Transduction Channels in Sensory Cells

Transduction Channels in Sensory Cells Edited by S. Frings and J. Bradley

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

Further Titles of Interest M. Futai, Y. Wada, J. Kaplan (eds.)

Handbook of ATPases – Biochemistry, Cell Biology, Pathophysiology 2004

ISBN 3-527-30689-7

C. M. Niemeyer, C. A. Mirkin (eds.)

Nanobiotechnology – Concepts, Applications and Perspectives 2004

ISBN 3-527-30658-7

Y. Yawata

Cell Membrane – The Red Blood Cell as a Model 2003

ISBN 3-537-30463-0

G. Krauss

Biochemistry of Signal Transduction and Regulation, 3rd edition 2003

ISBN 3-527-30591-2

O. von Bohlen und Halbach, R. Dermietzel (eds.)

Neurotransmitters and Neuromodulators 2002

ISBN 3-527-30318-9

Transduction Channels in Sensory Cells Edited by Stephan Frings and Jonathan Bradley

Prof. Dr. Stephan Frings University of Heidelberg Department of Molecular Physiology Im Neuenheimer Feld 230 69120 Heidelberg Germany [email protected]

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Dr. Jonathan Bradley Johns Hopkins University School of Medicine Department of Neuroscience 725 N. Wolfe St. Baltimore, MD 21205 USA [email protected]

Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data: A catalogue record for this book is available from the British Library. Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at http://dnb.ddb.de ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by lax. Cover Illustration Hirmer Verlag, Mu¨nchen Composition Mitterweger & Partner GmbH, Plankstadt Printing betz-druck GmbH, Darmstadt Bookbinding Großbuchbinderei J. Scha¨ffer GmbH & Co. KG, Gru¨nstadt Printed in the Federal Republic of Germany. Printed on acid-free paper ISBN 3-527-30836-9

V

Table of Contents Preface XIII List of Contributers 1

1.1 1.2 1.3 1.4 1.4.1 1.4.2 1.4.2.1 1.4.2.2 1.4.3 1.4.3.1 1.4.3.2 1.4.4 1.4.4.1 1.4.4.2 1.4.4.3 1.4.5 1.4.6 1.4.6.1 1.4.6.2 1.4.6.3

XV

The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans 1 Laura Bianchi and Monica Driscoll

Abstract 1 Introduction 2 Features of the C. elegans Model System 3 Mechanosensation Is a Major Mechanism by Which C. elegans Senses Its Environment 4 Gentle Body Touch 5 The Touch Receptor Neurons 5 Ultrastructural Features of the Touch Receptor Neurons 5 Touch Cell-specific Microtubules 5 The Extracellular Mantle 6 Genetic and Molecular Analysis of Body Touch 7 mec-4 and mec-10 Ion Channel Subunits Form Na+ Channels 7 MEC-4 at the Molecular Level 7 The Candidate Mechanotransducing Channel is a Heteromultimeric Complex 9 MEC-4 and MEC-10 Form a Functional Ion Channel 10 MEC-2 Is a Stomatin-like Protein That May Help Tether the MEC-4/ MEC-10 Channel to the Membrane Bilayer and/or the Cytoskeleton 10 MEC-6 Is a Transmembrane Paraoxonase-like Protein That Controls MEC Channel Activity 11 Intracellular Proteins Needed for Touch Transduction 13 Extracellular Proteins Needed for Touch Transduction 14 MEC-1 14 MEC-5 15 MEC-9 15

VI

Table of Contents

1.4.7 1.4.7.1 1.4.7.2 1.4.8 1.4.8.1 1.4.8.2 1.5 1.5.1 1.5.2 1.5.2.1 1.5.2.2 1.5.2.3 1.6

The MEC Channel Functions Specifically in Neuronal Responses to Gentle Touch 16 Test of a Key Hypothesis 16 Additional Insights Revealed by Imaging In Vivo Ca2+ Changes in Responding Touch Neurons 18 Summary: A Molecular Model for Gentle-touch Sensation 19 How Touch Is Sensed to Elicit a Specific Behavioral Response 19 Notes on the Working Model 19 The C. elegans Degenerin Family: A Global Role of Degenerin Channels in Mechanotransduction? 20 unc-105 20 unc-8 and del-1 21 A Stomatin Partner for the UNC-8 Channel Suggests a Common Composition of Degenerin Channels 21 Trp Channels May Also Contribute to Mechanosensory Functions in C. elegans 23 Fly and Mouse Neuronal DEG/ENaCs Influence Mechanotransduction, Supporting Conserved Roles for This Family of Proteins 24 Concluding Remarks 25 Acknowledgments 26 References 26

2

Transduction Channels in Hair Cells Robert Fettiplace

2.1 2.2 2.2.1 2.2.2 2.2.3 2.3 2.3.1 2.4 2.4.1 2.4.2 2.5 2.5.1 2.6 2.6.1 2.7

Introduction 31 Gating Mechanism: Channel Kinetics 32 Tip Links and Gating Springs 34 Gating Compliance 36 Three-state Channel Schemes 36 Ionic Selectivity 38 Blocking Compounds 40 MET Channel Adaptation 41 Ca2+ Regulation of Adaptation 42 The Function of Adaptation 44 Single-channel Conductance 45 Number of MET Channels Per Stereocilium 47 The MET Channel as a Member of the TRP Family Properties of TRPV Channels 49 Conclusions 49 Acknowledgments 51 References 51

31

48

Table of Contents

3

Acid-sensing Ion Channels 57 Kenneth A. Cushman and Edwin W. McCleskey

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.7.1 3.7.2 3.7.3 3.8 3.9 3.10

Introduction 57 ASICs and the DEG/ENaC Superfamily Amino Acid Structure 61 Assembly Into Channels 61 Pharmacology 62 Gating 63 Proposed Sensory Functions 65 Pain/Nociception 65 Mechanosensation 66 Taste 67 CNS ASICs 67 Stroke 68 Other pH-activated Channels 68 References 69

4

Chemosensory Transduction in Caenorhabditis elegans 73 Noelle LEtoile

4.1 4.1.1 4.1.2 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.4.1 4.4.2 4.4.3 4.4.3.1 4.4.4 4.4.5 4.5 4.6

Introduction 73 The organism C. elegans 73 Introduction to the Channels 75 The Chemosensory Organs 75 The Amphid Organ 75 Phasmid Organ 78 Inner Labial 79 The Sensory Signaling Circuit 79 Behavioral Assays 79 Chemotaxis 80 Repulsion 81 Thermotaxis 82 Social Feeding or Bordering 82 How Is The Response to Each Stimulus Generated? 83 The Chemotaxis Olfactory Response 83 Chemotaxis to Water-soluble Compounds 85 Repellents 85 The ASH Polymodal Sensory Neuron 85 Thermotaxis 86 Feeding Behavior 87 Structure of the TAX, Cyclic Nucleotide-gated Channels of the Worm Channel Regulation 93 References 94

58

88

VII

VIII

Table of Contents

5

5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.2 5.2.1 5.3 5.4 5.5 5.5.1 5.5.2 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15

Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents 99 Johannes Reisert and Jonathan Bradley

Abstract 99 Introduction 99 Tissue 99 Olfactory Receptor Neurons 100 Sustentacular Cells 101 Basal Cells 102 Recording Odor-induced Electrical Activity 102 The Electroolfactogram 102 Odorant Responses of Single Isolated Olfactory Receptor Neurons 103 Components of the Transduction Pathway 106 Cloning of G Proteins Expressed in the OE 108 Gaolf 108 Adenylyl Cyclase 108 Odorant Receptors 109 Cyclic Nucleotide-gated Channel in OE 110 Cloning of a CNG Channel Expressed in the OE 114 Negative Feedback by Ca2+ on the CNG Channel 115 The Olfactory Ca2+-activated Cl– Channel 118 Activation of the Cl– Conductance 119 Single Channel Properties and Channel Densities 121 Regulation of Cl– Channel Activity 122 Amplification of the CNG Current and Generation of the Cl– Current 123 Open Questions 126 References 127

6

Transduction Channels in the Vomeronasal Organ Emily R. Liman and Frank Zufall

6.1 6.2 6.3

Introduction 135 Anatomy of the Vomeronasal System 136 Sensory Responses Involve Generation of Action Potentials and Ca2+ Entry 137 Two Families of G-protein-coupled Receptors Mediate VNO Transduction 139 Signaling Downstream of G Proteins May Involve a PLC 140 Second Messengers for VNO Transduction: Functional Studies 140 Identification of the TRPC2 Ion Channel as a Candidate Transduction Channel for VNO Sensory Signaling 141 TRPC2 Is Essential for Pheromone Transduction 144 Mechanism of TRPC2 Activation 144 TRPC2 Knockout Mice: Behavioral Defects 146 Loss of VNO Signaling Components in Human Evolution 147 Summary: Is TRPC2 the VNO Transduction Channel? 149 Acknowledgements 150

6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12

135

Table of Contents

7

Transduction Mechanisms in Taste Cells Kathryn Medler and Sue C. Kinnamon

7.1 7.2 7.2.1 7.2.1.1 7.2.1.2 7.2.2 7.2.2.1 7.2.2.2 7.2.2.3 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 7.3.5.1 7.3.5.2 7.4

Introduction 153 Ionic Stimuli 155 Salt 155 Epithelial Sodium Channel 155 Amiloride-insensitive Pathway 158 Sour 159 Proton-permeable Channels 161 Proton-gated Channels 161 Proton-blocked Channels 162 Complex Stimuli 163 GPCR Signaling in Taste Cells 163 Store-operated Channels and TRPM5 Cyclic Nucleotide-regulated Channels Ligand-gated Channels 171 Miscellaneous Channels 173 Fat-modulated Channels 173 Water-activated channels 173 Conclusions 174

8

Invertebrate Phototransduction: Multimolecular Signaling Complexes and the Role of TRP and TRPL Channels 179 Armin Huber

8.1 8.2 8.3 8.4 8.5 8.5.1 8.5.2 8.5.3 8.5.4 8.6 8.7

153

164 169

Abstract 179 Introduction 180 Structure of the Drosophila Compound Eye and Its Visual Pigments 181 The Drosophila Phototransduction Cascade Is a Prototypical G-proteincoupled Signaling Pathway 184 Essential Components of the Transduction Pathway Are Organized into a Multimolecular Signaling Complex 186 TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors 189 Identification and Characterization of TRP and TRPL 189 Possible Gating Mechanism 191 Transduction Channels in the Visual Systems of Other Invertebrates 193 Drosophila TRP Is the Founding Member of the TRP Family of Ion Channels 194 Light-dependent Relocation of TRPL Alters the Properties of the Photoreceptive Membrane 196 Concluding Remarks and Outlook 198 Acknowledgments 199

IX

X

Table of Contents

9

The Transduction Channels of Rod and Cone Photoreceptors U.B. Kaupp and D. Tra¨nkner

9.1 9.2 9.2.1 9.2.2 9.2.3 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.4 9.5 9.6 9.6.1 9.6.2 9.7 9.7.1 9.8 9.8.1 9.9 9.9.1 9.9.2

Introduction 207 Brief Overview 207 Ligand Sensitivity 207 Ion selectivity 208 Modulation 208 Function of CNG Channels in Phototransduction and Adaptation 209 Rod and Cone Photoreceptors 209 CNG Channels in Pinealocyte Photoreceptors 212 CNG Channels in Parietal Eye Photoreceptors 213 CNG Channels in Hyperpolarizing Photoreceptors of Invertebrates 214 Structure of Subunits 215 Transmembrane Topology and Subunit Stoichiometry 215 Interaction of CNG Channels With Other Proteins 219 The Glutamic Acid-rich Part (GARP) of B1 219 Interaction with the Na+/Ca2+-K+ Exchanger 220 Modulation 221 Modulation by Ca2+ 221 Phosphorylation 223 Retinal 223 Visual Dysfunction Caused by Mutant CNG Channel Genes 224 Mutations Associated with Retinitis Pigmentosa 225 Mutations Associated with Achromatopsia or Cone Dystrophy 227 Appendix 229

10

Ion Channels and Thermotransduction Michael J. Caterina

10.1 10.2

Introduction 235 Physiological Studies Provide Evidence for the Existence of Thermally Gated Ion Channels 236 Molecular Characterization of a Heat-gated Ion Channel, TRPV1 239 TRPV2 Is an Ion Channel Activated by Extremely Hot Temperatures 241 TRPV3 and TRPV4 Are Warmth-activated Channels 242 TRPM8 and ANKTM1 Are Activated by Cool and Cold Temperatures, Respectively 242 Non-TRP Channels Implicated in Mammalian Temperature Sensation 243 Temperature-sensing Proteins in Non-mammalian Species 244 Mechanisms of Temperature Transduction 245 Conclusions 246

10.3 10.4 10.5 10.6 10.7 10.8 10.9 10.10

207

235

Table of Contents

11

Pain Transduction: Gating and Modulation of Ion Channels 251 Peter A. McNaughton

11.1 11.2 11.2.1 11.2.2 11.2.3 11.2.4 11.2.5 11.2.6 11.2.7 11.3 11.3.1 11.3.1.1 11.3.1.2 11.3.2 11.3.3 11.3.4 11.3.4.1 11.3.4.2 11.3.4.3 11.3.4.4 11.3.4.5 11.3.5 11.3.6 11.3.7 11.4

Introduction 251 Ion Channels Gated by Noxious Stimuli 253 Ion Channels Gated by Noxious Heat 253 Ion Channels Gated by Noxious Cold 254 Ion Channels Gated by Acid 254 ATP-gated Ion Channels 255 Ion Channels Gated by Mechanical Stimuli 256 Initiation of Action Potentials by Noxious Stimuli 256 Summary Diagram of a Nociceptive Terminal 257 Sensitization by Inflammatory Mediators 257 Short-term Sensitization: Mediators and Pathways 258 Bradykinin and the PKC Pathway 258 Prostaglandins and the PKA Pathway 260 Nerve Growth Factor 262 Direct Modulation of TRPV1 by Protons 263 Other Modulators of Nociceptor Sensitivity 264 ATP 264 Proteases 264 Bv8/Prokineticin 264 Glutamate 265 Norepinephrine 265 Long-term Sensitization 265 Gene Expression Regulated by NGF 266 Gene Expression Regulated by GDNF 266 Conclusions 267

12

Transduction and Transmission in Electroreceptor Organs Robert C. Peters and Jean-Pierre Denizot

12.1 12.1.1 12.1.2 12.2 12.2.1 12.2.2 12.2.3 12.2.4 12.2.5 12.3 12.4 12.4.1

271

Abstract 271 Introduction 272 Transduction at Electroreceptor Cells 272 Favorite Species 273 Types of Electroreceptor Organs 274 General 274 The Sensory Mucous Glands in Monotremes 274 The Microampullary Organs 275 The Tuberous Organs 275 The Ampullae of Lorenzini 275 How is Transduction at Electroreceptor Cells Studied? 276 Current Views on Transduction and Transmission in Electroreceptor Organs 276 Spontaneous Activity and Modulation of Afferent Activity 276

XI

XII

Table of Contents

12.4.2 12.4.3 12.4.3.1 12.4.3.2 12.4.3.3 12.4.3.4 12.4.3.5 12.4.4 12.4.5 12.4.5.1 12.4.5.2 12.4.5.3 12.4.5.4 12.5 12.6

Monotreme Mucous Gland Electroreceptor Organs 277 Microampullary Organs in Freshwater Organisms 278 General 278 Patch-clamp Experiments 280 Indirect Pharmacological Evidence 281 The Synaptic Paradox 282 The Transduction Model Revisited 283 Tuberous Organs in Freshwater Fishes 285 Ampullae of Lorenzini in Marine Fish 286 General 286 Ampullae in Plotosus 287 Ampullae in Elasmobranchs 287 The Synaptic Paradox 287 Mucus and Transduction 290 Conclusions and Open Ends 291 Acknowledgments 293

Preface

She loves him, observes the tourist upon beholding the image of the Pharaoh and his wife in the Egyptian Museum in Cairo. And indeed, the intimate scene depicting Tutankhamun and his queen Ankhesenamun (shown on the cover of our book) confers that impression even more than 3000 years after its creation. For the sensory physiologist who recognizes Ankhesenamun’s gesture as a mechanosensory gentle touch, the sensation of a hand touching a shoulder is in molecular terms no simple process. In fact, more than 20 years of hard experimental work was necessary to shed some light on the molecular steps that convert, or transduce, physical contact into an electrical signal interpretable by the nervous system. As Laura Bianchi and Monica Driscoll outline in the first chapter of this book, the path toward understanding touch was paved by a creature much less noble than Tutankhamun: the soil nematode Caenorhabditis elegans. Painstaking genetic analysis of the worm’s response to being experimentally touched and probed with an eyelash led to identification of the transduction channels in mechanosensory neurons, known collectively now as the degenerin family of ion channels. It is these proteins that translate mechanical stimuli into electrical signals that can be processed by the sensory neurons and eventually are interpreted by the organism as a sensory experience. In all sensory cells transduction channels show fascinating adaptations to their task of reporting sensory stimuli. Imagine this: if Ankhesenamun speaks to her husband, or when the Pharaoh listens to his musicians playing cymbals and harp, tiny protein filaments tug at the transduction channels in his inner ear to excite mechanosensory hair cells and to produce a neuronal auditory signal. Robert Fettiplace describes in his chapter the biophysical examination of these exquisitely sensitive transduction channels.

XIV

Preface

In the world of chemoreception, transduction channels appear to be as numerous as the qualities of chemical stimuli. Acid-sensing ion channels respond to the simplest of all chemicals. They are opened by protons and probably serve multiple functions in the body, including the generation of heartache when ischemia turns things sour within the myocardium. Other chemoreception modalities are more conducive to Pharaoh’s bliss. In particular, the metabotropic transduction cascades of taste and smell form the molecular basis of sensory pleasures, which were so highly cherished by the ancient Egyptians that the hieroglyphic determinative for happiness was a nose. The chemoreception chapters in our book describe the state of knowledge about transduction channels in chemosensory cells. Here we meet an entire zoo of different transduction channels, including cyclic nucleotide-gated cation channels, calcium-activated chloride channels, and a channel family that plays an increasingly prominent role in sensory physiology: the transient receptor potential channels. Transient receptor potential channels mediate sensory transduction in systems as diverse as mouse pheromone receptors, insect ommatidia, and human thermoreceptors, apparently acting as one of nature’s multiple-purpose transduction components. Looking at his wife is probably what makes the Pharaoh really happy. And, indeed, the beautiful daughter of Nefertiti must have been an exceptional visual experience. Just look at how the rays of the sun seem to caress her and her husband with tiny hands of light! The old Egyptians surely had a way of representing sensory perception in art. In modern days, we have learned to understand how photoelectrical transduction works in the light-sensitive cells of the retina. Dimitri Tra¨nkner and Benjamin Kaupp describe the pivotal role that transduction channels play in such different visual tasks as looking at stars at night or beholding bright and colorful images such as the one on the book cover. And Armin Huber explains the ingenious method that flies use to achieve high temporal resolution in vision: the formation of multimolecular signaling complexes to rapidly drive transduction channels. If you have ever wanted to know why you can rarely catch the fly that annoys you, read this chapter. It won’t help you in catching the insect, but you will understand why the bug is so fast. In the concluding chapter, Robert C. Peters and Jean-Pierre Denizot discuss a sensory modality that Tutankhamun and Ankhesenamun did not use to perceive the world: electroreception. If not the Pharaoh, another denizen of Egypt is a master of electroreception. A small fish with a long nose, the elephant nose (Gnathonemus petersii), finds his way through murky waters by means of emitting and perceiving electrical signals. The elephant nose is one of the best-studied weakly electric fish, and it has slowly revealed how it does it. It is fascinating to read about the sensory equipment that electric fish employ to feel their way in the dark! Thus, the authors of this book cover many sensory modalities and explain the generation of receptor currents in a wide range of sensory cells. They address their chapters to students of biology, physiology, and medicine, as well as to scientists interested in signal transduction, sensory physiology, and perception. And who knows – even some aficionados of Egyptian archaeology may wish to know more about the Pharaoh’s senses. Baltimore and Heidelberg March 2004

Stephan Frings Jonathan Bradley

XV

List of Contributors

Laura Bianchi Department of Molecular Biology and Biochemistry Rutgers University New Brunswick, NJ 08855 USA Jonathan Bradley Department of Neuroscience Johns Hopkins University School of Medicine 725 N. Wolfe St. Baltimore, MD 21205 USA [email protected] Michael J. Caterina Department of Biological Chemistry and Neuroscience Johns Hopkins University School of Medicine 725 N Wolfe St. Baltimore, MD 21205 USA [email protected] Kenneth A. Cushman Vollum Institute Oregon Health & Science University Portland, OR 97239 USA

Jean-Pierre Denizot UNIC – Unit de Neurosciences Intgratives et Computationnelles CNRS UPR 2191 1 Avenue de la Terrasse 91198 Gif-sur-Yvette France Monica Driscoll Department of Molecular Biology and Biochemistry Rutgers University Nelson Biological Laboratories 604 Allison Road Piscataway, NJ 08855 USA [email protected] Robert Fettiplace Department of Physiology University of Wisconsin Medical School 1300 University Avenue Madison, WI 53706 USA [email protected] Stephan Frings Department of Molecular Physiology University of Heidelberg Im Neuenheimer Feld 230 69120 Heidelberg Germany [email protected]

XVI

List of Contributors

Armin Huber Department of Cell and Neurobiology Institute of Zoology University of Karlsruhe Haid-und-Neu-Str. 9 76131 Karlsruhe Germany [email protected] U. Benjamin Kaupp Institut fu¨r Biologische Informationsverarbeitung Forschungszentrum Ju¨lich Leo-Brand-Straße 52425 Ju¨lich Germany [email protected] Sue C. Kinnamon Department of Biomedical Sciences Colorado State University Ft. Collins, CO 80523 USA [email protected] Noelle D. L’Etoile Department of Psychiatry University of California Center for Neuroscience 1544 Newton Ct. Davis, CA 95616 USA [email protected] Emily R. Liman Department of Biological Science University of Southern California 3614 Watt Way Los Angeles, CA 90089 USA [email protected]

Edwin W. McCleskey Vollum Institute Oregon Health & Science University Portland, OR 97239 USA [email protected] Peter McNaughton Department of Pharmacology University of Cambridge Tennis Court Road Cambridge CB2 1PD UK [email protected] Kathryn Medler Department of Biomedical Sciences Colorado State University Ft. Collins, CO 80523 USA Robert C. Peters Functional Neurobiology Utrecht University Padualaan 8 3584 CH Utrecht The Netherlands [email protected] Johannes Reisert Department of Neuroscience Johns Hopkins University School of Medicine 725 N. Wolfe St. Baltimore, MD 21205 USA [email protected]

List of Contributors

Dimitri Tra¨nckner Institut fu¨r Biologische Informationsverarbeitung Forschungszentrum Ju¨lich Leo-Brand-Straße 52425 Ju¨lich Germany present address: University of California, HHMI 9500 Gilman Drive La Jolla, CA 92093 USA

Frank Zufall Department of Anatomy and Neurobiology University of Maryland School of Medicine Baltimore, MD 21201 USA

XVII

1

1

The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans Laura Bianchi and Monica Driscoll

Abstract

One of the looming mysteries in signal transduction today is the question of how mechanical signals, such as pressure or stretch, are sensed. Elegant electrophysiological studies in organisms ranging from bacteria to mammals support that mechanotransduction can be mediated by ion channels that gate in response to mechanical stimuli. Despite the importance of the molecular identification of these ion channels for elaborating mechanisms of mechanotransduction, genes encoding mechanosensitive ion channels eluded cloning efforts for a long time. Breakthroughs in the understanding of mechanosensitive channels have come from genetic analyses of touch sensation in Caenorhabditis elegans and Drosophila. In C. elegans, screens for touch-insensitive mutants identified two genes, mec-4 and mec-10, that encode channel subunits implicated in touch sensation and are postulated to be the core of a mechanotransducing ion channel complex. mec-4 and mec-10 encode proteins with similarity to subunits of the mammalian amiloride-sensitive epithelial Na+ channel (ENaC) that mediates sodium reabsorption in the kidney and lung. mec-4 is expressed exclusively in six neurons that laser ablation studies have identified as gentle-touch receptors, and mec-10 is expressed in these six neurons plus two pairs of touch receptors that are thought to sense harsher touch. The same genetic screens that identified mec-4 and mec-10 identified other genes required for normal touch sensation in the nematode. MEC-5, a novel collagen, and MEC-9, a protein that includes multiple Kunitz-type protease inhibitor repeats and EGF repeats, are extracellular matrix proteins that may interact with MEC-4/MEC-10 channel subunits on the extracellular side of the neuron to help exert gating tension on the channel. Inside the touch receptor, a specialized cytoskeleton is assembled that features 15-protofilament microtubules composed of MEC-12 a-tubulin and MEC-7 b-tubulin subunits. This cytoskeleton may be linked to tether MEC-4/MEC-10 on the intracellular side. When a mutant hyperactivated MEC-4(d) subunit is heterologously expressed in Xenopus oocytes, voltage-independent Na+ currents are produced that can be modulated in both amplitude and properties by two other proteins also identified by genetic screens as required for touch transduction: MEC-2, a stomatin-like protein, and Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

2

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

MEC-6, a protein that shares similarity with mammalian paraoxonases. The C. elegans genome encodes 28 members of the MEC-4 and MEC-10 channel family, called the degenerin family. We discuss here the global role of degenerins in mechanosensation, reporting findings on the function of three other degenerins (UNC-8, DEL-1, and UNC-105) in mechanosensitive and stretch-sensitive behaviors in the nematode, and we review studies addressing the role of mammalian homologues in touch sensation.

1.1

Introduction

The sense of touch is so profoundly important to our daily life that – when you actually think of it – the degree to which we take this sense for granted is unthinkable. We fully depend on our sense of touch to make and drink our morning coffee, to flip through the newspaper, to dress, and to move to the places where we type, phone, compute, pass paper, fold, sell, and manufacture things. Virtually no activities required for daily life (feeding, drinking, moving, protecting, communicating) can transpire without touch or mechanical sensation. Moreover, without touch sensation we would be unable to ensure the viability of our young. In addition to the obvious reasons for this, it is becoming increasingly clear that touch plays a critical role in both physical and emotional development. For example, hospitalized preterm infants show accelerated weight gain, enhanced activity, and faster development if they are gently stroked daily for 15 minutes – resulting in faster hospital discharge [5]. Despite widespread and fundamental importance, touch is the least understood of the senses, at both the cellular and molecular levels. The sense of touch is initiated by the perception of a mechanical stimulus such as pressure and the conversion of this signal into electrical signaling. Groundbreaking electrophysiological studies characterized ion channels that could be gated in response to pressure or stretch rather than voltage changes or ligand binding [39, 41, 58]. Such channels could be identified in specialized mechanoreceptors [24, 48], yet the genes encoding mechanically gated ion channels that mediate the senses of touch and hearing eluded cloning efforts for years (some genes, such as those encoding the hearing channel, remain unidentified even to this day; see Chapter 2). Technically, this might have been predicted, as there are no known reagents that specifically associate with mechanosensitive channel subunits at high affinity that could facilitate protein isolation and there is a remarkable paucity of mechanically gated channels even in specialized mechanotransducing structures such as the vertebrate cochlea. Moreover, given that these channels are likely to be tethered to accessory proteins that exert gating tension, reconstitution in heterologous systems is extremely difficult. Although the cloning of mechanically gated MscL and MscS channels from bacteria constituted major breakthroughs in the field of mechanical signaling [39], the MscL and MscS channel classes have no clear eukaryotic homologues, and thus their identification did not facilitate an immediate revolution in our understanding of mammalian mechanotransduction.

1.2 Features of the C. elegans Model System

Exciting advances in our understanding of the sense of touch have instead emerged from invertebrate genetics. Both nematode and fly mutants defective in touch sensation have facilitated the cloning of ion channels thought to act directly as mechanotransducing channels. More specifically, the DEG/ENaC Na+ channel subunits (named for the C. elegans degenerins and the related mammalian epithelial amiloride-sensitive Na+ channel) have been directly implicated in touch sensation in both invertebrates and vertebrates. Likewise, members of the transient receptor potential (TRP) channel family are mechanotransducing channels implicated in touch [84] and possibly hearing [50, 69] (see Chapter 2). Here we focus on reviewing the genetic, molecular, electrophysiological, and calcium-imaging studies conducted using the simple nematode C. elegans that have greatly advanced our understanding of touch sensation through the identification and the characterization of mechanically gated DEG/ENaC ion channels and accessory proteins. We discuss how these and TRP channels may work together to contribute to touch sensation and note how data from invertebrates has stimulated a successful search for analogous processes in higher organisms.

1.2

Features of the C. elegans Model System

The 1-mm long simple soil worm Caenorhabditis elegans is a facile system for experimental manipulation that features many developmental and behavioral pathways strikingly conserved between nematodes and mammals. C. elegans can be easily reared on an E. coli diet in the laboratory. This animal completes a reproductive life cycle in just 2.5 days at 25 8C, during which it progresses through embryonic development and four larval stages (L1-L4) before reaching sexual maturity. The most common sexual form is the hermaphrodite (XX), although males (X0), which can arise spontaneously by nondisjunction, can be easily propagated in the laboratory for use in genetic studies. The C. elegans body and eggshell are transparent so that each cell can be visualized by Normarski microscopy. In fact, the entire map of all cell divisions during the development of the animal has been constructed [44, 75]. The nervous system includes only 302 neurons, for which the pattern of synaptic connections, including circuits for specific mechanosensory behaviors, has been deduced using serial section electron microscopy [85]. Laser ablation experiments have helped define the importance of specific identified neurons in mechanosensory behaviors [14, 47]. The major advantage of using C. elegans as a model system for studying biological processes is that it is a powerful genetic system [8]. Mutations that affect development and behavior, including those affecting touch sensation, have been generated and mapped to specific genes. Sequence analysis of the C. elegans genome is complete [20], and powerful methods for generation of transgenic animals [26] and dsRNAimediated transcript disruption (RNAi) [27, 77] are routine. Despite the considerable advantages that C. elegans offers for studying gene function in vivo, this model has had certain limitations for electrophysiological analysis of chan-

3

4

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

nel function, especially for channels expressed in neurons. The tiny neurons (1–2 lm diameter) are embedded in poorly accessible tissues confined in a pressurized cuticle. However, recent technical improvements established in the field have led to the development of electrophysiological methods for characterizing channel function in C. elegans [35]. In addition, a recently developed method for culturing C. elegans cells now allows routine electrophysiological recordings from neurons, muscles, and other cell types [19, 76]. Finally, sophisticated methods for monitoring intracellular calcium concentration changes during channel activity in living and behaving nematodes have been developed and have led to important findings [49, 76]. Taken together, the identification of specialized mechanosensory neurons, the cloning of genes required for mechanosensitive responses, and the study of their function both in vivo and in vitro have led to significant insight into the molecular mechanisms of mechanotransduction in C. elegans.

1.3

Mechanosensation Is a Major Mechanism by Which C. elegans Senses Its Environment

C. elegans does not have a sense of sight and must evaluate its environment primarily by chemosensation and mechanosensation (see Chapter 4). C. elegans can respond to a range of mechanical stimuli encountered virtually anywhere on its body. The bestcharacterized mechanosensitive behavior is the movement away from a gentle brush of an eyelash hair delivered to the body, generally referred to as gentle-touch sensation [13]. Other mechanosensitive behaviors include response to head-on collision with an object (the nose touch response), response to light touch to the side of the nose (head withdrawal response), response to harsh touch delivered by a metal wire, and response to tapping on the plate on which the worms are reared. The process by which males mate most likely involves touch-mediated recognition of the hermaphrodite vulva. Mechanical stimuli also impact on locomotory behaviors, foraging, feeding, egg laying, and defecation circuits. Because the avoidance of gentle touch is the behavior most intensively investigated, here we will first focus on summarizing how the study of gentle touch has produced a detailed molecular model for a mechanically gated ion channel. Later, we will review what is known about the identities of genes that influence other mechanosensory behaviors, and we will consider emerging molecular themes in touch sensation.

1.4 Gentle Body Touch

1.4

Gentle Body Touch 1.4.1

The Touch Receptor Neurons

In the laboratory, C. elegans moves across an agar plate on its side with a readily observed sinusoidal motion. When stroked with an eyelash hair on the anterior body, the animal will reverse its direction and move backwards; if touched on the posterior body, it will move forward [13]. The neurons required for the sensation of the gentle-touch stimuli have been identified by laser ablation studies and genetic disruption. These six touch receptor neurons were initially called microtubule cells because their processes are filled with distinctive 15-protofilament microtubules. Their processes are embedded in the hypodermis adjacent to the cuticle (the worm “skin”) and run longitudinally along the body wall, a distribution that enables them to more or less “cover” the touch sensory field of most of the body. Two embryonically generated PLM neurons (posterior lateral microtubule cells) are situated in the posterior body, on the right and left sides; two embryonically generated ALM neurons (anterior lateral microtubule cell) are situated in the anterior, on the right and left sides. In the first larval stage, AVM (anterior ventral microtubule cell) and PVM (posterior ventral microtubule cell) are added to the body plan. Laser ablation of individual ALMs, PLMs, and AVM established roles for these neurons in gentle touch [14]. Although PVM looks identical to the other touch neurons, it does not initiate a behavioral response to gentle touch on its own, and thus it has been postulated to modulate other behavioral circuits that can be influenced by touch [14] (Fig. 1.1A).

1.4.2

Ultrastructural Features of the Touch Receptor Neurons Touch Cell-specific Microtubules Touch receptor processes are filled with bundles of wide-diameter (15-protofilament, pf) microtubules that are uniquely assembled in this group of six neurons [15, 16]. Most other nematode cells include 11-pf microtubules (Fig. 1.1C). The 15-pf microtubules are required for touch receptor function: if microtubules are disrupted by the microtubule assembly inhibitor colchicine or by genetic mutations, touch sensitivity is completely lost [13, 15]. Individual microtubules are not long enough to extend from end to end of the touch neuron. Rather, single microtubules (10-20 lm long) overlap with each other to span the full length of the touch cell processes (about 400-500 lm). Interestingly, the distal microtubule end is diffusely stained and is always situated outside of the microtubule bundle, often positioned adjacent to the plasma membrane. This ultrastructural feature suggests that the oriented microtubule network might associate with plasma membrane proteins, such as the mechanosensitive ion channels that sense touch, a hypothesis that remains to be tested [16] (see discussion of mechanotransduction model below). 1.4.2.1

5

6

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

A

C cuticle touch receptor process mantle

hypodermis 15-pf microtubules

B

D cuticle

mantle

touch receptor process

15-pf microtu

hypodermis

Fig. 1.1 C. elegans neurons that sense gentle body touch. (A) Diagram showing the position of the six neurons that in C. elegans sense the gentle stroke of an eyelash hair on the body; anterior body is to the left. There are two fields of touch sensitivity defined by the position of the touch neurons processes along the body axis. The ALMs and AVM sense touch to the anterior field, whereas PLMs sense touch to the posterior field. (B) Touch neurons are here visualized in a living nematode, by expression of the Green Fluorescent Protein under the control of the mec-4 promoter, which is active exclusively in these neurons. Arrows point to touch receptor cell bodies. (C) Electron micrograph of a cross-section of a touch receptor neuron process. The touch cell process, which is surrounded by the mantle and

embedded in the hypodermis, is filled with 15-pf microtubules and is in very close proximity to the cuticle. This anatomical arrangement is thought to ensure the transmission of the mechanical forces applied on the cuticle down to the touch neuron process. (D) Schematic representation of a touch receptor neuron EM cross-section, depicting its most important components. The darkly stained region, depicted here as a bar-shaded rectangle connecting the mantle and the cuticle, is the fibrous organelle (not visible in the electron micrograph). Such specializations occur periodically along the length of the touch receptor process and may serve to attach the process to the cuticle. Adapted from [78]

The Extracellular Mantle Touch receptor processes are surrounded by a specialized extracellular matrix, called the mantle, which appears to help maintain the touch receptor process in close association with the cuticle [13]. Cuticular structures resembling muscle attachment sites are positioned periodically along the length of the touch receptor process in close contact with the mantle and may be sites at which the touch receptor process is fixed to the cuticle (Fig. 1.1D). Although genetic mutations support that the integrity of the mantle is critical for touch receptor function, mutations in him-4 cause touch neurons to stray away from the cuticle, yet the mutants still sense touch [82]. Since detachment is variable in the him-4 background, it is possible that adequate contact is maintained 1.4.2.2

1.4 Gentle Body Touch

for some touch sensation; alternatively, any deflection of even a “loose” mechanoreceptor neuron might be sufficient to activate the behavioral avoidance response.

1.4.3

Genetic and Molecular Analysis of Body Touch

In pioneering studies on the genetics of touch sensation, Martin Chalfie and colleagues mutagenized animals and screened their progeny for the failure to respond to the gentle brush of an eyelash hair [11, 13]. The mutants selected exhibited grossly normal locomotion and were still able to respond to the prod of a metal wire, so that defects appeared to specifically alter gentle-touch sensitivity. Hundreds of touch-insensitive mutants, many of them designed as mec (mechanosensory abnormal), defined several genes that contribute specifically to touch cell development and function. It should be emphasized that since the criteria for mutant isolation demanded that other aspects of nematode locomotion and harsh-touch sensation be unaffected by the mutations, genes that encode proteins used for gentle-touch sensation but also used in other locomotory activities would not have been identified in this screen. Likewise, genes that encode functionally redundant proteins would be missed. Nonetheless, the genes identified in this screen provided a major breakthrough in our understanding of the molecules needed for touch sensation. mec-4 and mec-10 Ion Channel Subunits Form Na+ Channels mec-4 and mec-10 loss-of-function mutants are touch-insensitive, yet their touch receptor neurons appear to develop normally and share all apparent ultrastructural features of wild-type (WT) touch receptor neurons [13]. Cloning revealed that mec-4 and mec-10 encode homologous proteins related to subunits of the amiloride-sensitive, voltageindependent Na+ channel, which mediates Na+ reabsorption in vertebrate kidney, intestine, and lung epithelia (the ENaC channel [9, 10, 12, 22, 45, 52]). The mec-4 channel subunit is expressed only in the six touch receptor neurons (Fig. 1.1B [55]), and the mec-10 channel subunit is expressed in the six touch receptor neurons as well as in two other neuron pairs that may mediate stretch-sensitive or harsh-touch responses (FLPL/R and PVDL/R [45]). Because the MEC-4 and MEC-10 subunits are expressed exclusively in touch neurons and are clearly needed for the function of these neurons, and because no other channel genes were identified among touch-insensitive mutants, it was proposed that the MEC-4 and MEC-10 subunits assemble in vivo to create a mechanically gated channel that responds directly to touch. Progress toward addressing this hypothesis is outlined in more detail below. 1.4.3.1

MEC-4 at the Molecular Level There are many more mec-4 mutations than there are mec-10 mutations (perhaps suggesting that mec-4 plays a more central role in gentle touch), and thus MEC-4 structure/function is better understood. MEC-4 is a 768-amino-acid membrane protein that includes two membrane-spanning domains (MSDI, MSDII; see Fig. 1.2B). The chan1.4.3.2

7

8

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

Fig. 1.2 Degenerin MEC-4 structure/function. (A) Dendrogram of the 28 degenerins encoded by the C. elegans genome. The 28 genes encoding postulated degenerin subunits were identified by searching the C. elegans database, compiled by the C. elegans Genome Sequencing Consortium, for predicted proteins sharing homology with known degenerins. Black background indicates the most characterized degenerins, including MEC-4. (B) Structural features of a single MEC-4 subunit (likely four subunits form a channel). The MEC-4 polypeptide spans the membrane twice, leaving the Nand C-termini in the cytosol. The second membrane-spanning domain, which is longer than required for a single transmembrane pass, may loop back in the membrane to participate in the formation of the pore. Ala713, which when replaced by a bulkier amino acid results in necrotic cell death, is

indicated by the skull and crossbones icon. MEC-4 protein also features three cysteine-rich domains (CRDI, II, III) that are thought to be involved in protein-protein interactions, perhaps anchoring MEC-4 to extracellular matrix proteins. Other important domains include the putative extracellular regulatory domain, the neurotoxin-related domain, and the intracellular regulatory domain. (C) Model for mec-4(d)-induced toxicity. WT MEC-4 channels are able to open and close, but MEC-4(d) channels, which encode substitutions for a conserved alanine adjacent to MSDII, are thought to be “locked” in an open conformation due to steric hindrance. This is thought to result in excessive Na+ influx that triggers necrotic-like cell death, which manifests itself in the early stages as cell swelling (lower right panel)

1.4 Gentle Body Touch

nel subunit is positioned in the membrane such that relatively short N- and C-terminal domains project into the cytosol and a single large central loop extends extracellularly [52] (this is typical of all DEG/ENaC family members). The MEC-4 extracellular domain also includes three cysteine-rich domains (CRDI, CRDII, and CRDIII) and one region similar to venom neurotoxins (NTD) [79]. Understanding of structure/function relations in MEC-4 is still at an early stage, but studies on this and other members of the DEG/ENaC superfamily have highlighted three conserved regions important for function: (1) MSDII contributes to the channel pore [42]; (2) a short but highly conserved intracellular stretch adjacent to MSDI influences ion permeation and selectivity [36, 43]; and (3) the Cys-rich extracellular loop domains are important for function in some way, possibly mediating protein-protein interactions that may help tether the MEC-4 channel to the specialized extracellular matrix of the touch neuron. An unusual type of mec-4 mutation acts dominantly to induce swelling and neurodegeneration of the touch neurons. Substitution of large side-chain amino acids for a highly conserved Ala residue situated adjacent to channel pore MSDII (AA713; see Fig. 1.2B [22, 52]) generates MEC-4 mutant subunits (named MEC-4(d)) that induce necrotic-like death of the touch receptor neurons (Fig. 1.2C [11, 13]). The channel pore must be intact for neurodegeneration to occur [42], suggesting that ion influx is critical in the toxicity mechanism. Since large side-chain amino acids at the conserved “d” position are toxic but small ones are not, it was originally proposed that these large substitutions favor the channel-open conformation and hyperactivate ion influx [22, 42]. Indeed, the A713V substitution markedly enhances whole-cell currents when MEC-4(d) channel activity is measured in the Xenopus oocyte expression system [34]. Since other C. elegans family members (e.g., deg-1 and mec-10) can be altered by analogous amino acid substitutions to induce neurodegeneration [17, 45], the C. elegans branch of the gene family has been named the “degenerin” family (Fig. 1.2A). A mutant variant of neuronally expressed mammalian DEG/ENaC member MDEG (ASIC2), engineered to encode a large side-chain amino acid at the corresponding position, induces swelling and death when introduced in to Xenopus oocytes and hamster embryonic kidney cells [83]. The small amino acid normally situated at the “DEG” site can be modified with chemical reagents only when the channel is activated, supporting that conformational changes associated with an open channel involve this residue [1].

1.4.4

The Candidate Mechanotransducing Channel is a Heteromultimeric Complex

The subunit compositions and stoichiometry of DEG/ENaC channels remain somewhat uncertain. Electrophysiological assays of the rat ENaC channel reconstituted in Xenopus oocytes determined that at least three homologous subunits (a-, b-, and crENaC) must be co-expressed to form a channel with pharmacological properties similar to the in vivo channel [10]. Stoichiometries of four to nine subunits per ENaC channel have been supported [3, 6, 7, 21, 28, 51, 70]. Genetic interactions suggest that

9

10

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

MEC-4 and MEC-10, which cannot functionally complement one another and thus appear to perform distinct functions in vivo, form a heteromeric channel in touch neurons: engineered mec-10(d) subunit (harboring the substitution analogous to channel-activating, death-inducing MEC-4 substitution A713V) requires functional mec-4 to be toxic [37, 45]. MEC-4 and MEC-10 Form a Functional Ion Channel MEC-4 and MEC-10 co-assemble in Xenopus oocytes to form a Na+-selective channel sensitive to the ENaC-blocking agent amiloride. This channel exhibits a high permeability to lithium, as do other members of the DEG/ENaC superfamily. Interestingly, while MEC-4 can form channels of low conductance on its own, MEC-10 is not functional when expressed alone. However, the co-introduction of the MEC-10 subunit to the MEC-4(d) channel in oocytes affects the Kd for amiloride, consistent with MEC-10 being included in the same channel as MEC-4 [34]. Still, the introduction of MEC-10 in the oocyte system does little to change most properties of the MEC-4 channel, and the “MEC-10(d)” mutant subunit cannot conduct current on its own. Thus, the MEC-4 subunit appears most critical for channel properties. Importantly, in Xenopus oocytes the MEC-4/MEC-10 channel has not been demonstrated to be gated by mechanical forces (membrane stretch induced by hypotonic solutions), probably due to the lack of intracellular and extracellular proteins, normally present in vivo, that are essential for channel gating (see below). The MEC-4(d) subunit conducts much more current than the MEC-4(+) subunit (at least 10 times larger currents), consistent with the idea that the “d” substitutions next to the channel pore hyperactivate the channel. Most electrophysiological studies have therefore concentrated on the activated MEC-4(d) channel, which conducts markedly more robust current than the MEC-4(+) subunit. 1.4.4.1

1.4.4.2 MEC-2 Is a Stomatin-like Protein That May Help Tether the MEC-4/MEC-10 Channel to the Membrane Bilayer and/or the Cytoskeleton

MEC-2 May Participate in Several Protein Interactions in the Touch Channel Complex Mechanosensitive ion channels are thought to be gated by forces exerted upon the channels via associated protein attachments. MEC-2 is a candidate protein that might help exert gating tension on the MEC-4/MEC-10 channel from the membrane and/or from the intracellular side. mec-2 encodes a 481-amino-acid protein expressed in the touch receptor neurons and in a few additional neurons in the head [46]. There are three candidate protein interaction domains in MEC-2: (1) a cytoplasmically situated N-terminal domain (positioned between aa 42 and aa 118) needed for the localization of MEC-2::lacZ enzymatic activity to the touch receptor process; (2) a central domain that exhibits 65 % identity to the human red blood cell protein stomatin, an integral membrane protein that associates with the RBC cytoskeleton and affects ion balance via an unknown mechanism [74] (note that the stomatin-related domain includes a hydrophobic stretch that is membrane-associated, but most of this domain is thought to project into the cytoplasm); and (3) a C-terminal intracellular proline-rich region that

1.4 Gentle Body Touch

is similar to SH3-binding domains that mediate protein interactions. That MEC-2 might serve as part of a link between the channel and the specialized microtubule cytoskeleton is suggested by the facts that specific MEC-12 mutant forms (although not the MEC-12 null mutant) prevent MEC-2::lacZ activity from being localized out to the touch neuron processes (the normal MEC-2::lacZ distribution for the wild-type background [46]) and specific combinations of mutant MEC-2 stomatins and hyperactivated channel subunit MEC-10(d) alter the severity of MEC-10(d)-induced neurodegeneration [37]. MEC-2/stomatin can physically associate with MEC-4 and MEC-10 in heterologous assay systems ([34] and our unpublished results), but associations with MEC-12 a-tubulin have not been documented. MEC-2 also appears to homo-oligomerize in vivo because specific combinations of two distinct MEC-2 mutant proteins in heteroallelic strains can influence touch responses in different ways [23]; other stomatin family members have been shown to oligomerize [71]. In sum, MEC-2 might set up a membrane microdomain raft environment critical to the function of the touch channel complex. As noted above, whether MEC-2 actually participates in a link to the cytoskeleton remains to be determined. MEC-2 Enhances MEC-4/MEC-10 Channel Activity Based on the sequence similarity MEC-2 shares with stomatin, (known to influence ion balance in RBCs [72, 73]), it was postulated that MEC-2 may affect the function of the MEC-4/MEC-10 channel. Assays of the MEC-4(d) channel expressed in Xenopus oocytes confirm this hypothesis [18, 34] (Fig. 1.3). Co-expression of MEC-2 with MEC4 and MEC-10 (both WT and (d) variants) increases current amplitude up to 40-fold over channels lacking MEC-2, yet the presence of MEC-2 does not change the total number of channels present at the oocyte surface. While the MEC-2 stomatin-like domain alone is sufficient to boost channel activity, the intact protein is required for the full effect. Interestingly, human stomatin is also able to modulate the MEC-4/MEC-10 channel complex, suggesting that stomatins regulate DEG/ENaC channels by a conserved mechanism. MEC-2 seems to participate in, or influence the formation of, the channel pore, since co-expression of MEC-2 alters lithium permeability of the channel complex (channels become less permeable to Li+ but remain permeable to Na+). The mechanism by which MEC-2 regulates MEC-4/MEC-10 channel activity remains unclear and may involve effects on single-channel conductance and/or open probability.

MEC-6 Is a Transmembrane Paraoxonase-like Protein That Controls MEC Channel Activity Recessive mec-6 alleles can confer insensitivity to gentle touch and also can completely block mec-4(d)-induced death, suggesting that mec-6 is needed in some way for in vivo MEC channel activity [13, 18, 40]. mec-6 encodes a 377-amino-acid transmembrane protein with a small intracellular N-terminus and a large extracellular C-terminus that has the potential for glycosylation. Over a stretch of 250 amino acids at its C-terminus, MEC-6 exhibits 25 % identity and 45 % similarity to vertebrate paraoxonase/ arylesterases. Vertebrate serum paraoxonase (PON1), and perhaps other mammalian 1.4.4.3

11

12

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

Fig. 1.3 Summary of currents generated by expression of combinations of MEC proteins in Xenopus oocytes. Current amplitude at –85 mV is shown here for the subunit compositions indicated by the grid. Complementary RNA corresponding to each subunit was synthesized in vitro and injected into Xenopus oocytes (10 ng per oocyte for mec-4(d), mec-10(d) and mec-2 and 1 ng per oocyte for and mec-6). Currents were recorded in voltage-clamp 4–10 days after injection and data are expressed as mean  SE. Adapted from [18]

paraoxonases, are esterases and are thought to protect against cellular damage from toxic agents such as oxidized lipids in the plasma like low-density lipoproteins (LDL). As such, mammalian paraoxonases have been implicated in the prevention of atherosclerosis and coronary heart disease. Nematode MEC-6 may modify some component of the MEC channel to influence its activity, although no evidence for such a role has yet been identified. MEC-6 is expressed broadly in neurons, muscles, and the canal cell (the kidney of the worm) in an expression pattern that overlaps with multiple degenerin family members, and its activity appears to be needed for toxicity of mec-4(d) expressed in many different cells [40]. Thus, mec-6 is likely required for function of many degenerin ion channels encoded by the C. elegans genome. Whether mammalian paraoxonases are needed for DEG/ENaC channel activity has yet to be determined.

1.4 Gentle Body Touch

MEC-6 Potentiates Channel Activity Co-expression of MEC-6 and MEC-4(d) in Xenopus oocytes increases current amplitude up to 24-fold greater than when MEC-4(d) is expressed alone [18] (Fig. 1.3). By contrast, MEC-6 does nothing to alter the inability of MEC-10(d) to form functional channels on its own, suggesting that the MEC-4 subunit is always required for the formation of conducting channels. Co-immunoprecipitation of heterologously expressed subunits in CHO cells suggests that MEC-6 can associate with MEC-4, MEC-10, and MEC-2, as would be expected if all subunits act together in a single functional channel complex. MEC-4, MEC-10, and MEC-2 protein levels at the plasma membrane are not affected by the addition of MEC-6 to an oocyte expression system. Thus, MEC-6 is likely to increase current amplitude acting at the single channel level, perhaps increasing conductance and/or open probability, but this still remains to be established. MEC-6 can affect channel selectivity lowering lithium permeability, indicating that it may directly participate in the channel pore or that it may affect pore structure through interaction with other sites. Interestingly, effects of MEC-2 and MEC-6 on current amplitude are synergistic: when these subunits are co-expressed with MEC-4(d), they increase current amplitude up to 400-fold (Fig. 1.3). MEC-10(d) does not have a major impact on current magnitude or properties when MEC-2, MEC-4(d), and MEC-6 are co-expressed (except for the Kd of amiloride). The synergy afforded by MEC-2 and MEC-6 suggests that these two subunits might increase current amplitude through different mechanisms, and the transmembrane topologies suggest that MEC-6 has potential for more interaction with extracellular MEC-4/MEC-10 domains, whereas MEC-2 would have to interact with intracellular channel domains. Since they both include membrane-associated regions, both MEC-2 and MEC-6 could influence the transmembrane pore via the membrane as well. Taken together, the model of the touch channel that emerges is that core-conducting subunit MEC-4 associates with degenerin MEC-10, stomatin-related MEC-2, and paraoxonase-related MEC-6. These findings reveal mutual interactions between channel subunits and accessory proteins that form a heteromultimeric complex whose regulation and mechanism of function is just starting to be deciphered.

1.4.5

Intracellular Proteins Needed for Touch Transduction

The touch receptor processes are filled with bundled 15-protofilament microtubules. Mutations in two genes, mec-7 and mec-12, disrupt the production of these microtubules, often eliminating them entirely [11, 13]. Still, even in the absence of the 15-pf microtubules, the touch receptor processes grow out normally and become filled with 11-pf microtubules [15]. Such touch receptors do not function, suggesting that the extensively cross-linked 15-pf microtubules contribute a specific role in touch transduction. mec-7 encodes a b-tubulin [64] expressed at high levels in the touch receptor neurons [38, 65]. Like other tubulins, MEC-7 is highly conserved: it differs from other b-tubu-

13

14

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

lins by only seven amino acids in the variable C-terminal domain. mec-7 mutations isolated in the screen for touch-insensitive mutants range in severity from recessive to strongly dominant, and many of the amino acid changes that disrupt MEC-7 function have been defined [64, 65]. Mutations that disrupt touch sensitivity affect the domain for GTP binding and hydrolysis, sites for heterodimerization with a-tubulin, and sites for higher-order microtubule assembly. mec-12 encodes an a-tubulin expressed in the touch receptor neurons but also expressed in several other neurons that do not assemble 15-pf microtubules [38]. MEC12 is acetylated (Lys40) in the touch neurons but not in other neurons, suggesting that the modification may be important for 15-pf microtubule assembly [30]. Acetylation of MEC-12 also occurs in cultured touch neurons (Bianchi and Driscoll, unpublished observations), suggesting that correct processing of this protein and assembly of 15-pf microtubules do not require the presence of the extracellular mantle proteins normally produced by the hypodermis, such as collagen MEC-5 [23]. As is the case for mec-7, many mec-12 mutations are semi-dominant or dominant and are likely to disrupt subunit interactions or protofilament assembly. Work on mec-7 and mec-12 strongly supports that unique a and b tubulins assemble to form the 15-pf microtubules required for touch receptor function. Whether these specialized microtubules play a direct role in the function of the mechanotransducing complex, perhaps by participating in a direct linkage to the channel, remains to be determined.

1.4.6

Extracellular Proteins Needed for Touch Transduction MEC-1 mec-1 is needed for proper formation of the mantle, the touch neuron’s specialized extracellular matrix. In mec-1 mutants [13] the mantle is almost completely absent and touch neuron processes fail to attach to the cuticle. The failure to attach the touch receptor to the cuticle might be the reason touch cannot be transduced, although the observation that touch cell processes detach sporadically in a him-4 mutant background without affecting touch sensitivity [82] does somewhat challenge this working hypothesis. mec-1 encodes a 1999-amino-acid polypeptide with an N-terminal signal sequence followed by a Kunitz-type domain, two EGF domains, 14 additional Kunitz-type domains, and a C-terminus of 160 amino acids. The Kunitz and EGF domains are likely to be protein-protein interaction domains. Mutations in the N-terminal region through the sixth Kunitz domain affect wrapping of the touch neuron by the surrounding hypodermis and disrupt touch sensitivity, whereas mutations in the C-terminal region affect only touch transduction. This indicates that the C-terminus of MEC-1 is essential for mechanosensory function but not for attachment and engulfment of the touch cell processes. 1.4.6.1

1.4 Gentle Body Touch

MEC-5 mec-5 encodes a novel collagen that is secreted by hypodermal cells and is a component of the touch neuron mantle [23]. mec-5 mutations affect the mantle in a subtle manner. While normally the mantle can be stained with peanut lectin, in mec-5 mutants this staining fails ([13]; E. Hedgecock and M. Chalfie, unpublished). The central portion of the mec-5 protein is made up of Pro-rich Gly-X-Y repeats. mec-5 mutations (many of which are temperature-sensitive, ts) cluster toward the carboxy terminus of the protein and affect these repeats. No touch-insensitive mutations map to the amino- and carboxy-termini; therefore, the role of these unique sequences in MEC-5 function is not known. Genetic interactions indicate that mec-5 influences MEC-4/MEC-10 channel function (for example, mec-4 and mec-10 mutations can enhance the mec-5(ts) mutant phenotype [37]). Thus, MEC-5 could physically interact with the touch receptor channel, perhaps acting to provide gating tension via an extracellular link. The probable importance of collagen/degenerin interactions in channel function is underscored by studies of unc-105, another degenerin family member that is expressed in muscle [53]. Semi-dominant gain-of-function mutations in unc-105 cause severe muscle hypercontraction [57]. Specific mutations in let-2/sup-20, which encodes a type IV basement membrane collagen, suppress the unc-105(sd) phenotype [53, 57]. Taken together, these observations suggest that degenerin/collagen interactions, which still await experimental verification, may emerge as a common theme in the function of this channel class. 1.4.6.2

MEC-9 mec-9 encodes a protein that is secreted by the touch receptor neurons [23] and it is likely to be part of the mantle. Despite this, mec-9 mutations do not alter mantle ultrastructure in a detectable manner [13]. The mec-9 gene encodes two transcripts, the larger of which encodes an 834-amino-acid protein (MEC-9L) that is expressed exclusively by the touch neurons. The predicted MEC-9L protein appears specialized for protein interactions and contains several domains related to the Kunitz-type serine protease inhibitor domain, the Ca2+-binding EGF repeat, the non-Ca2+-binding EGF repeat, and a glutamic acid-rich domain. (Agrin, a protein that localizes acetylcholine receptors, has a similarly specialized domain structure, including multiple EGF and Kazal-type serine protease inhibitor repeats [63].) Mutations that disrupt MEC-9 function affect the two Ca2+-binding EGF repeats, the sixth EGF repeat, and the third Kunitz-type domain, thus highlighting these regions as important in touch transduction. mec-9 mutations are dominant enhancers of a mec-5(ts) allele, indicating that MEC-5 and MEC-9 proteins may interact in the mantle [23, 37]. 1.4.6.3

15

16

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

1.4.7

The MEC Channel Functions Specifically in Neuronal Responses to Gentle Touch Test of a Key Hypothesis Although genetic studies provided clear breakthroughs in the identification of molecules required for gentle body touch sensation, the technical challenges that precluded direct electrophysiological recording from tiny nematode touch receptor neurons stalled testing of the critical hypothesis that the MEC-4 channel functioned specifically in the process of touch transduction. This next step in analysis of touch mechanisms was critical because although the genetic data indicate that each MEC protein is needed for a behavioral response to gentle touch, they cannot distinguish whether a given protein might be a true component of the mechanotransducing complex rather than being needed generally for the integrity of neuronal function or for neuronal signaling downstream of the initial perception event. This concern was particularly strong for the channel subunits, which could function in setting the resting membrane potential or propagating electrical signals, and for MEC-2 stomatin, which might be associated with a general leakiness of membranes (analogous to what occurs in hereditary stomatocytosis in which RBCs lack stomatin [29]). Proteins required for general maintenance of touch neurons would play a secondary role rather than being true touch-transducing molecules. Elegant answers to the question of specificity of function have been recently obtained using a gene-based, fluorescent calcium-binding reporter named “cameleon” [76]. In cameleon, the fluorophores cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP) are linked by calmodulin and the calmodulin-binding peptide M13 [25, 56]. In low calcium, the two fluorophores are positioned far from each other, resulting in low fluorescence resonance energy transfer (FRET). In high calcium, calcium-bound calmodulin associates with M13 and the two fluorophores are brought close together, enabling FRET. Ratiometric CFP/YFP signals over time thus give a report of in vivo calcium fluxes. The power of the cameleon reagent is that it is genetically encoded and thus signals from live transgenic animals expressing the cameleon in touch neurons can be imaged to ask whether touch neurons mount calcium transients in response to touch, providing for the first time a physiological readout of touch receptor function. Experiments based on cameleon-reported Ca2+ transients revealed three key findings: (1) C. elegans touch receptor neurons respond to gentle body touch with transient calcium influx. (Although the MEC-4 channel appears to be a Na+ channel [34], the Ltype voltage-gated Ca2+ channel EGL-19 and the regulatory subunit UNC-36 are needed for normal calcium transients. An EGL-19-containing channel is thus likely gated by membrane depolarization associated with Na+ channel activation.) (2) mec-4, mec-2, and mec-6 null mutants lack any cameleon-detected response to gentle-touch stimuli (Fig. 1.4A). Thus, the MEC-4/MEC-2/MEC-6 channel is needed for calcium transients provoked by gentle-touch stimuli. (3) The touch neurons mount a distinct cameleondetected response to a harsh-touch stimulus, and mec-4 and mec-2 null mutants are normal for this response (mec-6 not tested) (Fig. 1.4A). Furthermore, electrophysiological studies on cultured neurons revealed that mec-4 and mec-2 mutant neurons 1.4.7.1

1.4 Gentle Body Touch

Body Touch Sensation Gentle Touch 20

20

5 10

20

30

40

10

fast

slow

5 0

0

10

20

30

40

Ratio change (%)

10 fast slow

00

15

15 Ratio change (%)

Ratio change( %)

A

20

buzz

buzz

15

0

0

mec-4(null)

WT

10

20

30

40

Time (sec)

Time (sec)

Time(sec)

buzz

10 fast slow 5

mec-2(null)

16

40

14 12

35

20 15

10

5 00

10

-5

20

30

Time(sec)

WT

40

30

8 6 4 2 00

10

20

30

40

Ratio change (%)

10

Ratio change (%)

Ratio change (%)

Harsh Touch 25

25 20 15 10 5 00

10

20

30

40

Time (sec)

Time (sec)

mec-4(null)

mec-2(null)

B MEC-5

MEC-10

MEC-6

EGL-36 MEC-9 MEC-4

EGL-19

- - + + +

MEC-2

MEC-12, MEC-7 Fig. 1.4 Molecular model for body touch sensation. (A) Studies using the genetically encoded calcium sensor cameleon revealed that touch neurons undergo transient changes in intracellular calcium concentration during touch stimulation [76]. As shown here in the upper left panel, large transients are generated upon stimulation by a constantly moving probe (buzz) and small ones are stimulated by slow or fast pokes. Middle and right upper panels show lack of calcium transients in touch neurons from mec-4 and mec-2 null backgrounds. In the lower panels calcium transients were apparent in all three strains when worms were stimulated by harsher touch, which was delivered at the time indicated by the arrow. (B) Cartoon of a touch-transducing complex in C. elegans touch-

receptor neurons (see text for discussion of the properties of the MEC proteins depicted) and the possible mechanism of calcium transient activation. In the absence of mechanical stimulation, the channel is closed and therefore the sensory neuron is at rest. Application of a mechanical force to the body of the worm results in distortion of a network of interacting molecules that opens the channel allowing Na+ influx. This in turn causes depolarization of the neuron, which activates the voltagegated L-type calcium channel EGL-19 (with likely accessory subunit UNC-36) that produces rapid and transient changes in intracellular calcium concentration, initiating the perceptory integration of the stimulus

17

18

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

have resting currents indistinguishable from wild-type neurons and also depolarize in response to high K+ like wild-type neurons. These last points clearly establish that touch neurons maintain normal physiological functions in the absence of mec-4 or mec-2 – in other words mec-4 and mec-2 are needed specifically for gentle-touch responses and are not generally needed for other touch neuron activities. Although cameleon data cannot establish that MEC-4 and MEC-2 are the actual touch transducers, these experiments position them very close to the actual transduction event and thus validate the attention given to them as components of a premier model of a touchtransducing channel. Additional Insights Revealed by Imaging In Vivo Ca2+ Changes in Responding Touch Neurons Use of the cameleon Ca2+ reporter also revealed previously unknown aspects of touch receptor biology. First, touch neurons appear to be rapidly adapting receptors that sense motion better than continuous pressure. A “press” stimulus in which a probe is moved a unit distance into the worm cuticle, held steady, and then pulled away elicits Ca2+ transients only when the stimulus is moved in or moved out—the neurons do not appear sensitive to continuous pressure. When the probe is moved rapidly back and forth against the body (a “buzz” stimulus), a large Ca2+ influx is observed (Fig. 1.4A). Thus, it appears that the touch neuron is tuned to mechanically sense motion, such as the brush of an eyelash hair, the original gentle-touch stimulus applied to the study of touch sensitive behavior. A second observation is that touch neurons respond when probed in their sensory fields and do so in a roughly “all or none” fashion. Gentle touch near a sensory process virtually always elicits a calcium transient, and touch far away from the process virtually never elicits a response. Interestingly, neurons occasionally respond when the animal is touched slightly posterior to their process, but when they do respond, the magnitude of the response is roughly the same as when the touch is delivered directly over the neuron process, suggesting that neurons do not respond in a markedly graded fashion according to the position of the stimulus. A third important point is that the six “gentle” body touch neurons can sense and respond to more than gentle touch. Harsh stimuli (such as a higher velocity jab with a rigid probe) elicit neuronal Ca2+ influxes that are distinct from those induced by gentletouch stimuli, and these responses to harsh touch do not depend upon the function of MEC-4 or MEC-2 (Fig. 1.4A). This strongly suggests that touch neurons can respond to distinct mechanical stimuli using distinct molecular machinery. What might the harsh-touch sensory channel be? Current data suggest that of the 20 or so degenerin channel subunits encoded by the C. elegans genome, only degenerins mec-4 and mec-10 are expressed in the touch neurons. It is possible that a monomeric MEC-10 channel responds to harsh touch, a hypothesis that can easily be tested when mec-10 null alleles are generated. Alternatively, a distinct channel type might be used. 1.4.7.2

1.4 Gentle Body Touch

1.4.8

Summary: A Molecular Model for Gentle-touch Sensation How Touch Is Sensed to Elicit a Specific Behavioral Response Genetic, molecular, and functional analysis of cloned touch cell structural genes suggests a working model of the touch receptor mechanotransducing complex (Fig. 1.4B; see [23, 37, 46, 52] for discussion). The central component of this model is the candidate mechanosensitive ion channel that includes multiple MEC-4 and MEC-10 subunits. These subunits assemble to form a channel pore that is lined by hydrophilic residues in MSDII of MEC-4. Accessory subunits MEC-2 and MEC-6 associate with the core channel and may influence channel pore properties, as they do in the oocyte expression system. The MEC-4 and MEC-10 channel subunits have a transmembrane topology in which the Cys-rich domains extend into the specialized extracellular matrix surrounding the touch cell (the mantle) and the amino- and carboxy- channel termini project into the cytoplasm. MEC-5 collagen and MEC-1 and MEC-9 interaction domain-rich proteins are components of the extracellular mantle that might serve to exert gating tension on the extracellular domains of the channel. Inside the touch neuron, a unique microtubule network may associate (likely via protein linkages) with the channel to help tether the intracellular channel domains. MEC-2 may also play a role in channel tethering or localization to a microdomain environment required for proper gating. Regulated gating is expected to depend on mechanical forces exerted on the channel by some or all these proteins. A touch stimulus could deform the microtubule network, or could perturb the mantle connections (or both), to deliver the gating stimulus. Na+ influx would activate the touch receptor to signal the appropriate locomotory response via a characterized neuronal circuit. Cameleon-based reporting of physiological responses to touch stimuli have revealed that one of the signals that is triggered by activation of the putative mechanosensory channel is elevation of intracellular calcium via a specific L-type voltage-gated calcium channel, EGL-19 (Fig. 1.4B). Upon conversion of a mechanical stimulus into an electrochemical response, the touch receptor neurons activate a simple reflex circuit [14]. The touch cells activate the interneurons (AVD for backward and PVC for forward locomotion) that in turn activate motor neurons. While the touch cells form gap junctions with agonist interneurons, they form apparent chemical synapses with the antagonist interneurons. This reciprocal pattern of connectivities (for example, AVD activated and PVC inhibited for backward movement) enables locomotion in the appropriate direction to be stimulated at the same time that locomotion in the inappropriate direction is inhibited. 1.4.8.1

Notes on the Working Model The working model for touch transduction accommodates all cloned MEC proteins to postulate a complex that may gate like the mechanosensory channels that respond to sound in the hair cells of the vertebrate inner ear (reviewed in [33]). Aspects of the model that remain to be tested include: 1.4.8.2

19

20

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans *

*

*

Whether the channel is itself directly mechanically gated. Such demonstration will require direct assay of the channel in an in vivo context (cell culture or direct recording from neurons in the nematode). The channel cannot be mechanically gated as reconstituted in Xenopus oocytes (this was to be expected, as putative accessory proteins were not co-expressed and would be unlikely assemble properly even if they were [34]). Whether specific MEC proteins associate with the channel. Once interactions are identified, definition of residues involved in interactions will fill in molecular details of the model. Whether all MEC proteins are directly involved in the channel complex rather than carrying out other functions required for touch receptor function or signal propagation. In addition, deciphering single channel properties and determining precisely how subunits influence channel gating will extend our understanding of molecular mechanisms of touch.

1.5

The C. elegans Degenerin Family: A Global Role of Degenerin Channels in Mechanotransduction?

The C. elegans genome encodes a total of 28 degenerins. Our preliminary surveys of expression patterns suggest that degenerins are expressed in specific but overlapping cell groups. Initial work suggests that at least some of these ion channel subunits (unc105, unc-8, and del-1) may function in mechanical signaling, suggesting a potential common functional theme for members of this channel class in C. elegans.

1.5.1

unc-105

Semi-dominant unc-105 alleles induce hypercontraction of body wall muscles [57]. The unc-105(sd) mutations encode amino acid substitutions in the extracellular domain [54] that render the mutant UNC-105 subunits hyperactive [31], suggesting that hypercontraction is the result of excess ion influx. Interestingly, specific alleles of sup-20/let2, an essential type IV basement membrane collagen [68], suppress unc-105(sd) hypercontraction [53, 57], indicating that sup-20/let-2 collagen may function in muscle cells similarly to mec-5 collagen in touch neurons. The working model is that unc-105 may function as a stretch-responsive channel in body wall muscle that is gated via attachment to collagen in the extracellular matrix [53]. Interestingly, UNC-105 null mutations have no apparent phenotype [57], suggesting that another degenerin coexpressed with unc-105 could redundantly supply the same function in muscle or that the unc-105 null phenotype may be a subtle behavioral defect not readily detectable.

1.5 The C. elegans Degenerin Family: A Global Role of Degenerin Channels in Mechanotransduction?

1.5.2

unc-8 and del-1

Unusual semi-dominant gain-of-function unc-8 alleles induce transient neuronal swelling [67] and severe uncoordination [8, 57], with a particularly pronounced defect in backward locomotion. unc-8 encodes a degenerin expressed in several motor neuron classes, in some interneurons, and in ASH polymodal sensory neurons [80]. Another degenerin family member, del-1(for degenerin-like), is co-expressed with unc-8 in a subset of neurons (the VA and VB motor neurons and the FLP harsh touch neurons) and is likely to assemble into a channel complex with UNC-8 in these cells [80]. What is the role of the UNC-8 degenerin channel in locomotion? Important clues came from the unc-8 null mutant phenotype and from the neuronal anatomy of motor neurons expressing unc-8. unc-8 null mutants have locomotory defects: although they move in a sinusoidal trajectory (the normal locomotion mode for C. elegans), the tracks they carve in the bacterial lawn on which they live and travel through are markedly reduced in amplitude. Thus, unc-8 null mutants do not bend as deeply as wild-type worms (Fig. 1.5A). Interestingly, some motor neurons that co-express unc-8 and del-1 (namely, the VA and VB motor neurons) have processes that have been hypothesized to be stretch-sensitive because they lack synapses to muscle or other nerves over most of their processes (originally by R. L. Russell and L. Byerly and discussed in [85] and Fig. 1.5B). Given the homology of UNC-8 and DEL-1 to candidate mechanically gated channels, we have suggested that these subunits co-assemble into a stretch-sensitive channel that might be localized to the sensory regions of the motor neuron process. In this model UNC-8/DEL-1 channels respond to displacement of the bending neuronal process during locomotion, and when they do so, they enhance signaling at the distant neuromuscular junction. The increased strength and duration of the muscle contraction they contribute is needed for a full-sized body bend. In the absence of the stretch activation signal delivered by the UNC-8/DEL-1 channel, the body bend still occurs, but with reduced amplitude (Fig. 1.5B). Although this proprioception model is based on neuroanatomy, behavior of mutant UNC-8 subunits, and proposed functional homology with MEC-4/MEC-10 channel complexes, details of the working hypothesis still await experimental verification.

1.5.2.1 A Stomatin Partner for the UNC-8 Channel Suggests a Common Composition of Degenerin Channels Genetic screens for mutations that affect anesthetic sensitivity in C. elegans identified several genes that alter sensitivity to specific anesthetics. Among these was unc-1, a gene that encodes a stomatin-like protein [61, 62]. unc-1 mutants share some uncoordinated traits with unc-8(sd) mutants, which prompted testing of unc-8 mutants for anesthetic responses and for genetic interactions with unc-1. Importantly, allele-specific interactions between unc-1/stomatin and unc-8/degenerin were identified in this survey, strongly suggesting that the channel and the stomatin, which are co-expressed in many cells, physically interact [62]. Biochemical evidence for physical association has not yet been published.

21

22

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

1.5 The C. elegans Degenerin Family: A Global Role of Degenerin Channels in Mechanotransduction?

3 Fig. 1.5 Modulation of locomotion by the stretch-responsive channel UNC-8 in motor neurons (proprioception model). (A) Examples of tracks left on an E. coli lawn by WT and unc-8(lf ) nematodes. While WT animals inscribe wide sinusoidal marks, unc-8(lf ) animals produce much narrower tracks due to defects in the VB motor neurons’ function caused by the absence of UNC-8 channels. (B) Features and function of VB motor neurons in body stretch sensation. Two VB motor neurons in the ventral nerve cord are depicted here. A typical VB motor neuron makes synapses to muscle near to the cell body and possesses a long nerve terminal that is synapse-free and proposed to have stretch-sensory functions. The stretch-sensitive channels are postulated to be situated in these “undifferentiated” processes. The anterior signaling by VB is proposed to be potentiated by opening of the UNC-8 channels that monitor body stretch due to local muscle contraction. Following potentiation, this motor neuron will signal to the anterior muscles to become fully contracted. At the same time another motor neuron in the middle of the body does not experience stretch and therefore remains idle. Sequential activation of motor neurons distributed along the ventral nerve cord may amplify and propagate the sinusoidal body wave. (C) Cartoon of a stretch-sensitive channel complex in C. elegans VB motor neurons (see text for hypothesized channel subunits’ function). The channel complex, by analogy with the touch-sensitive MEC-4 channel expressed in touch neurons, is hypothesized to sense stretching forces and includes the degenerin subunits UNC-8 and DEL-1 and stomatin UNC-1. Stomatin UNC-24 is needed for UNC-1 to be properly localized and may be part of the channel complex. In the absence of stretch stimulation, the channel is closed and therefore the motor neuron is at rest. When the worm bends its body, the motor neuron experiences stretch forces along its undifferentiated nerve terminal where UNC-8 channels are located. Stretch forces may distort the network of interacting molecules that opens the UNC-8 channel allowing Na+ influx. Adapted from [80]

UNC-1 protein expression and distribution is dependent upon a second stomatin domain protein, UNC-24 [66]. The unc-24 gene encodes a protein that consists of a stomatin-like domain and a lipid-transfer domain [4]. Lipid-transfer proteins are mostly soluble proteins that transfer a wide variety of phospholipids and sterols from donor membranes, mainly vesicles, to acceptor membranes. Because UNC24 is membrane-anchored at the stomatin domain, its lipid-transfer domain can act only locally, possibly regulating the composition of small areas of the lipid bilayer. UNC-1 protein is normally distributed in a punctate pattern in neuronal processes. In unc-24 null mutants, UNC-1-related puncta are greatly diminished in number in unc-24 null animals. Interestingly, in almost all neurons there are only two puncta located on each side of the nucleus. These data indicate a role for UNC-24 in UNC-1 localization to the process and suggest that UNC-24 may heteromultimerize with UNC-1 and possibly other proteins in the UNC-8 channel complex (Fig. 1.5C). It is noteworthy that UNC-24 is expressed in touch neurons and in FLPs and PVDs that mediate harsh-touch responses, such as the ones delivered by a wire prod. Thus, UNC24 is likely to regulate multiple degenerin channels and contribute to multiple mechanosensory functions.

Trp Channels May Also Contribute to Mechanosensory Functions in C. elegans In the laboratory, C. elegans moves through a bacterial lawn on a petri dish with a readily observable sinusoidal motion. When the worm collides head-on with an object such as an eyelash hair, it initiates backward motion known as nose touch avoidance. Three classes of mechanosensory neurons act in parallel to mediate the avoidance 1.5.2.2

23

24

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

response, including ASH, FLP, and OLQ [47]. ASH neurons mediate two other wellcharacterized behaviors: osmotic avoidance (nematodes respond to high osmotic strength solutions by avoiding them) and adverse response to noxious stimuli. TRP channels (see Tab. 8.3) have been implicated in all ASH-mediated behavioral responses. Based on sequence conservation, the TRP superfamily can be divided into three subfamilies named TRPC, TRPV, and TRPM. The C. elegans TRPV subfamily includes OSM-9 and OCR-1, -2, -3, and -4. osm-9 and ocr-2 mutants are defective in all ASH-mediated behaviors. In addition, osm-9 also acts in other neuronal types, where its functions may be different. For example, in AWA neurons OSM-9 is proposed to form a G protein-regulated transduction channel required for odorant sensitivity, and in AWC it has a role in olfactory adaptation. Thus, studies on TRPV channels implicate this channel family in mechanisms of mechanotransduction in C. elegans but also document their involvement in other sensory behaviors, suggesting that they may be gated by stimuli besides mechanical forces [81]. One possibility is that TRPV channels may set the physiological stage so that the other channels (for example, DEG/ ENaCs) can function. Studies in Drosophila [50, 84] and zebrafish [69] have strongly implicated TRP channels in mechanosensory transduction. In a Drosophila screen for touch-insensitive mutants, mutations affecting a gene named nompC (no mechanoreceptor potential) were isolated. nompC encodes a TRP channel expressed in Drosophila mechanosensory organs, including bristles, and mutants lack or have abnormal mechanosensory currents in bristle neurons during bristle deflection. Zebrafish larvae treated with morpholino antisense oligonucleotides for the homologue of nompC are deaf, and sensory hair cells no longer respond to sound stimulation [69] (see Chapter 2).

1.5.2.3 Fly and Mouse Neuronal DEG/ENaCs Influence Mechanotransduction, Supporting Conserved Roles for This Family of Proteins C. elegans DEG/ENaCs are clearly directly implicated in mechanotransduction. A key question that arises, then, is whether this role is conserved across species such that mammalian DEG/ENaCs may contribute to mechanoperception. Recent studies in Drosophila and mouse support such a conserved role.

Fly The Drosophila DEG/ENaC Pickpocket is expressed early in dendritic varicosities of the da/md abdominal peripheral neurons that function in the adult as mechanoreceptors [1]. The PPK1 channel plays an essential role in controlling rhythmic locomotion, suggesting a potential role in mechanical signaling [2]. Mouse Based on sequence similarity, mammalian DEG/ENaCs can be grouped into two subfamilies: the ENaC subfamily that includes a, b, c, and d subunits involved in Na+ reabsorption in specialized kidney and lung epithelia [9] and the ASIC subfamily. The mammalian acid-gated ASIC subfamily includes five members, with two known

1.6 Concluding Remarks

splice variants for ASIC2, that are expressed in both the central and peripheral nervous system (see Chapter 3). ASIC2a, one of the ASIC2 splice variants, is expressed in medium and large diameter mechanosensory neurons of the dorsal root ganglia (DRG) and is localized to nerve termini that are known to function as cutaneous mechanosensors [32, 59]. Functional analysis of these mechanoreceptors from an ASIC2 knockout mouse revealed that two types of low-threshold (i.e., they sense light touch) mechanosensitive fibers (the rapidly adapting [RA] and, to a lesser extent, the slowly adapting [SA]) are not able to respond to stronger displacement force stimuli by increasing the frequency of action potentials, as fibers from WT mice do. This finding has biological relevance, since the dynamic sensitivity of RA and SA mechanosensors is thought to play an important role in how we perceive gentle touch. ASIC3 has also been implicated in mechanosensation. ASIC3 is also localized to several types of mechanosensory nerve terminals, and its expression pattern partly overlaps with ASIC2a [60]. In particular, both channels are expressed in RA mechanoreceptors. Studies of nerve-skin preparations from the ASIC3 knockout mouse revealed that both the threshold sensitivity and the response frequency of AM fiber mechano-nociceptors (which respond to high-threshold stimuli-like pinching) are reduced. Interestingly, RA mechanoreceptors from ASIC3 knockout mouse respond to stronger displacement forces, but they do so with doubled firing frequency as compared to WT. To conclude, ASIC3 is involved in responding to both gentle-touch and painful stimuli and it seems to do so in distinct ways. These findings support a conserved role for DEG/ENaC family members in touch sensation, but also highlight the higher complexity of mammalian touch perception mechanisms. In fact, ASIC2a and ASIC3 knockout effects are subtle. In both knockout mice no mechanoreceptor responses are completely eliminated, and for some mechanoreceptors that normally express the ASIC2 channel, responses are unaffected. Explanations for such differences may include redundancy (other members of the family may be expressed in the same mechanoreceptors and may functionally overlap with ASIC2a and ASIC3), modulatory function of ASIC2a and ASIC3 in a heteromultimeric channel complex (similar to MEC-10), and a more crucial role of other channel families (TRP for example) in touch sensation in mammals.

1.6

Concluding Remarks

DEG/ENaC and TRP ion channels have both been implicated in mechanosensory transduction mechanisms in invertebrates and vertebrates and both have been suggested to be the primary transducers of the mechanical stimuli. Looming questions in the field of molecular mechanotransduction include the following. Are TRPs and DEG/ENaCs functionally redundant, or do they “sense” distinct stimuli? Does one channel type merely set the physiological stage so that the other can function? Do the DEG/ENaCs and TRPs physically interact or cooperate to sense force? Addressing these questions will be clearly feasible once DEG/ENaC and TRP channel function, regulation, and gating are studied in detail in both invertebrates and vertebrates.

25

26

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans

Studying the members of the two channel families that are co-expressed in the same mechanosensory cells should clarify relative roles in mechanotransduction.

Acknowledgments

We thank our colleagues cited herein, especially D. Hall for providing touch receptor electron micrographs. Some of the work reviewed here was supported by the National Institutes of Health NINDS (1R01-NS37955).

References 1

2

3

4

5 6

7

8

Adams, C. M., Anderson, M. G., Motto, D. G., Price, M. P., Johnson, W. A., and Welsh, M. J. Ripped pocket and pickpocket, novel Drosophila DEG/ENaC subunits expressed in early development and in mechanosensory neurons. J Cell Biol 1998, 140, 143–152. Ainsley, J. A., J. M. Pettus, D. Bosenko, C. E. Gerstein, N. Zinkevich, M. G. Anderson, C. M. Adams, M. J. Welsh, and W. A. Johnson, Enhanced locomotion caused by loss of the drosophila DEG/ENaC protein Pickpocket1. Curr Biol, 2003, 13(17), 1557–1563. Ausiello, D. A., Stow, J. L., Cantiello, H. F., de Almeida, J. B., and Benos, D. J. Purified epithelial Na+ channel complex contains the pertussis toxin-sensitive Gai-e protein. J Biol Chem 1992, 267, 4759–9472. Barnes, T. M., Jin, Y., Horvitz, H. R., Ruvkun, G., and Hekimi, S. The Caenorhabditis elegans behavioral gene unc-24 encodes a novel bipartite protein similar to both erythrocyte band 7.2 (stomatin) and nonspecific lipid transfer protein. J Neurochem 1996, 67, 46–57. Beachy, J. M. Premature infant massage in the NICU. Neonatal Netw 2003, 22, 39–45. Benos, D. J., Saccomani, G., and SaribanSohraby, S. The epithelial sodium channel: subunit number and location of the amiloride binding site. J Biol Chem 1987, 262, 10613–10618. Berdiev, B. K., K. H. Karlson, B. Jovov, P. J. Ripoll, R. Morris, D. Loffing-Cueni, P. Halpin, B. A. Stanton, T. R. Kleyman, and Ismailov, I. I. Subunit stoichiometry of a core conduction element in a cloned epithelial amiloride-sensitive Na+ channel. Biophys J, 1998, 75(5), 2292–2301. Brenner, S. The genetics of Caenorhabditis elegans. Genetics 1974, 77, 71–94.

9

10

11

12 13

14

15

16

17

Canessa, C. M., Horsiberger, J. D., and Rossier, B. C. Functional cloning of the epithelial sodium channel relation with genes involved in neurodegeneration. Nature 1993, 349, 588–593. Canessa, C. M., Schild, L., Buell, G., Thorens, B., Gautschi, I., Horisberger, J. D., and Rossier, B. C. Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature 1994, 367, 412–413. Chalfie, M., and Au, M. Genetic control of differentiation of the Caenorhabditis elegans touch receptor neurons. Science 1989, 243, 1027–1033. Chalfie, M., Driscoll, M., and Huang, M. Degenerin similarities. Nature 1993, 361, 504. Chalfie, M., and Sulston, J. Developmental genetics of the mechanosensory neurons of Caenorhabditis elegans. Dev. Biol. 1981, 82, 358–370. Chalfie, M., Sulston, J. E., White, J. G., Southgate, E., Thomson, J. N., and Brenner, S. The neural circuit for touch sensitivity in Caenorhabditis elegans. J Neurosci 1985, 5, 956–964. Chalfie, M., and Thomson, J. N. Structural and functional diversity in the neuronal microtubules of Caenorhabditis elegans. J Cell Biol 1982, 93, 15–23. Chalfie, M., and Thomson, N. J. Organization of neuronal microtubules in the nematode Caenorhabditis elegans. J Cell Biol 1979, 82, 278–289. Chalfie, M., and Wolinsky, E. The identification and suppression of inherited neurodegeneration in Caenorhabditis elegans. Nature 1990, 345, 410–416.

1.6 Concluding Remarks 18

19

20

21

22

23

24

25

26

27

28

29

Chelur, D. S., G. G. Ernstrom, M.B. Goodman, C. A. Yao, L. Chen, R. O’Hagan and M. Chalfie, The mechanosensory protein MEC-6 is a subunit of the C. elegans touch-cell degenerin channel. Nature, 2002 420(6916), 669–673. Christensen, M., Estevez, A., Yin, X., Fox, R., Morrison, R., McDonnell, M., Gleason, C., Miller, D. M., and Strange, K. A primary culture system for functional analysis of C. elegans neurons and muscle cells. Neuron 2002, 33, 503–514. The C. elegens sequence consortium, Genome sequence of the nematode C. elegans: a platform for investigating biology. The C. elegans Sequencing Consortium. Science, 1998, 282(5396), 2012–2018. Dijkink, L., A. Hartog, C. H. van Os, and R. J. Bindels, The epithelial sodium channel (ENaC) is intracellularly located as a tetramer. Pflugers Arch, 2002, 444(4), 549–555. Driscoll, M., and Chalfie, M. The mec-4 gene is a member of a family of Caenorhabditis elegans genes that can mutate to induce neuronal degeneration. Nature 1991, 349, 588–593. Du, H., Gu, G., Williams, C., and Chalfie, M. Extracellular proteins needed for C. elegans mechanosensation. Neuron 1996, 16, 183–194. Dumont, R. A., and Gillespie, P. G. Ion channels: hearing aid. Nature 2003, 424, 28–29. Fan, G. Y., H. Fujisaki, A. Miyawaki, R. K. Tsay, R. Y. Tsien, and M. H. Ellisman, Videorate scanning two-photon excitation fluorescence microscopy and ratio imaging with cameleons. Biophys J, 1999, 76(5), 2412–2420 Fire, A. and Mello, C. C. DNA Transformation. Methods in Cell Biology. Caenorhabditis elegans: Modern Biological Analysis of an Organism. H.F. Epstein and D.C. Shakes (eds), Academic Press, Inc., San Diego. 1995, 48, 451–482. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E., and Mello, C. C. Potent and specific genetic interference by doublestranded RNA in Caenorhabditis elegans. Nature 1998, 391, 806–811. Firsov, D., I. Gautschi, A. M. Merillat, B. C. Rossier, and L. Schild, The heterotetrameric architecture of the epithelial sodium channel (ENaC). Embo J, 1998, 17(2), 344–352. Fricke, B., Lints, R., Stewart, G., Drummond, H., Dodt, G., Driscoll, M., and von During, M. Epithelial Na+ channels and stomatin are expressed in rat trigeminal mechanosensory neurons. Cell Tissue Res 2000, 299, 327–334.

30

31

32

33

34

35

36

37

38

39

40

41

Fukushige, T., Siddiqui, Z. K., Chou, M., Culotti, J. G., Gogonea, C. B., Siddiqui, S. S., and Hamelin, M. MEC-12, an a-tubulin required for touch sensitivity in C. elegans. J Cell Sci 1999, 112, 395–403. Garcia-Anoveros, J., Garcia, J. A., Liu, J. D., and Corey, D. P. The nematode degenerin UNC-105 forms ion channels that are activated by degeneration- or hypercontractioncausing mutations. Neuron 1998, 20, 1231–1241. Garcia-Anoveros, J., Samad, T. A., ZuvelaJelaska, L., Woolf, C. J., and Corey, D. P. Transport and localization of the DEG/ENaC ion channel BNaC1alpha to peripheral mechanosensory terminals of dorsal root ganglia neurons. J Neurosci 2001, 21, 2678–2686. Gillespie, P. G., and Walker, R. G. Molecular basis of mechanosensory transduction. Nature 2001, 413, 194–202. Goodman, M. B., Ernstrom, G. G., Chelur, D. S., O’Hagan, R., Yao, C. A., and Chalfie, M. MEC-2 regulates C. elegans DEG/ENaC channels needed for mechanosensation. Nature 2002, 415, 1039–1042. Goodman, M. B., Hall, D. H., Avery, L., and Lockery, S. R. Active currents regulate sensitivity and dynamic range in C. elegans neurons. Neuron 1998, 20, 763–772. Grunder, S., Jaeger, N. F., Gautschi, I., Schild, L., and Rossier, B. C. Identification of a highly conserved sequence at the N-terminus of the epithelial Na+ channel alpha subunit involved in gating. Pflugers Arch 1999, 438, 709–715. Gu, G., Caldwell, G. A., and Chalfie, M. Genetic interactions affecting touch sensitivity in Caenorhabditis elegans. Proc Natl Acad Sci USA 1996, 93, 6577–6582. Hamelin, M., Scott, I. M., Way, J. C., and Culotti, J. G. The mec-7 beta-tubulin gene of Caenorhabditis elegans is expressed primarily in the touch receptor neurons. Embo J 1992, 11, 2885–2893. Hamill, O. P., and Martinac, B. Molecular basis of mechanotransduction in living cells. Physiol Rev 2001, 81, 685–740. Harbinder, S., Tavernarakis, N., Herndon, L., Kinnel, M., Xu, S., Fire, A., and Driscoll, M. Genetically targeted cell disruption in Caenorhabditis elegans. Proc Natl Acad Sci USA 1997, 94, 13128–13133. Harteneck, C., Plant, T. D., and Schultz, G. From worm to man: three subfamilies of TRP channels. Trends Neurosci 2000, 23, 159–166.

27

28

1 The Molecular Basis of Touch Sensation as Modeled in Caenorhabditis elegans 42

43

44

45

46

47

48

49

50

51

52

53

Hong, K., and Driscoll, M. A transmembrane domain of the putative channel subunit MEC4 influences mechanotransduction and neurodegeneration in C. elegans. Nature 1994, 367, 470–473. Hong, K., Mano, I., and Driscoll, M. In vivo structure-function analyses of Caenorhabditis elegans MEC-4, a candidate mechanosensory ion channel subunit. J Neurosci 2000, 20, 2575–2588. Horvitz, H. R., and Sultson, J. E. Isolation and genetic characterization of cell-lineage mutants of the nematode Chaenorhabditi elegans. Genetics 1980, 96, 435–454. Huang, M., and Chalfie, M. Gene interactions affecting mechanosensory transduction in Caenorhabditis elegans. Nature 1994, 367, 467–470. Huang, M., Gu, G., Ferguson, E. L., and Chalfie, M. A stomatin-like protein necessary for mechanosensation in C. elegans. Nature 1995, 378, 292–295. Kaplan, J. M. and Horvitz, H. R. A dual mechanosensory and chemosensory neuron in Caenorhabditis elegans. Proc Natl Acad Sci USA 1993, 90, 2227–2231. Kennedy, H. J., Evans, M. G., Crawford, A. C., and Fettiplace, R. Fast adaptation of mechanoelectrical transducer channels in mammalian cochlear hair cells. Nature Neurosci 2003, 6, 832–836. Kerr, R., Lev-Ram, V., Baird, G., Vincent, P., Tsien, R. Y., and Schafer, W. R. Optical imaging of calcium transients in neurons and pharyngeal muscle of C. elegans. Neuron 2000, 26, 583–594. Kim, J., Chung, Y. D., Park, D. Y., Choi, S., Shin, D. W., Soh, H., Lee, H. W., Son, W., Yim, J., Park, C. S., Kernan, M. J., and Kim, C. A TRPV family ion channel required for hearing in Drosophila. Nature 2003, 424, 81–84. Kosari, F., S. Sheng, J. Li, D. O. Mak, J. K. Foskett, and T. R. Kleyman, Subunit stoichiometry of the epithelial sodium channel. J Biol Chem, 1998. 273(22), 13469–13474. Lai, C. C., Hong, K., Kinnell, M., Chalfie, M., and Driscoll, M. Sequence and transmembrane topology of MEC-4, an ion channel subunit required for mechanotransduction in Caenorhabditis elegans. J Cell Biol 1996, 133, 1071–1081. Liu, J., Schrank, B., and Waterston, R. Interaction between a putative mechanosensory membrane channel and a collagen. Science 1996, 273, 361–364.

54

55

56

57

58

59

60

61

62

63

64

Liu, K. S., and Sternberg, P. W. Sensory regulation of male mating behavior in Caenorhabditis elegans. Neuron 1995, 14, 79–89. Mitani, S., Du, H., Hall, D. H., Driscoll, M., and Chalfie, M. Combinatorial control of touch receptor neuron expression in Caenorhabditis elegans. Development 1993, 119, 773–783. Miyawaki, A., O. Griesbeck, R. Heim, and R. Y. Tsien, Dynamic and quantitative Ca2+ measurements using improved cameleons. Proc Natl Acad Sci USA, 1999, 96(5), 2135–2140. Park, E., and Horvitz, H. R. Mutations with Dominant Effects on the Behavior and Morphology of the Nematode C. elegans. Genetics 1986, 113, 821–852. Patel, A. J., Lazdunski, M., and Honore, E. Lipid and mechano-gated 2P domain K(+) channels. Curr Opin Cell Biol 2001, 13, 422–428. Price, M. P., Lewin, G. R., McIlwrath, S. L., Cheng, C., Xie, J., Heppenstall, P. A., Stucky, C. L., Mannsfeldt, A. G., Brennan, T. J., Drummond, H. A., Qiao, J., Benson, C. J., Tarr, D. E., Hrstka, R. F., Yang, B., Williamson, R. A., and Welsh, M. J. The mammalian sodium channel BNC1 is required for normal touch sensation. Nature 2000, 407, 1007–1011. Price, M. P., McIlwrath, S. L., Xie, J., Cheng, C., Qiao, J., Tarr, D. E., Sluka, K. A., Brennan, T. J., Lewin, G. R., and Welsh, M. J. The DRASIC cation channel contributes to the detection of cutaneous touch and acid stimuli in mice. Neuron 2001, 32, 1071–1083. Rajaram, S., Sedensky, M. M., and Morgan, P. G. Unc-1: a stomatin homologue controls sensitivity to volatile anesthetics in Caenorhabditis elegans. Proc Natl Acad Sci U S A 1998, 95, 8761–8766. Rajaram, S., Spangler, T. L., Sedensky, M. M., and Morgan, P. G. A stomatin and a degenerin interact to control anesthetic sensitivity in Caenorhabditis elegans. Genetics 1999, 153, 1673–1682. Rupp, F., D. G. Payan, C. Magill-Solc, D. M. Cowan, and R. H. Scheller, Structure and expression of a rat agrin. Neuron, 1991, 6(5), 811–823. Savage, C., Hamelin, M., Culloti, J. G., Coulson, A., Albertson, D., and Chalfie, M. mec-7 is a beta tubulin gene required for the production of 15 protofilament microtubules in Caenorhabditis elegans. Genes Dev 1989, 3, 870–881.

1.6 Concluding Remarks 65

66

67

68

69

70

71

72 73

74

75

Schild, L., Canessa, C. M., Shimkets, R. A., Gautschi, I., Lifton, R. P., and Rossier, B. C. A mutation in the epithelial sodium channel causing Liddle disease increases channel activity in the Xenopus laevis oocyte expression system. Proc Natl Acad Sci USA 1995, 92, 5699–5703. Sedensky, M. M., Siefker, J. M., and Morgan, P. G. Model organisms: new insights into ion channel and transporter function. Stomatin homologues interact in Caenorhabditis elegans. Am J Physiol Cell Physiol 2001, 280, C1340–C1348. Shreffler, W., Magardino, T., Shekdar, K., and Wolinsky, E. The unc-8 and sup-40 genes regulate ion channel function in Caenorhabditis elegans motorneurons. Genetics 1995, 139, 1261–1272. Sibley, M.H., J.J. Johnson, C.C. Mello, and Kramer, J. M. Genetic identification, sequence, and alternative splicing of the Caenorhabditis elegans alpha 2(IV) collagen gene. J Cell Biol, 1993, 123(1), 255–264. Sidi, S., Friedrich, R. W., and Nicolson, T. NompC TRP channel required for vertebrate sensory hair cell mechanotransduction. Science 2003, 301, 96–99. Snyder, P. M., Cheng, C., Prince, L. S., Rogers, J. C., and Welsh, M. J. Electrophysiological and biochemical evidence that DEG/ ENaC cation channels are composed of nine subunits. J Biol Chem 1998, 273, 681–684. Snyers, L., Umlauf, E., and Prohaska, R. Oligomeric nature of the integral membrane protein stomatin. J Biol Chem 1998, 273, 17221–17226. Stewart, G. W. Stomatin. Int J Biochem Cell Biol 1997, 29, 271–274. Stewart, G. W., Argent, A. C., and Dash, B. C. Stomatin: a putative cation transport regulator in the red cell membrane. Biochim Biophys Acta 1993, 1225, 15–25. Stewart, G. W., and Turner, E. J. The hereditary stomatocytoses and allied disorders: congenital disorders of erythrocyte membrane permeability to Na and K. Baillieres Best Pract Res Clin Haematol 1999, 12, 707–727. Sulston, J. E., Schierenberg, E., White, J. G., and Thomson, J. N. The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev Biol 1983, 100, 64–119.

76

77

78

79

80

81

82

83

84

85

Suzuki, H., Kerr, R., Bianchi, L., FrokjarJensen, C., Slone, D., Xue, J., Gerstbrein, B., Driscoll, M., and Schafer, W. R. In vivo imaging of C. elegans mechanosensory neurons demonstrates a specific role for the mec4 channel in the process of gentle touch sensation. Neuron 2003, 39, 1005–1017. Tabara, H., Grishok, A., and Mello, C. C. RNAi in C. elegans: soaking in the genome sequence. Science 1998, 282, 430–431. Tavernarakis, N., and Driscoll, M. Molecular modeling of mechanotransduction in the nematode Caenorhabditis elegans. Annu Rev Physiol 1997, 59, 659–689. Tavernarakis, N., and Driscoll, M. Caenorhabditis elegans degenerins and vertebrate ENaC ion channels contain an extracellular domain related to venom neurotoxins. J Neurogenet 2000, 13, 257–264. Tavernarakis, N., Shreffler, W., Wang, S., and Driscoll, M. unc-8, a DEG/ENaC family member, encodes a subunit of a candidate mechanically gated channel that modulates C. elegans locomotion. Neuron 1997, 18, 107–119. Tobin, D. M., Madsen, D. M., Kahn-Kirby, A., Peckol, E. L., Moulder, G., Barstead, R., and Bargmann, C. I. Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans. Neuron 2002, 35, 307–318. Vogel, B. E., and Hedgecock, E. M. Hemicentin, a conserved extracellular member of the immunoglobulin superfamily, organizes epithelial and other cell attachments into oriented line-shaped junctions. Development 2001, 128, 883–894. Waldmann, R., Champigny, G., Voilley, N., Lauritzen, I., and Lazdunski, M. The mammalian degenerin MDEG, an amiloride-sensitive cation channel activated by mutations causing neurodegeneration in Caenorhabditis elegans. J. Biol. Chem. 1996, 271, 10433–10436. Walker, R. G., Willingham, A. T., and Zuker, C. S. A Drosophila mechanosensory transduction channel. Science 2000, 287, 2229–2234. White, J. G., Southgate, E., Thomson, J. N., and Brenner, S. The structure of the nervous system of Caenorhabditis elegans. Philos Trans R Soc Lond 1986, 314, 1–340.

29

31

2

Transduction Channels in Hair Cells Robert Fettiplace

2.1

Introduction

Hair cells are the sensory neurons of the vertebrate inner ear, comprising the organs of hearing (the cochlea) and balance (three semicircular canals, a saccule, and the utricle). They also serve as receptors for the lateral line neuromasts on the skin of fish and primitive amphibia. A hair cell detects mechanical stimuli by deflection of its hair bundle, an array of between 20 and 300 elongated actin-filled stereocilia projecting from the top of the cell [1]. In the cochlea, hair bundles are mechanically vibrated by motion of an overlying gelatinous flap known as the tectorial membrane to which they are attached. The stereocilia are arranged in several ranks of increasing height and are interconnected by extracellular filaments [2, 3] to ensure that all move in synchrony during bundle deflection. One class of filament, the tip link, extending from the apex of each stereocilium to the side of its taller neighbor [4, 5], is essential for transduction [6]. Thus, deflection of the bundle towards its taller edge, rotating the rigid stereocilia about their basal insertion into the cell apex [7], is proposed to exert tension on the tip links [4, 8]. This transmits force directly or indirectly to mechanoelectrical transduction (MET) channels located near the tips of the stereocilia (Fig. 2.1 [9]). When opened by hair bundle displacement, the transduction channels allow influx of cations to generate a depolarizing receptor potential at a cell resting potential of about –50 to –70 mV. Mainly because of the paucity of hair cells in the inner ear epithelia, the MET channel has been difficult to clone and a substantive clue to its molecular identity has only recently appeared [10]. It has been provisionally classified as a member of a new branch of the TRP (transient receptor potential) family, TRPN1. However it has not yet been expressed in a heterologous system and, unlike the mechanosensitive bacterial MscL channel [11], there is little information about its behavior under controlled conditions. As a consequence, the biophysical properties of the channel have been wholly derived from measuring transduction currents in intact hair cells during Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

32

2 Transduction Channels in Hair Cells

Fig. 2.1 Tip links and hair cell transduction. (A) Transmission electron micrograph showing the tip link from the apex of the shorter stereocilium of a guinea pig outer hair cell bundle to the electron-dense plaque on the side of the neighboring taller stereocilium. Note the “tenting” of the membrane, suggesting that the tip link (arrowed) tugs on the stereociliary membrane, and also the electron-dense contact region between the two stereocilia, which has been suggested as an alternative location for the MET channel (courtesy of Y. Katori, D.N. Furness, and C.M. Hackney). (B) Schematic of top of the hair cell, illustrating how deflection of the hair bundle causes the stereocilia to bend at their rootlets where a subset of actin filaments insert into the cuticular plate. Stereociliary displacement exerts force via the tip link to open the MET channel and, in the intact hair cell epithelium, allows influx of K+ and Ca2+

manipulation of the hair bundle. The interpretation of such responses (for example, in terms of gating kinetics or actions of Ca2+) may be complicated by the way in which the mechanical stimulus is coupled to the channel. The behavior of the macroscopic current may therefore reflect intrinsic channel properties as well as the micromechanics of channel connections. This review will consider both the performance and molecular identity of the MET channel. Topics to be covered include how it is gated and whether its sensitivity and kinetics are sufficient for the needs of transduction, especially in the cochlea. What are its intrinsic properties, ionic selectivity, pharmacological profile, and single-channel conductance, and do these identify or exclude candidates for channel structure? Finally, what is the role of Ca2+ in channel operation? Ca2+ permeates readily and also modulates channel gating at both external and internal faces, but the mechanism and significance of its action are still incompletely understood. 2.2

Gating Mechanism: Channel Kinetics

The fine mechanical sensitivity and fast kinetics that are hallmarks of hair cell transduction have been amply documented in both non-mammalian vertebrates [12, 13] and mammals [14–16]. The probability of channel opening, assayed by recording MET currents in individual hair cells, can be fully modulated by between 100 and 500 nm displacements of the tip of the hair bundle, depending on the preparation. A maximum displacement of 100 nm, although smaller than the diameter of a single

2.1 Introduction

Fig. 2.2 Activation of transduction currents in auditory hair cells. Rapid deflection (Dx) of the hair bundle with a piezoelectric stimulator connected to a rigid glass fiber evokes fast onset METcurrents. In the turtle (A), the activation time constant decreases from
400 ls to 50 ls with increasing stimulus amplitude. In the rat (B) the time course of current activation is faster than in the turtle and is

limited by the stimulus onset, time constant approximately 50 ls. The mammalian responses also show adaptation at a faster rate than in the turtle, dotted line denoting fit with a time constant of 300 ls. In both experiments the fluid bathing the hair bundle contained 0.05 mM Ca2+ and the hair cell had a holding potential of –80 mV at room temperature (
22 8C)

stereocilium, would be achieved in the cochlea only at the very highest sound levels [17]. Step deflections of the bundle with abrupt onset evoke MET currents that develop with a sub-millisecond time course (Fig. 2.2). The insignificant delay (< 20 ls) between the mechanical stimulus and channel opening argues that external force is fed directly to the MET channels without any of the biochemical amplification that occurs in other sensory receptors [12]. As might be expected if the stimulus directly affects the channel’s opening rate, the MET current activates with a principal time constant that depends upon stimulus amplitude. The activation time constant decreases with the size of bundle displacement from
400 ls for small stimuli to less than 50 ls for saturating ones (Fig. 2.2 [12, 13]). The channel’s activation kinetics are rapid even though measured in non-mammalian vertebrates, i.e., frog or turtle. However, when corrected for differences in body temperature, they may be too slow to explain the speed of transduction in auditory hair cells of mammals with high-frequency hearing. An indirect measure of transduction in the mammalian cochlea can be obtained from the cochlear microphonic, the summed extracellular voltage that is dominated by outer hair cell transduction currents [18, 19]. Cochlear microphonics can be observed for tonal stimuli with frequencies of at least 50 kHz [20], implying that channel gating in the mammalian cochlea can occur on a cycle-by-cycle basis up to this frequency. Such a frequency response would require the channel to activate with an unprecedented time constant of (2p. 50)–1 ms or 3 ls. It is possible that there are kinetic variations attributable to differences in channel subunit composition among species and that the MET channel in mammalian cochlear hair cells is optimized for the rapid gating needed for high-

33

34

2 Transduction Channels in Hair Cells

frequency hearing. Because of the bandwidth of the patch-clamp recording system and the limited speed with which a piezoelectric stimulator can deflect the bundles, the kinetic performance of MET currents in mammals is not known for certain. However, recent measurements suggest that they activate at room temperature in less than 50 ls even for small displacements, over five- to tenfold faster than in frogs or turtles under the same conditions (Fig. 2.2; [15]). The fast kinetics and high sensitivity of the hair cell MET channel distinguish it from those mechanically sensitive channels (for example, MscL) gated by changes in tension of the lipid bilayer [21]. These properties probably require the hair cell channel protein to be tethered to both the intracellular cytoskeleton and the extracellular tip link so that force can be applied directly across the channel without delays imposed by viscous damping.

2.2.1

Tip Links and Gating Springs

The tip links, stretching obliquely between the vertex of each stereocilium and the lateral membrane of its taller neighbor (Fig. 2.1; [4, 5]) are central to hair cell transduction. They explain theoretically the polarization of transduction, whereby only those movements of the hair bundle along its axis of symmetry are detected [22]. Deflections towards the taller edge of the bundle that tug on the tip links open the MET channels, whereas deflections of opposite polarity relax the links and close the channels. The importance of the tip links is supported by two lines of experimental evidence. Firstly, exposing hair cells to sub-micromolar Ca2+, buffered with BAPTA or EGTA, destroys the links and simultaneously abolishes transduction [6, 23, 24]. Secondly, experiments to localize the MET channels show that they are confined to the tips of the stereocilia in close proximity to tip-link insertion. These experiments include defining the location of maximum susceptibility to iontophoresis of the channel blocker dihydrostreptomycin [25] or visualizing the Ca2+ influx through open channels using intracellular calcium dyes and confocal microscopy [26, 27]. An alternative location for the channels is in the contact region just below the stereociliary tips where short lateral connections can be seen between the membranes of adjacent stereocilia (Fig. 2.1; [28]). Placement at this site is supported by immunogold labeling with an antibody raised against the binding site for amiloride, which blocks the transduction channels. A key premise of models for transduction is that the MET channels are opened by force applied by elastic elements, the “gating springs” that are stretched when the hair bundle is deflected (Fig. 2.3 [8]). One end of the spring is driven by bundle motion, and the other end is attached to the channel’s hypothetical gate. The gating springs have previously been identified with the tip links. The structure of the tip links is not known [29]; however, their coiled double-helical structure suggests that they are inextensible [30] and are more likely rigid connections for force transmission. Alternative sites for the series compliance of the gating springs lie in the intracellular cytoskeletal attachments of the channel or in the channel itself.

2.1 Introduction

Fig. 2.3 Cartoon of MET channel activation and adaptation. Hair bundle displacement tensions the tip link (T), which acts via gating springs that deliver force to the channel to increase its probability of being open. The gating springs are depicted as connecting to the channel both externally (Gs) and internally (Gs’). Ca2+ entering the stereocilium through the open channel may promote channel closure and evoke adaptation by binding at the inner face of the channel and/or by altering the stiffness of the internal gating spring Gs’. Gs’ may represent a cytoskeletal attachment

In the model’s simplest form [8], the channel occupies two states, closed and open (C$O), with the transition rates modulated by stimulus energy. The open probability of the channel (pO) is then a sigmoidal function of bundle displacement (X): pO(X) = [1 + exp (–z(X – XO)/kBT)]–1

(1)

where kB is the Boltzmann constant, T is absolute temperature, XO is the displacement for half-activation, and z is the single-channel gating force applied at the top of the bundle. The single-channel gating force is a measure of the channel’s mechanical sensitivity and has a value in frog or turtle of
0.3 pN [31, 32]. It can be used to specify the range of displacements over which the channel activates. Thus, the difference in bundle position from where the channel starts to activate (pO = 0.12) to where it is almost saturated (pO = 0.88) is equal to 4 kBT/z [31]. The operating range defined in this manner varies from 50 to 250 nm for different hair cells of the turtle cochlea [13, 32], frog saccule [8, 33], and mouse cochlea [34]. In the theoretical analysis, z is equal to c KGSd where KGS is the stiffness of the gating spring, d is a constant identified with the distance by which the gate moves during channel opening (
5 nm), and c is a dimensionless factor (
0.1) that depends on bundle geometry [35]. c symbolizes a transformer ratio, the reduction in displacement at the tip link compared to that at the top of the bundle. Variation in c in hair cells from different end organs and species can in theory generate a range of values for the single-channel gating force (0.1–0.8 pN [31]). Since c is inversely proportional to bundle height, a change in bundle dimensions provides a means of altering the sensitivity of transduction, matching it to the needs of the hair cell in a particular end-organ. The hair bundle may therefore be regarded as an accessory structure

35

36

2 Transduction Channels in Hair Cells

for converting forces and displacements occurring during inner ear stimulation to the molecular scale appropriate for deforming the MET channel protein.

2.2.2

Gating Compliance

An important and surprising feature of the gating-spring model for channel activation is that it can be used to derive the effective stiffness, Kch, of the gating spring-channel complex: Kch = KGS – z2pO (1 – pO) / kBT

(2)

The second (negative) term represents a reduction in stiffness referred to as “gating compliance” [8] that depends on the channel opening. Thus, the change in conformation of the channel between closed and open states introduces a phenomenological compliance in series with the gating spring. This signifies reversibility in transduction: a mechanical stimulus opens the channel, but, conversely, opening the channel can itself generate force. Such reversibility contrasts with the rectification in those sensory receptors (for example, photoreceptors) where a signal is amplified in an enzymatic cascade prior to activation of the transduction channel. An implication of Eq. (2) is that the force-displacement relationship of the hair bundle is nonlinear over the range of bundle positions where the channel is activated: it predicts that the bundle will not behave like a simple Hookian spring. The existence of such nonlinearity has been verified in measurements of bundle stiffness (Fig. 2.4; [8]). These are performed by applying force stimuli, usually with a flexible glass fiber more compliant than the bundle, and monitoring the resultant motion by photodiode imaging of the bundle or the attached fiber [8, 32, 34, 36, 37]. Experimental observation of the appropriate nonlinearity supports the model for gating and furthermore allows independent estimates of several of the theoretical parameters (z, KGS, d). Thus, the gating-spring model implemented in terms of a two-state (C$O) channel is a powerful tool for extrapolating from experimental measurements of macroscopic transduction current and bundle stiffness to some fundamental properties of the MET channel.

2.2.3

Three-state Channel Schemes

Several aspects of MET channel gating are not fully described by the simple two-state channel scheme but are better fit with a three-state channel, with two closed states and one open state (C1$C2$O). The most significant discrepancy is that the activation curve (pO–X relationship) for the MET channel deduced from recording MET currents in response to bundle displacement often does not show the symmetrical shape expected of a single Boltzmann equation (Eq. (1)). Instead, it has an asymmetric form in which the current saturates more gradually for positive displacements than for nega-

2.1 Introduction

Fig. 2.4 Nonlinear stiffness of the hair bundle. (A) The hair bundle was displaced with a flexible glass fiber, delivering a family of force steps (top) that generated MET currents (middle) and associated bundle displacements (bottom). The largest force step was 175 pN. Note the adaptation in the MET current for small stimuli. (B) The force-displacement relationship (top) and the current-displacement relationship (bottom, filled circles) were de-

rived from the records in (A); both MET currents and bundle movements were measured 1 ms after the onset of the force step. Hair bundle stiffness was calculated by differentiating the force-displacement results and is plotted against displacement (bottom, open circles). Note the reduction in stiffness in the region over which the MET channel activates. Taken from [32] with permission

tive ones, where a sharp corner is seen (Fig. 2.4). This type of asymmetry can be better described with a three-state channel [12–14]. In one version of the three-state model [31], the gating spring was postulated to be less compliant when the channel is in the first closed state (C1) than in either of the other two states (C2 or O). This was realized by assuming that a portion of the spring is immobilized or “latched” when the channel is in C1. A second refinement, a necessary correlate of the model, is that the gating springs eventually slacken for large negative displacements and therefore the energy of the channel becomes independent of displacement. The model reproduced the observed asymmetry of the channel’s activation curve. It also predicted a difference in bundle stiffness between positive and negative displacements that is sometimes found experimentally. However, there may be other explanations for this result (for example, an effect of intracellular Ca2+ on the stiffness of the gating spring). An alternative thermodynamic treatment of the three-state scheme [38] does not require the gating-spring stiffness to differ between states but assumes that the three

37

38

2 Transduction Channels in Hair Cells

states are engaged at different bundle positions, rather than being modulated over the entire stimulus range. Differential engagement was incorporated to account for the finite minimum open probability at bundle positions more negative than –50 nm. This treatment too successfully described the asymmetry in the pO-X relationship and predicted nonlinearity in the bundle mechanics. Differences in energies between the closed and open states in this model were found to be
5 kBT, somewhat smaller than the >10 kBT estimated from the two-state model [39]. An extension of the threestate model where intracellular Ca2+ affects the energy gaps between the different states predicts Ca2+-dependent adaptation in the channel opening [38]. The role of Ca2+ in channel adaptation will be discussed in Section 2.4.1.

2.3

Ionic Selectivity

The MET channel discriminates little between monovalent cations but is highly permeable to divalent cations, especially Ca2+ [40, 41]. The permeability ratios for monovalent cations determined from measuring reversal potentials for the transduction current are Li+, 1.14; Na+, 1.0; K+, 0.96; Rb+, 0.92; and Cs+, 0.79 [23, 41]. The equivalent permeabilities for divalent cations relative to Na+ are Ca2+, 3.8; Sr2+, 2.3; Ba2+, 2.2; Mg2+, 2.0; and Na+, 1.0, and they indicate only a modest preference for Ca2+ over monovalent cations [41]. However, they too were based on interpreting reversal potentials using the constant field equation, which assumes no interaction between different ionic species in the pore. This is unlikely to be the case, given the blocking action of external calcium on the flux of monovalent ions [23, 42] and the anomalous mole fraction effects seen with ion mixtures [43]. A more pertinent measure of the channel’s selectivity is obtained from the fraction of current carried by the different ions. This indicates an effective permeability ratio for Ca2+ over monovalent ions of more than 100:1 [42, 44]. The proportion of transduction current contributed by Ca2+ was inferred in the turtle [42] by loading hair cells with 1 mM of a calcium-sensitive dye and using the change in dye fluorescence to assay the Ca2+ influx. The calcium fluorescence was calibrated under conditions where Ca2+ was the sole charge carrier. This technique had previously been applied to the cyclic nucleotide-gated channel [45] (see Chapter 5). In saline with 2.8 mM Ca2+ and 130 mM Na+, over half the current was found to be carried by Ca2+ [42], and even in an endolymph-like solution containing only 50 lM Ca2+, Ca2+ carried at least one-tenth of the current (Fig. 2.5). Hair cells in vivo form a tight epithelium, separating two fluids of different ionic composition. The basolateral synaptic pole of the hair cells is exposed to Na+-rich perilymph similar to normal extracellular fluid, whereas the apical transducing pole is bathed in endolymph with high K+ and low Ca2+. The Ca2+ concentration in endolymph, in which the hair bundles are immersed, has been estimated in different preparations as between 20 and 60 lM [23, 46, 47]. The high flux of Ca2+ through the channel when exposed to low-Ca2+ endolymph is important physiologically because of the ion’s role in regulating adaptation of the MET channels. Although the channel is

2.3 Ionic Selectivity

Fig. 2.5 Fraction of MET current carried by Ca2+ in different external Ca2+ concentrations. (A) Fractional Ca2+ current determined by two methods: from measurements of Ca2+ influx using a cytoplasmic Ca2+-fluorescent indicator with Na+ (filled circles) or K+ (filled triangles) as the external monovalent ion; from the ratios of the MET current in Tris+ to the current in Na+ (open circles) assuming that Tris+ does not permeate the channel.

Note that in 0.06 mM Ca2+ (a concentration similar to that in endolymph) the divalent ion still carries 10–20 % of the total current. (B) Mean MET current carried by Na+ (filled circles) and Ca2+ (open circles). The dependence of Na+ current on extracellular Ca2+ was fitted with a single inhibitory-site model with inhibition constant KI = 1 mM. Modified from [42]

equally permeable to both Na+ and K+, substitution of K+ for Na+ as the chief monovalent ion increased the fractional current carried by Ca2+ and augmented its intracellular effect on adaptation (Fig. 2.5; [42]). In assessing the contributions of Ca2+ in vivo, there are conflicting requirements because the ion has a dual action on the channel: Ca2+ entry regulates its gating, but extracellular Ca2+ also blocks the channel and hence diminishes the total transduction current. Thus, the composition of endolymph, rich in K+ and unique among body fluids with a low level of Ca2+, may be specialized to compromise between the largest monovalent current and the sufficient Ca2+ influx to the control adaptation. A variety of larger organic cations can traverse the channel with permeability ratios relative to Na+ of 0.27 for choline, 0.16 for TMA, and 0.14 for TEA [41]. A recent finding is that the cationic styryl dye FM1-43 (molecular weight = 451) can also pass through the channel, its fluorescence on binding to intracellular membranes providing a sensitive indication of rapid influx [48, 49]. Channel permeability to such a large molecule has been explained by its elongated and linear structure with a cross-section not much greater than that of TEA [48]. This discovery is technically significant because intracellular accumulation of fluorescent FM1-43 can be used as a monitor of normal MET channel function without the need for invasive electrical recording. This method has been exploited to screen rapidly for genetic mutants lacking transduction [10].

39

40

2 Transduction Channels in Hair Cells

2.3.1

Blocking Compounds

Both Ca2+ and Mg2+ block the MET channel at millimolar extracellular concentrations (Fig. 2.5; [23, 42]), as do other inorganic cations such as Gd3+ [50] and La3+ [41]. The channel is also blocked by a broad spectrum of organic compounds (Tab. 2.1), including the calcium channel antagonists D600 [51, 52], cis-diltiazem [51], and FM1-43 [48]. Familiar antagonists of hair cell transduction, such as aminoglycoside antibiotics like dihydrostreptomycin (DHS), are also polycations. These antagonists share a number of properties that indicate that they may compete with Ca2+ for a binding site in the external mouth of the pore. They are effective only from the extracellular surface, and their block is voltage-dependent and diminished by depolarization that opposes the entry of the cation into the pore [41, 53]. Furthermore, the block is subject to competition from extracellular Ca2+: raising the Ca2+ concentration alleviates channel block with FM1-43 [48] and increases the KI for DHS (Tab. 2.1). Despite the blocking action of extracellular Ca2+, the current-voltage (I-V) relationship for the MET channel does not show pronounced rectification in control salines containing 1–4 mM Ca2+ [13, 14, 41, 55] and is approximately ohmic for membrane potentials between –50 and +50 mV. I-V relations with a modest outward curvature at voltage extremes of 100 mV have been fit [14, 55] to a single energy barrier model: I = k(exp((1-d)(V–Vrev)/Vs) – exp(-d(V–Vrev)/Vs))

(3)

where Vrev is the reversal potential (
0 mV), Vs is a measure of the steepness of rectification (
50 mV), d is the fractional distance of the energy barrier from the outside boundary of the membrane’s electric field (
0.5), and k is a proportionality constant. Tab. 2.1

Blocking agents of the hair cell transduction channel

Blocker

KI (lM)

nH

Reference

Ca2+ Gd3+ Tetracaine cis-Diltiazem D600 Dihydrostreptomycin Dihydrostreptomycin Dihydrostreptomycin Dihydrostreptomycin

1000 10 608 236 111 8 (0.25 mM Ca) 23 (2.5 mM Ca) 44 (5 mM Ca) 0.8 (low CF*, 0.25 mM Ca) 6 (high CF, 0.25 mM Ca) 50 53 24 6 2.3 2.4

1 1.1

42 50 51 51 51, 52 53 54 53 56

Amiloride Amiloride Amiloride Benzamil Curare FM1-43

1 0.9 1 1 1.7 2 1.6 1.0 1.2

57 55 56 55 58 48

KI: half-blocking concentration; nH: Hill coefficient; CF: characteristic frequency

2.4 MET Channel Adaptation

However, the I-V relationship does acquire significant outward rectification in the presence of external blockers such as DHS and amiloride, with greater attenuation for inward current as the cationic blocker is swept into the channel at negative membrane potentials [41, 53, 55]. In its ionic selectivity and pharmacological signature, the MET channel resembles the cyclic nucleotide-gated (CNG) channel. Both are non-selective cation channels with similar permeability sequence for alkali cations and high selectivity for Ca2+. Neither shows the same degree of discrimination for Ca2+ over monovalent cations found with voltage-dependent Ca2+ channels, but, nevertheless, their high permeability to Ca2+ has important physiological consequences. In each case Ca2+ influx through channels activated by the sensory stimulus modulates transduction and promotes adaptation (see Chapter 5). Both types of channels are also subject to block by
1 mM Ca2+ at the external face and are affected by some of the same antagonists, such as amiloride, diltiazem, D600, and tetracaine (for CNG channels: [59, 60]). This might imply a similar pore structure, but there are significant dissimilarities. The MET channel is blocked by aminoglycoside antibiotics and has a five- to tenfold larger unitary conductance. Most conspicuously, it is significantly permeable to organic cations such as choline, TMA, and TEA, none of which pass through the CNG channel (PCat/PNa < 0.019; [61]). Indeed the MET channel is anomalous compared to other channels in displaying a high selectivity for Ca2+ while still being substantially permeable to organic ions. For example, its choline permeability is twice that for the nicotinic acetylcholine receptor [62]. This implies a wide pore, a property that may be linked to the unusually large single-channel conductance (>100 pS).

2.4

MET Channel Adaptation

The high Ca2+ permeability of the MET channel implies that its opening by a mechanical stimulus promotes influx and intracellular accumulation of Ca2+. The local rise in Ca2+ acts as a feedback signal to regulate channel opening just as it does in other sensory receptors; control may be exerted either through resetting the mechanical input or by directly interacting with the channel to affect its gating. Such regulation is referred to as adaptation [63, 64] and is manifested as a decline in the transduction current during a sustained deflection of the bundle. This reflects a parallel translation of the pO-X relation along the displacement axis in the direction of the stimulus to keep the channels near their maximal sensitivity. Two main mechanisms can be distinguished based on their speed and range: slow adaptation, with a time constant (sA) of 10–100 ms and a working range of
1 lm [65, 66], and fast adaptation, with a sA of 0.3-5.0 ms [13, 67] and a working range of
0.1 lm. Slow adaptation has been proposed to operate through the action of an unconventional myosin that alters the mechanical input, possibly by adjusting the tension in the tip links [68]. In one scenario, the upper attachment point of the tip link is ferried up and down the stereocilium by attachment through the plasma membrane to an array of myosins that track along sub-membranous actin filaments [69]. In mouse vestibular

41

42

2 Transduction Channels in Hair Cells

hair cells, myosin-1c has been implicated as the isoform driving slow adaptation [70]. Modification of the ATP-binding site on myosin-1c renders it susceptible to block by ADP analogues [71], and introducing such analogues through the recording pipette abolishes slow adaptation. In mouse cochlear hair cells, a distinct isoform, myosin7a, may perform a supportive role in slow adaptation [72]. In contrast, fast adaptation is unaffected by agents that interfere with the myosin ATPase [70, 73,], and furthermore it has kinetics too fast to be compatible with the slow cycle time of an ATPase. However, despite the differences in underlying mechanism, fast and slow adaptation mechanisms are both controlled by Ca2+ entering the stereocilia through the MET channels [13, 33, 42]. Furthermore, Ca2+ action in one or both components of adaptation may be implemented through binding to calmodulin [74], which has been shown immunohistochemically to be concentrated at the tips of the stereocilia [75, 76]. Besides these two types of adaptation, there is evidence for at least one other slower process affecting the operating point of the MET channels. Extracellular application of a membrane-permeant form of cyclic AMP causes a positive shift in the pO-X relation along the displacement axis [67]. This shift can be as large as 1 lm with no effect on fast adaptation or on the slope of the pO-X relation. The pathway may involve stimulation by cAMP (produced by action of a Ca2+-dependent adenyl cyclase) of protein kinase A, which in turn phosphorylates the channel or the myosin motor [77]. 2.4.1

Ca2+ Regulation of Adaptation

The Ca2+ dependence of fast adaptation has been extensively documented in turtle auditory hair cells. Experiments manipulating either extracellular or intracellular Ca2+ have shown it to have a dual effect on the adaptation time constant (sA) and the open probability (pO) of the MET channels at rest. Lowering extracellular Ca2+ slows sA and shifts the pO–X relation in the negative direction, thus increasing the resting pO [13, 78]. Similar effects are produced by loading hair cells with high concentrations of calcium buffer (1–5 mM BAPTA), which argues for an intracellular action of Ca2+. Because relatively large concentrations of BAPTA, which has a fast Ca2+-binding rate, were needed to affect adaptation, the site of calcium’s action is probably not far from the channel. By comparing the extent of the shift in the pO-X relation with different BAPTA concentrations, the distance Ca2+ diffuses to its target was assessed as no more than 15–35 nm from the mouth of the channel [78]. This distance was obtained by computing Ca2+ gradients along the stereocilia from the tip where the channel was assumed to be located. The Ca2+ gradient was steeper with higher buffer concentrations. Finally, the rate of adaptation (1/sA) is proportional to the amount of Ca2+ entering the stereocilia, assayed by Ca2+ imaging bundles filled with the fluorescent dye Ca-Green-1 [42]. The proportionality extended up to rates of 1.4 ms–1, which implies that the subsequent molecular events underlying adaptation must occur in less than 1 ms. The simplest hypothesis based on this collection of results is that fast adaptation requires a direct interaction of Ca2+ with the MET channels (possibly via a regulatory subunit) to modulate their probability of opening (Fig. 2.3; [13, 73, 78]. As extra sup-

2.4 MET Channel Adaptation

Fig. 2.6 Model of activation of MET channel. (A) Kinetic scheme in which the channel is converted from closed (C) to open (O) states by bundle displacement (x) modulating rate constants K1, K2, and K3 and can also bind two Ca2+ ions. Ki are of form exp[ai(X – Xa1)], where X is bundle displacement and ai and Xa1 are constants. (B) Open probability of the channel (Popen) for hair bundle displacements of different amplitude calculated using the kinetic scheme in (A). Ca2+ ions bind to the closed state with sequential affinities of 20 lM

and 10 lM and to the open state with affinities of 2 lM and 5 lM. Ca2+ is cleared from the internal mouth of the channel with a time constant of 0.1 ms. Note that channel activation accelerates for stimuli of increasing amplitudes and that lowering the channel current carried by Ca2+ (ICa) from 4 pA to 1 pA slows both activation and adaptation rates. The bundle’s resting position is indicated by the stimulus level following the step. Note that lowering Ca2+ increases Popen at this resting position

port for this notion, intracellular Ca2+ has been found to alter the time constant of channel activation as well as adaptation [79], indicating that its action is closely linked with channel gating. If Ca2+ acts by binding to calmodulin, then kinetics demands that calmodulin be constitutively bound to the channel. The interaction of Ca2+ with the MET channel can be expressed in terms of a kinetic scheme [Fig. 2.6; 80]) with the binding of multiple (two or more) Ca2+ ions being required to achieve the requisite sensitivity. This scheme can reproduce the effects of Ca2+ on the activation and adaptation kinetics of the channel solely by altering the Ca2+ influx (Fig. 2.6). An alternative site for Ca2+ to act is on the gating springs to reduce their stiffness (KGS) (Fig. 2.3; [81]). This latter mechanism acting alone would reduce the gating force z, and also the slope of the pO-X relationship, which would extend the channel’s operating range.

43

44

2 Transduction Channels in Hair Cells

2.4.2

The Function of Adaptation

The accepted role of adaptation in sensory receptors is to preserve transducer sensitivity for small changes in stimulus about a larger background level. The limited operating range of the MET channels requires that transduction be endowed with one or more adaptation mechanisms to maintain the channels near their maximal sensitivity. In the inner ear, adaptive processes exist peripherally to the hair cells to prevent large static displacement being imposed on the cells, shielding them from over-stimulation and damage. These include the helicotrema, the shunt between the two perilymphatic compartments of the cochlea, which acts like a high-pass filter for sound frequencies below about 100 Hz [82, 83]. More precise control is exerted at the hair cell level by the slow myosin-based adaptation that can adjust the mechanical input to the MET channels. However, fast adaptation with its small dynamic range and rapid kinetics may have a more subtle function. A possible clue to the role of fast adaptation comes from the turtle auditory papilla, where the adaptation time constant (sA) varies inversely with hair cell characteristic frequency (CF) [67, 78]. (The CF is the sound frequency to which a given hair cell is most sensitive, and in all vertebrate cochleae, it changes systematically with location to generate a tonotopic map.) How the variation in sA originates will be considered in delineation of the single channel properties, but it is partly attributable to a change in stereociliary influx of Ca2+. The corner frequency (1/2psA) of the high-pass filter contributed by fast adaptation is approximately two-thirds of the CF. This suggests that the adaptation may in some way contribute to hair cell frequency selectivity, a cell’s ability to discriminate different frequency components in a sound stimulus. This idea is reinforced by the observation that in saline containing physiological (50 lM) Ca2+ concentrations, fast adaptation may display under-damped resonance at frequencies in the turtle’s auditory range [78]. In the mammalian cochlea, where the CFs are much higher than those in the turtle, sA is correspondingly smaller, with a value of 100 ls or less [15]. In the turtle cochlea the channel’s activation time constant also increases with CF, [79] and the activation and fast adaptation time constants therefore confer on transduction a variable band-pass filter matched to the CF. This filter is unlikely to be the major source of hair cell frequency selectivity in the turtle, which instead stems from a sharply-tuned electrical resonance [84] produced by interplay of a voltage-dependent Ca2+ current and a Ca2+-activated K+ current [85]. The hair cell’s Ca2+-activated K+ channels vary in number and kinetics with location along the cochlea in order to generate a range of CFs [86, 87]. In contrast, the transduction filter may provide a mechanism for actively restricting the bandwidth to improve the signal-to-noise ratio of transduction within the frequency range encoded by the hair cell [88]. This must be done on a cycleby-cycle basis and to be optimal should therefore vary with CF. (An intriguing developmental question is how the center frequency of the filter supplied by the MET channels is matched to the frequency of electrical resonance endowed by the Ca2+-activated K+ channels.) It has been previously argued [89, 90] that using an active filter to narrow the stimulus bandwidth is a way of extending the physical detection limits of sensory

2.5 Single-channel Conductance

transduction in the face of intrinsic thermal noise. Another way of viewing this active filter is in terms of the mechanical stimulus. As a result of the gating compliance, closing the MET channels through adaptation elicits a mechanical-output, active motion of the hair bundle [8, 36, 91, 92]. Thus, a positive bundle displacement opens the MET channels, leading to a rise in intracellular Ca2+, which recloses the channels producing negative recoil. Such active motion may effectively amplify the cell’s response for external stimuli near threshold [93, 94]. The presence of fast adaptation in mammalian outer hair cells [15] implies that the mechanism is retained as part of the amplification in the mammalian cochlea [94, 95].

2.5

Single-channel Conductance

A distinguishing feature of the MET channel in assigning it to a particular protein family is single-channel conductance, but this property has been difficult to define. Conventional methods of recording single channels in membrane patches, cell-attached or detached, are unavailable probably because of the sub-micron diameter of the stereocilia and the need to preserve extracellular mechanical links to observe gating with physiological stimuli. Approaches employing analysis of current fluctuations [96] or inference from measurement of the macroscopic current and the number of active stereocilia [26] are at best indirect. The only systematic method devised has been to destroy or inactivate the majority of channels in a bundle and to monitor the one or two channels remaining in whole-cell recording mode [23, 97]. The number of channels was reduced by brief exposure to sub-micromolar Ca2+ concentrations, which most likely works by severing the majority of the tip links [6, 24]. The technique carries the assumption that the remaining channels were not significantly modified by the isolation procedure. Some justification for this assumption is provided by the finding that the residual channel behaved similarly to the macroscopic transduction current: it was gated by small displacements of the hair bundle, displayed fast activation and adaptation, and was blocked by suitable doses of dihydrostreptomycin [97]. Conspicuously, changing external Ca2+ had a dual action on channel properties resembling the effects on the macroscopic current (Fig. 2.7). Thus, reducing Ca2+ increased the channel amplitude by removal of block and also slowed the time course of channel activation and adaptation. The latter outcome was manifested in both the slowing of the ensemble average current and the increase in the channel’s mean open time. Two main results emerged from the single-channel experiments. Firstly, the unitary conductance was found to be surprisingly large, between 80 and 160 pS in 2.8 mM extracellular Ca2+; the range of values nearly doubled to 150–300 pS on removing the blocking action of Ca2+ by lowering its concentration to 50 lM (Fig. 2.7 [97]). Since the KI for channel block by Ca2+ is
1 mM (Fig. 2.5; [42]), measurements in 50 lM Ca2+ should reflect the maximum conductance of the channel in its unblocked state. Secondly, variation in channel conductance, in either high or low Ca2+, was correlated with the CF of the hair cell: cells with higher CFs possessed channels with bigger conduc-

45

46

2 Transduction Channels in Hair Cells

Fig. 2.7 Single MET channels and their modulation by Ca2+. (A) Four single-channel responses recorded in a turtle hair cell for hair bundle deflections (Dx) of 150 nm in 2.8 mM extracellular Ca2+. Middle trace is the ensemble average of 140 channel responses. At the bottom is the open-time histogram of channel events in the absence of bundle stimulation. C and O denote the closed and open levels of the channel. (B) Two single-channel responses in the same cell in 0.05 mM extracellular Ca2+. Middle trace is the ensemble average of 250

channel responses, and at the bottom is the histogram of open-times without stimulation. Reducing extracellular Ca2+ had two effects: it increased the mean single-channel amplitude from 7 pA to 15 pA and it slowed channel kinetics. The latter effect was indicated by the slower time course of activation and adaptation of the ensemble average current and the increase in mean open time (sO) (holding potential, –80 mV). Unitary conductance in the Ca2+-unblocked state is 190 pS

tance. MET channels of
100 pS conductance have also been observed in hair cells of chicks [41] and mammals [34]. It will be important to determine whether this conductance varies with hair cell CF, especially in the mammalian cochlea. The change in channel conductance with CF in the turtle cochlea is unlikely to result from differences in Ca2+ homeostasis within the stereocilia because virtually identical conductance ranges were found in high and low extracellular Ca2+ [97]. Changes in MET channel conductance are most simply explained if hair cells in different regions of the cochlea contain channels with unique structure or sub-unit composition. Further support for this notion comes from the finding that MET channels in hair cells with higher CFs also differ functionally in having faster activation kinetics [56, 79]. The tonotopic distribution of MET channel properties, conductance and ki-

2.5 Single-channel Conductance

netics, is unexpected, but it parallels gradients in the properties of the Ca2+-activated K+ channel that underlie variation of the electrical resonant frequency in the turtle cochlea [87, 98]. The latter variation may stem from differential expression of alternatively spliced isoforms of the Ca2+-activated K+ channel a-subunit combined with a cochlear gradient in an accessory b-subunit [86, 99, 100].

2.5.1

Number of MET Channels Per Stereocilium

The single-channel conductance can be used to infer the number of channels per stereocilium if the maximum size of the macroscopic transduction current and the number of stereocilia in the hair bundle are known. If the MET channels are uniformly distributed across the bundle, each stereocilium possesses between one and two channels irrespective of hair cell CF [96]. Increases in the stereociliary complement [101] and single-channel conductance with CF together produce a severalfold change in the peak amplitude of the macroscopic MET current between turtle hair cells tuned to low and high frequency [67]. Inferring the number of channels per stereocilium from the total current assumes that all available channels are active during macroscopic recordings. The channels or their mechanical connections (for example, the tip links) may be damaged when the cochlea is isolated, which would lead to an underestimate in the total MET current and hence in the number of channels per stereocilium. Denk et al. [26] used two-photon imaging of stereociliary Ca2+ transients to assess how many of the stereocilia were actively contributing to a macroscopic transduction current. From those experiments they too concluded that each stereocilium possesses two channels. Based on the observation that bundle stimulation generated Ca2+ signals in both the shortest and tallest stereociliary rows, it was argued [26] that channels occur at both ends of the tip link. Such an arrangement implies a negative cooperativity between pairs of channels at the two ends of the tip link where opening of a channel at one end of the link relieves the force on the channel at the opposite end of the link. However, there is no experimental support for an interaction of this kind. Furthermore, tip-link destruction in sub-micromolar Ca2+ during the single-channel experiments should have left two channels for each intact tip link, but often one channel remained after exposure to low Ca2+ [97]. These results suggest that the MET channels occur at only one end of the tip link, probably at its insertion into the top of the stereocilium. The tonotopic gradient in the time constant of fast adaptation may also be explicable in terms of the single-channel properties. Adaptation was evident in the ensemble averages of single-channel activity (Fig. 2.7) from which a time constant could be extracted. Systematic variation in this time constant was seen in the ensemble average currents: cells with higher CFs adapting more rapidly [97], just as with the macroscopic current. Since the rate of fast adaptation is directly proportional to Ca2+ influx [42], the increase in adaptation rate with hair cell CF may be partly explained by an increase in channel size: doubling the channel conductance doubles the amount of Ca2+ entering and thus halves the adaptation time constant. Changes in other channel properties

47

48

2 Transduction Channels in Hair Cells

may also contribute in varying the adaptation time constant. There is no evidence that the increase in channel conductance is accompanied by a concomitant augmentation of Ca2+ permeability, which would theoretically speed up adaptation [56]. However, spectral analysis of transducer current noise has suggested a difference in channel kinetics between high- and low-frequency hair cells [56]. This difference is endorsed by measurements of activation kinetics obtained by fitting the onset of the MET current in response to rapid bundle deflections [79]. Speeding up the activation and deactivation kinetics of the channel might produce additional acceleration of adaptation rate over that realizable by changes in single-channel conductance.

2.6

The MET Channel as a Member of the TRP Family

There are several distinctive features of the MET channel gleaned from measurements on intact hair cells that may aid in its molecular classification. These include a high selectivity for Ca2+ over other cations, a broad spectrum of blocking agents, large unitary conductance, and regulation by intracellular Ca2+. These attributes eliminate several channel contenders [102]. One is the epithelial Na+ channel (ENaC), which has subunits orthologous to the MEC-4 and MEC-10 proteins that form a mechanoreceptor channel in touch neurons of the nematode worm Caenorhabditis elegans [103, 104] (see Chapter 1). Like the hair cell MET channel, ENaC is blocked by amiloride. However, known forms of ENaC are Na+-selective channels with low Ca2+ permeability and small unitary conductance (13–40 pS [105]. Moreover, the properties of their block by amiloride are substantially different from the MET channel, with a 100-fold greater affinity for the drug (KI =
0.5 lM) and a Hill coefficient of 1.0 [106]. Nevertheless, a member of the ENac family may be a constituent of the transduction channel in vertebrate cutaneous mechanoreceptors [107]. The most likely channel candidate on the basis of present evidence belongs to the TRP (transient receptor potential) superfamily, some members of which possess channel properties resembling the hair cell MET channel [108]. The MEC channel subunits were obtained from genetic screens of mutants lacking a touch response in C. elegans. Such screens also generated a separate channel, CeOSM-9, from animals with defects in osmotic avoidance and nose touch [109]. CeOSM-9 was the first TRP channel implicated in mechanosensitivity, but two relatives were subsequently cloned from Drosophila melanogaster (see Chapter 8). One is DmNOMPC, the mutation of which largely abolishes receptor potentials in the touch-sensitive bristle organs [110] but only mildly affects hearing in Drosophila [111]. The other is DmNanchung (DmNAN), which is localized to the ciliary neurons in Johnston’s organ and whose mutation deafens the insects [112]. Both DmNOMPC and DmNAN have structures consistent with TRP channels with six transmembrane domains, a pore region between S5 and S6, intracellular N- and C-termini, and multiple ankyrin repeats at the N-terminus (29 in DmNOMPC and 5 in DmNAN). The two proteins have been assigned to different subclasses of the TRP superfamily: DmNAN to TRPV, similar to CeOSM-9, and DmNOMPC to a new sub-class, TRPN [113]. A vertebrate NOMPC with 62 % simi-

2.7 Conclusions

larity at the amino acid level to the Drosophila version has recently been identified in hair cells of the zebra fish, Danio rerio [10]. Removal of NOMPC gene function by injection of morpholino antisense oligonucleotides into larvae caused initial deafness and imbalance that later reversed as the morpholino was diluted by the endogenous transcript. There is as yet no evidence for the occurrence of NOMPC in hair cells of other vertebrates, including mammals.

2.6.1

Properties of TRPV Channels

Although no channel information exists for NOMPC, DmNAN, when expressed in cultured Chinese hamster ovary (CHO) cells, forms Ca2+-permeable channels that confer responsiveness to osmotic stress, consistent with a role in mechanotransduction [112]. More detailed information on channel attributes is available for other TRP members [108], which behave as nonspecific cation channels highly permeable to Ca2+ with sizable single-channel conductance. The precise properties depend on channel type but may be illustrated for the TRPV subfamily [114–123] to which DmNAN belongs. This family can be subdivided in terms of Ca2+ selectivity: for TRPV1, TRPV3, and TRPV4 PCa/PNa is
10:1, whereas for TRPV5 and TRPV6 it is
100:1. The less Ca2+-selective TRPV1, TRPV3, and TRPV4 channels exhibit larger single-channel conductance (110, 137, 90, or 300 pS) compared to TRPV5 and TRPV6 (78, 50 pS). The functional category to which DmNAN belongs is not yet known, but clearly there are TRPV members with permeation properties encompassing those of the hair cell MET channel. Although there have been no systematic studies of TRP channel permeability to organic cations, TRPV1, like the hair cell MET channel, can pass the large FM1-43 dye [49]. Within the TRPV family only TRPV4 and DmNAN have been recognized as being mechanosensitive. Nevertheless, other channels in the TRP superfamily are capable of mechanotransduction, including the yeast vacuolar channel, Yvc1p [124], the polycystic kidney disease gene product, PKD-2 [125], and a renal stretchinhibitable channel [126]. These too are Ca2+ permeable and have large (>100 pS) unit conductance. As a final mark of similarity, TRP channel gating can be modulated by divalent ions. For example, TRPV5 is blocked in a voltage-dependent manner by extracellular Ca2+ [123], Mg2+, and low concentrations of La3+ [122]. It is also desensitized by elevation of intracellular Ca2+ [127]. TRPV4 is modulated by extracellular and intracellular Ca2+ [128] and is blocked by Gd3+ [119].

2.7

Conclusions

Measurements in intact hair cells have provided considerable insight into the properties of the transduction channels, including their fast sub-millisecond gating and exquisite nanometer mechanical responsiveness, their cationic permeability with preference for Ca2+, and their large single-channel conductance. The permeability properties

49

50

2 Transduction Channels in Hair Cells

are consistent with the MET channel being a member of the TRP superfamily, a hypothesis reinforced by the identification of TRP channels underlying mechanosensitivity in the hearing systems of Drosophila and zebra fish. Whether or not a TRP channel eventually proves to be the hair cell transduction in all vertebrate inner ears, it is likely that multiple isoforms or accessory proteins will be needed to explain the distribution of MET channel properties encountered in the turtle. The availability of single-channel measurements in intact hair cells provides a norm for comparison with future cloned channels to define subunit composition, as was the case for olfactory cyclic nucleotide-gated channels [129]. From recordings in intact hair cells there is also evidence that Ca2+ regulates the channel and its mechanical input to mediate multiple components of adaptation. Neither the mechanism nor the role of adaptation is entirely clear. Fast adaptation may directly alter MET channel gating, whereas slow adaptation may reset the mechanical stimulus via the action of one or more unconventional myosins. For auditory hair cells, fast adaptation varies with hair cell CF, which may be important for maximizing the signal-to-noise ratio of transduction in a frequency band around CF. By generating active hair bundle motion, it may also amplify the extrinsic mechanical stimulus [94]. Cloning the MET channel may lead to an understanding of how the channel interacts with Ca2+ and with other subcellular components, including myosins. It may in addition reveal the existence of multiple channel isoforms with unitary conductance or kinetics specialized for operation in different frequency ranges. If the MET channel is localized to the osmiophilic cap seen in electron micrographs of the tips of the stereocilia (Fig. 2.1), it could well be part of a larger protein complex analogous to the transduciosome proposed in Drosophila photoreceptors [130, 131]. Other proteins would include those anchoring the channel to the internal cytoskeleton and to the extracellular links, one or more of the different myosin isoforms (1C, 3, 6, 7A and 15 [132–134] localized to hair cells), and a plasma membrane CaATPase [135], which is crucial for Ca2+ extrusion from the bundle. All may be cemented together by the PDZ -domain protein INAD, which localizes TRP channels in Drosophila photoreceptors [131, 136, 137] (see Chapter 8). INAD was recently identified in mammalian hair cells [134]. There is a significant lack of knowledge about how the mechanical input (for example, increased tension in the tip links) is coupled to channel gating, which may involve one or more accessory proteins. The molecular identities of the proteins that comprise the “gating spring” are also currently unknown. However, the existence of mechanical linker proteins has been established for the mechanotransducer complex in C. elegans. Touch mutations have shown that proper mechanical gating of the DEG/ENac channel requires multiple intracellular and extracellular accessory proteins [138] (see Chapter 1). Application of molecular genetics to mammalian hearing may similarly identify not only the hair cell MET channel but also the set of accessory proteins essential for the correct targeting and operation of the channel [139, 140].

2.7 Conclusions

Acknowledgments

This work was supported by the National Institutes on Deafness and other Communicative Disorders Grant RO1 DC 01362. I would like to thank Carole Hackney for commenting on the manuscript; Tony Ricci, Helen Kennedy, and Andrew Crawford, who took part in the experiments in Figures 2.2 and 2.7; and Gaurang Patel for help with modeling in Fig. 2.6.

References 1 2

3

4

5

6

7

8

9

10

11

Tilney LG, Tilney MS Functional organization of the cytoskeleton. Hear Res 1986, 22, 55–77. Bagger-Sjoback D, Wersall J The sensory hairs and tectorial membrane of the basilar papilla in the lizard Calotes versicolor. J Neurocytol 1973, 2, 329–350. Little KF, Neugebauer D-C Interconnections between the stereovilli of the fish inner ear. II Systematic investigation of saccular hair bundles of Rutilus rutilus (Teleostei). Cell Tissue Res 1985, 284, 473–479. Pickles JO, Comis SD, Osborne MP Crosslinks between stereocilia in the guinea pig organ of Corti, and their possible relation to sensory transduction. Hear Res 1984, 15, 103–112. Furness DN, Hackney CM Cross-links between stereocilia in the guinea pig cochlea. Hear Res 1985, 18, 177–188. Assad JA, Shepherd GM, Corey DP Tip-link integrity and mechanical transduction in vertebrate hair cells. Neuron 1991, 7, 985–994. Flock A, Flock B, Murray E. Studies on the sensory hairs of receptor cells in the inner ear. Acta Otolaryngol 1977, 83, 85–91. Howard J, Hudspeth AJ Compliance of the hair bundle associated with gating of mechanoelectrical transduction channels in the bullfrog’s saccular hair cell. Neuron 1988 1, 189–199. Hudspeth AJ Extracellular current flow and the site of transduction by vertebrate hair cells. J Neurosci 1982, 2, 1–10. Sidi, S., Friedrich, R.W. and Nicolson, T. NompC TRP channel required for vertebrate sensory hair cell mechanotransduction. Science 2003, 301, 96–99. Sukharev SI, Blount P, Martinac B, Kung C. Mechanosensitive channels of Escherichia coli: the MscL gene, protein, and activities. Annu Rev Physiol. 1997, 59, 633–657.

12

13

14

15

16

17

18

19

20

21

22

Corey DP, Hudspeth AJ Kinetics of the receptor current in bullfrog saccular hair cells. J Neurosci 1983, 3, 962–976. Crawford AC, Evans MG, Fettiplace R Activation and adaptation of transducer currents in turtle hair cells. J Physiol 1989, 419, 405–434. Kros CJ, R€ usch A, Richardson GP Mechanoelectrical transducer current in hair cells of the cultured neonatal mouse cochlea. Proc R Soc Lond B 1992, 249, 185– 193. Kennedy HJ, Evans MG, Crawford AC, Fettiplace R Fast adaptation of mechanoelectrical transducer channels in mammalian cochlear hair cells. Nat Neurosci 2003, 6, 832–836. Vollrath MA, Eatock RA Time course and extent of mechanotransducer adaptation in mouse utricular hair cells: comparison with frog saccular hair cells. J Neurophysiol 2003, 90, 2676–2689. Robles L, Ruggero MA Mechanics of the mammalian cochlea. Physiol Rev 2001, 81, 1305–1352. Dallos P, Cheatham MA Production of cochlear potentials by inner and outer hair cells. J Acoust Soc Am 1976, 60, 510–512. Patuzzi RB, Yates GK, Johnstone BM. The origin of the low-frequency microphonic in the first cochlear turn of guinea-pig. Hear Res 1989, 39, 177–188. Pollak G, Henson OW, Novick A Cochlear microphonic audiograms in the pure tone bat, Chilonycteris parnelli parnelli. Science 1972, 176, 66–68. Hamill OP, Martinac B. Molecular basis of mechanotransduction in living cells. Physiol Rev 2001, 81, 685–740. Shotwell SL, Jacobs R, Hudspeth AJ. Directional sensitivity of individual vertebrate hair cells to controlled deflection of their hair bundles. Ann N Y Acad Sci 1981, 374, 1–10.

51

52

2 Transduction Channels in Hair Cells 23

24

25

26

27

28

29

30

31

32

33

34

Crawford AC, Evans MG, Fettiplace R The actions of calcium on the mechano-electrical transducer current of turtle hair cells. J Physiol 1991, 434, 369–398. Hackney, C.M & Furness, D.N. Hair cell ultrastructure and mechanotransduction: morphological effects of low extracellular calcium on stereociliary bundles in the turtle cochlea. In Active Hearing, Flock A, Ottoson D, Ulfendahl M (eds.), Pergamon, London, pp 103–111. 1995. Jaramillo F, Hudspeth AJ Localization of the hair cell’s transduction channels at the hair bundle’s top by iontophoretic application of a channel blocker. Neuron 1991, 7, 409–420. Denk W, Holt JR, Shepherd GM, Corey DP Calcium imaging of single stereocilia in hair cells: localization of transduction channels at both ends of tip links. Neuron 1995, 15, 1311–1321. Lumpkin EA, Hudspeth AJ Detection of Ca2+ entry through mechanosensitive channels localizes the site of mechanoelectrical transduction in hair cells. Proc Natl Acad Sci USA 1995, 2, 10297–10301. Hackney CM, Furness DN, Benos DJ, Woodley JF, Barratt J Putative immunolocalization of the mechanoelectrical transduction channels in mammalian cochlear hair cells. Proc R Soc Lond B 1992, 248, 215–221. Goodyear R, Richardson G A novel antigen sensitive to calcium chelation that is associated with the tip links and kinocilial links of sensory hair bundles. J Neurosci 2003, 23, 4878–4887. Kachar B, Parakkal M, Kurc M, Zhao Y, Gillespie PG High-resolution structure of haircell tip links. Proc Natl Acad Sci USA 2000, 97, 13336–13341. Markin VS, Hudspeth AJ Gating-spring models of mechanoelectrical transduction by hair cells of the internal ear. Annu Rev Biophys Biomol Struct 1995, 24, 59–83. Ricci AJ, Crawford AC, Fettiplace R Mechanisms of active hair bundle motion in auditory hair cells. J Neurosci 2002, 22, 44–52. Assad JA, Hacohen N, Corey DP Voltage dependence of adaptation and active bundle movement in bullfrog saccular hair cells. Proc Natl Acad Sci USA 1989, 86, 2918–2922. Ge´le´oc GS, Lennan GW, Richardson GP, Kros CJ A quantitative comparison of mechanoelectrical transduction in vestibular and auditory hair cells of neonatal mice. Proc R Soc Lond B 1997, 264, 611–621.

35

36

37

38

39

40

41

42

43

44

45

46

47

48

Howard J, Roberts WM, Hudspeth AJ Mechanoelectrical transduction by hair cells. Annu Rev Biophys Biophys Chem 1988, 17, 99–124. Crawford AC, Fettiplace R The mechanical properties of ciliary bundles of turtle cochlear hair cells. J Physiol 1985, 364, 359–379. Russell IJ, Kossl M, Richardson GP Nonlinear mechanical responses of mouse cochlear hair bundles. Proc R Soc Lond B 1992, 250, 217–27. van Netten SM, Kros CJ Gating energies and forces of the mammalian hair cell transducer channel and related hair bundle mechanics. Proc R Soc Lond B 2000, 267, 1915–1923. Hudspeth AJ Hair-bundle mechanics and a model for mechanoelectrical transduction by hair cells. In Sensory transduction, DP Corey, SD Roper (eds.), Rockefeller University Press, New York, pp 357–370. 1992 Corey DP, Hudspeth AJ Ionic basis of the receptor potential in a vertebrate hair cell. Nature 1979, 281, 675–677. Ohmori H Mechano-electrical transduction currents in isolated vestibular hair cells of the chick. J Physiol 1985, 359, 189–217. Ricci AJ, Fettiplace R Calcium permeation of the turtle hair cell mechanotransducer channel and its relation to the composition of endolymph. J Physiol 1998, 506, 159–173. Lumpkin EA, Marquis RE, Hudspeth AJ The selectivity of the hair cell’s mechanoelectricaltransduction channel promotes Ca2+ flux at low Ca2+ concentrations. Proc Natl Acad Sci USA 1997, 94, 10997–11002. Jorgensen F, Kroese AB. Ca selectivity of the transduction channels in the hair cells of the frog sacculus. Acta Physiol Scand 1995, 155, 363–376. Frings S, Seifert R, Godde M, Kaupp UB. Profoundly different calcium permeation and blockage determine the specific function of distinct cyclic nucleotide-gated channels. Neuron 1995, 15, 169–179. Bosher SK, Warren RL Very low calcium content of cochlear endolymph, an extracellular fluid. Nature 1978, 273, 377–378. Salt AN, Inamura N, Thalmann R, Vora A Calcium gradients in inner ear endolymph. Am J Otolaryngol 1989, 10, 371–375. Gale JE, Marcotti W, Kennedy HJ, Kros CJ, Richardson GP FM1-43 dye behaves as a permeant blocker of the hair-cell mechanotransducer channel. J Neurosci 2001, 21, 7013–7025.

2.7 Conclusions 49

50

51

52

53

54

55

56

57

58

59

60

61

Meyers JR, MacDonald RB, Duggan A, Lenzi D, Standaert DG, Corwin JT, Corey DP. Lighting up the senses: FM1-43 loading of sensory cells through nonselective ion channels. J Neurosci 2003, 23, 4054–4065. Kimitsuki T, Nakagawa T, Hisashi K, Komune S, Komiyama S Gadolinium blocks mechanoelectric transducer current in chick cochlear hair cells. Hear Res 1996, 101, 75–80. Ricci, A.J. Pharmacological clues to the nature of the mechanoelectric transducer channel. Assoc Res Otolaryngol Abstr 2003, 26, 841. Baumann M, Roth A The Ca++ permeability of the apical membrane in neuromast hair cells. J Comp Physiol A 1986, 158, 681–688. Kroese AB, Das A, Hudspeth AJ Blockage of the transduction channels of hair cells in the bullfrog’s sacculus by aminoglycoside antibiotics. Hear Res 1989, 37, 203–217. Kimitsuki T, Ohmori H Dihydrostreptomycin modifies adaptation and blocks the mechanoelectric transducer in chick cochlear hair cells. Brain Res 1993, 624, 143–150. R€ usch A, Kros CJ, Richardson GP Block by amiloride and its derivatives of mechanoelectrical transduction in outer hair cells of mouse cochlear cultures. J Physiol 1994, 474, 75–86. Ricci A Differences in mechanotransducer channel kinetics underlie tonotopic distribution of fast adaptation in auditory hair cells. J Neurophysiol 2002, 87, 1738–1748. Jorgensen F, Ohmori H. Amiloride blocks the mechano-electrical transduction channel of hair cells of the chick. J Physiol 1988, 403, 577–588. Glowatzki E, Ruppersberg JP, Zenner HP, R€ usch A Mechanically and ATP-induced currents of mouse outer hair cells are independent and differentially blocked by d-tubocurarine. Neuropharmacol 1997, 36, 1269–1275. Frings S, Lynch JW, Lindemann B Properties of cyclic nucleotide-gated channels mediating olfactory transduction. Activation, selectivity and blockage. J Gen Physiol 1992, 100, 45–67. Fodor AA, Gordon SE, Zagotta WN Mechanism of tetracaine block of cyclic nucleotidegated channels. J Gen Physiol 1997, 109, 3–14. Picco C, Menini A The permeability of the cGMP-activated channel to organic cations in retinal rods of the tiger salamander. J Physiol 1993, 460, 741–758.

62

63 64

65

66

67

68

69 70

71

72

73

74

75

Dwyer TM, Adams DJ, Hille B The permeability of the endplate channel to organic cations in frog muscle. J Gen Physiol 1980, 75, 469–492. Eatock RA Adaptation in hair cells. Annu Rev Neurosci 2000, 23, 285–314. Fettiplace R, Ricci AJ Adaptation in auditory hair cells. Curr Opin Neurobiol 2003, 13, 446–451. Assad JA, Corey DP An active motor model for adaptation by vertebrate hair cells. J Neurosci 1992, 12, 3291–3309. Shepherd GM, Corey DP. The extent of adaptation in bullfrog saccular hair cells. J Neurosci 1994, 14, 6217–6229. Ricci AJ, Fettiplace R The effects of calcium buffering and cyclic AMP on mechano-electrical transduction in turtle auditory hair cells. J Physiol 1997, 501, 111–124. Howard J, Hudspeth AJ Mechanical relaxation of the hair bundle mediates adaptation in mechanoelectrical transduction by the bullfrog’s saccular hair cell. Proc Natl Acad Sci USA 1987, 84, 3064–3068. Gillespie PG, Corey DP Myosin and adaptation by hair cells. Neuron 1997, 19, 955–958. Holt JR, Gillespie SK, Provance DW, Shah K, Shokat KM, Corey DP, Mercer JA, Gillespie PG. A chemical-genetic strategy implicates myosin-1c in adaptation by hair cells. Cell 2002, 108, 371–381. Gillespie PG, Gillespie SK, Mercer JA, Shah K, Shokat KM Engineering of the myosinibeta nucleotide-binding pocket to create selective sensitivity to N(6)-modified ADP analogs. J Biol Chem 1999, 274, 31373–31381. Kros CJ, Marcotti W, van Netten SM, Self TJ, Libby RT, Brown DM, Richardson GP, Steel KP Reduced climbing and increased slipping adaptation in cochlear hair cells of mice with Myo7a mutations. Nature Neurosci 2002, 5, 41–47. Wu YC, Ricci AJ, Fettiplace R Two components of transducer adaptation in auditory hair cells. J Neurophysiol 1999, 82, 2171–2181. Walker RG, Hudspeth AJ Calmodulin controls adaptation of mechanoelectrical transduction by hair cells of the bullfrog’s sacculus. Proc Natl Acad Sci USA 1996, 93, 2203–2207. Furness DN, Karkanevatos A, West B, Hackney CM An immunogold investigation of the distribution of calmodulin in the apex of the cochlear hair cells. Hear Res 2002, 173, 10–20.

53

54

2 Transduction Channels in Hair Cells 76

77

78

79

80

81

82

83

84

85

86 87

88

89

Cyr JL, Dumont RA, Gillespie PG: Myosin 1-c interacts with hair-cell receptors through its calmodulin- binding IQ domains. J Neurosci 2002, 22, 2487–2495. Ge´le´oc G, Corey DP Modulation of mechanoelectrical transduction by protein kinase A in utricular hair cells of neonatal mice. Assoc Res Otolaryngol Abstr 2001, 24, 242. Ricci AJ, Wu YC, Fettiplace R The endogenous calcium buffer and the time course of transducer adaptation in auditory hair cells. J Neurosci 1998, 18, 8261–8277. Fettiplace, R., Crawford, A.C. and Ricci, A.J. The effects of calcium on mechanotransducer channel kinetics in auditory hair cells. In Biophysics of the Cochlea: from molecule to model, AW Gummer (ed.), World Scientific, Singapore, pp. 65–72. 2003. Choe Y, Magnasco MO, Hudspeth AJ A model for amplification of hair-bundle motion by cyclical binding of Ca2+ to mechanoelectrical-transduction channels. Proc Natl Acad Sci USA 1998, 95, 15321–15326. Martin P, Bozovic D, Choe Y, Hudspeth AJ. Spontaneous oscillation by hair bundles of the bullfrog’s sacculus. J Neurosci 2003, 23, 4533–48. Franke R, Dancer A. Cochlear mechanisms at low frequencies in the guinea pig. Arch Otorhinolaryngol 1982, 234, 213–8. Cheatham MA, Dallos P. Inner hair cell response patterns: implications for low-frequency hearing. J Acoust Soc Am 2001, 110, 2034–44. Crawford AC, Fettiplace R An electrical tuning mechanism in turtle cochlear hair cells. J Physiol 1981, 312, 377–412. Art JJ, Fettiplace R Variation of membrane properties in hair cells isolated from the turtle cochlea. J Physiol 1987, 385, 207–42. Fettiplace R, Fuchs PA Mechanisms of hair cell tuning. Ann Rev Physiol 1999, 61, 809–34. Wu YC, Art JJ, Goodman MB, Fettiplace R. A kinetic description of the calcium-activated potassium channel and its application to electrical tuning of hair cells. Prog Biophys Mol Biol 1995, 63, 131–58. Dinklo T., van Netten, M., Marcotti, W. & Kros, C.J. Signal processing by transducer channels in mammalian outer hair cells. In Biophysics of the Cochlea: from molecule to model, AW Gummer (ed.), World Scientific, Singapore, pp. 73–79. 2003. Bialek W Physical limits to sensation and perception. Ann Rev Biophys Biophys Chem 1987, 16, 455–478.

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

Block S.M. Biophysical principles of sensory transduction. In Sensory transduction, DP Corey, SD Roper (eds.) Rockefeller University Press, New York., pp 1–17, 1992. Benser ME, Marquis RE, Hudspeth AJ Rapid, active hair bundle movements in hair cells from the bullfrog’s sacculus. J Neurosci 1996, 16, 5629–5643. Ricci AJ, Crawford AC, Fettiplace R Active hair bundle motion linked to fast transducer adaptation in auditory hair cells. J Neurosci 2000, 20, 7131–7142. Martin P, Hudspeth AJ Active hair-bundle movements can amplify a hair cell’s response to oscillatory mechanical stimuli. Proc Natl Acad Sci USA 1999, 96, 14306–14311. Fettiplace R, Ricci AJ, Hackney CM Clues to the cochlear amplifier from the turtle ear. Trends Neurosci 2001, 24, 169–175. Hudspeth A Mechanical amplification of stimuli by hair cells. Curr Opin Neurobiol 1997, 7, 480–486. Holton T, Hudspeth AJ The transduction channel of hair cells from the bull-frog characterized by noise analysis. J Physiol 1986, 375, 195–227. Ricci, A.J., Crawford, A.C. & R. Fettiplace, R. Tonotopic variation in the conductance of the hair cell mechanotransducer channel. Neuron 2003, (in press). Art JJ, Wu YC, Fettiplace R. The calcium-activated potassium channels of turtle hair cells. J Gen Physiol 1995, 105, 49–72. Jones EM, Gray-Keller M, Fettiplace R. The role of Ca2+-activated K+ channel spliced variants in the tonotopic organization of the turtle cochlea. J Physiol 1999, 518, 653–65. Ramanathan K, Michael TH, Jiang GJ, Hiel H, Fuchs PA. A molecular mechanism for electrical tuning of cochlear hair cells. Science 1999, 283, 215–7. Hackney CM, Fettiplace R, Furness DN. The functional morphology of stereociliary bundles on turtle cochlear hair cells. Hear Res 1993, 69, 163–75. Strassmaier M, Gillespie PG The hair cell’s transduction channel. Curr Opin Neurobiol 2002, 12, 380–386. HuangM,ChalfieMGeneinteractionsaffecting mechanosensory transduction in Caenorhabditis elegans. Nature 1994, 367, 467–470. Goodman MB, Ernstrom GG, Chelur DS, O’Hagan R, Yao CA, Chalfie M MEC-2 regulates C. elegans DEG/ENaC channels needed for mechanosensation. Nature 2002, 415, 1039–1042.

2.7 Conclusions 105 Ismailov II, Awayda MS, Berdiev BK, Bubien

106

107

108

109

110

111

112

113 114

115

116

117

JK, Lucas JE, Fuller CM, Benos DJ Triplebarrel organization of ENaC, a cloned epithelial Na+ channel. J Biol Chem 1996, 271, 807–16. Sariban-Sohraby S, Benos DJ The amiloridesensitive sodium channel. Am J Physiol 1986, 250, C175–190. Price MP, McIlwrath SL, Xie J, Cheng C, Qiao J, Tarr DE, Sluka KA, Brennan TJ, Lewin GR, Welsh MJ The DRASIC cation channel contributes to the detection of cutaneous touch and acid stimuli in mice. Neuron 2001, 32, 1071–83. Minke B, Cook B TRP channel proteins and signal transduction. Physiol Rev 2002, 82, 429–72. Colbert HA, Smith TL, Bargmann CI OSM-9, a novel protein with structural similarity to channels, is required for olfaction, mechanosensation, and olfactory adaptation in Caenorhabditis elegans. J Neurosci 1997, 17, 8259–69. Walker RG, Willingham AT, Zuker, CS A Drosophila mechanosensory transduction channel. Science 2000, 87, 2229–2234. Eberl DF, Hardy RW, Kernan MJ. Genetically similar transduction mechanisms for touch and hearing in Drosophila. J Neurosci 2000, 20, 5981–8. Kim J, Chung YD, Park D-Y, Choi S, Shin, DW, Soh H, Lee HW., Son W, Yim J, Park CS, Kernan MJ, Kim C A TRPV family ion channel required for hearing in Drosophila. Nature 2003, 424, 81–84. Corey DP New TRP channels in hearing and mechanosensation. Neuron 2003, 39, 585–8. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 1997, 389, 816–24. Premkumar LS, Ahern GP. Induction of vanilloid receptor channel activity by protein kinase C. Nature 2000, 408, 985–90. Mohapatra DP, Wang SY, Wang GK, Nau C. A tyrosine residue in TM6 of the vanilloid receptor TRPV1 involved in desensitization and calcium permeability of capsaicin-activated currents. Mol Cell Neurosci 2003, 23, 314–324. Xu H, Ramsey IS, Kotecha SA, Moran MM, Chong JA, Lawson D, Ge P, Lilly J, SilosSantiago I, Xie Y, DiStefano PS, Curtis R, Clapham DE. TRPV3 is a calcium-permeable temperature-sensitive cation channel. Nature 2002, 418, 181–186.

118 Strotmann R, Harteneck C, Nunnenmacher

119

120

121

122

123

124

125

126

127

128

K, Schultz G, Plant TD OTRPC4, a nonselective cation channel that confers sensitivity to extracellular osmolarity. Nature Cell Biol 2000, 2, 695–702. Liedtke W, Choe Y, Marti-Renom MA, Bell AM, Denis CS, Sali A, Hudspeth AJ, Friedman JM, Heller S Vanilloid receptor-related osmotically activated channel (VR-OAC), a candidate vertebrate osmoreceptor. Cell 2000, 103, 525–535. Voets T, Prenen J, Vriens J, Watanabe H, Janssens A, Wissenbach U, Bodding M, Droogmans G, Nilius B. Molecular determinants of permeation through the cation channel TRPV4. J Biol Chem 2002, 277, 33704–33710. Vennekens R, Hoenderop JG, Prenen J, Stuiver M, Willems PH, Droogmans G, Nilius B, Bindels RJ. Permeation and gating properties of the novel epithelial Ca2+ channel. J Biol Chem 2000, 275, 3963–3969. Nilius B, Vennekens R, Prenen J, Hoenderop JG, Bindels RJ, Droogmans G Whole-cell and single channel monovalent cation currents through the novel rabbit epithelial Ca2+ channel ECaC. J Physiol 2000, 527, 239–248. Yue L, Peng JB, Hediger MA, Clapham DE CaT1 manifests the pore properties of the calcium-release-activated calcium channel. Nature 2001, 410, 705–709. Zhou XL, Batiza AF, Loukin SH, Palmer CP, Kung C, Saimi Y The transient receptor potential channel on the yeast vacuole is mechanosensitive. Proc Natl Acad Sci USA 2003, 100, 7105–7110. Nauli SM, Alenghat FJ, Luo Y, Williams E, Vassilev P, Li X, Elia AE, Lu W, Brown EM, Quinn SJ, Ingber DE, Zhou J Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells. Nat Genet 2003, 33, 129–137. Suzuki M, Sato J, Kutsuwada K, Ooki G, Imai M Cloning of a stretch-inhibitable nonselective cation channel. J Biol Chem 1999, 274, 6330–6335. den Dekker E, Hoenderop JG, Nilius B, Bindels RJ The epithelial calcium channels, TRPV5 & TRPV6: from identification towards regulation. Cell Calcium 2003, 33, 497–507. Watanabe H, Vriens J, Janssens A, Wondergem R, Droogmans G, Nilius B Modulation of TRPV4 gating by intra- and extracellular Ca2+. Cell Calcium 2003, 33, 489–495.

55

56

2 Transduction Channels in Hair Cells 129 B€ onigk W, Bradley J, M€ uller F, Sesti F,

130

131

132

133

134

Boekhoff I, Ronnett V, Kaupp UB and Frings S The native rat olfactory nucleotide-gated channel is composed of three distinct subunits. J Neurosci 1999, 19, 5332–5347. Scott K, Zuker CS Assembly of the Drosophila phototransduction cascade into a signalling complex shapes elementary responses. Nature 1998, 395, 805–8. Huber A Scaffolding proteins organize multimolecular protein complexes for sensory signal transduction. Eur J Neurosci 2001, 14, 769–776. Hasson T, Gillespie PG, Garcia JA, MacDonald RB, Zhao Y, Yee AG, Mooseker MS, Corey DP: Unconventional myosins in innerear sensory epithelia. J Cell Biol 1997, 137, 1287–1307. Belyantseva IA, Azevedo RB, Fridell RA, Friedman TB, Kachar B Assoc Res Otoloaryngol Abstr 2002, 25, 157. Walsh T, Walsh V, Vreugde S, Hertzano R, Shahin H, Hsika S, Lee MK, Kanaan M, King M-C, Avraham K From flies’ eyes to our ear: mutations in a human class III myosin causes nonsyndromic hearing loss DFNB30. Proc Natl Acad Sci USA 2002, 99, 7518–7523.

135 Dumont RA, Lins U, Filoteo AG, Penniston

136

137

138

139

140

JT, Kachar B, Gillespie PG Plasma membrane Ca2+-ATPase isoform 2a is the PMCA of hair bundles. J Neurosci 2001, 21, 5066–5078. Shieh BH, Zhu MY. Regulation of the TRP Ca2+ channel by INAD in Drosophila photoreceptors. Neuron 1996 16, 991–8. Chevesich J, Kreuz AJ, Montell C Requirement for the PDZ domain protein, INAD, for localization of the TRP store-operated channel to a signaling complex. Neuron 1997 18, 95–105. Tavernarakis N, Driscoll M Molecular modeling of mechanotransduction in the nematode Caenorhabditis elegans. Ann Rev Physiol 1997, 59, 659–89. Steel KP, Kros CJ A genetic approach to understanding auditory function. Nat Genet 2001, 27, 143–9. Boda B, El-Amraoui A, Bahloul A, Goodyear R, Daviet L, Blanchard S, Perfettini I, Fath KR, Shorte S, Reiners J, et al: Myosin VIIa, harmonin and cadherin 23, three Usher I gene products that cooperate to shape the sensory hair bundle. EMBO J 2002, 21, 6689–6699.

57

3

Acid-sensing Ion Channels Kenneth A. Cushman and Edwin W. McCleskey

3.1

Introduction

Krishtal and Pidoplichko discovered acid-sensing ion channels in 1980 by using voltage clamp recordings from rat sensory neurons [1]. This and a series of subsequent papers [2–4] from the same group established the essential features of these molecules: (1) the channels open when pH drops below 7.0, (2) they pass Na+ over K+ about as well as voltage-gated Na+ channels do, (3) amiloride blocks them at rather high concentrations (tens of micromolars), and (4) distinct subtypes of the channel evident

Fig. 3.1 Different desensitization kinetics of acid-evoked currents observed in trigeminal ganglia cells. Currents recorded in response to pH stimulus of 6.2. From [2] Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

58

3 Acid-sensing Ion Channels

in different sensory neurons differ in their pH sensitivity and rates of activation and desensitization (Fig. 3.1). Krishtal argued that these acid-evoked currents were caused by unique ion channels, whereas another group thought they might be acid-modified Ca++ channels [5]. The debate was loud but the audience was small, and the subject fell from view until 1997. That was when Lazdunski and colleagues cloned an acid-sensing, amiloride-sensitive, Na+-selective channel and coined the acronym ASIC (acid-sensing ion channel) [6]. The work showed that ASICs were a subfamily of mammalian epithelial Na+ channels (ENaCs) and their relatives in C. elegans, degenerins (see Chapters 1 and 4), channels that are all Na+ selective and amiloride sensitive but that differ in their gating mechanisms. From the first, ASICs were proposed to be transducers for acid-evoked pain [3]. They also have been implicated in mechanosensation, taste transduction, and learning and memory. A variety of reviews discuss their various properties and proposed functions [7–9].

3.2

ASICs and the DEG/ENaC Superfamily

ASICs belong to the superfamily of epithelial sodium channels (ENaCs) and degenerins (DEGs) (Fig. 3.2). All channels in the DEG/ENaC family are sodium selective and voltage insensitive and are blocked by amiloride (albeit at very different concentrations). Some are constitutively open, protons gate some, a peptide (FMRFamide) gates one [10], and some may be mechanosensitive. The ASICs are defined by homology to each other rather than by their acid sensitivity. ASIC1 and ASIC3 are readily opened by small pH changes, but ASIC2 needs extreme pH (5.0) to open and ASIC4 may not be acid sensitive at all. The variability in acid sensitivity suggests that ASICs may prove to have functions that are unrelated to their name. ENaCs are among the most critical ion channels to mammalian biology. Expressed in epithelial cells, they control sodium reabsorption for fluid homeostasis. For example, by controlling sodium flux across kidney epithelia, they control the amount of blood in the body [11]. Amiloride can block ENaCs at tens of nanomolars, making it clinically useful for control of blood pressure. Concentrations at least 1000-fold greater are needed to block ASICs. The dominant “degenerin” mutations cause the neurons that express this protein in C. elegans to swell and lyse [12–14], due to a constitutive activity of the ion channel [15]. Most of the DEGs are expressed in mechanosensory cells, leading to “mechanosensory abnormal” (MEC) or uncoordinated (UNC) phenotypes in the mutants and raising the possibility that these channels are mechanosensors (see Chapter 1). The first ASIC cloned was ASIC2a, originally called MDEG (mammalian degenerin) and BNC1 (Brain Na+ Channel 1) [16, 17]. It has homology to DEGs (20–29 % identity) and is constitutively active with a degenerin mutation. Neither group that cloned it connected it with Krishtal’s acid-gated sodium channels; indeed, ASIC2a requires quite extreme acid (pH 5) to open.

3.2 ASICs and the DEG/ENaC Superfamily

Fig. 3.2 Phylogenetic tree of DEG/ENaC family members. Sequences aligned using ClustalW

ASIC1a was cloned by two groups: Corey’s, which named it BNaC2 (Brain Na+ Channel 2) [18], and Lazdunski’s, which realized its acid sensitivity and named it ASIC [6]. ASIC1a in rats is a 526-amino-acid protein that is expressed throughout the peripheral and central nervous system. An extracellular pH drop below 6.9 activates a transient amiloride-sensitive sodium current, with a half maximal activating pH (pH0.5) near 6.5 (Tab. 3.1) [6, 19, 20]. ASIC1a has a significant Ca++ permeability and is blocked by higher levels of Ca++. This channel has kinetics, expression pattern, pharmacology, and selectivity similar to some of the currents described by Krishtal and Pidoplichko. The Lazdunski group then cloned and characterized rat ASIC3 [21]. It was originally called DRASIC (dorsal root ASIC) because its mRNA was found only in dorsal root ganglion sensory neurons in rats. ASIC3 is more sensitive to acid than is ASIC1a, with a pH0.5 of 6.7 and a steeper activation curve [22]. ASIC3 desensitizes much faster than ASIC1a. Calcium ions carry current through ASIC3 in the absence of Na+ [22], but there is no reported Ca++ permeability in the presence of Na+. Lazdunski’s group then showed that ASIC2a is also a proton-gated ion channel [23]. ASIC2a is much less sensitive to protons than are either ASIC3 or ASIC1a, with a pH0.5 of 4.35 [24]. They also cloned a splice variant of ASIC2 with a different amino terminus, ASIC2b (MDEG2). ASIC2b is inactive as a homomer but modulates the kinetics, selectivity, and sensitivity of both ASIC2a and ASIC3 when co-expressed

59

60

3 Acid-sensing Ion Channels Tab. 3.1.

Summary of properties of ASIC homomers and heteromers

Channel

Other Names

PNa/PK

PNa/PCa

pH0.5

Sustained Location Current?

Modulating Peptides

ASIC1a

ASIC, ASICa, BNaC2 ASICb BNC1, BNaC1, MDEG1 MDEG2 DRASIC

7.8-13

2.5-18.5

5.3-6.6

N

C/S

FMRFRFRP1/2

14 10

high 20

5.8-5.9 4.35

N

S C/S

Inactive 13,5

– high

– 6.5-6.7

– Y

C/S S

– 4.8

–

C

ASIC2a/2b ASIC2a/3

Inactive – 7.2 36 Doesn’t form Doesn’t form * >1 for bothhigh

3.9 4.3-6.5

Y Y

ASIC3/2b

*

6.5-6.7

Y

ASIC1b ASIC2a ASIC2b ASIC3

ASIC4 SPASIC ASIC1a/2a ASIC1a/2b ASIC1a/3

NPFFFMRFNPSFRFRP1/2

NPFFFMRF

* = Sodium-selective initial current, nonselective sustained current S = Sensory neurons C = Central nervous system

[23]. ASIC2a and ASIC2b are both expressed throughout the central and peripheral nervous systems [20, 23]. Other labs have since cloned further genes and splice variants in this family. ASIC1b (ASIC-b) and ASIC-b2 are splice variants of ASIC1 [25–27]. ASIC1b lacks the Ca++ permeability of ASIC1a, has a decreased sensitivity to protons (pH0.5=5.84) [19], and is restricted to sensory neurons. ASIC-b2 does not form functional homomers but does associate with ASIC1b to decrease its affinity for protons. ASIC4 was the last member of the ASIC family to be cloned [28, 29]. It is expressed in the CNS and has not been shown to form a functional ion channel or to modulate the acid sensitivity of other ASICs. The human ASICs are different from the rat ASICs described above. They show different expression patterns, isoforms, and pharmacology. Human ASIC3, for example, has three isoforms and is in the CNS, PNS, and internal organs such as lung and kidney [30].

3.4 Assembly Into Channels

3.3

Amino Acid Structure

ASICs have two predicted transmembrane domains in their structure, with most of the protein being in the extracellular loop. There are two conserved cysteine-rich domains (CRDs) in the extracellular loop that are common to all DEG/ENaC family members. These CRDs in ENaCs play a role in proper expression and activation of the channel [31]. The two transmembrane domains are presumed to be analogous to the pore-forming alpha helices of other ion channels [32, 33]. The second (C-terminal) transmembrane domain and a short region (the “P-loop”) just upstream from it are considered the primary structures contributing to pores. This hypothesis has not been investigated in ASICs, but systematic mutation of residues in this domain of ENaCs alters Na+ selectivity [34, 35], in agreement with expectation. However, mutation of the residues just upstream of transmembrane domain 1 of ASIC2 also clearly modifies Na+ selectivity, arguing that this region also contributes to the pore [36].

3.4

Assembly Into Channels

There is no doubt that multiple ASIC proteins must assemble in order to form a functional channel, but the stoichiometry is debated. One group provides data arguing for a nonameric structure, whereas another argues for a tetrameric structure [37, 38]. A similar conflict exists in the ENaC field [39, 40]. It seems likely that ASICs assemble with the same stoichiometry as ENaCs given their conserved topology. Unlike ENaCs, whose functional channels must contain different channel subtypes [41], ASICs can form homomeric ion channels. ASIC1a, -1b, -2a, and -3 can form functional homomeric channels. ASIC2b can form heteromers with other ASICs to modify their activity but is inactive by itself. ASIC4 has been shown neither to form a functional ion channel nor to modulate other subunits. ASIC1a and ASIC2a share common patterns of expression in the central nervous system, suggesting they may colocalize to form heteromers. It has been shown in oocytes that the two subunits can form a heteromeric channel [42]. The heteromer is less selective than either homomer and has a decreased permeability to Ca++ and an intermediate proton sensitivity. This heteromer has also been observed in cultured hippocampal neurons [43]. ASIC2b is electrically silent as a homomer but can change the properties of ASIC2a and ASIC3 when co-expressed. ASIC2a loses sensitivity to protons and creates a sustained nonselective current when ASIC2b is present [23, 44]. Co-expression of ASIC2b with ASIC3 does not change the pH sensitivity of ASIC3; however, the normal sodium-selective sustained current seen with ASIC3 at low pH becomes nonselective [23]. ASIC2b has no significant effect when co-expressed with ASIC1a. ASIC2a and ASIC3 are co-expressed in many dorsal root ganglion (DRG) neurons. When expressed together in oocytes, they form a functional heteromeric channel with

61

62

3 Acid-sensing Ion Channels

novel properties [45]. There is a large increase in sustained current that is sodium selective, like the sustained current seen in ASIC3 homomers. The ASIC currents seen in mouse DRG are different from those evoked from cloned channels, suggesting the presence of heteromers in vivo. Benson reproduced the currents evoked from medium to large DRG cells by expressing three subunits together: ASIC1a, -2a, and -3 [46]. It is not clear whether the three subunits all combine to form one channel or whether native currents are due to a mix of heteromers composed of two subunits. In individual knockouts of ASIC1, ASIC2a, or ASIC3, DRG ASIC currents matched heteromeric currents of the remaining two subunits. ASIC3 seems necessary for the fast desensitization kinetics seen in DRG, since the ASIC3 knockouts had slow kinetics. ASIC1 in mouse seems significantly different from the rat clone. Rat ASIC1 is substantially less sensitive to pH than rat ASIC3, whereas the two mouse clones both activate at and just below pH 7.

3.5

Pharmacology

Like all members of the DEG/ENaC family, ASICs are blocked by amiloride. The amiloride block is voltage dependent [47], with an IC50 ranging from 10–100 lM depending on the subunit composition of the channel. ENaCs have a much higher affinity for amiloride, typically on the order of 100 nM [48]. At high enough concentrations, amiloride can block L- and T-type Ca++ channels [49, 50], Na+/Ca++ exchangers, Na+/H+ exchangers, Na+ pumps, Ca++ pumps [51], and other channels. The low affinity of amiloride for ASICs, coupled with its promiscuity as a blocker, makes it a weak pharmacological tool. The mechanosensing ion-channel blocker Gd3+ also blocks ASIC2a/3 heteromers [45], but once again, this is not a very selective blocker. Residues contributing to amiloride binding are well established in ENaCs; amiloride is thought to bind to two sites, one in the extracellular domain and one near the pore [48, 52, 53]. Insight into the reason that ASICs are so much less sensitive to amiloride than are ENaCs arises from a study that mutated the residues in ASICs that are analogous to the ENaC amiloride-binding site near the pore. The mutation eliminated block and unmasked a second action of amiloride: enhancement of the ASIC current [47]. The interpretation is that ASICs also have two amiloride-binding sites, one that activates and one, like in ENaCs, that blocks. The only known ASIC-specific blocker is Psalmotoxin 1 (PcTx1), which comes from the venom of a South American tarantula [54, 55]. PcTx1 is specific for ASIC1a homomers and blocks with an IC50 of 0.9 nM. This potent toxin has proven useful to distinguish ASIC1a homomers from heteromers in the CNS [43]. The divalent ion Zn++ potentiates current in ASIC2a-containing channels [56]. Zn++ does not affect ASIC3 or ASIC1a homomers, but it does potentiate heteromers formed with ASIC2a. The potentiation is greater at higher pH values, with a greater than sevenfold enhancement at pH 6.0 for ASIC2a and with an EC50 of 120 lM. The presence of Zn++ (300 lM) caused a leftward shift of the activation curve of ASIC1a/2a

3.6 Gating

heteromers (pH0.5 from 5.5 to 6.0). Two histidine residues are essential for this potentiation; when either one is mutated to alanine, the effect disappears. ASIC currents are very sensitive to Ca++ [2]. Ca++ decreases both single-channel conductance and open probability [57, 58]. Dropping extracellular Ca++ increases the current through ASICs and shifts its activation curve to more basic pHs [59]. The steadystate inactivation curves of ASICs are also shifted by Ca++, with higher levels of Ca++ preventing inactivation [19]. Ca++ has a central role in gating the channel, as will be discussed further. FMRFamide (Phe-Met-Arg-Phe-amide) gates one DEG/ENaC channel called FaNaCh, and ASICs have strong homology to it. This led to the question of whether FMRFamide or other peptides can modulate ASICs. Mammals do not have FMRFamide, but they have related peptides such as neuropeptide FF (NPFF). NPFF is involved in pain modulation via both opioid-dependent and -independent pathways [60]. Although the peptides do not gate ASICs, they do slow the rate of desensitization and introduce a sustained current [61]. Interestingly, the peptides need to be applied prior to proton activation to have an effect. NPFF heavily modulates ASIC3 and ASIC2a/3 heteromers in oocytes but has no effect on ASIC1a or -2a (Tab. 3.1). FMRFamide modulates all the above channels except homomeric ASIC2a. Interestingly, neither NPFF nor FMRFamide modulates human ASIC3 [62]. NPSF, another mammalian RFamide peptide, also potentiates ASIC3 and DRG acid-evoked currents [63]. Experiments with other RFamide-related peptides show slowed desensitization and increased peak amplitude in DRG and heterologously expressed ASIC1 and ASIC3 [64]. The ability of RFamide peptides to induce sustained currents in some ASICs raises the idea that they can increase the sensitivity of ASICs in sustained pain states.

3.6

Gating

Our lab has recently proposed an idea for ASIC gating (Fig. 3.3) that, at this early stage, should be considered controversial. ASIC gating is highly sensitive to calcium concentration in the extracellular medium [2, 19, 57, 58, 65]. This raises the possibility that the proton-binding site is a titratable calcium-binding site, more or less like calcium chelators such as EGTA. We tested this hypothesis on ASIC3 and, consistent with it, found (1) that Ca++ and H+ compete at the gating site and (2) that eliminating extracellular Ca++ opens the channel without any change in extracellular pH [65]. Like a Ca++ chelator, the affinity of the channel for Ca++ diminishes as pH drops. It has an apparent affinity of around 150 nM at pH 9, of 10 lM at pH 7.4, and of 100 lM at pH 7.0. Thus, protons essentially catalyze release of Ca++ from this binding site, and we suggest that this release is the event that opens the channel. At physiological Ca++ (mM) concentrations, negligibly few channels will be free of Ca++ at pH 7.4, whereas nearly 10 % will be free at pH 7.0, more or less the fraction of channels open at this pH. Multiple protons must bind for effective Ca++ release, conveniently explaining the very steep proton activation curve of ASIC3.

63

64

3 Acid-sensing Ion Channels Fig. 3.3 Schematic showing proposed gating mechanism for ASIC3. At rest, ASIC3 has a high affinity for Ca++, which blocks the channel. When H+ binds, it decreases the affinity for Ca++. When Ca++ is released, the channel is relieved from block and opens. From [65]

A mathematical model that fit all our gating data also quantitatively fit Ca++ block data on single channels. This led us to suggest that the Ca++ ion involved in gating simply blocks the pore – in other words, that the channel does not open due to a proton-induced conformation change but, rather, due to a relief of Ca++ block. In contrast, desensitization clearly appears to be a slow, H+-driven conformation change. The model seems appealing and explains a variety of data on ASIC3, but it needs to be critically examined on other ASICs. At present, one study in the literature seems supportive. Askwith et al. found that decreasing temperature slows ASIC desensitization with little effect on activation rate [66]. Because conformational changes are affected by temperature far more than channel block is, this supports the idea that channel activation is not a conformation change. Other studies have investigated gating mechanisms of ASICs using mutations or chemical reagents to reveal important residues. A study on ASIC2a found that when Gly430 is mutated to a bulkier amino acid, the proton sensitivity can increase more than two orders of magnitude [24]. Besides an increased proton sensitivity, the mutant channel also fails to desensitize. Gly430 is the same residue that causes the degenerin phenotype when mutated [16]. This residue in ASIC2a is also implicated in gating due to its MTS reactivity in open but not closed states [67]. It is accessible only when the channel is open, suggesting a conformational change exposing the residue. Clearly, this residue is very important for channel function, but its role in gating the channel is uncertain. ENaCs are not gated by ligands or voltage, but are constitutively active. They are regulated either through controlling surface expression or by modifying their open probability. ENaCs are also blocked by Ca++ [68], which can be removed with mechanical forces to open the channel [69].

3.7 Proposed Sensory Functions

3.7

Proposed Sensory Functions

The ability to sense pH makes ASICs candidates for transducing acid-evoked pain and taste sensations. Their relation to degenerins makes them appealing as possible mechanosensing ion channels. Expression patterns and channel properties may be consistent with these ideas, but definitive tests in whole animals are lacking. 3.7.1

Pain/Nociception

Krishtal and Pidoplichko found that 74 % of DRG neurons expressing an acid-sensitive current were smaller than 26 lm [3], demonstrating expression in neurons that make Ad and C axons that carry the bulk of nociceptive information [70]. Immunocytochemistry shows that ASIC1 co-expresses with vanilloid receptors, Substance P, and CGRP, all markers for nociceptors [71]. The distribution is consistent with ASICs playing a role in pain but does not mean this need be their only function. Experiments on humans have shown that amiloride diminishes the pain associated with injection of a pH 6.0 solution [72]. The VR1 antagonist capsazepine has no effect on the intensity of pain perceived at pH 6.0 but does reduce it at pH 5.0, although to a lesser extent than amiloride. This provides evidence for the role of ASICs in transducing pain from acidic stimuli. It suggests that VR1 plays little or no part in sensing acid within the physiological range. Inflammation causes transcript levels of ASICs to increase [73] through transcriptional control that involves serotoninergic and nerve growth factor pathways [74]. Nonsteroid anti-inflammatory drugs (NSAIDs) such as aspirin and ibuprofen suppress this increase. In addition to suppressing the increase in transcript levels, NSAIDs also directly inhibit the currents produced by ASICs in both DRG and ASIC3 or ASIC1a expressing COS cells. The sensitivity of ASIC expression to persistent pain conditions is further circumstantial evidence for a role in pain. Benson et al. fluorescently tagged sensory neurons that innervate heart muscle and found that they all expressed ASIC3-like current at exceedingly high levels [22, 75]. As a sensory system, the heart is unusual in that the only conscious sensation that arises from it is pain (angina) and the only trigger for this is ischemia, when the heart receives insufficient oxygen for its metabolic demand. Thus, the expression is consistent with ASIC3 being a sensor for ischemic pain by detecting lactic acid. Moreover, ASIC3 is clearly sensitive enough to detect the fairly small (1/2 pH unit) changes in extracellular acidity that occurs in ischemic muscle. ASIC3 also responds better to lactic acid than to other forms of acid [59], making it better for detecting the lactic acidosis that accompanies ischemia than the carbonic acidosis typical of most metabolic acidoses. Together these observations make a circumstantial case that ASIC3 is a sensor for angina and related ischemic pain. The idea is difficult to test in transgenic mice because there is no rodent model of ischemic pain. The clearest demonstration for a role of an ASIC in pain has come from Sluka, who created a persistent pain paradigm triggered by muscle acidity in mice. Two muscle

65

66

3 Acid-sensing Ion Channels

injections of acid some time apart create a long-lived hypersensitivity to touch of nearby skin [76]. This very reproducible syndrome simply does not occur in mice that lack ASIC3 [77]. The condition may be analogous to some forms of temporomandibular joint syndrome, in which muscle or joint problems cause chronic pain throughout the face. More traditional assays provide mixed messages. Two labs have made ASIC3 knockout mice that differ in their behavior. With i.p. injections of 0.6 % acetic acid, the Zimmer lab knockout shows an increase (not decrease) in writhing over wild type [78]. The Welsh lab knockout exhibits no change in paw licking after acid injections and a decrease in mechanical hyperalgesia after intramuscular acid injection [79]. With regard to noxious heat, the Zimmer knockout is more sensitive to temperatures above 50 8C in a hot-plate test [78]; the Welsh knockout shows a decreased response of heat sensitive C-fibers to noxious heat and normal paw withdrawal to radiant heat [79]. The Zimmer lab knockout differs from the wild type only in tests at high intensity, suggesting a role for ASIC3 in high-intensity pain in different modalities. The Welsh lab knockout demonstrates a role for ASIC3 in acid and carrageenan-evoked hyperalgesia. In both cases, neither knockout shows a severe phenotype, indicating that there are other players in all of these types of pain. It seems fair to summarize the data on knockout mice to indicate that they give subtle effects in pain assays and that different labs report different results. A crucial concern is that these experiments cannot address the hypothesis that ASICs are sensors for ischemic pain such as angina because there is no mouse model for such pain.

3.7.2

Mechanosensation

ASICs are considered as possible mechanosensing channels because of their homology to degenerins, which are expressed on mechanosensing C. elegans neurons (see Chapter 1). ASIC2a and ASIC3 are expressed on mechanosensory terminals in rat skin [79, 80]. ASIC2a knockouts show decreased sensitivity of rapidly adapting mechanoreceptors [81]. ASIC3 knockouts have altered mechanosensation; however, in this case the rapidly adapting mechanoreceptors become more sensitive [79]. Neither knockout has dramatically altered mechanosensation like the degenerin mutants. However, the degenerin mutation causes loss of the entire mechanosensing cell, not just a single molecule. How might mechanosensitivity arise? Is acid sensitivity relevant? Most models of mechanosensation surmise that the detecting channel is tethered to other proteins in the extracellular and intracellular media. The DEG proteins are thought to form ion channels in a multi-subunit complex necessary for transducing mechanical forces. Many proteins associate with DEGs to form this complex, but very few interacting proteins have been found that associate with ASICs. One protein that associates with ASIC3 is CIPP (channel-interacting PDZ domain protein). CIPP interacts with the carboxy terminus of ASIC3 using one of its four PDZ domains. When expressed with ASIC3 in COS cells, CIPP increases the current fivefold, and shifts

3.8 CNS ASICs

the activation curve towards more basic values [82]. CIPP may act as a scaffold protein to link ASIC3 to other intracellular proteins, since it has four PDZ domains.

3.7.3

Taste

Foods are perceived as sour due to an acid receptor in taste cells (see Chapter 7). Currents generated by acid in taste cells generally have a low amiloride and proton sensitivity. Recently it has been shown that both ASIC2a and -2b are present in a population of rat taste cells and that the heteromer is insensitive to amiloride at low pH [44, 83, 84]. ASIC2a/2b heteromers have a proton sensitivity similar to that seen for sour taste. Two other proton-activated channels have also been implicated in sour taste transduction: VR1 and HCN channels. They can both be activated by acid and are expressed in taste cells. It is unclear to what extent these channels participate in sour taste transduction or whether all are involved.

3.8

CNS ASICs

Immunolocalization studies show that ASIC1a is expressed in areas of high synaptic concentration such as cortex, olfactory bulb, hippocampus, amygdala, and cerebellum [85, 86]. This bolsters the hypothesis that ASICs have a synaptic function. ASICs may be activated in the CNS due to synaptic release. Vesicles become acidified when they are loaded with their neurotransmitter. When vesicles empty their transmitter, they also unload protons, thereby rapidly acidifying the synaptic cleft [87]. Alternatively, there may be some as yet undiscovered ligand of ASICs. ASICs in the CNS may enhance learning and memory formation. The Welsh group knocked out ASIC1 in mice and found deficits in LTP and learning in the knockouts [85, 88]. The mutant mice have decreased facilitation of EPSPs during high-frequency stimulation, deficits in spatial memory, impaired LTP, and impaired eye blink and fear conditioning. Even with all these defects, the mice appear normal and are capable of learning. Interestingly, all proton-gated currents are abolished in the CNS of knockout animals, suggesting that ASIC1a is necessary for proper expression of ASIC2a, the other common CNS ASIC. No role for ASICs in normal synaptic transmission has been demonstrated. Blockade of ASICs with amiloride or desensitization with low pH causes no evident effect on synaptic transmission [86]. Furthermore, the study did not find enrichment of ASIC1 in postsynaptic density (PSD) fractions of cell homogenates [86], unlike other synaptic proteins (e.g., NMDA receptors). However, ASIC2a is enriched in synaptic fractions from cerebellum [89].

67

68

3 Acid-sensing Ion Channels

3.9

Stroke

Since ASICs are proton-gated, they may have deleterious side effects in the CNS during pathological conditions such as stroke and seizure. During the ischemic conditions of a stroke or seizure, the pH can drop as low as 6.2 due to lactic acidosis [90], increasing extracellular lactate and proton concentrations. In addition to the pH change, the extracellular Ca++ concentration can drop from 1.2 mM to 0.1 mM [91], due to the translocation of Ca++ into cells. Synaptic release occurring during these pathologies can also release large amounts of Zn++ into the extracellular medium. All of these ionic conditions – increased H+, Zn++, and lactate and decreased Ca++ – are known potentiators of ASIC currents. If ASICs are activated strongly in the CNS, they could conceivably lead to neuronal death, much as constitutively active degenerins kill cells expressing them [12]. Hyperactive ASICs are capable of loading a cell with sodium, creating an osmotic stress and thereby killing the cell. ASIC1a has a significant Ca++ permeability, which could let excess Ca++ into the cell, thereby killing it via Ca++-activated apoptotic pathways. For either of these to occur, substantial activation would be necessary. The transient nature of ASIC currents may protect against such damage, but small sustained ASIC currents occur at many pH levels. Allen and Atwell subjected cerebellar purkinje cells to many different conditions occurring during CNS ischemia [92]. The ASICs in the cerebellar slices are activated when pH falls below 6.8 with a pH0.5 of 6.4. Many ischemic factors enhanced both the maximal and sustained ASIC current in these cells. Arachidonic acid, cell swelling, and lactate all potentiate the acid-gated current, often inducing a sustained component. In a rat model of global ischemia, Johnson et al. found that ASIC2a is upregulated in surviving cells [93]. Since the ASIC1a/2a heteromers are less sensitive to acid than ASIC1a, the increase in ASIC2a may act to suppress activity of ASIC1a homomers by increasing the proportion of ASIC1a in heteromers. Using a pilocarpine model of epilepsy, both ASIC1a and 2b are downregulated [94]. This may be a neuroprotective mechanism to prevent neuronal death.

3.10

Other pH-activated Channels

ASICs may be the most sensitive, but they are not the only channels activated by protons. The vanilloid receptor (VR1 or TRPV1), a member of the transient receptor potential (TRP) family of ion channels, is opened by many types of stimuli, including heat, protons, and capsaicin [95]. Protons activate VR1 receptors by lowering their temperature threshold. The pH needs to drop below 5.9 at room temperature to activate the channel. The current through VR1 receptors is easily distinguished from ASICs since it does not desensitize and is a nonselective cation channel. VR1 is thought to transduce noxious heat and capsaicin sensitivity in vivo [96] (see Chapters 10 and 11). Hyperpolarization-activated and cyclic nucleotide-gated channels (HCNs)

3.10 Other pH-activated Channels

are also activated by protons. These cation channels are normally opened by hyperpolarization; however, both cyclic nucleotides and extracellular protons shift the activation curve to more positive voltages [97, 98]. At very low pH values, this shifts to above the resting potential of the cell, thereby activating the channel. As sensory transduction channels, HCN1 and HCN4 are thought to participate in sour taste transduction. There is also a family of potassium channels that shows proton gating. The TASKs (TWIK-related acid-sensitive K+ channels) are members of the tandem pore potassium channel family. Instead of being activated by low pH, TASK1 and TASK3 are inactivated by acid. The IC50 for protons of TASK1 and TASK3 are pH 7.3 and 6.3, respectively, putting them within the physiological range of pH changes, especially TASK1 [99]. Low pH excites cells expressing these channels by eliminating a resting potassium conductance.

References 1

2

3

4

5

6

7

8

9

10

Krishtal, O.A. and V.I. Pidoplichko, A receptor for protons in the nerve cell membrane. Neuroscience, 1980. 5(12): p. 2325–7. Krishtal, O.A. and V.I. Pidoplichko, Receptor for protons in the membrane of sensory neurons. Brain Research, 1981. 214: p. 150–154. Krishtal, O.A. and V.I. Pidoplichko, A “receptor” for protons in small neurons of trigeminal ganglia: possible role in nociception. Neuroscience Letters, 1981. 24: p. 243–246. Korkushko, A.O. and O.A. Krishtal, Blocking of proton-activated sodium permeability of the membranes of trigeminal ganglion neurons in the rat by organic cations. Neirofiziologiia, 1984. 16(4): p. 557–561. Morad, M., Proton-induced transformation in gating and selectivity of the calcium channel in neurons. Proton passage across cell membranes, 1988: p. 187–200. Waldmann, R., et al., A proton-gated cation channel involved in acid-sensing. Nature, 1997. 386(6621): p. 173–7. Krishtal, O., The ASICs: Signaling molecules? Modulators? Trends in Neurosciences, 2003. 26(9): p. 477–483. Kellenberger, S. and L. Schild, Epithelial sodium channel/degenerin family of ion channels: a variety of functions for a shared structure. Physiological Reviews, 2002. 82: p. 735–767. de la Rosa, D.A., et al., Structure and Regulation of Amiloride-Sensitive Sodium Channels. Annual Review of Physiology, 2000. 62(1): p. 573–594. Lingueglia, E., et al., Cloning of the amiloridesensitive FMRFamide peptide-gated sodium channel. Nature, 1995. 378: p. 730–733.

11

12

13

14

15

16

17

18

Benos, D.J. and B.A. Stanton, Functional domains within the degenerin/epithelial sodium channel (Deg/ENaC) superfamily of ion channels. J Physiol (Lond), 1999. 520(3): p. 631–644. Chalfie, M. and E. Wolinsky, The identification and suppression of inherited neurodegeneration in Caenorhabditis elegans. Nature, 1990. 345(6274): p. 410–6. Hall, D.H., et al., Neuropathology of degenerative cell death in Caenorhabditis elegans. J Neurosci, 1997. 17(3): p. 1033–45. Garcia-Anoveros, J., et al., The nematode degenerin UNC-105 forms ion channels that are activated by degeneration- or hypercontractioncausing mutations. Neuron, 1998. 20(6): p. 1231–41. Hong, K. and M. Driscoll, A transmembrane domain of the putative channel subunit MEC-4 influences mechanotransduction and neurodegeneration in C. elegans. Nature, 1994. 367(6462): p. 470–473. Waldmann, R., et al., The mammalian degenerin MDEG, an amiloride-sensitive cation channel activated by mutations causing neurodegeneration in Caenorhabditis elegans. J Biol Chem, 1996. 271(18): p. 10433–6. Price, M.P., P.M. Snyder, and M.J. Welsh, Cloning and expression of a novel human brain Na+ channel. J Biol Chem, 1996. 271(14): p. 7879–82. Garcia-Anoveros, J., et al., BNaC1 and BNaC2 constitute a new family of human neuronal sodium channels related to degenerins and epithelial sodium channels. Proc Natl Acad Sci U S A, 1997. 94(4): p. 1459–64.

69

70

3 Acid-sensing Ion Channels 19

20

21

22

23

24

25

26

27

28

29

30

31

Babini, E., et al., Alternative splicing and interaction with di- and polyvalent cations control the dynamic range of acid-sensing ion channel (ASIC) 1. J. Biol. Chem., 2002: p. M205877200. Alvarez de la Rosa, D., et al., Functional implications of the localization and activity of acidsensitive channels in rat peripheral nervous system. Proc Natl Acad Sci U S A, 2002. 99(4): p. 2326–31. Waldmann, R., et al., Molecular cloning of a non-inactivating proton-gated Na+ channel specific for sensory neurons. J Biol Chem, 1997. 272(34): p. 20975–8. Sutherland, S.P., et al., Acid-sensing ion channel 3 matches the acid-gated current in cardiac ischemia-sensing neurons. Proc Natl Acad Sci U S A, 2001. 98(2): p. 711–6. Lingueglia, E., et al., A modulatory subunit of acid sensing ion channels in brain and dorsal root ganglion cells. J Biol Chem, 1997. 272(47): p. 29778–83. Champigny, G., et al., Mutations Causing Neurodegeneration in Caenorhabditis elegans Drastically Alter the pH Sensitivity and Inactivation of the Mammalian H+-gated Na+ Channel MDEG1. J. Biol. Chem., 1998. 273(25): p. 15418–15422. Chen, C.-C., et al., A sensory neuron-specific, proton-gated ion channel. PNAS, 1998. 95(17): p. 10240–10245. Bassler, E.L., et al., Molecular and functional characterization of acid-sensing ion channel (ASIC) 1b. J Biol Chem, 2001. 276(36): p. 33782–7. Ugawa, S., et al., Cloning and functional expression of ASIC-b2, a splice variant of ASIC- b. Neuroreport, 2001. 12: p. 2865–2869. Grunder, S., et al., A new member of acid-sensing ion channels from pituitary gland. Neuroreport, 2000. 11(8): p. 1607–11. Akopian, A., et al., A new member of the acidsensing ion channel family. Neuroreport, 2000. 11(10): p. 2217–2222. Babinski, K., K.-T. Le, and P. Seguela, Molecular Cloning and Regional Distribution of a Human Proton Receptor Subunit with Biphasic Functional Properties. J Neurochem, 1999. 72(1): p. 51–57. Firsov, D., et al., Mutational Analysis of Cysteine-rich Domains of the Epithelium Sodium Channel (ENaC). IDENTIFICATION OF CYSTEINES ESSENTIAL FOR CHANNEL EXPRESSION AT THE CELL SURFACE. J. Biol. Chem., 1999. 274(5): p. 2743–2749.

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

North, R.A., Families of ion channels with two hydrophobic segments. Curr Opin Cell Biol, 1996. 8(4): p. 474–83. Doyle, D.A., et al., The Structure of the Potassium Channel: Molecular Basis of K+ Conduction and Selectivity. Science, 1998. 280(5360): p. 69–77. Sheng, S., et al., Characterization of the Selectivity Filter of the Epithelial Sodium Channel. J. Biol. Chem., 2000. 275(12): p. 8572–8581. Sheng, S., et al., Epithelial Sodium Channel Pore Region. STRUCTURE AND ROLE IN GATING. J. Biol. Chem., 2001. 276(2): p. 1326–1334. Coscoy, S., et al., The Pre-transmembrane 1 Domain of Acid-sensing Ion Channels Participates in the Ion Pore. J. Biol. Chem., 1999. 274(15): p. 10129–10132. Snyder, P.M., et al., Electrophysiological and biochemical evidence that DEG/ENaC cation channels are composed of nine subunits. J Biol Chem, 1998. 273(2): p. 681–4. Coscoy, S., et al., The Phe-Met-Arg-Phe-amideactivated Sodium Channel Is a Tetramer. J. Biol. Chem., 1998. 273(14): p. 8317–8322. Firsov, D., et al., The heterotetrameric architecture of the epithelial sodium channel (ENaC). EMBO J., 1998. 17(2): p. 344–352. Eskandari, S., et al., Number of Subunits Comprising the Epithelial Sodium Channel. J. Biol. Chem., 1999. 274(38): p. 27281–27286. Canessa, C.M., et al., Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature, 1994. 367: p. 463–467. Bassilana, F., et al., The acid-sensitive ionic channel subunit ASIC and the mammalian degenerin MDEG form a heteromultimeric H+gated Na+ channel with novel properties. J Biol Chem, 1997. 272(46): p. 28819–22. Baron, A., R. Waldmann, and M. Lazdunski, ASIC-like, proton-activated currents in rat hippocampal neurons. J Physiol, 2002. 539(Pt 2): p. 485–94. Ugawa, S., et al., Amiloride-Insensitive Currents of the Acid-Sensing Ion Channel-2a (ASIC2a) / ASIC2b Heteromeric Sour-Taste Receptor Channel. J. Neurosci., 2003. 23(9): p. 3616–3622. Babinski, K., et al., Mammalian ASIC2a and ASIC3 Subunits Co-assemble into Heteromeric Proton-gated Channels Sensitive to Gd3+. J. Biol. Chem., 2000. 275(37): p. 28519–28525. Benson, C.J., et al., Heteromultimers of DEG/ ENaC subunits form H+-gated channels in mouse sensory neurons. PNAS, 2002. 99(4): p. 2338–2343.

3.10 Other pH-activated Channels 47

48

49

50

51

52

53

54

55

56

57

58

59

Adams, C.M., P.M. Snyder, and M.J. Welsh, Paradoxical stimulation of a DEG/ENaC channel by amiloride. J Biol Chem, 1999. 274(22): p. 15500–4. McNicholas, C.M. and C.M. Canessa, Diversity of Channels Generated by Different Combinations of Epithelial Sodium Channel Subunits. J. Gen. Physiol., 1997. 109(6): p. 681–692. Garcia, M., et al., Amiloride analogs inhibit Ltype calcium channels and display calcium entry blocker activity. J. Biol. Chem., 1990. 265(7): p. 3763–3771. Tang, C.M., F. Presser, and M. Morad, Amiloride selectively blocks the low threshold (T) calcium channel. Science, 1988. 240(4849): p. 213–215. Murata, Y., et al., Non-selective effects of amiloride and its analogues on ion transport systems and their cytotoxicities in cardiac myocytes. Japanese Journal of Pharmacology, 1995. 68(3): p. 279–85. Ismailov, I.I., et al., Identification of an Amiloride Binding Domain within the alpha -Subunit of the Epithelial Na+ Channel. J. Biol. Chem., 1997. 272(34): p. 21075–21083. Schild, L., et al., Identification of Amino Acid Residues in the alpha , beta , and gamma Subunits of the Epithelial Sodium Channel (ENaC) Involved in Amiloride Block and Ion Permeation. J. Gen. Physiol., 1997. 109(1): p. 15–26. Escoubas, P., et al., Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels. J Biol Chem, 2000. 275(33): p. 25116–21. Escoubas, P., et al., Recombinant production and solution structure of PcTx1, the specific peptide inhibitor of ASIC1a proton-gated cation channels. Protein Sci, 2003. 12(7): p. 1332–1343. Baron, A., et al., Zn2+ and H+ are coactivators of acid-sensing ion channels. J Biol Chem, 2001. 276(38): p. 35361–7. de Weille, J. and F. Bassilana, Dependence of the acid-sensitive ion channel, ASIC1a, on extracellular Ca(2+) ions. Brain Res, 2001. 900(2): p. 277–81. Korkushco, A.O., O.A. Krishtal, and N.I. Chernevskaya, Steady-state characteristics of the proton receptor in the somatic membrane of rat sensory neurons. Neirofiziologiia, 1983. 15: p. 632–638. Immke, D.C. and E.W. McCleskey, Lactate enhances the acid-sensing Na+ channel on ischemia-sensing neurons. Nat Neurosci, 2001. 4(9): p. 869-70.

60

61

62

63

64

65

66

67

68

69

70

71

72

Roumy, M. and J.-M. Zajac, Neuropeptide FF, pain and analgesia. European Journal of Pharmacology, 1998. 345(1): p. 1–11. Askwith, C.C., et al., Neuropeptide FF and FMRFamide potentiate acid-evoked currents from sensory neurons and proton-gated DEG/ ENaC channels. Neuron, 2000. 26(1): p. 133–41. Catarsi, S., K. Babinski, and P. Seguela, Selective modulation of heteromeric ASIC protongated channels by neuropeptide FF. Neuropharmacology, 2001. 41(5): p. 592–600. Deval, E., et al., Effects of neuropeptide SF and related peptides on acid sensing ion channel 3 and sensory neuron excitability. Neuropharmacology, 2003. 44(5): p. 662–671. Xie, J., et al., ASIC3 and ASIC1 mediate FMRFamide-related peptide enhancement of H+-gated currents in cultured dorsal root ganglion neurons. J Neurophysiol, 2003: p. 00707.2002. Immke, D.C. and E.W. McCleskey, Protons open Acid-sensing ion channels by catalyzing relief of ca(2+) blockade. Neuron, 2003. 37(1): p. 75–84. Askwith, C.C., et al., DEG/ENaC ion channels involved in sensory transduction are modulated by cold temperature. PNAS, 2001. 98(11): p. 6459–6463. Adams, C.M., et al., Protons activate brain Na+ channel 1 by inducing a conformational change that exposes a residue associated with neurodegeneration. J Biol Chem, 1998. 273(46): p. 30204–7. Berdiev, B.K., et al., Actin modifies Ca2+ block of epithelial Na+ channels in planar lipid bilayers. Biophys J, 2001. 80(5): p. 2176–86. Ismailov, II, et al., Mechanosensitivity of an epithelial Na+ channel in planar lipid bilayers: release from Ca2+ block. Biophys J, 1997. 72(3): p. 1182–92. Waddell, P.J. and S.N. Lawson, Electrophysiological properties of subpopulations of rat dorsal root ganglion neurons in vitro. Neuroscience, 1990. 36(3): p. 811–822. Olson, T.H., et al., An acid sensing ion channel (ASIC) localizes to small primary afferent neurons in rats. Neuroreport, 1998. 9(6): p. 1109–1113. Ugawa, S., et al., Amiloride-blockable acid-sensing ion channels are leading acid sensors expressed in human nociceptors. J. Clin. Invest., 2002. 110(8): p. 1185–1190.

71

72

3 Acid-sensing Ion Channels 73

74

75

76

77

78

79

80

81

82

83 84

85

Voilley, N., et al., Nonsteroid anti-inflammatory drugs inhibit both the activity and the inflammation-induced expression of acid-sensing ion channels in nociceptors. J Neurosci, 2001. 21(20): p. 8026–33. Mamet, J., et al., ProInflammatory Mediators, Stimulators of Sensory Neuron Excitability via the Expression of Acid-Sensing Ion Channels. J Neurosci, 2002. 22(24): p. 10662–70. Benson, C.J., S.P. Eckert, and E.W. McCleskey, Acid-Evoked Currents in Cardiac Sensory Neurons : A Possible Mediator of Myocardial Ischemic Sensation. Circ Res, 1999. 84(8): p. 921–928. Sluka, K.A., A. Kalra, and S.A. Moore, Unilateral intramuscular injections of acidic saline produce a bilateral, long-lasting hyperalgesia. Muscle Nerve, 2001. 24(1): p. 37–46. Sluka, K.A., et al., Chronic hyperalgesia induced by repeated acid injections in muscle is abolished by the loss of ASIC3 but not ASIC1. Pain, . In Press. Chen, C.C., et al., A role for ASIC3 in the modulation of high-intensity pain stimuli. Proc Natl Acad Sci U S A, 2002. 99(13): p. 8992–7. Price, M.P., et al., The DRASIC cation channel contributes to the detection of cutaneous touch and acid stimuli in mice. Neuron, 2001. 32(6): p. 1071–83. Garcia-Anoveros, J., et al., Transport and localization of the DEG/ENaC ion channel BNaC1alpha to peripheral mechanosensory terminals of dorsal root ganglia neurons. J Neurosci, 2001. 21(8): p. 2678–86. Price, M.P., et al., The mammalian sodium channel BNC1 is required for normal touch sensation. Nature, 2000. 407(6807): p. 1007–11. Anzai, N., et al., The multivalent PDZ domaincontaining protein CIPP is a partner of acidsensing ion channel 3 in sensory neurons. J Biol Chem, 2002. 277(19): p. 16655–61. Ugawa, S., et al., Receptor that leaves a sour taste in the mouth. Nature, 1998. 395: p. 555–556. Lin, W., T. Ogura, and S.C. Kinnamon, AcidActivated Cation Currents in Rat Vallate Taste Receptor Cells. J Neurophysiol, 2002. 88(1): p. 133–141. Wemmie, J.A., et al., Acid-Sensing Ion Channel 1 Is Localized in Brain Regions with High Synaptic Density and Contributes to Fear Conditioning. J. Neurosci., 2003. 23(13): p. 5496–5502.

86

87

88

89

90

91

92

93

94

95

96

97

98

99

de la Rosa, D.A., et al., Distribution, subcellular localization and ontogeny of ASIC1 in the mammalian central nervous system. J Physiol (Lond), 2003. 546(1): p. 77–87. Krishtal, O.A., et al., Rapid extracellular pH transients related to synaptic transmission in rat hippocampal slices. Brain Research, 1987. 436(2): p. 352–356. Wemmie, J.A., et al., The acid-activated ion channel ASIC contributes to synaptic plasticity, learning, and memory. Neuron, 2002. 34(3): p. 463–77. Jovov, B., et al., Immunolocalization of the acidsensing ion channel 2a in the rat cerebellum. Histochemistry and Cell Biology, 2003. 119(6): p. 437–46. Nedergaard, M., et al., Dynamics of interstitial and intracellular pH in evolving brain infarct. Am J Physiol, 1991. 260(3 Pt 2): p. R581–8. Kristian, T., et al., Calcium metabolism of focal and penumbral tissues in rats subjected to transient middle cerebral artery occlusion. Exp Brain Res, 1998. 120(4): p. 503–9. Allen, N.J. and D. Attwell, Modulation of ASIC channels in rat cerebellar Purkinje neurons by ischaemia-related signals. J Physiol (Lond), 2002. 543(2): p. 521–529. Johnson, M.B., et al., Global ischemia induces expression of acid-sensing ion channel 2a in rat brain. Journal of Cerebral Blood Flow and Metabolism, 2001. 21: p. 734–740. Biagini, G., et al., Regional and Subunit-Specific Downregulation of Acid-Sensing Ion Channels in the Pilocarpine Model of Epilepsy. Neurobiology of Disease, 2001. 8(1): p. 45–58. Tominaga, M., et al., The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron, 1998. 21(3): p. 531–43. Caterina, M.J., et al., Impaired Nociception and Pain Sensation in Mice Lacking the Capsaicin Receptor. Science, 2000. 288(5464): p. 306–313. Chen, S., J. Wang, and S.A. Siegelbaum, Properties of Hyperpolarization-Activated Pacemaker Current Defined by Coassembly of HCN1 and HCN2 Subunits and Basal Modulation by Cyclic Nucleotide. J. Gen. Physiol., 2001. 117(5): p. 491–504. Stevens, D.R., et al., Hyperpolarization-activated channels HCN1 and HCN4 mediate responses to sour stimuli. Nature, 2001. 413(6856): p. 631–635. Lesage, F., Pharmacology of neuronal background potassium channels. Neuropharmacology, 2003. 44(1): p. 1–7.

73

4

Chemosensory Transduction in Caenorhabditis elegans Noelle L’Etoile

4.1

Introduction 4.1.1

The organism C. elegans

In 1963 Sydney Brenner decided to “tame” the worm Caenorhabditis elegans. He envisioned the worm, a small (1.5 mm long), transparent, free-living soil nematode, to be the metazoan equivalent of the bacteriophage, the perfect beast with which to “dissect the genetic specification of the nervous system.” His stated goal was to understand, in molecular detail, how a nervous system is designed to generate behavior [1]. It is not hard to see why he chose C. elegans: this small beast is an appealing genetic organism; hundreds of animals can inhabit one petri dish, dining on standard laboratory strains of E. coli. With a generation time of only three days, genetic manipulations are just a bit slower than in yeast and large populations can be grown in a matter of days. Since it is a hermaphrodite under standard cultivation conditions, the populations can be essentially cloned and recessive mutations are easily uncovered. The hermaphrodite can also be mated into by males, which allows for mapping of mutations. Males are easily produced by high temperatures or ethanol exposure, and they are maintained by mating. But the key quality that lured Sydney Brenner to the worm, Caeno (recent) rhabditis(rod) elegans (nice), was its photogenic aspect: high-quality electron micrographs could be obtained and used to reconstruct its entire nervous system. In 1986 the hermaphrodite’s 302-neuron nervous system was reconstructed by White and Sulston using serial electron micrographs of two worms [2]. Each chemical synapse and gap junction was noted and the communicating cells were identified. When the worms were compared, relatively little variation of either cell body position or synapse number was observed. In this way, a potential circuit diagram of the worm’s nervous system was described. To create a “functional” circuit diagram, two criteria had to be met: first, an assay for neuronal function had to be established, and second, the contribution of each neuron to the behavior had to be assessed. To this end, each sensory neuron and many of the Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

74

4 Chemosensory Transduction in Caenorhabditis elegans

interneurons were individually ablated by laser. The laser-operated animals were subjected to behavioral studies and their deficits noted. This led to a rough description of some sensory circuits [3]. The behavioral repertoire of such a small nervous system might have proven too restricted to be interesting. The invariance of cell body position and cell number also gives rise to the speculation that C. elegans behavior might prove to be completely “hard-wired,” leaving little room for plasticity. Though C. elegans has neither eyes nor ears, the range of behaviors generated by this simple nervous system is proving to be quite astonishing. The number of “new” behaviors grows as the number of researchers examining the link between genetics and behavior grows. Behavioral responses to external stimuli such as temperature, volatile chemicals, water-soluble chemicals, pheromones, light and harsh touch, and bacteria, which are ill-defined components of other worms, can all be observed reproducibly in the laboratory. The behaviors that arise in response to simple stimuli can be modulated, often by the pairing of food or starvation with a specific second stimulus, to give rise to complex behaviors. For example, “social feeding” or “swarming” arises from the pairing of food with other worms [4]. Pairing of a specific temperature with food results in a preference for or “memory” of the food-associated temperature and aversion to a temperature that is paired with starvation [5–7]. Pairing of starvation with a salt or an odor will cause aversion to the paired compound [8, 9]. A rather bizarre behavior is induced by crowding and starvation: nictation. Nictating worms writhe about each other and form long strands, perhaps in an attempt to catch a ride to greener pastures on a passing insect. Nothing is known about genes that regulate or produce this behavior. Though not a behavior, the decision to go into the alternate, dauer larval stage is a sensory-driven decision that relies on some of the same molecules as the afore-mentioned behaviors. The larval C. elegans can, if starved, crowded, and placed at high enough temperatures, decide to enter the dauer larval state. In this state, it becomes desiccation resistant by sealing off its buccal cavity and growing a thick cuticle. Its metabolism drops, it stores fat, moves little, and pumps not at all. Though the lifespan of a worm in good conditions is a little less than three weeks (20–28 days), as dauers, worms can live for more than six months [3]. Perhaps in an effort to maximize its ability to sense favorable conditions, the dauer’s olfactory sensory cilia expand to twice the size of the equivalent non-dauer larva’s cilia [10]. Once conditions become more hospitable, the dauer will shed its cuticle, resume gonad development, and become a fertile adult. The sensory signal generated by some of these disparate chemical stimuli will be the subject of this chapter. The lens through which we will view these behaviors is the channel that is most directly stimulated by sensory stimulus. The primary signal transduction channels expressed in C. elegans sensory neurons are of only two classes, the tax channels and the osm-related channels.

4.2 The Chemosensory Organs

4.1.2

Introduction to the Channels

Though many behavioral screens have yielded mutations in the CNG TAX channel, the tax channels’ discovery was initiated by the genetic screens for salt chemotaxisdefective mutants carried out in Dr. Richard Russel’s lab [11]. The screens, designed to probe the molecular basis for chemosensation, identified two mutants, tax-2 and tax-4. These were later mapped and cloned by the Bargmann and Ohshima labs and identified as the beta and alpha subunits of the C. elegans cyclic nucleotide-gated channel [12, 13]. The sequence of the C. elegans genome contains four additional subunits that have yet to be characterized. Another primary sensory transduction channel was identified as a result of screens for osmotic (osm) avoidance-defective mutants (Thomas lab), light nose-touch avoidance-defective mutants (Kaplan lab), and volatile chemotaxis-defective mutants (Bargmann lab). This channel consists of the transient receptor potential channel type V (TRPV; see Tab. 8.3) subunit OSM-9 [14] in complex with at least one other TRPV subunit, OCR-2 (OSM-9 capsaicin receptor related). The OCR-1, -2, -3, and -4 subunits were identified by sequence homology to OSM-9 [15–18], and null mutants of ocr-1 and ocr-2 were generated by imprecise excision of a TC1 transposon [16]. This chapter will describe (1) the organs the worm uses to sense chemical and thermal stimuli; (2) the assays employed to monitor and quantify the sensory response; (3) the cells and molecules responsible for generating the behavioral response; and (4) the channels’ structure and how their activity is modulated. A subject that will be touched on only briefly is the role of the sensory channel in development.

4.2

The Chemosensory Organs

Though a terrestrial nematode, C. elegans lives at the air-water interface and senses its chemical environment by taste and smell. C. elegans has three organs devoted to sensing its external chemical environment: the amphid and inner labial organs in the anterior, and the phasmid organ in the posterior. The amphids and phasmids are bilaterally symmetric organs that lie on the left and right sides of the worm, while there are six radially symmetric inner labials. These “organs” are defined by the presence of an opening to the exterior of the animal that is supported by structural epithelial cells, such as the socket cell, that hold the exposed sensory cilia in place [19, 20].

4.2.1

The Amphid Organ

The amphid (anterior) organ contains 12 pairs of sensory neurons, eight of which are directly exposed to the environment and four of which are embedded within a glial-like sheath cell (see Fig. 4.1). The exposed amphid neurons (ADF, ADL, ASE, ASG, ASH,

75

76

4 Chemosensory Transduction in Caenorhabditis elegans

Fig. 4.1 The sensory organs of C. elegans. Top: schematic of the worm with amphid, inner labial, and phasmid organs labeled. Bottom left: cartoon representation of the sensory endings of the amphid neurons. ASE, ASG, ASH, ASI, ASK, ASJ, AFD,

and ADL cilia are bounded by the socket cell and exposed to the environment (Adapted from [75]). Structure of the sensory endings of these amphid neurons. Adapted from [21]

ASI, ASJ, and ASK) mostly sense micromolar concentrations of water-soluble compounds, while the sheathed neurons (AWA, AWB, and AWC) sense volatile compounds in the picomolar range while AFD senses heat [21, 22]. Using these cells, the worm responds to both attractive and repulsive stimuli. Some amino acids, cyclic nucleotides, basic pH, salts, and a host of volatile compounds (at low concentration) are actively sought out by the worm, while D-tryptophan, acid pH, high osmotic strength, heavy metals, detergents, and high concentrations of volatiles are actively avoided. Although a single amphid sensory neuron may be able to sense a number of compounds, the behavioral response generated by that neuron will lead to either attraction or repulsion, never both [23]. Cells within the amphid organ can be grouped into two categories: those that probably use the cyclic nucleotide-gated channel comprised of the alpha and beta subunits TAX-4 and TAX-2 as their primary signal transduction channel [12, 13], and those that are predicted to use both the OSM-9 and OCR-2 subunits of the transient receptor potential (TRP) channel [16] (see Tab. 4.1). ASH and AWA express both TRP channel subunits, and mutation of either subunit abolishes both ASH-mediated repulsion and AWA-mediated chemotaxis [14]. AWC, AFD, ASE, and ASK all express TAX- 2 and TAX-4, and responses generated by these

4.2 The Chemosensory Organs Tab. 4.1

C. elegans chemosensory organs, neurons, stimuli, and channels

Organ

Neuron

Class of stimulus sensed

Channel used

Channels expressed

Amphid

AWA

Attractive-volatile

TRP

AWC

Attractive-volatile

TAX

AWB ADL

Repulsive-volatile Repulsive-volatile, water-soluble Repulsive osmolarity, nose-touch, water-soluble Attractive salts

TAX ?

OSM-9, OCR-1, OCR-2 TAX-2, TAX-4, OSM-9 TAX-2, TAX-4 OSM-9, OCR-1, OCR-2 OSM-9, OCR-2

TAX

AFD PHA

Repulsive water-soluble/ Attractive amino acid Dauer pheromone/ Attractive water-soluble (minor) Dauer pheromone (minor)/ Attractive water-soluble (minor) Dauer pheromone/ Attractive water-soluble (minor) Dauer recovery promoting factors, e.g., food Temperature Repulsive water-soluble

TAX ?

PHB

Repulsive water-soluble

?

IL1 and IL2 (six each) URX/ BAG

IL1 may be mechanosensory IL2 may be chemosensory Internal cues that promote “social” feeding

?

TAX-2, TAX-4, OSM-9 TAX-2, TAX-4, OSM-9 TAX-2, TAX-4, OSM-9 TAX -2, TAX-4 TAX-2(?), TAX-4, OSM-9(?) TAX-2(?), OCR-2 TAX-4, OSM-9(?) OSM-9, OCR-2

TAX

TAX-2, TAX-4

ASH ASE ASK ADF ASG ASI ASJ

Phasmid

Inner labials Cilia exposed to worm’s interior

TRP TAX

? ? ? TAX

TAX-2, TAX-4, OSM-9 TAX-2, TAX-4, OSM-9 OSM-9, OCR-2

cells are defective in tax-2 and tax-4 mutants [12, 13]. AWC and ASE express the CNG channel subunits along with one TRP channel subunit, OSM-9. osm-9 mutants chemotax well to AWC and ASE-sensed compounds; therefore, OSM-9 is not the primary sensory signaling channel [8, 24, 25]. In fact, though hour-long exposures to these compounds will adapt the wild-type worm’s responses, osm-9 mutants never adapt; their response to AWC and ASE-sensed compounds is as robust after exposure as before [8, 25]. It is important to note that these adaptation defects appear after prolonged exposure and do not interfere with the rapid desensitization that is required for normal sensation of a gradient. Thus, expression of both channel subunits predicts which channel is used as the primary signal transduction channel [16], and OSM-9 would seem to promote sensory adaptation when it is expressed by itself in the context of TAX-2 and TAX-4 expression. With a very limited set of at most 32 sensory neurons, C. elegans is able to respond to a wide array of stimuli. The strategy by which the nematode diversifies its ability to

77

78

4 Chemosensory Transduction in Caenorhabditis elegans

respond to its environment without expanding its array of sensory neurons is to have multiple receptors on each neuron. Interestingly, there seems to be bit of redundancy: at least two pairs of neurons respond to most classes of stimuli. In general one pair will express the TAX CNG channel and the other will express the TRP channel (Tab. 4.1). Two pairs of cells sense volatile compounds and cause a chemotaxis response. One pair, the AWCs, uses TAX-4/2 [12, 13], and the other, the AWAs, uses the TRP channels [24]. Three pairs detect repulsive volatiles: the AWB pair uses TAX channels [26], while ASH [14] and ADL [14] use the TRP channel [16]. Though ASH, ADL, and ASK each detect water-soluble repellents, ASH and ADL express the TRP channel, while ASK uses the TAX channel [27]. Four pairs of neurons seem to be important for regulating the dauer response: ASG, ASI, ASJ, and ADF. Of these, ADF expresses the TRP channel and the others use the TAX channel [14]. What the worm achieves by this division of labor is unclear. Independent modulation of cells that utilize either class of channel may allow for a greater regulation of behavior by either developmental or environmental influences.

4.2.2

Phasmid Organ

The phasmid organ contains two sensory neurons, PHA and PHB (Fig. 4.1). Until recently, little was known about the function of the phasmid organ. Using a new assay, the “dry drop test,” Hilliard et al. [28] examined the phasmids’ function. In this assay, a repellant is allowed to soak into the agar immediately in front of a moving worm. The worm runs into the spot containing a high concentration of the repellant and backs away from it. If an additional spot of repellant is placed behind the worm, it will back into the second spot and move forward. Thus, the “dry drop test” separates amphid from phasmid responses. The duration of the backing response is 3 s when a repellent compound is encountered solely by the head of the worm. The second dry drop encountered by the tail decreases the duration of the backing response to 2 s. The decrease indicates that neurons in the tail promote the forward movement of the worm. This makes sense if the phasmids respond to repellant stimuli by making the worm move away from an encounter with a noxious compound that is localized at the rear of the worm. The contribution of the phasmids to this behavior was uncovered first by weakening the repellent-induced backing response. This was accomplished by laser-ablating a subset of the amphid neurons responsible for sensing repellents. This allowed Hilliard et al. [28] to observe a weakened backing response that was dependent upon an intact phasmid support cell. This was mimicked by a TAX-4 mutation. The phasmids are interesting in that they express both TRP subunits (OCR-2 and OSM-9) [16] and require both TAX channel subunits for proper axon outgrowth [12].

4.3 Behavioral Assays

4.2.3

Inner Labial

The six inner labials each contain two types of neurons, IL1 and IL2. Not much is known about the function of the inner labials. They have sensory cilia that are exposed to the environment. It is hard to predict which channels these neurons use because they seem to express only the OSM-9 subunit of the TRP channel [16, 24].

4.2.4

The Sensory Signaling Circuit

Sensory neurons relay their information to each other via chemical synapses and gap junctions. They also synapse onto interneurons that, in turn, integrate and relay sensory information to the head motor neurons. It is the head motor neurons that allow the animal to swim in response to a sensory cue [2]. The disparity between the resolution with which we can describe the physical circuit and the resolution with which we describe the actual flow of information is appalling. The traditional approaches of laser-ablation studies may need to be combined with cellspecific reporters of neuronal function such as the cameleon GFP pioneered by the Schafer lab [29] or the pericam GFP [30] used by the Axel lab to study olfactory processing in Drosophila in order to achieve an understanding of the path that information takes through the worm’s “brain.” Perhaps, too, the Mariqc lab’s use of electrophysiology [31] will allow a description of the circuits that generate chemosensory behavior. Another approach is to reconstitute the circuit to determine the minimal number and type of connection that would allow the worm to perform a specific behavior. This could be accomplished in the glutamate transporter minus, mutant eat-4, background [32] by placing the cDNA for the transporter, EAT-4, downstream of promoters specific for particular sensory neurons. For example, to reconstitute the olfactory chemosensory circuit, AWCs, specific interneurons and specific head motor neurons may all have to express the glutamate transporter to enable the eat-4 null worm to chemotax toward an AWC-sensed odor. The minimal number would give a clue as to the true working circuit diagram.

4.3

Behavioral Assays

In order to create a circuit diagram and to understand the molecular underpinnings of behavior, robust behavioral assays had to be developed. Initial investigations into C. elegans chemosensory behaviors focused on chemotaxis towards food or specific chemicals and thermotaxis towards its preferred temperature [5, 11, 21]. The following descriptions of the assays used to probe the role of the TAX and TRP channels in generating a behavioral response are meant to allow the reader both to appreciate

79

80

4 Chemosensory Transduction in Caenorhabditis elegans

the limitations of these assays and to evaluate some of the data presented in the subsequent sections.

4.3.1

Chemotaxis

To assess chemotaxis of a single worm, a slightly dry square plate is used. The worm is placed onto the center of the plate, the diluted compound of interest is placed at one end, and the diluent is placed at the other end [33, 34] (Fig. 4.2, top left). The worm is allowed to roam the plate for one hour. A grid is drawn on the plate and numbered such that the center, vertical line has the value 0 and the next flanking lines are 2 in the direction of the odor and –2 in the direction of the diluent. The tracks left by the worm as it navigated the chemical gradient are noted, and each time a track crosses a line, the value of the line is added. Thus, a journey that takes the worm straight to the chemical would accrue a value of 6 and one that takes it directly away from the odor –6. In the absence of a chemical cue, the value is generally 0. Such assays are generally carried out on laser-operated individuals.

Fig. 4.2 Behavioral assays for sensory signaling. Top left: single-worm attraction or repulsion assay. Top right: population-attraction assay. Bottom right: population-repulsion assay. Bottom right: thermosensation assay

4.3 Behavioral Assays

Population assays (Fig. 4.2, top right) are more statistically powerful and can be used to compare animals with different genotypes. These assays are useful in genetic mapping studies and when subtle changes in behavior need to be assessed. Population assays are carried out by placing a large number of animals (100) onto a 10-cm plate containing a thin film of agar and then placing a point source of the compound of interest and its diluent equidistant from the initial placement of the worms. An anesthetic is placed at the point source of the compound and at the counter–point source of the diluent control. After one hour, the worms at each point and throughout the plate are counted and a chemotaxis index is derived [35]. The point source sets up a gradient whose shape can be measured by conductivity for a soluble compound [36] but is really unknowable for a volatile one. The assays, however, are quite robust; mutants have been isolated in behavioral screens where those animals that fail to seek out the point source of the attractive chemical are isolated. In a careful study, Pierce-Shimomura et al. [36] concluded that C. elegans exhibits a biased random walk towards the attractive point source. This is a slight modification of the bacterial run and tumble where tumbles are initiated by a decrease in attractant concentration and runs are promoted by increases in attractant concentration. These authors found a slight bias in how a worm reoriented itself after the equivalent of a tumble (coined a pirouette). In the absence of a gradient, the reorientation of the worm after a spontaneous pirouette was 1808 from the initial direction of the worm’s movement. However, if the worm was traveling up a concentration gradient, the worm’s head was more likely to be pointed into the gradient than away from it, regardless of the direction in which it was initially traveling.

4.3.2

Repulsion

Repulsion from volatile chemicals has been tested using a plate assay as described for single-worm assays ([33], Fig. 4.2, top left). The negative chemotaxis examined in this assay requires sensation of an extended spatial gradient. This is less important for the response to soluble repellents. To assay repulsion from soluble repellents such as high osmolarity (8 M glycerol) noxious chemicals (SDS), the “drop test” has been used [16, 28]. In this assay, a droplet of chemical is placed just ahead of a moving worm so that the worm enters the drop of its own accord. Repellants such as quinine cause rapid backing of the worm once it enters the drop. The worm then changes its direction completely. Water-soluble repellents are sensed by the head and tail of the animal. The animal compares the stimulation at the head to that at the tail in deciding in which direction to move. Thus, a repellant at the head causes the worm to back. If the same repellent is sensed by the tail, the worm “realizes” it is surrounded and that moving backward may be worse than forward; thus, the response is somewhat randomized. A modification of this assay, the “dry drop” test, is described along with the phasmid organ. To assay populations of worms for their ability to respond to a water-soluble repellent such as 4 M fructose, a 2.5-cm diameter ring of the compound is placed onto a 6

81

82

4 Chemosensory Transduction in Caenorhabditis elegans

cm normal growth plate and the population of worms is placed at the center of the ring (Fig. 4.2, bottom left). The number of worms that cross the ring in 10 minutes is counted ([37, 38]). Wild-type worms will usually back in response to the ring of noxious compound, but osm mutants will ignore it.

4.3.3

Thermotaxis

Temperature is of great importance to a small, poikilothermic animal such as C. elegans. To analyze individuals and populations, the radial thermal gradient assay can be performed [5, 6, 39]. If it is fed at a specific temperature and presented with a radial temperature gradient, the worm will seek out the thermocline that matches the temperature at which it last dined. The animal will then track the thermocline. The radial temperature gradient (Fig. 4.2, bottom right) is produced by placing a small (closed) vial of frozen ammonium acetate (the melting temperature is 17 8C) at the center of a petri dish containing a thin film of agar and then moving the dish to 25 8C. To assess the thermal preference of a population of worms, the percent of worms at the center (17 8C), the middle (20 8C), and the edge (25 8C) of the dish is determined. Wild-type worms will accumulate in the region that most closely matches their cultivation temperature [5, 6, 39]. The precision with which the worm tracks the thermocline indicates that C. elegans can detect a thermal gradient of less than 0.1 8C. Once a worm is starved at a given temperature, it will avoid that temperature. The “memory” of temperature can be dissected from the ability to sense temperatures. If the major thermosensory neuron, AFD, is ablated, most animals become athermotactic, that is, they will move randomly on a temperature gradient, while some will be cryophilic. If the AIZ interneuron is ablated, they become thermophilic but will still track the new hotter isotherm [6]. This indicates that interneuron-ablated animals can sense temperature gradients; they just fail to determine what the “right” temperature is. Other population assays employ an extended gradient [5, 40]. These will not be examined here.

4.3.4

“Social Feeding” or Bordering

To quantify the gathering of worms at the borders of the lawn, a large number of animals (100) are placed onto a 2.5-cm diameter lawn of bacteria that has been grown for two days so that it has a defined border. After three hours, the fraction of animals at the border of the lawn is calculated [41]. To establish whether a single worm will feed socially, that worm is “marked” by a GFP transgene and mixed with non-fluorescent “social” animals. If it joins the clumps of social animals, it is deemed “social” [4].

4.4 How Is The Response to Each Stimulus Generated?

4.4

How Is The Response to Each Stimulus Generated?

This section will describe what is known about the signal transduction pathway from receptor-stimulus interaction to the channel opening. 4.4.1

The Chemotaxis Olfactory Response

Together, the AWC and AWA olfactory neuron pairs are responsible for chemotaxis towards more than 60 known single compounds [35]. AWA expresses ODR-10, the receptor for diacetyl, the odor of buttery popcorn [42]. Thus far, of the 1000 or so GPCRs predicted by the C. elegans genome (Genome Sequencing Consortium [43–45]), no other GPCR has been identified as an odor receptor. One other GPCR, STR-2, is expressed in a suggestive pattern: it is expressed in one of the two AWC neurons. It is randomly expressed in either the left or right AWC, never both [46]. Each olfactory neuron is thought to respond to at least 15 volatile chemicals [46]. One model for olfactory signaling starts with odor binding to a GPCR on an AWA or AWC (Fig. 4.3) neuron and stimulating the G alpha ODR-3 [42, 47]. There are 20 G-alphas, 2 G-betas, and 2 G-gammas expressed in the worm [48]. Of these, the G-alpha ODR-3 seems to be primarily responsible for olfactory signal transduction. An odr-3 null is unable to respond to any volatile compound, nor can it detect repellents [47]. Thus, ODR-3 would seem to be a central player in C. elegans chemosensation. There is redundancy of signaling through the G-alphas, however, as some responses are blunted but not eliminated in the odr-3 null and other responses can be restored to an odr-3 null by loss of a second G-alpha such as GPA-5. In AWC, ODR-3 is expressed along with GPA-2, GPA-13, and GOA-1. In AWA, ODR-3, GPA-5, and GPA-6 are all expressed [48]. The target of ODR-3 activity is not known, as its sequence does not predict whether it is a Gi or a Gs [47]. Downstream of ODR-3, differences between signal transduction in AWA and AWC become apparent. While in AWC, cGMP production via the guanylyl cyclases ODR-1 and DAF-11 is required for response to all odors sensed by these neurons, no guanylyl cyclase is required within AWA [49–51]. Since there is no precedent for a G-alpha stimulating guanylyl cyclases, it is hard to predict whether changes in cGMP levels are produced by activation of the guanylyl cyclases in an undiscovered way or by the inhibition of a phosphodiesterase. In either case, cGMP levels are predicted to rise in response to GPCR signaling, and the increased cGMP is thought to bind and increase the open-probability of the cGMP-gated channel. TAX-2 and TAX-4 are both required for response to all odors sensed by AWC [12, 13]. The cGMP-dependent protein kinase (PKG) EGL-4 is also thought to become activated by odor signaling. The active PKG then downregulates the AWC neuron. One target for its activity may be TAX-2, the beta subunit of the cyclic nucleotide-gated channel. Mutation of a serine residue at the C-terminus to the non-phosphorylatable alanine (S727A) renders the tax channel unable to adapt to short odor exposures [52].

83

84

4 Chemosensory Transduction in Caenorhabditis elegans Fig. 4.3 Models for sensory signal transduction within the olfactory AWC (top), AWA (second from top), gustatory ASE (third from top), and thermosensory AFD (bottom) neurons

4.4 How Is The Response to Each Stimulus Generated?

In the AWA neuron (Fig. 4.3), odor binding stimulates the G-alpha ODR-3 [47] and this is thought to open, by some unknown mechanism, the TRP-like channel comprised of OSM-9 and OCR-2 [8, 16]. There are no known mutants that are defective for AWA adaptation. Many mysteries remain: each neuron responds to at least 15 odors and each odor response is independently adapted. How is this achieved? What are the multiple guanylyl cyclases doing? What are the multiple G-alphas doing?

4.4.2

Chemotaxis to Water-soluble Compounds

The ASE pair of neurons is responsible for attraction to many salts [50]. ASER allows chemotaxis towards chloride and potassium, while ASEL provides the worm with the ability to sense sodium [36]. Though the ASE right and left neurons respond to different attractants, they both require theTAX-2/4 channel to generate their response [12, 13]. They both express the G-alpha GPA-3 (and no other) [48]. While ASEL expresses two guanylyl cyclases, gcy-6 and gcy-7, ASER expresses only gcy-5 [36, 53]. Electrophysiological studies on the ASE neuron indicated that ASE depolarizes in response to salt exposure, probably as a result of the TAX-2/4 channel opening [54]. Adaptation to salt was shown to require the G-gamma GPC-1 [25]. The involvement of the Ggamma argues that salt sensation requires a GPCR. The primary signal transduction channel could be an ENaC/Deg-type channel and the G-alpha, G-gamma guanylyl cyclases, and the cyclic nucleotide-gated channel might act peripherally. One model for salt sensation (Fig. 4.3) is proposed to start with salt binding to a GPCR and stimulating the G-alpha GPA-3. Once again, no one knows exactly how the levels of cGMP are increased in response to salt stimulation, but the guanylyl cyclases GCY-6 and 7 in ASEL and GCY-5 in ASER may provide the cGMP. The increased cGMP levels lead to channel opening, depolarizing the neuron and leading to chemotaxis. G-gammas are postulated to attenuate GPCR signaling by stimulating beta arrestin kinases to phosphorylate the ligand-bound GPCR, which leads to beta-arrestin binding and inactivation of the receptor [55]. Interestingly, the TRP channel OSM-9 is also required for salt adaptation [25].

4.4.3

Repellents The ASH Polymodal Sensory Neuron When a worm encounters a noxious stimulus such as a water-soluble repellent (bitter compounds, heavy metals, or SDS) [28], high osmotic strength [50], a repulsive volatile (octanol) [26], or a touch on the nose [56], it rapidly backs away and often reverses its course. ASH is the single neuron pair responsible for most of the worm’s escape behavior from these disparate noxious stimuli [28, 50, 56]. How all these disparate cues are sensed by ASH is still a mystery, as no receptor has been identified. Though 4.4.3.1

85

86

4 Chemosensory Transduction in Caenorhabditis elegans

mechano-, osmo-, and chemosensation would seem very different, ASH requires one channel comprised of OSM-9 and OCR-2 to respond to each stimulus [16]. The Galpha ODR-3 is also required for all ASH responses [16]. The requirement for ODR-3 makes it seem likely that GPCRs initiate the signaling response to most if not all ASH-sensed stimuli. Worms that express the VR1 capsaicin receptor in ASH respond to capsaicin by backing and reversing. Tobin et al. found that this response did not require ODR-3, indicating that the VR1 channel is directly gated by capsaicin while the native C. elegans TRP channel is probably not directly gated by noxious stimuli. Instead, the G-alpha ODR-3 may initiate DAG signaling and the IP2 pathway may be required to gate this channel. Calcium entry, once again, seems to dampen the response to external stimuli sensed by ASH. The calcineurin A, tax-6 mutant worms reverse in response to concentrations of glycerol (2 M) that wild-type worms cannot sense [39]. The ASH-mediated responses require EAT-4, the glutamate transporter [16]. This indicates that signaling from ASH (or an intermediary neuron) to the backward command interneurons requires glutamate release. Hart et al. [57] and Mariq et al. [58] showed that mechano- and osmosensation could be differentially affected by mutation of a single non-NMDA-type glutamate receptor, GLR-1. Mechanosensation required GLR-1 while osmosensation did not. Mellem et al. [31] suggested that osmosensation activates both NMDA (NMR-1) and non-NMDA glutamate receptor (GLR-1 and GLR2), while mechanosensation activates only the non-NMDA receptor GLR-1/2. They postulate that osmostimulation evokes a larger glutamate release than does mechanostimulation. The larger release may activate both the synaptic GLR-1/2 and the extra synaptic NMR-1, while the smaller mechano-induced release activates only the synaptic GLR-1/2.

4.4.4

Thermotaxis

The ability to find the correct temperature is essential to the worm. At least two sensory neurons are required for the worm to locate the isotherm that matches the temperature at which it was raised (Fig. 4.3). Laser-ablation studies identified AFD as the major neuron required to drive the worm towards the hotter climes. Thus, ablation of AFD makes some wild type worms athermotaxic and others cryophilic [6]. Ablation of the AIY interneuron caused cryophilic behavior, while ablation of the interneuron AIZ caused thermophilic behavior. Thus, it is the interplay between the two interneurons that results in the worm’s ability to seek the appropriate temperature. Laser ablations, however, could not identify the neuron “X,” postulated to signal to AIZ. How temperature generates a signal within AFD is really unknown, though we have hints at the identity of some players. cGMP and calcium are probably major second messengers within AFD. cGMP is implicated because tax-2/4 mutants are athermotactic, meaning that they seek neither hot nor cold [13]. It really is not known whether TAX-2/4 opens or closes in response to heat. In support of heat causing channel opening and calcium entry, the action of the calcineurin A is to dampen the response of

4.4 How Is The Response to Each Stimulus Generated?

AFD [39]. This is what would be postulated to happen if stimulation is to cause the neuron to become less sensitive to the stimulus and allow it to react to a wider range of temperatures. In support of a major role for cGMP in AFD signaling, Yu et al. [53] found that the guanylyl cyclase GCY-8 is expressed in AFD and may provide the cGMP to open the channel in response to changes in temperature. Several key players are noticeable for their absence from AFD. Though expression studies using engineered GFP reporters must be taken with a grain of salt, such studies failed to observe G-alpha or TRP-like channel expression in AFD. The absence of a G-alpha might indicate that the temperature receptor is not a GPCR. Mori [7] postulated that a receptor guanylyl cyclase such as GCY-8 might be the temperature receptor. Since the TRP channels provide the sensory signal for temperature in mammals, they would seem to be the most obvious candidates for an AFD receptor, but to date they seem neither to be expressed in AFD nor involved in temperature-sensing behavior. Killing AFD caused cryophilic phenotypes, while a tax-2/4 mutant is athermotaxic. This leads to the hypothesis that both AFD and the cryotaxic sensory neuron “X” use the tax channels to signal temperature. Not much more is known about neuron X. 4.4.5

Feeding Behavior

Worms exhibit two strategies when they feed: in a pattern reminiscent of the binary feeding strategies of Drosophila larvae [59], they dine either alone or en masse. The strain adopted by Sydney Brenner, N2, isolated from a compost heap in Bristol, England, dines alone. This makes them easy to pick from a plate and they behave nicely, i.e., they do not tear into the agar or “burrow.” About half of the wild strains actually are much messier [41]; they aggregate at the lawn’s edge and burrow beneath the agar, making them harder to capture with a “pick” (the platinum wire we use to move individuals about). This intriguing behavior spurred identification of its genetic basis. A single gene, originally called bor-1 for its bordering defects (Randy Cassada, personal communication), was shown to be responsible for the phenotype. An inactivating mutation within the neuropeptide Y receptor was shown to cause the bordering, burrowing, and “social feeding” behavior [41]. Half the wild-type strains examined had a polymorphism at a critical residue of the neuropeptide Y receptor that was predicted to decrease its function. Thus, the “solitary” N2 strain has a more active form of the receptor and the “social” strain has a less active form [41]. In C. elegans biology, as is soon to be true of biology in general in the post-genomic era, the genetic basis for behavior is often known and guesses can be made about how the gene is involved with the behavior, but there is a huge gap between the gene and the behavior. What does the gene do? Which neurons require the gene, and what do these neurons actually do for the animal? In the case of the “social feeding” phenotype, Mario de Bono and Cori Bargmann decided to determine which other genes were responsible for the bordering behavior; thus, they looked for suppressors of the bordering and burrowing phenotype. It turned

87

88

4 Chemosensory Transduction in Caenorhabditis elegans

out that both the tax and osm channel mutants suppressed this phenotype. That is, TAX-2/4 and OSM-9/OCR-2 are required to interpret the environmental cues that promote bordering, burrowing, and swarming [4]. Juliet Coates found that the TAX channels were required in neurons that had their sensory cilia exposed to the interior of the worm. Furthermore, these cells (AQR, PQR, and URX) express the soluble guanylyl cyclase GCY-32 [53, 60]. Other soluble guanylyl cyclases have been shown to dimerize and produce cGMP via the binding of the gas NO. The worm genome does not have a plausible NO synthase, and the gas that binds GCY-32 therefore remains a mystery. One idea is that the concentration of gas depends on the ratio of bacteria to oxygen in the environment. In the moist environment of the soil, gasses are trapped in specific microenvironments and can indicate the bacterial “load” of that space. Thus, the CO2 and O2 content of a space can indicate a source of food. Worms that are in an airtight environment with an abundance of bacteria can “suffocate” in the anoxic conditions produced by the bacteria. One way for a group of worms to get out of an environment that is teetering on the brink of having too little oxygen is to swarm together to burrow their way out. It would be helpful if worms that liked such environments would enter them together so that they could help each other exit as well. The idea is that they are truly social animals and can go about almost like a living “drill bit” that can bore into and out of high-risk, high-gain environments. de Bono et al. [4] determined that the TRP channel comprised of OSM-9 and OCR-2 functions in the ASH and ADL neurons that respond to water-soluble repellents to promote burrowing behavior in the “social feeding,” npr-1 mutant strain. The environmental cues that ASH and ADL respond to by promoting burrowing are unknown but may be byproducts of bacterial metabolism [4]. Though the decision to enter the alternative, dauer larva stage is not a behavior per se, but rather a developmental decision, it is regulated in part by signals generated by the TAX channels. This is not surprising since many of the cues that lead to the dauer decision are food related and the tax channels are primarily used for food location and sensation. Tax channels are also required for the dauer promoting pathways; thus, channel mutants are potent inhibitors of the dauer decision in some genetic backgrounds.

4.5

Structure of the TAX, Cyclic Nucleotide-gated Channels of the Worm

The alignment of the TAX channel subunits with several representative mammalian channels is shown in Fig. 4.4A [61]. Several key features are revealed in the alignment. The TAX-2 and TAX-4 sequences share key motifs with the vertebrate CNG channels: six transmembrane domains (S1–6), one pore-forming loop (P), and a cyclic nucleotide-binding domain. Examination of the selectivity filter [62] of TAX-4 and TAX-2 indicated that TAX-4 is an alpha subunit and TAX-2 a beta subunit (see Fig. 4.4B). Coburn and Bargmann [12] showed that tax-2 defective mutants can be rescued by over-expression of the TAX-4 cDNA but not the TAX-2 cDNA. This is consistent with the evidence that homomeric alpha subunits can reconstitute channel activity

4.5 Structure of the TAX, Cyclic Nucleotide-gated Channels of the Worm

when expressed heterologously, but beta subunits cannot [63–66]. Corroborating evidence in Komatsu et al. showed that HEK cells transfected with TAX-4 alone were able to generate a cGMP-induced current, but cells transfected with TAX-2 alone were not [67]. The cyclic nucleotide-binding domains of TAX-2 and TAX-4 were predicted to bind cGMP. The cyclic nucleotide-binding domain consists of two alpha helices followed by an anti-parallel beta roll, which is followed by a third alpha helix (the C helix). Aspartic acid 604 within the bovine alpha subunit seems to be responsible for cGMP selectivity in some cases [68]. The corresponding residue within TAX-2 and TAX-4 is also an aspartic acid. The specificity for cGMP was confirmed by Komatsu et al. [67], who showed that HEK-expressed TAX-2/4 channels are half-maximally activated by about 10 lm cGMP, while it took two orders of magnitude more cAMP to reach EC50. cGMP’s EC50 for the TAX-2/4 channel was in the same range as cGMP and cAMP are for the mammalian olfactory channels; this is 10-fold less than cGMP’s EC50 for the rod channels [64, 67]. This may support the idea that in C. elegans, the TAX channels open and the sensory neurons depolarize in response to stimulation. Thus far, the evidence for this is based on electrophysiology of a few neurons including ASE [54], AWC, and AWA [69]. There are four additional predicted cyclic nucleotide-gated channel subunits in the C. elegans genome. F14H8.6 (this channel is the fourth gene on the cosmid F14H8), C23H5.7, Y76B12C.1, and F38E11.12 have been added to the alignment in Fig. 4.4A. cDNA has been identified for F14H8.6, Y76B12C.1, and F38E11.12 (http:// www.wormbase.org). By sequence analysis, F14H8.6 and F38E11.12 seem to be alpha subunits (Fig. 4.4B). F14H8.6 is predicted to be liganded by cGMP, but it is hard to predict the ligand specificity of F38E11.12. Y76B12C.1 and C23H5.7 are predicted beta subunits that could be equally responsive to cAMP or cGMP. Four kilobases of sequence upstream of the start site of the alpha subunit F38E11.12 (formerly known as F38E11.8) were used to drive a GFP expression reporter. Expression was observed in the AWC, AFD, ASE, AWB, and ASI sensory neurons ([27], Tab. 4.2). TAX-2 and TAX-4 are expressed in these same neurons, as well as ASJ, ASK, ASG, URX, and BAG (Tab. 4.2). What might these additional subunits be doing? Recent data from Zhong et al. [70] indicate that a leucine zipper within the C-terminus of the alpha subunits of the rod and olfactory channels trimerizes to give a channel stoichiometry of three alphas to one beta. An examination of the alignment in Fig. 4.4C shows that TAX-4 has the same trimeric leucine zipper (also note Fig. 2a in [70]). TAX-2, like vertebrate beta subunits, does not share the amino acids shown to be critical for co-immunoprecipitation of the trimeric leucine zipper found in the rod alpha subunits. This might indicate that the TAX channels share the mammalian cyclic nucleotide-gated channel’s A3:B1 stoichiometry by containing three subunits of TAX-4 and one of TAX-2 per functional channel. The other alpha subunits, F14H8.6 and F38E11.12, both share the conserved leucine zipper, while the predicted beta subunits do not. Since TAX-4 and F38E11.12 are expressed together in AWC, AFD, ASE, AWB, and ASI, they could form alternate channels [27]. F38E11.12 may act like CNGA4 [71] and change the sensitivity of the TAX channels for factors such as calmodulin that modulate their activity. The fact that these

89

90

4 Chemosensory Transduction in Caenorhabditis elegans

Fig. 4.4 (A) Alpha and beta subunit alignments. (B) Selectivity filter. (C) Leucine zipper alignment

channel subunits have not been identified in behavioral screens leads to the speculation that their function is subtle and may be required only for modulating sensory behavior. The alignment of the “new” subunits reveals another interesting feature: each subunit has an extra 46 amino acids between S5 and the pore-forming helix. It has been suggested that this region could form part of the selectivity filter for the channel

4.5 Structure of the TAX, Cyclic Nucleotide-gated Channels of the Worm

Fig. 4.4

continued

91

92

4 Chemosensory Transduction in Caenorhabditis elegans Tab. 4.2

C. elegans cyclic nucleotide-gated channel subunits

Subunit

Alpha/beta

Cyclic nucleotide

Expression

Mutant defects

TAX-4

Alpha (confirmed)

CGMP (confirmed)

Confirmed by cDNA

Chemotaxis-defective: AWC-sensed odors

GFP reporters show: Repulsion: AWB-sensed AWC, AWB, ASE, ASK, odors ASI, ASJ, AFD, ASG, Chemotaxis defective: URX, BAG, PHA ASE-sensed salts Thermotaxis defective (AFD) Suppresses daf-11 dauer constitutivity (multiple neurons) ASJ and ASE axons show aberrant morphology Suppresses npr-1 “social feeding” (URX) Phasmid axons show aberrant morphology TAX-2

Beta (confirmed)

CGMP (confirmed)

Confirmed by cDNA GFP reporters show: AWC, AWB, ASE, ASK, ASI, ASJ, AFD, ASG, URX, BAG, PQR

Chemotaxis-defective: AWC-sensed odors. Repulsion: AWB-sensed odors. Chemotaxis defective: ASE-sensed salts. Thermotaxis defective (AFD). Suppresses daf-11 dauer constitutivity (multiple neurons). ASJ and ASE axons show aberrant morphology. Suppresses npr-1 “social feeding” (URX). Phasmid axons show aberrant morphology

4.6 Channel Regulation Tab. 4.2

C. elegans cyclic nucleotide-gated channel subunits

Subunit

Alpha/beta

Cyclic nucleotide

Expression

Mutant defects

F14H8.6

Beta (proposed)

cGMP (proposed)

Confirmed by cDNA

No known mutations in this gene

F38E11.12

Beta (proposed)

Unable to predict

Y76B12C.1

Alpha (proposed)

Unable to predict

C23H5.7

Alpha (proposed)

Unable to predict

No expression data Confirmed by cDNA GFP reporters show: AWC, AWB, AFD, ASI expression Confirmed by cDNA

No known mutations in this gene

No known mutations in this gene

No expression data No cDNA confirmation No known mutations in available this gene

(Tsung-Yu Chen and Chul Sun Park, personal communication). The extra sequence has a net negative charge of 6, which might increase the affinity for cations, allow Mg to block the pore more completely, and increase the voltage dependence of the channel.

4.6

Channel Regulation

The classical modulator of olfactory channels is the calcium-binding protein calmodulin [68, 72–74]. The TAX channels could be subject to regulation by calmodulin, but it is difficult to predict whether any of the subunits do, in fact, contain a calmodulinbinding site. There are, however, multiple predicted phosphorylation sites for both cA/GMP-dependent protein kinases and calcium-dependent protein kinases. As discussed in the olfactory signaling section of this chapter, the cGMP-dependent protein kinase, EGL-4, was shown to be required for adaptation of the AWC neuron to odors [52]. Phosphorylation of serine 727 within TAX-2’s C-terminus was shown to be required for effective adaptation to 30 minutes or less of AWC-sensed odor exposure [52]. Longer odor exposures led to adaptation of the olfactory response that was the same as wild-type. This indicates that after brief exposures to odor, a kinase may phosphorylate TAX-2, making it less responsive to stimulation. The position of the phosphorylation site, in the Cterminus but after the cN-binding domain, makes it hard to predict how phosphorylation might regulate channel activity. This phosphorylation site is not conserved in the other beta subunits (Fig. 4.4A).

93

94

4 Chemosensory Transduction in Caenorhabditis elegans

Analysis of the phosphatase calcineurin A (tax-6) mutants indicates that the AWC response in tax-6 mutants is hyperadapted [39]. tax-6 mutants are unable to chemotax toward AWC-sensed odors. Kuhara et al. were able to show that the adaptation-defective mutant osm-9 is able to suppress the chemotaxis defects of a tax-6 mutant [39]. Thus, brief exposures to odor may cause rapid down-regulation of AWC’s response to that odor via OSM-9. Usually, calcineurin keeps this rampant adaptation in check [39]. Thus, the kinase EGL-4 and the phosphatase TAX-6 may work in opposition to each other. If they work on the same residue, cGMP produced in response to odor may decrease the sensitivity of the channel to cGMP (making the cell less responsive to odor), while calcium entry through this or another channel may negate this and reset the sensitivity of the channel (and make the cell responsive after a lag). It would be the ratio between the activities of these molecules that would give the AWC neuron its dynamic range. The fact that the double tax-6 osm-9 mutant can chemotax indicates that the phosphatase is not required for the dynamic alterations within AWC that allow for gradient sensation. Prolonged exposures to odor must drive additional changes, perhaps by the action of the PKG in the nucleus [52]. Regulation of the TAX channel in other neurons has not been examined directly. Regulation of the TRP channels may occur by alteration of the subunit composition of the channel: neurons that express both OSM-9 and OCR-2 use this channel for neuronal signaling, while neurons that express only the OSM-9 subunit use it for adaptation of the sensory signal. In both AWC and ASE, the primary chemosensory signal is provided by TAX-2/4, while OSM-9 allows for neuronal adaptation to odor and salt, respectively [8, 14, 25].

References 1 2

3

4

5

6

Brenner, S., The genetics of Caenorhabditis elegans. Genetics, 1974. 77(1): p. 71–94. White, J.G., et al., The structure of the ventral nerve cord of Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci, 1976. 275(938): p. 327–48. Sulston, J. and J. Hodgkin, Methods: The nematode Caenorabditis elegans. AKA C. elegans I. Cold Spring Harbor Laboratory Press, 1988: p. 587–606. de Bono, M., et al., Social feeding in Caenorhabditis elegans is induced by neurons that detect aversive stimuli. Nature, 2002. 419(6910): p. 899–903. Hedgecock, E.M. and R.L. Russell, Normal and mutant thermotaxis in the nematode Caenorhabditis elegans. Proc Natl Acad Sci U S A, 1975. 72(10): p. 4061–5. Mori, I. and Y. Ohshima, Neural regulation of thermotaxis in Caenorhabditis elegans. Nature, 1995. 376(6538): p. 344–8.

7

8

9

10

11

Mori, I., Genetics of chemotaxis and thermotaxis in the nematode Caenorhabditis elegans. Annu Rev Genet, 1999. 33: p. 399–422. Colbert, H.A. and C.I. Bargmann, Odorantspecific adaptation pathways generate olfactory plasticity in C. elegans. Neuron, 1995. 14(4): p. 803–12. Saeki, S., M. Yamamoto, and Y. Iino, Plasticity of chemotaxis revealed by paired presentation of a chemoattractant and starvation in the nematode Caenorhabditis elegans. J Exp Biol, 2001. 204(Pt 10): p. 1757–64. Albert, P.S. and D.L. Riddle, Developmental alterations in sensory neuroanatomy of the Caenorhabditis elegans dauer larva. J Comp Neurol, 1983. 219(4): p. 461–81. Dusenbery, D.B., R.E. Sheridan, and R.L. Russell, Chemotaxis-defective mutants of the nematode Caenorhabditis elegans. Genetics, 1975. 80(2): p. 297–309.

4.6 Channel Regulation 12

13

14

15

16

17

18

19

20

21

22

23

24

25

Coburn, C.M. and C.I. Bargmann, A putative cyclic nucleotide-gated channel is required for sensory development and function in C. elegans. Neuron, 1996. 17(4): p. 695–706. Komatsu, H., et al., Mutations in a cyclic nucleotide-gated channel lead to abnormal thermosensation and chemosensation in C. elegans. Neuron, 1996. 17(4): p. 707–18. Colbert, H.A., T.L. Smith, and C.I. Bargmann, OSM-9, a novel protein with structural similarity to channels, is required for olfaction, mechanosensation, and olfactory adaptation in Caenorhabditis elegans. J Neurosci, 1997. 17(21): p. 8259–69. Harteneck, C., T.D. Plant, and G. Schultz, From worm to man: three subfamilies of TRP channels. Trends Neurosci, 2000. 23(4): p. 159–66. Tobin, D., et al., Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans neurons. Neuron, 2002. 35(2): p. 307–18. Caterina, M.J., et al., The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature, 1997. 389(6653): p. 816–24. Bargmann, C.I., Neurobiology of the Caenorhabditis elegans genome. Science, 1998. 282(5396): p. 2028–33. Ward, S., et al., Electron microscopical reconstruction of the anterior sensory anatomy of the nematode Caenorhabditis elegans.?2UU. J Comp Neurol, 1975. 160(3): p. 313–37. Ware, R., et al., Caenorhabditis elegans: sensory input and motor output. J. Comp. Neurol., 1975. 162: p. 71–110. Ward, S., Chemotaxis by the nematode Caenorhabditis elegans: identification of attractants and analysis of the response by use of mutants. Proc Natl Acad Sci U S A, 1973. 70(3): p. 817–21. Dusenbery, D.B., Behavior of free-living nematodes., in Nematodes as Biological Models, B. Zuckerman, Editor. 1980, Academic Press: New York. p. 127–158. Riddle, D.L., C. elegans II. Cold Spring Harbor monograph series 33. 1997, Plainview, N.Y.: Cold Spring Harbor Laboratory Press. xvii, 1222. Colbert, H.A. and C.I. Bargmann, Environmental signals modulate olfactory acuity, discrimination, and memory in Caenorhabditis elegans. Learn Mem, 1997. 4(2): p. 179–91. Jansen, G., D. Weinkove, and R.H. Plasterk, The G-protein gamma subunit gpc-1 of the nematode C.elegans is involved in taste adaptation. Embo J, 2002. 21(5): p. 986–94.

26

27

28

29

30

31

32

33

34

35

36

37

38

39

Troemel, E.R., et al., Divergent seven transmembrane receptors are candidate chemosensory receptors in C. elegans. Cell, 1995. 83(2): p. 207–18. Coburn, C.M., Cyclic Nucleotide Gated Channels in C. elegans, in Thesis. 1996, Univ. California, San Francisco: San Francisco. Hilliard, M.A., C.I. Bargmann, and P. Bazzicalupo, C. elegans responds to chemical repellents by integrating sensory inputs from the head and the tail. Curr Biol, 2002. 12(9): p. 730–4. Suzuki, H., et al., In vivo imaging of C. elegans mechanosensory neurons demonstrates a specific role for the MEC-4 channel in the process of gentle touch sensation. Neuron, 2003. 39(6): p. 1005–17. Wang, J.W., et al., Two-photon calcium imaging reveals an odor-evoked map of activity in the fly brain. Cell, 2003. 112(2): p. 271–82. Mellem, J.E., et al., Decoding of polymodal sensory stimuli by postsynaptic glutamate receptors in C. elegans. Neuron, 2002. 36(5): p. 933–44. Lee, R.Y., et al., EAT-4, a homolog of a mammalian sodium-dependent inorganic phosphate cotransporter, is necessary for glutamatergic neurotransmission in caenorhabditis elegans. J Neurosci, 1999. 19(1): p. 159–67. Troemel, E.R., B.E. Kimmel, and C.I. Bargmann, Reprogramming chemotaxis responses: sensory neurons define olfactory preferences in C. elegans. Cell, 1997. 91(2): p. 161–9. Sagasti, A., et al., Alternative olfactory neuron fates are specified by the LIM homeobox gene lim4. Genes Dev, 1999. 13(14): p. 1794–806. Bargmann, C.I., E. Hartwieg, and H.R. Horvitz, Odorant-selective genes and neurons mediate olfaction in C. elegans. Cell, 1993. 74(3): p. 515–27. Pierce-Shimomura, J.T., T.M. Morse, and S.R. Lockery, The fundamental role of pirouettes in Caenorhabditis elegans chemotaxis. J Neurosci, 1999. 19(21): p. 9557–69. Culotti, J.G. and R.L. Russell, Osmotic avoidance defective mutants of the nematode Caenorhabditis elegans. Genetics, 1978. 90(2): p. 243–56. Vowels, J.J. and J.H. Thomas, Multiple chemosensory defects in daf-11 and daf-21 mutants of Caenorhabditis elegans. Genetics, 1994. 138(2): p. 303–16. Kuhara, A., et al., Negative regulation and gain control of sensory neurons by the C. elegans calcineurin TAX-6. Neuron, 2002. 33(5): p. 751–63.

95

96

4 Chemosensory Transduction in Caenorhabditis elegans 40

41

42

43

44

45

46

47

48

49

50

51

52

Yamada, Y. and Y. Ohshima, Distribution and movement of Caenorhabditis elegans on a thermal gradient. J Exp Biol, 2003. 206(Pt 15): p. 2581–93. de Bono, M. and C.I. Bargmann, Natural variation in a neuropeptide Y receptor homolog modifies social behavior and food response in C. elegans. Cell, 1998. 94(5): p. 679–89. Sengupta, P., J.H. Chou, and C.I. Bargmann, odr-10 encodes a seven transmembrane domain olfactory receptor required for responses to the odorant diacetyl. Cell, 1996. 84(6): p. 899–909. Genome sequence of the nematode C. elegans: a platform for investigating biology. The C. elegans Sequencing Consortium. Science, 1998. 282(5396): p. 2012–8. Robertson, H.M., Two large families of chemoreceptor genes in the nematodes Caenorhabditis elegans and Caenorhabditis briggsae reveal extensive gene duplication, diversification, movement, and intron loss. Genome Res, 1998. 8(5): p. 449–63. Robertson, H.M., Updating the str and srj (stl) families of chemoreceptors in Caenorhabditis nematodes reveals frequent gene movement within and between chromosomes. Chem Senses, 2001. 26(2): p. 151–9. Troemel, E.R., A. Sagasti, and C.I. Bargmann, Lateral signaling mediated by axon contact and calcium entry regulates asymmetric odorant receptor expression in C. elegans. Cell, 1999. 99(4): p. 387–98. Roayaie, K., et al., The G alpha protein ODR-3 mediates olfactory and nociceptive function and controls cilium morphogenesis in C. elegans olfactory neurons. Neuron, 1998. 20(1): p. 55–67. Jansen, G., et al., The complete family of genes encoding G proteins of Caenorhabditis elegans. Nat Genet, 1999. 21(4): p. 414–9. L’Etoile, N.D. and C.I. Bargmann, Olfaction and odor discrimination are mediated by the C. elegans guanylyl cyclase ODR-1. Neuron, 2000. 25(3): p. 575–86. Bargmann, C.I. and H.R. Horvitz, Chemosensory neurons with overlapping functions direct chemotaxis to multiple chemicals in C. elegans. Neuron, 1991. 7(5): p. 729–42. Birnby, D.A., et al., A transmembrane guanylyl cyclase (DAF-11) and Hsp90 (DAF-21) regulate a common set of chemosensory behaviors in caenorhabditis elegans. Genetics, 2000. 155(1): p. 85–104. L’Etoile, N.D., et al., The cyclic GMP-dependent protein kinase EGL-4 regulates olfactory adaptation in C. elegans. Neuron, 2002. 36(6): p. 1079–89.

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

Yu, S., et al., Guanylyl cyclase expression in specific sensory neurons: a new family of chemosensory receptors. Proc Natl Acad Sci U S A, 1997. 94(7): p. 3384–7. Goodman, M.B., et al., Active currents regulate sensitivity and dynamic range in C. elegans neurons. Neuron, 1998. 20(4): p. 763–72. Kameyama, K., et al., Activation by G protein beta gamma subunits of beta-adrenergic and muscarinic receptor kinase. J Biol Chem, 1993. 268(11): p. 7753–8. Kaplan, J.M. and H.R. Horvitz, A dual mechanosensory and chemosensory neuron in Caenorhabditis elegans. Proc Natl Acad Sci U S A, 1993. 90(6): p. 2227–31. Hart, A.C., S. Sims, and J.M. Kaplan, Synaptic code for sensory modalities revealed by C. elegans GLR-1 glutamate receptor. Nature, 1995. 378(6552): p. 82–5. Maricq, A.V., et al., Mechanosensory signalling in C. elegans mediated by the GLR-1 glutamate receptor. Nature, 1995. 378(6552): p. 78–81. Osborne, K.A., et al., Natural behavior polymorphism due to a cGMP-dependent protein kinase of Drosophila. Science, 1997. 277(5327): p. 834–6. Coates, J.C. and M. de Bono, Antagonistic pathways in neurons exposed to body fluid regulate social feeding in Caenorhabditis elegans. Nature, 2002. 419(6910): p. 925–9. Combet, C., et al., NPS@: network protein sequence analysis. Trends Biochem Sci, 2000. 25(3): p. 147–50. Flynn, G.E. and W.N. Zagotta, A cysteine scan of the inner vestibule of cyclic nucleotide-gated channels reveals architecture and rearrangement of the pore. J Gen Physiol, 2003. 121(6): p. 563–82. Finn, J.T., M.E. Grunwald, and K.W. Yau, Cyclic nucleotide-gated ion channels: an extended family with diverse functions. Annu Rev Physiol, 1996. 58: p. 395–426. Zagotta, W.N. and S.A. Siegelbaum, Structure and function of cyclic nucleotide-gated channels. Annu Rev Neurosci, 1996. 19: p. 235–63. Biel, M., X. Zong, and F. Hofmann, Cyclic nucleotide gated channels. Adv Second Messenger Phosphoprotein Res, 1999. 33: p. 231–50. Kaupp, U.B., Family of cyclic nucleotide gated ion channels. Curr Opin Neurobiol, 1995. 5(4): p. 434–42. Komatsu, H., et al., Functional reconstitution of a heteromeric cyclic nucleotide-gated channel of Caenorhabditis elegans in cultured cells. Brain Res, 1999. 821(1): p. 160–8.

4.6 Channel Regulation 68

69

70

71

Varnum, M.D., K.D. Black, and W.N. Zagotta, Molecular mechanism for ligand discrimination of cyclic nucleotide-gated channels. Neuron, 1995. 15(3): p. 619–25. Nickell, W.T., et al., Single ionic channels of two Caenorhabditis elegans chemosensory neurons in native membrane. J Membr Biol, 2002. 189(1): p. 55–66. Zhong, H., et al., The heteromeric cyclic nucleotide-gated channel adopts a 3A:1B stoichiometry. Nature, 2002. 420(6912): p. 193–8. Bradley, J., D. Reuter, and S. Frings, Facilitation of calmodulin-mediated odor adaptation by cAMP-gated channel subunits. Science, 2001. 294(5549): p. 2176–8.

72

73

74

75

Chen, T.Y. and K.W. Yau, Direct modulation by Ca(2+)-calmodulin of cyclic nucleotide-activated channel of rat olfactory receptor neurons. Nature, 1994. 368(6471): p. 545–8. Kurahashi, T. and A. Menini, Mechanism of odorant adaptation in the olfactory receptor cell. Nature, 1997. 385(6618): p. 725–9. Leinders-Zufall, T., M. Ma, and F. Zufall, Impaired odor adaptation in olfactory receptor neurons after inhibition of Ca2+/calmodulin kinase II. J Neurosci, 1999. 19(14): p. RC19. Perkins, L. A. et al., Mutant sensory cilia in the nemat ode Caenorhabditis elegans. Dev Piol, 1986. 117 (2): p. 465–87.

97

99

5

Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents Johannes Reisert and Jonathan Bradley

Abstract

Over the past 40 years much data have been generated from studies on olfaction, starting with experiments using electrophysiology and, over the last decade, molecular biology. From these works a picture is beginning to emerge as to how odorant binding to olfactory receptor neurons (ORNs) is transduced into an electrical signal. Volatile chemicals (on the order of 300 Da or less) act as stimulants in olfaction by recognition and binding to receptor proteins expressed on the surface of ORNs. ORNs are bipolar cells that extend single dendrites to an epithelial border from which cilia protrude into a nasal mucus, covering the olfactory epithelium in the nasal cavity. Therefore, unlike rods and cones in the eye, or hair cells in the ear, ORNs are in direct contact with the external environment. Olfactory signal transduction and the generation of the depolarizing receptor current occur in the cilia, and hence are a function of the unique extraciliary environment of the nasal mucus. ORNs employ a mechanism that allows them to cope with this particular ionic composition and still provide reliable electrical readout of odor binding. That is, the receptor potential combines ion conduction of two distinct channels: a cAMP-gated Ca2+ channel and a Ca2+-activated Cl– channel, both of which contribute to excitation. Concentrating on vertebrates, we will introduce the olfactory system and then describe the characteristics of the two channels in question and their interaction to generate the odor-induced receptor current.

5.1

Introduction 5.1.1

Tissue

In vertebrates the sensory tissue, or olfactory epithelium (OE), is located in the nasal cavity. In mammals the inhaled air first passes over a respiratory epithelium (RE) where it is warmed and moistened. The lateral walls of the posterior portion of the Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

100

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

nasal cavity

basal cell

Fig. 5.1 Schematic diagram of the olfactory epithelium. Illustration of the three basic cell types in the olfactory epithelium. Basal cells are located along the mucus basement membrane. Soma of the olfactory sensory neurons are located in the central portion of the epithelium and have a bipolar morphology with a single dendrite extending to the lumen and an axon supporting extending to the olfactory bulb. The tip of the dendrite cell forms a knob from which emanate the olfactory cilia. Supporting cells constitute the third type of cell in the epithelium

olfactory receptor cell

cavity are elaborated into a series of complex folds called turbinal extensions (Fig. 6.1A). These have a surface area of a few square centimeters in man and more than 100 cm2 in dogs. The OE is located on these ecto- and endoturbinals and is divided into two histologically distinct layers isolated by a basement membrane. The deeper layer, or lamina propria, is glandular and contains vascular and connective tissue. The Bowman’s glands have ducts that traverse the basement membrane and neuroepithelium to the nasal lumen and secrete components of a protective mucus layer. These include secretory forms of the immunoglobulin types A and M, bacteriostatic and bactericidal proteins including lactoferrin and lysozyme, and detoxifying enzymes [1]. The lamina propria layer also contains the nerve fascicles of the receptor neurons, located in the upper neuroepithelial layer. The neuroepithelium is 100–200 lm thick and is pseudostratified, consisting of three cell types with the nuclei of each cell type in fairly discrete layers (see Fig. 5.1).

5.1.2

Olfactory Receptor Neurons

The olfactory receptor neurons (ORNs) constitute the predominant cell type in the OE and have nuclei scattered in a band six or eight nuclei wide through the central portion of the epithelium. There are about 106 ORNs in the rat, and they constitute 75–80 % of all the epithelial cells [2]. ORNs are bipolar with unmyelinated axons [3] that group into fascicles in the lamina propria surrounded by a single Schwann cell and project directly to the CNS, signaling in the olfactory bulb. A single dendrite terminates at the nasal lumen and ends in a structure referred to as a dendritic knob. Emanating from each dendritic knob and extending into the lumen of the nasal cavity are 5–20 immotile cilia 50–200 lm in length (in mammals) that taper from 2–3 lm at their base to about 0.1 lm at their tip. ORNs are a unique neuronal cell type in that they continually turn over on average every 30–60 days and are replenished from a population of mitotic basal cells [4]. Although the ORNs are derived from the basal cells and differenti-

5.1 Introduction

ate into mature sensory neurons, several features classify them as “juvenile” in comparison to neurons elsewhere. First, the time course of death of the ORNs in response to axotomy is rapid in comparison to other neurons [5]. Neurons in adult animals typically survive axotomy or die over the course of days or weeks, even when the site of axotomy is relatively close to the soma [6]. In contrast, ORNs, like neuronal populations in embryos and neonates [7], respond to axotomy with profound and rapid death. Immunohistochemically, ORNs also resemble “juvenile” neurons elsewhere in that they retain a pattern of intermediate filament- and microtubule-associated protein expression, which is characteristic of immature neurons [8–10]. Also resembling “juvenile” neurons elsewhere, ORNs retain a high concentration of intracellular chloride [11], an important feature of olfactory signal transduction (see below).

5.1.3

Sustentacular Cells

Sustentacular, or supporting cells, span the width of the epithelium and have their nuclei located most apically, forming a single layer. In the rat, sustentacular cells constitute 10-12 % of all the epithelial cells [2]. These cells act to physically protect and electrically isolate the receptor neurons, and also contribute secretions to the mucus [12]. Each sustentacular cell is surrounded by the dendrites and somata of 2–8 sensory neurons [4]. Sustentacular cells have a relatively hyperpolarized resting membrane potential (–80 to –120 mV) and a low membrane impedance (10 M
) [13]. Injection of current into these cells fails to generate an action potential, and studies with injection of fluorescent dye indicate that they are not electrically coupled [12, 13]. In addition to physically protecting the underlying sensory neurons, sustentacular cells contain detoxification enzymes including an olfactory-specific isoform of cytochrome P450 (P-450olf1) (phase I biotransformation enzyme) and UDP-glucuronyl transferase (UGTolf) (phase II biotransformation enzyme) at levels that rival their concentration in the liver [14, 15]. In phase I biotransformation, cytochrome P-450 catalyzes the hydroxylation of the substrate, and in phase II UGT catalyzes transfer of glucuronic acid from UDP-glucuronate to hydroxyl groups, converting a hydrophobic molecule into a hydrophilic, membrane-impenetrable one that is readily excreted. Beyond their obvious role in the destruction of inhaled toxins, it has been suggested that these enzymes play a role in the metabolism of odorants. Hydroxyl-containing odorants are substrates for UGTolf, and UGTolf is two- to fivefold more active than the liver enzyme on odorants [14]. Moreover, glucuronidation abolishes the activity of odorants in an in vitro assay for olfactory signal transduction [14]. Although to date no direct role in olfaction has been demonstrated for these enzymes, these data perhaps suggest a role in signal termination by shortening the tissue residence time of odorants and facilitating their removal through chemical modification. Mucus by itself has been proposed to concentrate the odorants, which is favored by their air-mucus partition coefficients [16].

101

102

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

5.1.4

Basal Cells

The third basic cell type in the OE is the stem cells or basal cells located at the base of the olfactory epithelium along the basement membrane. In contrast to mature neurons, basal cells do not express the cytoplasmic olfactory marker protein (OMP) [17] but do express N-CAM (a general marker for mature CNS neurons) [18]. Two types of basal cells have been described, which are referred to as globose basal cells and horizontal basal cells. Globose and horizontal cells can be distinguished ultrastructurally and immunohistochemically. Globose cells, as the name implies, have rounded nuclei. These cells stain lightly with toluidine blue and are located just one cell layer above the basal lamina. Horizontal cells can be distinguished from globose cells because they have elongated nuclei, stain darkly with toluidine blue, lie adjacent to the basal lamina, and are also the only cells in the OE that express cytokeratin [19]. Evidence that the basal cells are a self-renewing source of receptor neurons comes from the observations that both globose and horizontal cells undergo mitosis and incorporate 3H-thymydine, with 90 % of the labeled cells ending up in the receptor cell zone [4]. Bulbectomy (removal of the olfactory bulb, which is the target of the receptor neurons) increases the production of neurons and specifically enhances mitotic activity of the globose cells [19]. These data suggest that the globose cells are the direct precursors of new neurons. Using a replication-incompetent retrovirus that expresses human placental alkaline phosphatase, the fate of cells from both types of basal cells was determined to high resolution [20]. The results show that the globose cells are the major source of new olfactory receptor neurons and that some progeny of the globose cells divide, or transiently amplify, in the globose cell zone of the OE before migrating up to the receptor cell zone and differentiating into mature receptor neurons. This is consistent with observations in vitro of an immediate neuronal precursor that is keratin- and N-CAM negative, migratory, and non-neurite bearing [21], characterized as a transient amplifying cell dependent on fibroblast growth factors [22].

5.2

Recording Odor-induced Electrical Activity 5.2.1

The Electroolfactogram

The electroolfactogram (EOG) was one of the early electrophysiological experiments done with the OE [23]. An EOG is essentially a field recording with an electrode placed on the mucus near the cilia during the application of odorants. It is thought to represent a summation of the receptor neuron generator potentials and shows a monophasic negative-voltage transient involving sodium (but is unaffected by tetrodotoxin) and a requirement for extracellular calcium [24]. An EOG cannot be recorded from respiratory epithelium and is abolished when the epithelium is lesioned with zinc sulfate or Triton X-100 to remove the cilia [25]. Killing the receptor neurons by transection of the

5.3 Odorant Responses of Single Isolated Olfactory Receptor Neurons

olfactory nerve also abolishes the EOG [26]. Recovery of the EOG is accompanied by repopulation of the epithelium with receptor neurons and re-growth of cilia. The EOG amplitude is dependent on the concentration of odorant applied, with larger amplitude for higher concentrations. However, when odorant is repeatedly applied, there is a reduction in the relative amplitude. Later, recording from isolated ORNs, this was shown to involve Ca2+ influx reflecting adaptation (see below). Analysis of EOG responses indicates that many regions of the epithelium can respond to a given odorant, although some areas are activated to a greater extent than others [27–31]. Thus, it seems likely that some functional heterogeneity exists across the OE. How does the biochemical process of an odorant interaction on the surface of the cilia change the membrane potential of the sensory neuron and get converted into an electrical signal sent to the brain? Several events came together to explain this. Electrophysiologically, it had been observed that cAMP played a role in olfaction [32]. In these early experiments, Minor and Sakina recorded EOGs from explants of frog OE that they could place in a tube or gutter and over which Ringer solution flowed. They were able to record slow (odorant-like) depolarizations of the ORNs when membrane-permeable dibutyryl derivatives of cAMP were introduced into the flow of the Ringer. In addition, they could see odorant-potentiating effects with phosphodiesterase inhibitors and odorant-depressing effects with phosphodiesterase activators. Minor and Sakina even proposed what we now call a “signal transduction pathway” for odorant detection [32]: “Cyclic 3’, 5’- AMP is considered to play the role of mediator in the mechanism of excitation of the olfactory receptor; during interaction between an odiferous substance and the receptor, adenylyl cyclase is activated and the concentration of cyclic 3’, 5’-AMP increases; this, in turn, causes depolarization of the receptor cell membrane.”

5.3

Odorant Responses of Single Isolated Olfactory Receptor Neurons

A more detailed study of the olfactory response was made possible with the use of metal- or KCl-filled electrodes [29, 33, 34]. Action potentials, fired by individual olfactory receptor cells, could now be recorded extracellularly. A comprehensive study of odor-induced action potential responses was performed by Getchell and Shepherd [30, 35]. They reported that salamander olfactory receptor cells generate spikes in response to odor presentations; furthermore, a higher stimulus concentration led to an increase in spike firing frequency. Similar observations were made by many others [30, 36–39]. The maximal odor-induced spike firing frequency ranges from 20 to 40 Hz. But while the frequency monotonically increases with higher odor concentrations, the number of fired spikes does not show such uniform changes. At high odor concentrations, the spike train is often shorter than at lower concentrations, and fewer spikes are fired in response to higher odor concentrations [33, 39–41]. At these high concentrations the amplitude of individual spikes in an odor-induced train progressively decreases with time, until the spikes are not detectable anymore. This phenomenon was observed as soon as odor-induced spike trains could be recorded [33, 36, 37, 42–44] and has been

103

104

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

suggested to arise from progressive inactivation of voltage-activated Na+ channels during prolonged depolarization [39]. Not only can the spike amplitude decline during odor presentation, but also in some cases spikes reemerge with progressively increasing spike amplitude [30, 40, 42, 45]. For a 10-s odor exposure, an initial high-frequency burst is often followed by spiking of lower frequency, with the spiking ceasing temporarily or completely during the stimulation [30, 33, 46]. For stimulation durations lasting over 30 s, repetitive bursts of spikes have been reported that are driven by underlying oscillations of the intracellular voltage [47]. Seventy percent of frog ORNs respond with an oscillatory response pattern, with bursts of action potentials appearing around every 6 s [45], and every 1 s in mouse ORNs [48]. In addition to excitatory responses, odor-induced inhibitory responses in the form of a reduction of the basal spike frequency were recorded by Getchell and Shepherd (see also [29, 35, 38, 49]). Resting potentials of olfactory receptor cells were shown to be in the range of –40 to –70 mV by recording the intracellular potential either with intracellular electrodes or by means of the whole-cell current-clamp configuration [42, 43, 50–53]. Vertebrate olfactory receptor cells depolarize upon stimulation in a graded manner with associated generation of action potentials [42, 51, 54–56]. Alternatively, odor-induced hyperpolarization has been reported for the toad and the mudpuppy, which causes the above-mentioned reduction in basal spike rate [57, 58]. ORNs are electronically very compact cells and have a high input resistance of several G
[47, 52] Consequently, even a small depolarizing current can lead to a depolarization sufficiently large to generate action potentials [39, 47, 59–62]. Still, olfactory receptor cells are surprisingly quiet at rest, with a low basal spike rate of between 0.05 and 3 Hz [30, 36, 38, 42, 63, 64], an indication that the transduction mechanism itself is quiet when not stimulated. A more precise study of how the odor-induced receptor current depends on stimulus concentration was made possible in the late 1980s with the introduction of whole-cell patch-clamp techniques into olfactory research [51, 54, 65]. A detailed analysis has been done for olfactory receptor cells isolated from the tiger salamander [66] and exposed to odor for 1.2 s (see Fig. 5.2). When stimulated at low concentrations, the receptor cell responded with a transient inward receptor current, which terminated shortly after the end of the stimulation. An increase in stimulus concentration increased the response in a graded manner, and at high concentrations the receptor current outlasted the stimulation by several seconds. The amphibian dose-response relationship is narrow, and saturation of the receptor current occurs within a 10fold increase of stimulus concentration (see Fig. 5.2Ab and 5.2Bb). This is reflected by the average Hill coefficient of 3–4. The sigmoidal shape of the dose-response curves implies a high level of cooperativity in the response. Especially at low odor concentrations, olfactory receptor cells integrate over the stimulus presentation for up to 1.2 s; hence, longer stimulations lead to larger receptor currents [66, 67].

5.3 Odorant Responses of Single Isolated Olfactory Receptor Neurons

Fig. 5.2 Odor-induced responses in isolated salamander ORNs. Increasing concentration of acetophenone (A) or amyl acetate (B) were applied for 1.2 s as indicated by the solution monitor on top (S), and response families were recorded using the whole-cell voltage-clamp technique. The holding potential was –55 mV. The peak receptor current is plotted against the acetophenone (Ab) or amyl acetate (Bb) concentration and fitted with the Hill function (Hill coefficient n = 3, K1/2 = 17 lM and n = 4.2, K1/2 = 53 lM). Modified from [66] with permission from the Journal of Physiology, London

The simultaneous recording of receptor current and spike firing is possible using the suction pipette technique [68], where the cell body of an ORN is sucked into the tip of a recording pipette. The current recorded in this configuration has two components, the slow transduction current and the fast biphasic current spikes that represent action potentials. This technique was used to record odor-induced responses from mouse ORNs (Fig. 5.3). Compared with the responses of frog ORNs, the spike firing rate of mouse ORNs depends even more steeply on odor concentration, and the maximal spike firing rate is reached with only a threefold increase in odor concentration (Fig. 5.3C). Interestingly, this is an increase too small to saturate the receptor current, which still continues to rise even when spike firing has saturated (Fig. 5.3C) [48]. During longer odor exposures, the receptor current response is transient and begins to decline even in the presence of odor. At low to medium concentrations the receptor current can actually be reduced to values near zero [45]. For high concentrations, the receptor current increases quickly at the onset of stimulation, declines slowly thereafter, and remains present for the duration of the stimulus [45, 69, 70].

105

106

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

Fig. 5.3 The dose-response relation of a mouse olfactory receptor cell. (A) The suction pipette technique was used to record responses to the odor cineole (15) at the indicated concentration. (B) Rising phase of the receptor current during which action potentials are fired. (C) Dose dependence of the spike frequency, number of spikes fired, and the peak receptor current. Modified from [48] with permission of the Journal of Physiology, London

Olfactory receptor cells respond only when odor is delivered to the cilia [51, 71], and the response increases linearly with the length of the ciliary bundle exposed to the odor [68]. These findings demonstrate that the odor-sensing part of olfactory receptor cells is restricted to the cilia. A delay between odorant delivery and electrical response has been reported to be as short as 50 ms when measured with the suction pipette technique [68]. However, a range of 140 to 660 ms has been measured for different cells [41, 51, 65]. This time delay is consistent with the notion that the generation of the receptor current comprises several biochemical steps and that a second messenger system is involved in olfactory transduction.

5.4

Components of the Transduction Pathway

G-proteins are substrates for ADP-ribosylation. The adenylyl cyclase (AC) stimulating G-protein Gas is ADP-ribosylated by cholera toxin and the AC inhibiting G-protein Gai is ADP-ribosylated by pertussis toxin [72], as is the phospholipase C activating G-protein Gao. Pace and Lancet [73] were able to biochemically identify a Gas like

5.4 Components of the Transduction Pathway

Ca2+

Na+

R

Gα β γ

AC ATP cAMP CaMKII Ca2+

Ca2+/CaM

Ca2+ Cl-

Na+ Fig. 5.4 Schematic diagram of the olfactory signal transduction pathway. Odorants interact with a seven-helix receptor (R) activating a G-protein (G), which stimulates adenylyl cyclase (AC) to produce cAMP. cAMP opens the CNG channel, allowing Ca2+ influx. Ca2+ activates a Cl– channel, and Cl– efflux results in amplification of the initial depolarization. A Na+/Ca2+ exchanger uses the Na+ gradient to extrude Ca2+, helping to terminate signaling. Ca2+ also feeds back negatively to downregulate the CNG channel’s sensitivity to cAMP and the activity of AC by interaction with calmodulin (CaM) and CaM-kinaseII (CamKII), respectively

G-protein in OE cilia membranes 0.5 kDa larger than the liver Gas using cholera toxin and 32P labeled ADP-ribose. Pretreatment of the cilia membranes with cholera toxin decreased the odorant stimulation of AC in a dose dependent manner, establishing a relation between a Gas like G-protein and odorant stimulation of AC. Unlike the cholera toxin catalyzed ADP-ribosylated protein, the three proteins ADP-ribosylated by pertussis toxin were not OE specific. These data, and the previously described electrophysiological data, directed molecular studies to clone and characterize nearly all the members of a signal transduction pathway in which receptors of the seven helix class activate a Gas like G-protein which would stimulate an AC, increasing cAMP levels. Inturn the increase in cAMP concentration would then directly gate open a cyclic nucleotide-gated (CNG) channel, initiating depolarization of the ORN and the firing of action potentials. The receptor current later was shown to comprise a second inward conductance; the initial influx of Ca2+ through the CNG channel triggers a Cl– conductance, which amplifies the original predominantly Ca2+ driven depolarization. This pathway is diagrammed in Fig. 5.4.

107

108

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

5.5

Cloning of G Proteins Expressed in the OE 5.5.1

Gaolf

To clone cDNAs for G-proteins expressed in the OE, a library was screened with an oligonucleotide against a portion of an 18-amino-acid sequence common to all known G-proteins at the time. Clones for five different G-proteins were isolated. The majority of clones were Gas; the rest were Gao Gai1, Gai2, and the new species Gai3 [74]. In order to determine whether these cDNA clones represented mRNAs expressed in the ORNs, Northern blots were done with RNA from rats that had or did not have a bulbectomy. The results showed that none of the G-proteins had their signal reduced due to bulbectomy, which results in death of the ORNs by removal of their target, suggesting non-sensory cell expression of these genes in the OE [75]. A second screen done at low stringency was initiated using as a probe a mixture of the five G-proteins plus a degenerate sequence against a conserved region of the GTP-binding domain. One of five clones identified encoded a class that weakly hybridized to Gas. This new G-protein, named Gaolf, is 88 % identical to Gas at the amino acid level and shows robust expression in the OE. Gaolf can activate AC in S49 cyc– kin– cells, a cell line devoid of Gas, by incubation with GTP-c-S or AlF4–, potent activators of G-proteins. Not surprisingly, Gaolf is also activated as potently as Gas by cholera toxin. Gaolf can also couple to the b2 adrenergic receptor in transfected S49 cyc– kin– cells about as well as Gas [76]. Bulbectomy results in loss of Gaolf message, and peptide-specific antisera against Gaolf stain the OE cilia and axon bundles [77]. Ultrastructural/immunohistochemical electron microscopy studies using rapid freezing followed by rapid freeze-etching or freeze-substitution localize both Gaolf and Gas to the long, thin, distal portions of the cilia [78]. These data suggested that Gaolf and possibly Gas are the AC-stimulating Gproteins that mediate olfaction. The functional significance of Gaolf in olfactory transduction was convincingly demonstrated using mice that had disrupted expression of the protein. These animals lacked odorant-induced olfactory EOGs [79]. The exclusive expression of Gaolf in the OE, however, has been challenged with the report of Gaolf expression in the basal ganglia [80] and its implication in dopamine D1 receptor signaling through AC [81].

5.5.2

Adenylyl Cyclase

A wide variety of odorants were shown to stimulate cAMP production in the OE [82], and these increases were linearly correlated to the EOG amplitude evoked in response to odorant stimulation, an indication that signal transduction utilizes mainly cAMP as the second messenger [83]. In addition, adenylyl cyclase activity was shown to be highly enriched in olfactory cilia [84], which led to immunoidentification [85] and finally cloning of an adenylyl cyclase from the rat OE [86]. Bakalyar and Reed screened a

5.6 Odorant Receptors

rat olfactory cDNA library at low stringency with a mixture of coding sequence from the type I and type II adenylyl cyclase enzymes. Two overlapping clones were combined to generate a cDNA with an open reading frame that encodes 1144 amino acids. This enzyme is referred to as adenylyl cyclase type III (AC III). The deduced amino acid sequence appears topologically similar to the 1134-amino-acid type I enzyme. Both proteins have two hydrophobic regions: one near the NH2 terminus and the other between amino acid residues 600 and 850, each containing six potential transmembrane regions. Northern blot analysis indicated that AC III mRNA is enriched in the OE and that AC III mRNA disappeared after bulbectomy. Similarly, protein expression on the cilia reduced concomitantly with bulbectomy as detected by staining with a polyclonal antibody. Immunoelectron microscopy also localized ACIII to the cilia [87]. When expressed stably in HEK293 cells, AC III had almost no basal activity (4.70.1 pmol min–1 mg protein–1), as compared to control (4.00.3 pmol min–1 mg protein–1). In contrast, AC I has a relatively high basal activity (125.68.8 pmol min–1 mg protein–1). A similar level of basal activity is observed for AC II under these conditions. When the various AC isoforms were stimulated with forskolin in vitro, AC III proved to be most active of the three. The functional basal concentration of cAMP in ORNs has been shown to be just below that required for opening of the CNG channels [88]. This is probably an equilibrium set at rest in the cells between the activities of AC III and phosphodiesterases. It has been proposed that this can act to set the cAMP concentration at or very near threshold for activation of the transduction pathway [88]. During long odor exposures lasting several seconds, AC III activity can be downregulated by Ca2+/calmodulin kinase II (CaMK II)-dependent phosphorylation [89, 90]. Inhibition of CaMK II prevents adaptation of the odor-induced response [91]. A Ca2+/calmodulin-dependent phosphodiesterase is also present in olfactory cilia and is proposed to be involved in response termination [92, 93]. As with Gaolf, animals engineered to be deficient in expression of AC III lack odorant-induced olfactory EOGs and nurse poorly or fail olfaction-based behavioral tests [94].

5.6

Odorant Receptors

Single-cell spike recordings [27] demonstrate that response selectivity of ORNs is broad. Whole-cell recording experiments indicate that there is roughly a 50 % chance of finding a cell responsive to a mix of three odorants, and 50 % of those cells will respond to more than one of the three odorants [66]. However, although they appear broadly tuned, they do show specificity even when tested with high or low concentrations of an ineffective odorant. These characteristics, of being responsive to more than one stimulus, and having a relatively low sensitivity differ from those of photoreceptors and hormone receptors. Such distinctions likely reflect fundamental differences between these superficially similar systems, suggesting that the OE acts as a low-specificity detector array. In fact, a controversy has developed [69, 95, 96] over the suggestion that, as photoreceptors can signal detection of single photons [97], ORNs can

109

110

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

signal binding of single odorant molecules, i.e., odorants can induce quantal events (currents) in the ORNs [70]. Menini et al. recorded in the whole-cell configuration from salamander ORNs with low concentrations of odorant (0.5 lM cineole) for long periods of time (>30 min). Their claim that small current responses seen during these recordings correspond to quantal evoked-odor activation of the above-outlined transduction cascade is not substantiated by the data. The biggest problem with their interpretation of this data as representing quantal responses is that their type of analysis assumes a linear relationship between current and the number of quantal events. As outlined above, the transduction current is highly nonlinear, which critically undermines an estimation of quantal-event amplitudes. A very large gene family of closely related olfactory-specific, seven-helix receptors has been identified using a polymerase chain reaction (PCR) probe. The probe was generated by amplification of olfactory epithelial cDNA with oligonucleotides based on seven-helix receptor sequences from transmembrane regions (TM) II and VII from hormone receptors [98]. The original 10 complete cDNA sequences and 8 partial sequences cloned from rat have been expanded through genome-mining techniques such that the full complement of receptor genes from mouse (approximately 1300, of which 20–30 % are pseudogenes) [99] and human (approximately 900, of which 60–70 % are pseudogenes) [100, 101] are known and available on hybridization matrix chips (at least for mouse).

5.7

Cyclic Nucleotide-gated Channel in OE

More than 10 years after the EOG studies of Minor, a cGMP-gated conductance mediating phototransduction in rod and cone outer segments was discovered [102, 103]. These data led Nakamura and Gold [104] to do patch-clamp experiments on the cilia of toad olfactory neurons and look for a cAMP-gated conductance. In excised patches, their results directly demonstrated an outwardly rectifying, nonselective cation conductance with little sensitivity to voltage (in the absence of divalent cations) that could be gated by cAMP (concentration for half maximal activation, K1/2, of 2.4 lM). Unexpectedly, all the patches also showed a 1.7 higher sensitivity to cGMP (K1/21.6 lM). The Hill coefficients were 1.8 for cAMP and 1.7 for cGMP, indicating cooperativity of ligand binding. They proposed that an odorant-stimulated increase in cyclic nucleotide concentration has a direct activating effect on a cation conductance that initiates a depolarizing response to odorants. Nakamura and Gold added the caveat that this model does not apply to odorants that do not appear to stimulate adenylyl cyclase [82] or to those that stimulate polyphosphoinositide turnover [105]. Besides cAMP, IP3 and cGMP have been implicated as second messengers in olfactory transduction. For further discussion on these other pathways, we would like to point the interested reader towards two comprehensive reviews [95, 106]. Following the pioneering work of Nakamura and Gold, several whole-cell and excised patch recording studies were conducted [107–110], further characterizing the cAMP-gated channel in terms of its pharmacology and activation behavior. The chan-

5.7 Cyclic Nucleotide-gated Channel in OE

nel densities have been determined to be 8 per lm2 [111] or 70 per lm2 [112] in rat and frog, respectively, although much higher values have been reported (see [113, 114]). One of the most detailed studies [107] characterized the CNG channels in the dendritic knob, dendritic stalk, and soma of ORNs from rat and frog. The characteristics determined were rectification, activation by cyclic nucleotides, selectivity for monovalent cations, inhibition by cytosolic acidification, and inhibition by organic blockers. Summarizing their results, it was determined that the single-channel conductance, in the absence of divalent cations, was 8–35 pS for rat and frog (see also [111, 112, 114, 115]). There was a graded increase in patch current with the application of cAMP or cGMP to the cytosolic face of the membrane, saturating at 10–30 lM. In agreement with Nakamura and Gold [104], the K1/2 for cAMP was 2.5 lM (+50 mV) and 4 lM (–50 mV), with a Hill coefficient (n) of 1.8 for rat ORNs. For cGMP the values were (rat) K1/2 1.0 lM (+50 mV), 1.8 lM (–50 mV), n 1.3 (Fig. 5.5). There was weak outward rectification that was strongly dependent on cyclic nucleotide concentration and influenced by membrane voltage. There was no rundown of the current, and the conductance remained constant over the course of 30 min. The selectivity for monovalent cations was determined by applying voltage ramps in the presence or absence of saturating concentrations of cAMP (40 lM) with an extracellular concentration of 150 mM Na+ in the pipette and successive exposure of the intracellular membrane to similar concentrations of Na+, K+, Li+, Rb+, or Cs+. The different reversal potentials revealed by the I-V relationships under each biionic condition yielded the following sequence of permeation ratios relative to Na+: Na+(1):K+(0.81):Li+(0.74):Rb+(0.74):Cs+(0.52). The values from rat channels differ slightly form those measured for newt: Na+(1): Li+(0.93): K+(0.93): Rb+(0.91): Cs+(0.72) [116]. The fact that the current carried by Li+ is smaller than the Na+ current, despite the equilibrium selectivity for Li+, is an anomalous permeation property that has also been observed for the cGMP-gated channel in rod photoreceptors [117]. Reducing the pH of the solution on the cytosolic side of the inside-out patch from 7.0 to 5.0 inhibited the cAMP-induced current by 60 % and completely blocked it at pH 4 in 18 s (solution change <100 ms). This effect was titrated (at +60 mV) and was found to have a pK of 5.0 but showed no voltage dependence and was reversible. No such effect was observed when the solution on the outside of outside-out patches was acidified, and the effect was independent of cAMP concentration, indicating that the titratable group interacts with residues involved in permeation rather than in ligand binding. It is unlikely that acidification of this magnitude occurs in the ORNs (in photoreceptor cells cytosolic pH does not change by more than 0.002 units during light excitation) [118], but these results indicate the presence of negatively charged residues near the channel mouth, with possible effects on ion permeation. Divalent-free solutions are not physiological. The inclusion of physiologically relevant concentrations of divalent cations greatly alters the current recorded from olfactory CNG channels [119]. Addition of Ca2+ or Mg2+ to the external face of the channels induces a rapid flickering of the open channels, thus reducing the macroscopic current activated by saturating concentrations of ligand. This effect was found to be voltage dependent, increasing steadily with hyperpolarization with both Ca2+ and Mg2+, and the I-V relationship of the channel is strongly outwardly rectifying [119, 120]. However,

111

112

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

the effect differs between the two ions around –70 mV, where with Ca2+ the block begins to decrease with further hyperpolarization, indicating permeation of Ca2+ [119] (see below). This is also a trait of other Ca2+-selective channels such as the voltage-gated Ca2+ channels and NMDA receptors. Importantly, Frings and colleagues [120, 121] have shown by simultaneous whole-cell recording and fura-2-based Ca2+ imaging that the olfactory CNG channel functions as a Ca2+ channel in the presence of physiological levels of extracellular Ca2+, with >70 % of the fractional current (Pf) being carried by Ca2+ at (Ca2+)o = 3 mM. Thus, much of the CNG channel’s contribution to the receptor current is the divalent cation Ca2+ and not monovalent Na+ flux. The reduction of Na+ flux through the CNG channel also prevents Na+ from accumulating in the small ciliary space, which is important for maintaining the Na+ gradient across the cilia [122]. When the effects of several organic blockers (amiloride, the phenylalkylamine D600 [methoxyverapamil], and the benzothiazepine diltiazem) were assessed, it was found that all could block to some degree from the inside. Some data suggest that amiloride can also block from the outside [47]. There is a question about whether this effect might actually be due to amiloride getting to the inside via the basolateral membrane of the cell, based on the long latency of effect. Similarly, data on block of the retinal rod and cone CNG channels by extracellular and intracellular l-cis-diltiazem are consistent with a single binding site on the cytoplasmic side of the channel [123]. There is no data on external block of the olfactory CNG channel by phenylalkylamines or benzothiazepines. In excised patches from rat or frog ORNs, the compounds showed a voltage-dependent block; there was a 10–12 higher sensitivity at more positive membrane potentials, presumably due to the positively charged compound being driven into the channel. In patches from frog ORNs, the Ki for amiloride decreased from 400 lM at –100 mV to 17 lM at +100 mV. The Ki at +60 mV for amiloride on the rat channels was 71 lM. This is two orders of magnitude higher than that of Na+ epithelial channels [124] (Ki 0.1–0.5 lM) but is similar to Na+/H+ exchangers [125] and T-type Ca2+ channels [126] (3–80 lM). On the other hand, the olfactory CNG channels are more sensitive to amiloride than are other Na+ transporters. For instance, the Na+/Ca2+ exchanger Ki is 1 mM [127], Na/K-ATPase Ki is 3 mM [128], Na-glucose transporter Ki is 2 mM [129], and voltage-sensitive Na+ channels Ki is 0.6 mM [129]. In parallel with their similar sensitivity to amiloride, both the CNG channel [130] and the T-type Ca2+ channel [136] have an even higher sensitivity to the amiloride derivative 3’-4’-dichlorobenzamil, suggesting a possible relationship between these channels. D600 and diltiazem, other Ca2+ channel blockers, also block the olfactory CNG channels in a voltage-dependent manner but have Ki values up to two orders of magnitude higher than on Ca2+ channels. Frings et al. [107] determined Kis for D600 of 200 lM at –20 mV and 12 lM at +100 mV for the CNG channels from frog ORNs. The ability of diltiazem to block the CNG channels showed stereospecificity, with the l-cis enantiomer being more potent (Ki 70 lM) than d-cis diltiazem (Ki 200 lM) at +50 mV. This same stereospecificity is also true for the photoreceptor CNG channel [131]. Ca2+channels also show stereospecific diltiazem block, but the d-cis enantiomer is more potent than the l-cis one. The diltiazem stereoselectivity implies that this block of the CNG channels is not “nonspecific” but rather involves

5.7 Cyclic Nucleotide-gated Channel in OE

Fig. 5.5 Activation by cyclic nucleotides of native rat CNG channels. I-V curves from rat knob membrane patches with different concentrations of (top left) cAMP or (botton left) cGMP, with symmetrical divalent-free solutions. (top right) dose-response relationships of activation by cyclic nucleotide derived from I-V curves. Modified from [107] with permission of the Rockefeller University Press

interaction with a binding site on the cytoplasmic side of the channel. For photoreceptor CNG channels, diltiazem binding and interaction with cGMP appear to involve separate domains; when purified photoreceptor channels incorporated into artificial membranes are digested with trypsin (producing a 63-kDa product), they retain activation by cGMP but lose sensitivity to l-cis diltiazem [132].

113

114

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

5.8

Cloning of a CNG Channel Expressed in the OE

Cloning of the rat olfactory CNG channel (rOCNC1, CNG2, and CNGA2) [133, 134] was facilitated by the earlier cloning of the bovine retinal rod CNG channel (bRCNC1, CNG1, and CNGA1), which followed a biochemical approach of purification and amino acid sequencing of protein from bovine retina [135] (a confusing use of pet names for these subunits has been clarified to an accepted nomenclature [136]). To isolate clones of the olfactory CNG channel, Dhallan et al. screened an olfactory cDNA library at low stringency with the retinal bCNGA1. The resultant rat clone, rCNGA2, was shown to be a nonspecific cation channel gated by cAMP or cGMP. cDNAs from bovine and catfish OE encoding a homologous channel with similar function were also cloned [134, 137]. Topologically, the channel resembles the pore-forming subunit of voltage-gated potassium channels, with six transmembrane domains, a signature pore-forming sequence, and an 80- to 100-residue intracellular carboxy terminal domain resembling the cyclic nucleotide-binding domain of the regulatory subunit of protein kinase A and the catabolite activator protein (CAP) of bacteria (for a more detailed presentation of CNG channel structure function analysis, see [138]). Although CNGA2 formed functional CNG channels when expressed heterologously, there was a discrepancy in the rectification properties and sensitivity to cAMP relative to the native channel. Subsequent cloning and expression analysis demonstrated the existence of a “modulatory” subunit of the rat olfactory CNG channel, rCNGA4 [139, 140] (see alignment in Fig. 4.4A). Modulatory is a functional label because when expressed alone, CNGA4 does not form functional CNG channels (but see [141]). Heterologous expression of CNGA2 together with CNGA4 produced heteromeric channels more similar in cAMP sensitivity to native channels but still about threefold less sensitive. Sometime later, yet a third subunit, rCNGB1b [142, 143], was definitively characterized to be a component of the native channel [142]. Heterologous expression of rCBGB1b together with rCNGA2 and rCNGA4 further increased cAMP sensitivity, essentially mimicking that of the native channel. In addition rectification ratios, the presence of a subconductance state, and co-localization of expression in situ all argue strongly that the native channel is a heteromeric complex composed of a combination of CNGA2, CNGA4, and CNGB1b [124] (see Fig. 5.6). Unclear is the stoichiometric relationship of the subunits in the supposed tetrameric channel, although it has been suggested that the stoichiometry is (CNGA2)2-(CNGA4)(CNGB1b) [144].

5.9 Negative Feedback by Ca on the CNG Channel Fig. 5.6 Ligand sensitivity of native and heterologously expressed olfactory CNG channels. Dose-response relations fit to the Hill equation for the activation of macroscopic currents by cAMP at +40 mV. Reproduced from [142] with permission of the Society of Neuroscience

5.9

Negative Feedback by Ca2+ on the CNG Channel

Ca2+ not only is a component of the receptor current but also acts to terminate the response and control adaptation. An odor-induced receptor current, which is transient and adapts in Ringer, fails to adapt when Ca2+ influx through the CNG channels is minimized by removing external Ca2+ [145, 146] or when internal Ca2+ buffering is increased using BAPTA-AM [91]. Olfactory cilia contain the Ca2+-binding protein calmodulin at a concentration of 1 lM [147]. In retinal rod photoreceptors, it has been shown that Ca2+ bound to calmodulin (Ca-CaM) reduces the cGMP sensitivity of the native rod CNG channel from 19 lM to 32 lM, effectively reducing the cation influx by two- to sixfold [148]. Therefore, when light stimulates hydrolysis of cGMP and the channel closes, Ca2+ influx is reduced, while the Na+/Ca2+/K+ co-transporter continues to extrude Ca2+ and internal Ca2+ drops. This relieves the inhibitory binding of Ca-CaM and can promote recovery by stimulating the channels to rebind cGMP. The prominence of this feedback in phototransduction, however, is not considered to be major [149]. In olfaction, feedback inhibition by Ca-CaM on the CNG channel is more important. It was shown that Ca2+ could act directly on the native olfactory CNG channel in excised patches to reduce the open probability, with a threshold for inhibition

115

116

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

Fig. 5.7 Effect of Ca-CaM on activation by cyclic nucleotides of native rat CNG channels. I-V curves from rat knob membrane patches with different concentrations of cAMP in symmetrical divalent-free solution in the (A) absence or (B) presence of Ca-CaM. (C) Doseresponse relationships (fit to the Hill equation) of activation by cyclic nucleotide derived from I-V curves. From [205] with permission

around 100 nM Ca2+ [146]. Later, using a two-pulse odor protocol or with flash photolysis of “caged cAMP” (the latter generates a receptor current directly by gating the CNG channel without activation of the G-protein-coupled cascade), Kurahashi and Menini [67] showed that for short (100 ms) activations, adaptation can be explained solely by negative feedback by Ca2+ on the CNG channel. Thus, the odorant-induced intracellular elevation of Ca2+ is thought to promote adaptation because Ca-CaM can directly reduce the sensitivity of the native CNG channel for cAMP (Fig. 5.7) [150–152]. The principal subunit CNGA2 has been extensively characterized as a binding target of Ca-CaM. Expressed homomeric CNGA2 channels are inhibited by cytoplasmic application of Ca-CaM [150, 153–156] (see [157] for review) and the rate of channel inhibition by Ca-CaM is limited by the rate of Ca-CaM binding [151]. Co-expression of the modulatory subunits CNGA4 and CNGB1b together with CNGA2, however, increases the rate of inhibition by Ca-CaM by two orders of magnitude [151]. These data beg the question of how it is that CNGA2 should

5.9 Negative Feedback by Ca on the CNG Channel

be the binding target of Ca-CaM [150, 153–156] yet CNGA4 and CNGB1b control the rate of inhibition/binding of Ca-CaM [151]. The identified calmodulin-binding site on CNGA2 [153] is a classic basic amphiphilic alpha helix (Baa motif) [158] regulating channel inhibition in a strictly Ca2+-dependent manner by binding only a complex of Ca-CaM [153–155]. Bradley et al. [159] tested the relevance of this site in mediating Ca-CaM inhibition of CNGA2-A4-B1b channels. Quite surprisingly, they found that this site has no role in this process (Fig. 5.8, top). In the absence of a role for CNGA2 in inhibition of native channels by Ca-CaM, the question then becomes how does calmodulin associate with the native channels, and by what mechanism does calmodulin facilitate Ca2+-dependent feedback inhibition? To answer these questions, CNGB1b was first examined. The gene cngb1 is expressed as a number of splice variants in retina, sperm, and olfactory epithelium (see [138] and references therein). Variant CNGB1a is expressed in retinal rod photoreceptors and contains two calmodulin-binding sites (CaM1 and CaM2). CaM1, termed an “unconventional” CaM-binding site, is located in the NH2-terminus of CNGB1a and confers a weak ability of Ca-CaM to alter cGMP sensitivity of native heteromeric rod CNGA1B1a channels [160, 161]. CaM2, a Baa motif, resides in the COOH-terminus of CNGB1a and has not been ascribed any function so far [160, 161]. In CNGB1b, the variant of cngb1 expressed in ORNs, a large NH2-terminal glutamic-acid-rich domain of CNGB1a, is replaced by 74 unique amino acids [142, 143]. Otherwise, CNGB1a and CNGB1b are identical. Accordingly, the CaM1 and CaM2 sites de-

Ca-CaM

Inorm

1

wt A2∆CaM-A4-B1b

0 0

5

10 time (s)

15

20

Ca-CaM A2-A4(L292E)-B1b A2-A4-B1b(L183E)

1

Inorm

Fig. 5.8 Analysis of heteromeric channel inhibition by Ca-CaM in excised inside-out patches. Top: deletion of the Baa-type Ca-CaM binding site in CNGA2 has no effect on heteromeric channel inhibition by Ca-CaM. Comparison of A2-A4-B1b channels (black trace) and A2(CaM)A4-B1b channels (gray trace). Bottom: only the IQ-type CaM binding sites in CNGA4 and CNGB1b are necessary for Ca-CaM inhibition of heteromeric channels. Deletion of the Baa Ca-CaM binding site, CaM2, in CNGB1b has no effect on heteromeric channel inhibition by Ca-CaM (black trace). Singlepoint mutations in the IQ-type CaM binding sites in CNGA4 or CNGB1b completely abolish heteromeric channel inhibition by CaCaM (dark gray and light gray traces, respectively)

0

A2-A4-B1b∆CaM2 0

20

40 time (s)

60

80

117

118

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

scribed for CNGB1a are also part of CNGB1b. Co-expression of CNGA2 and CNGA4, with a CNGB1b mutant lacking CaM2, did not alter the inhibitory effect of Ca-CaM (Fig. 5.8, bottom, black trace, A2-A4-B1b D CaM2). Between residues 183 and 193 of CNGB1b (LQELVKMEKER), CaM1 resembles a “generalized” IQ-type CaM-binding motif ({I,L,V}QxxxRxxxx{R,K}) [161], understood to mediate binding of Ca2+-free calmodulin [162, 163] (apocalmodulin). Mutation of a single key residue [164] (L183E) in this site specifically eliminated Ca-CaM inhibition of heteromeric CNG channels (Fig. 5.8, bottom, light gray trace, A2-A4-B1b(L183E)). Thus, the integrity of a binding site for apocalmodulin in CNGB1b is necessary for Ca-CaM inhibition of native channels. What about CNGA4, which had also been shown to be necessary for fast inhibition of native channel by Ca-CaM [151]? An IQ-type motif (LQHVNKRLERR), very similar in sequence to that of the CaM1 site in CNGB1a/b, located in CNGA4 between the sixth transmembrane domain (S6) and the cAMP-binding site in the cytoplasmic COOH-terminus was found. This region, termed the “C-linker,” is thought to be an integral part of the gating machinery of CNG channels (for review, see [138]). A single amino acid substitution (L292E) in this IQ site rendered heteromeric channels completely insensitive to inhibition by Ca-CaM, (Fig. 5.8, bottom, medium gray trace, A2-A4(L292E)-B1b). Heteromeric channels with mutated IQ motifs in both CNGA4 and CNGB1b, but with the CNGA2 Baa motif intact (CNGA2-A4(L292E)B1b(L183E)), also failed to respond at all to Ca-CaM, supporting the conclusion that CNGA2 has no function in channel inhibition [159]. In olfactory CNG channels, CNGA4 and CNGB1b serve to increase cAMP sensitivity, so that a few lM cAMP are enough to open the heteromeric channels [142, 143]. Calmodulin appears to function as a cAMP sensitivity switch for the channel, blocking the effects of either CNGA4 or CNGB1b, or both, when triggered by Ca2+.

5.10

The Olfactory Ca2+-activated Cl– Channel

From a historical perspective, olfactory transduction was seen as a system that worked analogously to vertebrate phototransduction: activation of a G protein-coupled receptor triggers an electrical response by generating a current through a CNG channel. This view was altered when Kleene and Gesteland [165] reported the presence of a Ca2+activated Cl– conductance in olfactory cilia. This finding provided an explanation for a long-standing observation that the olfactory response persists even in the absence of external Na+ [166–169]. Three questions were raised: what is the Ca2+ source to gate the Cl– channel, when is the conductance opened, and is the ensuing Cl– current excitatory or inhibitory? Apparently olfactory cilia do not contain Ca2+ stores because application of pharmacological agents that cause Ca2+ release from intracellular stores fail to generate ciliary Ca2+ transients [170]. Additionally, Ca2+ signaling in olfactory receptor cells is regulated separately in the cilia and in the cell body, which allows the generation of compartmentalized Ca2+ transients [171]. This eliminates the cell body as a source of Ca2+

5.11 Activation of the Cl– Conductance

in the cilia. The origin of ciliary Ca2+ has been shown to be exclusively the CNG channel. Odor stimulation increases the ciliary Ca2+ concentration by opening CNG channels [171, 172], and this primary Ca2+ signal generates an additional secondary Cl– current. This current is inward, and therefore excitatory, and carries 80 % of the odor-induced current in rodents [173] and 36–65 % in amphibians [174, 175]. Two functions for this unusual Cl– current have been proposed. First, it provides a nonlinear, low-noise amplification of the small primary Ca2+ current [173, 176]. Second, it reduces the dependency of the receptor current on external mucosal Na+, which might vary depending on the environment, a situation more relevant for amphibians [174] and fish [177] than for terrestrial mammals. Many of the characteristics of the odor-induced response therefore depend on the interplay between the CNG channel (the Ca2+ source) and the Cl– channel and on their regulation. Furthermore, since spike frequency is a function of the receptor current, the interaction of these two channels will affect action potential generation (see above). To support the excitatory Cl– current across the ciliary membrane, the Cl– distribution in the cilia and the mucus has to be such as to allow the Cl– reversal potential to be positive with respect to the resting potential of the cell. Using electrophysiological methods, the Cl– reversal potential has been shown to be around 0 mV in Xenopus [175] and –45 mV in mudpuppy [178]. Alternative approaches are to measure the Cl– distribution directly using fluorescent Cl– sensitive dyes, energy-dispersive xray microanalysis, or ion-selective microelectrodes. With these methods intracellular Cl– concentrations were found to be high and to range from 40 to 80 mM [179–181]. Furthermore, the Cl– concentration in the mucus is lower than in interstitial fluid (93 mM in toad mucus and 55 mM in rat mucus [179, 182]), thereby aiding Cl– efflux from the cilia. After being first discovered in frog cilia [165], the Ca2+-activated Cl– current in olfactory receptor cells has since been found in newt [174], Xenopus [175], salamander [183], trout [177], mouse [48], and rat [111, 173, 184]. In all species the Cl– current is excitatory, underscoring its importance in olfactory transduction. Interestingly, since the first (electrophysiological) description of the Cl– channel in 1991, its molecular identity has remained elusive.

5.11

Activation of the Cl– Conductance

In their instrumental description of the Ca2+-activated Cl– conductance, Kleene and Gesteland [165] sealed whole frog olfactory cilia into patch pipettes and obtained current-voltage relationships (Fig. 5.9, top). Increasing Ca2+ concentrations were applied to the cytoplasmic side of the cilia, which gave rise to increasing currents with a K1/2 for Ca2+ of 4.8 lM and a Hill coefficient of 2 (Fig. 5.9, bottom). On average the maximal Cl– current of a single cilium was –50 pA at –50 mV [176], with average ciliary lengths of 47 lm. Surprisingly, the maximal current is not correlated with ciliary length [185]. In symmetrical NaCl solution, the current reversed close to 0 mV and showed slight inward rectification, as is the case in the rat Cl– channel [111]. When internal Cl– was

119

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

+400

300 150 4,10 3 2 1 0

pA 100

mV +100

0-2 3 4 10 150 300

400

100 Normalized conductance

120

80 60 40 20 0 0.1

1

10

100

2+

[ Ca ] i , uM Fig. 5.9 Activation of the frog ciliary Cl– conductance by Ca2+. Top: a frog ORN cilium was sealed inside a patch pipette and excised from the dendritic knob. The cytoplasmic side was exposed to increasing Ca2+ concentrations, and current-voltage relations were recorded (Ca2+ concentration in lM next to each I-V relation). Bottom: Ca2+ dependence of the Cl– conductance. The conductance was measured between 0 and –50 mV and normalized to the value obtained at 300 lM Ca2+ (data of seven cilia). A Hill fit (solid line) yielded a Hill coefficient of 2 and a K1/2 of 4.8 lM. Reproduced from [165] with permission of the Society of Neuroscience

5.12 Single Channel Properties and Channel Densities

increasingly replaced by gluconate–, the reversal potential shifted to values that closely followed the predictions of the Nernst equation for a Cl– channel [165], confirming Cl– as the charge carrier. Replacing Na+ with choline+ or Tris+ did not alter the Cl– current magnitude [165], and Na+ hardly permeates the channel (PNa/PCl = 0.034 [184]). Membrane patches excised from dendritic knobs of rat olfactory receptor cells were used to study mammalian Cl– channels [111, 184]. When patches were held at –40 mV and rapidly exposed to Ca2+ for 10 s, Cl– currents quickly activated then slowly inactivated by 40 % over 10 s. The current magnitude depended steeply on the Ca2+ concentration, and fitting the Hill function to the peak current yielded a K1/2 of 2.2 lM and a Hill coefficient of 2.8 [111], values close to the ones found in the frog [165]. Nonneuronal Ca2+-activated Cl– channels display similar Ca2+ sensitivities (for review, see [186]), while a higher K1/2 of 26 lM with a low Hill coefficient (n) of 1 has previously been reported for the rat Cl– channel [196]. External Ca2+ does not seem to affect the Cl– current magnitude [111]. Activation by Ca2+ was voltage dependent, and at +40 mV a lower K1/2 (1.5lM, n = 3) [111] was observed, as has been the case for Ca2+-activated Cl– channels in, e.g., Xenopus oocytes [187, 188], rat parotid acinar cells [189], or smooth muscle cells [190]. Interestingly, sensitivity of the frog Cl– channel seems not be voltage dependent [165], nor has it been demonstrated previously in the rat [184]. Mg2+, which is found at high concentrations inside cells, cannot substitute for Ca2+ as the activating ion. Concentrations as high as 2 mM, or even 10 mM, did not elicit a Cl– current in patches excised from frog or rat ORNs [111, 186], which indicated that Mg2+ does not play a physiological role in gating the Cl– channel. Ba2+ activated the channel only poorly, while 1 mM Sr2+ evoked 80 % of the maximal Ca2+-activated current [111, 191].

5.12

Single Channel Properties and Channel Densities

Both the frog and the rat Cl– channel have a conductance too small to be observed directly as single channels in excised patches; therefore, noise analysis was used as the investigative tool. Larsson et al. [112] excised whole frog olfactory cilia and monitored the current noise associated with the macroscopic currents elicited by increasing Ca2+ concentrations applied to the cytoplasmic side of the cilium. Conscientiously, they took into consideration that, since they recorded from a long slender cilium, voltage clamp will be imperfect. They obtained a single-channel conductance of just 0.8 pS, with a maximal open-probability of 1 [112, 176]. With the known length and diameter of the cilium, the channel density was calculated to be around 80 channels per lm2. The single-channel conductance in the rat was determined by making use of the observation that in some excised patches the Cl– current declined slowly after removal of cytoplasmic Ca2+ application, probably due to slow solution exchange between the patch and the superfusate. Current and noise levels during the slow falling phase could therefore be determined, and noise analysis yielded a single-channel conductance of

121

122

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

1.3 pS with a maximal open-probability of nearly 1 [111]. This range of single-channel conductance from 0.8 to 1.3 pS for the frog and rat channels is similar to values reported in smooth muscle [190, 192–194]. The Cl– channel density in rat cilia was not determined by dividing by the plasma membrane area. Instead, the theory of buffered Ca2+ diffusion was used. Here, an excised patch was held in symmetrical cholineCl solution with 1 mM CaCl2 in the pipette solution. Application of cAMP to the cytoplasmic side of the patch yielded opening of the CNG channel and to Ca2+, but not to choline+, influx, since the latter is impermeant. The incoming Ca2+ will radially diffuse away, and the Ca2+ concentration will decrease with increasing distance from the source (the CNG channel). The speed of decline is determined by the cytoplasmic Ca2+ buffer in the superfusate of the patch and can be calculated using the theory of buffered Ca2+ diffusion [195, 196] (which describes the diffusion of Ca2+ in the presence of buffers). Depending on the distance between the CNG and a Cl– channel, both of which are present in the excised ciliary membrane, the Cl– channel will “see” a certain Ca2+ concentration and generate a Cl– current governed by its Ca2+ sensitivity. With a previously calculated Ca2+ doseresponse curve as a standard, the level of Ca2+ can be determined, which in turn allows calculation of the distance between the CNG and the Cl– channels. With this method the rat Cl– channel density was found to be 62 channels per lm2 [111], which is similar to the frog [112].

5.13

Regulation of Cl– Channel Activity

Some ion channels permeable to Cl– have their open probability and sensitivity to Ca2+ up- or downregulated by various kinases such as Ca2+/calmodulin-dependent protein kinase II (CaMK II) or PKC (for a comprehensive review, see [186]). Little is know, however, about such regulation of the olfactory Ca2+-activated Cl– channel. In excised membrane patches from rat ORNs the Cl– current reduces with time, or “runs down,” and has been attributed to the progressive loss of functional channels [111], which has been described in a variety of other tissues (for review, see [197]). The cause of the rundown is not understood, but it could be the loss of a soluble factor or dephosphorylation. On the other hand, rundown has not been observed for the frog Cl– channel recorded in excised cilia [165, 198] Ca2+/calmodulin plays an important role in regulating the sensitivity of the CNG channel (see 5.9). It is therefore interesting to investigate whether calmodulin affects the Cl– channel. Neither the frog nor the rat Cl– channel in an excised cilium patch showed any change in current level when calmodulin was added to the cytoplasmic solution [111, 199]. This indicates that cytosolic calmodulin does not regulate the channel directly but does not rule out that it might do so indirectly via, e.g., CaMK II, as is the case for AC III. Termination of the odor-induced response requires the shutoff of all steps in the transduction cascade to allow the Cl– channel, the final step, to close. Moreover, the Ca2+ that accumulates in the cilia and keeps the Cl– channel open has to be removed. A

5.14 Amplification of the CNG Current and Generation of the Cl– Current

Na+/Ca2+ exchanger is present in olfactory receptor cells [200, 201] and is responsible for ciliary Ca2+ removal and response termination [202]. In Na+-free solutions (i.e., Na+ replaced by Li+) Na+/Ca2+ exchange cannot function and the return to pre-stimulus Ca2+ levels is slowed. Consequently, a prolonged current is generated which is carried by Cl– [45, 202]. This demonstrates an interesting role reversal for Na+ relative to its “normal” electrophysiological function; Na+ is not essential for the generation of the receptor current but rather is necessary for the termination of the response. A Ca2+ATPase has also been described in a plasma membrane-rich fraction of olfactory epithelium from salmon [203].

5.14

Amplification of the CNG Current and Generation of the Cl– Current

The gating of Cl– channels is dependent on the intraciliary Ca2+ concentration, which in turn depends on the opening of CNG channels and intraciliary cAMP concentration. The interplay of these two channels in the ciliary membrane (in conjunction with the Na+/Ca2+ exchanger) greatly determines the shape of the odor-induced response and, ultimately, the cells’ firing pattern to the brain. Kleene [198] investigated how cAMP can lead to a Cl– current in ORNs. Application of cAMP to the cytoplasmic side of an excised cilium sealed inside a patch pipette led to

)

*

 F)

Fig. 5.10 Origin of the Cl– current. (A) A frog cilium was sealed inside a patch pipette and excised. The pipette solution contained 1 mM Ca2+, while the bath solution contained 100 lM cAMP; both solutions had similar Cl– concentrations. The membrane potential was stepped to potentials from +80 mV to –60 mV in 20-mV increments. Only at negative potentials was a biphasic current observed. (B) The secondary current component could be suppressed by applying the Cl– channel blocker niflumic acid Modified from [198] with permission from Elsevier

 IA?

123

124

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

the generation of a biphasic current at negative, but not at positive, holding potentials (Fig. 5.10A). The additional current component observed at negative potentials could be blocked by the Cl– channel blocker niflumic acid (Fig. 5.10B), required the presence of Ca2+ in the extracellular solution, and increased when the cytoplasmic Ca2+ buffering capacity was reduced [198]. Thus, at negative potentials opening of the CNG channel first caused a mixed Na+ and Ca2+ influx with the accumulating Ca2+ activating the secondary Cl– current. What is the relative contribution of these two current components to the overall current? In the experiment in Fig. 5.10, the two current phases were of equal magnitude [176], which indicates that the CNG current is amplified twofold by the Cl– channel. This is consistent with the observation that during the odor response around half of the receptor current is carried by Cl–. Similarly, in an excised frog cilium the saturating CNG and Cl– currents are of comparable magnitude in the presence of physiological concentrations of Ca2+ and Mg2+ [176]. While extracellular divalents do not seem to affect the Cl– channel [111], they greatly reduce the single-channel conductance of the CNG channel to 0.56 pS (compared to 15 pS in divalent free conditions; see 5.7), a value close to the one for the Cl– channel of 0.5 pS [176]. The observation that both single-channel conductances are small has an important consequence. Individual channels contribute only small currents to the overall response, a mechanism that reduces noise due to the opening of individual channels. But do the sequential activation of both channels and the twofold amplification of the primary CNG channel by the Cl– channel lead to an increase in noise? Careful individual evaluation of the noise levels during the two phases of inward current recorded from frog cilia (see Fig. 5.10) demonstrated that little noise is added due to the activation of Cl– channels [176]. The main source of noise seems to be the CNG channel (Fig. 5.10B), with its maximal open-probability reaching only 0.68 in contrast to the maximal open-probability of 1 for the Cl– channel. Taken together, extracellular Ca2+ blocks the CNG channel and hence reduces noise, but influx of Ca2+ amplifies the CNG current without a further increase in noise [176]. In rat ORNs Cl– channels carry around 80 % of the odor-induced current [173], and in the presence of the divalent cations Ca2+ and Mg2+, membrane patches excised from rat ORNs have a Cl– current 30 times larger than the CNG current [111], quite different from frog olfactory cilia, where both channels conduct similar amounts of current. While the Cl– channel densities in frog and rat are comparable, rat CNG channels seem to be present at a 10-fold lower density [111, 112]. With this lower density of CNG channels in the rat ciliary membrane, it is interesting to ask how many CNG channels are responsible for contributing the Ca2+ to activate an individual Cl– channel. Or, put differently, how far does Ca2+ travel once it enters the cilia through a single CNG channel before it is bound to a Ca2+ buffer? Within 20 nm, Ca2+ entering through a channel will diffuse freely before it can be bound to a buffer; therefore, Ca2+ buffers do not affect the Ca2+ concentrations near Ca2+ channels [204]. With this precedent, the magnitude of the Ca2+-activated Cl– current should not be affected by Ca2+ buffers if the Cl- channels are within 20 nm of the Ca2+-permeable CNG channels. Membrane patches excised from the dendritic knob of rat ORNs were used to investigate this question. A patch was held in symmetrical

5.14 Amplification of the CNG Current and Generation of the Cl– Current

125

A CholCl 1mM HEDTA

CholCl 1mM HEDTA

Test solution

Current (pA)

0

-10

-20

-30

0

2

4

6

8

10

12

Time (s)

B

1

2 3 45

6

r1

C % contribution to [Ca2+]local free

Fig. 5.11 Interaction between the CNG and the Cl– channel. (A) A patch excised from a rat dendritic knob was held in symmetrical cholineCl solutions with 1 mM Ca2+ added to the pipette solution. Application of 67 lM Ca2+ to the cytoplasmic side activated a saturated Cl– current (lower trace), while 100 lM cAMP in 1 mM HEDTA (middle trace) or 0.2 mM HEDTA (upper trace) generated currents of smaller magnitude. This current could be suppressed by niflumic acid, identifying it as a Cl– current (not shown). (B) Model for the arrangement of Cl– (squares) and CNG (circles) channels in the ciliary membrane. CNG channels at increasing distances (r1–r6) will contribute Ca2+ to the Cl– channel in the center. (C) Level of Ca2+ contribution from CNG channels positioned in circles 1–6. Modified from [111] with permission from the Rockefeller University Press

100

2

1

3

4

5

circle 6

8

9

90

1 mM HEDTA 80 70 60

0.2 mM HEDTA

50 1

2

3

4

5

6

7

10 11

number of channels contributing to [Ca2+]local free

126

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents

choline Cl solution with the inclusion of 1 mM Ca2+ in the pipette. Under these ionic conditions, any observed current may originate only from flux of Cl– or Ca2+. In Fig. 5.11A, Cl– currents were evoked by applying either a saturating concentration of Ca2+ (lower trace) or cAMP in 0.2 mM (middle trace) or 1 mM (upper trace) of the Ca2+ buffer HEDTA. With 1 mM of HEDTA, a smaller Cl– current was observed than with 0.2 mM of HEDTA; thus, the increased Ca2+ buffer concentration can reduce the effective Ca2+ concentration at the Cl– channel once it enters through the CNG channel. From this observation that Cl– currents with different magnitudes were observed at two Ca2+ buffer concentrations, it was concluded that the CNG and the Cl– channel do not form transduction complexes. Therefore, instead of a configuration in which a CNG and a Cl– channel are very close to each other, the known ratio of Cl– to CNG channels of 4:1 (based on Cl– channel rundown; see 5.13) was used to construct an equidistant matrix for the distribution of Cl– and CNG channels in the patch membrane (Fig. 5.11B) [111]. The theory of buffered Ca2+ diffusion [195, 196] yielded two parameters: (1) the channel densities of the CNG and the Cl– channel and (2) the Ca2+ diffusion coefficient of 90 lm2 s–1 [111] for the cytosolic side of the patch. Depending on the internal Ca2+ buffer concentration, Ca2+ can diffuse considerable distances, and up to 50 % of the Ca2+ at a given Cl– channel originates not from the nearest CNG channel but from channels farther way (Fig. 5.11C). Thus, a Cl– channel sums the Ca2+ that enters through multiple CNG channels, a mechanism that can reduce noise originating from fluctuations of individual CNG channels, further enforcing the notion that these channels form a functional network with an amplification component. On the other hand, the number of CNG channels contributing to the Ca2+ concentration at a given Cl– channel is limited to around 10, which indicates that within a long, slender cilia, the spread of a Ca2+ signal might be spatially limited. Interestingly, odor-induced localized increases in intraciliary Ca2+ have been reported [171].

5.15

Open Questions

With Ca2+ playing such a critical role in olfactory transduction, many questions remain to be addressed. For instance, while the Cl– channel samples Ca2+ from multiple CNG channels, does the Ca2+ that leads to desensitization of the CNG channel via Ca2+/CaM originate from multiple CNG channels or a single CNG channel? How does the additional depolarization due to the Cl– current affect the Ca2+ current through the CNG channel and, therefore, intracellular Ca2+ levels? On the other hand, an increased level of depolarization could reduce Na+/Ca2+ exchange, maintain Ca2+ at an increased level, and therefore prolong the odor response. ORNs convey their action potentials directly to mitral/tufted cells in the olfactory bulb. But how do the CNG and the Cl– channels contribute to spike generation? Is the small Ca2+ current through the CNG channel sufficiently large to trigger action potentials, and how does the Cl– channel affect spike firing, which is, after all, the way the nose communicates with the brain?

5.15 Open Questions

Although much is known about individual components of the olfactory signal transduction, we are only beginning to understand the intricacies of the complex interaction within the transduction cascade and how this shapes the odor-induced response. Acknowledgement We thank the Howard Hughes Medical Institute and King-Wai Yau for their support, and Stephan Frings for reading the manuscript.

References 1

2

3 4

5

6

7

8

9

Getchell, M.L., Getchell, T.V., Immunohistochemical localization of components of the immune barrier in the olfactory mucosae of salamanders and rats. Anat. Rec., 1991. 231: p. 358–374. Farbman, A.I., ed. Cellular interactions in the development of the vertebrate olfactory system. Molecular Neurobiology of the Olfactory System, ed. F.L. Margolis, Getchell, T.V. 1988, Plenum, N.Y. 319–332. Gasser, H.S., Olfactory nerve fibres. Journal of General Physiology, 1956. 39: p. 473–496. Graziadei, P.P.C., and Monti Graziadei, G.A., Neurogenesis and neuron regeneration in the olfactory system of mammals. I. Morphological aspects of differentiation and structural organization of the olfactory sensory neurons. J. Neurocytol., 1979. 8: p. 1–18. Masukawa, L.M., B. Hedlund, and G.M. Shepherd, Changes in the Electrical-Properties of Olfactory Epithelial-Cells in the Tiger Salamander After Olfactory Nerve Transection. Journal of Neuroscience, 1985. 5(1): p. 136–141. Berkelaar, M., Clarke, D.B., Wang, Y.C., Bray, G.M., Aguayo, A.J., Axotomy results in delayed death and apoptosis of retinal ganglion-cells in adult-rats. J. Neurosci., 1994. 14: p. 4368–4374. Snider, W.D., Elliott, J.L., Yan, Q., Axotomyinduced neuronal death during development. J. Neurobiol., 1992. 23: p. 1231–1246. Schwob, J.E., Farber, N.B., Gottlieb, D.I., Neurons of the olfactory epithelium in adult-rats contain vimentin. J. Neurosci., 1986. 6: p. 208–217. Ophir, D., Lancet, D., Expression of intermediate filaments and desmoplakin in vertebrate olfactory mucosa. Anat. Rec., 1988. 221: p. 754–760.

10

11

12

13

14

15

16

17

18

19

Viereck, C., Tucker, R.P., Matus, A., The adultrat olfactory system expresses microtubule-associated proteins found in the developing brain. J. Neurosci., 1989. 9: p. 3547–3557. Ben-Ari, Y., Excitatory actions of gaba during development: the nature of the nurture. Nat Rev Neurosci, 2002. 3(9): p. 728–39. Okano, M., Tagaki, S.F., Secreation and electrogenesis of the supporting cell in the olfactory epithelium. J. Physiol. Lond., 1974. 242: p. 353–370. Masukawa, L.M., Hedlund, B. , Shepherd, G.M., Electrophysiologicl properties of identified cells in the in vitro olfactory epithelium of the tiger salamander. J. Neurosci., 1985. 5: p. 128–135. Lazard, D., Zupko, K., Poria, Y., Nef, P., Lazarovits, J., Lazarovits, J., Horn, S. and M. Kehn, Lancet, D., Odorant signal termination by olfactory UDP-glucuronosyl transferase. Nature, 1991. 349: p. 790–793. Nef, P., Heldman, J., Lazard, D., Margalit, T., Jaye, M., Hanukoglu, I., Lancet, D., Olfactoryspecific cytochrome-p-450: cDNA cloning of a novel neuroepithelial enzyme possibly involved in chemoreception. J. Biol. Chem., 1989. 264: p. 6780–6785. Amoore, J.E., Buttery, R.G., Partition coefficients and comparative olfactomery. Chem. Sens. Flavor, 1978. 3: p. 57–71. Margolis, F.L., A brain protein unique to the olfactory bulb. Proc. Natl. Acad. Sci., 1972. 69: p. 1221–1224. Key, B., Akeson, R.A., Olfactory neurons express a unique glycosylated form of the neural adhesion molecule (N-CAM). J. Cell Biol., 1990. 110: p. 1729–1743. Levey, M.S., Chikaraishi, D.M., Kauer, J.S., Characterization of potential precursor populations in the mouse olfactory epithelium using immunocytochemistry and autoradiography. J. Neurosci., 1991. 11: p. 3556–3564.

127

128

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents 20

21

22

23

24

25

26

27

28

29

30

31

32

33

Caggiano, M., Kauer, J.S., Hunter, D.D., Globose basal cells are neuronal progenitors in the olfactory epithelium: a lineage analysis using a replication-incompetent retrovirus. Neuron, 1994. 13: p. 339–352. Gordon, M.K., Mumm, J.S., Davis, R.A., Holcomb, J.D., Calof, A.L., Dynamics of mash1 expression in-vitro and in-vivo suggest a non-stem cell site of mash1 action in the olfactory receptor neuron lineage. Mol. Cell Neurosci., 1995. 6: p. 363–379. Dehamer, M.K., Guevara, J.L., Hannon, K., Olwin, B.B., Calof, A.L., Genesis of olfactory receptor neurons in vitro : Regulation of progenitor cell divisions by fibroblast growth-factors. Neuron, 1994. 13: p. 1083–1097. Ottoson, D., Analysis of the electrical activity of the olfactory epithelium. Acta Physiol. Scand., 1956. 122: p. 1–83. Leveteau, J., et al., Role of divalent cations in EOG generation. Chemical Senses, 1989. 14(5): p. 611–620. Bronshtein, A.A. and A.V. Minor, Regeneration of olfactory flagella and restoration of the electroolfactogram following application of Triton X-100 to the olfactory muscosa of frogs. Tsitologiya, 1977. 19: p. 33–39. Simmons, P.A., Getchell, T.V., Neurogenesis in the olfactory epithelium: loss and recovery of transepithelial voltage transients following olfactory nerve section. J. Neurophysiol., 1981. 45: p. 516–528. Sicard, G. and A. Holley, Receptor cell responses to odorants: similarities and differences among odorants. Brain Res, 1984. 292(2): p. 283–96. Duchamp, A., Revial, M.F., Holley, A., MacLeod. P., Odor discrimination by frog olfactory receptors. Chem.Senses Flavor, 1974. 1: p. 213–233. Gesteland, R.C., Lettvin, J.Y., Pitts, W.H., Chemical transmission in the nose of the frog. J. Physiol. London., 1965. 181: p. 525–559. Getchell, T.V., Shepherd, G.M., Responses of olfactory receptor cells to step pulses of odor at different concentrations in the salamander. J. Physiol. London, 1978. 282: p. 521–540. Kauer, J.S. and D.G. Moulton, Responses of olfactory bulb neurones to odour stimulation of small nasal areas in the salamander. Journal of Physiology, 1974. 243(3): p. 717–737. Minor, A.V., Sakina, N.L., Role of cyclic adenosine-3’,5’ -monophosphate in olfactory reception. Neurofysiologiya, 1973. 5: p. 415–422. Shibuya, T. and S. Shibuya, Olfactory epithelium: unitary responses in the tortoise. Science, 1963. 140: p. 495–496.

34

35

36

37

38

39 40

41

42

43

44

45

46

47

48

Getchell, T.V., Analysis of unitary spikes recorded extracellularly from frog olfactory cells and axons. Journal of Physiology-London, 1973. 234: p. 533–551. Getchell, T.V. and G.M. Shepherd, Adaptive properties of olfactory receptors analysed with odour pulses of varying durations. Journal of Physiology-London, 1978. 282: p. 541–560. Mathews, D.F., Response pattern of single neurons in the tortoise olfactory epithelium and olfactory bulb. Journal of General Physiology, 1972. 60: p. 166–180. Baylin, F., Temporal pattern and selectivity in the unitary responses of olfactory receptors in the tiger salamander to odor stimulation. Journal of General Physiology, 1979. 74: p. 17–36. O’Connell, R.J. and M.M. Mozell, Quantitative stimulation of frog olfactory receptors. Journal of Neurophysiology, 1969. 32: p. 51–63. Trotier, D., Intensity coding in olfactory receptor cells. Seminars in Cell Biology, 1994. 5: p. 47–54. van Drongelen, W., Unitary responses of near threshold responses of receptor cells in the olfactory mucosa of the frog. Journal of PhysiologyLondon, 1978. 277: p. 423–435. Reisert, J. and H.R. Matthews, Adaptation of the odour-induced response in frog olfactory receptor cells. Journal of Physiology-London, 1999. 519(Sep): p. 801–813. Trotier, D. and P. MacLeod, Intracellular recordings from salamander olfactory receptor cells. Brain Research, 1983. 268(2): p. 225–237. Masukawa, L.M., B. Hedlund, and G.M. Shepherd, Electrophysiological Properties of Identified Cells in the Invitro Olfactory Epithelium of the Tiger Salamander. Journal of Neuroscience, 1985. 5(1): p. 128–135. Trotier, D., Physiology of transduction in olfaction and taste. Seminars in the Neurosciences, 1990. 2: p. 69–76. Reisert, J. and H.R. Matthews, Responses to prolonged odour stimulation in frog olfactory receptor cells. Journal of Physiology, 2001. 534(Pt 1): p. 179–191. Baylin, F. and D.G. Moulton, Adaptation and cross-adaptation to odor stimulation of olfactory receptors in the tiger salamander. Journal of General Physiology, 1979. 74: p. 37–55. Frings, S. and B. Lindemann, Odorant response of isolated olfactory receptor cells is blocked by amiloride. Journal of Membrane Biology, 1988. 105(3): p. 233–243. Reisert, J. and H.R. Matthews, Response properties of isolated mouse olfactory receptor cells. Journal of Physiology-London, 2001. 530: p. 113–122.

5.15 Open Questions 49

50

51

52

53

54

55

56

57

58

59

60

61

Morales, B., et al., Inhibitory K+ current activated by odorants in toad olfactory neurons. Proceedings of the Royal Society of London Series B-Biological Sciences, 1994. 257(1350): p. 235–242. Pun, R.Y.K. and R.C. Gesteland, Somatic Sodium Channels of Frog Olfactory Receptor Neurons Are Inactivated At Rest. Pflugers Archiv-European Journal of Physiology, 1991. 418(5): p. 504–511. Firestein, S., G.M. Shepherd, and F.S. Werblin, Time course of the membrane current underlying sensory transduction in salamander olfactory receptor neurons. Journal of Physiology-London, 1990. 430(V): p. 135–158. Trotier, D., A patch-clamp analysis of membrane currents in salamander olfactory receptor cells. Pflugers Archiv-European Journal of Physiology, 1986. 407(6): p. 589–595. Kawai, F., T. Kurahashi, and A. Kaneko, T-type Ca2+ channel lowers the threshold of spike generation in the newt olfactory receptor cell. Journal of General Physiology, 1996. 108(6): p. 525–535. Anderson, P.A.V. and B.W. Ache, Voltage and current-clamp recordings of the receptor potential in olfactory receptor cells In situ. Brain Research, 1985. 338(2): p. 273–280. Anderson, P.A.V. and K.A. Hamilton, Intracellular recordings from isolated salamander olfactory receptor neurons. Neuroscience, 1987. 21(1): p. 167–173. Getchell, T.V., Analysis of intracellular recordings from salamander olfactory epithelium. Brain Research, 1977. 123: p. 275–286. Dionne, V.E., Chemosensory responses in isolated olfactory receptor neurons from Necturus maculosus. Journal of General Physiology, 1992. 99(3): p. 415–433. Morales, B., P. Labarca, and J. Bacigalupo, A Ciliary K+ Conductance Sensitive to Charibdotoxin Underlies Inhibitory Responses in Toad Olfactory Receptor Neurons. FEBS Letters, 1995. 359(1): p. 41–44. Leinders-Zufall, T., G.M. Shepherd, and F. Zufall, Regulation of cyclic nucleotide-gated channels and membrane excitability in olfactory receptor cells by carbon monoxide. Journal of Neurophysiology, 1995. 74(4): p. 1498–1508. Maue, R.A. and V.E. Dionne, Patch-clamp studies of isolated mouse olfactory receptor neurons. Journal of General Physiology, 1987. 90(1): p. 95–125. Lynch, J.W. and P.H. Barry, Action potentials initiated by single channels opening in a small neuron (rat olfactory receptor). Biophysical Journal, 1989. 55(4): p. 755–768.

62

63

64

65

66

67

68

69

70

71

72 73

74

75

Madrid, R., et al., Tonic and phasic receptor neurons in the vertebrate olfactory epithelium. Biophysical Journal, 2003. 84(6): p. 4167–4181. Getchell, T.V., Unitary responses in frog olfactory epithelium to sterically related molecules at low concentrations. Journal of General Physiology, 1974. 64: p. 241–261. Frings, S., S. Benz, and B. Lindemann, Current recording from sensory cilia of olfactory receptor cells insitu. 2. Role of mucosal Na+, K+, and Ca2+ ions. Journal of General Physiology, 1991. 97(4): p. 725–747. Firestein, S. and F. Werblin, Odor-induced membrane currents in vertebrate olfactory receptor neurons. Science, 1989. 244(4900): p. 79–82. Firestein, S., C. Picco, and A. Menini, The relation between stimulus and response in olfactory receptor cells of the tiger salamander. Journal of Physiology-London, 1993. 468: p. 1–10. Kurahashi, T. and A. Menini, Mechanism of odorant adaptation in the olfactory receptor cell. Nature, 1997. 385(6618): p. 725–729. Lowe, G. and G.H. Gold, The spatial distributions of odorant sensitivity and odorant-induced currents in salamander olfactory receptor cells. Journal of Physiology-London, 1991. 442(OCT): p. 147–168. Lowe, G. and G.H. Gold, Olfactory transduction is intrinsically noisy. Proceedings of the National Academy of Sciences of the United States of America, 1995. 92(17): p. 7864–7868. Menini, A., C. Picco, and S. Firestein, Quantal-like current fluctuations induced by odorants in olfactory receptor cells. Nature, 1995. 373(6513): p. 435–437. Kurahashi, T., Activation by odorants of cationselective conductance in the olfactory receptor cell isolated from the newt. Journal of PhysiologyLondon, 1989. 419(DEC): p. 177–192. Gilman, A.G., G-proteins and duel control of adenylate cyclase. Cell, 1984. 36: p. 577–579. Pace, U., Lancet, D., Olfactory GTP-binding protein: Signal-transducing polypepide of vertebrate chemosensory neurons. Proc. Natl. Acad. Sci. USA, 1986. 83: p. 4947–4951. Jones, D.T., Reed, R.R., Molecular cloning of five G-bindin protein cDNA species form rat olfactory neuroepithelium. J. Biol. Chem., 1987. 262: p. 14241–14249. Jones, D.T., Barbosa, E., Reed, R.R., Expression of G-protein a subunits in rat olfactory neuroepithelium: Candidates for olfactory signal transduction. Cold Spring Harbor Symp. Quan. Bio., 1988. LIII: p. 349–353.

129

130

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents 76

77

78

79 80

81

82

83

84

85

86 87

88

89

Jones, D.T., Masters, S.B., Bourne, H.R., Reed, R.R., Biochemical-characterization of 3 stimulatory GTP binding proteins: The large and small forms of Gs and the olfactory-specific Gprotein, Golf. J. Biol. Chem., 1990. 265: p. 2671–2676. Jones, D.T., and Reed, R.R., G-olf: an olfactory neuron-specific G-protein involved in odorant signal transduction. Science, 1989. 244: p. 790–795. Menco, P.M.B., Ultrastructural aspects of olfactory transduction and perireceptor events. Seminars Cell Bio., 1994. 5: p. 11–24. Belluscio, L., et al., Mice deficient in G(olf) are anosmic. Neuron, 1998. 20(1): p. 69–81. Drinnan, S.L., Hope, B.T., Snutch, T.P., Vincent, S.R., Gaolf in the basal ganglia. Mol. Cell Neurosci., 1991. 2: p. 66–70. Herve, D., Levistrauss, M., Mareysemper, I., Verney, C., Tassin, J.P., Glowinski, J., Girault, J.A., Golf and Gs in rat basal ganglia: Possible involvement of Golf in the coupling of dopamineD1 receptor with adenylyl cyclase. J. Neurosci., 1993. 13: p. 2237–2248. Sklar, P.B., Anholt, R.R.H., and Snyder, S.H., The odorant-sensitive adenylate cyclase of olfactory receptor cells: differential stimulation by different classes of odorants. J. Biol. Chem., 1986. 261: p. 15538–15543. Lowe, G., T. Nakamura, and G.H. Gold, Adenylate cyclase mediates olfactory transduction for a wide variety of odorants. Proceedings of the National Academy of Sciences of the United States of America, 1989. 86(14): p. 5641–5645. Pace, U., Hanski, E., Salomon, Y., and Lancet, D., Odorant-sensitive adenylate cyclase may mediate olfactory reception. Nature, 1985. 316: p. 255–258. Pfeuffer, E., et al., Olfactory adenylyl cyclase. Identification and purification of a novel enzyme form. J Biol Chem, 1989. 264(31): p. 18803–7. Bakalyar, H.A. and R.R. Reed, Science, 1990. 250: p. 1403–1406. Menco, B.P.M., et al., Ultrastructural localization of olfactory transduction components: the G-protein subunit Golf-Alpha and type III adenylyl cyclase. Neuron, 1992. 8(3): p. 441–453. Pun, R.Y. and S.J. Kleene, Contribution of cyclic-nucleotide-gated channels to the resting conductance of olfactory receptor neurons. Biophysical Journal, 2003. 84(5): p. 3425–3435. Wei, J., et al., Phosphorylation and inhibition of olfactory adenylyl cyclase by CaM kinase II in neurons: a mechanism for attenuation of olfactory signals. Neuron, 1998. 21(3): p. 495–504.

90

91

92

93

94

95

96 97

98

99

100

101

102

103

104

Wei, J., G. Wayman, and D.R. Storm, Phosphorylation and inhibition of type III adenylyl cyclase by calmodulin-dependent protein kinase II in vivo. Journal of Biological Chemistry, 1996. 271(39): p. 24231–24235. Leinders-Zufall, T., M. Ma, and F. Zufall, Impaired odor adaptation in olfactory receptor neurons after inhibition of Ca2+/calmodulin kinase II. Journal of Neuroscience, 1999: p. 19:RC19 (1–6). Borisy, F.F., et al., Calcium/calmodulin-activated phosphodiesterase expressed in olfactory receptor neurons. Journal of Neuroscience, 1992. 12(3): p. 915–923. Yan, C., et al., Molecular cloning and characterization of a calmodulin-dependent phosphodiesterase enriched in olfactory sensory neurons. Proceedings of the National Academy of Sciences of the United States of America, 1995. 92(21): p. 9677–9681. Wong, S.T., et al., Disruption of the type III adenylyl cyclase gene leads to peripheral and behavioral anosmia in transgenic mice. Neuron, 2000. 27(3): p. 487–497. Gold, G.H., Controversial issues in vertebrate olfactory transduction. Annu Rev Physiol, 1999. 61: p. 857–71. Gold, G., Lowe, G., Single odorant molecules? Nature, 1995. 376: p. 27. Baylor, D.A., T.D. Lamb, and K.W. Yau, Responses of retinal rods to single photons. Journal of Physiology, 1979. 288: p. 613–634. Buck, L., and Axel, R., A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell, 1991. 65: p. 175–187. Zhang, X. and S. Firestein, The olfactory receptor gene superfamily of the mouse. Nat Neurosci, 2002. 5(2): p. 124–33. Glusman, G., et al., The complete human olfactory subgenome. Genome Res, 2001. 11(5): p. 685–702. Rouquier, S., et al., Distribution of olfactory receptor genes in the human genome. Nat Genet, 1998. 18(3): p. 243–50. Fesenko, E.E., Koleniskov, S.S., Lyubarsky, A.L., Induction by cGMP of cationic conductance in plasma membrane of retinal rod outer segment. Nature, 1985. 313: p. 310–313. Haynes, L.W., Yau, K.W., Cyclic GMP-sensitive conductance in outer segment membrane of catfish cones. Nature, 1985. 317: p. 61–64. Nakamura, T., and Gold, G.H., A cyclic nucleotide-gated conductance in olfactory receptor cilia. Nature, 1987. 325: p. 342–344.

5.15 Open Questions 105 Huque, T., Bruch, R.C., Odorant and guanine

118 Yoshikami, S., Hagins, W.A., Cytoplasmic pH

nucleotide-stimulated phosphoinositide turnover in olfactory cilia. Biochem. Biophys. Res. Commun., 1986. 137: p. 36–42. Schild, D. and D. Restrepo, Transduction mechanisms in vertebrate olfactory receptor cells. Physiological Reviews, 1998. 78(2): p. 429–466. Frings, S., Lynch, J.W., Lindemann, B., Properties of cyclic nucleotide-gated channels mediating olfactory transduction: activation, selectivity, and blockage. J. Gen. Physiol., 1992. 100: p. 45–67. Firestein, S., B. Darrow, and G.M. Shepherd, Activation of the sensory current in salamander olfactory receptor neurons depends on a G-protein-mediated cAMP second messenger system. Neuron, 1991. 6: p. 825–835. Firestein, S., F. Zufall, and G.M. Shepherd, Single odor-sensitive channels in olfactory receptor neurons are also gated by cyclic nucleotides. J. Neurosci., 1991. 11: p. 3565–3572. Zufall, F., S. Firestein, and G.M. Shepherd, Analysis of single cyclic nucleotide-gated channels in olfactory receptor cells. J. Neurosci., 1991. 11: p. 3573–3580. Reisert, J., et al., The Ca-activated Cl Channel and its control in rat olfactory receptor neurons. Journal of General Physiology, 2003. 122(3): p. 349–364. Larsson, H.P., S.J. Kleene, and H. Lecar, Noise analysis of ion channels in non-space-clamped cables: Estimates of channel parameters in olfactory cilia. Biophysical Journal, 1997. 72(3): p. 1193–1203. Kurahashi, T., Kaneko, A., High-density campgated channels at the ciliary membrane in the olfactory receptor cell. Neuroreport, 1991. 2: p. 5–8. Kaur, R., et al., IP3-gated channels and their occurrence relative to CNG channels in the soma and dendritic knob of rat olfactory receptor neurons. Journal of Membrane Biology, 2001. 181(2): p. 91–105. Kolesnikov, S.S., A.B. Zhainazarov, and A.V. Kosolapov, Cyclic nucleotide-activated channels in the frog olfactory plasma membrane. FEBS Letters, 1990. 266: p. 96–98. Kurahashi, T., The response induced by intracellular cyclic-AMP in isolated olfactory receptor cells of the newt. Journal of Physiology-London, 1990. 430(V): p. 355–371. Menini, A., Currents carried by monovalent cations through cyclic GMP-activated channels in excised patches from salamander rods. J. Physiol., 1990. 424: p. 167–185.

in rod outer segments and high energy phosphate metabolism during phototransduction. Bio. Phys. J., 1985. 47: p. 101a (Abstr). Zufall, F. and S. Firestein, Divalent cations block the cyclic nucleotide-gated channel of olfactory receptor neurons. J Neurophysiol, 1993. 69(5): p. 1758–68. Dzeja, C., et al., Ca2+ permeation in cyclic nucleotide-gated channels. Embo Journal, 1999. 18(1): p. 131–144. Frings, S., Seifert, R., Godde, M., Kaupp, U.P., Profoundly different calcium permeation and blockage determine the specific function of distinct cyclic nucleotide-gated channels. Neuron, 1995. 15: p. 169–179. Lindemann, B., Predicted profiles of ion concentrations in olfactory cilia in the steady state. Biophysical Journal, 2001. 80(4): p. 1712–21. Haynes, L.W., Block of the cyclic GMP-gated channel of vertebrate rod and cone photoreceptors by l-cis-diltiazem. J. Gen. Physiol., 1992. 100: p. 783–801. Benos, D.J., Amiloride: a molecular probe of sodium transport in tissues and cells. Amer. J. Physiol., 1982. 242: p. C131–C145. Zhuang, Y.X., Cragoe, E.J., Shaikewitz, T., Glaser, L., Cassel, D., Characterization of potent Na/H exchange inhibitors from the amiloride series in a431 cells. Biochem., 1984. 23: p. 4481–4488. Tang, C.M., Presser, F., Morad, M., Amiloride selectively blocks the low threshold (T) calciumchannel. Science, 1988. 240: p. 213–215. Bielefeld, D.R., Hadley, R.W., Vassilev, P.M., Hume, J.R., Membrane electrical-properties of vesicular Na-Ca exchange inhibitors in single atrial myocytes. Circul. Res., 1986. 59: p. 381–389. Soltoff, S.P., Mandel, L.J., Amiloride directly inhibits the Na,K-ATPase activity of rabbit kidney proximal tubules. Science, 1983. 220: p. 957–959. Kleyman, T.R., Cragoe, E.J., Amiloride and its analogs as tools in the study of ion-transport. J. Membr. Bio., 1988. 105: p. 1–21. Nicol, G.D., Schnetkamp, P.P.M., Saimi, Y., Cragoe, E.J., Bownds, M.D., A derivative of amiloride blocks both the light-regulated and cyclic gmp-regulated conductances in rod photoreceptors. J. Gen. Physiol., 1987. 90: p. 651–669. Stern, J.H., Kaupp, U.B., Macleish, P.R., Control of the light-regulated current in rod photoreceptors by cyclic-GMP, calcium, and l-cis-diltiazem. Proc. Natl. Acad. Sci., 1986. 83: p. 1163–1167.

106

107

108

109

110

111

112

113

114

115

116

117

119

120

121

122

123

124

125

126

127

128

129

130

131

131

132

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents 132 Hurwitz, R., Holcombe, V., Affinity purifica-

133

134

135

136 137

138

139

140

141

142

143

144

145

tion of the photoreceptor GMP-gated cation channel. J. Biol. Chem., 1991. 226: p. 7975–7977. Dhallan, R.S., Yau, K-W., Schrader, K.A., Reed, R.R., Primary structure and functional expression of a cyclic nucleotide-activated channel from olfactory neurons. Nature, 1990. 347: p. 184–187. Ludwig, J., et al., Primary structure of cAMPgated channel from bovine olfactory epithelium. FEBS Lett., 1990. 270: p. 24–29. Kaupp, U.B., Niidome, T., Tanabe, T., Terada, S., Bonigk, W., Stuhmer, W., Cook, W.J., Kangawa, K., Matsuko, H., Hirose, T., Miyata, T., Numa, S., Primary structure and functional expression from complementary cDNA of the rod photoreceptor cyclic GMP-gated channel. Nature, 1989. 342: p. 762–766. Bradley, J., et al., Nomenclature for ion channel subunits. Science, 2001. 294(5549): p. 2095–6. Goulding, E.H., et al., Molecular cloning and single-channel properties of the cyclic nucleotidegated channel from catfish olfactory neurons. Neuron, 1992. 8: p. 45–58. Kaupp, U.B. and R. Seifert, Cyclic nucleotidegated ion channels. Physiol Rev, 2002. 82(3): p. 769–824. Bradley, J., Li, J., Davidson, N., Lester, H.A., Zinn, K, Heteromeric olfactory cyclic nucleotidegated channels: A subunit that confers increased sensitivity to cAMP. PNAS, 1994. 91: p. 8890–8894. Liman, E.R., Buck, L.B., A second subunit of the olfactory cyclic nucleotide-gated channel confers high sensitivity to cAMP. Neuron, 1994. 13: p. 611–621. Broillet, M.C. and S. Firestein, b subunits of the olfactory cyclic nucleotide-gated channel form a nitric oxide activated Ca2+ channel. Neuron, 1997. 18(6): p. 951–958. B€onigk, W., et al., The native rat olfactory cyclic nucleotide-gated channel is composed of three distinct subunits. J Neurosci, 1999. 19(13): p. 5332–47. Sautter, A., et al., An isoform of the rod photoreceptor cyclic nucleotide-gated channel beta subunit expressed in olfactory neurons. Proc Natl Acad Sci U S A, 1998. 95(8): p. 4696–701. Zheng, J. and W.N. Zagotta, Determine cyclic nucleotide-gated channel stiochiometry with FRET. Biophys J, 2003. 84(2): p. 138A. Kurahashi, T. and T. Shibuya, Ca2+-dependent adaptive properties in the solitary olfactory receptor cell of the newt. Brain Research, 1990. 515(1–2): p. 261–268.

146 Zufall, F., G.M. Shepherd, and S. Firestein,

147

148

149

150

151

152

153

154

155

156

157

158

159

Inhibition of the olfactory cyclic nucleotide gated ion channel by intracellular calcium. Proc R Soc Lond B Biol Sci, 1991. 246(1317): p. 225–30. Anholt, R.R.H., Rivers, A.M., Olfactory transduction - cross-talk between 2nd-messenger systems. Biochem., 1990. 29: p. 4409–4054. Hsu, Y.-T. and R.S. Molday, Modulation of the cGMP-gated channel of rod photoreceptor cells by calmodulin. Nature, 1993. 361: p. 76–79. Fain, G.L., et al., Adaptation in vertebrate photoreceptors. Physiol Rev, 2001. 81(1): p. 117–151. Chen, T.-Y., Yau, K-W., Direct modulation by Ca2+-calmodulin of cyclic nucleotide-activated channel of rat olfactory receptor neurons. Nature, 1994. 368: p. 545–548. Bradley, J., D. Reuter, and S. Frings, Facilitation of calmodulin-mediated odor adaptation by cAMP-gated channel subunits. Science, 2001. 294(5549): p. 2176–8. Munger, S.D., et al., Central role of the CNGA4 channel subunit in Ca2+-calmodulin-dependent odor adaptation. Science, 2001. 294(5549): p. 2172–5. Liu, M., et al., Calcium-calmodulin modulation of the olfactory cyclic nucleotide-gated cation channel. Science, 1994. 266(5189): p. 1348–54. Varnum, M.D. and W.N. Zagotta, Interdomain interactions underlying activation of cyclic nucleotide- gated channels. Science, 1997. 278(5335): p. 110–3. Grunwald, M.E., et al., Molecular determinants of the modulation of cyclic nucleotide-activated channels by calmodulin. Proc Natl Acad Sci U S A, 1999. 96(23): p. 13444–9. Zheng, J., M.D. Varnum, and W.N. Zagotta, Disruption of an Intersubunit Interaction Underlies Ca2+-Calmodulin Modulation of Cyclic Nucleotide-Gated Channels. J Neurosci, 2003. 23(22): p. 8167–8175. Trudeau, M.C. and W.N. Zagotta, Calcium/ calmodulin modulation of olfactory and rod cyclic nucleotide-gated ion channels. J Biol Chem, 2003. 7: p. 7. O’Neil, K.T. and W.F. DeGrado, How calmodulin binds its targets: sequence independent recognition of amphiphilic alpha-helices. Trends Biochem Sci, 1990. 15(2): p. 59–64. Bradley, J., Bonigk, W., Yau, K. W., Frings, S., Calmodulin permanently associates with rat olfactory CNG channels under native conditions. Nat Neurosci, 2004 7: p. 705–710.

5.15 Open Questions 160 Grunwald, M.E., et al., Identification of a do-

161

162

163

164

165

166

167

168

169

170

171

172

main on the beta-subunit of the rod cGMP-gated cation channel that mediates inhibition by calcium-calmodulin. J Biol Chem, 1998. 273(15): p. 9148–57. Weitz, D., et al., Calmodulin controls the rod photoreceptor CNG channel through an unconventional binding site in the N-terminus of the beta-subunit. Embo J, 1998. 17(8): p. 2273–84. Bahler, M. and A. Rhoads, Calmodulin signaling via the IQ motif. FEBS Lett, 2002. 513(1): p. 107–13. Erickson, M.G., et al., FRET two-hybrid mapping reveals function and location of L-type Ca2+ channel CaM preassociation. Neuron, 2003. 39(1): p. 97–107. Zuhlke, R.D., et al., Calmodulin supports both inactivation and facilitation of L-type calcium channels. Nature, 1999. 399(6732): p. 159–62. Kleene, S.J. and R.C. Gesteland, Calcium-activated chloride conductance in frog olfactory cilia. Journal of Neuroscience, 1991. 11(11): p. 3624–3629. Yoshii, K. and K. Kurihara, Role of cations in olfactory reception. Brain Research, 1983. 274: p. 239–248. Suzuki, N., Effects of different ionic invironments on the responses of single olfactory responses in the lamprey. Comperative Biochemistry and Physiology. A Comperative Physiology, 1978. 61: p. 461–467. Kleene, S.J. and R.Y.K. Pun, Persistence of the olfactory receptor current in a wide variety of extracellular environments. Journal of Neurophysiology, 1996. 75(4): p. 1386–1391. Tucker, D. and T. Shibuya, A physiologic and pharmacologic study of olfactory receptors. Cold Spring Harb Symp Quant Biol, 1965. 30: p. 207–15. Zufall, F., T. Leinders-Zufall, and C.A. Greer, Amplification of odor-induced Ca2+ transients by store-operated Ca2+ release and its role in olfactory signal transduction. Journal of Neurophysiology, 2000. 83(1): p. 501–512. Leinders-Zufall, T., et al., Imaging odor-induced calcium transients in single olfactory cilia: Specificity of activation and role in transduction. Journal of Neuroscience, 1998. 18(15): p. 5630–5639. Leinders-Zufall, T., et al., Calcium entry through cyclic nucleotide-gated channels in individual cilia of olfactory receptor cells: Spatiotemporal dynamics. Journal of Neuroscience, 1997. 17(11): p. 4136–4148.

173 Lowe, G. and G.H. Gold, Nonlinear amplifi-

174

175

176

177

178

179

180

181

182

183

184

cation by calcium-dependent chloride channels in olfactory receptor cells. Nature, 1993. 366(6452): p. 283–286. Kurahashi, T. and K.-W. Yau, Co-existence of cationic and chloride components in odorant-induced current of vertebrate olfactory receptor cells. Nature, 1993. 363(6424): p. 71–74. Zhainazarov, A.B. and B.W. Ache, Odor-induced currents in Xenopus olfactory receptor cells measured with perforated-patch recording. Journal of Neurophysiology, 1995. 74(1): p. 479–483. Kleene, S.J., High-gain, low-noise amplification in olfactory transduction. Biophysical Journal, 1997. 73(2): p. 1110–1117. Sato, K. and N. Suzuki, The contribution of a Ca2+-activated Cl– conductance to amino-acidinduced inward current responses of ciliated olfactory neurons of the rainbow trout. Journal of Experimental Biology, 2000. 203(2): p. 253–262. Dubin, A.E. and V.E. Dionne, Action potentials and chemosensitive conductances in the dendrites of olfactory neurons suggest new features for odor transduction. Journal of General Physiology, 1994. 103(2): p. 181–201. Reuter, D., et al., A depolarizing chloride current contributes to chemoelectrical transduction in olfactory sensory neurons in situ. Journal of Neuroscience, 1998. 18(17): p. 6623–6630. Nakamura, T., H. Kaneko, and N. Nishida, Direct measurement of the chloride concentration in newt olfactory receptors with the fluorescent probe. Neuroscience Letters, 1997. 237(1): p. 5–8. Kaneko, H., T. Nakamura, and B. Lindemann, Noninvasive measurement of chloride concentration in rat olfactory receptor cells with use of a fluorescent dye. American Journal of Physiology, 2001. 280(6): p. C1387–1393. Chiu, D., T. Nakamura, and G.H. Gold, Ionic Composition of Toad Olfactory Mucus Measured With Ion-Selective Microelectrodes. Chemical Senses, 1988. 13(4): p. 677–678. Firestein, S. and G.M. Shepherd, Interaction of anionic and cationic currents leads to a voltage dependence in the odor response of olfactory receptor neurons. Journal of Neurophysiology, 1995. 73(2): p. 562–567. Hallani, M., J.W. Lynch, and P.H. Barry, Characterization of calcium-activated chloride channels in patches excised from the dendritic knob of mammalian olfactory receptor neurons. Journal of Membrane Biology, 1998. 161(2): p. 163–171.

133

134

5 Vertebrate Olfactory Signal Transduction and the Interplay of Excitatory Anionic and Cationic Currents 185 Kleene, S.J., R.C. Gesteland, and S.H. Bryant,

195 Neher, E., Concentration profiles of intracellular

An electrophysiological survey of frog olfactory cilia. Journal of Experimental Biology, 1994. 195: p. 307–328. Frings, S., D. Reuter, and S.J. Kleene, Neuronal Ca2+-activated Cl– channels - homing in on an elusive channel species. Progress in Neurobiology, 2000. 60(3): p. 247–289. Kuruma, A. and H.C. Hartzell, Bimodal control of a Ca2+-activated Cl– channel by different Ca2+ signals. Journal of General Physiology, 2000. 115(1): p. 59–80. Callamaras, N. and I. Parker, Ca2+-dependent activation of Cl– currents in Xenopus oocytes is modulated by voltage. American Journal of Physiology Cell Physiology, 2000. 278(4): p. C667–675. Arreola, J., J.E. Melvin, and T. Begenisich, Activation of calcium-dependent chloride channels in rat parotid acinar cells. Journal of General Physiology, 1996. 108(1): p. 35–47. Piper, A.S. and W.A. Large, Multiple conductance states of single Ca2+-activated Cl– channels in rabbit pulmonary artery smooth muscle cells. Journal of Physiology, 2003. 547(Pt 1): p. 181–196. Nakamura, T., et al., Gated Conductances in Native and Reconstituted Membranes From Frog Olfactory Cilia. Biophysical Journal, 1996. 70(2): p. 813–817. Hirakawa, Y., et al., Ca2+-dependent Cl– channels in mouse and rabbit aortic smooth muscle cells: regulation by intracellular Ca2+ and NO. American Journal of Physiology, 1999. 277: p. H1732–1744. Van Renterghem, C. and M. Lazdunski, Endothelin and vasopressin activate low conductance chloride channels in aortic smooth muscle cells. Pflugers Archiv, 1993. 425(1–2): p. 156–163. Kl€ockner, U., Intracellular calcium ions activate a low-conductance chloride channel in smoothmuscle cells isolated from human mesenteric artery. Pflugers Archiv, 1993. 424(3–4): p. 231–237.

calcium in the presence of a diffusible chelator, in Experimental Brain Research, U. Heinemann, et al., Editors. 1986, Springer: Heidelberg. p. 80–96. Bauer, P.J., The local Ca concentration profile in the vicinity of a Ca channel. Cell Biochemistry and Biophysics, 2001. 35(1): p. 49–61. Becq, F., Ionic channel rundown in excised membrane patches. Biochimica et Biophysica Acta, 1996. 1286: p. 53–63. Kleene, S.J., Origin of the chloride current in olfactory transduction. Neuron, 1993. 11(1): p. 123–132. Kleene, S.J., Both external and internal calcium reduce the sensitivity of the olfactory cyclic-nucleotide-gated channel to cAMP. Journal of Neurophysiology, 1999. 81(6): p. 2675–2682. Noe, J., et al., Sodium/calcium exchanger in rat olfactory neurons. Neurochemistry International, 1997. 30(6): p. 523–531. Jung, A., et al., Sodium/calcium exchanger in olfactory receptor neurons of Xenopus laevis. Neuroreport, 1994. 5(14): p. 1741–1744. Reisert, J. and H.R. Matthews, Na+-dependent Ca2+ extrusion governs response recovery in frog olfactory receptor cells. Journal of General Physiology, 1998. 112(5): p. 529–535. Lo, Y.H., T.M. Bradley, and D.E. Rhoads, High-Affinity Ca2+,Mg2+-ATPase in plasma membrane-rich preparations from olfactory epithelium of atlantic salmon. Biochimica Et Biophysica Acta-Biomembranes, 1994. 1192(2): p. 153–158. Neher, E., Vesicle pools and Ca2+ microdomains: new tools for understanding their roles in neurotransmitter release. Neuron, 1998. 20(3): p. 389–399. Reuter, D., Ph. D. Thesis. 2000.

186

187

188

189

190

191

192

193

194

196

197

198

199

200

201

202

203

204

205

135

6

Transduction Channels in the Vomeronasal Organ Emily R. Liman and Frank Zufall

6.1

Introduction

The vomeronasal organ (VNO) is a sensory structure found in the nasal cavity of most vertebrate animals that detects chemosignals, including some pheromones. Pheromones are defined as chemicals released by an animal that elicit stereotyped behavioral or neuroendocrine responses in other animals of the same species [1]. In mammals, pheromones play an important role in regulating reproductive and social behaviors. For example, pheromones can induce suppression of estrus, aggressive behavior, suckling, and mating, effects that depend on the nature of the pheromones released as well as the gender, age, and neuroendocrine status of the recipient. The VNO mediates many, but not all, of the effects of pheromones in vertebrates, and it may also play a role in detecting chemosignals that are not pheromones [2, 3]. For example, in the snake the VNO plays a role in prey detection [4]. In recent years, great progress has been made in understanding mechanisms by which chemosensory stimuli are detected and transduced by the VNO (reviewed in [5]). We now know the identity of most of the elements of the transduction cascade, including the identity of at least one component of the transduction channel, the ion channel TRPC2 [45]. The identification of TRPC2 in the VNO and the subsequent generation of TRPC2 knockout mice have not only provided insight into mechanisms of sensory transduction but also have allowed further study into the functional significance of the vomeronasal system. Thus, at the end of this chapter, information from TRPC2 knockout mice is used to evaluate the function of the vomeronasal system in the mouse and its evolution in humans.

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

136

6 Transduction Channels in the Vomeronasal Organ

6.2

Anatomy of the Vomeronasal System

Two anatomically distinct organs sense pheromones: the main olfactory epithelium (MOE) and the VNO. The VNO is a paired tubular structure at the rostral end of the nasal cavity (see Fig. 6.1) that receives input either from the nasal or oral cavity, depending on species. Because the VNO is a closed tube, non-volatile chemicals are introduced into the lumen of the VNO by an active pumping mechanism [6]. Recent in vivo recording shows that in mice the VNO is most active when an animal is in direct contact with a conspecific and is most sensitive to chemicals present in the oral or anogenital region [7]. In contrast, the main olfactory system responds to odorants and pheromones that are volatile and are passed over the epithelium during the respiratory cycle. The VNO sends projections to the accessory olfactory bulb (AOB), which lies adjacent to the main olfactory bulb (MOB). However, projections from the AOB, unlike those from the MOB, do not reach cortical areas; instead, signals are sent to hypothalamic regions involved in the control of stereotyped behavior and neuroendocrine responses. In cross-section the VNO appears as a crescent-shaped structure (see Fig. 6.1), and within this crescent two broad zones can be distinguished. The apical region contains sensory neurons that express the signaling molecular Gai2 and a set of putative receptor molecules (V1Rs; discussed below), whereas basal neurons express Gao and V2R receptors. Projections from the apical neurons are to the anterior part of the AOB, whereas basal neurons project to the posterior part of the AOB [5]. It is likely that some of the projections remain segregated at higher brain regions [8, 9] and that these two zones play differential roles in regulating behavior. The VNO sensory neurons are bipolar cells that send a single dendrite to the lumen of the vomeronasal epithelium and a single axon to the AOB (see Fig. 6.1). The den-

Fig. 6.1 Anatomy of the mouse vomeronasal system. (A) Schematic drawing of a parasagittal section shows the VNO and the turbinates of the main olfactory epithelium (OE). The VNO projects to the accessory olfactory bulb (AOB), whereas the OE projects to the olfactory bulb (OB). (B) In a section taken at the position indicated in (A), both

the OE on the roof of the nasal cavity and the VNO at the base (depicted as crescents) can be seen. (C) A close-up of the epithelium of the VNO shows bipolar vomeronasal sensory neurons, which send dendrites containing microvilli to the lumen of the epithelium. From [69]

6.3 Sensory Responses Involve Generation of Action Potentials and Ca2+ Entry

drite ends in a tuft of microvilli, which serve to increase the surface area over which pheromone receptors are expressed. The observation that VNO neurons contain microvilli, whereas olfactory neurons are ciliated, suggests a distinct evolutionary origin for the two sets of cells. This is supported by the recent molecular data showing different sets of signaling molecules in the two cell types. As will be discussed later, a major difference between sensory transduction in the VNO and that in the MOE is that VNO sensory transduction is mediated by a PLC signaling pathway leading to the opening of an ion channel in the transient receptor potential (TRP) family, whereas transduction in the MOE is mediated by cyclic nucleotides that act directly on a cyclic nucleotide-gated (CNG) channel. It may not be a coincidence that other microvillous sensory neurons, such as the fly photoreceptor (see Chapter 8), the vertebrate hair cell, and the vertebrate taste cell, also express TRP channels, whereas ciliated neurons such as the vertebrate photoreceptor express CNG channels. The reason for this correlation between structure and signaling is not known, and it may have a functional significance or instead be an evolutionary vestige.

6.3

Sensory Responses Involve Generation of Action Potentials and Ca2+ Entry

An important question that was until recently unresolved is whether sensory transduction in the VNO leads to an excitatory response followed by the induction of action potentials, as in the main olfactory system, or to the suppression of activity, as in the visual system. This is a key issue because some papers claimed that sensory transduction led to suppression of activity [10, 11] and only recently have these results been refuted. Given the extreme diversity of receptors expressed by VNO sensory neurons, finding the answer to this question required that methods be devised to record simultaneously from large populations of VNO sensory neurons [12, 13]. Summed field potentials recorded from the microvillous layer of intact mouse VNO sensory epithelium was first employed by Leinders-Zufall et al. [12] to identify chemically defined ligands acting on vomeronasal neurons and their receptors. These studies clearly showed that pheromonal ligands generate negative deflections in the VNO field potential, consistent with an excitation of the VNO neurons. Subsequently, this method proved highly useful for the functional phenotyping of mice with targeted mutations in genes that encode the TRPC2 channel, V1r receptors, or adenylyl cyclase III [3, 14, 15] (see below). A 64-microelectrode array was used by Holy et al. [13] to record extracellular action potential activity from large subsets of sensory neurons in an explant of mouse VNO epithelium. This study demonstrated that as many as 40 % of the recorded neurons responded by increasing their spike frequency rate to the complex pheromonal stimuli that are present in urine. Moreover, subsets of VNO neurons were activated selectively by either male or female urine, whereas the response of other cells was independent of the sex of the donor animal [13]. This method has also proven useful for the characterization of mice with specific defects in VNO transduction [50].

137

138

6 Transduction Channels in the Vomeronasal Organ

Comparable results were also obtained by other researchers using extracellular single-electrode recording in tissue slices from rat VNO [16]. To examine responses of individual vomeronasal neurons to chemically defined ligands, Leinders-Zufall et al. [12] used extracellular loose-patch recordings to register activity from single, optically identified neurons in mouse VNO tissue slices. These experiments clearly demonstrated that VNO neurons respond to sensory stimulation with an increase in the rate of action potential firing. Therefore, it is now clear that VNO neurons in several species generate an excitation in response to chemostimulation. To record from a large number of spatially defined cells simultaneously, LeindersZufall et al. [12] used confocal imaging to measure Ca2+ responses of VNO slices. They showed for the first time that sensory neurons exhibit an increase in intracellular Ca2+ in response to chemostimulation and that these Ca2+ signals are eliminated after removal of extracellular Ca2+, indicating that they were triggered primarily by Ca2+ entry. These experiments also showed that the detection thresholds for pheromones by individual VNO neurons are at remarkably low concentrations, near 10–11 M, and that the individual cells are selectively tuned to detect only one or a few chemicals [12]. Subsequent studies using snake and rat VNO neurons have confirmed that chemostimulation produces an increase in intracellular Ca2+ in these cells [17, 18]. However, in the snake VNO, removal of extracellular Ca2+ did not fully suppress the chemoresponse [17]. Recent patch-clamp recordings from single mouse VNO neurons have revealed the ionic basis underlying this excitatory response (Fig. 6.2) [19]. These findings show that mouse VNO neurons generate a depolarizing receptor potential that, in turn, causes the cells to discharge action potentials (Fig. 6.2). Underlying this graded depolarization is a transient inward current, the sensory current, which represents the earliest electrical event in the transduction process. Since the input resistance of VNO neurons is

Fig. 6.2 Sensory responses of single mouse VNO neurons to focal stimulation of the dendritic tip with dilute urine (DU, 1/100), a rich source of natural pheromones. (A) Under current clamp, sensory stimulation produces a depolarizing receptor potential leading to robust action potential discharges. (B) Under voltage clamp, stimulation generates a rapidly activating and then deactivating inward current, the sensory current (holding potential, –70 mV). Recordings were obtained from visually identified neurons using an acute VNO tissue slice preparation. From [19]

6.4 Two Families of G-protein-coupled Receptors Mediate VNO Transduction

very high, only a few picoamperes of inward current are needed to generate action potentials [20]. What is the nature of the signaling cascade that leads to the generation of this sensory current? This question has been approached with both molecular and physiological methods.

6.4

Two Families of G-protein-coupled Receptors Mediate VNO Transduction

An early insight into VNO transduction came from the observation that components of MOE transduction were absent from the VNO [21–23]. Thus, a whole new set of signaling molecules remained to be found. Most elusive of these, perhaps, were the receptor molecules. A priori there was little information concerning the number or chemical makeup of these receptors. Thus, a major breakthrough in our understanding came with the cloning of the first family of VNO receptor genes by Dulac and Axel in 1995 [21]. In cloning the VNO receptor genes, Dulac and Axel made only two assumptions: that cells would differ in the receptor transcripts that they expressed and that the expression of these receptors would be restricted to the VNO. By cloning mRNAs that were differentially expressed between two single VNO neurons, Dulac and Axel identified a putative seven-transmembrane protein which, at the time, showed little homology to known G-protein-coupled receptors and was uniquely expressed in the VNO [21]. It is now known that there are 137 related genes with intact reading frames in the mouse genome [24], including the original receptor family (originally named VNs and subsequently named V1Rs) and a second group of receptors (originally named V3Rs and subsequently named V1Rds), [25] and that these receptors show structural similarity to a class of taste receptors (T2Rs). V1R receptors are restricted in expression to the apical region of the VNO sensory epithelium [21]. A distinct family of receptors (V2Rs) that are expressed in the basal region of the VNO sensory epithelium was subsequently identified [26–28]; these receptors, of which there are 100 in the mouse, contain a large amino terminus and show structural similarity to metabotropic glutamate receptors and T1Rs taste receptors. How these receptors interact with their ligands is not known. Recently it was shown that non-classical major histocompatibility complex (MHC) class Ib molecules interact with V2Rs and participate in receptor trafficking and possibly ligand binding [29]. Evidence that either V1Rs or V2Rs are involved in chemosensory transduction remains sparse. V2R receptor protein has been detected in sensory microvilli [29], and a mouse in which a large cluster of V1Rs was deleted shows some behavioral abnormalities consistent with a partial loss of pheromone transduction [15]. Given the large repertoire of VNO receptors, linking specific receptors with ligands and with behavioral responses remains an important challenge for the future. A step in this direction was made with a report that VNO neurons expressing the V1Rb2 receptor respond specifically to nanomolar concentrations of 2-heptanone [30].

139

140

6 Transduction Channels in the Vomeronasal Organ

6.5

Signaling Downstream of G Proteins May Involve a PLC

What is the nature of the signaling cascade that mediates VNO sensory transduction? Several lines of investigation have yielded insight into this question. Immunohistochemistical studies have shown that two G proteins, Gao and Gai2, are expressed in distinct and non-overlapping zones of the VNO epithelium and that their distribution coincides exactly with that of the two classes of receptors [31, 32]. This suggests that V1Rs may couple to Gai2 and V2Rs to Gao. Biochemical experiments show activation of these G proteins to pheromonal stimuli [33], and the observation that Gai2 and Gao proteins are present in microvilli provides further support for this hypothesis [34]. More recently, mice that carry a targeted deletion of Gai2 have been found to show defects in pheromone-mediated behaviors, consistent with a role for Gai2 in VNO transduction [35]. The coupling between G protein and downstream effectors is perhaps the least understood step in the signal transduction cascade. There is general consensus that a phospholipase C (PLC) isoform plays a critical role in sensory transduction. Phospholipase C hydrolyzes PIP2 into IP3 and DAG. Biochemical studies have shown that IP3 is generated in response to sensory activation of the VNO in snake [36], hamster [37], and pig [38] and that physiological responses of VNO receptor neurons to urine stimuli are blocked by a selective inhibitor of PLCs (U71332) [13, 18, 19]. However molecular studies have mysteriously failed to reveal the identity of the putative PLC. As the G alpha subunits Gi and Go are not typically involved in signaling through a PLC, it has been suggested that bc subunits, specifically Gc2 and Gc8, are responsible for activation of the PLC signaling pathway. Evidence to support this comes from the observation that antibodies directed against Gc8 and Gc2 block the generation of IP3 in response to urinary pheromones and a2 globulin, respectively [39].

6.6

Second Messengers for VNO Transduction: Functional Studies

To determine mechanisms of sensory transduction, a number of laboratories have studied functional responses of VNO sensory neurons to putative second messengers. No responses to cAMP or cGMP were detected in mouse VNO neurons [20], indicating that transduction is distinct from that of either the main olfactory system or the visual system of vertebrates. Responses to a wide range of second messengers that are generated downstream of PLC (Ca2+, IP3, DAG, and PUFAs) have been reported, but a connection to sensory transduction as described below has been made for only two of them: DAG and its metabolite arachidonic acid. In particular the following second-messenger responses have been reported: 1. Ca2+: Ca2+ has been shown to activate a 25-pS cation channel in excised patches from hamster [40] and mouse VNO neurons (Liu and Liman, unpublished results). This channel is equally permeable to Na+, K+, and Cs+ but is not detectably

6.7 Identification of the TRPC2 Ion Channel as a Candidate Transduction Channel

permeable to Ca2+. The channel is activated by only high concentrations of intracellular Ca2+ (K1/2 0.5 mM) and it is blocked by micromolar levels of adenine nucleotides (ATP and cAMP). The molecular identity of this channel is not known, although it may belong to the TRPM class of ion channels, several of which are Ca2+ activated [70–73]. This conductance may play a role in amplifying the primary sensory response but is unlikely to mediate primary sensory transduction. 2. IP3: Dialysis of IP3 through a pipette has been shown to activate a conductance in VNO neurons from turtle [41], rat [42], and snake [43]; however, none of these reports showed that IP3 directly activates a membrane channel in these cells. Furthermore, other researchers have observed only small and slow responses to intracellular dialysis of mouse VNO neurons with IP3 ([19] and Liman, unpublished observations), inconsistent with the previous studies. In addition, pharmacological block of the IP3 receptor does not interfere with the response of mouse neurons to pheromones in urine [18]. 3. Arachidonic acid (AA): AA, a polyunsaturated fatty acid that is generated as a metabolite of DAG, induced an increase in Ca2+ and a slowly activating inward current in rat VNO neurons, although direct activation of a membrane channel by AA has not been demonstrated [18]. Application of the DAG lipase inhibitor RHC-80267, which inhibits the synthesis of AA from DAG, blocked Ca2+ responses to urinary pheromones in the experiments by Spehr et al. [18], leading the authors to conclude that AA plays a role as a second messenger for VNO transduction. However, RHC80267 failed to block activation of the sensory current in mouse VNO neurons [19]. Thus, it is still unclear whether an AA-activated conductance plays a role in VNO sensory transduction. 4. Diacylglycerol (DAG): Analogues of DAG activate a 42-pS channel in excised insideout patches from the dendritic tip of VNO sensory neurons [19] (Fig. 6.3). The channel is permeable to Na+, Cs+, and Ca2+ but not to N-methyl-D-glucamine (with relative permeabilities of PCa/PNa = 2.7 and PCs/PNa = 1.5). A current with similar properties is activated by DAG in whole-cell recordings, showing a linear I-V relation with a reversal potential close to 0 mV [19]. These currents are blocked by external Ca2+ and by 2-aminoethoxydiphenyl borate (2-APB), which blocks some TRP channels [44]. These properties match those of the pheromone-induced current, providing support that the DAG-activated channel mediates pheromone transduction [19]. This notion is further supported by the result that both the DAG-activated channel and the sensory response are significantly impaired in mice with a targeted deletion in the TRPC2 gene (see below).

6.7

Identification of the TRPC2 Ion Channel as a Candidate Transduction Channel for VNO Sensory Signaling

An essential element of the signal transduction cascade is the ion channel that ultimately converts the chemical signaling into an electrical response. At present, the best candidate for this channel is the molecule TRPC2 (originally called TRP2), which was

141

142

6 Transduction Channels in the Vomeronasal Organ

Fig. 6.3 Identification of a diacylglycerol-activated cation channel in inside-out membrane patches from the dendritic tip of VNO neurons. (A) Addition of the endogenous DAG analogue 1-stearoyl-2-arachidonoyl-sn-glycerol (SAG, 100 lM) to the bathing medium produces an increase in channel activity that gives rise to a sustained inward current. This effect is accompanied by an increased noise level that reflects fluctuations in the activity of single channels. The presence of ATP, GTP, Ca2+, or Mg2+ in the bath is not required for activation of this current (holding potential, –80 mV). (B, C) Examples of single-channel events before (B) and after (C) application of SAG (10 lM). The SAGactivated channel has a unitary conductance of 42 pS in divalent cationfree solutions with Na+ as the sole cation and shows none of the hallmarks of capacitative Ca2+ entry channels (holding potential, –80 mV). From [19]

identified by Liman et al. in 1999 [45]. TRPC2 was identified in a PCR screen for ion channels in the VNO that were related to dTRP, the ion channel that mediates phototransduction in the fly (see Chapter 8). Because dTRP acts downstream of a PLC signaling cascade, it was hypothesized that a related channel would mediate PLC-dependent signaling in the VNO. As will be discussed below, this hypothesis has now been confirmed. The TRPC2 message encodes a protein of 885 amino acids in the rat and 890 amino acids in the mouse. It is distantly related to other mammalian TRPC channels (30 % amino acid identity), and, like other TRP channels, contains six trans-

6.7 Identification of the TRPC2 Ion Channel as a Candidate Transduction Channel

Fig. 6.4 Structure of the TRPC2 ion channel. (A) The rat TRPC2 protein contains 885 amino acids with an ankyrin repeat domain in the N-terminus and a coiled-coil domain in the C-terminus. The six transmembrane (tm) domains are thought to fold like those of a K+ channel, with a proposed pore region between the fifth and sixth tm domains. TRPC2 is likely to form a tetramer, like the K+ channel. (B) Phylogenetic tree of mammalian TRPC channels. TRPC2 is the most divergent of the mammalian TRPC channels

membrane domains (Fig. 6.4). Like voltage-gated K+ channels [46, 47], it is assumed to assemble as a tetramer. TRPC2 is expressed uniquely and abundantly in the VNO, with expression restricted to VNO sensory neurons [45] (Fig. 6.5). Since the original cloning of TRPC2 from the VNO, additional, longer splice isoforms have been identified [48, 49]. At present the significance of these is under dispute, and additional reports have confirmed that the isoform originally identified in the VNO is specifically expressed only in the VNO [49]. Further evidence that the short isoform is indeed the only TRPC2 protein expressed in the VNO comes from an examination of the protein by Western blot analysis using an antibody raised against the C-terminus of the TRPC2 protein. This shows that a

Fig. 6.5 Expression of TRPC2 in the VNO. (A) Labeling of a section of VNO from an adult rat with a digoxigenic antisense probe directed against TRPC2 reveals strong expression (blue reaction product) of the TRPC2 mRNA in VNO sensory neurons (N). (B) TRPC2 protein, detected by labeling with an antiTRPC2 antibody (red), is localized to the luminal surface of the epithelium. (C) In singly dissociated VNO sensory neurons, TRPC2 immunoreactivity (red) is clearly seen in the tuft of microvilli at the distal end of the dendrite. Scale bars: (B) 100 lm (C) 5 lm From [45]

143

144

6 Transduction Channels in the Vomeronasal Organ

protein of identical molecular weight is expressed in the VNO and in HEK cells transfected with the short isoform of TRPC2 [45]. To determine whether TRPC2 is likely to mediate VNO transduction, Liman et al. [45] localized the protein in sections of VNO and in singly dissociated VNO neurons (Fig. 6.5). These experiments showed a striking restriction of the TRPC2 protein to sensory microvilli, strongly supporting the view that the channel plays a role in sensory transduction [45]. This is further supported by immuno-EM localization of TRPC2 [34].

6.8

TRPC2 Is Essential for Pheromone Transduction

If TRPC2 is a major component of the VNO transduction channel, then targeted deletion of the channel should impair VNO sensory transduction. Indeed, two groups have independently generated TRPC2 -/- mice and shown that in these mice there is little or no electrical response to putative pheromones [14, 50]. Stowers et al. tested TRPC2 -/mice with a 64-microelectrode array and found a complete absence of electrical responses to dilute urine. Leypold et al. measured field potentials in response to dilute urine and reported a severe reduction in the amplitude and sensitivity of the response, although a small residual field potential was observed. Furthermore, as described below, TRPC2 -/- mice show severe behavioral defects in response to pheromones, further supporting the notion that TRPC2 is essential for pheromone transduction.

6.9

Mechanism of TRPC2 Activation

Identification of the mechanism by which TRPC2 is gated has been impeded by the fact that, in heterologous cell types, TRPC2 is retained in the endoplasmic reticulum [49]. TRPC2 is structurally related to ion channels that mediate Drosophila phototransduction. Considerable literature has provided a detailed understanding of phototransduction in Drosophila, which therefore serves as a model for understanding VNO transduction [51] (see Chapter 8). In the Drosophila rhabdomere, light-induced activation of a PLC leads to the opening of a light-activated conductance that is composed of three types of TRP channels, dTRP, dTRP-like, and TRPc. Recent evidence suggests that these TRP channels are activated by DAG or its metabolites (polyunsaturated fatty acids), possibly in combination with the reduction in phosphatidyl inositol 4,5 bisphosphate (PIP2) [52, 53]. Other mammalian homologues of the Drosophila TRPs are also activated by downstream products of PLC. Two of these (hTRPC3 and hTRPC6) appear to be activated directly by DAG [54], while others may be activated by association with an IP3 receptor [55, 56]. A major insight into the mechanism of activation of TRPC2 channels has come from comparing second messenger-gated ion channels in VNO sensory neurons from wild-

6.9 Mechanism of TRPC2 Activation

type (wt) and TRPC2 mutant mice [19]. This work revealed a novel Ca2+-permeable cation channel that exists in high density in the plasma membrane at the dendritic tips of wt VNO neurons and that is defective in TRPC2 -/- mice (Fig. 6.6). The channel can be gated effectively by the endogenous DAG analogue 1-stearoyl-2-arachidonoylsn-glycerol (SAG), as well as the other DAG analogues including OAG and DOG, but not by a monoacylglycerol or by IP3. Channel activation by DAG is independent of protein kinase C, Ca2+, and Mg2+, suggesting that DAG exerts a direct effect on the channel, not unlike its effects on hTRPC3 and hTRPC6 [54] . In TRPC2 -/- mice, the DAG-gated current is conspicuously absent, with a reduction in the maximum current in response to SAG of  8 % of wt levels. Thus, it is very likely that the TRPC2 is a major component of the DAG-activated channel. The presence of a small but significant residual conductance in TRPC2-/- sensory neurons suggests that other DAG-activated channel subunits may exist in these cells. One possibility is that TRPC2 forms a heteromultimeric channel complex with these predicted channel subunits, which otherwise function only poorly on their own. The mechanism of activation of TRPC2 deduced from the work of Lucas et al. [19] is inconsistent with the proposal that TRPC2 is gated by store-dependent Ca2+ mobilization or by a possible postulated interaction of TRPC2 with an IP3 receptor [48, 57]. These mechanisms are further unlikely given the strict localization of TRPC2 to sensory microvilli that are devoid of membranous compartments that could serve as Ca2+ stores [34, 45].

Fig. 6.6 TRPC2-/- VNO neurons display a striking defect in the activation of the DAG-gated channel. (A) Representative families of whole-cell currents to a series of depolarizing and hyperpolarizing voltage steps (as indicated in the figure) recorded from isolated wild-type VNO neurons under control conditions. Experiments are performed in the presence of 1 lM tetrodotoxin to block voltage-gated Na+ channels; voltage-activated K+ channels are blocked by using a Cs+-based pipette solution. Dotted line, zero current level. (B) A prominent DAG-gated conductance is observed in these cells following the addition of 100 lM SAG to the bath solution. (C, D) In VNO neurons from TRPC2-/- mice, SAG application fails to activate a large conductance. However, a drastically diminished residual response to SAG still exists in these cells (D). From [19]

145

146

6 Transduction Channels in the Vomeronasal Organ

6.10

TRPC2 Knockout Mice: Behavioral Defects

Previous understanding of the function of the VNO in animal behavior came primarily from studies in which the VNO was surgically ablated in animals and the effects of the ablation were assessed by behavioral assays after the animals recovered [58]. However, this approach is limited because few researchers are experienced in the surgical procedures and because the animals must be exposed to major surgery, which might mask subtle effects of the VNO ablation. The genetic ablation of TRPC2, which severely impairs the VNO sensory response, thus provides an excellent model system in which to determine the biological role of the VNO. These animals have a nonfunctional VNO, although in other ways they appear to be normal. For example, contrary to a report that the TRPC2 plays a role in sperm acrosome reaction [59], these mice are fertile and indistinguishable from wild-type littermates in the number of offspring. The behavior of the TRPC2 knockout mouse has been examined independently by two labs that have reached broadly similar conclusions [14, 50]. Two key results have emerged. First, TRPC2 is essential for pheromone-evoked male-male aggression (Fig. 6.7). In a resident-intruder assay, which tests for male-male aggression, TRPC2-/males fail to initiate attack behavior, although they are physically and neurologically capable of displaying aggressive interactions [14, 50]. Interestingly, presumably as a result of this deficit in displaying aggressive behavior in response to male pheromones, TRPC2-/- males usually fail to establish dominance and instead display urine-marking behavior typical of subordinate males [14]. Aggressive behavior is also severely attenuated in lactating female TRPC2-/- mice that are confronted with a

Fig. 6.7 TRPC2 knockout mice show behavioral abnormalities. (A) Average attack frequency of a resident male mouse of the genotype indicated in response to a castrated male intruder swabbed or not swabbed with pheromones from a sexually mature mouse. Note that male pheromones elicit attack behavior in the wt but not the TRPC-/mouse. (B) Mounting behavior of wt and TRPC2 -/- mice to same intruder mice. There is a dramatic increase in mounting behavior in the TRPC2-/- mouse From [50]

6.11 Loss of VNO Signaling Components in Human Evolution

male intruder, indicating that signals transduced by the VNO initiate aggressive behavior in both males and females [14]. Second, a striking defect is seen in the sexual behavior of TRPC2-/- males (Fig. 6.7). Although TRPC2-/- males mate normally with females, they display increased sexual behavior towards other males, i.e., mounting other males at a much higher rate [14, 50]. This behavior has not been observed in animals in which the VNO was surgically ablated, possibly due to secondary effects of the surgery [58]. This unexpected result has been interpreted as evidence that TRPC2-mediated signaling may be essential for gender discrimination [50]. One possible model consistent with these data is that mounting is an innate behavior that is inhibited by male pheromones acting through the VNO. TRPC2-/- males, therefore, persist in mounting other males. Because of the absence of major defects in male-female sexual behavior in TRPC2-/- mice, pheromones or other sensory cues essential for mating may not be detected by the VNO, but rather by other sensory systems such as the main olfactory epithelium. On the basis of these results, it is now clear that TRPC2 is essential for the detection of male-specific cues in the VNO that, in turn, regulate the expression of complex behavioral repertoires including aggressive and sexual behaviors.

6.11

Loss of VNO Signaling Components in Human Evolution

Whether humans have a functional VNO has until recently been a contentious question; a small pit is found in the nasal septum of most humans, but it has been difficult if not impossible to determine whether this pit contains functional VNO sensory neurons [60]. Recent evidence from molecular studies now strongly suggests that the human VNO is vestigial. Notably, of the several hundred VNO receptor genes in the human genome, nearly all are pseudogenes and only five contain intact reading frames [61]. Moreover, the TRPC2 gene, which is essential for VNO function in the mouse, is a pseudogene in humans [45, 62]. Thus, unless one is to hypothesize that a whole new set of genes is used for VNO transduction in humans, it is hard to escape the conclusion that the human VNO is vestigial. When did the human VNO become vestigial and why? Because the TRPC2 gene is expressed uniquely in the VNO and is essential for VNO function, the loss of a TRPC2 gene can serve as a marker for the loss of VNO function. Based on this reasoning, Liman and Innan [63] and Zhang and Webb [64] examined sequences of the TRPC2 gene from a large number of primate species. The human TRPC2 gene has six mutations that generate premature stop codons, resulting in a severely truncated protein. The earliest mutation is a nonsense mutation that is shared by all old world (OW) monkeys and apes and that is predicted to generate a protein that is missing much of its C-terminus [63, 64] (Fig. 6.8). Because this mutation occurs in a wellconserved region of the protein, it is likely to impair functioning. Thus, based on the observation that this mutation is found in all OW monkeys and apes but not in new world (NW) monkeys, we can date the loss of a functional TRPC2 gene in the human

147

148

6 Transduction Channels in the Vomeronasal Organ Fig. 6.8 The earliest mutation in the TRPC2 gene occurred in the common ancestor of OW monkeys and apes. (A) Sequence of the TRPC2 from monkeys was examined for the presence of stop codons or frameshift mutations. Mutations are indicated by the numbers 1–9, placed at the point in the phylogenetic tree where the mutation is inferred to have occurred. Note that mutation 9 was found in OW monkeys and gibbon but not in other apes, indicating that it either arose twice or that there was a reversion event (indicated by a white 9 on a black background). (B) Schematic representation of the TRPC2 gene, indicating the position of each mutation. Black bars represent transmembrane domains (From [63]).

lineage to 25–40 million years ago. This dating is further supported by an examination of selective pressure on the TRPC2 gene, which was also relaxed at this time [63]. If humans have lost a functioning VNO, how can we explain the presence of V1R genes with intact open reading frames (ORFs) in the human genome? Zhang and Webb [64] have considered the possibility that these genes are relics of an incomplete process of pseudogenization of this large family of receptors. Indeed, mathematical models show that approximately five receptors are expected to be intact if the initial repertoire was similar in size to that of the mouse (140 intact ORFs) and selective pressure was relaxed 23 million years ago. This timing is consistent with previous anatomical studies showing that a well-developed VNO and a morphologically distinct AOB were also lost in the common ancestor of OW monkeys and apes [65, 66]. Why did the ancestor of apes and OW monkeys lose a functioning VNO, while NW monkeys retained a functioning VNO? Interestingly, around the same time that the VNO was lost, the common ancestor of OW monkeys and apes acquired trichromatic vision through a duplication of the green opsin gene [67, 68]. Trichromatic vision, which employs three cone types, provides improved discrimination among colors from green to red and has therefore been postulated to have arisen as an adaptation to foraging for fruits or plants among a dense green canopy [68]. This enhanced visual capacity may also have proven useful for discriminating reproductive and social status

6.12 Summary: Is TRPC2 the VNO Transduction Channel?

among our distant ancestor, as seen in the highly colorful sexual skin of many OW monkeys and apes. Thus, vision that allowed discrimination at a distance may have replaced VNO-dependent pheromone signaling in mediating social interactions in humans.

6.12

Summary: Is TRPC2 the VNO Transduction Channel?

The evidence reviewed here suggests a model for VNO sensory transduction (Fig. 6.9) whereby binding of pheromones or other signaling molecules to V1R or V2R receptors leads to activation of a PLC that hydrolyzes PIP2 to IP3 and DAG. DAG then acts directly on the ion channel TRPC2, which allows an influx of Ca2+ and Na+ into the cell. Ca2+ may further act on Ca2+-activated nonselective channels that further amplify the electrical signal. The sensory current leads to a depolarization and the generation of actions potentials. The following evidence supports that TRPC2 is a major component of the VNO transduction channel: 1. It is in the right place. TRPC2 appears to be exclusively expressed in the VNO and the protein is restricted to the sensory microvilli [45]. 2. It is structurally related to other channels known to mediate sensory transduction in other systems. 3. Targeted deletion of TRPC2 largely eliminates sensory responses to pheromones [14, 50]. A possible concern with these experiments is that the loss of sensory responses is due to a nonspecific effect, such as cell death or developmental arrest. This does not appear to be the case, as although there is some loss of neurons in TRPC2-/- mice, the remaining neurons appear healthy and can fire action potentials in response to depolarization. 4. Targeted deletion of TRPC2 leads to the specific absence of a DAG-activated current [19]. With properties similar to that of the current activated in response to activation of the sensory signaling pathway.

2+

Na+

Ca

Pheromone

Na+

Ca2+

TRPC2 Fig. 6.9 Proposed mechanism of VNO sensory transduction. V1R and V2R receptors couple through Gai2 and Gao to a PLC, leading to hydrolysis of PIP2 into IP3 and DAG. DAG acts directly on the TRPC2 ion channel to allow an influx of Ca2+ and Na+

VR G

PLC

2

+++

DAG +

Na

IP3

Ca2+

149

150

6 Transduction Channels in the Vomeronasal Organ

To further confirm the role of TRPC2 in VNO sensory transduction, it will be important to see that sensory responses in TRPC2-/- mice can be rescued by expression of TRPC2 and to determine whether in a heterologous cell type, TRPC2 makes a DAGactivated channel.

Acknowledgements

We thank C. Dulac for critical reading of this chapter. The authors’ research is supported by grants from the National Institutes of Health/National Institute on Deafness and other Communication Disorders to E.R.L. and F.Z. References 1

2

3

4

5

6

7

8

9

10

Karlson, P. and M. Luscher, ’Pheromones’: a new term for a class of biologically active substances. Nature, 1959. 183: p. 55–56. Sam, M., et al., Neuropharmacology. Odorants may arouse instinctive behaviours. Nature, 2001. 412(6843): p. 142. Trinh, K. and D.R. Storm, Vomeronasal organ detects odorants in absence of signaling through main olfactory epithelium. Nat Neurosci, 2003. 6(5): p. 519–25. Halpern, M., The organization and function of the vomeronasal system. Annual review of Neuroscience, 1987. 10: p. 325–62. Dulac, C. and A.T. Torello, Molecular detection of pheromone signals in mammals: from genes to behaviour. Nat Rev Neurosci, 2003. 4(7): p. 551–62. Meredith, M., et al., Vomeronasal pump: significance for male hamster sexual behavior. Science, 1980. 207(4436): p. 1224–6. Luo, M., M.S. Fee, and L.C. Katz, Encoding pheromonal signals in the accessory olfactory bulb of behaving mice. Science, 2003. 299(5610): p. 1196–201. Martinez-Marcos, A. and M. Halpern, Differential projections from the anterior and posterior divisions of the accessory olfactory bulb to the medial amygdala in the opossum, Monodelphis domestica. Eur J Neurosci, 1999. 11(11): p. 3789–99. von Campenhausen, H. and K. Mori, Convergence of segregated pheromonal pathways from the accessory olfactory bulb to the cortex in the mouse. Eur J Neurosci, 2000. 12(1): p. 33–46. Moss, R.L., et al., Urine-derived compound evokes membrane responses in mouse vomeronasal receptor neurons. J Neurophysiol, 1997. 77(5): p. 2856–62.

11

12

13

14

15

16

17

18

19

20

Moss, R.L., et al., Electrophysiological and biochemical responses of mouse vomeronasal receptor cells to urine-derived compounds: possible mechanism of action. Chem Senses, 1998. 23(4): p. 483–9. Leinders-Zufall, T., et al., Ultrasensitive pheromone detection by mammalian vomeronasal neurons. Nature, 2000. 405(6788): p. 792–6. Holy, T.E., C. Dulac, and M. Meister, Responses of vomeronasal neurons to natural stimuli. Science, 2000. 289(5484): p. 1569–72. Leypold, B.G., et al., Altered sexual and social behaviors in trp2 mutant mice. Proc Natl Acad Sci U S A, 2002. 99(9): p. 6376–81. Del Punta, K., et al., Deficient pheromone responses in mice lacking a cluster of vomeronasal receptor genes. Nature, 2002. 419(6902): p. 70–4. Inamura, K., et al., Laminar distribution of pheromone-receptive neurons in rat vomeronasal epithelium. J Physiol (Lond), 1999. 517(Pt 3): p. 731–9. Cinelli, A.R., et al., Calcium transients in the garter snake vomeronasal organ. J Neurophysiol, 2002. 87(3): p. 1449–72. Spehr, M., H. Hatt, and C.H. Wetzel, Arachidonic acid plays a role in rat vomeronasal signal transduction. J Neurosci, 2002. 22(19): p. 8429–37. Lucas, P., et al., A Diacylglycerol-Gated Cation Channel in Vomeronasal Neuron Dendrites Is Impaired in TRPC2 Mutant Mice: Mechanism of Pheromone Transduction. Neuron, 2003. 40: p. 551–561. Liman, E.R. and D.P. Corey, Electrophysiological characterization of chemosensory neurons from the mouse vomeronasal organ. J Neurosci, 1996. 16(15): p. 4625–4637.

6.12 Summary: Is TRPC2 the VNO Transduction Channel? 21

22

23

24

25

26

27

28

29

30

31

32

33

34

Dulac, C. and R. Axel, A novel family of genes encoding putative pheromone receptors in mammals. Cell, 1995. 83(2): p. 195–206. Berghard, A., L.B. Buck, and E.R. Liman, Evidence for distinct signaling mechanisms in two mammalian olfactory sense organs. Proc Natl Acad Sci U S A, 1996. 93(6): p. 2365–9. Wu, Y., R. Tirindelli, and N.J. Ryba, Evidence for different chemosensory signal transduction pathways in olfactory and vomeronasal neurons. Biochem Biophys Res Commun, 1996. 220(3): p. 900–4. Rodriguez, I., et al., Multiple new and isolated families within the mouse superfamily of V1r vomeronasal receptors. Nat Neurosci, 2002. 5(2): p. 134–40. Pantages, E. and C. Dulac, A novel family of candidate pheromone receptors in mammals. Neuron, 2000. 28(3): p. 835–45. Herrada, G. and C. Dulac, A novel family of putative pheromone receptors in mammals with a topographically organized and sexually dimorphic distribution. Cell, 1997. 90(4): p. 763–73. Matsunami, H. and L.B. Buck, A multigene family encoding a diverse array of putative pheromone receptors in mammals. Cell, 1997. 90: p. 775–784. Ryba, N.J. and R. Tirindelli, A new multigene family of putative pheromone receptors. Neuron, 1997. 19(2): p. 371–9. Loconto, J., et al., Functional expression of murine V2R pheromone receptors involves selective association with the M10 and M1 families of MHC class Ib molecules. Cell, 2003. 112(5): p. 607–18. Boschat, C., et al., Pheromone detection mediated by a V1r vomeronasal receptor. 2002. 5(12): p. 1261–1262. Berghard, A. and L. Buck, Sensory transduction in vomeronasal neurons: Evidence for Gao, Gai, and adenylyl cyclase II as major components of a pheromone signalling cascade. J Neurosci, 1996. 16: p. 909–918. Halpern, M., L.S. Shapiro, and C. Jia, Differential localization of G proteins in the opossum vomeronasal system. Brain Research, 1995. 677(1): p. 157–61. Krieger, J., et al., Selective activation of G protein subtypes in the vomeronasal organ upon stimulation with urine-derived compounds. J Biol Chem, 1999. 274(8): p. 4655–62. Menco, B.P., et al., Ultrastructural localization of G-proteins and the channel protein TRP2 to microvilli of rat vomeronasal receptor cells. J Comp Neurol, 2001. 438(4): p. 468–89.

35

36

37

38

39

40

41

42

43

44

45

46

47

Norlin, E.M., F. Gussing, and A. Berghard, Vomeronasal phenotype and behavioral alterations in G alpha i2 mutant mice. Curr Biol, 2003. 13(14): p. 1214–9. Luo, Y., et al., Identification of chemoattractant receptors and G-proteins in the vomeronasal system of garter snakes. J Biol Chem, 1994. 269(24): p. 16867–77. Kroner, C., et al., Pheromone-Induced Second Messenger Signaling In the Hamster Vomeronasal Organ. Neuroreport, 1996. 7(18): p. 2989–2992. Wekesa, K.S. and R.R. Anholt, Pheromone regulated production of inositol-(1, 4, 5)-trisphosphate in the mammalian vomeronasal organ. Endocrinology, 1997. 138(8): p. 3497–504. Runnenburger, K., H. Breer, and I. Boekhoff, Selective G protein beta gamma-subunit compositions mediate phospholipase C activation in the vomeronasal organ. Eur J Cell Biol, 2002. 81(10): p. 539–47. Liman, E.R., Regulation by voltage and adenine nucleotides of a Ca2+-activated cation channel from hamster vomeronasal sensory neurons. J Physiol, 2003. 548(Pt 3): p. 777–87. Taniguchi, M., M. Kashiwayanagi, and K. Kurihara, Intracellular injection of inositol 1,4,5trisphosphate increases a conductance in membranes of turtle vomeronasal receptor neurons in the slice preparation. Neuroscience Letters, 1995. 188(1): p. 5–8. Inamura, K., M. Kashiwayanagi, and K. Kurihara, Inositol-1,4,5-trisphosphate induces responses in receptor neurons in rat vomeronasal sensory slices. Chemical Senses, 1997. 22(1): p. 93–103. Taniguchi, M., D. Wang, and M. Halpern, Chemosensitive conductance and inositol 1,4,5trisphosphate-induced conductance in snake vomeronasal receptor neurons. Chem Senses, 2000. 25(1): p. 67–76. Clapham, D.E., L.W. Runnels, and C. Strubing, The TRP ion channel family. Nat Rev Neurosci, 2001. 2(6): p. 387–96. Liman, E.R., D.P. Corey, and C. Dulac, TRP2: a candidate transduction channel for mammalian pheromone sensory signaling. Proc Natl Acad Sci U S A, 1999. 96(10): p. 5791–6. MacKinnon, R., Determination of the subunit stoichiometry of a voltage-activated potassium channel. Nature, 1991. 350(6315): p. 232–5. Liman, E.R., J. Tytgat, and P. Hess, Subunit stoichiometry of a mammalian K+ channel determined by construction of multimeric cDNAs. Neuron, 1992. 9(5): p. 861–71.

151

152

6 Transduction Channels in the Vomeronasal Organ 48

49

50

51

52

53

54

55

56

57

58

59

60

61

Vannier, B., et al., Mouse trp2, the homologue of the human trpc2 pseudogene, encodes mTrp2, a store depletion-activated capacitative Ca2+ entry channel. Proc Natl Acad Sci U S A, 1999. 96(5): p. 2060–4. Hofmann, T., et al., Cloning, expression and subcellular localization of two novel splice variants of mouse transient receptor potential channel 2. Biochem J, 2000. 351(Pt 1): p. 115–22. Stowers, L., et al., Loss of sex discrimination and male-male aggression in mice deficient for TRP2. Science, 2002. 295(5559): p. 1493–500. Hardie, R.C. and P. Raghu, Visual transduction in Drosophila. Nature, 2001. 413(6852): p. 186–93. Chyb, S., P. Raghu, and R.C. Hardie, Polyunsaturated fatty acids activate the Drosophila light-sensitive channels TRP and TRPL. Nature, 1999. 397(6716): p. 255–9. Hardie, R.C., Regulation of TRP channels via lipid second messengers. Annu Rev Physiol, 2003. 65: p. 735–59. Hofmann, T., et al., Direct activation of human TRPC6 and TRPC3 channels by diacylglycerol. Nature, 1999. 397(6716): p. 259–63. Kiselyov, K., et al., Functional interaction between InsP3 receptors and store-operated Htrp3 channels. Nature, 1998. 396(6710): p. 478–82. Yuan, J.P., et al., Homer binds TRPC family channels and is required for gating of TRPC1 by IP3 receptors. Cell, 2003. 114(6): p. 777–89. Brann, J.H., et al., Type-specific inositol 1,4,5trisphosphate receptor localization in the vomeronasal organ and its interaction with a transient receptor potential channel, TRPC2. J Neurochem, 2002. 83(6): p. 1452–60. Wysocki, C.J. and J.J. Lepri, Consequences of removing the vomeronasal organ. J Steroid Biochem Mol Biol, 1991. 39(4B): p. 661–9. Jungnickel, M.K., et al., Trp2 regulates entry of Ca2+ into mouse sperm triggered by egg ZP3. Nat Cell Biol, 2001. 3(5): p. 499–502. Meredith, M., Human vomeronasal organ function: a critical review of best and worst cases. Chem Senses, 2001. 26(4): p. 433–45. Rodriguez, I. and P. Mombaerts, Novel human vomeronasal receptor-like genes reveal speciesspecific families. Curr Biol, 2002. 12(12): p. R409–11.

62

63

64

65

66

67

68

69

70

71

72

73

Wes, P.D., et al., TRPC1, a human homolog of a Drosophila store-operated channel. Proc Natl Acad Sci U S A, 1995. 92: p. 9652–9656. Liman, E.R. and H. Innan, Relaxed selective pressure on an essential component of pheromone transduction in primate evolution. Proc Natl Acad Sci U S A, 2003. 100(6): p. 3328–32. Zhang, J. and D.M. Webb, Evolutionary deterioration of the vomeronasal pheromone transduction pathway in catarrhine primates. Proc Natl Acad Sci U S A, 2003. 100(14): p. 8337–41. Bhatnagar, K.P. and E. Meisami, Vomeronasal organ in bats and primates: extremes of structural variability and its phylogenetic implications. Microsc Res Tech, 1998. 43(6): p. 465–75. Meisami, E. and K.P. Bhatnagar, Structure and diversity in mammalian accessory olfactory bulb. Microsc Res Tech, 1998. 43(6): p. 476–99. Hunt, D.M., et al., Molecular evolution of trichromacy in primates. Vision Res, 1998. 38(21): p. 3299–306. Surridge, A.K., D. Osorio, and N.I. Mundy, Evolution and selection of trichromatic vision in primate. Trends in Ecology and Evolution, 2003. 18(4): p. 198–205. Liman, E.R., Pheromone transduction in the vomeronasal organ. Current Opinion in Neurobiology, 1996. 6: p. 487–493. Liu, D., and E.R. Liman, Intracellular Ca2+ and the phospholipid PIP2 regulate the taste transduction ion channel TRPM5. Proc Natl Acad Sci USA, 2003. 100: p. 15160–15165. Prawitt, D. et al., TRPM5 is a transient Ca2+activated cation channel responding to rapid changes in [Ca2+]i. Proc Natl Acad Sci USA, 2003. 100: p. 15166–15171. Launay, P. et al., TRPM4 is a Ca2+-activated nonselective cation channel mediating cell membrane depolarization. Cell, 2002. 109: p. 397–407. Hofmann, T., et al., TRPM5 is a voltage-modulated and Ca(2+)-activated monovalent selective cation channel. Curr Biol, 2003. 13: p. 1153–1158.

153

7

Transduction Mechanisms in Taste Cells Kathryn Medler and Sue C. Kinnamon

7.1

Introduction

All organisms respond to chemicals found in their external environment. These responses detect nutrients, conspecifics, predators, or potentially harmful conditions. The ability to discriminate items with nutritional value from items that are harmful depends on chemical detectors called taste receptor cells. Taste receptor cells are specialized neuroepithelial cells that are housed as taste buds in the lingual epithelium of the oral cavity. Taste buds consist of 50 to 150 taste receptor cells that extend apical processes into the oral cavity to detect stimuli and form synapses with gustatory neurons to send information to the brain (Fig. 7.1). Taste cells have properties of both neurons and epithelial cells and are responsible for converting chemical stimuli into electrical signals. Like neurons, they are capable of generating receptor and/or action potentials and can form classical chemical synapses with gustatory neurons [1–4]. Taste cells express TTX-sensitive voltage-gated Na+ channels and multiple types of voltage-gated K+ and Ca2+ channels [5–7]. However, like epithelial cells, taste cells express several types of epithelial ion channels and have a limited life span of about two weeks. As a result of turnover, gustatory neurons must disconnect from dying taste cells to form new synaptic contacts with emerging taste cells [1, 4]. In mammals, taste buds are found in specialized protrusions called papillae. These papillae consist of fungiform, foliate, and circumvallate on the tongue, with additional taste buds present on the soft palate, scattered throughout the oral cavity and in the “geschmacksstreifen,” a row of taste buds between the hard and soft palate. Taste buds are innervated by the facial, (VII), glossopharyngeal (IX), or vagus (X) cranial nerves, depending on their location in the tongue. A branch of the facial nerve, the chorda tympani, innervates the fungiform papillae and the anterior portion of the foliate papillae, while the glossopharyngeal nerve innervates the rest of the foliate and the circumvallate papillae. The greater superficial petrosal branch of the facial nerve innervates the taste buds found in the soft palate and geschmacksstreifen [8, 9]. Taste receptor cells detect a wide range of chemically diverse stimuli that comprise the different taste qualities: sweet, salty, sour, bitter, and umami, the detection of Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

154

7 Transduction Mechanisms in Taste Cells

Fig. 7.1 Morphology of the peripheral taste system. The upper panel shows a schematic of a mammalian tongue, showing fungiform, circumvallate, and foliate papillae and their innervation. A taste bud is illustrated in the bottom panel. The taste pore is the site of interaction with taste stimuli, while afferent gustatory neurons carry information from the taste cells to the brain

amino acids, primarily glutamate. These taste qualities can be divided into two broader categories, ionic and complex, based on the chemical structure of the taste stimulus. Ionic stimuli include Na+ and H+, which elicit salty and sour tastes, respectively. These ions can interact directly with apically located ion channels to depolarize taste cells, by either permeating or blocking the channel, or they can diffuse through tight junctions to interact with basolateral ion channels. More chemically complex stimuli, including sugars, amino acids, and most bitter stimuli, interact with apically located G-proteincoupled receptors (GPCR) or, in some cases, ligand-gated channels. Because taste is tied to the nutritional needs of an organism, there is considerable species diversity in transduction mechanisms. This necessitates multiple mechanisms to detect different stimuli. In this chapter, we will describe both ion channels directly modulated by taste stimuli and ion-channel targets of second messengers generated as a result of GPCR signaling. When possible, we will discuss their role in taste trans-

7.2 Ionic Stimuli

duction. Describing all the channels that have been proposed to mediate transduction in different species is beyond the scope of this review. Instead, we will focus on the best-characterized channels, for which there is either direct molecular or physiological evidence for their role in transduction.

7.2

Ionic Stimuli

Two taste qualities, salty and sour, detect the presence of ions in the oral cavity. Due to their simple structure and charge, ions are capable of passing through ion channels on the apical membrane of taste cells to directly depolarize taste cells. Since ions can also pass through tight junctions between taste cells, they are able to interact at the basolateral membrane in addition to the apical pore.

7.2.1

Salt

Salt taste is the detection of cations, such as Na+, K+, and Li+. This leads to the ingestion of NaCl and other required minerals, thus helping to maintain ion and water homeostasis. The specific ionic needs vary across species depending on the ionic content of their diet and surrounding environment. When NaCl is present in sufficient levels in the oral cavity, Na+ permeates apically located ion channels, causing membrane depolarization and activation of voltage-gated Na+ and K+ channels. In addition to interacting with apically located channels, NaCl and other monovalent salts diffuse passively across tight junctions, where they can potentially interact with basolateral channels. In addition, this paracellular shunt pathway generates a hyperpolarizing field potential around the taste cells that depresses the receptor potential and makes it less likely to fire action potentials. The field potential is influenced by the anion present. For example, the chloride ion (Cl–) of NaCl is very permeable and thus provides a shunting current that minimizes the hyperpolarizing field potential, while less conductive anions do not. This causes a difference in their saltiness perception as compared to NaCl [10, 11]. Studies to determine the role of the anion in salt transduction have found no evidence of Cl– co-transporters, exchangers, or channels at the apical membrane, further supporting the idea that anions influence salt perception via the paracellular pathway [12]. Epithelial Sodium Channel Salt taste transduction can be divided into an amiloride-sensitive component and an amiloride-insensitive component. Amiloride is a diuretic that blocks a resting Na+ conductance found in many species. When the channel was first characterized in taste cells, it was initially called the amiloride-sensitive sodium channel (ASSC), but it is now generally accepted that this channel is the epithelial sodium channel (ENaC), 7.2.1.1

155

156

7 Transduction Mechanisms in Taste Cells

a member of the degenerin family of channels that is widely expressed in epithelial tissues throughout the body [13]. The ENaC is a heteromultimer consisting of a-, b-, and c-ENaC subunits (Fig. 7.2) [14]. In expression systems, the a subunits are sufficient to induce channel activity, but the b and c subunits are needed to confer maximal Na+ current [15]. However, in taste cells, it is thought that all three subunits are needed to form a functional channel. Immunocytochemical analyses in rat taste cells demonstrate labeling in nearly all the fungiform papillae and in about half of the foliate and vallate papillae. The labeling intensity is significantly lower in the vallate papillae compared to the fungiform, especially for the b and c subunits [16–18]. Interestingly, treatment with aldosterone increases the apical immunoreactivity of the b and c ENaC subunits in all papillae and increases both the number of amiloride-sensitive taste cells and the amplitude of the response. Additionally, aldosterone treatment induces an amiloride-sensitive current in about half the vallate taste cells, which are normally unresponsive to amiloride. The upregulation of b and c ENaC subunits by aldosterone seems to induce or increase the channel activity [16], which is consistent with its role in regulating ENaCs in other tissues. Several hormones that are involved in osmotic regulation and Na+ balance have been found to regulate ENaC expression. Like aldosterone, arginine vasopressin (AVP) increases Na+ transport in taste tissue. AVP is thought to act on ENaCs via the V2-type vasopressin receptors, as its effects are mimicked by cAMP. These hormones increase the number of ENaCs expressed at the apical membrane, presumably through a receptor-mediated increase in intracellular cAMP levels [19]. Meanwhile, both atrial natriuretic peptide and oxytocin cause decreases in Na+ transport in lingual epithelia. These data suggest that hormones may regulate the sensitivity of salt detection [20]. In taste cells, the ENaC is constitutively active and has a very high affinity for amiloride, with a Ki in the sub-micromolar concentration range. The amiloride dose-response curve can be fitted with a single binding isotherm, indicating a possible 1:1 stoichiometry for the drug-channel interaction [18]. In rat taste cells, it has been shown that

Fig. 7.2 Membrane topology of each subunit (a, b, and c) ENaC. M1 and M2 indicate helical transmembrane domains; CRD1 and CRD2 indicate cysteine-rich domains; glycosylation sites are indicated (6 on a-ENaC, 12 on b-ENaC; and 5 on

c-ENaC). The filled bar within CRD1 of a-ENaC indicates the relative position of a six-amino-acid region known to bind amiloride. Reprinted from [14] with permission

7.2 Ionic Stimuli

Na+ inhibits its own influx through the ENaC. This Na+ self-inhibition is restricted to the amiloride-sensitive response and is slow, taking up to 15 s to become maximally inhibited. It is thought that Na+ binds the ENaC receptor on its extracellular face and reduces the Na+ permeability of the channel. While its role is not completely understood, Na+ self-inhibition may be involved in adaptation to Na+ salts or in the regulation of salt intake [21, 22]. The amiloride-sensitive conductance is believed to be the main contributor to salt transduction in animals that have inherently low salt diets. Studies in hamsters found that over half of all taste cells had some ENaC activity and that within the amiloridesensitive cells, ENaCs accounted for an average of 48 % of the total current generated (Fig. 7.3) [23]. This is true for most rodent species, while in humans the amilorideinsensitive response is predominant, with a relatively minor amiloride-sensitive response. However, to some degree, the salt response in all species studied is comprised of both amiloride-sensitive and -insensitive components. Aquatic animals, especially freshwater varieties, have a very strong driving force to maintain ionic homeostasis. Despite the different environmental pressures on the maintenance of water and ion balance in these animals, there are similarities in their salt transduction mechanisms, specifically the role of the ENaC. In frogs, more than 50 % of taste cells had a stationary inward Na+ current at –80 mV that was partially or fully blocked by amiloride. This conductance was slightly modulated by voltage but was not voltage gated [24]. However, there are some differences in the ENaC expressed in frog taste cells when compared with ENaCs expressed in other systems. The unitary conductance of the ENaC in frog skin epithelia is 6–10 pS compared to less than 2 pS for the taste ENaC. Additionally, the selectivity ratio for Na+ to K+ is less for the frog

Fig. 7.3 Amiloride inhibits the NaCl-induced currents. (A) Short circuit current resulting from NaCl solutions presented to rat tongue epithelium in the absence and presence of 10–4 M amiloride. (B) Time course of the open-circuit potential after

0.001 M NaCl was replaced with 1.5 M NaCl in the presence and absence of 10–4 M amiloride. Reprinted Fig. 1 with permission from Heck et al. (1984) Science 223:403–405 (Copyright 1984, AAAS)

157

158

7 Transduction Mechanisms in Taste Cells

taste channel than for ENaCs in mammalian taste cells and ENaCs in other tissues. Based on reversal potentials, the selectivity of this channel in frogs was K+>Na+>Li+>Cs+ [25]. Amiloride-insensitive Pathway The amiloride-insensitive pathway is not well characterized and is likely comprised of multiple mechanisms. One proposed mechanism is that the ENaC and/or nonselective cation channels on the basolateral membrane are permeable to Na+ that has diffused through tight junctions. Another possible mechanism is that an amiloride-insensitive Na+ channel is expressed on the apical membrane of taste cells. Physiological evidence supports both mechanisms, although mechanisms vary considerably in different species. Although the ENaC is normally restricted to the apical membrane in most transporting epithelia, basolateral expression has been proposed to mediate a portion of the amiloride-insensitive salt response in taste cells. Since amiloride cannot pass through tight junctions, it would not inhibit these channels. Voltage-clamp studies in rat found amiloride-sensitive channels on the basolateral membrane, though their properties differ from the apically located channels. Basolateral channels are less sensitive to amiloride, with a Ki of 52 lM compared with a Ki of 0.2 lM for apically located channels. ENaCs at the apical membrane are highly selective for Na+, with a Na+/ Cs+ permeability ratio greater than 10, while the channels at the basolateral membrane are not as selective for Na+ (PNa/PK =3.7) [26]. Two types of amiloride-insensitive responses have been characterized in mice in addition to the amiloride-sensitive response. The first response has a slight inward rectification, has a reversal potential close to the equilibrium potential for nonselective cation channels, and is blocked by Cd2+. The other response shows no rectification and has a reversal potential close to the equilibrium potential for Cl– at the basolateral 7.2.1.2

Fig. 7.4 The amiloride-insensitive response can be either enhanced or inhibited by cetylpyridinium chloride (CPC). (A) Effect of 250 lM CPC on the integrated chorda tympani response to 300 mM NaCl and 300 mM NaCl + 100 lM amiloride. CPC enhanced the response by the same magni-

tude in each case. (B) Effect of 2 mM CPC on the integrated chorda tympani response to 300 mM NaCl and 300 mM NaCl + 100 lM amiloride. CPC suppressed the entire amiloride-insensitive part of the response. Reprinted from [28] with permission

7.2 Ionic Stimuli

membrane, but not at the apical membrane. The Cl– channel blocker, NPPB, potentiates the salt response and abolishes the low reversal potential, indicating that NPPB blocks a Cl– channel. These data suggest that at least three components contribute to the salt response in mice: the amiloride-sensitive Na+ channel, an NPPB-sensitive Cl– channel, and a Cd2+-sensitive nonselective cation channel. The amiloride-sensitive Na+ channel and the nonselective cation channel are located at the apical membrane, while the Cl– channel is located on the basolateral membrane [27]. Evidence for apically located, amiloride-insensitive Na+ channels in rats comes from studies using a modulator, cetylpyridium chloride (CPC). At low concentrations, CPC enhances the amiloride-insensitive Na+ response, but at high concentrations the response is suppressed (Fig. 7.4). CPC also modulates responses to KCl and NH4Cl, suggesting that the channel may be nonselective for cations. Since this response is modulated by voltage fields applied across the epithelium, it is likely mediated by an apical conductance [28]. Further work will be required to characterize the properties of this conductance. In the frog, channels other than the ENaC are involved in transducing salt stimuli. Both Na+ and Cl– permeate apical channels believed to be gated by NaCl, leading to the generation of receptor potentials. This is in addition to cationic channels present on the basolateral membrane that also contribute to generation of the response to salty stimuli [29]. There is also evidence for two Na+-dependent K+ conductances in frog taste cells. Two different conductances, 35.8 pS and 9.4 pS, are activated by high internal NaCl. It is thought that these channels contribute to the stabilization of the resting potential and that they may be involved in salt transduction [30].

7.2.2

Sour

Sour taste is the detection of protons in the oral cavity. This taste quality serves to detect unripe or spoiled food in order to avoid ingestion of potentially harmful levels of acids. Multiple receptors have been proposed to serve as sour receptors, but to date, none have emerged as definitive. Because protons modulate most ionic conductances, nearly all taste cells respond to acid stimulation with changes in membrane conductance [31] and intracellular acidification [32]. However, it is not clear that all of these cells communicate this information to the nervous system. In slice preparations from mice, focal applications of acidic taste stimuli result in the acidification of the entire taste bud, but only a few taste cells also respond with increases in intracellular Ca2+. Cells showing acid-induced increases in intracellular Ca2+ also respond to KCl depolarization, and responses are blocked by Ba2+ and Cd2+, suggesting influx via voltagegated Ca2+ channels (Fig. 7.5) [33]. It seems probable that the taste cells expressing voltage-gated Ca2+ channels are the taste cells that generate signals to the nervous system in response to sour stimuli. Sour perception is affected by both extracellular pH (pHo), and the accompanying anions. At the same pH, weak acids produce larger nerve responses and a greater sour taste sensation when compared to strong acids [34]. Lyall et al. [32] hypothesize that

159

160

7 Transduction Mechanisms in Taste Cells Fig. 7.5 Ca2+ response to citric acid depends on Ca2+ influx through voltage-gated Ca2+ channels. (A) Superimposed Ca2+ responses in a mouse taste cell stimulated repeatedly with citric acid (pH 3, arrowheads) either before or after 2 mM Ba2+ in the bath. Mean (SEM) amplitudes of Ca2+ responses to citric acid shown in the bar graphs (*P<0.05 for control Ca2+ responses versus responses in the presence of 2 mM Ba2+, n=7 cells in each column). (B) Superimposed Ca2+ responses evoked by repeated focal stimulation with citric acid (pH 3, arrowheads) before and after 0.5 mM Cd2+ in the bath. Mean (SEM) amplitudes of responses shown in the bar graphs (*P<0.05 for control Ca2+ responses versus responses in the presence of Cd2+, n=4 cells in each column). Reprinted from [33] with permission

weak acids permeate taste cell membranes as neutral molecules but, upon entering the cell, dissociate and change pHi. This decrease in intracellular pH is believed to be a proximate stimulus for sour transduction, since intracellular acidification by CO2 at neutral extracellular pH is sufficient to activate taste nerve fibers [32]. However, taste cells also respond to strong, fully dissociated acids with decreases in pHi, suggesting that channels permeable to protons are expressed in taste cells. Further, several proton-gated and -blocked channels are expressed in taste cells, which could result in depolarization and voltage-gated Ca2+ influx in response to acidic stimuli. The determination of which channels function as sour receptors will likely depend on their

7.2 Ionic Stimuli

co-expression in the taste cells that increase intracellular Ca2+ in response to acidic stimuli.

Proton-permeable Channels In hamsters, the epithelial Na+ channel has been shown to conduct protons in the absence of extracellular Na+. Since salivary Na+ concentration is low in hamsters, it is likely that this channel contributes to sour transduction. Behavioral tests in hamsters have confirmed a role of the ENaC in sour transduction [35]. However, in mice amiloride does not suppress citric acid-induced responses, suggesting that this mechanism may be specific for hamsters [27, 36]. In mice, the Cl– channel blocker, NPPB, suppresses the citric acid-induced response. The reversal potential for the NPPB-sensitive component of the current is –2 mV, close to the reversal potential for Cl– ions [37]. In addition to the NPPB-sensitive Cl– channel, an amiloride-insensitive, H+ permeable cationic channel also contributes to the acid response in mice. Some studies suggest that the Cl– conductance is located on the basolateral membrane, while the proton permeable conductance is located on the apical membrane [38]. 7.2.2.1

7.2.2.2 Proton-gated Channels

The expression of several other channels has been implicated in sour transduction. Two isotypes of the hyperpolarization-activated channels (HCN), HCN1 and HCN4, have been identified in rat vallate papilla. These receptors are gated by extracellular protons, and lowering extracellular pH causes a dose-dependent flattening of the HCN activation curve, suggesting a role in sour taste [39]. In rat taste cells, molecular and immunocytochemical analyses have identified the expression of the proton-gated, acid-sensing ion channels, ASIC 2a and 2b [40, 41]. ASICs are Na+-selective, H+-gated channels that belong to the degenerin/ENaC superfamily [42, 43] (see Chapter 3). Expression studies of these channels in oocytes determined that the taste ASIC2a is a proton-gated, amiloride-sensitive channel with maximal inward currents at pH <2.0. This channel shows decreasing amiloride sensitivity with decreasing extracellular pH, suggesting that it may function as the amilorideinsensitive component of acid-induced responses in rat taste cells [40]. Physiological characterization of the acid-induced current in rat taste cells supports a role for ASICs in the detection of sour stimuli. Whole-cell recordings of rat vallate taste cells found that acids induce transient inward currents that are pH dependent and partially suppressed by amiloride. The reversal potential is near the Na+ equilibrium potential, suggesting that the acid-gated conductance is selective for Na+. However, additional acidic responses are generated that do not resemble ASIC currents [31] (Fig. 7.6). In frogs, sour taste is transduced primarily by apically located, acid-gated cation channels that are permeable to both monovalent and divalent cations. Two distinct responses to acidic stimuli have been found in frogs. Some taste cells depolarize in response to acidic stimuli, while about 30 % of the sour-sensitive taste cells generate

161

162

7 Transduction Mechanisms in Taste Cells

Fig. 7.6 Responses to citric acid in rat taste cells can closely resemble ASIC currents as seen in expression systems, though other responses also occur. (A) Functional expression of ASIC 2a and ASIC 2b. Representative whole-cell currents of ASIC 2 homomeric and heteromeric channels expressed in X. laevis oocytes. Acidic pH solutions were present in the bath during times indicated by bars. Reprinted with permission from Ugawa et al. (2003) J. Neuroscience 23(9): 3616–3622, Copyright 2003, the Society for Neuroscience. (B) Citric acid-induced currents in rat vallate taste cells. A large inward current in response to citric acid (pH 5) desensitized in response to prolonged application of citric acid. Note that the off response of inward current (arrow) developed upon removal of the stimulus, that a delayed outward current (double arrow) followed the inward currents, and that it is similar to (E). (C–E) Responses to application of citric acid (pH 5) that did not resemble the ASIC current profile. Responses included those with slow desensitization kinetics (C), a sustained response accompanied by a conductance decrease (D), and a slowly activating outward current with a decrease in conductance (E). Reprinted from [31] with permission

a hyperpolarizing response. This hyperpolarization is due to an increase in a protonsensitive K+ conductance that may be gated directly by extracellular H+ ions [44]. 7.2.2.3 Proton-blocked Channels

In mudpuppies, voltage-dependent K+ channels are differentially localized to the apical membrane of taste cells. The apical K+ conductance comprises several different channels, including large-conductance, Ca2+-dependent K+ channels and ATP-blocked K+ channels. These channels have a small open-probability at rest, so that block of the conductance generates a depolarizing receptor potential. Several stimuli act directly to block these channels, including acids and some bitter compounds such as quinine [45, 46]. In contrast, voltage-gated K+ channels appear to be restricted from the apical membrane in mammalian taste cells [47]. However, recent studies have characterized the

7.3 Complex Stimuli

expression of acid-sensitive, two-pore domain K+ channels (K2P or KCNK) in taste cells. Molecular and immunocytochemical analysis of rat taste cells detect multiple members of this gene family, including TWIK 1 (tandem of P domains in weak inward rectifier K+ channels), TASK 1-3 (TWIK-related acid-sensitive K+ channels), TREK 1-2 (TWIK-related K+ channels), and TRAAK 1 (TWIK-related arachidonic acid-activated K+ channels) [48, 49]. A subset of rat taste cells responds to small decreases in extracellular pH from 7.4 to 7.2 with a sustained decrease in membrane conductance, resulting in membrane depolarization. This conductance has pharmacological and physiological properties similar to that of expressed K2P channels, suggesting that they may contribute to sour transduction [49]. Since some two-pore domain channels are blocked by intracellular pH, these channels may be the missing link that couples a decrease in pHi to membrane depolarization and activation of voltage-gated Ca2+ channels.

7.3

Complex Stimuli

Most stimuli that elicit bitter, sweet, and umami (the detection of amino acids) taste are too bulky to permeate ion channels or to diffuse through tight junctions to interact with basolateral channels. Instead, these tastants primarily activate apically located, Gprotein-coupled receptors (GPCRs), leading to generation of second messengers and modulation of downstream effector channels. While GPCRs are critical for the transduction of most complex stimuli, some alternative pathways have been described. Many of these are species specific and not all are fully characterized. While a complete discussion is beyond the constraints of this review, some of these alternate pathways are described below. Bitter compounds include alkaloids, tannins, methylxanthines, peptides, and some non-sodium salts. Since many toxins taste bitter, it is believed that bitter taste evolved as a warning system to prevent ingestion of harmful compounds. Natural stimuli that elicit a sweet taste include sugars and some proteins. These stimuli usually signify a food source that is high in calories and thus elicit a feeding response in most animals. Amino acids act as taste stimuli for the consumption of protein rich foods. The taste of monosodium glutamate has been termed “umami,” which may represent a separate taste quality distinct from bitter, sweet, salty, and sour [2, 3, 50]. Other amino acids such as arginine and proline elicit a feeding response in several aquatic species, though their roles in other vertebrates have not been studied [51–53].

7.3.1

GPCR Signaling in Taste Cells

Two families of GPCRs have been identified that act as taste receptors: the T2Rs that function as bitter receptors [54, 55] and the T1Rs that mediate sweet and amino acid taste [56–58]. Members of the T1R family specifically dimerize to form heteromeric

163

164

7 Transduction Mechanisms in Taste Cells

receptors. T1R1 and T1R3 combine to detect amino acids, while T1R2 and T1R3 dimerize to detect sweet stimuli [59–61]. A truncated form of mGluR4 is also expressed in taste cells and may function in umami detection [62]. While the transduction mechanism for each of these complex taste stimuli is not fully characterized, it has been shown that each of these taste qualities depend to some extent on gustducin, a heterotrimeric G protein found primarily in taste cells [63–65]. Alpha-gustducin activates a phosphodiesterase that reduces cytosolic cAMP levels, while its beta-gamma partners activate the phospholipase C (PLC) signaling pathway, producing inositol trisphosphate (IP3) and diacylglycerol (DAG) [50, 66–68]. Recent evidence suggests that PLC signaling may be the most important leg of the pathway, since knockout of PLCb2 abolishes behavioral and physiological responses to bitter, sweet, and umami stimuli [69]. The ion channels believed to mediate these pathways will be discussed below. 7.3.2

Store-operated Channels and TRPM5

PLC-mediated hydrolysis of PIP2 results in the production of two second messengers, DAG and IP3. While the functional role of DAG in taste cells is unknown, IP3 activates the Type III IP3 receptor on the endoplasmic reticulum (ER) [70], causing release of Ca2+ from intracellular stores and Ca2+ influx via a store-operated channel (SOC) (Fig. 7.7) [71]. The SOC has been hypothesized to be TRPM5, a transient receptor potential channel that is co-expressed with other PLC signaling components in taste cells [72]. However, conflicting data involving heterologous expression of TRPM5 have called this hypothesis into question. Support for the role of TRPM5 as a SOC comes from expression studies in oocytes indicating that TRPM5 is a Ca2+-permeable cationic channel that activates upon depletion of intracellular Ca2+ stores (Fig. 7.8). When TRPM5 is expressed with a dominant negative construct, Ca2+ influx in response to store depletion is abolished [72], suggesting that TRPM5 functions as a store-operated channel. However, studies in modified HEK293 cells show that TRPM5 is activated as a result of receptor activation, but not by Ca2+, IP3, or depletion of internal Ca2+ stores (Fig. 7.9) [69]. Recent studies characterizing the properties of TRPM5 in HEK293 cells [73, 74] do not support the conclusion that TRPM5 is either a store-operated channel or a receptor-operated Ca2+ channel. These expression studies found that TRPM5 is modulated by voltage, is directly activated by Ca2+, and, most surprising, is selective against divalent cations (Figures 7.10 and 7.11). Both studies also demonstrate a strong outward rectification of the current, indicating voltage dependence of the channel [73, 74]. Hofmann et al. found that TRPM5 activation is not significantly reduced in the presence of the calmodulin inhibitor calmidazolium, indicating that Ca2+ may be acting directly on the channel. Liu and Liman demonstrated that TRPM5 is desensitized via a Ca2+-dependent process that can be partially reversed by PIP2 (Fig. 7.12). Monovalent selectivity precludes TRPM5 from acting as a mechanism for Ca2+ influx as either a SOC or a receptoractivated channel. While TRPM5 is critical to the detection of multiple taste qualities [69], its physiological role in this process remains obscure.

7.3 Complex Stimuli Fig. 7.7 Evidence for a store-operated channel (SOC) in mouse taste cells. (A) Dependency of the Ca2+ responses to denatonium on external Ca2+. Stimulation with 1 mM denatonium induced a transient increase in [Ca2+]i followed by a sustained response in gustducin-expressing mouse taste cells. The sustained response disappeared in Ca2+-free saline, suggesting that Ca2+ influx is required. In the presence of denatonium, changing the solution from Ca2+-free to normal saline induced increases in [Ca2+]I. (B) Depletion of Ca2+ stores induces Ca2+ influx. Direct depletion of intracellular Ca2+ stores with thapsigargin also induced increases in [Ca2+]I when external Ca2+ was added to gustducinexpressing mouse taste cells. The same cells showed sustained increases in Ca2+ in response to stimulation with denatonium. Reprinted from [71] with permission

It is interesting to note that TRPM5 has different properties depending on the expression system used. TRPs can combine to form heteromultimers [75] or can be incorporated into macromolecular assembles that may change their functional properties [76]. It is possible that differences in the properties of TRPM5 are due to differences in the expression systems used and in some of the endogenous proteins present in each. Further work is needed to determine whether TRPM5 forms heteromultimers in taste cells and whether these confer Ca2+ permeability upon the channel. The physiological role of TRPM5 in taste cells is unknown, and its characterization may lead to insight about how taste cells communicate with the nervous system. Since TRPM5 appears to be selective for monovalent cations, it seems likely that channel opening results in membrane depolarization, which generally results in opening voltage-gated Ca2+ channels to affect neurotransmitter release. However, we have demonstrated in mice that taste cells using the gustducin/PLC signaling pathway do not have voltage-gated Ca2+ channels [77], and other work in rats has shown that classical chemical synapses are not present in these cells [78]. Thus, how TRPM5 couples to transmitter release will be critical to understanding how bitter, sweet, and umami taste information is communicated to the nervous system (Fig. 7.13). Interestingly, frog taste cells have a conductance with properties similar to a SOC. A nonselective cation channel is activated by both Ca2+ and IP3, suggesting that an IP3dependent increase in intracellular Ca2+ can depolarize taste cells and cause neurotransmitter release. This response is reduced but not abolished in Ca2+-free external solution. However, removal of both Na+ and Ca2+ completely eliminates this current

165

166

7 Transduction Mechanisms in Taste Cells

Fig. 7.8 TRPM5 facilitates Ca2+ influx in X. laevis oocytes. (A) Average traces (n=3, mean  SEM) of oocytes injected with water or TRPM5 cRNA, incubated with thapsigargin (Tg) and then exposed to extracellular Ca2+ (horizontal gray bar). (B) Average fluorescent intensity of the different groups of oocytes analyzed (n=3, mean  SEM). Reprinted from [72] with permission

[79]. It is intriguing to speculate that this cation conductance is an amphibian homologue of TRPM5 and that in taste cells, the TRPM5 conductance is dependent on both Ca2+ and Na+ ions. Further analysis is needed to confirm this hypothesis.

7.3 Complex Stimuli

Fig. 7.9 TRPM5 encodes a functional ion channel activated by GPCR signaling. Whole-cell patchclamp analysis (black traces) was used to measure TRPM5 activity in transfected HEK293 cells. Changes in [Ca2+]I were simultaneously monitored using Fluo-3 (gray traces). (a) Carbachol stimulation of cells expressing Ga15 leads to robust increases in [Ca2+]I but no significant changes in membrane conductance. (b) Cells transfected with TRPM5 alone show neither channel activity nor [Ca2+]I responses to carbachol. (c) However, when Ga15 and TRPM5 are co-expressed, GPCR stimulation elicits both a robust [Ca2+]I response and a large transient increase in membrane conductance. Similar results were obtained in cells co-expressing

Fig. 7.10 TRPM5 is activated by micromolar concentrations of Ca2+. 40 lM Ca2+ activates an inward current in a patch excised from a TRPM5-transfected CHOK1 cell. Neither 10 lM IP3 or 100 lM OAG elicited a current in the same patch. Reprinted from [74] with permission (Copyright 2003, National Academy of Sciences, U.S.A.)

Ga15, TRPM5, and the mouse T2R5 bitter taste receptor when stimulated with cycloheximide (inset). Thus, TRPM5 encodes an ion channel activated by GPCR stimulation. (d) Application of thapsigargin to cells expressing Ga15 and TRPM5 releases [Ca2+]I from internal stores but fails to activate TRPM5. (e) Intracellular application of BAPTA abolishes [Ca2+]I increases but fails to prevent the carbachol-induced activation of TRPM5. (f) Intracellular release of IP3 by UV-mediated uncaging of caged IP3 (arrowheads) results in increases in [Ca2+]I but no activity from TRPM5. Note that subsequent application of carbachol triggers TRPM5 activation. Reprinted from [69] with permission (Copyright 2003, Elsevier)

167

168

7 Transduction Mechanisms in Taste Cells Fig. 7.11 TRPM5 is a voltage-modulated, monovalent-selective cation channel. TRPM5-expressing cells were perfused in the whole-cell mode with activating concentrations of Ca2+. Voltage ramps were applied from 60 mV to –100 mV after exchange of the standard bath solution with solutions containing (1) 140 mM NaCl, (2) 140 mM KCl, (3) 100 mM CaCl2, or (4) 140 mM NMDG-Cl. Replacement of the extracellular cations with CaCl2 or NMDG-Cl resulted in a pronounced inhibition of the inward component and a shift in the reversal potential to a more negative value. Reprinted from [73] with permission

Fig. 7.12 PIP2 partially restores TRPM5 channel activity following desensitization. (A) Enhancement of the current to 40 lM Ca2+ by PIP2 before (n=6) and after desensitization (n=8). Current amplitudes are the averaged peak magnitude recorded within 2 s from the start of Ca2+ exposure. Asterisk indicates a significant difference between enhancement of control and desensitized currents (P<0.05). (B) Dose-response relations before desensitization (open triangles), after desensitization (filled circles), and after desensitization in the presence of 10 lM PIP2 (open circles) (mean  SEM, n=3). In these experiments, PIP2 was present before and during a 6-s application of Ca2+. Currents were normalized to the maximum current obtained in each patch. Reprinted from [74] with permission (Copyright 2003, National Academy of Sciences, U.S.A.)

Fig. 7.13 Proposed model of signaling cascade in response to bitter stimuli. Activation of gustducin (a-gust) results in activation of phosphodiesterase (PDE) and a decrease in cAMP levels. The bc partners activate phospholipase C (PLC b2) to generate DAG and IP3. IP3 activates IP3 R3, causing release of Ca2+ from intracellular stores. Ca2+ can then activate the TRPM5 channel, which is permeable to Na+ and/or Ca2+. Ca2+ may also activate another store-operated channel (SOC) to allow for the influx of Ca2+. It is not yet clear whether TRPM5 and the SOC are the same channel or two different channels

7.3 Complex Stimuli

7.3.3

Cyclic Nucleotide-regulated Channels

In addition to stimulating the PLC signaling pathway, many complex stimuli modulate intracellular cyclic nucleotide monophosphate (cNMP) levels. Both bitter and umami stimuli elicit decreases in cAMP levels via activation of a-gustducin [68, 80, 81], while sugars have been shown to increase both cAMP and cGMP levels in taste cells [82–85]. Cyclic nucleotides have been hypothesized to activate cyclic nucleotide-gated (CNG) channels, analogous to their role in the olfactory system. Support for this hypothesis comes primarily from molecular studies in rats, where a CNG channel was cloned that shares 82 % homology with human cone CNG channels [86, 87]. The heterologously expressed taste channel opens in response to cAMP and cGMP [87], though very dif-

Fig. 7.14 Sweeteners and cyclic nucleotides block a resting K+ current. Holding current was recorded at –80 mV during bath exchanges of sweeteners and cyclic nucleotides. Potassium in the bath was elevated to +20 mM, shifting the K+ equilibrium potential (EK) to approximately –50 mV. Membrane conductance changes were monitored by applying brief hyperpolarizing pulses every 2–5 s. All recordings from (A) and (B) were from the same taste cell. (A) Bath application of 0.1 mM NC01 (artificial sweetener) reversibly reduced the holding current, with a concomitant decrease in membrane conductance. After washing and recovery, this response was mimicked by bath application of 8cpt-cAMP (1 mM) and db-cGMP (1 mM). (B) Simultaneous application of NC01 (0.1 mM) did not enhance the effect of 8cpt-cAMP (1 mM). (C) Simultaneous application of NC01 did not enhance the effect of TEA, a K+ channel blocker. Reprinted from [89] with permission

169

170

7 Transduction Mechanisms in Taste Cells

ferent responses were generated to each nucleotide [88]. To date, however, there is no physiological evidence for the functional expression of CNG channels in mammalian taste cells. In contrast, several physiological studies indicate the presence of cNMP-inhibited channels in taste cells. Whole-cell patch-clamp recordings of sweet-responsive hamster taste cells showed that cAMP and cGMP both inhibit a resting K+ conductance, resulting in taste cell depolarization [89, 90] (Fig. 7.14). The depolarization persists in the presence of protein kinase A inhibitors, suggesting that the nucleotides may be blocking the channels directly rather than via phosphorylation [91]. The channels that mediate these responses have not been identified molecularly, nor have they been studied with single-channel recording methods. Two interesting cNMP-inhibited channels have been described in frog taste cells. A Ca2+-permeable channel is directly blocked by cAMP in inside-out patches, independent of protein phosphorylation (Fig. 7.15) [92]. Since a-gustducin activation results in a decrease of cNMP levels, this channel may be constitutively closed by endogenous cNMP levels but then opened in response to a gustducin-induced decrease in cNMP levels. Channel opening would then cause cell depolarization. However, this channel

Fig. 7.15 Cyclic nucleotides (cNMPs) suppress current in inside-out patches taken from the apical membrane of frog TRCs. (a) Effects of different cNMPs on the integral patch current at –20 mV. (b) Patch-current variation at 0 mV induced by various cGMP concentrations. (c, d) I-V curves of inside-out patches in the absence and presence of cNMP. (e) Dose-response curves. The

continuous and dotted lines correspond to the Hill equation with coefficients 1 and 2, respectively. The data indicate a Hill coefficient close to 1. Filled triangles: 8-BrcAMP; filled squares and filled circles: cAMP; open squares and open triangles: cGMP. Reprinted from [92] with permission (Copyright 1995, Macmillan Magazines Limited)

7.3 Complex Stimuli

Fig. 7.16 Single-channel K+ currents recorded from inside-out excised membrane patches at 0 mV. (a) Superfusion of the cytosolic surface with 5 mM ATP closed the dominant 74-pS K+ channel. A 44-pS K+ channel, also present but seldom open, was not closed. Data for (b),(c), and (d) are from the same patch. (b) Control with 5 mM ATP before adding cAMP kinase. In this patch, at least four K+ channels of 44-pS conductance were active. (c) Addition of 6.6 lM cAMP kinase reduced the number of opening events. (d) Recovery: 7 min after washout of cAMP kinase. (e) Compressed time course of the response to cAMP kinase. Before addition of the kinase, the mean patch current was 1.31 pA. At the asterisk, cAMP kinase was added to a final concentration of 3 lM. The mean patch current decreased to 0.14 pA. During washout, the patch current slowly recovered. By 4 min, the mean patch current was 0.38 pA, and then the patch broke. Reprinted from [93] with permission (Copyright, 1988, Macmillan Magazines Limited)

has not been identified in gustducin-expressing mammalian taste cells, and its functional role in frog taste cells is unclear. Another channel that is modulated by cNMPs in frog taste cells is a 44-pS K+ channel that is blocked by cAMP-dependent phosphorylation (Fig. 7.16) [93]. This channel was reported to be a downstream target of sweet transduction, but frog taste cells show a poor response to sweet compounds.

7.3.4

Ligand-gated Channels

In bullfrogs, quinine directly activates a cation conductance in addition to blocking a K+ conductance. This 10-pS channel is selective for Na+ and the unitary conductance is suppressed by external Ca2+. Interestingly, this conductance is also reduced by cAMP, suggesting that quinine may decrease cAMP in addition to directly activating this

171

172

7 Transduction Mechanisms in Taste Cells

channel [94, 95]. Further work will be required to determine the relationship between cAMP and the quinine-activated channel. There is evidence in several species that sugars can gate channels directly. In canines, sugars activate a Na+-preferring cation conductance that can be partially suppressed by amiloride and ouabain [96]. This conductance does not depend on second messengers but is instead directly activated by sugars. It has been found in dogs, humans, and rabbits, but not in rats or macaques [96]. In the flesh fly, sucrose activates a nonselective cation channel with a 30-pS conductance and a very short mean open time (0.23 ms). This channel is specific for sucrose as compared to fructose, which is likely gated by a different ionic receptor [97]. In addition to activation of GPCRs, amino acids have been found to gate three ligand-gated channels in catfish. L-alanine activates a GPCR pathway, while L-arginine activates two nonselective cation channels. Both channels are high-affinity receptors that are antagonized by D-arginine; one channel has a conductance of 40–60 pS, while the second channel has a conductance of 75–100 pS [52]. Ligand-binding studies have found two L-Arg binding sites; one site has a KD of 20 nM while the other has a K D of 1.5 lM [98, 99]. However, no correlation has been determined between the apparent K D values and the different conductances. Finally, L-proline has been shown to activate a low-affinity, nonselective cation channel with a conductance of 50–65 pS that is blocked by D-proline (Fig. 7.17) [51–53]. To date, these channels have not been identified in other species.

Fig. 7.17 Single-channel currents produced by L-Pro and L-Arg. (A) Single-channel currents produced by 1 mM L-Pro (inset). From the I-V plot, the slope conductance was 48 pS and the current reversed at 6 mV. (B) Single-channel currents caused by 100 lM L-Arg (inset). Slope conductance was 45 pS and reversed at 5 mV. Reprinted from [52] with permission

7.3 Complex Stimuli

7.3.5

Miscellaneous Channels Fat-modulated Channels Recent studies have shown that fatty acids in the oral cavity can directly modulate taste cells. In rat fungiform taste cells, extracellular application of cis-polyunsaturated fatty acids inhibits delayed rectifying K+ channels in some cells while enhancing the current in others. Saturated, monounsaturated, and trans-polyunsaturated fatty acids are ineffective. There is no evidence of G protein involvement; therefore, free fatty acids likely act directly on the K+ channels or closely associated proteins. Overall, these fatty acids prolong stimulus-induced depolarizations in taste receptor cells [100]. Comparisons between fat-preferring and fat-avoiding rat strains found that K+ channels are significantly more sensitive to inhibition by polyunsaturated fatty acids in the rats that avoid fat. These data suggest an inverse correlation between dietary preferences for fat and the peripheral gustatory sensitivity to polyunsaturated fats [101, 102]. 7.3.5.1

7.3.5.2 Water-activated channels

Application of water to the tongue elicits impulses on the glossopharyngeal nerve for mammals, fish, insects, and amphibians. This response, called the water response, mediates a mouth-closing reflex in aquatic species that prevents ingestion of too much water. Mammals also strictly regulate their osmotic balance. However, little is currently known about the mechanisms underlying water detection by the gustatory system. In frogs, two mechanisms have been characterized in water-sensitive cells. In approximately 70 % of water-sensitive taste cells, Cl– secretion induces receptor potentials that are inhibited by anion channel blockers. In the remaining 30 % of watersensitive cells, receptor potentials are generated by blockage of a resting K+ conductance [103]. In mammalian taste cells, hypoosmotic solutions increase cell capacitance and activate a reversible conductance. Ion-substitution experiments indicate that these volume- or stretch-activated channels are selective for Cl–. This conductance is inhibited by some but not all anion channel blockers. This may be the well-characterized, swelling-activated Cl– current found in a variety of cell types [104]. Recently, aquaporin-2, a vasopressin-regulated water channel found in the kidney, has been identified in taste cells. Other channels, including the CIC-2 Cl– channel and the renal outer medullary K+ channel, have been found in taste buds. These channels respond to the natriferic hormones aldosterone and vasopressin and are involved in regulating Cl– and K+ homeostasis, respectively [105]. It seems likely that the gustatory system acts as the initial site of salt and water regulation in both aquatic and land animals.

173

174

7 Transduction Mechanisms in Taste Cells

7.4

Conclusions

The gustatory system has evolved multiple mechanisms to detect and respond to the surrounding environment. The chemical complexity of taste stimuli results in a heterogeneous array of cellular responses that culminates in signals that can be transmitted to gustatory afferents. These mechanisms include direct stimulus-gated and -blocked channels, stimulus-permeable channels, and channels activated and blocked as a result of GPCR signaling mechanisms. While many of these channels have been identified, their precise role in transduction remains obscure. Determining how these channels contribute to the coding of specific taste qualities is an important goal of future studies. References 1

2 3 4

5

6

7

8

9

10

Finger, T.E., Simon S.A. Cell Biology of taste epithelium, in The Neurobiology of Taste and Smell, T.E. Finger, Silver W.L., Restrepo D., Editor. 2000, Wiley-Liss: New York. 287–314. Lindemann, B. Taste reception. Physiol Rev, 1996. 76(3): 718–66. Lindemann, B. Receptors and transduction in taste. Nature, 2001. 413(6852): 219–25. Farbman, A.I. Renewal of taste bud cells in rat circumvallate papillae. Cell Tissue Kinet, 1980. 13(4): 349–57. Chen, Y., X.D. Sun, and S. Herness. Characteristics of action potentials and their underlying outward currents in rat taste receptor cells. J Neurophysiol, 1996. 75(2): 820–31. Herness, M.S. and X.D. Sun. Voltage-dependent sodium currents recorded from dissociated rat taste cells. J Membr Biol, 1995. 146(1): 73–84. Behe, P., et al. Membrane currents in taste cells of the rat fungiform papilla. Evidence for two types of Ca currents and inhibition of K currents by saccharin. J Gen Physiol, 1990. 96(5): 1061–84. Smith, D.V., Davis B.J. Neural representation of taste, in The Neurobiology of Taste and Smell, T.E. Finger, Silver W.L., Restrepo D., Editor. 2000, Wiley-Liss: New York. 353–394. Bryant, B.P., Silver, W.L. Chemesthesis: The common chemical sense, in The Neurobiology of Taste and Smell, T.E. Finger, Silver W.L., Restrepo D., Editor. 2000, Wiley-Liss: New York. 73–100. Ye, Q., G.L. Heck, and J.A. DeSimone. The anion paradox in sodium taste reception: resolution by voltage-clamp studies. Science, 1991. 254(5032): 724–6.

11

12

13 14

15

16

17

18

19

Simon, S.A., et al. Transcellular and paracellular pathways in lingual epithelia and their influence in taste transduction. Microsc Res Tech, 1993. 26(3): 196–208. Elliott, E.J. and S.A. Simon. The anion in salt taste: a possible role for paracellular pathways. Brain Res, 1990. 535(1): 9–17. Lindemann, B. Sodium taste. Curr Opin Nephrol Hypertens, 1997. 6(5): 425–9. Benos, D.J. and B.A. Stanton. Functional domains within the degenerin/epithelial sodium channel (Deg/ENaC) superfamily of ion channels. J Physiol, 1999. 520 Pt 3: 631–44. Canessa, C.M., et al. Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature, 1994. 367(6462): 463–7. Lin, W., et al. Epithelial Na+ channel subunits in rat taste cells: localization and regulation by aldosterone. J Comp Neurol, 1999. 405(3): 406–20. Lindemann, B., et al. Occurrence of ENaC subunit mRNA and immunocytochemistry of the channel subunits in taste buds of the rat vallate papilla. Ann N Y Acad Sci, 1998. 855: 116–27. Kretz, O., et al. Differential expression of RNA and protein of the three pore-forming subunits of the amiloride-sensitive epithelial sodium channel in taste buds of the rat. J Histochem Cytochem, 1999. 47(1): 51–64. Gilbertson, T.A., S.D. Roper, and S.C. Kinnamon. Proton currents through amiloridesensitive Na+ channels in isolated hamster taste cells: enhancement by vasopressin and cAMP. Neuron, 1993. 10(5): 931–42.

7.4 Conclusions 20

21

22

23

24

25

26

27

28

29

30

31

32

33

Gilbertson, T.A. Peripheral mechanisms of taste, in The Scientific Basis of Eating, R.W.A. Linden, Editor. 1998. 1–28. Gilbertson, T.A. and H. Zhang. Characterization of sodium transport in gustatory epithelia from the hamster and rat. Chem Senses, 1998. 23(3): 283–93. Gilbertson, T.A. and H. Zhang. Self-inhibition in amiloride-sensitive sodium channels in taste receptor cells. J Gen Physiol, 1998. 111(5): 667–77. Doolin, R.E. and T.A. Gilbertson. Distribution and characterization of functional amiloridesensitive sodium channels in rat tongue. J Gen Physiol, 1996. 107(4): 545–54. Bigiani, A. Amiloride-sensitive sodium currents in identified taste cells of the frog. Neuroreport, 2001. 12(7): 1315–21. Avenet, P. and B. Lindemann. Fluctuation analysis of amiloride-blockable currents in membrane patches excised from salt-taste receptor cells. J Basic Clin Physiol Pharmacol, 1990. 1(1–4): 383–91. Mierson, S., M.M. Olson, and A.E. Tietz. Basolateral amiloride-sensitive Na+ transport pathway in rat tongue epithelium. J Neurophysiol, 1996. 76(2): 1297–309. Miyamoto, T., et al. Acid and salt responses in mouse taste cells. Prog Neurobiol, 2000. 62(2): 135–57. DeSimone, J.A., et al. A novel pharmacological probe links the amiloride-insensitive NaCl, KCl, and NH(4)Cl chorda tympani taste responses. J Neurophysiol, 2001. 86(5): 2638–41. Miyamoto, T., Okada, Y., Sato T. Cationic and anionic channels of apical receptive membrane in a taste cell contribute to generation of salt-induced receptor potential. Comp Biochem Physiol, 1993. 106A(3): 489–493. Miyamoto, T., et al. Properties of Na+-dependent K+ conductance in the apical membrane of frog taste cells. Brain Res, 1996. 715(1–2): 79–85. Lin, W., T. Ogura, and S.C. Kinnamon. Acidactivated cation currents in rat vallate taste receptor cells. J Neurophysiol, 2002. 88(1): 133–41. Lyall, V., et al. Decrease in rat taste receptor cell intracellular pH is the proximate stimulus in sour taste transduction. Am J Physiol Cell Physiol, 2001. 281(3): C1005–13. Richter, T.A., A. Caicedo, and S.D. Roper. Sour taste stimuli evoke Ca2+ and pH responses in mouse taste cells. J Physiol, 2003. 547(Pt 2): 475–83.

34

35

36

37

38

39

40

41

42 43

44

45

46

47

Sowalsky, R.A. and A.C. Noble. Comparison of the effects of concentration, pH and anion species on astringency and sourness of organic acids. Chem Senses, 1998. 23(3): 343–9. Gilbertson, D.M. and T.A. Gilbertson. Amiloride reduces the aversiveness of acids in preference tests. Physiol Behav, 1994. 56(4): 649–54. Miyamoto, T., et al. Salty and sour transduction. Multiple mechanisms and strain differences. Ann N Y Acad Sci, 1998. 855: 128–33. Miyamoto, T., et al. Sour transduction involves activation of NPPB-sensitive conductance in mouse taste cells. J Neurophysiol, 1998. 80(4): 1852–9. Ohtubo, Y., et al. Optical recordings of taste responses from fungiform papillae of mouse in situ. J Physiol, 2001. 530(Pt 2): 287–93. Stevens, D.R., et al. Hyperpolarization-activated channels HCN1 and HCN4 mediate responses to sour stimuli. Nature, 2001. 413(6856): 631–5. Ugawa, S., et al. Cloning and functional expression of ASIC-beta2, a splice variant of ASIC-beta. Neuroreport, 2001. 12(13): 2865–9. Ogura, T., A. Mackay-Sim, and S.C. Kinnamon. Bitter taste transduction of denatonium in the mudpuppy Necturus maculosus. J Neurosci, 1997. 17(10): 3580–7. Waldmann, R., et al. H+-gated cation channels. Ann N Y Acad Sci, 1999. 868: 67–76. Waldmann, R. and M. Lazdunski. H+-gated cation channels: neuronal acid sensors in the NaC/DEG family of ion channels. Curr Opin Neurobiol, 1998. 8(3): 418–24. Bobkov, Y.V. and S.S. Kolesnikov. Extracellular protons activate K+ current in a subpopulation of frog taste receptor cells. Neurosci Lett, 1999. 264(1–3): 25–8. Cummings, T.A. and S.C. Kinnamon. Apical K+ channels in Necturus taste cells. Modulation by intracellular factors and taste stimuli. J Gen Physiol, 1992. 99(4): 591–613. Kinnamon, S.C., V.E. Dionne, and K.G. Beam. Apical localization of K+ channels in taste cells provides the basis for sour taste transduction. Proc Natl Acad Sci U S A, 1988. 85(18): 7023–7. Furue, H. and K. Yoshii. In situ tight-seal recordings of taste substance-elicited action currents and voltage-gated Ba2+ currents from single taste bud cells in the peeled epithelium of mouse tongue. Brain Res, 1997. 776(1–2): 133–9.

175

176

7 Transduction Mechanisms in Taste Cells 48

49

50

51

52

53

54

55 56

57

58

59

60 61 62

63

64

Burks CA, H.D., Rao S, Lin W, Kinnamon SC, Gilbertson TA. Rat taste buds express multiple members of the KCNK family of two-pore domain potassium channels. AChemS 25th Annual Meeting, Sarasota, FL, 2003. Lin, W., Rao, S., Kinnamon, S.C., Gilbertson, T.A. Evidence for expression of TASK-like K+ channels in rat taste cells. AChemS 24th Annual Meeting, Sarasota, FL, 2002. Margolskee, R.F. The molecular biology of taste transduction. Bioessays, 1993. 15(10): 645–50. Brand, J.G., et al. Transduction mechanisms for the taste of amino acids. Physiol Behav, 1991. 49(5): 899–904. Kumazawa, T., J.G. Brand, and J.H. Teeter. Amino acid-activated channels in the catfish taste system. Biophys J, 1998. 75(6): 2757–66. Teeter, J.H., J.G. Brand, and T. Kumazawa. A stimulus-activated conductance in isolated taste epithelial membranes. Biophys J, 1990. 58(1): 253–9. Chandrashekar, J., et al. T2Rs function as bitter taste receptors. Cell, 2000. 100(6): 703–11. Adler, E., et al. A novel family of mammalian taste receptors. Cell, 2000. 100(6): 693–702. Montmayeur, J.P. and H. Matsunami. Receptors for bitter and sweet taste. Curr Opin Neurobiol, 2002. 12(4): 366–71. Li, X., et al. Human receptors for sweet and umami taste. Proc Natl Acad Sci U S A, 2002. 99(7): 4692–6. Damak, S., et al. Detection of sweet and umami taste in the absence of taste receptor T1r3. Science, 2003. 301(5634): 850–3. Max, M., et al. Tas1r3, encoding a new candidate taste receptor, is allelic to the sweet responsiveness locus Sac. Nat Genet, 2001. 28(1): 58–63. Nelson, G., et al. Mammalian sweet taste receptors. Cell, 2001. 106(3): 381–90. Nelson, G., et al. An amino-acid taste receptor. Nature, 2002. 416(6877): 199–202. Chaudhari, N. and S.D. Roper. Molecular and physiological evidence for glutamate (umami) taste transduction via a G protein-coupled receptor. Ann N Y Acad Sci, 1998. 855: 398–406. Hoon, M.A., et al. Functional expression of the taste specific G-protein, alpha-gustducin. Biochem J, 1995. 309 ( Pt 2): 629–36. McLaughlin, S.K., P.J. McKinnon, and R.F. Margolskee. Gustducin is a taste-cell-specific G protein closely related to the transducins. Nature, 1992. 357(6379): 563–9.

65

66

67

68

69

70

71

72

73

74

75

76

77

78

Wong, G.T., K.S. Gannon, and R.F. Margolskee. Transduction of bitter and sweet taste by gustducin. Nature, 1996. 381(6585): 796–800. Huang, L., et al. Ggamma13 colocalizes with gustducin in taste receptor cells and mediates IP3 responses to bitter denatonium. Nat Neurosci, 1999. 2(12): 1055–62. Ming, D., Y. Ninomiya, and R.F. Margolskee. Blocking taste receptor activation of gustducin inhibits gustatory responses to bitter compounds. Proc Natl Acad Sci U S A, 1999. 96(17): 9903–8. Yan, W., et al. Bitter taste transduced by PLCb2-dependent rise in IP3 and a-gustducindependent fall in cyclic nucleotides. Am J Physiol Cell Physiol, 2001. 280(4): C742–51. Zhang, Y., et al. Coding of sweet, bitter, and umami tastes: different receptor cells sharing similar signaling pathways. Cell, 2003. 112(3): 293–301. Clapp, T.R., et al. Immunocytochemical evidence for co-expression of Type III IP3 receptor with signaling components of bitter taste transduction. BMC Neurosci, 2001. 2(1): 6. Ogura, T., R.F. Margolskee, and S.C. Kinnamon. Taste receptor cell responses to the bitter stimulus denatonium involve Ca2+ influx via store-operated channels. J Neurophysiol, 2002. 87(6): 3152–5. Perez, C.A., et al. A transient receptor potential channel expressed in taste receptor cells. Nat Neurosci, 2002. 5(11): 1169–76. Hofmann, T., et al. TRPM5 is a voltage-modulated and Ca2+-activated monovalent selective cation channel. Curr Biol, 2003. 13(13): 1153–8. Liu, D. and E.R. Liman. Intracellular Ca2+ and the phospholipid PIP2 regulate the taste transduction ion channel TRPM5. PNAS, 2003. 100(25): 15160–15165. Minke, B. and B. Cook. TRP channel proteins and signal transduction. Physiol Rev, 2002. 82(2): 429–72. Montell, C. Physiology, phylogeny, and functions of the TRP superfamily of cation channels. Sci STKE, 2001. 2001(90): RE1. Medler, K.F., R.F. Margolskee, and S.C. Kinnamon. Electrophysiological characterization of voltage-gated currents in defined taste cell types of mice. J Neurosci, 2003. 23(7): 2608–17. Clapp, T.R., Yang, R., Stoick, C.L., Kinnamon, S.C., Kinnamon, J.C. Morphologic characterization of rat taste receptor cells that express components of the phospholipase C signaling pathway. J Comp Neurol, 2004. 468: 311–321.

7.4 Conclusions 79

80

81

82

83

84

85

86

87

88

89

90

91

92

Okada, Y., et al. Inositol 1,4,5-trisphosphate activates non-selective cation conductance via intracellular Ca2+ increase in isolated frog taste cells. Eur J Neurosci, 1998. 10(4): 1376–82. Spielman, A.I., et al. Generation of inositol phosphates in bitter taste transduction. Physiol Behav, 1994. 56(6): 1149–55. Spielman, A.I., et al. Rapid kinetics of second messenger production in bitter taste. Am J Physiol, 1996. 270(3 Pt 1): C926–31. Naim, M., et al. Adenylate cyclase responses to sucrose stimulation in membranes of pig circumvallate taste papillae. Comp Biochem Physiol B, 1991. 100(3): 455–8. Striem, B.J., et al. Saccharin induces changes in adenylate cyclase activity in liver and muscle membranes in rats. Life Sci, 1990. 46(11): 803–10. Striem, B.J., et al. Sweet tastants stimulate adenylate cyclase coupled to GTP-binding protein in rat tongue membranes. Biochem J, 1989. 260(1): 121–6. Bernhardt, S.J., et al. Changes in IP3 and cytosolic Ca2+ in response to sugars and nonsugar sweeteners in transduction of sweet taste in the rat. J Physiol, 1996. 490 ( Pt 2): 325–36. Misaka, T., et al. Molecular cloning and taste bud-specific expression of a novel cyclic nucleotide-gated channel. Ann N Y Acad Sci, 1998. 855: 150–9. Misaka, T., et al. Taste buds have a cyclic nucleotide-activated channel, CNGgust. J Biol Chem, 1997. 272(36): 22623–9. Lee, H.M., et al. Electrophysiological characteristics of rat gustatory cyclic nucleotide– gated channel expressed in Xenopus oocytes. J Neurophysiol, 2001. 85(6): 2335–49. Cummings, T.A., C. Daniels, and S.C. Kinnamon. Sweet taste transduction in hamster: sweeteners and cyclic nucleotides depolarize taste cells by reducing a K+ current. J Neurophysiol, 1996. 75(3): 1256–63. Cummings, T.A., J. Powell, and S.C. Kinnamon. Sweet taste transduction in hamster taste cells: evidence for the role of cyclic nucleotides. J Neurophysiol, 1993. 70(6): 2326–36. Varkevisser, B. and S.C. Kinnamon. Sweet taste transduction in hamster: role of protein kinases. J Neurophysiol, 2000. 83(5): 2526–32. Kolesnikov, S.S. and R.F. Margolskee. A cyclic-nucleotide-suppressible conductance activated by transducin in taste cells. Nature, 1995. 376(6535): 85–8.

93

94

95

96

97

98

99

100

101

102

103

104

105

Avenet, P., F. Hofmann, and B. Lindemann. Transduction in taste receptor cells requires cAMP-dependent protein kinase. Nature, 1988. 331(6154): 351–4. Tsunenari, T., et al. A quinine-activated cationic conductance in vertebrate taste receptor cells. J Gen Physiol, 1996. 108(6): 515–23. Tsunenari, T., T. Kurahashi, and A. Kaneko. Activation by bitter substances of a cationic channel in membrane patches excised from the bullfrog taste receptor cell. J Physiol, 1999. 519 Pt 2: 397–404. Mierson, S., et al. Sugar-activated ion transport in canine lingual epithelium. Implications for sugar taste transduction. J Gen Physiol, 1988. 92(1): 87–111. Murakami, M. and H. Kijima. Transduction ion channels directly gated by sugars on the insect taste cell. J Gen Physiol, 2000. 115(4): 455–66. Bryant, B.P., S. Harpaz, and J.G. Brand. Structure/activity relationships in the arginine chemoreceptive taste response of the channel catfish. Chem Senses, 1989. 14: 805–815. Kalinoski, D.L., et al. Specific L-arginine taste receptor sites in the catfish, Ictalurus punctatus: biochemical and neurophysiological characterization. Brain Res, 1989. 488(1–2): 163–73. Gilbertson, T.A., et al. Fatty acid modulation of K+ channels in taste receptor cells: gustatory cues for dietary fat. Am J Physiol, 1997. 272(4 Pt 1): C1203–10. Gilbertson, T.A. Role of the taste system in ingestive behavior. Studies in NaCl and fatty acid transduction. Ann N Y Acad Sci, 1998. 855: 860–7. Gilbertson, T.A., et al. Dietary fat preferences are inversely correlated with peripheral gustatory fatty acid sensitivity. Ann N Y Acad Sci, 1998. 855: 165–8. Okada, Y., T. Miyamoto, and T. Sato. The ionic basis of the receptor potential of frog taste cells induced by water stimuli. J Exp Biol, 1993. 174: 1–17. Gilbertson, T.A. Hypoosmotic stimuli activate a chloride conductance in rat taste cells. Chem Senses, 2002. 27(4): 383–94. Gilbertson, T.A., Chen, W., Rao, S. Hansen D.R. Identification of natriferic hormone responsive elements in taste cells: Implications for the regulation of salt and water taste. AChemS 25th Annual Meeting, Sarasota, FL, 2003.

177

179

8

Invertebrate Phototransduction: Multimolecular Signaling Complexes and the Role of TRP and TRPL Channels Armin Huber

Abstract

The cation channels TRP and TRPL operate in Drosophila photoreceptor cells. They are activated by a G-protein-mediated signaling pathway in response to light-activation of the visual pigment rhodopsin. Drosophila TRP is the founding member of a large and diverse family of ion channels. In addition to photoreception, TRP channels mediate other sensory modalities such as smell, taste, thermoreception, mechanoreception, and nociception, as well as physiological responses of non-neuronal cells. TRP is a core component of a multimolecular signaling complex that is assembled by the PDZ domain protein INAD, which also binds other key members of the signal transduction pathway. The assembly of multimolecular signaling complexes may be regarded as a means to compartmentalize signaling molecules into a specific signaling pathway, the phototransduction cascade. This may enhance signaling specificity and signaling speed. Although most major components of the phototransduction cascade operating in photoreceptor cells of the Drosophila compound eye have been identified in the last 30 years, the exact gating mechanism of TRP and TRPL is still enigmatic. In response to light, diacylglycerol as well as inositol 1,4,5-trisphosphate are generated from the membrane phospholipid phosphatidylinositol 4,5-bisphosphate by a Gq protein-activated phospholipase Cb that is mandatory for visual excitation. Several pieces of evidence suggest that diacylglycerol or its derivates, rather than the inositol 1,4,5trisphosphate branch of the pathway, are crucial for TRP and TRPL activation. While the number of TRP molecules in the photoreceptive membrane is not affected by light, the concentration of TRPL in this membrane changes dramatically, at a timescale of hours, when the flies are transferred from darkness to light or vice versa. In the light, TRPL translocates from the rhabdomeral photoreceptor membrane to an intracellular storage compartment, thus changing the ion channel composition of the photoreceptive membrane. In addition to TRPL, the G-protein alpha subunit and the regulatory protein arrestin 2 undergo light-dependent relocation in the photoreceptor cell. Together, the relocation of signaling proteins seems to constitute a mechanism for long-term adaptation.

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

180

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

8.1

Introduction

Visual systems that detect light have been developed by the vast majority of species in the animal kingdom and also by some algae such as Chlamydomonas, as well as in Halobacteria. Evolution has generated two major types of photoreceptors. In ciliary photoreceptors, represented by the rods and cones of the vertebrate eye, an outer segment, which contains the machinery for converting the absorption of a photon into an electrical response, can be distinguished from the rest of the cell. Microvillar photoreceptors, on the other hand, are typical for most, but not all, invertebrate eyes. These receptors form a stack of microvilli along one side of the cell, and the proteins required for phototransduction are compartmentalized into this region. The best-studied invertebrate photoreceptors are those of the compound eye of the fruit fly Drosophila melanogaster, and this chapter will focus mainly on this model system. In both vertebrate and invertebrate visual systems, light is absorbed by rhodopsins, prototypical G-protein-coupled receptors, which are by far the most abundant proteins incorporated into either the disk membranes of outer segments or the microvillar membranes of invertebrate photoreceptor cells. The high density of rhodopsins in the photoreceptive cell compartment assures a high probability of catching a photon and is the molecular basis for the high sensitivity of photoreceptors. An important requirement for the signal transduction cascades downstream of rhodopsin is to distinguish the effective absorption of photons from background noise. Although vertebrate and invertebrate photoreceptor cells use different types of phototransduction cascades, which terminate in either the closure of cyclic nucleotide-gated transduction channels (see Chapter 9) or the opening of TRP channels (this chapter), this requirement is fulfilled perfectly by both systems, as each allows the detection of a single photon. A second requirement more important in vision than, for example, in the chemical senses is signaling speed. Fast activation and inactivation kinetics of the visual responses are essential for the detection of movement. In terms of signaling speed, Drosophila photoreceptors surpass vertebrate photoreceptors by a factor of 10 to 100 [1]. Indeed, with an activation time of about 20 ms and an inactivation time of about 100 ms [2], the signaling cascade underlying fly vision is among the fastest G-protein-coupled signaling pathways described to date. It will be discussed below that the basis for such extraordinarily fast signal transduction may lie in the architecture and molecular organization of the signaling elements within the photoreceptor microvilli. Thirdly, since photoreceptors must be able to operate in light intensities varying over several orders of magnitude, light adaptation, that is, adjustment of the visual responses to the background light, is a fundamental principle of this sensory modality. Towards this end, recent findings shall be presented showing that the ion channel equipment of the Drosophila photoreceptor membrane is subject to light-dependent regulation.

8.2 Structure of the Drosophila Compound Eye and Its Visual Pigments

8.2

Structure of the Drosophila Compound Eye and Its Visual Pigments

The Drosophila compound eye is composed of 750–800 single eyes, called ommatidia, which correspond to the facets observed on the eye surface (Fig. 8.1; for reviews, see [3–5]). Each ommatidium contains 8 microvillar photoreceptor cells and 12 accessory cells, including pigment cells, which optically shield one ommatidium from another. The photoreceptor cells form a densely packed stack of microvilli, the rhabdomere, a light-guiding structure that extends through the full-length of the ommatidium. The rhabdomeres are separated from the rest of the cell by a stalk, and they project into an extracellular cavity filled with extracellular matrix. The microvillar membranes of the rhabdomere and the stalk membrane correspond to the apical region of polarized epithelial cells, and they are structurally and functionally separated from the basolateral region by adherens junctions. Mutations in proteins required to organize a continuous adherens junction, for example, in the transmembrane protein Crumbs, result in marked morphological defects of the rhabdomeres [6, 7]. Besides its role in photoreceptor cells, Crumbs is also required for the establishment of cell polarity in ectodermal epithelial cells where its function was discovered first [8]. Intriguingly, mutations in the human Crumbs-homologue CRB1, which localizes to the inner segment of mammalian photoreceptors, cause the hereditary eye diseases retinitis pigmentosa (RP12) and Leber’s congenital amaurosis [9, 10]. By analogy to polarized epithelial cells, one would assume that the apical membrane compartment of Drosophila photoreceptors, that is, the rhabdomere and the rhabdomeral stalk, contains the same protein composition. However, the proteins making up the phototransduction machinery seem to be restricted to the rhabdomere alone. Formation of multimolecular complexes (see below) that can interact with the actin cytoskeleton of the rhabdomere may be involved in generating the specific molecular composition of this cellular compartment. The eight Drosophila photoreceptor cells can be divided into three different classes: the six outer photoreceptors R1–6 and two inner cells, R7 and R8. R7 and R8 form a fused rhabdomere that is located in the center of the ommatidium. The R7 cell contributes the apical part, and the R8 cell forms the basal part of this central rhabdomere. The different classes of photoreceptor cells are distinguished not only by their position within the ommatidium but also by the visual pigments they express. R1–6 cells contain the same rhodopsin, Rh1 [11, 12]. In non-overlapping subsets, R7 and R8 cells express four different rhodopsins: Rh3, Rh4, Rh5, and Rh6. The expression of rhodopsin Rh3 or rhodopsin Rh4 in the R7 cell is paired with Rh5 or Rh6 expression in the adjacent R8 cell, respectively [13]. Thus, two types of ommatidia exist, containing the rhodopsin combination of either Rh3/Rh5 or Rh4/Rh6 in the inner photoreceptor cells. Besides in the compound eye, Rh1, Rh3, Rh4, and Rh6 are expressed in a larval photoreceptor, the Bolwig’s organ [14, 15]. Rh6 has also been detected in the so-called Hofbauer-Buchner eyelet, a small photosensitive structure located in the adult head, adjacent to the optic ganglia where it is formed from the Bolwig’s nerve during metamorphosis [15, 16]. A possible function of this eyelet is the entrainment of circadian rhythms. Another Drosophila rhodopsin, Rh2, is not ex-

181

182

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

Fig. 8.1 Structure of the Drosophila compound eye. (A) Scanning electron micrograph of the compound eye. (B) Schematic drawing of the eye. (C) Scheme of a longitudinal section through one

single eye or ommatidium. (D, E) Schemes of cross-sections taken at the levels indicated by arrows (modified after [27])

pressed in the compound eye at all, but in three ocelli, small eyespots on the vertex of the fly head [14]. Finally, the gene encoding a seventh Drosophila rhodopsin, Rh7, has been discovered by the Drosophila genome project [17]. The expression pattern of Rh7 has not yet been determined. Sequence analysis reveals that Rh7 lacks a His-Glu-Lys motif in an intracellular domain that is highly conserved between invertebrate rhodopsins [18], suggesting that Rh7 may not be a bona fide rhodopsin. The Drosophila rhodopsins Rh3–Rh6 expressed in the inner photoreceptor cells have distinct absorption properties with absorption maxima ranging from UV light (331 nm) to green light (515 nm; see Tab. 8.1), thus allowing this visual system to discriminate colors [19]. From the diverse rhodopsins expressed in the photoreceptor cells of the compound eye, signal transduction converges onto the same downstream components. That is, except for rhodopsins, the phototransduction proteins, including the ion cannels TRP and TRPL, are expressed from the same genes in all eight photoreceptor cells. It is therefore possible to restore a light response of the outer photoreceptor cells R1–6 in a mutant lacking Rh1 rhodopsin by ectopic expression of any of the other rhodopsins (Rh2–Rh6) in these cells (e.g., [19]).

8.2 Structure of the Drosophila Compound Eye and Its Visual Pigments Tab. 8.1

Proteins mediating Drosophila vision

Protein

Mutant

Function

Refs.

Rhodopsins Rh1 (486 nm)*

ninaE

11, 12

Rh2 (418 nm)* Rh3 (331 nm)*

– –

Rh4 (355 nm)*

–

Rh5 (442 nm)* Rh6 (515 nm)*

– rh61

Visual pigment of R1–6 cells, also expressed in Bolwig’s organ Expressed in ocelli Expressed in a subset of R7 cells, in marginal R8 cells, and in Bolwig’s organ Expressed in a subset of R7 cells and in Bolwig’s organ Expressed in a subset of R8 cells Expressed in a subset of R8 cells, Bolwig’s organ, and Hofbauer-Buchner eyelet Not determined

Chaperone for rhodopsin Vitamin A formation Uptake of carotenoids

21, 22 23 24

a-, b-, and c-subunit of the rhodopsinactivated G-protein

29, 30 31, 32 33

PIP2 hydrolysis

34

Rh7 – Proteins required for rhodopsin maturation Cyclophilin ninaA Carotene-dioxygenase ninaB ninaD Class B scavenger G-Protein Gaq dgq Gbe gbe Gce – Effector enzyme Phospholipase Cb norpA Phototransduction channels TRP trp TRPL trpl TRPc ? – Regulatory proteins arr1 Arrestin 1 arr2 Arrestin 2 Rhodopsin phosphatase Rhodopsin kinase

rdgC –

Unconventional myosin Protein kinase C Scaffolding protein Unknown identity PIP2 regeneration DAG kinase CDP-DAG-synthase PI transfer protein

ninaC inaC inaD inaF rdgA cds rdgB

Mediate Ca2+ and Na+ influx

Rhodopsin inactivation, rhodopsin internalization

14, 55 14, 56, 57 14, 58 59, 60 15, 61 17

35, 36 37, 38 39

40–43

Dephosphorylation of rhodopsin Phosphorylation of activated rhodopsin (gene not yet identified) Links the INAD signaling complex to actin? Phosphorylation of TRP and INAD Tethering TRP, PLCb, and ePKC Required for TRP function

44, 45

Conversion of DAG to phosphatidic acid Synthesis of CDP-DAG Transfer of PI from inner membranes to microvillar membrane

63 64 65

*Absorption maxima as determined by Salcedo et al. [19].

54 51, 52 53 62

183

184

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

8.3

The Drosophila Phototransduction Cascade Is a Prototypical G-protein-coupled Signaling Pathway

Eukaryotic cells have a limited repertoire of signaling pathways. During evolution sensory cells recruited different types of these pathways for sensory signal transduction and adapted them to meet specific requirements. This adaptation included coupling of the pathway to a receptor molecule which is able to detect the adequate stimulus and, at its other end, coupling to an ion channel capable of inducing a change in the membrane potential which allows spatial propagation of the signal to the synapse of the sensory cell. For Drosophila vision, the molecular identification of the components of this signaling pathway has taken about 30 years and is not yet entirely completed. Understanding of Drosophila vision at the molecular level has been greatly aided by the possibility to generate mutants showing altered visual responses or no responses at all [20]. To date, Drosophila mutants are available for almost every component of the phototransduction cascade. Many of the visual mutants have been named according to their electrophysiological phenotype, as determined by electroretinogram (ERG) recordings. For example, a norp mutant shows “no receptor potential,” a trp mutant displays a “transient receptor potential,” and nina (neither inactivation nor afterpotential) mutants lack two specific features of the ERG: inactivation by a strong light stimulus and an afterpotential, which is usually observed after activation of a large number of rhodopsin molecules that cannot readily be inactivated after termination of the light stimulus. The nina mutants typically contain a drastically reduced amount of rhodopsin in the photoreceptive membrane. The ninaE gene encodes the major Drosophila rhodopsin Rh1, and ninaA encodes a chaperone required for rhodopsin maturation [11, 12, 21, 22]. ninaD and ninaB mediate the uptake and formation of the retinal chromophore, respectively [23, 24]. Finally, ina (inactivation no afterpotential) mutants, which also lack the afterpotential, typically show defects in regulatory phototransduction proteins. Another class of mutants (rdg, retinal degeneration) was isolated on the basis of massive degeneration of photoreceptor cells. In many cases mutants showing the same physiological or morphological phenotype have to be assigned to different complementation groups, such as ninaA, ninaB, ninaC, etc. (see Tab. 8.1). The genetic analysis, combined with biochemical and physiological studies, revealed that the phototransduction cascade in Drosophila is a prototypical G-protein-coupled signaling pathway [1, 25–28] (Fig. 8.2). When a rhodopsin molecule is activated by light-absorption, it transmits the signal to a heterotrimeric Gq protein consisting of Gaq, Gbe, and Gce [29–33] (Tab. 8.1). As a result of G-protein-activation, Gaq couples to the norpA-encoded effector enzyme phospholipase Cb (PLCb) [34]. Activated PLCb hydrolyzes the membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) to form the second messengers diacylglycerol and inositol 1,4,5-trisphosphate. The phototransduction cascade terminates in the opening of cation channels composed of ion channel subunits of the TRP protein family, which results in the depolarization of the photoreceptor cell membrane [35–39].

8.3 The Drosophila Phototransduction Cascade Is a Prototypical G-protein-coupled Signaling Pathway

Fig. 8.2 Hypothetical scheme of the phototransduction cascade in Drosophila photoreceptor cells. Rhodopsin activates a heterotrimeric Gq protein, which in turn activates a phospholipase Cb (PLCb). PLCb generates the second messengers diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3) from phosphatidylinositol 4,5-bisphosphate (PIP2). DAG or polyunsaturated fatty acids (PUFA) derived from DAG may open the cation channels TRP, TRPL, and TRPc. IP3 may initiate Ca2+ release from submicrovillar cisternae (SMC). Phosphorylation of TRP by a protein kinase C (ePKC) may participate in channel regulation. PLCb, ePKC, and TRP are bound by the scaffold protein INAD and form a multimolecular protein complex (modified after [27])

Besides these major players in phototransduction, a number of regulatory proteins are involved in mediating response termination, supply of PIP2, and Ca2+-regulation of the cascade. A list of the proteins mediating Drosophila vision, which are described here, is contained in Tab. 8.1. At least four regulatory proteins interact with rhodopsin: two arrestins (Arr1 and Arr2), which bind to the activated state of rhodopsin and thereby quench prolonged G-protein activation [40–43]; a rhodopsin kinase, which has so far escaped identification at the molecular level; and a rhodopsin phosphatase, which is encoded by the rdgC locus [44, 45]. In contrast to the inactivation mechanism of vertebrate rhodopsin, phosphorylation of fly rhodopsin is not required for arrestin binding; rather, arrestin binding precedes rhodopsin phosphorylation and seems to regulate the phosphorylation and dephosphorylation of the receptor [44, 46, 47]. The function of arrestin is not limited to uncoupling activated rhodopsin from the G-protein; it is also essential for the internalization of rhodopsin molecules. Activated rhodopsin-arrestin complexes are specifically removed from the microvillar membrane. Intriguingly, the internalized rhodopsin-arrestin complexes may induce cell death (apoptosis) if they persist for too long [48, 49]. The enzymes required to regenerate PIP2 from diacylglycerol are a diacylglycerol kinase encoded by rdgA, a CDP-DAG synthase, the phosphatidylinositol transfer protein RDGB, and phosphatidyl-inositol kinases (see [50]). RDGA converts diacylglycerol to phosphatidic acid, which then is converted to phosphatidylinositol via the CDP-activated intermediate CDP-DAG. Phosphatidylinositol finally becomes phosphorylated to regenerate phosphatidylinositol bisphosphate. The rdgB gene product is thought to shuffle phosphatidylinositol from an intracellular compartment, where it is synthesized to the microvillar membrane.

185

186

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

In addition, a number of photoreceptor proteins have been described that mediate Ca2+- or Ca2+-calmodulin-dependent feedback of the signaling cascade. These include the eye-specific protein kinase C (inaC) [51, 52], the scaffolding protein INAD, [53] and the calmodulin-binding protein NINAC [54], which is an unconventional myosin that may interact with the actin cytoskeleton of the microvillus.

8.4

Essential Components of the Transduction Pathway Are Organized into a Multimolecular Signaling Complex

For a long time it has been assumed that the signaling proteins of G-protein-coupled signaling cascades float freely in the cell membrane and transfer the signal from activated molecules to their downstream targets via random collisions of the signaling proteins. However, the finding that essential components of the Drosophila phototransduction cascade are assembled into a multimolecular signaling complex showed that these signaling pathways may be organized in much more highly ordered molecular structures (for reviews, see [66–70]). The crucial protein of the signaling complex is INAD. INAD is composed of five protein-protein interaction domains, so-called PDZ domains, that form a scaffold to which other proteins can bind. PDZ domains are protein modules of about 90 amino acids that fold into a three-dimensional structure that provides a binding pocket for specific target sequences, such as S/T-X-V/I, typically located at the C-terminus of the target protein [71]. The name PDZ refers to the three proteins in which this module was identified first, PSD-95 (postsynaptic density protein), dlg (Drosophila disks large), and the tight-junction protein ZO-1. As initially revealed by co-immunoprecipitation, the major interaction partners of INAD are the norpA-encoded phospholipase Cb, the TRP ion channel, and the regulatory protein kinase C, INAC (Fig. 8.3) [72–75]. These proteins are tethered together in the photoreceptor membrane and constitute a functional unit (Fig. 8.4). When the scaffolding protein INAD is missing, the other three proteins are no longer confined to the rhabdomeral membrane and they become degraded in older flies [74–78]. Likewise, INAD requires the ion channel TRP for proper localization to the rhabdomere. The studies of Li and Montell [79] and Tsunoda et al. [80] on the subcellular localization of the complex components in inaD and trp mutants revealed that TRP and INAD are targeted to the rhabdomeral membrane independently but require each other to be retained in the photoreceptive cell compartment. The proper localization of the signaling complex is not affected by mutations in inaC and norpA, indicating that ePKC and PLCb are not necessary for linking the complex to the rhabdomeral membrane. There is evidence that ePKC and PLCb are bound to INAD before they reach the rhabdomeral membrane [80], while TRP, which can reach the rhabdomeral membrane independently of INAD, may become incorporated into the signaling complex at a later stage. In addition to the major components of the INAD signaling complex, other key players of the phototransduction cascade have been shown to interact with INAD [75, 81, 82]. These proteins are rhodopsin, the TRPL ion channel, the unconventional

8.4 Essential Components of the Transduction Pathway Fig. 8.3 Co-immunoprecipitation of the INAD signaling complex. The left panel shows a silverstained gel of a protein extract obtained from Drosophila head membranes that was immunoprecipitated with anti-TRP antibodies (lane 3). The four protein bands detected in the immunoprecipitate were identified as the core components of the INAD signaling complex, TRP, PLCb, ePKC, and INAD, by Western blot analysis (right panel). The prominent bands between 45 and 60 kDa in lane 3 correspond to the antibody used for immunoprecipitation (modified after [28])

myosin NINAC, and calmodulin. While rhodopsin, TRPL, and NINAC interact with PDZ domains of INAD, calmodulin appears to bind to the linker region between PDZ1 and PDZ2 of INAD. These latter INAD-binding proteins differ from the major components of the INAD signaling complex in two ways. Firstly, while INAD, TRP, ePKC, and PLCb are present in the complex at about the same molar ratio [73], the other ligands seem to be present in the complex at a much lower level. This may suggest that they are not permanently associated with the complex but rather interact dynamically. A dynamic interaction with the signaling complex was shown for the visual G-protein alpha subunit, which binds to the complex in its activated form only [83]. Secondly, the correct localization to the rhabdomere and the stability of rhodopsin, TRPL, NINAC, and calmodulin do not depend on the presence of INAD [75,

Fig. 8.4 The INAD signaling complex. (A) Model for the association of the major components of the INAD signaling complex. (B) Four TRP, PLCb, ePKC, and INAD molecules may form a larger complex, containing one functional, tetrameric TRP

ion channel. Several of these complexes may then associate and build up higher organized structures (transducisomes), representing a functional phototransduction unit (reprinted from [70] with permission from Blackwell Science Ltd.)

187

188

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

82]. The finding that the unconventional myosin NINAC interacts with INAD has led Wes et al. [82] to suggest that the signaling complex may be linked to the actin filament present in the center of each microvillus of the photoreceptor membrane. Since disruption of the NINAC-INAD interaction results in a defect in the termination of the photoresponse, a possible coupling of the INAD signaling complex to the microvillar cytoskeleton may be of functional importance for response termination [82]. A further level of complexity in the molecular organization of the INAD signaling complex is added by the possibility that core complexes consisting of one molecule of each INAD, TRP, PLCb, and ePKC may multimerize to larger units that are referred to as signalplex [69] or transducisome [74] (Fig. 8.4). Such multimeres may be achieved, for example, by PDZ-PDZ interactions between INAD molecules [81], by the formation of tetrameric TRP channels, or by the simultaneous binding of PLCb to two INAD molecules. Detailed studies of the interactions of INAD with its binding partners using, for example, yeast two-hybrid assays, GST pull-down assays, co-immunoprecipitation of recombinantly expressed INAD constructs with its binding partners, and analysis of Drosophila mutants with mutations in specific PDZ domains provided information about which protein binds to which of the five PDZ domains of INAD. It turned out that the specificity of binding of some ligands is not limited to one PDZ domain of INAD. For example, TRP can interact with either PDZ3 or PDZ4, and ePKC was reported to bind to PDZ1, PDZ2, PDZ3, or PDZ4 [81, 84]. PLCb may interact simultaneously with PDZ1 through its C-terminal F-C-A motif and with PDZ5 through an internal sequence in the region of the G-protein-binding site [76, 77]. A crystal structure is available for the interaction of the C-terminal heptapeptide of PLCb with PDZ1 [85]. The structure analysis revealed the presence of a disulfide bridge between INAD and the cysteine residue in the F-C-A binding motif of PLCb. This covalent interaction is responsible for a high-affinity interaction between PLCb and INAD. The results of Kimple et al. [85] further suggest that PLCb molecules form homodimers that associate with INAD into heterotetramers, thus providing yet another means to establish higher-order signaling complexes. What is the functional significance of assembling proteins into signaling complexes? It is a fundamental principle of eukaryotic cells to compartmentalize different functions to different cellular sites. Sensory cells are particularly nice examples of this principle, because most of them have developed specific structures devoted exclusively to signal perception and transduction. Examples of such structures include the outer segments of vertebrate and the rhabdomeres of invertebrate photoreceptor cells or the cilia and stereovilli of olfactory cells and mechanosensitive hair cells, respectively. The identification of multimolecular signaling complexes suggests that compartmentalization occurs not only at the cellular level but also at the molecular level. The clustering of molecules that are part of the same signaling pathway may hinder unwanted crosstalk between different signaling cascades and thereby ensure high specificity of phototransduction [73, 74]. This should result in a high signal-to-noise ratio and facilitate the detection of minimal signals such as a single photon. A second advantage of the signaling complex may concern signaling speed. It seems reasonable to assume that bringing the components of a signal transduction cascade closely together, and

8.5 TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors

thereby minimizing or even eliminating diffusion times of the signaling molecules, would accelerate signaling speed. Indeed, the light responses of a number of Drosophila mutants in which the interaction of TRP, ePKC, or PLCb with INAD is abolished show prolonged deactivation times [53, 76, 84]. However, since these proteins require interaction with INAD for correct localization in the rhabdomeral membrane, it is not quite clear whether the observed physiological defects result from mislocalization of the signaling proteins or from failure to incorporate into the signaling complex. At least one TRP mutant that lacks the INAD-binding site shows light responses indistinguishable from wild-type in young flies, in which TRP is still confined to the rhabdomeral membrane but develops a deactivation defect when TRP becomes mislocalized with age [79]. The question of whether the signaling complexes increase signaling specificity has not been addressed experimentally so far.

8.5

TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors 8.5.1

Identification and Characterization of TRP and TRPL

A spontaneously occurring Drosophila trp mutant was isolated by Cosens and Manning [86] and was termed “transient receptor potential” because the light-response decayed to baseline during prolonged illumination [87]. Subsequent cloning of the trp gene revealed that it encoded a putative transmembrane protein that, at that time, had no significant homologies to known proteins [35, 88]. It was not immediately recognized that the trp gene encodes a phototransduction channel. This possibility was considered unlikely, because it was assumed that elimination of the phototransduction channel should result in a completely blind fly rather than in a fly showing a transient light response. A major breakthrough in the understanding of TRP function was achieved when elegant whole-cell voltage-clamp recordings of the light-sensitive conductance from photoreceptor cells of trp mutants revealed that the Ca2+ permeability in these cells was greatly reduced [36]. These findings led Hardie and Minke to conclude that there were at least two phototransduction channels in Drosophila photoreceptor cells, the TRP-dependent channels with high Ca2+ permeability and a second ion channel with a much lower Ca2+ permeability. Indeed, a second ion channel showing about 40 % amino acid identity to TRP was identified as a calmodulin-binding protein and was termed TRP-like (TRPL) [37]. A second line of evidence suggesting that TRP is an ion channel came from the finding that application of La3+, a known nonspecific Ca2+ channel blocker, to the extracellular space of wild-type photoreceptor cells phenocopied the trp mutation [36, 89]. At micromolar concentrations La3+ specifically inhibits TRP but not TRPL. The generation and analysis of trpl null mutants and trp;trpl double mutants revealed that the photoresponse is abolished when neither TRP nor TRPL is functionally expressed in the photoreceptor cells [38, 90]. The lack of TRPL alone has a relatively subtle but distinct influence on the light response: it alters the adaptation properties during dim background light and the dynamic range of de-

189

190

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

tectable light-intensities [91, 92]. A third TRP homologue, TRPc, expressed in Drosophila photoreceptor cells was isolated [39]. TRPc may participate in the photoresponse by forming heteromultimeric channels with TRPL, but it cannot form functional lightactivated channels on its own. The availability of trp and trpl null mutants allowed a genetic separation of the light-induced currents carried by either TRP or TRPL, which furthered a detailed characterization of the TRP and TRPL channels in vivo [38, 90]. Besides its higher Ca2+ permeability and susceptibility to La3+ blockage, TRP differs from TRPL in the single-channel conductances (Tab. 8.2). Because it proved very difficult to obtain excised patches from microvillar membranes containing single channels, single-channel conductance was deduced from noise analysis of whole-cell patchclamp recordings. These revealed a rather small single-channel conductance of ca. 8 pS for TRP and a higher conductance of ca. 35 pS for TRPL. On the other hand, the number of TRPL channel subunits in photoreceptor cells is about 10 times lower than that of TRP [81, 93] and, depending on the light condition in which the flies are kept, TRPL may not be located in the rhabdomere at all (see below). In addition, TRPL is efficiently inhibited by a rise in intracellular Ca2+. Therefore, under physiological conditions, a large fraction of the light-induced current is carried by TRP, despite its low single-channel conductance. The effect of Ca2+ on TRP channels is biphasic; initially the rise in intracellular Ca2+ amplifies the activation of TRP channels, but Ca2+ has an inhibitory effect when applied at a later phase of the response. The positive and subsequent negative feedback of Ca2+ on TRP may determine the kinetics of the formation of elementary responses to single photons that sum up to generate the macroscopic response to brighter light stimuli [1].

Tab. 8.2

Characteristics of Drosophila TRP and TRPL TRP

TRPL

Molecular weight Single-channel conductance Reversal potential at 1.5 mM external Ca2+ Ca2+ selectivity PCa:PNa Channel blocker Estimated amount present in eye Subcellular localization

143 kDa 8 pS 15 mV

128 kDa 35 pS 0 mV

100:1 La3+ 1:10 of rhodopsin

4:1 – 1:100 of rhodopsin

Rhabdomere

Mutant phenotype

Transient response to light, retinal degeneration

Relocates between rhabdomere and cell body Weak phenotype: absence of TRPL from the rhabdomere affects adaptation to dim light and steepness of the intensity-response curve

8.5 TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors

8.5.2

Possible Gating Mechanism

The norpA-encoded phospholipase Cb is mandatory for the light response of Drosophila photoreceptors. This enzyme hydrolyzes PIP2 in the rhabdomeral membrane and generates the second messengers IP3 and DAG. IP3 is well known for its function to release Ca2+ from internal stores such as the endoplasmic reticulum, which can subsequently result in the opening of plasma membrane ion channels by a mechanism termed store-operated Ca2+ entry [94, 95]. Therefore, a store-depletion mechanism was considered an attractive hypothesis for TRP and TRPL gating (e.g., [96, 97]). According to this model, IP3 would bind to an IP3-receptor in the submicrovillar cisternae, a putative Ca2+ store located at the base of the rhabdomere. Ca2+ release from this store would subsequently result in opening of the phototransduction channels, for example, via direct conformational coupling between the IP3-receptor and the TRP channels. The transient response of trp mutants could then be explained by a failure to refill the Ca2+ stores via Ca2+ influx from the extracellular space, because Ca2+ influx is dramatically reduced in the absence of TRP. However, the store-depletion hypothesis has largely been abandoned, at least for Drosophila photoreception, because abundant evidence against it has accumulated. These include studies showing that application of caged IP3 did not open the channels [98], nor did artificial Ca2+ store depletion induced by application of thapsigargin, a drug that blocks the smooth endoplasmic reticulum Ca2+ pump [99]. Most importantly, genetic disruption of the only IP3-receptor gene in Drosophila had no detectable effect on the light response [100, 101]. TRP gating by store-operated Ca2+ entry via activation of a ryanodine receptor could also be excluded [102]. The lack of results supporting TRP gating through store-operated Ca2+ entry prompted research directed at elucidating the function of the other branch of phospholipid signaling, the generation of DAG. DAG is best known for its role in activating protein kinase C. Indeed, one protein kinase C, ePKC, is part of the INAD signaling complex and has been shown to phosphorylate INAD and TRP [103–105]. A mutation in the inaC gene encoding ePKC leads to a defect in response inactivation and in light adaptation, but not in excitation [51, 106]. Thus, it is possible that ePKC-dependent phosphorylation of TRP participates in ion-channel regulation, but it seems unlikely that this reaction underlies TRP gating. There are several lines of evidence suggesting that DAG or polyunsaturated fatty acids (PUFAs), which might be released from DAG by a DAG-lipase, act directly on the channels. Reversible activation of TRP and TRPL by PUFAs (arachidonic, linolenic, and linoleic acid) has been reported when they were applied to isolated photoreceptors or to excised patches of Drosophila S2 cells containing recombinantly expressed TRPL [107]. However, evidence for the presence of a DAG lipase in Drosophila photoreceptors, which would be an essential component of the phototransduction cascade if PUFAs were the physiological activators of TRP and TRPL, is missing so far. Direct or indirect activation of TRP channels by DAG itself has been shown for the vertebrate channels TRPC3 and TRPC6 [108], and also for recombinantly expressed TRPL [109], but not for the Drosophila TRP channels in photoreceptor cells. A second line of evidence for an important role of the DAG

191

192

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

branch in the excitation process comes from analysis of the rdgA mutant. rdgA encodes a DAG kinase that converts DAG to phosphatidic acid [63]. In this mutant, the TRP and TRPL channels are constitutively active from the late pupal stage onwards, and this results in fast and light-independent degeneration of the photoreceptor cells [110]. The degeneration phenotype of rdgA was partially rescued in rdgA;trp and rdgA;norpA double mutants, consistent with the assumption that the primary cause for degeneration is constitutive Ca2+ influx through TRP channels [110, 111]. The light-response mediated by TRPL channels can be measured in the rdgA;trp double mutant. It shows a distinct deactivation defect. These results can be explained by assuming that DAG generated from PIP2 cannot be removed in the rdgA mutant and thus triggers prolonged or even constitutive activation of the channels. A drawback to this line of argument is the finding that degeneration in the rdgA mutant is light independent. Thus, one has to assume that enough DAG accumulates in the dark to constitutively activate TRP and TRPL by spontaneous activity of the phototransduction cascade. Biochemical analysis of the DAG content did not reveal significant accumulation of DAG in the eyes of the rdgA mutant [112]. Further, immunolocalization of the RDGA protein showed that it is predominantly associated with the submicrovillar cisternae, suggesting that conversion of DAG to phosphatidic acid may take place mainly outside the rhabdomeral membrane [113]. As another possibility for TRP and TRPL gating, the light-dependent decrease in the PIP2 content of the photoreceptive membrane could contribute to the activation of the channels (as discussed in [114]). Some evidence for this assumption comes from studies of recombinant TRPL activity that could be quenched by application of PIP2 [109]. An elegant study by Hardie et al. [115] used genetically targeted Kir2.1 channels as biosensors for monitoring the PIP2 concentration in Drosophila photoreceptors. This study revealed that PIP2 is completely depleted in the microvillar membrane of the trp mutant after prolonged illumination, while the TRPL-mediated light response declines to baseline. This finding argues against activation of TRPL by PIP2 depletion in vivo, although it does not preclude that TRP is regulated in that way. The light-induced loss of PIP2 in the trp mutant provides a possible explanation for the trp phenotype, if one accepts that the availability of PIP2 in the microvillar membrane is mandatory for visual excitation. A regulatory function of PIP2 has been established for vertebrate TRPM7 and TRPV1, although PIP2 regulates these channels in opposite directions. In the case of TRPM7, hydrolysis of PIP2 by a PLC, which is directly associated with TRPM7, results in channel inhibition [116]. TRPV1, on the other hand, is inhibited by the presence of PIP2, which binds to a motif in the C-terminal domain of TRPV1 [117]. It has not yet been determined whether Drosophila TRP and TRPL contain a PIP2-binding domain. Besides by phospholipids, there is at least one other way by which TRP and TRPL can be activated. Metabolic stress, induced by anoxia in intact Drosphila via application of N2 or induced in situ by mitochondrial uncouplers such as 2, 4-dinitrophenol, resulted in robust activation of TRP and TRPL [118]. This activation is independent of PLCb because it can be achieved in wild-type as well as in the norpA mutant. Both anoxia and mitochondrial uncouplers lead to depletion of ATP in the photoreceptor cell, which in turn seems to trigger channel opening. Indeed, omitting ATP and NAD

8.5 TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors

from the recording pipette of photoreceptor cells led to a delayed slow inward current following a few light pulses of medium intensity [118]. Why would depletion of ATP trigger channel opening? In light of the DAG/PUFA-hypothesis, this may be explained simply as a failure to remove DAG from the photoreceptor membrane by ATP-dependent phosphorylation by DAG kinase [114]. On the other hand, it has been argued that PUFAs do not activate the TRP and TRPL channels directly but rather act on the channels indirectly through their known function as mitochondrial uncouplers [119]. Then, ATP might be required more directly to keep the channels in a closed state, for example, through phosphorylation of the channels themselves. When discussing putative gating mechanisms for TRP, it should be kept in mind that this ion channel is incorporated into the INAD signaling complex. Possibly, conformational changes of the entire complex, initiated by activation of PLCb, which is also a member of the complex, may affect the behavior of the channel. It has been shown that Calliphora and Drosophila TRP expressed in Sf9 cells alters its gating properties when co-expressed with INAD from a store-dependent into a store-independent channel [120, 121].

8.5.3

Transduction Channels in the Visual Systems of Other Invertebrates

The findings suggesting that Drosophila phototransduction employs TRP ion channels, which are not activated by the IP3 branch but rather by the DAG branch of phospholipid signaling, must not be generalized for all invertebrate photoreceptors. True orthologs of Drosophila TRP and TRPL can be identified for the related fly species Calliphora ([73], B€ahner et al., unpublished) and in the annotated genome sequence of the mosquito Anopheles gambiae (The Anopheles Genome Sequencing Consortium; database accession numbers: XP 310766, XP 307827, and XP 317008). The Calliphora and Anopheles TRP channels show high sequence identities of more than 70 % in the region comprising the six transmembrane domains when compared to the respective Drosophila TRP channel. It seems reasonable to assume that these channels are gated in much the same way as Drosophila TRP and TRPL. Three splice variants of a Limulus polyphemus TRP gene have been identified, two of which are expressed in the horseshoe crab ventral eye (Bandyopadhyay and Payne, personal communication; database accession numbers AAN38979, AAN38980, and AAN38981). The Limulus TRP channels show significant but much lower identity (33–40 %) to Drosophila TRP. Likewise, a TRP homologue with moderate identity to Drosophila TRP has been isolated from squid (Loligo forbesi) photoreceptor membranes [122]. For the Limulus ventral eye photoreceptor, there is strong evidence that light-induced generation of IP3 results in Ca2+ release from internal stores and that application of IP3 or Ca2+ generates a receptor current (see [123]). Light-triggered Ca2+ release from submicrovillar ER has also been reported for bee photoreceptors (see [124]). It thus seems likely that invertebrate photoreceptors have developed more than one mechanism for generating the light-activated conductance. Indeed, since cyclic nucleotide-gated channels have been cloned from Limulus and have been shown to be located in the transducing lobes

193

194

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

of ventral eye and lateral eye photoreceptors [125, 126], it is possible that these photoreceptors employ cyclic nucleotide-gated channels for vision in addition to or instead of TRP-like channels. Likewise, abundant evidence suggests that the light response of invertebrate ciliary photoreceptors found, for example, in mollusks such as Pecten is mediated by cyclic nucleotide-gated channels [123].

8.5.4

Drosophila TRP Is the Founding Member of the TRP Family of Ion Channels

The light-activated cation channel TRP of Drosophila photoreceptors was the first identified member of an important ion channel family. Other members of the TRP family mediate sensory modalities other than vision, for example, olfaction, thermosensation, mechanoreception, nociception, and taste reception (Tab. 8.3). Additional TRP channels are present in non-neuronal cells, where they may be involved in vasorelaxation, cell cycle control, or even in tumor suppression (for recent reviews on the TRP family, see [127–133]). TRP proteins are cation channels with a more or less pronounced selectivity for Ca2+ as compared to Na+ (see Tab. 8.2). The proposed structural model

Fig. 8.5 Model of Drosophila TRP. (A) Structural features of the TRP channel include six transmembrane regions (1–6) and a putative ion pore between transmembrane domains 5 and 6. Four ankyrin repeats near the N-terminus and the TRP domain, a calmodulin-binding site (CaM), and two possible INAD-binding motifs (STXV, SGWL) near the Cterminus are indicated. (B) Hypothetical arrangement of four TRP subunits forming a functional channel; 1–6 denote transmembrane domains

8.5 TRP and TRPL, the Transduction Channels of Drosophila Photoreceptors Tab. 8.3

TRP channels mediating sensory transduction

Subfamily Ion channel TRPC

TRPV

TRPM

TRPN

Sensory modality

Sensory system

Reference 35, 36 37, 38 39

DmTRP DmTRPL DmTRPc ?

Vision

Drosophila photoreceptors

TRPC2

Olfaction

Mammalian vomeronasal organ

135–137

TRPV1 TRPV2 TRPV3 TRPV4

Thermosensation (moderate to hot), Nociception

Mammalian skin: free nerve endings from dorsal root and trigeminal ganglia, keratinocytes; inner organs, aorta

138 139 140–142 143, 144

DmNanchung

Hearing

145

CeOSM-9 CeOCR-2

Various sensory functions

Drosophila mechanoreceptors in Johnston’s organ C. elegans sensory neurons

TRPM5

Taste

Mammalian sweet, bitter, and amino acid taste receptor cells

147, 148

TRPM8

Thermosensation (cold)

Mammalian skin: free nerve endings 149, 150 from dorsal root and trigeminal ganglia

DmNOMPC

Mechanoreception Drosophila mechanoreceptors in sensory bristles

151

Zebrafish NOMPC

Mechanoreception Zebra fish sensory hair cells

152

DmPainless

Thermosensation (hot), nociception

ANKTM1 (= TRPN1)

Thermosensation Mammalian skin: free nerve endings (cold), nociception from dorsal root ganglia, fibroblasts

154

DmANKTM1

Thermosensation (warm)

Not determined

155

Drosophila larval multi-dendritic sensory neurons

146

153

for TRP channels assumes six transmembrane regions with a pore-forming domain between transmembrane segments 5 and 6 (Fig. 8.5). The N- and C-terminal domains of the protein are located intracellularly. In analogy to voltage-gated Ca2+ channels, the dihydropyridine receptors, to which TRP channels show limited sequence homology, it is assumed that functional channels are composed of four subunits [37]. On the basis of sequence homologies between TRP proteins, which is in general highest within the six transmembrane domains and in the putative pore region, the channels have been assigned to four different subfamilies: TRPC, TRPV, TRPM, and TRPN (Tab. 8.3). Additional structural features of TRP proteins include ankyrin repeats found in

195

196

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

the N-terminal region of TRPs, which are absent in the TRPM subfamily, and a conserved motif, the TRP domain, adjacent to the sixth transmembrane region, which is absent only in the TRPN subfamily. Ankyrin repeats are protein-protein interaction domains of about 30 amino acid residues, which can bind to a variety of molecular targets [134]. However, a specific target protein has not yet been identified for the ankyrin repeats of TRP channels. The TRP domain comprises about 25 amino acids, including a highly conserved E-W-K-F-A-R-motif referred to as the TRP box [129]. The function of the TRP domain is also still unknown.

8.6

Light-dependent Relocation of TRPL Alters the Properties of the Photoreceptive Membrane

The physiological properties of a neuron are largely determined by the ion channel equipment of the cell membrane. Consequently, regulating the amount and composition of ion channels at the cell surface is a means to alter the characteristics of a neuron. It has been shown that the number of TRPL channel subunits in the microvillar photoreceptor membrane, but not of TRP channels, undergoes dramatic light-dependent alterations [92]. The microvillar membrane contains a high level of TRPL in the dark and a low level in the light. When flies are switched from darkness to light, the amount of microvillar TRPL decreases from 100 % to 30 % within one hour and down to 4 % within 12–16 hours. When the flies are returned to darkness, the microvillar membrane is completely refilled with TRPL within one hour. The steady-state levels of trpl mRNA and of total TRPL present in the eye remain unaltered. Therefore, the changes in the microvillar TRPL content seem to result from a light-regulated translocation of TRPL molecules from the rhabdomere to an intracellular storage compartment. This is also supported by immunocytochemical studies showing TRPL labeling in the rhabdomeres of dark-raised flies and labeling in the cell body of light-raised flies (Fig. 8.6). The relocation of TRPL changes the ratio of TRP:TRPL in the microvillar membrane from ca. 10:1 in the dark to ca. 200:1 in the light. The physiological consequences of changing the equipment of the microvillar membrane with different ion channel subunits include a shift in the reversal potential of the light-induced current and an altered sensitivity of the light-induced current for the TRP channel inhibitor La3+. As expected, when TRPL is outside the rhabdomere, the reversal potential shifts towards the reversal potential of TRP channels and the light-response is significantly reduced by application of La3+. The reduction of microvillar TRPL results in a failure of the photoreceptor cells to adapt to dim background light, as also has been reported for a trpl null mutant [91]. In addition, the presence or absence of TRPL in the photoreceptive membrane affects the intensity range in which the photoreceptor cells respond to light. This light-intensity range is wider when the flies are kept in the dark, while the photoreceptors of flies kept in the light, which are equipped with little microvillar TRPL, have a smaller light-intensity range but better resolution within this range. Thus, the light-dependent relocation of TRPL seems to be a long-term adaptation mechanism to fine-tune the visual response.

8.6 Light-dependent Relocation of TRPL Alters the Properties of the Photoreceptive Membrane

Fig. 8.6 Relocation of TRPL between rhabdomere and cell body. (A) Immunocytochemical localization of TRPL in cross-sections through the compound eye of Drosophila kept in the dark or light for 16 hours. TRPL (green) is located in the rhabdomeres in dark-raised flies but in the cell body in light-raised flies. The rhabdomeres are labeled with the lectin WGA (red). In addition, the rhodopsin Rh3 expressed in a subset of R7 cells is detected (blue). (B) A scheme indicating the relocation and turnover of TRPL

Two recent papers investigated a light-dependent relocation of the G-protein alpha subunit (Gaq) and of the major arrestin (Arr2) in Drosophila photoreceptors [156, 157]. Upon illumination, a fraction of Gaq switches from a membrane-bound state to a soluble state and relocates from the rhabdomere to the cell body. This relocation depends on the Gbc-subunit and seems to be associated with morphological changes at the base of the microvilli [156]. Upon illumination, Arr2 moves in the other direction, that is, from cell body to rhabdomere. Arr2 translocation depends on the presence of phosphoinositides, most likely PIP3, which bind to a C-terminal domain of Arr2 [157]. Mutation of a critical lysine residue in this motif interfered with Arr2 translocation and altered long-term adaptation properties of the visual response, suggesting that the relocation of Arr2 is another mechanism for long-term adaptation. The light-dependent translocation of Gaq is also likely to facilitate adaptation of the flies to different environmental light conditions.

197

198

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes

The cellular mechanisms underlying the relocation of TRPL, Gaq, and Arr2 may be quite different. Gaq changes from a membrane-associated state to a soluble state, possibly due to regulated depalmitoylation [156], and then may move away from the rhabdomere by passive diffusion or by active transport within the cytosol. Arr2 is internalized from the rhabdomere together with rhodopsin in membrane vesicles [48, 49], although this may not be the only means by which Arr2 translocates from the rhabdomere to the cell body. Arr2 switches from a membrane-bound state in illuminated photoreceptors to a soluble state in photoreceptors containing no activated rhodopsin [158]. Since most Arr2 is associated with the rhabdomeres after illumination, the light-induced association of Arr2 with metarhodopsin-containing rhabdomeral membranes seems to override the light-induced vesicular internalization of Arr2-metarhodopsin complexes. Unlike Gaq and Arr2, TRPL is a transmembrane protein that cannot assume a soluble state. Therefore, TRPL has to rely on vesicular transport for relocation within the cell, probably by light-regulated endocytosis of this ion channel.

8.7

Concluding Remarks and Outlook

The Drosophila phototransduction cascade is one of the best-studied G-protein-coupled signaling pathways, and it is a powerful model system for the analysis of transduction channels of the TRP ion channel family. What are the main unresolved issues about the phototransduction channels of Drosophila photoreceptors? Maybe the hottest topic concerns the exact gating mechanism of these ion channels. It seems clear that the lipid messengers generated from PIP2 are crucially involved in the gating of the channels, but why does metabolic stress, that is, the depletion of ATP, also result in channel opening? Still little is known about what modulates the activity of TRP and TRPL and how the incorporation of TRP into the INAD signaling complex affects its properties. Although association with the complex does not seem to be a principle requirement for TRP activation, it may underlie fast feedback regulation, for example, through TRP phosphorylation by ePKC, which is also a member of the signaling complex. Why are there at least two light-activated transduction channels in Drosophila photoreceptors? There have been controversial discussions about whether TRP and TRPL form exclusively homomeric channels, that is, TRP tetramers and TRPL tetramers, or whether they also form heteromeric TRP-TRPL channels, and this issue has remained unresolved so far. It is clear, however, that two functionally distinct channels exist in wild-type photoreceptors. If heteromeric channels exist, what is the stoichiometry of the channel subunits? One reason for the expression of two channels may lie in the possibility to fine-tune the visual response by light-dependent relocation of the TRPLcontaining channels. Details on the cell biological mechanism underlying TRPL relocation are not known. How is the endocytotic machinery of a photoreceptor cell instructed to specifically internalize TRPL in response to light, but not TRP or other rhabdomeral signal transduction proteins? Where is TRPL stored inside the cell, and how does TRPL “know” that it is time to return to the rhabdomere when the fly is put

8.7 Concluding Remarks and Outlook

back into darkness? Given the wide panel of biological functions attributed to the TRP protein family, finding the answers to these questions may have impact on quite a few neuro- and cell biology topics that reach far beyond an understanding of invertebrate vision.

Acknowledgments

The author is grateful to Noel Da Silva, Monika B€ahner, and Nina Meyer for help in the preparation of figures and to Reinhard Paulsen, Joachim Bentrop, and Irene Huber for critical comments on the manuscript. The work of the author is supported by Deutsche Forschungsgemeinschaft (Hu 839/2-1, Pa 274/6-2) and by the German-Israeli Foundation for Scientific Research and Development (I-724-2.13/2002). References 1

2

3

4

5

6

7

Hardie, R. C. and P. Raghu. Visual transduction in Drosophila. Nature 2001, 413, 186–193. Ranganathan, R., W. A. Harris, and C. S. Zuker. The molecular genetics of invertebrate phototransduction. Trends Neurosci. 1991, 14, 486–493. Hardie, R. C. Functional organization of the fly retina. In Progress in sensory physiology 5, Autrum, H., Ottoson, D., Perl, E. R., Schmidt, R. F., Shimazu, H., and Willis, W. D. (eds.), Springer-Verlag, New York, 1983, p. 1–81. Hardie, R. C. The photoreceptor array of the dipteran retina. Trends Neurosci. 1986, 9, 419–423. Wolff, T. and D. F. Ready. Pattern formation in the Drosophila retina. In The development of Drosophila melanogaster , Bate, M. and Arias, A. M. (eds.), Cold Spring Habour Laboratory Press, Plainview, New York, 1993, p. 1277–1325. Pellikka, M., G. Tanentzapf, M. Pinto, C. Smith, C. J. McGlade, D. F. Ready, and U. Tepass. Crumbs, the Drosophila homologue of human CRB1/RP12, is essential for photoreceptor morphogenesis. Nature 2002, 416, 143–149. Izaddoost, S., S. C. Nam, M. A. Bhat, H. J. Bellen, and K. W. Choi. Drosophila crumbs is a positional cue in photoreceptor adherens junctions and rhabdomeres. Nature 2002, 416, 178–183.

8

9

10

11

12

13

Wodarz, A., U. Hinz, M. Engelbert, and E. Knust. Expression of crumbs confers apical character on plasma-membrane domains of ectodermal epithelia of Drosophila. Cell 1995, 82, 67–76. den Hollander, A. I., J. B. ten Brink, Y. J. M. De Kok, S. van Soest, L. I. van den Born, M. A. van Driel, D. J. R. van de Pol, A. M. Payne et al. Mutations in a human homologue of Drosophila crumbs cause retinitis pigmentosa (RP12). Nature Genetics 1999, 23, 217–221. den Hollander, A. I., J. R. Heckenlively, L. I. van den Born, Y. J. M. De Kok, S. D. VeldeVisser, U. Kellner, B. Jurklies, M. J. van Schooneveld et al. Leber congenital amaurosis and retinitis pigmentosa with coats-like exudative vasculopathy are associated with mutations in the crumbs homologue 1 (CRB1) gene. American Journal of Human Genetics 2001, 69, 198–203. Zuker, C. S., A. F. Cowman, and G. M. Rubin. Isolation and structure of a rhodopsin gene from D. melanogaster. Cell 1985, 40, 851–858. O’Tousa, J. E., W. Baehr, R. L. Martin, J. Hirsh, W. L. Pak, and M. L. Applebury. The Drosophila ninaE gene encodes an opsin. Cell 1985, 40, 839–850. Chou, W. H., A. Huber, J. Bentrop, S. Schulz, K. Schwab, L. V. Chadwell, R. Paulsen, and S. G. Britt. Patterning of the R7 and R8 photoreceptor cells of Drosophila: evidence for induced and default cell-fate specification. Development 1999, 126, 607–616.

199

200

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes 14

15

16

17

18

19

20

21

22

23

24

Pollock, J. A. and S. Benzer. Transcript localization of four opsin genes in the three visual organs of Drosophila; RH2 is ocellus specific. Nature 1988, 333, 779–782. Yasuyama, K. and I. A. Meinertzhagen. Extraretinal photoreceptors at the compound eye’s posterior margin in Drosophila melanogaster. Journal of Comparative Neurology 1999, 412, 193–202. Helfrich-Forster, C., T. Edwards, K. Yasuyama, B. Wisotzki, S. Schneuwly, R. Stanewsky, I. A. Meinertzhagen, and A. Hofbauer. The extraretinal eyelet of Drosophila: Development, ultrastructure, and putative circadian function. Journal of Neuroscience 2002, 22, 9255–9266. Rubin, G. M., M. D. Yandell, J. R. Wortman, M. G. Gabor, C. R. Nelson, I. K. Hariharan, M. E. Fortini, P. W. Li et al. Comparative genomics of the eukaryotes. Science 2000, 287, 2204–2215. Gartner, W. and P. Towner. Invertebrate Visual Pigments. Photochemistry and Photobiology 1995, 62, 1–16. Salcedo, E., A. Huber, S. Henrich, L. V. Chadwell, W. H. Chou, R. Paulsen, and S. G. Britt. Blue- and green-absorbing visual pigments of Drosophila: ectopic expression and physiological characterization of the R8 photoreceptor cell-specific Rh5 and Rh6 rhodopsins. J. Neurosci. 1999, 19, 10716–10726. Pak, W. L., J. Grossfield, and K. S. Arnold. Mutants of the visual pathway of Drosophila melanogaster. Nature 1970, 227, 518–520. Shieh, B. H., M. A. Stamnes, S. Seavello, G. L. Harris, and C. S. Zuker. The ninaA gene required for visual transduction in Drosophila encodes a homologue of cyclosporin A-binding protein. Nature 1989, 338, 67–70. Schneuwly, S., R. D. Shortridge, D. C. Larrivee, T. Ono, M. Ozaki, and W. L. Pak. Drosophila-nina gene encodes an eye-specific cyclophilin (cyclosporine-A binding-protein). Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 5390–5394. von Lintig, J., A. Dreher, C. Kiefer, M. F. Wernet, and K. Vogt. Analysis of the blind Drosophila mutant ninaB identifies the gene encoding the key enzyme for vitamin A formation in vivo. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 1130–1135. Kiefer, C., E. Sumser, M. F. Wernet, and J. von Lintig. A class B scavenger receptor mediates the cellular uptake of carotenoids in Drosophila. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10581–10586.

25

26 27

28

29

30

31

32

33

34

35

Zuker, C. S. The biology of vision of Drosophila. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 571–576. Montell, C. Visual transduction in Drosophila. Annu. Rev. Cell Dev. Biol. 1999, 15 231–268. Paulsen, R., M. B€ahner, A. Huber, M. Schillo, S. Schulz, R. Wottrich, and J. Bentrop. The molecular design of a visual cascade: Molecular stages of phototransduction in Drosophila. In Vision: The approach of biophysics and neurosciences, Musio, C. (ed.), World Scientific, Singapure, New Jersey, London, Hong Kong, 2001, p. 41–59. Paulsen, R., M. B€ahner, J. Bentrop, M. Schillo, S. Schulz, and A. Huber. The molecular design of a visual cascade: Assembly of the Drosophila phototransduction pathway into a supramolecular signaling complex. In Vision: The approach of biophysics and neurosciences, Musio, C. (ed.), World Scientific, Singapure, New Jersey, London, Hong Kong, 2001, p. 60–73. Lee, Y. J., M. B. Dobbs, M. L. Verardi, and D. R. Hyde. dgq: a Drosophila gene encoding a visual system-specific G alpha molecule. Neuron 1990, 5, 889–898. Scott, K., A. Becker, Y. Sun, R. Hardy, and C. Zuker. Gq alpha protein function in vivo: genetic dissection of its role in photoreceptor cell physiology. Neuron 1995, 15, 919–927. Yarfitz, S., G. A. Niemi, J. L. McConnell, C. L. Fitch, and J. B. Hurley. A G beta protein in the Drosophila compound eye is different from that in the brain. Neuron 1991, 7, 429–438. Dolph, P. J., S. H. Man, S. Yarfitz, N. J. Colley, J. R. Deer, M. Spencer, J. B. Hurley, and C. S. Zuker. An eye-specific G beta subunit essential for termination of the phototransduction cascade. Nature 1994, 370, 59–61. Schulz, S., A. Huber, K. Schwab, and R. Paulsen. A novel Gc isolated from Drosophila constitutes a visual G protein c subunit of the fly compound eye. J.Biol.Chem. 1999, 274, 37605–37610. Bloomquist, B. T., R. D. Shortridge, S. Schneuwly, M. Perdew, C. Montell, H. Steller, G. Rubin, and W. L. Pak. Isolation of a putative phospholipase C gene of Drosophila, norpA, and its role in phototransduction. Cell 1988, 54, 723–733. Montell, C. and G. M. Rubin. Molecular characterization of the Drosophila trp locus: a putative integral membrane protein required for phototransduction. Neuron 1989, 2, 1313–1323.

8.7 Concluding Remarks and Outlook 36

37

38

39

40

41

42

43

44

45

46

47

Hardie, R. C. and B. Minke. The trp gene is essential for a light-activated Ca2+ channel in Drosophila photoreceptors. Neuron 1992, 8, 643–651. Phillips, A. M., A. Bull, and L. E. Kelly. Identification of a Drosophila gene encoding a calmodulin-binding protein with homology to the trp phototransduction gene. Neuron 1992, 8, 631–642. Niemeyer, B. A., E. Suzuki, K. Scott, K. Jalink, and C. S. Zuker. The Drosophila light-activated conductance is composed of the two channels TRP and TRPL. Cell 1996, 85, 651–659. Xu, X. Z., F. Chien, A. Butler, L. Salkoff, and C. Montell. TRPgamma, a Drosophila TRPrelated subunit, forms a regulated cation channel with TRPL. Neuron 2000, 26, 647–657. Hyde, D. R., K. L. Mecklenburg, J. A. Pollock, T. S. Vihtelic, and S. Benzer. Twenty Drosophila visual system cDNA clones: one is a homolog of human arrestin. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1008–1012. Smith, D. P., B. H. Shieh, and C. S. Zuker. Isolation and structure of an arrestin gene from Drosophila. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1003–1007. Dolph, P. J., R. Ranganathan, N. J. Colley, R. W. Hardy, M. Socolich, and C. S. Zuker. Arrestin function in inactivation of G proteincoupled receptor rhodopsin in vivo. Science 1993, 260, 1910–1916. LeVine, H., D. P. Smith, M. Whitney, D. M. Malicki, P. J. Dolph, G. F. Smith, W. Burkhart, and C. S. Zuker. Isolation of a novel visualsystem-specific arrestin: an in vivo substrate for light-dependent phosphorylation. Mech. Dev. 1990, 33, 19–25. Vinos, J., K. Jalink, R. W. Hardy, S. G. Britt, and C. S. Zuker. A G protein-coupled receptor phosphatase required for rhodopsin function. Science 1997, 277, 687–690. Steele, F. R., T. Washburn, R. Rieger, and J. E. Otousa. Drosophila retinal-degeneration-C (Rdgc) encodes a novel serine threonine protein phosphatase. Cell 1992, 69, 669–676. Byk, T., M. Bar-Yaacov, Y. N. Doza, B. Minke, and Z. Selinger. Regulatory arrestin cycle secures the fidelity and maintenance of the fly photoreceptor cell. Proc.Natl.Acad.Sci.U.S.A. 1993, 90, 1907–1911. Plangger, A., D. Malicki, M. Whitney, and R. Paulsen. Mechanism of arrestin 2 function in rhabdomeric photoreceptors. J.Biol.Chem. 1994, 269, 26969–26975.

48

49

50

51

52

53

54

55

56

57

58

59

Alloway, P. G., L. Howard, and P. J. Dolph. The formation of stable rhodopsin-arrestin complexes induces apoptosis and photoreceptor cell degeneration. Neuron 2000, 28, 129–138. Kiselev, A., M. Socolich, J. Vinos, R. W. Hardy, C. S. Zuker, and R. Ranganathan. A molecular pathway for light-dependent photoreceptor apoptosis in Drosophila. Neuron 2000, 28, 139–152. Hardie, R. C. Regulation of TRP channels via lipid second messengers. Annual Review of Physiology 2003, 65, 735–759. Smith, D. P., R. Ranganathan, R. W. Hardy, J. Marx, T. Tsuchida, and C. S. Zuker. Photoreceptor deactivation and retinal degeneration mediated by a photoreceptor-specific protein kinase C. Science 1991, 254, 1478–1484. Schaeffer, E., D. Smith, G. Mardon, W. Quinn, and C. Zuker. Isolation and characterization of two new Drosophila protein kinase C genes, including one specifically expressed in photoreceptor cells. Cell 1989, 57, 403–412. Shieh, B. H. and B. Niemeyer. A novel protein encoded by the InaD gene regulates recovery of visual transduction in Drosophila. Neuron 1995, 14, 201–210. Montell, C. and G. M. Rubin. The Drosophila ninaC locus encodes two photoreceptor cell specific proteins with domains homologous to protein kinases and the myosin heavy chain head. Cell 1988, 52, 757–772. Cowman, A. F., C. S. Zuker, and G. M. Rubin. An opsin gene expressed in only one photoreceptor cell type of the Drosophila eye. Cell 1986, 44, 705–710. Zuker, C. S., C. Montell, K. Jones, T. Laverty, and G. M. Rubin. A rhodopsin gene expressed in photoreceptor cell R7 of the Drosophila eye: homologies with other signal-transducing molecules. J. Neurosci. 1987, 7, 1550–1557. Fryxell, K. J. and E. M. Meyerowitz. An opsin gene that is expressed only in the R7 photoreceptor cell of Drosophila. EMBO J. 1987, 6, 443–451. Montell, C., K. Jones, C. Zuker, and G. Rubin. A second opsin gene expressed in the ultraviolet-sensitive R7 photoreceptor cells of Drosophila melanogaster. J.Neurosci. 1987, 7, 1558–1566. Chou, W. H., K. J. Hall, D. B. Wilson, C. L. Wideman, S. M. Townson, L. V. Chadwell, and S. G. Britt. Identification of a novel Drosophila opsin reveals specific patterning of the R7 and R8 photoreceptor cells. Neuron 1996, 17, 1101–1115.

201

202

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes 60

61

62

63

64

65

66

67

68

69 70

71

72

Papatsenko, D., G. Sheng, and C. Desplan. A new rhodopsin in R8 photoreceptors of Drosophila: evidence for coordinate expression with Rh3 in R7 cells. Development 1997, 124, 1665–1673. Huber, A., S. Schulz, J. Bentrop, C. Groell, U. Wolfrum, and R. Paulsen. Molecular cloning of Drosophila Rh6 rhodopsin: the visual pigment of a subset of R8 photoreceptor cells. FEBS Lett. 1997, 406, 6–10. Li, C., C. Geng, H. T. Leung, Y. S. Hong, L. L. Strong, S. Schneuwly, and W. L. Pak. INAF, a protein required for transient receptor potential Ca(2+) channel function. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 13474–13479. Masai, I., A. Okazaki, T. Hosoya, and Y. Hotta. Drosophila retinal degeneration A-gene encodes an eye-specific diacylglycerol kinase with cysteine-rich zinc-finger motifs and ankyrin repeats. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 11157–11161. Wu, L., B. Niemeyer, N. Colley, M. Socolich, and C. S. Zuker. Regulation of PLC-mediated signalling in vivo by CDP-diacylglycerol synthase (see comments). Nature 1995, 373, 216–222. Vihtelic, T. S., D. R. Hyde, and J. E. O’Tousa. Isolation and characterization of the Drosophila retinal-degeneration-B (Rdgb) gene. Genetics 1991, 127, 761–768. Ranganathan, R. and E. M. Ross. PDZ domain proteins: scaffolds for signaling complexes. Curr. Biol. 1997, 7, R770–R773. Tsunoda, S., J. Sierralta, and C. S. Zuker. Specificity in signaling pathways: assembly into multimolecular signaling complexes. Curr. Opin. Genet. Dev. 1998, 8, 419–422. Tsunoda, S. and C. S. Zuker. The organization of INAD-signaling complexes by a multivalent PDZ domain protein in Drosophila photoreceptor cells ensures sensitivity and speed of signaling. Cell Calcium 1999, 26, 165–171. Montell, C. TRP trapped in fly signaling web. Curr. Opin. Neurobiol. 1998, 8, 389–397. Huber, A. Scaffolding proteins organize multimolecular protein complexes for sensory signal transduction. Eur. J. Neurosci. 2001, 14, 769–776. Sheng, M. and C. Sala. PDZ domains and the organization of supramolecular complexes. Annual Review of Neuroscience 2001, 24, 1–29. Shieh, B. H. and M. Y. Zhu. Regulation of the TRP Ca2+ channel by INAD in Drosophila photoreceptors. Neuron 1996, 16, 991–998.

73

74

75

76

77

78

79

80

81

82

Huber, A., P. Sander, A. Gobert, M. Ba¨hner, R. Hermann, and R. Paulsen. The transient receptor potential protein (Trp), a putative store-operated Ca2+ channel essential for phosphoinositide-mediated photoreception, forms a signaling complex with NorpA, InaC and InaD. EMBO J. 1996, 15, 7036–7045. Tsunoda, S., J. Sierralta, Y. Sun, R. Bodner, E. Suzuki, A. Becker, M. Socolich, and C. S. Zuker. A multivalent PDZ-domain protein assembles signalling complexes in a G- protein-coupled cascade. Nature 1997, 388, 243–249. Chevesich, J., A. J. Kreuz, and C. Montell. Requirement for the PDZ domain protein, INAD, for localization of the TRP store-operated channel to a signaling complex. Neuron 1997, 18, 95–105. Shieh, B. H., M. Y. Zhu, J. K. Lee, I. M. Kelly, and F. Bahiraei. Association of INAD with NORPA is essential for controlled activation and deactivation of Drosophila phototransduction in vivo. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 12682–12687. van Huizen, R., K. Miller, D. M. Chen, Y. Li, Z. C. Lai, R. W. Raab, W. S. Stark, R. D. Shortridge et al. Two distantly positioned PDZ domains mediate multivalent INAD- phospholipase C interactions essential for G protein-coupled signaling. EMBO J. 1998, 17, 2285–2297. Huber, A., G. Belusic, N. Da Silva, M. Ba¨hner, G. Gerdon, K. Draslar, and R. Paulsen. The Calliphora rpa mutant lacks the PDZ domainassembled INAD signalling complex. Eur. J. Neurosci. 2000,12, 3909–3918. Li, H. S. and C. Montell. TRP and the PDZ protein, INAD, form the core complex required for retention of the signalplex in Drosophila photoreceptor cells. J. Cell Biol. 2000, 150, 1411–1422. Tsunoda, S., Y. Sun, E. Suzuki, and C. Zuker. Independent anchoring and assembly mechanisms of INAD signaling complexes in Drosophila photoreceptors. J. Neurosci. 2001, 21, 150–158. Xu, X. Z., A. Choudhury, X. Li, and C. Montell. Coordination of an array of signaling proteins through homo- and heteromeric interactions between PDZ domains and target proteins. J.Cell Biol. 1998, 142, 545–555. Wes, P. D., X. Z. Xu, H. S. Li, F. Chien, S. K. Doberstein, and C. Montell. Termination of phototransduction requires binding of the NINAC myosin III and the PDZ protein INAD. Nat. Neurosci. 1999, 2, 447–453.

8.7 Concluding Remarks and Outlook 83

84

85

86

87

88

89

90

91

92

93

94

B€ahner, M., P. Sander, R. Paulsen, and A. Huber. The visual G protein of fly photoreceptors interacts with the PDZ domain assembled INAD signaling complex via direct binding of activated Gaq to phospholipase Cb. J. Biol. Chem. 2000, 275, 2901–2904. Adamski, F. M., M. Y. Zhu, F. Bahiraei, and B. H. Shieh. Interaction of eye protein kinase C and INAD in Drosophila. Localization of binding domains and electrophysiological characterization of a loss of association in transgenic flies. J.Biol.Chem. 1998, 273, 17713–17719. Kimple, M. E., D. P. Siderovski, and J. Sondek. Functional relevance of the disulfide-linked complex of the N-terminal PDZ domain of InaD with NorpA. EMBO J. 2001, 20, 4414–4422. Cosens, D. J. and A. Manning. Abnormal electroretinogram from a Drosophila mutant. Nature 1969, 224, 285–287. Minke, B., C. Wu, and W. L. Pak. Induction of photoreceptor voltage noise in the dark in Drosophila mutant. Nature 1975, 258, 84–87. Wong, F., E. L. Schaefer, B. C. Roop, J. N. LaMendola, D. Johnson-Seaton, and D. Shao. Proper function of the Drosophila trp gene product during pupal development is important for normal visual transduction in the adult. Neuron 1989, 3, 81–94. Hochstrate, P. Lanthanum mimicks the trp photoreceptor mutant of Drosophila in the blowfly Calliphora. J.Comp Physiol A 1989, 166, 179–187. Reuss, H., M. H. Mojet, S. Chyb, and R. C. Hardie. In vivo analysis of the Drosophila lightsensitive channels, TRP and TRPL. Neuron 1997, 19, 1249–1259. Leung, H. T., C. Geng, and W. L. Pak. Phenotypes of trpl mutants and interactions between the transient receptor potential (TRP) and TRP-like channels in Drosophila. J. Neurosci. 2000, 20, 6797–6803. B€ahner, M., S. Frechter, N. Da Silva, B. Minke, R. Paulsen, and A. Huber. Light-regulated subcellular translocation of Drosophila TRPL channels induces long-term adaptation and modifies the light-induced current. Neuron 2002, 34, 83–93. Paulsen, R., M. B€ahner, and A. Huber. The PDZ assembled “transducisome” of microvillar photoreceptors: the TRP/TRPL problem. Eur. J. Physiol. 2000, 439, 181–183. Putney, J. W. and G. S. Bird. The signal for capacitative calcium-entry. Cell 1993, 75, 199–201.

95 96

97

98

99

100

101

102

103

104

105

106

Berridge, M. J. Capacitative calcium-entry. Biochemical Journal 1995, 312, 1–11. Hardie, R. C. and B. Minke. Phosphoinositide-mediated phototransduction in Drosophila photoreceptors: the role of Ca2+ and trp. Cell Calcium 1995, 18, 256–274. Minke, B. and Z. Selinger. Role of Drosophila TRP in inositide-mediated Ca2+ entry. Mol.Neurobiol. 1996, 12, 163–180. Hardie, R. C. and P. Raghu. Activation of heterologously expressed Drosophila TRPL channels: Ca2+ is not required and InsP3 is not sufficient. Cell Calcium 1998, 24, 153–163. Ranganathan, R., B. J. Bacskai, R. Y. Tsien, and C. S. Zuker. Cytosolic calcium transients: spatial localization and role in Drosophila photoreceptor cell function. Neuron 1994, 13, 837–848. Acharya, J. K., K. Jalink, R. W. Hardy, V. Hartenstein, and C. S. Zuker. InsP3 receptor is essential for growth and differentiation but not for vision in Drosophila. Neuron 1997, 18, 881–887. Raghu, P., N. J. Colley, R. Webel, T. James, G. Hasan, M. Danin, Z. Selinger, and R. C. Hardie. Normal phototransduction in Drosophila photoreceptors lacking an InsP(3) receptor gene. Mol. Cell Neurosci. 2000, 15, 429–445. Sullivan, K. M. C., K. Scott, C. S. Zuker, and G. M. Rubin. The ryanodine receptor is essential for larval development in Drosophila melanogaster. Proc.Natl.Acad.Sci.U.S.A. 2000, 97, 5942–5947. Huber, A., P. Sander, M. Ba¨hner, and R. Paulsen. The TRP Ca2+ channel assembled in a signaling complex by the PDZ domain protein INAD is phosphorylated through the interaction with protein kinase C (ePKC). FEBS Lett. 1998, 425, 317–322. Huber, A., P. Sander, and R. Paulsen. Phosphorylation of the InaD gene product, a photoreceptor membrane protein required for recovery of visual excitation. J.Biol.Chem. 1996, 271, 11710–11717. Liu, M., L. L. Parker, B. E. Wadzinski, and B. H. Shieh. Reversible phosphorylation of the signal transduction complex in Drosophila photoreceptors. J. Biol. Chem. 2000, 275, 12194–12199. Hardie, R. C., A. Peretz, E. Suss-Toby, A. Rom-Glas, S. A. Bishop, Z. Selinger, and B. Minke. Protein kinase C is required for light adaptation in Drosophila photoreceptors. Nature 1993, 363, 634–637.

203

204

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes 107 Chyb, S., P. Raghu, and R. C. Hardie. Poly-

108

109

110

111

112

113

114

115

116

117

118

unsaturated fatty acids activate the Drosophila light-sensitive channels TRP and TRPL. Nature 1999, 397, 255–259. Hofmann, T., A. G. Obukhov, M. Schaefer, C. Harteneck, T. Gudermann, and G. Schultz. Direct activation of human TRPC6 and TRPC3 channels by diacylglycerol. Nature 1999, 397, 259–263. Estacion, M., W. G. Sinkins, and W. P. Schilling. Regulation of Drosophila transient receptor potential-like (TrpL) channels by phospholipase C-dependent mechanisms. Journal of Physiology-London 2001, 530, 1–19. Raghu, P., K. Usher, S. Jonas, S. Chyb, A. Polyanovsky, and R. C. Hardie. Constitutive activity of the light-sensitive channels TRP and TRPL in the Drosophila diacylglycerol kinase mutant, rdgA. Neuron 2000, 26, 169–179. Hardie, R. C., F. Martin, G. W. Cochrane, M. Juusola, P. Georgiev, and P. Raghu. Molecular basis of amplification in Drosophila phototransduction: Roles for G protein, phospholipase C, and diacylglycerol kinase. Neuron 2002, 36, 689–701. Inoue, H., T. Yoshioka, and Y. Hotta. Diacylglycerol kinase defect in a Drosophila retinal degeneration mutant rdgA. J. Biol. Chem. 1989, 264, 5996–6000. Masai, I., E. Suzuki, C. S. Yoon, A. Kohyama, and Y. Hotta. Immunolocalization of Drosophila eye-specific diacylgylcerol kinase, rdgA, which is essential for the maintenance of the photoreceptor. J. Neurobiol. 1997, 32, 695–706. Hardie, R. C. TRP channels in Drosophila photoreceptors: the lipid connection. Cell Calcium 2003, 33, 385–393. Hardie, R. C., P. Raghu, S. Moore, M. Juusola, R. A. Baines, and S. T. Sweeney. Calcium influx via TRP channels is required to maintain PIP2 levels in Drosophila photoreceptors. Neuron 2001, 30, 149–159. Runnels, L. W., L. X. Yue, and D. E. Clapham. The TRPM7 channel is inactivated by PIP2 hydrolysis. Nature Cell Biology 2002, 4, 329–336. Prescott, E. D. and D. Julius. A modular PIP2 binding site as a determinant of capsaicin receptor sensitivity. Science 2003, 300, 1284–1288. Agam, K., M. von Campenhausen, S. Levy, H. C. Ben Ami, B. Cook, K. Kirschfeld, and B. Minke. Metabolic stress reversibly activates the Drosophila light-sensitive channels TRP

119

120

121

122

123

124

125

126

127

128

129

130

and TRPL in vivo. J. Neurosci. 2000, 20, 5748–5755. Minke, B. and K. Agam. TRP gating is linked to the metabolic state and maintenance of the Drosophila photoreceptor cells. Cell Calcium 2003, 33, 395–408. Harteneck, C., S. N. Kuchta, A. Huber, R. Paulsen, and G. Schultz. The PDZ-scaffold protein INAD abolishes apparent store-dependent regulation of the light-activated cation channel TRP. Faseb J.l 2002, 16, U103–U119. Harteneck, C. Proteins modulating TRP channel function. Cell Calcium 2003, 33, 303–310. Monk, P. D., A. Carne, S. H. Liu, J. W. Ford, J. N. Keen, and J. B. C. Findlay. Isolation, cloning, and characterisation of a trp homologue from squid (Loligo forbesi) photoreceptor membranes. J. Neurochem. 1996, 67, 2227–2235. Nasi, E., M. del Pilar Gomez, and R. Payne. Phototransduction mechanisms in microvillar and ciliary photoreceptors of invertebrates. In Handbook of Biological Physics, Stavenga, D., de Grip, W., and Pugh, E. (eds.), Elsevier, 2000, p. 389–448. Walz, B. and O. Baumann. Structure and cellular physiology of Ca2+ stores in invertebrate photoreceptors. Cell Calcium 1995, 18, 342–351. Chen, F. H., M. Ukhanova, D. Thomas, G. Afshar, S. Tanda, B. A. Battelle, and R. Payne. Molecular cloning of a putative cyclic nucleotide-gated ion channel cDNA from Limulus polyphemus. J. Neurochem. 1999, 72, 461–471. Chen, F. H., A. Baumann, R. Payne, and J. E. Lisman. A cGMP-gated channel subunit in Limulus photoreceptors. Vis. Neurosci. 2001, 18, 517–526. Harteneck, C., T. D. Plant, and G. Schultz. From worm to man: three subfamilies of TRP channels. Trends Neurosci. 2000, 23, 159–166. Clapham, D. E., L. W. Runnels, and C. Strubing. The TRP ion channel family. Nat. Rev. Neurosci. 2001, 2, 387–396. Montell, C. Physiology, phylogeny, and functions of the TRP superfamily of cation channels. Sci.STKE. 2001, www.stke.org/cgi/ content/full/OC_sigtrans;2001/90/re1. Montell, C., L. Birnbaumer, and V. Flockerzi. The TRP channels, a remarkably functional family. Cell 2002, 108, 595–598.

8.7 Concluding Remarks and Outlook 131 Minke, B. and B. Cook. TRP channel proteins

132

133

134

135

136

137

138

139

140

141

142

143

and signal transduction. Physiological Reviews 2002, 82, 429–472. Montell, C. Thermosensation: hot findings make TRPNs very cool. Curr. Biol. 2003, 13, R476–R478. Montell, C. The venerable inveterate invertebrate TRP channels. Cell Calcium 2003, 33, 409–417. Sedgwick, S. G. and S. J. Smerdon. The ankyrin repeat: a diversity of interactions on a common structural framework. Trends Biochem. Sci. 1999, 24, 311–316. Stowers, L., T. E. Holy, M. Meister, C. Dulac, and G. Koentges. Loss of sex discrimination and male-male aggression in mice deficient for TRP2. Science 2002, 295, 1493–1500. Leypold, B. G., C. R. Yu, T. Leinders-Zufall, M. M. Kim, F. Zufall, and R. Axel. Altered sexual and social behaviors in trp2 mutant mice. Proc. Natl. Acad. Sci. U.S.A 2002, 99, 6376–6381. Liman, E. R., D. P. Corey, and C. Dulac. TRP2: a candidate transduction channel for mammalian pheromone sensory signaling. Proc. Natl. Acad. Sci. U.S.A 1999, 96, 5791–5796. Caterina, M. J., M. A. Schumacher, M. Tominaga, T. A. Rosen, J. D. Levine, and D. Julius. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 1997, 389, 816–824. Caterina, M. J., T. A. Rosen, M. Tominaga, A. J. Brake, and D. Julius. A capsaicin-receptor homologue with a high threshold for noxious heat. Nature 1999, 398, 436–441. Peier, A. M., A. J. Reeve, D. A. Andersson, A. Moqrich, T. J. Earley, A. C. Hergarden, G. M. Story, S. Colley et al. A heat-sensitive TRP channel expressed in keratinocytes. Science 2002, 296, 2046–2049. Xu, H., I. S. Ramsey, S. A. Kotecha, M. M. Moran, J. A. Chong, D. Lawson, P. Ge, J. Lilly et al. TRPV3 is a calcium-permeable temperature-sensitive cation channel. Nature 2002, 418, 181–186. Smith, G. D., M. J. Gunthorpe, R. E. Kelsell, P. D. Hayes, P. Reilly, P. Facer, J. E. Wright, J. C. Jerman et al. TRPV3 is a temperature-sensitive vanilloid receptor-like protein. Nature 2002, 418, 186–190. G€ uler, A. D., H. S. Lee, T. Iida, I. Shimizu, M. Tominaga, and M. Caterina. Heat-evoked activation of the ion channel, TRPV4. J. Neurosci. 2002, 22, 6408–6414.

144 Watanabe, H., J. Vriens, S. H. Suh, C. D.

145

146

147

148

149

150

151

152

153

154

Benham, G. Droogmans, and B. Nilius. Heatevoked activation of TRPV4 channels in a HEK293 cell expression system and in native mouse aorta endothelial cells. J. Biol. Chem. 2002, 277, 47044–47051. Kim, J., Y. D. Chung, D. Y. Park, S. Choi, D. W. Shin, H. Soh, H. W. Lee, W. Son et al. A TRPV family ion channel required for hearing in Drosophila. Nature 2003, 424, 81–84. Tobin, D., D. Madsen, A. Kahn-Kirby, E. Peckol, G. Moulder, R. Barstead, A. Maricq, and C. Bargmann. Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans neurons. Neuron 2002, 35, 307–318. Zhang, Y., M. A. Hoon, J. Chandrashekar, K. L. Mueller, B. Cook, D. Wu, C. S. Zuker, and N. J. Ryba. Coding of sweet, bitter, and umami tastes: different receptor cells sharing similar signaling pathways. Cell 2003, 112, 293–301. Perez, C. A., L. Huang, M. Rong, J. A. Kozak, A. K. Preuss, H. Zhang, M. Max, and R. F. Margolskee. A transient receptor potential channel expressed in taste receptor cells. Nat. Neurosci. 2002, 5, 1169–1176. McKemy, D. D., W. M. Neuhausser, and D. Julius. Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature 2002, 416, 52–58. Peier, A. M., A. Moqrich, A. C. Hergarden, A. J. Reeve, D. A. Andersson, G. M. Story, T. J. Earley, I. Dragoni et al. A TRP channel that senses cold stimuli and menthol. Cell 2002, 108, 705–715. Walker, R. G., A. T. Willingham, and C. S. Zuker. A Drosophila mechanosensory transduction channel. Science 2000, 287, 2229–2234. Sidi, S., R. W. Friedrich, and T. Nicolson. NompC TRP channel required for vertebrate sensory hair cell mechanotransduction. Science 2003, 301, 96–99. Tracey, W. D., Jr., R. I. Wilson, G. Laurent, and S. Benzer. painless, a Drosophila gene essential for nociception. Cell 2003, 113, 261–273. Story, G. M., A. M. Peier, A. J. Reeve, S. R. Eid, J. Mosbacher, T. R. Hricik, T. J. Earley, A. C. Hergarden et al. ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 2003, 112, 819–829.

205

206

8 Invertebrate Phototransduction: Multimolecular Signaling Complexes 155 Viswanath, V., G. M. Story, A. M. Peier, M. J.

157 Lee, S. J., H. Xu, L. W. Kang, L. M. Amzel, and

Petrus, V. M. Lee, S. W. Hwang, A. Patapoutian, and T. Jegla. Opposite thermosensor in fruitfly and mouse. Nature 2003, 423, 822–823. 156 Kosloff, M., N. Elia, T. Joel-Almagor, R. Timberg, T. D. Zars, D. R. Hyde, B. Minke, and Z. Selinger. Regulation of light-dependent Gq alpha translocation and morphological changes in fly photoreceptors. EMBO J. 2003, 22, 459–468.

C. Montell. Light adaptation through phosphorinositide-regulated translocation of Drosophila visual arrestin. Neuron 2003, 39, 121–132. 158 Bentrop, J. and R. Paulsen. Light-modulated ADP-ribosylation, protein phosphorylation and protein binding in isolated fly photoreceptor membranes. Eur. J. Biochem. 1986, 161, 61–67.

207

9

The Transduction Channels of Rod and Cone Photoreceptors U.B. Kaupp and D. Tra¨nkner

9.1

Introduction

The transduction channels of rod and cone photoreceptors are ion channels gated directly by cGMP. They belong to the family of cyclic nucleotide-gated (CNG) channels. CNG channels are involved in both hyperpolarizing and depolarizing light responses. The sign of the light response is determined by the signaling pathway that controls the cGMP concentration (increase or decrease by light) and by the ion selectivity of the channel (K+ selective versus nonselective). In most vertebrate photoreceptors, closing of nonselective cation CNG channels evokes a hyperpolarizing light response. The opening of similar cation channels in the parietal eye of some lizards evokes a depolarizing light response. Finally, in ciliary photoreceptors of molluscan eyes, the opening of K+-selective CNG channels produces a hyperpolarizing light response. Excellent reviews on phototransduction and light adaptation can be found in [1] and [2]. The function and properties of CNG channels have been the subject of several comprehensive reviews [3–9]. Here we will focus on recent advances that have furthered our understanding of the physiological function of CNG channels in photoreceptors – both vertebrate and invertebrate.

9.2

Brief Overview 9.2.1

Ligand Sensitivity

Cyclic nucleotides directly activate CNG channels by binding to a site on the channel protein. The dependence of channel activation on the ligand concentration is steep, indicating that several, most probably four, molecules of the ligand are required to fully open the channels. All CNG channels respond to some extent to both cAMP and Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

208

9 The Transduction Channels of Rod and Cone Photoreceptors

cGMP. In rods and cones, CNG channels sharply discriminate between cAMP and cGMP, whereas channels in chemosensitive cilia of olfactory sensory neurons (OSN) by and large respond equally well to both ligands (see Chapter 5). The ability to discriminate between ligands is commonly referred to as ligand selectivity. Selectivity can be achieved by differential control of either ligand affinity or efficacy or a combination of both. Ligand affinity is a measure of how tightly cyclic nucleotides bind to the channel. Efficacy refers to the ability to open the channel once the ligand has been seated in the binding cavity. The molecular basis of ligand affinity, efficacy, and selectivity is discussed in several reviews [9–11].

9.2.2

Ion selectivity

CNG channels of rods and cones are nonselective cation channels that poorly discriminate between alkali ions and even allow the passage of divalent cations, in particular Ca2+ ions (for a review, see [9]). Therefore, at rest (–60 mV) and when bathed in a physiologic ion milieu, CNG channels conduct mixed inward currents carried by Na+ and Ca2+ ions. In order to permeate, the ions bind to a site inside the channel pore. The dwell time at this binding site is significantly longer for Ca2+ than for monovalent cations. As an important result, Ca2+ blocks the current of the more permeant Na+ ions. The pronounced Ca2+ permeability and the concomitant blockage of Na+ current by divalent cations are crucially important for the channel’s function and underlie, for example, the ability of rod photoreceptors to detect single photons and to adapt to steady illumination. The interaction of Ca2+ with the channel pore and the consequences for the physiology of cells are discussed in [9].

9.2.3

Modulation

Unlike ligand-gated neurotransmitter receptors, CNG channels do not desensitize in the continuous presence of the ligand. The cooperative and sustained activation predestines CNG channels to serve as a molecular switch that faithfully tracks the cGMP concentration in a cell. While the channels do not desensitize, their activity is nonetheless modulated, notably by the Ca2+-binding protein calmodulin and by phosphorylation. The potential mechanisms of these modulations and their physiologic significance for light adaptation are discussed.

9.3 Function of CNG Channels in Phototransduction and Adaptation

9.3

Function of CNG Channels in Phototransduction and Adaptation 9.3.1

Rod and Cone Photoreceptors

Rod and cone photoreceptors respond to a light stimulus with a brief hyperpolarization by closing CNG channels in the surface membrane of the outer segment (for review, see [3]). Estimates of CNG channel density in rods from various species range between ca. 100 and 400 lm–2 [12–18]. The density of the CNG channel in cones is about 10-fold lower [19]. Considering that the surface membrane area of the cone outer segment is approximately 10-fold larger than that of the rod outer segment, the total numbers of CNG channels seem comparable. In the dark, channels are activated by the binding of cGMP, allowing a steady cation current (“dark current”) to flow into the outer segment. Light triggers a sequence of enzymatic reactions that lead to the hydrolysis of cGMP. When CNG channels close, the inward current ceases and the cell hyperpolarizes. The enzyme cascade comprises the photo pigment rhodopsin (R), the G protein transducin (T), and a phosphodiesterase (PDE). Light stimulation decreases the cytoplasmic Ca2+ concentration, [Ca2+]i, which (1) initiates the recovery from the light response by enhancing the synthesis of new cGMP molecules and (2) adjusts the sensitivity of the transduction machinery, a process known as light adaptation (for review, see [1, 2]). The CNG channel is crucially important for the control of [Ca2+]i because it provides the only source for Ca2+ influx into the outer segment. In rods, between 10 % and 18 % of the dark current (30 pA) is carried by Ca2+ [20–24]. Ca2+ entry through open CNG channels is balanced by Ca2+ extrusion through a Na+/Ca2+-K+ exchange mechanism (reviewed in [25, 26]; Fig. 9.1). In light, when CNG channels close, but the exchanger continues to clear Ca2+ from the cytosol, the balance is disturbed between Ca2+ entry and Ca2+ extrusion. The resulting decline in [Ca2+]i provides a negative feedback mechanism that controls at least three biochemical processes. First, the activity of the guanylyl cyclase (GC) that synthesizes cGMP is stimulated as Ca2+ levels decrease. The Ca2+ sensitivity of the GC is relayed by two small Ca2+-binding proteins, designated guanylyl cyclase-activating proteins (GCAP1 and GCAP2). At rest, when [Ca2+]i is 300–500 nM, the GCAPs prevail in the inactive form with Ca2+ ions bound. In light, when [Ca2+]i is lowered to 50–100 nM, Ca2+ dissociates from GCAPs; the Ca2+-free form then stimulates GC activity (for review, see [27, 28]). Second, the lifetime of active PDE is shortened through the phosphorylation of light-activated rhodopsin (R*) by the rhodopsin kinase (RhK). This reaction is mediated by another small Ca2+-binding protein, recoverin (for review, see [27, 29]). Finally, the ligand sensitivity of the CNG channel increases as [Ca2+]i decreases. The regulation of ligand sensitivity by Ca2+ is mediated by a third Ca2+-dependent protein, calmodulin (CaM) [30, 31]. All three reactions by various degrees help to restore the dark state and to adjust the light sensitivity of the cell [32, 33]; reviewed in [1]. A similar transduction scheme exists in cones, the photoreceptors responsible for vision in bright light. Fundamentally the same events underlie phototransduction in rods and cones, and the two photoreceptor types utilize similar protein isoforms of the

209

210

9 The Transduction Channels of Rod and Cone Photoreceptors

Fig. 9.1 Ca2+ feedback mechanism in rod and cone photoreceptors involving cyclic nucleotide-gated channels. PDE, phosphodiesterase; GCAP, guanylyl cyclase-activating protein

enzyme cascade. However, the light sensitivity of cones is 30- to 100-fold lower than that of rods, and cones adapt over a wider range of light intensities than do rods (reviewed in [1]). It has been suggested that differences in the Ca2+ homeostasis underlie the distinct light sensitivity and adaptation range of the two photoreceptor types. Important elements that control the dynamics and size of the changes in [Ca2+]i are cell volume, the rate of Ca2+ clearance by the Na+/Ca2+-K+ exchanger, the Ca2+-buffering capacity of the cytoplasm, and Ca2+ entry through CNG channels (see [34] for a thorough discussion). Several observations demonstrate that the CNG channels in rods and cones differ in ion permeation, ligand sensitivity, and modulation by Ca2+ ions. The relative ion permeability PCa/PNa of CNG channels is more than three times larger in cones than in rods (21.7 and 6.5, respectively [35, 36]), and, under physiologic ionic conditions, the fraction of the dark current carried by Ca2+ is about twofold larger in cones than in rods [24, 37]. The Na+/Ca2+-K+-exchange current in cones is roughly one order of magnitude faster than that in rods [20, 37–40]; for review, see [34]. From these observations it has been inferred that the light-stimulated changes in [Ca2+]i are far larger and faster in cones than in rods. The cGMP sensitivity of the CNG channel and its modulation by Ca2+ are also different in intact outer segments of rods and cones. At elevated [Ca2+]i (i.e., in the dark state), the K1/2 for cGMP can be as large as 550 lM in cones (mean K1/2 = 335.5 lM; [41]) compared to rods (37.8–40 lM; [42, 43]). In truncated or electropermeabilized rods, the Ca2+-dependent modulation of the ligand sensitivity is only 1.5- to 2-fold,

9.3 Function of CNG Channels in Phototransduction and Adaptation

similar to the effect of Ca2+/CaM on detached membrane patches [41–43]. In contrast, the range of the CNG channel modulation in intact cones is much broader than in rods and is not well mimicked by Ca2+/CaM in detached patches from the outer segment [44]. This has led to the hypothesis that an unknown factor, which is lost upon patch excision, is responsible for the larger range of modulation of the ligand sensitivity in cones [41]. In summary, the cGMP sensitivity, its modulation by [Ca2+]i, and the Ca2+ permeation are profoundly different in CNG channels of rods and cones, supporting the notion that the CNG channel is a pivotal determinant of the dynamics of Ca2+ homeostasis in vertebrate photoreceptor cells. CNG channels in cones serve a second function that is absent in rods. Light produces a graded hyperpolarization in rods and cones that is up to 35 mV in amplitude. Not all of this response range is effectively transmitted to the postsynaptic bipolar and horizontal cells. The highly nonlinear input-output relation of the rod synapse is largely accounted for by the voltage dependence of presynaptic Ca2+ channels. At the dark resting voltage of –35 mV, a fraction of the Ca2+ channel is open, and the continuous Ca2+ entry sustains a tonic release of the neurotransmitter glutamate from the synaptic terminal. The Ca2+ channels are characterized by an activation threshold of ca. –45 mV [45]. Therefore, when a rod is hyperpolarized to values more negative than ca. –45 mV, the Ca2+ channels close and synaptic transmission ceases [46, 47]. In contrast to rods, synaptic transmission in cone photoreceptors continues as the light-induced voltage response grows to –70 mV [48–50]. Whereas the small overlap of the voltage range of Ca2+ channel activation and the voltage range produced by light can explain signal clipping at the rod synapse, it fails to explain the broader voltage range over which synaptic transmission operates in cones. This conundrum has been partially solved by the discovery of CNG channels in the inner segment and synaptic terminal of cones [51, 52]. The density of CNG channels in the inner segment is low, whereas in the cone terminal these channels appear to come in clusters [52]. If the clusters were located near release sites, CNG channels would be ideally suited to control the Ca2+-dependent release of glutamate. In fact, experimental maneuvers that activate CNG channels also trigger exocytotic events and release glutamate from the cone terminal [51, 52]. The cGMP sensitivities measured in patches of membrane excised from either the outer segment or the axon terminal are indistinguishable, suggesting that CNG channels from both locales are built from identical or similar subunits. The cGMP sensitivity of the CNG channels in the synapse is as unusually low as that of CNG channels in the cone outer segment of the fish retina (K1/2 = 206 lM and 335.5 lM, respectively [41, 52]). We note, however, that the high K1/2 value in fish cones required an intact cone photoreceptor, whereas the K1/2 of synaptic channels was determined in excised patches. CNG channels could serve two different functions in the cone synapse. First, these channels might extend the voltage range over which synaptic transmission operates by providing a sustained Ca2+ influx even at very negative voltages. Second, nitric oxide (NO) is a good candidate to serve as a retrograde neurotransmitter that is released onto cone terminals from other retinal cells [52]. An NO synthase (NOS) is found predominantly in the inner segment of rods and cones and in processes of bipolar cells in the outer plexiform layer of the retina [53–55]. Furthermore, a soluble form of guanylate

211

212

9 The Transduction Channels of Rod and Cone Photoreceptors

cyclase (sGC) is found in the inner segment of cones and is stimulated by NO [53]. Thus, CNG channels may play an important role in the modulation of synaptic transmission by NO in the axon terminals of cones. CNG channels could serve—at least in some vertebrates—a third function. In the chick, the cGMP sensitivity of the cone CNG channel is under the control of a circadian rhythm [56–58]. During the subjective night, the sensitivity is approximately twofold higher than during the subjective day. Circadian modulation of ligand sensitivity is driven, at least in part, by rhythms in the activities of two protein kinases: the Erk form of mitogen-activated protein kinase and the Ca2+/CaM-dependent protein kinase II (CaMKII). Perturbation of these signaling pathways causes phase-dependent changes in the cGMP sensitivity of the cone CNG channel. The mechanism of the sensitivity regulation has not been determined. It may involve phosphorylation of A or B subunits by Erk or CaMKII or phosphorylation of other signaling molecules that interact with the channels and lead to modulation of their sensitivity.

9.3.2

CNG Channels in Pinealocyte Photoreceptors

The pineal gland regulates various physiological functions by nocturnal secretion of the hormone melatonin. Light sensitivity of the pineal gland has been retained in most vertebrates, except mammals. Pinealocytes, the light-sensitive cells, display hyperpolarizing responses to brief pulses of light [59, 60] and express several retinal proteins including arrestin, recoverin, rhodopsin kinase, phosducin, GC, and a cGMP-specific PDE (for review, see [61, 62]). Dryer and Henderson [63] recorded CNG channel activity from excised inside-out patches of dissociated photoreceptors from the chick pineal gland. These CNG channels in extra-retinal photoreceptors feature all the hallmarks of CNG channels in retinal photoreceptors [63, 64]. Activation is half-maximal between 10 and 50 lM cGMP. Even fully activated channels display frequent brief transitions to the closed state. For this reason, the open probability Po becomes less than unity at saturating cGMP concentrations. The brief closing events are more frequent at negative than at positive membrane potentials. Similar properties have been reported for CNG channels from retinal photoreceptors [19, 65, 66]. Moreover, expression of several CNG channel subunits in the pineal gland has been confirmed by in situ hybridization and immunohistochemistry. These results collectively show that the light response in pinealocytes of lower vertebrates is produced by activation of a cGMP-signaling pathway, which leads to the closure of cGMP-selective ion channels. Chick pineal cells display a circadian rhythm in cGMP concentration [67, 68]. It is therefore conceivable that CNG channels are involved in regulating the output of the intrinsic circadian oscillator.

9.3 Function of CNG Channels in Phototransduction and Adaptation

9.3.3

CNG Channels in Parietal Eye Photoreceptors

Some lizards do have a parietal eye, or third eye, on top of their head. The parietal eye seems likely to convey information about changes in light intensity and spectral composition during dusk and dawn. The parietal-eye photoreceptors morphologically resemble rod and cones of the vertebrate retina, yet they depolarize in response to a flash of light [69]. This suggested that parietal-eye photoreceptors, like rhabdomeric photoreceptors of the invertebrate eye, might utilize a phosphoinositide-signaling cascade rather than the cGMP-signaling pathway of retinal rods and cones. It came as a surprise when Finn and his collaborators [70] convincingly demonstrated that the outer segment membrane of the parietal-eye photoreceptors harbors a high density of CNG channels with all the hallmarks of CNG channels from rods and cones: the channels are selectively activated by cGMP; cAMP is much less effective; the channels are nonselective among monovalent cations and are permeable to Ca2+ ions; channels are blocked by L-cis-diltiazem; and Ca2+/CaM reduces the cGMP-activated current by reducing the ligand sensitivity. Which type of CNG channel is expressed in the parietal eye? CNG channels of retinal cones are significantly more Ca2+ permeable than those of rods [35, 71–73]. The relative selectivity for Ca2+ over alkali cations has been determined from reversal potentials, Vrev, under well-defined ionic conditions in excised patches from rod and cone [35] and parietal-eye [70] photoreceptors. In cones of striped bass, PCa/PNa = 21.7; in rods of tiger salamander, PCa/PNa = 5.9 [73]; and in the parietal eye, PCa/PNa = 8.1–10.3 [70]. Thus, at least with respect to Ca2+ permeability, the CNG channel in parietal-eye

Fig. 9.2 Summary of different transduction schemes in various types of photoreceptors involving CNG channels (except rhabdomeric photoreceptors of invertebrate eyes, which use Trp channels)

213

214

9 The Transduction Channels of Rod and Cone Photoreceptors

photoreceptors behaves more like the CNG channel of rods than that of cones. However, permeability ratios of native CNG channels are not invariant but depend on the cGMP concentrations [73]. When comparing relative ion permeabilities, this complication must be kept in mind. The depolarizing light response is produced by an increase in the cytosolic cGMP concentration that is controlled by an unusual cGMP-signaling pathway [74] (Fig. 9.2). In the dark, cGMP is synthesized continuously by GC activity and is rapidly degraded by PDE activity. The elevated PDE activity in the dark seems to rest on a constitutively active G protein, whereas the mechanism that keeps GC active in the dark is not known. Light acts by inhibiting the PDE through another G protein, permitting the cGMP concentration to rise and CNG channels to open.

9.3.4

CNG Channels in Hyperpolarizing Photoreceptors of Invertebrates

The retina of some molluscan eyes is composed of two layers of photoreceptors: depolarizing rhabdomeric-type cells, similar to those found in most other invertebrates, and ciliary photoreceptors that hyperpolarize in light [75] (Fig. 9.2). The mechanism underlying the hyperpolarizing light responses has been studied in two scallop genera, Pecten and Lima. Light stimulation under voltage-clamp activates an outward current that is accompanied by a decrease in the cellular input resistance. The Vrev of the lightstimulated current lies near EK [76], demonstrating that the light-dependent channel is highly K+ selective and that the hyperpolarizing light response is brought about by opening K+ channels rather than by closing nonselective cation channels, as in retinal rods and cones. In a series of incisive experiments, Gomez and Nasi [77] convincingly demonstrated that (1) the IP3/Ca2+-signaling pathway is not crucial for phototransduction; (2) the photoreceptors rely on cGMP as the internal messenger of the transduction cascade; and (3) the light-dependent channel is opened by cGMP. The latter two observations imply that light elevates cGMP, although it is unknown whether this involves the inhibition of a PDE or the stimulation of a GC. In contrast to the Ca2+-permeable CNG channels of retinal photoreceptors and OSNs, the Pecten channel is virtually impermeable to Ca2+ ions, and the K1/2 values for blockage by extracellular Ca2+ and Mg2+ are one to two orders of magnitude higher [78]. The significant K+ selectivity, the lack of Ca2+ permeability, and the weak divalent block suggest that the pore architecture is more like that of K+ channels than that of CNG channels. It is interesting to note that a cGMP-sensitive K+ channel also seems to underlie the light response of a photosensitive neuron in the abdominal ganglion of a marine mollusk [79]. This cell generates slow, depolarizing light responses due to the closure of K+ channels that are kept open in the dark by cGMP. The protein(s) forming the cGMP-dependent K+ channel is unknown. Its molecular identification is eagerly awaited, as it will certainly further our understanding of the molecular mechanisms that govern ion selectivity in CNG channels.

9.5 Transmembrane Topology and Subunit Stoichiometry

9.4

Structure of Subunits

The rod CNG channel consists of two different subunits: A1 and B1a. The A1 subunit is a 63-kD protein, whereas the B1 subunit is much larger (240 kD) and exhibits a bipartite structure [80, 81]. The subunits of the cone CNG channels are designated A3 and B3 [82] (see also [9] for discussion of nomenclature and subunit isoforms). The transmembrane topology and functional domains of CNG channel subunits are depicted in Fig. 9.3. The core unit consists of six membrane-spanning segments. A pore region of 20–30 amino acids is located between S5 and S6. The S4 segment resembles the voltage-sensor motif found in the S4 segment of voltage-gated K+-, Na+-, and Ca2+-channels, but it probably serves a different function because CNG channels cannot be activated by a voltage step in the absence of the ligand. The C-terminal region harbors the cNMP-binding domain comprising 80–100 amino acid residues. This domain displays characteristic sequence similarity with the cNMP-binding domain of protein kinases A and G, the catabolite-activating protein CAP, a subfamily of guanine-exchange factors (GEF), and hyperpolarization-activated and cyclic nucleotide-gated (HCN) channels. The region located between S6 and the cNMP-binding domain has been designated C-linker. It relays the changes in conformation upon binding of the ligand to the gate of the channel (for review, see [83]). Sequence comparisons of the S4 motif, the pore region, and the cNMP-binding domain between CNG channels and other proteins can be found in the review of Kaupp and Seifert [9]. The overall topology of A and B subunits is similar and they share all functional domains – with one exception. The B1 subunit of rods is significantly larger than A1, A3, and B3 due to a unique N-terminal domain, which has been designated GARP (glutamic acid-rich protein) because it is rich in glutamic acid residues. The cone B3 subunit lacks the GARP part, suggesting that it serves a rod-specific function (see section 9.6.1). The A1 and A3 subunits undergo posttranslational proteolytic processing of their Nterminus [84–86]. In bovine A1, a segment of 92 amino acid residues is removed [84]. The role of this proteolytic processing is not known. It may be required for targeting of the channel to the plasma membrane of the outer segment or for interaction with other photoreceptor proteins. The B1 subunit does not undergo proteolytic processing [81]. The A1 subunit of bovine rods is glycosylated at residue N327 [87]. The glycosylation pattern of A3 has not been studied. The B1 subunit does not seem to be glycosylated [81].

9.5

Transmembrane Topology and Subunit Stoichiometry

The current model for the membrane topology of the A and B subunits of CNG channels is illustrated in Fig. 9.4. The core structural unit consists of six membrane-spanning segments, designated S1–S6, followed by a nucleoside 3’:5’ cyclic monophosphate (cNMP) binding domain near the C-terminus. A pore region of about

215

216

9 The Transduction Channels of Rod and Cone Photoreceptors

Fig. 9.3 Organization of functional domains. Transmembrane segments are indicated by numbers 1–6. CaM, calmodulin-binding site; cNMP, binding site for cyclic nucleotides; P, pore region; LZ, leucine zipper motif; Glu, glutamic acid-rich part; R1–R4, protein-protein interaction domains in GARPs. Lower left: similarity dot matrix between the B1a subunit of chick and bovine. Lower right: Sequence alignment of the four repeats R1–R4

20–30 amino acids is located between S5 and S6. The S4 segment in CNG channels resembles the voltage-sensor motif found in the S4 segment of voltage-gated K+-, Na+-, and Ca2+ channels. Since the voltage-sensor motif, the six transmembrane segments, and the pore region are also characteristic features of voltage-gated channels, it has been suggested that CNG channels and voltage-gated channels are members of a

9.5 Transmembrane Topology and Subunit Stoichiometry

gene super family of cation channels, which evolved from a common primordial channel [88, 89]. Experimental evidence in support of the topological model was first obtained from immunogold labeling for electron microscopy of the CNG channel from rod photoreceptors. In these studies, both the N- and C-termini were localized to the cytoplasmic side of the rod outer segment (ROS) plasma membrane [84], and a glycosylated segment connecting S5 to the pore region was localized to the extracellular side [87]. The topological model for this and other A subunits received further support from a genefusion approach using enzyme reporters [90]. Although similar studies are lacking for the B subunits, because of the significant sequence similarity and similar structural features, A and B subunits most likely share a common transmembrane topology. In a similar vein, it has been taken for granted that rod and cone CNG channels form tetramers like their K+ channel cousins. In fact, three different experimental strategies revealed that the rod CNG channel exists as tetramer comprising three A1 and one B1 subunit [9, 91–93] (see also [94]). The asymmetrical arrangement of subunits came as a surprise, as Shaker-type K+ channels form symmetrical complexes [95]. What kind of

Fig. 9.4 Model of the transmembrane topology of A1 and B1 subunits for rod CNG channels

217

218

9 The Transduction Channels of Rod and Cone Photoreceptors

intersubunit interactions direct the assembly of three A subunits and one B subunit? The A1 subunit carries a leucine zipper motif C-terminal of the cNMP-binding domain that is lacking in the B1 subunit [93]. This leucine zipper motif probably represents an assembly domain that directs the preferential formation of A-B heteromers [96]. Zhong and his coworkers [96] proposed a two-step model for the assembly process. In a first step, homomeric trimers of A subunits are formed, followed by the incorporation of one B subunit into the A trimer. While the leucine zipper domain appears to be important for the oligomerization, it is by no means the only interaction site. Mutants lacking this motif form functional channels, although at much lower efficacy in HEK293 cells [96]. Moreover, mutants lacking the C-terminal part immediately after the cNMP-binding site readily form homomeric channels when expressed in Xenopus oocytes [97, 98]. The leucine zipper motif is also observed in the A3 subunit; thus, the cone CNG channel may also comprise three A3 subunits and one B3 subunit, although this suggestion has not yet been experimentally tested (Fig. 9.5). The trimer motif also raises interesting questions as to the oligomeric state of the A1 or A3 homomeric channels. Cross-linking studies and sedimentation equilibrium analysis show that soluble peptides or tagged fusion proteins carrying the leucine zipper motif in fact form trimers [93]. However, are the functional A1 or A3 homomeric channels of heterologous studies trimers? Mutants in which the original leucine zipper motif of the channel has been replaced by leucine zippers known to form dimers, trimers, or tetramers form functional homomeric channels with largely similar properties [93]. This result indicates that homomeric A channels adopt only a single oligomeric state irrespective of the nature of the leucine zipper. Electrophysiological studies with mutants that “report” the number of subunits argue that A-only channels most likely form tetramers [99].

Fig. 9.5 Top: Subunit stoichiometry of CNG channels from rods and cones. Bottom: Amino acid residues forming the intrapore ion-binding site in homomeric and native CNG channels

9.6 Interaction of CNG Channels With Other Proteins

9.6

Interaction of CNG Channels With Other Proteins 9.6.1

The Glutamic Acid-rich Part (GARP) of B1

The B1 subunit of the CNG channel of rod photoreceptors features a unique bipartite structure [81]. The membrane-spanning, so-called b’ part is homologous to A subunits, whereas the large cytoplasmic N-terminal domain is lacking in all other CNG channel subunits. Even more intriguing, except for a few amino acid residues, this N-terminal domain is identical to two soluble glutamic acid-rich proteins (GARPs) [100–102] that seem to be specifically expressed in rod photoreceptors. We refer to the soluble proteins as GARP1 and GARP2 and to the respective channel domain as GARP part [102]. GARP1 is twice as large as GARP2, and therefore these two proteins have also been designated full GARP and truncated GARP (f-GARP and t-GARP [101]). The soluble GARP1/GARP2 probably represent alternatively spliced forms [103, 104]. Splice variants of the rod B1 lacking the GARP part are found in OSNs [105, 106] and testes [107]. The B3 subunit expressed in cone photoreceptors is much shorter than B1 [108] and has no GARP-related sequences, suggesting that GARPs serve a rod-specific function. In fact, specific anti-GARP antibodies stain the outer segment of rods but not of cones [102]. The common N-terminal region of GARPs carries four short proline-rich repeats R1 to R4 (15 aa) that are highly conserved among each other and that represent the most conserved structural elements between GARPs from different species (see dot blot in Fig. 9.3). Peptide affinity chromatography suggests that these repeats are involved in proteinprotein interaction between GARP and the PDE, the GC, and the retina-specific ATPbinding cassette (ABCR) transporter [102]. The ABCR transporter [109, 110] probably translocates all-trans retinal complexed with phosphatidylethanolamine as a Schiff’s base [111]. The ABCR is better known as a “rim” protein because it is confined to the rim of the disc membrane [110, 112]. Reportedly, the GC is also located near the disc margin, although the electron microscopic evidence is less compelling than for ABCR [113]. The soluble GARPs are tightly associated with the margin of the disc, probably by binding to one or several of the rim-confined proteins. The interaction between GARPs and ABCR and GC has recently been called into question [114]. Using immunoprecipitation techniques, Poetsch and his collaborators showed that both ABCR and GC cannot be pulled down with GARP-specific antibodies, regardless of whether the potential binding partners have been covalently attached to each other by cross-linking reagents. Instead, peripherin, another protein located at the disc rim, seems to interact with GARP2 and the GARP’ part of the B1a channel subunit. What might be the function of the GARP part and soluble GARPs? The tethering of the GARP part to peripherin at the disc rim enforces a circular arrangement of channels in juxtaposition to the disc rim, and the rings are stacked along the length of the outer segment (Fig. 9.6). The repeat distance of stacked rings is equal to the repeat distance of discs (3 nm). Such a non-uniform distribution might account for the high

219

220

9 The Transduction Channels of Rod and Cone Photoreceptors

Fig. 9.6 Interaction of the rod CNG channel with other proteins. Left: Interaction with peripherin and the Na+/Ca2+-K+ exchanger. Right: Predicted

ringlike distribution of CNG channels in the outer segment membrane of rods

variability of channel density in the rod outer segments determined by patch-clamp recording [13], although the authors do not favor this explanation. In order to reach for peripherin, the GARP part must adopt an elongated structure to traverse the 10-nm gap between membrane and disc rim. The high density of negatively charged Glu residues of GARP could form a charged “wire” that channels Ca2+ ions from the mouth of the channel pore to Ca2+-dependent proteins on the disc surface, e.g., recoverin and GCAPs (Fig. 9.6). Cone outer segments form a disc-like structure by invaginations of the plasma membrane. As a consequence, Ca2+ enters the cell through the entire surface of the disc-like invaginations – provided that the CNG channels are uniformly distributed in the cone plasma membrane. Thus, there is no need to “channel” Ca2+ ions from the cell surface to the center of discs. GARP2, the most abundant GARP species, has been proposed to interact with PDE and thereby inhibit its activity [102]. In addition, GARP2 could serve as a cap for those peripherin molecules that are not engaged in binding of the B subunit [114].

9.6.2

Interaction with the Na+/Ca2+-K+ Exchanger

The CNG channel of rods also undergoes an interaction with the Na+/Ca2+-K+ exchanger that involves the A1 rather than the B1 subunit [114–117] (Fig. 9.6). The molar ratio of exchanger to channel is at least two [115, 118], implying that two or more exchanger

9.7 Modulation

molecules are bound per channel. In cone photoreceptors, the exchanger also interacts with the CNG channel [119]. The interaction is not isoform-specific: the rod and cone exchangers can interact with each other, and the rod exchanger can interact with cone A3 and vice versa [119]. The juxtaposition of the channel, which promotes Ca2+ influx, and the Na+/Ca2+-K+ exchanger, which promotes Ca2+ efflux, suggest that Ca2+ dynamics inside the cell might be localized to microdomains in the vicinity of the channel. For comprehensive reviews on the dynamics of [Ca2+] in microdomains, see [120–123].

9.7

Modulation

CNG channels of rods and cones do not desensitize in the continued presence of the ligand. However, their activity seems to be modulated by Ca2+, phosphorylation, and retinal. Although the modulatory effects are robust and well established in vitro, their physiological significance in the intact photoreceptor is not entirely clear.

9.7.1

Modulation by Ca2+

The ligand sensitivity of rod and cone CNG channels is modulated by binding of CaM or by as yet unidentified Ca2+-binding proteins. The native CNG channel of rod photoreceptors is exquisitely sensitive to regulation by Ca2+/CaM (K1/2 (CaM) of 1–2 nM; [30, 124, 125]), yet the decrease of cGMP sensitivity, by comparison with the olfactory channel, is modest (maximally twofold increase of K1/2 for cGMP activation [30, 43, 124–126]). The Ca2+ dependence of the modulation of the native channel (K1/2 (Ca2+) = 48 nM [43]) and the binding of CaM to the target peptide (K1/2 (Ca2+) = 117 nM [125]) are similar and well within the range of Ca2+ concentrations during the light response [20, 22, 127, 128]. In contrast to the olfactory A2 subunit (see Chapter 5), the A1 subunit of rod CNG channels does not bind CaM, and homomeric channels composed of A1 subunits are not modulated by CaM [30, 81, 129]. CaM sensitivity is conferred to heteromeric rod channels by an unconventional CaM-binding site in the N-terminal region of the B1a subunit [125, 130, 131]. The function of a second CaM-binding site in the C-terminal region is not known [125, 130]. While CaM modulation of the rod CNG channel is well established in vitro, its significance as an adaptive mechanism has been questioned on various grounds [22, 33, 42, 43] (for review, see [31]). Two of the most nagging problems have been that the less than twofold decrease in cGMP sensitivity of the native rod channel by CaM seems small compared with the large changes in the cell’s sensitivity during light adaptation, and efforts to demonstrate in vivo that a living rod is using that mechanism yielded mixed results [31, 42, 43, 132–134]. Even an unidentified endogenous factor has been implied, because exogenous CaM does not always fully recapitulate the change in the Ca2+ dependence of the cGMP sensitivity that is observed after patch excision or after

221

222

9 The Transduction Channels of Rod and Cone Photoreceptors

truncation of the ROS [42, 126]. It is unlikely that this factor is another Ca2+-binding protein. Removal of endogenous CaM from the cytosol of ROS by immunoprecipitation also removes all Ca2+-dependent modulatory activity [135]. Cone CNG channels in excised patches from some species are weakly modulated by CaM (heterologously expressed chicken A3 [136], native cone CNG channels of striped bass [44]), while cones from other species were found to be insensitive to CaM (native catfish [137], heterologously expressed human and bovine A3 [131, 136, 138, 139]). These observations do not appear to support a role for CaM in sensitivity regulation of the cone’s light response, but matters seem to be more complex. The A3 subunit of the cone CNG channel, unlike the A1 subunit of rods, comprises in its N-terminal region a fairly conserved CaM target motif that, in various binding assays, shows CaM binding [131, 136, 140]. Moreover, the B3 subunit of cones, like the B1 subunit of rods, carries two distinct CaM-binding sites in the N- and C-terminal region [139]. CNG channels lacking one of the two sites remained sensitive to regulation by Ca2+/CaM, but deletion of both sites abolished Ca2+/CaM sensitivity. The extent to which Ca2+/CaM modulates the cGMP sensitivity of heterologously expressed A3/ B3 channels is similar to that of rod CNG channels. These seemingly paradoxical findings might be reconciled by the idea that the cGMP sensitivity of the CNG channel of cones is controlled by another Ca2+-binding protein and that CaM can act as a weak partial agonist in some but not all species. Experimental support for this hypothesis has been provided by Rebrik and Korenbrot [41]. In electropermeabilized cones, the K1/2 (cGMP) increased from 84.3 lM in the absence of Ca2+ to 335 lM in the presence of 20 lM Ca2+ (the range is 67–550 lM). The Ca2+dependent modulation of K1/2 progressively and irreversibly vanished during recording of ligand sensitivity in low Ca2+ medium. In membrane patches detached from cone outer segments, the K1/2 is reduced only 1.5-fold in a solution free of Ca2+ ions, and the modulation is quickly lost after exposure of the patch to the Ca2+-free solution [44]. Finally, CaM does not mimic in detached membrane patches the effect of the diffusible factor. These results argue that a soluble factor, which is lost during perfusion, reversibly interacts with the CNG channel in a Ca2+-dependent manner. In electropermeabilized and truncated rods, the extent of Ca2+-dependent modulation of K1/2 is significantly smaller than in cones, and modulation in vivo is essentially recapitulated by CaM in excised patches [41–43]. Recently, Rebrik and Korenbrot [141] reported a clear difference in Ca2+ modulation between rod and cone CNG channels of the ground squirrel. These authors used caged BAPTA, a compound that releases the Ca2+ chelator BAPTA upon irradiation with UV light, to lower the free [Ca2+] inside intact photoreceptors. A flash of light at a constant concentration of 8-Br-cGMP activates an inward current through CNG channels in cones but not in rods. The rods of the ground squirrel, like those of other mammals, light-adapt very little and only near saturation of their light response. In contrast, cones adapt over five orders of magnitude of background light intensity. Thus, the lack of Ca2+ modulation in rods compared to cones testifies to the different importance of light adaptation for rods and cones in mammalian photoreceptors.

9.8 Phosphorylation

9.8

Phosphorylation

Repeated measurement of the dose-response curve of channel activation in excised patches of ROS membranes disclosed a slow increase in cGMP sensitivity over time [142]. The decrease in K1/2 was usually two- to threefold, but could be as large as 10-fold. The enhancement of ligand sensitivity was slowed down by both ATP and inhibitors of Ser/Thr phosphatases and was accelerated by purified type 1 phosphatase, suggesting that phosphorylation might control the conversion of channels between states of high and low ligand sensitivity. When the membrane patch was excised into a Ca2+-free medium (containing 1–2 mM EGTA), the time-dependent decrease of K1/2 was irreversibly abolished. This observation suggests that an unknown factor, possibly a Ca2+-dependent phosphatase tightly adhering to the membrane patch, is permanently inactivated or lost in Ca2+-free medium. The A1 subunit of the rod CNG channel, when heterologously expressed in Xenopus oocytes, displays a seemingly similar decrease of K1/2 after patch excision [143]. Unlike in the native rod CNG channel, inhibitors of Ser/Thr phosphatases are without effect, whereas orthovanadate and pervanadate—inhibitors of phospho-tyrosine phosphatases (PTP2)—slow down the progressive sensitivity enhancement [143, 144]. The effect is reduced but not entirely abolished by the mutation Y498F in the b1 strand of the cNMP-binding site, consistent with the idea that this residue in oocytes is phosphorylated by a protein tyrosine kinase (PTK). In Ca2+-free medium, the time-dependent effect of PTPs on A1 subunits in oocyte membranes persists, whereas in native ROS membranes no change in ligand sensitivity occurs [142, 143]. Genistein, a PTK inhibitor, dramatically slows channel activation and reduces maximal currents by twofold [145, 146]. These effects occur in the absence of ATP and were taken as evidence that genistein, in addition to inhibiting tyrosine phosphorylation, also promotes an allosteric inhibitory interaction between the PTK and the channel that does not involve phosphorylation [145, 146]. The action of genistein was also observed in native channels of rods, cones, and OSNs, raising the possibility that CNG channels are part of a regulatory complex that contains PTKs [145]. Insulin-like growth factor (IGF-1), a paracrine factor released from the retinal pigment epithelium, alters phototransduction by increasing the cGMP sensitivity of CNG channel in rods [147]. The IGF-1 signaling pathway involves a PTP that removes a phosphate group from residue Y498 of bovine A1 [147].

9.8.1

Retinal

All-trans-retinal directly inhibits cloned A1 channels expressed in Xenopus oocytes at nanomolar concentrations [148]. 11-cis-retinal and all-trans-retinal also inhibit the channel, although at somewhat higher concentrations. Inhibition probably occurs by a decrease of the open probability. Dean and his collaborators [148] have suggested that all-trans-retinal may be a potent regulator of the channel in rods during the re-

223

224

9 The Transduction Channels of Rod and Cone Photoreceptors

sponse to bright light, when there is a large surge in the concentration of all-transretinal. To firmly establish this hypothesis, experiments on intact rods under bright illumination regime are required. The K1/2 value of the A3 subunit of the cone CNG channel in HEK293 cells shifts from 19 lM to 56 lM upon treatment with phorbol esters [140]. The change in ligand sensitivity involves phosphorylation of two serine residues (S577 and S579) in the cNMP-binding domain. The d isoform of PKC is specifically expressed in cone outer segments and might mediate the phosphorylation. Phosphorylation and its effect on ligand sensitivity have not been studied in the native CNG channel of cone photoreceptors. In the chick, the cGMP sensitivity of the cone CNG channel is under the control of a circadian rhythm [56]. During the subjective night, the sensitivity is approximately twofold higher than during the subjective day. Circadian modulation of ligand sensitivity is driven, at least in part, by rhythms in the activities of two protein kinases: the Erk form of mitogen-activated protein kinase and the Ca2+/CaM-dependent protein kinase II (CaMKII). Perturbation of these signaling pathways causes phase-dependent changes in the cGMP sensitivity of the cone CNG channel. The mechanism of the sensitivity regulation has not been determined. It may involve phosphorylation of A or B subunits by Erk or CaMKII or phosphorylation of other signaling molecules that interact with the channels and lead to modulation of their sensitivity.

9.9

Visual Dysfunction Caused by Mutant CNG Channel Genes

Mutations in the genes for CNG channels of photoreceptors have been associated with the hereditary visual diseases retinitis pigmentosa, achromatopsia, and cone dystrophy. Retinitis pigmentosa (RP) is a clinically and genetically heterogeneous group of diseases characterized by night blindness, a progressive loss of the peripheral visual field, and eventual loss of central vision resulting in blindness. These symptoms reflect early dysfunction and degeneration of rod photoreceptors, followed by a slower degeneration of cone photoreceptors that proceeds from the periphery to the center of the visual field. RP is the most common inherited form of blindness, affecting approximately one out of 3000–4000 people worldwide. Dominant, recessive, X chromosomelinked, and digenic patterns of inheritance have been exemplified by families with RP (for review, see [149]). Mutations in the genes CNGA1 or CNGB1, encoding the CNG channel subunits A1 and B1 of rod photoreceptors, account for a small percentage of cases of autosomal recessive RP [150–152]. Achromatopsia is a recessively inherited, non-progressive disease characterized by the total loss of color vision, photophobia, and poor visual acuity resulting from the dysfunction of cone photoreceptors. The prevalence of achromatopsia has been estimated to be 1:30,000 (for review, see [153, 154]), with approximately 20–30 % of all cases caused by mutations in CNGA3 and 40–50 % of cases caused by mutations in CNGB3 [155]. CNGA3 and CNGB3 encode the CNG channel subunits A3 and B3 of

9.9 Visual Dysfunction Caused by Mutant CNG Channel Genes

cone photoreceptors. Incomplete forms of achromatopsia, characterized by residual cone function and color vision, have so far been exclusively associated with mutations in CNGA3 [156]. Moreover, in some instances mutations in CNGA3 result in cone dystrophy, a disease related to achromatopsia, but are characterized by the progressive loss of cone function and sometimes also progressive loss of rod function [156]. The steps that lead to photoreceptor degeneration are not completely understood and, in particular, the role of CNG channel subunits in this process remains to be elucidated (for review, see [149]). However, the recent success in the identification of CNG channel mutants that are responsible for hereditary visual diseases, and in the study of their defects by heterologous expression, has provided insight into the molecular basis of photoreceptor dysfunction. The results of molecular genetic analyses of patients with RP, achromatopsia, and cone dystrophy, as well as the in vitro experiments with the respective mutant CNG channel subunits, will be discussed in the following sections.

9.9.1

Mutations Associated with Retinitis Pigmentosa

So far, six different mutant CNGA1 alleles have been identified in RP patients [150, 151]. Three of these mutant alleles carry stop codons that terminate translation within the cytoplasmic N-terminus of A1 (R28X, E76X, and K139X). Another mutant CNGA1 allele has deleted most of if not the entire protein-coding region and therefore is also not expected to encode functional channels. The remaining two mutant alleles encode either a single amino acid substitution (S316F) or an amino acid substitution and a truncated C-terminus (R654D-X) (Fig. 9.7A). Dryja and collaborators expressed S316F and R654D-X in human embryonic kidney cells and found that these mutants hardly form functional homomeric channels [150]; cGMP-dependent currents carried by single or a few CNG channels were detected only in either three out of 85 (S316F) or one out of 83 membrane patches (R654D-X) tested. Similar results were obtained with the bovine R654D-X homologue (R656D-X) expressed in Xenopus laevis oocytes by Mallouk and collaborators [157]; co-expression of R656D-X and B1 did not rescue impaired channel trafficking. Surprisingly, neither truncation of the C-terminus (R656X) nor the substitution R656D alone impaired surface expression. A mutational screen identified two positions in R656D-X that are critical for channel trafficking: substituting leucine (L) or aspartate (D) in the C-terminal sequence KLKQD restored normal surface expression. Furthermore, substitution of the C-terminal five amino acids of wildtype A1 for KLKQD produced a mutant with impaired surface expression. In conclusion, Mallouk and co-workers [157] suggested that the sequence KLKQD, if positioned at the C-terminal end, serves as a signal that causes retention of the CNG channel subunit in the endoplasmic reticulum (ER). In contrast to the results reported by Dryja and Mallouk and their collaborators, Trudeau and Zagotta observed normal surface expression and channel formation of the human A1 mutant R654D-X in Xenopus oocytes [97]. Impaired trafficking to the membrane occurs only when R654D-X is coexpressed with B1. Using a biochemical pull-down assay, Trudeau and Zagotta [97]

225

226

9 The Transduction Channels of Rod and Cone Photoreceptors

Fig. 9.7 Mutations associated with retinitis pigmentosa, achromatopsia, or cone dystrophy. (A) Mutations in the A1 subunit of rod CNG channels associated with retinitis pigmentosa. (B) Mutations in the A3 subunit of cone CNG channels associated with achromatopsia or cone dystrophy. Mutations found exclusively in patients with incomplete achromatopsia or cone dystrophy (except T224R and V451fs; see text) are highlighted bold. (C) Mutations in the B3 subunit of cone CNG channels associated with achromatopsia. X, stop-codon mutations; fs, frame-shift mutation; del, amino acid deletion

9.9 Visual Dysfunction Caused by Mutant CNG Channel Genes

demonstrated stable binding of a peptide that represents the C-terminus of A1 (amino acids 609–693) to a peptide that represents the region between the GARP part and the first transmembrane domain of B1 (amino acids 677–764). Removing the C-terminal portion from the A1 peptide, which corresponds to amino acids 657–693 in the wildtype A1 subunit, abolished this binding. Trudeau and Zagotta [97] concluded from their results that the C-terminal amino acids 657–693 of wild-type A1—which are deleted in R654D-X—mask an ER retention signal within the amino acids 677–764 of B1. In co-expression experiments with B1 deletion mutants, this retention signal was confined to a segment of 10 amino acids that precedes the first transmembrane domain of B1 (YQFPQSIDPL). At present it is not clear how the contradictory observations reported by Dryja and Mallouk and their collaborators [150, 157] on the one hand and by Trudeau and Zagotta [97] on the other hand can be reconciled. Only a single mutation in the CNGB1 gene has as yet been associated with RP [152]. The mutation results in the substitution of a glycine (G993V) that is highly conserved in the CNBD of cyclic nucleotide-binding proteins, including hyperpolarization-activated, cyclic nucleotide-modulated (HCN) channels. According to the crystal structure of the CNBD in HCN2, this glycine resides in a turn between two b strands (b2 and b3) that contribute to a b roll inside which the ligand binds [158]. The substitution of valine for glycine at this site might impair the relative orientation of the two b strands and therefore might result in a nonfunctional CNBD and thereby nonfunctional rod CNG channels.

9.9.2

Mutations Associated with Achromatopsia or Cone Dystrophy

In a genetic screen of families from Europe and the United States, Kohl, Wissinger, and Johnson and their collaborators [156, 159, 160] identified 51 different mutant CNGA3 alleles that have been associated with achromatopsia or cone dystrophy. The majority of these alleles (40 of 51) encode single amino acid substitutions (Fig. 9.7B), indicating that there is little tolerance for sequence variations in A3 with respect to cone CNG channel function. Complete achromats are likely to carry mutant CNGA3 alleles that encode proteins, that are either nonfunctional or at least severely impaired in function. Some of the mutant CNGA3 alleles carried by complete achromats are also present in incomplete achromats or in patients with cone dystrophy (i.e., C191Y, R223W, R277C, R283W, R436W, and F547L). However, each incomplete achromat is expected to carry at least one CNGA3 allele that encodes functional A3 subunits, which might have altered properties but will permit some degree of cone function. Similarly, patients with cone dystrophy are likely to carry mutants that will permit cone function until the cells degenerate. Thus, mutant CNGA3 alleles carried only by incomplete achromats or patients with cone dystrophy are good candidates for encoding functional A3 subunits and thereby permitting residual cone function; these alleles, except T224R and V451fs (encoding an A3 polypeptide without a CNBD), are highlighted bold in Fig. 9.7B. The functional characterization of the A3 mutants encoded by these mutant alleles in vitro will help us to understand why the

227

228

9 The Transduction Channels of Rod and Cone Photoreceptors

maintenance and degree of cone function vary among patients with incomplete achromatopsia or cone dystrophy. In particular, examination of mutant CNG channels in their native subunit composition – produced in co-expression experiments – will be highly informative in this respect. Tra¨nkner and collaborators [161] have analyzed the molecular basis of an incomplete form of achromatopsia with considerably preserved cone function present in two sisters who both carry the mutations T224R and T369S. The sisters are able to discriminate saturated colors but confuse desaturated colors. Psychophysical and electroretinographic analyses show that the cone system is characterized by lower light sensitivity and by perturbed signal transfer from cones to postsynaptic neurons. Patch-clamp analysis of heterologously expressed subunits revealed that only the T369S mutant produces functional channels. The T369S channel displays altered ion permeation, gating, and ligand sensitivity and weaker blockage by extracellular Ca2+ ions. Co-expression of wild-type B3 with T369S restored most of the native properties, except for the weaker Ca2+ blockage. The strong influence of wild-type B3 on the T369S/ B3 channel was unexpected, given that photoreceptor CNG channels comprise only a single B subunit [91–93]. The properties of T369S/B3 channels suggest that the mild form of achromatopsia in the sisters results from relatively subtle changes in ion flux through the cone CNG channel. A genomic region that includes the CNGB3 gene has been identified as a locus for achromatopsia in people from Pingelap, an atoll island of Micronesia [162, 163]. About 5 % of the Pingelap population is affected by achromatopsia, also known as Pingelapese blindness (OMIM 262300), as noted in Oliver Sacks’ book The Island of the Colorblind [164]. The high frequency of the disease has been traced back to 1775, when a typhoon decimated the population of Pingelap, leaving only a handful of survivors who repopulated the island. Molecular genetic analysis identified a missense mutation in the CNGB3 gene (S435F) as the genetic basis of Pingelapese achromatopsia [165, 166] (Fig. 9.7C). Eight additional mutant CNGB3 alleles, identified in families from Europe, the United States, and Chile [160, 165–167], have as yet been associated with complete achromatopsia. Six of these mutant alleles encode B3 subunits that are lacking at least the pore region and the CNBD-containing cytoplasmic C-terminus due to stop-codon (R203X and E336X) or frame-shift mutations (K148E+K149fs, E199fs, P273fs, and T383fs). The remaining two mutant CNGB3 alleles comprise a missense mutation (F525N) or a splice-site mutation in intron 13 (not shown in Fig. 9.7C). Among these mutations, T383fs is notable because it accounts for more than 80 % of all CNGB3 mutations encountered in complete achromats [168]. The consequences of T383fs and the Pingelapese mutation S435F for cone CNG channel function have been analyzed by Peng and collaborators using heterologous expression of channel subunits in Xenopus oocytes [169]. Co-expression of wildtype A3 with mutant T383fs produced channels with functional properties not different from homomeric A3 channels. This result suggests that T383fs does not contribute to the formation of functional cone CNG channels and raises the possibility that those complete achromats that are homozygous for the T383fs allele (e.g., patients II:1, II:3, and II:5 in family CHRO56 [165]) might express homomeric A3 channels only. In contrast to T383fs, S435F does contribute to functional heteromeric channels. The

9.9 Visual Dysfunction Caused by Mutant CNG Channel Genes

A3/S435F channels display relatively mild functional alterations, compared with heteromeric wild-type channels. These alterations include an enhanced sensitivity for cGMP (twofold) and cAMP (four- to fivefold) and a 20 % decrease in single-channel conductance. These relatively mild changes in channel function seem to be inconsistent with the complete loss of cone function in the achromats of Pingelap. However, the authors observed a small but significant decrease in the density of heteromeric channels comprising mutant S435F, compared with wild-type channels. Possibly, surface expression of A3/S435F is more severely impaired in vivo, thereby causing the cone dysfunction. This possibility is supported by the observation that mutant S435F does not contribute to functional channels if expressed heterologously in human embryonic kidney cells (D. Tra¨nkner, K. Ko¨ppen, B. Wissinger, and U.B. Kaupp, unpublished observation).

Appendix

Consistent with the current nomenclature [170], the short term for each mutant protein names the amino acid of the wild-type protein that is either substituted or absent in the mutant, followed by the position of this amino acid in the protein sequence. The position number is followed by the amino acid present in the mutant at this site (missense mutation), by “del” if this amino acid is deleted in the mutant, or by “X” if translation is terminated in the mutant at this site (stop-codon mutation). In the case of frame-shift mutations (except R564D-X), the position number is followed by “fs”; to avoid complexity, we have omitted the position of the stop codon in the frame-shift mutant in the text. However, the full description of the frame shift is given here: the CNGA1 mutation R564D-X corresponds to R564fsX565; the CNGA3 mutations I50fs, V451fs, and I482fs correspond to I50fsX59, V451fsX453, and I482fsX487, respectively. Finally, the CNGB3 mutations K148E+K149fs, E199fs, P273fs, and T383fs correspond to K148E+K149fsX177, E199fsX201, P273fsX286, and T383fsX395, respectively. To avoid confusion due to different short terms for mutants used here and in the original reports, the original short terms are listed as follows: the CNGA1 mutations E76X, K139X, and R654D-X were originally termed Glu76End, Lys139End, and Arg654(1-bp del), respectively [150]. The CNGA3 mutation I50fs was originally termed G49fs [156]; the CNGA3 mutations R23X, Q196X, R221X, I482fs, and G548R were originally termed Arg23stop, Gln196stop, Arg221stop, Ile482fs, and Gly548Arg, respectively [160]. The CNGB3 mutation K148E+K149fs was originally termed K148E+N149fsX177 [167]; the CNGB3 mutations E199fs and F525N were originally termed Glu199fs and Phe525Asn, respectively [160].

229

230

9 The Transduction Channels of Rod and Cone Photoreceptors

References 1

2 3 4 5 6 7

8 9 10 11

12 13 14 15 16 17 18 19 20 21 22 23 24

E.N.Jr. Pugh and T.D. Lamb, in D.G. Stavenga, W.J. DeGrip and E.N.Jr. Pugh (editors), Handbook of Biological Physics, Elsevier Science B.V., North-Holland, 2000, 183 pp. G.L. Fain, H.R. Matthews, M.C. Cornwall and Y. Koutalos, Physiol. Rev., (2001) 81, 117–151. K.-W. Yau and D.A. Baylor, Annu. Rev. Neurosci., (1989) 12, 289–327. W.N. Zagotta and S.A. Siegelbaum, Annu. Rev. Neurosci., (1996) 19, 235–263. A. Menini, Curr. Opinion Neurobiol., (1999) 9, 419–426. M.J. Richards and S.E. Gordon, Biochemistry, (2000) 39, 14003–14011. R.S. Molday and U.B. Kaupp, in D.G. Stavenga, W.J. DeGrip and E.N.Jr. Pugh (editors), Molecular Mechanisms in Visual Transduction, Elsevier, North Holland, 2000, 143 pp. G.E. Flynn, J.P.Jr. Johnson and W.N. Zagotta, Nat. Rev. Neurosci., (2001) 2, 643–652. U.B. Kaupp and R. Seifert, Physiol. Rev., (2002) 82, 769–824. J. Li, W.N. Zagotta and H.A. Lester, Q. Rev. Biophys., (1997) 30, 177–193. M. Punta, A. Cavalli, V. Torre and P. Carloni, Proteins: Struct., Funct., Genet., (2003) 52, 332–338. A.L. Zimmerman and D.A. Baylor, Nature, (1986) 321, 70–72. J.W. Karpen, D.A. Loney and D.A. Baylor, J. Physiol., (1992) 448, 257–274. R.D. Bodoia and P.B. Detwiler, J. Physiol., (1985) 367, 183–216. P. Gray and D. Attwell, Proc. R. Soc. Lond. B, (1985) 223, 379–388. N.J. Cook and U.B. Kaupp, Photobiochem. Photobiophys., (1986) 13, 331–343. N.J. Cook, W. Hanke and U.B. Kaupp, Proc. Natl. Acad. Sci. USA, (1987) 84, 585–589. A. Caretta, A. Cavaggioni and R.T. Sorbi, Eur. J. Biochem., (1985) 153, 49–53. L.W. Haynes and K.-W. Yau, J. Physiol. , (1990) 429, 451–481. L. Lagnado, L. Cervetto and P.A. McNaughton, J. Physiol., (1992) 455, 111–142. K. Nakatani and K.-W. Yau, Nature, (1988) 334, 69–71. M.P. Gray-Keller and P.B. Detwiler, Neuron, (1994) 13, 849–861. J.P. Younger, S.T. McCarthy and W.G. Owen, J. Neurophysiol., (1996) 75, 354–366. T. Ohyama, D.H. Hackos, S. Frings, V. Hagen, U.B. Kaupp and J.I. Korenbrot, J. Gen. Physiol., (2000) 116, 735–753.

25 26 27 28 29

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

P.A. McNaughton, Physiol. Rev., (1990) 70, 847–883. E.N.Jr. Pugh and T.D. Lamb, Vision Res., (1990) 30, 1923–1948. K.-W. Koch, Behav.Brain Sci., (1995) 18, 3– K. Palczewski, A.S. Polans, W. Baehr and J.B. Ames, BioEssays, (2000) 22, 337–350. I.I. Senin, K.-W. Koch, M. Akhtar and P.P. Philippov, in K. Palczewski and W. Baehr (Editors), Photoreceptors and Calcium, Landes Bioscience, 2002, Y.-T. Hsu and R.S. Molday, Nature, (1993) 361, 76–79. R.S. Molday, Curr. Opinion Neurobiol., (1996) 6, 445–452. Y. Koutalos, K. Nakatani, T. Tamura and K.-W. Yau, J. Gen. Physiol., (1995) 106, 863–890. Y. Koutalos, K. Nakatani and K.-W. Yau, J. Gen. Physiol., (1995) 106, 891–921. J.L. Miller, A. Picones and J.I. Korenbrot, Curr. Opinion Neurobiol., (1994) 4, 488–495. A. Picones and J.I. Korenbrot, Biophys. J., (1995) 69, 120–127. G.B. Wells and J.C. Tanaka, Biophys. J., (1997) 72, 127–140. R.J. Perry and P.A. McNaughton, J. Physiol., (1991) 433, 561–587. A.L. Hodgkin, P.A. McNaughton and B.J. Nunn, J. Physiol., (1987) 391, 347–370. W.H. Cobbs and E.N.Jr. Pugh, Nature, (1985) 313, 585–587. S. Hestrin and J.I. Korenbrot, J. Neurosci., (1990) 10, 1967–1973. T.I. Rebrik and J.I. Korenbrot, J. Gen. Physiol., (1998) 112, 537–548. M.S. Sagoo and L. Lagnado, J. Physiol. , (1996) 497, 309–319. K. Nakatani, Y. Koutalos and K.-W. Yau, J. Physiol., (1995) 484, 69–76. D.H. Hackos and J.I. Korenbrot, J. Gen. Physiol., (1997) 110, 515–528. C.R. Bader, D. Bertrand and E.A. Schwartz, J. Physiol., (1982) 331, 253–284. D. Attwell, S. Borges, S.M. Wu and M. Wilson, Nature, (1987) 328, 522–524. J.H. Belgum and D.R. Copenhagen, J. Physiol., (1988) 396, 225–245. D.A. Baylor, M.G.F. Fuortes and P. O’Bryan, J. Physiol., (1971) 214, 265–294. M.G.F. Fuortes, E.A. Schwartz and E.J. Simon, J. Physiol., (1973) 234, 199–216. R.A. Normann and I. Perlman, Vision Res., (1979) 19, 391–394.

9.9 Visual Dysfunction Caused by Mutant CNG Channel Genes 51 52 53

54

55

56 57 58 59 60 61 62 63 64 65 66 67

68 69 70 71 72 73 74 75 76

F. Rieke and E.A. Schwartz, Neuron, (1994) 13, 863–873. A. Savchenko, S. Barnes and R.H. Kramer, Nature, (1997) 390, 694–698. K.-W. Koch, H.-G. Lambrecht, M. Haberecht, D. Redburn and H.H.H.W. Schmidt, EMBO J., (1994) 13, 3312–3320. D.E. Kurenny, L.L. Moroz, R.W. Turner, K.A. Sharkey and S. Barnes, Neuron, (1994) 13, 315–324. B.A. Liepe, C. Stone, J. Koistinaho and D.R. Copenhagen, J. Neurosci., (1994) 14, 7641–7654. G.Y.P. Ko, M.L. Ko and S.E. Dryer, Neuron, (2001) 29, 255–266. G.Y.P. Ko, M.L. Ko and S.E. Dryer, J. Neurosci., (2003) 23, 3145–3153. G.Y. Ko, M.L. Ko and S.E. Dryer, J. Neurosci., (2004) 24, 1296–1304. A. Pu and J.E. Dowling, J. Neurophysiol., (1981) 46, 1018–1038. S. Tomotsu and Y. Morita, J.Comp.Physiol.A, (1986) 159, 1–5. R.N. Lolley, C.M. Craft and R.H. Lee, Neurochem. Res., (1992) 17, 81–89. H.-W. Korf, Annals New York Acad. Sci. , (1994) 719, 13–42. S.E. Dryer and D. Henderson, Nature, (1991) 353, 756–758. S.E. Dryer and D. Henderson, J. Comp. Physiol. A, (1993) 172, 271–279. G. Matthews and S.-I. Watanabe, J. Physiol., (1988) 403, 389–405. A.C. Newton, Curr. Biol., (1995) 5, 973–976. J.S. Takahashi, N. Murakami, S.S. Nikaido, B.L. Pratt and L.M. Robertson, Rec. Progr. Horm. Res., (1989) 45, 279–352. N.L. Harrison and M. Zatz, J. Neurosci., (1989) 9, 2462–2467. E. Solessio and G.A. Engbretson, Nature, (1994) 364, 442–445. J.T. Finn, E.C. Solessio and K.-W. Yau, Nature, (1997) 385, 815–819. S. Frings, R. Seifert, M. Godde and U.B. Kaupp, Neuron, (1995) 15, 169–179. L.W. Haynes, J. Gen. Physiol., (1995) 106, 507–523. D.H. Hackos and J.I. Korenbrot, J. Gen. Physiol., (1999) 113, 799–817. W.-H. Xiong, E.C. Solessio and K.-W. Yau, Nature Neurosci., (1998) 1, 359–365. J.S. McReynolds and A.L.F. Gorman, Science, (1974) 183, 658–659. M. Gomez and E. Nasi, J. Gen. Physiol., (1997) 109, 371–384.

77 78 79 80

81

82 83 84

85

86

87 88 89 90

91 92 93 94 95

96 97 98

M. del Pilar Gomez and E. Nasi, Neuron, (1995) 15, 607–618. E. Nasi and M. del Pilar Gomez, J. Gen. Physiol., (1999) 114, 653–671. T. Gotow, T. Nishi and H. Kijima, Brain Res., (1994) 662, 268–272. U.B. Kaupp, T. Niidome, T. Tanabe, S. Terada, W. B€ onigk, W. St€ uhmer, N.J. Cook, K. Kangawa, H. Matsuo, T. Hirose, T. Miyata and S. Numa, Nature, (1989) 342, 762–766. H.G. K€ orschen, M. Illing, R. Seifert, F. Sesti, A. Williams, S. Gotzes, C. Colville, F. M€ uller, A. Dose´, M. Godde, L. Molday, U.B. Kaupp and R.S. Molday, Neuron, (1995) 15, 627–636. J. Bradley, S. Frings, K.-W. Yau and R. Reed, Science, (2001) 294, 2095– G.E. Flynn and W.N. Zagotta, Neuron, (2001) 30, 689–698. R.S. Molday, L.L. Molday, A. Dose´, I. ClarkLewis, M. Illing, N.J. Cook, E. Eismann and U.B. Kaupp, J. Biol. Chem., (1991) 266, 21917–21922. R.S. Molday, D.M. Reid, G. Connell and L.L. Molday, in P.A. Hargrave, K.P. Hofmann and U.B. Kaupp (editors), Signal Transduction in Photoreceptor Cells, Springer-Verlag, Berlin Heidelberg, 1992, 180 pp. W. B€ onigk, W. Altenhofen, F. M€ uller, A. Dose, M. Illing, R.S. Molday and U.B. Kaupp, Neuron, (1993) 10, 865–877. P. Wohlfart, W. Haase, R.S. Molday and N.J. Cook, J. Biol. Chem., (1992) 267, 644–648. L. Heginbotham, T. Abramson and R. MaKKinnon, Science, (1992) 258, 1152–1155. L.Y. Jan and Y.N. Jan, Nature, (1990) 345, 672– D.K. Henn, A. Baumann and U.B. Kaupp, Proc. Natl. Acad. Sci. USA, (1995) 92, 7425–7429. D. Weitz, N. Ficek, E. Kremmer, P.J. Bauer and U.B. Kaupp, Neuron, (2002) 36, 881–889. J. Zheng, M.C. Trudeau and W.N. Zagotta, Neuron, (2002) 36, 891–896. H. Zhong, L.L. Molday, R.S. Molday and K.W. Yau, Nature, (2002) 420, 193–198. A.L. Zimmerman, Neuron, (2002) 36, 997–999. Y. Jiang, A. Lee, J. Chen, M. Cadene, B.T. Chait and R. MacKinnon, Nature, (2002) 417, 523–526. H. Zhong, J. Lai and K.-W. Yau, Proc. Natl. Acad. Sci. USA, (2003) 100, 5509–5513. M.C. Trudeau and W.N. Zagotta, Neuron, (2002) 34, 197–207. N. Bennett, M. Ildefonse, F. Pages and M. Ragno, Biophys. J., (2002) 83, 920–931.

231

232

9 The Transduction Channels of Rod and Cone Photoreceptors 99 100

101 102

103

104 105

106

107

108 109

110 111 112

113

114 115 116 117 118

D.T. Liu, G.R. Tibbs and S.A. Siegelbaum, Neuron, (1996) 16, 983–990. Y. Sugimoto, K. Yatsunami, M. Tsujimoto, H.G. Khorana and A. Ichikawa, Proc. Natl. Acad. Sci. USA, (1991) 88, 3116–3119. C.A. Colville and R.S. Molday, J. Biol. Chem., (1996) 271, 32968–32974. H.G. K€orschen, M. Beyermann, F. M€ uller, M. Heck, M. Vantler, K.-W. Koch, R. Kellner, U. Wolfrum, C. Bode, K.P. Hofmann and U.B. Kaupp, Nature, (1999) 400, 761–766. M.D. Ardell, I. Aragon, L. Oliveira, G.E. Porche, E. Burke and S.J. Pittler, FEBS Lett., (1996) 389, 213–218. M.D. Ardell, D.L. Bedsole, R.V. Schoborg and S.J. Pittler, Gene, (2000) 245, 311–318. A. Sautter, X. Zong, F. Hofmann and M. Biel, Proc. Natl. Acad. Sci. USA, (1998) 95, 4696–4701. W. B€onigk, J. Bradley, F. M€ uller, F. Sesti, I. Boekhoff, G.V. Ronnett, U.B. Kaupp and S. Frings, J. Neurosci., (1999) 19, 5332–5347. B. Wiesner, J. Weiner, R. Middendorff, V. Hagen, U.B. Kaupp and I. Weyand, J. Cell Biol., (1998) 142, 473–484. A. Gerstner, X. Zong, F. Hofmann and M. Biel, J. Neurosci., (2000) 20, 1324–1332. R. Allikmets, N.F. Shroyer, N. Singh, J.M. Seddon, R.A. Lewis, P.S. Bernstein, A. Peiffer, N.A. Zabriskie, Y. Li, A. Hutchinson, M. Dean, J.R. Lupski and M. Leppert, Science, (1997) 277, 1805–1807. M. Illing, L.L. Molday and R.S. Molday, J. Biol. Chem., (1997) 272, 10303–10310. H. Sun, R.S. Molday and J. Nathans, J. Biol. Chem., (1999) 274, 8269–8281. D.S. Papermaster, B.G. Schneider, M.A. Zorn and J.P. Kraehenbuhl, J. Cell Biol., (1978) 78, 415–425. X. Liu, K. Seno, Y. Nishizawa, F. Hayashi, A. Yamazaki, H. Matsumoto, T. Wakabayashi and J. Usukura, Exp. Eye Res., (1994) 59, 761–768. A. Poetsch, L.L. Molday and R.S. Molday, J. Biol. Chem., (2001) 276, 48009–48016. P.J. Bauer and M. Drechsler, J. Physiol., (1992) 451, 109–131. R.S. Molday and L.L. Molday, Vision Res., (1998) 38, 1315–1323. A. Schwarzer, H. Schauf and P.J. Bauer, J. Biol. Chem., (2000) 275, 13448–13454. D.M. Reid, U. Friedel, R.S. Molday and N.J. Cook, Biochemistry, (1990) 29, 1601–1607.

119 K. Kang, P.J. Bauer, T.G. Kinjo, R.T. Sze-

120 121 122 123 124 125

126 127 128 129

130 131

132 133 134 135 136

137 138 139

140

141 142

rencsei, W. B€ onigk, R.J. Winkfein and P.P.M. Schnetkamp, Biochemistry, (2003) 42, 4593–4600. E. Neher, in Anonymous Exp. Brain Res., Springer, Heidelberg, 1986, p. 80. E. Neher, Cell Calcium, (1998) 24, 345–357. M. Naraghi and E. Neher, J. Neurosci. , (1997) 17, 6961–6973. P.J. Bauer, Cell Biochem. Biophys., (2001) 35, 49–61. P.J. Bauer, J. Physiol., (1996) 494, 675–685. D. Weitz, M. Zoche, F. M€ uller, M. Beyermann, H.G. K€ orschen, U.B. Kaupp and K.-W. Koch, EMBO J., ( 1998) 17, 2273–2284. S.E. Gordon, J. Downing-Park and A.L. Zimmerman, J. Physiol., (1995) 486, 533–546. J.I. Korenbrot and D.L. Miller, Vision Res., (1989) 29, 939–948. S.T. McCarthy, J.P. Younger and W.G. Owen, Biophys. J., (1994) 67, 2076–2089. T.-Y. Chen, M. Illing, L.L. Molday, Y.-T. Hsu, K.-W. Yau and R.S. Molday, Proc. Natl. Acad. Sci. USA, (1994) 91, 11757–11761. M.E. Grunwald, W.-P. Yu, H.-H. Yu and K.-W. Yau, J. Biol. Chem., (1998) 273, 9148–9157. M.E. Grunwald, H. Zhong, J. Lai and K.-W. Yau, Proc. Natl. Acad. Sci. USA, (1999) 96, 13444–13449. M.P. Gray-Keller and P.B. Detwiler, Behav. Brain Sci., (1995) 18, 475–476. P.B. Detwiler and M.P. Gray-Keller, Curr. Opinion Neurobiol., (1996) 6, 440–444. Y. Koutalos and K.-W. Yau, Trends Neurosci., (1996) 19, 73–81. R. Warren and R.S. Molday, Adv. Exp. Med. Biol., (2002) 514, 205–223. W. B€ onigk, F. M€ uller, R. Middendorff, I. Weyand and U.B. Kaupp, J. Neurosci., (1996) 16, 7458–7468. L.W. Haynes and S.C. Stotz, Vis. Neurosci., (1997) 14, 233–239. W.-P. Yu, M.E. Grunwald and K.-W. Yau, FEBS Lett., (1996) 393, 211–215. C.H. Peng, E.D. Rich, C.A. Thor and M.D. Varnum, J. Biol. Chem., (2003) 278, 24617–24623. F. M€ uller, M. Vantler, D. Weitz, E. Eismann, M. Zoche, K.-W. Koch and U.B. Kaupp, J. Physiol., (2001) 533, 399–409. T.I. Rebrik and J.I. Korenbrot, J. Gen. Physiol., (2004) 123, 63–75. S.E. Gordon, D.L. Brautigan and A.L. Zimmerman, Neuron, (1992) 9, 739–748.

9.9 Visual Dysfunction Caused by Mutant CNG Channel Genes 143 E. Molokanova, B. Trivedi, A. Savchenko and

157 N. Mallouk, M. Ildefonse, F. Pages, M. Ragno

R.H. Kramer, J. Neurosci., (1997) 17, 9068–9076. E. Molokanova, F. Maddox, C.W. Luetje and R.H. Kramer, J. Neurosci., (1999) 19, 4786–4795. E. Molokanova, A. Savchenko and R.H. Kramer, J. Gen. Physiol., (2000) 115, 685–696. E. Molokanova and R.H. Kramer, J. Gen. Physiol., (2001) 117, 219–233. A. Savchenko, T.W. Kraft, E. Molokanova and R.H. Kramer, Proc. Natl. Acad. Sci. USA, (2001) 98, 5880–5885. D.M. Dean, W. Nguitragool, A. Miri, S.L. McCabe and A.L. Zimmerman, Proc. Natl. Acad. Sci. USA, (2002) 99, 8372–8377. E.A. Pierce, BioEssays, (2001) 23, 605–618. T.P. Dryja, J.T. Finn, Y.-W. Peng, T.L. McGee, E.L. Berson and K.-W. Yau, Proc. Natl. Acad. Sci. USA, (1995) 92, 10177–10181. E. Paloma, A. Martinez-Mir, B. Garcia-Sandoval, C. Ayuso, L. Vilageliu, R. GonzalezDuarte and S. Balcells, J. Med. Genet., (2002) 39, E66– C. Bareil, C.P. Hamel, V. Delague, B. Arnaud, J. Demaille and M. Claustres, Hum. Genet., (2001) 108, 328–334. L.T. Sharpe and K. Nordby, in R.F. Hess, L.T. Sharpe and K. Nordby (editors), Night Vision: Basic, Clinical and Applied Aspects, Cambridge University Press, Cambridge, 1990, 253 pp. L.T. Sharpe, A. Stockman, H. J€agle and J. Nathans, in K. Gegenfurtner and L.T. Sharpe (Editors), Color Vision: From Genes to Perception, Cambridge University Press, Cambridge, 1999, p. 3. S. Kohl, B. Baumann, T. Rosenberg, U. Kellner, B. Lorenz, M. Vadala’, S.G. Jacobson and B. Wissinger, Am. J. Hum. Genet., (2002) 71, 422–425. B. Wissinger, D. Gamer, H. Jagle, R. Giorda, T. Marx, S. Mayer, S. Tippmann, M. Broghammer, B. Jurklies, T. Rosenberg, S.G. Jacobson, E.C. Sener, S. Tatlipinar, C.B. Hoyng, C. Castellan, P. Bitoun, S. Andreasson, G. Rudolph, U. Kellner, B. Lorenz, G. Wolff, C. Verellen-Dumoulin, M. Schwartz, F.P.M. Cremers, E. Apfelstedt-Sylla, E. Zrenner, R. Salati, L.T. Sharpe and S. Kohl, Am. J. Hum. Genet., (2001) 69, 722–737.

and N. Bennett, J. Membr. Biol., (2002) 185, 129–136. W.N. Zagotta, N.B. Olivier, K.D. Black, E.C. Young, R. Olson and E. Gouaux, Nature, (2003) 425, 200–205. S. Kohl, T. Marx, I. Giddings, H. J€agle, S.G. Jacobson, E. Apfelstedt-Sylla, E. Zrenner, L.T. Sharpe and B. Wissinger, Nature Genet., (1998) 19, 257–259. S. Johnson, M. Michaelides, I.A. Aligianis, J.R. Ainsworth, J.D. Mollon, E.R. Maher, A.T. Moore and D.M. Hunt, J. Med. Genet., (2004) 41, E20– D. Tr€ankner, H. J€agle, S. Kohl, E. ApfelstedtSylla, L.T. Sharpe, U.B. Kaupp, E. Zrenner, R. Seifert and B. Wissinger, J. Neurosci., (2004) 24, 138–147. J.D. Winick, M.L. Blundell, B.L. Galke, A.A. Salam, S.M. Leal and M. Karayiorgou, Am. J. Hum. Genet., (1999) 64, 1679–1685. A. Milunsky, X.L. Huang, J. Milunsky, A. DeStefano and C.T. Baldwin, Clin. Genet., (1999) 56, 82–85. O. Sacks, The Island of the Colorblind , Vintage Books, Random Hous, New York, NY, 1998. S. Kohl, B. Baumann, M. Broghammer, H. J€agle, P. Sieving, U. Kellner, R. Spegal, M. Anastasi, E. Zrenner, L.T. Sharpe and B. Wissinger, Hum. Mol. Genet., (2000) 9, 2107–2116. O.H. Sundin, J.-M. Yang, Y. Li, D. Zhu, J.N. Hurd, T.N. Mitchell, E.D. Silva and I.H. Maumenee, Nature Genet., (2000) 25, 289–293. C.V. Rojas, L. Santa Mar’a, J.L. Santos, F. Corte´s and M.A. Alliende, Eur. J. Human Gen., (2002) 10, 638–642. S. Kohl, H. J€agle, E. Zrenner, L.T. Sharpe and B. Wissinger, Invest. Ophthalmol. Vis .Sci., 42 (2001) S324.(Abstract) C. Peng, E.D. Rich and M.D. Varnum, J. Biol. Chem., (2003) 278, 34533–34540. J.T. den Dunnen and S.E. Antonarakis, Hum. Genet., (2001) 109, 121–124.

144

145 146 147

148

149 150

151

152

153

154

155

156

158

159

160

161

162

163

164 165

166

167

168

169 170

233

235

10

Ion Channels and Thermotransduction Michael J. Caterina

10.1

Introduction

Life as we know it is compatible with a relatively narrow range of ambient temperatures. Extremes of temperature can compromise the efficiency of biochemical processes or produce life-threatening cellular damage and must therefore be avoided. On the other hand, modest inhomogeneities in temperature can provide useful information relating to an individual’s physiological status and the presence of neighboring individuals or food. Consequently, species from bacteria to mammals have evolved mechanisms that allow them to spatially and temporally measure their temperatures and that of their surroundings. In mammals, the evaluation of surface and core body temperatures is mediated largely by specialized neurons located in the peripheral nervous system and preoptic/anterior hypothalamus, respectively. At the cellular level, the transduction of thermal stimuli into neuronal excitation appears to ultimately result from temperaturemediated changes in ion channel activity. Recently, several mammalian ion channels have been identified whose activities exhibit exceptional and nonlinear sensitivity to temperature within the biologically meaningful range. Each of these channels is characterized by a distinct temperature response profile that suits it to respond maximally over a particular segment of the temperature spectrum. The cell type-specific localization of these thermosensitive channels and the phenotypic analysis of mutants in which they have been genetically disrupted have begun to provide convincing evidence for their roles in peripheral thermosensation. Moreover, the involvement of homologous ion channels in invertebrate thermosensation argues that the use of these proteins represents an ancient strategy for environmental perception.

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

236

10 Ion Channels and Thermotransduction

10.2

Physiological Studies Provide Evidence for the Existence of Thermally Gated Ion Channels

Some of the earliest evidence for the existence of discrete mechanisms underlying thermosensation came from the observation that heating or cooling of localized sensory spots in the skin could evoke distinct thermal sensations in humans [1]. Almost a century later, direct electrophysiological recordings from peripheral neurons innervating mammalian tongue [2–4] and skin [5–11] demonstrated definitively that certain peripheral sensory neurons vary their firing rates as a function of both absolute temperature and acute temperature changes. While temperature influences the firing rates of all neurons to some extent, temperature-sensitive neurons were defined as those whose firing rates were unusually sensitive to this parameter. Careful analysis also revealed that different temperature-sensitive neurons respond best over different temperature ranges (Fig. 10.1, top panel). Warm and cold thermoreceptors, respectively, exhibit a tonic firing rate at normal surface temperatures but increase or decrease their firing rates in response to temperature changes over the innocuous (i.e., non-painful) range. Warm thermoreceptors begin to fire as the temperature of their receptive field rises above 30 8C and reaches a peak at 41–46 8C, with a relatively sharp decrease thereafter [6–9]. In contrast, cold thermoreceptors fire only below 40 8C, with increased activity at lower temperaturees, a peak at 25–30 8C, and gradually decreased activity at even colder temperatures [3, 6, 8, 9]. It should be noted, however, that these values are only representative and that considerable variability in the thermal response profiles of thermoreceptors has been observed as a function of species and anatomical location. In primates, warm thermoreceptors are small-diameter, unmyelinated C fibers, while cold thermoreceptors appear to be medium-diameter, lightly myelinated Ad fibers [11, 12]. In contrast, both classes of thermoreceptors appear to be C fibers in subprimate mammals [7, 9]. A distinct subpopulation of so-called heat nociceptors is silent at mildly elevated temperatures but becomes activated once skin temperature reaches the painful range (> 42 8C) [5, 9]. Heat nociceptors, like warm and cold thermoreceptors, fall into both the C-fiber and Ad-fiber classes and project to the superficial layers of the spinal cord dorsal horn. Finally, cold nociceptors, which are also of the C- and Ad-fiber classes, respond to painfully cold temperatures [9, 13]. Interestingly, certain so-called “polymodal” nociceptors exhibit responsiveness to temperature over multiple ranges, such as cold and painful heat, as well as responsiveness to noxious chemical and mechanical stimuli [4, 9]. As further evidence of complexity, at least three distinct populations of heat nociceptors can be distinguished on the basis of fiber diameter, activation threshold, response latency, and sensitization or desensitization during repeated heat exposure [14, 15]. The findings described above established the existence of discrete anatomical substrates underlying peripheral thermotransduction. However, only recently has an understanding of the physiological basis of these transduction events begun to emerge. Using patch-clamp methods, several groups of scientists were able to demonstrate in the 1990s that increases in ambient temperature resulted in the opening of nonselective cation channels in the cell bodies of a subpopulation of neurons isolated from

10.1 Introduction

Relative Activity

peripheral dorsal root ganglia [16–19] (Fig. 10.1, middle panel; Fig. 10.2). In contrast, such currents were never observed in non-sensory neurons, such as those derived from the sympathetic superior cervical ganglion. The very nonlinear temperature dependence of these currents led to the recognition that specific temperatures, like specific chemicals, could selectively trigger the opening of ion channels. Interestingly, the activation threshold exhibited by heat-gated ion channels (42 8C) was very close to that for heat-evoked pain in humans and most other mammals. Furthermore, the subset of neurons responsive to these temperatures exhibited morphological and functional features suggesting that they were nociceptors. Together, these finding led to the

Temperature Sensitive Neurons in vivo

0

10

20

30 40 Temperature ( C)

50

60

Sensory Neurons in vitro

Keratinocytes in vitro

TRPV2

ANKTM1 TRPM8

TRPV1 TRPV3

Recombinant Channels Expressed Heterologously

TRPV4

Fig. 10.1 Temperature-response characteristics of thermosensitive mammalian cells and thermosensitive ion-channel proteins. Top: Typical normalized firing rates of several distinct classes of cutaneous thermosensitive neurons recorded in vivo (purple, cold nociceptor; blue, cold thermoreceptor; gold, warm thermoreceptor; orange, moderate-threshold heat nociceptor; red, highthreshold heat nociceptor). Middle: Temperatureresponse profiles of thermosensitive rodent sensory neurons or immortalized mouse keratinocytes

recorded in vitro. Width of shaded bars represents relative cation influx at a given ambient temperature. Bottom: Recombinant thermosensitive ion channels of the TRP family, expressed heterologously in human embryonic kidney cells or Xenopus laevis oocytes, mediate ionic influx patterns resembling those of native thermosensitive ion channels. Width of shaded bars represents relative cation influx at a given ambient temperature. See text for references

237

238

10 Ion Channels and Thermotransduction

notion that the channels being assayed might represent the hypothetical “heat receptors” responsible for the initiation of heat-evoked pain. Temperature-evoked changes in neuronal firing has also been demonstrated in dissociated neurons or organotypic slice preparations derived from the anterior/preoptic hypothalamus. Indeed, both warm-sen-

Fig. 10.2 Electrophysiological responses of cultured rat dorsal root ganglion neurons to noxious heat stimuli. (A) Whole-cell current responses exhibited by a heat-insensitive neuron (trace 1) versus a heat-sensitive neuron (trace 2) during an increase in ambient temperature from 28 8C to 49 8C (top trace) at a holding potential of –70 mV. (B) Temperature-response profile of heat-evoked currents in representative heat-insensitive (trace 1) or heatsensitive (trace 2 and filled circles) neurons. Traces indicate instantaneous current responses recorded during a continuous heat ramp like that illustrated in panel (A). Filled circles represent peak responseamplitudes recorded during step pulses to each

of the indicated temperatures. (C) Top: Heatevoked single-channel currents recorded in cellattached configuration from cultured dorsal root ganglion neurons (DRG, upper three traces) or cultured superior cervical ganglion neurons (SCG, lower trace) at the indicated temperatures. Bottom: Temperature dependence of cell-attached singlechannel openings in a heat-sensitive dorsal root ganglion neuron. (Panels (A) and (B) reproduced with permission from [17], Copyright 1996, Proceedings of the National Academy of Sciences. Panel (C) reproduced with permission from [18], Copyright 1999, The Society for Neuroscience)

10.3 Molecular Characterization of a Heat-gated Ion Channel, TRPV1

sitive and cold-sensitive hypothalamic neurons have been described. However, the biophysical basis of the temperature-dependent firing rates recorded from these neurons remains somewhat controversial. Some findings suggest that thermosensitivity results from subtle temperature-dependent changes in the kinetics of activation and inactivation of “housekeeping” voltage-gated sodium and potassium channels, with modulation by GABA-gated chloride channels [20]. However, there is also evidence for the existence of temperature-sensitive nonselective cation channels, similar to those expressed in peripheral thermosensitive neurons, that respond discontinuously to temperature changes with discrete thresholds for activation [21, 22]. These do not represent mutually exclusive possibilities; as in the peripheral nervous system, multiple temperature-sensitive events might contribute to hypothalamic thermotransduction.

10.3

Molecular Characterization of a Heat-gated Ion Channel, TRPV1

The first “temperature receptor” to be characterized at a molecular level was vanilloid receptor 1 (VR1), later renamed TRPV1 [23] (Fig. 10.1, bottom panel; Fig. 10.3). This ion-channel protein was initially identified as a receptor for capsaicin, the hydrophobic vanilloid compound that lends “hot” peppers their pungency. Accordingly, non-neuronal cells engineered to express TRPV1 exhibit a robust influx of sodium and calcium ions in response to capsaicin administration. Consistent with its responsiveness to capsaicin, TRPV1 was found to be most highly expressed in a subset of small- to medium-diameter sensory neurons in dorsal root and trigeminal ganglia that appeared to correspond to C-fiber and Ad nociceptors. Further functional analysis of TRPV1, however, led to the observation that this channel could be activated by a number of physical and chemical stimuli, in addition to vanilloid compounds. Among these was an elevation in ambient temperature [23, 24]. In fact, heterologously expressed TRPV1 exhibits a temperature-response profile very similar to that of the heat-gated channels found in cultured sensory neurons, with a threshold of 42 8C. Furthermore, capsaicin- and heat responsiveness are highly correlated among cultured sensory neurons, and the TRPV1 antagonist, capsazepine, can inhibit responses to both stimuli [18, 19, 25]. Together, these data suggest that TRPV1 is a likely mediator of thermal nociception. Validation of this notion has come from the analysis of TRPV1 knockout mice [26, 27]. These animals are viable and fertile, with no obvious anatomical or behavioral deficits. However, sensory neurons cultured from them exhibit a dramatic decrease in the prevalence of heat responsiveness, and the few heat-responsive neurons that remain are characterized by an activation threshold (53 8C) much higher than that of TRPV1 (see below). In behavioral assays of thermal avoidance, TRPV1 knockout mice exhibit longer response latencies than wild-type controls but do eventually respond. Even more strikingly, in the absence of TRPV1, inflammatory thermal hyperalgesia – the augmented sensitivity to heat stimuli that results from tissue inflammation – is nearly eliminated. In contrast, thermal hyperalgesia resulting from nerve ligation is

239

240

10 Ion Channels and Thermotransduction

normal in these animals. Together, these findings support the participation of TRPV1 in heat-evoked pain responses but also point towards the existence of additional mechanisms of heat transduction. TRPV1 is a member of the transient receptor potential (TRP) family of ion-channel proteins. This family, whose prototypical member was identified as a component of the Drosophila melanogaster phototransduction machinery, is now known to have at least 25 mammalian members subdivided into 6 major subfamilies [28] (see Tab. 8.3). Each TRP channel subunit consists of six transmembrane domains, cytoplasmic amino and carboxyl termini, and a pore-loop region between transmembrane domains

+&2

D

FDSVDLFLQ

2



&XUUHQW'HQVLW\S$S)

A 1 +

+2

B

N

a a

a

 

 

  

TRPV1









Control

         

a

C

7HPSHUDWXUH &

TRPV1

C

E

TRPV1 +/+

7DLOZLWKGUDZDO ODWHQF\V

 



TRPV1 -/-



 

 

Fig. 10.3 The capsaicin receptor TRPV1 is a heatgated ion channel required for the normal perception of heat-evoked pain. (A) Chemical structure of capsaicin, the principal pungent ingredient in “hot” peppers. (B) Predicted topological structure of the capsaicin receptor TRPV1 (a, ankyrin repeat domain). (C) In situ hybridization illustrates the selective expression of TRPV1 in a subpopulation of small- to medium-diameter sensory neurons in rat dorsal root ganglion (arrow). Arrowhead indicates large-diameter, TRPV1-negative neuron. (D) Heatevoked activation of recombinant TRPV1 expressed heterologously in HEK293 human embryonic kid-







  7HPSHUDWXUH &



ney cells, assayed by whole-cell voltage clamp at –60 mV. Control cells were transfected with plasmid vector alone. (E) Latency for tail withdrawal from a water bath set at the indicated temperatures in wildtype mice (black bars) or mice lacking TRPV1 (gray bars). Note longer withdrawal latencies in the TRPV1-null mice at higher stimulus temperatures. (Panels (C) and (D) reproduced with permission from [23] and [31], respectively, Copyright 1997 and 1999, Nature Publishing Group. Panel (E) reproduced with permission from [26], Copyright 2000, The American Association for the Advancement of Science.)

10.4 TRPV2 Is an Ion Channel Activated by Extremely Hot Temperatures

5 and 6. Some family members, including TRPV1, contain several ankyrin repeat domains in their amino termini. These 33-amino-acid motifs are found in a variety of proteins and have been shown to mediate interactions with other protein or nonprotein macromolecules. Their functions in the TRP family, however, have yet to be determined. Four TRP channel subunits are thought to comprise a functional ion channel, with a single central pore lined by the pore loops. Many of the TRP channels are nonselective cation channels, though some exhibit a significant preference for divalent cations. Also, while some exist as homomultimers, others form heteromultimeric channels [29, 30]. As described below, the identification of TRPV1 as a temperature-gated channel has allowed the application of homology-based approaches to the identification of additional TRP channels that bear this peculiar property.

10.4

TRPV2 Is an Ion Channel Activated by Extremely Hot Temperatures

One temperature-sensitive TRPV1 homologue is TRPV2, a protein with 50 % identity to TRPV1 [31] (Fig. 10.1). TRPV2 is expressed much more broadly than TRPV1, in both neuronal and non-neuronal tissues. In rat dorsal root ganglia, TRPV2 is expressed most highly within a subpopulation of 16 % of neurons. These neurons have medium- to large-diameter cell bodies and appear to correspond to Ad nociceptors. Other sensory neuron subpopulations either fail to express detectable TRPV2 protein or express it at a much lower level. Functional expression of TRPV2 in non-neuronal cells has revealed that it is insensitive to vanilloid compounds but can be activated by very high temperatures (i.e., >53 8C) (Fig. 10.1, bottom panel). Approximately 10 % of cultured rat or mouse sensory ganglion neurons respond to heat in this temperature range, and these cells tend to be capsaicin-insensitive, with larger cell bodies than neurons exhibiting the more moderate heat threshold of 42 8C [25]. In addition, it is this high-threshold population whose heat responsiveness appears to have been preserved in TRPV1 knockout mice [26]. Neurophysiological studies have demonstrated that a subset of myelinated nociceptors in vivo exhibit a similar threshold for activation by heat [14, 15]. Together, these data suggest that TRPV2 may mediate the thermal responsiveness of these so-called Type I Ad nociceptors. Other studies have demonstrated that TRPV2 in non-neuronal cells can be activated at room temperature, either by the activation of growth factor receptors or by cell stretch-mediated TRPV2 recruitment to the cell surface [32, 33]. It remains to be seen whether such relocalization plays a role in heat-evoked TRPV2 activation in sensory neurons.

241

242

10 Ion Channels and Thermotransduction

10.5

TRPV3 and TRPV4 Are Warmth-activated Channels

Two additional TRPV homologues exhibit heat responsiveness at temperatures slightly lower than those required to activate TRPV1 or TRPV2 (Fig. 10.1, bottom panel). TRPV3 is encoded by a gene located adjacent to the trpv1 locus [34–37]. This protein, when expressed heterologously, can be activated by warm temperatures, with a threshold of approximately 32–39 8C, and continues to encode temperature up to at least 50 8C. TRPV3 is also capable of forming heteromultimers with TRPV1 and appears thereby to be capable of enhancing the latter protein’s responsiveness to capsaicin in heterologous expression systems [37]. The responsiveness of TRPV1/TRPV3 heteromultimers to heat has yet to be explored. TRPV4 was identified originally as an ion channel that could be activated by hypotonic cell swelling [38–41]. Later studies revealed responsiveness to a synthetic phorbol ester (4a-phorbol didecanoate) and certain naturally occurring cytochrome P450 metabolites of arachidonic acid [42, 43]. In addition, TRPV4 can be activated by warm temperatures, albeit with a threshold (25–34 8C) slightly lower than that of TRPV3 [44, 45]. While there is evidence for the expression of both TRPV3 and TRPV4 in sensory neurons, the most impressive cutaneous expression of both proteins occurs in epidermal keratinocytes [35, 44]. Indeed, keratinocyte-derived cell lines express both proteins and exhibit warmth-evoked opening of a nonselective cation channel with properties resembling those of TRPV4 [46]. These findings raise the intriguing notion that TRPV3 and TRPV4 might contribute to the perception of warm temperatures not only in sensory neurons, but also in keratinocytes. According to this model, which remains untested, keratinocyte activation would result in the release of a soluble signal that is in turn recognized by epithelial sensory nerve terminals. Another location where TRPV4 might participate in thermotransduction is in the anterior/preoptic hypothalamus, which contains neurons sensitive to both warmth and osmolarity [44]. Despite their similarities, TRPV3 and TRPV4 are not functionally identical. For example, repetitive heating of TRPV3 results in a response that sensitizes, with progressively larger response amplitudes [36, 37]. In contrast, heat causes a progressive desensitization of TRPV4 [44, 45]. Also, while TRPV3 is reportedly expressed in the basal keratinocyte layer [35], TRPV4 immunoreactivity is most pronounced in more superficial, postmitotic layers [44]. Finally, the heat responsiveness of TRPV4, unlike that of TRPV3, can be modulated by osmolarity [44]. Together, these findings suggest the existence of distinct physiological roles for these two warmth-gated channels.

10.6

TRPM8 and ANKTM1 Are Activated by Cool and Cold Temperatures, Respectively

Recently, it was demonstrated that cold, like heat, can activate a subpopulation of sensory neurons in culture [47–55]. This activation appears to have a complex ionic basis, involving both the activation of nonselective cation channels and the inhibition of potassium channels, either of which can lead to membrane depolarization. In addi-

10.7 Non-TRP Channels Implicated in Mammalian Temperature Sensation

tion, like heat-evoked responses, cold-evoked responses are heterogeneous in their temperature dependence, with both low-threshold (30 8C) and high-threshold (20 8C) responses being evident in histologically distinct neuronal subpopulations (Fig. 10.1). Just as capsaicin can produce a perception of “burning” pain, certain chemical agents, such as menthol, evoke a cooling sensation without an objective reduction in skin temperature [56]. This observation, in parallel with candidate-based characterization of TRP channels, led to the identification of TRPM8, an ion channel that could be activated alternatively by mildly cold temperatures (<25 8C) or by exposure to either menthol or a chemically distinct “cooling compound,” icilin [52, 57]. TRPM8 was originally identified as a protein that is upregulated in prostate cancer cells [58]. However, recognition of its responsiveness to cold led to the demonstration that it is also expressed within a subset of small-diameter sensory neurons in dorsal root and trigeminal ganglia. TRPM8 is in a subfamily (TRPM) distinct from the TRPV channels described above. TRPM8 has a much longer cytoplasmic amino terminal domain than TRPV channels and lacks ankyrin repeats. Another cold-gated ion channel that was identified by candidate screening is ANKTM1 [55]. This channel, a member of the TRPN subfamily, was originally identified in human fibroblasts as a protein with 14 ankyrin repeat domains and a channel-like structure [59]. Later analysis revealed it to be related to the Drosophila NompC protein, which has been implicated in mechanosensation [60]. Interestingly, ANKTM1 was found to be responsive to temperatures even colder (threshold <20 8C) than those required to activate TRPM8. Analysis of the expression patterns of these two cold-gated ion channels has revealed that they are expressed in mutually exclusive subpopulations of small-diameter sensory neurons, with ANKTM1 more likely than TRPM8 to be co-expressed with TRPV1. The prevailing hypothesis is that TRPM8 is involved in the perception of innocuous cool temperatures, while ANKTM1 is involved in cold-evoked pain sensation [55]. Moreover, while the temperature-response characteristics of these two molecules are quite similar to those of cultured cold-sensitive neurons (Fig. 10.1), none of these in vitro responses faithfully recapitulates the properties of cold thermoreceptors or cold nociceptors in vivo. This apparent discrepancy raises the possibility that the precise temperature-response properties of TRPM8 and ANKTM1 are altered by modulatory factors in vivo and/or are complemented by additional cold-transduction mechanisms.

10.7

Non-TRP Channels Implicated in Mammalian Temperature Sensation

In addition to the TRP channels described above, a number of other ion channels have been shown to be modulated by temperature in interesting ways. TREK1, a member of the two-pore potassium channel family, is activated by warm temperatures [61]. As a result, cooling of TREK1-expressing cells results in a dramatic decrease in potassium efflux under physiological conditions. Several members of the BNaC/ASIC family of

243

244

10 Ion Channels and Thermotransduction

sodium-selective ion channels, which are thought to be either constitutively active or activated by low extracellular pH, exhibit enhanced current responses at cold temperatures [62]. Both TREK1 and cold-sensitive BNaC/ASIC channels are expressed in sensory neurons. However, their relevance to temperature sensation has yet to be determined. As described above, there is also evidence for the involvement of voltage-gated ion channels in temperature detection, particularly in the hypothalamus. For example, temperature-dependent reactivation of certain voltage-gated potassium channels may regulate temperature-dependent firing rates of hypothalamic warmth-sensitive neurons [20]. In addition, a voltage-gated sodium current with a steep temperature dependence has been observed in this brain region, though the molecular identity of the responsible channel has yet to be established [63]. Finally, there may be a role for the ATP-gated ion-channel subtype P2X3 in thermosensation. Mice lacking this channel were found to have an unexpected defect in cutaneous warmth-evoked activation of second-order neurons in the spinal cord dorsal horn [64]. It is unclear whether this defect reflects a direct role for P2X3 in warmth transduction or a modulatory role in signaling thermotransduction events to the spinal cord. One attractive possibility, however, is that ATP represents the hypothetical messenger that signals warmth from keratinocytes to neighboring sensory nerve terminals [35].

10.8

Temperature-sensing Proteins in Non-mammalian Species

As described above, thermosensation is not a function unique to mammals. Homeothermic (e.g., mammals and birds) and non-homeothermic (e.g., amphibians and fish) vertebrates alike evaluate temperature to guide physiological and/or behavioral homeostatic responses. An ortholog of TRPV1 has been identified in birds that is responsive to heat but insensitive to capsaicin [65]. In fact, mammals are the only class of animals that exhibit aversive responses to vanilloid compounds [66]. These and other data suggest that mammalian responsiveness to vanilloids evolved after the acquisition of heat sensitivity by TRPV1. Invertebrates also exhibit responsiveness to ambient temperature that may be manifested as thermotaxis or thermal avoidance. Genetic studies have begun to reveal the molecular basis of these behaviors in bacteria [67], nematodes [68–70], and fruit flies [70, 71]. Ion channels responsive to hot and cold temperatures have been observed in a number of primitive invertebrates, including sponges [73] and paramecia [74]. In the nematode Caenorhabditis elegans, two cyclic nucleotide-gated channels, TAX2 and TAX4, have been implicated in thermotaxis [69]. Whether these channels are directly modulated by temperature or whether they respond to signals generated by temperature-sensitive processes has yet to be established. In addition, C. elegans expresses several ion channels of the TRPV subfamily [30, 75]. However, no roles for these proteins in thermosensation have been demonstrated. Two TRP channels have been implicated in Drosophila thermotransduction. The first, Painless, is structurally similar to

10.9 Mechanisms of Temperature Transduction

mammalian ANKTM1, with eight ankyrin repeat domains in its amino terminus [76]. Fly larvae lacking this protein are deficient in withdrawal responses to mechanical stimuli as well as to moderately hot temperatures (46 8C). However, no in vitro physiological studies of this protein have yet addressed whether recombinant Painless protein exhibits thermal responsiveness. A second candidate temperature-sensitive channel found in Drosophila is an ortholog of ANKTM1 [77]. Interestingly, in contrast to its mammalian counterpart, which is cold activated, Drosophila ANKTM1 is warmth activated, with a threshold of 31 8C.

10.9

Mechanisms of Temperature Transduction

One fundamental question regarding the temperature-transducing ion channels described above regards the molecular mechanism by which channel activity is regulated by thermal energy. Some temperature-sensitive TRP channels exhibit preserved temperature sensitivity in the context of an excised membrane patch, suggesting that soluble second messengers and cytoskeletal interactions are not important determinants of this property [24, 36, 45]. Single-channel recordings have also revealed that, in general, temperature “gating” of ion channels involves a temperature-dependent change in channel-open probability, with only a modest contribution from changes in unitary conduction amplitude [24, 45, 46, 78]. For example, while the open probability of TRPV1 at 22 8C is virtually zero, this parameter begins to increase at temperatures above approximately 35 8C and climbs precipitously after 43 8C, matching the whole-cell temperature-response curve of TRPV1-expressing cells. A similar pattern is exhibited by native heat-gated ion channels (Fig. 10.2, right panel). The distinct thermal response profiles exhibited by different temperature-gated ion channels, even when they are co-expressed in the same cell, further indicate that the primary protein sequence dictates, to a certain extent, a given channel’s response characteristics [31, 52]. However, there is growing evidence that thermosensitive channels do not act in isolation but rather respond to the combined influence of thermal and nonthermal environmental variables. For example, extracellular protonation and cytoplasmic phosphorylation of TRPV1 can augment its responsiveness to heat [24, 79, 80]. Furthermore, interaction of the cytoplasmic carboxyl terminus of TRPV1 with the membrane phospholipid phosphatidyl inositol bisphosphate (PIP2) appears to inhibit the thermal responsiveness of this channel [81, 82]. Accordingly, removal of the PIP2binding domain in TRPV1 results in a leftward shift in the TRPV1 temperature-response profile, while cleavage or sequestration of PIP2 produces both a leftward shift and a shallowing of temperature dependence. This property of TRPV1 is a likely contributor to the thermal hyperalgesia that results from the production of proinflammatory mediators such as bradykinin and nerve growth factor in injured skin. Whether phospholipid interaction accounts completely for the temperature sensitivity of TRPV1 and whether such a mechanism contributes to the gating of other thermosensitive TRP channels remains to be seen.

245

246

10 Ion Channels and Thermotransduction

Another issue concerns the very different physical nature of the stimuli that can activate temperature-sensitive TRP channels. For example, while the mechanistic relationship between the gating of heat- and cold-sensitive TRP channels is unclear, the fact that two homologous channel proteins, the mouse and fruit fly ANKTM1 orthologs, respond to cold and warm temperatures, respectively, suggests a certain degree of conservation between their thermal gating mechanisms, despite the expectation that heat and cold might produce opposite effects on protein structure [77]. In addition, it is worth noting that thermosensitivity, osmosensitivity, mechanosensitivity, and modulation by lipid molecules represent recurrent themes within the TRP channel family, particularly within the TRPV subfamily, and sometimes even in the same molecule. This curious association most likely reflects a common denominator in the way these channels respond to the distinct physical and chemical forces that impinge upon a cell. Regardless of its mechanistic basis, such polymodal responsiveness may explain why nature appears to have turned repeatedly to TRP channels to solve the wide variety of sensory challenges outlined in this textbook.

10.10

Conclusions

Temperature-sensitive ion channels of the TRP family constitute a diverse and multifunctional set of candidate thermotransduction molecules. The continued analysis of these channel proteins promises to provide valuable insights into how mammalian and non-mammalian species interpret and respond to their thermal environments. References 1

2

3

4

5

6

Blix, M. Experimentelle Bertrage zur Losung der Frage uber die spezifiche Energie der Hauptnerven., Z. Biol. 1884, 20, 141–156. Hensel, H., and Zotterman, Y. The response of the cold receptors to constant cooling., Acta Physiol. Scand. 1951, 22, 96–113. Hensel, H., and Zotterman, Y. Quantitative Beziehungen zwischen der Entladung einzelner Kaltefasern und der Temperatur., Acta Physiol. Scand. 1951, 23, 291–319. Dodt, E., and Zotterman, Y. The discharge of specific cold fibers at high temperatures (paradoxical cold.), Acta Physiol. Scand. 1952, 26, 358–365. Iggo, A. Cutaneous heat and cold receptors with slowly-conducting (C) afferent fibres., Quart. J. Exp. Physiol. 1959, 44, 362–370. Hensel, H., Iggo, A., and Witt, I. A quantitative study of sensitive cutaneous thermoreceptors with C afferent fibers, J. Physiol. (London) 1960, 153, 113–126.

7

8

9

10

11

Iriuchijima, J., and Zotterman, Y. The specificity of afferent cutaneous C fibres in mammals., Acta Physiol. Scand. 1960, 49, 267–278. Iggo, A. Cutaneous thermoreceptors in primates and sub-primates., J. Physiol. (London) 1969, 200, 403–430. Bessou, P., and Perl, E. R. Response of cutaneous sensory units with unmyelinated fibers to noxious stimuli., J. Neurophysiol. 1969, 32, 1025–1043. Darian-Smith, I., Johnson, K. O., and Dykes, R. “Cold” fiber population innervating palmar and digital skin of the monkey., J. Neurophysiol. 1973, 36, 325–346. Dubner, R., Sumino, R., and Wood, W. I. A peripheral “cold” fiber population responsive to innocuous and noxious thermal stimuli applied to the monkey’s face., J. Neurophysiol. 1975, 38, 1373–1389.

10.10 Conclusions 12

13

14

15

16

17

18

19

20

21

22

23

Hensel, H., and Iggo, A. Analysis of cutaneous warm and cold fibers in primates, Pflugers Arch. 1971, 329, 1–8. Simone, D., and Kajander, K. C. Responses of cutaneous A-fiber nociceptors to noxious cold, J. Neurophysiol. 1997, 77, 2049–2060. Dubner, R., Price, D. D., Beitel, R. E., and Hu, J. W. Peripheral neural correlates of behavior in monkey and human related to sensorydiscriminative aspects of pain. In Pain in the Trigeminal Region (Anderson, D. J., and B, M., Eds.) pp 57–66, Elsevier, Amsterdam (1977) Treede, R., Meyer, R. A., Srinivasa, R. N., and Campbell, J. N. Evidence for two different heat transduction mechanisms in nociceptive primary afferents innervating monkey skin, J. Physiol 1995, 483, 747–758. Cesare, P., and McNaughton, P. A novel heatactivated current in nociceptive neurons and its sensitization by bradykinin., Proc. Natl. Acad. Sci., U.S.A. 1996, 93, 15435–15439. Reichling, D. B., and Levine, J. D. Heat transduction in rat sensory neurons by calcium-dependent activation of a cation channel, Proc. Natl. Acad. Sci., U.S.A. 1997, 94, 7006–7011. Nagy, I., and Rang, H. P. Similarities and differences between the responses of rat sensory neurons to noxious heat and capsaicin, J. Neurosci. 1999, 19, 10647–10655. Kirschstein, T., Busselberg, D., and Treede, R. D. Coexpression of heat-evoked and capsaicinevoked inward currents in acutely dissociated rat dorsal root ganglion neurons, Neurosci. Lett. 1997, 231, 33–36. Boulant, J. A. Hypothalamic neurons: Mechanisms of sensitivity to temperature, Ann. N.Y. Acad. Sci. 1998, 856, 108–115. Hori, A., Minato, K., and Kobayashi, S. Warming-activated channels of warm-sensitive neurons in rat hypothalamic slices., Neurosci. Lett. 1999, 275, 93–96. Abe, J., Okazawa, M., Adachi, R., Matsumura, K., and Kobayashi, S. Primary cold-sensitive neurons in acutely dissociated cells of rat hypothalamus, Neurosci. Lett. 2003, 342, 29–32. Caterina, M. J., Schumacher, M. A., Tominaga, M., Rosen, T. A., Levine, J. D., and Julius, D. The capsaicin receptor: a heat-activated ion channel in the pain pathway., Nature 1997, 389, 816–824.

24

25

26

27

28

29

30

31

32

33

34

Tominaga, M., Caterina, M. J., Malmberg, A. B., Rosen, T. A., Gilbert, H., Skinner, K., Raumann, B. E., Basbaum, A. I., et al. The cloned capsaicin receptor integrates multiple pain-producing stimuli, Neuron 1998, 21, 1–20. Nagy, I., and Rang, H. Noxious heat activates all capsaicin-sensitive and also a sub-population of capsaicin-insensitive dorsal root ganglion neurons, Neurosci. 1999, 88, 995–997. Caterina, M. J., Leffler, A., Malmberg, A. B., Martin, W. J., Trafton, J., Petersen-Zeitz, K. R., Koltzenburg, M., Basbaum, A. I., et al. Impaired nociception and pain sensation in mice lacking the capsaicin receptor (see comments), Science 2000, 288, 306–13. Davis, J. B., Gray, J., Gunthorpe, M. J., Hatcher, J. P., Davey, P. T., Overend, P., Harries, M. H., Latcham, J., et al. Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia, Nature 2000, 405, 183–7. Montell, C. Physiology, phylogeny, and functions of the TRP superfamily of cation channels, Sci STKE 2001, 2001, RE1. Xu, X. Z. S., Li, H. S., Guggino, W. B., and Montell, C. Coassembly of TRP and TRPL produces a distinct store-operated conductance, Cell 1997, 89, 1155–1164. Tobin, D., Madsen, D., Kahn-Kirby, A., Peckol, E., Moulder, G., Barstead, R., Maricq, A., and Bargmann, C. Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans neurons, Neuron 2002, 35, 307–18. Caterina, M. J., Rosen, T. A., Tominaga, M., Brake, A. J., and Julius, D. A capsaicin receptor homologue with a high threshold for noxious heat, Nature 1999, 398, 436–441. Kanzaki, M., Zhang, Y. Q., Mashima, H., Li, L., Shibata, H., and Kojima, I. Translocation of a calcium-permeable cation channel induced by insulin-like growth factor-I., Nat Cell Biol 1999, 1, 165–170. Muraki, K., Iwata, Y., Katanosaka, Y., Ito, T., Ohya, S., Shigekawa, M., and Imaizumi, Y. TRPV2 Is a Component of Osmotically Sensitive Cation Channels in Murine Aortic Myocytes, Circ Res 2003. Peng, J. B., Chen, X. Z., Berger, U. V., Vassilev, P. M., Tsukaguchi, H., Brown, E. M., and Hediger, M. A. Molecular cloning and characterization of a channel-like transporter mediating intestinal calcium absorption, J Biol Chem 1999, 274, 22739–46.

247

248

10 Ion Channels and Thermotransduction 35

36

37

38

39

40

41

42

43

44

45

Peier, A. M., Reeve, A. J., Andersson, D. A., Moqrich, A., Earley, T. J., Hergarden, A. C., Story, G. M., Colley, S., et al. A heat-sensitive TRP channel expressed in keratinocytes, Science 2002, 296, 2046–9. Xu, H., Ramsey, I. S., Kotecha, S. A., Moran, M. M., Chong, J. A., Lawson, D., Ge, P., Lilly, J., et al. TRPV3 is a calcium-permeable temperature-sensitive cation channel, Nature 2002, 418, 181–6. Smith, G. D., Gunthorpe, M. J., Kelsell, R. E., Hayes, P. D., Reilly, P., Facer, P., Wright, J. E., Jerman, J. C., et al. TRPV3 is a temperaturesensitive vanilloid receptor-like protein, Nature 2002, 418, 186–90. Strotmann, R., Harteneck, C., Nunnenmacher, K., Schultz, G., and Plant, T. D. OTRPC4, a nonselective cation channel that confers sensitivity to extracellular osmolarity, Nature Cell Biology 2000, 2, 695–702. Liedtke, W., Choe, Y., Marti-Renom, M. A., Bell, A. M., Denis, C. S., Sali, A., Hudspeth, A. J., Friedman, J. M., et al. Vanilloid receptorrelated osmotically activated channel (VROAC), a candidate vertebrate osmoreceptor, Cell 2000, 103, 525–535. Wissenbach, U., Bodding, M., Freichel, M., and Flockerzi, V. Trp12, a novel Trp related protein from kidney, FEBS Lett. 2000, 485, 127–134. Delany, N. S., Hurle, M., Facer, P., Alnadaf, T., Plumpton, C., Kinghorn, I., See, C. G., Costigan, M., et al. Identification and characterization of a novel human vanilloid receptor-like protein, VRL-2, Physiol. Genomics 2001, 4, 165–174. Watanabe, H., Davis, J. B., Smart, D., Jerman, J. C., Smith, G. D., Hayes, P., Vriens, J., Cairns, W., et al. Activation of TRPV4 channels (hVRL-2/mTRP12) by phorbol derivatives, J Biol Chem 2002, 277, 13569–77. Watanabe, H., Vriens, J., Prenen, J., Droogmans, G., Voets, T., and Nilius, B. Anandamide and arachidonic acid use epoxyeicosatrienoic acids to activate TRPV4 channels, Nature 2003, 424, 434–8. Guler, A. D., Lee, H., Iida, T., Shimizu, I., Tominaga, M., and Caterina, M. Heat-evoked activation of the ion channel, TRPV4, J Neurosci 2002, 22, 6408–14. Watanabe, H., Vriens, J., Suh, S. H., Benham, C. D., Droogmans, G., and Nilius, B. Heatevoked activation of TRPV4 channels in an HEK293 cell expression system and in native mouse aorta endothelial cells, J Biol Chem 2002, 26, 26.

46

47

48

49

50

51

52

53

54

55

56 57

58

Chung, M. K., Lee, H., and Caterina, M. J. Warm temperatures activate TRPV4 in mouse 308 keratinocytes, J Biol Chem 2003, 278, 32037–46. Suto, K., and Gotoh, H. Calcium signaling in cold cells studied in cultured dorsal root ganglion neurons, Neuroscience 1999, 92, 1131–5. Okazawa, M., Takao, K., Hori, A., Shiraki, T., Matsumura, K., and Kobayashi, S. Ionic basis of cold receptors acting as thermostats, J Neurosci 2002, 22, 3994–4001. Reid, G., and Flonta, M. Cold transduction by inhibition of a background potassium conductance in rat primary sensory neurones, Neurosci Lett 2001, 297, 171–4. Reid, G., and Flonta, M. L. Physiology. Cold current in thermoreceptive neurons, Nature 2001, 413, 480. Reid, G., and Flonta, M. L. Ion channels activated by cold and menthol in cultured rat dorsal root ganglion neurones, Neurosci Lett 2002, 324, 164–8. McKemy, D. D., Neuhausser, V. M., and Julius, D. Identification of a cold receptor reveals a general role for TRP channels in thermosensation, Nature 2002, 416, 52–58. Nealen, M. L., Gold, M. S., Thut, P. D., and Caterina, M. J. TRPM8 mRNA is expressed in a subset of cold-responsive trigeminal neurons from rat, J Neurophysiol 2003, 90, 515–20. Thut, P. D., Wrigley, D., and Gold, M. S. Cold transduction in rat trigeminal ganglia neurons in vitro, Neuroscience 2003, 119, 1071–83. Story, G. M., Peier, A. M., Reeve, A. J., Eid, S. R., Mosbacher, J., Hricik, T. R., Earley, T. J., Hergarden, A. C., et al. ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures, Cell 2003, 112, 819–29. Goldscheider, A. (1898) Gesammelte abhandlungen, Johann Ambrosius Barth, Leipzig. Peier, A. M., Moqrich, A., Hergarden, A. C., Reeve, A. J., Andersson, D. A., Story, G. M., Earley, T. J., Dragoni, I., et al. A TRP channel that senses cold stimuli and menthol, Cell 2002, 108, 705–715. Tsavaler, L., Shapero, M. H., Morkowski, S., and Laus, R. Trp-p8, a novel prostate-specific gene, is up-regulated in prostate cancer and other malignancies and shares high homology with transient receptor potential calcium channel proteins, Cancer Res 2001, 61, 3760–9.

10.10 Conclusions 59

60

61

62

63

64

65

66

67

68

69

70

71

Jaquemar, D., Schenker, T., and Trueb, B. An ankyrin-like protein with transmembrane domains is specifically lost after oncogenic transformation of human fibroblasts, J Biol Chem 1999, 274, 7325–33. Walker, R. G., Willingham, A. T., and Zuker, C. S. A Drosophila mechanosensory transduction channel, Science 2000, 287, 2229–34. Maingret, F., Lauritzen, I., Patel, A. J., Heurteaux, C., Reyes, R., Lesage, F., Lazdunski, M., and Honore, E. TREK-1 is a heatactivated background K(+) channel, Embo J 2000, 19, 2483–91. Askwith, C. C., Benson, C. J., Welsh, M. J., and Snyder, P. M. DEG/ENaC ion channels involved in sensory transduction are modulated by cold temperature, Proc Natl Acad Sci U S A 2001, 98, 6459–63. Kiyohara, T., Hirata, M., Hori, T., and Akaike, N. Hypothalamic warm-sensitive neurons possess a tetrodotoxin-sensitive sodium channel with a high Q10, Neurosci Res (N Y) 1990, 8, 48–53. Souslova, V., Cesare, P., Ding, Y., Akopian, A. N., Stanfa, L., Suzuki, R., Carpenter, K., Dickenson, A., Boyce, S., Hill, R., NebeniusOosthuizen, D., Smith, A. J. H., Kidd, E. J., and Wood, I. N. Warm-coding deficits and aberrant inflammatory pain in mice lacking P2X3 receptors. Nature 2000. 407, 1015–1017. Jordt, S. E., Tominaga, M., and Julius, D. Acid potentiation of the capsaicin receptor determined by a key extracellular site, Proc Natl Acad Sci U S A 2000, 97, 8134–8139. Szallasi, A. The vanilloid (capsaicin) receptor: Receptor types and species specificity, Gen. Pharmac. 1994, 25, 223–243. Imae, Y. In Sensing and Response in Microorganisms (Eisenbach, M., and Malaban, M., Eds.) pp 73, Elsevier Science Publishing Company, Inc., New York (1985) Hedgecock, E. M., and Russell, R. L. Normal and mutant thermotaxis in the nematode Caenorhabditis elegans, Proc Natl Acad Sci U S A 1975, 72, 4061–5. Mori, I., and Ohshima, Y. Neural regulation of thermotaxis in Caenorhabditis elegans, Nature 1995, 376, 344–8. Wittenburg, N., and Baumeister, R. Thermal avoidance in Caenorhabditis elegans: an approach to the study of nociception, Proc Natl Acad Sci U S A 1999, 96, 10477–82. Sayeed, O., and Benzer, S. Behavioral genetics of thermosensation and hygrosensation in Drosophila, Proc Natl Acad Sci U S A 1996, 93, 6079–84.

72

73

74

75

76

77

78

79

80

81

82

Liu, L., Yermolaieva, O., Johnson, W. A., Abboud, F. M., and Welsh, M. J. Identification and function of thermosensory neurons in Drosophila larvae, Nat Neurosci 2003, 6, 267–73. Zocchi, E., Carpaneto, A., Cerrano, C., Bavestrello, G., Giovine, M., Bruzzone, S., Guida, L., Franco, L., et al. The temperature-signaling cascade in sponges involves a heat-gated cation channel, abscisic acid, and cyclic ADPribose, Proc Natl Acad Sci U S A 2001, 98, 14859–64. Kuriu, T., Nakaoka, Y., and Oosawa, Y. Coldsensitive Ca2+ influx in Paramecium, J. Membrane Biol. 1996, 154, 163–167. Colbert, H. A., Smith, T. L., and Bargmann, C. I. Osm9, a novel protein with structural similarity to ion channels, is required for olfaction, mechanosensation and olfactory adaptation in Caenorhabditis elegans, J. Neurosci. 1997, 17, 8259–8269. Tracey, W. D., Jr., Wilson, R. I., Laurent, G., and Benzer, S. painless, a Drosophila gene essential for nociception, Cell 2003, 113, 261–73. Viswanath, V., Story, G. M., Peier, A. M., Petrus, M. J., Lee, V. M., Hwang, S. W., Patapoutian, A., and Jegla, T. Opposite thermosensor in fruitfly and mouse, Nature 2003, 423, 822–3. Cesare, P., Moriondo, A., Vellani, V., and McNaughton, P. A. Ion channels gated by heat, Proc. Natl. Acad. Sci. 1999, 96, 7658–7663. Numazaki, M., Tominaga, T., Toyooka, H., and Tominaga, M. Direct phosphorylation of capsaicin receptor VR1 by protein kinase Cepsilon and identification of two target serine residues, J Biol Chem 2002, 277, 13375–8. Bhave, G., Zhu, W., Wang, H., Brasier, D. J., Oxford, G. S., and Gereau, R. W. t. cAMPdependent protein kinase regulates desensitization of the capsaicin receptor (VR1) by direct phosphorylation, Neuron 2002, 35, 721–31. Chuang, H. H., Prescott, E. D., Kong, H., Shields, S., Jordt, S. E., Basbaum, A. I., Chao, M. V., and Julius, D. Bradykinin and nerve growth factor release the capsaicin receptor from PtdIns(4,5)P2-mediated inhibition, Nature 2001, 411, 957–62. Prescott, E. D., and Julius, D. A modular PIP2 binding site as a determinant of capsaicin receptor sensitivity, Science 2003, 300, 1284–8.

249

251

11

Pain Transduction: Gating and Modulation of Ion Channels Peter A. McNaughton

11.1

Introduction

The sensation of pain informs us that potentially or actually damaging stimuli (noxious stimuli) are impinging on our bodies. All animals, and even many single-celled organisms that could scarcely be dignified by the name animal, show some form of avoidance reaction in response to noxious stimuli. Pain has therefore been a feature of many forms of life since early in evolution, and we would expect pain to be a highly evolved and complex sensation, responsive to a multitude of harmful stimuli – either those present in the external environment or those generated within the organism itself. The word “pain” causes some difficulties because it covers a multitude of quite separate concepts. Pain refers to the processes involved in detection of the physical effects of an injury (e.g., the pain of a fracture); to the perceptual and emotional reaction to an injury (the feeling of pain); and to the similar sensations experienced in response to a purely emotional event (e.g., the pain of parting). Sherrington [1] removed the emotional connotations of the word “pain” from the physical process of detection of a painful stimulus by coining the word “nociceptor” (a receptor for noxious stimuli) to refer to those primary sensory neurons that are activated by stimuli that we would regard as painful. The existence of nociceptors as a separate class of sensory neurons was doubted for many years; it was proposed instead that the sensation of pain could result from the strong stimulation of sensory receptors responsible for detecting non-painful stimuli, an idea which was given support by the discovery of second-order neurons in the deeper layers of the dorsal horn of the spinal cord that respond to a wide range of stimuli, both non-noxious and noxious [2]. However, Ed Perl, by recording directly from primary afferent nerve fibers [3], was able to demonstrate clearly that a separate class of primary sensory afferents responding only to high-threshold stimulation did indeed exist and had properties quite distinct from those of receptors responding to non-noxious stimulation. Thus, the existence of a distinct class of nociceptive neurons, as postulated by Sherrington, was finally confirmed. Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

252

11 Pain Transduction: Gating and Modulation of Ion Channels

In the field of pain, however, few things are simple, and most of those working in the field would agree that: 1. Even clearly non-noxious stimulation can cause a sensation of pain. Sometimes this can be attributed to the excitation of nociceptors whose threshold has been lowered into the region of non-noxious stimuli by the process of sensitization (see further discussion below), but in other circumstances it is clearly due to a painful sensation generated by the stimulation of non-nociceptive primary sensory neurons. 2. Non-noxious stimulation can modify the sensation of pain, and in particular mild mechanical or thermal stimulation can reduce or abolish the sensation of pain. 3. Even strong stimulation of nociceptors, caused, for example, by a severe injury, can in some circumstances (e.g., stressful situations such as battle) fail to cause any sensation of pain at all. This article will largely avoid discussing these overlaps between nociceptive and nonnociceptive function, most of which arise from processes occurring at a spinal or higher level, and will focus on the mechanisms by which painful stimuli are transduced in primary nociceptive nerve terminals. A notable exception will be the process of sensitization, mentioned under (1) above, which has its origin in processes occurring within peripheral nociceptive nerve terminals. Nociceptors, like all somatic sensory neurons, are “long” neurons in which the sensory terminal must be depolarized and action potentials elicited in order to communicate to the distant central synaptic terminals. The generator current activated in nociceptive nerve terminals by all noxious stimuli must therefore be inward, as it is in all somatic sensory neurons. In many other respects, however, nociceptive neurons differ from other sensory neurons, although most of the functional differences are readily explicable in terms of the tasks that nociceptors are designed to carry out. Most obviously, a painful stimulus is almost by definition large in magnitude, so the cellular amplification pathways that form a prominent part of (for instance) olfactory and visual sensory transduction are not often seen in nociceptive transduction (most stimuli in fact act directly to gate an ion channel). Our current understanding of ion channels important in nociception is discussed below. Secondly, while most sensory receptors respond with exquisite sensitivity to only a single sensory modality (for instance, a narrow band of wavelengths of light in the case of visual receptors), nociceptors are typically polymodal, responding to a wide range of stimuli that may include strong mechanical stimuli, heat, extreme cold, and a range of noxious chemical stimuli. Polymodal behavior makes sense in the context of pain, where the organism will be more concerned with detecting a noxious stimulus and taking appropriate avoiding action, rather than with analyzing the precise nature of the stimulus. Finally, in all other sensory receptors, adaptation is seen in response to the maintained presentation of a stimulus, a property that enables the receptors to operate over a wide range of stimulus intensities. The response of nociceptors, in contrast, typically increases, or sensitizes, in response to a prolonged or intense stimulus, a property that ensures that a noxious stimulus is not ignored by the organism. Our current understanding of the molecular basis of sensitization is discussed below.

11.2 Ion Channels Gated by Noxious Stimuli

11.2

Ion Channels Gated by Noxious Stimuli 11.2.1

Ion Channels Gated by Noxious Heat

Our current understanding of temperature-sensitive ion channels has been reviewed in Chapter 10 of this book by Michael Caterina and will be presented only in outline here. Many nociceptors, both unmyelinated and myelinated, respond to elevated temperatures with an action potential frequency that increases exponentially as the temperature increases above ca. 45 8C, which corresponds closely to the temperature at which a human observer would report a transition of the quality of the sensation from pleasant warmth to mild heat pain [3–5]. An inward membrane current with similar properties can be recorded from a subset of small sensory neurons [6]. The ion channel through which the current passes is activated at temperatures above 45 8C, turns on with a delay of ca. 35 ms following a temperature change, and is highly permeable to calcium and to monovalent cations. There is also a second neuronal population with a heat threshold of around 51 8C [7]. The heat-sensitive ion channel now thought to be responsible for the 45 8C threshold heat response was cloned by an expression cloning strategy using responsiveness to capsaicin, the active extract of chili peppers, as a probe [8]. The 6TM channel that was obtained is a member of the Trp channel family, now named TRPV1 (VR1 originally). The channel passes an inward current in response to temperatures above 45 8C, and is also activated by low pH [9]. The sites of binding of capsaicin (intracellular loop 2–3) and protons (extracellular, adjacent to the channel mouth) have now been identified [10, 11], but the molecular rearrangements involved in the gating of the channel by heat remain mysterious. In most aspects so far investigated, the properties of TRPV1 correspond closely to those seen in the corresponding population of nociceptive neurons (i.e., those with a heat threshold at ca. 45 8C). A second heat-sensitive channel named TRPV2 (VRL1 originally), showing substantial homology to TRPV1, was subsequently cloned [12]. This channel is capsaicin-insensitive, has a higher heat threshold, and seems likely to be the channel expressed in the high-threshold population of nociceptors. Mice in which TRPV1 has been deleted are still sensitive to noxious heat, albeit at a slightly higher temperature [13, 14]. The explanation may be that they sense heat using TRPV2, but it is possible that other, as yet undiscovered, noxious heat sensors also contribute. The most obvious defect in the TRPV1-/- mice was that the heat hyperalgesia observed following inflammation was almost completely abolished (discussed further below).

253

254

11 Pain Transduction: Gating and Modulation of Ion Channels

11.2.2

Ion Channels Gated by Noxious Cold

Most neurons that can be identified as nociceptors by other criteria also respond to noxious cold temperatures (below around 20 8C) but not to innocuous cold [15]. The ion channel likely to be responsible was identified by searching the human genome database for channels containing a 6TM motif plus the N-terminal domain ankyrin repeat characteristic of Trp channels [16]. ANKTM1 (also called TRPA1) is activated by cooling below 17 8C but is not activated by menthol, which potently activates TRPM8, the channel responsible for the sensation of non-noxious coolness [17, 18]. ANKTM1 is expressed in a subset of the neurons that express TRPV1, which explains why noxious cold is described as “burning.”

11.2.3

Ion Channels Gated by Acid

Acid-sensing ion channels (ASICs) have been discussed in Chapter 3 of this book by Ken Cushman and Ed McCleskey, and, as with the temperature-sensitive ion channels, only a summary of those aspects relevant for nociception will be presented here. The first ASIC (now called ASIC1) was cloned in a search for additional members of the amiloride-sensitive ion channel family, which includes a number of epithelial Na channels [19]. ASIC1 generates a transient inward current in response to even mild acidification and is expressed in sensory neurons, suggesting a role in signaling the acidification associated with anoxia and metabolic overload. The main difficulties with this idea are the extremely transient nature of the ASIC1 response and the powerful adaptation to even mild acidification, which makes a role in sustained signaling of acidotic pain seem improbable. ASIC1 is also widely expressed in neurons of the CNS, where its function is unknown. A second member of the same family, BNC1 [20], is now recognized as acid-sensitive and has been renamed ASIC2. Like ASIC1, it is widely expressed in neurons, including small sensory neurons, but it generates a more sustained inward current in response to acidification, which may make it a more promising candidate for a physiological sensor of the pain associated with acidification. A third member of the family, ASIC3 [21], was cloned from its homology with ASIC1 and is expressed mainly in small sensory neurones (it was initially named DRASIC—DRG specific ASIC—in recognition of this). The response of ASIC3 to acidification is, like ASIC1, extremely transient, though there is some evidence for a sustained component activated at very low pH [22]. Finally, the last member of the family as currently known, ASIC4, has strong homology with ASIC1 and is expressed in sensory neurons but does not respond to acid and thus seems unlikely to have a role in signaling acid pain [23]. The main alternative possible role for the ASIC family suggested to date is in mechanosensation, a suggestion that arises by analogy with the role of the related mec4 gene in mechanosensation in C. elegans. ASIC2a and ASIC3 mouse knockouts do show small alterations in non-noxious mechanosensation [24, 25], but the effects

11.2 Ion Channels Gated by Noxious Stimuli

are not large and the evidence for an involvement of the ASIC family in mechanosensation is not at present overwhelming. In support of this view, a recent study of mechanically gated ion currents in isolated sensory neurons from ASIC2 and ASIC3 knockout mice [26] found that these currents were similar to wild-type in both low-threshold neurons (presumed detectors of non-noxious mechanical stimulation) and in a class of neurons responding only to high-threshold mechanical stimulation (presumed mechanical nociceptors).

11.2.4

ATP-gated Ion Channels

It would be sensible for pain to be signaled by the release of the contents of damaged cells, in which case ATP is a prime candidate to activate nociceptive nerve terminals, as it is present at high concentrations within cells and is practically absent from the external medium. Injection of ATP does indeed cause pain [27]. ATP is known to act at two main classes of membrane receptors: ion channels of the P2X family and G protein-coupled receptors of the P2Y family. The first are implicated in direct activation of nociceptors and are discussed here, while P2Y receptors are involved in sensitization and are discussed below. Cloning of the first member of the P2X family opened up the way for homology cloning of seven other members, of which P2X3 would appear to be of particular significance for pain transduction, as it is expressed only in nociceptors [28]. ATP or its analogues rapidly activate an inward current through P2X3, but the current is only transient, a property that is, as was the case with the ASIC family, not obviously consistent with the need to generate a sustained sensation of pain in response to ongoing cell damage. Nociceptors from the autonomic nervous system generate a sustained inward current in response to ATP, a property that can be explained by their co-expression as a heteromeric channel of P2X2, which generates a sustained inward current, and of P2X3, which generates a transient inward current [29]. Most nociceptors from the somatic nervous system, however, express only P2X3 and generate a transient inward current in response to ATP and its analogues [30]. While these observations are internally self-consistent, they do not favor the idea that P2X3 receptors might signal somatic pain in a sustained manner. Consistent with this, acute pain signaling was relatively normal in P2X3-/- mice, although the detection of warm stimuli was affected, an observation that is still not fully explained [31, 32]. Interestingly, the main defect present in these mice was the absence of a normal bladder-emptying reflex, consistent with the idea that expression of P2X3 ion channels in nociceptive neurons of the bladder is essential for bladder emptying when a state of painful fullness has been reached [31].

255

256

11 Pain Transduction: Gating and Modulation of Ion Channels

11.2.5

Ion Channels Gated by Mechanical Stimuli

The first stimulus most people would think of in the context of pain is a strong mechanical stimulus, but ironically the ion channels involved in the generation of this most intuitively obvious of the pain sensations have proven the most elusive. Recent studies have identified inward membrane currents activated in isolated DRG neurons by direct mechanical stimulation with a probe [33, 34] or following stretching of the membrane by swelling in hypotonic solutions [35]. Two populations of neurons can be distinguished: those with a low threshold, which are capsaicin-insensitive and therefore are presumably non-nociceptive, and those with a higher threshold, which are in general sensitive to capsaicin and are therefore presumably nociceptive neurons [34]. These two populations are likely to correspond to the two main observed classes of in vivo mechanoreceptors, which detect non-noxious and noxious stimuli, respectively [36]. The molecular identity of the mechanically gated ion channel is still unclear, but two main possibilities are currently being considered. As discussed above in Section 11.2.3, the homology between the ASIC family and the MEC4 ion channel, known to be mechanically gated in C. elegans (see Chapter 1), has led to the proposal that ASICs form part of the mechanoreceptive complex, and there is some support for this proposal from gene knockout studies (see above). The mechanosensitive current is inhibited by calcium [34], a property that is reminiscent of the strong inhibition of ASIC-dependent currents by calcium [37], but in other respects the mechanosensitive currents do not obviously exhibit the properties of ion channels of the ASIC family. In particular, the amplitude of mechanosensitive currents is not correlated with that of acid-gated currents in DRG neurons [34]; the mechanosensitive currents are not altered by acidification [34]; and knockout of either ASIC2 or ASIC3, or both, has no effect on mechanically gated ion currents [26]. The second possibility is that a member of the TRP family may be involved. TRPV4 (originally called OTRPC4 or VR-OAC) [38, 39] generates an inward current in response to cell swelling, though the action appears to be via phosphorylation of TRPV4 [40], rather than being a direct gating, which would be more consistent with the known rapid response to noxious mechanical stimuli. Downregulation of TRPV4 gene expression impaired the nociceptive response to injection of hypotonic solutions [41], and pressure sensation was reduced in TRPV4-/- mice [42]. While these data are intriguing, they do not as yet provide unambiguous evidence for involvement of TRPV4 in mechanosensation, and other possibilities must remain open.

11.2.6

Initiation of Action Potentials by Noxious Stimuli

To be detected, all noxious stimuli must generate action potentials in the afferent nerve fiber. The first hint that this process may be different from that in other sensory neurons came when a voltage-sensitive Na channel unique to nociceptive neurons was

11.3 Sensitization by Inflammatory Mediators

cloned by a difference cloning strategy [43]. This channel, initially called SNS (sensory neuron specific) but subsequently renamed NaV1.8, has two main differences from most other voltage-dependent Na channels: it is relatively resistant to tetrodotoxin, and the voltage at which it is first activated is more positive by about 20 mV than that of the widely expressed TTX-sensitive Na channels. As well as being more positive, the threshold of NaV1.8 is modulated in the negative direction by inflammatory mediators such as prostaglandins, thereby making the neuron more excitable (see below). A second TTX-insensitive Na channel, NaV1.9 (initially named NaN or SNS2), was subsequently identified [44]. This channel, in spite of having a high homology to the other members of the NaV family, seems not to be voltage-activated in the same way as other Na channels, though it may generate a “window current” that contributes to maintenance of the resting potential [45]. Surprisingly, this Na channel has recently been shown to generate an inward current when directly activated by the binding of BDNF to TrkB receptors [46].

11.2.7

Summary Diagram of a Nociceptive Terminal

Fig. 11.1 summarizes the ion channels currently known to be involved in excitation of the nociceptive nerve terminal. Typically several different ion channels are expressed in a single nociceptor, thus conferring the property of polymodality commonly seen in nociceptors, but there is often in practice a more restricted expression than that shown in Fig. 11.1, leading to a somewhat greater degree of specificity in individual nerve terminals. An influx of current – usually carried mainly by Na+ ions, though Ca2+ ions may also contribute – through any of the ion channels leads to depolarization and consequent activation of voltage-sensitive Na channels, generating an action potential that propagates to the CNS. The sensitivity of the nociceptor to noxious stimuli can be modulated in two main ways: either by enhancing the generator current activated in response to a noxious stimulus or by changing the threshold of the Na channel so that the nociceptor is more readily excited. These possibilities are discussed below.

11.3

Sensitization by Inflammatory Mediators

Following tissue damage or inflammation, the sensitivity of nociceptive nerve terminals in vivo increases, a process known as sensitization [47, 48]. In the sensitized state, painful stimuli are enhanced (a phenomenon known as hyperalgesia), and ordinarily non-painful stimuli are perceived as painful (allodynia). Sensitization is caused by the release of chemical mediators from surrounding damaged or inflamed tissue, because when nociceptive neurons are isolated they show only adaptation to repeated application of stimuli such as heat or capsaicin [49]. The mediators capable of causing sensitization include prostaglandins, bradykinin, protons, ATP, and even nerve growth

257

258

11 Pain Transduction: Gating and Modulation of Ion Channels

Fig. 11.1 Schematic diagram of ion channels activated by noxious stimuli in nociceptor nerve terminals. Ion channels responsible for activating inward current in response to noxious stimuli are shown at the left, along with the stimuli believed to be primarily responsible for their activation. Voltage-sensitive sodium channels activated by the consequent depolarization are shown diagrammatically on the right. Note that NaV1.9 is not voltage-activated in the conventional sense but instead appears to be activated by direct interaction with BDNF bound to TrkB

factor, and new physiologically important factors are being discovered at a high rate. The majority of these factors act via G protein-coupled receptors and intracellular signaling pathways, though some may act directly (e.g., protons act directly on TRPV1). Sensitization has both short-term components, on the timescale of seconds and due to a local action of inflammatory mediators on nociceptive nerve terminals, and longterm components, due to upregulation of gene transcription and a consequent change in the protein expression in the nerve terminal. The pathways involved are better understood in some cases than in others.

11.3.1

Short-term Sensitization: Mediators and Pathways Bradykinin and the PKC Pathway Bradykinin is a nonapeptide released by proteolysis of a precursor protein, kininogen, when proteolytic enzymes are released from damaged cells and acts mainly at B2 receptors [50, 51]. Kininase I cleaves the C-terminal arginine to give des-arg9 bradykinin, which is a potent activator of B1 receptors but is largely ineffective at B2 receptors [52]. Both B1 and B2 receptors are coupled to Gq and thence to PLCb, leading to cleavage of 11.3.1.1

11.3 Sensitization by Inflammatory Mediators

Fig. 11.2 Pathways leading to sensitization of TRPV1. (A) Binding of bradykinin to the B2 receptor leads to activation of phospholipase C and release of IP3 and diacylglycerol. Of the five PKC isoforms expressed in nociceptive neurons, only PKCe is translocated to the membrane by bradykinin. Activation of PKCe by bradykinin leads to phosphorylation of TRPV1 and enhancement of the membrane current gated by a heat stimulus. (B) Targets of sensitization on TRPV1 (numbering of sites for rat TRPV1). The binding site for activation by capsaicin is shown. Target sites include intracellular serines that are phosphorylation sites for PKCe and PKA and extracellular glutamates whose protonation also modulates gating (see text)

259

260

11 Pain Transduction: Gating and Modulation of Ion Channels

PIP2 and release of IP3 and DAG. Bradykinin potently enhances the heat-activated current [6], an observation that was made before the identification of TRPV1, but the downstream target is now identified as TRPV1 because a similar effect is seen in HEK293 cells transfected with TRPV1 [49, 53]. The following lines of evidence suggest that the pathway leading to sensitization of TRPV1 involves protein kinase C: PMA, a selective activator of PKC, has similar effects to bradykinin; staurosporine, a PKC inhibitor, blocks the effects of bradykinin; and phosphatase inhibitors prolong the effects of bradykinin [6]. There are 11 isoforms of PKC, of which five (PKCbI, bII, c, e, and f) are expressed in neonatal DRG neurons, but only one isoform, PKCe, is involved in the sensitization of TRPV1 by bradykinin because infusion of a constitutively active form of PKCe has similar effects in enhancing the heat-activated membrane current and because a selective inhibitor of PKCe suppresses the action of bradykinin [54]. Bradykinin causes a selective translocation of PKCe to the neuronal cell membrane, a marker that is a sensitive indicator of PKCe activation [54]. A later study showed that PKCa, which is expressed in more mature DRG neurons but not in neonatal neurons, might also be able to activate TRPV1 [55]. The proposed pathway linking B2-receptor activation to TRPV1 activation is shown in Fig. 11.2A. Residues phosphorylated by PKCe have now been identified by site-directed mutagenesis [56, 57]. Serines 502 and 800 in rat TRPV1 are in the first intracellular loop and in the C-terminal tail, respectively (see Fig. 11.2B), and replacement of both these residues by non-phosphorylatable alanines was found to completely abolish the enhancement of TRPV1-dependent current caused by activation of PKCe. All of the effects of PKCe on TRPV1 would therefore seem to be mediated by phosphorylation of these two residues. A second mechanism for the action of bradykinin on TRPV1, independent of PKC activation, has also been proposed [58]. Removal of PIP2 from the membrane, either by hydrolysis or with an antibody, reduced the action of bradykinin and NGF in expression systems, suggesting that PIP2 associates with and inhibits TRPV1 and that removal of PIP2 is the crucial factor in enhancing TRPV1 following activation of PLC. The relative importance of this second mechanism for the action of bradykinin in sensory neurons, as opposed to in expression systems, has yet to be investigated, but the observation that application of kinase inhibitors removes most or all of the enhancement caused by bradykinin (see above) suggests that it plays a lesser role than the PKCe pathway shown in Fig. 11.2A.

Prostaglandins and the PKA Pathway Arachidonic acid is released from membrane lipids by the action of phospholipase A2 in response to a wide range of damaging and stressful stimuli and is the precursor for a vast range of proinflammatory mediators, including leukotrienes, lipoxins, thromboxanes, and prostaglandins. PGE2 and PGI2, which are both produced downstream of the cyclooxygenase (COX) enzymes, are particularly important mediators of hyperalgesia, and the inhibition of their production by COX inhibitors such as aspirin explains the potent analgesic effect of these inhibitors. When PGE2 is applied to nociceptive neu11.3.1.2

11.3 Sensitization by Inflammatory Mediators

rons, the threshold of the TTX-insensitive Na current shifts in the negative direction, leading to a decrease in the threshold for eliciting an action potential and, consequently, to an increase in excitability in response to a wide range of noxious stimuli [59]. Activation of adenylate cyclase with forskolin has a similar effect, suggesting that the cAMP/PKA pathway is important in this action of PGE2 [60]. In agreement with this idea, it was found that the effect could be abolished by mutating five serine residues, all of which are consensus sites for PKA and are located on the first major intracellular loop of NaV1.8 [61]. Thus, an important part of the action of PGE2 can be attributed to phosphorylation of NaV1.8 by PKA, as shown in Fig. 11.3. An alternative possibility was suggested by studies in which PKA and PKC were each blocked by selective inhibitors [62]. Activation of PKC alone increased the TTX-resistant Na current (but apparently without the shift of threshold caused by PKA activation), and the increase was not blocked by PKA inhibitors, while the increase in current caused by PKA activation was blocked by PKC inhibitors. These experiments suggest that the pathway linking PGE2 application to an increase in current through NaV1.8 may be PKA!PKC!NaV1.8. This result is not necessarily incompatible with the finding mentioned above, i.e., that mutating consensus sequences for PKA abolishes modulation of NaV1.8, as many potential phosphorylation sites are surrounded by adequate consensus sequences for both PKA and PKC. It seems likely that NaV1.8 is targeted by both kinases, but the relative importance of each for the observed increase in TTXinsensitive Na current and shift in excitation threshold has not been completely disentangled.

Fig. 11.3 Intracellular pathways and targets of the cAMP/PKA pathway that cause nociceptor sensitization. An indicative proalgesic agent (PGE2, binding to a receptor of the EP family) and an analgesic agent (an opiate, binding to the l opiate receptor) are shown, but it seems likely that a number of other agents, both proalgesic and analgesic, will activate the same pathways

261

262

11 Pain Transduction: Gating and Modulation of Ion Channels

PKA also targets TRPV1. While phosphorylation by PKC directly enhances the heator capsaicin-gated current, as described above, phosphorylation by PKA has the more subtle effect of reversing the desensitization caused by prolonged or repeated activation of TRPV1 [63]. Although more than one site is phosphorylated by PKA, the reversal of desensitization is abolished by mutating just a single serine, 116 in the rat TRPV1 sequence. The action of PKA on TRPV1 depends on anchoring by an A-kinase anchoring protein (AKAP) [64, 65], suggesting the formation of a signaling complex in which AKAP anchors PKA to TRPV1 in order to achieve rapid and reproducible physiological modulation of the membrane current gated by heat (Fig. 11.3). The cAMP produced by activation of adenylate cyclase may also have a direct effect in modulating the voltage sensitivity of Ih ion channels, which pass an inward current and are activated by hyperpolarization. Members of this family have a cyclic nucleotide-binding site near the C-terminal, and binding of cAMP shifts the activation range to more positive voltages, which has the effect of increasing the inward current and therefore of hastening depolarization to threshold for initiation of the next action potential. Ih plays a crucial role in many tissues, including heart tissue, where it mediates the pacemaker potential, and in neurones, where it determines the firing frequency by modulating the rate at which the membrane potential depolarizes between action potentials [66]. Four channel subunits, HCN 1–4, have been cloned. Ih has been reported in nociceptors, where it may play an important role in modulating firing in response to agents that modulate cAMP, such as prostaglandins, sympathomimetic monoamines [67], or opiates [68]. Opiates have well-known and potent central analgesic actions, but it has recently been shown that they also have a peripheral analgesic action [69]. Binding of agonist to the l opiate receptor expressed on nociceptors activates Gi, thus causing an inhibition of cAMP production and consequently antagonizing the proalgesic actions of (for example) prostaglandins [70]. A peripheral analgesic action for cannabinoids has also been suggested [71], but whether the cAMP pathway is involved in the same way as it is for opiates has yet to be established.

11.3.2

Nerve Growth Factor

Activation of mast cells by a stimulus such as bacterial lipopolysaccharide (LPS) causes release of nerve growth factor (NGF), which is a potent, though perhaps surprising, inflammatory mediator. NGF injection causes a rapid and long-lasting thermal hyperalgesia and a mechanical hyperalgesia with a slower onset [72]. NGF is involved in physiologically significant hyperalgesia in vivo, since removal of NGF reverses both the thermal and the mechanical hyperalgesia caused by an injection of LPS [73, 74]. There is general agreement that the long-term effects of NGF are due to upregulation of gene expression (see below), but the short-term effects, which can be seen in vivo within minutes of injection of NGF, must be due to activation of local signaling pathways within the nerve terminal. Rapid effects of NGF on TRPV1 in vitro have been characterized in isolated neurons using whole-cell patch-clamp [75]

11.3 Sensitization by Inflammatory Mediators

and calcium imaging [76] and show that an enhancement of the current passing through TRPV1 is induced within 2 min of exposure to NGF. TRPV1 gating is enhanced by NGF in around 40 % of TRPV1-expressing neurons, a proportion that agrees with immunocytochemical experiments in which it was found that about 50 % of nociceptive neurons express TrkA receptors for NGF [77]. The identity of the pathways activated by NGF and leading to sensitization of TRPV1 is currently the subject of some disagreement. Activation of TrkA receptors by NGF recruits many signaling molecules that can bind to the intracellular phosphorylated tyrosine residues within TrkA by means of Src homology (SH2) domains. Three proteins in particular have been identified based on their specific binding to phosphorylated Trk receptors: Shc, which activates the Ras/MEK pathway; phospholipase C gamma-1 (PLCc1), which cleaves PtdIns-4,5-P2 to IP3 and DAG; and phosphatidylinositol3-kinase (PI3K), which 3-phosphorylates PtdIns-4,5-P2 [78]. Many other signaling molecules are possible secondary mediators of the physiological effects of TrkA activation. Mendell’s group has proposed that PKA is an essential downstream element, as PKA inhibitors abolished the NGF-induced sensitization in patch-clamp studies, while inhibitors of PKC and MAP kinase were without effect [75]. David Julius’ group has proposed a quite different mechanism, independent of phosphorylation: activation of PKCc by TrkA, which is known to lead to breakdown of PIP2, may remove PIP2 from binding sites on TRPV1, releasing it from PIP2-mediated inhibition [58, 79]. This mechanism would unify the sensitizing actions of a number of agents, including bradykinin, NGF, ATP acting at P2Y receptors, and others that have in common that they activate PLC. In our lab we have used calcium imaging of intact neurons to investigate signaling pathways activated by TrkA, and our conclusions are again different: we find that wortmannin, a selective inhibitor of PI3 kinase, totally abolishes sensitization, while inhibition of PLC-c is without effect, suggesting that PI3 kinase, and not PLCc, is the crucial early step [76]. MEK inhibition had no effect on NGF-induced sensitization, showing that the MEK/MAPK pathway is not involved, at least downstream of MEK, while Ras inhibition had a small inhibitory effect on sensitization, consistent with a modulation of PI3K by Ras, as shown in other studies [80]. We have also found that inhibitors of PKC and CAM kinase abolish sensitization, suggesting an involvement of these kinases at a later stage, though the details of the later stages of the signaling pathway are currently unclear. The observation that kinase inhibitors abolish NGF-induced sensitization suggests that phosphorylation of TRPV1 is an essential final common pathway, as with sensitization in response to other inflammatory mediators discussed above, but the kinases involved, and the residues on TRPV1 that they phosphorylate, have yet to be identified.

11.3.3

Direct Modulation of TRPV1 by Protons

The acid produced by inflammation or during ischemic exercise, where pH levels may fall as low as 5.4, is an important contributor to pain [81]. As discussed above, protons may cause pain directly by an action at either TRPV1 or members of the ASIC family,

263

264

11 Pain Transduction: Gating and Modulation of Ion Channels

or both, but protons can additionally sensitize the response of TRPV1 to heat, without directly activating it, by shifting to lower temperatures the relationship between channel activation and heat, in much the same way as bradykinin does [9]. The relevant residues have been identified as two glutamates near the mouth of the channel, one of which (E600) is involved in setting the sensitivity of the channel to heat, while a second (E648) is involved in direct activation of the channel [11]. It is interesting to note that two actions on opposite sides of the membrane – protonation at an external site, and phosphorylation by PKC at an internal site – can have rather similar effects in terms of modulating the sensitivity of the TRPV1 ion channel to heat [82].

11.3.4

Other Modulators of Nociceptor Sensitivity

In such a highly evolved system as nociception, a wide variety of factors released by injury or inflammation would be expected to cause hyperalgesia, and the few factors outlined above are only the best understood among many. The following sections outline what is known of some others, but this list is sure to be far from complete.

ATP The best-characterized action of ATP is in activating P2X ion channels and thus directly exciting nociceptors, but a sensitizing action arising from an action at P2Y receptors may be more important in vivo. Through an action at the P2Y1 G proteincoupled receptors, ATP activates the same pathway as bradykinin, i.e., PLCb, releasing DAG and leading to activation of PKCe and consequent phosphorylation of TRPV1 as shown in Fig. 11.2 [56]. 11.3.4.1

Proteases Proteases released following cell damage activate a family of protease-activated receptors (PARs) in which proteolysis of the extracellular N-terminal domain leads to unmasking of a tethered agonist peptide and consequent receptor activation. In intact animals, prolonged hyperalgesia results from injection of PAR activators, and PAR2 seems to be particularly involved in generation of the sensation of itch [83]. Downstream pathways involve PLC, leading to intracellular calcium signals and consequent exocytosis of neuropeptides, and ERK activation may also play a role [84, 85]. 11.3.4.2

Bv8/Prokineticin Bv8 is a small protein isolated from amphibian skin whose mammalian homologues make up the prokineticin family. The receptors for these proteins, PKR1 and PKR2, are expressed in mammalian DRG neurons, and injection of Bv8 leads to a profound and long-lasting thermal and mechanical hyperalgesia [86]. Application of Bv8 to isolated DRG neurons causes calcium release from intracellular stores, suggesting that an activation of the PLC pathway may be the basis of the sensitization. While these ob11.3.4.3

11.3 Sensitization by Inflammatory Mediators

servations are tantalizing, there is as yet no evidence that prokineticins are actually released during inflammation, nor is there evidence regarding what component of physiological hyperalgesia may be due to activation of PKRs. Glutamate The possibility that glutamate may play a role in hyperalgesia was raised by the observation that the metabotropic glutamate receptors mGluR1 and mGluR5 are expressed on C-fiber afferents and that injection of agonists of these receptors causes thermal hyperalgesia [87]. The observation that antagonists partially suppress the pain caused by formalin injection suggests a physiological role for glutamate in pain. Current understanding of the intracellular pathway is that PLC is activated by mGluR but that the main downstream effect is not to activate PKC but instead to cause release of prostaglandins and TRPV1 sensitization via a PKA-dependent pathway [88]. 11.3.4.4

Norepinephrine Activation of b2 receptors by norepinephrine (noradrenaline) or other adrenergic agonists causes a hyperalgesia that is reduced in PKCe knockout mice or when PKCe is inhibited, suggesting an involvement of PKCe downstream of b2 receptor activation [89]. The target of PKCe when activated by norepinephrine has not been clearly established and could be TRPV1, as is the case when bradykinin receptors are activated (see above, Fig. 11.2) and/or NaV1.8 [90]. More recent work has suggested that the ERK/ MAPK pathway may also be an important mediator of sensitization caused by b2 activation [91]. 11.3.4.5

11.3.5

Long-term Sensitization

Pain is usually transient in nature, disappearing quickly when a mild noxious stimulus is removed or more gradually as the inflammation caused by an injury dies away. Some forms of pain, however, notably those caused by nerve injury, can be much longer lasting, suggesting that long-term changes in gene expression underlie them. Long-lasting pain caused by nerve injury is called neuropathic pain, and its origin is currently the subject of much debate. Long-term changes in transmission to second-order neurons in the spinal cord are important [92], but phenotypic changes in the ability of primary sensory neurons to detect painful stimuli or to sensitize in response to inflammatory mediators also play a role. Growth factors play a crucial role in the control of nociceptor phenotypes. Neonatal neurons depend on NGF for survival, but in the adult animal, removal of NGF does not cause cell death but instead changes the expression of many proteins associated with the nociceptive phenotype. A reduction in NGF supply can be caused by nerve section, which removes the supply of target-derived NGF from innervated tissues. The supply of neurotrophins can also be increased in some circumstances: for instance, NGF is known to be released from mast cells and other sources during inflammation [93], and

265

266

11 Pain Transduction: Gating and Modulation of Ion Channels

both NGF and GDNF are released from invading macrophages when nerve axons degenerate following nerve damage [94, 95] and will bathe any surviving axons in the damaged nerve trunk in a neurotrophin-rich fluid. 11.3.6

Gene Expression Regulated by NGF

The expression of the neuropeptides substance P and CGRP was shown many years ago to be downregulated by NGF removal or conversely upregulated by addition of NGF [96, 97], while other neuropeptides such as VIP and CCK are upregulated by NGF removal [97]. One effect of an oversupply of NGF therefore may be to enhance peptidergic neurotransmission by substance P and CGRP at the first synapse, while deprivation of NGF, caused for instance by nerve section, would have the converse effect. TRPV1 expression is upregulated by NGF in a pathway that depends upon activation of Ras and the MEK/MAPK cascade [98]. Expression of the bradykinin B2 receptor is also upregulated by NGF [99]. One of the prominent physiological changes in nociceptors following nerve damage involves the generation of sustained action potential firing [100], and a neurotrophindriven change in sodium-channel expression may be a contributor to this. The expression of NaV1.8 is upregulated by NGF [101] and, conversely, is downregulated by nerve section [102]. This last observation would not appear to offer a ready explanation of neuropathic pain, in which we would be looking for an upregulation of expression of NaV1.8 following nerve damage in order to explain repetitive activity, but other evidence suggests that in neuropathic pain states NaV1.8 may in fact be upregulated in nerve trunks, if not in the neuronal cell body [103, 104]. A second channel, NaV1.3, is not normally expressed in sensory neurons, but it appears after nerve damage and has characteristics that may make it a candidate for the originator of the repetitive firing which is seen following nerve damage [102]. A second possible contributor to the repetitive firing observed following nerve damage is the hyperpolarization-activated inward current channel, Ih, which contributes to action potential firing by regulating that rate at which the neuronal membrane depolarizes between action potentials (see above). Recently it has been suggested that Ih may be functionally enhanced in neuropathic pain states, since a blocker of Ih, ZD7288, alleviated neuropathic pain [105]. Surprisingly, Ih protein and mRNA expression was found to be downregulated in an experimental neuropathic pain state, but the magnitude of the Ih current nonetheless increased, perhaps because the voltage dependence of the current is shifted in the depolarizing direction [105].

11.3.7

Gene Expression Regulated by GDNF

Only around half of nociceptors express TrkA receptors for NGF, while the other half express Ret receptors for GDNF [77]. Many of the nociceptor-specific proteins expressed by these two groups of neurons are different: the TrkA-expressing neurons

267

express the neuropeptides substance P and CGRP and terminate in outer layers of the dorsal horn (layer I and II outer), while the Ret-expressing neurons express surface receptors binding to the plant lectin IB4 and terminate in layer II (inner) of the dorsal horn. These two populations might be considered the “rods and cones” of the nociceptor system, but so far no very distinctive differences in their properties have been characterized [106, 107], at least not to match those conferring differential wavelength selectivity and time course on the visual rod and cone systems. The ability to bind IB4 provides a convenient method for distinguishing the two nociceptor populations, as it can be readily applied to both living and fixed neurons. Several nociceptor-specific proteins are localized to the IB4-positive population: both the P2X3 receptor for ATP and the NaV1.9 sodium channel are exclusively located in the IB4+ neuronal population [108, 109], and our group has identified B1 bradykinin receptors as also being exclusively located in IB4+ neurons [110], unlike B2 receptors, which are expressed mainly in TrkA+ neurons [99]. Ret forms functional receptors for members of the GDNF family – which, apart from GDNF itself, comprises the elegantly named members neurturin, artemin, and persephin – by combining with the GFRa receptor family, although there is some evidence that Ret may be active alone [111]. By analogy with the action of NGF in upregulating gene expression in TrkA-expressing neurons, we might expect that GDNF and other members of this family would have a similar effect in upregulating the proteins expressed in the IB4+ neurons, but to date this possibility has been inadequately investigated. We have recently found, however, that B1 expression is potently upregulated by GDNF [110]. B1 receptors are normally expressed in only a very small fraction of neurons (around 2 %), but following exposure to GDNF, functional B1 receptor expression appears in around 20 % of neurons. GDNF or related neurotrophins are therefore likely candidates to drive upregulation of B1 expression following injury or inflammation, leading to the switch in agonist responsiveness from a B2 to a B1 profile that has been identified in in vivo experiments [112]. 11.4

Conclusions

There has been an explosion of research into the cellular and molecular basis of pain transduction over the past few years. Many of the membrane ion channels responsible for the detection of painful stimuli have now been cloned, and we are making good progress in identifying the pathways responsible for modulating these ion channels and thus mediating the process of sensitization. Some areas are still mysterious and will no doubt form the subject of advances in understanding in future years. We still know little about the molecular basis for detection of strong mechanical stimuli (or, indeed, any mechanical stimuli). The means by which temperature-sensitive ion channels are gated still remains unknown at the molecular level. And much remains to be discovered about the pathways mediating sensitization: kinases seem to be important in most of the pathways investigated to date, but the details of how they are activated, and which are their critical target residues on the primary transducer ion channels, still remain in many cases to be elucidated.

268

11 Pain Transduction: Gating and Modulation of Ion Channels

References 1 2 3 4 5

6

7 8

9

10 11

12

13

14

15 16

17

C. S. Sherrington, The integrative action of the nervous system, Scribner, New York, 1906. P. D. Wall and R. Dubner, Annu. Rev. Physiol, 1972, 34, 315–336. P. Bessou and E. R. Perl, J. Neurophysiol., 1969, 32, 1025–1043. C. Belmonte and F. Giraldez, Journal of Physiology, 1981, 321, 355–368. R. D. Treede, R. A. Meyer, S. N. Raja and J. N. Campbell, Journal of Physiology, 1995, 483, 747–758. P. Cesare and P. A. McNaughton, Proceedings of the National Academy of Sciences of the United States of America, 1996, 93, 15435–15439. I. Nagy and H. Rang, Neuroscience, 1999, 88, 995–997. M. J. Caterina, M. A. Schumacher, M. Tominaga, T. A. Rosen, J. D. Levine and D. Julius, Nature, 1997, 389, 816–824. M. Tominaga, M. J. Caterina, A. B. Malmberg, T. A. Rosen, H. Gilbert, K. Skinner, B. E. Raumann, A. I. Basbaum and D. Julius, Neuron, 1998, 21, 531–543. S. E. Jordt and D. Julius, Cell, 2002, 108, 421–430. S.-E. Jordt, M. Tominaga and D. Julius, Proceedings of the National Academy of Sciences of the United States of America, 2000, 97, 8134–8139. M. J. Caterina, T. A. Rosen, M. Tominaga, A. J. Brake and D. Julius, Nature, 1999, 398, 436–441. M. J. Caterina, A. Leffler, A. B. Malmberg, W. J. Martin, J. Trafton, K. R. Petersen-Zeitz, M. Koltzenburg, A. I. Basbaum and D. Julius, Science, 2000, 288, 306–313. J. B. Davis, J. Gray, M. J. Gunthorpe, J. P. Hatcher, P. T. Davey, P. Overend, M. H. Harries, J. Latcham, C. Clapham, K. Atkinson, S. A. Hughes, K. Rance, E. Grau, A. J. Harper, P. L. Pugh, D. C. Rogers, S. Bingham, A. Randall and S. A. Sheardown, Nature, 2000, 405, 183–187. D. A. Simone and K. C. Kajander, Neurosci. Lett., 1996, 213, 53–56. G. M. Story, A. M. Peier, A. J. Reeve, S. R. Eid, J. Mosbacher, T. R. Hricik, T. J. Earley, A. C. Hergarden, D. A. Andersson, S. W. Hwang, P. McIntyre, T. Jegla, S. Bevan and A. Patapoutian, Cell, 2003, 112, 819–829. D. D. McKemy, W. M. Neuhausser and D. Julius, Nature, 2002, 416, 52–58.

18

19

20 21

22

23

24

25

26

27

28

29

30 31

32

A. M. Peier, A. Moqrich, A. C. Hergarden, A. J. Reeve, D. A. Andersson, G. M. Story, T. J. Earley, I. Dragoni, P. McIntyre, S. Bevan and A. Patapoutian, Cell, 2002, 108, 705–715. R. Waldmann, G. Champigny, F. Bassilana, C. Heurteaux and M. Lazdunski, Nature, 1997, 386, 173–177. M. P. Price, P. M. Snyder and M. J. Welsh, J. Biol. Chem., 1996, 271, 7879–7882. R. Waldmann, F. Bassilana, J. De Weille, G. Champigny, C. Heurteaux and M. Lazdunski, Journal of Biological Chemistry, 1997, 272, 20975–20978. R. Waldmann, G. Champigny, E. Lingueglia, Weille JR De, C. Heurteaux and M. Lazdunski, Annals of the New York Academy of Sciences, 1999, 868 A. N. Akopian, C.-C. Chen, Y. Ding, P. Cesare and J. N. Wood, NeuroReport, 2000, 11, 2217–2222. M. P. Price, S. L. McIlwrath, J. Xie, C. Cheng, J. Qiao, D. E. Tarr, K. A. Sluka, T. J. Brennan, G. R. Lewin and M. J. Welsh, Neuron, 2001, 32, 1071–1083. M. P. Price, G. R. Lewin, S. L. McIlwrath, C. Cheng, J. Xie, P. A. Heppenstall, C. L. Stucky, A. G. Mannsfeldt, T. J. Brennan, H. A. Drummond, J. Qiao, C. J. Benson, D. E. Tarr, R. F. Hrstka, B. Yang, R. A. Williamson and M. J. Welsh, Nature, 2000, 407, 1007–1011. L. J. Drew, D. K. Rohrer, M. P. Price, K. Blaver, D. A. Cockayne, P. Cesare and J. N. Wood, J Physiol, 2004, 556, 691–710 S. G. Hamilton and S. B. McMahon, Journal of the Autonomic Nervous System, 2000, 81, 187–194. C. C. Chen, A. N. Akopian, L. Sivilotti, D. Colquhoun, G. Burnstock and J. N. Wood, Nature, 1995, 377, 428–431. C. Lewis, S. Neidhart, C. Holy, R. A. North, G. Buell and A. Surprenant, Nature, 1995, 377, 432–435. C. Kennedy, T. S. Assis, A. J. Currie and E. G. Rowan, J. Physiol, 2003, 553, 683–694. D.A.Cockayne,S.G.Hamilton,Q.M.Zhu,P.M. Dunn, Y. Zhong, S. Novakovic, A. B. Malmberg, G. Cain, A. Berson, L. Kassotakis, L. Hedley, W. G. Lachnit, G. Burnstock, S. B. McMahon and A. P. Ford, Nature, 2000, 407, 1011–1015. V. Souslova, P. Cesare, Y. Ding, A. N. Akopian, L. Stanfa, R. Suzuki, K. Carpenter, A. Dickenson, S. Boyce, R. Hill, D. NebenuisOosthuizen, A. J. Smith, E. J. Kidd and J. N. Wood, Nature, 2000, 407, 1015–1017.

11.4 Conclusions 33 34 35

36

37 38

39

40

41

42 43 44

45

46 47

48

49

50 51 52 53

G.C. McCarter,D.B.Reichlingand J. D.Levine, Neuroscience Letters, 1999, 273, 179–182. L. J. Drew, J. N. Wood and P. Cesare, J Neurosci., 2002, 22, RC228. F. Viana, Pena E. De La, B. Pecson, R. F. Schmidt and C. Belmonte, Eur. J Neurosci., 2001, 13, 722–734. N. R. Srinivasa, R. A. Meyer, M. Ringkamp, J. N. Campbell, in: Textbook of pain, ed. R. Melzack and P. D. Wall, Churchill Livingstone, Edinburgh, 1999 D. C. Immke and E. W. McCleskey, Nat. Neurosci., 2001, 4, 869–870. R. Strotmann, C. Harteneck, K. Nunnenmacher, G. Schultz and T. D. Plant, Nat. Cell Biol., 2000, 2, 695–702. W. Liedtke, Y. Choe, M. A. Marti-Renom, A. M. Bell, C. S. Denis, A. Sali, A. J. Hudspeth, J. M. Friedman and S. Heller, Cell, 2000, 103, 525–535. H. Xu, H. Zhao, W. Tian, K. Yoshida, J. B. Roullet and D. M. Cohen, J. Biol. Chem., 2003, 278, 11520–11527. N. Alessandri-Haber, J. J. Yeh, A. E. Boyd, C. A. Parada, X. Chen, D. B. Reichling and J. D. Levine, Neuron, 2003, 39, 497–511. M. Suzuki, A. Mizuno, K. Kodaira and M. Imai, J. Biol. Chem., 2003, 278, 22664–22668. A. N. Akopian, L. Sivilotti and J. N. Wood, Nature, 1996, 379, 257–262. S. D. Dib-Hajj, L. Tyrrell, J. A. Black and S. G. Waxman, Proceedings of the National Academy of Sciences of the United States of America, 1998, 95, 8963–8968. R. I. Herzog, T. R. Cummins and S. G. Waxman, J. Neurophysiol., 2001, 86, 1351–1364. R. Blum, K. W. Kafitz and A. Konnerth, Nature, 2002, 419, 687–693. J. N. Campbell and R. A. Meyer, in: Neurobiology of nociceptors, ed. C. Belmonte and F. Cervero, Oxford University Press, 1996 R. A. Meyer, J. N. Campbell, N. R. Srinivasa, in: Textbook of Pain, ed. R. Melzack and P. D. Wall, Churchill Livingstone, Edinburgh, 1994 V. Vellani, S. Mapplebeck, A. Moriondo, J. B. Davis and P. A. McNaughton, J. Physiol, 2001, 534, 813–825. D. Regoli and J. Barabe, Pharmacol Rev, 1980, 32, 1–46. A. Dray and M. Perkins, Trends in Neurosciences, 1993, 16, 99–104. F. Marceau, J. F. Hess and D. R. Bachvarov, Pharmacological Reviews, 1998, 50, 357–386. L. S. Premkumar and G. P. Ahern, Nature, 2000, 408, 985–990.

54

55 56

57

58

59

60 61

62

63

64

65

66 67 68 69 70 71 72 73

74

P. Cesare, L. V. Dekker, A. Sardini, P. J. Parker and P. A. McNaughton, Neuron, 1999, 23, 617–624. Z. Olah, L. Karai and M. J. Iadarola, J Biol. Chem., 2002, 277, 35752–35759. M. Numazaki, T. Tominaga, H. Toyooka and M. Tominaga, J Biol Chem, 2002, 277, 13375–13378. G. Bhave, H. J. Hu, K. S. Glauner, W. Zhu, H. Wang, D. J. Brasier, G. S. Oxford and R. W. Gereau, Proc. Natl. Acad. Sci. U. S. A, 2003, 100, 12480–12485. H. H. Chuang, E. D. Prescott, H. Kong, S. Shields, S. E. Jordt, A. I. Basbaum, M. V. Chao and D. Julius, Nature, 2001, 411, 957–962. M. S. Gold, D. B. Reichling, M. J. Shuster and J. D. Levine, Proceedings of the National Academy of Sciences of the United States of America, 1996, 93, 1108–1112. S. England, S. Bevan and R. J. Docherty, Journal of Physiology, 1996, 495, 429–440. E. M. Fitzgerald, K. Okuse, J. N. Wood, A. C. Dolphin and S. J. Moss, Journal of Physiology, 1999, 516, 433–446. M. S. Gold, J. D. Levine and A. M. Correa, Journal of Neuroscience, 1998, 18, 10345–10355. G. Bhave, W. Zhu, H. Wang, D. J. Brasier, G. S. Oxford and R. W. Gereau, Neuron, 2002, 35, 721–731. C. Distler, P. K. Rathee, K. S. Lips, O. Obreja, W. Neuhuber and M. Kress, J. Neurophysiol., 2003, 89, 2499–2505. P. K. Rathee, C. Distler, O. Obreja, W. Neuhuber, G. K. Wang, S. Y. Wang, C. Nau and M. Kress, J. Neurosci., 2002, 22, 4740–4745. U. B. Kaupp and R. Seifert, Annu. Rev. Physiol, 2001, 63, 235–257. S. L. Ingram and J. T. Williams, Journal of Physiology, 1996, 492, 97–106. S. L. Ingram and J. T. Williams, Neuron, 1994, 13, 179–186. C. Stein, M. Schafer and H. Machelska, Nat. Med., 2003, 9, 1003–1008. M. S. Gold and J. D. Levine, Neuroscience Letters, 1996, 212, 83–86. A. Calignano, G. La Rana, A. Giuffrida and D. Piomelli, Nature, 1998, 394, 277–281. G. R. Lewin, A. M. Ritter and L. M. Mendell, Journal of Neuroscience, 1993, 13, 2136–2148. C. J. Woolf, B. SafiehGarabedian, Q. P. Ma, P. Crilly and J. Winter, Neuroscience, 1994, 62, 327–331. S. B. McMahon, D. L. H. Bennett, J. V. Priestley and D. L. Shelton, Nature Medicine, 1995, 1, 774–780.

269

270

11 Pain Transduction: Gating and Modulation of Ion Channels 75 76 77 78 79 80

81

82

83

84

85

86

87 88 89

90

91

92 93

X. Shu and L. M. Mendell, J Neurophysiol., 2001, 86, 2931–2938. J. K. Bonnington and P. A. McNaughton, J. Physiol, 2003, 551, 433–446. W. D. Snider and S. B. McMahon, Neuron, 1998, 20, 629–632. D. R. Kaplan and F. D. Miller, Curr Opin Neurobiol, 2000, 10, 381–391. E. D. Prescott and D. Julius, Science, 2003, 300, 1284–1288. P. Rodriguez-Viciana, P. H. Warne, R. Dhand, B. Vanhaesebroeck, I. Gout, M. J. Fry, M. D. Waterfield and J. Downward, Nature, 1994, 370, 527–532. K. H. Steen, P. W. Reeh, F. Anton and H. O. Handwerker, Journal of Neuroscience, 1992, 12, 86–95. P. Cesare, A. Moriondo, V. Vellani and P. A. McNaughton, Proceedings of the National Academy of Sciences of the United States of America, 1999, 96, 7658–7663. M. Steinhoff, U. Neisius, A. Ikoma, M. Fartasch, G. Heyer, P. S. Skov, T. A. Luger and M. Schmelz, J. Neurosci., 2003, 23, 6176–6180. N. Vergnolle, M. Ferazzini, M. R. D’Andrea, J. Buddenkotte and M. Steinhoff, Trends Neurosci., 2003, 26, 496–500. M. Steinhoff, N. Vergnolle, S. H. Young, M. Tognetto, S. Amadesi, H. S. Ennes, M. Trevisani, M. D. Hollenberg, J. L. Wallace, G. H. Caughey, S. E. Mitchell, L. M. Williams, P. Geppetti, E. A. Mayer and N. W. Bunnett, Nat. Med., 2000, 6, 151–158. L. Negri, R. Lattanzi, E. Giannini, A. Metere, M. Colucci, D. Barra, G. Kreil and P. Melchiorri, Br. J Pharmacol., 2002, 137, 1147–1154. G. Bhave, F. Karim, S. M. Carlton and R. W. Gereau, Nat. Neurosci, 2001, 4, 417–423. H. J. Hu, G. Bhave and R. W. Gereau, J. Neurosci., 2002, 22, 7444–7452. S. G. Khasar, Y.-H. Lin, A. Martin, J. Dadgar, T. McMahon, D. Wang, B. Hundle, K. O. Aley, W. Isenberg, G. McCarter, P. G. Green, C. W. Hodge, J. D. Levine and R. O. Messing, Neuron, 1999, 24, 253–260. S. G. Khasar, G. McCarter and J. D. Levine, Journal of Neurophysiology, 1999, 81, 1104–1112. K. O. Aley, A. Martin, T. McMahon, J. Mok, J. D. Levine and R. O. Messing, J. Neurosci., 2001, 21, 6933–6939. R. R. Ji, T. Kohno, K. A. Moore and C. J. Woolf, Trends Neurosci, 2003, 26, 696–705. Q.-P. Ma and C. J. Woolf, NeuroReport, 1997, 8, 807–810.

94

95 96 97

98

99

100 101

102 103

104

105

106 107 108

109

110 111

112

M. Trupp, M. Ryden, H. Jornvall, H. Funakoshi, T. Timmusk, E. Arenas and C. F. Ibanez, Journal of Cell Biology, 1995, 130, 137–148. P. Naveilhan, W. M. ElShamy and P. Ernfors, Eur J Neurosci, 1997, 9, 1450–1460. R. M. Lindsay and A. J. Harmar, Nature, 1989, 337, 362–364. V. M. K. Verge, P. M. Richardson, Z. Wiesenfeld-Hallin and T. Hokfelt, Journal of Neuroscience, 1995, 15, 2081–2096. R. Bron, L. J. Klesse, K. Shah, L. F. Parada and J. Winter, Mol. Cell Neurosci., 2003, 22, 118–132. Y.-J. Lee, O. Zachrisson, D. A. Tonge and P. A. McNaughton, Mol Cell Neurosci., 2002, 19, 186–200. P. D. Wall, S. Waxman and A. I. Basbaum, Exp. Neurol., 1974, 45, 576–589. S. G. Waxman, S. Dib-Hajj, T. R. Cummins and J. A. Black, Proceedings of the National Academy of Sciences of the United States of America, 1999, 96, 7635–7639. S. G. Waxman, Neurology, 2001, 56, 1621–1627. J. Lai, M. S. Gold, C. S. Kim, D. Bian, M. H. Ossipov, J. C. Hunter and F. Porreca, Pain, 2002, 95, 143–152. M. S. Gold, D. Weinreich, C. S. Kim, R. Wang, J. Treanor, F. Porreca and J. Lai, J. Neurosci., 2003, 23, 158–166. S. R. Chaplan, H. Q. Guo, D. H. Lee, L. Luo, C. Liu, C. Kuei, A. A. Velumian, M. P. Butler, S. M. Brown and A. E. Dubin, J. Neurosci., 2003, 23, 1169–1178. C. L. Stucky and G. R. Lewin, J Neurosci, 1999, 19, 6497–6505. S. Dirajlal, L. E. Pauers and C. L. Stucky, J Neurophysiol., 2003, 89, 513–524. Y. Ding, P. Cesare, L. Drew, D. Nikitaki and J. N. Wood, Journal of the Autonomic Nervous System, 2000, 81, 289–294. J. Fjell, P. Hjelmstrom, W. Hormuzdiar, M. Milenkovic, F. Aglieco, L. Tyrrell, S. Dib-Hajj, S. G. Waxman and J. A. Black, NeuroReport, 2000, 11, 199–202. V. Vellani, O. Zachrisson and P. A. McNaughton, J Neurosci, 2004, submitted. R. H. Baloh, H. Enomoto, E. M. Johnson, Jr. and J. Milbrandt, Curr Opin Neurobiol, 2000, 10, 103–110. A. J. Davis and M. N. Perkins, Neuropharmacology, 1994, 33, 127–133.

271

12

Transduction and Transmission in Electroreceptor Organs Robert C. Peters and Jean-Pierre Denizot Religious truth is based on repetition of a message, scientific truth is gained by repeating an experiment.1

Abstract

This chapter reviews the current ideas on transduction and transmission in electroreceptor organs gained through electrophysiological, pharmacological, and histochemical means. Focal in electroreceptor organ functioning is the mechanism of neurotransmitter release. The spontaneous activity of the primary afferents is presumably caused by a sustained release of neurotransmitter from the electroreceptor cells, which in turn is controlled by a depolarizing “bias current” through the electroreceptor cells. The stimulus, an electrical potential difference across the integument, modulates the spontaneously active synapses of the electroreceptor cells. However, many findings are inconsistent with this general model. Electroreceptor organs in monotremes, for example, are without sensory cells and have only an afferent nerve fiber. In some tuberous organs, there is evidence for an electrical connection between receptor cells and primary afferents by means of a chemical synapse. Further, there is often a discrepancy between the behaviors of spontaneous activity and stimulus-evoked activity of primary afferents, when both should change equally. Therefore, an alternative explanation for stimulus transduction will be discussed in addition to the part played by ion channels in transduction and transmission. In the upgraded model the spontaneous activity is a property of the primary afferent, and the chemical synapse is a “switch” to provide a low-resistivity path between receptor cell and primary afferent to the stimulus current.

1

For more about learning and cellular mechanisms, please refer to [1].

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

272

12 Transduction and Transmission in Electroreceptor Organs

12.1

Introduction 12.1.1

Transduction at Electroreceptor Cells

According to the strict definition of the concept “transduction,” electroreceptor cells – the sensory cells in electroreceptor organs – do not possess membrane transduction channels. The stimulus is electrochemical in nature, and so are the receptor potential and the action potential. No transduction of one form of energy into another one is needed. Nevertheless, the electrical stimulus causes ions to pass membrane channels and to disturb the electrochemical homeostasis of the electroreceptor cells. Therefore, ion channels and ion pumps are involved in transfer of the stimulus from the outer world to the nervous system. If the meaning of “transduction” can be expanded into “converting a change in energy balance between receptor cell and environment into a change in membrane potential,” electroreceptor organs do have transduction channels. Electroreceptor organs respond further to natural changes in their ionic environment that can be considered as chemical stimuli. At the level of the sensory cell, both electrical and chemical stimuli come together because the electrical stimulus is carried by ions in an aquatic environment. So far, however, no one has succeeded in investigating the contribution of electroreceptor organs to chemoreception. Today, more than half a century after Lissmann’s discovery of the electric sense [2], there are still many challenging questions regarding electroreceptor organ transduction and transmission.

Fig. 12.1 The four main categories of electroreceptor organs schematically. (A) Monotreme sensory mucous glands. (B) Microampullary organs in freshwater fishes and amphibians. (C) Tuberous organs in freshwater electric fish, knollenorgan. (D) Tuberous organs in freshwater fishes, mormyromast. (E) Ampullae of Lorenzini in marine fishes. Note that the monotreme sensory mucous glands have no secondary receptor cells and that the

primary afferent nerve fiber contains the ion channels necessary for generating the receptor/ generator potential and the action potentials. All other electroreceptor organs have secondary electroreceptor cells with synaptic bodies distally of the primary afferents. A.n., afferent nerve fiber; r.c., receptor cell; bl, basal lamina; L, lumen. After [26, 123, 124] (drawing by Pieter van Dorp van Vliet)

12.1 Introduction

Fig. 12.2 Some examples of apical and basal specializations of electroreceptor cells. (A) Apical cytoplasm of an ampullary organ sensory cell (SC) of Plotosus anguillaris with microvilli (arrows) (20,000; photo: Denizot). (B) Peripheral cytoplasm of a knollenorgan sensory cell (SC) of Gnathonemus petersii with microvilli (arrows) (15,200; photo: Denizot). (C) Demonstration of ATPase (ATPase Mg Ca) on the membrane of knollenorgan sensory cell microvilli of Gnathonemus petersii. Arrows indicate the reaction product on the plasma membrane (81,000). (D) Chemical synapse in a sensory cell of a type-B larval tuberous

organ of a Mormyridae. Note the synaptic ribbons (arrows) in the cytoplasm of the sensory cell (SC) and at the base the afferent nerve terminals (N) (21,200; photo: Bensouilah et al.). (E) Detail of a chemical synapse: the synaptic ribbon in the synaptic cleft is surrounded by synaptic vesicles with an electron-dense deposit (arrows) due to calcium precipitates. In the cleft the cytoplasmic membrane constitutes the presynaptic membrane (pr), while the membrane of the afferent nerve terminal (N) forms the postsynaptic membrane (ps) (sensory cell SC) (60,200; photo: Djebar et al.)

12.1.2

Favorite Species

Electroreception is studied by a relatively small group of investigators in a relatively small group of species. Some electroreceptive species have received particular attention because of their availability, laboratory proneness, or suitability for a specific research question. The catfish Kryptopterus bicirrhis, for instance, is transparent to the extent that electroreceptor cells are visible under the microscope, thus allowing for visual control of intracellular electrophysiology and fluorescence experiments in vivo [3–5]. The elephant-nosed Gnathonemus petersii can be bought in almost any pet shop, is easy to keep in the lab, and is a favorite subject for studies of the genera-

273

274

12 Transduction and Transmission in Electroreceptor Organs

tion of electric organs discharges, electrocommunication, electrolocalization of objects, and central neural processing [6–9]. The knife fish Eigenmannia virescens, like similar Gymnotiformes, has also been the subject of electrocommunication studies for decades [10, 11]. The marine catfish Plotosus and marine skates and rays have centimeters long ampullae of Lorenzini, which are very well suited to study the mechanism of stimulus transduction in vitro [12, 13]. The freshwater catfish Ictalurus allows extracellular recording of electroreceptor primary afferents in vivo [14–17].

12.2

Types of Electroreceptor Organs 12.2.1

General

For a comprehensive overview of all types of electroreceptor organs and electroreception history, the reader is referred to the earlier handbooks dealing with this topic [18–24]. Here it suffices to mention briefly the existence of four main categories of electroreceptor organs: (1) the sensory mucous glands in monotremes, (2) the microampullary organs in freshwater fishes and amphibians, (3) the tuberous organs in freshwater fishes, and (4) the ampullae of Lorenzini in marine fishes (Fig. 12.1). Characteristic for all types of electroreceptor organs is their topology. In all types the receptive structures are at the bottom of some kind of invagination or cavity – the ampulla – and are protected by mucoid substances [25]. Further, the apical membranes of the secondary receptor cells generally have microvilli (Figures 12.1b,c,d and 12.2a,b,c), but the ampullae of Lorenzini also present kinocilia (Fig. 12.1e). According to morphological criteria, the synapses, if present, are chemical (Fig. 12.2d,e) [26].

12.2.2

The Sensory Mucous Glands in Monotremes

The sensory mucous glands resemble mucous glands in that they have duct-like invaginations. Trigeminal nerve fibers innervating these mucous glands have specialized terminals consisting of highly convoluted plasma membranes inside a myelin capsule. Unmyelinated axonal spines protrude from the capsules in the direction of the lumen of the gland. Secondary receptor cells are absent (Fig. 12.1a). Such receptor organs are found on the bill of the platypus Ornithorhynchus anatinus and on the snout of the echidna (Tachyglossus aculeatus) and the long-nosed echidna (Zaglossus bruijnii) [27–30].

12.1 Introduction

12.2.3

The Microampullary Organs

The microampullary organs in freshwater fishes and amphibians are microscopically small invaginations of the skin that do not extend beyond the dermis [31]. One lateral line nerve fiber innervates some tens of secondary electroreceptor cells at the bottom of the ampulla (Fig. 12.1b). Microampullae are found in a variety of electrosensitive freshwater species such as the catfish Ictalurus nebulosus, the lamprey Petromyzon marinus, and the sturgeon Acipenser sp. [32] and also in urodeles such as Siren sp. [19, 22].

12.2.4

The Tuberous Organs

The tuberous organs in freshwater fishes can be described as microampullary organs where a plug of cells closes the ampulla lumen (Fig. 12.1c,d). The receptor cells do not connect to the outer world directly. The receptor organs have the appearance of skindeep “knollen.” There are various subtypes, usually innervated by several types of lateral line fibers. Tuberous organs occur in Mormyridae such as Gnathonemus petersii [6, 7], in Gymnarchidae such as Gymnarchus niloticus, and in Gymnotiformes such as Eigenmannia virescens [20]. In Mormyridae there are two subtypes: the knollenorgans (Fig. 12.1c) and the mormyromasts (Fig. 12.1d). They are tuned to discharges emitted by electric organs of the fish itself, respectively by conspecifics and other species with electric organs. Knollenorgans are also called rapid-timing units because they respond with a single spike to a single electric organ discharge. Mormyromasts are amplitudemodulated units because they code the amplitude of the electric organ discharge in a spike train. In other Osteoglossiformes and Gymnotiformes, similar types of receptor organs exist.

12.2.5

The Ampullae of Lorenzini

The ampullae of Lorenzini are centimeters long, jelly-filled subcutaneous canals in marine fishes. Ampullae of Lorenzini form a salt bridge between the receptor cells and the seawater. The receptor cells are clustered in capsules from where they are innervated by lateral line fibers (Fig. 12.1e). Ampullae of Lorenzini occur in elasmobranch fishes such as the dogfish Scyliorhinus canicula, the marine teleost catfish Plotosus lineata, and the living coelacanth Latimeria chalumnae [33–35]. Intermediate forms between long ampullae of Lorenzini and microampullary organs also occur [23, 36–38].

275

276

12 Transduction and Transmission in Electroreceptor Organs

12.3

How is Transduction at Electroreceptor Cells Studied?

The most straightforward method to gain insight into the electrophysiological behavior of electroreceptor cells and their transduction channels is the patch-clamp method [39]. Until now, however, only two preliminary patch-clamp papers have appeared [40, 41]. What is known about transduction channels in electroreceptor cells is based largely on indirect evidence gained through intracellular recordings of receptor cells [3, 42, 43], extracellular recordings of afferent nerves and nerve fibers [33, 44, 45], and morphological studies on the presence of ATPases in receptor cell membranes (Fig. 12.2c) [46]. Electrophysiological recordings have been performed in both intact specimens and preparations. Although it is never emphasized, excision of electroreceptor organs or pieces of skin containing electroreceptor organs changes the organ’s behavior. Usually, both the spontaneous activity and the stimulus-evoked response or sensitivity are reduced. This is seen for instance in microampullary organ preparations [47, 48]. Also, the “bias current” in ampullae of Lorenzini changes if the ampullae are excised [49].

12.4

Current Views on Transduction and Transmission in Electroreceptor Organs 12.4.1

Spontaneous Activity and Modulation of Afferent Activity

Characteristic of all electroreceptor organs is that the primary afferents innervating the receptor cells are spontaneously active. The response to an electrical stimulus consists of an increase or decrease of this spontaneous activity, depending on the polarity of the stimulus. If the exterior of the organism is positive with respect to the interior, the stimulus is called anodal. The inverse stimulus is called cathodal. In teleost fishes an anodal stimulus excites, that is, raises the activity of the primary afferents. A cathodal stimulus has the reverse effect. In elasmobranch fishes, monotremes, amphibians, and the more primitive freshwater organisms, a cathodal stimulus excites and an anodal stimulus inhibits. Most electroreceptor organs have electroreceptor cells in addition to the innervating nerve fibers. An exception is seen in the organs of Monotremata, where receptor cells are absent. The synapse between electroreceptor cells and primary afferents is chemical, as is evident from electron micrographs (Fig. 12.2d). The spontaneous activity of the primary afferents is presumably caused by sustained release of neurotransmitter. To explain the permanent release of neurotransmitter, a depolarizing “bias current” is postulated [71]. The bias current opens voltage-sensitive Ca channels in the presynaptic membrane, resulting in Ca influx and subsequent neurotransmitter release. Externally applied electrical stimuli are superimposed on the bias current and result in a modulation of the afferent spontaneous activity. If excitation is brought about by anodal stimuli, it is thought that the stimulus – an electrical potential difference over the receptor organ – drops almost

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

unattenuated over the basal membrane. If excitation is brought about by cathodal stimuli, the apical membranes of the receptor cells respond regeneratively to the stimulus, thus causing a depolarization of the apical membrane that is strong enough to spread to the basolateral membrane. In both cases voltage-sensitive calcium channels are opened, which is necessary for neurotransmitter release [50]. Although this model has been used for decades to describe electroreceptor organ functioning, there are too many inconsistencies today to leave the concept unchallenged [4, 51, 52].

12.4.2

Monotreme Mucous Gland Electroreceptor Organs

Platypus mucous gland electroreceptor organs, which are representative of monotreme electroreceptor organs in general, lack associated specialized sensory cells (Fig. 12.1a). Afferent trigeminal fibers come together in ampulla-like structures and are directly stimulated by the electrical currents. Bare nerve endings terminating in close proximity to the bottom of a central glandular duct are used to detect electrical stimuli and to generate generator potentials and action potentials. The electrosensory afferents are large and myelinated and surround the bottom of the invagination of the organ in a cuff-like manner. Terminal axonal filaments emerge from the axon to the central pore and provide presumably a low-resistance pathway for the electrical stimulus [28]. Lateral interactions occur between adjacent terminal filaments via a plexus that is directed circumferentially around the duct from the proximal portion of the terminal filament. These circumferential arbors form an interconnecting ring between all terminal filaments and may be used to improve the signal-to-noise ratio of the electroreceptor and thus enhance overall sensitivity [53]. Other types of nerve terminals within each sensory gland are involved in the opening and closing of the pore [29]. Cathodal stimuli excite, while anodal stimuli inhibit spontaneous afferent activity. No additional receptor cells are involved. Latency to pulses in platypus electroreceptor afferents depends on stimulus strength; an average value of 0.8 ms was found [28]. Echidna electroreceptor organs have also been found to be thermosensitive. Apparently the electrosensory afferents are stimulated directly, but nothing is known about the type of ion channel that conducts the stimulus current. The only evidence of the nature of the transduction channels is given by the ongoing spontaneous activity of the primary afferents and by their sensitivity to cathodal or lumennegative stimuli. Since no secondary electroreceptor cells are present, both the spontaneous activity and the sensitivity to cathodal stimuli have to be attributed to the properties of the ion channels in the membrane of the primary afferent. Since the primary afferent generates action potentials, it is likely that a basic set of voltage-sensitive Na and K channels is present, as well as Krest and Cl– channels. It is tempting, however, to think that the mucus in the sensory glands has an essential function. Mucus could serve as a kind of ion exchanger or be involved in binding divalent ions or buffering the micro-environment of the afferents [54]. The convoluted endings might also serve a particular goal. Perhaps the convolutions represent an unknown electrical pathway that inverses the sensitivity of the primary afferents.

277

278

12 Transduction and Transmission in Electroreceptor Organs

Inspiration for this idea is derived from the role of the lamellae in the Pacinian corpuscles, which mediate the directivity of the mechanical stimulus [55, 56]. Another explanation for the sensitivity to cathodal stimuli could be that the homeostatic repair mechanisms that counteract the stimulus take the place of, and override, the original electrical stimulus.

12.4.3

Microampullary Organs in Freshwater Organisms General The microampullary organs have secondary receptor cells at the bottom of the ampulla (Fig. 12.1b). They respond to stimuli between 0.1 and 100 Hz [57, 58]. They are innervated by a single primary afferent with a glutamatergic synapse [47, 59–69]. High Mg blocks postsynaptic activity [62, 66]. The latency between a pulsed stimulus and the primary afferent response is longer than 1 ms [70], which is considered an argument for the chemical nature of the synapse. The postsynaptic membrane contains no GABA receptors since the primary afferents are insensitive to Saffan [68]. Microampullary organs resemble the monotreme electroreceptor organs in that they look like small pores in the skin, visible with the aid of a dissection microscope. They differ in that the primary afferents contact receptor cells. Studies on transduction mechanisms have been performed most often on the catfishes Ictalurus and Kryptopterus. No transduction channels have been identified through patch-clamp experiments, although attempts have been made (Fig. 12.3) [41]. Microampullary organs lack regenerative responses in their receptor cells [9, 71]. Anodal stimuli are excitatory; cathodal stimuli are inhibitory. Stimuli are supposed to act directly on the presynaptic basolateral membranes of the receptor cells to alter calcium permeability and modulate the release of neurotransmitter. Indirect arguments suggest that the great voltage sensitivity does not reside in any properties of the Ca channels but rather in subsequent processes leading to neurotransmitter release [50]. The apical membranes in amphibian ampullary electroreceptors are—in contrast to those in teleosts—postulated to be regenerative since cathodal stimuli are excitatory. The apical membrane is usually specialized and contains microvilli, suggesting that membrane surface enlargement is important for electroreceptor cell functioning. Increased membrane surface facilitates transport through ion channels, pores, or transporter molecules. Electroreceptor cells often contain electron-dense bodies containing Fe or Cd (Fig. 12.4) [31, 72, 73]. This is probably the result of exposure of receptor cells to contaminated water and of detoxification of the receptor cells by metallothioneins. The basolateral membrane of the electroreceptor cells contains voltage-gated Ca channels. Support for this is derived from the morphological demonstration of chemical synapses at the basal membranes and their postulated analogy to neuromuscular synapses. 12.4.3.1

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs Fig. 12.3 Example of patchclamp recordings in electroreceptor cells of Kryptopterus bicirrhis. (A) Cell-attached patch single-channel recording lasting about 30 s. Holding potential Vh = +70 mV. Downward is outward current. Channel conductance ca. 140 pS. (B) Family of I-V curves in a perforated patch experiment with electroreceptor cells. Recording was filtered at 5 kHz; sampling rate 20 kHz; before the variable step, a hyperpolarizing voltage step was applied in order to close the LVA channels. Vh = –70 mV, Vhyp = –100 mV, Vstim = –130 + 14 steps of 25 mV. Courtesy of Mieke L.Struik [41]

279

280

12 Transduction and Transmission in Electroreceptor Organs

12.4.3.2 Patch-clamp Experiments

Struik succeeded in making some cell-attached single-channel recordings from isolated electroreceptor cells of Kryptopterus bicirrhis [41]. The recordings lasted only 30 s and did not allow characterization of the channel type (Fig. 12.3a). Perforated patch-clamp recordings also were made. Of 200 recordings only 8 yielded I-V characteristics demonstrating voltage dependence of I-V curves (Fig. 12.3b). However, further characterization of the kinetics was not achieved. Struik [41] concludes from these efforts that (1) the presence of ion channels has been demonstrated in electroreceptor cells, (2) the voltage-gated channel density is apparently very low, and (3) at least part of the population of ion channels is voltage dependent and consists most likely of Ca(V) channels related to neurotransmitter release [41]. All this agrees with the experience of other investigators that Ca channels usually occur in low density and are difficult to recognize among other channels in a membrane patch [74]. Simultaneous recordings of intracellular calcium and afferent spike activity demonstrated some enigmatic phenomena [5, 41]. If a microampulla is stimulated by a sinusoidal current of 0.01 Hz at 1 lA pp, the intracellular calcium concentration rises and follows the stimulus. However, the level of calcium never falls below the level of the unstimulated ampulla. Further, the activity of the primary afferent does not follow

Fig. 12.4 Electron micrograph of an electroreceptor cell of a microampullary organ in Ictalurus nebulosus. The activity of the electroreceptor was recorded in vivo under normal electrical stimulation while 10 mM CdCl2 was administered apically. After the experiment free Cd2+ ion deposits were marked with the Timm’s reaction [125]. Cadmium is demonstrated everywhere in the receptor cells outside the nucleus. No cadmium is demonstrated in the neighboring supporting cells. Apparently, the apical membrane is leaky, and the Cd2+ ions can reach the basal membrane where voltage-sensitive Ca channels are blocked and release of neurotransmitter is halted [72, 125]. (A) Crosssection through ampulla. Large arrow, lumen; small arrows, dense inclusions containing iron [73] and cadmium. Scale bar: 50 lm. (B) Detail of an electroreceptor cell. N, nucleus; CP, cytoplasm; DI, dense inclusion. The black dots in the cytoplasm represent cadmium ions. Scale bar: 3 lm (Ruud Zwart, Henk J. Herwig, and Rob C. Peters, unpublished)

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

the stimulus and calcium concentration. Nonlinearities have been explained earlier by depletion of the synapses [58]. The transient nonlinearity at strong negative stimulation in Struik’s experiment cannot be explained in this way. Negative stimuli do not cause depletion of neurotransmitter since they suppress neurotransmitter release. Administration of amiloride and TEA suppresses the intracellular calcium response reversibly [5]. The stimulus-evoked Fura-2 ratio changes, while Na and K channels were blocked by amiloride and TEA administration, demonstrated the presence of both Na and K channels and their necessity for electroreception.

12.4.3.3 Indirect Pharmacological Evidence

Since the early days of the discovery of electroreception, it has been postulated that Ca2+ ions play an important part in stimulus transduction. Calcium-channel blockers have been administered at both apical and basal membranes of receptor cells. Na, K, and Cl concentrations of the external medium also have been manipulated. The Cachannel blockers cadmium, verapamil, bepridil, and menthol all suppress spontaneous activity and evoke responses to a certain extent [75, 76]. Moderate changes in the ion concentration of the environment have a relatively large effect on spontaneous activity and evoked responses [16, 62, 77–79]. TTX and TEA administered apically have different effects. TTX suppresses afferent activity above a dose of 1 lM, while TEA has a transient suppressing effect both when administered and when cleared. Unequivocal interpretation of these results is not possible since the site of attack is not clear. The cadmium-administration experiments present the clearest results. If cadmium is administered at the apical membranes of the electroreceptor cells, the response to the stimulus disappears, whereas the spontaneous activity remains present. Morphological experiments demonstrating the presence of intracellular cadmium showed that Cd2+ ions had invaded the receptor cells (and the dense bodies) but did not enter the nucleus and supporting cells. Apparently, Cd2+ions blocked the voltage-sensitive calcium channels at the basal membrane. This leads to several conclusions. First, the apical membrane does not contain voltage-sensitive Ca channels, whereas the basolateral membrane does. Blocking of the Ca channels in the basolateral membrane should cut off the release of neurotransmitter. Stimulus-evoked activity indeed disappears, but the spontaneous activity remains and is apparently not merely caused by neurotransmitter release [72]. Efforts to damage the cytoskeleton of the electroreceptor cells with cytochalasin resulted in changes in the frequency responses that could be explained by changes of apical membrane surface [80–82]. The presence of microvilli at the apical membrane suggests transport processes. Ions such as Cd2+ and other substances such as DASPEI, methylene blue, FURA2, and neutral red seem to enter the receptor cell rather easily [4, 5, 41, 52]. The leakiness of the receptor cells is contradicted by measurements of the input impedance [83]. Since the apical membrane seems to be leaky, it follows that the basolateral membrane represents the high resistance. Nevertheless, there are also indications of a low-resistance pathway from the primary afferents to the ampulla lumen. A metal microelectrode inserted in the lumen of the microampullae picks up action potentials from the

281

282

12 Transduction and Transmission in Electroreceptor Organs

primary afferents [4, 17]. Positioning of the electrode at different spots shows that the receptor cell represents a resistive leak from the afferent to the ampulla lumen [4].

12.4.3.4 The Synaptic Paradox

For more than 30 years it has been postulated that both spontaneous activity and stimulus-evoked responses of the ampullary primary afferent are the direct result of neurotransmitter release. Release of neurotransmitter in turn represents the degree of depolarization of the electroreceptor cells [50]. Inconsistencies with this model popped up frequently but were never integrated. Nevertheless, there is a considerable amount of data that asks for an upgrade of the old model. The inconsistencies of the model are the following. 1. In order to explain the spontaneous activity of the primary afferent, sustained release of neurotransmitter, caused by a “bias current” that permanently depolarizes the presynaptic membrane, has been postulated. Such a DC bias current is inconsistent with the frequency characteristics of the ampullary organs. Either the receptor cell or primary afferent should adapt to a DC bias current through the receptor cell, as it does to external electrical DC stimuli, or the primary afferent should respond to an external DC stimulus. Moreover, what process causes the bias current? So far the evidence for a spontaneous activity-correlated bias current in microampullary organ is negative. Efforts to record the postulated bias current directly in microampullary organs failed [4] or did not support the postulate [84]. Bretschneider et al. [85] measured bias currents in Kryptopterus ampulla openings with the vibrating probe method [86], but they could not demonstrate that currents through the ampulla lumen were correlated with spontaneous activity under normal conditions. Bias currents were –72, +39, and +108 pA in media with 0.2, 1, and 2 mM calcium, respectively (Fig. 12.5). However, bias currents in normal medium were both outward and inward. Estimates of the membrane potential in electroreceptor cells, e.g., in Kryptopterus, are not reliable because the membrane potential is heavily shunted upon impalement by microelectrodes [43, 87, 88]. This is unfortunate because it prevents reliable estimates of the operating range of ion channels. 2. The old model does not account for the effects of variations of the ionic composition of the aquatic environment on both the spontaneous activity and the evoked responses. 3. There is no explanation for the spontaneous activity of the primary afferent in the absence of receptor cells [4] or when release of neurotransmitter is blocked [72, 76, 89]. 4. No explanation is given for the paradoxical behavior of spontaneous activity and evoked responses if the primary afferents are transected [90–92], when the temperature is lowered [52, 70], or when cadmium chloride is administered externally [76]. In such cases the spontaneous activity falls, whereas the stimulus-evoked response remains unchanged (or vice versa). 5. If more than one ampulla connects to a single afferent fiber, the sensitivity to electrical stimuli increases proportionally, but the spontaneous activity remains at the

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

Fig. 12.5 Measurements of the postulated “bias current” that is supposed to be responsible for spontaneous activity in the absence of electrical stimulation. Inset: an ampullary organ of Kryptopterus bicirrhis with receptor cells and vibrating probe electrode [86]. P, probe tip black and its excursions during vibration gray; L, lumen and pore; RC, receptor cells; D, dendritic tree; F, nerve fiber. Bottom trace square-wave signal obtained from the output of the lock-in amplifier by alternating between positions left and right of the pore. Graph: Pore currents at the ampulla lumen in water with 0.2, 1, and 2 mM Ca2+. Note that the calcium content of environmental water shifts the distribution of the measured currents but that in normal water, with normal spontaneous activity, the ampullary current is both inward and outward (Courtesy of Franklin Bretschneider [85])

same level (as well as the variation) [4, 45, 93–96]. According to the old model, convergence of receptor cells onto the primary afferent should have an effect on both. 6. If the spontaneous activity is regular, the sensitivity to stimuli is low. This was found during 18 years of practical tutorials on electroreception in Ictalurus (cf. [17]). All of the items mentioned above point into another direction. Roth’s suggestion [4] that the primary afferent is spontaneously active without neurotransmitter release seems more realistic (cf. [97]), as well as his conclusion that the stimulus directly modulates the afferent activity through a low-resistance path. 12.4.3.5 The Transduction Model Revisited

The good things in the old model are the leaky, low-resistive, apical membrane of the receptor cells and the voltage-sensitive calcium channels in the high-resistive basolateral membrane. The synapse between electroreceptor cells and primary afferents is chemical and glutamatergic. The postsynaptic membrane does not contain GABAsensitive receptor channels. For a better fit of experimental data with a model, we had better not stress the “proportionality” of the chemical synapse, i.e., its role in controlling both spontaneous and stimulus-evoked activity [50]. It is more promising to take Roth’s proposition as

283

284

12 Transduction and Transmission in Electroreceptor Organs

starting point: a low-resistance pathway between ampullary lumen and primary afferent [4]. In the latter model, spontaneous activity is a property of the primary afferent. No chemical synapse is needed (cf. monotremes, [27]). It is possible, however, that postsynaptic NMDA receptors somehow help in sustaining the spontaneous activity in the resting condition. The role of the chemical synapse then would be to establish an electrical contact between electroreceptor cells and primary afferent. If on average presynaptic Ca channels are open and if at the same time postsynaptic receptor channel complexes are open, there exists a low-resistive pathway between the receptor cell and primary afferent. The chemical synapse thus would inevitably add noise to the spontaneous afferent activity and enhance its firing level. Its main role, however, would be to make electrical contact. The stimulus current then would directly modulate the postsynaptic activity by stimulating the neural afferent fiber membrane or the dendritic membrane. A comparable phenomenon has been described in blowfly taste hair dendrites [98, 99]. The frequency domain properties of the response are added by the channel properties of the receptor cell. In this way the receptor cells serve to funnel the stimulus, and the synapse is a switch to open the electrical pathway from the inner of the primary afferents to the ampullary lumen. A similar description might apply to the knollenorgan synapse [9], where the short latency does not match the properties of the chemical synapse. Now that it is known that secondary receptor cells are not

Fig. 12.6 Precipitates in lateral line organs after administration of 1 % neutral red solution and subsequent rinsing. (A) Ampullary organs and ampullary receptor cells in Kryptopterus bicirrhis, ventral fin. Scale bar: 100 lm. The dark dots are electroreceptor cells stained by brown neutral red precipitate. (B) For comparison, a cupula (arrow) with kinocilia of a free neuromast (mechanoreceptor) in Ictalurus nebulous. Scale bar: 200 lm. After [52]

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

needed for electroreception [27], the criteria which electroreceptor organs should meet can be reformulated. First, a cell, whether nerve ending or some other specialized cell, should be well accessible to an electrical current. This implies a low membrane resistance and leaky membranes. This in turn leads to a homeostatic challenge of the cell involved. Second, the homeostatic challenge has to be counteracted. It is likely that the various forms of electroreceptor organs represent to a large extent the speciesspecific adaptations to the homeostatic challenge. It is not unthinkable that ampullary specializations such as invaginations, mucus, and apical microvilli are meant to simulate an intracellular environment near the exposed receptor cell membrane. In this way, an unwanted leak of ions through the apical membrane could be avoided. There is experimental evidence for this view. If ampullary electroreceptors are exposed to a 1 % neutral red solution, a brownish precipitate is formed in the vicinity of the apical membrane (Fig. 12.6). Neutral red is usually used as a pH indicator in solutions with low ion concentrations, where it turns colorless or red. In fish saline, in seawater, or in solutions of 1 mM KCl, however, neutral red precipitates. Thus, the brown precipitate reveals local high concentrations of ions in the ampulla of freshwater species, probably between the microvilli.

12.4.4

Tuberous Organs in Freshwater Fishes

The last review of the operation of tuberous electroreceptor organs dates from 1986 [42]. From then on most studies on electroreception have dealt with morphology, behavior, and central processing of information. It is not surprising that so few investigators have tried to characterize transduction channels in tuberous organs. The tuberous organs are dispersed over, and buried in, the skin and yield one receptor cell per organ, e.g., the knollenorgan in Stomatorhinus and Pollimyrus, less than 10 cells per organ, e.g., in Campylomormyrus and Gnathonemus, or some tens of cells per organ, e.g., in Petrocephalus. They are tuned to stimulus frequencies of 100 to 10,000 Hz. No patch-clamp studies are known of ion channels in tuberous electroreceptor cells. Some investigators were able to successfully dissociate electroreceptor cells from their capsules and plate them out but failed to get a good seal (H.H. Zakon, personal communication). The few facts that are known about transduction channels are inferred from extracellular administration of drugs during electrophysiological recording of either the primary afferent activity or the activity of the receptor cell. Application of drugs showed that the receptor cells of Gymnotus and Gnathonemus are TTX resistant. TTX does not affect the regenerative oscillations of the receptor cells [100]. On the other hand, the ion composition affects the behavior of the receptor cells [101]. Transmission to the afferent nerve is electrical, as is inferred from the 0.3-ms latency in the knollenorgan and the antidrome invasion of spikes into the receptor cell [9, 102]. In spite of morphological evidence of the presence of chemical synapses, high Mg has no influence on the nerve threshold of knollenorgans [103], whereas it does affect mormyromast synapse functioning. Gap junctions between electroreceptor cells and afferent nerve fibers have been demonstrated in tuberous organs of gymnotiforms [104], but

285

286

12 Transduction and Transmission in Electroreceptor Organs

not in those of mormyriforms [104, 105]. The sensitivity of tuberous organs to Ca- and K-channel blockers prompts the presence of Ca and K conductances. The spontaneous activity and “threshold” of stimulus-evoked activity are independent (see [42] p. 164). Other indications for the nature of ion channels come from mathematical simulations. It proved possible to simulate both types of mormyromast receptor cells by incorporating only a calcium-activated K, a voltage-sensitive Ca, and leak channels in the basal membrane and leak channels in the apical membrane of mormyromast electroreceptor cells [106, 107]. Whereas not much is known about the nature of the ion channels in the apical membranes, ion pumps have been demonstrated by histochemical methods in the apical membranes [108], as well as a layer of mitochondria almost “touching” the apical surface of knollenorgans [26]. In the old model the apical membrane of the knollenorgan receptor cells serves only as a capacitance. It neglects possible contributions of ion pumps and the layer of mitochondria just beneath the apical membrane [26, 69]. Postsynaptic ion channels could be glutamatergic. L-glutamate could be the neurotransmitter involved, since it excites afferent fibers when applied to the inner surface of the skin [100] or at synapses [103]. Further sensitivity differences between single receptors and clusters of receptors have been demonstrated in Sternopygus [109]. There is also evidence that nitric oxide is somehow involved in chemical transmission [110]. Summarizing, only little progress has been made in the elucidation of the mechanisms of stimulus transduction in tuberous organs. There is morphological and pharmacological evidence of a chemical synapse, but, on the other hand, electrophysiological evidence suggests immediate electrical contact. Blocking the synapse with Mg is not always effective [103].

12.4.5

Ampullae of Lorenzini in Marine Fish General The ampullae of Lorenzini in marine fish are most appropriate for the study of membrane channels, transduction, and neurotransmission. Lorenzinian ampullae are relatively easy to dissect. A single ampulla can be put in a voltage-clamp setup with an air bridge so that the basal membrane is submerged in one compartment while the lumen is submerged in another compartment. Thus, the lumen is electrically representative of the apical membrane. Both apical and basal membranes can be perfused independently. The responses of apical and basal membranes thus can be recognized. The most work has been done on ampullae of the ray Raja and of the marine teleost Plotosus. For a review until 1990, the reader is referred to a number of excellent books on that topic [19–22, 24]. The following will mainly deal with insight based on more recently acquired data. 12.4.5.1

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

12.4.5.2 Ampullae in Plotosus

Successful voltage-clamp experiments have been performed on Plotosus ampullae, where the excitatory stimulus is anodal. In a series of convincing experiments, it was shown that the basal membrane of the ampullary receptor cells contain one type of Na/K pump [49], two types of voltage-sensitive Ca channels, and a calciumactivated K channel. The presence of the pump was demonstrated by adding ouabain to the basal membrane and by manipulating the K content of the saline. The Ca channels consist of a slow and a fast population. The apical membrane does not contain voltage-sensitive channels [13, 111]. It is inferred that an anodal stimulus in the lumen activates the voltage-sensitive Ca channels in the basal membrane, upon which neurotransmitter is released. The fast component of the Ca current triggers release of neurotransmitter, as well as repolarization by the K current. The slow Ca current is presumably related to the generation of spontaneous activity. In another study the nature of the Ca channels was characterized in a patch-clamp experiment [40]. Whole-cell current clamp recordings showed cell-attached resistances of 20–60 GOhm, which fell to 50–300 MOhm after puncture. This revealed a low apical resistance of the receptor cells. Voltage-clamp recordings showed current suppression after Cd2+ ion administration. Cell-attached patch recordings of basolateral membranes in 50 mM Ba in isotone KCl demonstrated two types of Ca channels: a long lasting, voltage-dependent channel with a conductance of 54 pS and a transient channel of 99 pS [40]. These findings are summarized in Fig. 12.7a.

12.4.5.3 Ampullae in Elasmobranchs

The other studies that gave detailed insight into the transduction and transmission channels of the ampullae of Lorenzini made use of skates and rays. In elasmobranch fishes the ampullae are excited by cathodal stimuli, in contrast to those of the teleost Plotosus. The ampullae of Raja erinacea and Raja ocellata were investigated in a voltageclamp setup with air gap, just like those of Plotosus. Both the basal and the apical faces of the receptor cells could be perfused with various drug-containing solutions. Manipulating the Na concentration near the basal membrane as well as administration of ouabain and clonazepam revealed the presence of Na/K pumps and Na/Ca exchangers in the basal membrane. Further L-type Ca, K, and calcium-gated Cl channels were demonstrated by the administration of, respectively, nitrendipine or Ca2+ ions or S(-)BAY K8644, TEA or Ba2+ ions or 4-AP, and DIDS or niflumic acid. The apical membrane contained L-type Ca channels, but no K channels or N-type Ca channels [12, 112–117]. There is ample evidence that the synapse is glutamatergic [24]. These findings are summarized in Fig. 12.7b.

12.4.5.4 The Synaptic Paradox

The response properties of the ampullae of Lorenzini can be related to properties of ion channels in the basal and apical membranes of the receptor cells, in both elasmobranch and teleost fishes. The results are in agreement with earlier predictions [50, 118]. However, there is still the question of how the spontaneous activity of the primary

287

288

12 Transduction and Transmission in Electroreceptor Organs

Fig. 12.7 Schematic summary of the various transduction channels and ion pumps that have been demonstrated in electroreceptor cells thus far. (A) Ampullae of Lorenzini in the teleost Plotosus. (B) Ampullae of Lorenzini in the elasmobranch Raja (after [12, 13]). (C, D) To illustrate the differences and similarities with cells with comparative topology and different function, the diagrams of mitochondria-rich cells or chloride cells are presented [126]. The most conspicuous difference between transport cells and electroreceptor cells is the synapse with the afferent nerve fiber in electroreceptor cells, which is absent in transport cells. Parallel lines are channels, and circles are pumps or exchangers

12.4 Current Views on Transduction and Transmission in Electroreceptor Organs

Fig. 12.7

Continued

afferents relates to the stimulus-evoked responses. If the bias current is considered as a DC stimulus, adaptation should occur within seconds, but if a bias current passes without adaptation, adaptation to a DC stimulus should not occur. Moreover, just as in the microampullary organs, there are experiments that demonstrate that spontaneous activity and stimulus-evoked responses are based on different cellular processes. This is seen in the Lorenzinian ampullae of the dogfish Scyliorhinus canicula. When Lorenzinian afferents are recorded in vivo in the unanesthetized fish [119], a discrepancy between resting discharge and evoked responses is seen. If the fish stops breathing (after tubocurarine administration) the spontaneous activity slowly drops and can be recovered by stirring the water in which the fish lies. The

289

290

12 Transduction and Transmission in Electroreceptor Organs

Fig. 12.8 Electrophysiological recording of the activity of primary afferents of ampullae of Lorenzini in Scyliorhinus canicula in vivo. The dogfish was not anesthetized but was immobilized by administration of D-tubocurarine. As a result of the administration of the muscle relaxant, the dogfish stopped breathing. Surprisingly the spontaneous activity dropped, whereas the stimulus-evoked response (0.5 Hz, 6 lV cm–1) remained unchanged. Stirring the water brought the spontaneous activity back (Peters, unpublished, technique as in [119]). Scale bar: 10 s; Y-axis arbitrary units

stimulus-evoked response, however, remains unaffected (Fig. 12.8). This can be understood if lack of oxygen slows down the resting discharge but not the stimulusevoked response. The stimulus apparently does not need oxygen-consuming processes to modulate the afferent activity, nor does the stimulus-inverting process at the apical membrane. If both spontaneous activity and stimulus-evoked responses were mediated via the same neurotransmitter releasing mechanism, both would suffer equally from lack of oxygen. Apparently this is not the case. An explanation of the paradox could be that in Lorenzinian ampullae as well, the spontaneous activity is a property of the afferent fiber and that the stimulus-evoked response of the receptor cells stimulates the afferent directly. The chemical synapse would then act only as a switch and not as a proportional throughput system.

12.5

Mucus and Transduction

All electroreceptor organs are characterized by skin-deep invaginations or cavities. The receptive cells, whether nerve cells or secondary receptor cells, are at the bottom and are protected by mucoid substances. In the ampullae of Lorenzini these features appear most pronounced. Further, the receptor cells have apical specializations such as microvilli and, in some species, supplementary kinocilia. The question arises in what way these features are related to electroreceptor cell functioning. Invaginations offer protection. Protection from what? Microvilli and kinocilia enlarge the membrane surface. To what purpose? The apical membrane of the receptor cells is in ampullary organs much smaller than the basolateral membrane. Why reduce the apical surface and enlarge it at the same time? Mucus or jelly has many functions [54]. An explanation that would apply to all types of electroreceptor organs is the following. The elec-

12.6 Conclusions and Open Ends

troreceptive cells form topologically the boundary between the milieu inte´rieur and the milieu exte´rieur. This puts a heavy load on the mechanisms regulating homeostasis. The accessibility of an electroreceptor cell to an electrical current implies homeostatic feedback. Such feedback mechanisms might participate in stimulus transfer. A homeostatic challenge would immediately activate ion pumps. The ion pumps would generate an inverse current that in itself might act as a new stimulus to the afferent fiber. It is not unthinkable that ion pumps are involved in the generation of the spontaneous activity and in the transduction itself. Another way to protect the receptor cell from too-strong electrochemical disturbances would be to replace the milieu exte´rieur with a medium with the properties of the milieu inte´rieur. In that case the homeostatic challenge would be shared by other cells not directly involved in stimulus detection. The K-rich jelly of the ampullae of Lorenzini [120, 121] would fulfill these demands. The microvilli and kinocilia in freshwater fishes would serve a similar purpose. Microvilli would reduce the effective diffusion surface through which the cell loses its ions, whereas at the same time the electrical resistance would be strongly reduced. The simulated milieu inte´rieur is found between the villi and would contain, for instance, many K+ ions. In kinocilia something else would be the case. The kinocilium would be the leaky structure where buffering of the milieu exte´rieur takes place. Inside the kinocilium we will find high conductivity but a very small leaky surface as seen from the cell body. In monotreme organs the mucus in the invagination might serve a similar purpose. The foregoing bears a strongly speculative character. On the other hand, until now there have been no functional explanations for the presence of microvilli or kinocilia, the presence of mucous, or the large number of mitochondria in afferent nerve ending and at particular spots in the receptor cell. It is time for a new approach.

12.6

Conclusions and Open Ends

Summarizing the experimental evidence listed above, the following conclusions emerge: 1. Apparently it is very difficult to perform patch-clamp studies on electroreceptor cells. The cause of lack of channel recordings might be that the channel density is low and that the cells deteriorate fast after isolation. Further, it proves difficult to maintain gigaseals, which might be due to the presence of microvilli and the glycocalyx. Also there are a few cells per receptor organ, except in the ampullae of Lorenzini. 2. Intracellular electrophysiological recordings of electroreceptor cells are not reliable due to an underestimation of the membrane potential as a consequence of electrode water mantle shunting. No reliable estimate of the activation range of ion channels is possible.

291

292

12 Transduction and Transmission in Electroreceptor Organs

3. There is evidence that the apical membrane is extremely permeable to ions and small molecules. There are no indications for specificity of those channels; on the contrary, all evidence points to non-specificity. 4. There is evidence that basal membranes are “conventional” membranes with voltage-sensitive Ca channels for the control of neurotransmitter release. 5. The spontaneous activity of the primary afferents is maintained by a process that is not indispensable for stimulus-evoked responses. 6. There are ATPases involved in the control of the electrochemical homeostasis of the electroreceptor cells. Whereas these ion pumps are often found in basal membranes [122], they have been demonstrated in apical membranes of electroreceptor cells (Fig. 12.2c) [46]. Their role in stimulus transduction has not yet been recognized. 7. The long accepted axiom that the spontaneous activity is caused by a “bias” current has to be revisited. A model where the spontaneous activity of the primary afferents is generated by the postsynaptic membrane itself is preferable. This matches better the frequency characteristics of the electroreceptor organs, the topology of mitochondria, and the firing patterns of afferent fibers after nerve transection, cooling, synapse blocking, and low oxygen. 8. Voltage-sensitive ion channels in apical membranes of electroreceptor cells can explain the sensitivity to cathodal stimuli. Other mechanisms such as feedback activity of ion pumps and unknown resistive pathways cannot be excluded. 9. Sensitivity to cathodal stimuli in monotremes is not mediated by secondary sensory cells. 10. The use of square pulses easily leads to the idea that spontaneously active primary afferents have thresholds. Experiments in Ictalurus nebulosus ampullary organs, however, have shown that there is no threshold. Recovery of the response depends only on the duration of the signal recovery procedure (cf. also [44]). It is better to speak of sensitivity of receptor organs represented by the slope of the input/output curve and of detection threshold in behavioral experiments. 11. The general model of electroreceptor functioning should be upgraded in the sense that the spontaneous activity is primarily based on properties of the primary afferents. The chemical synapses of electroreceptor cells, if present, function as electric switches to give electrical access to the primary afferents. The primary afferents are the real detectors. The function of the receptor cells is to channel the electric stimulus and to protect the sensitive afferents from homeostatic challenges. The ampullary invaginations and the tuberous cavities, as well as the apical microvilli and kinocilia of the receptor cells, serve the same goal. The mucus or jelly found in the invaginations must also buffer the homeostatic challenges.

12.6 Conclusions and Open Ends

Acknowledgments

The authors thank Prof. Dr. Stephan Frings and Dr. Jonathan Bradley for the invitation to write this chapter. Thanks also to all colleagues who were willing to expose their experimental past and to discuss the basics of electroreceptor functioning, in particular Dr. Mieke Struik, Dr. Remco Westerink, Dr. Franklin Bretschneider, Dr. Koos de Kramer, and Dr. Johanniek van der Molen. References 1

2

3

4

5

6

7

8

9

Kandel, E.R., The molecular biology of memory storage: A dialog between genes and synapses. Biosci. Rep. 2001, 21(5), 565–611. Lissmann, H.J., Continuous electric signals from the tail of a fish, Gymnarchus niloticus Cuv. Nature (Lond.) 1951, 176, 201. Dongen, A.M.J.v. and F. Bretschneider, Functioning of catfish electroreceptors: bursting discharge pattern of Kryptopterus electroreceptors elicited by microelectrode impalement. Comp. Biochem. Physiol. A 1984, 77(4), 647–650. Roth, A., Ampullary electroreceptors in catfish: afferent fiber activity before and after removal of the sensory cells. J. Comp. Physiol. 1973, 87, 259–275. Struik, M.L., H.G. Steenbergen, A.S. Koster, F. Bretschneider and R.C. Peters, Simultaneous measurements of calcium mobilization and afferent nerve activity in electroreceptor organs of anesthetized Kryptopterus bicirrhis. Comp. Biochem. Physiol. A 2001, 130(3), 607–613. Bell, C.C., Mormyromast electroreceptor organs and their afferent fibers in mormyrid fish. III. Physiological differences between two morphological types of fibers. J. Neurophysiol. 1990, 63(2), 319–332. Bell, C.C., Mormyromast electroreceptor organs and their afferent fibers in mormyrid fish. II. Intra-axonal recordings show initial stages of central processing. J. Neurophysiol. 1990, 63(2), 303–318. Szabo, T., Spontaneous electrical activity of cutaneous receptors in mormyrids. Nature 1962, 194(4828), 600–601. Bennett, M.V.L., Mechanisms of electroreception. In Lateral line detectors, P.H. Cahn, Editor, Indiana University Press, Bloomington, 1967, p. 313–393.

10

11

12

13

14

15

16

17

18

19

20

Heiligenberg, W., W. Metzner, C.J.H. Wong and C.H. Keller, Motor control of the jamming avoidance response of Apteronotus leptorhynchus: Evolutionary changes of a behavior and its neuronal substrates. J. Comp. Physiol. A 1996, 179(5), 653–674. Heiligenberg, W., Principles of electrolocation and jamming avoidance in electric fish. Springer, Berlin, 1977, 85. Lu, J. and H.M. Fishman, Ion channels and transporters in the electroreceptive ampullary epithelium from skates. Biophys. J. 1995, 69(6), 2467–2475. Sugawara, Y. and S. Obara, Receptor Ca-current and Ca-gated K-current in tonic electroreceptors of the marine catfish Plotosus. J. Gen. Physiol. 1989, 93(2), 343–364. Roth, A., Electroreception in the catfish, Amiurus nebulosus. Z. Vergl. Physiol. 1968, 61(2), 196–202. Roth, A., Wozu dienen die Elektrorezeptoren der Welse? J. Comp. Physiol. 1972, 79(2), 113–135. Peters, R.C. and R.H.S. Westerink, Catfish electroreceptor organ functioning during five days exposure to different calcium environments. Fish Physiol. Biochem. 1999, 21(1), 81–88. Peters, R.C., P.F.M. Teunis, F. Bretschneider and R. Van Weerden, Ampullary electroreceptors in neurophysiological instruction. J. Biol. Educ. 1988, 22(1), 61–66. Cahn, P.H., ed. Lateral line detectors. Indiana University Press, Bloomington, London, 1967, 496. Bullock, T.H. and W. Heiligenberg, eds. Electroreception. Wiley series in biology, ed. R.G. Northcutt, John Wiley & sons, New York etc., 1986, 722. Kramer, B., Electroreception and communication in fishes. Progress in Zoology, ed. W. Rathmayer. Vol. 42. Gustav Fischer, Stuttgart etc., 1996, 119.

293

294

12 Transduction and Transmission in Electroreceptor Organs 21

22

23

24

25

26

27

28

29

30

31

32

33

Fessard, A., ed. Electroreceptors and other specialized receptors in lower vertebrates. Handbook of Sensory Physiology. Vol. III/3, SpringerVerlag, Berlin, Heidelberg, New York, 1974, 333. Moller, P., ed. Electric Fishes, History and Behavior. Fish and Fisheries Series. Vol. 17, Chapman & Hall, London, Glasgow, etc., 1995, 584. Collin, S.P. and N.J. Marshall, eds. Sensory processing in aquatic environments. Springer, New York, 2003, 446. Akoev, G.N. and Y.N. Andrianov, Sensory Hair Cells. Synaptic transmission. Springer, Berlin etc., 1993, 194. Denizot, J.P., Etude histochimique des mucopolysaccharides du mormyromaste (type II de Cordier) chez Gnathonemus petersii, Mormyrides. Histochem. 1971, 28, 305–315. Szabo, T., Anatomy of the specialized lateral line organs of electroreception. In Electroreceptors and other specialized receptors in lower vertebrates, A. Fessard, Editor, Springer, Berlin, 1974, p. 13–58. Iggo, A., U. Proske, A.K. McIntyre and J.E. Gregory, Cutaneous electroreceptors in the platypus: a new mammalian receptor. In Transduction and cellular mechanisms in sensory receptors, W. Hamann and A. Iggo, Editors, Elsevier, Amsterdam, 1988, p. 133–138. Proske, U., J.E. Gregory and A. Iggo, Sensory receptors in monotremes. Philos. Trans. R. Soc. Lond. B Biol. Sci. 1998, 353(1372), 1187–1198. Manger, P.R., J.R. Keast, J.D. Pettigrew and L. Troutt, Distribution and putative function of autonomic nerve fibres in the bill skin of the platypus (Ornithorhynchus anatinus). Philos. Trans. R. Soc. Lond. B Biol. Sci. 1998, 353(1372), 1159–1170. Pettigrew, J.D., P.R. Manger and S.L.B. Fine, The sensory world of the platypus. Philos. Trans. R. Soc. Lond. B Biol. Sci. 1998, 353(1372), 1199–1210. Mullinger, A.M., The fine structure of ampullary electric receptors in Amiurus. Proc. R. Soc. London B 1964, 160, 345–359. Jørgensen, J.M., The morphology of the Lorenzinian ampullae of the sturgeon Acipencer rutheus (Pisces, Chondrostei). Acta Zool. Stockholm 1980, 61, 87–92. Dijkgraaf, S. and A.J. Kalmijn, Versuche zur biologischen Bedeutung der Lorenzinischen Ampullen bei den Elasmobranchiern. Z. vergl. Physiol. 1966, 53, 187–194.

34

35

36

37

38

39 40

41

42

43

44

45

46

Obara, S. and Y. Sugawara Electroreceptor mechanisms in teleost and non-teleost fishes. In Comparative physiology of sensory systems, L. Bolis, K. Keynes, and S.H.P. Maddrell, Editors, Cambridge University Press, London, 1984, p. 509–523. Bemis, W.E. and T.E. Heterington, The rostral organ of Latimeria chalumnae. Morphological evidence of an electroreceptive function. Copeia 1982, 2, 467–471. Whitehead, D.L., I.R. Tibbetts and L.Y.M. Daddow, Distribution and morphology of the ampullary organs of the salmontail catfish, Arius graeffei. J. Morphol. 1999, 239(1), 97–105. Whitehead, D.L., I.R. Tibbetts and L.Y.M. Daddow, Ampullary organ morphology of freshwater salmontail catfish, Arius graeffei. J. Morphol. 2000, 246(2), 142–149. Whitehead, D.L., I.R. Tibbetts and L.Y.M. Daddow, Microampullary organs of a freshwater eel-tailed catfish, Plotosus (tandanus) tandanus. J. Morphol. 2003, 255(2), 253–260. Neher, E. and B. Sakmann, The patch clamp technique. Scient. Amer. 1992, 266(3), 44–51. Sugawara, Y., Calcium-permeable channels in the isolated electroreceptor cells of Plotosus. J. Comp. Physiol. A 1993, 173(6), 745–746. Struik, M.L., A study on transduction and transmission in catfish ampullary electroreceptor organs, in Neuroethology. 2001, Utrecht University: Utrecht. p. 75. Bennett, M.V.L. and S., Obara Ionic mechanisms and pharmacology of electroreceptors. In Electroreception, T.H. Bullock and W. Heiligenberg, Editors, Wiley, New York, 1986, p. 157–181. Bretschneider, F., Transient measurement of the membrane potential of receptor cells of an ampullary electroreceptor in the transparent catfish. J. Physiol. (Lond) 1985, 366(0), C5. Tricas, T.C. and J.G. New, Sensitivity and response dynamics of elasmobranch electrosensory primary afferent neurons to near threshold fields. J. Comp. Physiol. A 1998, 182(1), 89–101. Peters, R.C., R.J. Brans, F. Bretschneider, E. Versteeg and A. Went, Converging electroreceptor cells improve sensitivity and tuning. Neurosci. 1997, 81(1), 297–301. Denizot, J.P., Adenosine triphosphatases in electroreceptor organs (ampullary organs and mormyromasts) of Gnathonemus petersii. Mormyridae. Histochem. J. 1982, 14(2), 239–255.

12.6 Conclusions and Open Ends 47

48

49

50

51

52

53

54 55

56

57

58

59

Andrianov, G.N., F. Bretschneider, R.C. Peters and P.F.M. Teunis, In vitro electroreceptor organs for pharmacological studies. J. Neurosci. Meth. 1992, 44(1), 1–6. Kraaij, D.A., F. Bretschneider and R.C. Peters, Comparison of electroreceptor organ functioning in vivo and in vitro. Prim. Sensory Neuron 1997, 2(1), 1–8. Sugawara, Y., Electrogenic Na-K pump at the basal face of the sensory epithelium in the Plotosus electroreceptor. J. Comp. Physiol. A 1989, 164, 589–596. Bennett, M.V.L. and W.T., Clusin, Transduction at electroreceptors:origins of sensitivity. In Membrane transduction mechanisms, R.A. Cone and J.E. Dowling, Editors, 1979, p. 91–116. Peters, R.C., E.L. Groen, M.M. Vandersluis, P.F.M. Teunis and K.A. Wilhelm, Denervation reveals 2 components of neurotransmission in electroreceptor synapses. Neurosci. 1988, 27(3), 1049–1053. Peters, R.C. and F. Bretschneider, Electroreceptive micro-ampullae in the African mudfish Clarias lazera (Cuv. & Val., 1840). In Sensory Physiology of aquatic lower vertebrates, T. Szabo, Cze´h, G., Editor, Pergamon, Oxford etc., 1981, p. 13–27. Manger, P.R., J.D. Pettigrew, J.R. Keast and A. Bauer, Nerve terminals of mucous gland electroreceptors in the platypus (Ornithorhynchus anatinus). Proc. R. Soc. Lond. B. Biol. Sci. 1995, 260(1357), 13–19. Shephard, K.L., Functions for fish mucus. Rev. Fish Biol. and Fish. 1994, 4(4), 401–429. Loewenstein, W.R. and M. Mendelson, Components of receptor adaptation in a Pacinian corpuscle. J. Physiol. (Lond) 1965, 177, 377–397. Loewenstein, W.R. and R. Skalak, Mechanical transmission in a Pacinian corpuscle. An analysis and a theory. J. Physiol. (Lond) 1966, 1966, 346–378. Peters, R.C. and R.J.A. Buwalda, Frequency response of the electroreceptors (“small pit organs”) of the catfish, Ictalurus nebulosus LeS. J. Comp. Physiol. 1972, 79, 29–38. Bretschneider, F., J.R. De Weille and J.F.L. Klis, Functioning of catfish Kryptopterus bicirrhis electroreceptors: fractional-order filtering and nonlinearity. Comp. Biochem. Physiol. A 1985, 80(2), 191–198. Bretschneider, F. and R.C. Peters, Transduction and transmission in ampullary electroreceptors of catfish. Comp. Biochem. Physiol. A 1992, 103(2), 245–252.

60

61

62

63

64

65

66

67

68

69

Andrianov, G.N., F. Bretschneider and R.C. Peters, Electrophysiological demonstration of N-methyl-D-aspartate receptors at the afferent synapse of catfish electroreceptor organs. Neurosci. 1997, 79(4), 1231–1237. Andrianov, G.N., R.C. Peters and F. Bretschneider, Identification of AMPA receptors in catfish electroreceptor organs. Neuroreport 1994, 5(9), 1056–1058. Teeter, J.H. and M.V.L. Bennett, Synaptic transmission in the ampullary electroreceptor of the transparent catfish Kryptopterus bicirrhis. J. Comp. Physiol. A 1981, 142(3), 371–378. Heijmen, P., J.-P. Denizot and R. Peters, Immunohistochemical distribution of glutamate in the ampullary electroreceptor organ of teleost fish. Comptes Rendus de l’Academie des Sciences Serie III Sciences de la Vie 1994, 317(8), 743–747. Andrianov, G.N., F. Bretschneider, J.P. Denizot, P.S. Heijmen and R.C. Peters, A neurotransmitter role of glutamate in the afferent synapse of electroreceptor organs in the catfish Ictalurus nebulosus. Soc. Neurosci. Abstr. 1994, 20, 1–2. Andrianov, G.N., F. Bretschneider and R.C. Peters, The actions of L-glutamate and its agonists on the ampullary electroreceptor organs of the catfish Ictalurus nebulosus. Comp. Bioch. Physiol. C 1992, 103(1), 65–71. Andrianov, G.N., F. Bretschneider and R.C. Peters, Demonstration of NMDA receptors in catfish electroreceptor organs. Pflueg. Arch. Eur. J. Physiol. 1995, 430(4), 9–12. Andrianov, G.N., F. Bretschneider and R.C. Peters, Mode of operation of ampullary electroreceptor organs of freshwater catfish. Prim. Sensory Neuron 1996(3), 231–242. Peters, R.C., B. Van den Hoek, F. Bretschneider and M.L. Struik, Saffan (R): A review and some examples of its use in fishes (Pisces : Teleostei). Neth. J. Zool. 2001, 51(4), 421–437. Denizot, J.P., D. Dememes, R.J. Wenthold and M. Bensouilah, Immunohistochemical study of glutamate in the specific lateral line sensory organ - electroreceptors - of gymnotid fish (teleost). Thirtieth Annual Colloquium of the Societe Francaise de Microscopie Electronique, Toulouse 1990, 69(2), 40a.

295

296

12 Transduction and Transmission in Electroreceptor Organs 70

71

72

73

74

75

76

77

78

79

80

81

Schaefer, K., H.A. Braun, F. Bretschneider, P.F.M. Teunis and R.C. Peters, Ampullary electroreceptors in catfish (Teleostei): temperature dependence of stimulus transduction. Pflueg. Arch. Eur. J. Physiol. 1990, 417, 100–105. Zakon, H.H., The electroreceptive periphery. In Electroreception, T.H. Bullock and W. Heiligenberg, Editors, Wiley, New York, 1986, p. 103–157. Zwart, R., Electrophysiological changes and histochemical demonstration of intracellular cadmium in catfish electroreceptors after exposure to cadmium. Neth. J. Zool. 1989, 38(4), 215–216. Eigenhuis, C., A combined electrophysiological and morphological study on individual electroreceptor organs of Ictalurus nebulosus, in Comparative Physiology. 1989, Rijksuniversiteit te Utrecht: Utrecht. p. 94. Hille, B., Ionic channels of excitable membranes. Sinauer Associates, Sunderland MA, 1992, 607. Schaefer, K., H.A. Braun, F. Bretschneider, P.F.M. Teunis and R.C. Peters, Ampullary electroreceptors in catfish (Teleostei): temperature dependence of stimulus transduction. Pflueg. Arch. Eur. J. Physiol. 1990, 417(1), 100–105. Peters, R.C., R. Zwart, W.J.G. Loos and F. Bretschneider, Transduction at electroreceptor cells manipulated by exposure of apical membranes to ionic channel blockers. Comp. Biochem. Physiol. C 1989, 94(2), 663–669. Roth, A., Zur Funktionsweise der Elektrorezeptoren in der Haut von Welsen (Ictalurus): Der Einfluss der Ionen im S€ usswasser. Z. Vergl. Physiol. 1971, 75(3), 303–322. Bauswein, E., Effect of calcium on the differentiating operation of the ampullary electroreceptor in Ictalurus nebulosus. J. Comp. Physiol. A 1977, 121(3), 381–394. Zhadan, G.G. and P.M. Zhadan, Effect of sodium potassium and calcium ions on electroreceptor function in catfish. Neirofiziologiya 1975, 7(4), 403–410. Heijmen, P.S. and R.C. Peters, Apically administered cytochalasin B and D decreases sensitivity of electroreceptor organs in the North-American catfish, Ictalurus nebulosus. J. Comp. Physiol. A 1994, 175(3), 279–287. Heijmen, P.S. and R.C. Peters, Sensory variability of ampullary electroreceptor organs in part explained by apical membrane properties. Prim. Sensory Neuron 1995, 1(2), 81–93.

82

83

84

85

86

87

88

89

90

91

92

Heijmen, P.S., M.A.H. Braks, A. Boele and R.C. Peters, Microvilli contribute to the sensitivity and the shape of the frequency characteristics in ampullary electroreceptor organs. Pflueg. Arch. Eur. J. Physiol. 1996, 430(4), 9–12. Bretschneider, F., J. Verwey and P. Heuts, Functioning of catfish electroreceptors - input impedance and stimulus efficiency. Comp. Biochem. Physiol. A 1991, 99(3), 295–299. Bretschneider, F., Influence of fresh-water composition on the functioning of a sensory receptor epithelium. Comp. Biochem. Physiol. A 1988, 90(4), 829–829. Bretschneider, F., R. Creton, M.R. Dohmen, P.S. Heijmen and J.P. Versluis, Offset currents in catfish electroreceptors measured with the vibrating probe. Pflueg. Arch. Eur. J. Physiol. 1995, 430(4), R160. Jaffe, L.F., Extracellular current measurements with a vibrating probe. Trends Neurosci. 1985, 8, 517–521. Lassen, U.V., A.-M.T. Nielsen, L. Pape and L.O. Simonsen, The membrane potential of Ehrlich ascites tumor cells microelectrode measurements and their critical evaluation. J. Membr. Biol. 1971, 6, 269–288. Lassen, U.V. and B.E. Rasmussen, Use of microelectrodes for measurement of membrane potentials. In Membrane transport in biology I, E. Tosteson, Editor, Springer, Berlin, 1978, p. 169–204. Struik, M.L., F. Bretschneider and R.C. Peters, Spontaneous nerve activity and sensitivity in catfish ampullary electroreceptor organs after tetanus toxin application. Pflueg. Arch. Eur. J. Physiol. 2002, 443(5–6), 903–907. Teunis, P.F., W. Vredevoogd, C. Weterings, F. Bretschneider and R.C. Peters, The emergence of electroreceptor organs in regenerating fish skin and concurrent changes in their transduction properties. Neurosci. 1991, 45(1), 205–212. Teunis, P.F.M., F. Bretschneider and R.C. Peters, Denervation changes the transmission properties of electroreceptor sensory synapses. Comp. Biochem. Physiol. A 1989, 94(4), 647–652. Peters, R.C., E.L. Groen, M.M.v.d. Sluis, P.F.M. Teunis and K.A. Wilhelm, Denervation reveals two components of neurotransmission in electroreceptor synapses. Neurosci. 1988, 27(3), 1049–1053.

12.6 Conclusions and Open Ends 93

94

95

96

97

98

99

100

101

102

103

104

Peters, R.C. and F. Mast, Information processing by ampullary receptors: parallel convergent inputs. Comp. Biochem. Physiol. 1983, 76A(1), 143–151. Peters, R.C. and S.v. Ieperen, Resting discharge and sensitivity of ampullary electroreceptors in Clarias gariepinus related to convergence ratio during ontogeny. Brain Behav. Evol. 1989, 34(1), 43–47. Teunis, P.F.M., F. Bretschneider and R.C. Peters, Convergence ratio and transduction in catfish electroreceptive organs. Comp. Biochem. Physiol. A 1990, 97(3), 405–410. Teunis, P.F.M., F. Bretschneider, C. Van Groeningen, R.C. Peters and J.J.M. Bedaux, Quantitative aspects of transduction in an electroreceptor organ studied by means of experimental manipulation of the interspike interval. Neurosci. 1991, 42(1), 283–290. Bretschneider, F., R.C. Peters, P.H. Peele and A. Dorresteijn, Functioning of catfish Ictalurus nebulosus electroreceptors. Statistical distribution of sensitivity and fluctuations of spontaneous activity. J. Comp. Physiol. A 1980, 137(3), 273–280. Kramer, J.J.d. and J.N.v.d. Molen, Current clamping amplifier. Med. Biol. Eng. Comput. 1979, 17, 407–409. Kramer, J.J.d. and J.N.v.d. Molen The influence of very small currents on the reception and transduction of chemical stimuli in insect contact chemoreceptors. In Olfaction and taste VI, J. LeMagnen and P. Mac Leod, Editors, Information retrieval, London, 1977, p 355. Zipser, B. and M.V.L. Bennett, Tetrodotoxin resistant electrically excitable responses of receptor cells. Brain Res. 1973, 62, 253–259. Peters, R.C. and J.P. Denizot, Miscellaneous features of electroreceptors in Gnathonemus petersii (G€ unther, 1862) (Pisces, Teleostei, Mormyriformes). Belgian J. Zool, 2004, 134 (2/1), bi–bb. Szabo, T., Activity of peripheral and central neurons involved in electroreception. In Lateral line detectors, P.H. Cahn, Editor, Indiana University Press, Bloomington, 1967, p. 295–311. Steinbach, A.B. and M.V.L. Bennett, Effects of divalent ions and drugs on synaptic transmission in phasic electroreceptors in a mormyrid fish. J. Gen. Physiol. 1971, 58(5), 580–598. Srivastava, C.B.L., Morphological evidence for electrical synapses of “gap” junction type in another vertebrate receptor. Experientia 1972, 28, 1029–1030.

105 Derbin, C. and T. Szabo, Ultrastructure of an

106

107

108

109

110

111

112

113

114

115

electroreceptor (knollenorgan) in the Mormyrid fish Gnathonemus petersii. J. Ultrastruct. Res. 1974, 22, 469–484. Shuai, J., Y. Kashimori, O. Hoshino, T. Kambara and G. von der Emde, Electroreceptor model of weakly electric fish Gnathonemus petersii: II. Cellular origin of inverse waveform tuning. Biophys. J. 1999, 76(6), 3012–3025. Shuai, J., Y. Kashimori and T. Kambara, Electroreceptor model of the weakly electric fish Gnathonemus petersii. I. The model and the origin of differences between A- and Breceptors. Biophys. J. 1998, 75(4), 1712–1726. Denizot, J.P., Adenosine triphosphatases (ATP-A) on sensory cell plasmic membranes of tuberous organs (electroreceptors) of Gnathonemus petersii (Mormyridae). Cytochemical Study. Cell. Mol. Biol. 1980, 26(6), 653–662. Sanchez, D.Y. and H.H. Zakon, The effects of postembryonic receptor cell addition on the response properties of electroreceptive afferents. J. Neurosci. 1990, 10(1), 361–369. Turner, R.W. and L.L. Moroz, Localization of nicotinamide adenine-dinucleotide phosphate- diaphorase activity in electrosensory and electromotor systems of a Gymnotiform teleost, Apteronotus leptorhynchus. J. Comp. Neurol. 1995, 356(2), 261–274. Sugawara, Y., Two Ca current components of the receptor current in the electroreceptors of the marine catfish Plotosus. J. Gen. Physiol. 1989, 93(2), 365–380. Broun, G.R., V.I. Govardovskii and V.L. Cherepnov, Action of calcium and potassium channels blockers on changes in transepithelial potential and spike responses of Lorenzinian ampullae in the black sea skate Raja clavata. Neirofiziologiya 1985, 17(5), 652–660. Lu, J. and H.M. Fishman, Linear properties of electroreceptive ampullary epithelium isolated from skates show underlying negative conductance behavior. Biophys. J. 1994, 66(2), 6–10. Lu, J. and H.M. Fishman, Interaction of apical and basal membrane ion channels underlies electroreception in ampullary epithelia of skates. Biophys. J. 1994, 67(4), 1525–1533. Lu, J. and H.M. Fishman, Localization and function of the electrical oscillation in electroreceptive ampullary epithelium from skates. Biophys. J. 1995, 69(6), 2458–2466.

297

298

12 Transduction and Transmission in Electroreceptor Organs 116 Lu, J. and H.M. Fishman, Electrical oscillati-

117

118

119

120

121

ons occur in the basal membranes of electroreceptive ampullary epithelium from skates. Biophys. J. 1995, 68(2), 12–16. Lu, J. and H.M. Fishman, Identification and role of Cl channels in basal membranes of electroreceptive ampullary epithelium from skates. Biophys. J. 1996, 70(2), A69. Broun, G.R. and V.I. Govardovskii, Electrical model of the electroreceptor of the ampulla of Lorenzini. Neirofiziologiia 1983, 15(3), 235–241. Peters, R.C. and H.-P. Evers, Frequency selectivity in the ampullary system of an elasmobranch fish (Scyliorhinus canicula). J. Exp. Biol. 1985, 118, 99–109. Murray, R.W. and T. Potts, The composition of the endolymph and other body fluids of elasmobranchs. Comp. Bioch. Physiol. 1961, 2, 65–75. Okitsu, S., S.-I. Umekita and S. Obara, Ionic composition of the media across the sensory epithelium in the ampullae of Lorenzini in the marine catfish, Plotosus. J. Comp. Physiol. 1978, 126, 115–121.

122 Marshall, W.S., S.E. Bryson and T. Luby,

123

124

125

126

Control of epithelial Cl– secretion by basolateral osmolality in the euryhaline teleost Fundulus heteroclitus. J. Exp. Biol. 2000, 203, 1897–1905. Waltman, B., Electrical properties and fine structure of the ampullary canals of Lorenzini. Acta Physiol. Scand. 1966, 66(Suppl. 264), 1–60. Dotterweich, H., Bau und Funktion de Lorenzinischen Ampullen. Zool. Jahrb. 1932, 50, 347–418. Danscher, G., Histochemical demonstration of heavy metals. A revised version of the sulphide-silver method suitable for both light and electronmicroscopy. Histochem. 1981, 71, 1–16. Marshall, W.S., Na+, Cl–, Ca2+ and Zn2+ transport by fish gills: retrospective review and prospective sysnthesis. J. Exp. Zool. 2002, 293, 264–283.

299

Index a

b

a-gustducin 164, 170 a-tubulin 14 acid-gated channel (ASIC) see ASIC action potentials – burst; vertebrate ORNs 104 – firing frequency, vertebrate ORNs 103, 105 Ad fibers 236, 239, 241 adenylyl cyclase – type III 108 aldosterone 156 allodynia 257 amiloride 34, 41, 57, 62, 67, 112, 155, 158 amiloride sensitive Na+ channel (ENaC) see EnaC amiloride-insensitive channel 158, 161 amino acid taste 163 – amino acid-activated channel 172 – T1R receptors (T1R1 and T1R3) 163 amphid organ 75 ampullae of Lorenzini 275, 286 angina 65 ANKTM1 242, 245 arachidonic acid 140, 141, 191 arginine vasopressin (AVP) 156 arrestin 185 – light-dependent translocation 197 ASIC 24, 58 – assembly into channels 61 – gating 63 – in CNS 67 – knockout mice 66, 67 – mechanosensation 66 – pain 65, 254 – pharmacology 62 – properties of 60 – splice variants 60 – taste 67, 161 – temperature sensation 244 ATP-gated ion channels (P2X) 255

b-tubulin 13 bacterial lipopolysaccharide 262 basal cells – vertebrate olfaction 102 Behavioral assays – chemotaxis 80 – feeding 82, 87 – pain 66 – repulsion 81 – thermotaxis 82 bitter 163 – PLCb2 164 – quinine-activated channel 171 – T2R receptors 139, 163 bradykinin 245, 258 bulbectomy 102, 108

c C fibers 236, 239 C. elegans – appealing genetic organism 73 – behavioral assays 79 – chemosensory cells 77 – chemosensory organs 75, 77 – chemotaxic response 83 – culturing cells 4 – electrophysiological recordings 4 – model system for touch 3 – monitoring intracellular calcium 4, 79 – repulsive response 85 – thermotaxis 82 – touch sensation screen 7 Ca2+ diffusion – vertebrate OSNs 126 Ca2+ extrusion 209 Ca2+ stores – vertebrate OSNs 118 Ca2+-activated Cl- channel – activation 119 – amplification of CNG current 119, 123

Transduction Channels in Sensory Cells. Edited by S. Frings and J. Bradley Copyright ª 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30836-9

300

Index – blocked by niflumic acid 124 – channel density 122 – in olfactory transduction 118 – K1/2 for Ca2+, frog, rat 119, 121 – maximal open probability 124 – regulation 122 – single channel properties 121 – voltage dependent 121 Ca2+-ATPase – vertebrate OSNs 123 Ca2+:calmodulin kinase II (CaMKII) – adaptation of odor-induced response 109 – circadian modulation 212, 224 – regulation of ACIII 109 caged cAMP 116 calmodulin 42, 43, 115, 187, 189, 208 calmodulin-binding site – Baa motif 117 – IQ-type 118 cameleon 16, 79 cAMP – caged cAMP 123 – second messenger vertebrate olfaction 103 – taste cells 169, 170 Cd2+ 159 cGMP – second messenger C. elegans chemosensation 83 – second messenger C. elegans thermotaxis 86 – second messenger phototransduction 209 cGMP-dependent protein kinase (PKG) 83 cholera toxin 106 Cl reversal potential – taste cells 161 – vertebrate OSNs 119 CNG, C. elegans – EC50 cGMP 89 – predicted subunits in C. elegans genome 89 – TAX channels 75, 244 CNG, invertebtate photoreceptors 213 CNG, olfactory sensory neurons (OSNs) – channel densities, rat and frog 111, 124 – cloning 114 – K1/2 cAMP, rat 111 – modulatory subunits 114 – negative feedback 115 – organic blockers 112 – permeation of Ca2+ 112 – single channel conductance, rat and frog 111

– stoichiometry 114 – toad 110 CNG, parietal eye photoreceptors 213 CNG, pinealocytes 212 CNG, rod and cone phototransduction – circadian modulation 224 – density 209 – glutamic acid-rich part of B1a (GARP) 219 – interaction with other proteins 219 – ion selectivity 208 – modulation by Ca2+ 221 – modulation by phosphorylation 223, 224 – modulation by retinal 223 – mutants that cause visual dysfunction 224 – structure of subunits 215 – topology and subunit stoichiometry 215 CNG, taste – taste cells 170 cochlea 31 colchicine 5 collagen 15, 20 – mec-5 15 – sup-20/let-2 20 Crumbs 181 cuticle 6 cyclic nucleotide-gated channel (CNG) see CNG cyclooxygenase (COX) 260 cytochrome P-450 100

d DAG 86, 140, 141, 144, 148, 164, 184, 191 DAG kinase (rdg2) 192 dark current 209 degenerin family (DEG) 2, 3, 9, 20, 24, 58, 66, 156 del-1 21 diacyl glycerol (DAG) see DAG dihydrostreptomycin 34 diltiazem 40, 112 dorsal root ganglia (DRG) see DRG DRG 25, 59, 61, 65, 237 Drosophila compound eye 181 Drosophila mutants 184 – phototransduction cascade 184 – retinal degeneration 184 Drosophila photoreceptor cells 181 – visual pigments they express 181 Drosophila phototransduction cascade 184 – multimolecular signaling complex 186, 188 – protypical G-protein coupled 184 Drosophila rhabdomere 144, 181

Index

e EGL-4 83, 93 EGL-19 16 Eigenmannia virescens 274 electroreceptor cells 276 – patch clamp experiments 280 – pharmacomogy 281 – synaptic paradox 282, 287 electroolfactogram (EOG) 102 electroreceptor cell transduction 272, 276, 283 electroreceptor organs 274 ENaC 1, 3, 7, 10, 24, 58, 155 – heteromultimer 156 – regulation of expression 156 – selevtivity ratio 158 – unitary conductance 157 endolymph 38

f fat-activated channels 173 feeding behavior 87 fluorescent calcium-binding reporter 16, 18 fluorescent resonance energy transfer 16 FMRFamide 58, 63

g Gai2 140, 148 Gao 140, 148 gating compliance 36 gentle body touch 18 Gnathonemus petersii 273 Golf 108 Gq 179, 184 – light-dependent translocation 197 – pain sensitization 258 guanylyl cyclase – EGL-4 83 – GCY-6 and GCY-7 85 – GCY-8 87 – GCY-32 88 – phototransduction 209 guanylyl cyclase activating protein (GCAP) 209 gustatory neurons 153

h hair cells 31 – Ca2+ regulation of adaptation 42 HCN 68, 161, 262 him-4 6, 14 Hofbauer-Buchner eyelet 181 hyperalgesia 257

hyperpolarization-activated channels (HCN) see HCN

i Ictalurus 274 INAD 186 inositol-1,4,5-trisphosphate (IP3) see IP3 insulin-like growth factor (IGF-1) 223 IP3 141, 164, 184 ischemia 65, 68

k Kryptopterus bicirrhis 273 Kunitz-type domain 14, 15

l L-type voltage-gated channel 16 La3+ 189, 196 lactic acid 65 lipid transfer protein 23 locomotory defects 21

m mantle 6, 14, 19 MEC-1 – formation of mantle 14 MEC-2 11, 16 – associates with MEC-4 and MEC-10 11 – influences MEC-4/MEC-10 channel activity 11 MEC-4 and MEC-10 1, 7, 16 – form a functional ion channel 10 – loss of function mutants 7 – not gated by mechanical forces 10 – protein structure 8 MEC-5 15 MEC-6 13, 16 – Paraoxonase 11 – potentiates MEC-4 channel activity 13 MEC-7 13 – b-tubulin 13 – GTP binding and hydrolysis 14 MEC-9 15 MEC-12 14 – a-tubulin 14 mechanically gated channels 2 – heteromeric complex 9 – microtubule network 5 – stoichiometry 9 mechanotransducing complex 19 mechanotransduction 1 – in C. elegans 1 – ion channel complex 1, 9 – TRPV4 256

301

302

Index membrane potential – resting potential; vertebrate ORNs 104 – sustentacular cells 101 MET channel – adaptation 41 – blockers 34, 40 – distribution 47 – gating kinetics 33 – ion selectivity 38 – permeability ratio for Ca2+ 38 – single-channel conductance 45 – three-state schemes 36 microampullary organ 275, 278 mucus – vertebrate olfaction 100, 101 mucus glands 274 myosin-1c 41

n Na+/Ca2+ exchanger 123 – Na+ dependence 123 Na+/Ca2+-K exchanger 209 – interaction with CNG channel 220 nerve growth factor (NGF) 262 neuropeptide FF 63 NINAC 188 nitric oxide synthase (NOS) 211 nociceptor modulators 264 – ATP 264 – Bv8/prokineticin 264 – glutamate 265 – norepinephrine 265 – proteases 264 nociceptor modulators, long-term 265 – GDNF 266 – NGF 266 nociceptor sensitization, inflammatory 257 – bradykinin and PKC 258 – nerve growth factor 262 – prostaglandins and PKA 260 – protons 263 nociceptors 236, 252 NompC 24, 48 NPPB 159, 161

o ocelli 182 OCR-2 75, 88 odorant receptors – quantal evoked-odor activation 109 – response selectivity 109 – sequences cloned 109 ODR-3 83, 86 ODR-10 83

olfactory epithelium 99 – illustration 100 olfactory marker protein 102 olfactory receptor neurons – oscillations of intracellular voltage 104 – vertebrate olfaction 100 ommatidium 181 opiates 262 Ornithorhynchus anatinus 274 OSM-9 24, 48, 75 – function in repulsive neurons 88 – promotes sensory adaptation 77 OSM-9 capsin receptor related (OCR-2) see OCR-2

p P2X3 244, 255 pain 251 papillae – types and innervation 153 paraoxonase 11 PDZ domain 186 pertussis toxin 106 phasmid organ 78 pheromones 135 – aggressive behavior 146 – gender discrimination 147 phospho-tyrosine phosphatase (PTP2) 223 phospholipase C 140 phospholipase C (PLC) see PLC phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) 140, 144, 164, 168, 184, 192, 245 photoreceptors – two types 180 phototransduction, rod and cone 209 PLC 140, 164, – PLCb 184, 188, 258 Plotosus 274 polyunsaturated fatty acids (PUFAs) 191 press stimulus 18 proprioception 21 prostaglandins 260 protein kinase C – Drosophila eye-specific (ePKC, INAC) 186, 191 protein tyrosine kinase (PTK) 223 protofilament microtubules 5, 13 – mec-7 and mec-12 13, 14 psalmotoxin 1 (PcTx1) 62

r receptor potential – adaptation 115, 116

Index – hair cell 31 – odorant dose-response 105 – vertebrate olfaction 99 – VNO 137 recoverin 209 retinitis pigmentosa 181, 225 rhodopsin kinase (RhK) 185, 209 rhodopsin phosphatase 185

s salt taste 154 – amiloride-insensitive pathway 158 – amiloride-sensitive conductance 157 sense of touch 2 – molecular model 17, 19 signal transduction – C elegans chemosensation 83 – vertebrate olfaction 100, 106 – VNO 140 sodium reabsorption 58 sour taste 159 – NPPB-sensitive Cl- channel 161 – proton blocked channels 162 – proton permeable or gated channels 161 stereocilia 31 – displacement of 33 stomatin 10, 11, 16, 21 store operated channel (SOC) 164, 165 – TRPM5 164 stroke 68 suction pipette recording – vertebrate ORNs 105 sustentacular cells 100 sweet taste – PLCb2 164 – sugar-activated channel 172 – T1R receptors (T1R2 and T1R3) 163

t Tachyglossus aculeatus 274 TASK 69, 163 taste bud 153 taste qualities 154 – salt 154 – sour 159 – sweet and bitter 163 – umami 163 taste receptor cells 153 – complex stimuli 163 – GPCR signaling 163 – ionic stimuli 154 taste receptor proteins 163 TAX channel 75, 83 – EC50 cGMP and cAMP 89

– Regulation 93 – structure 88 temperature sensitive neurons 236 – temperature response characteristics 237 temperature transduction 245 thermally-gated ion channels 236, 253 – in non mammalian species 244 – non TRP channels 243 – TRP channels 239 tip-link 31, 34 touch receptor neurons 5 – protofilament 5 – specialized extracellular matrix 6 – ultrastructural features 5 transducisome 188 – singnaling speed 180, 189 transient receptor potential channel (Trp) see TRP TREK 243 trichromatic vision 148 TRP channels – blocked by La3+ 189 – characteristics 190 – gating 191 – implicated in touch 3, 23 – in C. elegans behavior 77 – in detection of pheromones 135 – in Drosophila phototransduction 179, 184, 189 – in hair cells 31, 48 – in taste 164 – subfamilies 195 TRPC subfamily 135, 141 – hTRPC3 and hTRPC6 144, 191 – TRPC2, essential for pheromone transduction 144 – TRPC2, knockout mice 146, 148 – TRPC2, mechanism of activation 144 – TRPC2, pseudogene in humans 147 – TRPC2, structure and expression of 143 TRPL subfamily 179 – calmodulin binding 189 – characteristics 190 – gating 191 – in Drosophila phototransduction and adaptation 189 – Light-dependent translocation 179, 196 TRPM subfamily 141 – TRPM5 164 – TRPM8 242, 254 TRPN subfamily 31, 48 – ANKTM1 (TRPN1) 244, 254 – NOMPC 24, 48

303

304

Index TRPV subfamily 24, 48, 49, 68, 75 – TRPV1 239, 253 – TRPV2 241, 253 – TRPV3 and TRPV4 242, 256 TTX-insensitive Na channel 257, 261 – Nav1.8 (SNS) 257 – Nav1.9 (SNS2) 257 tuberous organ 275, 285 TWIK-rekated acid sensitive K+ channels (TASK) see TASK Two-pore potassium channel (TREK) see TREK

u umami 163 – mGluR4 164 – PLCb2 164 unc-1 21 unc-8 21 unc-24 23 unc-105 15, 20

VNO 135 – anatomy 136 – field potentials 137 – human evolution 147 VNO receptor genes 139, 148 VNO receptor neurons – Ca2+ responses 138 – detection threshold 138 – patch-clamp recordings 138 – spike frequency 137 VNO sensory transduction 140 – second messengers 140 – transduction channel (TRPC2) 141, 149 vomeronasal organ (VNO) see VNO

w water-activated channels 173 whole cell patch-clamp – Plotosus ampullae 287$ – vertebrate ORNs 105

z v visual dysfunction – achromatopsia 227 – retinitis pigmentosa 224

Zaglossus bruijnii 274 Zn2+ 62, 68