings can be found between the apex of NEB cells and the covering Clara-like cells (Van Lommel and Lauweryns, 1993). No direct contacts between the nerve endings in NEBs and the airway lumen, via the pores between the Clara-like cells that overlay the NEB cells, were ever observed. In favorable sections, nerve endings were seen to form “loops” over some NEB cells (Van Lommel and Lauweryns, 1993). The intraepithelial nerve terminals were reported to disappear after infranodosal, but not supranodosal, vagotomy (Van Lommel and Lauweryns, 1993).

240  |  Airway Chemoreceptors in the Vertebrates Much more rarely, intraepithelial nerve endings with small clear cholinergic-type synaptic vesicles could be observed (Van Lommel and Lauweryns, 1993). These morphological efferent-like nerve endings were regarded as axon-collaterals of the intraepithelial (sensory) nerve endings (Scheuermann et al., 1989; Van Lommel and Lauweryns, 1993). The two kinds of synaptic regions could be seen side by side in a single nerve terminal and were referred to as reciprocal synapses (for review see Scheuermann, 1987). Without a doubt, TEM has offered a good morphological characterization of the direct innervation of pulmonary NEBs. However, because of the preparation techniques intrinsic to TEM, only a limited number of NEBs has been examined, giving rise to incomplete and conflicting data and interpretations. More recently, techniques were developed that allow the simultaneous observation of numerous NEBs, which has made it possible to answer important questions, including that of the origin of the nerve fiber populations that selectively innervate NEBs. Immunocytochemical stainings with general neuronal and neuroendocrine markers, such as PGP9.5 (Lauweryns and Van Ranst, 1988b; Larson et al., 2003), NSE or SV2 (Pan et al., 2006), have been shown to be very useful for obtaining a general idea of the overall innervation of pulmonary NEBs but do not allow differentiation between NEB cells and their associated nerve terminals (Figures 2a, 3a, 4, 6a).

Selective Innervation of Neuroepithelial Bodies in Rat Lungs Until recently, NEBs have been suggested to be predominantly, if not exclusively, contacted by vagal nodose sensory nerve terminals (Cutz and Jackson, 1999; Van Lommel et al., 1999; Bollé et al., 2000; Widdicombe, 2001), and to receive sensory nerve endings only (Van Lommel et al., 1998). Considering what was known in the late 1990s, it was clear that an essential factor in the recognition of NEBs as airway receptors would be the full confirmation and characterization of their vagal nodose innervation. (For a summary of selective innervation of NEBs in rat lungs, see Figure 1.) Our current PNEC/NEB research, focusing on the overall innervation pattern, has been based on combinations of chemical or mechanical denervation, ultrasensitive immunohistochemistry, tracing studies and confocal microscopy.

Vagal Sensory Connections

To directly address the question whether pulmonary NEBs in rats were contacted by vagal sensory nerve terminals, the red fluorescent lipophilic carbocyanine tracer DiI (Honig, 1993) was injected in the nodose ganglion. In this way, indisputable evidence of a vagal sensory innervation of pulmonary NEBs was provided at the light microscopical level (Van Lommel et al., 1998; Adriaensen et al., 1998). DiItraced vagal afferent nodose nerve endings were seen to ramify between NEB cells in a candelabrum-like pattern and were correlated to the intraepithelial mitochondria-

Diverse and Complex Airway Receptors in Rodent Lungs  |  241

Figure 1: Diagram of the complex innervation pattern of neuroepithelial bodies (NEB) and of the main innervation of airway smooth muscle in rat lungs. The center of the figure represents a pulmonary NEB (colored yellow), covered by Clara-like cells (dark grey) and its extensive interactions with different nerve fibre populations (color-coded). Known characteristics of the represented neuronal populations and the NEB are included in the figure in the same color as the respective structures. The lower part of the figure shows airway smooth muscle that receives nerve terminals from postganglionic parasympathetic neurones located in an airway ganglion (cholinergic neurons = purple). Laminar nerve terminals of smooth muscle–associated airway receptors (SMARs; colored green) intercalate between the smooth muscle cells.   φ, diameter; aSMA, a smooth muscle actin; 5HT, serotonin; CALC, calcitonin; CB, calbindin D28k; CGRP, calcitonin gene-related peptide; CRT, calretinin; MBP, myelin basic protein; nNOS, neuronal nitric oxide synthase; NPY, neuropeptide Y; SP, substance P; TH, tyrosine hydroxylase; TRPV1, transient receptor potential vanilloid 1 receptor (capsaicin receptor); VAChT, vesicular acetylcholine transporter; VGLUT, vesicular glutamate transporter; VIP, vasoactive intestinal polypeptide.

242  |  Airway Chemoreceptors in the Vertebrates rich afferent nerve terminals (Van Lommel et al., 1998), the most constantly reported innervation of NEBs. Neuronal tracing is, however, a labor-intensive and time-consuming technique and hardly compatible with the routine identification of nerve fiber populations. Therefore, we further characterized the neurochemical coding of this vagal nodose population of nerve terminals in NEBs. Combinations of immunocytochemistry and unilateral infranodosal vagal denervation experiments revealed that the vagal nodose nerve fibers contacting pulmonary NEBs can be selectively labeled with antibodies against the calcium-binding protein CB (Figure 2a) (Brouns et al., 2000). Since CB labels NEB cells as well as the nodose fibers contacting them (Figure 2), it can be regarded as a very interesting routine marker for NEBs in rat lungs but does not allow a clear evaluation of the intraepithelial terminals of the vagal afferents. Recently, new markers became available that were demonstrated to selectively label populations of sensory nerve terminals. Multiple immunostaining with antibodies against vesicular glutamate transporters (VGLUTs), transmembrane proteins that load glutamate into synaptic vesicles, revealed that intraepithelial sensory nerve terminals in pulmonary NEBs invariably express VGLUTs (Brouns et al., 2004, 2006a, 2006b). Both VGLUT1 and VGLUT2 immunoreactivity (IR) (Figures 2b, 5, 6b) were detected in intraepithelial nerve terminals that appeared to extensively contact most cells in an NEB, and almost all of the NEBs (Brouns et al., 2004, 2006a, 2006b). Combinations of VGLUT1 and VGLUT2 immunostaining have, however, been technically difficult. For indisputable staining of intraepithelial sensory nerve terminals in rat pulmonary NEBs, VGLUT2 appeared to be the marker of choice. Since the expression of VGLUTs is, at present, regarded as an unequivocal identification of the glutamatergic nature of neurons (Takamori, 2006), the detection of VGLUT1 or VGLUT2 in profuse nerve fiber populations that selectively contact pulmonary NEBs in rats should be considered to reflect the capacity of these terminals to store and release glutamate as a neurotransmitter at the level of NEB cells. The vagal nature of the intraepithelial nerve terminals was demonstrated by the loss of intraepithelial VGLUT-immunoreactive (-ir) nerve terminals in NEBs in the ipsilateral lung after unilateral infranodosal vagotomy. Apart from a very small population, all vagal sensory VGLUT-ir nerve fibers also expressed CB IR. Further neurochemical identification of the vagal nodose nerve fiber population showed that two different VGLUT/CB-ir nerve endings have to be concerned, which both reveal extensive intraepithelial terminals in NEBs: (1) those additionally expressing P2X3 receptors and (2) those additionally expressing Na+/K+-ATPase a3 (Figure 6c). When both types of terminals are present in the same pulmonary NEB, they seem to typically innervate separate groups of NEB cells. P2X3 purinoreceptors (belonging to the family of ligand-gated ATP receptors) are present on the vagal nodose nerve fiber population that selectively contacts

Diverse and Complex Airway Receptors in Rodent Lungs  |  243

Figure 2: Pulmonary neuroepithelial body (NEB) in a rat bronchus, double immunostained for calbindin D28k (CB; a) and vesicular glutamate transporter 2 (VGLUT2; b). a. A group of CB-immunoreactive (ir) NEB cells is contacted by a CB-ir vagal sensory nerve fiber (arrow). b. The CB-ir nerve fiber (a) is also VGLUT2 ir (arrow), approaches the epithelium (E) and gives rise to an extensive arborization (arrowheads) of glutamatergic nerve terminals that surround the NEB cells. Maximum value projections of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). L, airway lumen.

NEBs (Brouns et al., 2000; Van Genechten et al., 2004). Combination of quinacrine histochemistry, to label ATP-storing cells, and P2X3 receptor-staining showed that the ATP receptor-expressing vagal sensory nerve terminals in rat lungs are specifically associated with quinacrine-stained NEB cells (Brouns et al., 2000). Intraepithelial vagal sensory nerve terminals expressing the plasma membrane sodium/potassium exchanging protein subunit Na+/K+-ATPase a3, a newly described marker for (mechano)sensory nerve terminals (Dobretsov et al., 2003), can be double stained with antibodies against VGLUTs (Figure 6). However, not all VGLUT-ir nerve terminals express Na+/K+-ATPase a3 (Figure 6). Na+/K+-ATPase a3-ir nerve endings appear to form unique “caps” over the apical pole of NEB cells (Figure 6). For the interpretation of pulmonary NEBs as sensory airway receptors, a functionally very important feature of the sensory component of the innervation of NEBs is the presence of myelin sheaths. Myelin basic protein (MBP) immunostaining showed that in adult rat lungs the vagal nodose sensory nerve fibers that contact NEBs are surrounded by myelin sheaths that are lost in the immediate neighborhood of the NEB, just before branching of the nerve terminals (Brouns et al., 2000, 2003, 2004, 2006a, 2006b). The latter myelinated vagal sensory nerve fibers have diameters ranging from 1 to 3.5 µm. Although myelinated nerve fibers had been reported in the vicinity of NEBs using TEM (Van Lommel and Lauweryns, 1993), MBP immunostaining provided the first evidence for a direct link between the myelinated fibers and the two populations of vagal nodose intraepithelial nerve terminals in NEBs. The vanilloid capsaicin is known to establish a long-lasting depletion of certain populations of afferent nerve fibers (for review see Holzer, 1991)—also in rat lungs

244  |  Airway Chemoreceptors in the Vertebrates (Cadieux et al., 1986; Martling et al., 1988; Shimosegawa and Said, 1991b). Systemic treatment of rats with capsaicin revealed no changes in the CB-ir vagal nodose innervation of NEBs as compared to control rats, strongly suggesting that a capsaicininsensitive population is concerned. Double staining for the capsaicin receptor “transient receptor potential vanilloid 1 (TRPV1)” and CB confirmed that the vagal nodose component of the innervation of NEBs in rats does not express the capsaicin receptor (Brouns et al., 2003).

CGRP-Immunoreactive Component

For many years now, CGRP-positive nerve fibers have been reported to contact CGRPir NEBs at all levels of rat intrapulmonary airways (Cadieux et al., 1986; Shimosegawa and Said, 1991b; Terada et al., 1992). It has been generally believed that those CGRP fibers represented the vagal sensory fibers that were long predicted. However, retrograde tracing from the lungs and denervation studies (Springall et al., 1987; own unpublished observations) revealed that the CGRP-ir nerve fiber population of rat pulmonary NEBs is non-vagal (Adriaensen et al., 1998; Brouns et al., 2003). CGRPir nerve fibers that selectively contact NEBs in rat lungs belong to a spinal sensory population that originates from thoracic dorsal root ganglia (DRG) (Springall et al., 1987; own unpublished observations). Double immunocytochemical stainings for CGRP and the above-mentioned markers for the vagal sensory connections of NEBs have proven that the vagal nodose sensory component of the innervation of NEBs consists of nerve fiber populations that are clearly different from the sensory CGRP-ir fibers (Adriaensen et al., 1998, 2006; Brouns et al., 2000, 2004). CGRP-positive nerve terminals in contact with rat pulmonary NEBs appear as thin varicose nerve endings. One of the important features of these CGRP-ir nerve endings is the observation that they all seem to colocalize substance P (SP), making CGRP/ SP double labeling a valid tool to differentiate between the individual CGRP+/SP+ nerve terminals and the NEB cells, which in rats store CGRP but not SP (Polak et al., 1993). Confocal microscopy revealed that in rats the CGRP+/SP+ nerve terminals do not penetrate the epithelium, but rather form a nervous plexus at the basal pole of the NEBs (Brouns et al., 2003; Van Genechten et al., 2004). The thin, varicose spinal CGRP+/SP+ nerve fibers that selectively contact NEBs were further shown to express VGLUT2 IR (Brouns et al., 2004), but with varying intensities both within single fibers and between fibers. Also, a very weak CB IR was observed in the CGRP+/SP+ nerve fibers. Since very sensitive immunolabeling procedures are required for the detection of VGLUT2 and CB, CGRP and/or SP immunostaining appears to be the method of choice for visualizing the spinal sensory connections of rat NEBs. After capsaicin treatment, the percentage of NEBs contacted by CGRP/SPpositive nerve terminals was dramatically reduced compared to control lungs, while the number of CGRP-ir NEBs remained unchanged (Brouns et al., 2003). All CGRP-

Diverse and Complex Airway Receptors in Rodent Lungs  |  245

ir nerve fibers in the vicinity of and contacting NEBs appeared to express TRPV1 and may therefore be considered capsaicin sensitive. In capsaicin-treated lungs, NEB cells, which do not express TRPV1, did not show obvious differences to control lungs (Shimosegawa and Said, 1991b; Brouns et al., 2003). All thin varicose CGRP+/SP+-fibers contacting NEBs were found to be unmyelinated. Taken together with all the above-mentioned features, the spinal sensory nerve fibers in contact with pulmonary NEBs reveal obvious C-fiber characteristics (Adriaensen et al., 2003). In addition, however, rat airways harbor a vagal CGRP/SP-ir nerve fiber population that, in contrast to the vagal nodose fibers that were described to contact NEBs, mainly originates from the jugular ganglia (Mazzone, 2005). The latter vagal C-fiber-like nerve terminals can be found in the epithelium of large diameter bronchi only, apparently without any specific relationship to NEBs (own unpublished observations).

Efferent Innervation

The functional morphology of rat pulmonary NEBs is clearly much more complex than could ever be predicted by electron microscopic studies alone (Van Lommel and Lauweryns, 1993). Detailed studies over the last 5 years have indicated that several other nerve fiber populations, some of which are not yet fully characterized, provide additional nervous connections to NEBs. Immunodetection of the NO-synthesizing enzyme neuronal nitric oxide synthase (nNOS) revealed that part of the pulmonary NEBs in rats are selectively innervated by nNOS-ir (nitrergic) nerve fibers (Brouns et al., 2002a, 2002b) . The latter were observed to give rise to extensive intraepithelial nerve terminals in pulmonary NEBs. Apparently, these nerve terminals originate from a non-cholinergic population of nitrergic neurons, located in the lamina propria of intrapulmonary bronchi and bronchioles (Brouns et al., 2002a). These intrinsic nNOS-ir neurons were shown to colocalize IR for vasoactive intestinal polypeptide (VIP) (Adriaensen et al., 2003). The nNOS+ neuronal cell bodies were shown to be invariably surrounded by a basket of CGRP-ir nerve terminals that seemed to originate from CGRP-ir fibers that spiral around the axons of the nitrergic neurons (Brouns et al., 2002a; Van Genechten et al., 2004). nNOS IR was absent from the spinal afferent and from the vagal nodose afferent nerve fiber populations that selectively contact NEBs. Quantitative analysis revealed that all NEBs receiving nNOS-ir terminals were also contacted by spinal sensory CGRP-ir nerve fibers, while only about half of them were additionally contacted by vagal nodose fibers (Brouns et al., 2002a). Since nitrergic neuronal cell bodies are always surrounded by a basket of CGRP-ir nerve terminals, presumably collaterals of the spinal afferent nerve fibers contacting NEBs, a direct relationship between the nitrergic and spinal sensory nerve fiber populations contacting NEBs has been suggested (Brouns et al., 2002a; Adriaensen et al., 2003; Van Genechten et al., 2004),

246  |  Airway Chemoreceptors in the Vertebrates whereas the correlation between nitrergic and vagal sensory nerve terminals is less obvious. In Fawn-Hooded rats, a model for primary pulmonary hypertension, the number of intrinsic pulmonary nitrergic neurons, and the percentage of pulmonary NEBs revealing a nitrergic innervation, have been shown to be significantly lower (Van Genechten et al., 2003, 2004). A considerable number of NEBs appeared to be contacted by profuse beaded VIPir intraepithelial nerve terminals (Adriaensen et al., 2003; Van Genechten et al., 2004). Although VIP IR was also shown to be localized in the intrinsic nitrergic neurons that give rise to the nitrergic terminals contacting NEBs, we believe that an additional population of VIP-expressing nerve endings with a so far unidentified origin may be involved in the selective innervation of NEBs. As already mentioned earlier, electron microscopy showed that NEBs in rat lungs are contacted by nerve endings that contain small electron-lucent cholinergic-like synaptic vesicles and often reveal synaptic contacts with NEB cells (Adriaensen and Scheuermann, 1993). Antibodies against VAChT were, therefore, used to visualize possible cholinergic nerve terminals in direct relation to NEBs (Pintelon et al., 2003; Van Genechten et al., 2004). Weakly VAChT-ir intraepithelial cell groups, characterized as NEBs after multiple immunostaining, indeed appeared to be contacted by VAChT-ir nerve fibers. Although the latter nerve fiber population is not yet fully characterized, we have evidence suggesting that cholinergic motor fibers, originating from preganglionic parasympathetic neurons, and hence evidently a so far not reported population, may be concerned. Preliminary data further revealed that tyrosine hydroxylase-ir nerve terminals selectively contact some of the rat pulmonary NEBs at their basal pole (Van Genechten et al., 2004; own unpublished observations). These nerve fibers are likely adrenergic and have their origin in sympathetic ganglia.

Selective Innervation of Neuroepithelial Bodies in Mouse Lungs In view of the interesting possibilities of using genetically modified mice for functional studies, detailed knowledge about sensory lung receptors in control mice is essential, but published information is largely lacking today. Neurochemical identification using antibodies against known selective markers for nerve terminals in rat lungs was performed in control and vagotomized mouse lungs and unravelled a complex innervation pattern of pulmonary NEBs in control mice. VGLUT immunostaining, proven to be an excellent marker for sensory nerve terminals in rat lungs, revealed also in mice extensive intraepithelial nerve endings protruding between pulmonary NEB cells. Combination of VGLUT1 and VGLUT2 immunostaining showed that only part of the VGLUT1-ir nerve terminals in NEBs are VGLUT2-ir, and that the staining intensity for VGLUT2 varied both within and between nerve fibers. Infranodosal vagal denervation resulted in the disappearance of

Diverse and Complex Airway Receptors in Rodent Lungs  |  247

VGLUT1-ir nerve terminals in the ipsilateral lung, illustrating a vagal origin for the glutamatergic nerve terminals in mouse pulmonary NEBs. The VGLUT1-ir nerve fibers giving rise to these intraepithelial arborizations could also be marked by their CB IR, making both VGLUT1 and CB immunostaining excellent markers for this vagal sensory nerve fiber population contacting mouse NEBs (Brouns et al., 2009). Multiple immunocytochemical staining and vagal denervation, however, disclosed another vagal sensory nerve fiber population with extensive intraepithelial nerve terminals in mouse NEBs: one that was ir for P2X3 receptors (Figure 3), but not for VGLUT or CB. Since quinacrine histochemistry, indicative for the accumulation of ATP in secretory vesicles, was shown to specifically stain NEB cells, ATP might be considered to act as a neurotransmitter/neuromodulator, via a P2X3 receptormediated pathway, for part of the vagal sensory innervation of mouse NEBs (Brouns et al., 2009). When the two different intraepithelial vagal sensory nerve fiber populations are present in the same NEB, they appear to innervate separate parts of the cluster of NEB cells. The VGLUT/CB-positive populations seem to form cap-like endings over the apical pole of at least a subpopulation of the NEB cells (Brouns et al., 2007). As detected by MBP immunohistochemistry, both vagal sensory nerve fiber populations that give rise to intraepithelial nerve endings in NEBs were myelinated. Multiple immunostainings for markers of the intraepithelial vagal sensory nerve fiber populations and for CGRP showed that the intraepithelial vagal sensory innervation

Figure 3: Pulmonary NEB in a bronchus of a mouse. Double immunocytochemical staining for protein gene product 9.5 (PGP9.5; a) and for the ATP receptor P2X3 (b). a. Extensive bundles of PGP9.5-ir nerve fibers (open arrows) in the vicinity of a PGP9.5-ir NEB give rise to many nerve fibers (arrows) that approach the NEB. b. One of the PGP9.5-ir nerve fibers appears to be P2X3 receptor-positive (arrow) and gives rise to complexly branching intraepithelial P2X3 receptor-ir nerve endings (arrowheads) that innervate part of the NEB. Maximum value projection of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). L, airway lumen; E, airway epithelium.

248  |  Airway Chemoreceptors in the Vertebrates of mouse NEBs does not express CGRP. CGRP-ir nerve fibers were observed in the subepithelial area of NEBs and were rarely found to protrude into the epithelium. Using very sensitive immunodetection, different types of CGRP-ir nerve fibers were seen to approach the NEBs: varicose fibers, comparable to those observed in contact with rat NEBs, and very thin (delicate) fibers. When CGRP immunostainings were combined with SP, a good marker for the spinal sensory nerve terminals in rat NEBs, three different CGRP-ir nerve fiber populations were identified: varicose CGRP+/ SP+, delicate CGRP+/SP+, and delicate CGRP+/SP-. Meticulous observation of single confocal optical sections revealed all of these nerve fiber populations situated close to the basal side of NEBs. Although proposed in literature (Verástegui et al., 1997b), in our hands direct contacts of CGRP-ir terminals with the basal pole of mouse NEBs, similar to those observed in rat NEBs, seem to be lacking. Infranodosal vagal denervation and immunostaining for CGRP clearly showed that the varicose CGRP-ir nerve terminals are vagal in origin, while the delicate CGRP-ir nerve fibers remained unaffected. In mouse lungs, cholinergic nerve terminals could also be identified by their VAChT IR. Beaded subepithelial nerve fibers that innervate the airway smooth muscle were prominent. In pulmonary NEBs, VAChT-ir nerve endings with a much smoother appearance were observed to protrude between the NEB cells. The smooth intraepithelial nerve terminals were never seen to be linked to the beaded subepithelial fibers. The exact nature and origin of the VAChT-ir nerve terminals in mouse NEBs are so far unknown. In mouse lungs, nitrergic nerve cell bodies were found in the adventitial layer of blood vessels, in large parasympathetic ganglia located in the hilar region, in smaller subepithelial ganglia in conducting airways, and even in the alveolar interstitium. Unlike in rats, only a limited number of the nitrergic cell bodies in smaller airway ganglia were surrounded by varicose CGRP-ir nerve fibers. Double labeling for VIP and nNOS revealed that some neuronal cell bodies in the hilar ganglia colocalized VIP and nNOS, while others were only VIP positive. In contrast to our observations of an extensive intraepithelial nitrergic innervation of NEBs in rat lungs, nNOS-ir nerve endings were only observed to contact the base of mouse pulmonary NEBs. Using nNOS/CGRP immunocytochemical double staining, mouse NEBs receiving nitrergic nerve terminals appeared to be invariably associated with varicose CGRP-ir nerve endings.

Functional Implications Neuroepithelial Bodies are Complex Airway Receptors The data presented in the current chapter demonstrate that pulmonary NEBs in rodents are abundantly contacted by different populations of sensory and motor nerve

Diverse and Complex Airway Receptors in Rodent Lungs  |  249

terminals. Although it is very likely that also in other mammals pulmonary NEBs will be complexly innervated, certain species differences can not be excluded (Pan et al., 2004; for review see Sorokin et al., 1997). Mainly because of their (vagal) sensory innervation, NEBs may be considered to represent an extensive population of complex intraepithelial sensory receptors and were therefore added to the list of “airway receptors” a few years ago (Widdicombe, 2001). The extremely complex innervation pattern of pulmonary NEBs suggests a great flexibility in neurally mediated communication and adaptation, potentially enabling multiple functions of NEBs in normal airway physiology in adults. Since the innervation pattern of NEBs has been reported to considerably change during development and in early neonatal life (Brouns et al., 2002a, 2003; Pan et al., 2004), NEBs probably even exert additional functions during these crucial periods of life. It should be stressed that none of the so far characterized nerve fiber populations contacts all pulmonary NEBs. Each population, however, accounts for at least a few hundred receptor points, in this way outnumbering other airway receptors and chemoreceptors in other organs (Sorokin and Hoyt, 1990). The observation that a single pulmonary NEB can be contacted by several nerve fiber populations adds to the assumption that NEBs may host complex functional properties. Moreover, since the innervation pattern apparently is not identical for all NEBs, the extensive group of NEB receptors may be functionally diverse. On the other hand, all NEBs consist of NEB cells/PNECs that harbor characteristic dense-cored secretory granules that store bioactive substances, such as monoamine, peptide and purine transmitters (for reviews see Sorokin and Hoyt, 1989, 1990; Adriaensen and Scheuermann, 1993; Sorokin et al., 1997; Adriaensen et al., 2003). Upon appropriate stimulation, these substances will be secreted and may then interact with NEB-associated nerve terminals, be taken up by nearby blood vessels and exert endocrine interactions, or have paracrine effects on neighboring non-endocrine epithelial cells, fibroblasts, immune cells, airway and vascular smooth muscle. Both light and electron microscopy revealed a close association of NEB cells with different types of nerve terminals. Especially NEBs in contact with sensory nerve fibers may be regarded as transducers of stimuli from their environment into nerve impulses. When local stimuli activate the release of secretory products from NEB cells, afferent nerve fibers in synaptic contact are believed to depolarize and develop a generator potential, which triggers an action potential that reaches the CNS once a certain threshold is reached. However, most transmitters can only influence their target by binding to specific receptors. Vagal sensory components of the innervation of pulmonary NEBs in rodents express P2X3 receptors (Brouns et al., 2000, 2009; Van Genechten et al., 2004). The extensive intraepithelial arborizations of these nerve terminals contact ATP-storing NEB cells, suggesting that ATP secreted by NEB cells may act as a neurotransmitter/neuromodulator in the vagal sensory transduction of NEBs. Given that P2X3 receptors are known as high-threshold receptors, the vagal

250  |  Airway Chemoreceptors in the Vertebrates sensory transduction from NEBs may only be carried out if NEBs receive very powerful, or even noxious, stimuli. The importance of the sensory innervation for pulmonary NEBs is even more stressed by the observation that NEBs may simultaneously be contacted by different populations of myelinated vagal sensory nerve endings. Since the principal role of the myelin sheath is “isolation” of nerve fibers to allow a faster axonal transmission of the nerve impulse, the vagal sensory component of the innervation of NEBs may therefore be regarded as a “fast” connection to the CNS. Spinal CGRP/SP-containing nerve fibers in very close proximity to, or even in contact with, rat pulmonary NEBs express TRPV1 (Brouns et al., 2003), the archetypal member of the vanilloid TRP family, and in this context used as a marker for capsaicinsensitive neuronal populations. Systemic treatment with capsaicin results in a longlasting depletion of the spinal sensory nerve terminals in contact with rat pulmonary NEBs (Cadieux et al., 1986; Shimosegawa and Said, 1991a; Brouns et al., 2003), resulting in a major decrease of the percentage of NEBs that receive CGRP-ir terminals (Brouns et al., 2003). It is, therefore, clear that the CGRP/SP/TRPV1-ir nerve fibers in rat lungs constitute an unmyelinated, capsaicin-sensitive population that reveal typical C-fiber characteristics but, at least for the largest part, do not have the vagal sensory origin that was reported for bronchopulmonary C-fiber receptors. The exact function of TRPV1 ion channels on the CGRP-ir nerve terminals in contact with NEBs is still a matter of speculation, but TRPV ion channels are believed to be potentially chemo- and mechanosensitive or may be involved in the transduction of noxious stimuli (Liedtke and Kim, 2005). In mice, a population of dorsal root ganglion neurons projecting to the lung was also reported to express TRPV1 (Dinh et al., 2004). The observation of varicose CGRP/SP-ir nerve terminals in close proximity to NEBs (Brouns et al., 2003, 2009) seems to be essential for the function of at least part of the NEBs. The NEB-related varicose SP/CGRP fibers appear to have different origins in rats and mice (Brouns et al., 2003, 2009). In addition, sensitive immunostaining techniques revealed in mice very delicate (SP)/CGRP-containing nerve fibers, in close proximity to pulmonary NEBs. Although extensive studies in rodents have demonstrated the presence of CGRP-ir nerve terminals in contact with NEBs, to our best knowledge, we reported for the first time different types of CGRPexpressing nerve fibers in mouse lungs (Brouns et al., 2009). Of great functional interest was the detection of VGLUTs in the majority of sensory nerve fiber populations that contact rodent pulmonary NEBs. The expression of VGLUTs implies the presence of glutamatergic synaptic vesicles, and therefore the possibility of nerve terminals to store and release glutamate as a neurotransmitter (Takamori et al., 2000). While it was already known for the central connections of airway-related primary afferents (Lawrence, 1995; Haxhiu et al., 2005; Lachamp et al., 2006), also the peripheral receptor terminals of these sensory neurons now appear to use the well-known excitatory transmitter glutamate. The presence of glutamatergic

Diverse and Complex Airway Receptors in Rodent Lungs  |  251

synaptic vesicles in sensory nerve endings seems contradictory at first sight but may be regarded as an indication of sensory nerve fibers being able to modulate their own activity, as suggested for the esophagus (Raab and Neuhuber, 2007), or to regulate the activity of the associated receptor cells, as suggested for the carotid body (Ichikawa, 2002), after or during transduction of adequate stimuli to afferent nerve discharges. Pulmonary NEBs contacted by glutamatergic vagal sensory nerve fibers may therefore be independent of additional motor fibers to regulate the excitability of the NEB receptor cells and/or control neurotransmitter release. It has been postulated for many years that a mechanism of local modulation of the neuroreceptor, the so-called “axon reflex” (Adriaensen and Scheuermann, 1993; Lauweryns and Van Lommel, 1986, 1987), might be involved in the activation of rat pulmonary NEBs (Van Lommel and Lauweryns, 1993). These potential “axon reflexes” were defined as mechanisms of local modulation of receptor cells by efferent terminals formed at the periphery of sensory nerve terminals. Besides neuroreceptor modulation, these axon reflexes could serve to integrate sensory input or to initiate a motor response via local neurons without involvement of the CNS. Detailed ultrastructural data suggest that vagal sensory intraepithelial nerve fibers in rodent NEBs carry both afferentlike (mitochondria-rich) and efferent-like (packed with clear synaptic vesicles) nerve endings (Van Lommel and Lauweryns, 1993). The observation of cholinergic (VAChTir) and glutamatergic (VGLUT-ir) terminals inside NEB cell clusters likely represent the light microscopical equivalents of the efferent ultrastructural characteristics of the intraepithelial vagal afferent nerve terminals. In both rats and mice, at least part of the NEBs are innervated by motor components that consist of populations that are clearly separate from the sensory endings. In rats, pulmonary NEBs may be contacted by intraepithelial intrinsic nNOS+/VIP+ nerve terminals and VIP+ terminals, and extrinsic (preganglionic) VAChT+ nerve terminals. Mouse NEBs receive intraepithelial VAChT-ir terminals and basal nitrergic terminals. The fact that motor nerve terminals appear to penetrate between the neuroendocrine cells of NEBs provides evidence for intraepithelial nerve fiber populations, which were only in recent reviews taken into consideration (Adriaensen et al., 2003; Adriaensen and Timmermans, 2004; Linnoila, 2006). The observation that at least a subpopulation of the NEBs actually receives both intrinsic and extrinsic motor endings suggests important additional control mechanisms for the secretory activity of pulmonary NEBs. For the intraepithelial nitrergic nerve terminals in rat NEBs, it was proposed that NO released from the terminals of intrinsic pulmonary nitrergic neurons may exert an inhibitory influence on the sensory dicharge of NEB cells in response to local stimuli. Except for local actions, such a system may keep NEB-receptors “quiet” as far as the CNS is concerned. In this way, signaling to the CNS via vagal or spinal afferent pathways may be limited to powerful stimuli that necessitate actions mediated by the CNS for the regulation of general lung function (Brouns et al., 2002a; Adriaensen et al., 2003).

252  |  Airway Chemoreceptors in the Vertebrates Evidently, pulmonary NEB cells and their associated nerve terminals are excellent candidates to register properties of the airway environment, translate and modulate signals, induce local action, or transduce the stimuli into nerve impulses that reach the CNS when necessary. Although each of the vagal afferents that selectively contact NEBs undoubtedly accounts for many hundreds of receptor sites in lungs, lung physiologists generally believe that no NEB-related activity has so far been measured in vagal afferents projecting to the lungs (Widdicombe, 2001). For many years, evidence has accumulated suggesting that vagotomy abolishes “all” cardio-respiratory reflexes from the lungs. Several recent publications, however, provide evidence for a spinal afferent component in the airway reflexes and/or cardio-respiratory responses to intrapulmonary chemical stimulation (Soukhova et al., 2003; Wang et al., 2003; Oh et al., 2006). The selective C-fiber afferent spinal CGRP/SP-ir innervation of NEBs in rat lungs may well be involved in these pathways. Although it has not yet been established whether stimuli act directly on the NEB cells or on the NEB-related nerve terminals, it is an intriguing observation that all intraepithelial receptor end organs always appear to coincide with the presence of NEBs. The location of pulmonary NEB complexes in the airway epithelium, shielded by Clara-like cells, with the processes of NEB cells in contact with the airway lumen makes NEBs excellent candidates to register properties of the airway environment. Although there is accumulating evidence for a complete functional oxygen-sensing system in PNECs in vitro (see below), it should be stressed that the exact nature of the physiological stimuli of NEB cells in healthy lungs is still unknown.

Neuroepithelial Bodies as Potential Chemoreceptors Since Fröhlich discovered the innervated “corpuscles” of neuroendocrine cells in the pulmonary epithelium in the 1940s, it has been suggested that these structures might function as chemoreceptors (Fröhlich, 1949). The anatomical location of NEBs, and their morphological organization in NEB-complexes, protected by Clara-like cells, with their apical cell processes making contact with the airway lumen, and their extensive innervation, indeed suggest an involvement in sensing alteration in environmental conditions. Although many potential stimuli have been proposed (Sorokin and Hoyt, 1990), a considerable number of studies have been performed—for cigarette smoke and hypoxia only. Indirect research in human smokers and animal models has demonstrated that cigarette smoke induces changes in the pulmonary neuroendocrine system (Tabassian et al., 1988, 1993; Aguayo, 1993). The detection of nicotinic acetylcholine receptors on PNECs (Sekhon et al., 1999; Plummer et al., 2000; Fu et al., 2003) was suggestive of the direct involvement of nicotine, or nicotine derivates such as nitrosamines, in changes in PNEC cells and possibly also in the development of tobacco-associated small lung cell carcinoma (Schuller et al., 2000). However, at the moment there are no

Diverse and Complex Airway Receptors in Rodent Lungs  |  253

data available concerning the sensitivity to nicotine or acute activation by nicotine of NEB cells in rats or mice. Over the last 10 to 15 years, substantial evidence has been collected that argues for an oxygen-sensing capacity of NEB cells as their primary chemoreceptor function. Since a comprehensive description of the oxygen-sensing mechanisms in NEBs will be outlined in another chapter of this book, we will only give a brief overview and will deal with the functional implications of the extensive data available for the innervation of rodent pulmonary NEBs.

Oxygen-Sensing Properties of NEBs in Rodents

As early as the 1970s, a possible relationship between oxygen sensing in the airways and NEBs was suggested (Lauweryns and Cokelaere, 1973). Experiments in neonatal rabbits revealed that exposure to short-term hypoxia, but not to hyperoxia or hypoxemia, resulted in increased exocytosis and decreased amine fluorescence in NEB cells, implying hypoxia-evoked 5-HT secretion (Lauweryns and Cokelaere, 1973). Recently, the improvement of in vitro models has been important for characterizing a functional oxygen-sensing system in NEBs (Cutz et al., 1985, 2004; Speirs et al., 1992; Speirs and Cutz, 1993; O’Kelly et al., 1998; Kemp et al., 2002). In NEBs (Youngson et al., 1993; Fu et al., 1999) and in the immortalized small cell lung carcinoma cell line H146 (O’Kelly et al., 1998, 1999, 2000; Hartness et al., 2001), acute inhibition of K+ channels by hypoxia seems to be central to chemosensing (Youngson et al., 1993; Cutz and Jackson, 1999; O’Kelly et al., 1999; Peers and Kemp, 2001; Kemp et al., 2003). The proposed signaling cascade involves closure of background K+ channels (Fu et al., 1999), consequent membrane depolarization and Ca2+ influx via voltage-gated Ca2+ channels (Cav), eventually triggering neurotransmitter exocytosis from NEB cells (Fu et al., 2002). Both in native NEBs and in PNEC cell lines, a membrane-associated NADPH oxidase has been identified as a potential molecular oxygen sensor (Youngson et al., 1993, 1997; Wang et al., 1996; Fu et al., 2000; O’Kelly et al., 2000), although other oxygen-sensing mechanisms may be involved (O’Kelly et al., 2001). The morphological identification of the P/Q (Cav2.1) voltage-gated Ca2+ channels in the apical membrane of rat pulmonary NEB cells (De Proost et al., 2006) is indicative for Ca2+ triggered exocytosis of transmitters from NEBs. Apart from some contradictory findings, the current consensus is that NEB cells are activated by and appear to adapt to chronic hypoxia (Cutz, 1997; Van Lommel, 2001). An increased number of NEBs has been reported in young rabbits kept in hypobaric chambers or bred at high altitude and in Sprague Dawley rats exposed to chronic normobaric hypoxia (Pack et al., 1986). Wistar rats exposed to hypoxic conditions for 1 to 3 wk showed elevated levels of intracellular CGRP (McBride et al., 1990; Roncalli et al., 1993), without a change in NEB cell number. It has been proposed that the generation of reactive oxygen species from mitochondria, having effects at the level of gene expression, underlies the responses of NEBs to chronic hypoxia (Peers and Kemp, 2001).

254  |  Airway Chemoreceptors in the Vertebrates Pulmonary NEBs in rodents may be regarded as inexhaustible local pools of vasoactive transmitters, such as the potent pulmonary vasodilator CGRP. In normoxic lung areas, the continuous release of CGRP from NEB cells might be at least partly responsible for the homeostatic control of blood vessel relaxation. Physiological airway hypoxia apparently inhibits CGRP secretion from rat NEBs (Springall and Polak, 1993), potentially resulting in local vasoconstriction and thus adjustment of pulmonary perfusion to ventilation. Specific populations of nerve fibers contacting the NEBs are believed to play a role in modulating the reactions of NEBs to hypoxia (Brouns et al., 2002a; Van Genechten et al., 2003, 2004). It has, for instance, been demonstrated that the spinal sensory nerve fibers in contact with NEBs are required for the modulatory effect on the pulmonary vascular tone mediated by endogenously released CGRP (Tjen-ALooi et al., 1998). Also, the intraepithelial nitrergic nerve terminals may be involved. All NEBs receiving an intraepithelial nitrergic innervation were seen to also reveal basal contacts with CGRP-ir spinal sensory nerve fibers, the presumable collaterals of which form baskets around the nitrergic neurons in the lamina propria. In this way, the nitrergic innervation, together with the spinal sensory nerve fibers, may be essential components in the hypoxic inhibition of CGRP release from NEBs (Brouns et al., 2002a; Adriaensen et al., 2003). For Fawn-Hooded rats, a model for primary pulmonary hypertension, it has been proposed that the strongly reduced intrinsic pulmonary nitrergic innervation of NEBs might be involved in the mechanism of the unexplained hypersensitivity of this rat strain to hypoxia (Van Genechten et al., 2003, 2004). For carotid bodies, chemosensory receptor-effector organs monitoring blood gasses, it has also been proposed that inhibition of neurotransmitter release from carotid body chemoreceptor cells is mediated by the release of nitric oxide from efferent nerve terminals (Campanucci and Nurse, 2007). Even during the normal pulmonary activity, parts of the lungs will be cyclically “hypoxic”. A system that prevents the receptors from continuously transmitting this “redundant” information to the CNS would be crucial for avoiding exaggerated central actions that would eventually lead to a general pulmonary vasoconstriction and hypertension similar to that seen in some pathological conditions or, for example, in the Fawn-Hooded rat. In the proposed mechanism, NO released from nitrergic nerve terminals in pulmonary NEBs may inhibit the sensory discharge of NEB cells in response to mild local hypoxia. It may be presumed that the P2X3 receptor-expressing vagal sensory component of the innervation of NEBs may be involved in the central interactions, once a certain threshold is reached.

Neuroepithelial Bodies as Potential Mechanoreceptors The possibility that at least subpopulations of pulmonary NEBs, in particular those connected to vagal sensory nerve terminals, may act as mechanoreceptors was suggested

Diverse and Complex Airway Receptors in Rodent Lungs  |  255

more than 30 years ago (Lauweryns and Peuskens, 1972; Wasano and Yamamoto, 1978). However, probably because of the more recent strong belief in NEBs as oxygen sensors (see above), even recent reviews dealing with the supposed functions of NEBs neglected this hypothesis (Linnoila, 2006). Intraepithelial nerve terminals in rodent pulmonary NEBs originate from myelinated vagal afferents and express Na+/K+-ATPase a3, VGLUTs, P2X3 receptors and calcium-binding proteins. Because this panel of markers has been described to rather selectively label mechanoreceptor terminals in other organs (Dütsch et al., 1998; Raab and Neuhuber, 2003; Dobretsov et al., 2003; Wang and Neuhuber, 2003; Wu et al., 2004), it is reasonable to assume that vagal nodose nerve terminals in NEBs may be involved in mechanosensory-like events. At least, a potential transduction system appears to be present in rodent pulmonary NEBs. As proposed for hollow organs (for review see Burnstock, 2006), an appropriate mechanical stimulus might induce release of ATP by exocytosis of the secretory granules from NEB cells. The released ATP may then bind to the P2X3 receptors that are present on a population of myelinated vagal afferent nerve fibers with terminals protruding between the NEB cells, and mechanosensory information may be transferred from the airways to the CNS (Brouns et al., 2000). Using the same panel of “mechanosensory” markers recently disclosed the presence, location, morphology and neurochemical coding of subepithelial receptor-like nerve endings that protrude between airway smooth muscle cells in both rat and mouse lungs (Brouns et al., 2006a, 2006b, 2007). The latter structures, referred to as “smooth muscle-associated airway receptors (SMARs)”, were characterized by their Na+/K+ATPase a3 (Figure 6c), vesicular glutamate transporter 1 (VGLUT1) and VGLUT2 IR (Figure 5, 6b), expression of the P2X3 receptor, and the presence of calcium-binding proteins. As a consequence, none of these markers can be regarded as exclusive for the vagal sensory innervation of NEBs or for SMARs. Nerve fibers giving rise to SMARs were also myelinated and had a vagal origin. In this way, the neurochemical coding and receptor-like appearance of SMARs appeared to be almost identical to at least part of the complex vagal sensory terminals in NEBs. NEBs and SMARs are located in the airway epithelium and the smooth muscle layer respectively. Since the airways of rats and mice harbor several thousands of NEBs (Van Genechten et al., 2004), and the smooth muscle layer is invariably located very close to the airway mucosa in these small mammals, NEBs and SMARs are regularly found in close apposition (Figures 4-6). This observation, together with the almost identical chemical coding of populations of vagal sensory terminals in NEBs and nearby SMARs (Brouns et al., 2006a, 2006b), clearly points out that interpretation of electrophysiological data based on “local” stimuli should be made with great caution. It is now generally accepted that one of the most prominent activators of “SAR”activity is an increase in transmural airway pressure (Sant’Ambrogio, 1982). In this way “SAR”-activity likely plays a role in the negative feedback mechanism that acts to limit increases in parasympathetic tone in the airway (Widdicombe and Nadel,

256  |  Airway Chemoreceptors in the Vertebrates 1963). Whereas SMARs, located in smooth muscle, seem to be perfectly positioned to perform such a function, an increasing intraluminal pressure would potentially stimulate mucosal sensors (such as NEBs) at least as effectively as receptors located in the muscle layer (Adriaensen et al., 2006). According to some authors, notable differences may exist between vagal pulmonary receptors in rats and those known in other species (Bergren and Peterson, 1993), while other reported data for rats and mice in general do not appear to be all that different from other species (Widdicombe, 2001; Ho et al., 2001; Yu et al., 2006; Zhang et al., 2006). Although direct experimental evidence is not yet available, our own observation that by far the largest number of myelinated vagal nodose afferents in rodent intrapulmonary airways selectively innervate NEBs strongly suggests that discharges from NEB-related myelinated vagal afferent fibers may be part of the already characterized activity of vagal myelinated receptors in the lower airways. Efforts for matching the available physiological and morphological data may, however, be hampered by potentially important species differences. The nearly identical neurochemical and morphological characteristics of SMARs and vagal nodose nerve terminals in pulmonary NEBs obviously complicates the straightforward correlation of these morphologically identified airway receptors with so-called “SAR”-activity. Based on inconclusive knowledge of sensory airway receptor morphology (Sant’Ambrogio, 1982; Widdicombe, 2001; Schelegle, 2003; Yu et al., 2003, 2004), SMAR-like nerve endings in the airway smooth muscle have for many years been presumed to give rise to “SAR”-activity. In view of the present data, both SMARs and at least a subpopulation of the vagal nodose nerve terminals in NEBs seem to be good candidates to represent the morphological counterparts of subsets of the physiologically identified myelinated vagal airway mechanoreceptors.

Figure 4: A PGP9.5-ir NEB is located in juxtaposition to extensive PGP9.5-ir subepithelial laminar nerve terminals of a smooth muscle-associated airway receptor (SMAR) in a rat bronchiole. Maximum value projection of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). L, airway lumen, E, airway epithelium.

Diverse and Complex Airway Receptors in Rodent Lungs  |  257

Figure 5: Rat bronchus. Intraepithelial VGLUT2-ir nerve terminals (arrowheads) that are known to surround NEB-cells are found adjacent to the VGLUT2-ir nerve terminals of a SMAR. Maximum value projection of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). L, airway lumen; E, airway epithelium.

Figure 6: Triple immunocytochemical staining for PGP9.5 (a), VGLUT2 (b) and Na+/K+ATPase a3 (c) in the rat lung. a. An innervated intraepithelial PGP9.5-ir NEB is found in the immediate neighborhood of a SMAR. b. Both the SMAR and the NEB-related intraepithelial nerve terminals show VGLUT2 immunoreactivity. c. Two separate Na+/K+-ATPase a3-ir nerve fibers give rise to the SMAR (open arrow) and to a subpopulation of the NEB-related VGLUT2-ir intraepithelial nerve terminals (arrow). Maximum value projections of confocal optical sections (PerkinElmer confocal UltraVIEW ERS). L, airway lumen; E, airway epithelium.

258  |  Airway Chemoreceptors in the Vertebrates Although direct functional data are not yet available, the nerve terminals associated with pulmonary NEB cells on their own seem to have everything for performing a mechanosensory function. If so, the question remains what might be the meaning and possible input of the neuroendocrine cells in the NEB complex.

Concluding Remarks and Future Prospects Exploring the location, morphology, and neurochemical coding of sensory receptor-like structures in the airways of rodents unravelled both intraepithelial receptor end organs, which were invariably co-localized with pulmonary NEB cells, and intramuscular SMAR endings. The above-mentioned data clearly revealed that pulmonary NEBs are constructed as integrated receptor complexes. Many different motor and sensory nerve fiber populations are in place to transduce and conduct the sensory information to the CNS and/or to modulate the activity (transmitter release) from NEB cells. The potential stimuli-translating NEB cells are well shielded from the airway lumen, leaving only thin, probably very specialized cell processes to monitor changes in the airway environment. The Clara-like cells that selectively cover NEB cells differ from “classical” Clara cells by their lack of the cytochrome p450-2F2 isoenzyme and their consequent resistance to certain pollutants (Reynolds et al., 2000). Together with the endocrine cells, these cells compose unique complexes. Recent reviews classify Clara-like cells as a stem celllike population (Meuwissen and Berns, 2005; Linnoila, 2006; Giangreco et al., 2007). Clara-like cells may, therefore, have more important functions in the NEB-complex than accounted for thus far. Evidently, NEBs in rodents are extensively innervated by vagal afferents. Although a direct link between the NEB-related vagal sensory populations and measurable activities in pulmonary vagal afferents is presently lacking, we strongly believe that the vagal afferent fibers that specifically contact NEBs indeed participate, and will eventually be recognized as so doing, in the plethora of known airway receptor activities. However, it appears to be extremely difficult to correlate electrophysiologically identified receptor activity to morphologically identified receptor end organs. Given the multiplicity of physiologically characterized airway receptors, and the presently limited number of morphologically identified airway receptors, we believe that a single morphological identified airway receptor may likely combine multiple sensory activities. Therefore, morphology needs to be linked to physiology, i.e., for any given receptor structure the functional properties should be identified. To achieve a better understanding of the sensory interactions between the airways and the CNS, it will be important to pay proper attention to the increasing amount of structural and neurochemical information regarding sensory airway receptors. Clearly, future combined morphological and electrophysiological studies in whole animals, lung

Diverse and Complex Airway Receptors in Rodent Lungs  |  259

slices and whole mount preparations will be essential to further unravel the complex maze of electrophysiologically and neurochemically identified airway receptors. Recently, in situ lung models were developed in which morphologically identified receptors can be specifically visualized, facilitating direct physiological studies of pulmonary NEBs (Pintelon et al., 2005, De Proost et al., 2008) and SMARs (De Proost et al., 2007). De Proost and colleagues (2008) further optimized the in situ live cell imaging model based on vibratome slices of live mouse lungs (Pintelon et al., 2005), allowing visualization of pulmonary NEBs and simultaneous monitoring of the intracellular Ca2+ concentration using a calciumindicator. The unique advantage of this model is the visualization of potential activation of all cells in a NEB and the simultaneous collection of information from many other surrounding cells and tissues in the lungs. This lung slice model, therefore, opens new perspectives for studying possible interactions between NEBs and neighboring cells. Finally, it should be stressed that conclusive functional evidence for the nature of the information carried by any of the multiple populations of afferents that selectively connect to rodent pulmonary NEBs is still lacking. However, the current essential knowledge on the organization and neurochemical characteristics of airway receptors in control mice (Brouns et al., 2009) will finally create possibilites to also take advantage of the use of genetically modified mice for ex vivo functional studies of pulmonary NEBs. In conclusion, there is now good evidence that pulmonary NEBs are likely able to accommodate various sensory activities, and that only the combination of morphological and physiological studies will be able to provide the key solutions to unravel their function in healthy lung. Given all available knowledge on organization of NEBs and on their complex innervation, it is likely that the NEB complexes are composed of modular compartments, and that their biological or physiological significance may be context- and time-dependent.

References Adriaensen, D., and Scheuermann, D.W. 1993. Neuroendocrine cells and nerves of the lung. Anat. Rec., 236: 70-85. Adriaensen, D., and Timmermans, J.-P. 2004. Purinergic signalling in the lung: important in asthma and COPD? Curr. Opin. Pharmacol., 4: 207-214. Adriaensen, D., Timmermans, J.-P., Brouns, I., Berthoud, H.R., Neuhuber, W.L., and Scheuermann, D.W. 1998. Pulmonary intraepithelial vagal nodose afferent nerve terminals are confined to neuroepithelial bodies: an anterograde tracing and confocal microscopy study in adult rats. Cell Tissue Res., 293: 395-405. Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.-P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat. Rec., 270A: 25-40. Adriaensen, D., Brouns, I., Pintelon, I., De Proost, I., and Timmermans, J.-P. 2006. Evidence for a role of neuroepithelial bodies as complex airway sensors: comparison with smooth muscle-associated airway receptors. J. Appl. Physiol., 101: 960-970.

260  |  Airway Chemoreceptors in the Vertebrates Aguayo, S.M. 1993. Pulmonary neuroendocrine cells in tobacco-related lung disorders. Anat. Rec., 236: 122-127. Bergren, D.R., and Peterson, D.F. 1993. Identification of vagal sensory receptors in the rat lung: are there subtypes of slowly adapting receptors? J. Physiol., 464: 681-698. Berkley, H.J. 1894. The intrinsic pulmonary nerves in mammalia. Johns Hopkins Hosp. Res., 4: 240-247. Bisby, M.A., and Fillenz, M. 1971. The storage of endogenous noradrenaline in sympathetic nerve terminals. J. Physiol., 215: 163-179. Bollé, T., Lauweryns, J.M., and Van Lommel, A. 2000. Postnatal maturation of neuroepithelial bodies and carotid body innervation: a quantitative investigation in the rabbit. J. Neurocytol., 29: 241-248. Brouns, I., Adriaensen, D., Burnstock, G., and Timmermans, J.-P. 2000. Intraepithelial vagal sensory nerve terminals in rat pulmonary neuroepithelial bodies express P2X3 receptors. Am. J. Respir. Cell Mol. Biol., 23: 52-61. Brouns, I., Van Genechten, J., Scheuermann, D.W., Timmermans, J.-P., and Adriaensen, D. 2002a. Neuroepithelial bodies: a morphologic substrate for the link between neuronal nitric oxide and sensitivity to airway hypoxia? J. Comp. Neurol., 449: 343-354. Brouns, I., Van Genechten, J., Timmermans, J.-P., Scheuermann, D.W., and Adriaensen, D. 2002b. Pulmonary neuroepithelial bodies are selectively contacted by a population of cholinergic motor fibers. FASEB J. 16: A453 (abstract). Brouns, I., Van Nassauw, L., Van Genechten, J., Majewski, M., Scheuermann, D.W., Timmermans, J.-P., and Adriaensen, D. 2002c. Triple immunofluorescence staining method with antibodies raised in the same species to study the complex innervation pattern of intrapulmonary chemoreceptors. J. Histochem. Cytochem., 50: 575-582. Brouns, I., Van Genechten, J., Hayashi, H., Gajda, M., Gomi, T., Burnstock, G., Timmermans, J.-P., and Adriaensen, D. 2003. Dual sensory innervation of pulmonary neuroepithelial bodies. Am. J. Respir. Cell Mol. Biol., 28: 275-285. Brouns, I., Pintelon, I., Van Genechten, J., De Proost, I., Timmermans, J.-P., and Adriaensen, D. 2004. Vesicular glutamate transporter 2 is expressed in different nerve fibre populations that selectively contact pulmonary neuroepithelial bodies. Histochem. Cell Biol., 121: 1-12. Brouns, I., De Proost, I., Pintelon, I., Timmermans, J.-P., and Adriaensen, D. 2006a. Sensory receptors in the airways: neurochemical coding of smooth muscle-associated airway receptors and pulmonary neuroepithelial body innervation. Auton. Neurosci., 126-127: 307-319. Brouns, I., Pintelon, I., De Proost, I., Alewaters, R., Timmermans, J.-P., and Adriaensen, D. 2006b. Neurochemical characterisation of sensory receptors in airway smooth muscle: comparison with pulmonary neuroepithelial bodies. Histochem. Cell Biol., 125: 351-367. Brouns, I., Pintelon, I., De Proost, I., Oztay, F., Timmermans, J.-P., and Adriaensen, D. 2007. Functional morphology and neurochemical characterization of diverse sensory receptors in mouse lungs. FASEB J. 21: A920. Brouns, I., Oztay, F., Pintelon, I., De Proost, I., Lembrechts, R., Timmermans, J.-P., Adriaensen, D. (2009) Neurochemical pattern of the compled innervation of neuropithelial bodies in mouse lungs. Histochem Cell Biol., 131: 55-7. Burnstock, G. 2006. Purinergic signaling. Br. J. Pharmacol., 147: S147-S181. Cadieux, A., Springall, D.R., Mulderry, P.K., Rodrigo, J., Ghatei, M.A., Terenghi, G., Bloom, S.R., and Polak, J.M. 1986. Occurrence, distribution and ontogeny of CGRP immunoreactivity in the rat lower respiratory tract: effect of capsaicin treatment and surgical denervations. Neuroscience, 19: 605-627.

Diverse and Complex Airway Receptors in Rodent Lungs  |  261 Campanucci, V.A., and Nurse, C.A. 2007. Autonomic innervation of the carotid body: role in efferent inhibition. Respir. Phsyiol. Neurobiol., doi: 10.1016/j.resp.2007.01.020. Carabba, V.H., Sorokin, S.P., and Hoyt, R.F.J. 1985. Development of neuroepithelial bodies in intact and cultured lungs of fetal rats. Am. J. Anat., 173: 1-27. Cole, G.A., Polak, J.M., Johnson, N.F., Marangos, P., and Pearse, A.G.E. 1982. Neuron specific enolase as a useful histological marker for the demonstration of asbestos-induced endocrine cell hyperplasia of the rat lung. J. Pathol., 137: 92-93. Crowe, R., and Burnstock, G. 1981. Comparative studies of quinacrine-positive neurones in the myenteric plexus of stomach and intestine of guinea-pig, rabbit and rat. Cell Tissue Res., 221: 93-107. Cutz, E. 1997. Studies on neuroepithelial bodies under experimental and disease conditions. In: Cellular and Molecular Biology of Airway Chemoreceptors. E. Cutz (Ed.). pp. 109-129. Cutz, E., and Jackson, A. 1999. Neuroepithelial bodies as airway oxygen sensors. Respir. Physiol., 115: 201-214. Cutz, E., Chan, W., Wong, V., and Conen, P.E. 1974. Endocrine cells in rat fetal lungs. Ultrastructural and histochemical study. Lab. Invest., 30: 458-464 (abstract). Cutz, E., Yeger, H., Wong, V., Bienkowski, E., and Chan, W. 1985. In vitro characteristics of pulmonary neuroendocrine cells isolated from rabbit fetal lung. I. Effects of culture media and nerve growth factor. Lab. Invest., 53: 672-683. Cutz, E., Fu, X.W., and Yeger, H. 2004. Methods to study neuroepithelial bodies as airway oxygen sensors. Meth. Enzymol., 381: 26-40. De Proost, I., Brouns, I., Pintelon, I., Timmermans, J.-P., and Adriaensen, D. 2006. Pulmonary expression pattern of voltage gated Ca 2+ channels: special reference to neuroepithelial bodies. FASEB J., 20: A788 (abstract). De Proost, I., Pintelon, I., Brouns, I., Timmermans, J.-P., and Adriaensen, D. 2007. Selective visualisation of sensory receptors in the smooth muscle layer of ex vivo airway whole mounts by styryl pyridinium dyes. Cell Tissue Res. 329:421-431. De Proost, I., Pintelon, I., Brouns, I., Kroese, A.B.A., Riccardi, D., Kemp, P.J., Timmermans, J.-P., Adriaensen D. (2008) Functional live cell imaging of the pulmonary neuroepithelial body microenvironment. Am. J. Resp. Cell Mol. Biol., 39: 180-189. Dinh, Q.T., Groneberg, D., Peiser, C., Mingomataj, E., Joachima, R.A., Witt, C., Arck, P.C., Klappa, F.H., and Fischer, A. 2004. Substance P expression in TRPV1 and trkA-positive dorsal root ganglion neurons innervating the mouse lung. Respir. Physiol. Neurobiol., 144: 15-24. Dobretsov, M., Hastings, S.L., Sims, T.J., Stimers, J.R., and Romanovsky, D. 2003. Stretch receptor-associated expression of alpha 3 isoform of the Na+,K+-ATPase in rat peripheral nervous system. Neuroscience, 116: 1069-1080. Dütsch, M., Eichhorn, U., Wörl, J., Wank, M., Berthoud, H.-R., and Neuhuber, W.L. 1998. Vagal and spinal afferent innervation of the rat esophagus: a combined retrograde tracing and immunocytochemical study with special emphasis on calcium-binding proteins. J. Comp. Neurol., 398: 289-307. Elftman, A.G. 1943. The afferent and parasympathetic innervation of the lungs and trachea of the dog. Am. J. Anat., 72: 1-27. Fröhlich, F. 1949. Die “Helle Zelle” der Bronchialschleimhaut und ihre Beziehungen zum Problem der Chemoreceptoren. Frankf. Z. Pathol., 60: 517-559. Fu, X.W., Nurse, C.A., Wang, Y.T., and Cutz, E. 1999. Selective modulation of membrane currents by hypoxia in intact airway chemoreceptors from neonatal rabbit. J. Physiol., 514: 139-150.

262  |  Airway Chemoreceptors in the Vertebrates Fu, X.W., Wang, D., Nurse, C.A., Dinauer, M.C., and Cutz, E. 2000. NADPH oxidase is an O2 sensor in airway chemoreceptors: evidence from K+ current modulation in wild-type and oxidase-deficient mice. Proc. Natl. Acad. Sci. USA, 97: 4374-4379. Fu, X.W., Nurse, C.A., Wong, V., and Cutz, E. 2002. Hypoxia-induced secretion of serotonin from intact pulmonary neuroepithelial bodies in neonatal rabbit. J. Physiol., 539: 503-510. Fu, X.W., Nurse, C.A., Farragher, S.M., and Cutz, E. 2003. Expression of functional nicotinic acetylcholine receptors in neuroepithelial bodies of neonatal hamster lung. Am. J. Physiol., Lung Cell Mol. Physiol., 285: 1203-1212. Giangreco, A., Groot, K.R., and Janes, S.M. 2007. Lung cancer and lung stem cells: strange bedfellows? Am. J. Respir. Crit. Care Med., 175: 547-553. Gomez-Pascual, A., Martin-Lacave, I., Moreno, A.M., Fernandez, A., and Galera, H. 1990. Neuroendocrine (NE) cells in rat neonatal lungs. A histochemical and immuncytochemical study. Anat. Histol. Embryol., 19: 158-163. Gosney, J.R., and Sissons, M. 1985. Widespread distribution of bronchopulmonary endocrine cells immunoreactive for calcitonin in the lung of the normal adult rat. Thorax, 40: 194198. Hage, E. 1976. Endocrine-like cells of the pulmonary epithelium. In: Chromaffin, Enterochromaffin and Related Cells. R.E. Coupland and R. Fujita (Eds.). Elsevier, pp. 317332. Hartness, M.E., Lewis, A., Searle, G.J., O’Kelly, I., Peers, C., and Kemp, P.J. 2001. Combined antisense and pharmacological approaches implicate hTASK as an airway O2 sensing K+ channel. J. Biol. Chem., 276: 26499-26508. Haxhiu, M.A., Kc, P., Moore, C.T., Acquah, S.S., Wilson, C.G., Zaidi, S.I., Massari, V.J., and Ferguson, D.G. 2005. Brain stem excitatory and inhibitory signaling pathways regulating bronchoconstrictive responses. J. Appl. Physiol., 98: 1961-1982. Ho, C.-Y., Gu, Q., Lin, Y.S., and Lee, L.-Y. 2001. Sensitivity of vagal afferent endings to chemical irritants in the rat lung. Respir. Physiol., 127: 113-124. Holzer, P. 1991. Capsaicin: cellular targets, mechanisms of action, and selectivity for thin sensory neurons. Pharmacol. Rev., 43: 143-201. Honig, M.G. 1993. DiI Labelling. Neurosci. Prot., 93-050-16-01-20. Hung, K.-S. 1984. Histology, ultrastructure, and development of the pulmonary endocrine cell. In: The Endocrine Lung in Health and Disease. K.L. Becker and A.F. Gazdar (Eds.). Saunders, Philadelphia, pp. 162-192. Hung, K.-S., and Loosli, C.G. 1974. Bronchiolar neuro-epithelial bodies in the neonatal mouse lungs. Am. J. Anat., 140: 191-200. Hung, K.-S., Hertweck, M.S., Hardy, J.D., and Loosli, C.G. 1973. Ultrastructure of nerves and associated cells in bronchiolar epithelium of the mouse lung. J. Ultrastr. Res., 43: 426-437. Ichikawa, H. 2002. Innervation of the carotid body: immunohistochemical, denervation, and retrogade tracing studies. Microsc. Res. Tech., 59: 188-195. Jeffery, P.K., and Reid, L. 1973. Intra-epithelial nerves in normal rat airways: a quantitative electron microscopic study. J. Anat., 114: 35-45. Jeffery, P.K., and Reid, L. 1975. New observations of rat airway epithelium: a quantitative and electron microscopic study. J. Anat., 120: 195-320. Kemp, P.J., Lewis, A., Hartness, M., Searle, G.J., Miller, P., O’Kelly, I., and Peers, C. 2002. Airway chemotransduction: from oxygen sensor to cellular effector. Am. J. Respir. Crit. Care Med., 166: S17-S24.

Diverse and Complex Airway Receptors in Rodent Lungs  |  263 Kemp, P.J., Searle, G.J., Hartness, M.E., Lewis, A., Miller, P., Williams, S., Wootton, P., Adriaensen, D., and Peers, C. 2003. Acute oxygen sensing in cellular models: relevance to the physiology of pulmonary neuroepithelial and carotid bodies. Anat. Rec., 270: 41-50. Lachamp, P., Crest, M., and Kessler, J.P. 2006. Vesicular glutamate transporters type 1 and 2 expression in axon terminals of the rat nucleus of the solitary tract. Neuroscience, 137: 73-81. Larsell, O. 1921. Nerve termination in the lung of the rabbit. J. Comp. Neurol., 33: 105-132. Larsell, O., and Dow, R.S. 1933. The innervation of the human lung. Am. J. Anat., 52: 125-146. Larson, S.D., Schelegle, E.S., Hyde, D.M., and Plopper, C.G. 2003. The three-dimensional distribution of nerves along the entire intrapulmonary airway tree of the adult rat and the anatomical relationship between nerves and neuroepithelial bodies. Am. J. Respir. Cell. Mol. Biol., 28: 592-599. Lauweryns, J.M., and Cokelaere, M. 1973. Hypoxia-sensitive neuro-epithelial bodies. Intrapulmonary secretory neuroreceptors modulated by the CNS. Z. Zellforsch. 521-540. Lauweryns, J.M., and Peuskens, J.C. 1972. Neuro-epithelial bodies (neuroreceptor or secretory organs?) in human infant bronchial and bronchiolar epithelium. Anat. Rec., 172: 471-481. Lauweryns, J.M., Van Lommel, A. (1986). Effect of various vagotomy procedures on the reaction to hypoxia of rabbit neuropithelial bodies: modulation by intrapulmonary axon reflexes. Exp. Lung Res., 11:319-339. Lauweryns, J.M., Van Lommel, A. (1987). Ultrastructure of nerve endings and synaptic junctions in rabbit intrapulmonary neuropithelial bodies: a single and serial section analysis. J. Anat., 151:65-83. Lauweryns, J.M., and Van Ranst, L. 1987. Calcitonin gene related peptide immunoreactivity in rat lung: light and electron microscopic study. Thorax, 42: 183-189. Lauweryns, J.M., and Van Ranst, L. 1988a. Immunocytochemical localization of aromatic L-amino acid decarboxylase in human, rat, and mouse bronchopulmonary and gastrointestinal endocrine cells. J. Histochem. Cytochem., 36: 1181-1186. Lauweryns, J.M., and Van Ranst, L. 1988b. Protein gene product 9.5 expression in the lungs of humans and other mammals. Immunocytochemical detection in neuroepithelial bodies, neuroendocrine cells and nerves. Neurosci. Lett., 85: 311-316. Lauweryns, J.M., Cokelaere, M., and Theunynck, P. 1972. Neuroepithelial bodies in the respiratory mucosa of various mammals. A light optical, histochemical and ultrastuctural investigation. Z. Zellforsch. Mikrosk. Anat., 135: 569-592. Lauweryns, J.M., de Bock, V., Verhofstad, A.A., and Steinbusch, H.W. 1982. Immunohistochemical localization of serotonin in intrapulmonary neuro-epithelial bodies. Cell Tissue Res., 226: 215-223. Lawrence, A.J. 1995. Neurotransmitter mechanisms of rat vagal afferent neurons. Clin. Exp. Pharmacol. Physiol., 22: 869-873. Liedtke, W., and Kim, C. 2005. Functionality of the TRPV subfamily of TRP ion channels: add mechano-TRP and osmo-TRP to the lexicon! Cell. Mol. Life Sci., 62: 2985-3001. Linnoila, R.I. 2006. Functional facets of the pulmonary neuroendocrine system. Lab. Invest., 86: 425-444. Luts, A., Uddman, R., Absood, A., Håkanson, R., and Sundler, F. 1991. Chemical coding of endocrine cells of the airways: presence of helodermin-like peptides. Cell Tissue Res., 265: 425-433.

264  |  Airway Chemoreceptors in the Vertebrates Luts, A., Uddman, R., Hakanson, R., and Sundler, F. 1994. Calcitonin, CGRP and helodermin in endocrine cells of the developing rat lung. Regul. Pept., 51: 121-129. Martling, C.-R., Saria, A., Fischer, J.A., Hökfelt, T., and Lundberg, J.M. 1988. Calcitonin gene-related peptide and the lung: neuronal coexistence with substance P, release by capsaicin and vasodilatory effect. Regul. Pept., 20: 125-139. Mazzone, S.B. 2005. An overview of the sensory receptors regulating cough. Cough, 1: 2. McBride, J.T., Springall, D.R., Winter, R.J., and Polak, J.M. 1990. Quantitative immunocytochemistry shows calcitonin gene-related peptide-like immunoreactivity in lung neuroendocrine cells is increased by chronic hypoxia in the rat. Am. J. Respir. Cell Mol. Biol., 3: 587-593. Meuwissen, R., and Berns, A. 2005. Mouse models for human lung cancer. Genes Dev., 19: 643-664. O’Kelly, I., Peers, C., and Kemp, P.J. 1998. O2-sensitive K+ channels in neuroepithelial bodyderived small cell carcinoma cells of the human lung. Am. J. Phsyiol. 275: L709-L716. O’Kelly, I., Stephens, R.H., Peers, C., and Kemp, P.J. 1999. Potential identification of the O2-sensitive K+ current in a human neuroepithelial body-derived cell line. Am. J. Physiol., 276: L96-L104. O’Kelly, I., Lewis, A., Peers, C., and Kemp, P.J. 2000. O2 sensing by airway chemoreceptorderived cells: protein kinase C activation reveals functional evidence for involvement of NADPH oxidase. J. Biol. Chem., 275: 7684-7692. O’Kelly, I., Peers, C., and Kemp, P.J. 2001. NADPH oxidase does not account fully for O2sensing in model airway chemoreceptor cells. Biochem. Biophys. Res. Commun., 283: 11311134. Oh, E.J., Mazzone, S.B., Canning, B.J., and Weinreich, D. 2006. Reflex regulation of airway sympathetic nerves in guinea-pigs. J. Physiol., 573: 549-564. Olson, L., Ålund, M., and Norberg, K.-A. 1976. Fluorescence microscopical demonstration of a population of gastro-intestinal nerve fibres with a selective affinity for quinacrine. Cell Tissue Res., 171: 407-423. Pack, R.J., Barker, S., and Howe, A. 1986. The effect of hypoxia on the number of aminecontaining cells in the lung of the adult rat. Eur. J. Respir. Dis., 68: 121-130. Pan, J., Yeger, H., and Cutz, E. 2004. Innervation of pulmonary neuroendocrine cells and neuroepithelial bodies in developing rabbit lung. J. Histochem. Cytochem., 52: 379-389. Pan, J., Luk, C., Kent, G., Cutz, E., and Yeger, H. 2006. Pulmonary neuroendocrine cells, airway innervation, and smooth muscle are altered in Cftr null mice. Am. J. Respir. Cell Mol. Biol., 35: 320-326. Peers, C., and Kemp, P.J. 2001. Acute oxygen sensing: diverse but convergent mechanisms in airway and arterial chemoreceptors. Respir. Res., 2: 145-149. Pintelon, I., Brouns, I., Van Genechten, J., Scheuermann, D.W., Timmermans, J.-P., and Adriaensen, D. 2003. Pulmonary expression of the vesicular acetylcholine transporter with special reference to neuroepithelial bodies. Auton. Neurosci., 106: 47 (abstract). Pintelon, I., De Proost, I., Brouns, I., Van Herck, H., Van Genechten, J., Van Meir, F., Timmermans, J.-P., and Adriaensen, D. 2005. Selective visualisation of neuroepithelial bodies in vibratome slices of living lung by 4-Di-2-ASP in various animal species. Cell Tissue Res., 321: 21-33. Plummer, H.K.I., Sheppard, B., and Schuller, H.M. 2000. Interaction of tobacco-specific toxicants with nicotinic cholinergic regulation of pulmonary neuroendocrine cells: implications for pediatric lung disease. Exp. Lung Res., 26: 121-135.

Diverse and Complex Airway Receptors in Rodent Lungs  |  265 Polak, J.M., Becker, K.L., Cutz, E., Gail, D.B., Goniakowska-Witalinska, L., Gosney, J.R., Lauweryns, J.M., Linnoila, I., McDowell, E.M., and Miller, Y.E. 1993. Lung endocrine cell markers, peptides, and amines. Anat. Rec., 236: 169-171. Raab, M., and Neuhuber, W.L. 2003. Vesicular glutamate transporter 2 immunoreactivity in putative vagal mechanosensor terminals of mouse and rat esophagus: indication of a local effector function. Cell Tissue Res., 312: 141-148. Raab, M., and Neuhuber, W.L. 2007. Glutamatergic functions of primary afferent neurons with special emphasis on vagal afferents. Int. Rev. Cytol., 256: 223-275. Reynolds, S.D., Giangreco, A., Power, J.H., and Stripp, B.R. 2000. Neuroepithelial bodies of pulmonary airways serve as a reservoir of progenitor cells capable of epithelial regeneration. Am. J. Pathol., 156: 269-278. Rogers, D.C., and Haller, C.J. 1978. Innervation and cytochemistry of the neuroepithelial bodies in the ciliated epithelium of the toad lung (Bufo marinus). Cell Tissue Res., 195: 395-410. Roncalli, M., Springall, D.R., Maggioni, M., Moradoghli-Haftvani, A., Winter, R.J.D., Zhao, L., Coggi, G., and Polak, J.M. 1993. Early changes in the calcitonin gene-related peptide (CGRP) content of pulmonary endocrine cells concomitant with vascular remodeling in the hypoxic rat. Am. J. Respir. Cell Mol. Biol., 9: 467-474. Sant’Ambrogio, G. 1982. Information arising from the tracheobronchial tree of mammals. Physiol. Rev., 62: 531-569. Schelegle, E.S. 2003. Functional morphology and physiology of slowly adapting pulmonary stretch receptors. Anat. Rec., 270: 11-16. Scheuermann, D.W. 1987. Morphology and cytochemistry of the endocrine epithelial system in the lung. Int. Rev. Cytol., 106: 35-88. Scheuermann, D.W., Adriaensen, D., and Timmermans, J.-P. 1989. Neuroepithelial endocrine cells in the lung of Ambystoma mexicanum. Anat. Rec., 225: 139-149. Schuller, H.M., Jull, B.A., Sheppard, B.J., and Plummer, H.K. 2000. Interaction of tobaccospecific toxicants with the neuronal alpha(7) nicotinic acetylcholine receptor and its associated mitogenic signal transduction pathway: potential role in lung carcinogenesis and pediatric lung disorders. Eur. J. Pharmacol., 393: 265-277. Sekhon, H.S., Jia, Y., Raab, R., Kuryatov, A., Pankow, J.F., Whitsett, J.A., Lindstrom, J., and Spindel, E.R. 1999. Prenatal nicotine increases pulmonary α7 nicotinic receptor expression and alters fetal lung development in monkeys. J. Clin. Invest., 103: 637-647. Sheppard, M.N., Johnson, N.F., Cole, G.A., Bloom, S.R., Marangos, P., and Polak, J.M. 1982. Neuron specific enolase (NSE) immunostaining—a useful tool for the light microscopical detection of endocrine cell hyperplasia in adult rats exposed to asbestos. Histochemistry, 74: 505-513. Shimosegawa, T., and Said, S.I. 1991a. Co-occurrence of immunoreactive calcitonin and calcitonin gene-related peptide in neuroendocrine cells of rat lungs. Cell Tissue Res., 264: 555-561. Shimosegawa, T., and Said, S.I. 1991b. Pulmonary calcitonin gene-related peptide immunoreactivity: nerve-endocrine cell interrelationships. Am. J. Respir. Cell. Mol. Biol., 4: 126-134. Sorokin, S.P., Hoyt, R.F. 1989. Neuroepithelial bodies and solitary small-granule cells. In: Lung Cell Biology. D. Massaro (Ed.). Marcel Dekker, New York, pp. 191-344. Sorokin, S.P., and Hoyt, R.F. 1990. On the supposed function of neuroepithelial bodies in adult mammalian lungs. News Physiol. Sci., 5: 89-95.

266  |  Airway Chemoreceptors in the Vertebrates Sorokin, S.P., and Hoyt, R.F. 1993. Proceedings of workshop on pulmonary neuroendocrine cells in health and disease. Anat. Rec., 236: 1-256. Sorokin, S.P., Hoyt, R.F., and Shaffer, M.J. 1997. Ontogeny of neuroepithelial bodies: correlations with mitogenesis and innervation. Microsc. Res. Tech., 37: 43-61. Soukhova, G., Wang, Y., Ahmed, M., Walker, J.F., and Yu, J. 2003. Bradykinin stimulates respiratory drive by activating pulmonary sympathetic afferents in the rabbit. J. Appl. Physiol., 95: 241-249. Speirs, V., and Cutz, E. 1993. An overview of culture and isolation methods suitable for in vitro studies on pulmonary neuroendocrine cells. Anat. Rec., 236: 35-40. Speirs, V., Wang, Y.V., Yeger, H., and Cutz, E. 1992. Isolation and culture of neuroendocrine cells from fetal rabbit lung using immunomagnetic techniques. Am. J. Respir. Cell. Mol. Biol., 6: 63-67. Springall, D.R., and Polak, J.M. 1993. Calcitonin gene-related peptide and pulmonary hypertension in experimental hypoxia. Anat. Rec., 236: 96-104. Springall, D.R., Cadieux, A., Oliveira, H., Su, H., Rayston, D., and Polak, J.M. 1987. Retrograde tracing shows that CGRP-immunoreactive nerves of rat trachea and lung originate from vagal and dorsal root ganglia. J. Auton. Nerv. Syst., 20: 155-166. Tabassian, A.R., Nylen, E.S., Giron, A.E., et al. 1988. Evidence for cigarette smoke induced calcitonin secretion from lungs of man and hamster. Life Sci., 42: 2323-2329. Tabassian, A.R., Snider, R.H., Nylen, E.S., Cassidy, M.M., and Becker, K.L. 1993. Heterogeneity studies of hamster calcitonin following acute exposure to cigarette smoke: evidence for a monomeric secretion. Anat. Rec., 236: 253-256. Takamori, S. 2006. VGLUTs: ‘Exciting’ times for glutamatergic research? Neurosci. Res., 55: 343-351. Takamori, S., Rhee, J.S., Rosenmund, C., and Jahn, R. 2000. Identification of a vesicular glutamate transporter that defines a glutamatergic phenotype in neurons. Nature, 407: 189-194. Terada, M., Iwanaga, T., Takahashi-Iwanaga, H., Adachi, I., Arakawa, M., and Fujita, T. 1992. Calcitonin gene-related peptide (CGRP)-immunoreactive nerves in the tracheal epithelium of rats: an immunohistochemical study by means of whole mount preparations. Arch. Histol. Cytol., 55: 219-233. Thompson, R.J., Doran, J.F., Jackson, P., Dhillon, A.P., and Rode, J. 1983. PGP 9.5-a new marker for vertebrate neurons and neuroendocrine cells. Brain Res., 278: 224-228. Tjen-A-Looi, S., Kraiczi, H., Ekman, R., and Keith, I.M. 1998. Sensory CGRP depletion by capsaicin exacerbates hypoxia-induced pulmonary hypertension in rats. Regul. Pept., 74: 1-10. Uddman, R., Luts, A., and Sundler, F. 1985. Occurrence and distribution of calcitonin generelated peptide in the mammalian respiratory tract and middle ear. Cell Tissue Res., 241: 551-555 (abstract). Van Genechten, J., Brouns, I., Scheuermann, D.W., Timmermans, J.-P., and Adriaensen, D. 2003. Reduced number of intrinsic pulmonary nitrergic neurons in Fawn-Hooded rats as compared to control rat strains. Anat. Rec., 272A: 446-453. Van Genechten, J., Brouns, I. Burnstock, G., Timmermans, J.-P., and Adriaensen, D. 2004. Quantification of neuroepithelial bodies and their innervation in Fawn-Hooded and Wistar rat lungs. Am. J. Respir. Cell. Mol. Biol., 30: 20-30. Van Lommel, A. 2001. Pulmonary neuroendocrine cells (PNECs) and neuroepithelial bodies (NEB): chemoreceptors and regulators of lung development. Paediatr. Respir. Rev., 2: 171176.

Diverse and Complex Airway Receptors in Rodent Lungs  |  267 Van Lommel, A., and Lauweryns, J.M. 1993. Neuroepithelial bodies in the Fawn Hooded rat lung: morphological and neuroanatomical evidence for a sensory innervation. J. Anat., 183: 553-566. Van Lommel, A., Lauweryns, J.M., and Berthoud, H.-R. 1998. Pulmonary neuroepithelial bodies are innervated by vagal afferent nerves: an investigation with in vivo anterograde Dil tracing and confocal microscopy. Anat. Embryol., 197: 325-330. Van Lommel, A., Bolle, T., Fannes, W., and Lauweryns, J.M. 1999. The pulmonary neuroendocrine system: the past decade. Arch. Histol. Cytol., 62: 1-16. Verástegui, C., Fernandez-Vivero, J., Prada, A., Rodriguez, F., Romero, A. GonzalezMoreno M., and de Castro, J.M. 1997a. Presence and distribution of 5HT-, VIP-, NPY-, and SPimmunoreactive structures in adult mouse lung. Histol. Histopathol., 12: 909-918. Verástegui, C., Oliveira, A.P., Fernández-Vivero, J., Romero, A., and de Castro, J.M. 1997b. Calcitonin gene-related peptide immunoreactivity in adult mouse lung. Eur. J. Histochem., 41: 119-126. Wang, D., Youngson, C.R., Wong, V., Yeger, H., Dinauer, M.C., Vega-Saenz, M.E., Rudy, B., and Cutz, E. 1996. NADPH-oxidase and a hydrogen peroxide-sensitive K+ channel may function as an oxygen sensor complex in airway chemoreceptors and small cell lung carcinoma cell lines. Proc. Natl. Acad. Sci USA, 93: 13182-13187. Wang, Y., Soukhova, G., Proctor, M., Walker, J., and Yu, J. 2003. Bradykinin causes hypotension by activating pulmonary sympathetic afferents in the rabbit. J. Appl. Physiol., 95: 233-240. Wang, Z.J., and Neuhuber, W.L. 2003. Intraganglionic laminar endings in the rat esophagus contain purinergic P2X2 and P2X3 receptor immunoreactivity. Anat. Embryol., 207: 363-371. Wasano, K. 1977. Neuroepithelial bodies in the lung of rat and mouse. Arch. Histol. Jpn., 40: 207-219. Wasano, K., and Yamamoto, T. 1978. Monoamine-containing granulated cells in the frog lung. Cell Tissue Res., 193: 201-209. Wasano, K., and Yamamoto, T. 1981. A scanning and transmission electron-microscopic study on neuroepithelial bodies in the neonatal mouse lung. Cell Tissue Res., 216: 481-490. Widdicombe, J.G. 2001. Airway receptors. Respir. Physiol., 125: 3-15. Widdicombe, J.G., and Nadel, J.A. 1963. Airway volume, airway resistance, and work and force of breathing: theory. J. Appl. Physiol., 18: 863-868. Wilkinson, K.D., Lee, K.M., Deshpande, S., Duerksen-Hughes, P., Boss, J.M., and Pohl, J. 1989. The neuron-specific protein PGP9.5 is a ubiquitin carboxyl-terminal hydrolase. Science, 246: 670-673. Wu, S.-X., Koshimizu, Y., Feng, Y.-P., Okamoto, K., Fujiyama, F., Hioki, H., Li, Y.-Q., Kaneko, T., and Mizuno, N. 2004. Vesicular glutamate transporter immunoreactivity in the central and peripheral endings of muscle-spindle afferents. Brain Res., 1011: 247-251. Youngson, C., Nurse, C. Yeger, H., Curnutte, J.T., Vollmer, C., Wong, V., and Cutz, E. 1997. Immunocytochemical localization on O2-sensing protein (NADPH oxidase) in chemoreceptor cells. Microsc. Res. Tech., 37: 101-106. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature, 365: 153-155. Yu, J., Lin, S.X., Zhang, J.W., and Walker, J.F. 2006. Pulmonary nociceptors are potentially connected with neuroepithelial bodies. Adv. Exp. Med. Biol., 580: 301-306.

268  |  Airway Chemoreceptors in the Vertebrates Yu, J., Wang, Y.F., and Zhang, J.W. 2003. Structure of slowly adapting pulmonary stretch receptors in the lung periphery. J. Appl. Physiol., 95: 385-393. Yu, J., Zhang, J., Wang, Y., Fan, F., and Yu, A. 2004. Neuroepithelial bodies not connected to pulmonary slowly adapting stretch receptors. Respir. Physiol. Neurobiol., 144: 1-14. Zhang, J.W., Walker, J.F., Guardiola, J., and Yu, J. 2006. Pulmonary sensory and reflex responses in the mouse. J. Appl. Physiol., 101: 986-992.

12 Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies E. Cutz1*, W.X. Fu2, H. Yeger1, J. Pan1 and C.A. Nurse3

Abstract The aim of this chapter is to review current evidence in regard to the function of pulmonary Neuroepithelial bodies (NEB’s) as hypoxia sensitive airway sensors in mammalian lungs. Typical NEB’s form clusters of innervated amine (serotonin, 5-HT) and peptide (i.e. bombesin, CGRP) producing cells widely distributed within the airway epithelium, with preferential location at airway branch points. Recent electrophysiological studies using the patch-clamp technique have demonstrated that NEB cells are transducers of hypoxia stimulus via a membrane bound molecular complex (”oxygen sensor”), characteristic of specialized cells that monitor and signal hypoxia in the body to maintain homeostasis. The O2 sensor in NEB cells has been partially characterized and consists of an H2O2 producing, multicomponent NADPH oxidase coupled to O2 sensitive K+ current. Under normoxia, H2O2 derived from the oxidase promotes K+ channel open state activity, whereas during hypoxia reduction in H2O2 leads to K+ channel closure, membrane depolarization and release of amine/peptide neurotransmitters. Studies of native NEB cells after isolation and culture or in fresh lung slices, related tumor cell line and k/o mouse models are reviewed focusing on cellular, molecular and electrophysiological approaches. The expression and function of the various components of "phagocytic" NADPH oxidase (gp91phox/NOX 2), recently identified NOX homologues and their possible role in O2 sensing is discussed. The mechanisms of chemotransmission of hypoxia and other stimuli from NEB cells to the brain stem via nodose neuron derived vagal afferents are reviewed. The

Division of Pathology, Department of Paediatric Laboratory Medicine, The Research Institute, CIHR Lung Development Group (CIHR grants * FRN15270; FRN 82153), The Hospital for Sick Children, Toronto, Ontario, Canada, Department of Laboratory Medicine and Pathobiology, University of Toronto. 2 Division of Neuroscience, Oregon Health Science University, Beaverton, Oregon, USA. 3 Department of Biology, McMaster University, Hamilton, Ontario, Canada. *Author for correspondence: Email: [email protected] Phone: (416)813-5966, Fax (416) 8135974 1

270  |  Airway Chemoreceptors in the Vertebrates functional role of NEB, particularly the possible involvement in control of respiration during neonatal adaptation and various perinatal disease processes is discussed.

Keywords: hypoxia airway sensors, peripheral chemoreceptors, O2 sensor molecular complex, NADPH oxidase, O2 sensitive potassium channels, neonatal control of respiration

Introduction The early history of pulmonary neuroendocrine cell (PNEC) system including NEB has been reviewed by Sorokin and Hoyt in their comprehensive review on pulmonary small granule cells [Sorokin and Hoyt, 1989]. The first description of these cells dates back to the 1930s when single and groups of “clear” cells with argyrophilic cytoplasm have been reported in the airway mucosa of human and animal lungs [Feyerter, 1938]. Initially, these cells were believed to be part of a diffuse paracrine system of cells distributed in different organs and serving local endocrine/paracrine function. The idea that NEB may represent airway sensors, monitoring changes in airway gas concentration was proposed in early neuroanatomical studies that highlighted their “corpuscular” structure, the presence of innervation, as well as histological features reminiscent of the taste buds [Feyerter, 1938; Sorokin and Hoyt, 1989]. The first experimental evidence indicating that NEB could function as hypoxia sensitive airway chemoreceptors, modulated by CNS was provided in a series of studies by Lauweryns and co-workers [1972, 1977]. They have shown in an neonatal rabbit model that exposure to acute hypoxia caused enhanced exocytosis of dense core vesicles and reduced amine fluorescence in NEB cells implying hypoxia mediated amine release [Laweryns et al., 1977]. Using various vagotomy procedures they also documented extensive afferent-like innervation of NEB’s derived from the vagus nerve, with the nerve cell bodies residing in the nodose ganglion [Lauweryns and Van Lommel, 1986]. Further advances in understanding the cellular and molecular biology of NEB cells have been facilitated with the development of in vitro models suitable for electrophysiological, biochemical and molecular analysis of its membrane and/or cytoplasmic components. The methods for isolation and culture of NEB as well as fresh lung slice model have allowed patch-clamp analysis of membrane currents in addition to direct studies on neurotransmitter synthesis and release [Youngson et al., 1993; Fu et al., 1999; 2002]. Availability of human tumor cell lines, representative of PNEC/NEB in native lungs, namely small cell lung carcinoma (SCLC) cell lines have been useful for studies using biochemical and molecular approaches not feasible in native cells due to their small number and wide distribution within the lung tissue [O’Kelly et al., 1998; 1999; 2000]. The principal aim of this chapter is to review and discuss recent advances in cellular and molecular biology of NEB, their possible function as airway O2 sensors,

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  271

particularly during neonatal adaptation and their potential involvement in a variety of perinatal disorders.

General Morphologic Features of Neb Morphologically, NEB are defined as innervated clusters of amine and peptidecontaining cells that are widely distributed within the airway mucosa of mammalian lungs [Sorokin and Hoyt, 1989; Cutz, 1997; Cutz et al., 2003]. The main structural features that support NEB function as chemoreceptors include: (i) corpuscular arrangement with preferential location at airway branch points; (ii) apical microvilli in contact with airway lumen; (iii) the presence of cytoplasmic dense core vesicles (DCV), the storage site of amine (serotonin, 5-HT) and a variety of peptides (i.e. bombesin, calcitonin [in human], calcitonin gene related peptide, CGRP [in rodents]), that act as neurotransmitters/neuromodulators; and (iv) sensory innervation via vagal afferents derived from the nodose ganglion [Adriaensen et al.,1998]. Many of the morphologic features of NEB are remarkably similar to glomus cells of the carotid body (CB) a well-defined arterial chemoreceptor, that monitor levels of pO2, pCO2 and pH in the blood [González et al., 1994]. Quantitative studies on the distribution and frequency of NEB revealed striking developmental changes with prominence in fetal/neonatal lungs and decline post-natally [Cho et al., 1989]. This suggests that NEB’s are most active during the perinatal period, perhaps complementing CB chemoreceptor function at the time when they are not fully mature [Van Lommel and Lanweryns, 1999]. NEB appear more prominent and numerous in neonates of animal species with relatively immature lungs at birth and hence may require an airway based sensor that detects hypoxia in advance of CB, to protect them against hypoxia [Van Lommel et al.,1997]. Furthermore the innervation of NEB at birth appears more mature compared to CB at the same stage of development [Bolle et al., 2000]. Recent studies using multilabel immunohistochemistry combined with confocal microscopy revealed complexity of NEB innervation [Adriaensen et al., 2003] (see Chapter by Brouns et al. for further details). In rat lungs, NEB innervation consist of three components, namely vagal afferent fibers immunoreactive for Calbindin D21K and P2/X2/3 purinoreceptors that project to the nodose ganglion; CGRP immunoreactive nerve fibers originating in the spinal ganglia; and n NOS immunoreactive nerve fibers derived from intrinsic peribronchial ganglia [Brouns et al., 2002a].There are likely wide species variations in NEB innervation since the rabbit seems to lack CGRP immunoreactive nerve fibers (unpublished observation).The development of innervation of NEB’s has been recently investigated in fetal/neonatal rabbit lungs. The earliest intraepithelial nerve endings in contact with emergent PNEC/NEB cells were observed at E18, when the lungs consist of a few primitive tubes surrounded by abundant mesenchyma [Pan et al.,2004]. Since not all early

272  |  Airway Chemoreceptors in the Vertebrates PNEC/NEB cells had neural contacts, this suggested that the differentiation of these cells can occur independently of airway innervation. Well defined corpuscular NEB with a few nerve fibers entering at the base of cell clusters were observed at E21 with an extensive intracorpuscular nerve network apparent at birth. These findings suggest that NEB become innervated early during lung development and that the full complement of innervation is reached at the time of birth to allow the function as airway sensors involved in neonatal adaptation to air breathing.

Cellular and Molecular Mechanisms of Oxygen (O2) Sensing A number of studies have confirmed that in O2-sensitive neurosecretory cells (such as CB glomus cells, pulmonary NEB and adrenal chromaffin cells), acute responses to hypoxia are mediated by ion channels [Lopez-Barneo et al., 1993; 2001]. In these cells a well-studied acute effect of hypoxia is the inhibition of O2 sensitive K+ channel causing membrane depolarization, activation of voltage gated Ca2+ channels followed by influx of Ca2+ triggering exocytosis of various neurotransmitters. These are electrically excitable cells that function as chemoreceptors monitoring global O2 tension to produce cardiorespiratory adjustments to low ambient pO2 . In contrast to other cell types, chemoreceptor cells have some tonic activity at normal pO2 levels (~90-100 mmHg), but they begin to be fully activated with moderate levels of hypoxia (<50-60 mmHg). Early studies in the field of O2 sensing predicted a universal O2 sensing mechanism [Lopez-Barneo et al., 1993]. However recent evidence indicates that instead of a monolithic O2 sensor, the hypoxia signal transduction pathway appears to be cell-type specific [Lopez-Barneo et al., 2001]. The primary event in the original “membrane model”, was the closure of O2-sensitive K+ channel with three possibilities: (a) the ion channel itself is an O2 sensor; (b) the ion channel is modulated by an independent O2 sensor via a diffusible cytoplasmic mediator; and (c) the ion channel is closely associated with a membrane bound O2 sensing protein, such as heme-linked NADPH oxidase, and the interaction occurs via a membrane delimited pathway. Evidence is accumulating in support of all three alternatives in different tissues and cells [LopezBarneo et al., 2001;Weir et al., 2005]. Our findings in pulmonary NEBs (see below) favor the third option where there is close interaction between a membrane-bound multicomponent NADPH oxidase and O2-sensitive K+ channels [K+(O2)][Youngson et al., 1993; Wang et al., 1996]) (Figure 1).

NADPH-Oxidase as a pO2 Sensor Protein in Pulmonary NEB An important aspect of the “membrane model” of pO2 chemotransduction is the identity of a membrane bound O2 sensor protein. A potential candidate was postulated

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  273

Figure 1: Schematic representation of membrane-bound O2 sensor postulated to operate in pulmonary NEB and other O2 sensing cells (i.e. carotid body [CB] glomus cells, pulmonary artery [PA] smooth muscle cells, adrenal medulla [AM] cells). (1) NEB cells express O2 sensing protein) representing multicomponent NADPH oxidase/ NOX2 consisting of membrane (gp91phox, p22phox) and cytoplasmic (p47, p67, rac2GTP) components. Under normoxia, NADPH oxidase tonicaly generates reactive oxygen species (ROS) [O2-; H2O2] that keep the effector Kv channel in oxidized or open state. During hypoxia, reduced production of activated oxygen species (AOS) including H2O2 leads to reduced or closed channel state modulated via cystein residues in O2-sensitive voltage activated K+ channels (Kv); (the gating mechanism for TASK channels is presently unknown). (2) Hypothetically, novel (so called low output oxidases NOX1, NOX4) identified in NEB cells (see Figure3), could play similar role in modulating O2-sensitive K+ channels (see text). In other O2-sensing cells, (CB, PA and AM). (3) Mitochondria rather than membrane bound NADPH oxidases are postulated to represent major O2 sensor in other O2-sensing cells (CB, PA and AM). (Modified from S.Archer and E. Michelakis, 2002 with permission).

by Acker et al. [1989], Acker and Xue [1995] and Cross et al. [1990] to be the hemelinked nicotinamide adenine dinucleotide phosphate oxidase (NADPH oxidase), similar to the one identified in neutrophils that consists of a catalytic membrane bound subunit, cytochrome b558, an integral membrane heterodimer containing gp91phox and p22phox, and regulatory subunits p47phox, p67phox and rac2 that constitute the functional NADPH oxidase [Quinn and Gauss, 2004]. The evidence in support of NADPH oxidase as a principal O2 sensor in NEB cells includes co-expression of mRNAs for both the membrane components of NADPH

274  |  Airway Chemoreceptors in the Vertebrates oxidase (gp91phox and p22phox) and hydrogen peroxide (H2O2) sensitive voltage gated K+ channel subunit KV3.3a in NEB cells of fetal rabbit and human neonatal lungs as well as in related small cell lung carcinoma (SCLC) cell lines [O’Kelly et al., 1998; 1999; 2000; Wang et al., 1996]. Immunohistochemical studies using specific antibodies against gp91phox, p22phox as well as cytosolic subunits p47phox, p67phox have localized these epitopes in the membrane and submembrane regions of NEB cells in culture [Youngson et al., 1997]. Of interest is our recent finding of localization of gp91phox in the apical cytoplasm of native NEB cells in-situ, the presumed site of the O2 sensor(Figure 2) (unpublished observation). Studies using a small cell lung carcinoma (SCLC) cell line H-146, an immortalized cell model for NEB cells, also supports the role of the oxidase as an O2 sensor [O’Kelly et al., 2000; Cutz et al., 2003]. This model confirmed the up-regulation of NADPH oxidase activity under normoxia by protein kinase C-dependent phosphorylation of p67phox and p47phox. Specific inhibitors of the electron transport chain and H-146 cells depleted of mitochondria (p0) retained their O2 sensing properties indicating that in NEB cell related tumor cell model, mitochondria are not required for acute O2 sensing [Searle et al.,2002].The definitive proof for the role of the oxidase in O2 sensing by NEB cells comes from a study of a mouse model with NADPH oxidase deficiency [OD; gp91phox k/o] [Fu et al., 2000]). In OD mice, exposure to acute hypoxia failed to reduce O2 sensitive K+ current in NEB cells (i.e., NEB failed to respond to hypoxia), whereas in control wild-type mice this response was intact. Furthermore, NADPH oxidase inhibitor diphenylene iodonium,

Figure 2: Immunohistochemical localization of NADPH oxidase protein gp91 phox/ NOX2 (green signal) in apical cytoplasm (arrow) of NEB cells in neonatal rat lung. Antibody against synaptic vesicle2 (SV2)was used to identify NEB cells (red signal) in lung sections. Double immunoflourescence on frozen sections using regionspecific rabbit antibody against gp91phox (N-terminal; courtesy Dr. M. C. Dagher, Univ. Grenoble). (Confocal microscopy; 1,000)

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  275

DPI (1uM), had no effect on K+ current, while external application of H2O2 (putative second messenger) increased K+ current indicating that in NEB of OD mice the redox modulation is not affected by the abrogation of the oxidase function. In addition, neonatal OD mice exhibited an abnormal pattern of respiration characterized by rapid shallow breathing compared to wild type controls when assessed by whole body plethysmography [Kazemian et al., 2002]. Interestingly, studies on other O2 sensing cells (i.e. CB, adrenal medullary cells, pulmonary artery smooth muscle cells) in the same OD mouse model revealed normal hypoxia responses [Archer et al., 1999; Thompson et al., 2002]. Taken together, the above studies provide strong evidence for a hypothesis of a membrane delimited mechanism for O2 sensing by NEB cells which may differ from other O2 sensing cells (Figure 1).

Homologues of NADPH Oxidase (Nox 1-4), Distribution and Function. Recent studies have identified homologues of NADPH oxidase in a variety of nonphagocytic cells [Lambeth, 2004]. These “low output” oxidases have been grouped into two families of gp91phox homologues: the Nox (NADPH oxidase homologues) and Duox (dual oxidases).The founder protein, gp91phox (Nox2) is predominantly expressed in phagocytic cells such as neutrophils and macrophages where it plays a critical role in host defense. The oxidase activity in these cells is strictly regulated, since inappropriate or excessive production of ROS results in tissue damage. Activation of phagocyte oxidase gp91phox/Nox 2 requires stimulus-induced membrane translocation of cytosolic regulators including the small GTPase Rac and the two cytosolic proteins p47phox and p67phox [Quinn and Gauss, 2004, Lambeth, 2004]. Nox 1 is expressed mostly in colonic epithelium where it is thought to play a role in local host defenses. This oxidase can form a complex with p22phox, but the complex is an inactive form by itself, however it can be activated when it is co-expressed with Noxo1 and Noxa1, homologues of p67phox and p47phox respectively [Sumimoto et al., 2004]. Nox 3 is expressed mainly in fetal kidney and the inner ear. In contrast to Nox 1 and Nox 2, expression of Nox 3 alone in various cell types can generate NADPH dependent production of superoxide, without cytosolic regulatory proteins p47phox, p67phox or their homologues [37, 38]. Nox 4, originally described as a renal oxidase (Renox), is widely expressed in various tissues, such as, the lung, placenta, pancreas, bone and blood vessels where it may be involved in various cellular processes such as cell proliferation, apoptosis and receptor signaling [Lambeth, 2004]. The regulatory factors and mechanisms of activation for Nox 4 have not yet been determined. In our preliminary studies using molecular profiling of NEB cells recovered by laser capture micro dissection from a human infant lung expression of all NOX proteins except for NOX 3 were identified (Figure 3).Although it is presumed that by analogy with NOX 2, NOX 1 and NOX 4 could also play a role in O2 sensing (Figure 1), their precise function in NEB cells is at present unknown.

276  |  Airway Chemoreceptors in the Vertebrates

Figure 3: O2 sensor gene expression in NEB cells of human infant lung using SuperArray human custom multigene 12 RT-PCR profiling kit. (A) PNEC/NEB cells immunostained with antibody against chromogranin A (CGA) (darkly stained) in epithelium of small airway in lung section from 3 mon control infant (10% formalin fixation, embedding in paraffin). (B) Laser cupture microdissection (LCM) using frozen sections of lung from the same case as in A. NEB cells were identified by immunofluorescence (IF) labeling with antibody against SV2. NEB cell clusters are shown before and after retrieval (capture). (C) NADPH oxidase gene profile of human NEB cells retrieved by LCM. All membrane components (NOX 1, 2&4, p22) except NOX 3, and all cytosolic (p40, p47, p67, rac1, NOXO1, NOXA1) components are abundantly expressed. (D) O2 sensitive K+ channel gene profile includes TASK1-3, Kv3.4 and Kv3.3 confirmed by pharmacological and electrophysiological studies. Note that Kv4.3, identified in rabbit NEB (Fu et al., 2007) is apparently not expressed in human NEB suggesting species variation (see text). [GAPDH-housekeeping gene; TH-tryptophane hydroxylase, NEB cell specific gene].

O2 Sensitive K+ Channels and their Interactions with NOX Proteins A wide variety of K+ channels belonging to different classes and families of K+ currents that are modulated by ambient O2 tension have been identified [for review see Patel and Honore, 2001]. We have reported co-expression of mRNAs encoding Shaw-related + KV3-3a K channel protein with gp91phox/NOX 2 and p22phox of NADPH oxidase in NEB cells of both fetal rabbit and neonatal human lungs as well as in several SCLC lines [Wang et al., 1996]. In our earlier electrophysiological studies using cultures of isolated NEB or in fresh lung slice model we have identified slowly inactivating

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  277

K+ current corresponding to Kv3.3a [Yongson et al., 1993; Wang et al., 1996] and a non inactivating outward delayed rectifier K+ current [Fu et al., 1999] (Figures 4 and 5). More recently we have identified in NEB cells of rabbit neonatal lung additional hypoxia-sensitive voltage activated K+ channels, namely Kv 3.4 and Kv 4.3 that belong to the Shaw and Shal-related family respectively [Fu et al., 2007](Figures 6 and 7). Interestingly immunohistochemical studies show that in rabbit NEB Kv3.4 epitope is restricted to the apical plasma membrane, a location expected for an airway O2 sensor (Figure 8).These voltage activated K+ channels, characterized by slowly inactivating A-type currents have been also identified in other O2 sensing cells particularly carotid body glomus cells [Sanchez et al., 2002]. In rabbit CB glomus cells, both Kv3.4 and Kv4.3 participate in O2-sensitive K+ current, and Kv4.1 and Kv4.3 contribute to heteromultimeric O2-sensitive K+ channel [Sanchez et al., 2002]. These studies suggest that multiple subtypes of hypoxia-sensitive voltage-gated K+ channels could be expressed in the same cell. The possibility of species variation in expression of various Kv alpha subunits is suggested by our finding that while Kv4.3 is abundantly expressed in NEB cells of rabbit neonatal lung (Fu et al., 2007), it is apparently lacking in NEB cells of a human infant lung (Figure 3D). The precise role of different O2-sensitive K+ channel subunits (Kv3.3a, Kv3.4, Kv4.3) in NEB cells is at present unknown and requires further investigation. It is of interest that both Kv3.3a and Kv3.4 are redox sensitive, therefore these Kv subunits could form heteromultimeric O2-sensitive K+ channels modulated by H2O2 generated by NADPH oxidase (NOX 2) [Vega-Saenz de Meira and Rudy, 1992]. Additional hypoxia-sensitive K+ current belonging to acid-sensitive K+ channel of the tandem P domain K+ channel family (TASK) have been identified in H146 cells, a tumor cell model for native NEB cells [Lewis et al., 2001; Hartness et al., 2001]. It has been suggested that TASK-1/3, that is voltage independent and active at a resting membrane

Figure 4: Neutral red staining of NEB cells in vivo (arrowheads) in fresh slice of neonatal rabbit lung.

278  |  Airway Chemoreceptors in the Vertebrates

Figure 5: Effect of hypoxia on non-activating, outward delayed rectifier K+ current (representative of Kv3.3) in NEB cell in neonatal rabbit lung slice preparation. (A) Outward non-inactivating delayed-rectifier K+ current evoked by depolarizing steps from -60 to +30 mV in control normoxia solution. (B) Outward current evoked by same voltage steps as in (A) was reversibly reduced ~40% by hypoxia (PO2~20 mmHg), with full recovery after wash (C). (D) I-V relationship for the currents in NEB cell demonstrated in A, B & C. (reproduced with permission from Fu et al., 1999).

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  279

Figure 6: Effect of blood-depressing substance I (BDS-I)on fast transient slowly inactivating K+ current (Kv 3.4) in NEB cells of neonatal rabbit lung. (A) Fast transient slowly inactivating K+ current evoked by depolarizing steps from -90 to +30mv in control Krebs solution. (B) Kv current was reduced after perfusion with 3µm BDS-I (venom of sea anemone ,Anemonia sulcata, specific blocker of Kv3.4). (C) Recovery of K+ current after BDS-I wash-out. (D) BDS-I sensitive K+ current was obtained by subtracting currents in B from those in A. (E) the peak current of K+ with I-V relationship was plotted under control condition (•), after perfusing 3 µM BDS-I (Ο), and the BDS-I sensitive K+ fast inactivating current (s). All recordings were obtained from the same NEB cell, I-V relationship of the mean ±S.E.M elicited in 5 NEB cells (reproduced with permission from Fu et al., 2007).

potential, rather then KV3-3a, mediates hypoxia-evoked depolarization in H-146 cells and possibly also in native NEB [Hartness et al., 2001; Kemp et al., 2002]. The evidence for H2O2 modulation of K+ channels is derived from studies of voltage activated K+ channels [Ruppersberg et al., 1991]. In this model H2O2 may serve as the second messenger in cells responsive to O2. The common feature of channels affected by H2O2 is a cysteine residue in the amino terminus shown to be highly sensitive to redox state [Ruppersberg et al., 1991; Vega-Saenz de Meira and Rudy, 1992].This amino terminus is believed to be intracellular and contains a site that is responsible for channel inactivation acting as a tethered “ball and chain” which occludes the internal mouth of the channel. An updated model of an O2 sensor in NEB proposed by Patel and Honore [2001] incorporates our earlier model, where H2O2 producing NADPH oxidase is closely associated with O2 sensitive K+ channel [Wang et al., 1996]. Possible involvement of Nox 4 as an O2 sensor that may regulate TASK-1 channel activity and hence play a role in Os2 sensing has been recently suggested by Lee et al. [2006]. These authors have shown that in HEK293 cells which endogenously express

280  |  Airway Chemoreceptors in the Vertebrates

Figure 7: Effects of spider venom,heteropodatoxin-2 (HpTx-2) on fast transient slowly inactivating K+ current (Kv4.3) in NEB cells. (A) K+ current evoked by depolarizing steps from -90 to +30mv in control condition. (B) K+ current was reduced by perfusion of 0.2 µm HpTx-2 (specific blocker of Kv4.3). (C) Recovery of the K+ current following HpTx-2 wash out. D, HpTx-2 sensitive K+ current was obtained by subtracting currents in B from those in A. (E) Peak K+current with I-V relationship plotted under control condition (•), after perfusion with 0.2 µm HpTx-2 (Ο), and BDS-I sensitive K+ current (∆). All recordings were obtained from the same NEB cell, I-V relationship of the means ±S.E.M elicited in 5 NEB cells (reproduced with permission from Fu et al., 2007).

Figure 8: Immunohistochemical localization of Kv3.4 in NEB cells of rabbit neonatal lung. Corpuscular NEB identified by immunostaning for serotonin (5-HT) in cytoplasm (green signal). Distinctive apical membrane localization of Kv3.4 epitope (red signal) in (a) longitudinal plane of section (arrow); (b) oblique or en face view with linear membrane pattern. (Confocal imaging using frozen section/paraformadehyde-zink fixative and rabbit polyclonal antibody against Kv3.4 [Alamone]).

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  281

Nox 4, the activity of transfected TASK-1 channel was inhibited by hypoxia. This hypoxia response was significantly augmented by co-transfection of Nox 4, but no change was observed when Nox 2/gp91phox was used. The O2 sensitivity of TASK-1 was abolished by Nox 4 siRNA and NADPH inhibitors suggesting that Nox 4 may represent O2 sensor protein partner for TASK-1. This suggests a possibility of a diversity of “O2 sensors”, even within the same cell type, matching specific Nox proteins with particular O2 sensitive K+ channel types (i.e. Nox 2/KV3.3a; Nox 4/TASK-1). Our updated model of hypoxia signal trasduction pathway for NEB cells and H146 cells is based on a model originally proposed by Kemp et al. [2002] and incorporates putative O2 sensing molecular complexes composed of different O2 sensitive K+ channels and respective NOX proteins (Figure 9).

Figure 9: The proposed signal transduction pathway for NEB cells and related human tumor cell model (H146 cell line). Native NEB cells and H146 cells share a common primary O2 sensor, NADPH oxidase. Recent findings suggest that both classical gp91phox/ NOX 2 and novel oxidase NOX 4 are expressed (see text). Hypoxia causes reduction in substrate–delimited production of reactive oxygen species (ROS) including H2O2. This in turn leads to reduction in the cellular redox potential that causes closure of different Kv ‘s and TASK channels. Recent experimental findings suggest that gp91phox/NOX 2 represents partner protein for Kv channels whereas NOX 4 may be associated with TASK channel (for details see text). In both NEB cells and H146 cells O2 sensitive K+ channel inhibition caused membrane depolarization, opening of voltage activated Ca2+ channels and subsequent transmitter release including amine (5-HT) and various peptides. (Modified from Kemp P.J. et al. with permission).

282  |  Airway Chemoreceptors in the Vertebrates

Mechanisms of Chemotransduction of Hypoxia Stimulus The candidate neurotransmitters that mediate fast chemosensory transmission of hypoxia signal from NEB cells to CNS via vagal sensory afferents include 5-HT, acetylcholine (Ach) and adenosine triphosphate (ATP), as recently proposed for the CB glomus cells [Zhang et al., 2000]. Evidence in support of 5-HT as a transmitter of hypoxia stimulus includes earlier in vivo and in vitro studies on NEB cells in a neonatal rabbit model [Lauweryns et al., 1977; Cutz et al., 1993]. In a more recent study, using fresh neonatal rabbit lung slice preparation and carbon fiber amperometry (a method that detects in real time the release of single molecules of 5-HT) has shown that hypoxia causes a dose-dependent 5-HT release from NEB cells within a physiologic range expected in the airway (i.e. pO2<95 mmHg) [Fu et al., 2002]. A large proportion of vagal afferent nerve fibers derived from the nodose neurons that innervate NEB cells express 5-HT-3 receptors that upon activation initiate neurotransmission of hypoxia at the synapse level [Higashi and Nishi, 1982]. In addition NEB cells also express ionotropic 5-HT-3 receptors that can act as autoreceptors providing positive feed back and thus amplify hypoxia signaling [Fu et al., 2001]. The evidence for cholinergic mechanisms in NEB include demonstration of acetylcholinesterase activity, the presence of small clear vesicles in efferent-like nerve endings demonstrated by electron microscopy, and positive immunoreactivity for vesicular acetylcholine transporter (VChat) [Lauweryns and Van Lommel, 1986; Brouns et al., 2002b]. NEB cells express different subtypes of nicotinic Ach receptors, which are critical in mediating the effects of nicotine derived from cigarette smoking, but do not seem to be directly involved in hypoxia signaling [Fu et al., 2003].The possible involvement of ATP and purinergic mechanisms is based on the demonstration of ATP in NEB cell cytoplasmic dense core granules and expression of P2/X receptors on nerve endings innervating NEB [Brouns et al.,2000] as well as on NEB cells proper [Fu et al., 2004].As with 5-HT receptors, P2X/2/3 receptors are abundantly expressed on vagal afferent nerve endings terminating on NEB cells that are derived from the nodose ganglion neurons and thus likely participate in neurotransmission of hypoxia stimulus from the airway to the brain stem [Brouns et al. 2000]. It was also found that heteromeric P2X2/3 receptors expressed on NEB cells may function as auto receptors involved in modulation of hypoxia and other stimuli [Fu et al., 2004]. To facilitate the study of cellular and molecular mechanisms of chemotransmission of hypoxia/hypercapnia stimuli in CB glomus cells, a reductionist co-culture model of a glomus cells-petrosal neurons has been developed [Nurse and Zhang, 1999]. In this model, clusters of isolated glomus cells become re-innervated and form functional synapses with petrosal neurons maintained in vitro that discharge action potentials upon exposure to hypoxia/hypercapnia, approximating closely the situation in vivo [Zhang and Nurse, 2004]. A similar model was developed using co cultures of NEB

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  283

cells isolated from rabbit fetal lung with nodose ganglion (NG) neurons derived from rabbit neonates that retain their plasticity and re innervate NEB cells in vitro (unpublished data). In preliminary experiments it was found that NG neurons when cultured alone showed no action potential discharges upon hypoxia exposure, however when NEB cells and NG neurons were cultured together it was observed hypoxia induced depolarization and/or increased spike discharge from NG neurons closely associated with NEB cells, indicating the presence of functional synapses. In addition, this hypoxia induced chemotransmission was blocked by specific blockers of 5-HT-3, nACh and purinergic receptors, confirming that 5-HT, Ach and ATP are the respective neurotransmitters of hypoxia stimulus in NEB cells. Although these findings need to be replicated in whole animal studies, our data provide strong evidence in support of the hypothesis that NEB function as airway sensors that detect and signal changes in pO2 levels from the airway to CNS.

Control of Respiration and the Role of Vagal Afferents At the present time direct evidence is lacking for transmission of hypoxia generated signal by NEB cells via vagal afferent fibers to the brain stem. Furthermore there are no conclusive data documenting the role of NEB cells in the control of respiration. Previous studies on the effects of hypoxia on vagal afferents have given negative results; apparently neither slowly adapting (SARs), rapidly adapting (RARs) nor pulmonary or bronchial C-fiber receptors were affected by hypoxia [Widdicombe, 2001].The possible explanation for these negative findings include the age (all studies were performed on adult animals) and the animal species used in these studies. The NEB cell responses in lungs of adult animals can be expected to be nil or minimal given the small numbers of NEB in the lungs of adults compared to neonates [Cho et al., 1989]. In addition, NEB are extremely rare in the lungs of guinea pigs, a species used as a principal model for airway C-fiber recordings [Carr and Undem, 2003]. It should be also noted that the majority (80-90%) of vagal sensory fibers originating from the lower airways remain unclassified [Coleridge and Coleridge, 1994]. On the other hand, studies of Kalhoff et al. [1994] have identified pulmonary vagal afferents involved in hypoxic breathing in the absence of arterial chemoreceptors .In carotid chemodenervated rabbits, moderate systemic hypoxia (PaO2~40 mmHg) led to tachypneic hyperventilation, providing that the vagus nerves were intact. Aortic chemoreceptors are apparently not involved in this species since severing the carotid sinus nerve abolishes the typical hypoxic ventilatory chemo reflex drive. Indirect evidence for the possible involvement of NEB in the control of breathing is provided by studies of k/o mouse models with altered NEB cell O2 sensor or where NEB cells are absent. As noted earlier, NEB cells in neonatal mice with gp91 phox/NOX 2 deficiency failed to respond to hypoxia in vitro, and

284  |  Airway Chemoreceptors in the Vertebrates in whole animal studies using plethysmography, showed rapid shallow breathing and abnormal responses to hypoxia challenge [Fu et al., 2000; Kazemian et al., 2002]. The disruption of Mash1 gene, that is selectively expressed in PNEC/NEB cells, (reviewed in more detail in Chapter 13) in mice leads to early neonatal death and total absence of PNEC/NEB cells in the lung (ie. NEB-less lungs) while analogous endocrine cells in the GI tract and pancreas remain intact [Borges et al., 1997]. Although the lungs of Mash1 null mice appear histologically normal, these mice die soon after birth because of respiratory failure perhaps due to the disruption NEB chemoreceptors required for neonatal adaptation. It is of interest to note that the lungs of heterozygous mice (Mash 1+/-) show significantly (<50%) reduced numbers of PNEC/NEB [Ito et al., 2000] that is accompanied by an abnormal breathing pattern in neonatal (Mash1+/-) mice during the perinatal period.

Potential Functional Role of NEB in Neonatal Lung The two main postulated roles for NEB in fetal/neonatal lungs include: a) chemoreceptor activity related to initiation of breathing at birth, a function complementary to conventional CB chemoreceptors, since CB are relatively inactive in the neonate [Donnely, 2000]; and b) regulation (directly or indirectly) of fetal/ neonatal pulmonary circulation [Lauweryns and Cockelaere, 1973], that has not yet been extensively investigated. It should be noted, however, that both activities are O2 dependent [Weir et al., 2005]. It has been previously suggested that NEB function as airway chemoreceptors that are part of peripheral chemoreceptor system involved in the control of breathing, particularly during the perinatal period [Cutz, 1997; Cutz et al., 2003]. The control of respiration is a complex activity by virtue of its vital function and is governed by a coordinated action of various anatomical structures with built-in redundancies and back-up systems (i.e. sensors, brain stem respiratory center, respiratory muscles) [Hansen and Corbet, 1992]. The sensors, or chemoreceptors play an essential role in the control of respiration since they detect changes in the internal milieu evoking compensatory reflex mechanisms and responses to maintain homeostasis. For example in the case of CB, the principal arterial chemoreceptor, hypoxia and other stimuli (pCO2, pH, glucose) activate the sensor which in turn triggers appropriate ventilatory responses (i.e. increased frequency and depth of breathing) [Weir et al., 2005]. Whether NEB play a similar role has been suspected for some time, but direct evidence is at present lacking. Although CB glomus cells and lung NEB cells share many structural and biochemical features in common (i.e. the presence of cytoplasmic neurosecretory granules containing amine/peptide(s); sensory innervation; O2 sensitivity), there are significant anatomical and biochemical differences. [for additional details and comparisons between NEB and CB see the chapter by Van Lommel]. Firstly, CB’s form

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  285

discrete vascularized paired structures innervated by the sinus nerve with the nerve cell bodies residing in petrosal ganglia that are of neural crest origin [Gonzalez et al., 1994]. In contrast, NEB’s are widely distributed within the airway epithelium as small cell aggregates forming corpuscular intraepithelial structures. Furtheremore, NEB’s are innervated by vagal afferents derived from the nodose ganglion that are of placode embryologic origin [Zhuo et al., 1997]. In their initial experiments, Lauweryns et al. (l977) noted that in addition to hypoxia, hypercapnia also stimulated the release of amine from NEB cells suggesting that they may represent pO2/pCO2 sensors since the lung is a primary site for the gas exchange. However, in subsequent experiments the effects of increased pCO2 on NEB amine content could not be replicated [Lauweryns et al., 1990]. This question was recently re-examined and it was found that indeed hypercapnia and acidosis causes release of 5-HT from NEB cells (unpublished observations). Studies by other investigators using an in vitro model have shown changes in CGRP contents in PNEC/NEB in response to hypercapnia [Ebina et al., 1997]. Furtheremore, increased CO2 (10% CO2 in serum free media) was found to potentiate the mitogenic effects of nicotine and its carcinogenic derivative, NNK, in normal and neoplastic PNEC via stimulation of autocrine and protein kinase C-dependent mitogenic pathways [Schuller, 1994]. Taken together, these findings support the notion that NEB cells in the mammalian lung (for details on CO2 chemoreception by NEB cells in avian lungs see Chapter 10) likely represent multimodal airway sensors that respond to changes in ambient concentrations of pO2, pCO2 and acidosis.

Summary, Conclusions and Clinical Relevance Over the last decade, a great deal of progress has been made in identification and characterization of O2-sensing mechanism in pulmonary NEB. Although the overall scheme of O2-sensing mechanism in these cells is similar to other O2 sensor coupled neurosecretory cells, there are some exceptions [Weir at al., 2005]. For example, there is now strong evidence for the direct involvement of NADPH oxidase/NOX 2 in O2-sensing in NEB cells and related H 146 cell line. This is in contrast to other O2 sensing cells (i.e. CB glomus cells, neonatal adrenal medulla cells, pulmonary artery smooth muscle cells) where this does not seems to be the case and where alternate mechanisms have been proposed [Archer and Michelakis, 2002;Weir et al., 2005].This complexity and diversity of O2-sensing mechanism likely reflect local modifications and adaptations to the level and range of hypoxia, developmental stage, or coupling to different effector mechanisms. The precise role and function of NEB under normal condition and in various pulmonary disorders remains largely unknown. There are a number of perinatal disorders including Bronchopulmonary Dysplasia (BPD), Congenital Central Hypoventilation Syndrome and Sudden Infant Death Syndrome

286  |  Airway Chemoreceptors in the Vertebrates where hyperplasia of these cells has been documented [reviewed in Cutz et al., 2007]. In addition, the O2 sensor in NEB cells by virtue of their location in the airway and direct exposure to the external environment is likely to be affected by a number of extrinsic factors (i.e. H2O2 and cytokines generated during airway inflammation, exposure to cigarette smoke, air pollutants etc). It is hoped that our expanding knowledge of the cellular and molecular mechanisms of O2-sensing will facilitate and stimulate further studies regarding the physiological function of NEB as well as their involvement in the pathophysiology of a variety of pulmonary disease processes.

References Acker, H., and Xue, D. 1995. Mechanisms of O2 sensing in the carotid body in comparison with other O2-sensing cells. NIPS 10:211-216. Acker, H., Dufau, E., Huber, J., and Sylvester, D. 1989. Indications to an NADPH oxidase as a possible pO2 sensor in the rat carotid body. FEBS Lett. 256:75-78. Adriaensen, D., Timmermans, J.P., Brouns, L., Berthoud, H.R., Neuhuber, W.L., and Scheuerman, D.W. 1998. Pulmonary intraepithelial vagal nodose afferent nerve terminals are confined to neuroepithelial bodies: an anterograde tracing and confocal microscopy study in adult rats. Cell Tiss Res 293:395-405. Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat Rec 270A:25-40. Archer, S., and Michelakis, E. 2002. The mechanism(s) of hypoxic pulmonary vasoconstriction: potassium channels, redox O2 sensors, and controversies. News Physiol Sci 17:131-137. Archer, S.L., Reeve, H.L., Michelakis, E., Puttagunta, L., Waite, R., Nelson, D.P., Dinauer, M., and Weir, E.K. 1999. O2 sensing is preserved in mice lacking the gp91phox subunit of NADPH oxidase. Proc Natl Acad Sci USA. 96:7944-7949. Bolle, T., Lauweryns, J. M., and Van Lommel, A. 2000. Post natal maturation of neuroepithelial bodies and carotid body innervation : a quantitative investigation in the rabbit. J. Neurocytol. 29:241-248. Borges, M. , Linnoila, R.I., Van De Velde, H.J.K., Chen, H., Nelkin, B.D., Marby, M., Baylin, S.B., and Ball, D.W. 1997. An achete-scute homologue essential for neuroendocrine differentiation in the lung. Nature. 386:852-855. Brouns, I., Adriaensen, D., Burnstock, G., and Timmermans, J.P. 2000. Intraepithelial vagal sensory nerve terminals in rat pulmonary neuroepithelial bodies express P2X/3 receptors. Am. J. Respir. Cell Mol. Biol. 23:52-61. Brouns, I., Van Genechten, J., Scheuermann, D.W., Timmermans, J.P., and Adriaensen, D. 2002a. Neuroepithelial bodies: a morphological substrate for the link between neuronal nitric oxide and sensitivity to airway hypoxia. J. Comp. Neurol. 449:343-354. Brouns, I., Van Genechten, J. Timmermans, J.P., Scheuermann, D.W., and Adriaensen, D. 2002b. Pulmonary neuroepithelial bodies are selectively contacted by a population of cholinergic motor fibers. FASEB 16:A453 (Abstract). Carr, M.J. and Undem, B.J. 2003. Pharmacology of vagal afferent nerve activity in guinea pig airways. Pulm Pharmacol Therap.16:45-52.

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  287 Cho, T., Chan, W., and Cutz, E. 1989. Distribution and frequency of neuro-epithelial bodies in post-natal rabbit lung. Quantitative study with monoclonal antibody against serotonin. Cell Tiss. Res. 255:353-362. Coleridge, H.M., and Coleridge, J.C.G. 1994. Pulmonary neural mechanisms of pulmonary defense. Ann Rev Physiol. 56:69-71. Cross, A.R., Henderson, L., Jones, O.T.G., Delpiano, M.A., Hentchel, J., and Acker, H. 1990. Involvement of an NAD(P)H oxidase as a PO2 sensor protein in the rat carotid body. Biochem. J. 272: 743-747. Cutz, E. (Ed.). 1997. Cellular and Molecular Biology of Airway Chemoreceptors. Landes Bioscience, Austin Texas. Cutz, E., Yeger, H., and Pan, G. 2007. Pulmonary neuroendocrine cell system in pediatric lung disease-recent advances. Ped Dev Pathol 10:419-435. Cutz, E., Spiers, V., Yeger, H., Newman, C., Wang, D., and Perrin, D.G. 1993. Cell biology of pulmonary neuroepithelial bodies-validation of an in vitro model. I. The effects of hypoxia and Ca2+ ionophore on serotonin content and exocytosis of dense core vesicles. Anat Rec 236:41-52. Cutz, E., Fu, X.W., Yeger, H., Peers, C., and Kemp, P.J. 2003. Oxygen sensing in pulmonary neuroepithelial bodies and related tumour cell model. In: Oxygen Sensing: Responses and Adaptation to Hypoxia, Lung Biology in Health and Disease, S. Lahiri, H. Prabhakar, G. Semenza (Eds.) Marcel Dekker, NY, pp. 567-602. Dauger, S., Renolleau, S., Vardon, G., Nepote, V., Mas, C., Simonneau, M., Gaultier, C., and Gallego, J. 1999. Ventilatory responses to hypercapnia and hypoxia in Mash-1 heterozygous newborns and adult mice. Pediatr Res 46:535-542. Donnely, D.F. 2000. Developmental aspects of oxygen sensing by the carotid body. J Appl Physiol 88:2296-2301. Ebina, M., Hoyt, R.F., McNelly, N.A., Sorokin, S.P. , and Linnoila, I. 1997. Effects of hydrogen and bicarbonate on endocrine cells in fetal rat organ culture. Am J Physiol 272:L178-L186. Feyerter, F. 1938. Uber diffuse endocrine epitheliale organe. J.A., Barth. Leipzig. Fu, X.W., Nurse, C.A., Wang, Y.T., and Cutz, E. 1999. Selective modulation of membrane currents by hypoxia in intact airway chemoreceptors from neonatal rabbit. J Physiol 514:139-150. Fu, X.W., Wang, D., Nurse, C.A., Dinauer, M.C., and Cutz, E. 2000. NADPH oxidase in an O2 sensor in airway chemoreceptors: Evidence from K+ current modulation in wild type and oxidase deficient mice. Proc Natl Acad Sci (USA) 97:4374-4379. Fu, X.W., Wang, D., Pan, Y., Farrager, S.M., Wong, V., and Cutz, E. 2001.Neuroepithelial bodies in mammalian lung express functional serotonin-type 3 receptor .Am J Physiol 281:L931L940. Fu, X.W., Nurse, C.A.,Wong, V., and Cutz, E. 2002. Hypoxia-induced secretion from intact pulmonary neuroepithelial bodies in neonatal rabbit lung. J Physiol 539:503-510. Fu, X.W., Nurse, C.A., Farragher, S.M., and Cutz, E. 2003. Expression of functional nicotinic acetylcholine receptors in neuroepithelial bodies of neonatal hamster lung. Am J Physiol Lung Cell Mol Physiol 285:L1203-L1212. Fu, X.W., Nurse, C.A., and Cutz, E. 2004. Expression of functional purinergic receptors in pulmonary neuroepithelial bodies and their role in hypoxia chemotransmission. Biol Chem 385:275-284. Fu, X.W., Nurse, C., and Cutz, E. 2007. Characterization of slowly inactivating Kv alpha current in rabbit pulmonary neuroepithelial bodies:effects of hypoxia and nicotine. Am J Physiol Lung Cell Mol Physiol 293:L892-L902.

288  |  Airway Chemoreceptors in the Vertebrates González, C., Almaraz, L., Obeso, A., and Rigual, R. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol Review.74:829-898. Hansen, T., and Corbet, A. 1992. Control of breathing. In :Diseases of the Newborn. 6th Ed, H.W. Taeusch, R.A. Ballard and M.E.Avery (Eds.) Saunders, New York, pp. 470-473. Hartness, M.E., Lewis, A., Searle, G.J., O’Kelly, I., Peers, C., and Kemp, P.J. 2001. Combined antisense and pharmacological approaches implicate hTASK as an airway O2 sensing K+ channel. J Biol Chem.276:26499-26508. Higashi, H., and Nishi, S. 1982. 5-hydroxytryptamine receptors on visceral primary affrent neurons of rabbit nodose ganglia. J Physiol 323:543-567. Ito, T., Ogura, T., Ogawa, N., Udaka, N., Hayashi, H., Inayama, Y., Yazawa, T., and Kitamura, H. 2000. Basic helix-loop-helix transcription factors regulate the neuroendocrine differentiation of fetal mouse pulmonary epithelium. Development 127:3913-3921. Kalhoff, H., Kiwull-Schone, H., and Kiwul, P. 1994. Pulmonsary vagal afferents involved in the hypoxic breathing without arterial chemoreceptors .In: Arterial Chemoreception, C. Eyzaguirre et al. (Eds.) Springer-Verlag, New York, pp. 350-356. Kazemian, P., Stephenson, R., Yeger, H., and Cutz, E. 2002. Respiratory control in neonatal mice with NADPH oxidase deficiency. Respir Physiol 126:89-101. Kemp, P.J., Lewis, A., Hartness, M.E., Searle, G.J., Miller, P., O’Kelly, I., and Peers, C. 2002. Airway chemotransduction. From oxygen sensor to cellular effector. Am Rev Respir Crit Care Med 166:S17-S24. Lambeth, J.D. 2004. Nox enzymes and the biology of reactive oxygen. Nature Rev Immunol 4: 181-189. Lauweryns, J.M., and Cokelaere, M. 1973. Hypoxia-sensitive neuroepithelial bodies. Intrapulmonary secretory neuroreceptors modulated by CNS Z Zellforsch 145:521-540. Lauweryns, J.M., and Van Lommel, A. 1986. Effects of various vagotomy procedures on the reaction to hypoxia of rabbit neuroepithelial bodies:modulation by intrapulmonary axon reflex? Exp Lung Res 11:319-339. Lauweryns, J.M. ,Cokelaere, M., and Theunynck, P. 1972. Neuroepithelial bodies in mammalian respiratory mucose of various mammals: light optical, histochemical and ultrastructural investigation. Z Zellforsch 135:569-592. Lauweryns, J.M., Cokelaere, M., Deleersnyder, M., and Liebens, M. 1977. Intraepithelial neuroepithelial bodies in newborn rabbits. Influence of hypoxia, hyperoxia, nicotine, reserpine, L-DOPA and 5-HTP.Cell Tissue Res 182:425-440. Lauweryns, J.M., Tierens, A., and Decramer, M. 1990. Influence of hypercapnia on rabbit intrapulmonary neuroepithelial bodies: microphlourometric and morphometric study.Eur Respir J 3:182-186. Lee, Y.M., Kim, B.J., Chun, Y.S., So, I., Choi, H., Kim, M.S., and Park, J.W. 2006. NOX4 as an oxygen sensor to regulate TASK-1 activity. Cellular Signalling 18:499-507. Lewis, A., Hartness, M.E., Chapman, C.G., Fearon, I.M., Meadows, H.J., Peers, C., and Kemp, I.M. 2001. Recombinant hTASK 1 is an O2 sensitive K+ channel. Biochem Biophys Res Comm. 285:1290-1294. Lopez-Barneo, J., Benot, A.R., and Urena, J. 1993. Oxygen-sensing and the electrophysiology of arterial chemoreceptor cells. News Physiol Sci .8:191-193. Lopez-Barneo, J., Pardal, R., and Ortega-Saenz, P. 2001. Cellular mechanisms of oxygen sensing. Ann Rev Physiol 63: 259-87.

Oxygen Sensing in Mammalian Pulmonary Neuroepithelial Bodies  |  289 Nurse, C.A., and Zhang, M. 1999. Acetylcholine contributes to hypoxic chemotransmission in co-cultures of rat type I cells and petrosal neurons .Respir Physiol 115:189-199. O’Kelly, I., Peers, V., and Kemp, P.J. 1998. O2-sensitve K+ channels in neuroepithelial bodyderived small cell carcinoma cells of the human lung. Am J Physiol 275: L709-L716. O’Kelly, I., Stephens, R.H., Peers, C., and Kemp, P.J. 1999. Potential identification of the O2sensitive K+ current in human neuroepithelial body derived cell line. Am J Physiol 276: L96L104. O’Kelly, I., Lewis, A., Peers, C., and Kemp, P.J. 2000. O2 sensing in airway chemoreceptorderived cells: protein kinase C activation reveals functional evidence for involvement of NADPH oxidase. J Biol Chem.275:7684-7692. Pan, J., Yeger, H., and Cutz, E. 2004. Innervation of pulmonary neuroendocrine cells and neuroepithrelial bodies in developing rabbit lung. J Histochem Cytochem 53:379-389. Patel, A.J., and Honore, E. 2001. Molecular physiology of oxygen-sensitive potassium channels. Eur Respir J 18: 221-227. Quinn, M.T., and Gauss, K.A. 2004. Structure and regulation of neutrophil respiratory burst oxidase: comparison with physiological oxidases. J Leukocyte Biol 76: 760-781. Ruppersberg, J.P., Stocker, M., Pongs, O., Heinemann, S.H., Frank, R., and Koenen, M. 1991. Regulation of fast inactivation of cloned mammalian Ik(A) channels by cysteine oxidation. Nature 352:711-714. Sanchez, D., Lopez-Lopez, J.R., Perez-Garcia, M.T., Sanz-Alfayate, G., Obeso, A.,Ganfornina, M.D., and Gonzalez, C. 2002. Molecular identification of Kv alpha subunits that contribute to the oxygen-sensitive K+ current in chemoreceptor cells of the rabbit carotid body. J Physiol 542:369-382. Schuller, H.M. 1994. Carbon dioxide potentiates the mitogenic effects of nicotine and its carcinogenic derivative NNK in normal and neoplastic neuroendocrine cells via stimulation of autocrine and protein kinase C-dependent mitogenic pathway. Neurotoxicology 15:877886. Searle, G.J., Hartness, M.E., Peers, C., and Kemp, P.J. 2002. Lack of contribution of mitochondrial electron transport to acute O2 sensing in model airway chemoreceptors. Biochem Biophys Res Commun 291:332-337. Sorokin, S.P., and Hoyt, R.F. 1989. Neuroepithelial bodies and solitary small-granule cells. In:Lung Cell Biology, D.Massaro (Ed.)Marcel Dekker, N.Y. pp. 91-344. Sumimoto, H., Ueno, N., Yamasaki, T., Taura, M., and Takeya, R. 2004. Molecular mechanisms underlying activation of superoxide-producing NADPH oxidases: roles of their regulatory proteins. Jpn J Infect Dis .57:S24-S25. Thompson, R.J., Farragher, S.M., Cutz, E., and Nurse, C.A. 2002. Developmental regulation of O2 sensing in neonatal adrenal chromaffin cells from wild-type and NADPH-oxidase deficient mice. Pflugers Arch Eur J Physiol 444: 539-548. Van Lommel, A., and Lauweryns, J.M. 1997. Postnatal development of the pulmonary neuroepithelial bodies in various species. J Autonom Nerv Syst 5:17-24. Van Lommel, A., Bolle, T., Faunes, W., and Lauweryns, J.M. 1999. The pulmonary neuroendocrine system: The past decade. Arch Histol Cytol 62:1-16. Vega-Saenz de Meira, E., and Rudy, B. 1992.Modulation of K+ channels by hydrogen peroxide. Biochem Biophys Res Commun 186:1681-1687. Wang, D., Youngson, C., Wong, V., Yeger, H., Dinauer, M.C., Vega-Saenz de Meira, E., Rudy, B., and Cutz, E. 1996. NADPH oxidase and Hydrogen peroxide-sensitive K+ channel may

290  |  Airway Chemoreceptors in the Vertebrates function as an oxygen sensor complex in airway chemoreceptors and small cell carcinoma cell lines. Proc. Natl. Acad. Sci (USA) 93:13182-87. Weir, E.K., LopezBarneo, J., Buckler, K.J., and Archer, S.L. 2005. Acute oxygen sensing mechanisms. New Engl J Med 353::2042-2055. Widdicombe, J. 2001. Airway receptors. Respir Physiol 125:3-15. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature(Lond) 365: 153-156. Youngson, C., Nurse, C.A., Yeger, H., Curnutte, J.T., and Cutz, E. 1997. Immunohistochemical localization of O2 sensing protein(NADPH) oxidase in chemoreceptor cells. Microsc Res Tech 37:101-106. Zhang, M., and Nurse, C.A. 2004. CO2/pH chemosensory signaling in co-cultures of rat carotid body receptors and petrosal neurons: role of ATP and ACh. J.Neurophysiol 92:3433-3445. Zhang, M., Zhong, H., Vollmer, C., and Nurse, C.A. 2000. Co-release of ATP and Ach mediates hypoxic signaling of rat carotid body chemoreceptors. J Physiol 525:143-158. Zhuo, H., Ichikawa, H., and Helke, C.J. 1997. Neurochemistry of the nodose ganglion. Progress in Neurobiol 52:79-107.

13 Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System in the Developing Mammalian Lung H.Yeger*, J. Pan and E. Cutz

Abstract The search for stem cells in every tissue and organ has become a subject of intense investigation in recent years and has led to new insights into how tissue lineages are derived and how tissues and organs are repaired following injury. Lung stem cells have been sought and a number of potential progenitor phenotypes identified. Interestingly, the resulting studies have indicated stem cells residing in several niches and with different properties and pluripotency. Studies on the pulmonary neuroendocrine cell system (PNEC) which includes single cells as well as PNEC clusters, neuroepithelial bodies (NEB), have yielded clear indication that NEB harbor progenitors and that PNEC may derive from a possible common undifferentiated but committed neuroepithelial like progenitor. What then directs these cells to become neuroendocrine and non-neuroendocrine is not yet fully understood, although early expression of neurogenic genes suggests distinct transcriptional programming. In this review we lay the groundwork for further investigations that may lead to better understanding of lung development and how the PNEC relate to other lung phenotypes. From an ontological point of view, organisms that ‘breathe’ utilize PNEC as their environmental sensors which are located within an epithelium that subserves transport functions. PNEC/NEB produce a variety of bioactive substances that help to maintain and modulate the surrounding epithelium and the invested connective tissue. The complex role of PNEC/NEB in lung is underscored by the following observations: i) early differentiation and widespread distribution of PNEC during lung development, ii) complex innervation of PNEC/NEB that peak during the neonatal period, and iii) the multitude of CIHR Group on Lung Development Division of Pathology and Department of Paediatric Laboratory Medicine, The Hospital for Sick Children, and Department of Laboratory Medicine & Pathobiology, University of Toronto, Ontario, Canada. (CIHR grants FRN15270 and FRN81253). *Author for Correspondence: Email: [email protected] Phone: (416) 813-5958, Fax: (416) 8135974.

292  |  Airway Chemoreceptors in the Vertebrates bioactive amines and peptides synthesized. There is still much to be learned about these aspects and about the specific PNEC/NEB functions that are required for overall lung function in health and disease.

Keywords: PNEC/NEB, stem cells and progenitors, developmental complexity

Introduction Lung development in humans closely parallels events occurring in other mammals. The respiratory system develops as an out-pouching of the foregut, which then forms a tube that grows progressively into the splanchopleuric mesenchyme. This forming lung bud undergoes an extensive series of tubular extensions and bifurcations surrounded by an interactive mesenchyme. By 28 d after fertilization the left and right bronchial buds (corresponding to left and right lungs) are established and by the 5th wk three bronchial stems on the right and two on the left determine the ultimate lobular structure of the lung. Essentially, lung development constitutes five developmental stages, 1) embryonic (3-7 wk) to form the broncho-pulmonary segments; 2) pseudoglandular (7-16 wk) to form up to 25 orders of branching up to the level of terminal bronchioles where the epithelium is pseudostratified squamous; 3) canalicular (16-24 wk) with rapidly advancing angiogenesis and differentiation of the mesenchyme surrounding tubules invested with a growing mesenchyme and consisting of a more cuboidal epithelium and formation of several orders of respiratory bronchioles; 4) terminal sac (24-36 wk) with initiation of primitive alveolar ducts and then gradual maturation of a functional blood/gas barrier and extensive vascularization of these structures; and 5) alveolar (36 wk to term/adult) with full maturation of the alveoli and proportional reduction of lung parenchyma as lungs switch to air breathing. Alveoleogenesis continues into the 3rd yr postnatally concomitant with lung expansion. Differentiation of the airway epithelium into mucociliary epithelium proximally and Clara cell dominant epithelium distally, starts at about wk 13 and follows that of the differentiating mesenchyme that forms bronchial smooth muscle and cartilage. The proximal vascular supply grows down with the differentiating mesenchyme and is also accompanied by an expanding neural network derived predominantly from the branches of the vagus nerve. Interestingly, the first cell type to differentiate within the early primitive airway epithelium are the pulmonary endocrine cells (PNEC) [Cutz et al., 2007; Sunday and Cutz, 2000]. These cells are distributed widely throughout the airway mucosa, first appearing as single PNEC and then later as PNEC clusters termed NEB. NEB are often located at the airway bifurcations and become innervated via vagal and other nerve trunks [Adriaensen et al., 2003; Brouns et al., 2003; Cutz et al., 2007]. Given the gradual and progressive nature of lung epithelial specialization and the unique investment of the epithelium with neuroendocrine cells, the question arises whether

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  293

specific stem cells exist within a lung that, at a minimum, would give rise to both the epithelial and mesenchymal lineages, representing endoderm and mesoderm. In fact, one could speculate whether a pluripotent stem cell might govern lung development starting from the pulmonary bud stage in the embryonic phase.

Pulmonary Stem Cell Niches (see Figure 1) Stem cells are operationally defined as progenitors with the capacity for self-renewal and ability to generate progeny that include transit-amplifying (TA) cells that eventually give rise to the functionally differentiated phenotypes of the tissue or organ. In a recent review on epithelial stem cells of the lung, Rawlins and Hogan [2006] distinguish between dedicated stem cells (self-renewing population) and facultative stem cells (differentiated cells that revert to the dedicated stem cell state under injury), progenitors, and self-renewing differentiated cells (divides in response to injury and maintains homeostasis). In fact, there is evidence for the existence in the lung of all these subtypes, thus making the identification of pluripotent stem cells in the lung more difficult compared to other organs. Stem cells are thought to reside within specialized niches that protect them from insults and provide an optimal microenvironment for survival and expansion as needed. Cells meeting these criteria retain mitotic labeling for prolonged periods (i.e., non cycling compartment) [Hong et al., 2001; Rawlins and Hogan, 2006]. It is not surprising therefore that such label retaining cells, LRC, (identified using a BrdU labeling technique) have been found in the lung in several distinct niches along the respiratory tract. LRC have been found within submucosal gland ducts and the cartilage-intercartilaginous junction [Rawlins and Hogan, 2006], bronchioalveolar duct junction [Giangreco et al., 2002; Kim, 2007; Kim et al., 2005] and in physical association with NEB at more proximal locations [Hong et al., 2001; Reynolds et al., 2000]. Note that NEB are found within the epithelium of intrapulmonary airways down to the level of the terminal bronchioles [Linnoila, 2006] so that their role as a stem cell niche may be broader. Stem cells like progenitors for proximal and distal lung regions have also been defined with other criteria, for example, the multipotent cytokeratin K14 and K5 expressing basal cell subtypes of the tracheobronchial epithelium [Borok et al., 2006; Rawlins and Hogan, 2006], and the Sca-1, CD34 + subpopulation within the bronchioalveolar region [Kim 2007; Borok et al., 2006]. In addition, subsets of type II (AT2) alveolar cells yield type 1 (AT1) alveolar cells and so are also progenitors within the alveolar region [Borok et al., 2006; Rawlins and Hogan, 2006; Griffiths et al., 2005]. Finally, it is yet unresolved whether lung side population cells (SP) (defined by their ability to exclude Hoechst 33342 dye via the BRCP1 drug pump used for FACS sorting) represent functional stem cells, as in other organs like bone marrow, or constitute nonpulmonary stem cell residents [Majka et al., 2005; Summer et al., 2004]. In the adult

294  |  Airway Chemoreceptors in the Vertebrates

Figure 1: Pulmonary Stem Cell Niches and PNEC/NEB. Illustrated are the multiple sites of progenitors and stem cells residing within the whole lung. Top: In the upper airways, trachea and bronchi, the basal cell of the mucociliary epithelium and progenitors lying within the ducts of the submucosal glands serve for purposes of tissue regeneration. The actual origin of these epithelia from a distinct stem cell is still unclear. Center: In the more distal lung cells with a Clara cell like phenotype (naphthalene resistant and LRC) are associated with NEB. These can regenerate the lung epithelium. However, actual progenitors for PNEC/NEB are still not identified. Bottom: The stem cell niche of the bronchioalveolar junction is thought to repopulate the bronchioalveolar region. Since NEB can be found at terminal bronchioles the relationship between bronchiolalveolar stem cells and terminal residing NEB is unresolved.

mouse lung the SP fraction contains both CD45- and CD45+ subpopulations with distinct marker profiles and differentiation potential [Liang et al., 2005]. There is some evidence suggesting that somatic stem cells from bone marrow and possibly lung mesenchyme can transdifferentiate into pulmonary epithelial subtypes like type 1 alveolar cells, in particular, following injury [Gomperts et al., 2006]. Further evidence has implicated bone marrow progenitors as facilitators of repair process helping to recruit and instruct local progenitors [Albera et al., 2005]. Evidence for the existence of pulmonary stem cells has come from basically two experimental approaches, a) selective injury of lung epithelium where, for example, naphthalene injections in mice eliminate Clara cells and regeneration occurs from

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  295

cells that reside within the NEB niche where one finds naphthalene resistant LRC with a Clara cell like phenotype [Hong et al., 2001; Reynolds et al., 2000], and b) Cre-lox mouse models using, for example, the K14 promoter to tag basal cell progeny [Hong et al., 2004]. Linnoila [2006] in reviewing the functional facets of the PNEC system discusses studies demonstrating that PNEC and tracheobronchial glands may originate from a different anlage than the distal lung epithelium and evidence that embryonic E13-E15 mouse lung airway cells concomitantly express CC10, CGRP and SpA, suggesting a common cellular origin for airway cells. Furthermore, studies have indicated that mature ciliated cells can de-differentiate following lung injury, switch to a squamous phenotype to cover the open wound, and then re-differentiate into a fully functional pseudostratified mucociliary epithelium [Park et al., 2006]. A more recent study disputed this original finding and showed with a more definitive Cre/Lox driven mouse model that de-differentiated ciliated cells did not re-differentiate during repair [Rawlins et al., 2007]. Further to this point, Hajj et al. [2007] using sorted CD151/ TF-positive human basal cells from large airways showed that only these and not the CD151/TF negative columnar cells of mucociliary epithelium could regenerate a fully differentiated and functional mucociliary epithelium in vitro and in vivo. Since this population also proliferated readily and showed telomerase activity these basal cells are considered airway surface transit-amplifying cells. These observations support basal cells as the means of rapid repair following injury. The identification of unique subpopulations of basal cells residing in niches along the major airways [Rawlins and Hogan, 2006] suggest that pulmonary epithelial cell plasticity plays a much larger role than previously appreciated. This is particularly important during lung repair and regeneration processes that are critical in air breathing lungs subjected to constant environmental insults. In an adult lung there may also be contributions from bone marrow epithelial progenitors that express the chemokine receptor CXCR4 responding to the natural chemokine CXCL12 found at sites of airway injury [Gomperts et al., 2006]; these cells reproduce a normal pseudostratified epithelium. However, depletion of CXCL12 defaults the injured epithelium to development of squamous metaplasia. Whether there is in fact a common pulmonary ‘stem cell’ is unknown since this type of cell, if it exists, might only be critically required during the embryonal and early fetal stages in lung development. Thereafter it is possible that sub-specialized progenitors and/or the process of transdifferentiation [Rawlins and Hogan, 2006], in concert with bone marrow derived mesenchymal stem cells, could contribute to the overall maintenance and regeneration of mature lungs. As in the case of the liver, the existing parenchyma may respond in a regenerative manner to tissue loss and removal. Recent identification of a multipotent stem cell at the bronchioalveolar junction, termed BASC (bronchioalveolar stem cell, reviewed in [Kim, 2007]) suggests that this region of the lung is repopulated by a cell that can give rise to the AT2, and thereby AT1 alveolar epithelium, and also Clara cells within the region. How this BASC relates to the Clara cell progenitor harbored by the NEB is unknown. Interestingly, NEB are

296  |  Airway Chemoreceptors in the Vertebrates distributed as far down as the terminal bronchioles and at the bronchioalveolar junction, thus raising the possibility that a similar interaction between NEB and progenitors occurs within this region and therefore along the entire airway of bronchiolar branching. This would then leave the trachea as a developmentally separate region of the airway where basal cells serve as progenitors, an arrangement seen in stratified epithelia of other organs. Finally, the BASC have been suggested as the targets for tumorigenic hits that yield non-small cell lung carcinomas (NSCLC) [Kim, 2007]. The fact that there is phenotype crossover between NSCLC and SCLC in significant fraction of lung cancers may be explained at the level of these stem cells or progenitors.

Differentiation of Pulmonary Neuroendocrine Cells during Lung Development What still remains as an open question is the actual origin of PNEC and NEB, and how their morphogenic programs are regulated. There is accumulating evidence that the embryonic lung starts off with a neuroepithelial or partial neuroendocrine like phenotype and then becomes increasingly subspecialized with the majority of cells differentiating into the various non-neuroendocrine epithelial subtypes [Haley et al., 1997; Sorokin et al., 1993]. A small fraction matures into the neuronal like PNEC with their content of amine (5HT, serotonin) and neuropeptides, not unlike similar neuroendocrine cells found in other organs but with their own distinct phenotype. Apropos is our observation that BrdU labels cells that express an epitope of stagespecific embryonic marker (SSEA-1) and that some of these then acquire 5HT [discussed below]. Experiments in mice have not resolved whether PNEC are replaced from a unique progenitor or replicate in a limited fashion after injury. An earlier study suggested a low level of proliferation [Cutz et al., 1985], while in certain respiratory diseases in humans PNEC appear to undergo hyperplasia but remained confined within the epithelial layer [Cutz et al., 2007]. However, in adults, PNEC are the source for development of small cell lung carcinomas and carcinoids while some nonsmall cell lung carcinomas retain a partial neuroendocrine phenotype [Linnoila et al., 2005; Linnoila, 2006], again supporting the idea that the undifferentiated pulmonary epithelium may have as its origin a neuropeithelial like phenotype.

The Neurogenic Molecular Regulation of Pnec/Neb Development In a recent comprehensive review of the functional facets of the PNEC system Linnoila [2006] emphasized the multifunctional roles for the neuroendocrine cells during lung

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  297

development and the relevance to a variety of diseases in neonates and adults. It is clear that the PNEC lineage arises separately from the other lung epithelial cells and PNEC do not convert to other types. As indicated above, however, in mouse studies the entire epithelium in E13-15 (pseudoglandular period) shows expression of CGRP, CC10 and SPA suggesting a neuroendocrine/Clara-like progenitor state. In humans the PGP9.5 marker (a ubiquitin C-terminal hydrolase) is universally expressed in the pseudoglandular phase [Linnoila, 2007; Haley et al., 1997]. Phylogenetically, PNEC are found in lower vertebrates indicating a highly conserved system. Since a close developmental relationship is maintained with Clara-like cells, and NEB harbor Claralike distal lung stem cells, the airways in the lung outside of the trachea and bronchus could share a common stem cell. So far it is not clear whether tracheal PNEC (NEB are not found at this location) are unique or are also aligned with Clara-like progenitors. In the mouse lung, where Clara cells are also found in the upper airways, perhaps tracheal PNEC are indeed closely aligned with the Clara cell compartment. This would then place the basal cells of the mucociliary epithelium into a separate class, and as evident from airway epithelial injury studies, suggest that basal cells are the undifferentiated stem cell state of the mucociliary epithelium. From a molecular neurogenic gene standpoint it is now evident that neurogenic genes like Achaete-scute-homolog-1 (MASH1 in the mouse and hASH1 in humans) and HES1 play related and opposite roles in governing neuroendocrine lineage fate in the lung with related family members like MATH1 operating in other neuroendocrine lineages [reviewed in Linnoila, 2006]. Mash1 deficient mice lack PNEC while HES1 knockout mice show precocious PNEC at E13 and PNEC hyperplasia at E18 [Borges et al., 1997]. Interestingly, CC10 positive (Clara) cells remain abundant in HES1 knockout mice supporting the idea of a presumed Clara/PNEC relationship but also suggest other factors mitigate PNEC differentiation. Upstream to MASH1 are members of the NOTCH family of neurogenic genes that appear to delineate neuroendocrine versus non-neuroendocrine cell fates. In the naphthalene model of airway injury and repair, SHH ligand and Gli1 signaling were expressed prominently in the regenerating epithelium but not in nascent CGRP positive PNEC cells consistent with the role of Hh signaling in determination of stem cell fates. Figure 2, left side. (unpublished McGovern et al.) shows that Mash1 is co-expressed with 5HT in developing NEB at E14, however single Mash1 positive cells also reside within the epithelium. When NEB are innervated Mash1 expression is retained (Figure 2, right side), implicating Mash1 in the developmental maturation of NEB. It is conceivable also that unassigned Mash1 expressing cells represent progenitors maintained in an apoptosis resistant state by Mash1 as recently suggested [Wang et al., 2007]. Could these cells then be the sought after progenitors? Taken together current evidence indicates that neurogenic programs and genes that play crucial roles in other organs also contribute to the stem cell fate determinations in the lung. However, the specific progenitor phenotype for PNEC commitment

298  |  Airway Chemoreceptors in the Vertebrates

Figure 2: Expression Patterns of Mash1 and PGP9.5 in Developing Lung of the Mouse.



Left: At E14 Mash1 nuclear immunostaining (red color) is noted in single cells residing within the developing airways and in PNEC clusters, NEB, where Mash-1 is co-expressed with PGP9.5 (bright green color) indicating a neuroendocrine phenotype. Right: At higher magnification a more mature NEB at E15 shows strong coexpression of Mash-1 and PGP9.5. Note also the detection of nerve processes expressing PGP9.5 subtending and entering the NEB, indicating innervation. Note the low level of PGP9.5 positivity in the airway epithelium.

has not as yet been identified. Nevertheless, working with the SV40ts58 transgenic Immortomouse we have recently found that continuous cell lines, growing at the permissive temperature of 33C, derived from dissociated postnatal lungs, express βIII-tubulin [Figure 3] suggesting a neuronal phenotype, a developmental state first surmised from older studies [Linnoila, 2006]. It has been reported that in rabbits, PNEC progenitors can be identified by expression of FORSE-1, an epitope related to Lewis-X and the stage-specific embryonic antigen (SSEA-1) [see Figure 4; from Pan et al., 2002]. FORSE-1 labeled the primitive epithelium and then became restricted to differentiating PNEC, suggesting that the early lung epithelium is pluripotent and then becomes specialized. It has also been reported that FORSE-1/5HT co-labeled cells were retained within the structure of the NEB and that these were also BrdU labeled suggesting a relationship to the LRC fraction mentioned above. It is therefore possible that FORSE-1 positive cells represent the bronchial ‘stem cells’ or progenitors that arise early on and then remain closely associated with NEB. Since FORSE-1 antigen expression seems to disappear from the surrounding lung epithelium close to birth it will be important to attempt the isolation of such FORSE-1 positive cells from early gestation (in the rabbit prior to E16) in order to assess their differentiation capacity. SSEA and related antigens may

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  299

Figure 3: A continuously growing cell line was derived from disassociated postnatal Immortomouse lung and maintained at the permissive temperature of 33°C, which allows continuous proliferation but not differentiation. Cells were immunostained for cytokeratin (not shown) and βIII-tubulin, a marker for neuronal specific tubulin suggesting that these undifferentiated lung cells are neuronal-like.

permit identification and isolation of different ‘stem cell’ phenotypes which have been identified thus far including the BASC [Kim, 2007]. Interestingly, in a recent report by Allahverdian et al. [2006], increased expression of sialyl Lewis X (sLeX), a fucose containing tetrasaccharide, was found in areas of damaged bronchial epithelium in human airways. Utilizing an in vitro model of bronchial epithelial monolayers, mechanical wounding produced a similar result and blocking with antibody to sLeX completely prevented epithelial repair. These studies implicate LeX containing surface glycoproteins during lung morphogenesis and regeneration, and suggest a link with PNEC differentiation. What drives this process is unknown but mesenchymal-epithelial interactions, the types that have been elaborated in nephrogenesis studies [Dressler, 2006], may also govern lung morphogenesis. In fact, regional mesenchymes can dictate expression of the ultimate epithelial phenotype in co-culture systems where proximal and distal mesenchymal cells are recombined in matched and mismatched pairs with corresponding anatomical level epithelia [Deimling et al., 2007; Perl et al., 2002; Shannon et al., 1998]. As indicated above, these in turn could depend on precisely timed expressions of the NOTCH and BMP

300  |  Airway Chemoreceptors in the Vertebrates

Figure 4: from Pan et al (2002); reproduced with permission.

families of proteins [Deimling et al., 2007; Warburton et al., 2000], with further refinements introduced by differential expression of neurogenic genes.

The Effects of Environmental Factors (Hypercapnia) on Pnec Differentiation and Function Emura and colleagues described a hamster in vitro model of differentiating lung epithelium [Emura, 2002; Emura et al., 1997; Emura et al., 1994; Germann et al., 1993]. They isolated a continuous cell line from a fetal hamster, designated M3E3/ C3, that undergoes Clara cell-like differentiation under one set of culture conditions and PNEC-like differentiation under a different set of culture conditions. The main difference between these two culture systems was that PNEC differentiation required hypercapnia and factors such as retinol. Potential for differentiation into Type II pneumocytes was also demonstrated using agarose overlay and hormone supplements [Germann et al., 1993], thus expanding the multipotency of this cell line model. It is of interest that the hamster has served as a model for tobacco relevant carcinogen induced tumorigenesis of small cell lung carcinoma (SCLC) [Schuller et al., 1995] and PNEC demonstrate expression of nicotinic receptors [Fu et al., 2003]. The fact that hypercapnia drives expression of the PNEC phenotype in the hamster supports the notion that pO2 and pCO2 may play important roles in lung morphogenesis [Cutz et al., 2007]. Models such as these, and the Immortomouse model mentioned above should help to dissect the developmental potential of putative progenitors.

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  301

The Role of Oxygen Concentration, pO2 Sensing and Mechanical Forces in Lung Development Lung development occurs under relative hypoxia and a continuous gradient of increasing oxygen concentration brought about by expansion of the vascular supply into the growing lung. During intra uterine life, airways are fluid filled, and undergo gradual hydrodynamic expansion, which stretches the developing lung tissues. It has been recently shown that cultures of isolated fetal PNEC and SCLC cell lines cultured on Flexcell membranes (to permit controlled stretching of cells), and placed under appropriate pressures, simulating intrauterine lung conditions including fetal breathing, release 5HT from the cytoplasmic pool, but not from an O2 sensitive and Ca2+ dependent vesicular pool, suggesting that during fetal lung growth mechanical stretch may be the key factor that activates secretion of bioactive amine [Pan et al., 2006]. Stretch activates specific mechano-sensitive channels within the plasma membrane [Pan et al., 2006]. With progressive gestation, pO2 levels rise and thus developing lung tissue experiences a changing O2 environment starting in relative hypoxia [euoxia] until birth at which time transition to air breathing places lung under ~20% O2. In this context it is note worthy that cyclic exposure to ozone postnatally compromises continuing morphogenesis of the tracheobronchial tree, thus demonstrating the detrimental effects of hyperoxia on lung structure [Fannuchi et al., 2006]. As indicated earlier, developing lung transitions through distinct morphological stages. It was found that formation of NEB during fetal life shows a peak incidence at E15 in the mouse and E21 in the rabbit, after which NEB numbers remain high until the postnatal period and then gradually decline as the lung matures and expands. Although mid-gestation NEB show evidence of early innervation, it is not known if they are functionally mature and express a pO2 sensor consisting of a NADPH oxidase complex coupled with an O2 sensitive K+ channel [Cutz et al., 2007]. However it is conceivable that hypoxia drives these O2 sensitive NEB to secrete vasoactive 5HT and bioactive neuropeptides, which in turn affect morphogenesis and multiple physiological functions. As noted above, there is also evidence suggesting that NEB may sense pCO2 but the precise mechanism is not known [Cutz et al., 2007; Linnoila, 2006]. In regards to the lung development timeframe, it is unclear whether a functional O2 sensor is expressed by early NEB cells present at E15 and E21 in the mouse and rabbit respectively. By analogy this time frame would be equivalent to ~wk 25 in humans which, interestingly, is compatible with survival of premature infants. And, if not quite competent, do these NEB drive morphogenic events via their bioactive secretory products? In agreement with previous studies [van Tuyl et al., 2005], it was observed recently that prolonged (6d) exposure of E12 mouse lung explants to hypoxia induces increased branching morphogenesis of airways [McGovern, Pan, Yeger, Cutz, unpublished observations]. However, this occurs at the expense of PNEC differentiation. Thus hypoxia abrogates PNEC differentiation, as well as Clara cell differentiation,

302  |  Airway Chemoreceptors in the Vertebrates suggesting that hypoxia drives airway epithelium toward an undifferentiated state, while at the same time stimulates proliferation of lung parenchyma. Since there is also loss of Mash1 neurogenic gene expression in fetal PNEC cultures exposed to hypoxia, it suggests that hypoxia switches developing lung epithelial cells back to a more primitive uncommitted state. Therefore it is surmised that pO2 levels critically pace lung development including airway cell differentiation and maturation, until such time when maturation programs lock in, since after E15 the PNEC maturation process is no longer sensitive to hypoxia. Again, of relevance to humans, it may explain the ability of premature infants to maintain relatively normal pulmonary function after reaching a critical threshold stage in lung development.

Cellular Adaptation to Hypoxia: The Role of Hypoxia Inducible Factors (Hifs) in Lung Development and Pnec Differentiation (see Figure 5) Recent discoveries of oxygen dependent gene regulation involve directing the organogenesis of major physiological systems such as vasculogenesis and erythropoiesis [Bunn et al., 1996; Maltepe et al., 1998, Semenza, 2001a]. Interestingly, local pO2 itself, rather than a genetically encoded developmental program or a central O2-measuring regulator, represents the main functional regulator of O2 homeostasis. The critical intermediary between hypoxia sensing and gene expression are hypoxia inducible factors (HIFs 1,2 &3), which appear to be master regulators of O2 homeostasis [Wenger, 2002]. HIF-1 is a heterodimeric, redox-sensitive protein composed of the two basic helixloop-helix (bHLH) Per-Arnt-Sim (PAS) subunits, HIF-1α and the constitutively expressed HIF-1β [Semenza, 2001b]. HIF-1α expression is induced by hypoxia in an O2 dependent fashion when cells are exposed to 6% O2 or less (i.e. pO2~ 40mmHg) [Semenza, 2001b]. The continuously translated HIF-1α is degraded rapidly by proteolysis in the presence of O2. An oxygen-dependent degradation domain (ODD) of HIF-1α appears to be modified by the hydroxylation of two proline residues, requiring O2 and Fe2+ for its activity [Ivan et al., 2001]. Further hydroxylation results in the binding of the von Hippel-Lindau tumor suppressor protein (pVHL) to the ODD, thereby initiating degradation of the alpha–subunit via the ubiquitinin-proteasome pathway. Low pO2 leads to stabilization of the HIF-1α, nuclear translocation, formation of a dimer with HIF-1β, and recruitment of transcriptional co-activators. This complex binds to an enhancer domain of the target gene, the hypoxia responsive element (HRE). In addition to well-established HIF-1α, two additional members of bHLH-PAS superfamily have been described, namely HIF-2α, which functionally resembles HIF-1α by its similarity in hypoxic stabilization, and a less well defined HIF-3α [Heidbreder et al., 2003].

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  303

Figure 5: Redox Model for O2 Sensing in PNEC/NEB and Their role in Lung Development (modified from Gonzalez et al., Respir Physiol Neurobiol 132:17, 2000).   This unifying model proposes convergence of acute (seconds) and chronic (minutes, hours) responses to hypoxia in specialized O2 sensing cells (i.e. PNEC/ NEB) with low threshold for hypoxia, becoming activated at pO2 ~70mmHg and increasing their effector activity with the intensity of hypoxia. In this model low pO2 is detected by O2 sensor (NADPH oxidase) resulting in reduced oxidase activity with decreased level of ROS production (i.e. H2O2) which in turn affects redox modification of O2 sensitive K+ channel (left panel) leading to Ca2+ mediated exocytosis of amine, peptide and neurotrophins from PNEC. During sustained hypoxia (right panel), as would be the case in developing lung, redox modification of HIFs leads to their stabilization and up-regulation of expression of neuroendocrine and neurogenic genes in PNEC leading to increased synthesis of relevant neurogenic proteins affecting lung organogenesis. Also included are gene deletion models * (K/O mice) to be tested in this proposal, involving different components of O2 sensing homeostatic loop. *K/O= knockout mouse model.

While HIF-1α is ubiquitous, HIF-2α and HIF-3α are more cell specific [Heidbreder et al., 2003; Wiesener et al., 2003]. For example, HIF-1α is required for proper development of the heart, blood and blood vessels and respiratory control centers [Iyer et al., 1998], whereas HIF-2α plays an important role in cardiovascular and respiratory systems [Compernolle et al., 2002]. The physiological role for the different classes of HIF alpha subunits is beginning to be defined. HIF-1α K/O mice die in utero because of cardiovascular malformations and neural tube defects [Iyer et al., 1998]. Interestingly, mice partially deficient in HIF-1α (HIF-1α +/-) have shown defective carotid body (CB) function and impaired ventilatory responses to hypoxia [Kline et al., 2002]. Mice with HIF-2α deficiency (HIF-2α –/-) were reported to die in utero

304  |  Airway Chemoreceptors in the Vertebrates due to bradycardia related to inadequate supply of catecholamines [Tian et al., 1998]. This study suggested that in wildtype mice HIF-2α, highly expressed in catecholamine producing cells (i.e. adrenal medulla, CB glomus cells) is stimulated by fetal hypoxia (euoxia), which in turn up-regulates catecholaminergic gene expression leading to adequate catecholamine synthesis and secretion to maintain normal cardiac function. Another recent study on HIF-2α K/O mice reported abnormal lung development due to immaturity of alveolar type II cells linked to reduced VEGF levels, a known target of HIF-2α transcription factor [Compernolle et al., 2002]. Thus HIF-2α appears to be an essential component of normal embryonal/fetal development including lung organogenesis. In postnatal lungs, HIF-2α expression is low or undetectable under normoxia, but shows a significant increase in animals exposed to hypoxia (8% O2, 6 h) [Wiesener et al., 2003]. Of particular interest is the finding that normobaric hypoxia is a strong stimulus for HIF-2α accumulation whereas hypoxemia had no effect in lung tissue. Furthermore, HIF-2α expression in lung localizes in airway epithelium and alveolar type II cells. In the human lung, two recent studies have started to delineate the temporal and spatial expression of HIFs during the first trimester [Groenman et al., 2007] and from 13.5 wk until term [Rajatapiti et al., 2007]. Essentially, substantial HIF-1α and 2α expression are noted in the developing airways and surrounding mesenchyme, HIF-3α diminishes with age, while HIF-2α expression is closely correlated with angiogenesis and the required VEGF-A expression necessary for the extensive vascularization of the lung. Notably, the constant expression of HIFs during the early phases of lung development suggests that organogenesis proceeds under low oxygen and then there is a gradual increase in oxygenation approaching birth. Thus far the roles of HIFs in PNEC/NEB and related SCLC have not been explored in detail.

Discussion We now have a better understanding of lung stem cells or progenitors that are responsible for further development of the 40+ lung phenotypes that exist in this large organ. Whether there is truly a single stem cell that gives rise to all other cell types within the epithelium during lung development has yet to be definitively explored. As in other organs where epithelial/mesenchymal interactions are critical steps in development it is certain that similar interactions play important roles in the lung. From the perspective of lung biology, PNEC/NEB produce mitogenic and other factors that affect the mesenchyme [Cutz et al., 2007] and thus PNEC/NEB operate at the interface between the epithelium and mesenchyme. Mouse and other animal models have helped to resolve that a neurogenic gene program underlies the differentiation and maturation of PNEC/NEB. What drives this neurogenic program in a select subpopulation within the embryonic lung is not known, although it would appear that

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  305

the first ingrowth from the foregut acquires a neuroendocrine-like (neuroepithelial) phenotype as it makes contact with a presumably specialized mesenchyme. This might then set the stage for further induction of neurogenic genes with Mash-1 identified thus far. Cells that maintain Mash-1 move towards the neuroendocrine side while expression of Hes-1 (within the same gene family) commits the non-neuroendocrine lineage. This patterning is identified with the parenchymatous area of the lung while that of the main bronchi and trachea lying outside this area appears to follow a different pattern where single PNEC are found. This then suggests that alveolization dictates a different set of rules governing epithelial maintenance and repair and involvement of PNEC/NEB. Additional clues to the origin and embryonic phenotype of the lung and PNEC/ NEB come from studies on the temporal and spatial expression of developmental genes expressed during embryogenesis (reviewed in [Kimura and Deutsch, 2007]). Thyroid transcription factor (TTF-1/ NKX2.1) first appears in the embryoid buds that will become thyroid and lung. TTF-1 defines the lung epithelium as compared to other epithelial organs. TTF-1 has been shown to directly regulate transcription of calcitonin gene expression. Interestingly, calcitonin is one of several peptides expressed by NEB. Clara cells and type II pneumocytes express TTF-1 while mucociliary epithelium does not. A recent report [Pelizzoli et al., 2007] shows that during forebrain development TTF-1 transcriptionally up-regulates the intermediate filament nestin, a marker of neuroepithelial stem cells (and also ES stem cells). TTF-1 expression prior to formation of lung buds could suggest that lung development mimics a variant on brain development where factors like ASH-1 (Mash-1 in mice) are instructive for the neuronal like component, i.e., PNEC/NEB. Fidelity of TTF-1 expression and restriction to the distal airway is shown by its expression in SCLC but not poorly differentiated squamous cell carcinoma of the lung, which carries the p63 marker of the basal cells [Kalhor et al., 2006], and in a subset of adenocarcinomas presumably derived from the terminal respiratory units [Tanaka et al., 2007]. Furthermore, constitutive expression of hAsh-1 and transfected SV40 large T antigen under the Clara cell CC-10 promoter in mice produces adenocarcinomas with focal neuroendocrine features while maintaining TFF-1 expression [Linnoila et al., 2005]. However, given that non-small cell lung carcinomas can exhibit features similar to small cell lung carcinomas [Hiroshima et al., 2006], but are still developmentally distinct, the ontological picture of lung development is one of an initiating neuroepithelium with endocrine like functions, and subsequent lineage diversion. To which, current thinking holds that the proximal airways develop along a separate route although being integrated into the whole lung [Perl et al., 2002; Shu et al., 2005]. Taken together it may not be farfetched to explore the idea that PNEC/NEB develop as the neural anlage of a developmentally broad neuroepithelial like lung and in concert with the submucosal ganglia function as a ‘lung brain’; an analogous system has been described for the GI tract [Schemann, 2005]. Since these internal organs connect via the vagus nerve, perhaps evolution has reiterated variations on neuroepithelium, however with specific functions dictated by physiological needs.

306  |  Airway Chemoreceptors in the Vertebrates We then have the scenario where the requirements for regeneration may have to be met by distinct stem cell niches to cover such a large organ that is under dynamic flux and subject to environmental insult. Physical and chemical damage to lung appears to recruit these stem cells, which interestingly are safeguarded by NEB and selected microenvironments. By the same token these stem cells and nearby progenitors may be susceptible to malignant transformation. As the understanding grows about molecular fine tuning of lung development and individual cell type functions we may eventually be able to provide clear insights and solutions for the different diseases that affect lung physiology and maintenance of the functional airway epithelium.

Acknowledgements The authors wish to thank the members of the laboratory who have helped to unravel the mystery of PNEC/NEB during the last three decades and collaborators who have contributed their much appreciated expertise.

References Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat Rec 270: 24-40. Albera, C., Polak, J.M., Janes, S., Griffiths, M.J., Alison, M.R., Wright, N.A., Navaratnarasah, S., Poulsom, R., Jeffery, R., Fisher, C., Burke, M., and Bishop, A.E. 2005. Repopulation of human pulmonary epithelium by bone marrow cells; a potential means to promote repair. Tissue Eng 11: 1115-21. Allahverdian, S., Wojick, K.R., and Dorscheid, D.R. 2006. Airway epithelial wound repair: role of carbohydrate sialyl Lewis X. Am J Physiol Lung Cell Mol Physiol 291: L8128-36. Borges, M., Linnoila, R.I., Van De Velde, H.J.K., Chen, H., Nelkin, B.D., Marby, M., Baylin, S.B., and Ball, D.W. 1997. An achaete-scute homologue essential for neuroendocrine differentiation in the lung. Nature 386: 852-855. Borok, Z., Li, C.G., Liebler, J., Aghamohammadi, N., Londhe, V.A., and Minoo, P. 2006. Developmental pathways and specifications of intrapulmonary stem cells. Pediatr Res 59: 84R-93R. Brouns, I., Van Genechten, J., Hayashi, H., Gajda, M., Gomi, T., Burnstock, G., and Timmermans J.P. 2003. Dual sensory innervation of pulmonary neuroepithelial bodies. Am J Respir Cell Mol Biol 228: 275-85. Bunn, H.F., and Poyton, R.O. 1996. Oxygen sensing and molecular adaptation to hypoxia. Physiol Rev 76: 839-85. Compernolle, V., Brusselmans, K., Acker, T., Hoet, P., Tiwa, M., Beck, H., Plaisance, S., Dewerchin, M., Van Veldhoven, P., Plate, K., Moons, L., Collen, D., and Carmeliet, P. 2002. Loss of HIF-2a and inhibitor of VEGF impair fetal lung maturation whereas treatment with VEGF prevents fatal respiratory distress in premature mice. Nature Med 8: 702-10.

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  307 Cutz, E., Yeger, H., Wong, V., Bienkowski, E., and Chan, W. 1985. Isolation and in vitro maintenance of neuroepithelial bodies from rabbit fetal lung. Lab Invest 53: 672-83. Cutz, E., Yeger, H., and Pan, J. 2007. Pulmonary neuroendocrine cell system in health and disease-recent advances. Pediatric Dev Pathol 10: 419-35. Deimling, J., Thompson, K., Tseu, I., Wang, J., Keijer, R., Tanswell, K., and Post, M. 2007. Mesenchymal maintenance of distal epithelial cell phenotype during late fetal lung development. Am J Physiol Lung Cell Mol Physiol 292: L725-41. Dressler, G.R. 2006. The cellular basis of kidney development. Annu Rev Cell Dev Biol 22: 509-29. Emura, M., Ochiai, A., Gobert-Bohlen A., Panning, B., and Dungworthi, D.L. 1994. Neuroendocrine phenotype differentiation in a hamster lung epithelial call line under low oxygen pressure or after transformation by dietling introsamine. Toxicol/Lett 72: 59-64. Emura, M. 1997. Stem cells of the respiratory epithelium and their in vitro cultivation. In Vitro Cell Dev Biol Anim 33: 3-14. Emura, M. 2002. Stem cells of the respiratory tract. Paediatr Respir Rev 3: 36-40. Emura, M., Ochiai, A., Gobert-Bohlen, A., Panning, B., and Dungworth, D.L. 1994. Neuroendocrine phenotype differentiation in a hamster lung epithelial cell line under low oxygen pressure or after transformation by diethylnitrosamine. Toxicol Lett 72: 59-64. Fanucchi, M.V., Plopper, C.G., Evans, M.J., Hyde, D.M., Van Winkel, L.S., Gershwin, L.J., and Schelegle, E.S. 2006. Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung Cell Mol Physiol 291: L644-50. Fu, X.W., Nurse, C.A., Farragher, S.M., and Cutz, E. 2003. Expression of functional nicotinic acetylcholine receptors in neuroepithelial bodies of neonatal hamster lung. Am J Physiol Lung Cell Mol Physiol 285: L1203-12. Germann, P.G., Ueberschar, S., Geruli, A., and Emura, M. 1993. In vitro induction of type II pneumocyte-related differentiation in aclonal fetal bronchiolo-alveolar eoithelial cell line (M3E3/C3). Exp Toxicol Pathol 45: 315-24. Giangreco, A., Reynolds, S.D., and Stripp, B.R. 2002. Terminal bronchioles harbor a unique airway stem cell population that localizes to the brochoalveolar duct junction. Am J Pathol 161: 173-82. Gomperts, B.N., Belpario, J.A., Rao, P.N., Randell, S.H., Fishbein, M.C., Burdick, M.D., and Strieter, R.M. 2006. Circulating progenitor epithelial cells traffic via CXCR4/CXCL12 in response to airway injury. J Immunol 176: 1916-27. Griffiths, M.J. Bonnet, D., and James, S.M. 2005. Stem cells of the alveolar epithelium. Lancet 366: 249-60. Groenman, F., Rutter, M., Cannigia, I., Tibboel, D., and Post, M. 2007. Hypoxia-inducible factors in the first trimester human lung. J Histochem Cytochem 55: 355-63. Hajj, R., Baranek, T., Le Naour, R., Lesimple, P., Puchelle, E., and Coraux, C. 2007. Basal cells of the human adult airway surface epithelium retain transit-amplifying cell properties. Stem Cells 25: 139-48. Haley, K.J., Drazen, J.M., Osthanondh, R., and Sunday, M.E. 1997. Comparison of the ontogeny of protein gene product 9.5, chromogranin A and proliferating nuclear antigen in developing human lung. Microsc Res Tech 37: 62-68. Heidbreder, M., Frohlich, F., Johren, O., Dendorfer, A., Quadri, F., and Dominiak, P. 2003. Hypoxia rapidly activates HIF-3alpha mRNA expression. FASEB J 17: 1541-43. Hiroshima, K., Iyoda, A., Shida, T., Shibuya, K., Iizasa, T., Kishi, H., Tanizawa, T., Fujisawa, T., and Nakatani, Y. 2006. Distinction of pulmonary large cell neuroendocrine carcinoma from small cell lung carcinoma: a morphological, immunohistochemical, and molecular analysis. Mod Pathol 19: 1358-68.

308  |  Airway Chemoreceptors in the Vertebrates Hong, K.U., Reynolds, S.D., Giangreco, A., Hurley, C.M., and Stripp, B.R. 2001. Clara cell secretory protein-expressing cells of the airway neuroepithelial body microenvironment include a label-retaining subset and are critical for epithelial renewal after progenitor cell depletion. Am J Respir Cell Mol Biol 24: 671-81. Hong, K.U., Reynolds, S.D., Watkins, S., Fuchs, E., and Stripp, BR. 2004. In vivo differentiation potential of tracheal basal cells: evidence for multipotent and unipotent subpopulations. Am J Physiol Lung Cell Mol Physiol 286: L643-49. Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M., Salic, A., Asara, J.M., Lane, W.S., and Kaelin, W.G Jr. 2001. HIF-alpha targeted for VHL-mediated destruction by proline hydroxylation: implication for O2 sensing. Science 292: 464-68. Iyer, N.V., Kotch, L.E., Agani, F., Leung, S.W., Laughner, E., Wenger, R.H., Gassman, M., Gearhart, J.D., Lawler, A.M., Yu, A.Y., and Semenza, G.L. 1998. Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1a. Genes Dev 12: 149-62. Kalhor, N., Zander, D.S., and Liu, J. 2006. TTF-1 and p63 for distinguishing pulmonary smallcell carcinoma from poorly differentiated squamous cell carcinoma in previously pap-stained cytologic material. Mod Pathol 19: 1117-23. Kim, C.F. 2007. Paving the road for lung stem cell biology: bronchioalveolar stem cells and other putative distal lung stem cells. Am J Physiol Lung Cell Mol Physiol, Aug 10, epub. Kim, C.F., Jackson, E.L., Wolfenden, A.E., Lawrence, S., Babar, I., Vogel, S., Crowley, D., Bonson, R.T., and Jacks, T. 2005. Identification of bronchioalveolar stem cells in normal lung and lung cancer. Cell 121: 823-35. Kimura, J., and Deutsch, G.H. 2007. Key mechanisms of early lung development. Pediat Dev Pathol 10: 335-47. Kline, D.D., Peng, Y.-J, Manalo, D.J., Semenza, G.L., and Prabhakar, N.R. 2002. Defective carotid body function and impaired ventilatory responses to chronic hypoxia in mice partially deficient for hypoxia-inducible factor 1alpha. Proc Natl Acad Sci USA 99: 821-26. Liang, S.X., Summer, R., Sun, X., and Fine, A. 2005. Gene expression profiling and localization of Hoechst-effluxing CD45- and CD45+ cells in the embryonic mouse lung. Physiol Genomics 23: 172-81. Linnoila, R.I. 2006. Functional facets of the pulmonary neuroendocrine system. Lab Invest 86: 425-44. Linnoila, R.I., Naizhen, X., Meuwissen, R., Berns, A., and DeMayo, F.J. 2005. Mouse lung neuroendocrine carcinomas: distinct morphologies, same transcription factors. Exp cell Res 31: 37-55. Majka, S.M., Beutz, M.A., Hagen, M., Izzo, A.A., Voelkel, N., and Helm, K.M. 2005. Identification of novel resident pulmonary stem cells: form and function of the lung side population. Stem Cells 23: 1073-81. Maltepe, E., and Simon, M.C. 1998. Oxygen, genes, and development: an analysis of the role of hypoxic gene regulation during murine vascular development. J Mol Med 7: 391-401. Pan, J., Copland, I., Post, M., Yeger, H., Cutz, E. 2006. Mechanical stretch induced serotonin release from pulmonary neuroendocrine cells: implications for lung development. Am J Physiol Lung Cell Mol Physiol. 290:L185-93. Pan, J., Yeger, H., and Cutz, E. 2002. Neuronal developmental marker FORSE-1 identifies a putative progenitor of the pulmonary neuroendocrine cell lineage during lung development. J Histochem Cytochem 50: 1567-78. Park, K.S., Wells, J.M., Zorn, A.M., Wert, S.E., Laubach, V.E., Fernandez, L.G., and Whitsett, J.A. 2006. Transdifferentiation of ciliated cells during repair of the respiratory epithelium. Am J Respir Cell Mol Biol 34: 151-57.

Precursors and Stem Cells of the Pulmonary Neuroendocrine Cell System  |  309 Pelizzoli, R., Tacchetti, C., Luzza, P., Strangio, A., Bellese, G., Zappia, E., and Guazzi, S. 2007. TTF-1/NKX2.1 up-regulates the in vivo transcription of nestin. In J Dev Biol 52: 55-62. Perl, A.K., Wert, S.E., Nagy, A., Lobe, C.G., and Whitsett, J.A. 2002. Early restriction of peripheral and proximal cell lineages during formation of the lung. Proc Natl Acad Sci USA 99: 10482-87. Rajatapiti, P., van der Horst, I., de Rooij, J., Tran, M., Maxwell, P., Tibboel, D., Rottier, R., and De Krijger, R. 2007. Expression of hypoxia- inducible factors in normal human lung development. Pediatr Dev Pathol Jul2;1 [Epub ahead of print]. Rawlins, E.L., and Hogan, B.L. 2006. Epithelial stem cells of the lung: privileged few or opportunities for many. Development 133: 2455-65. Rawlins, E.L., Ostrowski, L.E., Randell, S.H., and Hogan, B.L. 2007. Lung development and repair: contribution of the ciliated lineage. Proc Natl Acad Sci USA 104: 410-17. Reynolds, S.D., Giangreco, A., Power, J.H., and Stripp, B.R. 2000. Neuroepithelial bodies of pulmonary airways serve as a reservoir of progenitor cells capable of epithelial regeneration. Am J Pathol 156: 269-78. Schemann, M. 2005. Control of gastrointestinal motility by the “gut brain”—the enteric nervous system. J Pediatr Gastroenterol Nutr 41: S4-S6. Schuller, H.M., McGavin, M.D., Orloff, M., Reichert, A., and Porter, B. 1995. Simultaneous exposure to nicotine and hyperoxia causes tumors in hamsters. Lab Invest 73: 448-56. Semenza, G.L. 2001a. Hypoxia-inducible factor 1-oxygen homeostasis and disease pathophysiology. Trends Mol Med 7: 345-50. Semenza, G.L. 2001b. HIF-1 and mechanisms of hypoxia sensing. Curr Opin Cell Biol 13: 16771. Shannon, J.M., Nielsen, L.D., Gebb, S.A., and Randell, S.H. 1998. Mesenchyme specifies epithelial differentiation in reciprocal recombinants of embryonic lung and trachea. Dev Dyn 212: 482-94. Shu, W., Guttentag, S., Wang, Z., Andl, T., Ballard, P., Lu, M.M., Piccolo, S., Birchmeier, W., Whitsett, J.A., Miller, S.E., and Morrisey, E.E. 2005. Wnt/beta-catenin signaling acts upstream of N-myc, BMP4, and FGF signaling to regulate proximal-distal patterning in the lung. Dev Biol 283: 226-39. Sorokin, S.P., Ebina, M., and Hoyt, R.F. Jr. 1993. Development of PGP 9.5- and calcitonin gene-related peptide-like immunoreactivity in organ cultured fetal rat lungs. Anat Rec 236: 213-25. Summer, R., Kotton, D.N., Sun, X., Fitzsimmons, K., and Fine, A. 2004. Translational physiology: origin and phenotype of lung side population cells. Am J Physiol Lung Cell Mol Physiol 287: L477-83. Sunday, M.E., and Cutz, E. 2000. The role of neuroendocrine cells in fetal and post-natal lung. In: Endocrinology of the Lung, C.R. Mendelson (Ed.), Humana Press, Totowa, NJ, pp. 209336. Tanaka, H., Yanagisawa, K., Shinjo, K., and et al. 2007. Lineage-specific dependency of lung adenocarcinomas on the lung development of TTF-1. Cancer Res 67: 6007-11. Tian, H., Hammer, R.E., Matsumoto, A.M., Russell, D.W., and McKnight, S.L. 1998. The hypoxia-responsive transcription factor EPAS1 is essential for catecholamine homeostasis and protection against heart failure during embryonic development. Genes Dev 12: 3320-24. van Tuyl, M., Liu, J., Wang, J., Kuliszewski, M., Tibboel, D., and Post, M. 2005. Role of oxygen and vascular development in epithelial branching morphogenesis of the developing mouse lung. Am J Physiol Lung Cell Mol Physiol 288: L167-78.

310  |  Airway Chemoreceptors in the Vertebrates Wang, X.Y., Dakir el, H., Naizhen, X., Jensen-Taubman, S.M., DeMayo, F.J., and Linnoila, R.I. 2007. Achaete-scute homolog-1 linked to remodeling and preneoplasia of pulmonary epithelium. Lab Invest 87: 527-39. Warburton, D., Schwarz, M., Tefft, D., Flores-Delgado, G., Anderson, K.D., and Cardoso, W.V. 2000. The molecular basis of lung morphogenesis. Mech Dev 92: 55-81. Wenger, R.H. 2002. Cellular adaptation to hypoxia: O2-sensing protein hydroxylases, hypoxiainducible transcription factors, and O2-regulated gene expression. FASEB J 16: 1151-62. Wiesener, M.S., Jurgensen, J.S., Rosenberger, C., Scholze, C.K., Horstrup, J.H., Warnecke, C., Mandriota, S., Bechmann, I., Frei, U.A., Pugh, C.W., Ratcliffe, P.J., Bachmann, S., Maxwell, P.H., and Eckardt, K.U. 2003. Widespread hypoxia-inducible expression of HIF-2 alpha in distinct cell populations of different organs. FASEB J 17: 271-73.

14 Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators: Some New Findings and a Review of the Literature Alfons T.L. Van Lommel1*, Tania Bollé1 and Peter W. Hellings2

Abstract Neuroepithelial bodies (NEBs) are corpuscular arrangements of endocrine cells in the airway epithelium that are implicated in the regulation of lung organogenesis and that act as hypoxiasensitive chemoreceptors. Their strategic position in contact with the inhaled air, in combination with their endocrine nature, might also enable them to detect intraluminal antigens and to modulate the local immune response by secretion of specific immunomodulatory substances. In this paper, some ultrastructural aspects of the secretory response of NEBs to ovalbumin (OVA) were quantified in a mouse model of allergic asthma. Mice were sensitized to OVA and challenged with OVA or saline. Sensitization, alone or followed by OVA challenge, increased the number of exocytotic profiles and the volume percent of dense-cored vesicles (DCVs), indicating enhanced secretion as well as synthesis of bioactive substances. Therefore, NEBs appear to respond to systemic sensitization with synthesis and subepithelial secretion of messenger molecules, suggesting a role in airway homeostasis and allergic airway disease. A prime candidate immunomodulatory messenger molecule among several bioactive substances known to be secreted by NEBs is calcitonin gene-related peptide (CGRP). The immunomodulatory functions of CGRP are well established and include induction of eosinophil chemotaxis, reduction of antigen presentation, and modulation of cytokine and chemokine secretion. The Leuven Catholic University, Medical Faculty, Department of Morphology and Molecular Pathology, Minderbroedersstraat 12, 3000 Leuven, Belgium. 2 Department of Otorhinolaryngology, Kapucijnenvoer 33, 3000 Leuven, Belgium. *Author for Correspondence: [email protected] 1

312  |  Airway Chemoreceptors in the Vertebrates discussion presents an overview of the already known immunomodulatory effects of the various known NEB secretory products, including CGRP, and thus surveys the hypothetical NEB contribution to the pulmonary response to antigenic stimulation.

Keywords: Airways, bombesin/gastrin-releasing peptide (GRP), calcitonin generelated peptide (CGRP), immunomodulation, lungs, neuroepithelial bodies (NEBs), neuroendocrine cells (NECs), respiratory epithelium, serotonin

Introduction The various mucous membranes are gateways to the body’s interior and are equipped with diverse kinds of cells with sensory functions. Mucosal receptor cells are polarized cells that, upon stimulus reception at their luminal pole, engage in basal secretion of bioactive substances at their abluminal pole. The respiratory mucosa is no exception to this principle. For almost 40 years now, the corpuscular neuroepithelial bodies (NEBs) and the solitary neuroendocrine cells (NECs) they house have been the subject of systematic research by numerous teams all over the world. For years, there has been controversy over the exact relationship between the two forms. Recently, the discovery that most solitary NECs receive an innervation (Weichselbaum et al., 2005) has substantially narrowed the gap separating them. Therefore, when it is not absolutely necessary to make the distinction, the general term NEC will be used throughout this text to designate the two forms. In a nutshell, pulmonary NECs are stimulus-sensitive cells that synthesize bioactive substances in their subnuclear cytoplasm, store them in dense-cored secretory vesicles, and may release them by exocytosis at their abluminal cell pole. This essentially endocrine (paracrine is perhaps more exact) nature enables them to assume several functions. Among these is the supply of growth factors such as bombesin or gastrin-releasing peptide (Stahlman and Gray, 1997; Wang et al., 1996) with which they contribute, to what extent is not yet certain, to lung development in the embryo (Emanuel et al., 1999; Kresch et al., 1999; King et al., 1995; Aguayo et al., 1994). Pulmonary NECs appear at a very early stage in lung development (they are in fact the first pulmonary cell type to undergo differentiation) and have been reported to occur in greatest numbers in the prenatal period (Pan et al., 2004, 2002; McDowell et al., 1994; Cho et al., 1989). After birth, their growth-promoting properties come into play again in various kinds of lung injury, regeneration, and remodeling (Cullen et al., 2000), and they provide sheltered microenvironments for the storage of epithelial progenitor cells (Reynolds et al., 2000). In the neonatal organism, specific membrane receptors allow them to react to various intraluminal stimuli, notably ww, to which they respond with abluminal endocrine secretion of neurotransmitters, hormones, and vasomotor substances, thereby assuming the role of chemoreceptors, analogous to the arterial chemoreceptors (Kemp et al., 2003, 2002; Cutz and Jackson, 1999). Apart

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  313

from bombesin/gastrin-releasing peptide, the bioactive substances that were most consistently found to be synthesized and stored in their cytoplasm include serotonin (Cho et al., 1989) and calcitonin gene-related peptide (CGRP) (Avadhanam et al., 1997; Luts et al., 1994; Keith et al., 1991). The vasomotor and neurotransmitter properties of these substances are beyond dispute. The innervation of the NEBs has been the subject of particularly close scrutiny. Originally conceived as predominantly vagosensory, it proved to be much more complex (Brouns et al., 2002, 2003; Adriaensen et al., 2003). Finally, it looks as though a third function is to be added to the two already well established: immunomodulation. As is the case with the other two, this function is a consequence of the fundamentally endocrine nature of the NECs: luminal contact with antigens may result in abluminal secretion, or at least synthesis and storage, of immunomodulatory substances. The idea of pulmonary NECs as immunomodulators is not new, or even recent. It was first advanced some 20 years ago (Marchevsky et al., 1984), when it was demonstrated that the frequencies of argyrophilic solitary NECs in guinea pig airways were influenced by immunization with ovalbumin, followed or not by antigen challenge. These results were repeatedly confirmed (Bousbaa and Fleury-Feith, 1991; Bousbaa et al., 1994). It has also been suggested that pulmonary NEBs would chemoattract or influence various kinds of immune cells. Quantitative indications of this were obtained in neonatal dogs, cats, and hamsters, where it was demonstrated that various kinds of immune cells were lodged preferentially in the vicinity of a NEB, either in the airway epithelium itself, or in the underlying corium (Van Lommel et al., 1995). This paper is a report of some new findings in support of the theory that NEBs function as pulmonary immunomodulators. Mouse models of allergic asthma have proven to be interesting tools to unravel mechanisms underlying the allergic immune response (Hellings et al., 2003). Due to the standardization of sensitization and allergen exposure, we can take advantage of a mouse model to study the reactions and functions of NEBs in vivo. In the second place, this text is a review of what little we already know about the immunomodulatory role of the pulmonary NEBs. Lastly, we will try to outline what other roles the NEBs may play in the complex immunological behavior of the lungs.

Materials and Methods Ten adult Balb-c mice were sensitized with intraperitoneal injections of ovalbumin (OVA), 10 µg per individual, on alternate days from day 1 until day 12, as described previously (Hellings et al., 2003). From days 33 to 40, five sensitized mice were challenged (challenge group) daily by exposure to nebulized OVA (10 mg/ml albumin in physiological saline) during 5 min. The other five, the sensitization group, were challenged with saline. In addition, five sham sensitized and challenged individuals, the

314  |  Airway Chemoreceptors in the Vertebrates control group, were included. All animals were sacrificed on day 41, i.e., 24 h after the last aerosol treatment. These procedures were approved by the ethical commission of animal experiments of Leuven Catholic University. The animals were killed by cervical dislocation. The thorax was opened by incising the diaphragm, the cervical trachea was exposed and intubated, and the lungs were inflated with ice-cold glutaraldehyde 1.25% in 0.1 M phosphate buffer, until they filled the thoracic cavity. The lungs were removed; the lung lobes were separated, immersed in fresh fixative, and degassed in a vacuum chamber. The lung lobes were sliced in 1 mm3 cubes, immersed in fresh fixative for 2 h, rinsed in buffer, and postfixed by immersion in 1% osmium tetroxide in 0.1 M phosphate buffer for another 2 h. They were processed for electron microscopy by dehydration in graded alcohols, rinsing in propylene oxide, impregnation and embedding in di-epoxyresin (DER™, Fluka). Sections of 1 µm were cut, stained with toluidine blue, and examined light optically. Where NEBs were spotted, the tissue blocks were trimmed of superfluous tissue, and 50 nm sections were cut, mounted on grids, and stained with uranyl acetate and lead citrate. Sections were examined with a Philips CM 10 electron microscope at 50 kV. The abluminal parts of the NEBs were photographed. In each NEB, as long as possible a length of abluminal cell membrane was measured with a planimeter on photographs printed to a magnification of 28.000, and these lengths were converted into micrometers. The number of exocytotic profiles of secretory dense-cored vesicles (DCVs) was counted to quantify the exocytotic activity in various experimental conditions. Four kinds of profiles were regarded as indicative of exocytosis: (1) DCVs at a distance from the abluminal cell membrane, smaller than their own diameter, (2) DCVs that had docked with the cell membrane, (3) DCVs with their lumen in continuity with the extracellular space (omega profiles), and (4) empty vesicles at a distance from the abluminal cell membrane, smaller than their own diameter. The latter kind of profiles was regarded as recycled DCV membrane. The total number of profiles was pooled for each NEB and divided by the length, in micrometers, of the corresponding stretch of abluminal cell membrane. The second parameter to be quantified was the volume percent of the subnuclear cytoplasm, occupied by DCVs. To this end, an orthogonal point grid (density: 4 points per cm2) was superimposed on as large as possible an area of subnuclear cytoplasm, photographed and printed to a magnification of 28.000. For each NEB, the number of hits was recorded and expressed as a percentage of the total number of points covering the area under consideration. Statistical analysis was performed by means of a computer program: Graph Pad Prism 2.01. In all three groups of animals, data obtained of individual NEBs showed a normal frequency distribution for both parameters. Likewise, Bartlett’s test indicated that variances were comparable in all three experimental groups and for both parameters. Results of control, sensitized, and challenged animals were expressed as means +/- standard deviation. Statistical significance of differences was determined

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  315

by one-way analysis of variance, followed by Tukey’s multicomparison test, P values of 0.05 or less being regarded as statistically significant.

Results The microscopic aspect of the lungs of control as well as sensitized animals was normal. In challenged animals, lung tissue was heavily infiltrated with lymphocytes and other kinds of immune cells, to the extent that spotting NEBs was much more difficult. Figure 1 shows a representative example of NEB subnuclear cytoplasm containing large numbers of DCVs, a number of which are caught in various stages of exocytosis. As is clear from Table 1, sensitization led to a significant increase in both the number of exocytotic profiles and the volume percent of DCVs, as compared to the control group. In challenged animals, these values remained equally high but did not significantly deviate from the values of the sensitization group. Table 1:

Quantification of the response of pulmonary neuroepithelial bodies (NEBs) to exposure to antigens.

Groups/parameters

Exocytotic profiles DCV / micrometer (n)

Volume percent DCV (n)

Control

0.9 +/- 0.1

(17)

6.0 +/- 0.6 (13)

Sensitization

1.5 +/- 0.1

(22)

S

10.2 +/- 0.7 (17) S

Challenge

1.6 +/- 0.1

(17)

S

12.3 +/- 1.2 (13) S

Exocytotic profile counts at the basal cell membrane and determination of volume percentages in the subnuclear cytoplasm, of dense-cored vesicles (DCVs). Animals either sensitized with albumin (sensitization), or sensitized and challenged (challenge) were compared with controls. Results are expressed as means +/- standard deviation. The numbers of NEBs measured appear in parentheses. Statistical significances: S = significantly different from the control group, according to Tukey’s multicomparison test, at the 0.05 level.

Discussion Previous work on the hypothetical involvement of pulmonary NECs in immunological reactions mainly dealt with cell frequency counts, after specific histochemical staining, in the guinea pig lung. Sensitization with ovalbumin resulted in a 2- to 10-fold increase in argyrophilic NEC frequencies, depending on the airway segment, and on the parameter measured (cells per millimeter or cells per 100 nuclei). Sensitization followed, after 21 d, by challenge led to similar changes in the larger airways, but a 30-fold decrease in the bronchioles (Marchevsky et al., 1984). These results were interpreted as demonstrating enhanced storage or release, respectively, of bioactive substances upon contact with

316  |  Airway Chemoreceptors in the Vertebrates

Figure 1: A representative view of a neuroepithelial body abluminal pole. As a consequence of sensitization with ovalbumin, dense-cored vesicles (DCV) accumulate in the subnuclear cytoplasm, dock with the abluminal cell membrane, and undergo exocytosis (e). This response indicates secretion of bioactive substances into the corium upon antigenic stimulation. Mouse lung, TEM, 28.000 x.

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  317

antigen. The authors were unable to offer an explanation for the aberrant results in the bronchioles after challenge. They suggested this might be related to the method of administration of antigen during challenge, which was intracardiac. In the previously mentioned study, guinea pigs were sacrificed 10 min after challenge. Animals that had likewise been challenged 21 d after sensitization have been allowed to survive for longer periods (Bousbaa and Fleury-Feith, 1991), allowing observations in the later phases of challenge. In sensitized animals, the number of argyrophilic NECs per millimeter of bronchial wall increased by about 50%. Cell counts tended to remain in that range for up to 6 h after challenge, decreasing to control levels by 24 h after challenge. As the most likely explanation, it was again advanced that increased synthesis of argyrophilic substances as a result of sensitization was followed by postchallenge exocytotic release. In a later, parallel experiment (Bousbaa et al., 1994), the same counts were made after staining of guinea pig pulmonary NECs by incubation with chromogranin A antiserum. Again, NEC numbers rose upon contact with antigen: there was a 2- to 3-fold increase after sensitization, combined or not with challenge, followed by a post-challenge decrease to control levels in less than 24 h. Significantly, frequencies of NECs that had been incubated with anti-neuron specific enolase were not influenced. As chromogranin A is involved in vesicular storage of hormones, while neuron-specific enolase is a cytosolic enzyme, unaffected by vesicular traffic, these observations add significant weight to the hypothesis that pulmonary NECs react to antigen exposure by synthesis and subsequent release of bioactive substances. In control animals, there were about 10 times as many NSE-positive cells as there were chromogranin A-positive cells. Sensitization and challenge lowered this ratio to only 3 or 4, while total numbers of NSE-positive cells remained constant. Again, these observations nicely fit with the interpretation of increased detectability of existing cells by accumulation of bioactive materials. The new results we present here strengthen and extend the hypothesis, based on earlier work, that the processing of bioactive materials in pulmonary NEBs is affected by antigenic stimulation. Mouse pulmonary NEBs reacted to sensitization only or sensitization followed by challenge by raised frequencies of DCV exocytotic profiles, as well as elevated DCV volume percentages, at least at the post-sensitization and post-challenge times that were under consideration here. The first reaction points to enhanced endocrine secretion of bioactive substances, while the second one is most readily interpreted as enhanced synthetic activity. Synthesis and release could, in theory, neutralize each other. We observed a rise in both parameters, and the increase in DCV volume percent apparently is a net effect. The increase in DCV numbers elegantly accounts for the increased detectability of cells reported in previous work. It is not inconceivable that, at a later stage, synthesis is diminished. In that case, continued secretion would account for the reported decreased detectability of NECs. Tissue levels of specific NEB secretory products have also been investigated after antigenic stimulation, in the first place CGRP. In line with previous reports on raised

318  |  Airway Chemoreceptors in the Vertebrates levels of argyrophilic substances or chromogranin A, Tsukiji et al. (2004) observed increased expression of CGRP mRNA in rat lungs after sensitization and challenge with ovalbumin. The rise of the β-isoform in particular was marked. In contrast with cell frequency counts, expression levels of this isoform remained high for up to 120 h after challenge. The β-isoform was specifically located in pulmonary NEBs, pointing to a direct involvement of NEBs in pulmonary immune reactions. In ovalbuminsensitized mice, three successive challenges resulted in decrease of CGRP-positive NEB frequencies and mucosal nerve fiber densities, expressed as immunoreactive tissue area per unit length of epithelium, 24 to 48 h after the last challenge (Dakhama et al., 2002). Immunostaining with the pan-neuronal marker protein gene product 9.5 revealed no changes in airway nerve density. In guinea pig airways, frequencies of CGRPimmunoreactive nerve fibers were diminished as a consequence of sensitization and 6 h after challenge with toluene diisocyanate (Mapp et al., 1998). We may summarize that the level of CGRP in pulmonary NECs rises, parallel to a rise in cell frequencies, as a consequence of antigenic stimulation, and that this rise persists at least until the early stages of challenge. Apart from endogenous CGRP depletion, sensitization and challenge with ovalbumin was found to lead to airway hyperresponsiveness to inhalation of methacholine in mice (Dakhama et al., 2002). Administration of exogenous CGRP prevented this response or restored normal airway responsiveness, while a CGRP receptor antagonist induced airway hyperresponsiveness or prevented its restoration. Conflicting results were obtained in CGRP gene-disrupted mice, where antigeninduced airway hyperresponsiveness was attenuated (Aoki-Nagase et al., 2002). Next to their increased detectability and storage of bioactive substances, another phenomenon that has been linked to the hypothetical immunomodulatory activities of pulmonary NECs is their apparent chemotactic effects on various types of immune cells. Eosinophil infiltration after sensitization and challenge was reported several times, but it is difficult to link this specifically to NEC influences. However, in human lungs affected by eosinophilic granuloma, a fibrotic disease associated with smoking, a 10-fold increase of bombesin-positive NECs was recorded, relative to cell numbers in granuloma-free lungs of smokers or in idiopathic pulmonary fibrosis, a disease not associated with smoking (Aguayo et al., 1990). Thus, pathology hints at a specific link between eosinophil infiltration and NECs. In airways of autotransplanted dog lungs, a preferential association of mucosal mast cells with NEBs was recorded. In addition, neutrophilic granulocytes were regularly observed to lodge between NEB corpuscular cells in cats and hamsters (Figure 2) (Van Lommel et al., 1995). This seems to point at a chemotactic effect of NEB secretory products. This function may be analogous to that of the specialized antigen processing M-cells of the intestinal epithelium. M-cells of the gut capture intraluminal antigens and regulate their contacts with the underlying lymphoid tissue of Peyer’s patches, providing membrane pockets for lymphocytes (Kraehlenbuhl and Neutra, 2000). Similarly,

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  319

Figure 2: Part of a neuroepithelial body (DCV are visible at the abluminal pole) with several neutrophilic granulocytes (N) that have insinuated themselves between the corpuscular cells. It is theoretically possible that they did so in response to chemoattractant stimuli emanating from the NEB as a consequence of luminal antigenic stimulus reception. Hamster lung, TEM, 5.400 x.

the NEB basolateral cell membrane could function in the formation of intraepithelial pockets to provide docking sites for infraepithelial immune cells, effectively shortening the distance between those cells and the luminal, antigen-sampling, epithelial surface. One of the substances responsible for the chemotactic effects of NECs may well be CGRP, since it is well known that CGRP affects various kinds of immune cells.

320  |  Airway Chemoreceptors in the Vertebrates CGRP has chemotactic effects on eosinophils (Dunzendorfer et al., 1998). Pretreatment with CGRP significantly enhanced chemotactic responses of eosinophils to platelet-activating factor and leukotriene B4 (Numao and Agrawal, 1992). Administration of aerosolized CGRP to rats raised the numbers of eosinophils in bronchoalveolar lavage fluid (Bellibas, 1996). Eosinophil infiltration was observed to accompany CGRP depletion in mouse lungs 24 to 48 h after repeated antigenic challenge (Dakhama et al., 2002). Eosinophils are also recruited by factors such as very late antigen-4 or interleukin-5, and eosinophil infiltration was abolished by antisera against these factors. Significantly, so was CGRP depletion, implying an indirect role for CGRP. Macrophages are also influenced by CGRP. Mouse peritoneal macrophages produced less tumor necrosis factor α as a consequence of bacterial cell wall endotoxinmediated induction when treated with CGRP (Feng et al., 1997). In unstimulated peritoneal macrophages, CGRP augments the production of tumor necrosis factor and interleukin-1β (Yaraee et al., 2003). CGRP also augmented endotoxin-induced and granulocyte/macrophage colony-stimulating factor-induced release of interleukin-10 protein and expression of interleukin-10 mRNA. Release of interleukin-1β protein, and expression of interleukin-12 and interleukin-1β mRNA, on the other hand, were suppressed. Upregulation of B7-2 co-stimulatory molecule was suppressed by CGRP, an effect that was neutralized by antibodies against interleukin-10. Collectively, these data suggest that CGRP suppresses antigen presentation, and that this effect is partly mediated by changes in cytokine production (Torii et al., 1997). Antigen presentation by peritoneal macrophages was reduced in the presence of CGRP (Yaraee et al., 2005). Bacterial endotoxin induced interleukin-6 release from mouse peritoneal macrophages, and this effect was augmented by CGRP (Tang et al., 1998). CGRP profoundly inhibited the interferon γ induced production of peroxide by macrophages (Nong et al., 1989). It attenuated bacterial endotoxin-induced interleukin-12 release in mouse peritoneal macrophages. Interleukin-12 mRNA levels were also reduced (Liu et al., 2000). Bone marrow-derived macrophages expressed functional CGRP receptors whose activation caused upregulation of c-Fos and interleukin-6 mRNAs (Fernandez et al., 2001). In human peripheral blood mononuclear cells, CGRP inhibited antigen-induced proliferation, decreased expression of B7-1 and B7-2 co-stimulatory molecules, augmented interleukin-10 production, and inhibited interleukin-12 production (Fox et al., 1997), while it increased secretion of interleukin-1β, interleukin-6, and tumor necrosis factor α (Cuesta et al., 2002). In general, CGRP suppresses antigen-presenting cells. Langerhans cell antigen presentation, as judged by the occurrence of delayed-type hypersensitivity, or by decreased incorporation of DNA precursors in a responsive T helper cell clone, was inhibited, as was induction of the co-stimulatory molecule B7-2. CGRP also inhibited antigen presentation by a preparation of cultured epidermal cells containing Langerhans

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  321

cells, although it did not show stimulation of T-cell proliferation (Hosoi et al., 1993). In the presence of CGRP, cell surface expression of HLA-DR MHC class II antigen presentation molecules, as well as expression of the co-stimulatory molecule CD86, by human dendritic cells, was decreased. In addition, the dendritic cell-driven proliferative response of T-cells was dampened (Carucci et al., 2000). In the Langerhans cell-like cell line XS52, CGRP augmented the bacterial lipopolysaccharide- and granulocyte/ macrophage colony-stimulating factor-induced release of interleukin-10 protein and expression of interleukin-10 mRNA. Release of interleukin-1β protein and expression of interleukin-12 and interleukin-1β mRNA, on the other hand, were suppressed. Lipopolysaccharide-induced and granulocyte/macrophage colony-stimulating factorinduced upregulation of B7-2 was suppressed by CGRP. CGRP-mediated suppression of antigen presentation by a mixed preparation of epidermal and Langerhans cells was prevented by antibodies against interleukin-10 (Torii et al., 1997). Langerhans cells were observed to be contacted by CGRP-positive nerve fibers (Hosoi et al., 1993), arguing for an in vivo neural source of the CGRP acting on them. In a B lymphocyte cell line, CGRP inhibited lipopolysaccharide-induced expression of surface immunoglobulin and its associated mRNA (McGillis et al., 1993). Interleukin-7-induced colony formation by B cell precursors was inhibited in the presence of CGRP (Fernandez et al., 2000). Mitogen-induced proliferation of mouse splenic T cells was inhibited by CGRP (Boudard and Bastide, 1991). In a murine T helper cell clone, CGRP inhibited interleukin-2 production, as well as expression of interleukin-2 mRNAs, tumor necrosis factor α and β, and interferon γ (Wang et al., 1992). CGRP suppressed proliferation of thymic T cells, an effect partly attributed to raised apoptosis (Bulloch et al., 1998). Physiological concentrations of CGRP were found to induce β1 integrin-mediated adhesion of resting human T cells to fibronectin (Levite et al., 1998). Incubation with CGRP incited various types of mouse T helper cells to secrete cytokines, including interleukin-2, interferon γ, interleukin-4, and interleukin-10 (Levite, 1998). A human bronchial epithelial cell line, BEAS-2B, responded to exposure to CGRP by increasing its intracellular calcium, and by augmented synthesis and release of inflammatory cytokines (Veronesi et al., 1999). In general, CGRP can be seen to have both inflammatory action (e.g., eosinophil chemotaxis induction), and anti-inflammatory action (e.g., reduction of antigen presentation by macrophages, mononuclear cells, and dendritic cells, as well as inhibition of lymphocytes) on the various kinds of cells involved in immunological responses. CGRP is implicated in immunological reactions in more ways than by its actions on immune cells. Its vasodilatory properties enable it to potentiate the induction of inflammatory edema (microvascular permeability) by interleukin-1 (Buckley et al., 1991). The integumentary edema-promoting actions of inflammatory mediators, such as histamine, leukotriene B4, and serotonin, were inhibited by CGRP (Raud et al.,

322  |  Airway Chemoreceptors in the Vertebrates 1991). Since CGRP is a potent vasodilator, is chemotactic for eosinophils, and has various other immunomodulatory roles, its release may cause airway inflammation to intensify and spread. CGRP is one of the agents involved in neurogenic inflammation of the airways, the phenomenon whereby inflammatory mediators stimulate sensory nerve fibers, causing antidromic release of vasodilatory neuropeptides from these terminals, or from efferent fiber terminals that are in continuity with them, i.e., through a peripheral axon reflex (Barnes, 2001). Neurogenic inflammation depends on airway innervation: somewhere, and most probably at the sensory end, NEBs may well fit in. CGRP is present in NEBs and their sensory nerves, which are suited to accommodate axon reflexes since they display morphologically afferent as well as efferent nerve endings. CGRP and pulmonary NECs have been linked to asthma, but in the rare cases where such a link has been expressly sought, direct evidence is hard to come by. For instance, the relative density of CGRP-positive nerve fibers in the airway walls was unaffected by asthma and bronchitis (Chanez et al., 1998). Apart from CGRP, there are other substances that have been located in NEBs with fairly great regularity: the amine serotonin or 5-hydroxytryptamin (5-HT) in rabbits, and the peptide bombesin in amphibians or its mammalian counterpart bombesin-like peptide (BLP) or gastrin-releasing peptide (GRP), especially in humans (Stahlman and Gray, 1997; Wang et al., 1996). As it happens, both substances have been allocated immunomodulatory functions, inflammatory as well as anti-inflammatory. Gastrin-releasing peptide stimulated the natural killer activity (cytolysis) of murine splenic, axillary nodal, and thymic leukocytes on lymphoma cells, an effect that was most marked in adult animals (Medina et al., 1998a). It stimulated the chemotactic capacity, indicated by the adherence of cells to a filter on the other side of which is placed chemoattractant, of peritoneal, axillary nodal, and splenic lymphocytes of mice, while it inhibited chemotaxis in thymic lymphocytes. In general, these effects decreased with age (Medina et al., 1998b). Mice lymphocytes from axillary nodes, thymus, and spleen responded with enhanced proliferation to co-incubation with GRP, an effect that decreased with age. On the other hand, the stimulatory response to exposure to the mitogen concanavalin A was reduced (Medina et al., 1999). GRP may exert its chemotactic influences on lymphocytes indirectly, through modulation of nitric oxide secretion by adherent accessory cells (Medina et al., 2005). Tumor-derived bombesin inhibited human dendritic cell maturation, as assessed by down-regulation of the expression of the co-stimulatory molecules CD40, CD80, and CD86. Furthermore, interleukin-12 production by dendritic cells was also diminished by purified bombesin. Tumor-derived bombesin inhibited T cell proliferation, as judged by diminished incorporation of tritiated thymidine, as well as endocytosis of lucifer yellow. Dendritic cell apoptosis was induced by tumor-derived bombesin. These immunosuppressive effects were partly abrogated by a specific bombesin antagonist. Human dendritic cells expressed GRP receptor mRNA (Makarenkova et al., 2003). Mouse peritoneal macrophages in suspension adhered to substrate in greater numbers upon incubation with GRP. Chemotaxis was also stimulated, an effect that

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  323

increased with age. Phagocytosis of latex beads was stimulated, although here there was no rise with age. Superoxide production was likewise enhanced, but this effect decreased with age (De la Fuente et al., 2000). In mouse lungs, mast cell infiltration was increased under the influence of intratracheal BLP (Subramaniam et al., 2003). Serotonin, like CGRP, has been found to be an eosinophil chemoattractant, an effect that could be partially blocked by application of 5-HT2A receptor antagonists (Boehme et al., 2004). Serotonin intensified proliferation of the leukemic Jurkat T cell line, as well as mitogen-activated human T cells, via the 5-HT1A receptor. Receptor mRNA was detected in these cells, and receptor activation resulted in second messenger pathway activation (Aune et al., 1993). Inhibition of serotonin synthesis also inhibited interleukin-2-stimulated human T cell proliferation, an effect that was reversed by addition of 5-hydroxytryptophan, serotonin’s precursor. 5-HT3 and 5-HT1A receptor antagonists also reversed it. In mice, contact sensitivity response (cellular immunity) to oxazalone, but not antibody production (humoral immunity), in vivo was inhibited by 5-HT1A receptor antagonists. In addition, immunologically stimulated murine spleen cells produced less interleukin-2 and interferon γ in the presence of 5-HT1A but not 5-HT1C/2 receptor antagonists (Aune et al., 1994). Serotonin upregulated mitogeninduced proliferation of rat and mouse splenic B cells, an effect that was reproduced by a specific 5-HT1A receptor agonist, and blocked by a 5-HT1A receptor antagonist. There were also specific binding sites for the agonist (Iken et al., 1995). Serotonin, added to peripheral T lymphocyte cultures obtained from HIV-seropositive subjects, decreased their intracellular levels of cyclic AMP and increased their proliferation rate. These effects were mimicked by a specific activator of the 5-HT1A receptor. In addition, expression of mRNA for interleukin-2 and interferon γ was increased (Eugen-Olsen et al., 1997). Spleen, thymus, and peripheral blood lymphocytes have active, mRNAproducing genes for serotonin receptors, including the 5-HT1B, 5-HT1F, 5-HT2A, 5-HT2B, 5-HT6, and 5-HT7 (i.e., G-protein-coupled) subtypes. Mitogen-activated lymphocytes additionally expressed the 5-HT3 receptor subtype (cation channelcoupled) mRNA (Stefulj et al., 2000). In cultures of rat spleen cells, serotonin, in relatively high concentrations, was found to inhibit the effect of several mitogens on lymphocyte proliferation (Stefulj et al., 2001). Serotonin stimulated human peripheral blood mononuclear cells to secrete lymphocyte chemoattractant factor or interleukin-16, selectively attractive for CD4+ T cells, and blocked by 5-HT2 receptor antagonists. Proof of the nature of the attractant factor was obtained by neutralization with antibodies against interleukin-16 (Laberge et al., 1996). Airway hyperresponsiveness, as expressed by airway resistance, was measured in ovalbumin-sensitized mice after intravenous administration of metacholine. This effect was significantly reduced in the presence of ketanserin, a specific 5-HT2 receptor

324  |  Airway Chemoreceptors in the Vertebrates antagonist. Ketanserin likewise reduced eosinophil infiltration, judged by the number of eosinophils in bronchoalveolar lavage fluid (De Bie et al., 1998). The airway epithelium as a whole displays immunological responses; it produces a range of mediators in response to deleterious agents from the inhaled air. Primary mediators are set free when the airway epithelium is stimulated by antigens, toxins, particulates, or irritants, whereupon it releases a host of secondary mediators (Martin et al., 1997, Stick and Holt, 2003). Is it possible that a significant share in this response, overlooked thus far, is from NEBs, which are outspoken endocrine entities? The human bronchial epithelial cell line BET-1A, when incubated with epithelial lining fluid obtained from individuals affected by cystic fibrosis, showed a marked increase of interleukin-8 (a chemoattractant of neutrophils) mRNA. The active agent was identified as neutrophil elastase. Since this is a neutrophil product, this points at a self-perpetuating inflammatory process (Nakamura et al., 1992). Ozone exposure of the human bronchial epithelial cell line BEAS-S6 led to synthesis and secretion of multiple eicosanoids, lipid mediators derived from arachidonic acid through lipoxygenase or cyclooxygenase pathways: thromboxane B2, prostaglandin E2, leukotrienes C4, D4, and E4, and 12-hydroxyheptadecatrienoic acid (McKinnon et al., 1993). Airway epithelial cells were found to secrete granulocyte/macrophage colony-stimulating factors, as well as a number of cytokines and chemokines: interleukin-6, interleukin11, interleukin-8, RANTES (Regulated on Activation, Normal T-cell Expressed and Secreted), monocyte chemoattractant protein-4, eotaxin, MIP 1α (Macrophage Inhibitory Protein), interleukin-16, and growth factors: tumor growth factors β and α. Arachidonic acid was converted in a host of lipid mediators of inflammation. The most important metabolite of the 15-lipoxygenase pathway was 15-hydroxyeicosatetraenoic acid, while those of the cyclooxygenase pathway were prostaglandins E2 and E2α. Another lipid mediator was platelet-activating factor (Polito and Proud, 1998). Human bronchial epithelial cells that were cultured with suspensions of diesel exhaust particles increased their production of interleukin-8, granulocyte/macrophage colony-stimulating factor, as well as interleukin-8 mRNA (Takizawa et al., 2000). Ambient particulate matter also stimulated production in a culture of human bronchial epithelial cells, through activation of the extracellular signal-regulated kinase and the p38 mitogen-activated protein kinase (Reibman et al., 2002). Human interleukin-17 in vitro stimulated the induction of mRNA and the release of granulocyte/macrophage colony-stimulating factor protein in the human bronchial epithelial cell line 16HBE. This release was potentiated by co-incubation with tumor necrosis factor α (Laan et al., 2003). A human bronchial epithelial cell line reacted to exposure to tumor necrosis factor α, interleukin-1β, interleukin-4, interleukin-13, as well as to ambient particulate matter, with induction of mRNA and secretion of MIP-3α/CCL20 (Macrophage Inhibitory Protein / Chemotactic Cytokine), which activates dendritic cells. These agents also upregulated mitogen-activated protein kinase pathways in the epithelial cells (Reibman et al., 2003).

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  325

We conclude that the picture that emerges from the reported observations and inferences is one of pulmonary NECs monitoring the inhaled air for antigens and responding to this with paracrine subepithelial secretion of immunomodulatory and chemoattractive substances. Thus, they may be at the origin of the activation of the immune system upon pulmonary infection, or they may participate in its initiation and modulation. At the moment, there is quite substantial experimental evidence for various well-known NEC secretory products to be involved in these mechanisms. The extent to which the NECs as such participate is elusive at this stage. The preceding discussion can only be the briefest outline of the potential contribution of the pulmonary NECs to local immune responses and defenses and their participation in the most diverse aspects of inflammation, allergy, and asthma. We still have very little relevant information regarding these aspects of NEC behavior. It is clear that here is a new, vast and largely uncharted area of future research.

Acknowledgements The authors thank Chris Armee and Rolande Renwart for expertly processing our materials for electron microscopy, as well as Vital Noppen for assistance with digital photography and imaging.

References Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat Rec., 270: 25-40. Aguayo, S.M., King, T.E., Waldron, J.A., Sherritt, K.M., Kane, M.A., and Miller, Y.E. 1990. Increased pulmonary neuroendocrine cells with bombesin-like immunoreactivity in adult patients with eosinophilic granuloma. J. Clin. Invest., 86: 838-844. Aguayo, S.M., Schuyler, W.E., Murtagh, J.J., and Roman, J. 1994. Regulation of lung branching morphogenesis by bombesin-like peptides and neutral endopeptidase. Am. J. Respir. Cell Mol. Biol., 10: 635-642. Aoki-Nagase, T., Nagase, T., Oh-Hashi, Y., Shindo, T., Kurihara, Y., Yamaguchi, Y., Yamamoto, H., Tomita, T., Ohga, E., Nagai, R., Kurihara, H., and Ouchi, Y. 2002. Attenuation of antigen-induced airway hyperresponsiveness in CGRP-deficient mice. Am. J. Physiol. Lung Cell Mol. Physiol., 283: L963-970. Aune, T.M., McGrath, K.M., Sarr, T., Bombara, M.P., and Kelley, K.A. 1993. Expression of 5HT1a receptors on activated human T cells. Regulation of cyclic AMP levels and T cell proliferation by 5-hydroxytryptamine. J. Immunol., 151: 1175-1183. Aune, T.M., Golden, H.W., and McGrath, K.M. 1994. Inhibitors of serotonin synthesis and antagonists of serotonin 1A receptors inhibit T lymphocyte function in vitro and cellmediated immunity in vivo. J. Immunol., 153: 489-498.

326  |  Airway Chemoreceptors in the Vertebrates Avadhanam, K.P., Plopper, C.G., and Pinkerton, K.E. 1997. Mapping the distribution of neuroepithelial bodies of the rat lung. A whole mount immunohistochemical approach. Am. J. Pathol., 150: 851-859. Barnes, P.J. 2001. Neurogenic inflammation in the airways. Respir. Physiol., 15: 145-154. Bellibas, S.E. 1996. The effect of human calcitonin gene-related peptide on eosinophil chemotaxis in the rat airway. Peptides, 17: 563-564. Boehme, S.A., Lio, F.M., Sikora, L., Pandit, T.S., Lavrador, K., Rao, S.P., and Sriramarao, P. 2004. Serotonin is a chemotactic factor for eosinophils and functions additively with eotaxin. J. Immunol., 173: 3599-3603. Boudard, F., and Bastide, M., 1991. Inhibition of mouse T-cell proliferation by CGRP and VIP: effects of these neuropeptides on IL-2 production and cAMP synthesis. J. Neurosci. Res., 29: 29-41. Bousbaa, H., and Fleury-Feith, J. 1991. Effects of a long-standing challenge on pulmonary neuroendocrine cells of actively sensitized guinea pigs. Am. Rev. Respir. Dis., 144: 714-717. Bousbaa, H., Poron, F., and Fleury-Feith, J. 1994. Changes in chromogranin A-immunoreactive guinea-pig pulmonary neuroendocrine cells after sensitization and challenge with ovalbumin. Cell Tissue Res., 275: 195-199. Brouns, I., Van Nassauw, L., Van Genechten, J., Majewski, M., Scheuermann, D.W., Timmermans, J.P., and Adriaensen, D. 2002. Triple immunofluorescence staining with antibodies raised in the same species to study the complex innervation pattern of intrapulmonary chemoreceptors. J. Histochem. Cytochem., 50: 575-582. Brouns, I., Van Genechten, J., Hayashi, H., Gajda, M., Gomi, T., Burnstock, G., Timmermans, J.P., and Adriaensen, D. 2003. Dual sensory innervation of pulmonary neuroepithelial bodies. Am. J. Respir. Cell Mol. Biol., 28: 275-285. Buckley, T.L., Brain, S.D., Collins, P.D., and Williams, T.J. 1991. Inflammatory edema induced by interactions between IL-1 and the neuropeptide calcitonin gene-related peptide. J. Immunol., 146: 3424-3430. Bulloch, K., McEwen, B.S., Nordberg, J., Diwa, A., and Baird, S. 1998. Selective regulation of T-cell development and function by calcitonin gene-related peptide in thymus and spleen. Ann. N.Y. Acad. Sci., 840: 551-562. Carucci, J.A., Ignatius, R., Wei, J., Cypess, A.M., Schaer, D.A., Pope, M., Steinman, R.M., and Mojsov, S. 2000. Calcitonin gene-related peptide decreases expression of HLA-DR and CD86 by human dendritic cells and dampens dendritic cell-driven T-cell-proliferative responses via the type 1 calcitonin gene-related peptide receptor. J. Immunol., 164: 34943499. Chanez, P., Springall, D., Vignola, A.M., Moradoghi-Hattvani, A., Polak, J.M., Godard, P., and Bousquet, J. 1998. Bronchial mucosal immunoreactivity of sensory neuropeptides in severe airway diseases. Am. J. Respir. Crit. Care Med., 158: 985-990. Cho, T., Chan, W., and Cutz, E. 1989. Distribution and frequency of neuro-epithelial bodies in post-natal rabbit lung: quantitative study with monoclonal antibody against serotonin. Cell Tissue Res., 255: 353-362. Cuesta, M.C., Quintero, L., Pons, H., and Suarez-Roca, H. 2002. Substance P and calcitonin gene-related peptide increase IL-1β, IL-6 and TNFα secretion from human peripheral blood mononuclear cells. Neurochem. Int., 40: 301-306. Cullen, A., Emanuel, R.L., Torday, J.S., Asokananthan, N., Sikorski, K.A., and Sunday, M.E. 2000. Bombesin-like peptide and receptors in lung injury models: diverse gene expression, similar function. Peptides, 21: 1627-1638.

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  327 Cutz, E., and Jackson, A. 1999. Neuroepithelial bodies as airway oxygen sensors. Respir. Physiol., 115: 201-214. Dakhama, A., Kanehiro, A., Mäkelä, M.J., Loader, J.E., Larsen, G.L., and Gelfand, E.W. 2002. Regulation of airway hyperresponsiveness by calcitonin gene-related peptide in allergen sensitized and challenged mice. Am. J. Respir. Crit. Care Med., 165: 1137-1144. De Bie, J.J., Henricks, P.A.J., Cruikshank, W.W., Hofman, G., Jonker, E.H., Nijkamp, F.P., Van and Oosterhout, A.J.M. 1998. Modulation of airway hyperresponsiveness and eosinophilia by selective histamine and 5-HT receptor antagonists in a mouse model of allergic asthma. Br. J. Pharmacol., 124: 857-864. De la Fuente, M., Medina, S., Del Rio, M., Ferrandez, M.D., and Hernanz, A. 2000. Effect of aging on the modulation of macrophage functions by neuropeptides. Life Sci., 67: 21252135. Dunzendorfer, S., Meierhofer, C., and Wiedermann, C.J. 1998. Signaling in neuropeptideinduced migration of human eosinophils. J. Leukoc. Biol., 64: 828-834. Emanuel, R.L., Torday, J.S., Mu, Q., Asokananthan, N., Sikorski, K.A., and Sunday, M.E. 1999. Bombesin-like peptides and receptors in normal fetal baboon lung: roles in lung growth and maturation. Am. J. Physiol., 277: L1003-L1017. Eugen-Olsen, J., Afzelius, P., Andresen, L., Iversen, J., Kronborg, G., Aabech, P., Nielsen, J.O., and Hofmann, B. 1997. Serotonin modulates immune function in T cells from HIV-seropositive subjects. Clin. Immunol. Immunopathol., 84: 115-121. Feng, Y., Tang, Y., Guo, J., and Wang, X. 1997. Inhibition of LPS-induced TNF-α production by calcitonin gene-related peptide (CGRP) in cultured mouse peritoneal macrophages. Life Sci., 61: PL 281-287. Fernandez, S., Knopf, M.A., and McGillis, J.P. 2000. Calcitonin-gene related peptide (CGRP) inhibits interleukin-7-induced pre-B cell colony formation. J. Leukoc. Biol., 67: 669-676. Fernandez, S., Knopf, M.A., Bjork, S.K., and McGillis, J.P. 2001. Bone marrow-derived macrophages express functional CGRP receptors and respond to CGRP by increasing transcription of c-fos and IL-6 mRNA. Cell. Immunol., 209: 140-148. Fox, F.E., Kubin, M., Cassin, M., Niu, Z., Hosoi, J., Torii, H., Granstein, R.D., Trinchieri, G., and Rook, A.H. 1997. Calcitonin gene-related peptide inhibits proliferation and antigen presentation by human peripheral blood mononuclear cells: effects on B7, interleukin 10, and interleukin 12. J. Invest. Dermatol., 108: 43-48. Hellings, P.W., Kasran, A., Liu, Z., Vandekerckhove, P., Wuyts, A., Overbergh, L., Mathieu, C., and Ceuppens, J.L. 2003. Interleukin-17 orchestrates the granulocyte influx into airways after allergen inhalation in a mouse model of allergic asthma. Am. J. Respir. Cell Mol. Biol., 28: 42-50. Hosoi, J., Murphy, G.F., Egan, C.L., Lerner, E.A., Grabbe, S., Asahina, A., and Granstein, R.D. 1993. Regulation of Langerhans cell function by nerves containing calcitonin gene-related peptide. Nature, 363: 159-163. Iken, K., Cheng, S., Fargin, A., Goulet, A.C., and Kouassi, E. 1995. Serotonin upregulates mitogen-stimulated B lymphocyte proliferation through 5-HT1A receptors. Cell. Immunol., 163: 1-9. Kazemian, P., Stephenson, R., Yeger, H., and Cutz, E. 2001. Respiratory control in neonatal mice with NADPH oxidase deficiency. Respir. Physiol., 126: 89-101. Keith, I.M., Pelto-Huikko, M., Schalling, M., and Hökfelt, T. 1991. Calcitonin gene-related peptide and its mRNA in pulmonary neudoendocrine cells and ganglia. Histochemistry, 96: 311-315.

328  |  Airway Chemoreceptors in the Vertebrates Kemp, P.J., Lewis, A., Hartness, M.E., Searle, G.J., Miller, P., O’Kelly, I., and Peers, C. 2002. Airway chemotransduction. From oxygen sensor to cellular effector. Am. J. Respir. Crit. Care Med., 166: 517-524. Kemp, P.J., Searle, G.J., Hartness, M.E., Lewis, A., Miller, P., Williams, S., Wootton, P., Adriaensen, D., and Peers, C. 2003. Acute oxygen sensing in cellular models: relevance to the physiology of pulmonary neuroepithelial and carotid bodies. Anat. Rec., 270A: 41-50. King, K.A., Torday, J.S., and Sunday, M.E. 1995. Bombesin and (Leu8)phyllolitorin promote fetal mouse lung branching morphogenesis via a receptor-mediated mechanism. Proc. Natl. Acad. Sci. USA, 92: 4357-4361. Kraehlenbuhl, J.P., and Neutra, M.R. 2000. Epithelial M cells: differentiation and function. Annu. Rev. Cell Dev. Biol., 16: 301-332. Kresch, M., Christian, C., Zhu, L., Obe, M., Sanders, M.M., and Hussain, N. 1999. Bombesin inhibits apoptosis in developing fetal rat lung. Lung, 177: 241-251. Laan, M., Prause, O., Miyamoto, M., Sjöstrand, M., Hytönen, A.M., Kaneko, T., Lötvall, J., and Lindén, A. 2003. A role of GM-CSF in the accumulation of neutrophils in the airways caused by IL-17 and TNF-α. Eur. Respir. J., 21: 387-393. Laberge, S., Cruikshank, W.W., Beer, D.J., and Center, D.M. 1996. Secretion of IL-16 (lymphocyte chemoattractant factor) from serotonin-stimulated CD8+ T cells in vitro. J. Immunol., 156: 310-315. Levite, M. 1998. Neuropeptides, by direct interaction with T cells, induce cytokine secretion and break the commitment to a distinct T helper phenotype. Proc. Natl. Acad. Sci. USA, 95: 12544-12549. Levite, M., Cahalon, L., Hershkoviz, R., Steinman, L., and Lider, O. 1998. Neuropeptides, via specific receptors, regulate T cell adhesion to fibronectin. J. Immunol., 160: 993-1000. Liu, J., Chen, M., and Wang, X. 2000. Calcitonin gene-related peptide inhibits lipopolysaccharideinduced interleukin-12 release from mouse peritoneal macrophages, mediated by the cAMP pathway. Immunology, 101: 61-67. Luts, A., Uddman, R., Hakanson, R., and Sundler, F. 1994. Calcitonin, CGRP and helodermin in endocrine cells of the developing rat lung. Regul. Pept., 51: 121-129. Makarenkova, V.P., Shurin, G.V., Tourkova, I.L., Balkir, L., Pirtskhalaishvili, G., Perez, L., Gerein, V., Siegfried, J.M., and Shurin, M.R. 2003. Lung cancer-derived bombesin-like peptides down-regulate the generation and function of human dendritic cells. J. Neuroimmunol., 145: 55-67. Mapp, C.E., Lucchini, R.E., Miotto, D., Chitano, P., Jovine, L., Saetta, M., Maestrelli, P., Springall, D.R., Polak J., and Fabbri, L.M. 1998. Immunization and challenge with toluene diisocyanate decrease tachykinin and calcitonin gene-related peptide immunoreactivity in guinea pig central airways. Am. J. Respir. Crit. Care Med., 158: 263-269. Marchevsky, A.M., Keller, S., Fogel, J.R., and Kleinerman, J. 1984. Quantitative studies of argyrophilic APUD cells in airways. III. The effects of sensitization and anaphylactic shock. Am. Rev. Respir. Dis., 129: 477-480. Martin, L.D., Rochelle, L.G., Fischer, B.M., Krunkosky, T.M., and Adler, K.B. 1997. Airway epithelium as an effector of inflammation: molecular regulation of secondary mediators. Eur. Respir. J., 10: 2139-2146. McDowell, E.M., Hoyt, R.F., and Sorokin, S. 1994. Ontogeny of endocrine cells in the respiratory system of Syrian golden hamsters. II. Intrapulmonary airways and alveoli. Cell Tissue Res., 275: 157-167.

Pulmonary Neuroepithelial Bodies as Hypothetical Immunomodulators  |  329 McGillis, J.P., Humphreys, S., Rangnekar, V., and Cialella, J. 1993. Modulation of B lymphocyte differentiation by calcitonin gene-related peptide (CGRP). Cell. Immunol., 150: 405-416. McKinnon, K.P., Madden, M.C., Noah, T.L., and Devlin, R.B. 1993. In vitro ozone exposure increases release of arachidonic acid products from a human bronchial epithelial cell line. Toxicol. Appl. Pharmacol., 118: 215-223. Medina, S., Del Rio, M., Ferrandez, M.D., Hernanz, A., and De la Fuente, M. 1998a. Changes with age in the modulation of natural killer activity of murine leukocytes by gastrin-releasing peptide, neuropeptide Y and sulfated cholecystokinin octapeptide. Neuropeptides, 32: 549555. Medina, S., Del Rio, M., Victor, V.M., Hernanz, A., and De la Fuente, M. 1998b. Changes with ageing in the modulation of murine lymphocyte chemotaxis by CCK-8S, GRP, and NPY. Mech. Ageing Dev., 102: 249-261. Medina, S., Del Rio, M., De la Cuadra, B., Guayerbas, N., and De la Fuente, M. 1999. Age-related changes in the modulatory action of gastrin-releasing peptide, neuropeptide Y and sulfated cholecystokinin octapeptide in the proliferation of murine lymphocytes. Neuropeptides, 33: 173-179. Medina, S., Del Rio, M., Hernanz, A., Guaza, C., and De la Fuente, M. 2005. Nitric oxide released by accessory cells mediates the gastrin-releasing peptide effect on murine lymphocyte chemotaxis. Regul. Pept., 131: 46-53. Nakamura, H., Yoshimura, K., McElvaney, N.G., and Crystal, R.G. 1992. Neutrophil elastase in respiratory epithelial lining fluid of individuals with cystic fibrosis induces interleukin-8 gene expression in a human bronchial epithelial cell line. J. Clin. Invest., 89: 1478-1484. Nong, Y.-H., Titus, R.G., Ribeiro, J.M.C., and Remold, H.G. 1989. Peptides encoded by the calcitonin gene inhibit macrophage function. J. Immunol., 143: 45-49. Numao, T., and Agrawal, D.K. 1992. Neuropeptides modulate human eosinophil chemotaxis. J. Immunol., 149: 3309-3315. Pan, J., Yeger, H., and Cutz, E. 2002. Neuronal developmental marker FORSE-1 identifies a putative progenitor of the pulmonary neuroendocrine cell lineage during lung development. J. Histochem. Cytochem., 50: 1567-1578. Pan, J., Yeger, H., and Cutz, E. 2004. Innervation of pulmonary neuroendocrine cells and neuroepithelial bodies in developing rabbit lung. J. Histochem. Cytochem., 52: 379-389. Polito, A.J., and Proud, D. 1998. Epithelial cells as regulators of airway inflammation. J. Allergy Clin. Immunol., 102: 714-718. Raud, J., Lundeberg, T., Brodda-Jansen, G., Theodorsson, E., and Hedqvist, P. 1991. Potent anti-inflammatory action of calcitonin gene-related peptide. Biochem. Biophys. Res. Commun., 180: 1429-1435. Reibman, J., Hsu, Y., Chen, L.C., Kumar, A., Su, W.C., Choy, W., Talbot, A., and Gordon, T. 2002. Size fractions of ambient air particulate matter induce granulocyte macrophage colony-stimulating factor in human bronchial epithelial cells by mitogen-activated protein kinase pathways. Am. J. Respir. Cell Mol. Biol., 27: 455-462. Reibman J., Hsu, Y., Chen, L.C., Bleck, B., and Gordon, T. 2003. Airway epithelial cells release MIP-3α/CCL20 in response to cytokines and ambient particulate matter. Am. J. Respir. Cell Mol. Biol., 28: 648-654. Reynolds, S.D., Giangreco, A., Power, J.H.T., and Stripp, B.R. 2000. Neuroepithelial bodies of pulmonary airways serve as a reservoir of progenitor cells capable of epithelial regeneration. Am. J. Pathol., 156: 269-278.

330  |  Airway Chemoreceptors in the Vertebrates Stahlman, M.T., and Gray, M.E. 1997. Immunogold EM localization of neurochemicals in human pulmonary neuroendocrine cells. Microsc. Res. Tech., 37: 77-91. Stefulj, J., Jernej, B., Cicin-Sain, L., Rinner, I., and Schauenstein, K. 2000. mRNA expression of serotonin receptors in cells of the immune tissues of the rat. Brain Behav. Immun., 14: 219-224. Stefulj, J., Cicin-Sain, L., Schauenstein, K., and Jernej, B. 2001. Serotonin and immune response: effect of the amine on in vitro proliferation of rat lymphocytes. Neuroimmunomodulation, 9: 103-108. Stick, S.M., and Holt, P.G. 2003. The airway epithelium as immune modulator. The LARC ascending. Am. J. Respir. Cell Mol. Biol., 28: 641-644. Subramaniam, M., Sugiyama, K., Coy, D.H., Kong, Y., Miller, Y.E., Weller, P.F., Wada, K., Wada, E., and Sunday, M.E. 2003. Bombesin-like peptides and mast cell responses. Relevance to bronchopulmonary dysplasia? Am. J. Respir. Crit. Care. Med., 168: 601-611. Takizawa, H., Ohtoshi, T., Kawasaki, S., Abe, S., Sugawara, I., Nakahara, K., Matsushima, K., and Kudoh, S. 2000. Diesel exhaust particles activate human bronchial epithelial cells to express inflammatory mediators in the airways: a review. Respirology, 5: 197-203. Tang, Y., Feng, Y., and Wang, X. 1998. Calcitonin gene-related peptide potentiates LPS-induced IL-6 release from mouse peritoneal macrophages. J. Neuroimmunol., 84: 207-212. Torii, H., Hosoi, J., Beissert, S., Xu, S., Fox, F.E., Asahina, A., Takashima, A., Rook, A.H., and Granstein, R.D. 1997. Regulation of cytokine expression in macrophages and the Langerhans cell-like line XS52 by calcitonin gene-related peptide. J. Leukoc. Biol., 61: 216-223. Tsukiji, J., Sango, K., Udaka, N., Kageyama, H., Ito, T., Saito, H., Horie, H., Inoue, S., Kitamura, H., Hagiwara, E., Ikeda, H., Okubo, T., and Ishigatsubo, Y. 2004. Long-term induction of β−CGRP mRNA in rat lung by allergic inflammation. Life Sci., 76: 163-177. Van Lommel, A., van den Steen, P., and Lauweryns, J.M. 1995. Association of immune cells with neuroepithelial bodies in the lungs of neonatal dogs, cats and hamsters. Cell Tissue Res., 282: 519-522. Veronesi, B., Carter, J.D., Devlin, R.B., Simon, S.A., and Oortgiesen, M. 1999. Neuropeptides and capsaicin stimulate the release of inflammatory cytokines in a human bronchial epithelial cell line. Neuropeptides, 33: 447-456. Wang, D., Yeger, H., and Cutz, E. 1996. Expression of gastrin-releasing peptide receptor gene in developing lung. Am. J. Respir. Cell Mol. Biol., 14: 409-416. Wang, F., Millet, I., Bottomly, K., and Vignery, A. 1992. Calcitonin gene-related peptide inhibits interleukin 2 production by murine T lymphocytes. J. Biol. Chem., 267: 21052-21057. Weichselbaum, M., Sparrow, M.P., Hamilton, E.J., Thompson, P.J., and Knight, D.A. 2005. A confocal microscopic study of solitary pulmonary neuroendocrine cells in human airway epithelium. Resp. Res., 6: 115-126. Yaraee, R., Ebtekar, M., Ahmadiani, A., and Sabahi, F. 2003. Neuropeptides (SP and CGRP) augment pro-inflammatory cytokine production in HSV-infected macrophages. Int. Immunopharmacol., 3: 1883-1887. Yaraee, R., Ebtekar, M., Ahmadiani, A., and Sabahi, F. 2005. Effect of neuropeptides (SP and CGRP) on antigen presentation by macrophages. Immunopharmacol. Immunotoxicol., 27: 395-404.

15 Neuroepithelial Bodies and Carotid Bodies: A Comparative Discussion of Pulmonary and Arterial Chemoreceptors Alfons T.L. Van Lommel

Abstract This review is a comparative discussion of the most important aspects of two oxygen-sensitive chemoreceptor types: the arterial carotid bodies and the pulmonary neuroepithelial bodies. The two types are contrasted with regard to physiology and histochemistry, nerve terminals and synaptic junctions, neurotransmitters and membrane receptors, oxygen sensors and membrane channels, and embryology and development. Because the carotid bodies are compact, isolated organs, while the neuroepithelial bodies are diffuse and deeply integrated in the lungs, different experimental approaches have had to be applied to each of them. Nevertheless, it has become clear that the two function in very similar ways, in spite of their very different embryological background, the apparently different nature of their membrane oxygen sensors, and their orientation towards different body compartments. Just about the only functional aspect in which they fundamentally differ would appear to be the time course of their development: neuroepithelial bodies seem to develop in advance of the carotid bodies. Further progress in our understanding of both carotid and neuroepithelial bodies might greatly profit from a comparative experimental approach.

Keywords: Carotid body, chemoreceptors, embryology, hypoxia, membrane channels, membrane receptors, membrane sensors, nerve endings, neuroepithelial bodies, neurotransmitters, synapses, oxygen

Leuven Catholic University, Medical Faculty, Department of Morphology and Molecular Pathology, Minderbroedersstraat 12, 3000 Leuven, Belgium. Email: [email protected]

332  |  Airway Chemoreceptors in the Vertebrates

Introduction Oxygen may surely be regarded as being the most vital of the vital substances the body needs. To closely monitor changes in oxygen levels, or the corresponding carbon dioxide and pH levels, evolution has come up with several specialized types of receptor organs. Most of them monitor blood gas levels. This group includes arterial, lung vascular, and brain stem respiratory center receptors. The arterial chemoreceptors, i.e., the carotid and aortic bodies, consist of densely vascularized clusters of cells with an endocrine phenotype, closely associated with capillaries derived from the body’s main arteries, the carotid artery and aorta, respectively. They are sensitive to changes in oxygen tension, and corresponding changes in carbon dioxide tension and pH, in the arterial blood. Glossopharyngeal and vagal afferent nerve fibers connect them to the brain stem respiratory centers. The myocytes of the smaller arterial pulmonary vessels are directly sensitive to blood oxygen shortage (for review see Archer et al., 2000) and contribute to ventilation-perfusion matching in the lungs. The brain stem neurons of the respiratory centers monitor carbon dioxide and corresponding pH levels (for review see Bruce and Cherniak, 1987). The last type of receptor is the neuroepithelial bodies (NEBs) of the lungs. Neuroepithelial bodies, like the arterial chemoreceptors, are clusters of endocrine cells, connected to the brain stem via vagal sensory nerve fibers. The most profound difference with the other receptor types is that NEBs are not intimately associated with blood vessels. They are an integral part of the airway epithelia and apparently receive their stimuli from the inhaled air. The object of this review is a comparison between the pulmonary NEBs and the arterial chemoreceptors. There is a very close similarity between the two in several respects, not least with regard to cell physiology and ultrastructure. The basic sensory unit of both is the endocrine cell, containing dense-cored secretory vesicles and associated with sensory nerve fibers. This structure allows for transduction of chemical stimuli into nervous impulses that travel towards the central nervous system. Comparing the morphology, physiology, cell biology, and other aspects of the pulmonary and arterial chemoreceptors may prove instructive in both directions. Researchers interested in NEBs may learn from insights gained in carotid body (CB) research, and vice versa. Aortic bodies have received less attention than CBs, but it may be assumed that what applies to the CBs also applies to the aortic bodies. No mention will be made of the distinction between solitary pulmonary neuroepithelial cells and corpuscular NEBs. Since it is a recently established fact that the latter are innervated, the last great divide between the two has closed. Thus, we will focus on the term NEB. The history of NEB research has been determined by the fact that NEBs, in contrast to CBs, are a diffuse organ, deeply integrated in a larger organ, the lung. Even their recognition initially met with some resistance (the author has heard them called Non-Existent Bodies, but admittedly that was long ago) since, without specific staining methods, their diffuse nature makes them rather inconspicuous, and it takes a keen

Neuroepithelial Bodies and Carotid Bodies  |  333

observer to make them out at all. The compact, isolated nature of the CBs, with their largely separate vascularization and innervation, allowed straightforward experimental approaches that were completely unthinkable with respect to NEBs. In the end, largely similar conclusions regarding the function and significance of NEBs and CBs have been reached, but they were obtained via widely divergent routes.

Physiology and Histochemistry De Castro (1926) was one of the first researchers who subjected the CBs to a detailed histological examination. As a result, he was able to demonstrate the close association of glomus cell clusters with a dense network of capillaries and nerve fibers, and degeneration of glomus cell-associated nerve fibers upon section of the sinus branch of the glossopharyngeal nerve. This obviously suggested that CBs functioned as chemoreceptors, monitoring changes in blood gas concentration. Conclusive physiological proof that this was the case was not long in coming. It was provided by C. Heymans, a compatriot of 1938 Nobel Prize fame, in the 1920s and 1930s (for a summary and references see Heymans, 1955). When the blood gas pressures in the carotid artery were experimentally changed, it was found that this induced specific respiratory reflexes, and the afferent sinus nerve was found to display enhanced electric activity. Alternatively, section of the sinus nerve prevented the induction of respiratory reflexes by changes in blood gas values. In the early 1970s, when the first histological, histochemical, and electron microscopic investigations on the newly named pulmonary NEBs were performed (Lauweryns et al., 1972), it was immediately pointed out that there was a close ultrastructural resemblance between the CB glomus cells on the one hand and the NEB corpuscular cells on the other hand. This profound morphological similarity, added to the obvious fact that NEBs are an integral part of the airway epithelium, in close contact with the inhaled air, and thus ideally situated to detect changes in intraluminal oxygen levels, prompted their interpretation as hypoxia-sensitive chemoreceptors, similar to de Castro’s interpretation of the CB glomus cells, but there was no way of obtaining physiological proof of the kind mentioned above. For want of an alternative, the morphological and histochemical approaches would dominate the study of NEBs for many years. The first independent, indisputable functional proof of the chemoreceptor nature of NEB would be forthcoming only in the early 1990s, a full 20 years later, by the application of modern cell physiological techniques (Youngson et al., 1993). Meanwhile, and in spite of their limitations, the morphological and histochemical approaches led to a fairly deep insight in the nature and function of the pulmonary NEBs. Exposure to ventilation hypoxia, in combination with quantitative detection methods, showed that NEBs reacted to oxygen shortage (Lauweryns et al., 1977). The relative number of exocytotic profiles at their abluminal pole increased, while their

334  |  Airway Chemoreceptors in the Vertebrates cellular serotonin content decreased. While serotonin levels decreased, the argyrophilic NEB frequencies did not change (Keith et al., 1981). Obviously, this showed that NEBs responded to acute changes in oxygen levels by endocrine secretion of bioactive substances. Serotonin is well known as a vasoactive molecule, and immediately the NEBs were linked to hypoxia-induced vasoconstriction, a well-known lung physiological strategy to divert pulmonary blood from inadequately ventilated lung regions. Alternatively, serotonin might function as an afferent neurotransmitter, inducing cardiovascular and respiratory reflexes via the central nervous system. In later years, the effect of hypoxia on another well-established NEB secretory product, calcitonin gene-related peptide (CGRP), was investigated in a similar way. In sharp contrast with serotonin, CGRP storage was increased in hypoxia (Montuenga et al., 1992, Roncalli et al., 1993), and CGRP depletion of NEBs exacerbated hypoxiainduced pulmonary hypertension (Tjen-A-Looi et al., 1998). CGRP, in contrast to serotonin, is a pulmonary vasodilator. It may be that CGRP is chronically released in normoxia, and that hypoxia inhibits its release. This would entail vasodilatation in normoxia, and vasoconstriction in hypoxia. Thus, the antagonistic effects of serotonin and CGRP produce the same physiological end result. Since the lung is not only ventilated with air, but also intensely perfused with blood, as soon as the first evidence of the chemoreceptor nature of NEB was obtained, the point that the hypoxic stimulus could also be supplied by the pulmonary blood forced itself on the attention of the researchers. Evidence that NEBs are stimulated from the inhaled air initially rested solely on the results of rather complicated and highly invasive cross-perfusion studies (Lauweryns et al., 1978). Ventilation with normoxic air, in combination with perfusion of the lungs with hypoxemic blood, did not result in an endocrine secretory response of the NEBs. Later, and somewhat less invasively, the secretory response of NEBs in unilaterally ventilated lungs pointed to the same conclusion (Lauweryns et al., 1983). By ventilating one lung with pure oxygen, normal blood gas values could be combined with ventilatory hypoxia in the other lung. In this lung, the NEBs were found to show an endocrine response. A few comparable studies have been performed on the CBs (Blümcke et al., 1967; Mills and Slotkin, 1975). Unlike in NEB research, comparatively little weight was attached to them, since they were not vital to decide on the CB chemoreceptor nature. Hypoxia-induced serotonin release was confirmed more recently by the use of advanced cell physiological methods, and its mechanism was outlined in much more detail. Serotonin release, induced by administration of hypoxic superfusate, was measured amperometrically in intact rabbit NEBs in lung slices. The K+ channel blockers tetraethylammonium and 4-amino peridine induced serotonin secretion under normoxia, mimicking the effects of hypoxia (see the chapter on oxygen sensors and membrane channels). Serotonin secretion was abolished by medium free of Cd2+ (competes for entry with extracellular Ca2+) and Ca2+. Hypoxia-induced serotonin release was also suppressed by nifedipin, a specific L-type Ca2+ channel blocker.

Neuroepithelial Bodies and Carotid Bodies  |  335

Combination of hypoxia with a specific 5-HT3 receptor blocker inhibited serotonin release, indicating an autocrine regulatory effect (Fu et al., 2002). Using similar methods, essentially similar results were obtained for cathecholamine secretion in the CBs (Pardal et al., 2000).

Nerve Terminals and Synaptic Junctions It is self-evident that a receptor organ must be equipped with sensory nerves, and a lot of attention was consequently paid to the innervation of CBs and NEBs. The ultrastructural morphology of intraglomerular CB (Hansen, 1985; Pallot and Blakeman, 1982; Matsumoto et al., 1980; Morgan et al., 1975) and intracorpuscular NEB (Hung and Loosli, 1974; Wasano and Yamamoto, 1981; Lauweryns and Van Lommel, 1987; Van Lommel and Lauweryns, 1993a, 1993b) nerve endings was found to be highly typical, and very similar (Figures 1, 2). Large volume nerve endings predominantly filled with mitochondria (i.e., morphologically afferent) contrasted with smaller volume nerve endings loaded with synaptic vesicles (i.e., morphologically efferent). These types of terminals closely resembled the sensory and motor nerve endings of undisputed sense or effector organs, respectively. In addition, the apposed sensory cell and nerve terminal membranes also showed typical synaptic junctions. In the CBs, synapses with afferent as well as efferent polarity, as judged from the asymmetrical distribution of dense membrane coats and synaptic vesicles, could be distinguished. In combination, these afferent and efferent synapses constituted what has been called reciprocal synapses. In NEBs, only the afferent synaptic component has been convincingly observed, although efferent nerve terminals are present. The presence of afferent nerve endings and synapses was logical enough, but the participation of the efferent ones posed a problem that needed explanation. First of all: what was the relationship between the two? Were there separate motor and sensory terminal varicosities? Or were the endings continuous, forming in passing varicosities on the same fibers? Part of the difficulty lay in the loss of the third dimension in a histological section. Continuities between nerve terminals could only be made out when they happened to coincide with the section plane. Attempts have been made to resolve this issue by laborious spatial reconstructions based on serial sections. In the NEBs, unequivocal evidence did not surface for a substantial participation of an anatomically separate motor innervation (the originally postulated type 2 endings) (Lauweryns and Van Lommel, 1987; Van Lommel and Lauweryns, 1993b), and neither did it in the CBs (Kondo, 1976a; Biscoe and Pallot, 1972). Evidence thus favors the second alternative: varicosities in passing. A way out of the dilemma of receptors apparently equipped with motor nerves, or sensory nerves being presynaptic, has been offered by the axon reflex phenomenon. In a neuroanatomical sense, NEB corpuscular cell and CB glomus cell innervation was

336  |  Airway Chemoreceptors in the Vertebrates found to be sensory, since it derives from sensory ganglion neurons that send their central axons towards the central nervous system. However, depolarization of afferent nerve terminals not only induces afferent discharges that are directed towards the central nervous system. There is also antidromic invasion of efferent nerve endings by the generator potential, which may activate them in their turn. Thus, the confluence of afferent and efferent nerve endings, and the reciprocal synapses, allow axon reflexes: local peripheral regulatory mechanisms. They may strengthen or repress the autocrine inhibition attributed to a number of CB and NEB neurotransmitters. In the CBs,

Figure 1: An intracorpuscular neuroepithelial body nerve fiber showing an afferent terminal, loaded with mitochondria (MIT), in continuity with an efferent terminal with clear synaptic vesicles (SV). The surrounding corpuscular cell cytoplasm contains dense-cored vesicles (DCV). The distinct aspect of both nerve endings is probably due in large measure to a fortuitous orientation of the section plane. Nevertheless, it indicates that NEB intracorpuscular nerve endings display regions where they are post-synaptic (POST) and others where they are pre-synaptic (PRE) to the corpuscular cells, allowing for both afferent signal transduction by the corpuscular cells and efferent regulatory nervous activity. Hamster lung, TEM, 20.000 x.

Neuroepithelial Bodies and Carotid Bodies  |  337

Figure 2: An intraglomerular carotid body nerve ending, loaded with mitochondria (MIT), and with local accumulations of synaptic vesicles (SV). The adjacent membranes show synaptic densities. In the upper part of the picture, the glomus cell membrane seems to be irregularly thickened, while dense-cored vesicles (DCV) converge towards it. Thus, the nerve ending is post-synaptic to the glomus cell (POST). In the lower part of the picture, it is the neural cell membrane that appears to be irregularly thickened. Here, as well as in places where synaptic vesicles accumulate, the nerve ending is pre-synaptic to the glomus cell (PRE). Such reciprocal synapses allow for axon reflexes, i.e., afferent signal transduction by the glomus cells, accompanied by local efferent nervous regulation. Rat carotid body, TEM, 20.000 x.

the resolution of this dilemma was particularly timely, since it provided an elegant explanation of the phenomenon of efferent inhibition without having to invoke the participation of a separate motor component in the innervation of the glomus cells. Efferent inhibition was a consequence of electric stimulation of the carotid sinus nerve, and it decreased glomus cell chemoreceptor activity. It could henceforward be seen as antidromic conduction of regulatory signals by sensory nerves (McDonald and Mitchell, 1981).

338  |  Airway Chemoreceptors in the Vertebrates Selective denervation experiments provided neuroanatomical proof of the sensory nature of the NEB and CB innervation and gave additional, indirect evidence that there was no separate motor innervation to speak of: there was hardly any preferential survival of efferent nerve endings. In spite of an initial controversy, the majority of the nerve fibers innervating glomus cells were found to be located in the sensory glossopharyngeal petrosal ganglia (Hess and Zapata, 1972; Nishi and Stensaas, 1974; McDonald and Mitchell, 1975, 1981). Those innervating the intracorpuscular cells were located in the vagal nodose ganglia (Lauweryns et al., 1985; Van Lommel and Lauweryns, 1993a). In the CBs, intraglomerular nerve terminal degeneration after peripheral nerve section was very complete, more than 90% after several days. Postdenervation survival of NEB intracorpuscular nerve endings (up to 30% after 5 d) was not negligible and was explained primarily by survival of pulmonary vagal fibers crossing over from the intact contralateral vagus, and additionally by the contribution of intrinsic pulmonary vagal neurons and non-vagal, e.g., spinal, fibers. Again, we encounter the old problem referred to in the introduction: while the CBs are isolated structures, the NEBs are integrated in the lungs. The CB glomus cells are supplied virtually exclusively by the minute sinus branch of the glossopharyngeus, which is largely sensory. On the other hand, the NEBs receive a small fraction of the sensory fibers of the comparatively massive pulmonary vagus, which also contains autonomic motor fibers. As already hinted at, the situation is even further complicated by bilateral crossing over of pulmonary vagal fibers, and the contribution of non-vagal, i.e., mixed spinal nerves and autonomic motor nerves. In the NEBs, denervation experiments have been backed up with anterograde tracing studies involving fluorescent labels, which confirmed that NEBs are associated with sensory vagal nodose fibers. In fact, and this is rather important, most of the sensory vagal fibers to the airway epithelium proved to contact NEBs. Proof of the NEB nature of the end organ was provided by electron microscopy of re-embedded tissue containing a traced nerve fiber (Van Lommel et al., 1998) or by combining anterograde tracing with immunohistochemistry for CGRP (Adriaensen et al., 1998). In the CBs, glomus-cell associated nerve endings were similarly anterogradely labeled after injection of tritiated leucine in the petrosal ganglion and observed with electron microscopy (Fidone et al., 1975, 1977; Smith and Mills, 1976). Significantly, there was universal labeling of nerve endings, including small and vesicle-filled endings. Here again, this finding argued against a separate motor innervation of the glomus cells and thus in favor of the existence of reciprocal synaptic activity. This rather clear-cut picture has not entirely stood the test of time, however. Since the 1990s, attention has been drawn to the presence of motor neurons in several locations along the carotid sinus and glossopharyngeal nerves, and in the petrosal ganglion. Moreover, there is evidence that these neurons participate in efferent inhibition (Campanucci and Nurse, 2007; Campanucci et al., 2006). In comparable fashion, intrinsic motor nerves appear to be involved in NEB innervation (Adriaensen et al., 2003).

Neuroepithelial Bodies and Carotid Bodies  |  339

Neurotransmitters and Membrane Receptors If the CB glomus and NEB corpuscular cells are to convey stimuli to the sensory nerve endings they are associated with, they must make use of neurotransmitters, and the nerve endings must carry the appropriate membrane receptors. In addition, if sensory nerve stimulation is accompanied by regulatory antidromic motor activity in the sensory nerve terminals, they also must contain neurotransmitter substances with which to act on the glomus or corpuscular cells. The very fact that neurotransmitter substances have been located in these sensory nerve endings strengthens the concept of an efferent or antidromic regulatory neural mechanism. To make the picture complete, the glomus and corpuscular cells must carry membrane receptors tuned to the transmitters stored in the nerve endings. Thus, a picture emerges of rather complicated stimulatory and inhibitory circuits in the chemosensory cell-nerve ending unit. Actually, the situation is even further complicated by the intervention of autocrine regulatory mechanisms: glomus cells as well as corpuscular cells carry receptors for some of their own neurotransmitters. Many substances have been found in NEBs and CBs that are or could be secreted and that could act as transducing substances. It is a largely unexplained fact that there are important species differences in this respect: different species may store different substances, and the same substance may have antagonistic effects. The requirements for a neurotransmitter include the following: a synthesizing enzyme or active uptake mechanism must be present, the substance must be stored in the cytosol or in secretory vesicles, it must be released after proper stimuli, it must produce a specific post-synaptic effect, complementary post-synaptic membrane receptors must be present, and its actions must be mimicked by agonists and blocked by antagonists. Several substances meet a number or most of these requirements and thus may be regarded as neurotransmitters that convey stimuli from the glomus or corpuscular cells to the associated nerves and vice versa. Let us first consider the transmitter substances used by the sensory cells, and the membrane receptors present on them. In the CBs, one of the first substances for which definite proof, beyond mere presence, of a neurotransmitter function was obtained was acetylcholine. Again, proof of this was greatly facilitated by the isolated nature of the CBs, which allowed Loewitype perfusion experiments (Eyzaguirre et al., 1965). Of the amines, dopamine has been most frequently encountered in CB glomus cells, and it too was an early candidate for an afferent neurotransmitter substance in the CBs. Both are very likely co-released (Monti-Bloch and Eyzaguirre, 1980; Bairam and Marchal, 2003). Both may function as excitatory or inhibitory transmitters, depending on the species, and may also be implicated in autocrine actions on the glomus cells themselves. Acetylcholine was also found to be co-released with ATP upon stimulation of glomus cells and both excited afferent nerve fibers in co-cultures with petrosal ganglion neurons (Zhang et al., 2000),

340  |  Airway Chemoreceptors in the Vertebrates or in vitro (Varas et al., 2003). ATP also turned out to be an autocrine transmitter with an inhibitory effect on glomus cells (Varas et al., 2003; Buttigieg and Nurse, 2004; Xu et al., 2005). At present, the peptide neurotransmitter substance that has been demonstrated most often in the glomus cell cytoplasm is substance P (Wang et al., 1996; Kim et al., 2001). M2 muscarinic receptors were demonstrated on the glomus cells (Shirahata et al., 2004), reinforcing the theory that there exists autocrine cholinergic regulation of glomus cell activity. Dopamine D2 autoreceptors were also found to be present on glomus cells, and their stimulation inhibited dopamine release (Carroll et al., 2005). Dopamine D-1 receptors occurred in appreciably higher concentrations than D-2 receptors (Bairam et al., 1998). In accord with the observed autocrine function of ATP, the P2X2 receptor was located in the glomus cell membrane (He et al., 2006). In the NEBs, serotonin (Lauweryns et al., 1982; Luts et al., 1991) and CGRP (Lauweryns and Van Ranst, 1987; Keith et al., 1991; Luts et al., 1991) are the amine and peptide substances, respectively, that have most frequently been encountered in the corpuscular cell cytoplasm. NEB corpuscular cells were found to display P2X2 and P2X3 purinergic receptor subunits in hamster, where ATP induced depolarization and serotonin release, which was blocked by suramin (Fu et al., 2004). NEB corpuscular cells in hamsters also expressed mRNAs of the β2 subunit of nicotinic acetylcholine receptors and were immunoreactive to the α4, α7, and β2 subunits. Nicotine administration evoked NEB depolarization (Fu et al., 2003). Type 3 serotonin receptor mRNA and protein were found to be expressed in NEBs in lung slices of various mammals. Receptor agonists led to depolarization, while blockers prevented this response. This strongly indicates that serotonin functions as an autocrine regulator of NEB secretion, similar to CB dopamine (Fu et al., 2001). We will next consider the neurotransmitter substances and membrane receptors of the sensory nerve endings associated with glomus cells or corpuscular cells. In the CBs, nerve terminals more or less closely associated with glomus cells were found to show an immunocytochemical reaction to CGRP (Torrealba and Correa, 1995), tyrosine hydroxylase, which indicates synthesis of catecholamine and perhaps dopamine (Finley et al., 1992), SP co-localized with CGRP (Kummer et al., 1989), and nitric oxide synthase (Prabhakar et al., 1993). Direct, close synaptic contact with glomus cells was not always seen. Removal of the petrosal ganglion caused disappearance of CGRP-positive nerve fibers in the CBs (Torrealba, 1992). Retrograde labeling of the carotid sinus nerve with fluoro-gold was combined with immunohistochemistry of the petrosal ganglion for various neurotransmitters. This tracing technique labels the perikarya of the petrosal ganglion neurons that project to the CBs, most of them to the glomus cells, as was made out with the help of denervation and anterograde tracing. Transmitters that located in fluoro-gold-labeled neurons included CGRP and SP, often in co-localization, although many neurons showing this immunoreactivity

Neuroepithelial Bodies and Carotid Bodies  |  341

were not labeled with fluoro-gold (Ichikawa et al., 1993), and tyrosine hydroxylase and SP (Finley et al., 1992). In reports of immuno-electron observations many authors remark, apparently somewhat taken aback, that it is the smaller nerve endings that are labeled for neurotransmitters, not the large ones interpreted as afferents in morphological studies. In our opinion, this is not a problem at all, since the smaller efferent nerve endings are exactly those that are credited with antidromic activity and need neurotransmitter to do so. Backing up the already strong evidence that CB glomus cells use acetylcholine and ATP as neurotransmitters, the associated sensory nerve endings were found to carry various complementary receptors. The α7 subunit of the nicotinic acetylcholine receptor was located immunohistochemically in nerve fibers enveloping glomus cells (Shirahata et al., 1998). M1, but not M2, muscarinic acetylcholine receptors were also expressed in nerve fibers surrounding the glomus cells (Shirahata et al., 2004). Messenger RNAs were detected, and immunocytochemistry showed expression, of P2X2 and P2X3 receptors in petrosal ganglion neurons. In addition, isohydric hypercapnia caused excitation of petrosal neuron-glomus cell chemosensory units in co-culture, which was inhibited by suramin, a purinoreceptor blocker, pointing at involvement of ATP as neurotransmitter from glomus cells (Prasad et al., 2001). At least four types of P2X receptors (2b, 3, 4, and 7) were detected in glossopharyngeal neurons projecting to the rat CBs (Campanucci et al., 2006). Dopamine D1-receptor mRNA was detected in petrosal ganglion neurons (Bairam et al., 1998). CGRP is a widely occurring neurotransmitter in rat NEB intracorpuscular nerve fibers (Brouns et al., 2004). It was found in co-localization with nitric oxide synthase, indicating nitrergic neurotransmission (Brouns et al., 2003), and in co-localization with nitric oxide synthase and substance P (Adriaensen et al., 2003; Van Genechten et al., 2004). Also in rat NEB intracorpuscular nerve fibers, immunoreactivity for vesicular glutamate transporter 1 and 2 was detected (Brouns et al., 2006), which apparently is indicative of glutamate uptake, and thus of glutaminergic neurotransmission in intracorpuscular nerve endings. NEB intracorpuscular nerve fibers were found to carry purinergic P2X3 receptors (Adriaensen et al., 2003; Brouns et al., 2003, 2006), as well as P2X2 receptors colocalized with CGRP (Van Genechten et al., 2004). In conclusion, the CB glomus cell neurotransmitter spectrum, as it is presently known, is fairly complementary with the receptors present on the sensory nerve endings. While the glomus cells synthesize acetylcholine, dopamine, and ATP, the nerve endings carry nicotinic and muscarinic, as well as dopaminergic and purinergic membrane receptors. In addition, the glomus cells display muscarinic, dopaminergic, and purinergic autoreceptors. With regard to the neural neurotransmitters, the picture is less complete. Here, most of the neurotransmitter types located thus far are peptides. The corresponding receptors on the glomus cells have apparently not (yet?) been reported.

342  |  Airway Chemoreceptors in the Vertebrates In the NEBs, the picture is also still far from complete. Serotonin and a corresponding receptor have been found, but both are in the corpuscular cell and are thus involved in an autocrine mechanism. Purinergic receptors have been located in the corpuscular cell membrane as well as the neural membrane, but a purinergic neurotransmitter remains to be detected. Nicotinic receptors have been detected in the corpuscular cell membrane, but not the corresponding acetylcholine in the nerve fibers. It is remarkable that both sensory cells and nerve endings may make use of exactly the same transmitter substances or receptors. The most outstanding examples are CGRP and the purinergic receptors in the NEBs. SP, and perhaps dopamine, are similar examples in the CBs. Regarding the neurotransmitters and receptors they use, there are a number of similarities, as well as differences, between CBs and NEBs. CGRP occurs in the nerve fibers of both, as do the purinergic receptors and nitric oxide synthase. On the other hand, acetylcholine and ATP, while present in the CBs, have not yet been located in the NEBs. The above is not an exhaustive catalogue of neurotransmitters and receptors, however, only the most commonly reported ones having been mentioned. Moreover, it is to be expected that future research will unearth many more, and that most of the transmitters already known will turn out to have a complementary receptor, and vice versa. In recent years, nNOS has gained considerable experimental support as a neurotransmiter substance in both CB and NEB, but it seems to be largely confined to intrinsic efferent nerves (Adriaensen et al., 2003; Campanucci and Nurse, 2007).

Oxygen Sensors and Membrane Channels Cell physiological techniques provided deeper insight in the workings of the CBs and provided independent confirmation that the NEBs functioned as chemoreceptors. Direct cell physiological evidence in favor of sensitivity to hypoxia of NEBs was first obtained in NEBs of late fetal rabbit lungs, with the introduction of cell culture and cell physiological techniques such as whole cell patch clamp and single cell recording. Using these techniques, it was discovered that hypoxia acts by closing K+ membrane channels, thereby blocking outward K+ currents and inducing depolarization (Youngson et al., 1993). Voltage-gated Ca2+ channels subsequently open, and a cascade of intracytoplasmic mediators is set in motion that finally leads to docking of secretory vesicles at the cell membrane and exocytotic discharge of neurotransmitters in the extracellular space. It soon became apparent that there were different types of K+ channels involved. Messenger RNAs for Kv3.3a K+ channel (voltage-gated) subunits were detected by in situ hybridization in NEBs of late gestation fetal rabbit lungs, neonatal human lungs, and a number of small cell lung carcinoma cell lines, believed to be derived from NEB cells (Wang et al., 1996). In human NEB-derived H-146 cells

Neuroepithelial Bodies and Carotid Bodies  |  343

(a human small cell lung carcinoma cell line), the pharmacological profile of the K+ channel was found to be indicative of the TASK (two-pore acid-sensitive potassium, not voltage-gated) subfamily of K+ channel proteins. Messenger RNAs of this K+ channel gene could be isolated from these cells (O’Kelly et al., 1999). In the absence of Cd2+, K+ currents were in effect reduced by acidification. Incorporation of antisense TASK-1 RNA probes in H-146 cells led to complete annihilation of the corresponding protein, as indicated by immunocytochemistry. The same procedure led to abolishment of hypoxia-reduced K+ current inhibition. TASK-1 is very similar to TASK-3, but since Zn+ did not prevent hypoxic inhibition of K+ currents, the most likely candidate channel is TASK-3 (Hartness et al., 2001). K+ channels in NEBs were found to operate in close concert with a cell membrane oxygen sensor: nicotinamide adenine dinucleotide phosphate oxidase (NADPH oxidase). The p91 subunit (gp91phox) of this enzyme was demonstrated in NEB plasma membranes with immunocytochemistry (Youngson et al., 1993). Messenger RNAs for two subunits of NADPH oxidase, gp91phox and p22phox, were detected by in situ hybridization in NEBs of late gestation fetal rabbit lungs, neonatal human lungs, and a number of small cell lung carcinoma cell lines (Wang et al., 1996). In addition to the gp91 and p22 components, which had a surface membrane location, others (p67, p47, rac 2) were located in the cytoplasm (Youngson et al., 1997). Messenger RNAs of p47phox subunit of NADPH oxidase were also detected. NADPH oxidase activity can be regulated by protein kinase C dependent phosphorylation of these components (O’Kelly et al., 2000). Exposure to diphenylene iodinium, a specific inhibitor of NADPH oxidase, reduced outward K+ currents in a manner similar to exposure to hypoxia. In addition, NEBs were found to convert non-fluorescent dihydrorhodamine into fluorescent rhodamine, an indication of peroxide production by NADPH oxidase (Youngson et al., 1993). Rhodamine fluorescence was increased in NEB cultures exposed to a phorbol ester, an oxidase stimulator, and outward K+ currents were increased when cultured rabbit NEBs were depolarized in the presence of peroxide (Wang et al., 1996). In H-146 cells, hydrogen peroxide production decreased by perfusion with hypoxic bath solution, as judged by increased photobleaching of H2DCFDA (O’Kelly et al., 2000). Thus, it was hypothesized that NADPH oxidase functions as a sensor molecule by catalyzing the conversion of oxygen to superoxide, using electrons supplied by NADPH, an electron transport chain component. Superoxide is then converted to hydrogen peroxide. In hypoxic conditions, less peroxide is produced, whereupon K+ channels close. Reactive oxygen species (ROS) such as these probably interact with specific cysteine residues, sulfhydryl groups, or disulfide bridges in the K+ channel proteins, maintaining conformational changes that hold the channels open. In hypoxic conditions, fewer ROS are produced, and steric changes occur in the K+ channel proteins, inducing channel closure. In NADPH oxidase-deficient mice, NEB K+ currents were unaffected by hypoxia. Diphenylene iodinium (at least in low doses) had no effect on K+ currents. There

344  |  Airway Chemoreceptors in the Vertebrates still was increase of K+ currents by peroxide, however (Fu et al., 2000). Respiratory parameters were altered in NADPH oxidase-deficient mice (Kazemian et al., 2001). In light of the fact that CBs and lung arterioles may have another kind of oxygen sensor, this observation acquires profound significance. There is some evidence for more than one kind of NEB oxygen sensor. Phenyl arside and diphenylene iodonium, two structurally unrelated NADPH oxidase inhibitors, suppressed, but failed to abolish entirely, hypoxic inhibition of K+ currents in H146 cells (O’Kelly et al., 2001). The alternative oxygen sensor is apparently not mitochondrial, as has been proposed in the CBs. The relation of mitochondria with ROS production may not be as straightforward as in the case of NADPH oxidase. Somewhat paradoxically, hypoxia may actually increase mitochondrial ROS production, since it may slow down electron transfer in the respiratory chain. Rotenone and antimycin A, inhibitors of the electron transport chain, both mimic hypoxic reduction of K+ currents in H146 cells, while they should have opposing effects since they are upstream and downstream, respectively, of the principal chain site suspected of producing ROS. A structurally unrelated mitochondrial inhibitor, myxothiazol, had no effect. Lastly, H146 cells, which are free of mitochondria, show normal K+ current inhibition in hypoxia (Searle et al., 2002). Incidentally, NEB-associated nerve fibers contain nitric oxide synthase (Brouns et al., 2003; Van Genechten et al., 2004), which has been advanced as an oxygen sensor in the CBs (Prabhakar et al., 1993) In the CBs, chemostimuli likewise act by regulating K+ channel activity in the type I glomus cell membranes. Hypoxia, or decreasing oxygen pressure, causes K+ channel closure followed by membrane depolarization (Lopez-Barneo et al., 1988; Wyatt et al., 1995), and activation of voltage-gated Ca2+ channels and Ca2+-dependent transmitter release (Urena et al., 1994). Several types of K+ channel have been found to be operative, including TASK (Yamamoto et al., 2002; Buckler et al., 2000) and Kv (Perez-Garcia et al., 2004; Sanchez et al., 2002). Thus, there seem to be no important dissimilarities regarding the identity of the ion channels involved in hypoxic responses between NEBs and CBs. There seem to be more important differences with NEBs regarding the identity of the oxygen sensor in the CBs. In the CBs, there is fairly substantial evidence for more than one type of oxygen sensor, and the main one may not be NADPH oxidase. It is even probable that several types of oxygen sensor operate at once, or in different species, in different preparations, and at different ages. Candidate CB oxygen sensors include NADPH oxidase, cytochromes of the mitochondrial electron transport chain, heme oxygenase 2. They have in common their involvement with ROS or analogous molecules, which induce conformational changes in channel proteins, thereby determining whether these open or close. It is also possible that direct oxygen sensitivity of the membrane channels themselves contributes to the reaction of glomus cells to hypoxia. Although NADPH-oxidase immunoreactivity was detected in CB glomus cells, and the p22phox subunit was located by immunoelectron microscopy in dense-cored

Neuroepithelial Bodies and Carotid Bodies  |  345

vesicles, and gp91phox, p22phox, and rac 2 were also located in glomus cell cultures (as in NEBs), inactivation of the gp91phox subunit of the NADPH, while obliterating chemosensitivity in NEBs, appeared not to be effective (Youngson et al., 1997; Roy et al., 2000; He et al., 2002). On the other hand, genetic suppression of the p47phox regulatory subunit resulted in increased sinus nerve activity in hypoxia (Sanders et al., 2002). Thus, oxygen sensing in the CBs may partially rest on hypoxic reduction in production of ROS by NADPH oxidase, but NADPH oxidase as the main sensor has been largely discounted. Mitochondria, or more exactly the cytochromes they harbor, almost suggested themselves as sites of cellular oxygen sensing since the slightest fall in the optimal level of oxygen should directly affect electron transfer and oxidative phosphorylation. It is probable that electron transfer gives rise to ROS production. But disturbances in mitochondrial electron transfer are by no means linked to CB oxygen sensitivity in a simple, straightforward way, as already remarked in connection with NEBs. Initially, inhibitors of mitochondrial electron transfer were found to mimic the effects of hypoxia on glomus cells, such as a rise in intracellular Ca2+ and membrane depolarization (Buckler and Vaughan-Jones, 1998). When more kinds of mitochondrial electron transfer inhibitors were tested, some of them allowed an additive cellular response to hypoxia, while others did not (Ortega-Saenz et al., 2003; Wyatt and Buckler, 2004). At least in the CBs, the whole concept of oxygen sensing through diminished production of ROS has also been fundamentally questioned since N-acetylcysteine, a scavenger of ROS, did not affect the CB glomus cells, whereas, if the aforementioned concept of ROS involvement in reaction to hypoxia is correct, its actions should mimic those of hypoxia (Sanz-Alfayate et al., 2001). Heme oxygenase 2 (HO-2) may be another oxygen sensor. It generates carbon monoxide, was located in CB glomus cells, and its inhibition activated the glomus cells (Prabhakar et al., 1995). In theory, HO-2 inhibits the glomus cells, while hypoxia in its turn inhibits HO-2, decreasing the rate of carbon monoxide production, whereupon the glomus cells are activated through an effect on their membrane channels. Hypoxic inhibition of mitochondrial oxydative phosphorylation and a fall of cellular ATP levels may raise AMP concentrations sufficiently to induce activation of AMP protein kinase, which in turn may inhibit oxygen-sensitive K+ channels, leading to activation of glomus cells and neurotransmitter release (Wyatt and Evans, 2007).

Embryology and Development In this field also, research has been hampered by the pulmonary integration of the NEBs, while CBs are compact, isolated organs. The issue is even further complicated by the fact that NEBs are themselves regulators of lung development. The carotid body has a dual origin. An ectomesenchymal cell lineage, derived from the wall of the third branchial arch artery, forms a vascular and fibrous tissue rudiment that

346  |  Airway Chemoreceptors in the Vertebrates is subsequently invaded by a neuronal cell lineage from the superior cervical ganglion. It is this neuronal cell lineage that gives rise to the glomus cells. Both lineages ultimately derive from the neural crest. Embryological work providing definite proof of a neural crest origin has only been performed in birds, using cytochemical markers in quailchick chimeras (Pearse et al., 1973). Quail CB glomus cells store dopamine, while those of chickens store serotonin. Replacement of relevant parts of chicken rhombencephalic primordia with corresponding quail allografts in age-matched embryos resulted in the development of chicken CBs with glomus cells storing dopamine. A similar experiment made use of the fact that quail and chicken cell nuclei look quite distinct (Le Douarin et al., 1972). Conclusive proof of a neural crest origin of NEBs, long ago predicted by many, has never been provided. On the contrary, what little experimental data we have argue in favor of an endodermal derivation. In newborn hamster lungs exposed to tritiated thymidine during the final days of gestation, the nuclei of the airway epithelium show autoradiographic labeling. The younger, more peripheral airway branches show the heaviest label. Crucially, the NEBs follow this pattern: in the peripheral airways they are more heavily labeled than in the central ones, although in a given airway segment they are always less heavily labeled than the surrounding epithelial cells. These observations indicate that the NEBs develop in close parallel with the surrounding epithelium and do not derive from an external source (Hoyt et al., 1990). Fetal hamster lungs, maintained in explant culture from 2 d before the appearance of NEBs, showed normal appearance and developmental pattern of NEBs. Even after digestion of mesenchyma and nervous tissue, so that only epithelium remained in culture, NEBs developed normally (Ito et al., 1997). All this seems to argue conclusively against an extra-epithelial derivation. CB glomus cell and NEB corpuscular cell development are under genetic regulation. Development of the glomus cells and the vascular rudiment they lodge in is regulated by different genes. The glomus cells fail to develop in Mash 1-deficient mice (Kameda, 2005). The FORSE-1 (forebrain-surface-embryonic) epitope is a marker of progenitor cells in the very early mammalian forebrain. In rabbit fetal lung, it first labels the developing airway epithelium at day 16, then becomes restricted to NEBs, and decreases as serotonin immunoreactivity increases (Pan et al., 2002). Subsequent genetic regulation of NEB development is not yet fully understood, but there is evidence that the same genes that determine the development of nerve cells (i.e., promote neural cell commitment) are also active in embryonic NEBs, in great contrast to the other, non-neuroendocrine lung epithelial cells. NEBs are epithelial, but they show activation of a number of genes that also regulate the development of nerve cells. As is the case in the CBs, Mash 1 gene disruption (the mammalian analog of Drosophila Achaete-scute-homolog-1) leads to total absence of NEBs, in an otherwise entirely normal lung, of newborn mice (Borges et al., 1997; Ito et al., 2000). Apart from this, pulmonary NEBs may show developmental idiosyncrasies that put them apart from other neuroendocrine cells. Protein tyrosine phosphatase-σ deficiency, which

Neuroepithelial Bodies and Carotid Bodies  |  347

causes severe defects in neuroendocrine cells elsewhere, does not appear to affect NEBs at all (Batt et al., 2003). Contrary to CB glomus cell clusters, NEBs do not develop in isolation. There is evidence that they are under the influence of secretory products of adjacent epithelial cells, e.g., Clara cell 10kD protein. Knockout of the relevant gene resulted in significantly smaller NEBs than normal (Castro et al., 2000). In rabbit CBs, glomus cells can be recognized as such under the electron microscope at gestational day 16. Unmyelinated nerve fibers also appear at this time and gradually develop nerve endings that establish contacts with glomus cells. From here till term, the nerve endings acquire, consecutively, synaptic vesicles, efferent synaptic junctions, and afferent synaptic junctions. Significantly, large mitochondria-filled morphologically afferent nerve endings are not yet present at full term (Kariya et al., 1990). In rat CBs, glomus cells appear at gestational day 14, grow, increase in number, and accumulate dense vesicles until they reach their mature aspect shortly before birth. Unmyelinated nerve fibers appear around this time, but no synaptic junctions are present until 1617 d, i.e., at term, and even then they are not numerous. Large, morphologically afferent, mitochondria-filled nerve endings, of the type described in the adult rat CBs by the same author, are not specifically mentioned (Kondo, 1975). Just before birth, the number of afferent synapses begins to increase dramatically and continues to do so until several weeks after birth, while efferent synapses remain few in number (Kondo, 1976b). Post-natally, the number of mitochondria-filled afferent nerve endings increases at the expense of the vesicle-filled efferent ones in rabbit CBs (Bollé et al., 2000). There is still nuclear incorporation of the synthetic DNA precursor BrdU at birth and for a month thereafter, indicating glomus cell proliferation (Kameda, 2005; Wang and Bisgard, 2005). These observations seem to indicate that CB development is not fully terminated at birth. In particular it is the large, morphologically afferent nerve endings that appear to undergo considerable post-natal development. NEBs originate and differentiate according to a centrifugal pattern that follows the growing branches of the bronchial tree (Sorokin et al., 1982; Hoyt et al., 1990; McDowell et al., 1994). It has long been clear that NEBs assume a differentiated phenotype very early in lung development. In fact, they are virtually the first epithelial cell type to do so (Cutz et al., 1985). In the rabbit they appear at day 18-19 of gestation, i.e., in the pseudoglandular stage (Sorokin et al., 1982), and in the hamster lungs they appear at 12 d (McDowell et al., 1994). In the human fetus, NEBs are recognizable under the electron microscope by 8 wk (Cutz et al., 1985) or 10 wk (Stahlman and Gray, 1984) of pregnancy. CGRP immunoreactivity in NEBs appears rather late in the rat fetus (day 18 of gestation), peaks on day 20, and declines post-natally (Wada et al., 1988). Human NEBs show innervation by week 20 of gestation (Stahlman and Gray, 1984). Serotonin-positive NEBs are present on gestational day 18 in rabbit lungs. They are innervated from day 21 onwards. Their numbers, and the density of their innervation, peak at post-natal day 2 (Pan et al., 2004). Post-natally, the relative number of afferent and efferent nerve endings remains essentially constant, indicating

348  |  Airway Chemoreceptors in the Vertebrates that the innervation is mature at birth (Bollé et al., 2000). Upon administration of tritiated thymidine or BrdU, NEBs of hamsters near term accumulate appreciably less label than the surrounding epithelial cells (Hoyt et al., 1990, 1991). In rabbit NEBs at term, there was no uptake at all of tritiated thymidine (Hernandez-Vasquez et al., 1977). These findings indicate that NEBs are fairly mature at birth. During lung development, NEB frequencies increase, peak in the perinatal period, and decline subsequently, although they remain present in the adult lung. It is not yet certain whether these changes are absolute or relative. Here we run yet again into difficulties related to the pulmonary integration of the NEBs. The lung itself undergoes profound developmental changes pre-natally, but also during the time of birth, and continues its development (alveolarization) for a considerable time post-natally. Thus, it is by no means clear whether the changes in NEB frequencies are absolute, i.e., reflect post-natal involution, or relative, in the sense that the apparent decline in NEB frequencies is a dilution effect, caused by the growth of the surrounding lung tissues, especially alveoli. Efforts have been made to resolve this issue by taking recourse to quantitative procedures. In collapsed rabbit lungs, NEBs were detected by means of induced fluorescence of biogenic amines. Relative numbers of NEBs (per mm2 of lung tissue) dipped near term (30 d gestation), peaked shortly after birth, and declined postnatally to extremely low values. This is evidently a consequence of lung growth: collapsed lung volume increases substantially after birth. Total NEB numbers, estimated from collapsed lung volume, NEB diameter, and NEB density, dipped shortly before term but then re-established themselves and remained essentially unchanged post-natally at about 50,000 (Redick and Hung, 1984). When the effect of alveolarization was neutralized by considering the airway epithelium only, the number of NEBs per mm3 of airway epithelium in the rabbit peaked shortly after birth and then dramatically declined (Cho et al., 1989). Using a very sophisticated quantification technique, total absolute hamster NEB numbers of both lungs also declined significantly between 1 and 4 wk of age (Bollé et al., 1999). Thus, the question whether these changes in NEB frequencies are absolute or relative has not been entirely resolved. As an adaptation to low fetal arterial oxygen tensions, CBs are relatively insensitive to hypoxia at birth. They become more active and increase their sensitivity over the next days and weeks. Resetting, as this increase in CB chemosensitivity is called (Wasicko et al., 1999), is apparently causally related to the appreciable rise in arterial oxygen tension at birth: continued exposure to a hypoxic environment blunts this development (Sterni et al., 1999; Hanson et al., 1989), while exposure to hyperoxia in utero induces it (Blanco et al., 1988). The question as to the level(s) within the CBs where these developmental changes in chemotransduction occur is not yet entirely answered. In all probability, several levels are involved. Anatomical changes have been implicated, such as the increase of the number of afferent nerve terminals (Kondo, 1976b). There may be changes in transmitter synthesis and storage. Dopamine content is low at birth and increases post-natally, as does hypoxia-induced dopamine secretion (Bairam

Neuroepithelial Bodies and Carotid Bodies  |  349

et al., 1996). Concomitantly, tyrosine hydroxylase mRNA levels decrease and D2dopamine receptor mRNA levels increase (Gauda et al., 1996). There may be changes in cell membrane sensitivity to depolarization. Calcium influx in response to hypoxia increases with post-natal age (Wasicko et al., 1999), as does the autocrine inhibitory effect of dopamine on calcium influx (Carroll et al., 2005). Carotid sinus nerve afferent signaling frequencies in hypoxia increase in the course of post-natal development (Kholwadwala and Donelly, 1992), but it is uncertain whether this is a primary effect. The neural response of CBs to carbon dioxide also increases post-natally (Carroll et al., 1993). Glomus cells reactive to dopamine β hydroxylase were numerous at birth in the rat but decreased from post-natal week 2 onwards. In the glomus cell-associated nerves, the evolution was reversed. Tyrosine hydroxylase-reactive glomus cells and nerve fibers remained numerous (Oomori et al., 2002). Thus, post-natal maturation of the CBs probably involves numerous aspects of the chemotransduction mechanism at glomus cell and afferent nerve fiber levels. It is not known whether NEBs undergo such profound physiological changes after birth, but it is to be expected that future in vitro studies will help to resolve this issue.

Summary and Conclusions The main difficulty that one confronts in trying to compare NEBs and CBs is that, by nature of their very different anatomical organization and location, performing identical experiments has often been difficult or even impossible. One has been forced to draw conclusions from widely different kinds of experiments. Notwithstanding their rather different embryological descent, it is obvious that their cellular structure, the nature of their innervation, their pharmacological and cell physiological aspects, are all very similar, in an analogous if not in a homologous sense. Overall, the similarities appear to be more fundamental than the differences. There are tantalizing hints that the NEBs are similar enough to the CBs to be able to compensate for them. Congenital central hypoventilation syndrome is a pathology characterized by failure of the central respiratory control mechanisms. CBs were found to be half as large as normal, with decreased glomus cell number and frequency of dense-cored vesicles. In the same individuals, NEB size and frequency was increased twofold compared to healthy subjects, suggesting compensatory hyperplasia (Cutz et al., 1997). When the CBs are denervated, a peripheral ventilatory chemoreflex redevelops (Forster et al., 2000). It is unknown precisely where this reflex is situated, but as it is observed that the chemoreceptor properties of the carotid sinus region disappear permanently when the glomus cells are destroyed or removed (Verna et al., 1975; Smith and Mills, 1979), it is uncertain that this site is the CB itself. If one has to look for differences between CBs and NEBs, the two most obvious ones that suggest themselves appear to be their orientation towards different body

350  |  Airway Chemoreceptors in the Vertebrates compartments, and the different time course of their maturation. While CBs are sensitive to blood-borne stimuli, all data we have point to NEBs as monitors of changes in the composition of the inhaled air. On the other hand, judged by the quantitative as well as the qualitative data we presently have, NEBs appear to mature in advance of the CBs. The NEB strongpoint is that, by virtue of their contact with inhaled air, they can react much more quickly to hypoxia, and there is no delay period during which hypoxemia has to develop. This may be of particular importance in newborns. This theory is strengthened because it makes intuitive sense, and it is attractive because it suggests further testable hypotheses. It would be very instructive to further contrast CBs and NEBs, not so much in reviews, but in actual experimental conditions, insofar as this is possible. Efforts to do this have been all too rare.

Acknowledgements Gratitude is expressed to Chris Armee and Rolande Renwart for expertly processing our materials for electron microscopy, as well as to Vital Noppen for assistance with digital photography and imaging.

References Adriaensen, D., Brouns, I., Van Genechten, J., and Timmermans, J.P. 2003. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat. Rec., 270A: 25-40. Adriaensen, D., Timmermans, J.P., Brouns, I., Berthoud, H.R., Neuhuber, W.L., Scheuermann, D.W. 1998. Pulmonary intraepithelial vagal nodose afferent nerve terminals are confined to neuroepithelial bodies: an anterograde tracing and confocal microscopy study in adult rats. Cell Tissue Res., 293: 395-405. Archer, S.L., Weir, E.K., Reeve, H.L., and Michelakis, E. 2000. Molecular identification of O2 sensors and O2-sensitive potassium channels in the pulmonary circulation. Adv. Exp. Med. Biol., 475: 219-240. Bairam, A., Basson, H., Marchal, F., Cottet-Emard, J.M., Pequignot, J.M., Hascoet, J.M., and Lahiri, S. 1996. Effects of hypoxia on carotid body dopamine content and release in developing rabbits. J. Appl. Physiol., 80: 20-24. Bairam, A., Frenette, J., Dauphin, C., Carroll, J.L., and Khandjian, E.W. 1998. Expression of dopamine D1-receptor mRNA in the carotid body of adult rabbits, cats and rats. Neurosci. Res., 31: 147-154. Bairam, A., and Marchal, F. 2003. Carotid sinus nerve chemosensory response to dopamine and acetylcholine in catecholamine depleted cats. Respir. Physiol. Neurobiol., 134: 1-12. Batt, J., Cutz, E., Fladd, C., and Rotin, D. 2003. Apparent normal lung architecture in protein tyrosine phosphatase-sigma-deficient mice. Am. J. Physiol. Lung Cell. Mol. Physiol., 284: L214-L223. Biscoe, T.J., and Pallot, D. 1972. Serial reconstruction with the electron microscope of carotid body tissue. The type I cell nerve supply. Experientia, 28: 3-4.

Neuroepithelial Bodies and Carotid Bodies  |  351 Blanco, C.E., Hanson, M.A., and McCooke, H.B. 1988. Effects on carotid body chemoreceptor resetting of pulmonary ventilation in the fetal lamb in utero. J. Dev. Physiol., 10: 167-174. Blümcke, S., Rode, J., and Niedorf, H.R. 1967. The carotid body after oxygen deficiency. Z. Zellforsch. Mikrosk. Anat., 80: 52-77. Bollé, T., Van Lommel, A., and Lauweryns, J.M. 1999. Stereological estimation of number and volume of pulmonary neuroepithelial bodies (NEBs) in neonatal hamster lungs. Microsc. Res. Tech., 44: 190-194. Bollé, T., Lauweryns, J.M., and Van Lommel, A. 2000. Postnatal maturation of neuroepithelial bodies and carotid body innervation: a quantitative investigation in the rabbit. J. Neurocytol., 29: 241-248. Borges, M., Linnoila, R.I., Van de Velde, H.J.K., Chen, H., Nelkin, B.D., Mabry, M., Baylin, S.B., and Ball, D.W. 1997. An achaete-scute homologue essential for neuroendocrine differentiation of the lung. Nature, 386: 852-855. Brouns, I., Pintelon, I., De Proost, I., Alewaters, R., Timmermans, J.P., and Adriaensen, D. 2006. Neurochemical characterization of sensory receptors in airway smooth muscle: comparison with pulmonary neuroepithelial bodies. Histochem. Cell Biol., 125: 351-367. Brouns, I., Pintelon, I., Van Genechten, J., De Proost, I., Timmermans, J.P., and Adriaensen, D. 2004. Vesicular glutamate transporter 2 is expressed in different nerve fibre populations that selectively contact pulmonary neuroepithelial bodies. Histochem. Cell Biol., 121: 1-12. Brouns, I., Van Genechten, J., Hayashi, H., Gajda, M., Gomi, T., Burnstock, G., Timmermans, J.P., and Adriaensen, D. 2003. Dual sensory innervation of pulmonary neuroepithelial bodies. Am. J. Respir. Cell Mol. Biol., 28: 275-285. Bruce, E.N., and Cherniak, N.S. 1987. Central chemoreceptors. J. Appl. Physiol., 62: 389-402. Buckler, K.J., and Vaughan-Jones, R.D. 1998. Effects of mitochondrial uncouplers on intracellular calcium, pH and membrane potential in the carotid body type I cells. J. Physiol., 513: 819-833. Buckler, K.J., Williams, B.A., and Honore, E. 2000. An oxygen-, acid- and anaesthetic-sensitive TASK-like background potassium channel in rat arterial chemoreceptor cells. J. Physiol., 525: 135-142. Buttigieg, J., and Nurse, C.A. 2004. Detection of hypoxia-evoked ATP release from chemo receptor cells of the rat carotid body. Biochem. Biophys. Res. Commun., 322: 82-87. Campanucci, V.A., and Nurse, C.A. 2007. Autonomic innervation of the carotid body: role in efferent inhibition. Resp. Physiol. Neurobiol., 157: 83-92. Campanucci, V.A., Zhang, M., Vollmer, C., and Nurse, C.A. 2006. Expression of multiple P2X receptors by glossopharyngeal neurons projecting to rat carotid body O2-chemoreceptors: role in nitric oxide-mediated efferent inhibition. J. Neurosci., 26: 9482-9493. Carroll, J.L., Bamford, O.S., and Fitzgerald, R.S. 1993. Postnatal maturation of carotid chemoreceptor responses to O2 and CO2 in the cat. J. Appl. Physiol., 75: 2383-2391. Carroll, J.L., Boyle, K.M., Wasicko, M.J., and Sterni, L.M. 2005. Dopamine D2 receptor modulation of carotid body type I cell intracellular calcium in developing rats. Am. J. Physiol. (Lung Cell. Mol. Physiol.), 288: L910-L916. Castro, C.M., Yang, Y., Zhang, Z., and Linnoila, R.I. 2000. Attenuation of pulmonary neuroendocrine differentiation in mice lacking Clara cell secretory protein. Lab. Invest., 80: 1533-1540. Cho, T., Chan, W., and Cutz, E. 1989. Distribution and frequency of neuro-epithelial bodies in post-natal rabbit lung: quantitative study with monoclonal antibody against serotonin. Cell Tissue Res., 255: 353-362.

352  |  Airway Chemoreceptors in the Vertebrates Cutz, E., Gillan, J.E., and Bryan, A.C. 1985. Neuroendocrine cells in the developing human lung: morphologic and functional considerations. Pediatr. Pulmonol., 1: S21-S29. Cutz, E., Ma, T.K.F., Perrin, D.G., Moore, A.M., and Becker, L.E. 1997. Peripheral chemoreceptors in congenital central hypoventilation syndrome. Am. J. Respir. Crit. Care Med., 155: 358-363. De Castro, F. 1926. Sur la structure et l’ innervation de la glande intercarotidienne (glomus caroticum) de l’ homme et des mammifères, et sur un nouveau système d’ innervation autonome du nerf glossopharyngien. Trab. Lab. Invest. Biol. Univ. Madrid, 24: 365-432. Delpiano, M.A., and Acker, H. 1991. Hypoxia increases the cyclic AMP content of the cat carotid body in vitro. J. Neurochem., 57: 291-297. Eyzaguirre, C., Koyano, H., and Taylor, J.R. 1965. Presence of acetylcholine and transmitter release from carotid body chemoreceptors. J. Physiol., 178: 463-476. Fidone, S.J., Stensaas, L.J., and Zapata, P. 1975. Sensory nerve endings containing “synaptic” vesicles: an electron-microscopic autoradiographic study. J. Neurobiol., 6: 423-427. Fidone, S.J., Zapata, P., and Stensaas, L.J. 1977. Axonal transport of labeled material into sensory nerve endings of cat carotid body. Brain Res., 124: 9-28. Finley, J.C.W., Polak, J., and Katz, D.M. 1992. Transmitter diversity in carotid body afferent neurons: dopaminergic and peptidergic phenotypes. Neuroscience, 51: 973-987. Forster, H.V., Pan, L.G., Lowry, T.F., Serra, A., Wenninger, J., and Martino, P. 2000. Important role of carotid chemoreceptor afferents in control of breathing of adult and neonatal mammals. Resp. Physiol., 119: 199-208. Fu, X.W., Nurse, C.A., and Cutz, E. 2004. Expression of functional purinergic receptors in pulmonary neuroepithelial bodies and their role in hypoxia chemotransmission. Biol. Chem., 385: 275-284. Fu, X.W., Nurse, C.A., McFarragher, S.M., and Cutz, E. 2003. Expression of functional nicotinic acetylcholine receptors in neuroepithelial bodies of neonatal hamster lung. Am. J. Physiol. Lung Cell. Mol. Physiol., 285: L1203-L1212. Fu, X.W., Nurse, C.A., Wong, V., and Cutz, E. 2002. Hypoxia-induced secretion of serotonin from intact pulmonary neuroepithelial bodies in neonatal rabbit. J. Physiol., 539: 503-510. Fu, X.W., Wang, D., Nurse, C.A., Dinauer, M.C., and Cutz, E. 2000. NADPH oxidase is an O2 sensor in airway chemoreceptors: evidence form K+ current modulation in wild-type and oxidase-deficient mice. Proc. Natl. Acad. Sci. USA, 97: 4374-4379. Fu, X.W., Wang, D., Pan, J., Farragher, S.M., Wong, V., and Cutz, E. 2001. Neuroepithelial bodies in mammalian lung express functional serotonin type 3 receptor. Am. J. Physiol. Lung Cell. Mol. Physiol., 281: L931-L940. Gauda, E.B., Bamford, O., and Gerfen, C.R. 1996. Developmental expression of tyrosine hydroxylase, D2-dopamine receptor and substance P genes in the carotid body of the rat. Neuroscience, 75: 969-977. Hansen, J.T. 1985. Ultrastructure of the primate carotid body: a morphometric study of the glomus cells and nerve endings on the monkey (Macaca fascicularis). J. Neurocytol., 14: 13-32. Hanson, M.A., Kumar, P., and Williams, B.A. 1989. The effect of chronic hypoxia upon the development of respiratory chemoreflexes in the newborn kitten. J. Physiol., 411: 563-574. Hartness, M.E., Lewis, A., Searle, G., O’Kelly, I., Peers, C., and Kemp, P.J. 2001. Combined antisense and pharmacological approaches implicate hTASK as an airway oxygen sensing K+ channel. J. Biol. Chem., 276: 26499-26508.

Neuroepithelial Bodies and Carotid Bodies  |  353 He, L., Chen, J., Dinger, B., Sanders, K., Sundar, K., Hoidal, J., and Fidone, S. 2002. Characteristics of carotid body chemosensitivity in NADPH oxidase-deficient mice. Am. J. Physiol. Cell. Physiol., 282: 27-33. He, L., Chen, J., Dinger, B., Stensaas, L., and Fidone, S. 2006. Effect of chronic hypoxia on purinergic synaptic transmission in rat carotid body. J. Appl. Physiol., 100: 157-162. Hernandez-Vasquez, A., Will, J.A., and Quay, W.B. 1977. Quantitative characteristics of the Feyrter (APUD) cells of the neonatal rabbit lung in normoxia and chronic hypoxia. Thorax, 32: 449-456. Hess, A., and Zapata, P. 1972. Innervation of the cat carotid body: normal and experimental studies. Fed. Proc., 31: 1365-1382. Heymans, C. 1955. Actions of drugs on carotid body and sinus. Pharmacol. Rev., 7: 119-142. Hoyt, R.F., McNelly, N.A., McDowell, E.M., and Sorokin, S. 1991. Neuroepithelial bodies stimulate proliferation of airway epithelium in fetal hamster lungs. Am. J. Physiol. Lung Cell. Mol. Physiol., 260: L234-L240. Hoyt, R.F., McNelly, N.A., and Sorokin, S. 1990. Dynamics of neuroepithelial body (NEB) formation in developing hamster lung: light microscopic autoradiography after 3H-thymidine labeling in vivo. Anat. Rec., 227: 340-350. Hung, K.S., and Loosli, C.G. 1974. Bronchiolar neuro-epithelial bodies in the neonatal mouse lungs. Am. J. Anat., 140: 191-199. Ichikawa, H., Rabchevsky, A., and Helke, C. 1993. Presence and coexistence of putative neurotransmitters in carotid sinus baro- and chemoreceptor afferent neurons. Brain Res., 611: 67-74. Ito, T., Nogawa, H., Udaka, N., Kitamura, H., and Kanisawa, M. 1997. Development of pulmonary neuroendocrine cells of fetal hamster in explant culture. Lab. Invest., 77: 449457. Ito, T., Udaka, N., Yazawa, T., Okudela, K., Hayashi, H., Sudo, T., Guillemot, F., Kageyama, R., and Kitamura, H. 2000. Basic helix-loop-helix transcription factors regulate the neuroendocrine differentiation of fetal mouse pulmonary epithelium. Development, 127: 3913-3921. Kameda, Y. 2005. Mash 1 is required for glomus cell formation in the mouse carotid body. Dev. Biol., 283: 128-139. Kariya, I., Nakajima, T., and Ozawa, H. 1990. Ultrastructural study on cell differentiation of the rabbit carotid body. Arch. Histol. Cytol., 53: 245-258. Kazemian, P., Stephenson, R., Yeger, H., and Cutz, E. 2001. Respiratory control in neonatal mice with NADPH oxidase deficiency. Resp. Physiol., 126: 89-101. Keith, I.M., Pelto-Huikko, M., Schalling, M., and Hökfelt, T. 1991. Calcitonin gene-related peptide and its mRNA in pulmonary neuroendocrine cells and ganglia. Histochemistry, 96: 311-345. Keith, I.M., Wiley, L.A., and Will, J.A. 1981. Pulmonary neuroendocrine cells: decreased serotonin fluorescence and stable argyrophil-cell numbers in acute hypoxia. Cell Tissue Res., 214: 201-205. Kholwadwala, D., and Donelly, D.F. 1992. Maturation of carotid chemoreceptor sensitivity to hypoxia: in vitro studies in the newborn rat. J. Physiol., 453: 461-473. Kim, D.K., Oh, E.K., Summers, B.A., Prabhakar, N.R., and Kumar, G.K. 2001. Release of substance P by low oxygen in the rabbit carotid body: evidence for the involvement of calcium channels. Brain Res., 892: 359-369. Kondo, H. 1975. A light and electron microscopic study on the embryonic development of the rat carotid body. Am. J. Anat., 144: 275-293.

354  |  Airway Chemoreceptors in the Vertebrates Kondo, H. 1976a. Innervation of the carotid body of the adult rat. A serial ultrathin section analysis. Cell Tissue Res., 173: 1-15. Kondo, H. 1976b. An electron microscopic study on the development of synapses in the carotid body. Neurosci. Lett., 3: 197-200. Kummer, W., Fischer, A., and Heym, C. 1989. Ultrastructure of calcitonin gene-related peptideand substance P-like immunoreactive nerve fibres in the carotid body and carotid sinus of the guinea pig. Histochemistry, 92: 433-439. Lauweryns, J.M., Cokelaere, M., Deleersnyder, M., and Liebens, M. 1977. Intrapulmonary neuro-epithelial bodies in newborn rabbits. Influence of hypoxia, hyperoxia, hypercapnia, nicotine, reserpine, L-DOPA and 5-HTP. Cell Tissue Res., 182: 425-440. Lauweryns, J.M., Cokelaere, M., Lerut, T., and Theunynck, P. 1978. Cross-circulation studies on the influence of hypoxia and hypoxemia on neuro-epithelial bodies in young rabbits. Cell Tissue Res., 193: 373-386. Lauweryns, J.M., Cokelaere, M., and Theunynck, P. 1972. Neuro-epithelial bodies in the respiratory mucosa of various mammals. Z. Zellforsch., 135: 569-592. Lauweryns, J.M., De Bock, V., Guelinckx, P., and Decramer, M. 1983. Effects of unilateral hypoxia on neuroepithelial bodies in rabbit lungs. J. Appl. Phys. Resp. Environ. Exercise Physiol., 55: 1665-1668. Lauweryns, J.M., De Bock, V., Verhofstad, A.A.J., and Steinbusch, H.W.M. 1982. Immunohistochemical localization of serotonin in intrapulmonary neuro-epithelial bodies. Cell Tissue Res., 226: 215-223. Lauweryns, J.M., and Van Lommel, A. 1987. Ultrastructure of nerve endings and synaptic junctions in rabbit intrapulmonary neuroepithelial bodies: a single and serial section analysis. J. Anat., 151: 65-83. Lauweryns, J.M., Van Lommel, A.T., and Dom, R.J. 1985. Innervation of rabbit intrapulmonary neuroepithelial bodies. Quantitative and qualitative ultrastructural study after vagotomy. J. Neurol. Sci., 67: 81-92. Lauweryns, J.M., and Van Ranst, L. 1987. Calcitonin gene related peptide immunoreactivity in rat lung: light and electron microscopic study. Thorax, 42: 183-189. Le Douarin N., Le Lièvre, C., and Fontaine, J. 1972. Recherches expérimentales sur l’ origine embryologique du corps carotidien chez les oiseaux. C.R. Acad. Sci. Ser. D 275: 583-586. Lopez-Barneo, J., Lopez-Lopez, J.R., Urena, J., and Gonzalez, C. 1988. Chemotransduction in the carotid body: K+ current modulated by pO2 in type I chemoreceptor cells. Science, 241: 580-582. Luts, A., Uddman, R., Absood, A., Hakanson, R., and Sundler, F. 1991. Chemical coding of endocrine cells of the airways: presence of helodermin-like peptides. Cell Tissue Res., 265: 425-433. Matsumoto, S., Uchida, T., Nakajima, T., and Ozawa, H. 1980. Presence of reciprocal synapses in the rabbit carotid body. Arch. Histol. Jap., 43: 275-279. McDonald, D.M., and Mitchell, R.A. 1975. The innervation of glomus cells, ganglion cells and blood vessels in the rat carotid body: a quantitative ultrastructural analysis. J. Neurocytol., 4: 177-230. McDonald, D.M., and Mitchell, R.A. 1981. The neural pathway involved in “efferent inhibition” of chemoreceptors in the cat carotid body. J. Comp. Neurol., 201: 457-476. McDowell, E.M., Hoyt, R.F., and Sorokin, S.P. 1994. Ontogeny of endocrine cells in the respiratory system of Syrian golden hamsters. II. Intrapulmonary airways and alveoli. Cell Tissue Res., 275: 157-167.

Neuroepithelial Bodies and Carotid Bodies  |  355 Mills, E., and Slotkin, T.A. 1975. Catecholamine content of the carotid body in cats ventilated with 8-40% oxygen. Life Sci., 16: 1555-1562. Monti-Bloch, L., and Eyzaguirre, C. 1980. A comparative physiological and pharmacological study of cat and rabbit carotid bodies. Brain Res., 193: 449-470. Montuenga, L.M., Springall, D.R., Gaer, J., Winter, R.J.D., Zhao, L., Mcbride, J.T., Taylor, K.M., Barer, G., and Polak, J.M. 1992. CGRP-immunoreactive endocrine cell proliferation in normal and hypoxic rat lung studied by immunocytochemical detection of incorporation of 5’-bromodeoxyuridine. Cell Tissue Res., 268: 9-15. Morgan, M., Pack, R.J., and Howe, A. 1975. Nerve endings in rat carotid body. Cell Tissue Res., 175: 255-272. Nishi, K., and Stensaas, L.J. 1974. The ultrastructure and source of nerve endings in the carotid body. Cell Tissue Res., 154: 303-319. O’Kelly, I., Lewis, A., Peers, C., and Kemp, P.J. 2000. O2 sensing by airway chemoreceptorderived cells. Protein kinase C activation reveals functional evidence for involvement of NADPH oxidase. J. Biol. Chem., 276: 7684-7692. O’Kelly, I., Peers, C., and Kemp, P.J. 2001. NADPH oxidase does not account fully for O2-sensing in model airway chemoreceptor cells. Biochem. Biophys. Res. Comm., 283: 1131-1134. O’Kelly, I., Stephens, R.H., Peers, C., and Kemp, P.J. 1999. Potential identification of the O2sensitive K+ current in a human neuroepithelial body-derived cell line. Am. J. Physiol., 276 (Lung Cell. Mol. Physiol., 20): L96-L104. Oomori, Y., Murabayashi, H., Ishikawa, K., Miyakawa, K., Nakaya, K., and Tanaka, H. 2002. Neuropeptide Y- and catecholamine-synthesizing enzymes: immunoreactivities in the rat carotid body during postnatal development. Anat. Embryol., 206: 37-47. Ortega-Saenz, P., Pardal, R., Garcia-Fernandez, M., and Lopez-Barneo, J. 2003. Rotenone selectively occludes sensitivity to hypoxia in rat carotid body glomus cells. J. Physiol., 548: 789-800. Pallot, D.J., Al Neamy, K.W., and Blakeman, N. 1986. Quantitative studies of rat carotid body type I cells. Acta Anat., 126: 187-192. Pallot, D.J., and Blakeman, N. 1982. Quantitative ultrastructural studies of the cat carotid body II. The type I cell nerve endings. Acta Anat., 113: 151-158. Pan, J., Yeger, H., and Cutz, E. 2002. Neuronal developmental marker FORSE-1 identifies a putative progenitor of the pulmonary neuroendocrine cell lineage during development. J. Histochem. Cytochem., 50: 1567-1578. Pan, J., Yeger, H., and Cutz, E. 2004. Innervation of pulmonary neuroendocrine cells and neuroepithelial bodies in developing rabbit lung. J. Histochem. Cytochem. 52: 379-389. Pardal, R., Ludewig, U., Garcia-Hirschfeld, J., and Lopez-Barneo, J. 2000. Secretory responses of intact glomus cells in thin slices of rat carotid body to hypoxia and tetraethylammonium. Proc. Natl. Acad. Sci. USA, 97: 2361-2366. Pearse, A.G.E., Polak, J.M., Rost, F.W.D., Fontaine, J., Le Liévre, C., and Le Douarin, N. 1973. Demonstration of the neural crest origin of type I (APUD) cells in the avian carotid body, using a cytochemical marker. Histochem. Cell. Biol., 34: 191-203. Perez-Garcia, M.T., Colinas, O., Miguel-Velado, E., Moreno-Dominguez, A., and Lopez-Lopez, J.R. 2004. Characterization of the Kv channels of mouse carotid body chemoreceptor cells and their role in oxygen sensing. J. Physiol., 557: 457-471. Prabhakar, N.R., Dinerman, J.L., Agani, F.H., and Snyder, S.H. 1995. Carbon monoxide: a role in carotid body chemoreception. Proc. Natl. Acad. Sci. USA, 92: 1994-1997.

356  |  Airway Chemoreceptors in the Vertebrates Prabhakar, N.R., Kumar, G.K., Chang, C.H., Agani, F.H., and Haxhiu, M.A. 1993. Nitric oxide in the sensory function of the carotid body. Brain Res., 625: 16-22. Prasad, M., Fearon, I.M., Zhang, M., Laing, M., Vollmer, C., and Nurse, C.A. 2001. Expression of P2X2 and P2X3 receptor subunits in rat carotid body afferent neurones: role in chemosensory signalling. J. Physiol., 537: 667-677. Redick, M.L., and Hung, K.S. 1984. Quantitation of pulmonary neuroepithelial bodies in preand postnatal rabbits. Cell Tissue Res., 238: 583-587. Roncalli, M., Springall, D.R., Maggioni, M., Moradogli-Haftvani, A., Winter, R.J.D., Zhao, L., Coggi, G., and Polak, J.M. 1993. Early changes in the calcitonin gene-related peptide (CGRP) content of pulmonary endocrine cells concomitant with vascular remodelling in the hypoxic rat. Am. J. Respir. Cell Mol. Biol., 9: 467-474. Roy, A., Rozanov, C., Mokashi, A., Daudu, P., Al-Mehdi, A.B., Shams, H., and Lahiri, S. 2000. Mice lacking gp91 phox subunit of NAD(P)H oxidase showed glomus cell (Ca 2+)i and respiratory responses to hypoxia. Brain Res., 872: 188-193. Sanchez, D., Lopez-Lopez, J.R., Perez-Garcia, M.T., Sanz-Alfayate, G., Obeso, A., Ganfornina, M.D., and Gonzalez, C. 2002. Molecular identification of Kv α subunits that contribute to the oxygen-sensitive K+ current of chemoreceptor cells of the rabbit carotid body. J. Physiol., 542: 369-382. Sanders, K., Sundar, K., He, L., Dinger, B., Fidone, S., and Hoidal, J.R. 2002. Role of components of the phagocytic NADPH oxidase in oxygen sensing. J. Appl. Physiol., 93: 1357-1364. Sanz-Alfayate, G., Obeso, A., Agapito, M.T., and Gonzalez, C. 2001. Reduced to oxidized gluthatione ratios and oxygen sensing in calf and rabbit carotid body chemoreceptor cells. J. Physiol., 537: 209-220. Searle, G.J., Hartness, M.E., Hoareau, R., Peers, C., and Kemp, P.J. 2002. Lack of contribution of mitochondrial electron transport to acute O2 sensing in model airway chemoreceptors. Biochem. Biophys. Res. Comm., 291: 332-337. Shirahata, M., Hirasawa, S., Okumura, M., Mendoza, J.A., Okumura, A., Balbir, A., and Fitzgerald, R.S. 2004. Identification of M1 and M2 muscarinic acetylcholine receptors in the cat carotid body chemosensory system. Neuroscience, 128: 635-644. Shirahata, M., Ishizawa, Y., Rudisill, M., Schofield, B., and Fitzgerald, R.S. 1998. Presence of nicotinic acetylcholine receptors in the cat carotid body afferent system. Brain Res., 814: 213217. Smith, P.G., and Mills, E. 1976. Autoradiographic identification of the terminations of petrosal ganglion neurons in the cat carotid body. Brain Res., 113: 174-178. Smith, P.G., and Mills, E. 1979. Physiological and ultrastructural observations on regenerated carotid sinus nerves after removal of the carotid bodies in cats. Neuroscience, 4: 2009-2020. Sorokin, S.P., Hoyt, R.F., and Grant, M.M. 1982. Development of neuroepithelial bodies in fetal rabbit lungs. I. Appearance and functional maturation as demonstrated by high-resolution light microscopy and formaldehyde-induced fluorescence. Exp. Lung Res., 3: 237-259. Stahlman, M.T., and Gray, M.E. 1984. Ontogeny of neuroendocrine cells in human fetal lung. I. An electron microscopic study. Lab. Invest., 51: 449-463. Sterni, L.M., Bamford, O.S., Wasicko, M.J., and Carroll, J.L. 1999. Chronic hypoxia abolished the postnatal increase in carotid body type I cell sensitivity to hypoxia. Am. J. Physiol. (Lung Cell. Mol. Physiol.), 277: L645-L652. Timmermans, J.P., Brouns, I., Berthoud, H.R., Neuhuber, W.L., and Scheuermann, D.W. 1998. Pulmonary intraepithelial vagal nodose afferent nerve terminals are confined to neuroepithelial bodies: an anterograde tracing and confocal microscopy study in adult rats. Cell Tissue Res., 293: 395-405.

Neuroepithelial Bodies and Carotid Bodies  |  357 Tjen-A-Looi, S., Kraiczi, H., Ekman, R., and Keith, I.M. 1998. Sensory CGRP depletion by capsaicin exacerbates hypoxia-induced pulmonary hypertension in rats. Regul. Pept., 74: 1-10. Torrealba, F. 1992. Calcitonin gene-related peptide immunoreactivity in the nucleus of the tractus solitarius and the carotid receptors of the cat originates from peripheral afferents. Neuroscience, 47: 165-173. Torrealba, F., and Correa, R. 1995. Ultrastructure of calcitonin gene-related peptideimmunoreactive, unmyelinated afferents to the cat carotid body: a case of volume transmission. Neuroscience, 64: 777-785. Urena, J., Fernandez-Chacon, R., Benot, A.R., Alvarez de Toledo, G., and Lopez-Barneo, J. 1994. Hypoxia induces voltage-dependent Ca2+ entry and quantal dopamine secretion in carotid body glomus cells. Proc. Natl. Acad. Sci. USA, 91: 10208-10211. Van Genechten, J., Brouns, I., Burnstock, G., Timmermans, J.P., and Adriaensen, D. 2004. Quantification of neuroepithelial bodies and their innervation in fawn-hooded and wistar rat lungs. Am. J. Respir. Cell Mol. Biol., 30: 20-30. Van Lommel, A., and Lauweryns, J.M. 1993a. Neuroepithelial bodies in the fawn hooded rat lungs: morphological and neuroanatomical evidence for a sensory innervation. J. Anat., 183: 553-566. Van Lommel, A., and Lauweryns, J.M. 1993b. Ultrastructure and innervation of neuroepithelial bodies in the lungs of newborn cats. Anat. Rec., 236: 181-190. Van Lommel, A., Lauweryns, J.M., and Berthoud, H.R. 1998. Pulmonary neuroepithelial bodies are innervated by vagal afferent nerves: an investigation with in vivo anterograde DiI tracing and confocal microscopy. Anat. Embryol., 197: 325-330. Varas, R., Alcayaga, J., and Iturriaga, R. 2003. ACh and ATP mediate excitatory transmission in cat carotid body identified chemosensory units in vitro. Brain Res., 988: 154-163. Verna, A., Roumy, M., and Leitner, L.M. 1975. Loss of chemoreceptive properties of the rabbit carotid body after destruction of the glomus cells. Brain Res., 100: 13-23. Wada, C., Hashimoto, C., Kameya, T., Yamaguchi, K., and Ono, M. 1988. Developmentally regulated expression of the calcitonin gene related peptide (CGRP) in rat lung endocrine cells. Virch. Arch. B Cell. Pathol., 55: 217-223. Wang, Z.Y., and Bisgard, G.E. 2005. Postnatal growth of the carotid body. Resp. Physiol. Neurobiol., 149: 181-190. Wang, D., Youngson, C., Wong, V., Yeger, H., Dinauer, M., Vega-Saenz De Miera, E., Rudy, B., and Cutz, E. 1996. NADPH-oxidase and a hydrogen peroxide-sensitive K+ channel may function as an oxygen sensor complex in airway chemoreceptors and small cell lung carcinoma cell lines. Proc. Natl. Acad. Sci. USA, 93: 13182-13187. Wasano, K., and Yamamoto, T. 1981. A scanning and transmission electron-microscopic study on neuro-epithelial bodies in the neonatal mouse lung. Cell Tissue Res., 216: 481-490. Wasicko, M.J., Sterni, L.M., Bamford, O.S., Montrose, M.H., and Carroll, J.L. 1999. Resetting and postnatal maturation of oxygen chemosensitivity in rat carotid chemoreceptor cells. J. Physiol., 514: 493-503. Wyatt, C.N., and Buckler, K.J. 2004. The effect of mitochondrial inhibitors on membrane currents in isolated neonatal rat carotid body type I cells. J. Physiol., 556: 175-191. Wyatt, C.N., and Evans, A.M. 2007. AMP-activated protein kinase and chemotransduction in the carotid body. Resp. Physiol. Neurobiol., 157: 22-29. Wyatt, C.N., Wright, C., Bee, D., and Peers, C. 1995. O2-sensitive K+ currents in carotid-body chemoreceptor cells from normoxic and chronically hypoxic rats and their roles in hypoxic chemotransduction. Proc. Natl. Acad. Sci. USA, 92: 295-299.

358  |  Airway Chemoreceptors in the Vertebrates Xu, J., Xu, F., Tse, F.W., and Tse, A. 2005. ATP inhibits the hypoxia response in type I cells of rat carotid bodies. J. Neurochem., 92: 1419-1430. Yamamoto, Y., Kummer, W., Atoji, Y., and Suzuki, Y. 2002. TASK-1, TASK-2, TASK-3 and TRAAK immunoreactivities in the rat carotid body. Brain Res., 950: 304-307. Youngson, C., Nurse, C., Yeger, H., Curnutte, J.T., Vollmer, C., Wong, V., and Cutz, E. 1997. Immunocytochemical localization of O2-sensing protein (NADPH oxidase) in chemoreceptor cells. Microsc. Res. Tech., 37: 101-106. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature, 365: 153-155. Zhang, M., Zhong, H., Vollmer, C., and Nurse, C.A. 2000. Co-release of ATP and ACh mediates hypoxic signalling at rat carotid body chemoreceptors. J. Physiol., 525: 143-158.

Solitary Chemosensory Cells in the Airways of Mammals: Distribution, Immunocytochemistry, Fine Structure and Function 16. Solitary Chemosensory Cells in the Airways of Mammals  359-376 A. Sbarbati, M.P. Cecchini, C. Crescimanno, F. Merigo, D. Benati, M. Tizzano and F. Osculati 17. Solitary Chemosensory Cells: Phylogeny and Ontogeny Anne Hansen and Thomas E. Finger

377-388

18. Functional Importance of Pulmonary Neuroendocrine Cells 389-402 Staffan Skogvall 19. CO2/H+ Chemoreceptors in the Respiratory Passages of Vertebrates K.M. Gilmour and W.K. Milsom

403-426

16 Solitary Chemosensory Cells in the Airways of Mammals A. Sbarbati1*, M.P. Cecchini1, C. Crescimanno2, F. Merigo1, D. Benati1, M. Tizzano3 and F. Osculati4

Abstract Solitary chemosensory cells (SCCs) are present in the epidermis, oropharynx and gills of most primary aquatic vertebrates. They resemble taste bud cells but are distinguished from them by fine structural characteristics and by the fact that they do not aggregate into groups.   Mammalian SCCs were first described in rats in 1998 and their presence was confirmed by further studies. Immunoreactivity for the G-protein subunit α-gustducin was found in SCCs as in taste bud cells. A recent study revealed the presence of SCCs in the respiratory apparatus of a large mammalian species (Bos Taurus) in which the airway cytology is similar to that of humans.   The few findings in humans concern the nasal cavity, where cells with the morphology of SCCs have been found. These cells express Trp M5 and a subset of them also express gustducin, calbindin, and/or vesicular acetylcholine transporter. These early findings raise questions about the possible role of SCCs in the control of complex functions (e.g., airway surface liquid secretion or innate immunity), and about the involvement of chemoreceptors in respiratory diseases. Pharmacological action on SCCs could be important in the treatment of respiratory pathologies and might open new horizons in drug discovery.

Keywords: Solitary chemosensory cells (SCCs), large mammals, α-gustducin, defence, drug discovery Department of Morphological-Biomedical Sciences, Anatomy and Histology Section, University of Verona, Medical Faculty, Italy. 2 Faculty of Exercise and Sport Science, Kore University, Enna, Italy. 3 Rocky Mountain Taste and Smell Center, Department of Cell and Developmental Biology, University of Colorado, Denver, Aurora, USA. 4 IRCCS Centro Neurolesi “Bonino-Pulejo” Messina, Italy. 1

*Auhtor for Correspondence.

360  |  Airway Chemoreceptors in the Vertebrates

Introduction Some records of the existence of a system of differentiated epithelial sensory cells in fish can be found in the 19th century literature. For instance, it has been known since 1872 that the free pectoral fin rays of triglids did not bear taste buds; there were a few reports of the presence of sensory cells, among which the study by Morril (1895) appears to be the most accurate (Whitear, 1971). In 1876, Foettinger and Langgerhans independently described bipolar cells in the epidermis of Lampetra fluviatilis and Lampetra planeri (Fahrenholz, 1936), some of which correspond to the chemosensory cells investigated later (Whitear and Lane, 1983). In ranid tadpoles, Kölliker (1886) found Stiftchenzellen, which were the subject of later histological and fine-structural studies (Meyer, 1962; Whitear, 1976). Bipolar cells in teleost epidermis were stained intra-vitam with methylene blue by Whitear (1952), but it was not until these cells could be identified by electron microscopy that their resemblance to gustatory cells of the corresponding species could be appreciated and their innervation confirmed (Whitear, 1965, 1971). These studies showed that the skin, the gills and the oropharyngeal surfaces of primary aquatic vertebrates are provided with a diffuse system of chemoreceptors, not organized into discrete end organs, which are related to but distinct from the gustatory system. In recent years, several groups have contributed to a precise characterization of the function of SCCs (also called solitary chemoreceptor cells; Finger et al., 2003) in fish. This chemosense may be used for feeding or predator avoidance (Finger, 1997). The SCC system responds to mucus substances and may serve as a predator detector (Peters et al., 1991). Some studies (Osculati and Sbarbati, 1995) reported the same fine-structural resemblance of SCCs to taste bud cells that had already been described in fish, in the oral cavity of amphibia. Cellular elements with ultrastructural features resembling those of SCCs (i.e., bipolar cells with microvilli, apical vesicles, packed cisternae of smooth endoplasmic reticulum and nerve contact) were found in the frog (Rana esculenta) taste disk, in which the presence of three different neuroepithelial systems was demonstrated (Osculati and Sbarbati, 1995). For several years, SCCs were considered typical of aquatic vertebrates and were thought to have disappeared with the evolution from aquatic to terrestrial forms of vertebrates. Later, these cells were also described in a mammal, the rat (Sbarbati et al., 1998), and the presence of the taste cell-related G-protein subunit α-gustducin confirmed a similarity with taste cells (Sbarbati et al., 1998); these isolated cells were found in the vallate papillae of the rat tongue during the first days of extrauterine life (Figure 1). The presence of SCCs in mammals was subsequently confirmed by other studies (Zancanaro et al., 1999; El-Sharaby et al., 2001a, 2001b). A later investigation (Tizzano et al., 2006) was carried out in a large mammalian species—cows and bulls (Bos Taurus)—in which the airway cytology is similar to that

Solitary Chemosensory Cells in the Airways of Mammals  |  361

Figure 1: SCC in the developing gustatory epithelium of the rat tongue. Alpha-gustducin immunostaining (300x) (Sbarbati and Osculati, 2005b).

of humans. In this study the presence of SCCs was demonstrated in the epithelium of the arythenoids, in the trachea and the bronchi, and α-gustducin-immunoreactive SCCs were also frequently found (Figure 2, 3). Despite this morphological evidence, the first functional data were not obtained until 2003, when it was demonstrated that SCCs localized in the nasal cavity operate in the detection of irritants (Finger et al., 2003). Recently, the same group also proved that nasal SCCs proliferate and undergo rapid turnover (Gulbransen and Finger, 2005). In the last two years, research into SCCs has opened exciting new possibilities. It has been shown that SCCs are present throughout the airways, both in the larynx (Sbarbati et al., 2004a, 2004b) (Figure 4) and in the trachea/bronchi (Merigo et al., 2005). The presence of a diffuse chemosensory system in the airways raises questions about the role of chemoreceptors in the control of complex functions (e.g., airway surface liquid secretion) and about the involvement of chemoreceptors in respiratory diseases.

The Chemoreceptor Molecular Cascade The availability of markers for chemoreceptive elements made it possible to study these cells in various organs; the principal markers derive from the chemoreceptor molecular cascade. Olfactory receptor neurons and SCCs both utilize signal transduction cascades involving different G-proteins. Gustducin is a heterotrimeric guaninenucleotide binding protein (G protein), the existence of which was first demonstrated in rats (McLaughlin et al., 1992) and then confirmed in man (Takami et al., 1994).

362  |  Airway Chemoreceptors in the Vertebrates

Figure 2: Sections taken from circumvallate papillae (a-c), fungiform papillae (d-f ), arytenoids (g-i) and various regions of the respiratory tract (trachea, l-n; bronchi, o-q) of adult cows and bulls, showing α-gustducin-immunoreactive taste bud cells and SCCs. Scale bars=150μm (a), 100 μm (d, e, g, h), 50μm (b, c, f, i), 5μm (l, q) (Tizzano et al., 2006).

Figure 3: Immunoblotting of bovine tissue reacted with the polyclonal anti-α-gustducin antibody. 1, circumvallate papillae; 2, epiglottis; 3, arytenoids; 4, trachea; 5, kidney; 6, bronchi. (Tizzano et al., 2006).

Solitary Chemosensory Cells in the Airways of Mammals  |  363

Figure 4: BC/SCC in the rat larynx (13,000x) (Sbarbati and Osculati, 2005b).

Although in the original studies gustducin was considered to be specific to a subset of taste cells, immunoreactivity for α-gustducin was later found in brush cells of the digestive apparatus (Höfer and Drenckhahn, 1996; Höfer et al., 1996, 1999), in SCCs, and in the vomeronasal organ. So several studies have demonstrated that gustducin is a potent marker of chemosensitive cells, which are localized in several organs in the form of isolated elements, clusters or taste buds. In newborn rats, isolated α-gustducin immunoreactive cells were found within the epithelium (Figure 1). In the following days, small taste buds appeared, but isolated, bipolar-shaped α-gustducin immunoreactive cells were also found. In the vallate papilla of newborn rats, the presence of SCCs is paralleled by a rapid development of intrinsic neurones and nerves (Sbarbati et al., 2000, 2002). This seems to suggest that an anatomical and functional relationship exists between the chemosensory cells and the intrinsic nervous system of the developing gustatory apparatus. There is some evidence in humans for the presence, during early tongue ontogenesis, of individual slender cells that are immunopositive for cytokeratin 20, an intermediate filament protein found exclusively in taste buds and epidermal Merkel cells (Witt et al., 2003). Moreover, Hansen has found unconventional sensory cells in the human

364  |  Airway Chemoreceptors in the Vertebrates nasal epithelia; subsets of these cells express gustducin, calbindin and/or vesicular acetylcoline transporter (VACht) (Hansen et al., 2005; Hansen, 2006).

SCCs in the Human Nasal Cavity There have been few findings in humans so far. Recent studies revealed a possible receptor cell in human and rodent olfactory epithelium. Also, several potential chemosensory cell types were found in the respiratory epithelium using electron microscopy. Immunocytochemical experiments showed cell types positive for gustducin, calbindin, and/or VAchT that closely resembled rodent SCCs (Hansen et al., 2005). These cells have the morphology of SCCs and express Trp M5. Subsets of these cells express gustducin, calbindin, and/or VAChT. These findings suggest the existence of possible unconventional receptor cell types in the main olfactory epithelium and in the respiratory epithelium of rodents and humans (Hansen, 2006)

SCCs and Brush Cells An open question is the relationship between SCCs and brush cells (BCs). Over the past 50 years, hundreds of studies have described cells characterized by a brush of rigid apical microvilli with long rootlets that are found in the digestive and respiratory apparatuses. These cells have been given names such as brush cells, tuft cells, fibrillovesicular cells, multivesicular cells, and caveolated cells (Figure 5). The first description of BCs is generally attributed to Rhodin and Dalhamn (1956) in their study of the rat trachea. The presence of BCs was later confirmed in the airways of several species, including humans (Rhodin, 1959). Nine years after this discovery, it was realized that BCs are also present in the lung (Meyrick and Reid, 1968) and in the digestive apparatus (Luciano et al., 1968a, 1968b) (Figure 6). Then BCs were discovered to be mainly concentrated in the gall bladder (Luciano and Reale, 1969). However, isolated cells with the ultrastructural features of BCs had previously been identified in several organs and indicated with different names, so that a systematic review of the literature is difficult. New data suggest that the concept of BCs must be changed, and that their ultrastructure, together with their neurochemistry and molecular biology, could provide a more adequate definition. The description of gustducin (Höfer and Drenckhahn, 1998) and other bitter-taste related molecules in these elements localized in the digestive and respiratory apparatuses demonstrated a link between these cells and elements of taste buds. And recent results strongly support the idea that BCs may operate as solitary chemoreceptors, probably representing a subfamily of SCCs localized in specific microenvironments (Sbarbati and Osculati, 2005a).

Solitary Chemosensory Cells in the Airways of Mammals  |  365

Figure 5: Schematic images of a BC and a SCC (Sbarbati and Osculati, 2005b). FS, free surface; BL, basal lamina; R, roots of the microvilli, V, vesicles; F, filaments; G, Golgi complex.

Figure 6: Apical extremity of a BC/SCC in the human duodenum (40,000x) (Sbarbati and Osculati, 2005b).

366  |  Airway Chemoreceptors in the Vertebrates

Distribution of Sccs and Bcs in Mammals SCCs In fish, SCCs occur in the oropharynx, gills and skin and have often been found in association with taste buds. Among amphibia, a diffuse chemosensory system (DCS) has been described on the ventral skin of toads, and a structural resemblance of SCCs to taste buds cells has been reported in frogs (Osculati and Sbarbati, 1995). SCCs have also been described in mammals at specific sites in the digestive or respiratory apparatus (Sbarbati and Osculati, 2003; Sbarbati et al., 1998; Finger et al., 2003). They were found in the nasal cavity, on the rat tongue in the vallate papilla, in the rat airways, both in the larynx and the trachea/bronchi, and in large mammalian airways (Tizzano et al., 2006). In humans, they have so far been found in the nasal cavity (Hansen et al., 2005; Hansen, 2006).

BCs In the airways, after their description in the rat trachea (Rhodin and Dalhamn, 1956), the presence of BCs was confirmed in the rabbit (Leeson, 1961). In subsequent years, several studies found them in different mammalian species (Rhodin, 1966; Luciano et al., 1968a, 1968b; Jeffery and Reid, 1975; Ishida, 1977; Taira and Shibasaki, 1978; Christensen et al., 1987). They were seen in the nasal cavity, in particular in the non-sensory epithelium of the rat vomeronasal organ (Höfer et al., 2000), and in the alveolar epithelium of the lung (where they are also called type III pneumocytes). Several studies also confirmed the presence of BCs in the human respiratory apparatus (Rhodin, 1959). In the digestive apparatus, they have been found in the salivary ducts of rat salivary glands (Sato and Miyoshi, 1996, 1997; Sato et al., 2000), in the stomach (Luciano et al., 1980), and especially in the gall bladder (Luciano and Reale, 1969, 1979, 1990; Luciano et al., 1981). These studies demonstrated that BCs are the second most frequent cellular component of the epithelium of the mouse gall bladder. Intestinal BCs were first described in the rat by Luciano et al. (1968a, 1968b) and later found in the major pancreatic excretory ducts. From this “overview” it is clear that these cells are widely distributed throughout several organs. They seem to configure a DCS that covers large areas, having analogies with the chemosensory system described in aquatic vertebrates.

Morphofunctional Considerations SCCs are slender epithelial elements, recently discovered in mammals, which display cytological characteristics suggesting a chemosensory role and possess signalling

Solitary Chemosensory Cells in the Airways of Mammals  |  367

mechanisms typical of taste cells (Sbarbati et al., 1998, 2004a; Finger et al., 2003). In fish, SCCs form a system of differentiated sensory epithelial cells, which are not organized into discrete end organs and which may occur in the epithelia of the oropharynx, the gills, and the skin (Whitear, 1992). The elementary unit of this chemoreceptor structure differs from that of taste buds of the gustatory apparatus, consisting of a single, bipolar epithelial cell contacted by nerves and lacking a specialized connective bed. Glial-like epithelial cells can surround SCCs. In general, the form of the cell depends on its location: in a thin epidermis, the cell body may be inclined to one side. Where the epidermis is thick, the nucleus of the sensory cell often lies at the level of the second tier of epithelial cells from the surface, but in other situations the cell may be elongated, with its deep pole immediately above the basal layer of the epidermis. Usually the apical process is of sufficient length to raise the presumed receptive membrane above the mucus covering the surface of the epithelium (Figure 5). Common ultrastructural features of SCCs include spindle shape, basal synapses, abundant endoplasmic reticulum within the proximal part of the cell, and an apical microvillus. The distal process of an SCC contains a distinct Golgi apparatus and characteristic vesicles (Whitear and Kotrschal, 1988). Electrophysiological recordings supported the hypothesis that SCCs are chemosensory (Peters et al., 1991) and that they respond to predator-avoidance or food-related stimuli, although they do not respond to some typical taste stimuli (Silver and Finger, 1984). In mammals, a specific set of SCCs associated with the gustatory epithelium has been described. During the first days of post-natal life, the epithelia of the rat circumvallate papilla contain isolated cells with a bipolar shape, nerve contacts and neuroendocrine-type granules (Sbarbati et al., 1998). Ultrastructural features of these cells prove that they are epithelial elements, suggesting that they could be homologous to the SCCs described in aquatic vertebrates. The presence of SCCs in the rat circumvallate papilla during the first week of post-natal life was also studied using α-gustducin immunocytochemistry (Sbarbati et al., 1999). Within the non-olfactory nasal epithelium of mammals, SCCs are morphologically similar to the individual cells in taste buds, but unlike taste cells, they form distinct synapses on to cutaneous nerve fibres of the trigeminal nerve (Finger et al., 2003). Like olfactory receptor neurons, which are replaced throughout the life of the animal, SCCs undergo continuous proliferation in the adult animal (Gulbransen and Finger, 2005).

SCCs: Homology in Different Species To establish homology among SCCs in fish, amphibians and mammals is difficult, partly because these cells form heterogeneous systems. So far, findings in mammals

368  |  Airway Chemoreceptors in the Vertebrates have fully confirmed previous findings in fish about the general morphology of SCCs, despite the fact that in mammals SCCs seem to be used as internal rather than as external chemoreceptors. In the oral cavity, the homology between SCCs described in the different species is evident, even if the relationship with the taste system requires further clarification. It is interesting to note that SCCs develop before taste buds in mammals, and it has been shown that in fish also they appear earlier in ontogenesis than taste buds (Hansen et al., 2002; Kotrschal et al., 1997).

Specific Laryngeal Sensory Epithelium A specific laryngeal sensory epithelium that includes arrays of solitary chemoreceptor cells has recently been described in the supraglottic region of the rat (Sbarbati et al., 2004a). These SCCs lie in this specific epithelium together with taste buds. Recently, Finger et al. (2005) demonstrated that taste buds are clearly innervated by nerve fibres immunoreactive for purinergic receptors and that stimulation of taste buds in vitro evokes release of ATP. So ATP fulfils the criteria for a neurotransmitter linking taste buds to the nervous system. On the other hand, laryngeal solitary chemoreceptor cells are not innervated by purinergic nerve fibres, although such fibres do innervate nearby epithelium. This indicates that nerve fibres that innervate laryngeal SCCs use a different neurotransmitter and/or receptor system (Finger et al., 2005). With immunocytochemistry, laryngeal immunoreactivity for α-gustducin was found to be mainly localized in SCCs (Sbarbati et al., 2004a) (Figure 7).

Laryngeal Chemosensory Clusters A new form of chemosensory clusters reported by Sbarbati et al. (2004b) are multicellular organizations that differ from taste buds, being generally composed of 2-3 chemoreceptor cells (Sbarbati et al., 2004b). Compared with lingual taste buds, chemosensory clusters show reduced height and smaller diameter. In laryngeal chemosensory clusters, immunocytochemistry using antibodies against either α-gustducin or phospholipase β2 (PLC β2) identified a similar cytotype. PLC β2 is expressed in a subset of cells within mammalian taste buds. This enzyme is believed to be a marker for gustatory sensory receptor cells (Rossler et al., 1998; Kim et al., 2006). The demonstration of the existence of chemosensory clusters strengthens the hypothesis of a phylogenetic link between gustatory and solitary chemosensory cells. Due to their structure and location, chemosensory clusters seem to represent the missing link between buds and SCCs. Laryngeal chemosensory clusters appear to be

Solitary Chemosensory Cells in the Airways of Mammals  |  369

Figure 7: Staining pattern obtained in the specific laryngeal sensory epithelium by α-gustducin (A, B1-3), phospholipase C β2 (PLCβ2; C1-3), and protein gene product (PGP) 9.5 (D1-3) immunocytochemistry. Light microscopy images were obtained from free-floating sections that were subsequently observed by electron microscopy. Scale bars=50µ in A; 5µ in B1, B3, C3; 10µ in B2; 2.5µ in C1, C2, D2, D3; 15µ in D1 (Sbarbati et al., 2004a).

a transitional structure between the rostrally located buds and SCCs, which are more distally located in specific areas of the larynx (Sbarbati et al., 2004a). The existence of such clusters strengthens the analogy between the chemoreceptor system of the larynx and that previously reported in the skin and oral cavity of fish (Sbarbati et al. 2004a).

370  |  Airway Chemoreceptors in the Vertebrates

Secretory Cells of the Airway and the Chemoreceptive Cascade Airway secretion is maintained by specialized non-ciliated epithelial cells whose phenotype varies with their topographical location. In recent studies (Merigo et al., 2007), immunohistochemical analyses were performed to evaluate whether the secretory cells themselves possess chemoreceptive capability. Antibodies against α-gustducin and PLC β2 were employed as chemoreceptive markers and, as secretory markers, CC10 and CC26 antibodies against Clara-cell-specific secretory proteins. A marker of chloride secretion, cystic fibrosis transmembrane regulator, was used to characterize the secretory cell type. This investigation found α-gustducin localized in non-ciliated cells of the epithelium lining the trachea and bronchioles of adult rats, where it was co-expressed with CC10 and CC26. So non-ciliated epithelial cells of the rat airway express components of distinct signalling mechanisms, which suggests that secretory events are driven by a molecular mechanism activated by the binding of luminal substances to G-proteincoupled receptors. The authors assume that those receptors play a role in activating signalling pathways linked to the production of secreta.

Possible Role of DcS in Antimicrobial Defence The presence of these chemoreceptors in epithelia of entodermal origin suggests the existence of a DCS sharing common signalling mechanisms with the “classic” taste organs. The elements of this taste cell-related DCS display a site-related morphologic polymorphism, and in the past they have been indicated with a variety of names. The demonstration of a DCS in the airways raises questions about the role of chemoreceptors in the control of complex functions (e.g., airway surface liquid secretion) and about the involvement of chemoreceptors in respiratory diseases. The chemoreceptive capacity of the DCS seems to protect against exogenous substances. In addition, recently published data suggest that the DCS could have an important role in defence against bacteria. The elements of this system are located in an optimal position to intercept the exchange of information between bacteria (Sbarbati and Osculati, 2006). There is a growing body of evidence that several bacterial species operate a quorum-sensing strategy (Kolter, 2005). Briefly, these bacteria coordinate their activities using extracellular signals, i.e., auto-inducers or pheromones (Hardman et al., 1998). When such compounds reach a sufficient concentration (i.e., when the total population is large enough), the bacteria activate genetic pathways often involved in the initiation of aggressive behaviour. Quorum sensing appears therefore to be a strategy used by bacteria to coordinate their activities, and it is based on the release of small molecules, generally proteins or acyl-lactones. These findings suggest that a “war

Solitary Chemosensory Cells in the Airways of Mammals  |  371

of communication” takes place on the mucosal surfaces of the digestive and respiratory systems, with two chemosensory systems in opposite camps (Sbarbati and Osculati, 2006; Sbarbati, 2006). Due to its structural and biochemical characteristics, the DCS appears to be able to intercept communication among bacteria and predict their movements. If messages are indeed detected in this way, it may be that the organism mounts a highly localized and efficient response to bacterial activation. This would be based on defences such as the quenching of auto-inducers, the dilution or removal of bacteria, or secretion of antibiotic agents, and it might precede and avoid the need for intervention by immune cells.

Conclusions Several questions remain about SCCs and about the physiology and morphology of the DCS. Links between the molecular mechanisms of the taste and secretory apparatuses have not yet been studied, and the existence of BCs not containing α-gustducin raises the possibility of alternative G-proteins. Such questions could be answered by a detailed chemical code for the different elements of the DCS. The DCS seems to be a potential target for new drugs, because there are several indications that information obtained by this system induces secretory reflexes. Therefore, modulation of the respiratory and digestive apparatuses by substances acting on their chemoreceptors could be important in the treatment of diseases such as cystic fibrosis and asthma and might open new horizons in drug discovery.

Acknowledgement The authors thank Christine Harris for revising the manuscript.

Referenes Christensen, T.G., Breuer, R., Hornstra, L.J., Lucey, E.C., Stone, P.J., and Snider, G.L. 1987. The ultrastructure of hamster bronchial epithelium. Exp. Lung Res., 13(3): 253-277. El-Sharaby, A., Ueda, K., Kurisu, K., and Wakisaka, S. 2001a. Development and maturation of taste buds of the palatal epithelium of the rat: histological and immunohistochemical study. Anat. Res., 263: 260-268. El-Sharaby, A., Ueda, K., and Wakisaka, S. 2001b. Differentiation of the lingual and palatal gustatory epithelium of the rat as revealed by immunohistochemistry of alpha-gustducin. Arch. Histol. Cytol., 64: 401-409. Fahrenholz, C. 1936. Die sensiblen Einrichtungen der Neunaugenhaut. Z. Mikrosk. Anat. Forsch., 40: 323-380.

372  |  Airway Chemoreceptors in the Vertebrates Finger, T.E. 1997. Evolution of taste and solitary chemoreceptor cell system. Brain Behav. Evol., 50: 234-243. Finger, T.E., Böttger, B., Hansen, A., Anderson, K.T., Alimohammadi, H., and Silver, W.L. 2003. Solitary chemoreceptor cells in the nasal cavity serve as sentinels of respiration. Proc. Natl. Acad. Sci. USA, 100: 8981-8986. Finger, T.E., Danilova, V., Barrows, J., Bartel, D.L., Vigers, A.J., Stone, L., Hellekant, G., and Kinnamon, S.C. 2005. ATP signaling is crucial for communication from taste buds to gustatory nerves. Science, 310: 1495-1499. Gulbransen, B.D., and Finger, T.E. 2005. Solitary chemoreceptor cell proliferation in adult nasal epithelium. J. Neurocytol., 34: 117-122. Hansen, A. 2006. Unconventional sensory cells in the nasal epithelia of rodents and humans. Chem. Senses, 31: E4. Hansen, A., Reutter, K., and Zeiske, E. 2002. Taste bud development in the zebrafish, Danio Rerio. Dev. Dyn., 223(4): 483-496. Hansen, A., Witt, M., and Hummel, T. 2005. Unconventional neurons in the nasal cavity of humans. Chem. Senses, 30: A55. Hardman, A.M., Tewart, G.S., and Williams, P. 1998. Quorum sensing and the cell-cell communication dependent regulation of gene expression in pathogenic and non-pathogenic bacteria. Antonie Van Leeuwenhoek, 74: 199-210. Höfer, D., and Drenckhahn, D. 1996. Cytoskeletal markers allowing discrimination between brush and other epithelial cells of the gut including enteroendocrine cells. Histochem. Cell Biol., 105(5): 405-412. Höfer, D., and Drenckhahn, D. 1998. Identification of the taste cell G-protein alpha-gustducin in brush cells of the rat pancreatic duct system. Histochem. Cell Biol., 110: 303-309. Höfer, D., Püschel, B., and Drenckhahn, D. 1996. Taste receptor-like cells in the rat gut identified by expression of α-gustducin. Proc. Natl. Acad. Sci. USA, 93: 6631-6634. Höfer, D., San, E., and Drenckhahn, D. 1999. Chemosensory perception in gut. New Physiol. Sci., 14: 18-23. Höfer, D., Shin D.W., Drenckhahn, D. 2000. Identification of cytoskeletal mackers for the different microvilli and cell types of the rat vomeronasal sensory epithelium. J. Neurocytol., 29:147-156. Ishida, H. 1977. Fine structural study on the postnatal development of the rat tracheal mucosa with special reference to the brush cells. Yokohama Med. Bull., 28: 123-147. Jeffery, P.K., and Reid, L. 1975. New observations of rat airway epithelium: a quantitative and electron microscopic study. J. Anat., 120: 295-320. Kim, J.W., Roberts, C., Maruyama, Y., Berg, S., Roper, S., and Chaudari, N. 2006. Faithful expression of GFP from the PLCβ2 promoter in a functional class of taste receptor cells. Chem. Senses, 31(3): 213-219. Kölliker, A. 1886. Histologishe Studien an Batrachierlauen. Z. Wiss, Zool., 43:1-40. Kolter, R. 2005. Surfacing views of biofilm biology. Trends Microbiol., 13(1): 1-2. Kotrschal, K., Whitear, M., and Finger, T.E. 1993. Spinal and facial innervation of the skin in the gadid fish Ciliata mustela (teleostei). J. Comp. Neurol., 331: 407-417. Kotrschal, K., Krautgartner, W.D., and Hansen, A. 1997. Ontogeny of the solitary chemosensory cells in the zebrafish, Danio Rerio. Chem. Senses, 22: 111-118. Leeson, T.S. 1961. The development of the trachea in the rabbit with particular reference to its fine structure. Anat. Anz., 110: 214-223.

Solitary Chemosensory Cells in the Airways of Mammals  |  373 Luciano, L., Reale, E., and Ruska, H. 1968a. On a “chemoreceptive” sensory cell in the trachea of the rat. Z. Zellforsch. Mikrosk. Anat., 85: 350-375. Luciano, L., Reale, E., and Ruska, H. 1968b. On a glycogen containing brush cell in the rectum of the rat. Z. Zellforsch. Mikrosk. Anat., 191: 153-158. Luciano, L., Reale, E., Ruska, H. et al. 1969. Brush cells in the alveolar epithelium of the rat lung. Z. Zellforsch. Mikrosk. Anat., 95(2): 198-201. Luciano, L., Reale, E. 1969. A new cell type (Brush cell) in the gallbladder epithelium of the mouse. J. Submicrose. Cytol., 1:37-44. Luciano, L., Reale, E. 1979. A new morphological aspect of the brush cells of the mouse gallbladder epithelium. Cell Tissue Res., 201: 37-44. Luciano, L., Reale E., and von Engelhardt, W. 1980. The fine structure of the stomach mucosa of the Llama II the fundic region of the hind stomach. Cell Tissue Res., 208(2): 207-228. Luciano, L., Castellucci, M., and Reale, E. 1981. The brush cells of the common bile duct of the rat. This section, freeze-fracture and scanning electron microscopy. Cell Tissue Res., 218(2): 403-420. Luciano, L. Reale, E. 1990. Brush cells of the mouse gallbladder. A correlative light and electronmicroscopical steady. Cell Tissue Res., 262:239-349. McLaugghlin, S.K., McKinnon, P.J., and Margolskee, R.F. 1992. Gustducin is a taste cell specific G protein close related to the transducin. Nature, 357: 563-569. Merigo, F., Benati, D., Tizzano, M., Osculati, F., and Sbarbati, A. 2005. α-gustducin immunoreactivity in the airways. Cell Tissue Res., 319: 211-219. Merigo, F., Benati, D., Di Chio, M., Osculati, F., and Sbarbati, A. 2007. Secretory cells of the airway express molecules of the chemoreceptive cascade. Cell Tissue Res., Feb. 327(2): 231-247. Meyer, M. 1962. Kegel und andere Sonderzellen der larvalen Epidermis von Froschlurchen. Z. Mikrosk. Anat. Forsch., 68: 79-131. Meyrick, B., and Reid, L. 1968. The alveolar brush cell in rat lung- a third pneumocyte. J. Ultrastruct. Res., 23(1): 71-80. Morrill, A.D. 1895. The pectoral appendages of Prionotus and their innervation. J. Morph., 11: 177-192. Osculati, F., and Sbarbati, A. 1995. The frog taste disk: aprototype of the vertebrate gustatory organ. Prog. Neurobiol., 46: 351-399. Peters, R.C., Kotrschal, K., and Krautgartner, W.D. 1991. Solitary chemoreceptor cells of Ciliata mustela (Gadidae, teleostei) are tuned to mucoid stimuli. Chem. Senses, 16: 31-42. Rossler, P., Kroner, C., Freitag, J., Noe, J., and Breer, H. 1998. Identification of a phospholipase Cβ subtype in rat taste cells. Eur. J. Cell Biol., 77: 253-261. Rhodin, J. 1959. Ultrastructure of the tracheal ciliated mucosa in rat and man. Ann. Otol. Rhinol. Laryngol., 68: 964-974. Rhodin, J.A. 1966. The ciliated cell-ultrastructure and function of the human tracheal mucosa. Am. Rev. Respir. Dis., 93 (suppl.): 1-5. Rhodin, J., and Dalhalmn, T. 1956. Electron microscopy of the tracheal ciliated mucosa in rat. Z. Zellforsch. Mikrosk. Anat., 44: 345-412. Sato, A., and Miyoshi, S. 1996. Tuft cells in the main excretory duct epithelia of the three major rat salivary glands. Eur. J. Morph., 34: 225-228. Sato, A., Miyoshi, S. 1997. Fine structure of teeft cells of the main excretory duct epithelium in the eat submandibular gland. Anat. Rec., 248:325-331. Sato, A., Suganuma, T., Ide, S., Kawano, J., and Nagato, T. 2000. Tuft cells in the main excretory duct of the rat submandibular gland. Eur. J. Morph., 38(4): 227-231.

374  |  Airway Chemoreceptors in the Vertebrates Sbarbati, A. 2006. Antimicrobial intelligence: how eukaryotes combat the quorum sensing strategy. 16th European Congress of Clinical Microbiology and Infectious Diseases. Keynote lecture. Nice, April 1-4. Clin. Microbiol. Infect., 12(4): K243. Sbarbati, A., and Osculati, F. 2003. Solitary chemosensory cells in mammals? Cell Tissue Organs, 175: 51-55. Sbarbati, A., and Osculati, F., 2005a. A new fate for old cells: brush cells and related elements. J. Anat., 206(4): 349-358. Sbarbati, A., and Osculati, F. 2005b. The taste cell-related diffuse chemosensory system. Prog. Neurobiol., 75: 295-307. Sbarbati, A., and Osculati, F. 2006. Allelochemical communication in vertebrates: kairomones, allomones and synomones. Cell Tissue Org., 183(4): 206-219. Sbarbati, A., Crescimanno, C., Benati, D., and Osculati, F. 1998. Solitary chemosensory cells in the developing chemoreceptorial epithelium of the vallate papilla. J. Neurocytol., 27: 631635. Sbarbati, A., Crescimanno, C., Bernardi, P., and Osculati, F. 1999. Alpha-gustducin immunoreactivesolitary chemosensory cells in the developing chemoreceptorial epithelium of the rat vallate papilla. Chem. Senses, 24: 469-472. Sbarbati, A., Crescimanno, C., Bernardi, P., Benati, D., Merigo, F., and Osculati, F. 2000. Postnatal development of the intrinsic nervous system in the circumvallate papilla-von Ebner gland complex. Histochem. J., 32(8): 483-488. Sbarbati, A., Merigo, F., Bernardi, P., Crescimanno, C., Benati, D., and Osculati, F. 2002. Ganglion cells and topographically related nerves in the vallate papilla-von Ebner gland complex. J. Histochem. Cytochem., 50: 709-718. Sbarbati, A., Merigo, F., Benati, D., Tizzano, M., Bernardi, P., Crescimanno, C., and Osculati, F. 2004a. Identification and characterization of a specific sensory epithelium in the rat larynx. J. Comp. Neurol., 475: 188-201. Sbarbati, A., Merigo, F., Benati, D., Tizzano, M., Bernardi, P., and Osculati, F. 2004b. Laryngeal chemosensory clusters. Chem. Senses, 29: 683-692. Silver, W.L., and Finger, T.E. 1984. Electrophysiological examination of a non-olfactory, nongustatory chemosense in the sea robin, Prionotus carolinus. J. Comp. Physiol., A154: 167174. Taira, K., and Shibasaki, S. 1978. A fine structure study of the non-ciliated cells in the mouse tracheal epithelium with special reference to the relation of “brush cells” and “endocrine cells”. Arch. Histol. Jpn., 41: 351-366. Takami, S., Getchell, T.V., McLaughlin, S.K., Margolskee, R.F., and Getchell, M.L. 1994. Human taste cells express the G protein α-gustducin and neuron-specific enolase. Mol. Brain Res., 22(1-4): 193-203. Tizzano, M., Merigo, F., and Sbarbati, A. 2006. Evidence of solitary chemosensory cells in a large mammal: the diffuse chemosensory system in Bos Taurus airways. J. Anat., 209(3): 333-337. Whitear, M. 1952. The innervation of the skin of teleost fishes. Q. J. Micnojc. Sci., 93:289-305. Whitear, M. 1965. Presumed sensory cells in fish epidermis. Nature, 208: 703-704. Whitear, M. 1971. Cell specialization and sensory function in fish epidermis. J. Zool. Lond., 163: 237-264. Whitear, M. 1976. Identification of the epidermal “Stiftchenzellen” of frog tadpoles by electron microscopy. Cell Tissue Res., 175: 391-402.

Solitary Chemosensory Cells in the Airways of Mammals  |  375 Whitear, M. 1992. Solitary chemoreceptor cells. In: Chemoreception in Fishes. T.J. Hara (Ed.). Chapman and Hall, New York, pp. 103-125. Whitear, M., and Lane, E.B. 1983. Oligovillous cells of the epidermis: sensory elements of lamprey skin. J. Zool. Lond., 199: 359-384. Whitear, M., and Kotrschal, K. 1988. The chemosensory anterior dorsal fin in rocklings (Gaidopsarus and Ciliata, Teleostei, Gadidae): activity, fine structure and innervation. J. Zool., 216: 339-366. Witt, M., Reutter, K., and Miller, I.J. 2003. Morphology of the peripheral taste system. In: Handbook of Olfaction and Gustation. R.L. Doty (Ed.). Marcel Dekker, Inc., New York, pp. 651-677. Zancanaro, C., Caretta, C.M., Merigo, F., Cavaggioni, A., and Osculati, F. 1999. alpha-gustducin expression in the vomeronasal organ of the mouse. Eur. J. Neurosci., 11(12): 4473-4475.

17 Solitary Chemosensory Cells: Phylogeny and Ontogeny Anne Hansen* and Thomas E. Finger

Abstract Vertebrates possess not only the chemical senses of taste and smell, but also an additional chemosensory system mediated by solitary chemosensory cells. This population of cells is not confined to a specialized endorgan but scattered in various epithelia. First seen in fishes, these chemosensory cells are now also described in terrestrial as well as aquatic vertebrates. Here we relate the history of solitary chemosensory cells, their distribution and development. Furthermore, the phyletic origin and continuity of solitary chemosensory cells are discussed as well as their relationship to the taste system.

Keywords: Solitary chemosensory cell, evolution, development, fish, reptile, mammal, human

Introduction All organisms have means for detecting the chemical composition of their environment. Single cells, e.g. bacteria, utilize molecular receptors coupled to intracellular cascades that ultimately terminate in an effector mechanism (van Houten, 2000). Metazoan animals usually possess specialized receptor organs or cells within the epithelium

Dept. Cell & Devel. Biology, Univ. Colorado, Denver. Sch. Medicine, Mail Stop 8108, PO Box 6511, Aurora, CO 80045. *Author for Correspondence.

378  |  Airway Chemoreceptors in the Vertebrates with connection to the nervous system. For example, arthropods possess numerous chemosensory sensilla scattered across the body surface including legs, thorax and even wings. Most Metazoa have evolved a taste system characterized by specific chemosensory endorgans associated with the mouthparts for sampling potential food items prior to ingestion. Likewise, most complex animals, including invertebrates as well as vertebrates, possess a second specialized chemosensory modality (the olfactory sense) which is utilized in a social context (recognition of conspecifics) as well as for more generalized awareness of the environment, e.g. for detection of potential predators, identification of “home” territory, or alerting to the presence of danger or food. In the case of vertebrates, the senses of taste and smell each involve the use of specialized sensory endorgans, respectively taste buds and the olfactory organs, including the vomeronasal organ. Because of the specialized structure of the endorgans, the senses of taste and smell are considered “special” senses (Landacre, 1911). Yet the chemoreceptive capability of vertebrates does not end with the conventional special senses of taste and smell. Recent work from our laboratory and others have confirmed the earlier descriptions of a widespread population of scattered, solitary chemosensory receptor cells present within the external and internal epithelia of vertebrates. This chapter will provide an overview of these systems; a more detailed summary of the situation in mammals can be found in the work of Sbarbati in this volume.

Solitary Chemosensory Cells - History Solitary chemosensory cells (SCCs) constitute a third chemosensory modality in addition to the widely studied senses of taste and smell. These cells do not necessarily accumulate in specific endorgans but are scattered as single cells in various epithelia. First mentioned by Kölliker (1886) (“Stiftchenzellen”) for tadpoles and by Morrill (1895) for sea robins, SCCs were postulated to be chemosensory cells by Whitear (1965) based on her transmission electron microscopic studies. Since then, these cells have been described for a vast variety of fish species, including hagfish, lampreys, lungfish, elasmobranchs, sturgeons, and modern teleosts [for reviews see Kotrschal, 1991, 1996; Whitear, 1992; Braun, 1996; Kapoor and Finger, 2003; Hansen and Reutter, 2004; Hansen, 2005]. SCCs are specialized bipolar epithelial cells and thus secondary neurons. In general, they are spindle-shaped but there is some variation usually in accordance with the thickness of the surrounding epithelium (Figure 1). In thinner epithelia, SCCs may be roundish or inclined to one side, sometimes including a short “tail” pointing towards the basal lamina. The apical ending reaching above the surface of the epithelium varies according to the species and stage of life. It may present itself as a dome-shaped stout villus (alligator, human) (Hansen, 2006a, 2006b; Hansen, 2007) or one single long villus (adult zebrafish) (Kotrschal et al., 1997) or as several small villi that sit on a common base (goldfish, mice) (Hansen et al., 1999; Finger et al., 2003) (Figures 2, 3). The cell

Solitary Chemosensory Cells: Phylogeny and Ontogeny  |  379

bodies of SCCs contain numerous mitochondria, a Golgi system and longitudinally oriented microtubules sometimes associated with microfilaments. Vesicles are often abundant and sometimes also occur in the apical villus. Electron density and size varies according to species (e.g. 50 – 70 nm in diameter in cyprinids and silurids). The nucleus is lobulated or embayed (Whitear, 1992). The SCCs do not extend beyond the basal lamina of the epithelium, but form functional contacts with intraepithelial nerve fibers belonging to cranial or spinal nerves depending on the location of the SCC (Lane and Whitear, 1977; Lane, 1977; Whitear and Kotrschal, 1988; Kotrschal and Whitear, 1988; Kotrschal and Finger, 1996; Kotrschal et al., 1998). Often the nerve fibers form embracing, repeated contacts with the SCCs with numerous vesicles of varied sizes accumulating within the SCC at the point of contact with the nerve fiber.

Figure 1: Confocal images of SCCs in the nasal epithelium of mice. Gustducin-positive SCCs in the respiratory areas of the nasal cavity are different in shape. Depending on the area they are scarce (A) or abundant (B). Scale bar for A and B = 20 µm. Image B courtesy of Brian Gulbransen, University of Colorado Denver, School of Medicine.

Distribution In fishes, SCCs are scattered in almost all epithelia: across the outer body surface (Figures 2A, 2B), on fins, gills, in the oropharyngeal cavity and in some species even in the olfactory epithelium (Hansen et al., 1999) (Figure 3). SCCs are not evenly distributed across the body and their absolute numbers vary immensely among species. In some species, the SCCs are densely packed within specialized regions of epithelium, e.g. the anterior dorsal fin of rocklings (Gaidropsarus) or the pectoral fin rays of sea robins (Prionotus). Kotrschal and Adam (1984) counted densities of up to 1.0 x 105 SCCs per mm2 in Gaidropsarus mediterraneus and estimated a total number of 3 to 6 million SCCs for a rockling of 20 cm total length. Numbers of SCCs in other fish species are much lower but in any case much higher than the total number of taste buds cells (Kotrschal, 1991; Peters et al., 1991). SCCs may even be related to or assemble into specialized endorgans [Schreiner Organs (Braun, 1998)] in hagfish.

380  |  Airway Chemoreceptors in the Vertebrates

Figure 2: Scanning electron micrographs of SCCs. A Young larva of Danio rerio. A SCC with a brush-like apical ending (arrow) between the cell borders of the epidermis. B Adult Danio rerio. A taste bud (TB) surrounded by SCCs with one thick villus. C Apical endings of SCCs (arrows) with microvilli-like protrusions on a common base in the fish Pimephales promelas. ci – cilia. D A monovillous apical ending of a SCC reaches in to the lumen of the nasal cavity between adjacent cilia of ciliated supporting cells (csc). Alligator mississipiensis.

Despite this variability in location and density, SCCs commonly populate characteristic areas of the epithelia. In particular, SCCs often occur within the nasal (non-olfactory) epithelium or along respiratory passageways (Braun, 1996). One function of SCCs in these locations may be to monitor the incoming water or air stream for possible irritants or toxins. Similarly, the SCCs described in mammals are situated within the nasal cavity or pharynx and appear to be involved in monitoring the incoming air stream for potential toxins (Finger et al., 2003).

Phyletic Origins and Continuity of SCCs SCCs of similar morphology, situation and innervation are present in diverse vertebrates including teleost fishes, elasmobranchs, lampreys and hagfish. A reasonable conclusion is that such cells are a primitive trait of vertebrates, but is it possible that this distributed chemosensory network may have evolved prior to the origin of the vertebrates? Answering this question necessitates understanding the phyletic origins of vertebrates and the relationship between cephalochordates (e.g. amphioxus), hagfish (myxinoids), and craniate vertebrates. The cephalochordates are the common ancestor

Solitary Chemosensory Cells: Phylogeny and Ontogeny  |  381

Figure 3: Transmission electron micrographs of SCCs. A SCC with one dome-shaped apical ending in the fish, Pollimyrus castelnaui. B SCC with microvilli-like protrusions (arrow) sitting on a common base. Note the indentation by the adjacent cells forming a “neck” (arrows) as also seen in C, Carassius auratus (goldfish). C SCC with domeshaped apical ending. The adjacent supporting cells (sc) form a “neck” by indenting into the SCC (arrows). Alligator mississipiensis.

for the craniate chordates which include both hagfishes and the vertebrates (lampreys, elasmobranchs, teleosts, amphibians and amniotes) (see Figure 4). While hagfish used to be grouped with lampreys as “agnathan” vertebrates, more recent formulations place hagfish as a sister group to the vertebrates which include the lampreys and gnathostome vertebrates (Rasmussen et al., 1998; Javier: http://www.tolweb.org/Craniata/14826).

382  |  Airway Chemoreceptors in the Vertebrates

Figure 4: Tree depicting the phyletic relationship from cephalochordates to vertebrates.

As mentioned above, SCCs are apparent in all of the craniate chordates including hagfishes and lampreys. But what of the cephalochordates? The epithelial sensory cells of other invertebrates are primary sensory cells complete with an apical, dendritic sensory process and a basal axon extending into the central nervous system. Although cephalochordates possess numerous primary sensory cells (Lacalli and Hou, 1999) including primitive olfactory neurons (Satoh, 2005), they are the first group to exhibit secondary sensory cells, i.e. modified epithelial cells with an apical sensory process, but lacking a basal axon. Rather, these secondary sensory cells synapse onto the peripheral process of neurons situated remotely in ganglia. In amphioxus, these secondary sensory cells [Type II receptors, (Holland and Yu, 2002)] possess both microvilli and a central cilium extending beyond the epithelial surface. Basally, these Type II receptors synapse onto nerve processes (Lacalli and Hou, 1999). Some, but not all of these features are reminiscent of SCCs (e.g. apical microvilli; basal synaptic contact with nerve fibers), but the function of the amphioxus Type II receptors is unknown. It may be that they are both mechanosensory (due to the presence of the central immotile cilium) and chemosensory (related to the apical microvilli). But in the absence of either functional or molecular evidence, these possibilities are purely speculative. Nonetheless, the general similarity in morphology between vertebrate SCCs and the Type II receptor cells of amphioxus may indicate a phyletic relationship between these cell types. The Type II sensory cells may be an early evolutionary form of the SCCs which are elaborated in the craniate lineage.

Solitary Chemosensory Cells: Phylogeny and Ontogeny  |  383

Since SCCs of similar morphology can be found in fishes, amphibians and rodents, it is reasonable to ask whether there is any evidence for phyletic continuity of these cell populations. Accordingly, a study was undertaken of the nasal passageways of a nonmammalian amniote vertebrate: the alligator, Alligator mississippiensis. Alligators have long nasal passageways which are utilized for chemodetection in both air and under water (Neill, 1971; Weldon et al., 1990). Nearly the entire nasal cavity of the American alligator is lined with olfactory (= sensory) epithelium. However, the density of OSNs is heterogeneous with the highest numbers of OSNs in the areas of the two turbinates and the lowest numbers close to the naris. Occasionally, cells with the morphology of SCCs occur within the olfactory epithelium. These cells are scarce but stand out in SEM and TEM preparations (Figures 2, 3). Their apical ending is dome-shaped. Supporting cells seem to indent into the upper part of the cell resulting in a “neck” of the presumed SCCs (Figure 3C). The dome is filled with dense cytoplasm and few small vesicles. The area below the “neck” on the other hand contains abundant small vesicles. The cells express PLCβ2, a member of the transduction cascade seen in rodent SCCs. In summary, SCCs are present in all craniate chordates ranging from hagfish to mammals, including lampreys, teleost fishes, amphibians and reptiles. Assuming the SCC populations are homologous in these diverse groups, SCCs must have been a primitive trait for the craniate chordates and perhaps extending back to the Type II sensory cells of cephalochordates. Only further study will reveal whether craniate SCCs represent a phylogenetically-related population of cells or even a primitive trait first appearing in the cephalochordates.

Development One way of adducing phylogenetic relationships is to study the developmental origins of cell populations and organs. If similar cell populations in diverse species originate from a common progenitor, then it is likely that the systems are homologous. Relatively little is known about the origins of the SCC populations in different vertebrates. For fish, the development of SCCs has been studied only in one species, the zebrafish. SCCs develop later than the olfactory system but prior to the taste buds. Surprisingly, the morphology of SCCs in embryos and larvae is different from that found in adult zebrafish (Figures 2A, 2B). SCCs first develop with an apex that is divided into several small villi. In the adult zebrafish the majority of SCCs has only a single villus (Kotrschal et al., 1997). It is unknown whether the oligovillous SCCs of the juvenile fish are replaced by the monovillous type of SCC in the adult or whether the same SCC changes its shape. In rodents, differentiated nasal SCCs appear at or just prior to the time when the nasal epithelium receives its innervation, E15.5/16 in mice (Gulbransen and Finger, 2006). This is prior to the appearance of differentiated taste buds in the oral cavity.

384  |  Airway Chemoreceptors in the Vertebrates Similarly, within the oral cavity, the appearance of single presumed SCCs antedates the appearance of differentiated taste buds (Sbarbati et al., 1998). Although the sample is very small (rodents and a teleost), SCCs appear to develop prior to taste buds in diverse vertebrates. This may indicate a general principle, but obviously far more study is necessary before any firm conclusions can be reached.

Relationship to the Taste System Several investigators have suggested that the evolutionary origin of taste buds may depend on the prior development of SCCs (Whitear, 1971; Finger, 1997; Sbarbati and Osculati, 2005). The presence of SCCs, but not taste buds, in hagfish lends credence to this idea in that SCCs appeared phylogenetically prior to taste buds. Likewise, the ontogenetic appearance of SCCs prior to taste buds may relate to the necessity for SCCs to form prior to taste buds. Sbarbati et al. (1998) shows that SCC-like cells appear in the region of the circumvallate papilla prior to the appearance of differentiated taste buds. The presence of SCCs in this location may be necessary for the subsequent differentiation of taste buds, or the SCC-like cells may even become incorporated into the maturing taste bud. SCCs share some properties with one of the three cell types of taste buds (e.g. Type II cells of mammals). In catfish, taste buds and SCCs appear to use a common receptor protein in the detection of certain amino acids (Finger et al., 1996). In rodents, both SCCs and Type II taste cells express T2R-family taste receptors (so-called “bitter” receptors), gustducin, PLCβ2 and TrpM5, as well as several other elements of this transduction cascade. However, unlike SCCs, Type II taste cells do not form synaptic contacts with nerve fibers (Clapp et al., 2001). Nor do SCCs express a taste budrelated ATPase or receive innervation from P2X-expressing nerve fibers (Finger et al., 2003). Even the aggregates of SCCs present in hagfish do not exhibit the ATPase of taste buds (Kirino et al., 2006) suggesting that SCCs do not utilize ATP as a crucial neurotransmitter. In contrast, the Type II cells of taste buds appear to release ATP in a non-synaptic fashion (Romanov et al., 2007) to activate P2X on nerve fibers. Thus despite similarities in the receptors and transduction cascades utilized by SCCs and taste bud Type II cells, the cells are not identical in terms of the effector system used to transmit the neural signal.

Conclusions SCCs are a primitive feature of the vertebrate, if not chordate, lineage that persists in contemporary mammals, reptiles and fishes. Epithelial secondary chemosensory cells are distributed widely in the body, but have a common characteristic morphology in the respiratory passageways of amniote vertebrates and in the epidermis of aquatic

Solitary Chemosensory Cells: Phylogeny and Ontogeny  |  385

anamniote vertebrates. SCCs and receptor cells of taste buds share some common biochemical and structural features, but distinct differences exist.

Acknowledgements This study was supported by grants NIDCD DC-06070 to T.E.F. and DC-007732 to A.H.

References Braun, C. B. 1996. The sensory biology of the living jawless fishes: a phylogenetic assessment. Brain Behav. Evol. 48:262-76. Braun, C. B. 1998. Schreiner organs: a new craniate chemosensory modality in hagfishes. J. Comp. Neurol. 392:135-63. Clapp, T. R., Stone, L. M., Margolskee, R. F., and Kinnamon, S. C. 2001. Immunocytochemical evidence for co-expression of type III IP3 receptor with signaling components of bitter taste transduction. BMC Neurosci. 2:6. Finger, T. E. 1997. Evolution of taste and solitary chemoreceptor cell systems. Brain Behav. Evol. 50:234-43. Finger, T. E., Bryant, B. P., Kalinoski, D. L., Teeter, J. H., and Böttger, B. 1996. Differential localization of putative amino acid receptors in taste buds of the channel catfish, Ictalurus punctatus. J. Comp. Neurol. 373:129-38. Finger, T. E., Böttger, B., Hansen, A., Anderson, K. T., Alimohammadi, H., and Silver, W. L. 2003. Solitary chemoreceptor cells in the nasal cavity serve as sentinels of respiration. PNAS. 100:8981-6. Gulbransen, B. D., and Finger, T. E. 2006. Embryonic development of nasal solitary chemoreceptor cells and associated nerve fibers in mice. Chem. Senses. 31:A127. Hansen, A. 2005. The system of solitary chemosensory cells. In: Fish Chemosenses, B. G .Kapoor and K. Reutter (Eds.). Oxford and IBH Publ. Co., New Delhi. pp. 165-74. Hansen, A. 2006a. Olfactory and other chemosensory receptor cells in the nasal cavity of the American alligator. Chem. Senses. 31:A15-A16. Hansen, A. 2006b. Unconventional sensory cells in the nasal epithelia of rodents and humans. Chem. Senses. 31:E4. Hansen, A. 2007. Two different chemosensory systems in the olfactory epithelium of the American alligator, Alligator mississippiensis BMC Neurosci. 8:64. Hansen, A., and Reutter, K. 2004. Chemosensory systems in fish: structural, functional and ecological aspects. In: The Senses of Fish: Adaptations for the Reception of Natural Stimuli, G. von der Emde, J. Mogdans and B.G. Kapoor (Eds.). Narosa Publishing House Pvt. Ltd., New Delhi. pp. 55-89. Hansen, A., Zippel, H. P., Sorensen, P. W., and Caprio, J. 1999. Ultrastructure of the olfactory epithelium in intact, axotomized, and bulbectomized goldfish, Carassius auratus. Microsc. Res. Techn. 45:325-38. Holland, N. D., and Yu, J. K. 2002. Epidermal receptor development and sensory pathways in vitally stained amphioxus (Branchiostoma floridae). Acta Zool. Stockholm. 83:309-19.

386  |  Airway Chemoreceptors in the Vertebrates Kapoor, B. G., and Finger, T. E. 2003. Taste and solitary chemoreceptor cells. In: Catfishes, G. Arratia, B. G. Kapoor, M. Chardon, and R., Diogo (Eds.). Science Publishers, Inc., Enfield. pp. 753-69. Kirino, M., Kiyohara, S., Hansen, A., and Finger, T. E. 2006. Ecto-ATPase in taste buds of fishes. Chem. Senses. 31:A23. Kölliker, A. 1886. Histologische Studien an Batrachierlarven. Z. wiss. Zool. 43:1-40. Kotrschal, K. 1991. Solitary chemosensory cells - taste, common chemical sense or what? Rev. Fish Biol. Fish. 1:3-22. Kotrschal, K. 1996. Solitary chemosensory cells: why do primary aquatic vertebrates need another taste system? TREE 11:110-3. Kotrschal, K., and Adam, H. 1984. Morphology and histology of the anterior dorsal fin of Gaidropsarus mediterraneus (Pisces Teleostei), a specialized sensory organ. Zoomorphology. 104:365-72. Kotrschal, K., and Whitear, M. 1988. Chemosensory anterior dorsal fin in rocklings (Gaidropsarus and Ciliata, Teleostei, Gadidae): somatotopic representation of the ramus recurrens facialis as revealed by transganglionic transport of HRP. J. Comp. Neurol. 268:109-20. Kotrschal, K., and Finger, T. E. 1996. Secondary connections of the dorsal and ventral facial lobes in a teleost fish, the rockling (Ciliata mustela). J. Comp. Neurol. 370:415-26. Kotrschal, K., Krautgartner, W. D., and Hansen, A. 1997. Ontogeny of the solitary chemosensory cells in the zebrafish, Danio rerio. Chem. Senses. 22:111-8. Kotrschal, K., Royer, S., and Kinnamon, J. C. 1998. High-voltage electron microscopy and 3-D reconstruction of solitary chemosensory cells in the anterior dorsal fin of the gadid fish Ciliata mustela (Teleostei). J. Struct. Biol. 124:59-69. Lacalli, T. C., and Hou, S. 1999. A reexamination of the epithelial sensory cells of amphioxus (Branchiostoma). Acta Zool. Stockholm. 80:125-34. Landacre, F. L. 1911. The theory of nerve components and the fore brain of vertebrates. Trans. Am. Microsc. Soc. 30:57-66. Lane, E. B. 1977. Structural aspects of skin sensitivity in the catfish Ictalurus [dissertation]. University of London. London. Lane, E. B., and Whitear, M. 1977. On the occurrence of Merkel cells in the epidermis of teleost fishes. Cell Tissue Res. 182:235-46. Morrill, A. D. 1895. The pectoral appendages of Prionotus and their innervation. J. Morphol. 11:177-92. Neill, W. T. 1971. The Last of the Ruling Reptiles: Alligators, Crocodiles, and Their Kin. Columbia University Press, New York. Peters, R. C., Kotrschal, K., and Krautgartner, W. D. 1991. Solitary chemoreceptor cells of Ciliata mustela (Gadidae, Teleostei) are tuned to mucoid stimuli. Chem. Senses. 16:31-42. Rasmussen, A. S., Janke, A., and Arnason, U. 1998. The mitochondrial DNA molecule of the hagfish (Myxine glutinosa) and vertebrate phylogeny. J. Mol. Evol. 46:382-8. Romanov, R. A., Rogachevskaja, O. A., Bystrova, M. F., Jiang, P., Margolskee, R. F., and Kolesnikov, S. S. 2007. Afferent neurotransmission mediated by hemichannels in mammalian taste cells. EMBO J. Doi:10.1038/sj.emboj.7601526. Satoh, G. 2005. Characterization of novel GPCR gene coding locus in amphioxus genome: gene structure, expression, and phylogenetic analysis with implications for its involvement in chemoreception. Genesis. 41:47-57. Sbarbati, A., and Osculati, F. 2005. The taste cell-related diffuse chemosensory system. Progr. Neurobiol. 75:295-307.

Solitary Chemosensory Cells: Phylogeny and Ontogeny  |  387 Sbarbati, A., Crescimanno, C., Benati, D., and Osculati, F. 1998. Solitary chemosensory cells in the developing chemoreceptorial epithelium of the vallate papilla. J. Neurocytol. 27:631-5. van Houten, J. 2000. Chemoreception in microorganisms. In: The Neurobiology of Taste and Smell. T. E.Finger, W. L. Silver, and D. Restrepo (Eds.). Wiley-Liss, New York. pp. 11-40. Weldon, P. J., Swenson, D. J., Olson, J. K., and Brinkmeier, W. G. 1990. The American alligator detects food chemicals in aquatic and terrestrial environments. Ethology. 85:191-8. Whitear, M. 1965. Presumed sensory cells in fish epidermis. Nature (Lond.) 208:703-4. Whitear, M. 1971. Cell specialization and sensory function in fish epidermis. J. Zool. Lond. 163:237-64. Whitear, M. 1992. Solitary chemosensory cells. In: Fish Chemoreception, T.J. Hara (Ed.). Chapman and Hall, London. pp. 103-25. Whitear, M., and Kotrschal, K. 1988. The chemosensory anterior dorsal fin in rocklings (Gaidropsarus and Ciliata, Teleostei, Gadidae): activity, fine structure and innervation. J. Zool. Lond. 216:339-66.

18 Functional Importance of Pulmonary Neuroendocrine Cells Staffan Skogvall

Abstract The pulmonary neuroendocrine cell (PNEC) is a cell type in the airway epithelium with endocrine and neural-like features. It appears solitary or in innervated clusters, called neuroepithelial bodies (NEBs). PNECs produce several bioactive molecules such as serotonin (5-hydroxytryptamine, 5-HT), calcitonin gene-related peptide and bombesin. Several lines of evidence suggest that PNECs have important functions. During fetal life, PNECs may be involved in regulation of lung development and, after birth, they may modulate immune function and participate in airway epithelial repairing. In addition, their airway chemoreceptor properties may, via a reflex mechanism, be important for regulation of respiratory and cardiac frequencies and blood pressure and, locally, for control of pulmonary blood flow and bronchomotor tone. It has conclusively been shown that solitary PNECs are able to control the spontaneous airway contraction in guinea pigs. Pathological changes in PNECs may be of importance in several lung diseases including asthma and chronic obstructive pulmonary disease.

Keywords: PNEC, NEB, chemoreceptor, hypoxia, bronchomotor tone, spontaneous tone

Introduction Some 60 years ago, solitary and grouped cells with a clear cytoplasm and ´neuroendocrine´ characteristics were discovered in the airways of mammals (Feyrter, Managing Director, PharmaLundensis AB, Bio Medical Center D10, 221 84 Lund, Sweden, Email: [email protected]

390  |  Airway Chemoreceptors in the Vertebrates 1946). These cells are called pulmonary neuroendocrine cells (PNECs), and innervated clusters of them are designated intrapulmonary neuroepithelial bodies (NEBs) (Lauweryns and Cokelaere, 1973). PNECs have since then been characterized in lungs of humans, various mammals and amphibians (Cutz et al, 1984, Sorokin and Hoyt Jr., 1989, Zaccone et al, 1994, Mauceri et al, 1999). PNECs are abundant but not homogenously distributed within the adult human airway epithelium, with densities ranging from 65/mm2 to denser patches of 250/mm2 (Weichselbaum et al, 2005). The widespread and scattered distribution has slowed the progress in unraveling their precise function. However, in experiments involving rabbit neonates it was found that NEBs react to airway hypoxia by increased exocytosis of their dense core vesicles and a decrease in cytoplasmic amine content, suggesting that these cells may represent oxygen-sensitive airway sensors (Cutz et al, 1984, Lauweryns et al, 1978). To date, the general consensus is that clusters of PNECs, i.e. NEBs, may function as hypoxiasensitive airway sensors that, together with arterial chemoreceptors such as the carotid body, affect the control of respiration (Cutz and Jackson, 1999). In addition, PNECs have been proposed to have other important functions including control of fetal pulmonary development, involvement in immune function, airway epithelial renewal, and regulation of pulmonary blood flow and bronchomotor tone (Linnoila, 2006).

Morphology Solitary PNECs are found in the mammalian respiratory epithelium of both the upper and lower airways, including the nose, while the corpuscular NEBs are confined to the intrapulmonary airways (Van Lommel, 2001). NEBs consist of tight clusters of PNECs which usually are heavily innervated. NEBs are mainly located at airway bifurcations and bronchio-alveolar junctions. Both solitary PNECs and grouped NEB cells lie on the basal lamina and reach from the basement membrane to the airway lumen with at least a process. PNECs and NEBs are secretory. Their ultrastructure is dominated by numerous dense core vesicles with a diameter of up to 120 nm. These vesicles accumulate in the basal and perinuclear cytoplasm, which points to the existence of an endocrine secretory mechanism which discharges substances into the subepithelial chorium. Several bioactive molecules have been identified in the PNEC cytoplasm, including serotonin (5-hydroxytryptamine, 5-HT), calcitonin gene related peptide (CGRP), bombesin including the human variant gastrin-releasing peptide (GRP) and, less consistently, neuron-specific enolase (NSE), calcitonin, colecystokinin, endothelin, peptide YY, helodermin and pituitary adenylate cyclase activating protein (Van Lommel et al, 1999, Scheuerman, 1997). When released these substances may interact with nearby epithelial cells, fibroblasts, smooth muscle cells, endothelial cells and nerve fibers. NEBs are more frequent in fetal and neonatal lungs compared with mature, adult pulmonary tissue (Van Lommel et al, 1999, Van Lommel and Lauweryns, 1997). It

Functional Importance of Pulmonary Neuroendocrine Cells  |  391

was, for a long time, believed that this difference could be explained by dilution of preexisting cells in the growing lungs. However, when absolute numbers of NEBs were counted the decrease over time was found to be real rather than dilutional (Bollé et al, 1999). Thus, NEBs decrease in frequency with age, and are rare in adult mammalian lungs. In one human study (Gosney et al, 1985) only three NEBs were observed after searching through preparations from five post-mortem specimens, while another study (Weichselbaum et al, 2005) only found two NEBs, which were confined to a single lung, after scanning several hundred whole mount PIECES from eight adult human lungs. A marked reduction of the number of NEBs after birth was also found in several other mammalian species (Van Lommel and Lauweryns, 1997). In contrast, a recent confocal microscopic study demonstrated that solitary PNECs may be abundant in adult human airway epithelium (Weichselbaum et al, 2005).

Innervation NEBs are usually heavily innervated. The innervation predominantly consists of afferent sensory fibers of vagal origin with cell bodies residing in the nodose ganglion (Lauweryn et al, 1985). These nerves are not affected by systemic capsaicin treatment, and ATP might act as a neurotransmitter through the P2X3 purinoreceptor (Brouns et al, 2000, Brouns et al, 2003). In addition, there is a rich network of CGRP/Substance P-immunoreactive sensory nerve fibers which originate in the dorsal root ganglia that are depleted by systemic capsaicin treatment. A third group of nerve fibers that occasionally innervate NEBs consists of nitrergic nerve terminals originating from intrinsic bronchial ganglia (Brouns et al, 2002). Interestingly, it has recently been demonstrated that solitary PNECs may also be innervated. For example, in early postnatal rabbit lungs, most single PNECs appear innervated by a complex network of interconnecting fine varicose nerve fibers originating in the submucosa (Pan et al, 2004). Furthermore, a recent confocal microscopic study showed that human PNECs may, in some cases, also be innervated (Weichselbaum et al, 2005). These observations differs from the traditional view that only NEBs are innervated, and indicate that there is a closer functional connection between single PNECs and NEBs than previously thought.

Functional Significance of Pnecs Lung Development Lung development in its early phases is dominated by growth and differentiation of the conducting airways, while the alveoli develop late in fetal life (Cutz, 1987). PNECs are

392  |  Airway Chemoreceptors in the Vertebrates the first specialized epithelial cell type to appear, followed by ciliated and secretory cells. The appearance of PNECs follows a centrifugal pattern first appearing in the larger, central airways, and subsequently in progressively narrower peripheral airways, as they develop (Van Lommel et al, 1999). After 8 wk of gestation in humans, primitive PNECs containing serotonin and neuron-specific enolase (NSE) can be detected, and after 10 wk the major neuropeptide in human lungs, the human version of bombesin (gastrinreleasing peptide, GRP), appears (Linnoila, 1994). The early appearance of PNECs may signal an important role of these cells in lung development. PNECs are present also in lower vertebrates, suggesting that they are phylogenetically an integral early part of pulmonary development (Zaccone et al, 1997, Goniakowska-Witalinska 1997). Evidence has been obtained which strongly support a paracrine regulation of lung development by NEBs, including regulation of branching morphogenesis (King et al, 1995) and cellular growth and maturation. For example, in the developing hamster lung, the epithelial cells next to NEBs are strongly labeled with 3H-thymidine, indicating cell proliferation. The amount of labeling decreases progressively with increasing distance from the NEBs (Hoyt et al, 1993). Thus, NEBs may act as foci of growth in the nonendocrine epithelium (Sorokin et al 1997). The mechanism of action may involve paracrine secretion of bioactive neuropeptides and growth factors synthesized in PNECs such as CGRP in rodents and, in humans, members of the bombesin peptide family. Interestingly, CGRP has been shown to stimulate growth of cultured guinea pig tracheal epithelial cells (White et al, 1993) and bombesin to increase growth and maturation of lungs in mice in utero and in human organ cultures (Sunday et al, 1990). In addition, blockade of bombesin by the monoclonal antibody MAb 2A11 strongly inhibited lung maturation in serum-free organ cultures (Sunday et al, 1993).

Airway Epithelial Repairing When the lungs are damaged, NEBs may revert to their role as regulators of lung growth (Van Lommel, 2001). Hence, by their mitogenic properties NEBs are thought to be involved in the repair of damaged lung tissue. Various forms of pulmonary insults are often associated with hyperplasia of NEBs. For example, in an adult mouse model of airway injury and repair involving naphthalene-ablation of Clara cells, epithelial regeneration occurs preferentially at airway branching points, which is the main localization for NEBs, and is accompanied by NEB hyperplasia (Stevens et al, 1997). It has also been demonstrated that the NEB microenvironment is critical for the maintenance of a reservoir of pollutant-resistant stem and progenitor cells that respond to epithelial injury (Linnoila, 2006).

Immunomodulation A close topographical association between NEBs and immune cells, predominantly mast cells, neutrophils, and eosinophils, has been demonstrated in various mammalian

Functional Importance of Pulmonary Neuroendocrine Cells  |  393

species (Van Lommel et al, 1995). These data suggest that PNECs may be involved in immune responses. This hypothesis is further supported by additional studies showing that sensitization with ovalbumin stimulates PNECs to synthesize and store secretory substances which are released upon challenge with antigen (Bousbaa et al, 1994). The released substances may have direct bronchocontracting or vasodilating effects, but there is also evidence of a chemotactic influence on various types of immune cells. Recent studies using baboons and genetically engineered mice have revealed a potential mechanistic link between PNEC-derived bombesin-like proteins (BLPs) and chronic inflammatory lung disease (Sunday et al, 2004). BLPs are elevated shortly after birth in premature babies who later develop bronchopulmonary dysplasia, a disease characterized by inflammation and scarring of the lungs. Interestingly, in baboon models of this condition, anti-BLP blocking antibodies abrogate bronchopulmonary dysplasia (Sunday et al, 1998) and reduce both the PNEC and mast cell hyperplasia observed in these lungs (Subramaniam et al, 2003). Also, intratracheally administered bombesin results in increased numbers of pulmonary mast cells in mice. Thus, taken together, these data suggest that PNECs may have pro-inflammatory functions (Bousbaa and Fleury-Feith, 1991).

Hypoxia-sensitive Chemoreceptor There are several findings that support the notion that NEBs may function as hypoxia-sensitive airway chemoreceptors. This includes a preferential location of NEBs at airway branching points, apical microvilli in contact with the airway lumen, cytoplasmic neurosecretory granules with increased exocytosis during hypoxia, afferent sensory innervation from the vagus nerve, and proximity to blood capillaries (Cutz and Jackson, 1999). Furthermore, chronic exposure of rats to hypoxia causes a significant increase in the number of solitary PNECs and enlargement of NEBs to more than double that of control (Pack et al, 1986). Also, exposure of NEB cell cultures, derived from fetal rabbit lung tissue, to hypoxia for 5-15 min results in a distinct and reversible reduction of intracellular 5-HT content, indicating cell activation (Cutz et al, 1993). It is believed that NEB chemoreceptor information is transduced to the central nervous system via a classical reflex mechanism, and that this input participates in the regulation of respiratory and cardiac frequencies and blood pressure. Vagosensory NEB nerve fiber innervation appear rather late during fetal development, which may indicate that the NEBs exert their chemoreceptor function only during the postnatal period. To clarify the cellular effects by hypoxia, whole-cell patch clamp experiments on isolated PNECs, derived from mammalian NEBs, have been performed. These studies show that hypoxia (PO2 25-30 mm Hg) results in a reversible reduction by 25-30 % in outward K+ current, with no change in inward Na+ and Ca2+ currents (Youngson et al, 1993). Similar hypoxic exposure to intact neonatal PNECs in fresh rabbit lung slices

394  |  Airway Chemoreceptors in the Vertebrates resulted in a 34 % reduction in outward K+ current without any alteration of Na+ and Ca2+ inward currents (Fu et al, 1999). The oxygen sensor in PNECs has been proposed to comprise an NADPH oxidase that produces the oxygen free radical product H2O2 and a closely associated H2O2-sensitive K+ channel (Youngson et al, 1993, Wang et al, 1996). A low oxygen concentration leads to a reduced formation of H2O2 and an inhibition of the K+ channel. This leads to a depolarization of the PNECs and an increased release of PNEC neurosecretory granules. On the other hand, a higher oxygen concentration may lead to an increase in the formation of H2O2, causing an activation of the K+ channels and a hyperpolarization of the PNECs, thereby reducing the release of transmitters. Further support for the oxygen-sensing function of NEB cells comes from the close structural similarity between the corpuscular cell-nerve terminal complexes of the PNECs and the glomus cell-nerve terminal complexes of the oxygen sensing carotid body (López-Barneo, 1996). Both chemoreceptors have the same hypoxia-sensitive membrane receptor, although the carotid body monitors arterial oxygen levels, while the PNECs are directly simulated from inhaled air. PNECs may therefore react more rapidly to insufficient lung ventilation while the carotid body will only be activated when hypoxemia develops.

Redistribution of Pulmonary Blood Flow Hypoxia-induced pulmonary vasoconstriction (HPV) is a mechanism whereby local hypoxia induces vasoconstriction in insufficiently ventilated peripheral lung regions. As a result, blood is shunted to better ventilated parts of the lungs where no hypoxia and no vasoconstriction occur (Van Lommel, 2001). This improves the efficiency of the gas exchange process. It has been found that hypoxia inhibits the outward K+ current in single pulmonary vascular smooth muscle cells, leading to membrane depolarization and calcium entry through voltage-dependent calcium channels (Weir and Archer, 1995). The increase in free cytosolic calcium leads to activation of the contractile apparatus in the smooth muscle cells. Thus, there are striking similarities, but also differences, in the O2-sensing mechanism in pulmonary vascular smooth muscle cells and PNECs. In addition to the internal hypoxia sensor in the pulmonary smooth muscle cells, it has been suggested that PNEC transmitters can modulate pulmonary blood flow by binding to receptors on smooth muscle cells (Van Lommel, 2001). Indeed, PNECs secrete potent vasoactive agents during hypoxia (e.g. vasoconstricting 5-HT and vasodilating CGRP), which may modulate pulmonary blood flow. The importance of CGRP was demonstrated experimentally when depletion of endogeneous pulmonary CGRP stores with capsaicin resulted in exacerbation of hypoxia-induce pulmonary hypertension in rats (Tjen-A-Looi et al, 1998).

Functional Importance of Pulmonary Neuroendocrine Cells  |  395

It has recently been shown that the classical transient receptor potential channel 6 (TRPC6) is essential for acute, but not chronic, hypoxic pulmonary vasoconstriction and alveolar gas exchange in mice (Weissmann et al, 2006). TRPC6-deficient mice were found to lack acute HPV and hypoxia-induced influx of calcium in spite of having a normal contractile response to the thromboxane mimetic U46619. Introduction of regional hypoventilation resulted in severe arterial hypoxemia in TRPC6-deficient but not wild type mice. However, it is at present not clear how this finding fits with the traditional model for HPV.

Control of Bronchomotor Tone It has for a long time been suspected that PNEC products influence bronchomotor tone and thereby control airway diameter. Some time ago, we were able to provide experimental evidence in support of this view (Skogvall et al, 1999). Thus, in vitro experiments in guinea pig tracheal preparations demonstrated that the spontaneous force development, i.e. the spontaneous tone, is highly sensitive to the oxygen concentration in the superfusing solution (Skogvall and Grampp, 1999). In 94 % oxygen, the preparations display a strong, non-oscillating, smooth type of spontaneous tone, while in 12 % oxygen, corresponding to the arterial oxygen concentration, the preparations instead develop an oscillating type of spontaneous tone with considerably lower average force development. To clarify how PNECs participate in the generation of the spontaneous tracheal tone, various additional experiments interfering with PNEC function were performed. One method was removal of the PNEC cells by epithelium denudation. In other experiments, changes in the concentration of H2O2, which is assumed to play a key role in the PNECs signal transduction, were achieved by pharmacological means. It was found that epithelium denuded preparations in 12 % O2 are unable to develop a normal, oscillating tone with complexes, and instead display a non-oscillating spontaneous tone which is ~500 % stronger than the average tone in intact preparations (Figure 1). To clarify if the change of the spontaneous tone in denuded preparations is due to a lack of oxygen sensitive PNECs, several experiments with substances that are expected to modulate the activity in these cells were performed. When intact preparations in 94 % O2 displaying a smooth type of spontaneous tone are exposed to the NADPH oxidase inhibitor diphenyleneiodonium chloride (DPI), the spontaneous tone is transformed to an oscillating type of tone with considerably less force (Figure 2). Similar experiments in denuded preparations show no change of the average force development and no oscillations. Exposure to H2O2 in intact preparations in 12 % O2 reversibly transforms the tone from a weak, oscillating complex tone to a much stronger smooth tone, if the preparations have been pre-treated with 3-amino-1,2,4triazole, which blocks the H2O2 degrading enzyme catalase (Figure 3).

396  |  Airway Chemoreceptors in the Vertebrates

Figure 1: Typical examples of spontaneous tone in intact and denuded preparations in 12 % oxygen, which corresponds to the normal systemic arterial oxygen concentration. Denuded preparations were unable to develop a normal oscillating spontaneous tone, and instead developed a much stronger ‘smooth’ spontaneous tone which was ~500 % stronger than the average tone in intact preparations. The experiment with the denuded preparation was concluded by addition of 10 µM terbutaline (Ter.) to find the baseline tension level.

Figure 2: Effects of different oxygen concentrations and the NADPH-oxidase inhibitor diphenyleneiodonium chloride (DPI) on the spontaneous tone. The preparations initially displayed a normal, oscillating spontaneous tone in 12 % O2. Exposure to 94 % O2 transformed the tone to a much stronger, non-oscillating type of tone. DPI (20 µM) resulted in a strong decline of the average tone and in development of oscillations.

Immunohistochemical evaluation of the epithelium in the tracheal preparations displayed the existence of only single PNECs, but no NEBs (Figure 4). Thus, it appears that also single PNECs have the ability to function as oxygen sensors. This is surprising, because it has generally been believed that it is the innervated NEBs that make up the oxygen sensor. However, it was recently demonstrated that also single PNECs may be

Functional Importance of Pulmonary Neuroendocrine Cells  |  397

Figure 3: Effects of H2O2 on the oscillating spontaneous tone in 12 % oxygen. Addition of 2 mM H2O2 after 30 min pre-treatment with 1 mM of the catalase inhibitor 3amino-1,2,4-triazole reversibly transformed the tone to a strong, non-oscillating spontaneous tone.

innervated (see also above) (Pan et al, 2004). Therefore, the finding that single PNECs may function as oxygen sensors need not be a very big conceptual change.

Figure 4: Fluorescence micrograph of a guinea pig tracheal section. Immunostaining for chromogranin A showed three solitary PNECs with flasklike shape (arrows) in the airway epithelium. No groups of PNECs (NEBs) were found in the five examined tracheal specimens. Calibration bar 50 µm.

In humans, NEBs decrease in frequency with age, and are rare in the adult lung. This raises the question whether it is the solitary PNECs that are mainly responsible for pulmonary oxygen detection in adult humans, rather than the NEBs. As previously mentioned, recent data have revealed an abundance of solitary PNECs in the adult human airway epithelium (Weichselbaum et al, 2005). Clearly, our finding that the guinea pig bronchomotor tone during hypoxia and hyperoxia is mainly controlled by substances released from PNECs, shows that also single PNECs have the ability to function as effective oxygen sensors. This is clear because these preparations only displayed solitary PNECs and no grouped NEBs.

398  |  Airway Chemoreceptors in the Vertebrates

Conclusion Much data support the view that PNECs and innervated groups of these cells, NEBs, have important functions in the body. This includes regulation of lung development during fetal life and, after birth, airway epithelial repair mechanism, immunomodulation and airway chemoreceptor function. The chemoreceptor information can be transduced to the central nervous system via a reflex mechanism, and this input may participate in the regulation of respiratory and cardiac frequencies and blood pressure. Other possible functions for the oxygen-sensing PNECs include redistribution of pulmonary blood flow and, especially, control of bronchomotor tone. It has conclusively been shown that solitary PNECs are able to control the spontaneous airway contraction in guinea pigs. A defect regulation of the PNEC function, leading to a reduced activity in the PNEC cells, may result in an increased bronchomotor tone. In addition, data suggest that PNECs have pro-inflammatory functions. Therefore, it can be suspected that pathological changes affecting PNECs may be important for diseases that are characterized by inflammation and bronchoconstriction, such as asthma and chronic obstructive pulmonary disease.

Acknowledgements The author is grateful to Dr Magnus Korsgren for interesting discussions and valuable suggestions during the preparation of this chapter.

References Bollé, T.A., van Lommel, A., and Lauweryns, J.M. 1999. Stereological estimation of number and volume of pulmonary neuroepithelial bodies (NEBs) in neonatal hamster lungs. Bousban, H., and Fleury-Feith, J. (Eds.). 1991. Effects of a long-standing challenge on pulmonary neuroendocrine cells of actively sensitized guinea pigs. Am Rev Respir Dis 144:714-717. Bousbaa, H., Poron, F., and Fleury-Feith, J. 1994. Changes in chromogranin A immunoreactive guinea pig pulmonary neuroendocrine cells after sensitization and challenge with ovalbumin. Cell Tissue Res 275:195-199. Brouns, I., Adriaensen, D., Burnstock, G., and Timmermans, J.P.. 2000. Intraepithelial vagal sensory nerve terminals in rat pulmonary neuroepithelial bodies express P2X3 receptors. Am J Respir Cell Mol Biol 23:52-61. Brouns, I., van Genechten, J., Scheuermann, D.W., Timmermans, J.P., and Adriaensen, D. 2002. Neuroepithelial bodies: a morphologic substrate for the link between neuronal nitric oxide and sensitivity to airway hypoxia? J Comp Neurol 449: 343-354. Brouns, I., van Genechten,J., Hayashi, H., Gajda, M., Gomi,T., Burnstock,G., Timmermans, J.P., and Adriaensen, D. 2003. Dual sensory innervation of pulmonary neuroepithelial bodies. Am J Respir Cell Mol Biol 28:275-285.

Functional Importance of Pulmonary Neuroendocrine Cells  |  399 Cutz, E., Gillian, J.E. and Track,N.S. 1984. Pulmonay endocrine cells in developing human lung and during neonatal adaptation. In: The Endocrine Lung in Health and Disease, K.L. Becker and A.F. Gazdar (Eds.) Saunders, PA, pp. 210-231. Cutz, E. 1987. Cytomorphology and differentiation of airway epithelium in developing human lung. In: Lung Carcinomas, E.M. McDowell (Ed.). 1st Ed. Churchill Livingstone: Edinburgh. Cutz, E., and Jackson, A. 1999. Neuroepithelial bodies as airway oxygen sensors. Respir Physiol 115:201-214. Cutz, E., Speirs,V., Yeger, H., Newman, C., Wang, D., and Perrin, D. 1993. Cell biology of pulmonary neuroepithelial bodies- validation of an in vitro model (I): the effects of hypoxia and Ca2+ ionophore on serotonin content and exocytosis of dense core vesicles. Anat Rec 236:41-52. Feyrter, F. 1946. Uber die these von den peripheren endokrinen drusen. Wien Z Innere Med Grenzgeb 10:9-36. Fu, X.W., Nurse, C.A., Wang, Y.T., and Cutz, E. 1999. Selective modulation of membrane currents by hypoxia in intact airway chemoreceptors from neonatal rabbit. J Physiol Lond 514:139-150. Goniakowska-Witalinska, L. 1997. Neuroepithelial bodies and solitary neuroendocrine cells in the lungs of amphibian. Microsc Res Tech 37:13-30. Gosney, J.R., Sissons, M.C., and O´Malley, J.A. 1985. Quantitative study of endocrine cells immunoreactive for calcitonin in the normal adult human lung. Thorax 40:866-869. Hoyt, R.F., Sorokin, S.P., McDowell, E.M., and McNelly N.A. 1993. Neuroepithelial bodies and growth of the airway epithelium in developing hamster lung. Anat Rec 236:15-22. Kemp, P.J., Lewis, A., Hartness, M.E., Searle,G.J., Miller, P., O’Kelly, I., and Peers, C. 2002 Airway chemotransduction: from oxygen sensor to cellular effector. Am J Respir Crit Care Med 166:17S-24S. King, K.A., Torday, J.S., and Sunday, M.E. 1995. Bombesin and [Leu8] phyllolitorin promote fetal mouse lung branching morphogenesis via a receptor-mediated mechanism. Proc Natl Acad Sci USA 92:4357-4361. Lauweryns, J.M., and Cokelaere, M. 1973. Hypoxia-sensitive neuro-epithelial bodies: intrapulmonary secretory neuroreceptors modulated by the CNS. Z Zellforsch Mikrosk Anat 145:521-540. Lauweryns, J.M., Cokelaere, M., Lerut,T., and Theunynck, P. 1978. Cross-circulation studies on the influence of hypoxia and hypoxaemia on neuroepithelial bodies in young rabbits. Cell Tissue Res 193:373-386. Lauweryn, J.M., Van Lommel, A.T., and Dom, R.J. 1985. Innervation of rabbit intrapulmonary neuroepithelial bodies. Quantitative and qualitative ultrastructural study after vagotomy. J. Neurol Sci 67:81-92. Leach, R.M., Hill, H.M., Snetkov, V.A., Robertson, T.P., and Ward, J.P.T. 2001. J Physiol (London) 536:211-224. Linnoila, R.I. 1994. Pulmonary endocrine cells in vivo and in vitro. In: Neuropeptides in Respiratory Medicine, M.A. Kaliner, P.J. Barnes, H.W. Kunz, and J.N. Baraniuk (Eds.). Marcel Dekker, New York. Linnoila, R.I. 2006. Functional facets of the pulmonary neuroendocrine system. Lab Invest 86:425-444. López-Barneo, J. 1996. Oxygen-sensing by ion channels and the regulation of cellular functions. Trends Neurosci 19:435-440.

400  |  Airway Chemoreceptors in the Vertebrates Mauceri, A., Fasulo, S., Ainis, L., Licata, A., Lauriano, E.R., Artinez, A., Mayer, B., and Zaccone, G. 1999. Neuronal nitric oxide synthase (nNOS) expression in the epithelial neuroendocrine cell system and nerve fibers in the gill of the catfish. Heteropneustes fossilis Acta Histochem 101: 437-48. Pack, R.J., Barker, S., and Howe, A. 1986. The effects of hypoxia on the number of aminecontaining cells in the lung of the adult rat. Eur J Respir Dis 68:121-130. Pan, J., Yeager, H., and Cutz, E. 2004. Innervation of pulmonary neuroendocrine cells and neuroepithelial bodies in developing rabbit lung. J Histochem Cytochem 52:379-389. Scheuerman, D.W. 1997. Comparative histology of pulmonary neuroendocrine cell system in mammalian lungs. Micr Res Tech 37:31-42 Shimoda, L.A., Sham, J.S., Liu, Q., and Sylvester, J.T. 2002. Acute and chronic hypoxic pulmonary vasoconstriction: a central role for endothelin-1? Respir Physiol Neurobiol 132:93-106. Skogvall, S., and Grampp, W. 1999. Physiological oxygen concentration gives an oscillating spontaneous tone in guinea-pig tracheal preparations. Acta Physiol Scan 165:81-93. Skogvall, S., Korsgren, M., and Grampp, W. 1999. Evidence that neuroepithelial endocrine cells control the spontaneous tone in guinea pig trachea. J Appl Physiol 86:789-978. Sorokin, S.P. and Hoyt, Jr., R.F. 1989. Neuroepithelial bodies and solitary small-granule endocrine cells. In: Lung Cell Biol, D. Massaro, Jr. (Ed.) Marcel Dekker, New York, pp. 191-344. Sorokin, S.P., Hoyt, Jr, R.F., and. Shaffer, M.J. 1997. Ontogeny of neuroepithelial bodies: correlation with mitogenesis and innervation. Microsc Res Tech 37:43-61. Stevens, T.P., McBride, J.T., Peake, J.L., Pinkerton, K.E., and Stripp, B.R. 1997. Cell proliferation contributes to PNEC hyperplasia after acute airway injury. Am J Physiol 272:L486-L493. Subramaniam, M., Sugiyama, K. Coy, D.H., Kong, Y., Miller, Y.E., Weller, P.F., Wada, K., Wada, E., and Sunday, M.E. 2003. Bombesin-like peptides and mast cell responses: relevance to bronchopulmonary dysplasia? Am J Respir Crit Care Med 168:601-611. Sunday, M.E., Hua, J., Dai, H.B., Nusrat, A., and Torday, J.S. 1990. Bombesin increases fetal lung growth and maturation in utero and in organ culture. Am J Respir Cell Mol Biol 3:199205. Sunday, M.E., Hua, J., Reyes, B., Masui, H., and Torday, J.S. 1993. Anti-bombesin monoclonal antibodies modulate fetal mouse lung growth and maturation in utero and in organ cultures. Anat Rec 236:25-32. Sunday, M.E., Yoder, B.A., Cuttitta, F., Haley, K.J, and Emanuel, R.L. 1998. Bombesin-like peptide mediates lung injury in a baboon model of bronchopulmonary dysplasia. J Clin Invest 102:584-594. Sunday, M.E., Shan, L., and Subramaniam, M. 2004. Immunomodulatory functions of the diffuse neuroendocrine system: Implications for bronchopulmonary dysplasia. Endocr Pathol 15:91-106. Tjen-A-Looi, S., Kraiczi, H., Ekman, R., and Keith, I.M. 1998. Sensory CGRP depletion by capsaicin exacerbates hypoxia-induced pulmonary hypertension in rats. Regul Pept 74:1-10. Van Lommel, A. 2001. Pulmonary neuroendocrine cells (PNEC) and neuroepithelial bodies (NEB): chemoreceptors and regulators of lung development. Paediatr Respir Rev 2:171-176. Van Lommel, A. and Lauweryns, J.M. 1997. Postnatal development of the pulmonary neuroepithelial bodies in various animal species. J Autonom Nerv Sys 65:17-24. Van Lommel, A., van den Steen, P., and Lauweryns, J.M. 1995. Association of immune cells with neuroepithelial bodies in the lungs of neonatal dogs, cats and hamsters. Cell Tissue Res 282:519-522.

Functional Importance of Pulmonary Neuroendocrine Cells  |  401 Van Lommel, A., Bollé, T., Fannes, W., and Lauweryns, J.M. 1999. The pulmonary neuroendocrine system: the past decade. Arch histol Cytol 62:1-16. Wang, D., Youngson, C., Wong, H., Dianauer, M.C., . Vega-Saenz de Miera, E., Rudy, B., and Cutz, E. 1996. NADPH-oxidase and a hydrogen peroxide-sensitive K+ channel may function as an oxygen sensor complex in airway chemoreceptors and small cell lung carcinoma cell lines. Proc Natl Acad Sci USA 93:13182-13187. Ward, J.P.T., and Aaronson, P.I. 1999. Mechanisms of hypoxic pulmonary vasoconstriction: can anyone be right? Respir Physiol 115: 261-271. Weichselbaum, M., Sparrow, M.P., Hamilton, E.J., Thompson, P.J., and Knight, D.A 2005. A confocal microscopic study of solitary pulmonary neuroendocrine cells in human airway epithelium. Respir Res 10;6:115. Weir, E.K. and Archer, S.L. 1995. The mechanism of acute hypoxic pulmonary vasoconstriction: the tale of two channels. FASEB J 9:183-189. Weissmann N., Dietrich A., Fuchs B., Kalwa H., Ay M., Dumitrascu R., Olschewski A., Storch U., Mederos y Schnitzler M., Ghofrani H.A., Schermuly R.T., Pinkenburg O., Seeger W., Grimminger F., Gudermann T., 2006. Classical transient receptor potential channel 6 (TRPC6) is essential for hypoxic pulmonary vasoconstriction and alveolar gas exchange. Proc Natl Acad Sci USA. 19093-8. White, S.R., Hershenson, M.B., Sigrist, K.S., Zimmermann, A., and Solway, J. 1993. Proliferation of guinea pig tracheal epithelial cells induced by calcitonin gene-related peptide. Am J Respir Cell Mol Biol 8:592-596. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. 1993. Oxygen sensing in airway chemoreceptors. Nature 365:153-155. Zaccone, G., Fasulo, S., and Ainis, L. 1994. Distribution patterns of the paraneuronal endocrine cells in the skin, gills and the airways of fishes as determined by immunohistochemical and histological methods. Histochem J 26:609-629. Zaccone, G., Fasulo, S., Ainis, L., and Licata, L. 1997. Paraneurons in the gills and airways of fishes. Microsc Res Tech 37:4-12.

19 CO2/H+ Chemoreceptors in the Respiratory Passages of Vertebrates K.M. Gilmour1* and W.K. Milsom2

Abstract CO2 and/or pH-sensitive chemoreceptors located in the respiratory passages play an important role in ventilatory control in most vertebrates. In fish, branchial chemoreceptors appear to detect changes in water CO2 levels specifically, and are linked to the stimulation of ventilation frequency and/or amplitude, as well as to the initiation of cardiovascular responses such as bradycardia. The situation appears to be very similar for bimodal breathers (air-breathing fish and amphibian tadpoles) except that in many cases the input from these receptors is transformed with increasing levels of aquatic CO2 such that it stimulates air breathing rather than gill ventilation. CO2-sensitive chemoreceptors (gustatory and potentially, olfactory) are also located in the orobranchial cavity, although the roles of these chemoreceptors are less well understood. In air-breathing ectotherms, olfactory receptors often inhibit breathing and prolong breath holding when environmental CO2 levels are high. Pulmonary CO2/H+-sensitive receptors [intra-pulmonary chemoreceptors (IPC) and/or pulmonary stretch receptors (PSR)], on the other hand, regulate breathing pattern in all vertebrates (endotherms and ectotherms, including lungfish) in a manner that reduces dead space ventilation and enhances the efficiency of CO2 excretion under conditions of environmental hypercarbia, and/or reduces CO2 loss from hyperpnea/polypnea. This may be particularly true for IPC because of their greater CO2 sensitivity. Because there are several different CO2-sensitive chemoreceptor groups with different degrees of CO2 sensitivity eliciting different reflex responses in all vertebrates, responses to hypercarbia (elevated environmental CO2 tension) versus hypercapnia (elevated blood CO2 tension) may differ, giving rise to what appear to be anomalous responses to environmental CO2.

Department of Biology, University of Ottawa, 30 Marie Curie, Ottawa, ON, K1N 6N5, Canada, Phone 613-562-5800 x6004, Fax 613-562-5486, Email: [email protected] 2 Department of Zoology, University of British Columbia, Vancouver, BC, Canada. * Author for Correspondence. 1

404  |  Airway Chemoreceptors in the Vertebrates Keywords: fish, amphibian, reptile, bird, mammal, oral cavity, orobranchial cavity, gills, airways, lungs, CO2/H+-sensitive chemoreceptors, ventilation

Introduction The phylogeny of CO2/pH chemoreception in vertebrates is a topic that has received much attention over the past few years giving rise to several reviews, each with its own unique focus (see for instance: Milsom, 1995a, 1995b, 1998, 2002; Gilmour, 2001; Perry and Gilmour, 2002; Burleson and Milsom, 2003, Milsom et al., 2004; Gilmour and Perry, 2007). Given this wealth of information, the present review extracts key information with a focus specifically on CO2/pH chemoreceptors within the respiratory passages, their roles in cardiorespiratory control, and evolutionary changes in the CO2/ pH chemoreceptor control of cardiorespiratory parameters.

Phylogenetic Perspectives From Feeding to Breathing; Origins of the Respiratory Passages One of the characteristics that define all vertebrates is the possession of pharyngeal slits, at least at some point during their development (Cameron et al., 2000). It is believed that these pharyngeal slits were likely to be first evolved from the corners of the mouth to aid in suspension feeding in primitive chordates and proto-chordates (Gutmann, 1981). This development allowed a one-way flow of water; in at the mouth and out through the pharyngeal slits. Initially, this form of suspension feeding depended solely on ciliary pumps to create the flow of water (Gilmour, 1979). As the primitive chordates enlarged, the flanks of the body weakened, favouring the evolution of supporting structures between successive slits that ultimately gave rise to cartilaginous pharyngeal arches. Initially, the walls of the slits were associated with mucus-bearing cilia that served to trap suspended particles – respiration was primarily cutaneous. Only secondarily did the walls defining the slits become associated with gills and begin to participate in respiratory gas exchange. At this point, water entering the mouth could bring suspended food and oxygen to the animal. This increase in feeding efficiency gave rise to more active lifestyles and the evolution of a muscular buccal pump that helped to produce the foodbearing current. Consequently, animals were able to attain larger masses and the ciliary mechanisms for moving water were lost (Sanderson and Wassersug, 1990). Transitional species probably became raptorial feeders, plucking individual particles selectively from suspensions or from surfaces. The supporting structures of the first pharyngeal slit moved forward and evolved into jaws, further increasing feeding efficiency, and giving rise to the origins of active predation and a shift away from sessile suspension feeding (Mallatt,

Evolutionary Trends in CO2/H+ Chemoreception  |  405

1996). With this development the pharyngeal slits were no longer required for feeding, but the active lifestyle demanded greater gas exchange than was provided by cutaneous exchange alone. With removal of the constraints placed upon the pharyngeal slits for feeding, however, true gills evolved such that the pharyngeal slits and buccal pump that originally evolved for feeding gave rise to gills for breathing, with water flow being driven by a buccal pump involving muscles primarily innervated by the trigeminal and facial nerves. Muscles in the walls of the pharyngeal arches, innervated by the glossopharyngeal and vagus nerves, acted as accessory muscles to stabilize and maintain the gill curtain (Mallatt, 1996; Kardong, 2002). It is this situation that is found today in the agnathans (hagfish and lampreys), as well as the cartilaginous and bony fish. The same basic pump, with modest modifications, is still present in extant amphibians (Kardong, 2002). In most fully terrestrial vertebrates (reptiles, birds and mammals), this buccal pump was replaced by an aspiration pump involving the muscles of the chest wall, the abdomen, and in mammals, the diaphragm. In some reptiles, a buccal pump is retained to assist ventilation, suggesting that the buccal pump was not immediately abandoned when costal aspiration breathing first evolved (Brainerd, 1999). The muscles of the aspiration pumps are innervated by spinal nerves. Unlike the buccal pump, feeding and ventilation are now uncoupled, removing further constraints and increasing the opportunities for diversification of feeding and ventilation mechanisms (Gans, 1970; Kardong, 2002). With the origin of air breathing, the origin of accessory air-breathing organs and eventually lungs also occurred. The evolution of air-breathing organs occurred several times within different lines of bony fish. Graham (1997) recently put forward a simplified classification scheme for structures utilized by fish for aerial gas exchange (air-breathing organs). He suggests that “even though air breathing has evolved numerous times and independently, the location of aerial exchange sites has remained largely under the conservative influence of structures ‘predisposed’ for air gulping and sites in the body where gas storage and the requisite vascularization could be developed” (Graham, 1997). While Graham’s scheme distinguishes clearly between gas bladders and lungs, of relevance to the discussion here is that both are of endodermal origin and arise from the embryonic foregut. That is, these structures and their connecting passages also arose from various parts of the alimentary canal.

From Feeding to Breathing; Origins of the Respiratory Chemoreceptors Given that the respiratory passages of all vertebrates have arisen from digestive passages, and that, with regards to CO2 and/or pH, taste, smell and cardiorespiratory chemoreception are arbitrary distinctions, it seems highly probable that airway chemoreceptors arose from digestive (olfactory, gustatory) chemoreceptors. The extent

406  |  Airway Chemoreceptors in the Vertebrates to which the cardiorespiratory chemoreceptors have taken on distinct roles or remain associated with multiple reflexogenic responses is a topic of interest that should be borne in mind throughout the following discussion of their cardiorespiratory effects.

Receptors of the Respiratory Passages and Receptor Responses in Water-breathing Fish The respiratory passages of water-breathing fish are comprised of the orobranchial cavity and gills. Several populations of CO2/pH-sensitive chemoreceptors have been identified in these areas, including branchial and extra-branchial CO2-sensitive chemoreceptors linked to the initiation of cardiorespiratory reflexes, as well as gustatory chemoreceptors that respond to CO2 and/or pH. It is notable that the branchial and extra-branchial populations of CO2-sensitive chemoreceptors have been identified primarily on the basis of cardiorespiratory reflexes to CO2/pH that are eliminated by nerve sectioning or gill extirpation, so that the chemosensory cells responsible have yet to be definitively identified. By contrast, cells identified on the basis of location and/or structure as gustatory chemoreceptors have been examined for responsiveness to CO2/ pH, but little is known of how this information is used by the animal.

Branchial CO2-sensitive Chemoreceptors Linked to Cardiorespiratory Reflexes A wealth of evidence supports the presence in fish of branchial CO2/pH-sensitive chemoreceptors that are involved in cardiorespiratory regulation (reviewed by Milsom, 1995a, 1995b, 1998, 2002; Gilmour, 2001; Perry and Gilmour, 2002; Gilmour and Perry, 2007). Cardiorespiratory responses to hypercarbia [elevated water partial pressure of CO2 (PCO2)1] vary among fish species, but frequently include increases in ventilation frequency and/or amplitude (summarized in Table 2 of Gilmour, 2001), as well as a reduction in heart rate (summarized in Table 3.2 of Gilmour and Perry, 2007). The elimination or marked attenuation of these CO2/pH-evoked cardiorespiratory responses following gill denervation by transection of the branchial branches of cranial nerves IX and X (Burleson and Smatresk, 2000; Reid et al., 2000; Sundin et al., 2000; McKendry et al., 2001; Florindo et al., 2004) or following gill extirpation (Perry and Reid, 2002) established the primary role of branchial receptors in mediating such CO2/pH-chemoreflexes. In the fish species examined to date, the branchial CO2/pHchemoreceptors respond chiefly to CO2 rather than to pH, and to stimuli delivered in the water flow across the gills (external orientation) rather than via the circulation 1

By strict definition, hypercapnia refers to increased CO2 concentration in arterial blood while hypercarbia refers to increased CO2 concentration in the respiratory medium.

Evolutionary Trends in CO2/H+ Chemoreception  |  407

(internal orientation). Thus, the injection of CO2-enriched water into the inspired water stream to preferentially stimulate externally-oriented receptors evoked Pco2dependent changes in ventilation and cardiovascular parameters that were typical of those observed during hypercarbic exposure (Perry and McKendry, 2001; Perry and Reid, 2002; Gilmour et al., 2005). These responses were not observed, however, with either manipulation of water pH in the absence of changes in Pco2, or injection of CO2-enriched saline into the venous circulation to preferentially stimulate internallyoriented receptors (Reid et al., 2000; Sundin et al., 2000; Perry and McKendry, 2001; Perry and Reid, 2002; Gilmour et al., 2005). Moreover, use of the carbonic anhydrase inhibitor acetazolamide or exposure to hyperoxic water to promote CO2 retention and trigger a respiratory acidosis revealed that cardiorespiratory variables were resistant to the elevation of blood Pco2 (hypercapnia) in the absence of exposure to (external) hypercarbia (McKendry and Perry, 2001; Gilmour et al., 2005). Despite the clear evidence for the existence of branchial externally-oriented CO2sensitive chemoreceptors that mediate ventilatory and cardiovascular responses to changes in water Pco2, virtually nothing is known of the cellular basis of CO2 sensing in fish gills. The neuroepithelial cells of the branchial epithelium are strong candidates to serve as CO2 sensors, but have yet to be examined for CO2 sensitivity or the presence of carbonic anhydrase, an enzyme that plays a key role in mammalian CO2 chemotransduction mechanisms (Iturriaga, 1993; Peers and Buckler, 1995; Lahiri and Forster, 2003; Putnam et al., 2004; Nurse, 2005). Neuroepithelial cells are typically scattered across the filamental epithelium of fish gills in a species specific fashion (Saltys et al., 2006; Zaccone et al., 2006; Coolidge et al., 2008). Like mammalian chemoreceptor cells, neuroepithelial cells are derived from the neural crest, receive a rich and complex innervation, and can be identified by the presence of numerous dense-cored vesicles that may be congregated near the basal lamina and/or nerve profiles (Dunel-Erb et al., 1982; Bailly et al., 1992; Goniakowska-Witalinska et al., 1995; Sundin et al., 1998; Zaccone et al., 2003; Jonz and Nurse, 2003; Saltys et al., 2006). The dense-cored vesicles are presumed to be secretory, a premise supported by the observation that neuroepithelial cells of several fish species are immunopositive for a synaptic vesicle protein, SV2, that is found in neuronal and endocrine cells (Jonz and Nurse, 2003; Saltys et al., 2006; Coolidge et al., 2008). Finally, direct evidence for a chemosensory capacity in zebrafish neuroepithelial cells – albeit for O2 rather than CO2 – has been obtained by whole-cell patch-clamp approaches (Jonz et al., 2004). This work suggests that zebrafish neuroepithelial cells sense O2 by means of a mechanism similar to that in mammalian O2-chemoreceptors, relying on an O2-sensitive background K+ channel (Jonz et al., 2004). The O2-chemoreceptors of mammals and other tetrapods frequently also exhibit sensitivity to CO2 (e.g. Smatresk, 1990; Hempleman et al., 1992; West and Van Vliet, 1993; Gonzalez et al., 1994; Peers and Buckler, 1995; Kusakabe, 2002), raising the possibility that the same situation will hold true for fish gill neuroepithelial cells. Clearly, however, definitive identification of the CO2-sensitive chemoreceptor cells within the gill epithelium is needed, together with characterization of the CO2chemotransduction mechanism.

408  |  Airway Chemoreceptors in the Vertebrates In the context of evolutionary trends in respiratory CO2 chemoreception, it is notable that the CO2-sensitive chemoreceptors of fish gills are homologous not to airway CO2 chemoreceptors in tetrapods, but rather to the blood-sensing chemoreceptors associated with the carotid and aortic vessels (e.g. mammalian carotid and aortic bodies, amphibian carotid labyrinth). Fish gill arches share a common embryological origin with the carotid and aortic arches of tetrapods (e.g. see Figure 1 of Burleson and Milsom, 2003). Thus, the branchial distribution of water-sensing CO2 chemoreceptors and water-/ blood-sensing O2 chemoreceptors that modulate cardiorespiratory function in fish is phylogenetically consistent with the distribution of peripheral chemoreceptors that monitor only blood CO2/pH and O2 in tetrapods (Milsom, 1998, 2002; Taylor et al., 1999; Burleson and Milsom, 2003). Moreover, whereas the branchial chemoreceptors of fish, particularly O2-sensitive chemoreceptors, are of critical importance in initiating cardiorespiratory adjustments to changing environmental conditions or metabolic demands (see reviews by Burleson et al., 1992; Milsom, 1998; Perry and Gilmour, 2002; Burleson and Milsom, 2003; Gilmour and Perry, 2007), this role in tetrapods largely has been superseded by the evolution of central CO2/pH-sensitive chemoreceptors (Milsom, 1998, 2002; Nattie, 1999; Taylor et al., 1999; Milsom et al., 2004).

Extra-branchial CO2-sensitive Chemoreceptors Linked to Cardiorespiratory Reflexes Total gill denervation has generally eliminated virtually all heart rate and ventilatory responses to hypercarbia in the few fish species on which such experiments have been performed (Burleson and Smatresk, 2000; Reid et al., 2000; Sundin et al., 2000; McKendry et al., 2001; Florindo et al., 2004). In a few cases, however, certain cardiorespiratory responses have persisted even following total branchial denervation, suggesting the existence of extra-branchial CO2/pH-sensitive chemoreceptors that play a role in mediating cardiorespiratory chemoreflexes (Hughes and Shelton, 1962; Reid et al., 2000; Milsom et al., 2002; Florindo et al., 2004). Since, with the possible exception of air-breathing fish (e.g. Sanchez et al., 2001a), there is little experimental support for central CO2 chemosensitivity in fish (Smatresk, 1994; Milsom, 1995a, 2002; Gilmour, 2001; Remmers et al., 2001; Gilmour and Perry, 2007), it is most likely that extra-branchial chemoreceptors in fish are located in the orobranchial cavity, innervated by cranial nerves V and VII, as has been suggested also for extra-branchial O2-chemoreceptors (Hughes and Shelton, 1962; Butler et al., 1977). To investigate the role of extra-branchial chemoreceptors in mediating CO2chemoreflexes, Milsom et al. (2002) assessed the ventilatory responses to hypercarbia in decerebrate, spinalectomized tambaqui before and after central transection of cranial nerves IX and X (total gill denervation), and transection of the branches of cranial nerves V and VII that innervate the orobranchial cavity. Although total gill denervation of tambaqui with intact brains eliminated all ventilatory responses to

Evolutionary Trends in CO2/H+ Chemoreception  |  409

hypercarbia, a significant increase in ventilation frequency persisted following both gill and orobranchial cavity denervation in decerebrate, spinalectomized fish (Milsom et al., 2002). These results suggest the existence of an extra-branchial chemosensory site that stimulates ventilation in response to hypercarbia, and that is masked in fish with intact brains by an extra-branchial chemosensory site that inhibits ventilation in response to hypercarbia. Because two small branches of cranial nerve VII had to be left intact for ventilatory responses to be detectable, it is possible that the excitatory response reflects orobranchial chemoreceptors with afferent fibres in these branches of cranial nerve VII. The inhibitory response may reflect CO2-sensitive chemoreceptors in the olfactory epithelium that, when stimulated by hypercarbia, inhibit ventilation (Milsom et al., 2002). In amphibians, reptiles and mammals, stimulation of olfactory CO2 chemoreceptors causes a decrease in ventilation (reviewed by Coates, 2001; see below). To date, however, the sensitivity of olfactory chemoreceptors to CO2 does not appear to have been examined in fish (Zielinski and Hara, 2007). There is clearly a need for such studies, which should be combined with an examination of the olfactory epithelium for carbonic anhydrase activity since carbonic anhydrase appears to serve as a marker for olfactory CO2 chemoreception in other vertebrate groups (Coates, 2001).

CO2/pH-sensitive Gustatory Chemoreceptors Gustatory chemoreceptors are sensory cells of epithelial origin that are incorporated into taste buds (see reviews by Hansen and Reutter, 2004; Hara, 2007). Taste buds in teleost fish are found in subpopulations that may be associated with the oral cavity, palate, gill arches, skin and barbels. Afferent sensory information from taste buds is transmitted to the brain via cranial nerves VII (cutaneous taste buds and taste buds of the anterior buccopharyngeal cavity), or IX or X (taste buds of the posterior buccorpharyngeal cavity and gills). Unlike the gustatory system of tetrapods, which functions in contact chemoreception, fish gustatory receptors can be stimulated by dilute solutions and so act as a distance chemoreception system. In addition to the gustatory system, fish possess solitary chemosensory cells that resemble gustatory receptors in being of epithelial origin, but are not organized into taste buds; these chemoreceptors are associated with the skin and oral cavity (reviewed by Hara, 2007). The sensitivity of gustatory receptors to CO2 has been tested in several fish species by recording neural activity in palatine branches of cranial nerve VII during exposure of the chemoreceptors to CO2 and/or pH stimuli applied in water or air (Konishi et al., 1969; Hidaka, 1970; Yoshii et al., 1980; Yamashita et al., 1989). The results of these studies indicate that fish possess CO2/pH-sensitive gustatory chemoreceptors that respond independently to CO2 and H+ stimuli. In rainbow trout, integrated responses to CO2 of a whole nerve bundle were unaffected by adaptation to amino acids or bile salts, and single fibre recordings revealed that fibres sensitive to CO2 failed to respond

410  |  Airway Chemoreceptors in the Vertebrates to amino acid or bile salt stimulants (Yamashita et al., 1989). Taken together, these findings suggest that gustatory chemoreceptors for CO2 are distinct from those that detect amino acids or bile salts, and therefore that fish possess one or more populations of CO2-specific gustatory chemoreceptors (Hara, 2007). The responses that are mediated by stimulation of CO2-sensitive gustatory chemoreceptors largely remain to be determined. Konishi et al. (1969) reported reductions in ventilation amplitude and frequency upon stimulation of the palatal chemoreceptors of carp with CO2, suggesting that CO2-sensitive gustatory chemoreceptors may be involved in mediating cardiorespiratory responses to CO2 and/or pH. Interestingly, however, the CO2-mediated inhibition of ventilation elicited by stimulating palatal gustatory receptors (Konishi et al., 1969) contrasts with the hyperventilatory responses typically evoked by exposure of branchial chemoreceptors to hypercarbic water (discussed above). Branchial chemoreceptors could include CO2sensitive gustatory receptors located on the gills (although to date the CO2 sensitivity of gill taste buds has not been examined) as well as the non-gustatory CO2-chemoreceptors (presumably neuroepithelial cells – see above), and the extent to which CO2-evoked cardiorespiratory reflexes reflect the activation of one receptor type versus the other remains unknown. Indeed, distinguishing between gustatory and neuroepithelial cell inputs from the gills will be difficult given that cranial nerves IX and X constitute the afferent sensory pathway for both – denervation studies in which the branchial branches of cranial nerves IX and X were transected will have eliminated both inputs (Burleson and Smatresk, 2000; Reid et al., 2000; Sundin et al., 2000; McKendry et al., 2001; Florindo et al., 2004). Assuming that gustatory CO2 chemoreceptors are located on the gills and are involved in mediating hyperventilatory responses (both large assumptions at present), the inhibition of ventilation following stimulation of palatal CO2 chemoreceptors may simply reflect differences in chemoreceptor location. A similar reflexive decrease in ventilation is observed in tetrapods upon stimulation of CO2-sensitive chemoreceptors present in the posterior oral cavity (e.g. laryngeal epithelium) (Heman-Ackah and Goding, 2000; Nishijima et al., 2004). Hara (2007) suggested that this shared response may stem from its presence in an ancestral, airbreathing fish.

Airway Receptors and Receptor Responses in Bimodal Breathers It might be expected that animals that employ both gill and lung ventilation simultaneously (bimodal breathers) would possess the full repertoire of receptors described for water-breathing fish (see above) and air-breathing vertebrates (see below). Unfortunately, this group, which includes air-breathing fish (chondrosteans, holosteans and teleosts as well as the sarcopterygian lungfish) as well as many larval

Evolutionary Trends in CO2/H+ Chemoreception  |  411

and a few adult amphibians, is the least studied in this respect and thus the story is extremely incomplete. In some air-breathing fish, increases in aquatic CO2 are without effect on gill ventilation (McMahon and Burggren, 1987; Todd, 1972; Johansen, 1966) or air breathing ( Johansen et al., 1968; Lomholt and Johansen, 1974). In many, however, there is significant indirect evidence for the presence of water-sensing CO2/pH chemoreceptors that are linked to cardiorespiratory reflexes and located somewhere within the orobranchial cavity and/or gills. Many species show significant increases in gill ventilation in response to low levels of aquatic CO2 (<3%) despite relatively constant levels of arterial Pco2 ( Johansen and Lenfant, 1968; Perry et al., 2005). In some of these species, further increases in aquatic CO2 (>2-10%) lead to inhibition of gill breathing and usually, stimulation of air breathing ( Johansen et al., 1967, 1970; Graham and Baird, 1982; Jesse et al., 1967; Rahn et al., 1971; Shelton et al., 1986; Graham, 1997; Sanchez and Glass, 2001; Sanchez et al., 2005), often also with no accompanying changes in arterial Pco2 ( Johansen and Lenfant, 1968). These results are consistent with those described for water-breathing fish in implicating a watersensing CO2 receptor whose input, in many cases, is transformed centrally such that increasing levels of stimulation result in a switch from excitation of gill ventilation, to excitation of air breathing and a concomitant inhibition of water breathing. To date there is no evidence to suggest where the CO2 receptors that produce these responses reside. Again, as discussed earlier, if peripheral O2-sensitive chemoreceptors are also CO2 sensitive in fish as they are in mammals (e.g. Smatresk, 1990; Hempleman et al., 1992; West and Van Vliet, 1993; Gonzalez et al., 1994; Peers and Buckler, 1995; Kusakabe, 2002), then some insight may be obtained from the distribution of watersensing O2 chemoreceptors in these species. O2 receptors sensing aquatic O2 that stimulate both gill and lung ventilation have been found on all gill arches in the African lungfish Protopterus, but are more predominant on the first two (filament free) arches ( Johansen and Lenfant, 1968; Lahiri et al., 1970). Many air-breathing fish show no response to increasing levels of CO2 in inspired air (Sanchez and Glass, 2001; Perry et al., 2005) while others show an increase (Babiker, 1979) or decrease ( Jesse et al., 1967) in pulmonary ventilation. The only receptors thus far that have been found to be sensitive to CO2 in lungs and other air-breathing organs are pulmonary stretch receptors (PSR, see next section). Slowly adapting pulmonary stretch receptors in the lungs of both the African (Protopterus) and South American (Lepidosiren) lungfish are inhibited by CO2 (Delaney et al., 1983) as are those in the gar (Lepisosteus oculatus) (Smatresk and Azizi, 1987) but not Amia, the bowfin (Milsom and Jones, 1985). The levels of CO2 required to inhibit discharge in the gar, however, were in excess of the physiological range. Interestingly, Sanchez and Glass (2001) report a profound increase in ventilation on return from aerial hypercarbia to breathing air in the South American lungfish (an ‘off response’ , see next section) suggestive of an interaction between excitatory and

412  |  Airway Chemoreceptors in the Vertebrates inhibitory inputs from different sources; i.e. the presence of more than one site of airway CO2 sensing. Similarly discrepant results have been reported for lungfish in aestivation. Thus, some researchers have found air breathing to be stimulated by exposure to hypercarbia during aestivation in adult Protopterus, when blood Pco2 levels were already greatly elevated (Smith, 1930; Delaney et al., 1974, 1976, 1977), whereas others have not (Perry et al., 2008).

Airway Receptors and Receptor Responses in Air-breathing Vertebrates Olfactory Chemoreceptors A series of studies has identified upper airway receptors sensitive to changes in CO2 in the nasal sensory epithelium, innervated by the olfactory nerve in the bullfrog (Rana catesbeiana, Sakakibara, 1978; Kinkead and Milsom, 1996; but see also Smyth, 1939; Coates and Ballam, 1990), the tegu lizard (Tupinambis nigropunctatus, Ballam, 1984, 1985; Coates and Ballam, 1987), the garter snake (Thamnophis sirtalis, Coates and Ballam, 1989), rats (Coates and Silvis, 1999) and humans (Youngentob et al., 1991; Alvaro et al., 1993). These receptors are relatively rare, are stimulated by CO2 levels below or near the animal’s end-tidal CO2 concentration, and produce a reflex inhibition of breathing (Getchell and Shepherd, 1978; Coates and Ballam, 1987, 1989, 1990; Ballam and Coates, 1989; Coates et al., 1991, 1998; Coates, 2001). Receptors with similar properties and/or reflex effects have been inferred in fish (see above); thus, at present there is no reason to believe that they are not present in all vertebrates. In frogs and rats, inhibition of carbonic anhydrase (CA) attenuates the response of CO2sensitive olfactory receptors to transient changes in CO2 and it has been suggested that CA activity can serve as a marker of CO2-sensitive receptors in the olfactory epithelium (Coates, 2001). In all air-breathing vertebrates, olfactory receptors appear to have no influence on resting ventilation when environmental CO2 levels are low. In species such as frogs, lizards and some snakes, these receptors inhibit breathing and prolong breath holding when environmental CO2 levels rise. Thus, it has been postulated that stimulation of olfactory chemoreceptors by changes in CO2/H+ might function to inhibit breathing when the ambient Pco2 is higher than levels in the systemic blood, reducing CO2 uptake while the animals seek fresh air (Coates and Ballam, 1987). As such, CO2/ H+-chemosensitive olfactory receptors may serve as the afferent limb of a defensive reflex that initiates reduced ventilation or breath holding while animals seek a better environment. Such responses have also been postulated to occur in fish as a defensive reflex (Burleson et al., 1992). Hypercarbia is not uncommon in aquatic environments and retention of such a reflex in some terrestrial animals would not be surprising. One

Evolutionary Trends in CO2/H+ Chemoreception  |  413

might predict that these receptors and this response would be more strongly developed in species that encounter, and can choose to avoid, hypercarbic environments. In situations where they cannot avoid the hypercarbia, however, these reflexes will be in conflict with excitatory cardiorespiratory responses. It has also been suggested that the receptors found in the nasal epithelium may function to detect changes in environmental CO2 originating from prey or predators, especially in confined burrows (Ballam, 1985; Coates and Ballam, 1987, 1989). Why such a function should lead to a net inhibition of breathing and change in breathing pattern, however, is not at all clear. While this response might reduce the frequency of body wall movements, making the waiting predator or alerted prey more difficult to detect, the size of each breath and associated body wall expansion that occurs when the animal does breathe is enhanced.

Vomeronasal Chemoreceptors The vomeronasal (or Jacobsen’s) organ is an accessory olfactory organ found in many tetrapods but absent in most turtles, crocodilians, birds, some bats and aquatic mammals. Its innervation runs parallel to, but remains entirely separate from, the main olfactory system. It is believed to participate primarily in sensing chemical signals important to social or reproductive behaviour, as well as for feeding and prey trailing (Halpern and Kubie, 1984; Cooper and Burghardt, 1990). However, the inhibition of respiratory frequency in response to upper airway CO2 is greater after vomeronasal nerve lesions in gartersnakes, suggesting that CO2-sensitive receptors in the vomeronasal organ of this species are excited by CO2 and elicit a reflex increase in breathing frequency (Coates and Ballam, 1989). No role for the modulatory effects of the vomeronasal receptors on respiration has yet been proposed. It has also been suggested that these receptors may be involved in the chemical sensing associated with feeding and prey tracking but this proposed function does not seem to be consistent with the fact that stimulation of these receptors excites ventilation (Halpern and Kubie, 1984; Coates and Ballam, 1989; Cooper and Burghardt, 1990).

The Pulmonary Chemoreceptors: Intra-pulmonary Chemoreceptors (IPC) and Pulmonary Stretch Receptors (PSR) Intrapulmonary chemoreceptors (IPC) have now been described in the lungs of several species of reptiles and birds. They are located within the lung, innervated by the vagus nerve and have a discharge that is inversely proportional to Pco2. The phylogenetic distribution of IPC suggests that this receptor group evolved after mammals split from the reptilian stock. IPC have not been found in amphibians (Milsom and Jones, 1977; Kuhlmann and Fedde, 1979; Furilla and Bartlett, 1988) or mammals (Kunz et al.,

414  |  Airway Chemoreceptors in the Vertebrates 1976) and their presence in turtles is questionable ( Jones and Milsom, 1979; Ishii et al., 1986; Sundin et al., 2001). IPC are present in birds (see Scheid and Piiper, 1986 for review) and all diapsid reptiles examined so far (lizards: Gatz et al., 1975; Fedde et al., 1977; snakes: Sundin et al., 2001; alligators, Alligator mississippiensis: Powell et al., 1988; Douse et al., 1989). Their CO2 sensitivity is nearly abolished by intracellular inhibitors of carbonic anhydrase and inhibitors of Na+/H+ antiport exchange (Hempleman et al., 2000, 2003), suggesting that they respond to intracellular pH rather than CO2 per se (see Hempleman and Posner, 2004). A unique feature of these receptors is their extremely fast response and sensitivity to the rate of CO2 change, resulting in a dynamic and responsive respiratory chemoreceptor (Osborne et al., 1977; Hempleman and Posner, 2004). All air-breathing vertebrates appear to possess slowly adapting pulmonary receptors. The primary stimuli for pulmonary stretch receptors are changes in lung volume, pressure or wall tension. Increasing pulmonary CO2, however, inhibits their discharge by varying degrees in amphibians, reptiles, birds and mammals (see Milsom, 1995a&b; 1998; for reviews). The effects of CO2 on receptor discharge range from insignificant to total inhibition, both between species and between individual receptors in a single animal in some species ( Jones and Milsom, 1979; Powell et al., 1988; Sundin et al., 2001). The inhibition is largely (but not completely) independent of the effects of the CO2 on pulmonary smooth muscle tone, and it is unknown whether these effects result from changes in pH or CO2 (Milsom, 1995a; Sundin et al., 2001). Where this mechanoreceptor discharge is CO2-sensitive, high levels of intrapulmonary CO2 also elevate tidal volume and reduce breathing frequency ( Jones and Milsom, 1982). One role of CO2/H+-chemosensitive pulmonary receptors (IPC and PSR) may lie in the regulation of breathing pattern to enhance the efficiency of CO2 excretion under conditions of environmental hypercarbia. Both PSR and IPC contribute to inspiratory termination (Banzett and Burger, 1977; Milsom et al., 1981). CO2 reduces discharge of both receptor groups and acts to prolong inspiration, increasing tidal volume and decreasing breathing frequency, responses that reduce dead space ventilation under any condition in which pulmonary CO2 is elevated. The extent to which this response is of benefit to an animal will be a consequence of the magnitude of the ventilatory dead space, and the extent to which changes in pattern are offset by net changes in the level of total ventilation. The opposite will also be true, and feedback from these receptors will also therefore act to prevent CO2 loss under conditions where breathing is elevated, such as during certain behavioural displays.

Interactions among the Different Receptor Groups ‘Steady-state responses’

In amphibians, it appears that there are only two groups of CO2 sensitive airway receptors, the olfactory receptors and the PSR. The net effect of inspiring CO2 on

Evolutionary Trends in CO2/H+ Chemoreception  |  415

these two groups is an inhibition of breathing due to olfactory receptor stimulation and the change in pattern resulting from inhibition of PSR discharge. Although there are too few data to draw firm conclusions, it appears that the net effect in frogs is slower, deeper breathing (enhanced ‘faveolar’ ventilation) during hypercarbia followed by a significant post-hypercarbic rebound that would serve to flush the lungs on return to breathing fresh air (Kinkead and Milsom, 1996). It is tempting to ascribe this scenario to a diving lifestyle, where internal CO2 build-up would be expected, but the response is primarily due to stimulation of the olfactory receptors that would continue to be ventilated by buccal pumping even during breath-holding. The chelonians and mammals arose from the stem reptiles at about the same time. Interestingly, olfactory receptors appear to have little effect on overall ventilation in either group, and the effect of CO2 on PSR discharge does not lead to much change in net ventilation. For these groups, there is little difference between ventilatory responses to hypercapnia and those to hypercarbia (Ballam, 1985; Coates and Silvis, 1999). The two groups in which we see the most profound difference between ventilatory responses to hypercarbia and hypercapnia are diapsid reptiles and birds, groups that both also have IPC. In diapsid reptiles, tremendous variation exists in the extent to which ventilation is affected by olfactory CO2 sensitive receptor stimulation, as well as in the magnitude to which the respiratory pattern is altered by increasing levels of intrapulmonary CO2. Increasing levels of environmental CO2 may inhibit total ventilation (tegu lizard, Ballam, 1985) or lead to modest (rattlesnake, Andrade et al., 2004) or robust (caiman, Tattersall et al., 2006) increases in total ventilation. Compared to the pattern of ventilation observed during the ‘off-response’ (see below), which is believed to more closely mimic the response to metabolically produced CO2, in all cases the pattern of breathing is altered in a manner that would increase relative alveolar ventilation. In animals with large saccular lungs, this response not only reduces dead space ventilation but also flushes the lungs and air sacs. On return to breathing air, the post-hypercarbic response (off-response) would further refresh air in the lungs. Finally, in situations where levels of CO2 fall (hypocapnia), this response would act to decrease tidal volume and increase breathing frequency, concentrating air flow to the more highly developed gas exchange portions of the lung that surround the opening of the trachea and primary bronchi. In birds, the sparse available data suggest that olfactory receptors play little or no role in the control of ventilation. IPC do appear to contribute to changes in the breathing pattern during hypercarbia, but given the mechanics of ventilation in birds, less effect on dead space ventilation would be expected to occur (Scheid and Piiper, 1986). The IPC in birds, however, appear to take on new roles. Birds are unique in that the portion of the lung associated with producing airflow and the gas exchange surface have become separated into, respectively, the air sacs and lungs. With this change, the lungs have become rigid, in turn allowing the blood gas barrier to become thinner and greatly increasing the anatomical diffusion capacity of the lung (Perry and Duncker,

416  |  Airway Chemoreceptors in the Vertebrates 1980). The IPC largely take over the role of the PSR in this rigid lung, monitoring the washout or dilution of CO2 in the respiratory passages during inspiration, which is a function of the rate and depth of each inspiration (Scheid and Piiper, 1986). It has also been suggested that IPC may contribute to the hypercapnic ventilatory response and, in particular, that these receptors might be able to monitor CO2 flux from venous blood into the pulmonary space and act as a venous CO2 receptor (Fedde et al., 1982). As such, IPC could act to minimize or even eliminate any increase in arterial Pco2 arising from increased metabolic CO2 production. This, however, might be a transient effect of changes in the breathing pattern in which tidal volume increases faster than breathing frequency decreases. Presently, there are data to both support (Fedde et al., 1982) and refute this hypothesis (Milsom et al., 1981). The extent to which IPC may perform a unique function may be in eliciting ventilatory responses to hypocapnia (Osborne et al., 1977). Here, the CO2 sensitivity of both PSR and IPC would act to decrease tidal volume and increase breathing frequency when CO2 levels are low, preventing alkalosis during the hyperpnea associated with thermal polypnea and behavioural displays.

Post-hypercapnic Hyperpnea - ‘Off-Response’

In many species, a return from hypercarbia to air is accompanied by a marked transient increase of ventilation relative to values during hypercarbic exposure. Thus, in the South American lungfish, Lepidosiren paradoxa (Sanchez and Glass, 2001), the anuran amphibian, Rana catesbeiana (Kinkead and Milsom, 1996), the lizards, Crotaphytus collaris, Lacerta viridis and Uromastyx aegypticus (Nielsen, 1961; Templeton and Dawson, 1963; Klein et al., 2002) and the snakes, Acrochordus javanicus, Coluber constrictor and Crotalus durissus (Glass and Johansen, 1976; Nolan and Frankel, 1982; Andrade et al., 2004), an immediate relative hyperpnea is seen when inspired hypercarbic gas is replaced with a normocarbic gas mixture, an effect termed the ‘offresponse’ or post-hypercapnic hyperpnea. This off-response has been interpreted to suggest that during conditions of environmental hypercarbia, the stimulating effect of systemic hypercapnia is, at least in part, masked by an inhibitory effect of tonically elevated airway CO2 (Boelaert, 1941; Nielsen, 1961; Templeton and Dawson, 1963; Nolan and Frankel, 1982; Coates and Ballam, 1989; Kinkead and Milsom, 1996; Klein et al., 2002; Andrade et al., 2004). When animals begin to breathe normocarbic air again, arterial CO2 levels will still be elevated for some time but the level of CO2 in the airways will be elevated only during expiration. Arterial levels of CO2 and endexpiratory levels of CO2 will fall slowly as whole body CO2 stores are lowered and CO2 is eliminated, whereas inspired CO2 levels will fall immediately. Thus, the inhibitory effect of tonically elevated airway CO2 is immediately relieved, whereas the excitatory effect of elevated systemic CO2 persists until a new steady state is attained. If this interpretation is correct, then the post-hypercapnic hyperpnea should be greater in instances where both the excitatory effect of the systemic hypercapnia

Evolutionary Trends in CO2/H+ Chemoreception  |  417

(i.e. the level of CO2) and the inhibitory effect on airway CO2–sensitive receptors (as indicated by the reduction in steady state breathing frequency during sustained hypercarbia) were greater. The literature contains reports suggesting both that there is a correlation (Templeton and Dawson, 1963; Andrade et al., 2004) and that there is no correlation (Nielsen, 1961; Nolan and Frankel, 1982) between the level of environmental hypercarbia and the magnitude of the response to CO2 removal. Interestingly, Klein et al. (2002) found that a correlation existed when the data were expressed in absolute terms, but not when the data were expressed in relative (proportionate) terms. Why consistent results have not been found in all studies is hard to say, but the diversity of data may reflect the kinetics of the receptor responses. It is possible that there is a maximum response that can be produced owing to receptor kinetics, and in some species this maximum is produced at lower levels of inspired CO2, masking any correlation.

Conclusions In conclusion, the CO2/pH-sensitive chemoreceptors of the respiratory passages play an important role in regulating ventilation in most vertebrates (see summary in Table 1). In strictly water-breathing fish, cardiorespiratory reflexes are keyed primarily to O2, but a modulatory role is played by water-sensing, CO2-sensitive, branchial chemoreceptors which function to stimulate ventilation when activated by hypercarbia. The cellular characterization of these receptors remains to be accomplished. The transition to air breathing is accompanied by a transition to a primarily CO2/pHkeyed ventilatory drive and the appearance of central CO2/pH chemosensitivity. Nevertheless, airway CO2/pH chemoreceptors continue to play a modulatory role in ventilatory control. In bimodal breathers, the available evidence supports the existence of both water-sensing CO2/pH chemoreceptors presumably located in the gills or orobranchial cavity, as well as CO2-sensitive pulmonary stretch receptors. Central coordination of different receptor inputs seems to favour the stimulation of gill ventilation at low levels of aquatic hypercarbia but a switch to air-breathing at higher Pco2 values. However, data on the CO2/pH-mediated control and coordination of ventilation in bimodal breathers are sparse and additional studies are needed to provide a more complete picture. In air-breathing vertebrates, several populations of CO2/ pH-sensitive chemoreceptor groups can be identified. CO2-sensitive chemoreceptors are found in the nasal epithelium; they may also be present in the nasal epithelium of water-breathing fish. The olfactory CO2-sensitive chemoreceptors are inhibitory, but their role remains poorly understood. Data on air-breathing ectotherms suggesting that these receptors prolong breath holding by inhibiting breathing under hypercarbic conditions, possibly as a defensive mechanism. Pulmonary CO2/pH-sensitive chemoreceptors consist of intra-pulmonary chemoreceptors, found only in birds

418  |  Airway Chemoreceptors in the Vertebrates Table 1:

Summary of the distribution, innervation and effect on ventilation of CO2/pHsensitive chemoreceptors of the respiratory passages across the vertebrates. fish lungfish

Branchial

Olfactory

IX, X stimulatory

I inhibitory?

IX, X mixed

?

Vomeronasal

IPC

PSR

X? ?

amphibians

I inhibitory

V stimulatory?

X pattern?

mammals

I inhibitory?

V stimulatory?

X pattern?

turtles

I inhibitory?

birds

I inhibitory?

X inhibitory

X pattern

crocodilians

I inhibitory?

X inhibitory

X pattern

lizards/ snakes

I inhibitory

X inhibitory

X pattern

X pattern

V stimulatory?

I, V, IX and X refer to the cranial nerves that innervate the CO2/pH-sensitive chemoreceptors in a particular location. The effect on total ventilation of the activation of these CO2/pH-sensitive chemoreceptors by hypercarbia is indicated as stimulatory (i.e. increases ventilation), inhibitory (i.e. decreases ventilation), mixed (i.e. both stimulatory and inhibitory responses have been reported) or pattern (changes in the balance of tidal volume and breathing frequency, with or without resultant changes in total ventilation). A question mark indicates that the ventilatory response to activation of the particular chemoreceptor is uncertain or unknown.

and reptiles, and pulmonary stretch receptors in which the pattern of response to the primary stimuli of changes in lung volume, pressure or wall tension is modulated by CO2. These receptor groups probably serve to reduce dead space ventilation and enhance the efficiency of CO2 excretion during hypercarbia. Thus, the CO2 and/or pH-sensitive chemoreceptors of the respiratory passages share a common purpose of modulating the breathing pattern.

Acknowledgements Original research was funded by the Natural Sciences and Engineering Research Council of Canada. We wish to thank the editors for the opportunity to contribute to this excellent volume.

Evolutionary Trends in CO2/H+ Chemoreception  |  419

References Alvaro, R.E., de Almeida, V., Kwiatkowski, K., Cates, D., Kryger, M., and Rigatto, H. 1993. A developmental study of the dose-response curve of the respiratory sensory reflex. Am. Rev. Respir. Dis. 148, 1013-1017. Andrade, D.V., Tattersall, G.J., Brito, S.P., Soncini, R., Branco, L.G., Glass, M.L., Abe, A.S., and Milsom, W.K. 2004.  The ventilatory response to environmental hypercarbia in the South American rattlesnake, Crotalus durissus. J. Comp. Physiol. B. 174, 281-291. Armstrong, B.W., Hurt, H.H., Blide, R.W., and Workman, J.M. 1961. The humoral regulation of breathing. Science 133, 1897-1906. Babiker, M.M. 1979. Respiratory behaviour, oxygen consumption and relative dependence on aerial respiration in the African lungfish (Protopterus anectens, Owen) and an air breathing teleost (Clarius lazera, C.). Hydrobiologia 65, 177-187. Bailly, Y., Dunel-Erb, S., and Laurent, P. 1992. The neuroepithelial cells of the fish gill filament: indolamine-immunocytochemistry and innervation. Anat. Rec. 233, 143-161. Ballam, G.O. 1984. Ventilatory response to inspired CO2 in the lizard, Tupinambis nigropunctatus. Comp. Biochem. Physiol. 78A, 757-762. Ballam, G.O. 1985. Breathing response of the tegu lizard to 1-4% CO2 in the mouth and nose or inspired into the lungs. Respir. Physiol. 62, 375-386. Ballam, G.O., and Donaldson, L.A. 1988. Effect of venous (gut) CO2 loading on intrapulmonary gas fractions and ventilation in the tegu lizard. J. Comp. Physiol. B. 158, 591-600. Ballam, G.O., and Coates, E.L. 1989. Effect of upper airway CO2 pattern on ventilatory frequency in tegu lizards. Am.J. Physiol. 26, R156-R161. Banzett, R.B., and Burger, R.E. 1977. Response of avian intrapulmonary chemoreceptors to venous CO2 and ventilatory gas flow. Respir. Physiol. 29, 63-72. Boelaert, R. 1941. Sur la physiologie de la respiration des Lacertiens. Arch. Int. Physiol. 51, 379437. Brainerd, E.L. 1999. New perspectives on the evolution of lung ventilation mechanisms in vertebrates. Exptl. Biol. Online. 4, 11-28. Burger, R.E. 1968. Pulmonary chemosensitivity in the domestic fowl. Fed. Proc. 27, 328. Burleson, M.L., and Smatresk, N.J. 2000. Branchial chemoreceptors mediate ventilatory responses to hypercapnic acidosis in channel catfish. Comp. Biochem. Physiol. 125A, 403414. Burleson, M.L., and Milsom, W.K. 2003. Comparative aspects of O2 chemoreception. Anatomy, physiology, and environmental applications. In: Oxygen Sensing. Responses and Adaptation to Hypoxia, S. Lahiri, G.L. Semenza and N.R. Prabhakar (Eds.), Marcel Dekker, New York, pp. 685-707. Burleson, M.L., Smatresk, N.J., and Milsom, W.K. 1992. Afferent inputs associated with cardioventilatory control in fish. In: Fish Physiology Vol. XIIB The Cardiovascular System, W.S. Hoar, D.J. Randall and A.P. Farrell (Eds.), Academic Press, San Diego, pp. 389-426. Butler, P.J., Taylor, E.W. and Short, S. 1977. The effect of sectioning cranial nerves V, VII, IX and X on the cardiac response of the dogfish Scyliorhinus canicula to environmental hypoxia. J. Exp. Biol. 69, 233-245. Cameron, C., Garey, J., and Swalla, B.J. 2000. Evolution of the chordate body plan: New insights from phylogenetic analysis of deuterostome phyla. PNAS 97, 4469-4474. Coates, E.L. 2001. Olfactory CO2 chemoreceptors. Respir. Physiol. 129, 219-230.

420  |  Airway Chemoreceptors in the Vertebrates Coates, E.L., and Ballam, G.O. 1987. Upper airway CO2 receptors in tegu lizards: localization and ventilatory sensitivity. J. Comp. Physiol. B. 157, 483-489. Coates, E.L., and Ballam, G.O. 1989. Breathing and upper airway CO2 in reptiles: role of the nasal and vomeronasal systems. Am. J. Physiol. 25, R91-R97. Coates, E.L., and Ballam, G.O. 1990. Olfactory receptor response to CO2 in bullfrogs. Am. J. Physiol. 27, R1207-R1212. Coates, E.L., and Silvis, M.L. 1999. Age-related changes in the ventilatory response to inspired CO2 in neonatal rats. Respir. Physiol. 118, 173-179. Coates, E.L., Furilla, R.A., Ballam, G.O., and Bartlett, D. 1991. A decrease in nasal CO2 stimulates breathing in the tegu lizard. Respir. Physiol. 86, 65-75. Coates, E.L., Wells, C.M.Q., and Smith, R.P. 1998. Identification of carbonic anhydrase activity in bullfrog olfactory receptor neurons: histochemical localization and role in CO2 chemoreception. J. Comp. Physiol. A. 182, 163-174. Coolidge, F.H., Ciuhandu, C.S., and Milsom, W.K. 2008. A comparative analysis of putative O2 sensing cells in the fish gill. J. Exp. Biol. 211, 1231-1242. Cooper, W., and Burghardt, G.M. 1990. Vomerolfaction and vomodor. J. Chem. Ecol. 16, 103105. Dealney, R.G., and Fishman, A.P. 1977. Analysis of lung ventilation in the aestivating lungfish Protopterus aethiopicus. Am. J. Physiol. 233, R181-R187. Delaney, R.G., Lahiri, S., and Fishman, A.P. 1974. Aestivation of the African lungfish Protopterus aethiopicus: Cardiovascular and respiratory functions. J. Exp. Biol. 61, 111-128. Delaney, R.G., Shub, C., and Fishman, A.P. 1976. Haematologic observations on the aquatic and aestivating African lungfish Protopterus aethiopicus. Copeia. 423-434. Delaney, R.G., Lahiri, S., Hamilton, R. and Fishman, A.P. 1977. Acid-base balance and plasma composition in the aestivating lungfish (Protopterus). Am. J. Physiol. 232, R10-R17. Delaney, R.G., Laurent, P., Galante, R., Pack, A.I., and Fishman, A.P. 1983. Pulmonary mechanoreceptors in the dipnoi lungfish Protpterus and Lepidosiren. Am. J. Physiol. 244, R418-R428. Douse, M.A., Powell, F.L., Milsom, W.K., and Mitchell, G.S. 1989. Temperature effects on pulmonary receptor responses to airway pressure and CO2 in Alligator mississippiensis. Respir. Physiol. 78, 331-344. Dunel-Erb, S., Bailly, Y., and Laurent, P. 1982. Neuroepithelial cells in fish gill primary lamellae. J. Appl. Physiol. 53, 1342-1353. Fedde, M.R., Kuhlmann, W.D., and Scheid, P. 1977. Intrapulmonary receptors in the tegu lizard: I. Sensitivity to CO2. Respir. Physiol. 29, 35-48. Fedde, M.R., Kiley, J.P., Powell, F.L., and Scheid, P. 1982. Intrapulmonary CO2 receptors and control of breathing in ducks: effects of prolonged circulation time to carotid bodies and brain. Respir. Physiol. 47, 121-140. Florindo, L.H., Reid, S.G., Kalinin, A.L., Milsom, W.K., and Rantin, F.T. 2004. Cardiorespiratory reflexes and aquatic surface respiration in the neotropical fish tambaqui (Colossoma macropomum): actue responses to hypercarbia. J. Comp. Physiol. B. 174, 319-328. Furilla, R.A., and Bartlett, D. 1988. Intrapulmonary receptors in the garter snake (Thamnophis sirtalis). Respir. Physiol. 74, 311-322. Gans, C. 1970. Strategy and sequence in the evolution of the gas exchanger of ectothermal vertebrates. Forma et Function 3, 61-104. Gatz, R.N., Fedde, M.R., and Crawford, E.C. 1975. Lizard lungs: CO2-sensitive receptors in Tupinambis nigropunctatus. Experientia 31, 455-456.

Evolutionary Trends in CO2/H+ Chemoreception  |  421 Getchell, T.V., and Shepherd, G.M. 1978. Adaptive properties of olfactory receptors analyzed with odour pulses of varying durations. J. Physiol. 282, 541-560. Gilmour, T.H.J. 1979. Feeding in pterobranch hemichordates and the evolution of gill slits. Can. J. Zool. 57, 1136-1142. Gilmour, K.M., 2001. The CO2/pH ventilatory drive in fish. Comp. Biochem. Physiol. A. 130, 219-240. Gilmour, K.M. and Perry, S.F. 2007. Branchial chemoreceptor regulation of cardiorespiratory function. In: Fish Physiology v 25 Sensory Systems Neuroscience, B. Zielinski and T.J. Hara (Eds.), Academic Press, San Diego, pp. 97-151. Gilmour, K.M., Milsom, W.K., Rantin, F.T., Reid, S.G., and Perry, S.F. 2005. Cardiorespiratory responses to hypercarbia in tambaqui (Colossoma macropomum): chemoreceptor orientation and specificity. J. Exp. Biol. 208, 1095-1107. Glass, M.L., and Johansen, K. 1976. Control of breathing in Acrochordus javanicus, an aquatic snake. Physiol. Zool. 49, 328-340. Goniakowska-Witalinska, L., Zaccone, G., Fasulo, S., Mauceri, A., Licata, A., and Youson, J.H. 1995. Neuroendocrine cells in the gills of the bowfin Amia calva. An ultrastructural and immunocytochemical study. Folia Histochemica et Cytobiologica 33, 171-177. Gonzalez, C., Almaraz, L., Obeso, A., and Rigual, R. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol. Rev. 74, 829-898. Graham, J.B. 1997. Air-breathing fishes: evolution, diversity and adaptation. Academic Press, San Diego. Graham, J.B., and Baird, T.A. 1982. The transition to air breathing in fishes: I. Environmental effects on the facultative air breathing of Ancistrus chagresi and Hypostomus plecostomus (Loricariidae). J. Exp. Biol. 96, 53-67. Gratz, R.K. 1979. Ventilatory response of the diamondback water snake, Natrix rhombifera to hypoxia, hypercapnia and increased oxygen demand. J. Comp. Physiol. B. 129, 105-110. Gutmann, W.F. 1981. Relationships between invertebrate phyla based on functional mechanical analysis of the hydrostatic skeleton. Amer. Zool. 21, 63-81. Halpern, M., and Kubie, J.L. 1984. The role of the ophidian vomeronasal system in speciestypical behavior. Trends Neurosci. 7, 472-477. Hansen, A., and Reutter, K. 2004. Chemosensory systems in fish: structural, functional and ecological aspects. In: The Sense of Fish. Adaptations for the Reception of Natural Stimuli, G. von der Emde, J. Mogdans and B.G. Kapoor (Eds.), Kluwer Academic, Boston, pp. 55-89. Hara, T. J. 2007. Gustation. In: Fish Physiology v 25 Sensory Systems Neuroscience, T.J. Hara and B. Zielinski (Eds.), Academic Press, San Diego, pp. 45-96. Heman-Ackah, Y.D., and Goding, G.S. 2000. Effects of intralaryngeal carbon dioxide and acetazolamide on the laryngeal chemoreflex. Ann. Otol. Rhinol. Laryngol. 109, 921-928. Hempleman, S.C., and Posner, R.G. 2004. CO2 transduction mechanisms in avian intrapulmonary chemoreceptors: experiments and models. Respir. Physiol. Neurobiol. 144, 203-214. Hempleman, S.C., Powell, F.L. and Prisk, G.K. 1992. Avian arterial chemoreceptor responses to steps of CO2 and O2. Respir. Physiol. 90, 325-340. Hempleman, S.C., Rodriguez, T.A., Bhagat, Y.A., and Begay, R.S. 2000. Benzolamide, acetazolamide, and signal transduction in avian intrapulmonary chemoreceptors. Am. J. Physiol. 48, R1988-R1995. Hempleman, S.C., Adamson, T.P., Begay, R.S., and Solomon, I.C. 2003. CO2 transduction in avian intrapulmonary chemoreceptors is critically dependent on transmembrane Na+/H+ exchange. Am. J. Physiol. 53, R1551-R1559.

422  |  Airway Chemoreceptors in the Vertebrates Hidaka, I. 1970. The effect of carbon dioxide on the carp palatal chemoreceptors. Bull. Japan. Soc. Sci. Fish. 36, 1034-1039. Hughes, B., and Shelton, G. 1962. Respiratory mechanisms and their nervous control in fish. Adv. Comp. Physiol. Biochem. 1, 275-364. Ishii, K., Ishii, K., and Dejours, P. 1986. Activity of vagal afferent fibers innervating CO2-sensitive receptors in the tortoise, Testudo hermanni. Jpn. J. Physiol. 36, 1015-1026. Iturriaga, R. 1993. Carotid body chemoreception: the importance of CO2-HCO3- and carbonic anhydrase. Biol. Res. 26, 319-329. Jesse, M.J., Shub, C., and Fishman, A.P. 1967. Lung and gill ventilation of the African lungfish Respir. Physiol. 3, 267-287. Johansen, K. 1966. Air breathing in the teleost Symbranchus marmoratus. Comp. Biochem. Physiol. 18, 383-395. Johansen, K., and Lenfant, C. 1968. Respiration in the African lungfish, Protopterus aethiopicus. II. Control of breathing. J. Exp. Biol. 49, 453-468. Johansen, K., Lenfant, C., and Grigg, G.C. 1967. Respiratory control in the lungfish, Neoceratodus fosteri (Krefft). Comp. Biochem. Physiol. 20, 835-854. Johansen, K., Lenfant, C., Schmidt-Nielsen, K., and Petersen, J.A. 1968. Gas exchange and control of breathing in the electric eel, Electrophorus electricus. Z. Vergl. Physiol. 61, 137-163. Johansen, K., Hanson, D., and Lenfant, C. 1970. Respiration in a primitive air breather, Amia calva. Respir. Physiol. 9, 162-174. Jones, D.R., and Milsom, W.K. 1979. Functional characteristics of slowly adapting pulmonary stretch receptors in the turtle (Chrysemys picta). J. Physiol. 291, 37-49. Jones, D.R., and Milsom, W.K. 1982. Peripheral receptors affecting breathing and cardiovascular function in non-mammalian vertebrates. J. Exp. Biol. 100, 59-91. Jonz, M.G., and Nurse, C.A. 2003. Neuroepithelial cells and associated innervation of the zebrafish gill: a confocal immunofluorescence study. J. Comp. Neur. 461, 1-17. Jonz, M.G., Fearon, I.M., and Nurse, C.A. 2004. Neuroepithelial oxygen chemoreceptors of the zebrafish gill. J. Physiol. 560, 737-752. Kardong, K.V. 2002. Vertebrates: Comparative Anatomy, Function, Evolution. 3rd Ed. McGraw Hill, New York. Kinkead, R., and Milsom, W.K. 1996. CO2-sensitive olfactory and pulmonary receptor modulation of episodic breathing in bullfrogs. Am. J. Physiol. 270, R134-R144. Klein, W., Andrade, D.V., Wang T., and Taylor, E.W. 2002. Effects of temperature and hypercapnia on ventilation and breathing pattern in the lizard Uromastyx aegyptius microlepis. Comp. Biochem. Physiol. A. 132, 847-859. Konishi, J., Hidaka, I., Toyota, M., and Matsuda, H. 1969. High sensitivity of the palatal chemoreceptors of the carp to carbon dioxide. Jpn. J. Physiol. 19, 327-341. Kuhlmann, W.D., and Fedde, M.R. 1979. Intrapulmonary receptors in the bullfrog: sensitivity to CO2. J. Comp. Physiol. 132A, 69-75. Kunz, A.L., Kawashiro, T., and Scheid, P. 1976. Study of CO2 sensitive vagal afferents in the cat lung. Respir. Physiol. 27, 347-355. Kusakabe, T. 2002. Carotid labyrinth of amphibians. Microscopy Research and Technique 59, 207226. Lahiri, S., and Forster, R.E. 2003. CO2/H+ sensing: peripheral and central chemoreception. Int. J. Biochem. Cell Biol. 35, 1413-1435. Lahiri, S., Szidon, J.P., and Fishman, A.P. 1970. Potential respiratory and circulatory adjustments to hypoxia in the African lungfish. Fed. Proc. 29, 1141-1148.

Evolutionary Trends in CO2/H+ Chemoreception  |  423 Lewis, S.M. 1975. Awake baboon’s ventilatory response to venous and inhaled CO2 loading. J. Appl. Physiol. 39, 417-422. Lomholt, J.P., and Johansen. K. 1974. Control of breathing in Amphipnous cuchia, an amphibious fish. Respir. Physiol. 21, 325-340. Mallatt, J. 1996. Ventilation and the origin of jawed vertebrates: A new mouth. Zool. J. Linn. Soc. 117, 329-404. McKendry, J.E., and Perry, S.F. 2001. Cardiovascular effects of hypercarbia in rainbow trout (Oncorhynchus mykiss): a role for externally oriented chemoreceptors. J. Exp. Biol. 204, 115125. McKendry, J.E., Milsom, W.K., and Perry, S.F. 2001. Branchial CO2 receptors and cardiorespiratory adjustments during hypercarbia in Pacific spiny dogfish (Squalus acanthias). J. Exp. Biol. 204, 1519-1527. McMahon, B.R., and Burggren, W.W. 1987. Respiratory physiology of intestinal air breathing in the teleost fish Misgurnus anguillicaudatus. J. Exp. Biol. 133, 371-393. Milsom, W.K. 1995a. Regulation of respiration in lower vertebrates: Role of CO2/pH chemoreceptors. In: Advances in Comparative and Environmental Physiology, Volume 21, Mechanisms of Systemic Regulation: Acid-Base Regulation, Ion Transfer and Metabolism, N. Heisler (Ed.), Springer, Berlin, pp. 61-104. Milsom, W. K. 1995b. The role of CO2/pH chemoreceptors in ventilatory control. Braz. J. Med. Biol. Res. 28, 1147-1160. Milsom, W.K. 1998. Phylogeny of respiratory chemoreceptor function in vertebrates. Zoology 101, 316-332. Milsom, W.K. 2002. Phylogeny of CO2/H+ chemoreception in vertebrates. Respir. Physiol. Neurobiol. 131, 29-41. Milsom, W.K., and Jones, D.R. 1977. Carbon dioxide sensitivity of pulmonary receptors in the frog. Experientia 33, 1167-1168. Milsom, W.K., and Jones, D.R. 1979. Pulmonary receptor chemosensitivity and the ventilatory response to inhaled CO2 in the turtle. Respir. Physiol. 37, 101-107. Milsom, W.K., and Jones, D.R. 1980. The role of vagal afferent information and hypercapnia in control of the breathing pattern in Chelonia. J. Exp. Biol. 87, 53-63. Milsom, W.K., and Jones, D.R. 1985. Characteristics of mechanoreceptors in the air breathing organ of the holostean fish, Amia calva. J. Exp. Biol. 117, 389-399. Milsom, W.K., Jones, D.R., and Gabbott, G.R.J. 1981. On chemoreceptor control of ventilatory responses to CO2 in unanesthetized ducks. J. Appl. Physiol. 50, 1121-1128. Milsom, W.K., Reid, S.G., Rantin, F.T., and Sundin, L. 2002. Extrabranchial chemoreceptors involved in respiratory reflexes in the neotropical fish Colossoma macropomum (the tambaqui). J. Exp. Biol. 205, 1765-1774. Milsom, W.K., Abe, A.S., Andrade, D.V., and Tattersall, G.J. 2004. Evolutionary trends in airway CO2/H+ chemoreception. Respir. Physiol. Neurobiol. 144, 191-202. Nattie, E. 1999. CO2, brainstem chemoreceptors and breathing. Prog. Neurobiology 59, 299331. Nielsen, B. 1961. On the regulation of the respiration in reptiles. I. The effect of temperature and CO2 on the respiration of lizards (Lacerta). J. Exp. Biol. 38, 301-314. Nishijima, K., Tsubone, H., and Atoji, Y. 2004. Contribution of free nerve endings in the laryngeal epithelium to CO2 reception in rats. Auton. Neurosci. 110, 81-88. Nolan, W.F., and Frankel, H.M. 1982. Ventilatory responses to CO2 at different body temperatures in the snake, Coluber constrictor. Experientia 38, 943-944.

424  |  Airway Chemoreceptors in the Vertebrates Nurse, C.A. 2005. Neurotransmission and neuromodulation in the chemosensory carotid body. Auton. Neurosci. 120, 1-9. Osborne, J.L., Burger, R.E., and Stoll, P.J. 1977. Dynamic responses of CO2-sensitive avian intrapulmonary chemoreceptors. Am. J. Physiol. 233, R15-R22. Peers, C., and Buckler, K.J. 1995. Transduction of chemostimuli by the type I carotid body cell. J. Membr. Biol. 144, 1-9. Perry, S.F., and Duncker, H.-R. 1980. Interrelationship of static mechnical factors and anatomical structure in lung evolution. J. Comp. Physiol. 138, 321-334. Perry, S.F., and McKendry, J.E. 2001. The relative roles of external and internal CO2 versus H+ in eliciting the cardiorespiratory responses of Salmo salar and Squalus acanthias to hpercarbia. J. Exp. Biol. 204, 3963-3971. Perry, S.F., and Gilmour, K.M. 2002. Sensing and transfer of respiratory gases at the fish gill. J. Exp. Zool. 293, 249-263. Perry, S.F., and Reid, S.G. 2002. Cardiorespiratory adjustments during hypercarbia in rainbow trout Oncorhynchus mykiss are initiated by external CO2 receptors on the first gill arch. J. Exp. Biol. 205, 3357-3365. Perry, S.F., Gilmour, K.M., Swenson, E.R., Vulesevic, B., Chew, S.F., and Ip, Y.K. 2005. An investigation of the role of carbonic anhydrase in aquatic and aerial gas transfer in the African lungfish (Protopterus dolloi). J. Exp. Biol. 208, 3805-3815. Perry, S.F., Euverman, R., Wang, T., Loong, A.M., Chew, S.F., Ip, Y.K., and Gilmour, K.M. 2008. Control of breathing in African lungfish (Protopterus dolloi): a comparison of aquatic and cocooned (terrestrialized) animals. Respire. Physiol. Neurobial 160, 8-7. Peterson, D.F., and Fedde, M.R. 1968. Receptors sensitive to carbon dioxide in lungs of chicken. Science 162, 1499-1501. Pough, F.H. 1969. Physiological aspects of the burrowing of sand lizards (Uma, Iguanidae) and other lizards. Comp. Biochem. Physiol. 31, 869-884. Powell, F.L., Milsom, W.K., and Mitchell, G.S. 1988. Effects of intrapulmonary CO2 and airway pressure on pulmonary vagal afferent activity in the alligator. Respir. Physiol. 74, 285-298. Putnam, R.W., Filosa, J.A., and Ritucci, N.A. 2004. Cellular mechanisms involved in CO2 and acid signaling in chemosensitive neurons. Am. J. Physiol. 287, C1493-C1526. Rahn, H., Rahn, K.B., Howell, B.J., Gans, C., and Tenney, S.M. 1971. Air breathing of the garfish (Lepisosteus osseus). Respir. Physiol. 11, 285-307. Reid, S.G., Sundin, L., Kalinin, A.L., Rantin, F.T., and Milsom, W.K. 2000. Cardiovascular and respiratory reflexes in the tropical fish, traira (Hoplias malabaricus): CO2/pH chemoresponses. Respir. Physiol. 120, 47-59. Remmers, J.E., Torgerson, C., Harris, M.B., Perry, S.F., Vasilakos, K., and Wilson, R.J.A. 2001. Evolution of central respiratory chemoreception: a new twist on an old story. Respir. Physiol. 129, 211-217. Sakakibara, Y. 1978. Localization of CO2 sensor related to the inhibition of the bullfrog respiration. Jpn. J. Physiol. 28, 721-735. Saltys, H.A., Jonz, M.G., and Nurse, C.A. 2006. Comparative study of gill neuroepithelial cells and their innervation in teleosts and Xenopus tadpoles. Cell Tiss. Res. 323, 1-10. Sanchez, A.P., and Glass, M.L. 2001. Effects of environmental hypercapnia on pulmonary ventilation of the South American lungfish. J. Fish Biol. 58, 1181-1189. Sanchez, A.P., Hoffmann, A., Rantin, F.T., and Glass, M.L. 2001a. Relationship between cerebrospinal fluid pH and pulmonary ventilation of the South American lungfish, Lepidosiren paradoxa (Fitz.). J. Exp. Zool. 290, 421-425.

Evolutionary Trends in CO2/H+ Chemoreception  |  425 Sanchez, A., Soncini, R., Wang, T., Koldkjaer, P., Taylor, E.W., and Glass, M.L. 2001b. The differential cardiorespiratory responses to ambient hypoxia and systemic hypoxaemia in the South American lungfish, Lepidosiren paradoxa. Comp. Biochem. Physiol. A, 130, 677-687 Sanchez, A.P., Giusti, H., Bassi, M., and Glass, M.L. 2005. Acid-base regulation in the South American lungfish Lepidosiren paradoxa: Effects of prolonged hypercarbia on blood gases and pulmonary ventilation. Physiol. Biochem. Zool. 78, 908-915. Sanderson, S.L., and Wassersug, R. 1990. Suspension-feeding vertebrates. Sci. Amer. 1990, 96-101. Scheid, P., and Piiper, J. 1986. Inert gas wash-out from tissue: Model analysis. Respir. Physiol. 63, 1-18. Scheid, P., Kuhlmann, W.D., and Fedde, M.R. 1977. Intrapulmonary receptors in the tegu lizard: II. Functional characteristics and localization. Respir. Physiol. 29, 49-62. Shelton, G., Jones, D.R., and Milsom, W.K. 1986. Control of breathing in ectothermic vertebrates. In: Handbook of Physiology, Section 3, The Respiratory System, Vol. II: Control of Breathing, part 2, N.S. Cherniack and J.G. Widdicombe (Eds.), American Physiological Society, Bethesda, MD, pp. 857-909. Smatresk, N.J. 1990. Chemoreceptor modulation of endogenous respiratory rhythms in vertebrates. Am. J. Physiol. 259, R887-R897. Smatresk, N.J. 1994. Respiratory control in the transition from water to air breathing in vertebrates. Amer. Zool. 34, 264-279. Smatresk, N.J., and Azizi, S.Q. 1987. Characteristics of lung mechanoreceptors in spotted gar, Lepisosteus oculatus. Am. J. Physiol. 252, R1066-R1072. Smith, H.W. 1930. Metabolism of the lungfish Protopterus aethiopicus. J. Biol. Chem. 88, 97130. Smyth, D.H. 1939. The central and reflex control of respiration in the frog. J. Physiol. 95, 305-327. Sundin, L., Holmgren, S., and Nilsson, S. 1998. The oxygen receptor of the teleost gill? Acta Zool. 79, 207-214. Sundin, L., Reid, S.G., Rantin, F.T., and Milsom, W.K. 2000. Branchial receptors and cardiorespiratory reflexes in the neotropical fish, Tambaqui (Colossoma macropomum). J. Exp. Biol. 203, 1225-1239. Sundin, L., Burleson, M., Wang, T., Reid, S., Salgado, H., Abe, A., Glass, M., and Milsom, W. 2001. Pulmonary receptors in reptiles: discharge patterns of receptor populations in snakes versus turtles. J. Comp. Physiol. B. 171, 103-111. Tottersall, G.J., Androde, D.V., Brito, S.P., Abe, A.S., and Milsom, W.K. 2006. Regulation of ventilation in the caiman, (Caiman crocodylus); Effect of inspired CO2 on pulmonary and upper airway chemoreceptors, J. Comp. Physiol. B. 176, 125-138. Taylor, E.W., Jordan, D., and Coote, J.H. 1999. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol. Rev. 79, 855-916. Templeton, J.R., and Dawson, W.R. 1963. Respiration in the lizard Crotaphytus collaris. Physiol. Zool. 36, 104-121. Todd, E.S. 1972. Hemoglobin concentration in a new air-breathing fish. Comp Biochem. Physiol. 42A, 569-573. West, N.H., and Van Vliet, B.N. 1993. Sensory mechanisms regulating the cardiovascular and respiratory systems. In: Environmental Physiology of the Amphibians, M.E. Feder and W.W. Burggren (Eds.), The University of Chicago Press, Chicago, pp. 151-182. Yamashita, S., Evans, R.E., and Hara, T.J. 1989. Specificity of the gustatory chemoreceptors for CO2 and H+ in rainbow trout (Oncorhynchus mykiss). Can. J. Fish. Aquat. Sci. 46, 17301734.

426  |  Airway Chemoreceptors in the Vertebrates Yoshii, K., Kashiwayanagi, M., Kurihara, K., and Kobatake, Y. 1980. High sensitivity of the eel palatine receptors to carbon dioxide. Comp. Biochem. Physiol. 66A, 327-330. Youngentob, S.L., Hornung, D.E., and Mozell, M.M. 1991. Determination of carbon dioxide detection thresholds in trained rats. Physiol. Behav. 49, 21-26. Zaccone, G., Ainis, L., Mauceri, A., Lo Cascio, P., Lo Giudice, L., and Fasulo, S. 2003. NANC nerves in the respiratory air sac and branchial vasculature of the Indian catfish, Heteropneustes fossilis. Acta Histochem 105, 151-163. Zaccone, G., Mauceri, A., and Fasulo, S. 2006. Neuropeptides and nitric oxide synthase in the gill and the air-breathing organs of fishes. J. Exp.Zool. 305A, 428-439. Zielinski, B., and Hara, T.J. 2007. Olfaction. In: Fish Physiology v 25 Sensory Systems Neuroscience, T.J. Hara and B. Zielinski (Eds.), Academic Press, pp. 1-43.