Handbook of Neurotoxicology, Volume 1

Handbook of Neurotoxicology Volume I Edited by Edward J. Massaro Humana Press HANDBOOK OF NEUROTOXICOLOGY SECTION...

Author: Edward J. Massaro

160 downloads 1833 Views 13MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Handbook of

Neurotoxicology Volume I Edited by

Edward J. Massaro

Humana Press

HANDBOOK OF NEUROTOXICOLOGY

SECTION EDITORS David J. Adams University of Queensland, St. Lucia, Australia

Daniel G. Baden UNC Wilmington, Wilmington, NC

Jeffrey R. Bloomquist Virginia Polytechnic Institute and State University, Blacksburg, VA

Marion Ehrich Virginia-Maryland Regional College of Veterinary Medicine, Blacksburg, VA

Tomás R. Guilarte Johns Hopkins University, Baltimore, MD

Alan Harvey University of Strathclyde, Glasgow, UK

HANDBOOK OF NEUROTOXICOLOGY Volume I Edited by

EDWARD J. MASSARO The National Health and Environmental Effects Research Laboratory, Research Triangle Park, Durham, NC

HUMANA PRESS TOTOWA, NEW JERSEY

© 2002 Humana Press Inc. 999 Riverview Drive, Suite 208 Totowa, New Jersey 07512 www.humanapress.com All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from the Publisher. The content and opinions expressed in this book are the sole work of the authors and editors, who have warranted due diligence in the creation and issuance of their work. The publisher, editors, and authors are not responsible for errors or omissions or for any consequences arising from the information or opinions presented in this book and make no warranty, express or implied, with respect to its contents. Cover illustration: Production Editor: Jessica Jannicelli. Cover design by Patricia F. Cleary. For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the above address or at any of the following numbers: Tel.: 973-256-1699; Fax: 973-256-8341; E-mail: [email protected] or visit our Website: http://humanapress.com Due diligence has been taken by the publishers, editors, and authors of this book to ensure the accuracy of the information published and to describe generally accepted practices. The contributors herein have carefully checked to ensure that the drug selections and dosages set forth in this text are accurate in accord with the standards accepted at the time of publication. Notwithstanding, as new research, changes in government regulations, and knowledge from clinical experience relating to drug therapy and drug reactions constantly occurs, the reader is advised to check the product information provided by the manufacturer of each drug for any change in dosages or for additional warnings and contraindications. This is of utmost importance when the recommended drug herein is a new or infrequently used drug. It is the responsibility of the health care provider to ascertain the Food and Drug Administration status of each drug or device used in their clinical practice. The publisher, editors, and authors are not responsible for errors or omissions or for any consequences from the application of the information presented in this book and make no warranty, express or implied, with respect to the contents in this publication. This publication is printed on acid-free paper. ∞ ANSI Z39.48-1984 (American National Standards Institute) Permanence of Paper for Printed Library Materials. Photocopy Authorization Policy: Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Humana Press Inc., provided that the base fee of US $8.00 per copy, plus US $00.25 per page, is paid directly to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [0-89603-795-9/02 $10.00 + $00.25]. Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1 Library of Congress Cataloging-in-Publication Data Handbook of neurotoxicology / edited by Edward J. Massaro. p. cm. Includes bibliographical references and index. ISBN 0-89603-795-9 (alk. paper) 1. Neurotoxicology--Handbooks, manuals, etc. I. Massaro, Edward J. RC347.5 .N4857 2001 616.8’047--dc21 2001039605

PREFACE Neurotoxicology is a broad and burgeoning field of research. Its growth in recent years can be related, in part, to increased interest in and concern with the fact that a growing number of anthropogenic agents with neurotoxic potential, including pesticides, lead, mercury, and the polytypic byproducts of combustion and industrial production, continue to be spewed into and accumulate in the environment. In addition, there is great interest in natural products, including toxins, as sources of therapeutic agents. Indeed, it is well known that many natural toxins of broadly differing structure, produced or accumulated for predatory or defensive purposes, and toxic agents, accumulated incidentally by numerous species, function to perturb nervous tissue. Components of some of these toxins have been shown to be useful therapeutic agents and/or research reagents. Unfortunately, the environmental accumulation of some neurotoxicants of anthropogenic origin, especially pesticides and metals, has resulted in incidents of human poisoning, some of epidemic proportion, and high levels of morbidity and mortality. Furthermore, an increasing incidence of neurobehavioral disorders, some with baffling symptoms, is confronting clinicians. It is not clear whether this is merely the result of increased vigilance and/or improved diagnostics or a consequence of improved health care. In any case, the role of exposure to environmental and occupational neurotoxicants in the etiology of these phenomena, as well as neurodegenerative diseases, is coming under increasing scrutiny and investigation. Recognition and utilization of environmental (in the broadest sense) information comprise the currency of life. Therefore, the effects of perturbation of these critical capacities deserve thorough investigation. The acquisition of information, and its processing, storage, retrieval, and integration leading to functional outputs, are fundamental nervous system functions. It should not be surprising, then, that structural, functional, and evolutionary research has revealed that even “simple” nervous systems are immensely complex. On the systems level, the intact nervous system is an exquisite example of integration within the context of a continuously evolving, apparently infinitely programmable and regulatable hierarchical input/output system of complex chemical structure. However, as the complexity of nervous systems has increased, so has their vulnerability to chemical and physical insult. In part, this is a consequence of loss of regenerative capacity. Living systems have evolved to function within relatively narrow ranges of environmental conditions. Perturbation beyond the limits of the range of a given system can result in irreversible damage manifested as loss of function or viability. Also, the nervous tissue of more highly evolved organisms is particularly refractory to regeneration. But, with complexity has come an increased capacity for compensability. Albeit often limited and difficult to achieve, through learning and recruitment, compensation can bypass irreversible damage allowing, to varying degrees, recovery of function. The developing brain, in particular, is endowed with immense plastic potential. Unfortunately, the efficiency of both homeostatic and compensatory mechanisms progressively diminishes as a function of aging. Indeed, a large body of literature indicates that humans generally lose memory with age and the magnitude and rate of loss are highly variable among individuals. In addition, data obtained through the medium of testing protocols, and supported by evidence obtained from functional neuroimaging studies, indicate that not all types of v

vi

Preface

memory are affected equally. Depending on the task, such studies show that, compared with younger adults, older adults can display greater or lesser activity in task-associated brain areas. Conceivably, the increases in activity may be the result of the input from compensatory mechanisms. In any case, age-related diminished mental capacity is a complex function of the interaction of genetic constitution and environmental factors. The type, magnitude, duration, and period of exposure in the life cycle to the latter can impact the functional status of the aging nervous system. Major windows of vulnerability occur during development, when target sizes are small and defense mechanisms immature, and in post-maturity, following decline of the functioning of compensatory and defense mechanisms along with increased duration of exposure. Intellectually, we may appreciate that thermodynamics dictates that, as a function of population size, environmental pollution will increase. However, do we appreciate that, in the short-run, if a connection between environmental pollution and nervous system damage exists, the incidence of nervous system damage will increase as the population increases? Likewise, as life span increases, exposure to neurotoxicants will increase and, it is not unreasonable, therefore, to predict that the incidence of neurodegenerative diseases also will increase. Are these phenomena self-limiting? If not, can we estimate the magnitude of these problems that ensuing generations will have to face? With time, sufficient funding, and manpower, it may be possible to solve many of these problems. Indeed, we must. If not, the consequences border on the Orwellian. With an eye to the future, the Handbook of Neurotoxicology has been developed to provide researchers and students with a view of the current status of research in selected areas of neurotoxicology and to stimulate research in the field. Obviously, the field is enormous and all areas of interest could not be covered. However, if the Handbook of Neurotoxicology, volumes 1 and 2 prove useful, other volumes will be forthcoming. Therefore, we invite your comments and suggestions. Edward J. Massaro

CONTENTS Preface ............................................................................................................................ v Companion Table of Contents ...................................................................................... xi Contributors ................................................................................................................ xiii

I. PESTICIDES Marion Ehrich and Jeffrey R. Bloomquist, Section Editors

A. Anticholinesterase Insecticides 1 2 3

Acute Toxicities of Organophosphates and Carbamates ................. 3 Janice E. Chambers and Russell L. Carr Organophosphate-Induced Delayed Neuropathy ........................... 17 Marion Ehrich and Bernard S. Jortner Nonesterase Actions of Anticholinesterase Insecticides ............... 29 Carey Pope and Jing Liu

B. Pesticides that Target Ion Channels 4 5 6

Agents Affecting Sodium Channels ............................................... 47 David M. Soderlund Agents Affecting Chloride Channels ............................................. 65 Jeffrey R. Bloomquist The Neonicotinoid Insecticides ...................................................... 79 Larry P. Sheets

C. Miscellaneous Pesticides with Action on the Nervous System 7

Miscellaneous Pesticides with Action on the Nervous System ..... 91 Dennis Blodgett, Marion Ehrich, and Jeffrey R. Bloomquist

II. METALS Tomás R. Guilarte, Section Editor 8 9

10

11

Molecular Mechanisms of Low-Level Pb2+ Neurotoxicity ......... 107 Michelle K. Nihei and Tomás R. Guilarte Elucidation of the Zinc-Finger Motif as a Target for Heavy-Metal Perturbations .............................................................................. 135 Nasser H. Zawia and Morad Razmiafshari Blood-Brain Barrier and Blood-CSF Barrier in Metal-Induced Neurotoxicities .......................................................................... 161 Wei Zheng Manganese in Health and Disease: From Transport to Neurotoxicity .. 195 Michael Aschner, James R. Connor, David C. Dorman, Elise A. Malecki, and Kent E. Vrana vii

viii

Contents

12

Aluminum Neurotoxicity .............................................................. 211 Andrzej Szutowicz

III. NATURAL TOXINS OF MICROBIAL ORIGIN David J. Adams and Daniel G. Baden, Section Editors 13 14

15

16 17

18 19

Ecology of Microbial Neurotoxins ............................................... 239 Lyndon E. Llewellyn Biosynthesis of Important Marine Toxins of Microorganism Origins....................................................................................... 257 Yuzuru Shimizu Biological Assay and Detection Methods for Marine “Shellfish” Toxins................................................................... 269 Neale R. Towers and Ian Garthwaite An Overview of Clostridial Neurotoxins ..................................... 293 Mark A. Poli and Frank J. Lebeda Molecular Mechanism of Action of Botulinal Neurotoxins and the Synaptic Remodeling They Induce In Vivo at the Skeletal Neuromuscular Junction ................................... 305 Frédéric A. Meunier, Judit Herreros, Giampietro Schiavo, Bernard Poulain, and Jordi Molgó Marine Mammals as Sentinels of Environmental Biotoxins ....... 349 Vera L. Trainer The Epidemiology of Human Illnesses Associated with Harmful Algal Blooms ............................................................................ 363 Lora E. Fleming, Lorraine Backer, and Alan Rowan

IV. NATURAL TOXINS OF ANIMAL ORIGIN Alan Harvey, Section Editor 20

21

22 23 24 25

Snake Neurotoxins that Interact with Nicotinic Acetylcholine Receptors .................................................................................. 385 Denis Servent and André Ménez Presynaptic Phospholipase A2 Neurotoxins from Snake Venoms ..................................................................................... 427 John B. Harris Dendrotoxins from Mamba Snakes .............................................. 455 J. Oliver Dolly and Giacinto Bagetta Neurotoxins from Spider Venoms ................................................ 475 Alfonso Grasso and Stefano Rufini Neurotoxins from Scorpion Venoms ............................................ 503 Marie-France Martin-Eauclaire Anthozoan Neurotoxins ................................................................ 529 William R. Kem

Contents

ix

26 27

28

Nemertine Neurotoxins ................................................................. 573 William R. Kem Secretagogue Activity of Trachynilysin, a Neurotoxic Protein Isolated from Stonefish (Synanceia trachynis) Venom ........... 595 Frédéric A. Meunier, Gilles Ouanounou, Cesar Mattei, Pascal Chameau, Cesare Colasante, Yuri A. Ushkaryov, J. Oliver Dolly, Arnold S. Kreger, and Jordi Molgó Neurotoxins of Cone Snail Venoms ............................................. 617 Robert Newcomb and George Miljanich

Index ........................................................................................................................... 653

CONTENTS OF THE COMPANION VOLUME Handbook of Neurotoxicology Volume II

I. DEVELOPMENTAL NEUROTOXICOLOGY James L. Schardein, Section Editor 1 2 3

4

Interpretation of Developmental Neurotoxicity Data Judith W. Henck Manifestations of CNS Insult During Development Susan A. Rice Developmental Neurotoxicology: What Have We Learned from Guideline Studies? Gregg D. Cappon and Donald D. Stump Risk Assessment of Developmental Neurotoxicants Hugh A. Tilson

II. DRUGS OF ABUSE Patricia A. Broderick, Section Editor 5

6 7

8

9 10 11

Electrophysiologic Evidence of Neural Injury or Adaptation in Cocaine Dependence Kenneth R. Alper, Leslie S. Prichep, E. Roy John, Sharon C. Kowalik, and Mitchell S. Rosenthal Neurotoxicology of Marijuana and Cannabinoids Eliot L. Gardner Dopamine and Its Modulation of Drug-Induced Neuronal Damage Donald M. Kuhn NMDA Antagonist-Induced Neurotoxicity and Psychosis: The Dissociative Stimulation Hypothesis Kevin Kiyoshi Noguchi Emerging Drugs of Abuse: Use Patterns and Clinical Toxicity Katherine R. Bonson and Matthew Baggott Mechanisms of Methamphetamine-Induced Neurotoxicity Jean Lud Cadet and Christie Brannock Neurotoxic Effects of Substituted Amphetamines in Rats and Mice: Challenges to the Current Dogma James P. O’Callaghan and Diane B. Miller

xi

xii

Contents of Companion Volume

12

Studies of Neuronal Degeneration Indicate that Fasciculus Retroflexus is a Wink Link in Brain for Many Drugs of Abuse Gaylord Ellison

13

Microsensors Detect Neuroadaptation by Cocaine: Serotonin Released in Motor Basal Ganglia Is Not Rhythmic with Movement Patricia A. Broderick

III. IMAGING Thomas E. Schlaepfer, Section Editor 14

15 16

17 18

Impact of Intoxication: Structural and Functional Modifications in the Brain Induced by Ethanol Response David J. Lyons, Cory S. Freedland, and Linda J. Porrino Structural and Functional Neuroimaging of the Effects of Opioids David Nutt and Mark Daglish Structural and Functional Neuroimaging of the Effects of Cocaine in Human and Nonhuman Primates Linda J. Porrino, David Lyons, Sharon R. Letchworth, Cory S. Freedland, and Michael A. Nader Functional Neuroimaging of Cannabinoid Effects Godfrey D. Pearlson Neuroimaging of MDMA-Induced Neurotoxicity Una D. McCann, Zsolt Szabo, and George A. Ricaurte

IV. NEUROBEHAVIORAL ASSESSMENT METHODS Joel L. Mattson, Section Editor 19 20

21

Tier 1 Neurological Assessment in Regulated Animal Safety Studies Joseph F. Ross Neurological Assessment: The Role of the Clinician in Clinical Neurotoxicology James W. Albers Human Neuropsychological Testing and Evaluation Stanley Berent and Christine L. Trask

CONTRIBUTORS DAVID J. ADAMS • Department of Physiology and Pharmacology, School of Biomedical Sciences, University of Queensland, St. Lucia, Australia MICHAEL ASCHNER • Department of Physiology and Pharmacology, and Interdisciplinary Program in Neuroscience, Wake Forest University School of Medicine, Winston-Salem, NC LORRAINE BACKER • National Center for Environmental Health, Centers for Disease Control and Prevention (CDC), Atlanta, GA DANIEL G. BADEN • Center for Marine Science, University of North Carolina at Wilmington, Wilmington, NC GIACINTO BAGETTA • Department of Pharmacobiology, University of Calabria at Cosenza, Italy DENNIS BLODGETT • Department of Biomedical Sciences and Pathobiology, Virginia-Maryland Regional College of Veterinary Medicine, Virginia Tech, Blacksburg, VA JEFFREY R. BLOOMQUIST • Department of Entomology, Virginia Polytechnic Institute and State University, Blacksburg, VA RUSSELL L. CARR • College of Veterinary Medicine, Mississippi State University, Mississippi State, MS JANICE E. CHAMBERS • College of Veterinary Medicine, Mississippi State University, Mississippi State, MS PASCAL CHAMEAU • Institut Fédératif de Neurobiologie Alfred Fessard, Laboratoire de Neurobiologie Cellulaire et Moléculaire, Centre National de la Recherche Scientifique, Gif sur Yvette, France CESARE COLOSANTE • Institut Fédératif de Neurobiologie Alfred Fessard, Laboratoire de Neurobiologie Cellulaire et Moléculaire, Centre National de la Recherche Scientifique, Gif sur Yvette, France JAMES R. CONNOR • Department of Neuroscience and Anatomy, Penn State College of Medicine, Hershey, PA J. OLIVER DOLLY • Department of Biochemistry, Imperial College of Science, Technology and Medicine, London, UK DAVID C. DORMAN • CIIT Centers for Health Research, Research Triangle Park, Durham, NC MARION EHRICH • Laboratory for Neurotoxicity Studies, Virginia-Maryland Regional College of Veterinary Medicine, Blacksburg, VA LORA E. FLEMING • National Institute of Environmental Health Sciences (NIEHS), Marine and Freshwater Biomedical Sciences Center at University of Miami Rosenstiel School of Marine and Atmospheric Science, Miami, FL IAN GARTHWAITE • Toxinology and Food Safety Research, AgResearch, Ruakura Research Centre, Hamilton, New Zealand ALFONSO GRASSO • Institute of Cellular Biology, Consiglio Nazionale delle Ricerche, Rome, Italy TOMÁS R. GUILARTE • Department of Environmental Health Sciences, Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD xiii

xiv

Contributors

JOHN B. HARRIS • School of Neurosciences and Psychiatry, Faculty of Medicine, University of Newcastle upon Tyne, UK ALAN HARVEY • Strathclyde Institute for Drug Research, University of Strathclyde, Glasgow, UK JUDIT HERREROS • Molecular Neuropathobiology Laboratory, Imperial Cancer Research Fund, London, UK BERNARD S. JORTNER • Laboratory for Neurotoxicity Studies, Virginia-Maryland Regional College of Veterinary Medicine, Virginia Tech, Blacksburg, VA WILLIAM R. KEM • Department of Pharmacology and Therapeutics, University of Florida College of Medicine, Gainesville, FL ARNOLD S. KREGER • Department of Epidemiology and Preventive Medicine, and Department of Dermatology, University of Maryland School of Medicine, Baltimore, MD FRANK J. LEBEDA • Toxinology and Aerobiology Division, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, MD LYNDON E. LLEWELLYN • Australian Institute of Marine Science, Townsville, Australia JING LIU • Department of Physiological Sciences, College of Veterinary Medicine, Oklahoma State University, Stillwater, OK ELISE A. MALECKI • Department of Neuroscience and Anatomy, Penn State College of Medicine, Hershey, PA MARIE-FRANCE MARTIN-EAUCLAIRE • Biochemistry Laboratory, IFR Jean-Roche, Université de la Méditerranée, Faculté de Médecine Secteur Nord, Marseilles, France EDWARD J. MASSARO • The National Health and Environmental Effects Research Laboratory, Research Triangle Park, Durham, NC CESAR MATTEI • Institut Fédératif de Neurobiologie Alfred Fessard, Laboratoire de Neurobiologie Cellulaire et Moléculaire, Centre National de la Recherche Scientifique, Gif sur Yvette, France ANDRÉ MÉNEZ • Département d’Ingénierie et d’Etudes des Protéines, Commissariat á l’Energie Atomique (CEA), Saclay, France FRÉDÉRIC A. MEUNIER • Department of Biochemistry, Imperial College of Science, Technology, and Medicine; and Molecular Neuropathobiology Laboratory, Imperial Cancer Research Fund, London, UK GEORGE MILJANICH • Elan Pharmaceuticals Inc., Menlo Park, CA JORDI MOLGÓ • Institut Fédératif de Neurobiologie Alfred Fessard, Laboratoire de Neurobiologie Cellulaire et Moléculaire, Centre National de la Recherche Scientifique, Gif sur Yvette, France ROBERT NEWCOMB • Elan Pharmaceuticals Inc., Menlo Park, CA MICHELLE K. NIHEI • Department of Environmental Health Sciences, Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD GILLES OUANOUNOU • Institut Fédératif de Neurobiologie Alfred Fessard, Laboratoire de Neurobiologie Cellulaire et Moléculaire, Centre National de la Recherche Scientifique, Gif sur Yvette, France MARK A. POLI • Toxinology and Aerobiology Division, United States Army Medical Research Instiute of Infectious Diseases, Fort Detrick, MD

Contributors

xv

CAREY POPE • Department of Physiological Sciences, College of Veterinary Medicine, Oklahoma State University, Stillwater, OK BERNARD POULAIN • Neurotransmission et Sécrétion Neuroendocrine, Centre de Neurochimie, Centre National de la Recherche Scientifique, Strasbourg, France MORAD RAZMIAFSHARI • Department of Environmental Toxicology, Community and Environmental Medicine, University of California at Irvine, CA ALAN ROWAN • Division of Environmental Epidemiology, Florida Department of Health, Tallahassee, FL STEFANO RUFINI • Dipartimento di Biologia, Universitá di Roma Tor Vergata, Rome, Italy GIAMPIETRO SCHIAVO • Molecular Neuropathobiology Laboratory, Imperial Cancer Research Fund, London, UK DENIS SERVENT • Département d’Ingénierie et d’Etudes des Protéines, Commissariat á l’Energie Atomique (CEA), Saclay, France LARRY P. SHEETS • Toxicology Department, Bayer Corporation, Stilwell, KS YUZURU SHIMIZU • Department of Biomedical Sciences, College of Pharmacy, University of Rhode Island, Kingston, RI DAVID M. SODERLUND • Department of Entomology, New York State Agricultural Experiment Station, Cornell University, Geneva, NY ANDRZEJ SZUTOWICZ • Chair of Clinical Biochemistry, Department of Laboratory Medicine, Medical University of Gdansk, Gdansk, Poland NEALE R. TOWERS • Toxinology and Food Safety Research, AgResearch, Ruakura Research Centre, Hamilton, New Zealand VERA L. TRAINER • Marine Biotoxin Program, National Marine Fisheries Service, Northwest Fisheries Science Center, Environmental Conservation Division, Seattle, WA YURI A. USHKARYOV • Department of Biochemistry, Imperial College of Science, Technology and Medicine, London, UK KENT E. VRANA • Department of Physiology and Pharmacology, and Interdisciplinary Program in Neuroscience, Wake Forest University School of Medicine, Winston-Salem, NC NASSER H. ZAWIA • Department of Biomedical Sciences, University of Rhode Island, Kingston, RI WEI ZHENG • Division of Environmental Health Sciences, Columbia University School of Public Health; and Department of Pharmacology, College of Physicians and Surgeons, Columbia University, New York, NY

Organophosphates and Carbamates

1

I Pesticides A. Anticholinesterase Insecticides

2

Chambers and Carr


3

1 Acute Toxicities of Organophosphates and Carbamates Janice E. Chambers and Russell L. Carr

1. INTRODUCTION The insecticides of the organophosphorus and carbamate classes are widely used and highly effective pest control agents. Although there are agents within these two classes that have other pesticidal uses, such as fungicidal or herbicidal applications, it is the insecticides (which also have utility as nematocides, acaricides, and helminthicides) that display the greatest neurotoxic properties. Any agent designed to kill pests is of potential danger to nontarget organisms, such as humans, if the molecular target for the pesticide also exists as an important entity in the nontarget organism. Such a common molecular target exists for the organophosphorus and carbamate insecticides. As will be discussed in greater detail below, members of these two insecticidal classes are inhibitors of acetylcholinesterase (AChE). The inhibition of AChE mediates most, if not all, of the clinical signs of toxicity during an acute intoxication. Because of the environmental and metabolic lability of these two classes of agrochemicals, they were important replacements for the persistent and bioaccumulative organochlorine insecticides, which were the predominant agricultural chemicals in the 1950s and 1960s. The use of the organophosphorus insecticides (less accurately but more commonly called organophosphates: OPs) and the carbamates has been an important component in the control of insects in agriculture, buildings, home gardens, and public health since the 1950s. While attempts have been made by the agrochemical industry to improve the pest vs nontarget organism selectivity, and these attempts have frequently been very effective, it remains a fact that some of the agents with high-use patterns are still moderately or highly toxic to mammals (1,2). Because of their intense use, it is inevitable that human exposures will occur, and, despite important safety precautions being in place, some of these exposures are likely to be high level and lifethreatening during accidents. This chapter provides a summary of the history, chemistry, metabolism, and mechanism of action of the OPs and carbamates. Further information can be obtained from the following references: Chambers (3), Chambers and Levi (4), Ecobichon (5), Ecobichon and Joy (6), Eto (7), Fest and Schmidt (8), Hayes, (9,10), Heath (11), and Kuhr and Dorough (12). From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

3

4

Chambers and Carr

The OPs are esters of phosphoric acid, phosphonic acid, or phosphorothioic acid, and were derived from the chemical technology that generated the nerve gases of World War II. Organic phosphorus chemistry dates to about 1820 when triethyl phosphate was synthesized, with tetraethyl pyrophosphate (TEPP) synthesized in 1854. However, the insecticidal properties of TEPP were not discovered until the 1930s by Gerhard Schrader of Germany. Schrader was primarily interested in the insecticidal properties of the OP compounds he synthesized, even though he synthesized two notorious nerve gases, sarin and tabun. He was responsible for the creation of a variety of compounds, including parathion in 1944 (which is probably the most noteworthy), and, because of these accomplishments, he is known as the “father” of the OP insecticides. Along with several of the other early OPs, parathion was highly toxic to mammals including humans, because of the acute neurotoxicity described below. Therefore, the early OPs created some human safety concerns that were less relevant than with many of the organochlorine insecticides, such as DDT, which they were replacing (13). In 1950, a less toxic OP came into existence with the synthesis of malathion. Since then, a wide variety of OPs have been synthesized and used worldwide. The presence of less toxic OPs along with the development of serious problems of resistance of insects to the formerly effective organochlorine insecticides led to the rise of the OPs as the dominant class of insecticides, with 100 million pounds of OPs used worldwide around 1970. Additionally, the emerging environmental movement, which led to the banning of many of the organochlorine insecticides in the early 1970s in the United States, contributed to the prominent use of OPs. Compared to the OPs, the carbamate insecticides do not have such infamous relatives (such as the nerve gases) or as colorful a history. They emerged as potential insecticidal agents in a similar time frame as the OPs, with the synthesis in 1956 of carbaryl, the first successful carbamate insecticide. The carbamate insecticide structures were derived from that of the carbamate physostigmine (eserine), an alkaloid from calabar beans known to be an anticholinesterase. The majority of the carbamate insecticides are N-methyl or N,N-dimethyl derivatives of esters of carbamic acid. There is substantial chemical diversity among both classes of insecticides. As indicated earlier, the OPs are esters of phosphoric acid, phosphonic acid, or phosphorothioic acid. The central phosphorus atom is pentavalent and tetracoordinate, with the coordinate covalent bond (typically represented as a double bond) to either an oxygen or a sulfur. Of the other three groups bonded to the phosphorus, the one which has the least stable bond is termed the leaving group, and it is the one that is released from the remainder of the molecule as it reacts with molecular targets, as described further below. Representative examples of some OP insecticides are given in Fig. 1. It should be noted that there is great variety in the nature of the leaving group, shown on the right side of each molecule in Fig. 1, with aliphatic, aromatic, and heterocyclic moieties of various degrees of complexity comprising the leaving groups. There is considerably less diversity within the other two R groups, with methyl or ethyl groups occurring in many of the insecticides. It should also be noted that those molecules that possess a sulfur bonded through the coordinate covalent bond (such as parathion, azinphosmethyl, malathion, and chlorpyrifos, which are phosphorothionates) are not potent cholinesterase inhibitors, and require metabolic activation by cytochromes P450 to their oxon metabolites. In this reaction, the sulfur attached to the central phosphorus


5

Fig. 1. Representative chemical structures of some organophosphorus insecticides.

atom is replaced by an oxygen in a reaction termed desulfuration. The oxons are appreciably more potent as anticholinesterases and are the forms that induce toxicity through inhibition of AChE. A representative desulfuration reaction, explained more fully below, is illustrated in Fig. 2. Those insecticides, such as dichlorvos and tetrachlorvinphos, which possess an oxygen bonded through the coordinate covalent bond, are active anticholinesterases and do not require metabolic activation to exert toxicity. In contrast, the carbamates, which also demonstrate substantial chemical diversity in their leaving groups, do not require metabolic activation. Some representative carbamate insecticides are illustrated in Fig. 3. Metabolism cannot be ignored when the neurotoxicity of these anticholinesterases is considered. Part of the reason that the OPs and carbamates largely replaced the persistent, bioaccumulative organochlorine insecticides is the fact that both the OPs and carbamates are less chemically stable and far more amenable to metabolism and environmental breakdown than are the highly stable organochlorine compounds. Of course, their chemical reactivity contributes substantially to their ability to react with biological molecules and cause toxic reactions. As described earlier, a number of the OP insecticides require P450-mediated activation through the desulfuration reaction to form the oxon metabolites. The desulfuration reaction is postulated to result following the formation of an unstable phosphooxythiiran intermediate (14), as illustrated in the metabolism of chlorpyrifos in Fig. 2. This hypothetical intermediate can yield two products: (1) the oxon through loss of the sulfur with replacement by oxygen in the

6

Chambers and Carr

Fig. 2. Cytochrome P450-mediated metabolism of the organophosphorus insecticide chlorpyrifos. Potential reactions include both bioactivation (left) and degradation (right).

desulfuration reaction discussed earlier, which is an activation reaction; and (2) the leaving group plus dialkyl phosphorothioic acid or dialkyl phosphoric acid, which is a detoxication reaction. Therefore, only a fraction of the parent compound will be bioactivated. There are several other detoxication options for the parent compound or the oxon. Two of the most important of these detoxication reactions are: (1) the phosphorylation by the oxon of other serine esterases (such as nontarget acetylcholinesterase, butyrylcholinesterase, or carboxylesterases), which stoichiometrically destroys the oxon molecules; and (2) the catalytic hydrolysis of the oxon or phosphates by calcium-dependent A-esterases (15). The efficiency of these detoxication reactions will determine in large measure how much of the active anticholinesterase molecules can be eliminated from the system before they have an opportunity to inhibit AChE in target tissues. Therefore, the compounds with more efficient detoxication pathways are likely to be the less toxic members of the class. A more extensive discussion of OP


7

Fig. 3. Representative chemical structures of some carbamate insecticides.

insecticide metabolism can be found in Chambers and Chambers (16), Ecobichon (5,13), and Kulkarni and Hodgson (17). Carbamates are also metabolically labile, with detoxication reactions occurring primarily through esterase-mediated hydrolysis or P450-mediated oxidation. There are no bioactivations required for toxicity, as is the case with many of the OP insecticides, because the parent carbamate molecules are active anticholinesterases. Carbamate metabolism is described in greater detail in Ecobichon (5,18) and Kuhr and Dorough (12). 2. CLINICAL EVIDENCE OF NEUROTOXICITY Both the OPs and the carbamates display a broad range of mammalian acute toxicity levels. Because of the high toxicity levels of some members of both classes, there have been some serious incidents of human poisoning, not only among the agriculturalworker community, but also among members of the general public. Although those persons involved with anticholinesterase agents occupationally, both in factories and in agricultural settings, have been the groups most frequently poisoned, accidental poisonings have occurred in which insecticides improperly stored have been ingested, typically by children, and insecticides contaminating food, clothing, or materials used as toys have caused serious, and sometimes lethal, poisonings (19). The high acute toxicity level of some of the OP insecticides has made accidental poisonings likely with these compounds. Routine exposures resulting from proper use of these anticholinesterase insecticides according to the regulations would not be expected to yield any clinical signs of toxicity.

8

Chambers and Carr

Because of the wide range of acute toxicity levels, the amount of an OP insecticide that would be fatal to a human being varies widely. A few representative examples of known or estimated human fatal doses of OP insecticides in grams are: coumaphos, 10; diazinon, 25; azinphosmethyl, 0.2; malathion, 60; methyl parathion, 0.15; and parathion, 0.15–0.30 (19). For the carbamate carbaryl, 0.25 g causes moderately severe symptoms of poisoning (5). Because of the extensive involvement of acetylcholine (ACh) in nervous-system function, the signs of anticholinesterase-induced hypercholinergic activity are diverse. Some of the signs of poisoning include muscle weakness, fasciculations, tremors, convulsions, and paralysis of the limbs, the back, and lastly the diaphragm. Autonomic signs also occur, including the appearance of the clinical syndrome called SLUD (abbreviation for salivation, lacrimation, urination, and defecation) as well as other autonomic effects such as on cardiovascular function or pupil diameter. The cause of death is usually respiratory failure, probably of both central nervous system (CNS) and peripheral nervous system (PNS) origin. Further descriptions of the clinical signs of toxicity can be found in Ecobichon (5), Marrs and Dewhurst (20), and Thienes and Haley (19). Because of the wide range in toxicity levels of the carbamate insecticides, they are also subjects of accidental poisonings. Generally carbamates yield less severe and more transient signs of toxicity than the OPs, but severe, and even lethal, carbamate intoxications have occurred, even with the moderately toxic carbamates (10,18). One of the most toxic of all insecticides is the carbamate aldicarb. As is generally true of the carbamates, aldicarb is reasonably water-soluble, and when it was inappropriately used on vegetable crops, it became incorporated into edible portions of crops that possess a high water content, such as watermelons and cucumbers (21). A number of people were hospitalized after consuming these contaminated crops because of the signs of anticholinesterase poisoning. 3. MECHANISMS OF ACTION AND ANTIDOTES As indicated earlier, the significant mechanism of action of the OPs and carbamates in acute intoxications is the inhibition of the widespread nervous system enzyme, AChE. AChE is the enzyme responsible for the rapid hydrolysis and therefore inactivation of the ubiquitous neurotransmitter ACh. ACh binds to two sites on AChE, with the quaternary nitrogen of choline attracted to the anionic site and the carbonyl of the ester binding at the serine residue in the esteratic site with subsequent loss of the choline during the hydrolytic reaction; subsequently, the acetate is rapidly hydrolyzed from the serine residue, with restoration of the AChE active site for a subsequent hydrolysis. A schematic of this reaction is illustrated in Fig. 4. A more detailed description of AChE structure and function can be found in Taylor and Radic (22). ACh mediates neurotransmission within both the CNS and PNS. It mediates transmission in a variety of regions of the brain, and in the PNS is also responsible for transmission to the skeletal muscles in the somatic nervous system and to cardiac and smooth muscles and other effectors in the autonomic nervous system. Specifically in the autonomic nervous system, ACh is the neurotransmitter at both the preganglionic and postganglionic junctions of the parasympathetic division and at the preganglionic


9

Fig. 4. Reactions involved in the hydrolysis of acetylcholine (ACh) by acetylcholinesterase (AChE). The quaternary nitrogen of the choline portion of ACh associates with the anionic site on glutamate 334 followed by the binding of the carbonyl of the ester portion of ACh binding to the hydroxyl group on serine 203. The choline portion of ACh is removed in a hydrolysis reaction leaving the acetate bound to the serine hydroxyl. The acetate is rapidly hydrolyzed freeing the enzyme’s active site.

10

Chambers and Carr

Fig. 5. Reactions involved in the inhibition of acetylcholinesterase (AChE) by an organophosphate compound (OP). The leaving group portion of the OP (indicated by R3) associates with the anionic site on glutamate 334 followed by the binding of the phosphate portion of the OP to the hydroxyl group on serine 203. The phosphorylation of AChE is persistent (days to weeks). The phosphorylated AChE can then undergo two possible reactions: (1) aging, in which one of the remaining R groups is removed rendering the AChE refractory to recovery; (2) spontaneous reactivation, in which the phosphate moiety is removed by hydrolysis rendering the AChE functional.


11

synapses and a few postganglionic junctions of the sympathetic division. Inhibition of AChE by these anticholinesterase insecticides (or their active metabolites in the case of some of the OPs) prevents the rapid inactivation of ACh, and the resultant accumulation of ACh results in hyperactivity within cholinergic pathways. Because ACh is such a widely distributed neurotransmitter throughout the nervous system, hypercholinergic activity results in an impact on numerous effectors, including voluntary muscles and various visceral organs and tissues, leading to the symptomatology noted earlier. In some cases, such as the heart, the stimulation of both the sympathetic and parasympathetic nervous systems results in opposing effects, as both tachycardia and bradycardia are observed. Impact on the respiratory system is ultimately the life-threatening phenomenon in a lethal level acute intoxication, with respiratory failure resulting from paralysis of the respiratory muscles, bronchoconstriction, stimulation of the secretion of bronchiolar mucus, and perturbation of the respiratory control center in the brain. This inhibition of AChE results from covalent-bond formation between the insecticide or metabolite and the serine hydroxyl group in the catalytic site of the AChE. In the case of the OPs, the OP molecule phosphorylates the serine residue with simultaneous loss of the leaving group; thus the OP molecule cannot be recovered intact following the phosphorylation reaction. The phosphorylated AChE is a relatively stable entity, persisting for hours to days before the phosphate moiety is spontaneously hydrolyzed that restores AChE activity. The nature of the phosphate determines the degree of persistence, with the diethyl phosphates being considerably more persistent (half-life of spontaneous reactivation greater than 2 d) than the dimethyl phosphates (half-life of spontaneous reactivation about 2 h). An additional reaction that can occur to the phosphorylated AChE is a nonenzymatically mediated loss of one of the alkyl groups, leaving the resultant phosphorylated AChE charged at physiological pH. This phenomenon has been termed “aging.” The phosphorylated and aged AChE is refractory to spontaneous reactivation, and therefore aging contributes even further to the persistence of the AChE inhibition, with recovery of AChE activity following aging only possible through de novo synthesis of AChE. These reactions are schematically represented in Fig. 5. Further discussion of OP inhibition of AChE can be found in Chambers (3), DuBois (23), DuBois et al. (24), Ecobichon (13), and Eto (7). This persistent phosphorylation perturbs nervous-system activity for a period of hours to days, and, if the exposure is sublethal, can elicit compensatory biochemical/ physiological mechanisms to counteract some of the hypercholinergic activity. One well-documented type of homeostatic compensation is the downregulation of central and peripheral muscarinic cholinergic receptors, which involves the initial internalization of cholinergic receptors with subsequent destruction of the receptors. The result of the presence of fewer cholinergic receptors is to attenuate the action of the excess ACh and reduce the level of toxic signs displayed (e.g., tremors) upon subsequent exposures to similar levels of the OP, a phenomenon termed tolerance (25). The biochemical compensation and the behavioral tolerance dissipate as the levels of AChE activity recover. Although there are many similarities between inhibition of AChE by carbamate and OP insecticides and the subsequent neurotoxicity, there are some very distinct differences as well. The carbamylation of the serine hydroxyl group of the AChE active site also results in the formation of a covalent bond, with resultant loss of the leaving group;

12

Chambers and Carr

Fig. 6. Reactions involved in the inhibition of acetylcholinesterase (AChE) by a carbamate. The carbamate forms a covalent bond with the hydroxyl group on serine 203. The carbamylation of AChE is transient (hours to days). The carbamate moiety is removed by hydrolysis rendering the AChE functional.

thus, with the carbamates, similar to the OPs, no recovery of the intact insecticide is possible following its reaction with AChE. A schematic representation of these reactions is given in Fig. 6. Further discussion can be found in Ecobichon (5,18) and Kuhr and Dorough (12). However, the carbamylated AChE is much less stable than the phosphorylated AChE, and spontaneous reactivation is much faster, in the range of minutes to hours. The inhibition of AChE by carbamates, however, results in the same impact on cholinergic pathways as does that resulting from OPs, and the signs of poisoning are similar. Therefore, impact on the cholinergic system, while possibly very severe and life-threatening, is relatively transient and is unlikely to be persistent enough to elicit appreciable homeostatic compensation or tolerance, as is observed with the OPs, unless the exposures to the carbamate were extremely frequent. As indicated earlier, the signs of toxicity are those of hyperexcitability within cholinergic pathways within the CNS and PNS, with the primary life-threatening effects


13

being those within the respiratory system. It should be noted that the toxicity of these anticholinesterases results not from the inhibition of AChE per se, but from the aberrant physiological events resulting from the accumulation of ACh in synapses and neuromuscular/effector junctions leading to hyperstimulation of cholinergic receptors. Therefore, therapy of anticholinesterase poisoning is predicated upon the strategy of counteracting the excessive stimulation of cholinergic receptors through use of cholinergic receptor antagonists, such as the belladonna alkaloid atropine. Atropine blocks the muscarinic receptors and attenuates the life-threatening actions of the accumulated ACh within critical synapses and neuromuscular/effector junctions. Thus, death from respiratory failure can be averted if atropine administration is initiated early enough and continued until sufficient AChE activity recovery or compensatory reactions take over. Use of artificial respiration may also be required. Atropine is effective on muscarinic receptors only and not at nicotinic receptors (such as those activating skeletal muscle), so not all signs of poisoning are eliminated during atropine therapy. Atropine is an effective antidote for both OP and carbamate poisoning. Diazepam can be effective in alleviating convulsions and fasciculations. A more extensive description of therapy can be found in Marrs and Dewhurst (20). An additional form of therapy effective with OPs only and not carbamates is the use of oxime reactivators. Oximes, such as the therapeutic agent 2-PAM (pralidoxime), are capable of undergoing a transphosphorylation reaction with the phosphorylated OP, with subsequent transfer of the phosphate moiety to the oxime; thus, reactivation of AChE is hastened. However, this reaction is only possible with nonaged phosphorylated AChE, so 2-PAM therapy can only be effective if initiated while a substantial fraction of the inhibited AChE is still not aged. Because 2-PAM contains a quaternary nitrogen, it is incapable of crossing the blood-brain barrier (BBB), and therefore 2PAM therapy is only capable of increasing the reactivation rate of phosphorylated AChE in the PNS. Additionally, 2-PAM does not counteract the signs of poisoning so it would need to be used in conjunction with atropine therapy. As noted earlier, oxime therapy is not effective with carbamylated AChE, and may actually worsen the effects from stabilization of the carbamylated enzyme. However, it should be remembered that carbamylated AChE has a relatively fast rate of recovery compared to phosphorylated AChE, and therefore there is less need of a therapeutic agent that accelerates the reactivation. 4. FUTURE DIRECTIONS There are over 35 OP insecticides registered for use in the United States at the time this chapter was written. There are fewer registered carbamate insecticides, less than a dozen. Both classes demonstrate a wide diversity of structures and a wide range of acute toxicity levels. However, they are all anticholinesterases and display similarities in signs of acute toxicity. Therefore, there has been a tendency in the past to assume that all OPs, at least, act similarly, displaying the same toxic endpoints and displaying a relatively similar toxicity profile. However, because of their diverse structures, they differ substantially in their lipophilicity and rates of metabolism, and therefore in their disposition. These differences in metabolism and disposition will likely create differences among compounds in toxicity levels, rate of onset of signs of toxicity, the range of persistence of AChE inhibition, the degree of homeostatic compensation elicited,

14

Chambers and Carr

and therefore in the persistence of the toxic responses. Because of these differences in persistence, it would be expected that both short-term and long-term ramifications of exposures to different members of the class would differ among the compounds. Although there has been a tendency to assume in the past that the effects of anticholinesterase poisoning were over when the AChE inhibition was gone, there are reports of long-term derangements of autonomic, neuromuscular, and psychiatric function following exposures (usually high level) to anticholinesterase insecticides (5,13). However, some of these reports are anecdotal and not well-described, so the existence of long-term effects that can be attributed specifically to anticholinesterase exposure remains controversial. The likelihood of long-term effects from acute exposures is an area that should still be investigated more critically. Another area that is still controversial is the relevance of alternate targets in OP toxicity. It is documented that some OP compounds can interact directly with ACh receptors (26). Other serine esterases and serine proteases can be phosphorylated by OP compounds. Some of these potential alternate targets may have a greater affinity for the OP compound than does AChE. Whether or not these alternate targets contribute to the overall toxic response, and, if they do, how greatly they contribute, are unknown. Such information on the relevance of alternate targets would be useful in more fully understanding the etiology of the observed toxicity and in predicting how extensive or persistent toxic responses might be. (For further discussion, see Chapter 3 by Pope and Liu.) As alluded to earlier, there is great chemical diversity among the two classes of anticholinesterase insecticides. This chemical diversity creates diversity in targetenzyme sensitivity, and even greater diversity in the likely pathways and efficiency of metabolism. A better understanding of the structure-activity relationships for target sensitivity and metabolism would be beneficial in predicting toxic responses. Another area lacking a sufficient data base at the present time is that of the quantitative and qualitative responses of sensitive subpopulations to anticholinesterase exposures, with infants and children probably being the subpopulations of greatest scientific and public concern. It is not clear to what extent the developing nervous system is impacted by exposure to these agents, whether any such impact is permanent and leads to functional or cognitive deficits, and through what mechanisms these agents might exert developmental neurotoxicity. Such information will be important in allowing us to better protect more vulnerable populations from adverse effects. Lastly, another area that lacks data is the impact of low-dose exposures to these compounds. Much of the data base exists on high-level exposures to laboratory animals that were of a sufficient level to induce substantial inhibition of AChE, possibly to life-threatening levels. There is some epidemiological evidence on people, primarily occupationally exposed people, and some anecdotal information from human accidental exposures, but the data base from humans is necessarily limited. If these insecticides are used according to the regulations governing their legal uses, human exposures should be small. Information about the nature and levels of any toxicity that might be induced by low exposure levels of these compounds will be useful to determine whether routine use of these compounds is likely to induce any adverse effects on human populations.


15

In summary, the OP and carbamate insecticides are widely used compounds that have a long history of applications and have had, and currently have, great utility in enhancing agricultural productivity and preserving public health. They are of concern because of the serious effects they can have on cholinergic neurophysiology. Some of the members of both classes are highly toxic and can be a threat to human life and health. Nevertheless, their extensive use over the last 40–50 yr has had major positive impacts on human productivity and health and suggests that these compounds still are of considerable benefit to society and warrant continued research to more fully understand the nature and extent of any adverse effects that they may elicit. 5. REFERENCES 1. Montgomery, J. H. (1993) Agrochemicals Desk Reference Environmental Data. Lewis Publishers, Chelsea, MI. 2. Worthing, C. R. and Walker, S. B. (eds.) (1987) The Pesticide Manual. 8th ed. British Crop Protection Council, UK. The Lavenham Press Limited, Lavenham, Suffolk, UK. 3. Chambers, H. W. (1992) Organophosphorus compounds: an overview, in Organophosphates: Chemistry, Fate, and Effects (Chambers, J. E. and Levi, P.E., eds.) Academic Press, San Diego, CA, pp. 3–17. 4. Chambers, J. E., Levi, P. E. (eds.) (1992) Organophosphates: Chemistry, Fate, and Effects. Academic Press, San Diego, CA. 5. Ecobichon, D. J. (1996) Toxic effects of pesticides, in Casarett & Doull’s Toxicology: The Basic Science of Poisons, 5th ed. (Klaassen, C. D., ed.), McGraw-Hill, NY, pp. 643–689. 6. Ecobichon, D. J. and Joy, R. M. (1994) Pesticides and Neurological Diseases, 2nd ed. CRC, Boca Raton, FL. 7. Eto, M. (1974) Organophosphorus Pesticides: Organic and Biological Chemistry. CRC, Cleveland, OH. 8. Fest, C. and Schmidt, K.-J. (1973) The Chemistry of Organophosphorus Pesticides. Springer-Verlag, NY. 9. Hayes, W. J., Jr. (1975) Toxicology of Pesticides. Waverly Press, Baltimore, Md. 10. Hayes, W. J., Jr. (1982) Pesticides Studied in Man. Williams & Wilkins, Baltimore, MD. 11. Heath, D. F. (1961) Organophosphorus Poisons. Anticholinesterases and Related Compounds. Pergamon, London, UK. 12. Kuhr, R. J. and Dorough, H. W. (1976) Carbamate Insecticides: Chemistry, Biochemistry and Toxicology. CRC, Boca Raton, FL. 13. Ecobichon, D. J. (1994a) Organophosphorus ester insecticides, in Pesticides and Neurological Diseases, 2nd ed. (Ecobichon, D. J. and Joy, R. M., eds.), CRC, Boca Raton, FL, pp. 171–249. 14. Neal, R. A. (1980) Microsomal metabolism of thiono-sulfur compounds: mechanisms and toxicological significance, in Reviews in Biochemical Toxicology 2 (Hodgson, E., Bend, J.R., and Philpot, R. M., eds.), Elsevier North Holland, New York, NY, pp. 131–171. 15. Chambers, J. E., Ma, T., Boone, J. S., and Chambers, H. W. (1994) Role of detoxication pathways in acute toxicity levels of phosphorothionate insecticides in the rat. Life Sci. 54, 1357–1364. 16. Chambers, J. E. and Chambers, H. W. (1991) Biotransformation of organophosphorus insecticides in mammals, in Pesticide Transformation Products Fate and Significance in the Environment (Somasundaram, L. and Coats, J. R., eds.), ACS Symposium Series, American Chemical Society, Washington, DC, pp. 32–42. 17. Kulkarni, A. P. and Hodgson, E. (1984) The metabolism of insecticides: the role of monooxygenase enzymes. Annu. Rev. Pharmacol. 24, 19–42.

16

Chambers and Carr

18. Ecobichon, D. J. (1994) Carbamic acid ester insecticides, in Pesticides and Neurological Diseases, 2nd ed. (Ecobichon, D. J. and Joy, R. M., eds.), CRC, Boca Raton, FL, pp. 251–289. 19. Thienes, C. H. and Haley, T. J. (1972) Clinical Toxicology, 5th ed. Lea & Febiger, Philadelphia, PA. 20. Marrs, T. C. and Dewhurst, I. (1999) Toxicology of Pesticides, in General and Applied Toxicology, 2nd ed. (Ballantyne, B., Marrs, T., and Syversen, T., eds.), Grove’s Dictionaries, New York, NY, pp. 1993–2012. 21. Goldman, L. R., Beller, M., and Jackson, R. J. (1990) Aldicarb food poisonings in California, 1985–1988. Toxicity estimates for humans. Arch. Environ. Health 45, 141–148. 22. Taylor, P. and Radic, Z. (1994) The cholinesterases: from genes to proteins, in Annual Review of Pharmacology and Toxicology, vol. 34 (Cho, A. K. Blaschke, T. F., Loh, H. H., Way, J. L. eds.), Annual Reviews, Palo Alto, CA, pp. 281–320. 23. DuBois, K. P. (1948) New rodenticidal compounds. J. Am. Pharm. Assoc. 37, 307–310. 24. DuBois, K. P., Doull, J., Salerno, P. R., and Coon, J. M. (1949). Studies on the toxicity and mechanisms of action of p-nitrophenyl-diethyl-thionophosphate (Parathion). J. Pharmacol. Exp. Ther. 95, 75–91. 25. Hoskins, B. and Ho, I. K. (1992) Tolerance to organophosphorus cholinesterase inhibitors, in Organophosphates: Chemistry, Fate, and Effects (Chambers, J. E., Levi, P. E., eds.), Academic Press, San Diego, CA, pp. 285–297. 26. Eldefrawi, A. T., Jett, D., and Eldefrawi, M. E. (1992) Direct actions of organophosphorus anticholinesterases on muscarinic receptors, in Organophosphates: Chemistry, Fate, and Effects (Chambers, J.E. and Levi, P.E., eds.), Academic Press, San Diego, CA, pp. 258–270.

OPIDN

17

2 Organophosphate-Induced Delayed Neuropathy Marion Ehrich and Bernard S. Jortner

1. INTRODUCTION/HISTORY Although the cholinesterase-inhibiting effects of organophosphate (OP) compounds were not utilized until the time of World War II, the ability of some of these chemicals to cause an irreversible, progressive delayed neuropathy was recognized as early as the 1890s, when a 15% solution of tri-ortho-cresyl phosphate (TOCP, or tri-ortho-tolyl phosphate, TOTP) was used to treat tuberculosis. A number of incidents of organophosphate-induced delayed neuropathy (OPIDN) have been reported since then, with TOTP identified as the neurotoxic contaminant of cresyl phosphates associated with these poisonings. Large numbers of humans were affected in some of these incidents, including over 50,000 Americans who ingested a TOTP-contaminated alcoholic extract of ginger during the era of prohibition (1930s), 10,000 Moroccans who ingested TOTP-contaminated cooking oil in the 1950s, and 600 Indians who consumed TOTPcontaminated rapeseed oil in 1988. Early studies determined that not every OP compound was capable of causing OPIDN and that all animal species were not uniformly susceptible. Clinical evidence of progressive, irreversible OPIDN has been observed in humans, water buffalo, sheep, cats, ferrets, chickens, and a number of other species; laboratory rodents (e.g., rats, mice), however, do not demonstrate progressive locomotor effects after exposure (1–7). TOTP has been associated with the largest number of cases of OPIDN observed in people, but OP compounds initially but no longer manufactured as pesticides (e.g., leptophos, mipafox, O-ethyl-O-p-nitrophenyl phenylphosphonothioate [EPN]) were also responsible for a number of accidental poisonings of humans and animals. To reduce risk for OPIDN in humans and susceptible animals of economic importance, US Environmental Protection Agency (EPA)-required testing procedures are now used and OP compounds capable of inducing OPIDN in the absence of significant (i.e., lethal) cholinesterase inhibition are not currently marketed for use as insecticides in the United States (8). Some compounds that cause OPIDN elict this syndrome at dosages that also cause acute toxicity as a result of cholinesterase inhibition (e.g., EPN, diisopropyl phosphorofluoridate [DFP]). A number of other delayed neuropathy-inducing OP comFrom: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

17

18

Ehrich and Jortner

Table 1 Examples of Neuropathy-Inducing OP Compounds Name or abbreviation

Chemical name

TOCP TOTP

Tri-ortho-cresyl phosphate Tri-ortho-tolyl phosphate

DFP

Diisopropyl phosphorofluoridate; difluorophosphate N,N’-diisopropyl phosphorodiamidofluoridate O-4-bromo-2,5-dichlorophenyl O-methyl phenyl phosphorothioate O-ethyl O-p-nitrophenyl phenylphosphonothioate O,S-dimethyl phosphorothioamidate

Mipafox Leptophos

EPN Methamidophos

Current or former use Lubricant, fuel additive, manufacture of plastics Nerve gas Insecticide Insecticide

Insecticide Insecticide

pounds are not notable inhibitors of cholinesterase (e.g., TOTP), and are used as lubricants, fuel additives, and in the manufacture of plastics. Today such products are formulated to decrease the neuropathy-inducing component (4,9). Tables of neuropathy-inducing OP compounds and their chemical structures are included in several previous reviews (1,2,5,10). An abbreviated list is provided in Table 1. Although no specific structural features that positively identify a neuropathy-inducing OP compound have been identified, structure activity studies have noted that the phosphorus must be in a pentavalent state, the atom attached with the coordinate covalent bond to the phosphorus must be an oxygen, at least one oxygen must bridge an R group to the phosphorus, and increased hydrophobicity can increase neuropathy-inducing capability among a series of neuropathy-inducing OP analogs (11). Another type of delayed neuropathy that may follow administration of OP compounds (termed type II delayed neurotoxicity) has been described. Compounds inducing this syndrome have a trivalent phosphorus atom. Differences between type II delayed neurotoxicity and classical OPIDN can be noted in time to onset and manifestations of clinical signs and in location and spectrum of neuropathological lesions. Further information on type II delayed neurotoxicity induced by trivalent OP compounds can be found in other reviews (2,11,12). 2. CLINICAL AND MORPHOLOGICAL EVIDENCE OF OPIDN IN MAN AND ANIMALS OPIDN can occur in humans and in a number of animal species following single or multiple exposures. The symptoms, which do not appear for days to weeks after exposure, are progressive and irreversible, although some improvement has been reported over time (2,13). Clinical features of OPIDN in man have been described. OPIDN begins with sensory loss in hands and feet, but exclusively sensory neuropathy is not a

OPIDN

19

Fig. 1. Development of clinical signs in chickens after administration of phenyl saligenin phosphate (PSP) and tri-ortho-tolyl phosphate (TOTP). Results are presented as mean ± SD, n = 3 –9. PSP im 2 mg/kg (❍ – ❍), 3 mg/kg (■ – ■), 10 mg/kg (∆ – ∆); TOTP 360 mg/kg po (• – •), 500 mg/kg po (■ – ■). Increasing clinical scores reflect progression of deficits. Reprinted with permission from ref. 15, ©Intox Press.

feature in humans. Motor alterations such as slowed conduction and, eventually, bilateral and symmetrical weakness progressing to flaccidity of the distal skeletal muscles of the lower and upper extremities occur. The patient notices tingling then loss of feeling in hands and feet, locomotor difficulties, and abnormal reflexes (4,10,11,14). In animals as well as man, there is a latent period between exposure and manifestations of OPIDN. The domestic chicken (hen) is the recognized animal model for OPIDN (8). Effects on the legs are noted, and the hen exhibits progressive incoordination and difficulty in walking (Fig. 1; 15). Eventually ability to walk is lost and the wings, too, become involved. There is an age-related susceptibility, in that these effects are not seen in chickens less than 55 d of age. Progressive ataxia is also seen in adults of other susceptible species (e.g., cats, sheep, water buffalo, horses, ferrets). Ataxia has not been a prominent feature of OPIDN in rodents (1,3,11,16–18). The neuropathologic changes in classical (type I) OPIDN are typified by those elicited in experimental animals such as the chicken, cat and ferret dosed with compounds such as TOTP, DFP, or phenyl saligenin phosphate (PSP). These relate well to the observed clinical deficits and consist of degeneration of distal regions of large, long myelinated axons as the primary lesion, which progresses to Wallerian-like degeneration of affected fiber regions (19). The primary lesion is thought to reside in a distal nonterminal axonal region, with subsequent somartofugal extension of the alterations to the terminal axons and their endings (20–23). Lesions generally become apparent at

20

Ehrich and Jortner

Fig. 2. Organophosphorus-induced delayed neuropathy. (A) Low-power view of darkly stained bilateral degeneration of fasciculus gracilis (straight arrows) and spinocerebellar tracts (curved arrows) in a transversely stained cervical spinal cord. (B) Higher-power longitudinal section showing swollen axons with attenuated myelin sheaths (arrow), dark staining masses of myelin-rich debris (arrowhead) and replacement of degenerated fibers by pale-stained regions of astrocytic proliferation.* Both sections from a hen dosed with 1 mg/kg DFP 21 d earlier, toluidine blue and safranin stain.

or close to the end of the symptom-free postdosing period, and increase in severity and proximal extent associated with progressing clinical deficits. Regions of pathologic involvement include bilateral distal regions of long peripheral nerves and of brain or spinal cord long tracts such as fasciculus gracilis, and spinocerebellar, spinolivary, rubrospinal reticulospinal, and medial pontine spinal (hens only) tracts (Fig. 2) (19,23–30). Neuronal-cell bodies are spared (19,24).

OPIDN

21

Fig. 3. Photographs are of tibial nerve branch to gastrocnemius muscle of hens dosed with a single neurotoxic doses of phenyl saligenin phosphate (A–C) or the dorsal metatarsal nerve of hen dosed with diisopropyl phosphofluoridate (D). (A) Tangential section of swollen degenerating axon with thin (attenuated) myelin sheath (arrow). Day 9. Toluidine blue and safranin stain (also used in C). (B) Intact myelinated fiber (arrow points to node of Ranvier) and an adjacent fiber in Wallerian-like degeneration (arrowhead). Day 14, teased fiber preparation, osmium tetroxide stain. (C) Tangential section showing various stages of myelinated fiber degeneration (arrows). Day 15. (D) Cross-section showing axonopathic changes (arrows) to advanced Wallerian-like degeneration (arrowhead). Increased endoneural space suggests edema. Reproduced from Toxicologic Pathology with the permission of The Society of Toxicologic Pathologists (31).

The morphologic features of the nerve fiber lesions include swelling of affected axons (generally the long, large fibers) leading to attenuation of their myelin sheaths (Figs. 2 and 3; 31). The affected axons may demonstrate proliferation of tubules and cisterns, vacuoles (also affecting inner myelin sheaths), disorganized masses of abnormal mitochondria, cytoskeletal elements, dense bodies, and membranous multilamellar bodies (20–23,32). This appears to progress to granular degeneration of axonal contents, yielding swollen, electron-lucent axoplasm (Fig. 3). Another axonal alteration,

22

Ehrich and Jortner

Fig. 4. NTE inhibition in SH-SY5Y cells exposed to OP compounds. Human neuroblastoma cells (1 × 107 cells/mL in saline) were exposed for 1 h to concentrations of 10–5 M. DFP, mipafox, TSP, triphenyl phosphite, and PSP are active NTE inhibitors and agents that induce OPIDN in animal models. TOTP and leptophos are protoxicants that require metabolic activation before NTE can be inhibited. The protoxicant parathion and its active oxon congener, paraoxon, as well as malathion and malaoxon, do not induce OPIDN in the hen model. Additional protoxicants are chlorpyrifos, fenthion, and fenitrothion. Dichlorvos is an active inhibitor of acetylcholinesterase, but is not likely to induce OPIDN in the hen model. Results are expressed as mean ± SEM of results from 3–7 different days on which assays were done. Reprinted with permission from ref. 35; ©Taylor & Francis.

common in affected myelinated tracts of the central nervous system (CNS), is atrophic, dark-staining axons containing only diminished, electron-dense, amorphous axoplasm (15). This change is associated with disaggregating myelin sheaths. Degradation of axonal contents is thought to be associated with enhanced axonal activity of calcium-activated proteases (33). As noted earlier, these lesions progress to Wallerian-like fragmentation of affected fiber regions, with phagocytosis and myelin ovoid formation (Fig. 3). The latter are prominent in peripheral nerve. In peripheral nerve there is also subsequent breakdown of phagocytized fiber debris and proliferation of columns of Schwann cells within their basal lamina sheaths. This provides an environment allowing significant myelinated fiber regeneration following OPIDN (13,34). The myelinated fiber degen-

OPIDN

23

eration in the central nervous system evolves more slowly, and is associated with proliferation of astrocytes (astrocytosis) in affected tracts (13) (Fig. 2). As might be expected, regeneration of spinal-cord or brain fibers is not a feature of OPIDN (13). Studies using the Fink-Heimer silver impregnation method revealed a greater spectrum of OPIDN spinal cord and brain lesions. These consisted of degenerating small axons and synaptic boutons in spinal gray matter and some medullary and cerebellar nuclei (29,30). 3. MECHANISMS OF ACTION/TREATMENTS The initial event that occurs in the nervous system within hours after exposure to neuropathy-inducing OP compounds is inhibition of a carboxylesterase called neuropathy target esterase (NTE, also known as neurotoxic esterase). OP compounds that do not induce OPIDN do not inhibit this enzyme (Fig. 4; 35). Inhibition of this enzyme requires the oxon form of the OP compound; P = S compounds and TOTP are protoxicants that require metabolism before NTE is inhibited and OPIDN can be induced. Not only must OP compounds inhibit NTE, but the inhibition must be significant (e.g., about 70% or more after acute administration; approx 50% after multiple exposures) and the interaction between the OP compound and NTE must be so strong that the inhibition essentially irreversible. Without these features, neuropathy will not develop. For most OP compounds that essentially irreversibly inhibit NTE, a leaving group causes the OP compound–NTE complex to become charged, a process known as “aging.” Pretreatment of experimental animals with reversible inhibitors of NTE prevents OP-induced inhibition and aging and protects exposed subjects from OPIDN (5,36). Although inhibition of NTE is a necessary antecedent to OPIDN, the precise relationship between NTE and OPIDN has not been defined. The physiological function of NTE is unknown, so the significance of OP-induced inhibition is unknown. NTE is present in brain, spinal-cord, and peripheral nerves, as well as in non-neural cells (e.g., lymphocytes), but no adverse effects of OP-induced inhibition have been noted outside the nervous system. Inhibition is notable in brain, but this tissue is not a major site of injury in OPIDN. Furthermore, NTE can be inhibited just as significantly in animals not demonstrating clinical evidence of this disorder (e.g., chicks, rodents) as it is in susceptible species. Continued NTE inhibition is not necessary for OPIDN, as activity may be back to pre-exposure levels before clinical signs appear and morphological evidence of the neuropathy develops. The relationship between NTE inhibition and OPIDN has been further complicated in recent years by the discovery that administration of a reversible NTE inhibitor after administration of a neuropathy-inducing OP compound results in exacerbation (promotion) of the OPIDN beyond that which would have been expected by the OP compound alone. This promotion can occur even when NTE is maximally inhibited by the first compound given, suggesting the possibility of an additional, nonesterase site of action for OPIDN promotion (11,37–41). NTE has been a difficult enzyme to purify. It is a serine esterase that is an integral membrane protein with a molecular weight of approx 155 kDa, although a soluble isoform present only in peripheral nerve has recently been described and characterized (42). Immunostaining of the recently purified enzyme indicated that was present in essentially all neurons and that immunostaining was not altered by treatment with neuropa-

24

Ehrich and Jortner

thy-inducing OP compounds. Its structure has reasonable similarity with that of an insect neuronal protein that interrupts the relationship between neurons and glial cells, leading to apoptosis of both. It has been, therefore, suggested that NTE plays a role in cell signaling during development (43). The precise temporal sequence of events that occur between NTE inhibition and onset of clinical and pathological manifestations of OPIDN is as yet unclear. Alterations in threshold excitability of peripheral nerves, axonal transport, neurotrophic factors, protease activity, and protein phosphorylation have all been reported in the interval between NTE inhibition and the onset of OPIDN (2,11,36,37,44). In addition to the modifications of OPIDN in the presence of reversible NTE inhibitors noted earlier, clinical, electrophysiological, and morphological endpoints indicative of OPIDN can be ameliorated by pretreatment with corticosteroids or calcium-channel blockers. These treatments did not, however, affect OP-induced NTE inhibition. They could, however, be affecting events that occur subsequent to NTE inhibition and prior to manifestations of OPIDN (37). 4. FUTURE DIRECTIONS Studies on OPIDN are likely to continue even though few pesticides currently in use are likely to cause the syndrome, even under extremely high exposures. There have been suggestions that OP compounds may have delayed effects following treatment for acute intoxication, after low-dose, long-term exposures, or following exposure to mixtures of OP insecticides and other chemicals (45–47). Much remains to be done to define the precise mechanism(s) responsible for the neurotoxicities reported under these conditions, and any potential relationship to the classical OPIDN described earlier. A recent report indicated that antibodies to nervous-system proteins appeared in OP-exposed subjects (48). Because the precise mechanism(s) and temporal sequence of events that lead to OPIDN are still unknown, the significance of this finding cannot be evaluated. Also, the lack of information on mechanisms makes treatment difficult when exposures occur and risk of the development of neuropathy is high. The symbiotic relationship of NTE inhibition and OPIDN currently leaves many unanswered questions. The recent synthesis of potent and specific NTE inhibitors and the recent work on the molecular biology of NTE have the potential to define its possible role in the nervous system, both in the absence and presence of neuropathy-inducing OP compounds (43,49). OPIDN is expressed only in some long myelinated axonal fibers while others remain unaffected. Why this occurs is unknown. The relationship of the pathology to metabolic events in the whole neuron are at present undefined. Cell-culture systems have potential to provide models for such studies. Culture systems can be chosen in which NTE and acetylcholinesterase are inhibited in a manner similar to that seen in exposed animals (Fig. 4), especially when systems for activating protoxicants are included (50,51). Prevention is always likely to be better than treatment of OPIDN. The establishment of regulatory guidelines for OP compounds that permit risk to be based on biochemical, clinical, and morphological features of OPIDN will continue to be important in

OPIDN

25

protection of the public (8). Review and refinement of the guidelines will continue as more information becomes available on the mechanism(s) associated with this disorder. ACKNOWLEDGMENTS The authors acknowledge support of funds from the US EPA, USDA, and VirginiaMaryland Regional College of Veterinary Medicine. Laboratory personnel, including post-doctoral research associates, graduate students, and laboratory technicians contributing to published studies on OPIDN from our laboratory include D. Barber, K. Dyer-Inzana, L. Shell, H. A. N. El-Fawal, A. Nostrandt, W. McCain, K. Carlson, D. Carboni, C. Massicotte, L. Correll, and S. Perkins. REFERENCES 1. Abou-Donia, M. B. (1981) Organophosphorus ester-induced delayed neurotoxicity. Ann. Rev. Pharmacol. Toxicol. 21, 511–548. 2. Abou-Donia, M. B. (1995) Organophosphorus pesticides, in Handbook of Toxicology (Chang, L. W. and Dyer, R. S., eds.), Marcel Dekker, NY, pp. 419–473. 3. Davis, C. S. and Richardson, R. J. (1980) Organophosphorus compounds, in Experimental and Clinical Neurotoxicology (Spencer, P. S. and Schaumburg, H. H., eds.), Williams & Wilkins, Baltimore, pp. 527–544. 4. Ecobichon, D. J. (1994) Organophosphorus insecticides, in Pesticides and Neurological Diseases (Ecobichon, D. J. and Joy, R. M., eds.), CRC Press, Boca Raton, FL, pp. 171–249. 5. Johnson, M. K. (1982) The target for initiation of delayed neurotoxicity by organophosphorus esters: biochemical studies and toxicological applications, in Reviews in Biochemical Toxicology, vol. 4 (Hodgson, E., Bend, J. R., and Philpot, R.M., eds.), Elsevier Biomedical, New York, pp. 141–212. 6. Metcalf, R. L. (1984) Historical perspective of organophosphorus ester-induced delayed neurotoxicity, in Delayed Neurotoxicity (Cranmer, J. M. and Hixon, E. J., eds.), Intox Press, Little Rock, AK, pp. 7–22. 7. Smith, M. I., Elvore, E., and Frazier, W. H. (1930) The pharmacological action of certain phenol esters, with special reference to the etiology of so-called ginger paralysis. Public Health Rep. 45, 2509–2524. 8. US EPA (1991) Pesticide assessment guidelines, subdivision E. Hazard evaluation: human and domestic animals. Addendum 10: Neurotoxicity, series 81, 82 and 83. National Technical Information Service, Springfield, VA. 9. Weiner, M. and Jortner, B. S. (1999) Organophosphate-induced delayed neurotoxicity of triarylphosphates. Neurotoxicology 20, 653–674. 10. Gallo, M. and Lawryk, N. J. (1991) Organic phosphorus pesticides, in Handbook of Pesticide Toxicology (Hayes, W. J. and Laws, E. R., eds.), Academic Press, San Diego, pp. 917–1123. 11. Ehrich, M. and Jortner, B. S. (2001) Organophosphorus-induced delayed neuropathy, in Handbook of Pesticide Toxicology (Krieger, R., ed.), Academic Press, San Diego, CA, in press. 12. Abou-Donia, M. B. and Lapadula, D. M. (1990) Mechanisms of organophosphorus esterinduced delayed neurotoxicity: type I and type II. Ann. Rev. Pharmacol. Toxicol. 30, 405–440. 13. Jortner, B. S., Shell, L., El-Fawal, H., and Ehrich, M. (1989) Myelinated nerve fiber regeneration following organophosphorus ester-induced delayed neuropathy. Neurotoxicology 10, 717–726.

26

Ehrich and Jortner

14. Moretto, A. and Lotti, M. (1998) Poisoning by organophosphorus insecticides and sensory neuropathy. J. Neurol. Neurosurg. Psychiatry 64, 463–468. 15. Jortner, B. S. and Ehrich, M. (1987) Neuropathological effects of phenyl saligenin phosphate in chickens. Neurotoxicology 8, 303–314. 16. Ehrich, M., Jortner, B. S., and Padilla, S. (1995) Comparison of the relative inhibition of acetylcholinesterase and neuropathy target esterase in rats and hens given cholinesterase inhibitors. Fundam. Appl. Toxicol. 24, 94–101. 17. Funk, K. A., Henderson, J. D., Liu, C. H., Higgins, R. J., and Wilson, B. W. (1994) Neuropathology of organophosphate-induced delayed neuropathy (OPIDN) in young chicks. Arch. Toxicol. 68, 308–316. 18. Padilla S. and Veronesi, B. (1988) Biochemical and morphological validation of a rodent model of organophosphorus-induced delayed neuropathy. Toxicol. Ind. Health 4, 361–371. 19. Cavanagh, J. B. (1954) The toxic effects of tri-ortho-cresyl phosphate on the nervous system. An experimental study in hens. J. Neurol. Neurosurg. Psychiatry 17, 163–172. 20. Bischoff, A. (1970) Ultrastructure of tri-ortho-cresyl phosphate poisoning in the chicken. II. Studies on spinal cord alterations. Acta Neuropathol. 15, 142–155. 21. Bouldin, T. W. and Cavanagh, J. B. (1979a) Organophosphorus neuropathy. I. A teasedfiber study of the spatio-temporal spread of axonal degeneration. Am. J. Pathol. 94, 241–252. 22. Bouldin, T. W. and Cavanagh, J. B. (1979b) Organophosphorus neuropathy. II. A finestructural study of the early stages of axonal degeneration. Am. J. Pathol. 94, 253–270. 23. Pineas, J. (1969) The pathogenesis of dying-back polyneuropathiesm Part I. An ultrastructural study of experimental tri-ortho-cresyl phosphate intoxication in the cat. J. Neuropath. Exp. Neurol. 28, 571–597. 24. Cavanagh, J. B. (1964) Peripheral nerve changes in ortho-cresyl phosphate poisoning in the cat. J. Pathol. Bact. 87, 365–383. 25. Cavanagh, J. B. and Patangia, G. N. (1965) Changes in the central nervous system in the cat as the result of tri-o-cresyl phoshate poisoning. Brain 88, 165–180. 26. Classen, W., Gretener, P., Rauch, M., Weber, E., and Krinke, G. J. (1996) Susceptibility of various areas of the nervous system of hens to TOCP-induced delayed neuropathy. Neurotoxicology 17, 597–604. 27. Krinke, G., Ullmann, L., Sachsee, K., and Hess, R. (1979). Differential susceptibility of peripheral nerves of the hen to tri-ortho-cresyl phosphate and to trauma. Agents Actions 9, 227–231. 28. Krinke, G. J., Classen, W. S., Rauch, M., and Weber, E. (1997) Optimal conduct of the neuropathology evaluation of organophosphorus induced delayed neuropathy in hens. Exp. Toxicol. Pathol. 49, 451–458. 29. Tanaka, D. and Bursian, S. J. (1989) Degeneration patterns in the chicken central nervous system induced by ingestion of the organophosphorus delayed neurotoxin tri-ortho-tolyl phosphate. A silver impregnation study. Brain Res. 484, 240–256. 30. Tanaka, D., Bursian, S. J., Lehning, E. J., and Aulerich, R. J. (1991) Delayed neurotoxic effect of bis (1-methylethyl) phosphorofluoridate (DFP) in the European ferret: a possible mammalian model for organophosphorus-induced delayed neurotoxicity. Neurotoxicology 12, 209–224. 31. Jortner, B. J. (2000) Mechanisms of toxic injury in the peripheral nervous system: neuropathologic considerations. Toxicol. Pathol. 28, 54–69. 32. Bischoff, A. (1967) The ultrastructure of tri-ortho-cresyl phosphate poisoning. I. Studies on the myelin and axonal alterations in the sciatic nerve. Acta Neuropatholog. 9, 158–174. 33. El-Fawal, H. A. N., Correll, L., Gay, L., and Ehrich, M. (1990) Protease activity in brain, nerve, and muscle of hens given neuropathy-inducing organophosphates and a calcium channel blocker. Toxicol. Appl. Pharmacol. 103, 133–142.

OPIDN

27

34. Glazer, E. J., Baker, T., and Riker, W. F. (1978) The neuropathology of DFP at cat soleus neuromuscular junction. J. Neurocytol. 7, 741–758. 35. Ehrich, M. (1988) Cell cultures for screening of antiesterase compounds, in Advances in Animal Alternatives for Safety and Efficacy Testing (Salem, H. and Katz, S. A., eds.), Taylor & Francis, Washington, DC, pp. 229–234. 36. Richardson, R. J. (1995) Assessment of the neurotoxic potential of chlorpyrifos relative to other organophosphorus compounds: a critical review of the literature. J. Toxicol. Environ. Health 44, 135–165. 37. Ehrich, M. (1996) Neurotoxic esterase. A predictor of potential for neuropathy, in Biomarkers for Agrochemicals and Toxic Substances (Blancato, J. N., Brown, R. N., Dary, C. C., and Saleh, M. A. eds.), American Chemical Society, Washington, DC, pp. 79–93. 38. Lotti, M. (1992) The pathogenesis of organophosphate polyneuropathy. CRC Crit. Rev. Toxicol. 21, 465–488. 39. Lotti, M. and Moretto, A. (1999) Promotion of organophosphate induced delayed polyneuropathy by certain esterase inhibitors. Chem. Biol. Interact. 119-120, 519–524. 40. Pope, C. N., Tanaka, D., and Padilla, S. (1993) The role of neurotoxic esterase (NTE) in the prevention and potentiation of organophosphorus-induced delayed neurotoxicity (OPIDN). Chem. Biol. Interact. 87, 395–406. 41. Milatovic, D., Moretto, A., Osman, K. A., and Lotti, M. (1997) Phenyl valerate esterases other than neuropathy target esterase and the promotion of organophosphate polyneuropathy. Chem. Res. Toxicol 10, 1045–1048. 42. Vilanova, E., Escudero, M. A., and Barril, J. (1999) NTE soluble isoforms: new perspective for targets of neuropathy inducers and promoters. Chem. Biol. Interact. 199-120, 525–540. 43. Glynn, P. (1999) Neuropathy target esterase. Biochem. J. 344, 325–631. 44. Pope, C. diLorenzo, K., and Ehrich, M. (1995) Possible involvement of a neurotrophic factor during the early stages of organophosphate-induced delayed neurotoxicity. Toxicol. Lett. 75, 111–117. 45. Abou-Donia, M. B., Wilmarth, K. R., Abdel-Rahman, A., Jensen, K. F., Oehme, F. W., and Kurt, T.L. (1996) Increased neurotoxicity following concurrent exposure to pyridostigmine bromide, DEET, and chlorpyrifos. Fund. Appl. Toxicol. 34, 201–222. 46. Haley, R. W., Horn, J., Roland, P. S., Bryan, W. W., Van Ness, P. C., Bonte, F. J., et al. (1997) Evaluation of neurologic function in Gulf War veterans. JAMA 277, 223–230. 47. Jamal, G. A. (1997) Neurological syndromes of organophosphorus compounds. Adverse Drug React. Toxicol. Rev. 16, 133–170. 48. McConnell, R., Delgado-Tellez, E., Cuadra, R., Torres, E., Keifer, M., Almandarez, J., et al. (1999) Organophosphate neuropathy due to methamidophos: biochemical and neurophysiological markers. Arch. Toxicol. 73, 296–300. 49. Wu, S. Y. and Casida, J. E. (1995) Ethyl octylphosphonofluoridate and analogs: optimized inhibitors of neuropathy target esterase. Chem. Res. Toxicol. 8, 1070–1075. 50. Ehrich, M., Correll, L., and Veronesi, B. (1997) Acetylcholinesterase and neuropathy target esterase inhibitions in neuroblastoma cells to distinguish organophosphorus compounds causing acute and delayed neurotoxicity. Fund. Appl. Toxicol. 38, 55–63. 51. Barber, D., Correll, L., and Ehrich, M. (1999) Comparison of two in vitro activation systems for protoxicant organophosphorous esterase inhibitors. Toxicol. Sci. 47, 16–22.

28

Ehrich and Jortner

Anticholinesterase Insecticides

29

3 Nonesterase Actions of Anticholinesterase Insecticides Carey Pope and Jing Liu

1. INTRODUCTION Organophosphorus (OP) and carbamate compounds have been utilized for a variety of purposes including use as therapeutic agents, agricultural chemicals, plasticizers, lubricants, flame retardants, and fuel additives. Many of the pesticides in use today belong to the OP or carbamate classes of compounds. Some OP compounds, the highly toxic nerve agents, have been developed for chemical warfare, whereas some carbamates have more recently been utilized as prophylactic drugs to prevent the devastating effects of nerve agent exposures (1,2). Although these agents exhibit a wide array of chemical structures, physicochemical properties, and toxicological potencies, the acute toxicity of most OP and carbamate pesticides is initiated by inhibition of the enzyme acetylcholinesterase (AChE, EC 3.1.1.7) in the peripheral and/or central nervous system (PNS/CNS)(3). Although some OP (e.g., glyphosate) and carbamate (e.g., thiuram) pesticides are not potent anticholinesterases, the OP and carbamate agents elicit acute toxicity by covalently binding to the active site serine residue on AChE and thereby inhibiting the catalytic degradation of acetylcholine (4). Under normal conditions, AChE rapidly degrades acetylcholine, terminating the signal for cholinergic neurotransmission. With extensive inhibition of AChE, acetylcholine accumulates in the synapse, resulting in excessive stimulation of acetylcholine receptors on postsynaptic cells and/or end organs. It is generally believed, however, that some degree of AChE inhibition can be tolerated before neurotransmission is affected. With more than about 50% inhibition, signs of “cholinergic” toxicity (autonomic dysfunction, exopthalmus, involuntary movements, muscle fasciculations, changes in heart rate, and in severe cases, respiratory depression) can be elicited (3). In addition, some OP agents can initiate organophosphorus-induced delayed neurotoxicity (OPIDN), a neuropathy of the CNS and PNS. The putative initial target site for this delayed toxicity is another esterase referred to as neurotoxic esterase (NTE, also known as neuropathy target esterase) (5). (Chapter 1 by Chambers and Carr and Chapter 2 by Ehrich and Jortner more extensively discuss these neurotoxicities.) From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

29

30

Pope and Liu

Considerable evidence indicates that many OP and carbamate esterase inhibitors have additional, nonesterase targets that may alter the toxic cascade following AChE inhibition. They may interact directly with additional macromolecular targets, potentially modifying cholinergic toxicity and exerting toxic actions separate from their effects on either AChE or NTE. These nonesterase actions of OP and carbamate insecticides, therefore, may contribute both to qualitative and quantitative differences in toxicity exhibited among these compounds. 2. EVIDENCE OF NEUROTOXICITY 2.1 Nonesterase Enzymes as Macromolecular Targets of Anticholinesterases The enzymes most often associated with the OP and carbamate insecticides are various types of esterases, i.e., target enzymes (AChE, neurotoxic esterase), “marker” enzymes (e.g., plasma cholinesterase) and detoxification enzymes (carboxylesterases, A-esterases). Some anticholinesterases can bind to and inhibit other types of enzymes, however. For example, the OP anticholinesterase diisopropylphosphorofluoridate (DFP) has been used by biochemists for decades to prevent degradation of proteins during isolation procedures (6). The pancreatic protease chymotrypsin is sensitive to inhibition by some OP anticholinesterases (7,8). A low-dosage exposure to the OP insecticide pirimiphos-methyl (10 mg/kg, LD50 about 2 g/kg) was reported to inhibit the activity of several liver proteases (e.g., proline endopeptidase, dipeptidyl aminotransferases I and IV, cathepsin D) (9). The neuropathic OP compound tri-orthocresyl phosphate (TOCP) was also shown to inhibit liver proteases following in vivo exposure in mice (10). Pruett and coworkers (11) compared the in vitro inhibitory potencies of a series of OP compounds towards target esterases (AChE, NTE), the digestive peptidase trypsin, and an enzyme involved in mitogen-induced activation of T lymphocytes. Out of 20 compounds evaluated, some were found to be potent inhibitors of all four enzymes. A more recent study (12) evaluated the comparative inhibitory actions of selected OP anticholinesterases, including profenofos, tribufos, and phenyl saligenin cyclic phosphonate, towards blood clotting factors and digestive enzymes. Thrombin was relatively sensitive to inhibition by tribufos, phenyl saligenin cyclic phosphonate, and profenofos. Other enzymes, e.g., trypsin and elastase, were also relatively sensitive to inhibition by some agents. It was concluded by the authors, however (based on comparatively higher sensitivity of the target enzymes), that toxic manifestations owing to direct inhibitory actions of these compounds on the various blood-clotting factors and digestive enzymes would be unlikely. Choline acetyltransferase, the synthetic enzyme for acetylcholine, was reported to be inhibited in vitro by the prototype OP pesticide parathion (13), but only at millimolar concentrations. The physiological relevance of anticholinesterase interactions with these enzymes is, therefore, generally unclear. Alteration of protein catabolism could, however, be an alternative action for some anticholinesterases (e.g., pirimiphos-methyl).


31

2.2. Cholinergic Receptors as Macromolecular Targets of Anticholinesterases It has been known for decades that the OP insecticides elicit toxicity through inhibition of AChE (14). Interestingly, direct actions of some anticholinesterases on acetylcholine receptors have also been known since the 1950s (15). Acetylcholine receptors have traditionally been subclassified into two major types, muscarinic and nicotinic, based on their sensitivity to the toxins muscarine and nicotine, respectively. Muscarinic receptors are members of the G-protein receptor superfamily coupled to second messenger transduction processes, whereas nicotinic receptors are pentameric structures with ionotropic functions. Five muscarinic cholinergic receptor subtypes (m1–m5) have been identified by genetic cloning (16). A diversity of subtypes of nicotinic receptors have been described based on subunit heterogeneity (17). In brain, the major subtype is composed of α4 and ß2 subunits, whereas pentameric receptors comprised exclusively of α7 subunits appear to be the target of the nicotinic receptor probe, α-bungarotoxin (18). Using single-cell recordings from Electrophorus cells, the OP compounds DFP, phospholine, and paraoxon all had rapidly reversible direct antagonistic actions on the nicotinic receptor (19). The block of nicotinic receptor function in these studies required considerably higher concentrations, however, than those needed to potentiate the effect of exogenously added acetylcholine (which had its presumable effect through OPinduced AChE inhibition). Eldefrawi and Eldefrawi (20) reported that several OP pesticides at high concentrations (100 µM), such as azinphos-methyl, dichlorvos, dicrotophos, and monocrotophos, could bind in vitro to nicotinic receptors of the electric organ of Torpedo. These same OP agents had little effect on total muscarinic receptor ([3H]quinuclidinyl benzilate; QNB) binding in either insect or mammalian brain. Bakry and coworkers (21) showed that other OP toxicants, including echothiophate and the nerve agent VX, could also bind directly to neuromuscular nicotinic receptors at high concentrations. Thus, these early studies suggested that some OP anticholinesterases could bind to nicotinic cholinergic receptors but apparently only at concentrations difficult or impossible to achieve in vivo. Carbamate anticholinesterases such as those used in the treatment of myasthenia gravis have also been shown to interact directly with nicotinic receptors (22). The binding of the carbamate physostigmine (eserine) has been shown to be to a site on nicotinic receptors distinct from the acetylcholine recognition site (23). Electrophysiological studies using the frog sciatic nerve-sartorius muscle preparation indicated that, at low concentrations, physostigmine acted on nicotinic receptors indirectly through AChE inhibition, whereas at higher concentrations physostigmine appeared to act as an open channel blocker. Patch-clamp studies further identified a novel agonist-binding site for physostigmine activated by much lower concentrations of the carbamate (24). Other carbamates (e.g., pyridostigmine and neostigmine) were also shown to activate nicotinic receptors through this allosteric site, but at much higher concentrations (25,26). Physostigmine has been shown to activate nicotinic receptors through this allosteric site even following desensitization of the receptors by acetylcholine (27). Together, these data suggest that nicotinic receptor activation through this allosteric site may occur selectively by some anticholinesterases, independent of AChE inhibition and its consequences on cholinergic neurotransmission.

32

Pope and Liu

Several studies have evaluated the ability of OP or carbamate anticholinesterases to modulate agonist-stimulated [3H]phencyclidine (PCP) binding to nicotinic receptors. PCP binds to the nicotinic channel in the open position; thus its binding can be an indicator of activation of the nicotinic receptor and used as a method to study agonistinduced desensitization (28). Mansour and coworkers (29) reported that neostigmine could activate [3H]PCP binding to reconstituted nicotinic receptors from Torpedo nobiliana. Moreover, neostigmine and physostigmine were relatively potent inhibitors of carbachol-stimulated [3H]PCP binding (Ki = 10 – 20 µM) whereas pyridostigmine and aldicarb were much less potent. DFP also reduced agonist-stimulated [3H[PCP binding to Torpedo and desensitization of nicotinic receptors (30). More recently, Katz and coworkers (31) reported that chlorpyrifos, chlorpyrifos-oxon, parathion, and paraoxon all decreased agonist-stimulated binding of [3H]thienyl-cyclohexylpiperidine (TCP, an analog of phencyclidine) to nicotinic receptors in Torpedo membranes in a concentration-dependent and reversible manner (IC50 values from 5–300 µM). All four OP toxicants increased TCP binding in the absence of agonist. Although neither of these OP agents had an effect on equilibrium binding to [α-125I]bungarotoxin in Torpedo membranes in vitro, they increased the apparent affinity of the membranes for the agonist carbachol. Collectively, these data indicate that some OP and carbamate anticholinesterases can desensitize nicotinic receptors. In general, however, concentrations required to affect nicotinic receptor binding appear considerably higher than would be expected to occur in vivo. In addition to actions on nicotinic receptors, a number of studies have reported that some anticholinesterases interact directly with muscarinic receptors. Using QNB, a muscarinic antagonist that binds to all known subtypes of muscarinic receptors with equal affinities, Volpe and coworkers (32) showed that dichlorvos, paraoxon, and tetraethylpyrophosphate inhibited QNB binding in membranes from bovine caudate at low (5–50 nM) concentrations. Katz and Marquis (33) reported that paraoxon at extremely low levels (as low as 10–15 M) could block radioligand (QNB) binding to muscarinic M2 and M3 receptors. The effects of several OP anticholinesterases on [3H]-N-methylscopolamine (NMS) binding has been investigated by Ehrich and coworkers using SH-SY5Y neuroblastoma cells (34). In contrast to QNB, which binds both surface and sequestered muscarinic receptors, NMS is thought to label preferentially surface oriented receptors due to its quaternary structure (35). Relatively high concentrations (about 1 mM) of paraoxon and phenyl saligenin cyclic phosphate (PSP) inhibited ligand binding under saturating conditions whereas lower concentrations (10 µM) were effective when the radioligand concentration was half-saturating. Selective use of antagonists suggested that these cells possessed muscarinic receptors most sensitive to M3 antagonists. Both paraoxon and PSP reduced basal inositol phosphate levels in a concentration-dependent manner, but the induced changes in inositol phosphates were not sensitive to muscarinic or nicotinic antagonists. Together, the results suggested that two OP agents, one primarily associated with acute neurotoxicity (paraoxon) and the other a known delayed neurotoxicant (PSP), could interact directly with muscarinic receptors. Furthermore, the inhibition of inositol phosphate production by the OP toxicants was apparently through interaction with another unknown site of action, independent of either target esterases or cholinergic receptors (34).


33

Bakry and coworkers (21) first evaluated the effects of OP agents on muscarinicreceptor binding using the putative subtype-selective muscarinic agonist, [3H]cisdioxolane (CD). Using Chinese hamster ovary (CHO) cells transfected with one of each of the muscarinic-receptor subtypes (m1–m5), it was shown that CD labels predominantly the m2-receptor subtype (36). Echothiophate and VX were shown to be extremely potent at blocking binding to CD. Echothiophate, paraoxon, and the nerve agents VX, soman, and tabun were also shown to block CD binding to rat cardiac tissues in vitro with submicromolar concentrations (37). Jett and coworkers (38) reported that paraoxon could inhibit CD binding and cAMP formation in an atropine-sensitive manner in rat striatal cells. Ward and coworkers (39) observed that paraoxon and malaoxon were relatively potent blockers of specific CD binding to both hippocampal and cortical membranes, whereas the parent compounds parathion and malathion were much less potent. Huff and coworkers (40) reported that chlorpyrifos-oxon was a potent inhibitor of specific CD binding in rat striatum (IC50 = 22 nM) and could inhibit forskolin-stimulated cAMP formation (IC50 = 155 nM), but in an apparently atropine-insensitive manner. Ward and Mundy (41) compared the abilities of paraoxon, malaoxon, and chlorpyrifos-oxon to alter muscarinic receptor-mediated phosphoinositide turnover and cAMP formation in slices of rat frontal cortex. All three OP agents inhibited CD binding and cAMP formation in a concentration-dependent manner at sub-micromolar concentrations (potencies: chlorpyrifos-oxon > paraoxon > malaoxon). Chlorpyrifos-oxon was quite potent at inhibiting forskolinstimulated cAMP formation (IC50 = 57 nM). Neither of the OP agents affected basal or carbachol-stimulated phosphoinositide turnover, however. These findings extended those of others (38,40), indicating that some OP compounds can inhibit cAMP formation through interaction with muscarinic receptors, presumably either the m2 or m4 subtype. Moreover, the lack of effects on phosphoinositide turnover indicated that changes in cAMP formation were direct actions and not a consequence of AChE inhibition. In addition to affecting muscarinic binding of QNB, NMS, and CD, more recent findings suggest that binding of the muscarinic agonist [3H]oxotremorine is also sensitive to relatively low concentrations of some OP and carbamate agents (42). Paraoxon and physostigmine were both reported to be relatively potent displacers of radiolabeled oxotremorine binding to rat brain membranes. Using a novel cell-function assay that measures changes in acidification of themedium using a microphysiometer, Cao and coworkers (43) evaluated the in vitro actions of selected OP anticholinesterases. The OP agents were added to either hepatocytes or neuroblastoma cells in culture and changes in metabolic rates evaluated. Parathion, chlorpyrifos, and their oxons stimulated hepatocyte but inhibited neuroblastoma metabolism, and after 24 h of exposure exhibited LC50 values in the low micromolar range (3.7–31 µM). The nerve agent VX inhibited metabolism in both cell types. Effects of both VX and paraoxon were further shown to be partially blocked by atropine, suggesting involvement of muscarinic receptors in this cytotoxic response. Thus, the studies on interactions of cholinesterase-inhibiting insecticides with cholinergic receptors suggest that both muscarinic- and nicotinic-receptor subtypes may be sensitive to lower concentrations of a variety of anticholinesterases. Furthermore, the results taken together indicate that such direct anticholinesterase-cholinergic

34

Pope and Liu

receptor interactions can alter ion flux (through nicotinic receptors) and modulate the levels of second messengers (through muscarinic receptors) in either presynaptic or postsynaptic cells and thereby alter cellular function at potentially relevant concentrations. 2.3. Noncholinergic Neurotransmitter Receptors as Targets of Anticholinesterases Some studies also indicate that noncholinergic receptors may be targets for anticholinesterases. El-Sebae and coworkers (44) reported that the OP agents cyanofenphos, leptophos, salithion, and TOCP could block [3H]norepinephrine binding in heart in vitro with potencies similar to that of the prototype beta-blocker, propranolol. Johnson and Michaelis (45) showed that DFP, dichlorvos, cyanophos and mipafox were inhibitors of the N-methyl-D-aspartate (NMDA) receptor-protein complex in rat-brain membranes labeled with the radioligand 3-((+)-2-carboxypiperazin-4-yl)-[1,2-3H]propyl1-phosphonic acid ([3H]CPP), with IC50 values around 10 µM. The blockade of [3H]CPP binding by DFP appeared irreversible. None of the anticholinesterases from this study (45) affected binding to other NMDA ligands, however, including kainiteand quisqualate-sensitive [3H]AMPA, strychnine-sensitive [3H]glycine, and [3H]TCP. Some OP anticholinesterases have been reported to be relatively potent inhibitors of GABAA receptor binding to t-[35S]butyl-bicyclophophorothionate in electric organ preparations from Torpedo californica (46). Reversible inhibition of radioligand binding to brain and heart adenosine receptors in vitro has also been demonstrated for soman, sarin, and tabun, with soman being most potent but still relatively weak (Ki = 37–57 µM) (47,48). Thus, a variety of noncholinergic neurotransmitter receptors may have affinity for some of the anticholinesterases. In most cases, however, the toxicological importance of these noncholinergic receptor interactions is unclear. 2.4. Effects of Anticholinesterases on Neurotransmitter Release Katz and Marquis (33) proposed that some anticholinesterases may bind to muscarinic autoreceptors regulating acetylcholine release in the brain. Presynaptic muscarinic autoreceptors have been shown to inhibit transmitter release by a negative feedback mechanism (49–51), and this adaptive mechanism could be important in the ultimate toxicity of AChE inhibition. Watson and coworkers (52) concluded that [3H]CD labels principally presynaptic muscarinic receptors. Thus, if some anticholinesterases can directly activate or block presynaptic muscarinic autoreceptors at physiologically relevant concentrations, this could provide a mechanism for selective modulation of cholinergic toxicity by those agents. Rats treated with high dosages of parathion and chlorpyrifos exhibited similar rates and maximal degrees of brain AChE inhibition (53) but different degrees of cholinergic toxicity, with parathion-treated rats exhibiting more extensive signs of toxicity (54). Qualitative differences in binding to the muscarinic agonist CD were noted between the two OP pesticide groups, with brain CD binding reduced following parathion exposure but increased following chlorpyrifos administration (54). Upregulation of brain CD binding sites was also reported in female rats following chlorpyrifos exposure (55). As CD-labeled receptors had been reported to be presynaptically located (52) and proposed to be coupled to muscarinic-autoreceptor function (33,56), it was


35

hypothesized that qualitative differences in modulation of those muscarinic receptors following OP exposure might be an indication of differential effects on presynaptic regulation of acetylcholine release. Muscarinic autoreceptor function was reduced in a time-dependent manner by both parathion and chlorpyrifos (56). Of importance, however, autoreceptor function was inhibited within 2 d after parathion exposure but was still functional at that timepoint following chlorpyrifos administration, i.e., at the time of peak incidence of toxicity (56). These findings suggested that some OP anticholinesterases can affect presynaptic muscarinic-autoreceptor function in a selective manner, potentially playing a role in their differential toxicity. Other cholinergic presynaptic processes may contribute to differential toxicity of parathion and chlorpyrifos following dosages causing similar changes in AChE activity. High-affinity choline uptake, the rate-limiting step in acetylcholine synthesis, is differentially reduced in a time-dependent manner by chlorpyrifos and parathion (57). Choline uptake has been reported to be regulated by cAMP (58,59), which in turn can be modulated by activation of muscarinic m2 and/or m4 receptors. If some anticholinesterases can differentially activate these subtypes of muscarinic receptors, changes in acetylcholine synthesis could alter the toxic consequences of cholinesterase inhibition. Thus, presynaptic muscarinic receptors may be coupled to two different neurochemical processes, acetycholine synthesis and acetylcholine release. The selective activation and/or inhibition of these receptors by some anticholinesterases could potentially lead to differential toxicity, in particular with high, acute exposures. At relatively low in vitro concentrations, paraoxon (0.3–3 µM) increased the frequency of miniature postsynaptic currents (MPCs) in cultured hippocampal cells induced by GABA (60). At higher concentrations (i.e., at concentrations far exceeding those necessary for inhibition of AChE), frequency, decay-times, and peak amplitudes of the GABA-mediated MPCs were substantially reduced. Paraoxon (300 nM) also markedly increased glutamate-stimulated MPC frequency. In contrast, nicotinic agonists (including acetylcholine, 1 mM) and antagonists were without effect on either GABA or glutamate-mediated MPCs. At higher concentrations, paraoxon inhibited in a noncompetitive and reversible manner GABAA, glycine, NMDA, and nicotinic receptors. The authors concluded that paraoxon at submicromolar concentrations enhances the frequency of neurotransmitter (GABA and glutamate) release in hippocampal neurons and at higher concentrations blocks multiple types of postsynaptic receptors, apparently acting as an open-channel blocker. As acetylcholine had no effect on MPCs induced by either GABA or glutamate, the possibility that paraoxon acted indirectly through AChE inhibition was dismissed. The nerve agent VX was both more potent and efficacious than paraoxon at increasing transmitter release from hippocampal neurons in culture, acting at levels as low as 1 nM (61). The nerve agent soman, however, had no effect on transmitter release in this system, even at concentrations higher than necessary to cause complete AChE inhibition. Dam and coworkers (62) reported that chlorpyrifos increased norepinephrine release from rat brain synaptosomes in vitro at relatively low concentrations (50 µg/mL, about 143 µM). The effect of chlorpyrifos on norepinephrine release was not sensitive to either the muscarinic antagonist atropine or the nicotinic antagonist mecamylamine. Thus, anticholinesterases have been reported to increase neurotransmitter release in a variety of settings in vitro and to modulate muscarinic receptor-mediated regulation of acetylcholine release in vivo. Again, these

36

Pope and Liu

findings suggested that other presynaptic processes may be targeted by some OP anticholinesterases, independent of their action on AChE. 2.5. Potential Nonesterase Actions of Anticholinesterases on Neuronal Development The OP insecticide chlorpyrifos has been shown to affect macromolecular (DNA, RNA, and protein) synthesis in the postnatally developing brain in a time-dependent and maturational state-dependent manner (63–66). Song and coworkers (67) reported that multiple components of the adenylyl cyclase signaling pathway were disrupted by postnatal chlorpyrifos exposures. These included reduction in adenylyl cyclase activity and alterations in G-protein function, even in noncholinergic (adrenergic) systems. Using the cultured rat-embryo model, chlorpyrifos was shown to alter the incidence and orientation of mitotic cells as well as induce cytoplasmic vacuolation with concentrations as low as 0.5 µg/mL (68). The results from these in vitro and in vivo studies suggested that chlorpyrifos could potentially alter developmental processes and induce cell death in the developing nervous system. AChE itself appears to have different actions aside from its critical role in cholinergic neurotransmission (69–71). Several reports indicate that AChE expression during development coincides with axonal outgrowth (72–75). In cultured dorsal-root ganglion cells, the level of AChE activity has been shown to correlate well with the extent of neurite outgrowth (76). The morphogenic role of AChE does not appear to depend on its catalytic activity, however (77,78). Nevertheless, some OP and carbamate anticholinesterases have been reported to affect developmental processes associated with this morphogenic function. Physostigmine, neostigmine, and edrophonium all caused retraction of growth cones in primary chicken neurons in vitro, with physostigmine being the most potent (79). OP compounds that caused delayed neuropathy inhibited neurite extension in C6 glioma cells and N-18 mouse neuroblastoma cells (80). Flaskos and coworkers (81) reported that tricresyl phosphate, triphenyl phosphite, and paraoxon inhibited cell growth in PC12 cells, and that subcytotoxic concentrations of tricresyl phosphate reduced neurite density. Similarly, two direct-acting OP inhibitors of NTE were reported to inhibit neurite outgrowth in PC12 cells at concentrations well below those inducing cytotoxicity, whereas the potent anticholinesterase chlorpyrifos-oxon only affected neurite extension at cytotoxic levels (82). The protoxicant chlorpyrifos was also shown to inhibit neurite outgrowth in PC12 cells (83), as was its major metabolite (3,5,6-trichloropyridinol), with this occurring in the absence of cholinesterase inhibition. While these are all in vitro studies, inhibition of neurite outgrowth, and alterations in growth-cone dynamics may be additional actions of some of the OP and carbamate insecticides. In vivo studies should be conducted to determine if such developmental processes are altered following exposure to these toxicants at relevant dosages. AChE has also been shown to be secreted in a variety of brain regions (85,86). In hippocampus, secreted AChE appears to be associated with cholinergic neurotransmission (85). It has been proposed that the secretion of catalytically active AChE may be an additional mechanism (along with muscarinic autoreceptor-mediated inhibition of release) to counteract elevated synaptic levels of acetylcholine. In contrast, secreted


37

AChE does not appear associated with cholinergic neurotransmission in other brain regions (87,88). Administration of AChE directly into the substantia nigra or cerebellum caused hyperpolarization, and this effect was not blocked by preinhibition of the AChE by soman (89,90). It was subsequently shown that the hyperpolarizing effect of AChE on nigral neurons could be mimicked by NMDA and blocked by NMDA receptor antagonists (91). Holmes and coworkers (92) reported that mouse recombinant G1 and G4 AChE increased both survival and neurite outgrowth of cultured midbrain neurons. 3. FUTURE DIRECTIONS The investigations cited above indicate that OP and carbamate insecticides have nonclassical, noncholinergic actions, but that physiological relevance and toxicological implications are still unclear for many of these effects. Studies with relevant dosages in vivo are needed to further elucidate differential toxicities. As with the effects of anticholinesterases on neurite extension described earlier, catalytic activity may not be important in the actions of secreted AChE. This would argue against the possibility that anticholinesterases could impair the physiological role of secreted AChE, and in fact it has been shown that one anticholinesterase (soman) did not affect the ability of AChE to hyperpolarize cerebellar and nigral cells (89,90). It remains possible, however, that other anticholinesterases may differentially alter the actions of secreted AChE. Although the physiological role(s) of secreted AChE is unclear, more studies to compare effects of other anticholinesterases on associated neuronal processes need to be performed. Furthermore, although any interaction of anticholinesterases with secreted AChE could not be considered a “nonesterase” action, such effects could be outside the classical role of AChE in cholinergic neurotransmission and therefore possibly contribute to the differential toxicity of these toxicants. Among the macromolecular targets for anticholinesterases, several that could potentially elicit toxicity through noncholinergic mechanisms are enzymes that are not esterases, cholinergic receptors, noncholinergic transmitter receptors, autoreceptors for neurotransmitter release, and targets for neuronal development. These areas deserve continued investigation. A variety of enzymes, receptors, and other macromolecules may be sensitive to direct interactions with OP and carbamate anticholinesterases. Both nicotinic and muscarinic cholinergic receptors appear to be targeted by some OP and carbamate insecticides, in some cases at very low concentrations. Noncholinergic neurotransmitter receptors can also be macromolecular targets for some OP and carbamate insecticides. Anticholinesterase-induced changes in neurotransmitter (e.g., glutamate, GABA) release occur in vitro at feasible concentrations with some agents. Owing to its putative morphogenic role in neuronal differentiation, OP or carbamate toxicant interactions with AChE during maturation of the nervous system may lead to subtle developmental changes. Thus, there is considerable evidence to suggest that the OP and carbamate insecticides may interact directly with macromolecules in addition to the cholinergic target-enzyme AChE or the putative OP neuropathy target, NTE. In all cases, however, it is imperative to consider the relative potencies of the OP and carbamate agents towards the target esterases and any additional sites of action, in vitro and in vivo, when evaluating potential hazards.

38

Pope and Liu

ACKNOWLEDGMENTS Work in Dr. Pope’s laboratory was partially supported by grant R825811 from the US EPA and by grant RO1ES09119 from NIEHS. REFERENCES 1. Sidell, F. R. and Borak, J. (1992) Chemical warfare agents: II. Nerve agents. Ann. Emerg. Med. 21, 865–871. 2. Ehrich, M. (1998) Organophosphates, in Encyclopedia of Toxicology (Wexler, P., ed.), Academic Press, San Diego, CA, pp. 467–471. 3. Nostrandt, A. C., Padilla, S., and Moser, V. C. (1997) The relationship of oral chlorpyrifos effects on behavior, cholinesterase inhibition, and muscarinic receptor density in rat. Pharmacol. Biochem. Behav. 58, 15–23. 4. Gallo, M. A. and Lawryk, N. J. (1991) Organic phosphorus pesticides, in Handbook of Pesticide Toxicology (Hayes, W. J. and Laws, E. R., eds.), Academic Press, San Diego, CA, pp. 917–1123. 5. Weiner, M. L. and Jortner, B. S. (1999) Organophosphate-induced delayed neurotoxicity of triarylphosphates. Neurotoxicology 20, 653–673. 6. Scopes, R. K. (1982). Maintenance of active enzymes, in Protein Purification, Principles and Practice, Springer Verlag, NY, pp. 185–200. 7. Hamilton, S. E., Dudman, A. P., DeJersey, J., Stoops, J. K., and Zerner, B. (1975) Organophosphate inhibitors: the reactions of bis(p-nitrophenyl)methyl phosphate with liver carboxylesterases and alpha-chymotrypsin. Biochim. Biophys. Acta 377, 282–296. 8. Johnson, M. K. and Clothier, B. (1980) Biochemical events in delayed neurotoxicity: is aging of chymotrypsin inhibited by saligenin cyclic phosphates a model for aging of neurotoxic esterase? Toxicol. Lett. 5, 95–98. 9. Mantle, D., Saleem, M. A., Williams, F. M., Wilkins, R. M., and Shakoori, A. R. (1997) Effect of pirimiphos-methyl on proteolytic enzyme activities in rat heart, kidney, brain and liver tissues in vivo. Clin. Chem. Acta 262, 89–97. 10. Saleem, M. A., Williams, F. M., Wilkins, R. M., Shakoori, A. R., and Mantle, D. (1998) Effect of tri-o-cresyl phosphate (TOCP) on proteolytic enzyme activities in mouse liver in vivo. J. Environ. Pathol. Toxicol. Oncol. 17, 69–73. 11. Pruett, S. B., Chambers, H. W., and Chambers, J. E. (1994) A comparative study of inhibition of acetylcholinesterase, trypsin, neuropathy target esterase, and spleen cell activation by structurally related organophosphorus compounds. J. Biochem. Toxicol. 9, 319–327. 12. Quistad, G. B. and Casida, J. E. (2000) Sensitivity of blood-clotting factors and digestive enzymes to inhibition by organophosphorus pesticides. J. Biochem. Mol. Toxicol. 14, 51–56. 13. Murumatsu, M. and Kuriyama, K. (1976) Effect of organophosphorus compounds on acetylcholine synthesis in brain. Jpn. J. Pharmacol. 26, 249–254. 14. DuBois, K. P., Doull, J., Salerno, P. R., and Coon, J. (1949) Studies on the toxicity and mechanisms of action of p-nitrophenyl diethyl thionophosphate (parathion). J. Pharmacol. Exp. Ther. 95, 79–91. 15. Frederickson, T. (1958) Further studies on fluoro-phosphorylcholines. Pharmacological properties of two new analogues. Arch. Int. Pharmacodyn. 115, 474–482. 16. Bonner, T. I. (1989) The molecular basis of muscarinic receptor diversity. Trends Neurosci. 12, 148–151. 17. McGehee, D. S. and Role, L. W. (1995) Physiological diversity of nicotinic acetylcholine receptors expressed by vertebrate neurons. Ann. Rev. Physiol. 57, 521–546. 18. Albuquerque, E. X., Alkondon, M., Pereira, E. F., Castro, N. G., Schrattenholz, A., Barbosa, C. T., et al. (1997) Properties of neuronal nicotinic acetylcholine receptors: phar-


19. 20. 21.

22. 23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

39

macological characterization and modulation of synaptic function. J. Pharmacol. Exp. Ther. 280, 1117–1136. Bartles, E. and Nachmansohn, D. (1969) Organophosphate inhibitors of acetylcholinereceptor and –esterase tested on the electroplax. Arch. Biochem. Biophys. 133, 1–10. Eldefrawi, M. E. and Eldefrawi, A. T. (1983) Neurotransmitter receptors as targets for pesticides. J. Environ. Sci. Health [B] 18, 65–88. Bakry, N. M., el-Rashidy, A. H., Eldefrawi, A. T., and Eldefrawi, M. E. (1988) Direct actions of organophosphate anticholinesterases on nicotinic and muscarinic acetylcholine receptors. J. Biochem. Toxicol. 3, 235–259. Seifert, S. A. and Eldefrawi, M. E. (1974) Affinity of myasthenia drugs to acetylcholinesterase and acetylcholine receptor. Biochem. Med. 10, 258–265. Albuquerque, E. X., Deshpande, S. S., Kawabuchi, M., Aracava, Y., Idriss, M., Rickett, D. L., and Boyne, A. F. (1985) Multiple actions of anticholinesterase agents on chemosensitive synapses: molecular basis for prophylaxis and treatment of organophosphate poisoning. Fundam. Appl. Toxicol. 5, S182–S203. Shaw, K. P., Aracava, Y., Akaike, A., Daly, J. W., Rickett, D.L., and Albuquerque, E. X. (1985) The reversible cholinesterase inhibitor physostigmine has channel-blocking and agonist effects on the acetylcholine receptor-ion channel complex. Mol. Pharmacol. 28, 527–538. Akaike, A., Ikeda, S. R., Brookes, N., Pascuzzo, G. J., Rickett, D. L., and Albuquerque, E. X. (1984) The nature of the interactions of pyridostigmine with the nicotinic acetylcholine receptor-ionic channel complex. II. Patch clamp studies. Mol. Pharmacol. 25, 102–112. Albuquerque, E. X., Akaike, A., Shaw, K. P., and Rickett, D. L. (1984) The interaction of anticholinesterase agents with the acetylcholine receptor-ionic channel complex. Fundam. Appl. Toxicol. 4, S27–S33. Kuhlmann, J., Okonjo, K. O., and Maelicke, A. (1991) Desensitization is a property of the cholinergic binding region of the nicotinic acetylcholine receptor, not of the receptor-integral ion channel. FEBS Lett. 279, 216–218. Eldefrawi, M. E., Eldefrawi, A. T., Aronstam, R. S., Maleque, M. A., Warnick, J. E., and Albuquerque, E. X. (1980) [3H]Phencyclidine: a probe for the ionic channel of the nicotinic receptor. Proc. Natl. Acad. Sci. USA 77, 7458–7462. Mansour, N. A., Valdes, J. J., Shamoo, A. E., and Annau, Z. (1987) Biochemical interactions of carbamates and ecothiophate with the activated conformation of nicotinic acetylcholine receptor. J. Biochem. Toxicol. 2, 25–42. Eldefrawi, M. E., Schweizer, G., Bakry, N. M., and Valdes, J. J. (1988) Desensitization of the nicotinic acetylcholine receptor by diisopropylfluorophosphate. J. Biochem. Toxicol. 3, 21–32. Katz, E. J., Cortes, V. I., Eldefrawi, M. E., and Eldefrawi, A. T. (1997) Chlorpyrifos, parathion, and their oxons bind to and desensitize a nicotinic acetylcholine receptor: relevance to their toxicities. Toxicol. Appl. Pharmacol. 146, 227–236. Volpe, L. S., Biagioni, T. M., and Marquis, J. K. (1985) In vitro modulation of bovine caudate muscarinic receptor number by organophosphates and carbamates. Toxicol. Appl. Pharmacol. 78, 226–234. Katz, L. S. and Marquis, J. K. (1989) Modulation of central muscarinic receptor binding in vitro by ultralow levels of the organophosphate paraoxon. Toxicol. Appl. Pharmacol. 101, 114–123. Ehrich, M., Intropido, L., and Costa, L. G. (1994) Interaction of organophosphorus compounds with muscarinic receptors in SH-SY5Y human neuroblastoma cells. J. Toxicol. Environ. Health 43, 51–63. Fisher, S. K. (1988) Recognition of muscarinic cholinergic receptors in human SK-N-SH neuroblastoma cells by quaternary and tertiary ligands is dependent upon temperature, cell integrity, and the presence of agonists. Mol. Pharmacol. 33, 414–422.

40

Pope and Liu

36. Huff, R. A. and Abou-Donia, M. B. (1994) cis-Methyldioxolane specifically recognizes the m2 muscarinic receptor. J. Neurochem. 62, 388–391. 37. Silveira, C. L., Eldefrawi, A. T., and Eldefrawi, M. E. (1990) Putative M2 muscarinic receptors of rat heart have high affinity for organophosphorus anticholinesterases. Toxicol. Appl. Pharmacol. 103, 474–481. 38. Jett, D. A., Abdallah, E. A. M., El-Fakahany, E. E., Eldefrawi, M. E., and Eldefrawi, A. T. (1991) High-affinity activation by paraoxon of a muscarinic receptor subtype in rat brain striatum. Pest. Biochem. Physiol. 39, 149–157. 39. Ward, T. R., Ferris, D. J., Tilson, H. A., and Mundy, W. R. (1993) Correlation of the anticholinesterase activity of a series of organophosphates with their ability to compete with agonist binding to muscarinic receptors. Toxicol. Appl. Pharmacol. 122, 300–307. 40. Huff, R. A., Corcoran, J. J., Anderson, J. K., and Abou-Donia, M. B. (1994) Chlorpyrifos oxon binds directly to muscarinic receptors and inhibits cAMP accumulation in rat striatum. J. Pharmacol. Exp. Ther. 269, 329–335. 41. Ward, T. R. and Mundy, W. R. (1996) Organophosphorus compounds preferentially affect second messenger systems coupled to M2/M4 receptors in rat frontal cortex. Brain Res. Bull. 39, 49–55. 42. Van Den Beukel, I., Dijcks, F. A., Vanderheyden, P., Vauquelin, G., and Oortgiesen, M. (1997) Differential muscarinic receptor binding of acetylcholinesterase inhibitors in rat brain, human brain and Chinese hamster ovary cells expressing human receptors. J. Pharmacol. Exp. Ther. 281, 1113–1119. 43. Cao, C. J., Mioduszewski, R. J., Menking, D. E., Valdes, J. J., Katz, E. J., Eldefrawi, M. E., and Eldefrawi, A. T. (1999) Cytotoxicity of organophosphate anticholinesterases. In vitro Cell Dev. Biol. Anim. 35, 493–500. 44. El-Sebae, A. H., Soliman, S. A., Ahmed, N. S., and Curley, A. (1981) Biochemical interaction of six OP delayed neurotoxicants with several neurotargets. J. Environ. Sci. Health [B.] 16, 465–474. 45. Johnson, P. S. and Michaelis, E. K. (1992) Characterization of organophosphate interactions at N-methyl-D-aspartate receptors in brain synaptic membranes. Mol. Pharmacol. 41, 750–756. 46. Gant, D. B., Eldefrawi, M. E., and Eldefrawi, A. T. (1987) Action of organophosphates on GABAA receptor and voltage-dependent chloride channels. Fundam. Appl. Toxicol. 9, 698–704. 47. Lau, W-M, Freeman, S. E., and Szilagyi, M. (1988) Binding of some organophosphorus compounds at adenosine receptors in guinea pig brain membranes. Neurosci. Lett. 94, 125–130. 48. Lau, W-M, Szilagyi, M., and Freeman, S. E. (1991) Effects of some organophosphorus compounds on the binding of a radioligand (8-cyclopentyl 1,3-[3H]dipropylxanthine) to adenosine receptors in ovine cardiac membranes. J. Appl. Toxicol. 11, 411–414. 49. Weiler, M. H. (1989) Muscarinic modulation of endogenous acetylcholine release in rat neostriatal slices. J. Pharmacol. Exp. Ther. 250, 617–623. 50. Feuerstein, T. J., Lehmann, J., Sauermann, W., van Velthoven, V., and Jackisch, R. (1992) The autoinhibitory feedback control of acetylcholine release in human neocortex tissue. Brain Res. 572, 64–71. 51. Kitaichi, K., Hori, T., Srivastava, L. K., and Quirion, R. (1999) Antisense oligodeoxynucleotides against the muscarinic m2, but not m4, receptor supports its role as autoreceptors in the rat hippocampus. Brain Res. Mol. Brain Res. 67, 98–106. 52. Watson, M., Roeske, W. R., Vickroy, T. W., Smit, T. L., Akiyama, K., Gulya, K., et al. (1986) Biochemical and functional basis of putative muscarinic receptor subtypes and its implications. Trends Pharmacol. Sci. (Suppl.) 2, 44–55. 53. Pope, C. N., Chakraborti, T. K., Chapman, M. L., Farrar, J. D., and Arthun, D. (1991)


54.

55. 56.

57.

58.

59.

60.

61.

62. 63. 64. 65.

66.

67.

68.

69. 70. 71.

41

Comparison of in vivo cholinesterase inhibition in neonatal and adult rats by three organophosphorothioate insecticides. Toxicology 68, 51–61. Chaudhuri, J., Chakraborti, T. K., Chanda, S., and Pope, C. N. (1993) Differential modulation of organophosphate-sensitive muscarinic receptors in rat brain by parathion and chlorpyrifos. J. Biochem. Toxicol. 8, 207–216. Liu, J. and Pope, C. N. (1996) Effects of chlorpyrifos on high-affinity choline uptake and [3H]hemicholinium-3 binding in rat brain. Fundam. Appl. Toxicol. 34, 84–90. Pope, C. N., Chaudhuri, J., and Chakraborti, T. K. (1995) Organophosphate-sensitive cholinergic receptors: possible role in modulation of anticholinesterase toxicity, in Enzymes of the Cholinesterase Family (Quinn, D. M., Balasubramanian, A. S., Doctor, B. P., and Taylor, P., eds.), Plenum, NY, pp. 305–312. Liu, J. and Pope, C. N. (1998) Comparative presynaptic neurochemical changes in rat striatum following exposure to chlorpyrifos or parathion. J. Toxicol. Environ. Health 53, 531–544. Cancela, J. M., Bertrand, N., and Beley, A. (1995) Involvement of cAMP in the regulation of high affinity choline uptake by rat brain synaptosomes. Biochem. Biophys. Res. Commun. 213, 944–949. Vogelsberg, V., Neff, N. H., and Hadjiconstantinou, M. (1997) Cyclic AMP-mediated enhancement of high-affinity choline transport and acetylcholine synthesis in brain. J. Neurochem. 68, 1062–1070. Rocha, E. S., Swanson, K. L., Aracava, Y., Goolsby, J. E., Maelicke, A., and Albuquerque, E. X. (1996a). Paraoxon: cholinesterase-independent stimulation of transmitter release and selective block of ligand-gated ion channels in cultured hippocampal neurons. J. Pharmacol. Exp. Ther. 278, 1175–1187. Rocha, E. S., Pereira, E. F. R., Swanson, K. L., and Albuquerque, E. X. (1996b) Novel molecular targets in the central nervous system for the actions of cholinesterase inhibitors: alterations of modulatory processes. Proceedings of the 1996 Medical Defense Bioscience Review III, pp. 1635–1643. Dam, K., Seidler, F. J., and Slotkin, T. A. (1999) Chlorpyrifos releases norepinephrine from adult and neonatal rat brain synaptosomes. Dev. Brain Res. 118, 129–133. Whitney, K. D., Seidler, F. J., and Slotkin, T. A. (1995) Developmental neurotoxicity of chlorpyrifos: cellular mechanisms. Toxicol. Appl. Pharmacol. 134, 53–62. Campbell, C. G., Seidler, F. J., and Slotkin, T. A. (1997) Chlorpyrifos interferes with cell development in rat brain regions. Brain Res. Bull. 43, 179–189. Johnson, D. E., Seidler, F. J., and Slotkin, T. A. (1998) Early biochemical detection of delayed neurotoxicity resulting from developmental exposure to chlorpyrifos. Brain Res. Bull. 45, 143–147. Dam, K., Seidler, F. J., and Slotkin, T. A. (1998) Developmental neurotoxicity of chlorpyrifos: delayed targeting of DNA synthesis after repeated administration. Dev. Brain Res. 108, 39–45. Song, X., Seidler, F. J., Saleh, J. L., Zhang, J., Padilla, S., and Slotkin, T. A. (1997) Cellular mechanisms for developmental toxicity of chlorpyrifos: targeting the adenylyl cyclase signaling cascade. Toxicol. Appl. Pharmacol. 145, 158–174. Roy, T. S., Andrews, J. E., Seidler, F. J., and Slotkin, T. A. (1998) Chlorpyrifos elicits mitotic abnormalities and apoptosis in neuroepithelium of cultured rat embryos. Teratology 58, 62–68. Robertson, R. T. and Yu, J. (1993) Acetylcholinesterase and neural development: new tricks for an old dog? News Physiol. Sci. 8, 266–272. Layer, P. G. and Willbold, E. (1995) Novel functions of cholinesterases in development, physiology and disease. Prog. Histochem. Cytochem. 29, 1–94. Small, D. H., Michaelson, S., and Sberna, G. (1996) Non-classical actions of cholinest-

42

72.

73. 74. 75.

76.

77.

78.

79. 80.

81. 82.

83.

84.

85. 86.

87.

88. 89.

Pope and Liu erases: role in cellular differentiation, tumorigenesis and Alzheimer’s disease. Neurochem. Int. 28, 453–483. Kostovic, I. and Goldman-Rakic, P. S. (1983) Transient cholinesterase staining in the mediodorsal nucleus of the thalamus and its connections in the developing human and monkey brain. J. Comp. Neurol. 219, 431–447. Kristt, D. A. (1983) Acetylcholinesterase in the ventral thalamus: transience and patterning during ontogenesis. Neuroscience 10, 923–939. Layer, P. G. (1990) Cholinesterases preceeding major tracts in vertebrate neurogenesis. BioEssays 12, 415–420. Robertson, R. T., Mostamand, F., Kageyama, G. H., Gallardo, K. A., and Yu, J. (1991) Primary auditory cortex in the rat: transient expression of acetylcholinesterase activity in developing geniculocortical projections. Brain Res. Dev. Brain Res. 58, 81–95. Dupree, J.I. and Bigbee, J.W. (1994) Retardation of neuritic outgrowth and cytoskeletal changes accompany acetylcholinesterase inhibitor treatment in cultured rat dorsal root ganglion neurons. J. Neurosci. Res. 39, 567–575. Layer, P. G., Weikert, T., and Alber, R. (1993) Cholinesterases regulate neurite growth of chick nerve cells in vitro by means of a non-enzymatic mechanism. Cell. Tissue Res. 273, 219–226. Sternfeld, M., Ming, G-L., Song, H-J., Sela, H, Timberg, R., Poo, M-M., and Soreq, H. (1998). Acetylcholinesterase enhances neurite growth and synapse development through alternative contributions of its hydrolytic capacity, core protein and variable C termini. J. Neurosci. 18, 1240–1249. Saito, S. (1998) Cholinesterase inhibitors induce growth cone collapse and inhibit neurite extension in primary cultured chick neurons. Neurotoxicol. Teratol. 20, 411–419. Henschler, D., Schmuck, G., van Aerssen, M., and Schiffmann, D. (1992) The inhibitory effect of neuropathic organophosphate esters on neurite outgrowth in cell cultures: a basis for screening for delayed neurotoxicity. Toxicol. Vitro 6, 327–325.84. Flaskos, J., McLean, W. G., and Hargreaves, A. J. (1994) The toxicity of organophosphate compounds toward cultured PC12 cells. Toxicol. Lett. 70, 71–76. Li, W. and Casida, J. E. (1998) Organophosphorus neuropathy target esterase inhibitors selectively block outgrowth of neurite-like and cell processes in cultured cells. Toxicol. Lett. 98, 139–146. Song, X., Violin, J. D., Seidler, F. J., and Slotkin. T. A. (1998) Modeling the developmental neurotoxicity of chlorpyrifos in vitro: macromolecular synthesis in PC12 cells. Toxicol. Appl. Pharmacol. 151, 182–191. Das, K. P. and Barone, S. Jr. (1999) Neuronal differentiation in PC12 cells is inhibited by chlorpyrifos and its metabolites: is acetylcholinesterase inhibition the site of action? Toxicol. Appl. Pharmacol. 160, 217–230. Appleyard, M. E. (1992) Secreted acetylcholinesterase: non-classical aspects of a classical enzyme. Trends Neurol. Sci. 15, 485–490. Greenfield, S. A. (1991) A noncholinergic action of acetylcholinesterase (AChE) in the brain: from neuronal secretion to the generation of movement. Cell Mol. Neurobiol. 11, 55–77. Greenfield, S. A., Chubb, I. W., Grunewald, R. A., Henderson, Z., May, J., Portnoy, S.,et al. (1984) A non-cholinergic function for acetylcholinesterase in the substantia nigra: behavioural evidence. Exp. Brain Res. 54, 513–520. Appleyard, M. E., Vercher, J. L., and Greenfield, S. A. (1988) Release of acetylcholinesterase from the guinea-pig cerebellum in vivo. Neuroscience 25, 133–138. Webb, C. P. and Greenfield, S. A. (1992) Non-cholinergic effects of acetylcholinesterase in the substantia nigra: possible role for an ATP-sensitive potassium channel. Exp. Brain Res. 89, 49–58.


43

90. Appleyard, M. and Jahnsen, H. (1992) Actions of acetylcholinesterase in the guinea-pig cerebellar cortex in vitro. Neuroscience 47, 291–301. 91. Webb, C. P., Nedergaard, S., Giles, K., and Greenfield, S. A. (1996) Involvement of the NMDA receptor in a non-cholinergic action of acetylcholinesterase in guinea pig subtantia nigra pars compacta neurons. Eur. J. Neurosci. 8, 837–841. 92. Holmes, C., Jones, S. A., Budd, T. C., and Greenfield. (1997) Non-cholinergic, trophic action of recombinant acetylcholinesterase on mid-brain dopaminergic neurons. J. Neurosci. Res. 49, 207–218.

44

Pope and Liu

Agents Affecting Sodium Channels

45

I Pesticides B. Pesticides that Target Ion Channels

46

Soderlund


47

4 Agents Affecting Sodium Channels David M. Soderlund

1. INTRODUCTION Voltage sensitive sodium channels are widely recognized as the site of action for two classes of insecticides, the diphenylethanes (e.g., DDT and analogs) and the pyrethroids. Diphenylethane insecticides have not been used in significant quantity in the United States for more than two decades, but pyrethroids are one of the most important and widely used classes of insecticides. This chapter considers the actions of pyrethroids on voltage sensitive sodium channels and other putative sites of action and assesses the significance of those effects in determining the toxic actions of these compounds. Pyrethroid insecticides are synthetic analogs of the pyrethrin I (Fig. 1), one of the six structurally related insecticidal constituents of pyrethrum extract, a natural insecticide that has been used for more than 200 years (1). The discovery and development of synthetic pyrethroid insecticides has been the subject of numerous reviews. The summary that follows is based primarily on a recent comprehensive review by Elliott (2) as well as on earlier sources cited therein. The principal drawback of the pyrethrins as insecticides is their instability in light and air, which limits their effectiveness in crop protection and other insect control contexts in which residual activity is essential. The development of synthetic pyrethroids is the result of efforts to modify the structure of the natural pyrethrins in order to increase photostability while retaining the potent and rapid insecticidal activity and relatively low acute mammalian toxicity of the pyrethrins. Most synthetic analogs were discovered by the sequential replacement of structural elements of the pyrethrins with novel structural moieties selected to conserve the molecular shape and electronic properties of the template structure. Both the historical development of pyrethroids as an insecticide class and the structural diversity of pyrethroids in current use are illustrated by the compounds shown in Fig. 1. Allethrin, one of the earliest of synthetic pyrethroids still in current use, represents one initial synthetic approach, the replacement of the pentadienyl side chain of pyrethrin I with a simpler, synthetically more accessible moiety having similar steric

From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

47

48

Soderlund

Fig. 1. Structures of pyrethrin I and five representative synthetic pyrethroids. For compounds used as isomer mixtures, the most toxic isomer is shown.

and electronic properties. The next significant step in pyrethroid development involved the replacement of the cyclopentenolone ring of the pyrethrin and allethrin alcohols with an alternative unsaturated heterocyclic moiety, resulting in resmethrin. This compound not only exhibited increased photostability but also was substantially more potent as an insecticide and lower in acute mammalian toxicity than pyrethrin I. The combination of these desirable properties in a single molecule provided a strong impetus to search for new compounds with greater activity and photostability. Permethrin proved to be the first synthetic pyrethroid with sufficient photostability for agricultural use. When compared to resmethrin, this compound contains structural replacements in both the alcohol moiety (3-phenoxybenzyl for 5-benzyl-3-furylmethyl) and the acid moiety (chlorines for methyl groups) that confer enhanced photostability without loss of insecticidal activity. Inclusion of an α-cyano substituent in the 3-phenoxybenzyl alcohol moiety, as in deltamethrin, produced compounds with much greater insecticidal potency than permethrin but with similar photostability. Synthetic pyrethroids related in structure to permethrin and deltamethrin constitute the largest chemical subfamily of pyrethroids in current use. The structural diversity of synthetic pyrethroids was further enhanced by the discovery that the 2,2-dimethylcyclopropanecarboxylic acid moiety of the pyrethrins and most synthetic compounds could be replaced by an α-isopropylphenylacetic acid moiety. This new series of compounds led to the discovery of the commercial insecticide fenvalerate. The first synthetic pyrethroids with sufficient environmental stability for the control of agricultural pests were introduced as commercial products in the late 1970s. Since


49

that time, pyrethroid use has grown to represent approx 25% of the dollar value of the world insecticide market, placing these compounds second in market value only to organophosphate insecticides (3). Pyrethroids are used not only in agricultural pest control but also in the control of household pests and vectors of animal and human disease. 2. PYRETHROID TOXICITY 2.1. Animal Studies 2.1.1. Acute Toxicity Pyrethroids are widely perceived as being safe insecticides, especially when compared to organophosphate and methylcarbamate insecticides. When administered orally in vegetable oils, most pyrethroids are moderately toxic (EPA Category II), with LD50 values in rats ranging from 50–500 mg/kg (4). However, many pyrethroids are very toxic (LD50 values 8600 >4300 >860 11 5.4b 12 3.9 1.6 0.6 0.5b, 1.2 >860b

aData bData

from Lawrence and Casida (7) except where noted. from Ghiasuddin and Soderlund (8), recalculated on the basis of µg/g brain weight.

mammalian toxicity, this effect is highly stereospecific: α-R epimers of compounds that retain the appropriate configurations for high toxicity in the acid moiety have no demonstrable toxicity when injected directly into the CNS (e.g., NRDC 156B, Table 1). The α-cyano substituent also indirectly alters structure–toxicity relationships in the acid moiety. The most dramatic effects are seen with the 1R,trans cyclopropanecarboxylates of 3-phenoxybenzyl alcohol (e.g., NRDC 163; Table 1), which exhibit extremely low toxicity to mammals; addition of an α-cyano substituent in the S configuration to these esters produces compounds (e.g., NRDC 158; Table 1) with significant neurotoxicity to rodents. 2.1.3. Two Syndromes of Intoxication Two studies published in the early 1970s identified two distinct syndromes associated with the acute toxicity of pyrethroids to rats. Verschoyle and Barnes (9) provided the first systematic description of the signs of pyrethroid intoxication in rats following oral and intravenous dosing. These authors noted the same syndrome of intoxication for pyrethrins, bioallethrin, resmethrin, and NRDC 108 (an analog of resmethrin) by either route of administration. This syndrome included hypersensitivity and aggression followed by stimulus induced bouts of general tremor, convulsive twitching, coma,

52

Soderlund

and death. The principal difference observed between oral and intravenous dosing was the speed of onset of intoxication. The publication of the discovery of deltamethrin (10), the first pyrethroid containing the α-cyano-3-phenoxybenzyl moiety, was coupled with a brief report describing the acute toxicity of deltamethrin to rats (11). This report noted that the signs of deltamethrin intoxication following either oral or intravenous administration, which involved salivation without lachrymation followed by jerking leg movements and progressive writhing convulsions (choreoathetosis), were distinctly different than those reported for other pyrethroids by these authors (9). A subsequent landmark study (6) described both the acute toxicity and the signs of intoxication of 36 pyrethroids following intravenous administration to rats, thereby establishing a taxonomy of pyrethroid intoxication in mammals that persists to the present. Of the 18 esters of various primary alcohols examined in this study, the 14 compounds with measurable toxicity all produced signs of intoxication corresponding to those first described for pyrethrins and pyrethroids (9), which was designated the T (tremor) syndrome. Of the 17 esters of α-cyano-3-phenoxybenzyl alcohol examined, 12 produced signs of intoxication like those first described for deltamethrin (11), which was designated the CS (choreoathetosis with salivation) syndrome, whereas 4 produced the T syndrome of intoxication. One α-cyano-3-phenoxybenzyl ester and one compound in which the α-cyano group was replaced by an α-ethynyl group were found to produce elements of both syndromes (tremor with salivation, designated TS). This classification of the signs of pyrethroid intoxication into two principal syndromes was subsequently confirmed in studies of the intracerebral toxicity of 29 pyrethroids to mice (7). An alternative nomenclature (Type I and Type II) has also been proposed for subgroups of pyrethroids based not only on the syndromes of intoxication produced in mammals (7) but also on their chemical structures, their signs of poisoning in insects, and their actions on insect nerve preparations (12). The Type I/Type II nomenclature has been widely adopted in the literature and is often used in a manner parallel to the T/CS nomenclature, so that Type I compounds are generally considered to produce the T syndrome of intoxication and Type II compounds are considered to produce the CS syndrome (7). 2.2. Toxicity to Humans 2.2.1. Acute Toxicity The principal source of information on human intoxication by pyrethroids is a comprehensive review of 573 cases of acute pyrethroid poisoning that were reported in the Chinese medical literature during the period 1983–1988 (13). These cases encompassed both occupational exposure as the result of mishandling during agricultural uses (229 cases) and accidental exposure, usually by ingestion of formulated insecticide products (344 cases). All but seven of these cases involved three pyrethroids: deltamethrin (325 cases), fenvalerate (196 cases), and cypermethrin (45 cases). The most common signs of systemic pyrethroid poisoning included dizziness, headache, nausea, anorexia, and fatigue, whereas more serious cases exhibited coarse muscle fasciculations, disturbance of consciousness, coma, and convulsive attacks. Seven deaths were reported, of which


53

one was owing to atropine intoxication following misdiagnosis of organophosphate insecticide poisoning. All patients received symptomatic and supportive therapies and most recovered within 6 d of exposure, although recovery of patients experiencing convulsions required up to 55 d. Subsequent reports of isolated cases of pyrethroid exposure and intoxication (14–16) are in substantial agreement with the findings reported in the Chinese literature (13). 2.2.2. Peripheral Effects Following Dermal Exposure The experience of workers involved in the handling of technical or formulated pyrethroids, either during manufacture or use, provides insight into the effects of pyrethroids on humans following cutaneous exposure. A large body of occupational exposure data, encompassing published clinical reports and unpublished reports obtained from industrial sources, has been assembled in two review articles (17,18). The most frequently reported symptom in worker exposure studies was paresthesia, which was characterized by numbness, itching, burning, or tingling of the skin following dermal exposure to a pyrethroid. These sensations generally occurred in the absence of erythema, edema, vesiculation, or other signs of overt skin irritation and were usually limited to the directly exposed areas of the skin. Pyrethroid induced paresthesia was transient and reversible within hours after exposure, but in some instances it lasted for up to 48 h. In studies of workers who experienced cutaneous sensations following pyrethroid exposure, no clinical signs of acute pyrethroid intoxication were observed (17,18). In addition, no exposure related differences were detected in hematology parameters or in heart, lung, liver, kidney, or nervous system function in these individuals. Further, electrophysiological assessment of peripheral nerve function did not detect any abnormalities related to occupational exposure to pyrethroids. Reports of occupational exposure are supported and amplified by the results of studies with human volunteers and experimental animals. In human volunteers, the effects of four pyrethroids were evaluated by application to the earlobe (19). With each compound, paresthesia developed within 30 min of exposure, peaked by 8 h, and dissipated by 24–32 h after exposure. In studies with guinea pigs, the onset of abnormal sensation caused by six pyrethroids was judged by an increase in scratching, licking, or biting behavior at the site of dermal application (20). The onset of symptoms usually occurred within 1 h after application of each of the six pyrethroids. The latency and duration of the behavioral response were affected by the formulation employed for any single pyrethroid, but the magnitude of the response was independent of formulation. The results of worker exposure data coupled with the results of controlled experiments with human volunteers and animals show that paresthesia appears to be an exclusively local effect. It occurs only at the site of dermal exposure, is not correlated with the appearance of a rash or other signs of classical skin irritation, and is not associated with any signs of systemic intoxication. Based on these observations and the proposed origins of paresthesia in the sensory nervous system (21), pyrethroid induced paresthesia has been postulated to be a direct excitatory effect of pyrethroids on small sensory nerve fibers in the skin rather than a response due to classical skin irritation.

54

Soderlund

3. MECHANISM OF ACTION 3.1. Effects of Pyrethroids on Sodium Channels 3.1.1. Electrophysiological Studies The actions of pyrethroid insecticides on sodium channels in invertebrate and vertebrate nerve preparations have been widely documented over the past four decades and are extensively and critically summarized in numerous reviews (17,18,22–27). Briefly, intracellular microelectrode recordings of action potentials demonstrate that pyrethroids produce a range of effects on nerve excitability depending on the structure of the pyrethroid employed (28,29). Natural pyrethrins and a structurally heterogeneous group of synthetic pyrethroids lacking the α-cyano-3-phenoxybenzyl alcohol moiety are characterized by the induction of long trains of action potentials (“burst discharges”) following a single stimulus with little or no effect on the nerve membrane resting potential. In contrast, pyrethroids that contain the α-cyano-3-phenoxybenzyl alcohol moiety typically do not produce repetitive firing but instead block the action potential upon repeated nerve stimulation and depolarize the resting-membrane potential. Examination of a wide variety of pyrethroid structures has also identified compounds with intermediate effects on neuronal excitability; these compounds typically produce bursts of action potentials of declining amplitude followed, after repetitive stimulation, by nerve block. Effects of pyrethroids on sodium channel function that underlie these effects on nerve excitability have been elucidated using voltage clamp and patch clamp techniques. Under voltage clamp, all neuroactive pyrethroids prolong the deactivation of sodium channels, which is evident as the production of a slowly decaying sodium tail current that flows following a depolarization–repolarization cycle (23,27). In most preparations, pyrethroids also retard the closing (inactivation) of sodium channels during a depolarizing pulse. Voltage clamp experiments show that the apparently divergent effects of pyrethroids on nerve excitability observed in intracellular microelectrode recordings are the result of similar effects on sodium channels (28,29). Compounds that produce burst discharges in intact nerves produce tail currents that decay rapidly, whereas compounds that produce use dependent block of action potentials produce extremely persistent tail currents that exhibit little or no decay for several seconds after repolarization and persist for several minutes. The differences in tail current kinetics between these groups compounds are conceptually consistent with their different effects on evoked action potentials: Transient prolongation of sodium channel inactivation and deactivation results in a depolarizing afterpotential and repetitive firing, whereas persistent prolongation of sodium inactivation and deactivation produces slow depolarization of the cell membrane and concomitant block of the action potential. It is also noteworthy that pyrethroids identified as having intermediate behavior in intracellular recordings produce tail currents under voltage clamp conditions with intermediate decay kinetics. Patch clamp studies of the action of pyrethroids on single sodium channels are limited to a much smaller group of compounds. Tetramethrin, a compound known to produce burst discharges in intact nerves and transient sodium current prolongation in voltage clamp experiments, increased the mean open time of individual sodium channels in patch clamp experiments approx 10-fold (30). In contrast, deltamethrin and fenvalerate, which are known to produce use dependent block of intact nerves and


55

persistent sodium current prolongation in voltage clamp experiments, increased the mean open times of individual sodium channels in patch clamp studies up to 200-fold and produced channels that remained open at the end of depolarizing pulses (31,32). Recordings of single sodium channels revealed that deltamethrin also delayed channel opening in response to a depolarizing pulse (31). These findings have been interpreted as evidence that pyrethroids stabilize multiple sodium channel states and slow the transitions between states. 3.1.2. Biochemical Studies Biochemical studies of the actions of pyrethroids on sodium channels have provided important information on the effects of these insecticides on channels in the mammalian CNS and have elucidated the relationship between the pyrethroid binding site and other identified binding domains of the sodium channel. Two experimental approaches have been employed: studies of the effects of pyrethroids and other sodium channeldirected toxins on radiosodium uptake into synaptic vesicles, and studies of the effects of pyrethroids on the binding of radioligands that label the sodium channel. Pyrethroids alone do not affect radiosodium uptake into brain synaptosomes, but they allosterically enhance sodium uptake that is stimulated by veratridine or batrachotoxin (8,33,34). Pyrethroids also allosterically enhance the binding of [3H]batrachotoxinin A-20-α-benzoate (BTX-B), an analog of batrachotoxin (35,36). Structure–activity relationships for the action of pyrethroids on rat and mouse brain sodium channels in sodium uptake and radioligand binding assays are in general agreement with structure–toxicity relationships. The enhancement of both veratridinedependent sodium uptake into brain synaptosomes and BTX-B binding to brain sodium channels is stereospecific for deltamethrin and the neurotoxic isomers of cypermethrin and fenvalerate (8,35,37). However, structural analogs of deltamethrin and cypermethrin lacking the α-cyano substituent were much less effective in these assays. The results of these studies implied the existence of a pyrethroid binding site on the sodium channel, designated Site 6 by Lombet et al. (36), that is distinct from the sites labeled by other radioligands. Initial attempts to label this site with a pyrethroid radioligand were unsuccessful (36,38) because of the extreme lipophilicity of potent pyrethroid ligands. More recent studies using a potent experimental pyrethroid as a radioligand have demonstrated high affinity saturable binding to brain sodium channels that exhibits the allosteric coupling to other binding domains predicted by previous studies, a result consistent with the existence of Site 6 as a pharmacologically distinct binding domain (39). However, the utility of this ligand for the detailed characterization of pyrethroid binding is still limited by its extreme lipophilicity and the high levels of nonspecific binding that are encountered. 3.1.3. Effects on Individual Sodium Channel Isoforms Most of what is known about the actions of pyrethroids on mammalian sodium channels has been learned using neuronal tissue preparations, which are now known to express multiple sodium channel α subunit isoforms. As a result, the action of pyrethroids has not been correlated with the expression of identified sodium channel isoforms in these tissues. However, a limited number of physiological studies suggest that sodium channel isoforms expressed in various mammalian tissues exhibit differential sensitivity to pyrethroids. Deltamethrin, but not cismethrin or permethrin, had a

56

Soderlund

direct positive inotropic effect on the mammalian heart in addition to an indirect effect mediated by the stimulation of catecholamine release (40–42). Five α-cyano pyrethroids that produce the CS syndrome of poisoning in the rat also caused repetitive action potentials in directly stimulated mammalian skeletal muscle but do not cause repetitive discharges evoked by sensory stimulation in the trigeminal reflex pathway (43–45). In contrast, three compounds that produce the T syndrome were inactive on skeletal muscle, but were active on sensory and motor neuronal elements of the rattrigeminal reflex pathway (43,45). Two pyrethroids classified as intermediate with respect to signs of intoxication were active on both nerve and muscle in these assays (45). These findings imply that pyrethroids affect multiple sodium channel isoforms, including those not expressed in the mammalian CNS, and that these isoforms may vary in their relative sensitivity to different pyrethroids. The clearest evidence of differential sensitivity to pyrethroids between sodium channel isoforms is found in the responses of the tetrodotoxin (TTX)-sensitive and TTX-resistant sodium channel populations in dorsal root ganglion cells to pyrethroids. The TTX-resistant current in these cells is much more sensitive than the TTX-sensitive current to allethrin (46), tetramethrin (47,48), and deltamethrin (49). It is now possible to examine the functional and pharmacological properties of individual cloned sodium channel isoforms by expressing them in unfertilized oocytes of the frog Xenopus laevis, which efficiently synthesizes sodium channel proteins from cloned cDNA injected into the nucleus or synthetic messenger RNA injected into the cytoplasm and inserts functional channels into the cell membrane (50). Initial experiments using cloned sodium channel isoforms in the Xenopus oocyte expression system examined the sensitivity of the rat brain IIa sodium channel isoform (51), which is abundantly expressed in the adult brain. Expression of the rat brain IIa sodium channel α subunit, either alone or in combination with the rat β1 subunit, produced functional sodium channels in oocytes that were sensitive to modification by [1R,cis,αS]cypermethrin (52). These studies demonstrated that the pyrethroid binding site was intrinsic to the α subunit. However, coexpression with the β1 subunit increased the apparent affinity of rat brain IIa channels to pyrethroids more than 20-fold, thus implying an allosteric effect of coassembly with the β1 subunit on the pyrethroid binding site. Even in the presence of the β1 subunit, rat brain IIa channels exhibited only modest sensitivity to cypermethrin and deltamethrin. The effects of deltamethrin and related compounds on rat brain IIa sodium channels exhibited the stereospecificity predicted by structure–toxicity relationships. Only deltamethrin itself and its 1R,trans,αS isomer, which is neurotoxic to rats (6), were effective in modifying currents carried by rat brain IIa sodium channels, whereas other isomers having the 1S or αR configurations, which have very low acute toxicities, were inactive at the highest concentrations attainable in this assay system (53). Rat brain IIa channels expressed in oocytes were completely insensitive to several pyrethroids that produce the T syndrome at the maximum nominal concentrations of pyrethroid that were attainable in this assay system (52). The complete insensitivity of rat brain IIa channels to these compounds is surprising in view of their well established neurotoxic effects observed following direct injection into the brain (4). It is therefore likely that the central neurotoxic effects of these pyrethroids, as well as other compounds that produce the T syndrome, are mediated by actions on one or more mole-


57

cular targets that are expressed in the CNS other than the brain IIa sodium channel isoform. The actions of pyrethroids on the rat-brain IIa channel isoform were compared to those on the rat SNS/PN3 sodium channel isoform, which is preferentially expressed in the peripheral nervous system (PNS) and is distinguished by its high level of resistance to TTX (54,55). In contrast to rat brain IIa sodium channels, rat SNS/PN3 sodium channels were highly sensitive to pyrethroids producing both the T (cismethrin) and CS (cypermethrin) syndromes of intoxication (56). The threshold cypermethrin concentration for the modification of SNS/PN3 channels was approx 60-fold lower than the threshold concentration for cypermethrin-dependent modification of rat brain IIa channels. The biophysical properties, TTX resistance, and pyrethroid sensitivity of the SNS/PN3 sodium channel isoform expressed in oocytes suggest that this isoform carries the TTX-resistant, pyrethroid-sensitive sodium current found in dorsal root ganglion neurons (46,47,49). 3.2. Effects of Pyrethroids on Other Targets 3.2.1. Effects on GABA Receptors The first report of an action of pyrethroids on GABA receptors described the stereospecific inhibition by deltamethrin but not its nontoxic α-R epimer of the binding of [3H]dihydropicrotoxinin to the convulsant (chloride channel) site of rat brain GABA receptors (57). The subsequent development of [35S]TBPS as an improved radioligand for the chloride channel site of GABA receptors led to the further characterization of the interaction of pyrethroids with the convulsant/chloride channel site (58,59). These studies, employing 37 pyrethroids, documented the inhibition of [35S]TBPS binding by the toxic isomers of four pyrethroids containing the α-cyano-3-phenoxybenyl moiety and by isomer mixtures of two other α-cyano compounds but not by the nontoxic isomers of α-cyano compounds or by any pyrethroids lacking the α-cyano substituent. The structure–activity correlations for pyrethroid-dependent inhibition of [35S]TBPS binding led to the widely-recognized hypothesis that α-cyano pyrethroids caused the CS intoxication syndrome by an action at the GABA receptor–ionophore complex (58). Further analysis of the action of pyrethroids on GABA receptors was undertaken using assays of GABA-receptor function. Several studies of the effects of pyrethroids on GABA-stimulated chloride-36 uptake into brain vesicles confirmed the action of neurotoxic isomers of α-cyano pyrethroids as antagonists at mammalian brain GABA receptors (60–63). However, the inhibition of chloride uptake in these studies was typically incomplete at maximally effective pyrethroid concentrations. Also, the incomplete stereoselectivity of pyrethroid action on GABA receptors in these assays was inconsistent with the profound stereospecificity of pyrethroid intoxication. Electrophysiological assays have been employed to assess the relative sensitivity of GABA receptors and voltage-sensitive sodium channels expressed in the same cell or neuronal pathway to pyrethroids. GABA receptors in cultured dorsal root ganglion neurons were much less sensitive to the actions of deltamethrin than were the populations of voltage sensitive sodium channels expressed in the same cells (64). Also, electrophysiological recordings from defined GABAergic pathways in the rat hippocampus (65–68) showed that the effects of pyrethroids were consistent with an augmentation of inhibition, resulting from sodium channel-mediated presynaptic excitation of

58

Soderlund

GABAergic neurons, rather than the antagonism of inhibition that would be expected of a blocker of the GABA-gated chloride channel. Thus, the results of functional assays imply that GABA receptors do not constitute a primary site of action for pyrethroids. 3.2.2. Effects on Voltage Sensitive Chloride Channels Studies of the action of cismethrin and deltamethrin on skeletal muscle showed that deltamethrin but not cismethrin increased muscle membrane resistance, which was suggested to result from a block of the chloride permeability of the muscle membrane (44). A subsequent study (69) confirmed the effect of deltamethrin but not cismethrin on the input resistance of rat diaphragm skeletal muscle fibers and showed that a reduction in extracellular chloride ion concentration prevented the effects of deltamethrin. This study also documented similar effects of deltamethrin on the input resistance of ratvagus nerve preparations that were prevented by low extracellular chloride or treatment of the preparation with ivermectin, which activates neuronal voltage-dependent chloride channels (70). Patch clamp studies of single neuronal voltage-dependent chloride channels in excised membrane patches from N1E-115 neuroblastoma cells documented the blockade of single channel conductance by deltamethrin and cypermethrin, but not cismethrin (71,72). Studies in vivo of the interactions between pyrethroids and agents known to act at voltage-sensitive chloride channels provide further insight into the involvement of this target in pyrethroid intoxication. These experiments involved co-administration of deltamethrin with ivermectin (which is known to activate voltage sensitive chloride channels and has limited access to the CNS), pentobarbital (a barbiturate that selectively activates voltage sensitive chloride channels), or phenobarbital (which exhibits sedative effects typical of barbiturates without activating voltage sensitive chloride channels) (73,74). Intraperitoneal pretreatment of rats with ivermectin reduced the degree of salivation caused by subsequent intravenous treatment with deltamethrin and also reduced the incidence of deltamethrin mortality and the motor signs of deltamethrin intoxication at this dose, but ivermectin only affected salivation at a lower dose of deltamethrin (73,74). Ivermectin also reduced the severity of the direct effects of deltamethrin on skeletal muscle excitability in urethane-anesthetized rats (73,74). In parallel experiments, pentobarbital significantly antagonized the motor signs of intoxication and reduced the degree of mortality in deltamethrin-treated rats but was less effective than ivermectin in reducing salivation (73,74). In contrast, an equi-sedative dose of phenobarbital (which lacks the selectivity of pentobarbital for voltage sensitive chloride channels) reduced the number of deaths in deltamethrin-treated rats but did not significantly affect salivation or the motor signs of intoxication (73,74). These studies were also extended to examine the effects of barbiturates on intoxication by cismethrin, a pyrethroid without demonstrable effects on voltage sensitive chloride channels in vitro. Pentobarbital did not affect the either the motor signs or number of deaths in rats treated intravenously with cismethrin, but phenobarbital produced a significant reduction in lethality under these conditions (73). These authors concluded that the effects of ivermectin and pentobarbital on the signs of deltamethrin intoxication reflected a specific antagonism of the action of deltamethrin on voltage sensitive chloride channels, whereas the effects of pentobarbital and phenobarbital on the lethality of both deltamethrin and cismethrin were ascribed to central effects on neuronal


59

excitability that were not specifically attributable to an action on chloride channels (73). For further discussion of the role of chloride channels in pesticide mode of action, see Chapter 5 by Bloomquist. 3.3. Antidotes for Pyrethroid Intoxication In light of the reversibility of sublethal pyrethroid intoxication and the lack of a specific pyrethroid antidote, treatment of pyrethroid poisoning has typically involved supportive therapy directed at the alleviation of symptoms. Experience in China (13) points to the value of gastric lavage to limit the absorbed dose of pyrethroid in cases of ingestion. This report also documented the use of atropine in 189 poisoning cases, which reduced pyrethroid-induced salivation in some cases but also led to atropineinduced intoxication and, in one case, lethal atropine poisoning. Muscle relaxants, anticonvulsants, and anesthetics have been evaluated in animal models of pyrethroid intoxication, but the results of such studies have not identified useful therapies for human intoxication (18). 4. CONCLUSIONS AND FUTURE DIRECTIONS Several lines of evidence implicate effects on sodium channels as a principal mode of toxic action of pyrethroids. First, the nature of pyrethroid action on sodium channels and the ensuing disruption of nerve function are mechanistically consistent with the excitatory signs of intoxication observed in vivo. Where the stereospecificity of action of sodium channels has been examined, it is in complete agreement with the stereospecificity of intoxication by pyrethroids. Moreover, a robust correlation exists between the duration of pyrethroid-modified sodium currents and acute toxicity that encompasses compounds producing both the T and CS syndromes of intoxication as well as those with intermediate effects (24). Finally, the high pyrethroid sensitivity of TTX-resistant sodium channels in peripheral neurons underlie the production of paresthesia by pyrethroids given the role ascribed to these channels in the production of anomalous sensory responses (21). Although the production of the T and CS syndromes is correlated with the relative duration of pyrethroid-modified sodium currents (24), it is possible that the differential sensitivity of individual sodium channel isoforms to pyrethroids may also contribute to the production of the two principal syndromes of intoxication. Additional research to characterize the actions of a range of pyrethroid structures representative of those producing the T, CS, and intermediate syndromes of intoxication on individual sodium channel isoforms may shed further light on the role of differential isoform sensitivity as a determinant of the signs of intoxication associated with the T and CS syndromes. Whereas pyrethroids have been shown to affect a variety of ion channels, receptors, and enzymes in in vitro assays, most of these actions are not well correlated with intoxication in vivo (22,24). One notable exception is the blockade of voltage sensitive chloride channels (see Subheading 3.2.2.), which appears to occur at toxicologically relevant concentrations and may underlie the production of some of the signs of intoxication associated with the CS syndrome. At present, observations of the action of pyrethroids on voltage sensitive chloride channels are limited to very few compounds; further research to determine structure–activity relationships for pyrethroids at this tar-

60

Soderlund

get would establish the broader significance of voltage sensitive chloride channels as secondary targets for the action of pyrethroids. ACKNOWLEDGMENTS Studies reviewed here from the author’s laboratory were supported by grant number R01 ES08962 from the National Institute of Environmental Health Sciences, National Institutes of Health. REFERENCES 1. Casida, J. E. and Quistad, G. B. (eds.) (1995) Pyrethrum Flowers: Production, Chemistry, Toxicology, and Uses. Oxford University Press, NY, p. 356. 2. Elliott, M. (1995) Chemicals in insect control, in Pyrethrum Flowers: Production, Chemistry, Toxicology, and Uses (Casida, J. E. and Quistad, G. B., eds.), Oxford University Press, NY, pp. 3–31. 3. Casida, J. E. and Quistad, G. B. (1998) Golden age of insecticide research: past, present, or future? Ann. Rev. Entomol. 43, 1–16. 4. Gray, A. J. and Soderlund, D. M. (1985) Mammalian toxicology of pyrethroids, in Insecticides (Hutson, D. H. and Roberts, T. R., eds.), Wiley, NY, pp. 193–248. 5. Elliott, M., Farnham, A. W., Janes, N. F., Needham, P. H., and Pulman, D. A. (1974) Insecticidally active conformations of pyrethroids, in Mechanism of Pesticide Action (Kohn, G. K., ed.), American Chemical Society, Washington, DC, pp. 80–91. 6. Verschoyle, R. D. and Aldridge, W. N. (1980) Structure-activity relationships of some pyrethroids in rats. Arch. Toxicol. 45, 325–329. 7. Lawrence, L. J. and Casida, J. E. (1982) Pyrethroid toxicology: mouse intracerebral structure-toxicity relationships. Pestic. Biochem. Physiol. 18, 9–14. 8. Ghiasuddin, S. M. and Soderlund, D. M. (1985) Pyrethroid insecticides: potent, stereospecific enhancers of mouse brain sodium channel activation. Pestic. Biochem. Physiol. 24, 200–206. 9. Verschoyle, R. D. and Barnes, J. M. (1972) Toxicity of natural and synthetic pyrethrins to rats. Pestic. Biochem. Physiol. 2, 308–311. 10. Elliott, M., Farnham, A. W., Janes, N. F., Needham, P. H., and Pulman, D. A. (1974) Synthetic insecticide with a new order of activity. Nature 248, 710–711. 11. Barnes, J. M. and Verschoyle, R. D. (1974) Toxicity of new pyrethroid insecticide. Nature 248, 711. 12. Gammon, D. W., Brown, M. A., and Casida, J. E. (1981) Two classes of pyrethroid action in the cockroach. Pestic. Biochem. Physiol. 15, 181–191. 13. He, F., Wang, S., Liu, L., Chen, S., Zhang, Z., and Sun, J. (1989) Clinical manifestations and diagnosis of acute pyrethroid poisoning. Arch. Toxicol. 63, 54–58. 14. Lessenger, J. E. (1992) Five office workers inadvertently exposed to cypermethrin. J. Toxicol. Environ. Health 35, 261–267. 15. Box, S. A. and Lee, M. R. (1996) A systemic reaction following exposure to a pyrethroid insecticide. Hum. Exp. Toxicol. 15, 389–390. 16. Gotoh, Y., Kawakami, M., Matsumoto, N., and Okada, Y. (1998) Permethrin emulsion ingestion: clinical manifestations and clearance of isomers. Clin. Toxicol. 36, 57–61. 17. Vijverberg, H. P. M. and van den Bercken, J. (1990) Neurotoxicological effects and the mode of action of pyrethroid insecticides. CRC Crit. Rev. Toxicol. 21, 105–126. 18. Clark, J. M. (1995) Effects and mechanisms of action of pyrethrin and pyrethroid insecticides, in Handbook of Neurotoxicology (Chang, L. W. and Dyer, R. S., eds.), Marcel Dekker, NY, pp. 511–546.


61

19. Flannigan, S. A. and Tucker, S. B. (1985) Variation in cutaneous sensation between synthetic pyrethroid insecticides. Contact Dermatitis 13, 140–147. 20. Cagen, S. Z., Malley, L. A., Parker, C. M., Gardiner, T. H., van Gelder, G. A., and Jud, V. A. (1984) Pyrethroid-mediated skin sensory stimulation characterized by a new behavioral paradigm. Toxicol. Appl. Pharmacol. 76, 270–279. 21. Rizzo, M. A., Kocsis, J. D., and Waxman, S. G. (1996) Mechanisms of paresthesiae, dysthesiae, and hyperesthesia: role of Na+ channel heterogeneity. Eur. Neurol. 36, 1–12. 22. Soderlund, D. M. and Bloomquist, J. R. (1989) Neurotoxic actions of pyrethroid insecticides. Ann. Rev. Entomol. 34, 77–96. 23. Narahashi, T. (1992) Nerve membrane Na+ channels as targets of insecticides. Trends Pharmacol. Sci. 13, 236–241. 24. Bloomquist, J. R. (1993) Neuroreceptor mechanisms in pyrethroid mode of action and resistance, in Reviews in Pesticide Toxicology (Roe, M. and Kuhr, R. J., eds.), Toxicology Communications, Raleigh, NC, pp. 181–226. 25. Soderlund, D. M. (1995) Mode of action of pyrethrins and pyrethroids, in Pyrethrum Flowers: Production, Chemistry, Toxicology, and Uses (Casida, J. E. and Quistad, G. B., eds.), Oxford University Press, NY, pp. 217–233. 26. Bloomquist, J. R. (1996) Ion channels as targets for insecticides. Ann. Rev. Entomol. 41, 163–190. 27. Narahashi, T. (1996) Neuronal ion channels as the target sites of insecticides. Pharmacol. Toxicol. 78, 1–14. 28. Lund, A. E. and Narahashi, T. (1983) Kinetics of sodium channel modification as the basis for the variation in the nerve membrane effects of pyrethroids and DDT analogs. Pestic. Biochem. Physiol. 20, 203–216. 29. Vijverberg, H. P. M., van der Zalm, J. M., van Kleef, R. G. D. M., and van den Bercken, J. (1983) Temperature—and structure-dependent interaction of pyrethroids with the sodium channels in frog node of Ranvier. Biochim. Biophys. Acta 728, 73–82. 30. Yamamoto, D., Quandt, F. N., and Narahashi, T. (1983) Modification of single sodium channels by tetramethrin. Brain Res. 274, 344–349. 31. Chinn, K. and Narahashi, T. (1986) Stabilization of sodium channel states by deltamethrin in mouse neuroblastoma cells. J. Physiol. 380, 191–207. 32. Holloway, S. F., Salgado, V. L., Wu, C. H., and Narahashi, T. (1989) Kinetic properties of single sodium channels modified by fenvalerate in mouse neuroblastoma cells. Pflugers Arch. 414, 613–621. 33. Soderlund, D. M., Bloomquist, J. R., Ghiasuddin, S. M., and Stuart, A. M. (1987) Enhancement of veratridine-dependent sodium channel activation by pyrethroids and DDT analogs, in Sites of Action for Neurotoxic Pesticides (Hollingworth, R. M. and Green, M. B., eds.), American Chemical Society, Washington, DC, pp. 251–261. 34. Bloomquist, J. R. and Soderlund, D. M. (1988) Pyrethroid insecticides and DDT modify alkaloid-dependent sodium channel activation and its enhancement by sea anemone toxin. Mol. Pharmacol. 33, 543–550. 35. Brown, G. B., Gaupp, J. E. and Olsen, R. W. (1988) Pyrethroid insecticides: stereospecific allosteric interaction with the batrachotoxinin-A benzoate binding site of mammalian voltage-sensitive sodium channels. Mol. Pharmacol. 34, 54–59. 36. Lombet, A., Mourre, C., and Lazdunski, M. (1988) Interactions of insecticides of the pyrethroid family with specific binding sites on the voltage-dependent sodium channel from mammalian brain. Brain Res. 459, 44–53. 37. Rubin, J. G., Payne, G. T., and Soderlund, D. M. (1993) Structure-activity relationships for pyrethroids and DDT analogs as modifiers of [3H]batrachotoxinin A 20-α-benzoate binding to mouse brain sodium channels. Pestic. Biochem. Physiol. 45, 130–140. 38. Soderlund, D. M., Ghiasuddin, S. M., and Helmuth, D. W. (1983) Receptor-like stereospecific binding of a pyrethroid insecticide to mouse brain membranes. Life Sci. 33, 261–267.

62

Soderlund

39. Trainer, V. L., McPhee, J. C., Boutelet-Bochan, H., Baker, C., Scheuer, T., Babin, D., et al. (1997) High affinity binding of pyrethroids to the α subunit of brain sodium channels. Mol. Pharmacol. 51, 651–657. 40. Forshaw, P. J. and Bradbury, J. E. (1983) Pharmacological effects of pyrethroids on the cardiovascular system of the rat. Eur. J. Pharmacol. 91, 207–213. 41. Berlin, J. R., Akera, T., Brody, T. M., and Matsumura, F. (1984) The inotropic effects of a synthetic pyrethroid decamethrin on isolated guinea pig atrial muscle. Eur. J. Pharmacol. 98, 313–322. 42. Daly, J. W., McNeal, E. T., and Gusovsky, F. (1987) Cardiotonic activities of pumiliotoxin B, pyrethroids and a phorbol ester and their relationships with phosphoinositide turnover. Biochim. Biophys. Acta 930, 470–474. 43. Forshaw, P. J. and Ray, D. E. (1986) The effects of two pyrethroids, cismethrin and deltamethrin, on skeletal muscle and the trigeminal reflex system in the rat. Pestic. Biochem. Physiol. 25, 143–151. 44. Forshaw, P. J., Lister, T., and Ray, D. E. (1987) The effects of two types of pyrethroid on rat skeletal muscle. Eur. J. Pharmacol. 134, 89–96. 45. Wright, C. D. P., Forshaw, P. J., and Ray, D. E. (1988) Classification of the actions of ten pyrethroid insecticides in the rat, using the trigeminal reflex and skeletal muscle as test systems. Pestic. Biochem. Physiol. 30, 79–86. 46. Ginsburg, K. S. and Narahashi, T. (1993) Differential sensitivity of tetrodotoxin-sensitive and tetrodotoxin-resistant sodium channels to the insecticide allethrin in rat dorsal root ganglion neurons. Brain Res. 627, 239–248. 47. Tatebayashi, H. and Narahashi, T. (1994) Differential mechanism of action of the pyrethroid tetramethrin on tetrodotoxin-sensitive and tetrodotoxin-resistant sodium channels. J. Pharmacol. Exp. Ther. 270, 595–603. 48. Song, J.-H. and Narahashi, T. (1996) Differential effects of the pyrethroid tetramethrin on tetrodotoxin-sensitive and tetrodotoxin-resistant single sodium channels. Brain Res. 712, 258–264. 49. Tabarean, I. V. and Narahashi, T. (1998) Potent modulation of tetrodotoxin-sensitive and tetrodotoxin-resistant sodium channels by the Type II pyrethroid deltamethrin. J. Pharmacol. Exp. Ther. 284, 958–965. 50. Lester, H. A. (1988) Heterologous expression of excitability proteins: route to more specific drugs? Science 241, 1057–1063. 51. Auld, V. J., Goldin, A. L., Krafte, D. S., Marshall, J., Dunn, J. M., Catterall, W. A., et al. (1988) A rat brain Na+ channel α subunit with novel gating properties. Neuron 1, 449–461. 52. Smith, T. J. and Soderlund, D. M. (1998) Action of the pyrethroid insecticide cypermethrin on rat brain IIa sodium channels expressed in Xenopus oocytes. NeuroToxicology 19, 823–832. 53. Smith, T. J. and Soderlund, D. M. (2000) Structure-activity relationships for the action of deltamethrin analogs on sodium channels expressed in Xenopus oocytes. Soc. Neurosci. Abst. 22, 60. 54. Akopian, A. N., Sivilotti, L., and Wood, J. N. (1996) A tetrodotoxin-resistant voltagegated sodium channel expressed in sensory neurons. Nature 379, 257–262. 55. Sangameswaran, L., Delagado, S. G., Fish, L. M., Koch, B. D., Jakeman, L. B., Stewart, G. R., et al. (1996) Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel specific to sensory neurons. J. Biol. Chem. 271, 5953–5956. 56. Smith, T. J. and Soderlund, D. M. (2000) Potent actions of the pyrethroid insecticides cismethrin and cypermethrin on rat tetrodotoxin-resistant peripheral nerve (SNS/PN3) sodium channels expressed in Xenopus oocytes. Pestic. Biochem. Physiol. 70, 52–61. 57. Leeb-Lundberg, F. and Olsen, R. W. (1980) Picrotoxinin binding as a probe of the GABA postsynaptic membrane receptor-ionophore complex, in Psychopharmacology and Bio-


58. 59.

60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.

74.

63

chemistry of Neurotransmitter Receptors (Yamamura, H. I., Olsen, R. W., and Usdin, E., eds.), Elsevier, NY, pp. 593–606. Lawrence, L. J. and Casida, J. E. (1983) Stereospecific action of pyrethroid insecticides on the γ-aminobutyric acid receptor-ionophore complex. Science 221, 1399–1401. Casida, J. E. and Lawrence, L. J. (1985) Structure-activity correlations for interactions of bicyclophosphorus esters and some polychlorocycloalkane and pyrethroid insecticides with the brain-specific t-butylbicyclophosphorothionate receptor. Environ. Health Perspec. 61, 123–132. Bloomquist, J. R. and Soderlund, D. M. (1985) Neurotoxic insecticides inhibit GABAdependent chloride uptake by mouse brain vesicles. Biochem. Biophys. Res. Commun. 133, 37–43. Bloomquist, J. R., Adams, P. M., and Soderlund, D. M. (1986) Inhibition of γ-aminobutyric acid-stimulated chloride flux in mouse brain vesicles by polychlorocycloalkane and pyrethroid insecticides. NeuroToxicology 7, 11–20. Abalis, I. M., Eldefrawi, M. E., and Eldefrawi, A. T. (1986) Effects of insecticides on GABA-induced chloride influx into rat brain microsacs. J. Toxicol. Environ. Health 18, 13–23. Ramadan, A. A., Bakry, N. M., Marei, A. S. M., Eldefrawi, A. T., and Eldefrawi, M. E. (1988) Action of pyrethroids on GABAA receptor function. Pestic. Biochem. Physiol. 32, 97–105. Ogata, N., Vogel, S. M., and Narahashi, T. (1988) Lindane but not deltamethrin blocks a component of GABA-activated chloride channels. FASEB J. 2, 2895–2900. Gilbert, M. E., Mack, C. M., and Crofton, K. M. (1989) Pyrethroids and enhanced inhibition in the hippocampus of the rat. Brain Res. 477, 314–321. Joy, R. M., Albertson, T. E., and Ray, D. E. (1989) Type I and Type II pyrethroids increase inhibition in the hippocampal dentate gyrus of the rat. Toxicol. Appl. Pharmacol. 98, 398–412. Joy, R. M., Lister, T., Ray, D. E., and Seville, M. P. (1990) Characteristics of the prolonged inhibition produced by a range of pyrethroids in the rat hippocampus. Toxicol. Appl. Pharmacol. 103, 528–538. Joy, R. M. and Albertson, T. E. (1991) Interactions of GABA A antagonists with deltamethrin, diazepam, pentobarbital, and SKF100330A in the rat dentate gyrus. Toxicol. Appl. Pharmacol. 109, 251–262. Forshaw, P. J. and Ray, D. E. (1990) A novel action of deltamethrin on membrane resistance in mammalian skeletal muscle and non-myelinated nerve fibers. Neuropharmacology 29, 75–81. Abalis, I. M., Eldefrawi, A. T., and Eldefrawi, M. E. (1986) Actions of avermectin B1a on the γ-aminobutyric acidA receptor and chloride channels in rat brain. J. Biochem. Toxicol. 1, 69–82. Forshaw, P. J., Lister, T., and Ray, D. E. (1993) Inhibition of a neuronal voltage-dependent chloride channel by the Type II pyrethroid deltamethrin. Neuropharmacology 32, 105–111. Ray, D. E., Sutharsan, S., and Forshaw, P. J. (1997) Action of pyrethroid insecticides on voltage-gated chloride channels in neuroblastoma cells. NeuroToxicology 18, 755–760. Ray, D. E., Lister, T., and Forshaw, P. J. (1999) A new basis for therapy against Type-II pyrethroid poisoning, in Progress in Neuropharmacology and Neurotoxicology of Pesticides and Drugs (Beadle, D. J., ed.), Royal Society of Chemistry, Cambridge, UK, pp. 204–214. Forshaw, P. J., Lister, T., and Ray, D. E. (2000) The role of voltage-gated chloride channels in type II pyrethroid insecticide poisoning. Toxicol. Appl. Pharmacol. 163,1–8.

64

Soderlund

Agents Affecting Chloride Channels

65

5 Agents Affecting Chloride Channels Jeffrey R. Bloomquist

1. INTRODUCTION The polychlorocycloalkanes (PCCAs), including the chlorinated cyclodienes, toxaphene, and lindane comprise the original group of insecticides acting on chloride channels. Although most of these compounds are no longer in use, they still have potential neurotoxicological impacts in humans and animals owing to their extreme chemical and environmental persistence (1). The newest member of convulsant insecticides targeting chloride channels is fipronil, a compound having improved properties and a chemical structure unrelated to PCCAs. Another class of chemistry affecting chloride channels is the avermectins, natural products with activity against insects, acarines, and parasitic nematodes (2). This review will focus on the neurotoxicity of the PCCAs, fipronil, and avermectins, as well as their interactions with the chloride channels of electrically excitable membranes. For discussion of the effects of pyrethroids on chloride channels, see Chapter 4 by Soderlund. 1.1. Chemistry of PCCAs and Fipronil A thorough review of the discovery and development of the PCCAs as insecticides is available from the treatise written by Brooks (3), along with a more recent review summarizing newer findings (4). The following is derived largely from these sources. The PCCA family of insecticides is a very old group. In 1825, Faraday described the chlorination of benzene in the presence of sunlight, which would be a technical preparation of hexachlorocyclohexane (HCH), an insecticidal material. The chemical composition of this mixture contains a number of isomers, with the γ-isomer (Fig. 1) being the most toxic and this isomer in its pure form (99%) comprises the commercial product, lindane. The bicyclic PCCAs (the cyclodienes) appeared during the late 1940s to mid-1950s. The original compound in this series was chlordene (Fig. 1), which served as a precursor for heptachlor and chlordane, the first commercial cyclodiene insecticides. Chlordene also served as the starting material for isobenzan (Fig. 1), one of the most toxic cyclodienes. The widely used compounds aldrin and dieldrin were produced by Diels-Alder reaction of hexachlorocyclopentadiene with norbornadiene, and the carbon skeleton of these compounds is folded into an endo, exo configuration. Other From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

65

66

Bloomquist

Fig. 1. Structures of PCCAs and related compounds discussed in the text. For the γ-isomer of HCH (lindane), a = axial and e = equatorial configurations of the chlorine atoms.

compounds in the cyclodiene series were isodrin and endrin, which are folded into an endo, endo configuration (Fig. 1). Chemists at Farbewerke-Hoechst AG produced the compound endosulfan (Fig. 1), which exists in two isomeric forms (α and β) because of the pyramidal nature of the sulfite ester group. Insecticidal materials structurally related to the cyclodienes are complex mixtures of chlorinated camphene (toxaphene, Fig. 1) or α-pinene (chlorothene and strobane). The technical mixture of compounds comprising toxaphene has five major components, including the component shown in Fig. 1 and its 3-exo-chloro, 8-chloro, 9-chloro, and 10-chloro derivatives (5). These compounds comprise about 23% of technical toxaphene (6). The most recently developed chloride channel-blocking compound employed as an insecticide is fipronil (Fig. 1), a substituted phenylpyrazole that was registered for worldwide use on a variety of crops by the mid-1990s (7). 1.2. Chemistry of the Avermectins The chemistry of the avermectins (8) and milbemycins (9) has been extensively reviewed. The avermectins and milbemycins are macrocyclic lactones produced by Streptomyces fungi. The primary avermectin molecule around which other compounds are typically defined is avermectin B1a (Fig. 2). The insecticide/miticide abamectin contains at least 80% of this material, along with other related compounds. Reduction of avermectin B1a at the 22, 23 double bond results in ivermectin (Fig. 2), a major anthelmintic. Newer compounds include semi-synthetic derivatives of avermectin B1a having essentially the same mode of action, but with different potencies against nema-


67

Fig. 2. Structures of avermectins and milbemycins referred to in the text.

todes, acarines, and insects. These compounds (Fig. 2) include the 4" substituted avermectins, emamectin (8) and eprinomectin (10). Doramectin (Fig. 2) is a cyclohexyl derivative of avermectin B1a, where the cyclohexyl group increases lipophilicity and the half-life within tissues (11). Other experimental avermectins include 4'-difluoro, 4"-difluoro, 13-difluoro, and 23-difluoro avermectins, which have anticonvulsant and anthelmintic activity (10). The major structural difference between the avermectins and the milbemycins is the lack of the oleandrose disaccharide moiety at the 13 position. Naturally occurring milbemycins include the A and D forms (Fig. 2), and important chemical substitutions of commercial milbemycins have included oxime derivatives, such as milbemycin A oxime (Fig. 2). 2. NEUROTOXICITY 2.1. Acute Toxicity in Animals The PCCA insecticides show considerable acute toxicity to mammals (reviewed in refs. 3,4,12,13). The approximate rank order of toxicity among mammalian species, from most to least susceptible is: dog, human, monkey, cat, guinea pig, rabbit, rat, hamster, and mouse. In rats, several compounds have oral LD50s ≤ 50 mg/kg, ranging up to about 600 mg/kg for chlordane. The rank order of oral toxicity in the rat is: isobenzan > isodrin = endrin > dieldrin = aldrin > heptachlor epoxide > α-endosulfan > lindane > heptachlor > β-endosulfan > chlordane. This hierarchy of toxicity indicates that epoxides are more toxic than the parent dienes, and therefore oxidation of the

68

Bloomquist

dienes, in vivo, constitutes a bioactivation step. Moreover, the endo, endo ring configuration increases toxicity in mammals, and the configuration of the alpha isomer of endosulfan results in greater toxicity than the beta isomer. High dermal toxicity is also expressed following exposure to cyclodienes, where the rat dermal LD50s of aldrin (98 mg/kg), dieldrin (75 mg/kg), endrin (15 mg/kg), for example, are within two fold of their oral LD50 values (14). In contrast, the rat oral and dermal LD50s for fipronil are 95 mg/kg, and >2000 mg/kg, respectively (7). Fipronil is oxidized, in vivo, to the corresponding sulfone, or photodegraded to the compound shown in Fig. 1, both of which are slightly more toxic to mammals than fipronil (15–17). Investigation of the insecticidal properties of HCH at Imperial Chemical Industries, Ltd. in the 1940s revealed that the gamma isomer possessed nearly all of the insecticidal activity (3). Although their wide therapeutic index allows them to be used for effective parasite control in animals and humans, the absolute toxicity of the avermectins is considerable. Oral or intraperitoneal administration of ivermectin to mice or rats gives LD50s of 25–30 mg/kg and 50–55 mg/kg, respectively (8). Similar toxicity values are observed for abamectin (8). Dermal toxicity of ivermectin is about 10-fold less in rats (LD50 >660 mg/kg), distinguishing it from the high dermal toxicity of the PCCAs. Replacement of the hydroxyl group in the 5 position (Fig. 2) with an oxime usually reduces potency or spectrum of activity in the avermectins and milbemycins. However, with this substitution, compounds such as milbemycin A oxime are better tolerated by dogs for nematode control, even in purebred collies that are especially sensitive to the neurotoxic action of these compounds (9). 2.2. Acute Signs of Intoxication and Gross Effects on Nerve and Muscle PCCAs cause convulsive signs of intoxication indicative of an action on the CNS. In rats treated with dieldrin, general neuronal excitability appears first, followed by exaggerated motor responses to sensory stimuli and violent paroxysmal convulsions (18). Similar signs of intoxication (tremors, convulsions, and seizures) are observed following treatment with fipronil (19). A central site of action for the cyclodienes is supported by the finding that intracerebral injection potentiates cyclodiene toxicity 7–33-fold when compared to intraperitoneal treatment (20). Surprisingly, lindane had virtually identical LD50s by either route of administration, suggesting both central and peripheral effects, the implications of which are discussed later in this review. A central action of dieldrin was confirmed in isolated frog spinal cord, where it augmented polysynaptic-reflex arcs (21). This effect was attributed to a reduction in postsynaptic inhibition, but at that time, no direct evidence was available to support this conclusion. Facilitated discharges in central nerve pathways were also observed in the visual and somatosensory cortex of cats treated with dieldrin (18) and in the limbic system of rats treated with lindane (22). In contrast to the uniform hyperexcitation caused the PCCAs, the avermectins elicit a more complicated poisoning syndrome. A just lethal dose of abamectin, given by intraperitoneal injection to mice causes hyperexcitability, incoordination, and tremor (23). These neuroexcitatory signs often give way to ataxia and coma-like sedation later in poisoning. Ivermectin (22,23-dihydro-avermectin B1a, Fig. 1) causes a similar syndrome in rodents, expressed primarily by ataxia and tremor (24). The idiosyncratic toxicity of avermectins to collie dogs is not due to changes in receptor binding or recep-


69

tor distribution, but has been suggested to arise from altered properties of the bloodbrain barrier (BBB), or to some unknown form of avermectin transport (25). In mice, intracerebral injection of abamectin causes a modest potentiation of toxicity ranging from 1.7- to 3.4-fold compared to intraperitoneal treatment (20). This finding is inconsistent with the brain being the primary site of action of avermectins in mammals, or that the BBB limits toxicity by reducing penetration into the brain (26). The ataxia and sedation observed in mammalian toxicity studies suggest that the overall effect of avermectins on the mammalian nervous system would be similar to the inhibitory effects on electrical activity of nerve and muscle cells observed in invertebrates (27). 2.3. Subchronic or Chronic Neurotoxicity in Animals Neurobehavioral and neurotoxicological effects of PCCAs have been reviewed by Hayes (28) and Ecobichon and Joy (13). For example, 2.5 mg/kg/d oral doses of dieldrin to rats for 12 wk resulted in decreased performance in a sound detection behavioral paradigm. Similarly, doses as low as 0.1 mg/kg/d decreased the rate of learning in monkeys given a task of visual successive nonspatial discrimination. Lifetime dietary exposures of 10 ppm caused irritability in rats, and levels at or above 20 ppm caused convulsions. Although a large number of such feeding studies have been done in the past, most evaluated carcinogenic or hepatotoxic effects and not alteration of the nervous system. Little chronic neurotoxicity data on avermectins is published in the peer-reviewed scientific literature. Subchronic exposure to daily oral treatments of emamectin (3.6 mg/kg/d) in pregnant rats had a number of effects in the offspring including: tremors, hindlimb splay, altered open field behavior, decreased postweaning weight gain, decreased auditory startle, and delayed developmental signs (29,30). However, no histological changes were observed in the CNS or PNS of drug-exposed pups or adults. The NOAEL in both reports was set at 0.6 mg/kg/d. 2.4. Poisoning Syndromes in Humans The symptoms of poisoning in humans following acute or chronic exposures to PCCAs has been thoroughly reviewed (13,28). The following discussion is taken from these excellent reviews. Most intense acute exposures occur following accidental or deliberate oral consumption. Human ingestion of lindane at approx 86 mg/kg caused malaise and dizziness, progressing to nausea, vomiting, and convulsions. Complete recovery occurred after 72 h. For cyclodienes, convulsions may be the first sign of intoxication following exposure, and bouts of convulsions may recur. Between convulsions, confusion, incoordination, and hyperexcitablity are often present. With few exceptions, individuals acutely poisoned with cyclodienes experience full reversion of symptoms, although it may take several months. Chronic exposures to PCCAs may lead to persistent neurological deficits. Spraymen often complain of headache, dizziness, and general malaise when applying cyclodienes over long periods of time (28). Long-term exposure to dieldrin in humans has caused neurological impairments lasting over 1 yr, along with a sensitization to subsequent acute exposures, and an increased responsiveness to epileptogenic stimuli (13). Effects of dieldrin can occur at low doses, since 30 µg/kg/d of dieldrin caused convulsions in some people. Thus, there are sensitive individuals for whom even mild insecticide expo-

70

Bloomquist

sure appears to be significant. In addition, children may be particularly sensitive to these compounds: Chronic exposure to aldrin and lindane has been claimed to cause mental retardation and hyperkinesia, and chronic endrin poisoning with decerebrate rigidity and brain damage (13). Human exposure to avermectins primarily occurs via their therapeutic uses as anthelmintics, and I am unaware of any documented cases of agricultural avermectin use resulting in either acute or chronic neurotoxicity in humans. Community-based treatment with ivermectin for onchocerciasis in Liberia resulted in 7699 treated individuals. Only 1.3% complained of any complications, but some dizziness was observed (31). A similar study covering several countries in west Africa (32) documented postural hypotension in 49 cases out of nearly 51,000 treated individuals. In both studies, neurological effects were correlated with microfilarial load, which suggests they are probably a side effect of microfilaria mortality and unrelated to drug action on the nervous system. However, it is interesting to note that deliberate oral ingestion of abamectin in attempted suicides also results in hypotension as a prominent feature of intoxication in humans (33). 3. MECHANISM OF ACTION 3.1. Effects of PCCAs and Avermectins at the GABAA Receptor The GABAA receptor-chloride ionophore complex is an important mediator of neuronal inhibition in the mammalian CNS (34). A number of studies on the electrophysiological and biochemical actions of insecticides on the GABAA receptor have appeared. This material has been extensively reviewed (4,27,35–37) and the following two sections are a summary of electrophysiological and biochemical effects on the GABAA receptor taken from these sources, along with specific citation of more recent or additional findings. 3.1.1. Electrophysiological Studies

The suggestion that blockage of the GABA-gated chloride channel might account for cyclodiene-dependent synaptic facilitation was originally put forward by Ghiasuddin and Matsumura (38). A number of subsequent studies using voltageclamped neurons found that nanomolar or micromolar concentrations of lindane or related cyclodienes strongly antagonized inward chloride current caused by application of GABA, with somewhat selective effects on peak current amplitude. Moreover, these compounds showed a noncompetitive or mixed type of GABA antagonism. In part, the blocking action of cyclodienes and other convulsants can be overcome, in vitro, by prolonged exposure to GABA (39). The mixed inhibition observed with these compounds was explained by stabilization of a closed or desensitized GABA-bound form of the channel. Single channel studies found three closed states of the GABAgated chloride channel whose time constants were prolonged by the action of dieldrin (40). Blockage of the chloride channel involves a critical Alanine residue at the 302 position in the M2 channel-forming helix of GABA receptor subunits (41). Although these effects are usually thought to be confined to central GABAA receptors, lindane is also known to block the action of GABA on the myenteric plexus ganglia of the guinea pig, thereby demonstrating an effect on the PNS (42).


71

Unlike the channel blockage observed with PCCAs, the major effect of avermectins is channel activation. Ivermectin increased the membrane conductance and evoked an inward current in cultured chick spinal neurons that was sensitive to the chloride channel blocker, picrotoxinin (43). Reducing the external chloride ion concentration of the saline bathing the neurons shifted the reversal potential for milbemycin D- or ivermectin-induced currents in a positive direction, consistent with chloride being the charge carrier. The results suggest that these compounds activate the GABAA receptor, although a greater percentage of cultured neurons were activated by GABA than milbemycin D or ivermectin. When applied to rat dorsal root ganglion neurons at low concentrations, avermectins also block the action of applied GABA and at high concentrations irreversibly activate the chloride channel (44). GABA-induced inward currents from chick brain GABA receptors expressed in Xenopus oocytes were potentiated by 1 µM abamectin (45). The proposed mechanisms underlying this potentiation included an increase of nondesensitizing current, a nearly 10-fold decrease in the Ka for GABA, and a reduction from 1.7 to 1.1 in the Hill coefficient for GABA. Similarly, ivermectin mimicked the effect of GABA on ileum preparations, suggesting that this tissue is a target for avermectins in the periphery (42). 3.1.2. Ion Flux and Radioligand Binding Studies GABA-stimulated 36Cl uptake and radioligand-binding experiments using mouse and rat brain vesicle preparations confirmed PCCA blockage of the GABA receptor and facilitated quantitative structure-activity analysis. The most potent compounds for inhibiting chloride flux were 12-ketoendrin, isobenzan, and endrin, with IC50 values of approx 1 µM. Dieldrin and heptachlor epoxide had intermediate potency (IC50s typically between 4 and 18 µM), and the least active compounds were aldrin, heptachlor, and lindane with IC50s > 20 µM. Moreover, the inhibition displayed noncompetitive kinetics, consistent with their neurophysiological effects. Fipronil blocks 36Cl uptake into mouse (46) and rat (15) brain vesicles with IC50 values >10 µM, consistent with its somewhat lower mammalian toxicity compared to the older cyclodienes. Cyclodienes also proved to be potent competitive inhibitors of [ 3H]-4'-ethynyl-4-n-propylbicycloorthobenzoate (EBOB) binding, typically in the nanomolar range, indicating that they bind to the same site on the chloride channel. In contrast, fipronil displaced [3H]EBOB binding with an IC50 4.3 µM in mouse brain, and was a noncompetitive inhibitor of EBOB binding in the housefly (46), suggesting that its binding is poorly reversible or allosteric to the EBOB site. Good overall correlations were obtained between the potency for blocking chloride uptake, displacement of [3H]EBOB binding from the convulsant site, and acute lethality for a range of PCCAs and trioxabicyclooctanes. However, all the structural classes of chloride channel blockers and all radioligands purporting to label the same site could not be grouped together because they fell on different regression lines when correlated with toxicity or binding activity, suggesting subtle differences in action or binding sites in mammals (35). In contrast to the predominant channel-blocking action of the cyclodienes, the avermectins have more complicated effects on the GABAA receptor (8,27). In 36Cl flux assays, abamectin displays some ability to activate the GABA receptor of rat brain vesicles in a bicuculline-sensitive manner, but acted as a pure noncompetitive antagonist at mouse brain GABAA receptors (IC50 for blocking GABA-stimulated chloride

72

Bloomquist

uptake = 93 nM). GABA antagonism is consistent with tremor in exposed animals, whereas ataxia and sedation suggest potentiation of GABA action or direct activation of the chloride channel. Avermectins are also known to displace [3H]EBOB binding in rat brain membranes, again suggesting an action on GABA receptors. In cultured cerebellar granule cell neurons, [3H]avermectin B1a bound to high (Kd = 5 nM) and low (866 nM) affinity sites that correlated with its ability to stimulate 36Cl uptake at 3–100 nM and to block uptake at 1–3 µM (47). These data suggested that avermectins bind to two different sites on the GABA-gated chloride channel, activating the channel upon binding to the high-affinity site and blocking it upon further binding to the low affinity site. This hypothesis, however, is inconsistent with electrophysiological studies, where avermectins tend to block the action of GABA at low concentrations, and then activate the channel irreversibly at higher concentrations. Additional research is required to reconcile this discrepancy. 3.2. Effects on Other Ligand-Gated Chloride Channels The glycine receptor is another inhibitory neurotransmitter receptor found in the brain and spinal cord that could serve as a target for chloride channel-directed compounds. Glycine-evoked currents were blocked noncompetitively by picrotoxin with an IC50 of 180 µM (48). In contrast, immunoaffinity-column removal of glycine receptors from CNS membranes had little effect on the binding of [35S]tert-butylbicyclophophorothionate (TBPS), a ligand with properties similar to EBOB. Thus, there appears to be little, if any, binding of TBPS to the glycine receptor (49). These data suggest that glycine receptors in mammals may be of relatively minor importance as a target site of PCCAs, because of their reduced sensitivity compared to GABAA receptors. A single study reported that abamectin noncompetitively displaces the binding of [3H]strychnine from spinal-cord membranes or from purified receptor with Ki values in the low micromolar range (50). This effect is less potent than other documented actions for abamectin, and any effects on glycine-receptor function remain to be determined. Effects on glutamate-gated chloride channels are important in the selectivity of the avermectins, where these compounds apparently target an invertebrate receptor that is absent in mammals. Studies on GABA-insensitive locust muscle fibers demonstrated that ivermectin increased the conductance of the muscle membrane and blocked the effect of applied agonist on a glutamate-gated chloride channel (51). Effects on this receptor have been studied in more detail on nematode (Caenorhabditis elegans) receptors expressed in Xenopus oocytes (52). These studies also found that avermectin analogs activated a glutamate-gated chloride channel with a threshold concentrations in the low nanomolar range. Effects on this receptor can account for the excellent selectivity and potent paralytic effects of avermectins observed in insects, acarines, and nematodes. 3.3. Effects on Voltage-Gated Chloride Channels A role for voltage-gated chloride channels in the action of PCCAs comes from toxicity and binding studies. In fish (Torpedo nobiliana) electric organ, a tissue that lacks GABA receptors, but does possess voltage-gated chloride channels (53), specific bind-


73

ing of [35S]TBPS was displaced by picrotoxinin and endrin with IC50 values of 100 µM and 0.75 µM (54). These values are 500-fold and 25-fold less potent, respectively, than their IC50 values for displacing TBPS binding in rat brain. In contrast, lindane was about fourfold more potent as an inhibitor of binding to Torpedo membranes than rat brain, and was a more effective inhibitor than either picrotoxinin or endrin (IC50 = 40 nM). Later studies confirmed a 10-fold increase in potency of lindane for inhibiting [35S]TBPS binding in electric organ compared to rat brain (55), and suggested that TBPS may bind to both the GABA receptor and a voltage-dependent chloride channel in rat brain. The toxicological relevance of effects on these channels remains to be established. However, lindane and TBPS show similar LD50 values in mice after intraperitoneal or intracerebral injection, which differentiates them from the cyclodienes and picrotoxinin and supports the possibility of effects on other sites, since most GABA receptors are confined to the CNS (20). Voltage-dependent chloride channels also serve as potential target sites for the avermectins. The effects of avermectins have been studied in assays of 36Cl efflux from brain vesicles. The efflux is insensitive to GABA, but can be stimulated by avermectins (55–58). GABA-insensitive efflux was ascribed to an action on voltage-gated chloride channels because it was blocked by 4,4'-di-isothiocyanotostilbene-2,2'-disulfonic acid (DIDS), a known blocker of these channels (56). Similar results were observed for abamectin-dependent efflux from mouse brain vesicles, but in this study, DIDS was ineffective as a channel blocker, as were picrotoxinin and TBPS (57). Structure-activity studies in mouse brain vesicles (58) compared inhibition of GABA-stimulated chloride uptake to stimulation of chloride efflux. Abamectin and emamectin were fivefold and twofold more active, respectively, as GABA antagonists, whereas other noncommercial avermectin analogs were more potent for stimulating GABA-independent chloride efflux. The potencies of avermectins for stimulating chloride efflux were not correlated with anthelmintic or insecticidal activity, but may contribute to the toxicity of these compounds in mammals (57,58) 3.4. Antidotes for Cyclodiene and Avermectin Intoxication Data on effective antidotes is available for PCCAs, but there is less information available on antidotes to avermectin poisoning. Treatment with barbiturates is the most effective therapy for blocking intoxication by PCCAs (13). Acute poisoning of a human with aldrin (ca. 26 mg/kg) was successfully treated with pentobarbital, which effectively suppressed convulsions when given over a 4-d period. Similarly, convulsions and gran mal seizures were controlled by phenobarbital in a child that was reported to consume 1.5 g of lindane. The known ability of barbiturates to increase the proportion of GABAA receptors to the longest-lived open state (34) is consistent with their antidotal effects on channel-blocking PCCAs. In cattle suffering from avermectin intoxication, picrotoxin was used as a candidate antidote to little ameliorative effect (59). This approach was predicated on the assumption that avermectins were activating the GABAA receptor, which would be blocked by picrotoxin. Clearly, the intoxication process is more complex than envisioned, and identifying an effective antidote for avermectin poisoning will require further experimentation.

74

Bloomquist

4. FUTURE DIRECTIONS Future studies on the chlorinated PCCAs should address their potential for chronic neurotoxicity at low exposure levels. These studies are justified by the extreme persistence of these molecules and their demonstrated neurotoxicity, especially to neonates. One study (60) has implicated organochlorine insecticides in chronic motor neuron disease, and confirmed intoxication in these cases by aldrin, lindane, and heptachlor. Another movement disorder where organochlorines may play a role is Parkinson’s disease (PD). Epidemiological studies have shown a consistent linkage between an increased risk for PD and exposure to insecticides (61–63). A recent study from Germany has specifically implicated exposure to organochlorine and organophosphorus insecticides in the etiology of PD (64). Two studies observed a correlation between the incidence of PD and brain residues of organochlorine insecticides. Fleming et al. (65) found that the occurrence of PD was significantly correlated (p = 0.03) with the presence of brain residues of the insecticide dieldrin (mean = 13 ppb). Similar results were reported by Corrigan et al. (66), who found significantly elevated levels of dieldrin in the caudate nucleus of Parkinson’s patients. When applied to mesencephalic neuron cultures, dieldrin caused cytotoxicity in dopaminergic neurons more so than GABAergic neurons, although relatively high concentrations were required: EC50 = 12 µM (67). Any contribution of GABAA receptors for mediating these effects remains to be explored. REFERENCES 1. Albrecht, W. (1987) Central nervous system toxicity of some common environmental residues in the mouse. J. Toxicol. Env. Health 21, 405–421. 2. Putter, I., MacConnell, J., Preiser, F., Haidiri, A., Ristich, S., and Dybas, R. (1981) Avermectins: novel insecticides, acaricides, and nematicides from a soil microorganism. Experientia 37, 963–964. 3. Brooks, G. T. (1974) Chlorinated Insecticides, vol. 1, 2. CRC Press, Cleveland, OH. 4. Bloomquist, J R. (1998) Chemistry and toxicology of the chlorinated cyclodienes and lindane. Rev. Toxicol. 2, 333–355 5. Lawrence, L. J. and Casida, J. E. (1984) Interactions of lindane, toxaphene and cyclodienes with brain-specific t-butylbicyclophosphorothionate receptor. Life Sci. 35, 171–178. 6. Turner, W., Engel, J., and Casida, J. (1977) Toxaphene components and related compounds: Preparation and toxicity of some hepta-, octa-, and nonachlorobornanes, hexa-, and heptachlorobornenes, and a hexachlorobornadiene. J. Agric. Food Chem. 25, 1394–1401. 7. Anonymous (1986) Fipronil. Worldwide Technical Bulletin. Rhone-Poulenc Inc., Research Triangle Park, NC, pp. 1–20. 8. Fisher, M. and Mrozik, H. (1992) The chemistry and pharmacology of avermectins. Ann. Rev. Pharmacol. Toxicol. 32, 537–553. 9. Shoop, W., Mrozik, H., and Fisher, M. (1995) Structure and activity of avermectins and milbemycins in animal health. Vet. Parasitol. 59, 139–156 10. Meinke, P., Shoop, W., Michael, B., Blizzard, T., Dawson, G., Fisher, M., and Mrozik, H. (1998) Synthesis of gem-difluoro-avermectin derivatives: potent anthelmintic and anticonvulsant agents. Bioorg. Med. Chem. Lett. 8, 3643–3646 11. Goudie, A., Evans, N., Gration, K., Bishop, B., Gibson, S., Holdom, K., et al. (1993) Doramectin-a potent novel endectocide. Vet. Parasitol. 49, 5–15. 12. Cole, L. M. and Casida, J. E. (1986) Polychlorocycloalkane insecticide-induced convul-


13. 14. 15. 16. 17.

18. 19. 20. 21.

22.

23.

24. 25.

26. 27. 28. 29.

30.

31.

32.

75

sions in mice in relation to disruption of the GABA-regulated chloride ionophore. Life Sci. 39, 1855–1862. Ecobichon, D. J. and Joy, R. M. (1994) Pesticides and Neurological Diseases, 2nd ed. CRC Press, Boca Raton, FL, pp. 122–170. Meister, R. T. (ed.) (2000) Farm Chemicals Handbook. Meister Publishing, Willoughby, OH. Gant, D., Chalmers, A., Wolff, M., Hoffman, H., and Bushey, D. (1998) Fipronil: action at the GABA receptor. Rev. Toxicol. 2, 147–156. Hainzl, D. and Casida, J. (1996) Fipronil insecticide: novel photochemical desulfinylation with retention of neurotoxicity. Proc. Natl. Acad. Sci. USA 93, 12,764–12,767. Hainzl, D., Cole, L., and Casida, J. (1998) Mechanisms for selective toxicity of fipronil insecticide and its sulfone metabolite and desulfinyl photoproduct. Chem. Res. Toxicol. 11, 1529–1535. Joy, R. M. (1982) Mode of action of lindane, dieldrin and related insecticides in the central nervous system. Neurobehav. Toxicol. Teratol. 4, 813–823. Hamernik, K. (1997) Fipronil, in Pesticide Residue in Food; Toxicology Evaluations, Food and Agriculture Organization/World Health Organization, Malta, pp. 109–156. Bloomquist, J. R. (1992) Intrinsic lethality of chloride-channel-directed insecticides and convulsants in mammals. Toxicol. Lett. 60, 289–298. Akkermans, L. M., van den Bercken, J., and Versluijs-Helder, M. (1975) Excitatory and depressant effects of dieldrin and aldrin-transdiol in the spinal cord of the toad (Xenopus laevis). Eur. J. Pharmacol. 34, 133–142. Wooley, D., Zimmer, L., Hasan, Z., and Swanson, K. (1984) Do some insecticides and heavy metals produce long-term potentiation in the limbic system? in Cellular and Molecular Neurotoxicology, Raven Press, NY, pp. 45–69. Bloomquist, J. R., Adams, P. M., and Soderlund D. M. (1987) Neurotoxic insecticides as antagonists of the GABAA receptor function, in Sites of Action for Neurotoxic Pesticides (Hollingworth, R. M. and Green, M. B., eds.), American Chemical Society, Washington DC, pp. 97–106. Lankas, G. R. and Gordon, L. R. (1989) Toxicology, in Ivermectin and Abamectin (Campbell, W. C., ed.) Springer-Verlag, New York, pp. 89–112. Schaeffer, J., Rohrer, S., Cully, D., and Arena, J. (1992) Avermectins: Idiosyncratic toxicity in a subpopulation of collie dogs, in Neurotox ’91, Molecular Basis of Drug and Pesticide Action (Duce, I. R., ed.), Elsevier Applied Science, London, pp. 71–76. Wang, C. C. and Pong, S.-S. (1982) Actions of avermectin B1A on GABA nerves, in Membranes and Genetic Diseases, Alan R. Liss, Inc., NY, pp. 373–395. Bloomquist, J. R. (1993) Toxicology, mode of action, and target site-mediated resistance to insecticides acting on chloride channels. Comp. Biochem. Physiol. 106C, 301–314. Hayes, W. J. (1982) Pesticides Studied in Man. Williams & Wilkins, Baltimore, pp. 211–283. Wise, L. D., Verbeke, D. R., Allen, H. L., Gerson, R. J., and Manson, J. M. (1992) Developmental neurotoxicity study of MK–0244 in Sprague-Dawley rats. Teratology 45(5), 525RN. Wise, L. D., Allen, H. L., Hoe, C. M., Verbeke, D. R., and Gerson, R. J. (1997) Developmental neurotoxicity evaluation of the avermectin pesticide, emamectin benzoate, in Sprague-Dawley rats. Neurotoxicol. Teratol. 19, 315–326. Pacque, M., Dukuly, Z., Greene, B., Munoz, B., Keyvan-Larijani, E., Williams, P., and Taylor, H. (1989) Community-based treatment of onchocerciasis with ivermectin: acceptability and early adverse reactions. Bull. WHO 67, 721–730. De Sole, G., Remme, J., Awadzi, K., Accorsi, S., Alley, E., Ba, O., et al. (1989) Adverse reactions after large-scale treatment of onchocerciasis with ivermectin: combined results from eight community trials. Bull. WHO 67, 707–719.

76

Bloomquist

33. Chung, K., Yang, C., Wu, M., Deng, J., and Tsai, W. (1999) Agricultural avermectins: an uncommon but potentially fatal cause of pesticide poisoning. Ann. Emerg. Med. 34, 51–57. 34. Delorey, T. and Olson, R. (1994) GABA and glycine, in Basic Neurochemistry, 5th ed. (Siegel, G., Agranoff, B., Albers, R., and Molinoff, P., eds.), Raven Press, NY, pp. 389–400. 35. Casida, J., Nicholson, R., and Palmer, C. (1988) Trioxabicyclooctanes as probes for the convulsant site of the GABA-gated chloride channel in mammals and arthropods, in Neurotox’88: Molecular Basis of Drug & Pesticide Action (Lunt, G.G., ed.), Elsevier Science Publishers B.V., Amsterdam, Netherlands, pp. 125–144. 36. Casida, J. (1993) Insecticide action at the GABA-gated chloride channel: recognition, progress, and prospects. Arch Insect Biochem. Physiol. 22, 13–23. 37. Deng, Y. (1995) Insecticide binding sites in the house fly head γ-aminobutyric acid gated chloride-channel complex, in Molecular Action of Insecticides on Ion Channels (Clark, J. M., ed.), ACS Sump. Ser. 591, American Chemical Society, Washington, DC, pp. 230–250. 38. Ghiasuddin, S. and Matsumura, F. (1982) Inhibition of gamma-aminobutyric acid (GABA)induced chloride uptake by gamma-BHC and heptachlor epoxide. Comp. Biochem. Physiol. 73C, 141–144. 39. Bloomquist, J. R., Grubs, R. E., Soderlund, D. M., and Knipple D. C. (1991) Prolonged exposure to GABA-gated chloride channels in the presence of channel-blocking convulsants. Comp. Biochem. Physiol. 99C, 397–402. 40. Ikeda, T., Nagata, K., Shono, T., and Narahashi, T. (1998) Dieldrin and picrotoxinin modulation of GABAA receptor single channels. NeuroReport 9, 3189–3195. 41. Ffrench-Constant, R., Zhang, H-J., and Jackson, M. (1995) Biophysical analysis of a single amino acid replacement in the resistance to dieldrin γ-aminobutyric acid receptor, in Molecular Action of Insecticides on Ion Channels (Clark, J. M., ed.), ACS Symp. Ser. 591, American Chemical Society, Washington, DC, pp. 192–204. 42. Coccini, T., Candura, S., Manzo, L., Costa, L., and Tonini, M. (1993) Interaction of the neurotoxic pesticides ivermectin and lindane with the enteric GABAA receptor-ionophore complex in the guinea-pig. Eur. J. Pharmacol. 248, 1–6. 43. Yamazaki, J., Matsumoto, K., Ono, H., and Fukuda, H. (1989) Macrolide compounds, ivermectin and milbemycin D., stimulate chloride channels sensitive to GABAergic drugs in cultured chick spinal neurons. Comp. Biochem. Physiol. 93C, 97–104. 44. Robertson, B. (1989) Actions of anaesthetics and avermectin on GABAA chloride channels in mammalian dorsal root ganglion neurones. Br. J. Pharmacol. 98, 167–176 45. Sigel, E. and Baur, R. (1987) Effect of avermectin B1A on chick neuronal γ-aminobutyrate receptor channels expressed in Xenopus oocytes. Mol. Pharmacol. 32, 749–752. 46. Cole, L., Nicholson, R., and Casida, J. (1993) Action of phenylpyrazole insecticides at the GABA-gated chloride channel. Pestic. Biochem. Physiol. 46, 47–54 47. Huang, J. and Casida, J. (1997) Avermectin B1a binds to high—and low-affinity sites with dual effects on the γ-aminobutyric acid-gated chloride channel of cultured cerebellar granule neurons. J. Pharmacol. Exp. Ther. 281, 261–266. 48. Akaike, N. and Kaneda, M. (1989) Glycine-gated chloride current in acutely isolated rat hypothalamic neurons. J. Neurophysiol. 62, 1400–1409. 49. Reinitz, A., Becker, C.-M., Betz, H., and Schmitt, B. (1987) The chloride channel blocking agent, t-butylbicyclophosphorothionate, binds to the γ-aminobutyric acid-benzodiazepine, but not to the glycine receptor in rodents. Neurosci. Lett. 76, 91–95. 50. Graham, D., Pfeiffer, F., and Betz H. (1982) Avermectin B1a inhibits the binding of strychnine to the glycine receptor of rat spinal cord. Neurosci. Lett. 29, 173–176. 51. Duce, I., Bhandal, N., Scott, R., and Norris, T. (1995) Effects of ivermectin on γaminobutyric acid and glutamate-gated chloride conductance in arthropod skeletal muscle,


52.

53. 54.

55.

56.

57.

58.

59. 60.

61. 62. 63.

64.

65. 66. 67.

77

in Molecular Action of Insecticides on Ion Channels (Clark, J. M., ed.), ACS Symp. Ser. 591, American Chemical Society, Washington, DC, pp. 250–263. Rohrer, S. and Arena, J. (1995) Structural and functional characterization of insect genes encoding ligand-gated chloride-channel subunits, in Molecular Action of Insecticides on Ion Channels (Clark, J. M., ed.), ACS Symp. Ser. 591, American Chemical Society, Washington, DC, pp. 264–283. Richard, E. A. and Miller, C. (1990) Steady-state coupling of ion-channel conformations to a transmembrane ion gradient. Science 247, 1208–1210. Abalis, I. M., Eldefrawi, M. E., and Eldefrawi, A. T. (1985) Binding of GABA receptor channel drugs to a putative voltage-dependent chloride channel in Torpedo electric organ. Biochem. Pharmacol. 34, 2579–2582. Thompson, R. G., Menking, D. E., and Valdez, A. A. (1990) Comparison of lindane, bicyclophosphate and picrotoxin binding to the putative chloride channel sites in rat brain and Torpedo electric organ. Neurotoxicol. Teratol. 12, 57–63. Abalis, I. M., Eldefrawi, A. T., and Eldefrawi, M. E. (1986) Actions of avermectin B1a on the γ-aminobutyric acidA receptor and chloride channels in rat brain. J. Biochem. Toxicol. 1, 69–82. Payne, G. T. and Soderlund, D. M. (1991) Activation of γ-aminobutyric acid insensitive chloride channels in mouse brain synaptic vesicles by avermectin B1a. J. Biochem. Toxicol. 6, 283–292. Payne, G. and Soderlund, D. (1993) Actions of avermectin analogues on γ-aminobutyric acid (GABA)-sensitive and GABA-insensitive chloride channels in mouse brain. Pestic. Biochem. Physiol. 47, 178–184. Button, C., Barton, R., Honey, P., and Rickford, P. (1988) Avermectin toxicity in calves and an evaluation of picrotoxin as an antidote. Aust. Vet. J. 65, 157–158. Fonseca, R., Resende, L., Silva, M., and Camargo, A. (1993) Chronic motor neuron disease possibly related to intoxication with organochlorine insecticides. Acta Neurol. Scand. 88, 56–58. Tanner, C. M. and Langston, J. W. (1990) Do environmental toxins cause Parkinson’s disease? A critical review. Neurology 40, 17–30. Butterfield, P., Valanis, B., Spencer, P., Lindeman, C., and Nutt, J. (1993) Environmental antecedents of young-onset Parkinson’s disease. Neurosci. Behav. 43, 1150–1158. Gorell, J., Johnson, C., Rybicki, B., Peterson, E., and Richardson, R. (1998) The risk of Parkinson’s disease with exposure to pesticides, farming, well water, and rural living. Neurology 50, 1346–1350. Siedler, A., Hellenbrand, W., Robra, B., Vieregge, P., Nischan, P., Joerg, J., et al. (1996) Possible environmental. occupational, and other etiologic factors for Parkinson’s disease: a case control study in Germany. Neurology 46, 1275–1284. Fleming, L., Mann, J., Bean, J., Briggle, T., and Sanchez-Ramos, J. (1994) Parkinson’s disease and brain levels of organochlorine pesticides. Ann. Neurol. 36, 100–103. Corrigan, F., Murray, L., Wyatt, C., and Shore, R. (1998) Diorthosubstituted polychlorinated biphenyls in caudate nucleus in Parkinson’s disease. Exp. Neur. 150, 339–342. Sanchez-Ramos, J., Facca, A., Basit, A., and Song, S. (1998) Toxicity of dieldrin for dopaminergic neurons in mesencephalic cultures. Exp. Neurol. 150, 263–271.

78

Bloomquist

Neonicotinoid Insecticides

79

6 The Neonicotinoid Insecticides Larry P. Sheets

1. INTRODUCTION 1.1. Historical Overview The neonicotinoid insecticides are a new generation of chemical agents that have recently been developed for commercial use. Their history can be traced to the late 1970s, when chemists at Shell Chemical Company investigated the heterocyclic nitromethylenes as potential insecticides (1,2). This research led to the discovery of nithiazine (WL 35651) (Fig. 1), which is the only representative of the early type of heterocyclic nitromethylene that has been registered for use as an insecticide. Poor photostability generally limits commercial applications for nithiazine and related (i.e., nitromethylene) compounds. In 1984, chemists at Nihon Bayer Agrochem explored the introduction of a 3-pyridylmethyl group on the nitromethylene heterocycle structure. The introduction of this moiety greatly increases insecticidal activity and ultimately led to the discovery of imidacloprid (3). Imidacloprid is the first representative from this group to be registered for use (1991 launch) and is currently the most important commercial product. Since then, other analogs of imidacloprid have been invented, including acetamiprid (4) and nitenpyram (5). Collectively, these chemicals are referred to as “neonicotinoids” (6) to distinguish them from the nicotinoids; the neonicotinoids being more highly effective insecticides and less toxic to vertebrate species. Representatives from this group are also referred to as “chloronicotinyls,” owing to the importance of the chlorine atom for insecticidal potency. The neonicotinoid insecticides are an important addition to the marketplace. They present a very favorable toxicological profile, being highly effective with relatively low risk to nontarget species, especially when compared to nicotine. Furthermore, the mode of action for representatives in this group is different from that of other insecticides, such as the organophosphates, carbamates, and pyrethroids. Thus, the neonicotinoids are new tools that are being used in the management of insect resistance. To understand the neurotoxic potential of these compounds, it is useful to review their mode of action.


79

80

Sheets

Fig. 1. Chemical structures of nicotine and the three major classes of neonicotinoid insecticides: nitromethylenes, chloropyridines, and chlorothiazoles.

1.2. Chemistry of Neonicotinoids The insecticidal activity of neonicotinoids is attributed to actions on nicotinic acetylcholine receptors (nAChR), which are ligand-gated ion channels. In insects, these receptors are located exclusively in central nervous system (CNS) tissues (ganglia and brain). The majority of these compounds (i.e., imidacloprid, acetamiprid, nitenpyram, and thiacloprid) possess a 6-chloro-3-pyridinyl moiety that confers added potency and selective action (Fig. 1). For related analogues, the binding affinity to the insect receptor is decreased several-fold if the chlorine atom is absent (6,7). Replacement of the chloropyridinyl moiety with a chlorothiazolyl group has led to the discovery of compounds that are considered a second generation of neonicotinoid insecticides (8). This substitution further reduces potency in assays with mammalian receptors but does not appear to reduce toxicity to mammals or activity at the insect nAChR (7,9). Compounds with the chlorothiazolyl group that have been developed for commercial use are clothianidin (TI-435) and thiamethoxam (CGA 293’343) (Fig. 1).


81

2. MODE OF ACTION 2.1. Insects Characterization of the structural and functional properties of nAChR subtypes in insect tissues has progressed rapidly in recent years, due to the development of highly specific ligands such as imidacloprid. The current state of knowledge regarding these receptors is summarized elsewhere (10). Neonicotinoid insecticides cause profound effects on nerve-impulse transmission through actions on insect postsynaptic nAChRs (11). Electrophysiological experiments in various neuronal preparations demonstrate that the response to treatment is biphasic: characterized by an initial increase in the frequency of spontaneous discharge, followed by a complete block to nerve-impulse propagation (2). A variety of neurotoxic signs is evident in insects that have been treated with a neonicotinoid insecticide, including evidence of nicotinic stimulation. For example, the administration of imidacloprid to the American cockroach (Periplaneta americana) causes uncoordinated abdominal quivering, wing flexing, tremors, and violent whole body shaking, followed by prostration and death (2). Insecticidal activity is greatly enhanced by synergists that inhibit oxidative degradation (7). 2.2. Mammals Mammalian tissues contain many subtypes of nAChRs that are formed from five homologous subunits, in combinations from nine α, four β, γ, δ, and ε subunits (10). The various receptor subtypes are located in peripheral tissues (e.g., the neuromuscular junction of skeletal muscle and autonomic ganglia), as well as in spinal cord and brain tissues. The neonicotinoid insecticides have much lower activity in vertebrate tissues than in insects, due to differences in binding to nAChR subtypes. Tomizawa and Casida (12) have summarized the evidence for the relatively low potency of imidacloprid, the most thoroughly studied representative, in vertebrate tissues as follows: (1) failure to recognize [3H]-imidacloprid-specific binding sites in brain from avian and mammalian species (7); (2) low potency as an inhibitor of [3H]-α-bungarotoxin (BGT) binding and low agonistic effect in muscle type nAChR from Torpedo electric organ (14); (3) little activity as an inhibitor of [3H]-nicotine binding to rat and mouse brain membranes (9,15); (4) very weak agonistic action in mouse N1E-115 neuroblastoma (neuronal type) and BC3H1 (motor endplate type) muscle cells (16,17); (5) low activity in ion channel activation compared to acetylcholine with rat α4β2 and α7 receptor subtypes expressed in Xenopus oocytes (18); and (6) weak or partial agonistic nature with recombinant chick α4β2 receptor (19). The toxicity of the neonicotinoid insecticides and related analogs in mammals is most closely related to potency at the α7 nAChR subtype, with a decreasing relationship reported sequentially at the α4β2, α3 and α1 nAChR, respectively (12). However, toxicity in mammals appears to involve complex actions at multiple-receptor subtypes, with relative subtype specificity conferred by minor structural changes (12). The actions of imidacloprid at these various subtypes has been described as a combination of agonist and antagonist effects (11). Given the combination of complex actions and relative

82

Sheets

subtype specificity, the toxic effects in the whole animal may vary among representative compounds. Poor penetration of the blood-brain barrier (BBB) is an additional factor that acts to reduce the toxicity of the neonicotinoid insecticides in vertebrate species (15). By having reduced access to receptors in the CNS, centrally mediated effects tend to occur only at higher levels of exposure, relative to ones that affect peripheral tissues. Studies with laboratory animals support this evidence of relatively low potency. For example, acute oral toxicity studies with acetamiprid (20), nitenpyram (21), imidacloprid (22), and thiamethoxam (8) provide LD50 estimates in the range of approx 200–1600 mg/kg. Doses in this range are considered high, relative to their potency in insects and the low application rates used to control insect pests. Nevertheless, at relatively high levels of exposure, neonicotinoid insecticides are neuroactive and produce neurotoxic effects. The effects that are evident following exposure depend on a number of factors, including the distribution to various target sites, relative potency at various receptor subtypes, and whether the principal effect is stimulation or inhibition. 3. NEUROTOXICITY 3.1. Adult Animals 3.1.1. Published Studies There is a paucity of information in the published literature about the neurotoxicity of the neonicotinoid insecticides in vertebrate species. Studies that have examined the toxicity of various compounds in mammals have generally limited the assessment to comparisons of acute doses that produce lethality, with no description of toxic signs or other evidence of neurotoxicity. One exception is a study by Chao and Casida (9), in which tremors were reported to occur in mice that had been treated with representative compounds. There are no published studies that have examined the neurotoxic potential of neonicotinoid insecticides following repeated exposure. The only exception to this generalization is a brief reference to the findings in acute and subchronic neurotoxicity studies with imidacloprid that were conducted to support product registrations (22). 3.1.2. Unpublished, Proprietary Studies The most extensive source of neurotoxicity data for the neonicotinoid insecticides, is the toxicology database that is generated by the crop protection industry to support the registration of compounds for commercial uses. Included in this database are studies in rats and other laboratory mammals following acute and subchronic (e.g., 90 d) exposure. For any food-use registration, the requisite database must also include chronic 1-yr exposure studies with rats, mice, and nonrodent (e.g., canine) species. In these studies, neurotoxicity is only one aspect of the potential health effects that is evaluated and neurotoxic potential is generally assessed by routine cage-side observations for clinical signs as well as microscopic examination of neural tissues (brain, spinal cord, and peripheral nerves) that are collected at study termination. The canine studies may include additional tests that are relevant to the assessment of neurotoxic potential, including a neurologic examination. The database for many of these insecticides also includes studies that are designed more specifically to assess neurotoxic potential. It is the results from these unpublished


83

studies that are evaluated here to profile the neurotoxic potential of the neonicotinoid insecticides. The source of this information is a combination of the author’s personal experience as the principal investigator/study director for studies with imidacloprid, thiacloprid, and clothianidin, and information kindly provided by the registrants of the other products. 3.1.3. Neurotoxicity Screening Battery (US EPA, OPPTS 870.6200) This test battery consists of acute and subchronic studies in young-adult male and female rats. In these studies, the technical grade active ingredient (typical purity, 98%) is administered at a minimum of three dose levels, plus a concurrent control. This includes one dose that produces no evidence of toxicity (i.e., a no-observed-effect level, NOEL) and a dose that produces clear evidence of toxicity, approximating the highest dose that can be tolerated without excessive mortality. The latter represents a severe test that is included to assist with the identification of target organs and overt toxicity associated with exposure to a near-lethal dose. The results at the high dose can also assist with the interpretation of potential effects at lower dose levels, where the effects may be more subtle or occur at a low incidence and might otherwise be considered incidental. At specific times relative to treatment, the animals are tested using a functional observational battery (FOB) and an automated test of spontaneous motor activity. The typical FOB used in these studies is based on the procedure described by Moser (23). Briefly, the FOB consists of a standardized series of neurobehavioral observations, neurologic tests, and measurements, including observations and tests that are performed outside the home cage. Standard observations include assessments of activity level, posture, gait, autonomic function, pupillary function, the presence of unusual behaviors (e.g., tremors or convulsions), and response to stimuli. Also included are measures of landing foot splay, grip strength, and body temperature. Motor activity is also measured by testing animals individually in an automated device (e.g., the figure-eight maze or an open field). At the end of the study, animals are anesthetized and perfused with a fixative and an assortment of neural tissues is collected for microscopic examination. The studies reviewed here include a minimum of the following tissues: brain (3–8 levels), spinal cord (3–4 levels), peripheral nerves (sciatic, tibial and sural), dorsal-root ganglia (including dorsal and ventral root fibers), the gasserian ganglion, retina, optic nerve, and representative skeletal muscle (e.g., gastrocnemius). The following is based on studies with acetamiprid, clothianidin, imidacloprid, thiacloprid, and thiamethoxam. These are reviewed collectively, because the principal effects are not distinctive among these representative compounds. Differences that are evident (e.g., the specific combination of clinical signs) appear to be associated with dose selection, rather than an actual difference in neurotoxic effects. On the other hand, these studies were not performed in a manner that would permit a detailed assessment of differential effects. 3.1.4. Acute Neurotoxicity For the acute neurotoxicity study, the test substance is administered as a single dose by gavage to rats that have been fasted overnight. The FOB and motor activity tests are conducted on four occasions: once prior to treatment, at the time of peak effects (within

84

Sheets

8 h following treatment), and again 7 and 14 d following treatment. Days 7 and 14 are included to examine the potential for causing effects that may persist or become apparent after a delay. The animals are also examined daily by cage-side observation and tissues are collected for microscopic examination 14 d following treatment. For the represented compounds, the time of peak effects in fasted animals ranges from 2–6 h following treatment. The most consistent finding at lower dose levels is evidence of decreased activity. At higher dose levels, tremors, impaired pupillary function (either dilated or pin-point pupils) and hypothermia are the most common effects observed. Finally, near-lethal dose levels produce an assortment of neurotoxic signs, including evidence of motor incoordination (e.g., an uncoordinated gait or impaired aerial righting), autonomic signs (e.g., lacrimation and urine staining), and CNS depression (e.g., markedly decreased motor activity and decreased response to stimuli). Deaths associated with treatment occur within 4–24 h following treatment. There is no evidence of neuropathology with any of these compounds. Clinical signs following an acute exposure resolve rather quickly in surviving animals, generally within 8–24 h following treatment. Lacrimal and urine stains are the principal effects that may persist for up to 4 d with some compounds. All findings in the FOB resolve by d 7, which is the first test occasion that follows the d 0 (time of peak effects) time point. Thus, these compounds produce a variety of neurotoxic signs following acute exposure, with complete recovery within several hours or a few days following treatment. There are no persistent neurotoxic effects or effects that are delayed in onset. Certain findings (e.g., tremors, impaired pupillary function, and hypothermia) that are evident at sublethal dose levels are likely associated with nicotinic stimulation. It is less clear whether certain other signs, especially those evident at lethal doses, are associated with nicotinic stimulation or represent nonspecific toxic effects. Evidence of decreased activity at low doses may represent a relatively specific effect, since it occurs at dose levels that produce no other evidence of toxicity. On the other hand, decreased activity is a common expression of toxicity that is evident with an assortment of treatments. 3.1.5. Subchronic Neurotoxicity

For these studies, the technical grade active ingredient is mixed in the diet and provided for ad libitum consumption for a period of 13 wk. A range of dietary concentrations is tested, including one that produces evidence of toxicity (e.g., clinical signs or a minimum 10% decrease in body weight gain) and a level expected to produce no evidence of toxicity, to establish a NOEL. The animals are assessed for evidence of toxicity or neurotoxicity using a FOB and an automated test of motor activity during wk 0 (pretreatment), 4, 8, and 13 of continued exposure. These time points are included to follow the progression of the animal’s condition; for example, to determine whether there is evidence of cumulative toxicity. Detailed cage-side observations are also performed weekly and, at study termination, animals are perfused and neural tissues are collected for microscopic examination. The results from these studies indicate that sustained dietary exposure produces little or no evidence of neurotoxicity. In general, there are no effects evident by cage-side observation, the FOB, or automated tests of motor activity. Furthermore, there is no


85

evidence of neuropathology. Each of these compounds produces a modest decrease in food consumption and an associated decrease in body weight gain at higher dietary concentrations. The reduced body weight gain generally limits the possibility of testing higher dietary concentrations. The principal effects seen with acute exposure (e.g., decreased activity, tremors, and hypothermia) are absent with sustained dietary exposure. The results from these studies indicate that cumulative toxicity is not a concern with the neonicotinoid insecticides. This outcome is consistent with their rapid metabolism and excretion in rats (8,21,22). 3.2. Developmental Neurotoxicity in Neonates 3.2.1. Published Studies There are no reports in the published literature of studies that have investigated the potential for a neonicotinoid insecticide to produce developmental neurotoxicity. 3.2.2. Unpublished Proprietary Studies Developmental toxicity (teratology) and multi-generation reproduction toxicology studies are required to support the registration of pesticides for food uses. However, these studies generally provide only a cursory indication of neurotoxic potential. In the teratology study, the offspring are delivered by cesarean section on gestation d 20 and the brain is examined for gross morphologic changes, without routine microscopic examination. For the multi-generation study, the effects of a treatment on the offspring are examined through observations for clinical signs and measurements of body weight gain and food consumption. Brains are weighed and undergo microscopic examination but other neural tissues are not routinely examined. A more rigorous assessment of a compound’s potential to produce effects on the developing nervous system is provided by studies that are conducted according to the US EPA guideline for a developmental neurotoxicity study (US EPA, OPPTS 870.6300). Included in this study design is a standardized battery of behavioral tests (e.g., automated tests of learning and memory, auditory startle habituation, and motor activity) and neuropathology. These tests are administered at specific ages, including around the time of weaning and in the young adult. An assortment of neural tissues is collected, and brain tissues undergo a qualitative microscopic examination for lesions, as well as a quantitative (i.e., morphometric) analysis. Studies with several of these compounds have recently been conducted or are currently in progress, but none is yet reported. The publication of these results will add substantially to the existing knowledge base for the neonicotinoid insecticides.

3.3. Antidotes No specific antidotes are known. In an acute poisoning case, the recommended treatment is symptomatic. In particular, it is important to support respiration if signs of paralysis appear and to monitor blood pressure and pulse rate, since bradycardia and hypotonia are possible. Since these compounds do not inhibit cholinesterase activity, treatment with a reactivating oxime (e.g., pralidoxime) is not indicated. Furthermore, symptoms of poisoning may be mediated by either stimulation or inhibition of nicotinic activity, or by other possible mechanisms. Therefore, treatment with a nicotinic antagonist might be either ineffective or contraindicated.

86

Sheets

4. CONCLUSIONS AND FUTURE DIRECTIONS The Food Quality Protection Act of 1996 includes a provision that alters the approach used in risk assessments. This change involves consideration of collective exposure to compounds that act via a common mechanism of toxicity. Since the mode of action for the neonicotinoids in vertebrate species differs from that of the organophosphorus and carbamate (both anticholinesterase) and pyrethroid (sodium channels) insecticides, they should be considered separately from these other classes of insecticides in risk assessment. While neurotoxicity is evident following acute exposure to the neonicotinoid insecticides, it should be appreciated that all effects are reversible and occur only at relatively high levels of exposure. In fact, representative neonicotinoid insecticides have among the highest margins of safety (24), because of the combination of low exposure and low toxicity in mammals. This makes the neonicotinoids among the safest insecticides that are currently available. Because of their favorable properties, the continued development of these compounds as commercial insect control products will likely continue in the future. ACKNOWLEDGMENT The assistance of Dr. Helen Cunny (Aventis), Dr. James Stevens (Novartis), and Mr. Koji Mizuta (Takeda Chemical Industries, Ltd.) in providing valuable data from their respective laboratories and useful comments on this manuscript is most gratefully acknowledged. REFERENCES 1. Soloway, S. B., Henry, A. C., Kollmeyer, W. D., Padgett, W. M., Powell, J. E., Roman, S. A., et al. (1978) Nitromethylene insecticides. Adv. Pestic. Sci. 4, 206–217. 2. Schroeder, M. E. and Flattum R. F. (1984) The mode of action and neurotoxic properties of the nitromethylene heterocycle insecticides. Pest. Biochem. Physiol. 22, 148–160. 3. Shiokawa, K., Tsuboi, S., Kagabu, S., and Moriya, K. (1986) Jpn. Kokai Tokkyo Koho JP 61-267575. 4. Takahashi, H., Mitsui, J., Takakusa, N., Matsuda, M., Yoneda, H., Suzuki, J., Ishimitsu, K., and Kishmoto, T. (1992) NI-25, a new type of systemic and broad spectrum insecticide. Brighton Crop Protection Conferences - Pest and Diseases 1, 89–96. 5. Minamida, I., Iwanaga, K., Tabuchi, T., Aoki, I., Fusaka, T., Ishizuka, H., and Okauchi, T. (1993) Synthesis and insecticidal activity of acyclic nitroethene compounds containing a heteroarylmethylamino group. J. Pesticide Sci. 18, 41. 6. Tomizawa, M. and Yamamoto, I. (1993) Structure-activity relationships of nicotinoids and Imidacloprid analogs. J. Pesticide Sci. 18, 91–98. 7. Liu, M.-Y., Lanford, J., and Casida, J. E. (1993) Relevance of [3H]Imidacloprid binding site in house fly head acetylcholine receptor to insecticidal activity of 2-nitromethyleneand 2-nitroimino-imidazolidines. Pestic. Biochem. Physiol. 46, 200–206. 8. Maienfisch, P., Brandl, F., Kobel, W., Rindlisbacher, A., and Senn, R. (1999) CGA 293’343: A novel, broad-spectrum neonicotinoid insecticide, in Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor (Yamamoto, I. and Casida, J., eds.), SpringerVerlag, Tokyo, Japan, pp. 177–209. 9. Chao, S. L. and Casida, J. E. (1997) Interaction of imidacloprid metabolites and analogs with the nicotinic acetylcholine receptor of mouse brain in relation to toxicity. Pest. Biochem. Physiol. 58, 77–88.


87

10. Tomizawa, M., Latli, B., and Casida, J. E. (1999). Structure and function of insect nicotinic acetylcholine receptors studied with nicotinoid insecticide affinity probes, in Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor (Yamamoto, I. and Casida, J., eds.), Springer-Verlag, Tokyo, Japan, pp. 271–292. 11. Nagata, K., Song, J. H., Shono, T., and Narahashi, T. (1998). Modulation of the neuronal nicotinic acetylcholine receptor-channel by the nitromethylene heterocycle Imidacloprid. J. Pharmacol. Exp. Ther. 285, 731–738. 12. Tomizawa, M. and Casida, J. E. (1999) Minor structural changes in nicotinoid insecticides confer differential subtype selectivity for mammalian nicotinic acetylcholine receptors. Br. J. Pharmacol. 127, 115–122. 13. Liu, M.-Y. and Casida, J.E. (1993) High affinity binding of [3H]Imidacloprid in the insect acetylcholine receptor. Pestic. Biochem. Physiol. 46, 40–46. 14. Tomizawa, M., Otsuka, H., Miyamoto, T., and Yamamoto, I. (1995) Pharmacological effects of Imidacloprid and its related compounds on the nicotinic acetylcholine receptor with its ion channel from the Torpedo electric organ. J. Pesticide Sci. 20, 49–56. 15. Yamamoto, I., Yabuta, G., Tomizawa, M., Saito, T., Miyamoto, T., and Kagabu, S. (1995) Molecular mechanism for selective toxicity of nicotinoids and neonicotinoids. J. Pesticide Sci. 20, 33–40. 16. Zwart, R., Oortgiesen, M., and Vijverberg, H. P. M. (1992). The nitromethylene heterocycle 1-(pyridin-3-yl-methyl)-2-nitromethylene-imidazolidine distinguishes mammalian from insect nicotinic receptor subtypes. Eur. J. Pharmacol. 228, 165–169. 17. Zwart, R., Oortgiesen, M., and Vijverberg, H. P. M. (1994). Nitromethylene heterocycles: selective agonists of nicotinic receptors in locust neurons compared to mouse N1E-115 and BC3H1 cells. Pestic. Biochem. Physiol. 48, 202–213. 18. Yamamoto, I., Tomizawa, M., Saito, T., Miyamoto, T., Walcott, E. C., and Sumikawa, K. (1998) Structural factors contributing to insecticidal and selective actions of neonicotinoids. Arch. Insect Biochem. Physiol. 37, 24–32. 19. Matsuda, K., Buckingham, S. D., Freeman, J. C., Squire, M. D., Baylis, H. A., and Sattelle, D. B. (1998) Effects of the alpha subunit on Imidacloprid sensitivity of recombinant nicotinic acetylcholine receptors. Br. J. Pharmacol. 123, 518–524. 20. Yamada, T., Takashi, H., and Hatano, R. (1999). A novel insecticide, Acetamiprid, in Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor (Yamamoto, I. and Casida, J., eds.), Springer-Verlag, Tokyo, Japan, pp. 149–176. 21. Akayama, A. and Minamida, I. (1999) Discovery of a new systemic insecticide, nitenpyram and its insecticidal properties, in Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor (Yamamoto, I. and Casida, J., eds.), Springer-Verlag, Tokyo, Japan, pp. 127–148. 22. Thyssen, J. and Machemer, L. (1999) Imidacloprid: toxicology and metabolism, in Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor (Yamamoto, I. and Casida, J., eds.), Springer-Verlag, Tokyo, Japan, pp. 213–222. 23. Moser, V. C. (1989). Screening approaches to neurotoxicity: A functional observational battery. J. Am. Coll. Toxicol. 8, 85–93. 24. Leicht, W. (1993). Imidacloprid: a chloronicotinyl insecticide. Pestic. Outlook 4, 17–21.

88

Sheets

Miscellaneous Pesticides

89

I Pesticides C. Miscellaneous Pesticides with Action on the Nervous System

90

Blodgett et al.


91

7 Miscellaneous Pesticides with Action on the Nervous System Dennis Blodgett, Marion Ehrich, and Jeffrey R. Bloomquist

1. INTRODUCTION Not all pesticides in use today belong to the classes discussed in previous chapters. A wide variety of substances currently marketed can cause clinical effects that suggest the products are neurotoxic. Mechanisms of action, especially with regard to the nervous system, are better defined for some than for others. This chapter presents a summary of various pesticides, including insecticides, rodenticides, repellants, fungicides, and fumigants that may have neurotoxic effects. If information is available, each is discussed in terms of: (1) an introduction with a historical overview; (2) clinical evidence of neurotoxicity; (3) mechanisms of action and recommended treatments, when known; and (4) future directions. Table 1 gives an overview of the miscellaneous pesticides discussed in the chapter along with several other miscellaneous pesticides with neurotoxic potential. 2. AVICIDES, FUMIGANTS, AND HERBICIDES 2.1 4-Aminopyridine (Avitrol®) 4-Aminopyridine (Fig. 1) has been marketed as an avicide, Avitrol®, since 1963 (2). It is used as a corn-based bait to control pest bird populations like blackbirds, sparrows, seagulls, starlings, and pigeons. Baits range from 0.5–3% as the hydrochloride salt and usually are not colored with a dye (12). 4-Aminopyridine also comes as 25 and 50% concentrated powders (2). Poisonings with accidental exposure to the bait have been reported in both dogs and horses (12). Relay poisoning by ingestion of dead birds has not been reported, but may be possible, especially in cats consuming a pigeon with a crop full of baited corn. Lethal doses in birds and mammals are typically in the range of 1–20 mg/kg (13). Human poisonings may result from consumption of as little as 60 mg (2). In addition to being an avicide, 4-aminopyridine is also an investigational drug for the treatment of some neurologic disorders including multiple sclerosis (MS), myasthenia gravis, Eaton-Lambert syndrome, and botulism (14). Oral therapeutic regimens range from 7.5–200 mg/d in single or divided doses.


91

92

Blodgett et al.

Table 1 Miscellaneous Pesticides and Their Action on the Nervous System Pesticide class

Specific toxicant

Mechanism of action

Avicide

4-Aminopyridine (Avitrol®)

(1,2)

Fumigant

Carbon disulfide

Fungicide

Dinitrophenol; dinitro-derivatives

Blockage of K+ channels; increased cholinergic transmission Polymerization of brain tubulin via carbon disulfide protein crosslinking Uncouplers of oxidative phosphorylation

Herbicides

2,4-D; other chlorophenoxy herbicides, Endothall; Paraquat; diquat Amitraz D-Limonene; linalool; citrus oil extracts Rotenone

Sarcolemma disruption → myotonia

(5)

Unknown Unknown in CNS α-2 Adrenergic agonist Unknown

— — (6) —

Complex I inhibitor of mitochondrial respiratory chain; reduction of dopamine synthesis in ventral midbrain Decreased neurotransmitter (5-hydroxy-tryptamine, 5-hydroxyindoleacetic acid, GABA and norepinephrine) concentrations in brain Unknown

(7,8)

Insecticides

Molluscicide

Metaldehyde

Repellant (insect) Rodenticides

DEET (N-N-diethyltoluamide) Bromethalin Fluoroacetate (1080) Strychnine Zinc phosphide; aluminum phosphide

Uncoupler of oxidative phosphorylation → brain edema Inhibition of cellular respiration (aconitase) Competitive antagonism of glycine at Renshaw cells Unknown in CNS

Reference

(3)

(4)

(9)

—

(10) (4) (11) —

Dogs and horses with 4-aminopyridine toxicosis exhibit hyperexcitability, salivation, muscle tremors, and clonic-tonic seizures (2,12). Additionally, cardiac arrhythmias and tachycardia may occur. Humans have burning of the throat, abdominal pain, vomiting, weakness, severe perspiration, acidosis, clonic-tonic seizures, and respiratory arrest (2,14). Early central nervous system (CNS) signs in humans include


93

Fig. 1. Chemical structures of miscellaneous pesticides with neurological effects.

disorientation, combativeness, and shouting incoherently (2). Avitrol is used in birds because of its ability to cause vocalization and thereby frighten away other birds feeding in the area. Birds that ingest enough Avitrol also exhibit seizures and death. Birds may take flight, seizure in flight, and crash to the ground. 4-Aminopyridine can block various potassium channels with different state dependences (1). The blockade or binding can occur either in the activated state or the closed state of the channels. Blockage of potassium channels augments neuronal conduction

94

Blodgett et al.

by hindering repolarization of the neuron following an action potential and allowing increased sodium and calcium ion influx (14). In addition to blockage of potassium channels, 4-aminopyridine also enhances cholinergic transmission at neuromuscular junctions and other synapses (2). Seizures are treated with diazepam and phenytoin (14). Avitrol continues to be a popular avicide because of its ability to frighten birds away while killing only a small percentage. 4-Aminopyridine may become a therapeutic drug for a variety of neurologic disorders. Because of its high toxicity for mammals and its low therapeutic index in humans, poisonings will continue to occur. 2.2. Carbon Disulfide Carbon disulfide (Fig. 1) is an industrial solvent and a heavily used fumigant for control of insect pests in stored products, such as grain. It has low toxicity to mammals, with an acute LD50 of 1545 mg/kg by intraperitoneal administration in 20-d-old rats (15). When mice are exposed by inhalation, carbon disulfide causes ataxia of hind limbs and a decrease in rearing and locomotor movement (16). Inhalation studies in rats found that carbon disulfide decreased conduction velocity and increased compound action potential amplitude in caudal tail nerve preparations (17). Similarly, neuromuscular integrity in rats decreased 50% in an acoustic startle paradigm after 12 wk of exposure to 500 ppm carbon disulfide (18). Pathological studies in exposed animals have demonstrated central and peripheral distal axonopathy with swollen axons and accumulation of neurofilament proteins within the axoplasm (16). Other studies have shown polymerization of brain tubulin via carbon disulfide protein crosslinking and that this action may be an important mechanism underlying axonal swelling (3). Similar pathology and protein cross linking occurs in animals following exposure to diethyldithiocarbamates structurally related to the fungicide (e.g., Maneb; Fig. 1)(19,20). These effects from diethyldithiocarbamate fungicides are mediated by carbon disulfide, liberated as a carbamate metabolite, in vivo. Carbon disulfide has been linked to environmentally induced parkinsonism in humans (21). Grain industry workers exposed to this material manifest a significant finger tremor of 5–7 Hz frequency, compared to aged-matched controls (22). This frequency is similar to the tremor frequency of idiopathic Parkinson’s disease (PD). Similarly, a self-selected group of granary workers were analyzed for neurological deficits, and these included cogwheel rigidity, decreased associated movements, and tremulousness, all to various extents (23). More detailed studies are needed to evaluate any link between carbon disulfide exposure and degeneration or dopamine loss in the nigrostriatal pathway in humans or animal models. 2.3. 2,4-D 2,4-D (2,4-dichlorophenoxyacetic acid; Fig. 1) is a common herbicide used both agriculturally and around the home for control of broad-leaf weeds. Other related herbicides with similar effects on plants and animals include MCPA (4-chloro-2methylphenoxy acetic acid), MCPP (2-(4-chloro-2-methyl phenoxy) propanoic acid), and dicamba (3,6-dichloro-2-methoxybenzoic acid) (24,25). Often, combinations of the above herbicides are used for broad-leaf weed control. Dogs are usually exposed to


95

open bottles of the herbicides or puddles of recently applied herbicide. Humans may mistake the container of herbicide for a beverage or may intentionally ingest the herbicide in suicide attempts. Acute oral LD50s for 2,4-D range from 300–700 mg/kg in most mammals (26). Although early deaths after massive doses of 2,4-D have been associated with ventricular fibrillation of the heart (26), myotonia is the main clinical sign associated with acute intoxications of these broad-leaf herbicides (24). In dogs, the major systems affected by these herbicides are gastrointestinal tract and skeletal muscles. Clinical signs include vomiting, diarrhea, ataxia, hypermetria, hind limb paresis, and paralysis. Skeletal muscle tone is rigid, consistent with myotonia. Convulsions are less likely, but sometimes occur in dogs with 2,4-D toxicosis. Clinical signs in humans include pharyngeal pain, lethargy, stupor, weakness, respiratory muscle paralysis, coma, convulsions, and possible cardiac arrhythmias (26,27). Coma is a fairly consistent finding in suicide attempts associated with chlorophenoxy herbicides. Myotonia induced by 2,4-D and related herbicides is thought to be associated with increased membrane resistance of the sarcolemma associated with decreased chloride conductance (5). The dysfunction is in the sarcolemma of the muscle and does not involve neural transmission or require an intact nerve transmission to the muscle (5,24). In addition to muscle disturbances, chlorophenoxyacetic acid herbicides cause disturbances in cerebral electrical activity (25,26). This may explain why coma is so common in 2,4-D toxicoses of humans. Chlorophenoxy herbicides continue to be some of the most widely used herbicides with many millions of pounds being applied each year (26). Because of their widespread use and availability, acute toxicoses from these compounds will continue to occur. 3. INSECTICIDES, MOLLUSCICIDES, AND REPELLANTS 3.1. Amitraz Amitraz (N’-(2,4-dimethylphenyl)-N-[[(2,4-dimethylphenyl)imino]methyl]-Nmethylmethanimidamide; Fig. 1) has been used to control mites and ticks on crops and domestic animals for the last 20 yr. Its acute toxicity to a variety of test species, including rodents, rabbits, dogs, and baboons, on single oral exposure, is reflected in LD50 values between 100 and 800 mg/kg (28). Sedation, bradycardia, mydriasis, hypothermia, tremors, ataxia, and increased susceptibility to electrically induced seizures were noted in laboratory animals and pets. Although case reports of amitraz toxicosis are few, effects in humans, most of whom were children, were similar, with drowsiness, loss of consciousness, respiratory depression, bradycardia, vomiting, hypothermia, and hypotension reported (28–30). Onset of clinical evidence of toxicosis occurred shortly after exposure and generally resolved within 2 d. Amitraz has been examined extensively for its neurotoxic effects in laboratory animals, including its effects on behavior, motor activity, operant behavior, and seizure threshold. At doses less than 50% of the LD50, CNS excitability, sensorimotor reactivity, and facilitated seizures were increased. Motor activity and performance on schedule-controlled tasks were decreased. Transient behavioral changes could be seen in offspring of exposed adults (31–34).

96

Blodgett et al.

Although in vitro experimentation suggested amitraz had effects on monamine oxidase, subsequent testing in laboratory animals indicated that heroic doses would be necessary to affect this enzyme in vivo. The specific alleviation of clinical signs by administration of α-2 antagonists such as yohimbine, and associated laboratory studies suggested that amitraz acts as a reversible α-2 agonist both in the central and in the peripheral nervous system (PNS) (6,30,32,35). No other formamidine insecticides have gained the popularity of amitraz. Use of amitraz is likely to continue, as it appears to have less toxicity than another formamidine insecticide, chlordimeform, which was found to induce tumors in rodents (28). 3.2. D-Limonene D-Limonene (1,8(9)-p-methadiene 1-methyl-4-isopropenyl-1-cyclohexene; Fig. 1) is a botanical insecticide extracted from citrus peels. It is effective against external pests of companion animals and has, therefore, been a popular component of a variety of over-the-counter pet products for many years. D-Limonene is considered relatively safe (LD50 >5 g/kg), and no reports of human poisonings have appeared in the literature (36). D-Limonene toxicosis has, however, been reported in pets. Cats appear particularly sensitive to overzealous dermal application of this insecticide. Signs of toxicosis include hypersalivation, muscle tremors, and ataxia. Deaths have occurred, but most cats with clinical signs recovered within hours (36). The mechanism(s) responsible for the clinical signs is unknown. Most cats respond to dermal baths. D-Limonene is currently under investigation as a potential antineoplastic agent. It appears to be well tolerated in humans, but capable of causing tumors in kidneys of male rats (37,38). 3.3. Rotenone Rotenone toxicity has been reviewed by Hayes (7), and much of the early work on this compound, summarized below, is taken from this review. Rotenone (Fig. 1) is harvested from the roots of plants belonging to the Derris and Lonchocarpus genera and is used as a piscicide and an insecticide for insect control on vegetables. Rotenone has a high acute toxicity to rats, with reported oral LD50s of 25–132 mg/kg, depending on the formulation particle size and manner of dispersion. Lethal poisoning in animals includes an initial respiratory stimulation, followed by depression, along with neurological signs of incoordination, muscle tremors, and tonic and clonic convulsions. Oral exposure in humans, usually associated with suicide, results in respiratory collapse with feeble pulse and dilated pupils. These signs are consistent with the well-established action of rotenone as a complex I inhibitor of the mitochondrial respiratory chain (7). Recent results have observed a specific neurotoxic effect of rotenone that has potential as a model of PD. This disease is characterized by degeneration of neurons in the nigrostriatal pathway and attendant loss of the neurotransmitter dopamine (39). Accordingly, specific lesions in the striatum and globus pallidus result from continuous intravenous infusion of rotenone to rats at a rate of 10–18 mg/kg/d over a 7–9 d period (40). However, no lesions were observed in the substantia nigra itself. These studies were essentially confirmed by subsequent work, where it was also reported that rotenone-infused rats became progressively bradykinetic and rigid, signs resembling PD (41). Other studies have shown that midbrain dopamine neurons are more sensitive


97

to rotenone in transmitter release assays than GABAergic neurons (8). Greater sensitivity to rotenone of dopamine-containing neurons of the ventral midbrain may result from the high metabolic demand of a Na+/K+ ATPase that drives postburst hyperpolarizations in these cells (42). It follows that a reduction of ATP synthesis would block the Na+/K+ ATPase pump, resulting in disruption of normal electrical activity and Na+ homeostasis of these cells, a process which is known to be neurotoxic (43). 3.4 Metaldehyde Metaldehyde (Fig. 1) is a cyclic polymer of acetaldehyde, used in bran-based baits for control of garden slugs. Metaldehyde is moderately toxic to rodents, having a mouse oral toxicity of 200 mg/kg, and rat oral toxicities of 227–690 mg/kg (9). The minimum oral lethal dose for dogs and humans is reported to be 100 mg/kg (44). Poisoning in dogs manifests as intense whole body tremor, and extreme hyperthermia, giving way to convulsions and coma. In humans, metaldehyde causes tremors, convulsions, and memory loss indicative of an action on the CNS (9). Despite the clear involvement of the nervous system in metaldehyde toxicosis, its mode of action is not well-defined. Initial hypotheses concerning its mode of action centered on the idea that metabolic liberation of acetaldehyde from the parent compound was involved. However, the LD50 of acetaldehyde in rats is reported to be 1930 mg/kg, a value much higher than that of metaldehyde (45). Moreover, pharmacokinetic studies in dogs found a correlation between toxicity and circulating levels of metaldehyde, but not acetaldehyde, in plasma (46). Similar studies in mice found 2000 mg/kg). However, DEET can be absorbed through the skin and by inhalation, with a few documented cases of death due to extensive skin absorption (47). Poisoning of children, with symptoms indicative of neurotoxicity, has been reported. These included seizures, restlessness, mood changes, behavioral changes, involuntary movements, loss of reflexes, ataxia, and convulsions (28,48). Females tend to be more sensitive to the effects of DEET than males, probably because of reduced metabolism (47). Of more recent concern is the suggestion that DEET, as part of the chemical combinations to which Gulf War veterans were exposed, contributed to neurological and neuropsychological abnormalities seen in some individuals (49,50). In the studies by Abou-Donia et al. (49), DEET increased the toxicity of other materials when tested

98

Blodgett et al.

in combination with chlorpyrifos and pyridostigmine bromide, presumably through effects on metabolism or by altering the properties of the blood-brain barrier (BBB). Neurotoxicity has been evaluated in rats, with mixed results. No treatment-related effects were reported using a Functional Observational Battery (FOB), swim maze, passive avoidance, or neuropathology following acute or multi-generational oral exposures, but decreased performance on the rotorod and increased locomotor activity were noted (51,52). High doses of DEET (1–3 g/kg) in rats caused ataxia, prostration, and electrophysiological signs of prolonged suppressed seizure activity (53). Associated pathological effects included a spongiform myelinopathy and neuronal cytoplasmic clefts concentrated in the cerebellar roof nuclei. 4. RODENTICIDES 4.1. Bromethalin Bromethalin (N-methyl-2,4-dinitro-N-[2,4,6-tribromophenyl]-6-[trifluoromethyl] benzeneamine; Fig. 1) was discovered in the mid–1970s in a search for new rodenticides because of resistance in the rodent population to warfarin-based products (10). It has been marketed since 1985 as a 0.01% bait formulated as tan or green, grain-based pellets (54). The reported LD50s for technical grade bromethalin in cats, rats, and dogs are 1.8, 2.0, and 4.7 mg/kg, respectively (10). The minimal lethal doses of bromethalin bait in the dog and cat are 25 and 4.5 gm/kg, respectively (54). Relay poisoning in dogs and cats from eating poisoned rodents has not been demonstrated. Both cats and dogs have been poisoned by access to bromethalin rodent baits. Clinical signs seen in dogs are very much dose-related (54). Dogs that ingest doses at or above the LD50 develop clinical signs within 24 h. Clinical signs at this higher dose in dogs consist of muscle tremors, hyperthermia, hyperesthesia, and focal motor or generalized seizures. Additional clinical signs could include loss of bark, coma, nystagmus, and anisocoria. A more common syndrome in dogs is associated with ingestion of toxic amounts but at less than the LD50 dose. Onset of clinical signs at the lower dose is delayed for several days. Clinical signs with the lower-dose syndrome include hindlimb ataxia, paresis, paralysis, depression, muscle tremors, and anisocoria. In cats, the syndrome more closely mimics that of the lower-dose syndrome of dogs (54). Clinical signs in cats typically develop within 3–7 d of ingestion of the bait. The most common signs in cats are ataxia, vocalization, hindlimb paralysis, depression, semicoma, coma, muscle tremors, and anisocoria. Animals with mild signs (e.g., ataxia, depression) typically recover in 1–2 wk (54). Animals with more severe signs usually die. Bromethalin is metabolized in the body to desmethylbromethalin (Fig. 1), which is a potent uncoupler of oxidative phosphorylation (10). The subsequent lack of ATP production by mitochondria of CNS tissue is theorized to result in decreased function of sodium-potassium ATPase in ion channel pumps. Loss of pump activity is believed to cause cerebral and spinal cord edema, decreased nerve-impulse transmission, and ultimately to lead to respiratory arrest and death. Edema of the white matter of the CNS is very evident in histologic sections and appears as spongiosis in myelin sheaths. Bromethalin also induces lipid peroxidation in the brain (54). Therapy to reduce CNS edema and increased cerebral spinal fluid pressure has been attempted with mixed success. Osmotic diuretics (e.g., mannitol and urea) and corticosteroids (e.g., dexametha-


99

sone) were useful in treating rats with bromethalin toxicosis (10). However, mannitol and dexamethasone were ineffective in treating bromethalin toxicosis in dogs and cats (54). Although bromethalin baits were developed initially to circumvent resistance of the rodent population to warfarin-based rodenticides, further advances have occurred in the development of anticoagulant-based rodenticides. New “second generation” anticoagulants have been developed that rodents are not resistant to at this point. These anticoagulants (e.g., brodifacoum, bromadiolone) are more potent and longer-lasting than warfarin. Their success in the rodenticide market has overshadowed the market share of bromethalin-based products and limited the use of bromethalin-based rodenticides. 4.2 Strychnine Strychnine (Fig. 1) is the principle alkaloid in the seeds of Strychnos nux-vomica, a tree native to India (55). The dried seeds contain 1–2% strychnine (56). Brucine is another alkaloid with similar properties, but less potency, that is found in approximately the same concentration as strychnine in the seeds (56). Nux vomica is another name sometimes used for strychnine. Strychnine has been used for many years as a stomachic, stimulant, tonic, and rodenticide (56). Strychnine was introduced into Germany as a rodenticide in the sixteenth century (55). Both veterinary and human practitioners have prescribed strychnine in previous years for multiple diseases. More recently, strychnine has been used to adulterate illicit drugs like heroin or cocaine and has been involved in several human poisonings in this regard. Strychnine, as a rodenticide, is often mixed with cereal grains or pellets. These baits typically range from 0.3– 1% strychnine and are often colored with a dye to denote contamination. Although a few poisonings occur in humans each year with accidental exposures or suicidal attempts, most poisonings are in dogs that gain access to the baits or are poisoned maliciously. Strychnine toxicoses are rare in cats because of the bitter taste of the alkaloid. Lethal doses of strychnine in most species range from 0.5–5 mg/kg. The neurologic syndrome seen with strychnine appears very similar in animals or humans (11). Usually within 15 min to 2 h after consumption of a toxic dose, muscle twitching or tightness appears in the face or neck. The victim appears anxious or nervous. In dogs, the face takes on the appearance of grinning and the body may appear to be in a sawhorse-like stance. Seizures appear abruptly thereafter and may be elicited by touch, sound, or light as the patient is extremely hyperesthetic. Typically, the seizures are characterized as tonic or tetanic with severe contracture of all muscle groups. Since extensor muscles are stronger than flexor muscles, extensor rigidity without flexion is the rule. In humans, the tetanic contaction of the body is so severe that the back is arched in such a way that the person is resting on the heels and back of the head (11). This opisthotonus also occurs in animals. The first seizure usually subsides within a few minutes, but without treatment each succeeding seizure lasts longer and the interval between seizures becomes shorter. Death results from anoxia because of paralysis of intercostal and diaphragm muscles. Strychnine has the ability to bind tightly to glycine receptors in the CNS with an affinity 300× greater than glycine (11). Strychnine competitively and reversibly

100

Blodgett et al.

antagonizes glycine at the postsynaptic neuronal sites in the Renshaw cells, the internuncial cells of the spinal cord and medulla. These Renshaw cells send out inhibitory signals to the somatic motor neurons. The lack of neuronal inhibition from glycine produces increased stimulation of all somatic motor neurons. Convulsions are typically treated in humans with diazepam. If diazepam is ineffective, sodium pentobarbital, shorter-acting barbiturates or mephenisin are sometimes used alternatively. In dog poisonings with strychnine, diazepam typically will not control convulsions for very long and sodium pentobarbital, sometimes in combination with methocarbamol, is used. Mephenisin and methocarbamol may more directly counteract strychnine poisoning than barbiturates (11,56). However, a barbiturate is also needed in conjunction with these muscle relaxants in order to control anxiety and to ameliorate the influence of sensory stimuli (56). Artificial respiration may be needed if hypoxia persists. The availability and use of strychnine as a rodenticide and tonic has greatly diminished over the years. Strychnine rodenticide baits are currently registered in the United States for “restricted use” by certified applicators. Occasional cases in both humans and dogs continue to appear, but the availability and use of other types of rodenticides have lessened the frequency of poisonings by strychnine. REFERENCES 1. Tseng, G. N., Jiang, M., and Yao, J. A. (1996) Reverse use dependence of Kv4.2 blockade by 4-aminopyridine. J. Pharmacol. Exp. Ther. 279, 865–876. 2. Spyker, D. A., Lynch, C., Shabanowitz, J., and Sinn, J. (1980) Poisoning with 4-aminopyridine: report of three cases. Clin. Toxicol. 16, 487–497. 3. Gupta, R. P. and Abou-Donia, M. B. (1997) Acrylamide and carbon disulfide treatments increase the rate of rat brain tubulin polymerization. Mol. Chem. Neuropathol. 30, 223–237. 4. Ellenhorn, M. J. and Barceloux, D. G. (1988) Pesticides, in Medical Toxicology: Diagnosis and Treatment of Human Poisoning, Elsevier Science, New York, NY, pp. 1069–1108. 5. Iyer, V., Whiting, M., and Fenichel, G. (1977) Neural influence on experimental myotonia. Neurology 27, 73–76. 6. Costa, L. G., Olibet, G., Wu, D. S., and Murphy, S. D. (1989) Acute and chronic effects of the pesticide amitraz on alpha 2-adreneoreceptors in mouse brain. Toxicol. Lett. 47, 135–143. 7. Hayes, W. J. (1982) Pesticides Studied in Man. Williams & Wilkins, Baltimore, pp. 81–86. 8. Marey-Semper, I., Gelman, M., and Levi-Strauss, M. (1993) The high sensitivity to rotenone of striatal dopamine uptake suggests the existence of a constitutive metabolic deficiency in dopamine neurons form the substantia nigra. Eur. J. Neurosci. 5, 1029–1034. 9. Von Burg, R. and Stout, T. (1991) Toxicology update metaldehyde. J. Appl. Toxicol. 11, 377–378. 10. van Lier, R. B. L. and Cherry, L. D. (1988) The toxicity and mechanism of action of bromethalin: a new single-feeding rodenticide. Fundam. Appl. Toxicol. 11, 664–672. 11. Ray, D. E. (1991) Pesticides derived from plants and other organisms. Chapter 13, in Handbook of Pesticide Toxicology, vol. 2 (Hayes, Jr., W. J. and Laws, Jr., E. R., eds.), Academic Press, San Diego, CA, pp. 615–619. 12. Ray, A. C., Dwyer, J. N., Fambro, G. W., and Reagor, J. C. (1978) Clinical signs and chemical confirmation of 4-aminopyridine poisoning in horses. Am. J. Vet. Res. 39, 329–331. 13. Schafer, E. W., Brunton, R. B., and Cunningham, D. J. (1973) A summary of the acute toxicity of 4-aminopyridine to birds and mammals. Toxicol. Appl. Pharmacol. 26, 532–538.


101

14. Pickett, T. A. and Enns, R. (1996) Atypical presentation of 4-aminopyridine overdose. Ann. Emerg. Med. 27, 382–384. 15. Green, E. C. and Hunter, A. (1985) Toxicity of carbon disulfide in developing rats: LD50 values and effects on the hepatic mixed-function oxidase enzyme system. Toxicol. Appl. Pharmacol. 78, 130–138. 16. Sills, R. C., Valentine, W. M., Moser, V., Graham, D. G., and Morgan, D. L. (2000) Characterization of carbon disulfide neurotoxicity in C57BL6 mice: behavioral, morphologic, and molecular effects. Toxicol. Pathol. 28, 142–148. 17. Herr, D. W., Vo, K. T., Morgan, D. L., and Sills, R. C. (1998) Carbon disulfide neurotoxicity in rats: VI. Electrophysiological examination of caudal tail nerve compound action potentials and nerve conduction velocity. Neurotoxicology 19, 129–146. 18. Clerici, W. J. and Fechter, L. D. (1991) Effects of chronic carbon disulfide inhalation on sensory and motor function in the rat. Neurotoxicol. Teratol. 13, 249–255. 19. Valentine, W. M., Amarnath, V., Amarnath, K., Rimmele, F., and Graham, D. G. (1995) Carbon disulfide mediated protein cross-linking by N,N-diethyldithiocarbamate. Chem. Res. Toxicol. 8, 96–102. 20. Johnson, D. J., Graham, D. G., Amarnath, V., Amarnath, K., and Valentine, W. M. (1998) Release of carbon disulfide is a contributing mechanism in the axonopathy produced by N,N-diethyldithiocarbamate. Toxicol. Appl. Pharmacol. 148, 288–296. 21. Tanner, C. M. (1992) Occupational and environmental causes of parkinsonism. Occup. Med. 7, 503–513. 22. Chapman, L. J., Sauter, S. L., Henning, R. A., Levine, R. L., Matthews, C. G., and Peters, H. A. (1991) Finger tremor after carbon disulfide-based pesticide exposures. Arch. Neurol. 48, 866–870. 23. Peters, H. A., Levine, R. L., Matthews, C. G., and Chapman, L. J. (1988) Extrapyramidal and other neurologic manifestations associated with carbon disulfide fumigant exposure. Arch. Neurol. 45, 537–540. 24. Beasley, V. R., Arnold, K. E., Lovell, R. A., and Parker, A. J. (1991) 2,4-D toxicosis 1: a pilot study of 2,4-dichlorophenoxyacetic acid- and dicamba-induced myotonia in experimental dogs. Vet. Hum. Toxicol. 33, 435–440. 25. Elo, H. A. and Ylitalo, P. (1979) Distribution of 2-methly–4-chlorophenoxyacetic acid and 2,4-dichlorophenoxyacetic acid in male rats: evidence for the involvement of the central nervous system in their toxicity. Toxicol. Appl. Pharmacol. 51, 439–446. 26. Gosselin, R. E., Smith, R. P., and Hodge, H. C. (1984) 2,4-D, in Clinical Toxicology of Commercial Products, 5th ed. Williams & Wilkins, Baltimore, MD, pp. III:130–134. 27. Durakovic, Z., Durakovic, A., Durakovic, S., and Ivanovic, D. (1992) Poisoning with 2,4dichlorophenoxyacetic acid treated by hemodialysis. Arch. Toxicol. 66, 518–521. 28. Knowles, C.O. (1991) Miscellaneous pesticides, in Handbook of Pesticide Toxicology, vol. 3 (Hayes, Jr., W. J. and Laws, Jr., E. R., eds.), Academic Press, San Diego, CA, pp. 1471–1526. 29. Aydin, K., Kurtoglu, S., Poyrazoglu, M. H., Uzum, K., Ustunbas, H. B., and Hallac, I. K. (1997) Amitraz poisoning in children: clinical and laboratory findings of eight cases. Hum. Exp. Toxicol. 16, 680–682. 30. Hugnet, C., Buronfosse, F., Pineau, X., Cadore, J. L., Lorgue, G., and Berny, P. J. (1996) Toxicity and kinetics of amitraz in dogs. Am. J. Vet. Res. 57, 1506–1510. 31. Gilbert, M. E. (1988) Formamidine pesticides enhance susceptibility to kindled seizures in amygdala and hippocampus of the rat. Neurotoxicol. Teratol. 10, 221–227. 32. Moser, V. C. and MacPhail, R. C. (1989) Investigations of amitraz neurotoxicity in rats. III. Effects on motor activity and inhibition of monamine oxidase. Fundam. Appl. Toxicol. 12, 12–22. 33. Moser, V. C. (1991) Investigations of amitraz neurotoxicity in rats. IV. Assessment of toxicity syndrome using a functional observational battery. Fundam. Appl. Toxicol. 17, 7–16.

102

Blodgett et al.

34. Palermo-Neto, J., Florio, J. C., and Sakate, M. (1994) Developmental and behavioral effects of prenatal amitraz exposure in rats. Neurotoxicol. Teratol. 16, 65–70. 35. Schaffer, D. D., Hsu, W. H., and Hopper, D. L. (1990) The effects of yohimbine and four other antagonists on amitraz-induced depression of shuttle avoidance responses in dogs. Toxicol. Appl. Pharmacol. 104, 543–547. 36. Hooser, S. B. (1990) D-limonene, linalool, and crude citrus oil extracts, in Toxicology of Selected Pesticides, Drugs, and Chemicals. The Veterinary Clinics of North America: Small Animal Practice (Beasley, V. R., ed.). W. B. Saunders Co., Philadelphia, PA, pp. 383–391. 37. Vigushin, D. M., Poon, G. K., Boddy, A., English, J., Halbert, G. W., Pagonis, C., et al. (1998) Phase I and pharmacokinetic study of D-limonene in patients with advanced cancer. Cancer Chemother. Pharmacol. 42, 111–117. 38. Whysner, J. and Williams, G. M. (1996) D-Limonene mechanistic data and risk assessment: absolute species-specific cytotoxicity, enhanced cell proliferation, and tumor promotion. Pharmacol. Ther. 71, 127–136. 39. Bowman, W. C. and Rand, M. J. (1980) Textbook of Pharmacology, 2nd ed. Blackwell, Oxford, UK, pp. 18.17–18.25. 40. Ferrante, R., Schulz, J., Kowall, N., and Beal, M. (1997) Systemic administration of rotenone produces selective damage in the striatum and globus pallidus, but not the substantia nigra. Brain Res. 753, 157–162. 41. Friedrich, M. J. (1999) Pesticide study aids Parkinson research. JAMA 282, 2200. 42. Johnson, S. W., Seutin, V., and North, A. R. (1992) Burst firing in dopamine neurons induced by N-methyl-D-aspartate: role of electrogenic sodium pump. Science 258, 665–667. 43. Lees, G. T. (1991) Inhibition of sodium-potassium-ATPase: a potentially ubiquitous mechanism contributing to central nervous system neuropathology. Brain Res. Bull. 16, 283–300. 44. Booze, T. and Oehme, F. (1985) Metaldehyde toxicity: a review. Vet. Hum. Toxicol. 27, 11–19. 45. Sparks, S., Quistad, G., Cole, L., and Casida, J. (1996) Metaldehyde molluscicide action in mice: Distribution, metabolism, and possible relation to GABAergic system. Pestic. Biochem. Physiol. 55, 226–236. 46. Booze, T. and Oehme, F. (1986) An investigation of metaldehyde and acetaldehyde toxicities in dogs. Fundam. Appl. Toxicol. 6, 440–446. 47. Qiu, H., Won Jun, H., and McCall, J. W. (1998) Pharmacokinetics, formulation, and safety of insect repellent N,N-diethyl–3-methylbenzamide (DEET): a review. J. Am. Mosq. Control Assoc. 14, 12–27. 48. Lipscomb, J. W., Kramer, J. E., and Leikin, J. B. (1992) Seizure following brief exposure to the insect repellent N,N-diethyl-m-toluamide. Ann. Emerg. Med. 21, 315–317. 49. Abou-Donia, M. B., Wilmarth, K. R., Abdel-Rahman, A., Jensen, K. F., Oehme, F. W., and Kurt, T. (1996) Increased neurotoxicity following concurrent exposure to pyridostigmine bromide, DEET, and chlorpyrifos. Fundam. Appl. Toxicol. 34, 201–222. 50. Kurt, T. L. (1998) Epidemiological association in US veterans between Gulf War illness and exposures to anticholinesterases. Toxicol. Lett. 102–103, 523–526. 51. Schoenig, G. P., Hartnagel, R. E. Jr., Schardein, J. L., and Vorhees, C. V. (1993) Neurotoxicity evaluation of N,N-diethyl-m-toluamide (DEET) in rats. Fundam. Appl. Toxicol. 21, 355–365. 52. Wright, D. M., Hardin, B. D., Goad, P. W., and Chrislip, D. W. (1992) Reproductive and developmental toxicity of N,N-diethyl-m-toluamide in rats. Fundam. Appl. Toxicol. 19, 33–42. 53. Verschoyle, R. D., Brown, A. W., Nolan, C., Ray, D. E., and Lister, T. (1992) A comparison of the acute toxicity, neuropathology, and electrophysiology of N,N-diethyl-m-


103

toluamide and N,N-dimethyl–2,2-diphenylacetamide in rats. Fundam. Appl. Toxicol. 18, 79–88. 54. Dorman, D. C. (1992) Bromethalin rodenticide toxicosis, in Current Veterinary Therapy XI: Small Animal Practice (Kirk, R. W. and Bonagura, J. D., eds.), W. B. Saunders, Philadelphia, PA, pp. 175–178. 55. Klaassen, C. D. (1996) Air pollutants, solvents, vapors, and pesticides, in Goodman and Gilman’s: The Pharmacologic Basis of Therapeutics, Ninth ed. (Limbird, L. E. and Hardman, J. G., eds.), McGraw-Hill, New York, pp. 1689–1690. 56. Gosselin, R. E., Smith, R. P., and Hodge, H. C. (1984) Strychnine, in Clinical Toxicology of Commercial Products, 5th ed. Williams & Wilkins, Baltimore, MD, pp. III:375–379.

104

Blodgett et al.

Low-Level Pb2+ Neurotoxicity

105

II Metals

106

Nihei and Guilarte


107

8 Molecular Mechanisms of Low-Level Pb2+ Neurotoxicity Michelle K. Nihei and Tomás R. Guilarte

1. INTRODUCTION Exposure to lead (Pb2+) has been known for centuries to produce both acute and chronic neurotoxicity (1). There is no known physiological role for Pb2+ and although blood Pb2+ levels provides an index of recent exposure, they do not reflect concentrations of Pb2+ in the brain that may be a consequence of earlier or long-term exposure. Levels of blood Pb2+ in humans appear to peak at 1–2 yr of age, declining to low levels during adulthood, then progressively increase again in the aging population (2). Millions of children continue to be exposed to hazardous levels of Pb2+. One in 10 children in the United States alone possess elevated blood-Pb2+ levels (2), and thousands more appear in Pb2+-treatment clinics with cases of acute Pb2+ poisoning that require clinical intervention (3–6). No current remedy is available for the many thousands possessing blood Pb2+ levels (10–30 µg/dL) known to impair cognition. Two decades of rigorous longitudinal studies have demonstrated that neurodevelopmental toxicity occurs at blood levels at and perhaps lower than 10 µg/dL (7–9). Even less well studied are the learning impairments found in adults that were chronically exposed to low levels of Pb2+ as children (10). Together, these observations underline the fact that environmental exposure to Pb2+ remains a serious public-health problem. The developing brain had been identified as a vulnerable target to Pb2+-induced insults (11–16). Efforts to mitigate the effects of Pb2+ have concentrated on reducing exposure in children and reducing the known toxic threshold for levels of Pb2+ in blood from 60 µg/dL (17) to 10 µg/dL (18). Children possess characteristics that make them more vulnerable than adults to Pb2+ neurotoxicity. First, the gastrointestinal system of the child absorbs a greater percentage of ingested Pb2+ compared to the adult (19). Second, in young children the bloodbrain barrier (BBB) is incomplete (20) and chemicals such as Pb2+ can damage the barrier itself, or enter and damage the developing nervous system. Third, the nervous system undergoes multiple, complex changes during development, many of which do not occur again in the mature nervous system (21,22). Therefore, the developing brain can be exposed to significant concentrations of Pb2+ during unusually vulnerable developmental periods such as synapse formation, gene and protein expression, and From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

107

108

Nihei and Guilarte

other diverse molecular changes associated with these processes. The sensitivity of the developing brain to Pb2+-induced neurotoxicity is an outcome of the many unique characteristics that comprise the maturing central nervous system (CNS). Two fundamental issues of concern arise from these observations. What are the molecular explanations for developmental neurotoxicity after exposure to Pb2+, and are these consequences persistent and detrimental in adults? Although the specific mechanisms of Pb2+ neurotoxicity remain elusive, molecular changes in ionotropic receptors have recently been identified that may be associated with cognitive deficits induced by exposure to Pb2+ (16,23,24). In addition, electrophysiological studies examining the effect of Pb2+ on synaptic plasticity has helped us to better appreciate the functional consequences of molecular differences in Pb2+-exposed experimental animals. The intent of the following review is to connect different realms of evidence that best illustrate our current understanding of the mechanisms of Pb2+ neurotoxicity. A critical aspect of research in the neurotoxicity of Pb2+ are the behavioral studies conducted under laboratory-controlled environments. In this way, Pb2+-induced impairments of learning have been linked to very low concentrations of Pb2+ in blood as well as defining specific concentrations of Pb2+ in the brain. Furthermore, confirming that behavioral effects exist in the Pb2+-treated animals legitimizes the route of exposure and dose(s) of Pb2+ used for physiological and molecular studies. We gain specific information about Pb2+-induced impairments and the possible pharmacological and molecular basis for observed changes in synaptic plasticity from electrophysiological studies examining the effects of Pb2+ on a cellular model for learning or, long term potentiation (LTP). Finally, we discuss the occasionally conflicting data, as well as the specific molecular findings that may begin to untangle the interwoven threads of information we have about the neurotoxicity induced by chronic exposure to low levels of Pb2+. 2. EFFECTS OF Pb2+ ON LEARNING IN EXPERIMENTAL ANIMALS To study molecular mechanisms of neurotoxicity produced by chronic exposure to environmentally relevant levels of Pb2+ an important step is to define an appropriate experimental model. Nonhuman primates exposed to low levels of Pb2+ during development exhibit deficits of different forms of learning and memory (25–29), but molecular studies in this animal model are prohibitively costly. Rats exposed to Pb2+ early in life are impaired in visual discrimination (30), active avoidance (31), fixed-interval schedule-control behavior (32,33), as well as acquisition (14,15) and retention (34) of spatial learning tasks. It is important to recognize that the dose of Pb2+ used in those studies produced blood-Pb2+ levels (16–40 µg/dL) that are consistent with Pb2+ levels found in environmentally exposed children (2,7,8). The behavioral impairments produced by low-level Pb2+ exposure are significant for learning and memory, specifically tasks that require normal functioning of the hippocampus. Morris and colleagues (35), showed that hippocampal lesions in the rat impaired spatial learning in a water maze learning task. Injection of N-methyl-D-aspartate (NMDA) receptor (NMDAR) antagonists (36,37) including Pb2+ (38) into the hippocampus also impaired performance in the water maze, but only if administered prior to acquisition of the task.


109

Effects of Pb2+ on learning behavior have been demonstrated in the absence of changes in sensory/motor or motivational function (14,15,33,19). For example, developmental exposure to Pb2+ produces deficits of spatial learning (14,15) that may be due to molecular alterations in the hippocampus (16,23,24). A number of water maze studies have now convincingly described Pb2+-induced deficits of spatial learning in rats exposed during development (14,15). The former study showed that male, rat pups exposed to Pb2+ over different periods of development (gestation and lactation, continuous, or postweaning), exhibited deficits of spatial learning as adults, except for the postweaning group despite their elevated Pb2+ levels in blood and brain. Notably, adult rats that were exposed to Pb2+ during development but no longer possessed significantly elevated levels of blood or brain Pb2+, remained impaired in learning the spatial task (15). These data, therefore, suggest that Pb2+-induced insults occurring early in life have long-lasting effects. It is also important to recognize that the subtle deficits of learning produced in adults by chronic exposure to low levels of Pb2+ are revealed only by stringent testing paradigms. For example, the “Morris Water Maze” often employs four to eight trials per day (14) and is able to discern more severe impairments in 21-dold (PN21) rats exposed to Pb2+. Acquisition impairments in adult rats exposed to Pb2+ during development is unambiguously demonstrated by a more difficult paradigm in which one trial is conducted per day (15). Modifications of behavioral methods to increase sensitivity may be needed for the detection of subtle but significant changes in cognition. Evidence that NMDARs are less sensitive to Pb2+ if exposure occurs later in life was also suggested by reversible changes in behavioral responses after administration of MK–801, a specific noncompetative antagonist (32). Postweaning exposure to Pb2+, however, produces persistent alterations of the dopaminergic system in the nucleus accumbens of mature rats (33,40,41). The importance of studying postnatal vs postweaning exposures is underscored by how differently Pb2+ exposure affects developing vs mature neuronal systems. Data such as these indicate that developmental exposure to Pb2+ results in persistent molecular changes that are manifest as deficits of learning in adults. 3. EFFECTS OF Pb2+ ON SYNAPTIC PLASTICITY IN EXPERIMENTAL ANIMALS Refinement of neural connections and the processing of information is thought to occur through activity-dependent changes in synaptic strength, such as LTP (reviewed previously refs. 42–44). LTP is a persistent enhancement of synapse strength in response to high-frequency electrical stimulation and is thought to represent a cellular model of learning and memory. LTP was first described in vivo when tetanic stimulus was found to activate afferents projecting from the entorhinal cortex through the perforant path, to granule-cell synapses of the dentate gyrus (45). Induced in the hippocampal dentate gyrus, LTP is an NMDAR-dependent event and selective NMDAR antagonists administered into the hippocampus impair both LTP (46) and spatial learning (47). Many ligand-gated ion channels and metabotropic receptors (see Subheading 5.) have critical roles in LTP, though activation of glutamate receptors is requisite for the induction of LTP and is important in activity-dependent synaptic plasticity during

110

Nihei and Guilarte

development (43,48). Glutamate receptors are a family of both ionotropic and metabotropic receptors, and in the hippocampus the ionotropic NMDAR is considered the coincidence detector for LTP due to its dependence on ligand binding as well as voltage-dependent channel opening for activation (43). Hippocampally mediated spatial learning (14,15), and LTP induced in vivo at the dentate gyrus and in the CA1, are both detrimentally affected by exposure to low levels of Pb2+ and both have NMDAR-dependent components that are impaired in Pb2+exposed animals. Chronic exposure to low levels of Pb2+ have been shown to change the threshold for induction (49,50) as well as the maintenance of LTP induced in vivo in the dentate gyrus (51,52). Rats exposed to Pb2+ during gestation and tested as adults exhibit impaired LTP induction, thus they require a higher train intensity to evoke LTP (49,53). However, Pb2+ exposure restricted to the lactational period was less disruptive to adult LTP than continuous exposure (54). In the latter study, measurement of the slope of the excitatory postsynaptic potential (EPSP), a measure of dendritic depolarization (increased neurotransmitter release), was impaired in rats exposed to Pb2+ for as little as 30 d in the early postnatal period (PN1-30). Maintenance of LTP is impaired by exposure to Pb2+ during development (51,52). Rats exposed to Pb2+ exhibited accelerated rates of LTP decay 1 wk after receiving the conditioning stimulus without altering baseline synaptic transmission and with no changes in the magnitude or incidence of LTP (51). The authors postulated that Pb2+ might interfere with the structural modifications thought to support the long-lasting form of LTP. In a different set of studies (55), Pb2+ exposure resulted in dose-dependent effects on LTP induced in vivo in the CA1. Rats possessing blood Pb2+ levels of 0–12.5 µg/dL showed normal induction and maintenance of LTP, whereas rats with blood Pb2+ levels of 20–30 µg/dL were unable to maintain LTP but induction was not impaired. LTP could not be induced or maintained in rats with blood Pb2+ levels of 27– 40 µg/dL. A reduction in paired pulse facilitation (50) suggested to these authors that Pb2+ may alter GABA receptor function in addition to the NMDAR. To better understand the physiological affects of Pb 2+ on neurotransmitter release in vivo, microdialysis was performed on rats possessing blood Pb2+ levels between 24–53 µg/ dL (56). Total and Ca2+-dependent glutamate release was reduced by Pb2+, but total GABA release was increased. Furthermore, the Pb2+-induced reductions in glutamate release are in agreement with the Pb2+-induced impairments of LTP in similarly exposed rats (52). Another form of activity-dependent synaptic plasticity that is NMDAR-dependent in the dentate gyrus, is long-term depression (LTD). Rats exposed chronically to low levels of Pb2+ also exhibit impairments of LTD in the dentate gyrus and the CA1 (57). LTP, LTD, and discussion of the mechanisms of learning and memory have been reviewed extensively elsewhere and will not be reviewed here in detail (44). The results from in vitro electrophysiology studies in the CA1 largely concur with the in vivo LTP findings in the dentate gyrus. Hippocampal LTP induced in vitro at synapses between the Schaffer collateral afferents and the CA1 pyramidal cells is both NMDAR-dependent and-independent (58). Impairment of LTP at Shaffer collateralCA1 synapses is found in vitro after perfusing the hippocampal slice with Pb2+ (59–61), or by exposing the animal during development and measuring potentiation in the slice


111

preparation (61). Rats exposed to Pb2+ pre- and postnatally that possess a mean blood Pb2+ of about 17 µg/dL at PN16 to PN20, exhibit significantly reduced LTP in the CA1 (62). By comparison, CA3 mossy fiber potentiation is not changed by exposure to Pb2+ and neither is paired pulse facilitation in the CA1 nor CA3 pyramidal-cell fields of the hippocampus (62,63). In contrast to the perforant path-dentate gyrus or the Shaffer collateral-CA1 pathways, LTP induced at the mossy fiber-CA3 synapse is exclusively NMDAR-independent (64). These data suggest that NMDA-dependent forms of synaptic plasticity are more susceptible to chronic low-level Pb2+-exposure than NMDAindependent forms. Thus, NMDAR-dependent LTP in the perforant path-dentate gyrus or the Shaffer collateral-CA1 pathway appears to be a sensitive target for Pb2+ effects, and Pb2+-induced changes in LTP may help to establish the molecular basis for Pb2+induced deficits of learning and memory. 4. MOLECULAR STUDIES 4.1. The Glutamatergic System Glutamate is the major excitatory neurotransmitter in the brain and mediates activity-dependent processes critical to both the developing and the mature brain. The actions of glutamate are mediated by distinct receptor subtypes separated by their affinity for NMDA, α-amino-3-hydroxy-5-methyl-4-isoxazoleproprionate (AMPA), kainate, and L-2-amino-4-phosphonobutyrate (L-AP4) (65,66). Cloning and expression studies have identified multiple polypeptide subunits that co-assemble as heteromeric complexes that make up different receptors within the NMDA, AMPA/ kainate, and metabotropic families (67–69). The AMPA/kainate receptors belong to a subfamily of ionotropic glutamate receptors that mediate fast excitatory synaptic transmission and permit the influx of Na+, K+, and Ca2+ following ligand binding and activation of the channel. Activation of the ionotropic NMDAR is slightly different in that along with glutamate binding, it also requires the coincident voltage-dependent release of Mg2+ from the channel in order to permit the influx of Ca2+. The receptor selective agonist, L-AP4 (65), recognizes the metabotropic glutamate receptors (mGluRs) coupled to G-proteins that activate inositol phospholipid hydrolysis and generate slow synaptic responses (70). At least seven subtypes of mGluRs have been identified showing different signal-transduction properties and pharmacologic profiles. Experimental evidence suggests that mGluRs are involved in synaptic transmission and developmental plasticity (71), however specific antagonists or agonists have yet to be identified limiting investigations into this arena (72). 4.1.1. N-Methyl-D-Aspartate Receptors (NMDAR)

Distinct genes encode five subunits of NMDA-type ionotropic glutamate receptors. Of these five subunits, presence of the NMDAR1 (NR1), appears to be obligatory in functional NMDARs found in vivo (66,73). Strains of mice in which the NR1 was selectively deleted in the CA1 pyramidal-cell field of the hippocampus show impairment of spatial learning and do not exhibit NMDAR-dependent LTP (48,74). There are diverse data that suggest that NMDARs in the developing brain are targets for Pb2+induced neurotoxicity (reviewed previously in ref. 75). Developmental reduction of specific NR1 splice variants gene expression is present in the dentate of developing

112

Nihei and Guilarte

rats exposed to low levels of Pb2+ (23). Recent evidence shows that adult rats exposed to the same low level of Pb2+ throughout life possess marked reductions in both gene and protein expression in the hippocampus (76). Notably, the changes in NR1 gene and protein expression were detected in littermates, exposed to the same levels of Pb2+and that exhibited deficits of spatial learning and LTP. The changes in mRNA expression vary within hippocampal regions, including the CA1 and show both increases as well as decreases at postnatal day 7, 14, 21, and 28 (23). In a similar fashion, NR1 protein expression examined by Western immunoblots of whole hippocampal homogenates show significant variation over the same time periods (16). Exposure to low levels of Pb2+ clearly impact the expression of mRNA and protein for the mandatory NR1 subunit, but compels further investigation into the basis for such changes. Alternative splicing of the NR1 subunit gene may in part explain some of the differences in response to Pb2+. Splicing of three exons from NR1 mRNA produces eight variants (77). The N-terminal, and two C-terminal cassettes (C1 and C2) may be either absent or present allowing for eight possible variants (78), each with different ligand affinities (79). The variants that contain the C1 cassette are found in receptor-rich domains associated with the plasma membrane and splice variants lacking this cassette appear throughout the cell. The amino acid sequences within the NR1 splice variant serve to localize the subunit to specific membrane domains in a manner that may be posttranslationally regulated by protein phosphorylation (80,81). In an interesting contradiction, variants lacking both C1 and C2 cassettes, the NR1-4 variants, were described to show the greatest response to phorbol ester, implicating phosphorylation by PKC (82), though lacking known consensus sequences on C1 and C2 for PKC phosphorylation (83). Guilarte and colleagues (24) demonstrated that alternative splicing of the NR1 gene was anatomically and temporally changed in the hippocampus of developing rats exposed to low levels of Pb2+, further suggesting that developing brains exposed to Pb2+ undergo atypical synaptic changes that may have structural and physiological consequences. The NR1 subunit aggregates with distinct NMDAR-2, or NR2, subunits to form tetra-(84) or possibly pentameric (85) receptor complexes that display a variety of physiological and pharmacological properties (77,86). Four genes code for the NMDAR2 subunits, NR2A-D, with the NR2A and 2B subunits the most abundant of NR2 subunits in brain tissue (86,87). The type and number of NR2 subunit that ultimately assembles with the NR1 is known to influence ligand affinities (77,88,89) as well as allosteric modulation by polyamines, Mg2+ (77), Zn2+ (90,91), and noncompetitive channel blockers such as MK-801. The combination of subunits that comprise each receptor subtype confers specific physiologic (92–96) and pharmacologic (97–99) properties to individual receptors and appears to contribute to their selective vulnerability to Pb2+ (100,101). A “developmental switch” known to occur in the rat brain is the appearance of the NR2A subunit around postnatal day 5–7, coinciding with a slow reduction in expression of the NR2B (102). The NR2B never disappears although peak expression is early in life. Exposure to 1500 ppm lead acetate through gestation and into development produced a delay in the expression of NR2A such that decreased levels of hippocampal NR2A protein were observed at postnatal days 14–21, but returned to normal around postnatal day 28 (16). Expression of NR2A protein in


113

the cortical homogenates of Pb2+-exposed rats was the same as controls. However, Western immunoblots of cortical homogenates do not exclude the possibility that NR2A expression may have changed in specific regions. NR2A protein expression in Western immunoblots of hippocampal homogenates was unchanged in developing 750 ppm PbAc-treated or control rats, however, in situ hybridization of the NR2A mRNA probe showed a dose-dependent reduction between the 750 and 1500 ppm PbAc treatment groups (23). NR2A mRNA was reduced throughout the hippocampal pyramidal- and granule-cell layers. One valuable feature of in situ hybridization is the anatomical definition of differences in brain structures. If a reduction in protein is restricted to a small cellular region of the hippocampus, homogenizing the entire structure may dilute differences in expression. Separation of dentate gyrus from the cornu ammonis may reveal NMDAR subunit protein changes that were not apparent when the hippocampus was analyzed as one structure. Many functional implications of developmentally reduced NR2A expression can be identified from recombinant and transgenic studies. Krupp and colleagues (103) showed that Ca2+-dependent inactivation of the NMDAR occurs predominantly via the NR2A subunit in recombinant systems. Fewer NR2A subunits incorporated into in vivo NMDAR assemblies may produce chronically prolonged activation of NMDAR currents and consequent excitotoxicity or down-regulation of Ca2+-dependent processes to prevent excitotoxicity. Mice lacking the murine equivalent of the NR2A receptor (GluRε1) show deficits of learning and impairments of LTP (104–106). NMDARs with the NR2A subunit exhibit faster kinetics of excitatory postsynaptic currents (96,107) and an apparent decrease in sensitivity to PKC (82). These data suggest that a relative loss of NR2A subunits may result in slower processing of synaptic events and a reduction of activity-dependent synaptic plasticity. Further dissecting the NMDAR, it is important to consider that activity of the receptor also depends on at least seven pharmacologically distinct sites that may be altered by Pb2+. The agonist binding site (1), that opens the high conductance channel, is activated by L-glutamate but is essentially ineffective unless there is binding of the coagonist, glycine (2). The glycine site is modulated by a divalent cation site (3) (100) that may be distinct from the inhibitory Zn2+-binding site (4) located near the mouth of the channel that creates a voltage-independent block. A site within the channel (5) where the noncompetitive antagonist, (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohep-ten-5,10-imine maleate (MK-801) binds, is different from the voltagedependent Mg2+ binding site (6). The polyamines, spermine and spermidine, bind to a regulatory site (7) that facilitates NMDAR mediated transmission (108). Selective effects of Pb2+ on the NMDAR have been described using whole-cell patch-clamp electrophysiology in hippocampal slices. Acute application of 5–20 µM Pb2+ to the hippocampus reduced the frequency of voltage-independent, NMDA-activated channel openings, without affecting AMPA/kainate receptors (109). Age- and region-dependent inhibition of [3H]MK-801 binding by Pb2+ was first shown in vitro (110). Binding studies showed that hippocampal membranes were more susceptible to Pb2+ effects than cortical membrane preparations. Whole-cell patch-clamp studies demonstrated that younger animals and younger dissociated hippocampal cultures both demonstrated greater sensitivity to the inhibition of NMDARs by Pb2+ (111). Other

114

Nihei and Guilarte

studies demonstrated that inhibition of [3H]MK-801 binding in cortical-membrane preparations by Pb2+ occurred more potently in the neonatal brain, than the adult brain (110). Adding Zn2+ to the inhibition curve data generated by Pb2+ and [3H]MK-801, reduced the inhibitory potency of Pb2+. This suggests that Pb2+ and Zn2+ interact at two sites, competitive at one site, and noncompetitive at the other (100,112,113). This effect was further delineated by experiments showing that Pb2+, added to inhibition-curve experiments generated by Zn2+ and [3H]MK-801, decreased the potency of Zn2+ inhibition. The interaction between Pb2+ and Zn2+ occurred at the NMDAR high-affinity site, without changing the apparent number of binding sites (Bmax), reducing both the potency of Zn2+ inhibition as well as decreasing the Bmax at the low-affinity site. At low Zn2+ concentrations (1–10 µM Zn2+, the high-affinity component) a voltage-independent, extracellular site appears to be the primary site of action (114,115). Binding of Zn2+ to the high-affinity site affected frequency of NMDAR channel openings. Higher concentrations of Zn2+ (>10 µM Zn2+, the low-affinity component) interact with a voltage-dependent (114,115), glycine-modulatory site (116) that appears to respond to Zn2+ binding by decreasing NMDAR single-channel amplitude (114,115). The behavior of Pb2+ on [3H]MK-801 binding to the NMDAR complex appeared much like inhibitory cations acting at Zn2+ allosteric sites or other Zn2+ modulatory sites, rather than the Mg2+ site. Whole-cell patch-clamp studies of cultured hippocampal cells suggested that NMDAR glycine affinity may be modulated by Pb2+ (117). Binding studies showed that Ca2+ and Mg2+ modulate the binding of glycine in a stimulatory manner, and this interfaces with an inhibitory divalent cation site activated by Zn2+ or by Pb2+ (100). Though a complex arrangement of possible interactions, these studies point to two major findings. First, that the NMDAR is potently and specifically inhibited by Pb2+ at multiple sites both in vivo and in vitro, and second, that Zn2+ may be an even more likely target than Ca2+ for competition by Pb2+. 4.1.2. α-Amino-3-Hydroxy-5-Methyl-4-Isoxazoleproprionate Receptors (AMPAR) The AMPARs are comprised of combinations of GluR1, 2, 3, and 4 subunits. RNA editing generates splice variants for these subunit families as well, permitting pharmacological and physiological diversity to the receptor population based on the assembly of the individual subunits. Electrophysiology studies have demonstrated that AMPAR currents do not change in response to concentrations of Pb2+ ranging from 0.1–100 µM (109,118,119). However, [3H]AMPA binding in brains from postweaned rats having blood-Pb2+ values ranging from 16–30 µg/dL suggested that a biphasic response may take place in Pb2+-exposed animals. [3H]AMPA binding was increased in rats administered Pb2+ for 2 wk, whereas rats administered Pb2+ under the same protocol for 8 mo, showed decreased [3H]AMPA binding (120). These findings were based on single ligand-binding concentrations and as such, it is difficult to determine if the data reflect changes in affinity (KD), or the apparent number of binding sites (Bmax). The authors posit that the biphasic AMPA receptor response to Pb2+ by these mature rats may be due to increased release of glutamate and downregulation of postsynaptic AMPAR, although they did not negate a direct effect of Pb2+ on AMPARs. A compelling possiblity for an effect by Pb2+ that involves AMPARs involves recent findings regarding silent synapses. Electrophysiologists (121,122) describe that during early development, alterations in synaptic strength by activity may be mediated by


115

NMDAR-containing synapses that do not express AMPARs. At normal resting potentials, the postsynaptic sides of these synapses are physiologically silent because of voltage-dependent Mg2+ blockade of the NMDAR channel, and are thus termed “silent” synapses. AMPA-receptor mediated responses at silent synapses can be stimulated by electrophysiologic induction of NMDAR-dependent LTP (123,124). Tetanic stimulation induces a rapid clustering of AMPARs in dendrites and movement of AMPARs into dendritic spines (125,126) showed that AMPA and NMDAR accumulation at synapses can be independently regulated by synaptic activity and by selective antagonists for each receptor subtype. Acute application of the AMPAR antagonist, CNQX, increased the size and number of AMPARs at previously silent synapses. Likewise, the NMDAR antagonist, APV, increased the size and number of NMDAR clusters and decreased the number of AMPARs ultimately increasing the number of silent synapses. The role of chronic exposure to an inhibitor of NMDAR such as Pb2+ in vivo in such a scenario could be multifold. One possibility is that developmental delay of the expression of NMDAR subunits, or pharmacological blockade of NMDAR by Pb2+, may interrupt normal development of synapses in vivo, and subsequent abnormalities in the number of silent synapses may translate into deficits of plasticity. 4.2. The Cholinergic System 4.2.1. Neuronal Nicotinic Acetylcholine Receptors (nAChR) Like the NMDAR, the neuronal nicotinic acetylcholine receptor (nAChR) is a ligandgated ion channel, and like the NMDAR, nAChRs can be modified at multiple sites that vary among closely related families of ion channels (127,128). Eight α (α2–α9) and three β (β2–β4) subunits have been identified (129,130) and although not all combinations are functional, multiple forms of physiologically and pharmacologically distinct homopentameric and heteropentameric receptors are possible (131). A compilation of studies have examined the electrophysiologic characteristics of nAChR subunit compositions (132–134) and their sensitivity to Pb2+ (109,111,135–138). Nanomolar concentrations of Pb2+ were shown to block nAChRs in mouse neuroblastoma cells and this effect was reversed by concentrations of Pb2+ that exceeded 1 µM (135). Ishihara and others (136) demonstrated that Pb2+ blocked nAChRs in dissociated hippocampal cultures with a peak IC50 of 3 µM, and that nAChRs that were highly permeable to Ca2+ were more sensitive to Pb2+ (α7 subunit-containing receptors) than other nAChR subtypes. Notably, mutant mice lacking the α7 nicotinic receptor subunit show normal spatial-learning performance (138) suggesting that the cognitive effects of low-level, chronic exposure to Pb2+ may not be mediated by its effects on nicotinic receptors. α-bungarotoxin insensitive nAChRs (presumed to be α3 and/or α4 in combination with β2 and/or β4) exhibit less Pb2+-induced inhibition at concentrations as high as 30 µM Pb2+. Zwart and coworkers (137) showed 1 nM–200 µM Pb2+ inhibited α4β2 and α3β4 nACHRs expressed in Xenopus oocytes, although in α3β2, (1–250 µM) Pb2+ potentiated inward currents evoked by acetylcholine. 4.2.2. Muscarinic Receptors The other subclass of acetylcholine receptors are defined by their selective affinity for muscarine. Five muscarinic receptors (M1–5) have been cloned and all coupled to G-proteins. M1,3 and 5 are coupled to PI hydrolysis, M2 and 4 are coupled to cAMP

116

Nihei and Guilarte

(139). Upon activation, muscarinic receptors act to close or open K+, Ca2+, or Cl– channels (140). Unfortunately, little is known about the specific function of each receptor subtype. Most muscarinic-receptor antagonists show no more than fivefold selectivity for one subtype over all the others. The classic muscarinic antagonists, atropine, QNB and the muscarinic agonist, N-methylscopolamine (NMS), do not distinguish between the subtypes and bind to all muscarinic receptors equally well. Autoradiography using the muscarinic agonist [3H]NMS showed decreased muscarinic-receptor expression in rat hippocampus after chronic exposure to Pb2+ (141). These, and other authors (142), also showed that Pb2+ produced no direct effect on the in vitro binding of [3H]NMS in tissue homogenates or tissue sections from rodents. Cory-Slechta and Pokora (143) report that postweaning exposures to Pb2+ produce altered sensitivity to the muscarinic agonist, arecoline, and the cholinergic antagonist, atropine. However, Lenox and others (144) found no changes in cyclic adenosylmonophosphate (cAMP) or cyclic guanosylmonophosphate (cGMP) to the muscarinic agonist, oxotremorine, and methylatropine in adult rats exposed to Pb2+ during development. Rossouw and colleagues (145) exposed postweaned rats to low levels of Pb2+ producing 0.44 ± 0.06 µg/g in brain tissue, and were unable to detect changes in muscarinic receptor binding. The visual cortex may be an exception in that Pb2+ may have a direct (146), selective inhibitory (147) effect on the muscarinic receptor-containing rod and bipolar cells of the retina (148,149). The preponderance of binding data suggests that Pb2+ does not directly affect muscarinic receptors, although other cholinergic components may be affected. Cholinergic innervation to the entire hippocampus originates in the medial septum and vertical diagonal band and there is evidence that exposure to Pb2+ (from conception to weaning, PN21) results in transient reductions of cholinergic markers in the septohippocampal pathway. A 20% reduction of choline acetyltransferase (ChAT) immunoreactive cells in the medial septum/vertical diagonal band was observed in rats exposed perinatally to 0.2% PbAc in drinking water (150). A 30% loss of both ChAT activity and [3H]hemicholinium-3 binding in the medial septum/vertical diagonal band was observed in these rats at PN81 and while ChAT immunoreactivity and activity had returned to normal by PN112, [3H]-hemicholinium-3 binding did not return to normal levels until PN200. Hemicholinium-3 is a ligand selective for the high-affinity, sodium-dependent, choline transporter located on cholinergic terminals. It is not clear whether the cholinergic-marker recovery reflected normal morphology and synaptic activity, or whether atypical functional changes occurred in these perinatally, Pb2+-exposed rats. In fact, Bielarczyk and others (151,152) suggested that chronic exposure to Pb2+ results in cholinergic changes in the brain that resemble lesion of the septohippocampal pathway. Alternatively, NMDARs could influence muscarinic cholinergic receptors indirectly in several ways, particularly during periods of development and changes in muscarinic receptors may be secondary to direct effects of Pb2+ on the glutamatergic system. NMDAR activation is thought to be important in neuronal migration, synaptogenesis, and differentiation (153–155), and NMDAR function in the hippocampus can be modulated by muscarinic agonists (156,157). Neuronal morphology and brain volumes of the predominantly glutamate receptor-rich granule-cell layer, mossy-fiber zone, and dentate molecular layer are altered in rats exposed postnatally (birth to weaning) to low levels of Pb2+ (158).


117

4.3. The Dopaminergic System Behavioral and biochemical studies have suggested that low level Pb2+-exposure results in a reduction of dopamine (DA) release (159–161). Others found no alterations of DA receptor binding in developmentally exposed rats (146). However, rats exposed to Pb2+ for 2 wk during the postweaning period having blood-Pb2+ levels ranging from 16–30 µg/dL exhibit a dramatic reduction in DA-like autoreceptors (D2) and DA transporter (DAT) binding (40), both located presynaptically, in the nucleus accumbens. This effect persisted for over 12 mo and as previously described (41), was not observed in the nigrostriatal dopaminergic pathway of the striatum, but innervation to the nucleus accumbens (162). In contrast to release studies (159–161), the authors suggest that the selective, sustained reduction of D2 receptors and DATs may be the result of chronically elevated DA that then produces a DA-like effect on fixed interval schedule-controlled behavioral performance. When injected with DA or the irreversible DA-antagonist, N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (EEDQ) directly into the nucleus accumbens, the fixed-interval performance rates demonstrated by Pb2+treated rats did not return to control levels (163) suggesting that maturing rats exposed to Pb2+ appear to have alterations of the mesolimbic DA system. The authors also point out that the data suggest that in addition to the dopaminergic system, Pb2+ appears to influence additional neuronal systems involved in the performance of fixed-interval responding paradigms. 4.4. The Gamma-Aminobutyrate (GABA) System Compared to the glutamatergic or cholinergic systems, little is known about the effect of Pb2+ on the γ-aminobutyrate (GABA)ergic system. Recent work by Braga and coworkers (164,165) showed that 0.1 µM Pb2+ increased tetrodotoxin-insensitive spontaneous release of glutamate and GABA from dissociated hippocampal-cell cultures. The authors postulate that Pb2+ was interacting at presynaptic intracellular sites because of the latency of onset of the elevated transmitter release (164). These same authors showed that nanomolar concentrations of Pb2+ blocked tetrodotoxin-sensitive release of glutamate or GABA evoked by spontaneous neuronal firing (165). Complete recovery of the blockade was observed when 4-aminopyridine, a K+-channel blocker, was applied suggesting that the blockade of transmitter release was mediated by the interference of voltage-gated Ca2+ channels by Pb2+. Recently, microdialysis studies in rats exposed in vivo to 0.2% PbAc in drinking water showed reduced hippocampal GABA and glutamate release (56). Decreases in glutamate release were pronounced for rats exposed during development, unlike reductions in GABA that were similar for a postweaning exposure group. Binding studies and additional in vivo exposure studies will promote better understanding of the effect of Pb2+ on the GABAergic system. 4.5. Calcium Homeostasis In vitro studies have demonstrated that Pb2+ and other heavy-metal ions can modulate components of signal-transduction pathways, particularly those that involve Ca2+ homeostasis and Ca2+-dependent processes such as ion channels, enzymes, and messenger proteins (167,168). For example, in bovine adrenal chromaffin and mouse neuroblastoma cells, Pb2+ can potently substitute for Ca2+ in the activation of small- and large-conductance Ca2+-dependent K+ channels (166,169). Strontium, barium, cad-

118

Nihei and Guilarte

mium, and Pb2+ were all found to activate the Ca2+-dependent messenger protein, calmodulin (170). Notably, the authors caution interpretation and remark that extrapolation of these data to the in vivo system is difficult since neither the concentration of free Pb2+, nor of free Ca2+ is known at the cellular/molecular level. Generally it has been assumed that free Ca2+ is 200 M–1cm–1).

146

Zawia and Razmiafshari

Fig. 6. Concentration-dependence DNA-binding of Zn(II) and Ca(II)-peptide complexes when reacted with a GC box consensus DNA element. Shown are autoradiograms of (A) Competition studies of a Zn(II)-peptide complex with hot and cold Sp1 oligonucleotide, (B) Ca(II)peptide complex and (C) Zn(II)-peptide complex. These reactions contained in each lane 32P-labeled oligonucleotide sequence recognized by Sp1. The concentrations of zinc and calcium used and the shifted bands are clearly marked.

Zinc-Finger Motif

147

Fig. 7. DNA-binding activity of metal(II)-peptide complexes when reacted with a GC box consensus DNA sequence. Shown are autoradiograms of (A) Pb(II)-peptide complexes and (B) Sn(II)-peptide complexes. These reactions contained in each lane a 32P-labeled oligonucleotide sequence recognized by Sp1. The concentrations of metals used and the shifted bands are clearly marked. Table 2 Chemical and Physical Properties of Divalent Metalsa

aTable

of Periodic Properties of The Elements, Sargent-Welch Scientific Company, 1992. C.K., 1966, Structure and Bonding, Vol. 1–4, 236. cLigand Field Stabilization Energy for a tetrahedral complex. dFace centered, cubic. bJorgensen,

In our metal competition studies, the effect of Pb(II) on the Zn(II)-peptide complex was evident from a LMCT absorbance at 300 nm (ε = 1204 M–1cm–1) and an absorbance maximum in the visible region (Table 1). These results suggest that Pb(II) is capable of displacing the Zn(II) and retaining the proper spectrum that exhibits similar characteristics as a Pb(II)-peptide complex spectrum (Fig. 4). Pb(II) is a member of

148


subgroup IVB of the periodic table; it is a typical heavy metal with a relatively high atomic weight, and a face-centered crystal structure with high electronegativity. This metal is an example where ligand field-stabilization energy (LFSE) can play no role, since Pb(II) is also a d10 ion (Table 2). However, Pb(II) is a much softer Lewis acid than Zn(II) and it has a greater preference for binding to the soft thiolate ligands. The bonding between metal ions and ligands is a result of the donation of electrons from the ligands to the metal ion, so we may expect the strength of the ligand-metal bond to increase with the electron-attracting power (electronegativity) of the metal ion. If an extremely soft ligand is coordinated to a soft metal, however, it may take care of all d-electrons available on this metal, leaving none to the less soft ligands. The result may be that the residual coordination capacity of the soft metal can be used only by hard ligands using mainly electrostatic forces for bonding (49,50). Thus, the coordination of very soft ligands (sulfur) to a soft metal (Pb) will decrease its softness character, or even turn it into a hard metal, which exhibits a stronger preference for hard ligands (nitrogen) and eventually creates a stronger tetrahedral coordination. Moreover, factors other than the changes in LFSE energy, such as ionic radius and polarizability, will alter metal ion-binding specificity. Pb(II) has almost double the ionic radius than Zn(II) and has a greater polarizability (Table 2). These factors are important in formation of a strong covalent bond, which requires that the valence atomic orbitals of the two atoms concerned overlap strongly. In order for a covalent bond to be formed, the bonding electrons of the ligand has to be available, and they will be more so, the stronger the polarizing influence of the metal ion. In addition, a decrease with difference in electronegativity (DIE) between metal ion and the ligand will imply an increase of the polarizability and consequently, a stronger degree of covalency of single bonds. The DIE for Zn(II) with sulfur and nitrogen ligands is 0.93 and 1.39, which corresponds to 20 and 39% ionic character of a single chemical bond, respectively. In contrast, the DIE for Pb(II) with sulfur and nitrogen is 0.25 and 0.71, which relates to 3.5 and 12% ionic character, respectively (Sargent-Welch Scientific Company, 1992). However, the addition of Cu(II), Cd(II) and Hg(II) to the Zn(II)-peptide complex seem to displace Zn(II) but does not retain the tetrahedral spectrum at 656 nm (Table 1). These divalent cations may have unfilled d-electron shell with a different LFSE that affects their ligand binding. However, copper, cadmium, and mercury have shown thiol coordination in the UV region, suggesting that they could bind at least one sulfur ligand of the cysteine residues (ε = 6125 M–1cm–1, 322 M–1cm–1, 589 M–1cm–1, respectively). Figure 5 represents the Mg(II)-peptide mixtures with negative absorbances and energies in both regions, indicative of failing to compete with the Zn(II) ions (Table 1). The next question to ask was: Do these metal-peptide complexes have biological activity? Since the peptide synthesized is part of the DNA-binding domain of Sp1, we wanted to investigate if these metal complexes can show specific DNA-binding properties. The binding and activity of ZFP have been shown to be modulated by heavy metals (11). Various proteins, which contain zinc-finger domains of the Cys2-His2 type, have been shown to bind specifically to DNA, RNA, and DNA-RNA hybrids (46). Also, previous DNA-binding studies employing gel-shift assays have suggested that single zinc-finger peptides bind to DNA only in the presence of zinc (51). Hence, for a

Zinc-Finger Motif

149

proper metal-peptide DNA complexation and specific binding, zinc ions and the peptide cooperatively must adopt the proper structure necessary for DNA-sequence recognition. Among all metal complexes studied, Pb(II) demonstrated a higher affinity for binding the double-stranded DNA (Fig. 7). The incorporation of Pb(II) within the zincfinger domain and its DNA-binding capability could be due to its higher ionic radius, polarizability, and its greater softness character (52), compared to Zn(II) (Table 2). Furthermore, these Pb-stabilized domains may use hydrophobic interactions in concert with the coordinate covalent Pb-S and Pb-N bonds to buttress the tetrahedral structure of the peptide and allow a more efficient DNA-complexation. Tin, (Sn) a nontransition metal, which is in the same group as Pb(II), did not exhibit DNA-binding at the concentration range studied (Fig. 7). This may be due to the fact that tin has coordinated to different ligands or it has failed to establish proper tetrahedral formation. Of particular interest in these studies is the increased mobility of the DNA-metalpeptide complex on the gel. We do not have a clear explanation as to why the DNAmetal-peptide is not retarded in the gel. However, the ability of zinc-finger proteins to cause winding and unwinding of DNA has been studied. Structural and biochemical analyses have shown that DNA was slightly unwound when bound to zinc-finger peptides (46,53,54). Also, structural studies of zinc-finger protein-DNA complexes have revealed that the DNA molecules are underwound relative to canonical B-form (46). Modeling studies confirm that the canonical linker is a bit too short to allow favorable docking of Zif268, which is closely related to Sp1 (53), on ideal B-DNA (55); the DNA must be slightly unwound to interact with zinc fingers in the mode seen in the Zif268 complex. Essentially, it appears that the helical periodicity of the zinc fingers does not quite match the helical periodicity of B-DNA (56). Although the zinc-finger motif may play a role in changing DNA shape and form, the most likely explanation is that the addition of specific metals to the peptide may change the packing structure of the peptide altering its mobility on the gel. This appears to be linked to the concentration of DNA present because an excess cold DNA competes and restores the accelerated band to its original position. In a global sense, the mechanism through which these metals perturb the integrity and function of transcription factors has yet to be elucidated. This approach would reveal the existence of a direct mechanism through which exposure to environmental metals would disturb the biological function of endogenous proteins and result in adverse health effects. Figure 8 summarizes a postulated mechanism through which these xenobiotic metal ions could interact with the apo-zinc-finger peptide and demonstrate potential DNA-binding characteristic. 4. INFLUENCE OF XENOBIOTIC METAL IONS ON THE DNA-BINDING ABILITY OF AN INTACT RECOMBINANT HUMAN SP1 PROTEIN The transcription factor Sp1 is involved in the regulation of a variety of genes and plays a role during growth and differentiation. Sp1 has a DNA-binding domain (Cys2His2 zinc fingers), which requires Zn(II) for its activity, and may be modulated by other transition metals. However, exposure of cells to the nontransition metal Pb(II), modulates Sp1 DNA-binding. To understand the mechanism by which Pb(II) could

150


Fig. 8. Postulated mechanism for the interaction of various divalent metals with the zincfinger motif and their metal-peptide complex reactions with the DNA consensus element. The expected consequences of such an interaction are also suggested.

alter Sp1 DNA-binding, transition and nontransition metal ions were added to a preparation of a purified recombinant human Sp1 protein (rhSp1). Sp1 DNA-binding was then evaluated using the electrophoretic mobility-shift essay (EMSA). Initiation of transcription by eukaryotic transcription factors involves an ordered assembly of large multiprotein-DNA complexes via protein-protein contacts. Sp1 contains three zinc finger-binding domains and recognizes the 6-bp consensus sequence 5'-GGGCGG (57). In this study, intact rhSp1 was utilized to examine the effects of various divalent metals (Zn, Cd, Hg, Pb, Sn, Ca, and Ba) on Sp1-DNA binding. Sp1 was initially identified in HeLa cells because of its ability to activate transcription from the SV40 early promoter (58,59). Studies in our laboratory (11) have shown that when HeLa-cell extracts were incubated with transition metals such as Zn(II) and Cd(II), a dramatic diminution of Sp1 DNA binding was observed at concentrations of 50–100 µM (Fig. 9). On the other hand, nontransition metals such as Ca(II) or Ba(II) had little effect on Sp1 DNA-binding even at high mM concentrations. Also, we observed that Pb(II), a nontransition metal, exhibited unique effects on Sp1 DNA-binding. Low concentrations of Pb (50–100 µM) enhanced Sp1 DNA-binding while higher concentrations (>100 µM) abolished such binding (Fig. 9). To determine whether the previous observations are due to direct effects on the Sp1 protein, the action of the above metals on Sp1 DNA-binding were tested in reactions, which contained a single rhSp1 protein. Figure 10 shows that the presence of concentrations of Zn(II) higher than 169 µM diminish Sp1 DNA-binding. However, similar concentrations of divalent nontransition metals of the class IIa variety (Ca, Ba) had little effect on Sp1 DNA-binding (data not shown). When transition metals such as Cd(II) and Hg(II) are incubated in the reaction medium which includes rhSp1 and the labeled DNA consensus sequence, an abolishment of Sp1 DNA-binding is observed at

Zinc-Finger Motif

151

Fig. 9. The effects of Pb, Cd, Ca, and Zn on the DNA-binding properties of Sp1 in HeLa cell extracts. The above metals alone or Zn were added to HeLa cell nuclear extracts to examine the ionic effects on Sp1 DNA-binding. The nuclear extracts were incubated with an oligonucleotide containing the Sp1 consensus sequence, and the reaction products were analyzed on a 4% nondenaturing polyacrylamide gel, using the gel shift mobility assay system. Shifted bands were analyzed and quantitated by laser densitometry (see ref. 11 for detailed methods).

low concentrations (>1.5 µM for Cd and > 5µM for Hg) (Fig. 10). Again Pb(II) exhibited a unique effect not seen with other nontransition metals, including its close neighbor Sn(II) (Fig. 10). Pb(II) at concentrations greater than 37 µM eliminated Sp1 DNA-binding (Fig. 10; Table 3). Other transition metals such as Mn(II), Fe(II), Co(II), and Cu(II) also exhibited diminished Sp1 DNA-binding above 90 µM (data not presented).

152


Fig. 10. Concentration-dependence DNA-binding ability of the purified rhSp1 with various metal cations when reacted with a 32P-labeled oligonucleotide sequence recognized by Sp1. Shown are autoradiograms of such reactions that contained (A) Zn(II), Cd(II), Hg(II), and (B) Zn(II), Sn(II), and Pb(II). The concentrations and ratios of metals used and the shifted bands are clearly marked. The arrows represent diminished/altered Sp1 protein-DNA complex.

Zinc-Finger Motif

Table 3 Interference of Divalent Metals with the DNA-Binding Activity of the INtact Sp1 Protein

153

154


Furthermore, incubation of the DNA probe alone with these metals did not alter its mobility. On the other hand, Cd(II) and Hg(II)-DNA mixtures postincubated with rhSp1 exhibited higher abolishment of the Sp1 DNA-binding (Table 3). To better understand the mechanism by which some metals may modulate Sp1 DNAbinding, we have investigated several metal-interaction scenarios. In the presence of excess zinc ions and Cd(II), prior to incubation with the DNA probe resulted in higher diminution of the Sp1 DNA-binding ability (>1.1 µM for Cd, >50 µM for Pb and Hg) (Table 3). In the absence of excess zinc ions, Hg(II) and Pb(II) exhibited similar inhibition of Sp1-DNA complex activity, whereas Cd(II) showed minor effect (>50 µM for Pb and Hg and >169 µM for Cd) (Fig. 10; Table 3). Nontransition metals such as Ba(II) and Ca(II) failed to alter the Sp1-DNA complex activity, however, other transition metals such as Mn(II), Co(II), and Cu(II) inhibited this complex in a concentration-dependent manner, whereas postincubation with Fe(II) required higher concentrations to alter Sp1DNA complex ability (data not presented). The addition of metals in vitro to nuclear extracts containing ZFP transcription factors is one way to assess whether divalent metal cations can alter the DNA-binding of transcription factors. In our earlier studies, we demonstrated that Pb(II) had the ability to selectively alter Sp1 DNA-binding both in vivo and in vitro (11). During the course of these studies, we added a series of metals to nuclear extracts from HeLa cells and examined Sp1 DNA-binding. Low concentrations of transition metals such as Zn(II) and Cd(II) dramatically reduced Sp1 DNA binding in a concentration-dependent manner, while nontransition metals such as Ca(II), Mg(II), and Ba(II) had minimal effects on Sp1 DNA-binding. These findings are consistent and predictable on the basis of the positions of these metals in the periodic table. Since the Sp1 present in these extracts already contains bound Zn(II), addition of exogenous zinc or other transition metals may be in excess of the optimal metal concentration needed to maintain Sp1 DNAbinding. It is possible that high zinc ion concentrations and/or the presence of additional xenobiotic metals interfere with the tetrahedral coordination of zinc in the zinc finger domain by altering its protein structure. An excess of these metals might form mixedligand complexes and induce structural changes of the zinc-finger conformation. The classical zinc-finger structure required for DNA binding would then be altered (10). In native Zn(II)Sp1, the three zinc-fingers do not contribute equivalently to the binding of Sp1 for the GC box, namely important base contacts arise from the second and third zinc fingers (60). In addition, the apparent cooperativity observed in the displacement of Zn(II) from the Sp1 DNA-binding domain suggested that the whole protein might unfold and lose most of its Zn(II) as soon as one site has been displaced. Cooperative rather than independent folding of the multiple zinc fingers in the Cys2-His2 type proteins where the number of fingers varies from 2–37 may be an important structural feature in the consideration of how these proteins bind and influence DNA topology (61). The biphasic effects of Pb(II) on Sp1 DNA-binding (Fig. 9) exhibited by HeLa cell nuclear extracts do not fit in with its position on the periodic table. Unlike Zn(II), Pb(II) is not a transition metal and therefore its effects may have been indirect and mediated by other cellular components present in the extract. To eliminate such a pos-

Zinc-Finger Motif

155

sibility, some of the above metals as well as Sn(II) (in the same class as Pb(II)), were incubated with rhSp1. Similar findings were again observed with rhSp1 as seen with the extracts. Transition metals diminished Sp1 DNA-binding, while nontransition metals did not (Figs. 9 and 10). Although no biphasic effects were observed with the purified rhSp1, Pb(II) again was the only nontransition metal with the ability to inhibit Sp1 DNA-binding at low concentrations. The inability of Sn(II) to modulate Sp1 DNAbinding further reinforced the unique properties of Pb(II) and its ability to directly alter Sp1 DNA-binding by acting on the Sp1 protein (Fig. 10). Although Pb(II) can alter Sp1 DNA-binding by interacting directly with the Sp1 protein, it was important to see whether the effects of Pb(II) are dependent on the presence of Zn(II) in the protein. Consistent with the findings of others (13,38), our previous experiments in this laboratory had demonstrated that when Zn(II) was removed from the protein, it lost its DNA-binding activity and reconstitution with Zn(II) partially restored such activity. Thiesen and Bach (10) who had used rhSp1 as an apo-protein and reconstituted it with Cd(II), Co(II), Cu(II), Mn(II), and Ni(II), found that reconstitutions with Cd(II) and Co(II) were active, while those with Ni(II) and Mg(II) were less active. Predki and Sarkar (41) following dialysis of the estrogen receptor and reconstitution with such metals reported similar findings. Therefore, although Pb(II) does not appear to mimic all the actions of transition metals, it still maintains the ability to directly influence Sp1 DNA binding. On the other hand, with the intact purified protein, when excess zinc ions were present, Pb(II) has shown to be more effective in disrupting the rhSp1-DNA complex activity than Cd(II) and Hg(II) and other transition metals tested. These results suggest that rhSp1 bound to the DNA protects the protein from metal influences at low concentrations of Cd(II) and Hg(II), but Pb(II) through unknown mechanism, could interfere with this complex (Figs. 9, 10; Table 3). Nontransition metals such as Ca(II) and Ba(II) had no effect on the rhSp1-DNA complex activity. Transition metals such as Cd(II) and Hg(II), however, have been more detrimental to the intact protein than Mn(II), Fe(II), Co(II), and Cu(II). Cadmium affected the binding ability of the protein, possibly due to its higher affinity for the sulfur groups of the cysteine residues. Table 3 summarizes the interference of several metals (Zn, Cd, Hg, Pb, and Ca) with the DNAbinding ability of the intact Sp1 protein. In conclusion, our data illustrate that transition metals such as Cd(II) and Hg(II) can directly interfere with the sequence-specific DNA-binding activity of Sp1 and possibly disrupt the metal core of zinc fingers. Pb(II) is the only nontransition metal that possesses some of these properties. Therefore Pb(II)-induced perturbations in developmental gene expression may be mediated via a direct effect of Pb(II) on the Sp1 protein. 5. SUMMARY Environmental metals such as lead (Pb) and mercury (Hg) are potentially toxic and can interfere with the metal-binding motifs of various critical proteins at the cellular level. The zinc finger is a major structural motif involved in protein-nucleic acid interactions and is present in the largest known superfamily of eukaryotic transcription factors such as Sp1. In order to maintain structural integrity, zinc ions coordinate this

156


finger-like structure through bonds created with cysteine and histidine residues. Studies by us had shown that exposure of animals or cells to Pb perturbed the DNA-binding of zinc-finger protein transcription factors (ZFP) such as Sp1 and Egr-1. These metalinduced perturbations can result in a variety of adverse effects including neurological disease and aggressive malignancies. Although, sufficient knowledge exists concerning the adverse health effects of exposure to environmental hazards such as Pb and Hg, the various cellular targets affected by these metals are not entirely understood. In these studies, we have utilized a zinc-finger peptide as a model to study the interactions of various environmental metals with this motif. This novel approach in neurotoxicology provides the advantage of designing apopeptides that resemble the binding sites of endogenous metals and complexing them in vitro with heavy metals suspected to bind to them. The properties of the peptide-metal complexes may then be compared to the action of these metals on an entire protein that normally contains the motif. Furthermore, details on the exact binding site of these metals to these small synthetic peptides may be examined by nuclear magnetic resonance (NMR) spectroscopy. Studies included in this chapter demonstrate that synthetic peptide models of the zinc-finger motif may be used to study the binding characteristics and effects of xenobiotic metals on structural alterations of zinc-finger proteins. These models can help reveal the mechanisms of action of certain metals, especially metal-induced alterations in gene expression. In summary, experiments included in this work have demonstrated that (1) The addition of heavy metals to Zn-peptide complexes can displace or substitute for zinc and retain the functional ability to bind to a cognate DNA sequence; (2) Group IIb metals, transition metals and the nontransition metal Pb, also affected the DNA-binding of the whole protein (rhSp1); (3) Nontransition metals such as Ca, Mg, and Ba did not bind the motif nor did they show an ability to promote the binding of the peptide to its cognate DNA consensus sequence or alter the DNA-binding of the purified protein. Therefore, environmental exposure to heavy metals may selectively alter developmental gene expression via direct alterations in the structure of zinc-finger proteins. 6. CONCLUSION Metals are an indispensable part of life and many are essential, participating in a vast array of critical functions, including control of gene expression through transcription factors. Transcriptional activation and repression events are responsible for coordinate regulation of gene expression at all levels of biological organization. Transcriptional events are susceptible to toxicant interference, and cellular injury as a result of environmental insults. Because of their ionic similarities to essential metals, toxic metals such as Cd(II), Hg(II), and Pb(II) have the tendency to interfere with biological processes that are normally regulated by essential metal cations such as Ca(II), Mg(II), Fe(II), and Zn(II) (7,14,15). Moreover, this ionic mimicry of xenobiotic metals toward the authentic endogenous metals is a likely possibility for their competition with Zn-mediated mechanisms. Therefore, it is possible that the effects of Cd(II), Hg(II), and Pb(II) on cellular processes are due to interactions at different cellular zincbinding site. Among important potential mediators for the long-term effects of heavymetal exposure are regulatory proteins such as the transcription factors.

Zinc-Finger Motif

157

We have presented experimental approaches to determine whether alterations in Sp1 function are a result of direct perturbations of its zinc ions in the zinc-finger conformation. We studied the absorption characteristics and DNA-binding properties of the protein as part of a nuclear extract, in purified form and as a synthetic motif that resembles the DNA-binding domain of a transcription factor. Evidence is thus presented demonstrating the ability of these metals to interact with the zinc-finger motif elucidating a potential mechanism through which altered binding of transcription factors to DNA may occur, resulting in abnormal gene expression. Moreover, these studies could illuminate the importance of the inhibitory effect of certain xenobiotic metals on zincfinger proteins leading to the integration of their physiochemical and biological activities. In addition, they may offer insight into developing approaches to minimize the neurotoxicity of environmental agents and help in designing novel proteins with desired properties. ACKNOWLEDGMENTS The contributions of Tiffany Crumpton and Rita Sharan are greatly appreciated. This research was supported by a NIEHS grant #129ES08104–01A and an EPA-GEM Consortium #T–902887 grant. REFERENCES 1. Holtzman, D., DeVries, C., Nguyen, H., Olson, J., and Bensch, K. (1984) Maturation of resistance to Pb encephalopathy: cellular and subcellular mechanism. Neurotoxicology 5, 97–124. 2. Shao, Z. and Suszkiw, J. B. (1991) Ca2+-surrogate action of Pb2+ on acetylcholine release from rat brain synaptosomes. J. Neurochem. 56, 568–574. 3. Verity, M. A. (1995) Metal Toxicology (Klaassen, C. D. and Goyer, R. A., eds.), Academic Press, New York, pp. 199–223. 4. Pabo, C. O. and Sauer, R. T. (1992) Transcription factors: structural families and principals of DNA recognition. Ann. Rev. Biochem. 61, 1053–1095. 5. O’Conner, T. R., Graves, R. J., de Murcia, G., Castaing, B., and Laval, J. (1993) Fpg protein of Escherichia coli is a zinc-finger protein whose cysteine residues have a structural and/or functional role. J. Biol. Chem. 682, 9063–9070. 6. Goyer, R. A. (1996) Toxic effects of metals in Casarett & Doull’s Toxicology: The Basic Science of Poison, 5th ed. (Klaassen, C. D., Doull, J., and Amdur, M. O., eds.), McGraw Hill, NY, 691–693. 7. Sunderman, F. W. Jr. and Barber, A. M. (1988) Finger-loops, oncogenes, and metals. Ann. Clin. Lab. Sci. 18, 267–288. 8. Berg, J. M. and Merkle, D. L. (1989) On the metal ion specificity of ‘zinc-finger’ proteins. J. Am. Chem. Soc. 111, 3759–3761. 9. Berg, J. M. (1989) Searching for metal-binding domains. J. Am. Chem. Soc., 6, 90–95. 10. Thiesen, H. J. and Bach, C. (1991) Transition metals modulate DNA-protein interactions of Sp1 zinc-finger domains with its cognate target site. Biochem. Biophys. Res. Commun. 176, 551–557. 11. Zawia, N. H., Sharan, R., Brydie, M., Oyama, T., and Crumpton, T.L. (1998) Sp1 as a target site for metal-induced perturbations of transcriptional regulation of developmental brain gene expression. Dev. Brain Res. 107, 291–298. 12. Zawia, N. H., Crumpton, T., Hilliard, A., Sharan, R., and RazmiAfshari, M. (1999) Manifestations of developmental exposure to lead and their implications to the neurodegene-

158

13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25. 26. 27.

28.

29. 30.

31. 32. 33. 34.

35.

Zawia and Razmiafshari rative processes of the aging nervous system, in Chemicals & Neurodegenerative Disease (Bondy, S. C.,ed.), Prominent Press, NY, pp 73–99. Kadonaga, J. and Tjian R. (1986) Affinity purification of sequence-specific DNA binding proteins. Proc. Natl. Acad. Sci. USA 83, 5889–5893. Finelli, V. N., Klauder, D. S., Karaffa, M. A., and Petering, H. G. (1975) Interaction of zinc and lead on deltaaminolevulinate dehydratase. Biochem. Biophys. Res. Commun. 65, 303–311. Suzuki, M., Gerstein, M. and Yagi, N. (1994) Stereochemical basis of DNA recognition by Zn fingers. Nucleic Acids Res. 22, 3397–3405. Goyer, R. A. (1995) Nutrition and metal toxicity. Am. J. Clin. Nutr. 61, 646S–650S. Petit, T. and LeBoutillier, J. (1979) Effects of lead exposure during development on neocortical dendritic and synaptic structure. Exp. Neurol. 64, 482–492. Petit, T., Alfano, D., and LeBoutillier, J. (1983) Early lead exposure and the hippocampus: a review and recent advances. Neurotox. 4, 79–94. Shelenberger, K. (1984) Effects of early lead exposure on neurotransmitter systems in the brain. A review with commentary. Neurotoxicol. 5, 177–212. Clarkson, T. W. (1993) Mercury: major issues in environmental health. Environ. Health Perspect. 100, 31–38. Atchison, W. D. and Hare, M. F. (1994) Mechanisms of methylmercury-induced neurotoxicity. FASEB J. 8, 208–214. Rodier, P. (1990) Critical periods for morphologic assessment. Cong. Anom. 32, 55–64. Silbergeld, E. K. (1992) Mechanisms of Pb neurotoxicity, or looking beyond the lamppost. FASEB J. 6, 3201–3206. McCauley, P., Bull, R., Tonti, P., Lutkenhoff, S., Meister, M., and Doerger Stober, J. (1982) The effect of prenatal and postnatal lead exposure on neonatal synaptogenesis in the rat cerebral cortex. J. Toxicol. Environ. Health 10, 639–651. Bull, R., McCauley, P., Taylor, D., and Croften, K. (1983) The effects of lead on the developing central nervous system of the rat. Neurotoxicol. 4, 1–18. Nichols, D. M. (1990) McLachlan, D. R. C. Issues of Pb toxicity, in Ad. In vivo Body Composition Studies (Yasumura, S., ed.), Plenum Press, New York, pp. 237–246. Choi, B. H., Lapham, L. W., Amin-Zaki, L., and Saleem, T. (1978) Abnormal neuronal migration, deranged cerebral cortical organization and diffuse white matter astrocytosis of human fetal brain. A major effect of methylmercury poisoning in utero. J. Neuropathol. Exp. Neurol. 37, 719–733. Sager, P. R. and Syversen, T. L. M. (1984) Differential responses to methylmercury exposure and recovery in neuroblastoma and glioma cells and fibroblasts. Exp. Neurol. 85, 371–383. Markovac, J. and Goldstein, G. W. (1988) Lead activates protein kinase C in immature rat brain microvessels. Toxicol. Appl. Pharmacol. 96, 14–23. Guilarte, T. R., Miceli, R. C., and Jett, D. A. (1995) Biochemical evidence of an interaction of lead at the zinc allosteric sites of the NMDA receptor complex: effects of neuronal development. Neurotoxicology 16, 63–67. Komulainen, H. and Bondy, S. C. (1987) Increased free intrasynaptosomal Ca2+ by neurotovic organmetalls: Distinctive mechanisms. Toxicol. Appl. Pharmacol. 88, 77–86. Miller, J., McLachlan, A. D., and Klug, A. (1985) EMBO J. 4, 1609. Rhodes, D. and Klug, A. (1993) Zinc-finger structure. Sci. Am. 268, 32–39. Lee, M. S., Gippert, G. P., Soman, K. V., Case, D. A., and Wright, P. E. (1989) Threedimensional solution structure of a single zinc-finger DNA-binding domain Science 245, 635–637. Dolle, P., Lufkin, T., Krumlauf, R., Mark, M. (1993) Local alteration of Krox–20 and Hox gene expression in the hindbrain suggest lack of rhombomeres 4 and 5 in homozygote null Hoxa–1 (Hox–1.6) mutant embryos. Proc. Natl. Acad. Sci. USA 90, 7666–7670.

Zinc-Finger Motif

159

36. Schneider-Maunoury, S., Topilko, P., Seitandou, T., and Levi, G. (1993) Disruption of Krox–20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell 75, 1199–1214. 37. Thiel, G., Lietz, M., and Leichter, M. (1999) Regulation of neuronal gene expression. Naturwissenschaften 86, 1–7. 38. Kadonaga, J. T., Carner, F., Masiarz, R., and Tjian, R. (1987) Isolation of cDNA encoding transcription factor Sp1 and functional analysis of the DNA binding domain. Cell 51, 1079– 1090. 39. Chung, J., Nartey, N. O., and Cherian, M. G. (1986) Mettlaothionein levels in the liver and kidney of Canadians: a potential indicator of environmental exposure to cadmium. Arch. Environ. Health 41, 319–323. 40. Ono, Y., Fujii, T., Igarashi, K., Kuno, T., Tanaka, C., Kikkawa, U., and Nishizuka, Y. (1989) Phorbol ester binding to protein kinase C requires a cysteine-rich zinc-finger-like sequence. Proc. Natl. Acad. Sci. USA 86, 4868–4871. 41. Predki, P. F. and Sarkar, B. (1992) Effect of replacement of “zinc-finger” zinc on estrogen receptor DNA interactions. J. Biol. Chem. 267, 5842–5846. 42. Lee, M. S., Mortishire-Smith, R. J., and Wright, P. E. The zinc-finger motif: conservation of chemical shifts and correlation with structure. FEBS Lett. 309, 29–32. 43. Desjarlais, J. R. and Berg, J. M. (1993) Use of a zinc-finger consensus sequence framework and specificity rules to design specific DNA-binding proteins. Proc. Natl. Acad. Sci. USA 90, 2256–2260. 44. Krizek, B. A., Amann, B. T., Kilfoil, V. J., Merkle, D. L., and Berg, J. M. (1991) A consensus zinc-finger peptide: design, high-affinity metal binding, a pH-dependent structure, and a His to Cys sequence variant. J. Am. Chem. Soc. 113, 4518–4523, 1991. 45. Schmiedeskamp, M., Rajagopal, P., and Klevit, P. E. (1997) NMR chemical shift perturbation mapping of DNA binding by a zinc-finger domain from the yeast transcription factor ADR1. Protein Sci. 6, 1835–1848. 46. Shi, Y. and Berg, J. M. (1996) DNA unwinding induced by zinc-finger protein binding. Biochemistry. 35. 3845–3848. 47. Bertini, I. and Luchinat, C. (1984) High spin cobalt(II) as a probe for the investigation of metalloproteins. Adv. Inorg. Chem. 6, 71–111. 48. Klemba, M. and Regan, L. (1995) Characterization of metal binding by a designed protein: single ligand substitution at a tetrahedral Cys2His2 site. Biochemistry 34, 10,094–10,100. 49. Shannon, R.D. (1976) Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta. Crysta. A32, 751–767. 50. Orgel, L. (1960) An introduction to transition metal chemistry, John Wiley and Sons, New York. 51. Lee, M. S., Gottesfeld, J. M., and Wright, P. E. (1991) Zinc is required for folding and binding of a single zinc finger to DNA. FEBS Lett. 279, 289–294. 52. Jorgensen, C. K. (1966) Structure and Bonding, Springer-Verlag, London, pp. 234–242. 53. Pavletich, N. P. and Pabo, C. O. (1991) Zinc-finger-DNA recognition: crystal structure of a Zif268-DNA complex at 2.1 Aº. Science 252, 809–816. 54. Nekludova, L. and Pabo, C. O. (1994) Distinctive DNA conformation with enlarged major groove is found in Zn-finger-DNA and other protein-DNA complexes. Proc. Natl. Acad. Sci. USA 91, 6948–6952. 55. Elrod-Erickson, M., Benson, T. E., and Pabo, C. O. (1998) High-resolution structures of variant Zif268-DNA complexes: implications for understanding zinc-finger-DNA recognition. Structure 6, 451–464. 56. Kim, J. S. and Pabo, C. O. (1998) Getting a handhold on DNA: design of poly-zinc-finger proteins with femtomolar dissociation constants. Proc. Natl. Acad. Sci. USA 95, 2812–2817.

160


57. Louie, A. Y. and Meade, T.J. (1998) A cobalt complex that selectively disrupts the structure and function of zinc fingers. Proc. Natl. Acad. Sci. USA 95, 6663–6668. 58. Dynan, W. and Tjian, R. (1983) The promoter-specific transcription factor Sp1 binds to upstream sequences in the SV40 promoter. Cell 35, 70–87. 59. Gidoni, D., Kadonaga, J. T., Barrera-Saldana, H., Takahashi, K., Chambon, P., and Tjian, R. (1985) Bi-directional SV40 transcription mediated by tandem Sp1 binding interactions. Science 230, 511–517. 60. Kriwacki, R. W., Schultz, S. C., Steitz, T. A., and Caradonna, J.P. (1992) Proc. Natl. Acad. Sci. USA 89, 9759–9763. 61. Kuwahara, J. and Coleman, J. E. (1990) Role of the zinc(II) ions in the structure of the three-finger DNA binding domain of the Sp1 transcription factor. Biochemistry 29, 8627–8631.

Brain Barriers and Metal Toxicity

161

10 Blood-Brain Barrier and Blood-CSF Barrier in Metal-Induced Neurotoxicities Wei Zheng

1. INTRODUCTION For a metal to enter the brain, it must pass the brain-barrier systems that safeguard brain chemical stability. These barriers exist both at the capillaries of the brain and the spinal cord, in essentially all parenchyma of the central nervous system (CNS), and at the choroid plexus in brain ventricles. If one assumes that the systemic compartment embraces most tissues and organs, via blood circulation, except the brain, then the CNS appears to comprise a unique compartment whose intrinsic circulation is nearly secluded from the blood circulation. Within this cerebral compartment, the interstitial fluid (ISF) flows between neurons and cerebrospinal fluid (CSF) circulates among major brain structures and ventricles. The direct continuity of ISF and CSF allows for the free exchange of substances within the extracellular space of the cerebral compartment. Thus, the barrier that separates the systemic compartment from ISF is defined as the blood-brain barrier (BBB), while the one that discontinues the circulation between systemic and CSF compartments is named blood-CSF barrier (BCB) (Fig. 1). As the brain must function in a chemically stable intramural environment, the brain relies heavily on the protective mechanisms that rigorously restrict access of substances from the systemic circulation to the cerebral compartments. The BBB and BCB are generally impermeable to plasma proteins and to most nonlipophilic organic chemicals. Some low molecular-weight lipid-soluble materials may pass across the barriers by passive diffusion. Other substances, such as electrolytes, nutrients, vitamins, neuroactive regulatory substances, and even water are selectively allowed to enter the cerebral compartment by active transport or by secretion mechanisms at the barriers. Because of the selectivity of the barriers to substances, the brain barriers are by no means the simple roadblocks that merely hinder the movement of material between the systemic and cerebral compartments. Rather, they actively participate in various aspects of brain functions, for normal healthy processes such as early stage of brain development, brain maturation, CNS homeostasis, and neuroendocrine regulation, and for abnormal processes, which include chemical-induced brain edema, brain developmental defects, and possibly initiation of the neurodegenerative disorders. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

161

162

Zheng

Fig. 1. The BBB and the BCB restrict access of substances from the blood circulation to ISF and CSF of the cerebral compartment. The ISF flows between neurons and CSF circulates among major brain structures and ventricles. Direct continuity between the ISF and CSF allows for the free exchange of nutrients, toxins, and products of metabolism within the extracellular space of the cerebral compartment. Thus, substances entering the CSF can come into contact with neurons and ultimately enter the ICF of neurons.

Because of their anatomical location, the brain barriers are destined to be the targets for the toxicities of both organic and inorganic chemicals. Certain heavy metals, such as lead, mercury, cadmium, and arsenic, by as yet undefined mechanisms, tend to accumulate in the substructure of barriers en route to the cerebral compartment. These metals can cause morphological damage to barrier structures, or induce subtle alterations in barrier functions. As a result, unlike damages in peripheral tissues, the dysfunction of brain barriers ultimately leads to substantial neurological and neurobehavioral disorders. This chapter outlines the current understanding of metal-induced neurotoxicities with respect to the brain-barrier systems. The structure and function of BBB and BCB will first be examined. For a more detailed discussion on the physiology of BBB and BCB, readers are referred to the publications by a number of leading investigators in this field (1–3). Much of this article will be devoted to stress the effect of toxic metals on these barriers following unwanted exposure, and the possible consequences of barrier injury in the context of neurological disorders. Finally, the vulnerability of brain barriers to the toxic insults in certain special circumstances will be discussed. 2. STRUCTURE AND FUNCTION OF THE BBB AND BCB 2.1. Structure of Blood-Brain Barrier The concept of a barrier between the blood and brain first emerged nearly a century ago, due to observations made by Paul Ehrlich (2). Following an intravenous injection of a macromolecular-weight dye, Ehrlich noticed a substantial staining in almost all of


163

the tissues and organs throughout the body, but not in the brain. When the same dye was introduced directly into the CSF, however, abundant staining in brain parenchyma was observed. These striking observations clearly indicate that there must be a permeability barrier that prevents the dye from entering the brain. Only with the advancement of electronic microscopic technology in the 1960s coupled with the use of horseradish peroxidase-staining techniques, was the ultrastructure of brain capillaries elucidated and the concept of brain barriers substantiated. Horseradish peroxidase is an electron dense marker (diameter of 5–6 nm) with a molecular weight of 40,000 Dalton. Following injection into the systemic circulation, the enzyme readily diffused from the peripheral blood vessels to non-neural tissues; but was retained by junctions in the intercellular clefts of brain-capillary endothelia (4). Electron microscopy allows the visualization of the vascular joints that hinder horseradish peroxidase. An important aspect of this observation is that it unveils the fundamental structural differences between capillaries in the systemic compartment vs those in the cerebral compartment. Under light microscope, the capillaries of the cerebral compartment are similar to those elsewhere in the body with flattened endothelial cells resting on a basement membrane. In peripheral capillaries, endothelial cells are fenestrated to allow large molecules in the blood to escape into the extracellular tissue space. There are no tight junctions between adjacent endothelial cells. The electrical resistance across the capillary wall is accordingly low, and is less than 100 ohms.cm2. Thus, the peripheral capillary is generally considered to be structurally leaky. In contrast, the endothelial cells in the CNS are not fenestrated and are bound by tight, intercellular junctions (zonulae occludentes) except in the choroid plexus. The endothelia are rested on the basement connective tissue, which continues with neuroglial footage. For small arteries and arterioles, a thin layer of the pia mater, which extends down into the CNS, surrounds the exterior of small blood vessels. Yet, farther down to the capillary level this type of structure disappears. Instead, the external surface of the capillaries is almost completely covered by the perivascular foot processes derived from astrocytes (Fig. 2). Primarily because of endothelial tight junctions, the electrical resistance of cerebral capillaries can reach as high as 1300 ohms.cm2. Thus, a complete BBB consists of three essential components: (1) the capillary endothelial cells with intercellular junctions; (2) the supporting basement membrane; and (3) the perivascular feet of astrocytes, among which the tight junctions between endothelia represent the key structure of the BBB. While the astrocytic footages probably contribute the least to the high resistance of the BBB, recent evidence suggests a role of astrocytes in inducing the formation of BBB. 2.2. Structure of BCB The BCB, which are also characterized by intercellular thigh junctions, is located in the epithelia of the choroid plexus (Fig. 3). Human choroid plexus first appears during the ninth week of ontogeny in the roof of the fourth ventricle; it subsequently appears in the lateral and third ventricles at approx 9–10 wk (5). During the 10th wk, the choroid plexus becomes granulated and begins to perform its secretory functions, although the entire structure is not matured or fully developed until the sixth gestational month (5,6).

164

Zheng

Fig. 2. A complete BBB consists of three essential components: (1) the capillary endothelial cells with intercellular junctions (zonulae occludentes), (2) the supporting basement membrane, and (3) the perivascular foot processes derived from astrocytes. The tight junctions between endothelia represent the key structure of the BBB.

Under the microscope, the choroid plexus also consists of three cellular layers: the apical epithelial cells, the underlying supporting connective tissue, and the inner layer of endothelial cells (Fig. 3). The apical layer consists of numerous closely packed, cuboidal or columnar epithelial cells. On the cell surface are many primary microvilli projecting into the cerebral ventricles. The choroidal epithelial cells possess the tight junctions near their apical surface, which seal one epithelial cell to another. These tight junctions constitute a structural basis for the BCB. The basal side of the epithelial cells rests on a stromal lamina derived from the pia mater. This thin stromal connective tissue contains various types of cells, namely free-lying pia-arachnoid cells, small irregular bundles of collagen fibers, and macrophages (6,7). Next to the connective tissue are the endothelial cells lining the choroidal capillaries. Unlike the rest of brain capillaries, the choroidal endothelial cells are extremely fenestrated. Large molecules such as proteins can pass from the blood through the fenestrated capillary and into the connective tissue. Yet, most of these materials are prevented from further entering the CSF by the tight junctions between the epithelial cells. 2.3. Function of BBB and BCB The primary function of the BBB and BCB, as their names imply, is to restrict the entrance of small substances from the blood to ISF and CSF. Both barriers impede the diffusion of water-soluble molecules, proteins, other macromolecules, and ions from the blood. For example, large molecular-weight molecules such as horseradish peroxidase, cytochrome C, and microperoxidase can pervade most non-neuro tissues and trespass plexus connective tissue as well as the epithelial basement membrane; but fail to reach the brain parenchyma and the CSF due to the presence of tight junctions (7,8).


165

Fig. 3. The choroid plexus consists of three layers. The apical columnar epithelial cells contact directly with CSF. The tight junctions between epithelial cells provide the structural basis for the BCB. Underneath the epithelial cells is a thin connective tissue. The inner endothelial cells lining the choroidal capillaries contact directly with blood. The endothelial cells are highly fenestrated.

However, it appears that small molecular weight-molecules that are lipophilic in nature had a preferred pathway in the BBB rather than in the BCB. Where the distance between brain barriers and neurons is concerned, the BBB is much closer to neighborhood neurons than is the choroid plexus. The close proximity of neurons and the BBB permits fast drug action or toxicity to occur. Also, in addition to blood → barrier cells → ISF → neuron pathway at the BBB, the lipophilic molecules can be transported by an intra-astrocytic transport mechanism, i.e., blood → barrier cells → astrocytes → neuron pathway. After diffusing into the astrocytic footage, the molecule can bind to an intracellular carrier in astrocytes and be transported to neurons without circulating through the ISF. This mechanism is not available at the BCB. For the BCB to deliver highly lipophilic molecules in the CSF, there must be an acceptor moiety, either a specific carrier or a nonspecific-binding ligand. Our own experience with in vitro BCB models shows that the lack of such an acceptor renders it difficult for a lipophilic molecule, such as α-tocopherol, to enter the CSF to reach neurobiologically significant concentrations, albeit it may well penetrate the barrier (Zheng and Blaner, unpublished data). Thus, the lipophilic molecules appear to circumvent the BBB better than the BCB. Unlike the traditional concept of “inert” barriers, both the BBB and BCB actively participate in the regulation of cerebral compartment homeostasis. The BBB transports a variety of materials to the CNS, such as amino acids (e.g., glutamine, arginine), sugar (e.g., glucose), purine bases (e.g., adenine), nucleosides (e.g., adenosine), as well as

166

Zheng

hormones (e.g., vasopressin) (9–11). Examples at the BCB include transport of amino acids (e.g., glycine, L-alanine) (12,13), hormones (e.g., thyroid hormones, melatonin, growth hormone) (14–16), and peptides (e.g., atriopeptin, vasopressin) (17). In addition, the choroid plexus secretes CSF. Human choroid plexus produces CSF at a rate of 0.3–0.4 mL/min and it comprises approx 80–90% of the total CSF (2,6). The secretion of CSF by the choroid plexus is an active process and is involved in the transport of Na+, Cl–, and HCO3– from the capillaries to the ventricles. Movement of these ions establishes an osmotic pressure across the cell surface of the epithelium. The plasma water component in the systemic compartment is then driven by this osmotic force into the ventricles. The BCB also plays a cleansing role in the cerebral compartment by transporting waste materials out of the CNS, i.e., in the opposite direction from the CSF to blood. Little is known on the function of the BBB on this regard. Since there is no lymphatic system in the CNS, the constant removal of wastes by the choroid plexus, in addition to the cerebral circulation via the arachnoid villi, provides a useful mechanism to prevent the build-up of metabolic byproducts. For example, an active transport system that has been shown to remove cephalosporin antibiotics from the CSF plays a major role in determining the pharmacokinetics of beta-lactam antibiotics in the CSF (18). Cimetidine, an histidine-2 (H2)-receptor antagonist, is eliminated from the CSF by a saturable, carrier-mediated active transport process in the choroid plexus (19,20). Ca2+ pumps located in the CSF-facing (epical) membrane (21) ensure a low concentration of Ca2+ in the CSF, even under extreme fluctuation of plasma Ca2+ concentration (22,23). Other examples include inorganic anions such as I– and SCN–(24,25); organic anions such as benzylpenicillin (26), bilirubin (27), and anionic pesticides (28,29); organic cations such as quaternary ammonium compounds (30–32); prostaglandins (33); and some peptides (34). 2.4. Comparison of BBB and BCB Given the fact that the BBB is evenly distributed throughout the brain and possesses the maximum proximity to neurons, one might conclude that the BBB, and not the BCB, may dominate the regulation of homeostasis in the cerebral compartment. However, with respect to the surface area of the barriers, the BCB does in fact play a significant role in governing brain chemistry. In comparison to its small tissue mass, the choroid plexus has a very large surface area (Table 1). Besides the primary microvilli, the entire choroid plexus is so pleated that it creates a secondary multiform macrovilli, which greatly increases the total surface area of the tissue. Estimates from tissues of 1mo old-rats suggest that the total apical surface area of the choroidal epithelium is approx 75 cm2 or about one-half the surface area of the BBB (155 cm2) (35). Thus, the broad surface area of the BCB ensures an efficient exchange of materials between the CSF and blood compartments. Another important factor in managing CNS homeostasis pertains to the rate of entrance of chemicals into the cerebral compartment. Because of the relatively slow CSF turnover and because of the spatial distance between the BCB and neuronal tissues, it has been speculated that the blood-CSF transfer for most solutes may be less dynamic


167

Table 1 Structural Comparison of the BBB and the BCB Parameter Anatomy

Tight junctions Location Surface area Blood flow rate Efficiency as a barrier

Blood-brain barrier • Endothelial cells of cerebral capillary • Basement membrane • Perivascular feet of astrocytes • Interstitial fluid (ISF) Between endothelia Whole brain 155 cm2 0.9–1.8 mL/min/g Highly efficient

Blood-CSF barrier • Endothelial cells of ventricular capillary • Basement membrane • Choroidal epithelium • Cerebrospinal fluid (CSF) Between epithelia Choroid plexus in lateral, 3rd and 4th ventricles 75 cm2 4–6 mL/min/g Efficient, but functionally leaky

than the blood-ISF exchange (36). This postulation, however, may be only partly true for highly lipid-soluble molecules, since they appear to gain access to neuronal tissue by a preferred lipophilic BBB (see discussion in Subheading 2.3.). The disposition of less lipid-soluble solutes in the brain relies largely on an equilibrium among several elements, which are coordinated by both barriers, i.e., blood flow at local area, ISF exchange rate, CSF production and circulation, and composition of both ISF and CSF. In fact, the blood flow to the choroid plexus ranges between 4–6 mL/min/g, which is about 3–5 times faster than those estimated to many other brain regions (0.9–1.8 mL/ min/g) (2,37–39) (Table 1). The fast blood flow to the choroid plexus is expected with respect to its physiological role in secretion of CSF to the cerebral compartment. The rapid blood flow at the BCB, on the other hand, warrants an efficient influx of chemicals into the CSF. It is possible that following entrance of a less lipid-soluble molecule, its CSF concentration may lag behind its ISF concentration for a short period of time. Nonetheless, this lag by no means undervalues the contribution of the BCB to the overall cerebral steady state of the transported molecule, because, unless the compound is tightly tissue-bound, its steady state in the cerebral compartment, i.e., among neuronal tissue, ISF, and CSF, must be established based on the functional integrity of both barriers. The fact that far more severe neurotoxic consequences result from intraventricular injection of certain neurotoxicants than from parental administration bespeaks the significance of the CSF circulation in drug action and in neurotoxicity (40–43). It is necessary to point out that the tight junctions between the epithelial cells in BCB seem less effective, or somewhat more “leaky,” than those between endothelial cells of the BBB (2). Taking electrical resistance across the barrier as an example, a well-established BCB model in a Trans-well culture device usually possesses the transepithelial resistance of 100–150 ohms⋅cm2 (44,45). In comparison, the trans-endothelial resistance of the endothelial cells cocultured with astrocytes in the similar device can achieve a trans-endothelial resistance as high as 500–800 ohms⋅cm2 (46,47). How

168

Zheng

this in vitro difference in trans-barrier resistance eventually resembles that in brain in vivo is unclear. Finally, from the toxicological point of view, the unique anatomical location of the choroid plexus within the brain seems likely to be associated with certain metal-induced neurotoxicities. For example, the hippocampus, a brain area associated with learning and memory, has been considered the primary target for lead (Pb)-induced learning deficits (48,49). Anatomically, the hippocampus extends from the cortex and folds inward to form part of the floor in the inferior horn of the lateral ventricles (50). Interestingly, the hippocampal formation is immediately adjacent to the lateral choroid plexus, where Pb accumulates to an extraordinary amount (44,51–53). 3. THE BBB AND BCB AS TARGETS IN METAL-INDUCED TOXICITIES As the gatekeepers between the blood circulation and the cerebral compartment, the brain barriers often become targets for a variety of xenobiotics. By far, at least nine metals have been found to accumulate in the BBB and BCB (Table 2). Clinically, poisoning with Pb, mercury (Hg), and arsenic (As) have been demonstrated to induce vascular destruction and cerebral hemorrhage (discussion follows). The damage to endothelial structure of the BBB is a fundamental cause of leakage of blood-borne materials to surrounding brain parenchyma. Heavy metals also accumulate in the choroid plexus. As early as in 1963, Berlin and Ullberg (54,55) observed the deposition of Cd and Hg in a brain area that corresponds to the choroid plexus. Animals exposed to Pb, Hg, Cd, and As retained much higher concentrations of these metal ions in the choroid plexus than in the brain cortex or CSF (52,53). The concentration of Pb in the choroid plexus, for example, was about 57 times greater than that in the brain cortex; Hg was 12 times greater, Cd, 33 times greater, and As, 13 times greater. CSF concentrations of Pb, Hg, and As were about 70, 95, and 40 times, respectively, less than those found in the choroid plexus. The toxic metals that act on the brain barriers are classified as general barrier toxicants, selective barrier toxicants, and sequestered barrier toxicants. General barrier toxicants usually cause the most substantial damage to barrier structure. The deposition of these metals in the barriers and the subsequent morphological alterations in the structure not only permit the metals themselves to diffuse into the brain, but also facilitate the entrance of other neuroactive toxicants to the brain. Metals in this category include Pb, Hg, Cd, and As. Selective barrier toxicants usually do not directly alter barrier’s permeability, nor do they produce a massive hemorrhage seen in clinics. However, en route to the brain, metals in this category selectively act on certain critical regulatory functions of the barriers, giving rise to profound neurotoxic consequences. This group of toxic metals includes manganese (Mn), copper (Cu), and conceptually, Pb in the BCB. Finally, the sequestered barrier toxicants may deposit in or be sequestered by the barriers as an essential CNS defense mechanism. The sequestration of these metals in the barriers do not provoke any known harmful consequences to the barriers in the current literature. Metals in this category include iron (Fe), silver (Ag), zinc (Zn), and gold (Au).


169

Table 2 Heavy Metals Sequestered or Transported by the Choroid Plexus Heavy metals

Species/model

References

Lead (Pb)

Rat/in vivo Guinea pig/in vivo Human/postmortem Rat/in vivo, in vitro Mouse/in vivo

(52,53) (69) (51); (67) (52); (108) (109,110); (55) (52); (101); (98) (54) (95) (44,136); (137,139); (41) (135) (52) (51) (167,168) (169) (170) (154) (160) (134,162)

Cadmium (Cd)

Mercury (Hg)

Manganese (Mn) Arsenic (As) Silver (Ag) Gold (Au) Tellurium (Te) Iron (Fe)

Rat/in vivo Mouse/in vivo Human/postmortem Rat/in vivo Mouse/in vivo Rabbit/in vivo Dog/in vivo Rat/in vivo Human/postmortem Dog/in vivo Rabbit/in vitro Rat/in vivo Human/postmortem

3.1. General Barrier Toxicants 3.1.1. Lead (Pb) Of the heavy metals causing neurotoxicity, Pb is of particular interest. Environmental exposure to Pb in children has been associated with cognitive deficits. Moreover, there appears to be no threshold with regard to the association between blood Pb concentration and cognitive deficits (56,57). The mechanism whereby Pb adversely affects childhood brain development remains uncertain. 3.1.1. Pb Exposure and BBB The BBB has long been known to be a target for Pb toxicity. One of the explicit clinical manifestations in acute Pb poisoning is brain swelling, accompanied with herniation, ventricular compression, and petechial hemorrhages (58–60). An increased incidence of cerebral hemorrhage, thrombosis, and arteriosclerosis has been described in a group of British battery workers exposed to extremely high levels of Pb (61). The brain edema and related increase in ISF space is suggestive of leakage of the BBB. Under these exposure conditions, Pb-induced microvascular damage is prevalent with leaky microvessels evident by a characteristic opening of the inter-endothelial tight junctions and by enhanced pinocytotic activity. Certain macromolecules such as horseradish peroxidase, trypan blue, and albumin can extravasate into the surrounding neuronal parenchyma in proximity to the microvessels with altered permeability (60,62,63).

170

Zheng

Chronic exposure to high concentrations of Pb also leads to cerebral vascular damage in young animals. Developing rats that develop Pb encephalopathy exhibit an extravasation of albumin, fibrinogen, and fibronectin, as seen in extensive staining in the cerebellar cortex, with diffuse spread to the white matter of the corresponding folium (64). Pb appears to have a unique affinity to cerebral endothelial cells and it accumulates in much greater concentrations than in other brain cell types (60,65). The changes in endothelial bud (or angioblast) may actually precede those involving neurons and glia. Thus, some have suggested that Pb encephalopathy probably results from the death of many of these buds (66). 3.1.1.2. PB EXPOSURE AND BCB Pb is known to accumulate in the choroid plexus to a greater extent that in brain endothelial cells. Friedheim et al. (51) found that Pb in human choroid plexus increased significantly with age, while Pb in the brain did not. Further, Manton et al. (67) reported a 100-fold increase of Pb in human choroid plexus compared with that in the brain cortex. A significant aspect of these findings is that an age-related accumulation of Pb in a particular tissue within the human brain is most likely associated with environmental exposures. A study in rats indicates that accumulation of Pb in the choroid plexus is both doseand time-dependent (52). Concentrations of Pb in the choroid plexus increased proportionally with increasing dose, while Pb concentrations in the cerebral cortex and CSF were not significantly changed. Following acute administration of Pb acetate (50 mg/ kg, ip), choroid plexus Pb concentrations continued to arise and did not plateau even after 24 h following dosing (52). A chronic Pb exposure study was also conducted in male rats exposed to Pb in drinking water at doses of 0, 50, or 250 µg Pb/mL (as Pb acetate) for 30, 60, or 90 d. At d 90, blood Pb in control rats was below the detection limit (5 mg/m3) as may occur in Mn mining, steel manufacturing, or welding (5–7). Mn is also present in methylcyclopentadienyl manganese tricarbonyl (MMT); an octane-enhancing fuel additive used in unleaded automotive gasoline. Since MMT is extremely unstable in light and rapidly degrades in air, exposure to its combustion products is of special concern. The combustion of MMT by the automobile engine results in the formation of a complex mixture of Mn salts. Different MMT combustion products are produced depending on the fuel composition and engine and catalytic converter thermodynamics. Modern automobiles equipped with catalytic converters emit Mn primarily in the phosphate form, although smaller amounts of sulfates and oxides may also be discharged (8,9). Using various environmental modeling approaches, it was estimated that air levels of Mn in most urban areas in the United States would increase less than 0.02 µg/m3 if MMT were used in all unleaded gasoline (8). Actual air Mn concentrations from Canadian cities in which MMT has been widely used for over 10 yr remain well below the current inhalation reference concentration (0.05 µg Mn/m3) for respirable Mn set by the US Environmental Protection Agency (EPA) (10–12).

Manganese in Health & Disease

197

2.2. Route of Exposure as a Critical Determinant of Neurotoxicity The route of exposure can influence the distribution, metabolism, and neurotoxicity of Mn (13,14). The oral route is considered to be less important for risk assessment purposes, since oral Mn is poorly absorbed from the gastrointestinal tract (net absorption < 5%), and brain and other tissue Mn levels remain relatively constant despite large fluctuations in oral Mn intake. In contrast, inhalation is more efficient than ingestion at delivering Mn to the brain. Pharmacokinetic factors that may contribute to the increased efficiency in brain Mn delivery observed after inhalation exposure include increased Mn absorption from the pulmonary tract and slower blood clearance of absorbed Mn (13). The liver plays a key role in maintaining normal organ Mn concentrations. Mn absorbed from the gastrointestinal tract is first transported to the liver where it is removed from the blood. Dose-dependent biliary excretion of divalent Mn serves to regulate the percentage of ingested Mn retained by the body and to limit increases in systemic tissue Mn concentrations. Liver Mn concentrations are commonly elevated following high-level oral exposure. In contrast, liver Mn concentrations often remain normal following Mn inhalation exposure. For example, Ulrich and coworkers (15) found no increases in rat or monkey liver Mn concentrations following subchronic inhalation exposure to Mn3O4 (11.6–1152 µg Mn/m3, 24 h/d, for 9 mo). Similarly, Vitarella and coworkers (16) did not observe increased liver Mn concentrations in rats following short-term (2-wk) inhalation exposure to Mn phosphate. These investigators did, however, report increased fecal Mn elimination rates and enhanced whole-body Mn clearance suggesting that enhanced biliary excretion was occurring. Under certain exposure conditions, inhalation exposure to Mn may result in increased liver Mn concentrations. For example, Morganti and coworkers (17a) reported increased liver Mn concentrations in male Swiss mice following subchronic (16–32 wk) inhalation exposure to much higher Mn concentrations (49–85 mg MnO2/m3, 7 h/d). Mn is capable of existing in a number of oxidation states, and Mn may undergo changes in valence states within the body. The valence of Mn in most enzymes is Mn+3 while most Mn taken into the body exists as either the Mn+2 or Mn+4 form. In vitro, ceruloplasmin can oxidize Mn+2 to Mn+3. Mn oxidation can result in a shift of Mn binding from α2-macroglobulin to transferrin. Clearance of Mn+2 bound to α2-macroglobulin is more rapid than clearance of Mn+3 bound to transferrin (17). The rate and extent of Mn oxidation or reduction in the body and associated protein binding is a key determinant of Mn retention in the body and brain Mn delivery (18). Depending on the liver’s ability to excrete excess Mn, exposure may or may not result in elevated liver concentrations of manganese. Chronic liver disease can interfere with first-pass biliary excretion of Mn in two ways: first, through cholestasis, and second, by portacaval shunting. Some patients with chronic liver disease with no unusual Mn oral intake or inhalation exposure were observed to have pallidal MR T1-weighted hyperintensities and increased sleep disturbance correlated with elevated erythrocyte Mn concentrations (19). Additionally, Mn exposure can exacerbate liver dysfunction: an iv bolus of bilirubin followed by Mn causes cholestasis in rats (20). Therefore, liver function is an important consideration in the monitoring and treatment of individuals possibly affected by Mn exposure.

198

Aschner et al.

2.3. Species and Age-Related Differences in Manganese Pharmacokinetics and Neurotoxicity A substantial literature exists regarding significant species differences in the neurotoxicity of Mn. Therefore, extrapolation among Mn-induced effects in rats, nonhuman primates, and human beings must be made with caution. The use of T1-weighted magnetic resonance imaging (MRI) techniques has revealed that similar to human beings, macaques given high doses of Mn develop elevated brain Mn concentrations localized to the striatum, globus pallidus, and substantia nigra (21,22). Furthermore, Mn-exposed monkeys have reduced levels of striatal and pallidal dopamine and 3,4-dihydroxyphenylacetic acid (DOPAC) and demonstrate dopaminergic neuronal losses analogous to Mn-poisoned humans (23). Monkeys also develop gait and other motor abnormalities that mimic those observed in affected human beings. In contrast, rats do not demonstrate selective Mn accumulation in the striatum (24). Only a limited number of studies have been conducted that evaluated the effect of Mn on neonatal or adult rat behavior. None of these studies, however, confirmed development of a behavioral syndrome comparable to that seen in Mn-poisoned humans and monkeys (25). The human substantia nigra becomes depigmented as the result of Mn neurotoxicity. One possible explanation for this selective depigmentation is the observation that Mn has a high affinity for neuromelanin and that its deposition is highest in melanin-containing tissues (26). The interplay between Mn and melanin may also explain some of the known species differences in Mn neurotoxicity. Neuromelanin levels increase considerably with higher phylogeny, and brain levels are highest in primate brain regions where the dopaminergic pathways are most active. Rats, including pigmented strains, have extremely low levels of substantia nigra neuromelanin and thus may not demonstrate the same patterns of Mn accumulation or extrapyramidal symptomatology. A wealth of evidence indicates that neonates are more susceptible than adults to many neurotoxicants. Neonatal rats appear to be at an increased risk for Mn-induced neurotoxicity due to their ability to develop higher brain Mn levels and more pronounced brain pathology than do adults in the face of equivalent or lesser Mn exposures (27,28). Factors influencing this increased susceptibility include increased Mn absorption from the gastrointestinal tract, an incompletely formed blood-brain barrier (BBB), and the virtual absence of biliary Mn excretory mechanisms until weaning. The brain takes up a significant proportion of the Mn retained during the early neonatal period. For example, Keen and coworkers (1a) report that approx 8% of the total oral Mn dose is retained, in rats, by the brain. In addition, Mn crosses the placental barrier and accumulates in the brain following gestational exposure (29). 2.4. Distribution to Other Tissues In humans, most tissue Mn concentrations range between 0.1 and 1 µg Mn/g wet weight. Tissues with high Mn concentrations include the liver, pancreas, and kidney. Lung Mn concentrations demonstrate dose- and time-dependent increases following inhalation exposure. Plasma and red blood cell samples typically have low Mn concentrations normally and even following short-term Mn exposure. The bone can account for up to 25% of the total Mn found in the body. In addition to its role in normal osteoblast metabolism, bone also effectively sequesters Mn. Dramatic elevations in bone Mn concentration may occur following excessive Mn exposure. For example, an


199

80–90-fold increase in bone Mn concentration is observed in Sprague-Dawley rats given intraperitoneal injections of 11 mg MnCl2 • 4H2O/kg/d for 1 mo (30). Animals given high levels of Mn also had decreased zinc and magnesium bone concentrations, suggesting that Mn interacts with other trace minerals critical for bone formation. When compared to liver, kidney, and other organs that eliminate Mn quickly (half-life 10–15 d), bone demonstrates prolonged retention of elimination half-lives > 50 d (31). Another tissue showing prolonged Mn retention is the brain where an apparent half life of elimination is estimated to be on the order of 50–75 d (32,33). Oral supplementation with a combination of Ca, Mn, Cu, and Zn has been shown to reduce loss of spinal bonemineral density in postmenopausal women (34). However, it has not been shown that Mn supplementation alone has this effect, so toxicity concerns should outweigh possible benefits of supplementation with Mn alone for this indication. 3. TRANSPORT OF MANGANESE INTO THE CENTRAL NERVOUS SYSTEM 3.1. Olfactory Transport There is evidence that uptake of Mn into the brain may occur by direct olfactory axonal-transport mechanisms. Olfactory transport of Mn has been demonstrated to occur in the rat, mouse, and pike following intranasal instillation (35,36), and in rats following inhalation exposure (37). In rats, intranasal instillation of Mn has been shown to result in direct movement of Mn to the olfactory bulb and the telencephalon via transport in secondary olfactory neurons. Once in the brain, Mn can continue to move across synaptic connections between neuronal-cell bodies and continue to be transported along their cell processes to sites distantly connected to the olfactory pathway (36). It is this neuronal-transport characteristic of Mn that has led to its use as a neuronal tracer (38,39). This route has been shown to be a rapid route of transport of Mn to brain structures in the olfactory pathway, but a slow route of delivery to the rat striatum. Several observations indicate that Mn undergoes active axonal transport. Mn transport to the olfactory bulb demonstrates concentration-dependent and saturable transport kinetics (40). Mn transport kinetics is altered following colchicine treatment suggested that the metal uses microtubule-associated fast axonal transport (41). The maximal transport velocity determined for 54Mn in pike exposed to the chloride salt was determined to be approx 70 mm/d which is consistent with a fast axonal-transport system (36). In mammals, fast axonal transport occurs at a rate of approx 200–400 mm/ d and is used by the neuron to transport organelles, lysosomes, nerve-growth factor, and selected small molecules. Until recently, little was known regarding the olfactory transport of Mn following inhalation exposure. Brenneman and coworkers (37) conducted studies in rats using short-term (90-min) inhalation exposure to radiolabeled 54MnCl2 aerosols (0.5 mg Mn/m3). These investigators used an animal model in which one nostril was occluded thus preventing olfactory transport of Mn to one side of the rat brain. Interestingly, these investigators showed that the olfactory route contributed the majority (>90%) of the 54Mn found in the olfactory pathway of the brain up to 8 d following acute inhalation exposure. Their conclusion was further supported by data using longer-term rat inhalation to a less soluble form of Mn (16). In the Vitarella study (16), rats were exposed

200

Aschner et al.

6-h/d for 7 d/wk (14 exposures) to Mn phosphate at 0, 0.03, 0.3, or 3 mg Mn/m3. The Mn concentration achieved in the olfactory bulb following the end of this 2-wk inhalation exposure was significantly higher than that observed in either the striatum or cerebellum, lending credence to the direct olfactory-transport theory. These findings suggest that the olfactory route may indeed be a significant pathway by which Mn may gain access to the brain. However, neither of these inhalation studies clearly demonstrated that direct olfactory uptake contributes significantly to increased striatal Mn concentrations. It is possible that repeated exposure or a longer postexposure delay might be required for detection of Mn at a site so distant from the olfactory pathway. The relevance of these findings to human Mn inhalation exposure and the risks for neurotoxicity are not known and are complicated by interspecies differences in nasal and brain anatomy and physiology. In the rat, the olfactory bulb accounts for a relatively large portion of the central nervous system (CNS), and the nasal olfactory mucosa covers approx 50% of the total nasal epithelium. These structures are proportionately smaller in humans, suggesting that this route of brain delivery may be less important in humans as compared to the rat. In addition, total airflow to the olfactory mucosa is much lower in humans than in rats. These differences likely predispose the rat, more so than humans, to olfactory deposition and potential olfactory transport of Mn. Additional research will be required to better clarify the potential significance of the olfactory route of delivery of Mn to the brain in humans exposed via inhalation. 3.2. Brain Uptake of Manganese Mn2+ does not exhibit high affinity for any particular endogenous ligand. It has almost no tendency to complex with-SH groups or amines, and it does not possess much variation in its stability constants for endogenous complexing ligands (log10k = 3, 4, 3, and 3, for glycine, cysteine, riboflavin, and guanosine, respectively, where k is the affinity constant). Approximately 80% of Mn in the plasma is bound to β1-globulin and albumin (42). A small fraction of Mn in plasma is in the trivalent oxidation state and bound to transferrin (Tf) (43). At normal plasma Fe concentrations (0.9–2.8 µg/ mL), Fe binding capacity (2.5–4 µg/mL), and Tf concentration (3 mg/mL, with 2 metalion-binding sites per molecule [Mr 77000], of which only 30% are occupied by Fe3+), Tf has 50 µ mole/L of unoccupied Mn3+ binding sites, and has therefore been implicated as a potential transporter for Mn across membranes. It is noteworthy that Tf receptors have been localized on the surface of the cerebral capillaries (44–46) and that endocytosis of Tf in capillaries of the BBB has been noted (46). Endocytosis of a Mn-Tf complex in cultured neuroblastoma cells (SHSY5Y) provides further support for receptor-mediated transport of Mn across membranes (47). The distribution of Tf receptors in the CNS vis-à-vis Mn accumulation is noteworthy. The thalamic nuclei, the pallidum, as well as the substantia nigra contain the highest brain Mn concentrations (48), as well as appreciable levels of Fe (49). The areas with dense Tf distribution (50) do not correspond to the distribution of brain Mn (or Fe). Yet, the fact that Mn-accumulates in brain areas that are efferent to high Tf-receptor density raises the intriguing possibility that Mn-rich brain sites accumulate Mn via axonal transport (35,39,51). Indeed, the nucleus accumbens and the caudate-putamen— two areas that are abundantly rich in Tf receptors—provide efferent fibers to areas that are Mn rich, such as ventral-pallidum, globus pallidus, and substantia nigra (52,53).


201

The competition between Fe and Mn for the same transport carrier is also noteworthy. For example, plasma Fe overload, significantly decreases the uptake of Mn across the BBB, whereas Fe deficiency is associated with increased CNS burden of Mn (54–56). High dietary Fe intake reduces the concentration of Mn in the CNS (57). In vivo, 6 h of intravenous administration of ferric-hydroxide dextran complex significantly inhibits Mn brain uptake as compared with its uptake in Fe-free dextran-treated rats (55,56). In an additional study, the transport of Mn across the rat BBB was characterized by a single capillary-pass technique (58). Initial rate measurements (at 15 s) of Mn accumulation in rat brains after intra-arterial injections indicated saturation kinetics. Common carotid injection of freshly mixed Mn2+ with Tf at a 1:10 molar ratio did not lead to a significant change in the initial rate of Mn brain levels compared with injection of Mn2+ alone. However, when Mn2+ was incubated at 25ºC in the presence of Tf at a 1:10 ratio for up to 5 d prior to common carotid injection, the initial rate of Mn uptake by brain was incubation-time-dependent, increasing linearly with prolonged incubations. These findings suggest that the saturable component of divalent Mn transport into brain represents but one of the transport mechanisms for Mn across the BBB, and that a second transport system for Mn may occur via a Tf-conjugated Mn-transport system (58). The increase in initial rate of Mn uptake in samples equilibrated with excess Tf for days (therefore presumably oxidized to Mn3+ and complexed in an ironbinding site of Tf) probably reflects binding of MnTf to brain capillary endothelial TfR. However, the internalization of MnTf into brain-capillary epithelia has yet to be demonstrated. Mn shares numerous similarities with Fe. For example: (1) Fe3+ and Mn2+ share a d5 electron configuration, allowing for both to adopt a wide variety of coordination geometries. (2) Both carry similar valence charges (2+ and 3+) in physiological conditions. (3) Both have similar ionic radius. (4) Both strongly bind Tf (47,55,56,59). (5) Intracellularly, both preferentially accumulate in mitochondria (60,61). (6) Both Fe- and Mn can catalyze autoxidation of dopamine in the presence of L-cysteine (62). Given these similarities, it is not surprising that Mn (at least in high concentrations) can interfere with Fe-regulated processes, and in particular certain mitochondrial enzymes (aconitase, NADH-ubiquinone reductase, and succinate dehydrogenase) that require Fe as a cofactor in their active catalytic center. A limited number of studies have addressed the transport kinetics of blood Mn into the CNS. Collectively, these studies suggest that Mn (MnCl2) enters the brain from the blood either across the cerebral capillaries and/or the cerebrospinal fluid (CSF). At normal plasma concentrations, Mn enters the CNS primarily across the capillary endothelium, whereas at high plasma concentrations, transport across the choroid plexus predominates (64,65). These findings are consistent with observations on the rapid appearance and persistent elevation of Mn in the choroid plexus (66,67). Radioactive Mn injected into the blood stream is concentrated in the choroid plexus within 1 h after injection, and 3 d postinjection, Mn is localized to the rat dentate gyrus and CA3 of the hippocampus (68). 3.3. Choroid Plexus Transport of Manganese The choroid plexus is potentially an important site for the homeostasis of Mn. This structure is where 54Mn from injected doses appears first in rodent brain (41,69). The

202

Aschner et al.

regulation of substrate entry into the brain via choroid plexus synthesis of CSF is different from that at the BBB. Capillaries in the choroid plexus are fenestrated, and substances must first be taken up by choroid plexus epithelium. This epithelium then secretes CSF. Eventual neuronal uptake of substances from CSF must proceed via the cells of ependyma. It is thought that tanycytes of the ependyma may play an important role in delivery of nontransferrin bound iron to specific nuclei in the hypothalamus (70). Further study is required on this potential mechanism of delivery of Fe and Mn to neurons. 3.4. Transport of Manganese in a Hypotransferrinemic Mouse Model It has long been appreciated that total body and specifically brain Mn concentrations are elevated with decreased stores of Fe (54,69,71,72). There is evidence for competition between Mn and Fe for intestinal absorption (73,74), probably by way of the transporter DMT-1 (otherwise known as DCT-1 or nramp-2)(75). However, brain Mn is elevated in some disorders of Fe overload, and transferrin receptor is elevated following Mn exposure (63). Because Mn3+ may be an important physiological oxidation state of this metal, and Mn3+ binds with reasonable affinity to Tf, we sought to explore the role of Tf in Mn transport. The hypotransferrinemic mouse (hpx/hpx) provides a unique model to study the role of transferrin in Mn transport in the brain. The hpx/hpx mouse is the result of a spontaneous mutation (76), and has a mRNA splicing defect resulting in virtually no synthesis of Tf protein (77). No difference was found in total brain uptake of sq, ip, or iv injections of 59FeCl3 or 54MnCl between +/+ and hpx/hpx mice (78–80). However, striking differences in 2 regional distribution were noticed for 59Fe, but only subtle differences were noticed in the case of 54Mn (80). This experiment demonstrated a striking effect of Tf on brain distribution of 59Fe: in hpx/hpx mice, iron remained in the choroid plexus 7 d after an iv injection, whereas in +/+ mice Fe was distributed to hippocampus, thalamus, and striatum (80). This observation is consistent with the Tf/TfR system being the predominant mechanism for Fe delivery to the brain, with a minor contribution coming from non-Tf-mediated Fe uptake by the choroid plexus. Although one report exists of low Tf mRNA in human choroid plexus (81), we hypothesize that choroid plexus synthesizes and secretes Tf to the CSF. Slightly lower levels of 54Mn in cerebral cortex and corpus collosum were observed in the hpx/hpx mice relative to +/+ controls. These subtle differences may be due to differences in the brain development of the hpx/hpx mice (78) and not strictly a Tf effect. Clearly Tf is not required for parenchymal Mn delivery to the extent that it is for Fe, at least in the hypotransferrinemic mouse. The most parsimonious interpretation of these data is that Tf is required for delivery of Fe to brain parenchyma, but that the Tf/TfR system is less important for delivery of Mn to brain parenchyma in mice. The data suggest that nonTf dependent mechanisms for Mn transport exist and that these mechanisms are unmasked in the hpx/hpx mutant. At this writing, no unique mammalian transporters are known for Mn. It remains to be seen what role, if any, iron transporters play in Mn homeostasis. There has been an explosion in recent years in candidate genes implicated in regulating iron transport


203

(69,82). The near future may bring similar breakthroughs for the field of Mn neurotoxicity. 4. MECHANISMS OF MANGANESE NEUROTOXICITY While many studies have suggested that industrial chemicals and pesticides may underlie idiopathic Parkinson’s disease (IPD) (2,7,83–87), its etiology remains elusive. Among the toxic metals, the relationship between Mn intoxication and IPD has long been recognized (5,48,54,84,88–90). The neurological signs of manganism have received close attention because they resemble several clinical disorders collectively described as “extrapyramidal motor system dysfunction,” and in particular, IPD and dystonia. Unlike Parkinsonism, manganism also produces dystonia, a neurological sign associated with damage to the globus pallidus (21,91). A comprehensive survey of patients afflicted by PD or manganism concludes that although similar in many respects, there are distinct differences between the two neurological disorders. Similarities between PD and manganism include the presence of generalized bradykinesia and widespread rigidity. Dissimilarities between Parkinson’s disease and manganism were also recognized, notably the following in manganism: (1) A less frequent resting tremor; (2) more frequent dystonia; (3) a particular propensity to fall backwards; (4) failure to achieve a sustained therapeutic response to levodopa; and (5) failure to detect a reduction in fluorodopa uptake by positron emission tomography (PET; for further details see ref. 21). Given these differences, it has been proposed that Mn intoxication is associated with preservation of the nigrostriatal dopaminergic pathway, and that chronic Mn intoxication causes parkinsonism-like effects by damaging output pathways downstream of the nigrostriatal dopaminergic pathway (21,91). The literature is replete with a number of potential mechanisms for Mn-induced neurotoxicity. These include the following: (1) a direct toxic effect of Mn in its divalent oxidation state (or perhaps Mn in a higher oxidation state; 92,93,93a) to dopamine-containing cells (94); (2) a Mn-induced decrease in the content of peroxidase and catalase within the substantia nigra (95); (3) production of superoxide (SO; O2.-), hydrogen peroxide (H2O2), or hydroxyl free (–OH) radicals by Mn, which in turn, “attack” dopamine, dopaminergic cells, and dopamine receptors (96–99); (4) production of 6-hydroxydopamine or other toxic catecholamines by Mn2+ and a decrease in protective thiols (96,100–102); (5) auto-oxidation of dopamine, leading to formation of toxic (semi) quinones, concomitantly depleting tissue dopamine (92,101,103–106); and (6) an excitotoxic mechanism in which the activation of glutamate-gated cation channels contributes to neuronal degeneration (107). As recently proposed by Verity (108), understanding of the pathogenesis of Mn neurotoxicity likely will have to incorporate a number of considerations/mechanisms: (1) the factors controlling Mn2+ uptake and distribution into the brain; (2) account for the apparent selectivity of dopaminergic neurons. (3) account for the role of mitochondrial dysfunction; and (4) account for the role of oxidative injury in the genesis of toxicity. Mn-induced neurotoxicity is likely a multi-factor process that in addition to coincident transport disturbances of iron may also affect the transport of aluminum and perhaps other metals as well. As evident from the previous discussion on mechanisms of injury, the selectivity of dopaminergic neurons will also need to be considered.

204

Aschner et al.

REFERENCES 1. Kawamura, R., Ikuta, H., Fukuzumi, S., Yamada, R., Tsubaki, S., Kodama, T., and Kurata, S. (1941) Intoxication by manganese in well water. Kitasato Arch. Exp. Med. 18, 145–169. 1a. Keen, C. L. and Zidenberg-Cherr, S. (1994) Manganese toxicity in humans and experimental animals, in Manganese in Health and Disease, (Klimis-Tavantzis, D. J., ed.), CRC Press, Boca Raton, pp. 193–205. 2. Ferraz, H. B., Bertolucci, P. H. F., Pereira, J. S., Lima, J. G. C., and Andrade, L. A. F. (1988) Chronic exposure to the fungicide maneb may produce symptoms and signs of CNS manganese intoxication. Neurology 38, 550–553. 3. Ensing, J. G. (1985) Bazooka: cocaine-base and manganese carbonate. J. Anal. Toxicol. 9, 45–46. 4. Davis, D. W., Hsiao, K., Ingels, R., and Shikiya, J. (1988) Origins of manganese in air particulates in California. JAPCA 38, 1152–1157. 5. Chandra, S. V., Shukla, G. S., Srivastawa, R. S., Singh, H., and Gupta, V. P. (1981) An exploratory study of manganese exposure to welders. Clin. Toxicol. 18, 407–418. 6. Roels, H., Lauwerys, R., Buchet, J. P., Genet, P., Sarhan, M. J., Hanotiau, I., et al. (1987) Epidemiological survey among workers exposed to manganese: effects on lung, central nervous system, and some biological indices. Am. J. Ind. Med. 11, 307–327. 7. Wang, J. D, Huang, C. C., Hwang, Y. H., Chiang, J. R., Lin, J. M., and Chen, J. S. (1989) Manganese induced parkinsonism: an outbreak due to un-repaired ventilation control system in a ferromaganese smelter. Br. J. Ind. Med. 46, 856–859. 8. Lynam, D. R., Roos, J. W., Pfeifer, G. D., Fort, B. F., and Pullin, T. G. (1999) Environmental effects and exposure to manganese from use of methylcyclopentadienyl manganese tricarbonyl (MMT) in gasoline. Neurotoxicology 20, 145–150. 9. Zayed, J., Hong, B., and L’Esperance, G. (1999) Characterization of manganese-containing particles collected from the exhaust emissions of automobiles running with MMT additive. Env. Sci. Technol. 33, 3341–3346. 10. Loranger, S. and Zayed, J. (1997) Environmental contamination and human exposure to airborne total and respirable manganese in Montreal. J. Air Waste Manag. Assoc. 47, 983–989. 11. Pellizazari, E. D., Clayton, C. A., Rodes, C. Mason, R. E., Piper, L., Fort, B. F., et al. (1999) Particulate matter and manganese exposures in Toronto, Canada. Atmos. Environ 33, 721–734. 12. Zayed, J., Thibault, C., Gareau, L., and Kennedy, G. (1999b) Airborne manganese particulates and methylcyclopentadienyl manganese tricarbonyl (MMT) at selected outdoor sites in Montreal. Neurotoxicology 20, 151–157. 13. Andersen, M. E., Gearhart, J. M., and Clewell III, H. J. (1999). Pharmacokinetic data needs to support risk assessments for inhaled and ingested manganese. Neurotoxicology 20, 161–171. 14. Roels, H., Meiers, G., Delos, M., Ortega, I., Lauwerys, R., Buchet, J. P., and Lison, D. (1997) Influence of the route of administration and the chemical form (MnCl2, MnO2) on the absorption and cerebral distribution of manganese in rats. Arch. Toxicol. 71, 223–230. 15. Ulrich, C. E., Rinehart, W., and Brandt, M. (1979) Evaluation of the chronic inhalation toxicity of a manganese oxide aerosol. III. Pulmonary function, electromyograms, limb tremor, and tissue manganese data. Am. Ind. Hyg. Assoc. J. 40, 349–353. 16. Vitarella D, Wong, B. A., Moss, O. R., and Dorman, D. C. (2000) Pharmacokinetics of inhaled manganese phosphate in male Sprague-Dawley rats following subacute (14-day) exposure. Toxicol. Appl. Pharmacol. 163, 279–285. 17. Gibbons, R. A., Dixon, S. N., Hallisk, K., Russell, A. M., Sansom, B. F., and Symonds, H. W. (1976) Manganese metabolism in cows and goats. Biochim Biophys Acta 444, 1–10.


205

17a.Morganti, J. B., Lown, B. A., Stineman, C. H., D’Agostino, R. B., and Massaro, E. J. (1985) Uptake, distribution and behavioral effects of inhalation exposure to manganese (MnO2) in the adult mouse. Neurotoxicology 6, 1–16. 18. Davidsson, L., Lonnerdal, B., Sandstrom, B., Kunz, C., and Keen, C. L. (1989) Identification of transferrin as the major plasma carrier protein for manganese introduced orally or intravenously or after in vitro addition in the rat. J. Nutr. 119, 1461–1464. 19. Malecki, E. A., Devenyi, A. G., Barron, T. F., Mosher, T. J., Eslinger, P. J., FlahertyCraig, C. V., et al. (1999b) Iron and manganese homeostasis in chronic liver disease: relationship to pallidal T1-weighted magnetic resonance signal hyperintensity. Neurotoxicology 20, 647–652. 20. Witzleben, C., Pitlick, P., Bergmeyer, J., and Benoit, R. (1968) Acute manganese overload: a new model of intrahepatic cholestasis. Am. J. Pathol. 53, 409–423. 21. Calne, D. B., Chu, N. S., Huang, C. C., Lu, C. S., and Olanow, W. (1994) Manganism and idiopathic parkinsonism: similarities and difference. Neurology 44, 1583–1586. 22. Nagatomo, S., Umehara, F., Hanada, K., Nobuhara, Y., Takenaga, S., Arimura, K., and Osame, M. (1999) Manganese intoxication during total parenteral nutrition: report of two cases and review of the literature. J. Neurol. Sci. 162, 102–105. 23. Olanow, C.W., Good, P.F., Shinotoh, H., Hewitt, K.A., Vingerhoets, F., Snow, B.J., et al. (1996) Manganese intoxication in the rhesus monkey: a clinical, imaging, pathologic, and biochemical study. Neurology 46, 492–498. 24. Brenneman, K. A., Cattley, R. C., Ali, S. F., and Dorman, D. C. (1999) Manganeseinduced developmental neurotoxicity in the CD rat: is oxidative damage a mechanism of action? Neurotoxicology 20, 477–487. 25. Boyes, W. K. and Miller, D. B. (1998) A review of rodent models of manganese neurotoxicity, Neurotoxicology 19, 468 [Abstract]. 26. Lyden, A., Larsson, B. S., and Lindquist, N. G. (1984) Melanin affinity of manganese. Acta Pharmacol. Toxicol. 55, 133–138. 27. Chandra, S. V. and Shukla, G. S. (1978) Manganese encephalopathy in growing rats. Environ. Res. 15, 28–37. 28. Dorman, D. C, Struve, M. F., Vitarella, D., Byerly, F. I., Goetz, J., and Miller, R. (2000) Neurotoxicity of manganese chloride in neonatal and adult CD rat following subchronic (21-day) high-dose oral exposure. J. Appl. Toxicol. 20, 179–187. 29. Kristensson, K., Eriksson, H., Lindth, B., Plantin, L. O., Wachtmeister, L., elAzazi, M., et al. (1986) Effect of manganese chloride on the developing nervous system, Acta Pharmacol. Toxicol. 59, 345–348. 30. Scheuhammer, A. M. and Cherian, M. G. (1983) The influence of manganese on the distribution of essential trace elements. II. The tissue distribution of manganese, magnesium, zinc, iron, and copper in rats after chronic manganese exposure. J. Toxicol. Environ. Health 12, 361–370. 31. Furchner, J. E., Richmond, C. R., and Drake, G. A. (1966) Comparative metabolism of radionuclides in mammals. 3. retention of manganese–54 in the mouse, rat, monkey and dog. Health Phys. 12, 1415–1423. 32. Cotzias, G. C., Horuichi, K., Fuenzalida, S., and Mena, I. (1968) Chronic manganese poisoning: Clearance of tissue manganese concentrations with persistence of the neurological picture. Neurology 18, 376–382. 33. Takeda, A., Sawashita, J., and Okada, S. (1995) Biological half-lives of zinc and manganese in rat brain, Brain Res. 695, 53–58. 34. Strause, L., Saltman, P., Smith, K., Bracker, M., and Andon, M. (1994) Spinal bone loss in postmenopausal women supplemented with calcium and trace minerals. J. Nutr. 124, 1060–1064. 35. Gianutsos, G., Morrow, G. R., and Morris, J. B. (1997) Accumulation of manganese in rat brain following intranasal administration. Fund. Appl. Toxicol. 37, 102–105.

206

Aschner et al.

36. Tjälve, H. and Henriksson, J. (1999) Uptake of metals in the brain via olfactory pathways. Neurotoxicology 20, 181–195. 37. Brenneman, K. A., Wong, B. A., Bucelloto, M. A., Costa, E. R., Gross, E. A., and Dorman, D. C. (2000) Direct olfactory transport of inhaled manganese (54MnCl2) to the rat brain: Toxicokinetic investigations in a unilateral nasal occlusion model. Toxicol. Appl. Pharmacol. 169, 238–248. 38. Pautler, R. G., Silva, A. C., and Koretsky, A. P. (1998) In vivo neuronal tract tracing using manganese-enhanced magnetic resonance imaging. Magn. Reson. Med. 40, 740–748. 39. Sloot, W. N. and Gramsbergen, J. B. (1984) Axonal transport of manganese and its relevance to selective neurotoxicity in the rat basal ganglia. Brain Res. 657, 124–132. 40. Henriksson, J., Tallkvist, J., and Tjälve, H. (1999) Transport of manganese via the olfactory pathway in rats: dosage dependency of the uptake and subcellular distribution of the metal in the olfactory epithelium and brain. Toxicol. Appl. Pharmacol. 156, 119–128. 41. Takeda A., Sawashita, J., and Okada, S. (1998) Manganese concentration in rat brain: manganese transport from the peripheral tissues. Neurosci. Lett. 242, 45–48. 42. Foradori, A. C., Bertinchamps, A., Gulibon, J. M., and Cotzias, G. C. (1967) The discrimination between magnesium and manganese by serum proteins. J. Gen. Physiol. 50, 2255–2266. 43. Aisen, P., Aasa, R., and Redfield, A. G. (1969) The chromium, manganese, and cobalt complexes of transferrin. J. Biol. Chem. 244, 4628–4633. 44. Fishman, J. B., Handrahan, J. B., Rubir, J. B., Connor, J. R., and Fine, R. E. (1985) Receptor-mediated trancytosis of transferrin across the blood-brain barrier. J. Cell Biol. 101, 423A. 45. Jeffries, W. A., Brandon, M. R., Hunt, S. V., Williams, A. F., and Mason, D. Y. (1984) Transferrin receptor on endothelium of brain capillaries. Nature 132, 162–163. 46. Partridge, W. M., Eisenberg, J., and Yang, J. (1987) Human blood-brain barrier transferrin receptor. Metabolism 36, 892–895. 47. Suarez, N. and Eriksson, H. (1993) Receptor-mediated endocytosis of a manganese complex of transferrin into neuroblastoma (SHSY5Y) cells in culture. J. Neurochem. 61, 127–131. 48. Barbeau, A., Inoué, N., and Cloutier, T. (1976) Role of manganese in dystonia. Adv. Neurol. 14, 339–352. 49. Hill, J. M. and Switzer, R. C., III. (1984) The regional distribution and cellular localization of iron in the rat brain. Neuroscience 11, 595–603. 50. Hill, J. M., Ruff, M. R., and Weber, R. J. (1985) Transferrin receptors in rat brain: Neuropeptide-like pattern and relationship to iron distribution. Proc. Natl. Acad. Sci. USA 82, 4553–4557. 51. Santamaria, A., Rios, C., Perez, P., Flores, A., Galvan-Arzate, S., Osorio-Rico L., and Solis F. (1996) Quinolinic acid neurotoxicity: in vivo increased copper and manganese content in rat corpus striatum after quinolinate intrastriatal injection. Toxicol. Lett. 87, 113–119. 52. Walaas, I. and Fonnum, F. (1979) The distribution and origin of glutamate decarboxylase and choline acetyltransferase in ventral pallidum and other basal forebrain regions. Brain Res. 177, 325–336. 53. Nagy, J. I., Carter, D. A., and Fibiger, H. C. (1978) Evidence for a GABA-containing projection from the enopenduncular nucleus to the lateral habenula in the rat. Brain Res. 145, 360–364. 54. Mena, I., Horiuchi, K., Burke, K., and Cotzias, G. C. (1969) Chronic manganese poisoning: individual susceptibility and absorption of iron. Neurology 19, 1000–1006. 55. Aschner, M. and Aschner, J. L. (1990) Manganese transport across the blood-brain barrier: relationship to iron homeostasis. Brain Res. Bull. 24, 857–860.


207

56. Aschner, M. and Aschner, J. L. (1991) Manganese neurotoxicity: cellular effects and blood-brain barrier transport. Neurosci. Biobehav. Rev. 15, 333–340. 57. Diez-Ewald, M., Weintraub, L. R., and Crosby, W. H. (1968) Inter relationship of iron and manganese metabolism. Proc. Soc. Exp. Biol. Med. 129, 448–151. 58. Aschner, M. and Gannon, M. (1994) Manganese (Mn) transport across the rat blood-brain barrier: saturable and transferrin-dependent transport mechanisms. Brain Res. Bull. 33, 345–349. 59. Ueda, F., Raja, K. B., Simpson, R. J., Trowbridge, I. S., and Bradbury, M. W. B. (1993) Rate of [59]Fe uptake into brain a cerebrospinal fluid and the influence thereon of antibodies against the transferrin receptor. J. Neurochem. 60, 106–113. 60. Grafstein, B. and Forman, D. S. (1980) Intracellular transport in neurons. Physiol. Rev. 60, 1167–1283. 61. Gavin, C. E., Gunter, K. K., and Gunter, T. E. (1994) Manganese and calcium efflux kinetics in brain mitochondria. Relevance to manganese toxicity. Biochem. J. 266, 329–334. 62. Shen, X. M. and Dryhurst, G. (1998) Iron—and manganese-catalyzed autoxidation of dopamine in the presence of L-cysteine: possible insights into iron—and manganesemediated dopaminergic neurotoxicity. Chem. Res. Toxicol. 11, 824–837. 63. Zheng, W., Zhao, Q., Slavkovich, V., Aschner, M., and Graziano, J. H. (1999) Alteration of iron homeostasis following chronic exposure to manganese in rats. Brain Res. 833, 125–132. 64. Murphy, V. A., Wadhwani, K. C., Smith, Q. R., and Rapoport, S. I. (1991) Saturable transport of manganese (II) across the rat blood-brain barrier. J. Neurochem. 57, 948–954. 65. Rabin, O., Hegedus, L., Bourre, J. M., and Smith, Q. R. (1993) Rapid brain uptake of manganese (II) across the blood-brain barrier. J. Neurochem. 61, 509–517. 66. London, R. E., Toney, G., Gabel, S. A., and Funk, A. (1989) Magnetic resonance imaging studies of the brains of anesthetized rats treated with manganese chloride. Brain Res. Bull. 23, 229–235. 67. Ingersoll, R. T., Montgomery, E. B., Jr., and Aposhian, H. V. (1995) Central nervous system toxicity of manganese. I. Inhibition of spontaneous motor activity in rats after intrathecal administration of manganese chloride. Toxicol. Appl. Pharmacol. 27, 106–113. 68. Takeda, A., Sawashita, J., and Okada, S. (1994) Localization in rat brain of the trace metals, zinc and manganese, after intracerebroventricular injection. Brain Res. 658, 252–254. 69. Malecki, E. A., Devenyi, A. G., Beard, J. L., and Connor, J. R. (1999c) Existing and emerging mechanisms for iron and manganese transport to the brain. J. Neurosci. Res. 56, 113–122. 70. Burdo, J., Martin, J., Menzies, S., Dolan, K., Romano, M., Fletcher, R., et al. (1999) Cellular distribution of iron in the brain of the Belgrade rat. Neuroscience 93, 1189–1196. 71. Shukla, A., Agarwal, K. N., and Shukla, G. S. (1989) Effect of latent iron deficiency on metal levels of rat brain regions. Biol. Trace Elem. Res. 22, 141–151. 72. Yokoi, K., Kimura, M., and Itokawa, Y. (1991) Effect of dietary iron deficiency on mineral levels in tissues of rats. Biol. Trace Elem. Res. 29, 257–265. 73. Thomson, A. B. R. and Valberg, L. S. (1972) Intestinal uptake of iron, cobalt, and manganese in the iron-deficient rat. Am. J. Physiol. 223, 1327–1329. 74. Davis, C. D., Malecki, E. A., and Greger, J. L. (1992) Interactions among dietary manganese, heme iron, and nonheme iron in women. Am. J. Clin. Nutr. 56, 926–932. 75. Gunshin, H., Mackenzie, B., Berger, U., Gunshin, Y., Romero, M., Boron, W., et al. (1997) Cloning and characterization of a mammalian proton-coupled metal-ion transporter. Nature 338, 482–488.

208

Aschner et al.

76. Bernstein, S. E. (1987) Hereditary hypotransferrinemia with hemosiderosis, a murine disorder resembling human atransferrinemia. J. Lab. Clin. Med. 110, 690–705. 77. Huggenvik, J. I., Craven, C. M., Idzerda, R. L., Bernstein, S., Kaplan, J., and McKnight, G. S. (1989) A splicing defect in the mouse transferrin gene leads to congenital atransferrinemia. Blood 74, 482–486. 78. Dickinson, T. K. and Connor, J. R. (1994) Histological analysis of selected brain regions of hypotransferrinemic mice. Brain Res. 635, 169–178. 79. Malecki, E. A., Devenyi, A. G., Beard, J. L., and Connor, J.R. (1998) Transferrin response in normal and iron-deficient mice heterozygotic for hypotransferrinemia: effects on iron and manganese accumulation. BioMetals 11, 265–276. 80. Malecki, E. A., Cook, B. M., Devenyi, A. G., Beard, J. L., and Connor, J. R. (1999a) Transferrin is required for normal distribution of 59Fe and 54Mn in brains of mice. J. Neurol. Sci. 170, 112–118. 81. Tu, G., Achen, M., Aldred, A., Southwell, B., and Schreiber, G. (1991) The distribution of cerebral expression of the transferrin gene is species specific. J. Biol. Chem. 266, 6201–6208. 82. Richardson, D. R. and Ponka, P. (1997) The molecular mechanisms of the metabolism and transport of iron in normal and neoplastic cells. Biochim. Biophys. Acta 1331, 1–40. 83. Cotzias, G. C. (1958) Manganese in health and disease. Physiol. Rev. 38, 503–532. 84. Gorell, J. M., Johnson, C. C., Rybicki, B. A., Peterson, E. L., Kortsha, G. X., Brown, G. G., and Richardson, R. J. (1997) Occupational exposures to metals as risk factors for Parkinson’s disease. Neurology 48, 650–658. 85. Rybicki, B. A., Johnson, C. C., Uman, J., and Gorell, J. M. (1993) Parkinson’s disease mortality and the industrial use of heavy metals in Michigan. Movement Disorders 8, 87–92. 86. Semchuk, K. M, Love, E. J., and Lee, R. G. (1992) Parkinson’s disease and exposure to agricultural work and pesticide chemicals. Neurology 42, 1328–1335. 87. Tanner, C. M. (1989) The role of environmental toxins in the etiology of Parkinson’s disease. Trends Neurosci. 12, 49–54. 88. Barbeau, A. (1985) Manganese and extrapyramidal disorders. Neurotoxicology 5, 13–16. 89. Mena, I., Court, J., Fuenzalida, S., Papavasiliou, P. S., and Cotzias, G. C. (1974) Modification of chronic manganese poisoning. Treatment with L-dopa or 5-OH tryptophane. N. Engl. J. Med. 282, 5–10. 90. Tepper, L. B. (1961) Hazards to health: manganese. N. Engl. J. Med. 264, 347–348. 91. Pal, P. K., Samii, A., and Calne, D. B. (1999) Manganese neurotoxicity: a review of clinical features, imaging and pathology. Neurotoxicology 20, 227–238. 92. Donaldson, J., McGregor, D., and Labella, F. S. (1982) Manganese neurotoxicity: a model for free radical mediated neurodegeneration? Can. J. Physiol. Pharmacol. 60, 1398–1405. 93. Archibald, F. S., and Tyree, C. (1987) Manganese poisoning and the attack of trivalent manganese upon catecholamines. Arch. Biochem. Biophys. 256, 638–650. 93a. Segura-Aguilar, J. and Lind, C. (1989) On the mechanism of the Mn3(+)-induced neurotoxicity of dopamine: prevention of quinone-derived oxygen toxicity by DT diaphorase and superoxide dismutase. Chem. Biol. Interact. 72, 309–324. 94. Parenti, M., Flauto, C., Parati, E., Vescovi, A., and Groppetti, A. (1988) Manganese neurotoxicity: effect of L-DOPA and pargyline treatments. Brain Res. 367, 8–13. 95. Ambani, L. M., Vanwoert, M. H., and Murphy, S. (1975) Brain peroxidase and catalase in Parkinson’s disease. Arch. Neurol. 32, 114–118. 96. Cohen, G. and Heikkila, R. E. (1974) The generation of hydrogen peroxide, superoxide radical, and hydroxyl radical by 6-hydroxydopamine, dialuric acid, and related cytotoxic agents. J. Biol. Chem. 249, 2447–2452. 97. Cohen, G. (1984) Oxy-radical toxicity in catecholamine neurons. Neurotoxicology 5,77–82.


209

98. Marinho, C. R. and Manso, C. F. (1993) O2 generation during neuromelanin synthesis. The action of manganese. Acta Med. Portug. 6, 547–554. 99. Sun, A. Y., Yang, W. L., and Kim, H. D. (1993) Free radical and lipid peroxidation in manganese-induced neuronal cell injury. Annal. NY Acad. Sci. 679, 358–363. 100. Graham, D. G., Tiffany, S. M., Bell, W. R., Jr., and Gutknecht, W. F. (1978) Autooxidation versus covalent binding of quinones as the mechanism of toxicity of dopamine, 6-hydroxydopamine, and related compounds toward C 1300 neuroblastoma cells in vitro. Mol. Pharmacol. 14, 644–653. 101. Graham, D. G. (1984) Catecholamine toxicity: a proposal for the molecular pathogenesis of manganese neurotoxicity and Parkinson’s disease. Neurotoxicology 5, 83–96. 102. Perry, T. L., Godin, D. V., and Hansen, S. (1982) Parkinson’s disease: a disease due to nigral glutathione deficiency? Neurosci. Lett. 33, 305–310. 103. Donaldson, J. (1987) The physiopathologic significance of manganese in brain: its relation to schizophrenia and neurodegenerative disorders. Neurotoxicology 8, 451–462. 104. Garner, C. D. and Nachtman, J. P. (1989) Manganese catalyzed auto-oxidation of dopamine to 6-hydroxydopamine in vitro. Chem. Biol. Interactions 69, 345–351. 105. Millar, D. M., Buttner, G. R., and Aust, S. D. (1990) Transition metals as catalysts of “autooxidation” reactions. Free Rad. Biol. Med. 8, 95–108. 106. Roy, B. P., Paice, M. G., Archibald, F. S., Misra, S. K., and Misiak, L. E. (1994) Creation of metal-complexing agents, reduction of manganese dioxide, and promotion of manganese peroxidase-mediated Mn(III) production by cellobiose: quinone oxidoreductase from Trametes versicolor. J. Biol. Chem. 269, 19,745–19,750. 107. Brouillet, E. P., Shinobu, L., McGarvey, U., Hochberg, F., and Beal, M. F. (1993) Manganese injection into the rat striatum produces excitotoxic lesions by impairing energy metabolism. Exp. Neurol. 120, 89–94. 108. Verity, M. A. (1999) Manganese neurotoxicity: a mechanistic hypothesis. Neurotoxicology 20, 489–497.

210

Aschner et al.

Aluminum Neurotoxicity

211

12 Aluminum Neurotoxicity Andrzej Szutowicz 1. INTRODUCTION 1.1. Environmental Sources of Aluminum Aluminum (Al) is the third most abundant element in the Earth’s outer crust, constituting about 8% of its weight. The highest content of Al is in the form of oxides and aluminosilicates found in clays and other secondary minerals (3–45%). In sedimentary rocks such as shales, sandstones, and carbonates, Al content varies from 8% to below 1%, respectively. Soil/sediment Al species form a large pool, from which it is released to surface waters. The water soluble pool of Al constitutes an extremely small fraction of the metal in the environment. Nevertheless, the latter serves as a main source of Al for living organisms (1). Aluminum is a strongly hydrolyzing metal forming in neutral pH insoluble hydroxides. In acidic surface waters, Al content may rise dramatically owing to formation of soluble complexes with various inorganic ligands such as F-, HCO3-, SO42-, PO33-, and H4SiO4 that are released from sediments. The composition of these Al species depends on the character of soils from which it is extracted. Acid rain and acid industrial wastes may significantly contribute to this phenomenon. Hence, acid conditions facilitate Al entry into the living biomass pool. Al is not an essential nutrient for living organisms. Its concentrations in plants and animals are usually low, varying from 0.002 to 0.00005%, respectively. Some plants, however, such as tea leaves, contain more than 0.1% Al by dry weight (1). Thus, most organisms tend to accumulate increased amounts of Al when it is present in drinking water. Environmental surface-water pollution with Al was blamed for toxic effects on fish associated with coaglulation of mucus on their gills, and disruption of osmotic balance. On the other hand, Al hydroxides may be an important factor regulating pH in acidic waters (1). 1.2. Al Absorption Daily meals and liquids consumed contain about 14 and 0.1 mg of Al, respectively. Insoluble Al species are solubilized by acid pH in the stomach. Absorption of Al occurs mainly in the jejunum and duodenum. The mucosal cells may take up Al presumably by calcium (Ca) channels, but intracellular complexation may create barrier to its


211

212

Szutowicz

absorption. There are also strong suggestions that Al enters portal circulation by the paracellular pathway. It is estimated that 0.2% (0.2 µg) of soluble and 0.05% (7 µg) of food Al is absorbed daily (2). Several small molecular-weight ligands modify Al absorption, retention, and in effect its toxicity. Dietary organic and inorganic anions such as citrate, ascorbate, carbonate, and to a lesser degree lactate, oxalate, and tartarate enhance Al absorption in experimental animals by formation of soluble complexes. Thus, beverages and fruits rich in citric acid were found to increase Al absorption in humans by several-fold (2,3). On the other hand, fluoride, silicate-forming insoluble complexes inhibit this process. High iron and Ca content in the diet suppress Al transport, presumably by interference with common transport mechanisms via transferrin receptors and Ca-channels, respectively. Neither rural or urban air nor Al-reach antiperspirants significantly contribute to metal absorption. For Al absorbed in intestines, the liver forms a first barrier protecting the brain from direct Al exposure (2). Several factors may increase Al intake significantly. It is estimated that patients using Al-containing antiacids or phosphate binders may absorb up to 5000 µg of Al daily. Industrial Al-rich dusts, fumes, and flakes were reported to increase metal absorption to several hundreds of micrograms either through lungs or directly into cerebrospinal fluid (CSF) through the olphactory route (2). Dialysis patients accumulate excessive amounts of Al if dialysis fluids contain too high concentrations of this metal (see Subheading 7.2.). The content of Al in preparations used for parenteral nutrition vary from 40–135 µg/l. In these patients yield of absorption is 100% and daily burden depends only on the volume of administrated nutrients. Premature infants receiving commercial intravenous feeding solutions were reported to accumulate excessive amounts of Al (see Subheading 7.2.). 1.3. Speciation of Al in Body Fluids In people with normal renal function, Al level in plasma does not exceed 0.5 µM (14 µg/L)(4). In these conditions, over 80% of plasma Al is bound to proteins, almost exclusively to transferrin. Concentrations of Fe3+ in plasma and its affinity to transferrin are one to two orders of magnitude higher than those of Al3+ (5). However, it does not affect formation of Al-transferrin complex since a large portion of the protein-binding sites (50 µM) is free of iron under physiologic conditions. Remaining Al forms different low molecular-weight complexes including Al(PO4)OH– (16%) and Al(Hcitrate)2– (2%) soluble anionic microforms. Each of the remaining microforms including Al(OH)3, Al(OH)4– and AlPO4 and many others constitute negligible (less than 1%) fractions of the metal in the extracellular compartment (5). The kidneys are a primary route of Al excretion. Al content in the urine is a good index of recent exposure to the metal. However, it does not reflect Al accumulation in tissues, including brain, as its elimination from intracellular binding sites is very slow. Ligand-exchange rates for Al complexes were found to be 105 and 108 times slower than those for Mg2+ and Ca2+, respectively (6). Only a small amount of Al is eliminated with bile. Drinks containing citric acid increase Al uptake but also facilitate its elimination in urine (2,3,6). Concentrations of Al in intracellular compartments are much higher than those present in plasma. The highest concentrations of Al accumulate in lungs and bones (about 400 µmol/kg wet weight). In other tissues, including brain, Al levels are much


213

lower varying from 10–40 µmol/kg (2). Intracellular phosphate concentration is five times higher than in serum (6). Also protein concentrations, particularly their phosphorylated forms, as well as ATP and other low molecular-weight ligand, are high, facilitating intracellular deposition of Al and slowing down its elimination. In such conditions, Al may compete with Mg2+ and Ca2+ for common intracellular-binding sites, leading to disruption of signal transduction and cell homeostasis (see Subheading 3.) (7). 2. Al TRANSPORT IN THE BRAIN 2.1. Effects on BBB Al-transferrin complex may penetrate the BBB epithelial cells through the surface high-affinity transferrin receptors, a system postulated to provide iron to neurons and glial cells (8). Animal experiments have shown that parenteral Al application may increase access of labeled sucrose, quinidine, β-endorphin, and various small peptides into the brain space (see ref. 9 for review). These changes seem to depend on the form of Al used in experiments. Increase in permeability was observed after injection of Al acetylacetonate or maltolate but not after lactate salts. It was concluded that these differences depended on the metal-coordination sphere (10). On the other hand, Al did not influence the permeability of the BBB for albumin. Studies performed on monolayer cultures of brain-vessel endothelial cells confirmed results on Al-evoked increase of permeability to low molecular-weight particles as well as attenuation of pinocytosis. It is possible that interaction of Al with plasma-membrane phospholipids may decrease their fluidity and increase lipophilicity, yielding a rise in their permeability (9,11,12). However, there is no clear indication that Al contributes to increased permeability of the BBB in humans suffering from encephalopathies caused by or associated with hyperaluminemia. Clearance of Al from various tissues including brain is very slow. Al is transported from brain to the blood in form of citrate complex by monocarboxylic acid transporter located at the BBB (13). 2.2. Transport into Neurons and Glial Cells Adhesion of Al into the outer surface of plasma membrane may trigger metal entry into neuronal cells by endocytosis. There is however, a general agreement that the bulk of Al transport into neurons and glial cells is mediated by transferrin (14,15). Culture studies have shown that at neutral pH Al-transferrin complexes are internalized into neuroblastoma and oligodendroglial cells 4–10 times faster than low molecular-weight Al species. Almost 70% of internalized Al was found in postmitochondrial fraction (cytosol, endoplasmic reticulum, lisosomes), 10% in mitochondria, and 20% in nuclear fraction (14). Intracellular concentrations of Al reached very high levels: two orders of magnitude higher than in the extracellular space (9). The uptake process was found to be energy-independent. Sequestration of Al by several intracellular-binding sites such as tubulin, tau protein, and other structural proteins, particularly by their phosphorylated forms, makes its high accumulation possible. Also phosphate and its low molecular-weight compounds, intracellular concentrations of which are several times higher than those in plasma may bind significant amounts of Al (6). At such high concentrations Al may easily displace Ca, Mg, and Fe from their intracellular-binding sites,

214

Szutowicz

thereby affecting several processes regulated by these cation complexes (6,9). The increase of intracellular free Ca2+ and Fe3+ cations may trigger excitotoxic events leading to cell damage. Like neurons, glial cells accumulate Al in the form of complex with transferrin but fail to do so with AlCl3 or Al-citrate (15). There is no explanation why in in vivo experiments Al is preferentially accumulated by neurons (9). In rats chronically intoxicated by subcutaneous applications of Al-glutamate, the highest concentrations of this metal were found in brain cortex, hippocampus, and amygdala, regions containing a high density of transferrin receptors and neurons with a glutamatergic phenotype (10). Similar distribution of Al deposits was reported in postmortem studies on the brains of Alzheimer’s disease (AD) patients (10). On the basis of these findings, the hypothesis was set forth that glutamic acid, forming a stable complex with Al, could no longer serve as an acceptor for ammonia, thus resulting in neuronal death (10,16). There are suggestions that primary olfactory neurons that are in contact with the nasal lumen may take up some aluminum dusts and volatile compounds from inhaled contaminated industrial air into the brain. It seems, however, that it cannot continue its passage into secondary neurons in the olfactory bulb (17). 2.3. Interactions with Cell Membranes Al has the ability to form mono- and polynuclear hydroxyl cationic species that are reported to bind with synthetic phospholipid bilayers, purified proteins, and biological membranes as well as with intact cells. Therefore, the binding of Al to cell membranes is likely to evoke both nonspecific and specific effects resulting from decrease of membrane fluidity as well as from functional changes of particular integral membrane proteins (6,17). Al is able to complex phosphate groups of membrane phospholipids. It has been postulated that polynuclear Al hydroxyl cations, due to their size, might be much more effective in decreasing membrane fluidity than the mononuclear Al ions (17). In addition, large polynuclear Al ions could act as crosslinkers between two neighboring membrane proteins, therefore they could affect transport and signal-transduction functions of the biological membranes (see Subheading 3.6.). There are also evidences that Al(PO4)OH– anion may be an active compound. In vitro experiments have shown that the inhibitory effect of Al on synaptosomal metabolism disappeared when phosphate was omitted from the incubation medium (18). 3. MOLECULAR BASIS OF AL NEUROTOXICITY 3.1. Interference with Ca Homeostasis The fact that Al binds with similar or greater affinity and stability than Ca to the same extra and intracellular binding sites implies a primary mechanism of its neurotoxic effects (6). Extracellular Al blocked fast phase of Ca influx through voltagegated Ca channels in synaptosomes. Half maximal effect was observed at 0.2 mM and total blockade at 1.0 mM Al concentration. It was found to be a competitive inhibitor of Ca binding to the calcium sites inside the channel (19). Fast phase of Ca entry plays a principal role in depolarization-evoked transmitter release. Thereby, Al would interfere with this basic neuronal function. Recent studies provided evidence that


215

Al(PO4)OH– anion may interact with verapamil-binding sites on voltage-dependent Ca-channels in synaptosomal plasma membrane, resulting in decrease of Ca-dependent acetylcholine release (Fig. 1) (18). On the contrary, after accumulation inside the cells, Al was found to increase [Ca2+] in cytoplasm due to its removal from protein and low molecular weight binding sites (20). The inhibition by intracellular Al of the ATPdependent 45Ca uptake by endoplasmic reticulum and synaptosomal plasma membranes points to a complementary mechanism leading to the rise of cytoplasmic [Ca2+] (21,22). In contrast to uptake studies, Al effectively stimulated Ca-ATPase activity in these membranes. The explanation for these findings is that Al by displacement of Ca2+ and Mg2+ from multiple sites increased enzyme activity but disrupted its transport properties (21). Al also increased accumulation of Ca in isolated brain mitochondria presumably due to inhibition of verapamil-sensitive Na/Ca antiporter. Like in extracellular compartment an active form of intracellular Al appeared to be anionic hydroxyl-phosphate complex (20). The increase of mitochodrial Ca was found to be an important factor in activating permeability-transition state of mitochondria, a commonly accepted initial step of apoptotic cell death (23). Studies performed on hybrid, septum-derived SN56 cholinergic neuroblastoma cells, revealed that Al may accumulate quickly inside the cells and was accompanied by an increase in Ca accumulation. In addition, both Al and Ca loads were 50 and 200% higher in cells, which have been previously highly differentiated with dibutyryl cAMP and all-trans-retinoic acid than in nondifferentiated ones (24). It indicates that also in brain in vivo phenotypic differences evoked by trophic, transmitter, and hormonal signals are likely to make neurons more or less susceptible to aluminum neurotoxicity. These findings demonstrate that two effects of Al on neurons may be mediated by disruption of Ca homeostasis (24) (Fig 1). On the contrary, other data seem to indicate that Al-evoked death of cultured hippocampal neurons is independent of calcium- or glutamate-receptor activation (25). On the other hand, chronic oral administration of Al caused no changes in Ca content in the brain despite a significant rise of Al concentrations in this tissue (26). 3.2. Al-Mediated Oxidative Stress It has been found that brains of rats treated chronically with Al-glutamate had higher levels of Al and higher production of thiobarbituric acid-reactive species. This phenomenon was explained as a result of displacement of Fe from its intracellular binding sites and activation of the Fenton reaction (27,28). In vitro studies on lipid membranes have shown that Al itself did not stimulate membrane lipid peroxidation. However, it markedly facilitated iron-mediated peroxidation of synthetic lipids (29). Myelin appeared to be a preferential target for Al-mediated oxidative damage. In a mouse model of oral Al intoxication during development, over 70% of brain lipid-peroxidation products was found in the myelin fraction (30). This finding is consistent with earlier work showing demeylinization in Al-intoxicated brains. Direct detrimental effect of Al or Fe on survival of cultured rat hippocampal neurons appeared to be minimal. However, Al markedly potentiated Fe-induced oxidative stress and neuronal death during 60 h culture (31). In neuroblastoma Neuro 2A cells, Al

216

Szutowicz

Fig. 1. Targets for aluminum effects on acetyl-CoA and acetylcholine metabolism in cholinergic neuron terminals. Thick solid and open arrows indicate metabolic and transport pathways; thin solid arrows and dashed lines with diamonds mark activatory and inhibitory influences, respectively. Numbers indicate: 1, glucose transport; 2, glycolytic pathway; 3, pyruvate dehydrogense; 4, direct acetyl-CoA transport out of mitochondria; 5, citrate transport out of mitochondria and ATP-citrate lyase; 6, choline acteyltransferase; 7, quantal ACh release; 8, nonquantal ACh release; 9, high affinity choline uptake; 10, voltage-dependent Ca transport; 11, mitochondrial Ca uniporter; 12, mitochondrial Na-Ca antiporter. Abbreviations: Ve, verapamil; Ca-lig., calcium complexes with low and high molecular weight compounds.


217

brought about an increase in iron uptake accompanied by inhibition of cell growth and excessive expression of neurofibrillary-tangle protein. The latter changes could mimic one of the principal hallmarks of AD. It was suggested that Fe overload-evoked oxidative stress may contribute to this phenomenon (32). Excessive production of nitric oxide (NO) and increased levels of nitrosyl radicals are important elements of excitotoxic neuronal injury in a number of encephalopathies. It is supported by studies on subchronic dietary application of excessive amounts of Al to the mice, which caused increase of thiobarbituric acid-reactive substances and lipid peroxides and decrease of glutathione level in the brain. Cotreatment with vitamin E prevented these changes from appearing (33). On the other hand, superoxide dismutase (SOD) activity in rat brain was not affected by the chronic intraperitoneal Al overload (34). Al was found to potentiate NO-evoked Ca overload in SN56 neuroblastoma cells (24). One may assume that by this mechanism Al may aggravate NO/Ca dependent excitotoxic damage of neurons. There is the possibility that Al may increase NO production itself. Acute 3-d and extended 3-wk treatment of rats with Al increased inducible nitric oxide synthase (NOS) activity in the cerebellum. Co-treatment with iron or treatment with iron alone did not change Al effects on enzyme activity. These results indicate that these effects of Al are independent of free-radical changes evoked by Alinduced increase of intracellular Fe (35). Amyloid-β is known to potentiate free-radical formation by stabilizing iron in its more harmful ionic form. Al aggravated these amyloid-β effects presumably by facilitation its aggregation and prolongation of ferrous ion stability (36). Al-stimulated amyloid-β aggregation may facilitate Ca entry into cells, thereby reinforcing its excitotoxic activity by stimulation oxygen- and nitrogen-radicals formation (24,37). 3.3. Aluminum-Evoked Modifications in Structural and Regulatory Proteins Single intracerebral application of Al was found to bring about morphological changes in rabbit-brain neuronal cytoskeleton, that were linked to motor abnormalities, convulsions and subsequent death of experimental animals (9,38). Neurofilaments, the basic structural elements of neurons, are composed of three 65(L), 160(M), and 200(H) kDa subunits (39). They are synthesized in perikaryon and continuously transported along the axon by slow axonal transport. Phosphorylation and dephosphorylation of neurofilament proteins plays an important role in their transport. The H subunit usually unphosphorylated in cell body becomes phosphorylated during its axonal transport. The reverse is the case for L and M subunits. In rabbits with aluminum encephalopathy, there were abnormally strong staining of neuronal perikaryons and axons with antibodies directed to their phosphorylated epitopes. It implied the existence of excessive phosphorylation of these cytoskeletal proteins and disturbances in slow axonal transport (40). On the other hand, fast axonal transport was not affected by Al. Alevoked neurofibrillary degeneration changes in rabbits were immunoreactive with antitau and anti-microtubule associated protein-2 (MAP2) antibodies indicating their similarity to Alzheimer’s neuropathology (41,42). On the contrary, rats were shown to be resistant to encephalopathogenic Al effects. Nevertheless, among a wide range of brain homogenate protein, two of them, MAP-2 and H neurofilament subunit, were

218

Szutowicz

found to be phosphorylated to a greater extent in Al-treated rats than in control rats (43). Cultured brain neurons develop neurofibrillary malformations when exposed to Al. However, only 5–20% of them develop changes that morphologically resembled neurofibrillary tangles seen in autopsied Alzheimer’s brains. In addition, they did not react with antibodies directed against MAP-2 and tau epitopes of Alzheimer’s tangles. Nevertheless, increased accumulation of neurofilament subunits H in perikarya of Altreated NB2a neuroblastoma cells was evident despite the absence of any changes in their morphology (44). In addition, neurofilament subunits when isolated from Altreated cells appeared to be resistant to degradation by phosphatases and calcium-dependent proteases (45). It has been shown that Al co-injected with human paired helical filaments (PHF) tau into rodent brains increased formation of their long-living aggregates with amyloid-β, ubiquitin, apolipoprotein-E, and alpha 1-antichymotripsin (46). It clearly suggests that Al contributes to disturbances in turnover and axonal transport of neurofilament proteins in those cells as well as their participation in plaques fromation (47). Under in vitro conditions, high concentrations of Al were found to bring about aggregation of neurofilament and MAP-phosphorylated proteins, possibly due to the occupation of their high and low-affinity Ca-binding sites. It also caused tubulin polymerization by formation of Al-GTP-tubulin complexes, which were more durable than natural Mg-GTP-tubulin complexes (48). A recently developed staining method of autopsy samples of Alzheimer’s disease (AD)-affected brains clearly demonstrated association of Al with neurofibrillary tangles phosphorylated tau protein (49). There is also evidence that in vitro aggregation of isolated human tau was lost on its dephosphorylation (49,50). Taken together, these data indicate that phosphorylation-dependent direct binding of Al with paired helical filaments takes place in AD brains. 3.4. Actions on the Level of Neuronal Nucleus A large fraction of Al present in AD and experimentally Al-overloaded animal brain was found to accumulate in neuronal nuclei. It has been found that as much as 65 Al ions/200 bp of DNA may be present in AD-affected neocortical nuclei (51). One must stress that fluorescent staining with morin revealed Al presence on chromatin of cells without signs of degeneration. However, only a small portion is bound to DNA. It is estimated that only 0.15 nmol of Al is bound per one mole of phosphate DNA (52). This is because of the fact that the affinity constant of DNA binding to Al is lowest of all commonly known intracellular ligands. Hence, DNA may serve merely as polyelectrolyte interacting with Al weakly and nonspecifically (6). Therefore, a high concentration of Al in cell nuclei may reflect its binding to nuclear phosphorylated proteins and low molecular-weight binding sites (phosphorylated histones, Pi, ADP, ATP) (6). It is possible that Al crosslinks not only nuclear proteins themselves but also proteins and nucleic acids. Other studies revealed that irrespective of its weak binding with DNA, Al inhibited digestion of intact chromatin by exogenous nucleases (53). The brain chromatin appeared to be more sensitive to this suppression than chromatin isolated from the liver. On the other hand, Al failed to inhibit digestion of naked DNA,which excludes the possibility of its direct inhibitory effect on nucleases. Hence, this finding was interpreted as caused by interference of Al with accessibility of exonucleases to their substrate (53). Histones form a family of lysine-rich nuclear pro-


219

teins that display a general gene-repressor activity. Al addition to isolated human cell nuclei increased binding of linker histones with DNA (54). An inverse relationship between histone binding to chromatin and gene activity was demonstrated in cultured neurons and neuroblastoma cells (55). These data are compatible with observations that in rabbits, intra-cerebroventricular administration of Al caused, within 2 d, a transient decrease of neuronal mRNAs for three neurofilament subunits and amyloid precursor protein (56). This general inhibitory effect disappeared on the 7th day postinjection. The only lasting inhibitory effect concerned that expression of mRNA for neurofilament H subunits in neurons bearing neurofibrillary tangles in their perikarions. This phenomenon was explained by feedback inhibition of gene expression by accumulated tangle proteins but not by Al itself (57). Similar selective suppressions of neurofilaments L and/or H mRNAs in vulnerable regions of central nervous system (CNS) were observed in AD, amyotrophic lateral sclerosis (ALS), and Parkinson’s disease (PD) (58–60). However, the causative role of Al in development of these pathologies may be questioned by the fact that in dialysis encephalopathy brains such changes were not observed despite of severe and frequently fatal impairment of brain function (61). 3.5. Effect on Protein Synthesis Effects of Al on expression of neurofilament proteins, described in Subheading 3.4., preclude its interference with synthesis process of specific neuronal proteins. Morin fluorescent Al-staining combined with other staining techniques revealed association of Al with nucleolus, interchromatin granules, rough endoplasmic reticulum, and free ribosomes, cellular components that play a role in protein synthesis (62). Both early and recent studies provide evidence that Al may cause a bulk increase of protein synthesis in neuroblastoma cells (24,63). However, some of the most recent evidence indicates that increases in intracellular neurofilament proteins content in Al-treated brains or neuroblastoma cells, is the result of hyperphosphorylation-dependent inhibition of proteolytic breakdown, rather than from increased synthesis (32; see Subheading 3.3.). Most of mRNA levels for cytoskeletal proteins decreased or remained unchanged upon intra-cerebrovenricular application of high doses of Al into rabbits (56,57). Only brain actin mRNA level in rabbit brain was found to be increased by Al in these conditions (56). On the other hand, chronic subcutaneous injections of low doses of Al into infant rabbits increased activity of mRNA as well as synthesis of immunoprecipitable calmodulin in cell-free systems of their brains (64). 3.6. Interference with Intracellular Signal-Transduction Pathways Al interferes with extra and intracellular Ca transport as well as with its binding to extra and intracellular ligands (6,22; see Subheading 3.1.). It precludes interaction of Al with several Ca-dependent steps of intracellular signaling pathways. Ca entry through voltage-dependent Ca channels is postulated to form Ca microdomains at depolarized presynaptic membrane, which triggers its fusion with synaptic vesicles and quantal transmitter release. Therefore, inhibition of Ca entry by Al into depolarized axonal terminals may be responsible for observed inhibition of release of various transmitters and transmitter evoked postsynaptic events (see Subheadings 5. and 6.). Disturbances of intraneuronal and intraterminal ion homeostasis are likely to be because

220

Szutowicz

of inhibition by Al of Ca- and Na,K-ATPase caused by stabilization of the phophorylated form of the enzyme (21,65). Al-enhanced phosphorylations dependent on cAMP and inhibited those catalyzed by protein kinase C (PKC). Oral or intra-cerebroventricular application of Al to rats increased cAMP level and phosphorylation of MAP2 in cerebral cortex and other brain regions (66,67). This rise could result from the inhibition of phophodiesterase by Al. On the other hand, basic synthesis of cAMP was not affected by Al. However, Al strongly increased agonist-activated cAMP synthesis in rat brain cortical slices (68). Activation of cAMP-dependent protein kinases was found to increase phosphorylation of some neurofilament protein subunits (69). This increased phosphorylation of structural proteins is likely to be caused by direct inhibition of phosphatase 2A activity by Al (70). Activation of Ca-calmodulin-dependent cAMP phosphodiesterase and other enzymes is an important factor regulating postsynaptic responses, including long-term potentiation. It has been claimed that Al, which binds calmodulin 10 times stronger than Ca, may exert a long-lasting inhibition in the activity of Ca-calmodulin-dependent cAMP phosphodiesterase (71). Studies with monoclonal antibodies (MAbs) recognizing different conformations of Ca-bound and Al-bound calmodulin complexes revealed that Al binds calmodulin in the presence of Ca+ leading to an inactive conformation (72). Recent in vivo studies support this hypothesis. They show that 4-wk intraperitoneal administration of Al increased phosphorylation of neuronal cytoskeletal proteins, which was prevented by coinjections of desferrioxamine. In addition, exogenous calmodulin-inhibited hyperphosphorylation of proteins in synaptosomes isolated from Al-treated animals (22). There are reports, however, that show no interactions between calmodulin and Al in vitro. (73). Receptor-mediated hydrolysis of phosphoinositides is a principal signal-transduction system for pre- and postsynaptic actions of several neurotransmitters such as acetylcholine, glutamate, and norepinephrine as well as neurotrophic factors (74). It has been shown that Al affects the function of particular components of phosphoinositide second messenger system including G-proteins, phospholipase C, and PKC (67,75,76). Al was shown to directly inhibit transducin activation and guanosine triphoshatase activity in retina (76) PKC activation by diacylglycerol occurs by its translocation from cytosol to inner surface of the plasma membrane. Al increased PKC enzyme binding to the membrane but inhibited the accumulation of inositol phosphates in brain slices (67,76). It could be explained by simultaneous direct inhibition by Al phopholipase C activity by Al (76,77). In addition, low 10–8 M Al concentrations were found to abolish Ca-dependent activation of PKC (78). This idea is supported by studies on human neuroblastoma cells in which Al strongly inhibited (IC50 15 µM) carbachol-stimulated production of inositol 1,4,5-triphosphate (79). On the other hand, no direct interaction between Al and Gp protein was reported (77). There was a differential susceptibility of phospholipase C in different brain regions to the inhibitory influence, which appeared to be independent of Ca concentrations in the extracellular compartment. Phospholipase C inhibition was greatest in cerebellum and weakest in the striatum (80). Significance of this differences for susceptibility of particular parts of the brain to Al toxicity remains unknown.


221

The glutamate-NO-cGMP pathway plays an important role in modulation of longterm potentiation a crucial step in learning and memory formation. Chronic administration of Al to rats in drinking water caused 50% inhibition of N-methyl-D-asparate (NMDA)-induced increase of extracellular cGMP as well as its basal level in cerebellum (81). Al also decreased expression of calmodulin and NOS proteins as well as inhibited basal activity of soluble guanylate cyclase. This inhibition was overcome by stimulation with S-nitroso-N-penicillamine, a NO donor (81). Similar Al-induced changes were also observed in primary cultures of cerebellar neurons (82). Also, prenatal exposure to Al reduced content of NOS and guanylate cyclase in whole cerebella and cultured cerebellar neurons (83). The increase in the level of inducible form of NOS by relatively acute (3-d) and chronic (3-wk) Al treatment may reflect a compensatory adaptative reaction to the inhibitory effect of Al on cGMP (84). The disruption of glutamate-NO-cGMP intracellular signalling by Al may contribute to neurological disturbances seen in dialysis encephalopathy (DE) and other neurodegenerative diseases. 3.7. Aluminum and Neuronal Death The selective loss of central cholinergic, serotoninergic, or some peptidergic neurons in the course of AD, DE, or other encephalopathies is well-documented (85). However links of neurofibrillary degeneration and neuronal death with excess Al are not well-established by available clinical and epidemiological data. Neurofibrillary degeneration in brains of Al-treated rabbits, which is apparently caused by its neurotoxic effects (see Subheading 3.3.) is claimed to be a model of AD pathology. On the other hand, DE studies indicate that severe loss of neuronal function and presumably neurons themselves may take place without neurofibrillary degeneration (61). This raises the possibility that Al may be a modifier rather than a causative factor in these human pathologies. Due to these uncertainties, the culture of brain- and clonal-cell lines were employed to investigate direct effects of Al on neuronal viability. It has been found that Al in clinically relevant 1–100 µM concentrations was highly toxic to rat hippocampal neurons grown on a glial layer, causing death of 50–70% cells. This effect was independent of extracellular Ca and glutamate receptor activation (25). On the other hand, high 1 mM concentrations of Al were found to exert neuroprotective actions. Similar results were reported for embryonic chick brain cultures in which Al caused inhibition of MAP2 and L neurofilament-protein expression (IC50>180 µM) followed by respective loss of neuronal-cell viability (IC50>280 µM). Astrocytes appeared to be more resistant to Al than neurons (86). On the other hand, neuroblastoma cells appear to be very insensitive to various Al species during 1–14 d of exposure (9,24,87). Cell number increased at the same rate, and protein content remained similar in cells grown in the absence or presence of Al. Also, formation of neurite extensions both in undifferentiated and cAMP or retinoic acid differentiated cells was not affected by this metal (24,44,88). Despite these findings, several discrete changes have been found in Al-treated neuroblastomas. They included: decreased resting potential in NIE-115 cells (9), accumulation of abnormal neurofibrillary proteins in perikarions of NB2a/dl neuroblastoma (44), and increased

222

Szutowicz

Ca and decreased acetyl-CoA content in SN56 hybrid septal neuroblastoma (24). In addition, Al is likely to aggravate the negative effects of other neurotoxins on neuronal integrity. For instance, Al increased amyloid-β inhibitory effect on formation of axonal extensions by differentiating SN56 neuroblastoma (24). 4. EFFECT OF ALUMINUM ON BRAIN-ENERGY METABOLISM 4.1. Changes in Glucose and Other Energy Substrate Metabolism The suppression of glucose metabolism is a characteristic, although not a specific feature of several encephalopathies, including AD, DE, vascular-ischemic dementias, and thiamine-deficiency encephalopathies (89). It indicates that Al may be one, but perhaps not the most important factor contributing to energy deficits. Nevertheless, different experimental settings provide some evidence on this aspect of Al toxicity. Al impaired 14CO2 production from [U14C]glucose in rat brain synaptosomes (90). In vitro studies of cytosolic and mitochondrial hexokinase from rat brain have shown its inhibition by Al. This inhibition was due to a competitive interaction of a Al-ATP complex that blocked phosphate transfer from ATP to glucose (91). Hexokinase is the ratelimiting enzyme for brain glycolysis. Therefore its inhibition by Al might cause energy shortages in affected brains by limiting pyruvate to the tricarboxylic acid cycle (89,90). One has to stress that Al-induced energy shortages may also result from inhibition of pyruvate utilization (18,20).This phenomenon was explained to be caused by the inhibition of monocarboxylate carrier and/or Na/Ca exchanger on the mitochondrial membranes. The latter was found to cause excessive accumulation of Ca in mitochondria, resulting in inhibition of pyruvate dehydrogenase activity (20,92). The same mechanism may apply to cAMP/retinoic acid differentiated SN56 neuroblastomas in which Al brought about marked depression of pyruvate dehydrogenase activity during 3-d culture. In addition, it reinforced inhibitory action of amyloid-β on enzyme activity down to 30% of control values presumably by induction of Ca overload (24). No significant changes were seen in nondifferentiated cells. This findings could explain the particular susceptibility of brain cholinergic neurons to Al. On the other hand, in vivo studies provide less convincing evidence on this aspect of Al toxicity. Intraperitoneal and intra-cerebroventricular injections of Al-tartarate caused transient, weak inhibition of 2-deoxyglucose uptake in selected brain areas that appeared on d 7 and did not exceed 20% (93). In addition, chronic 2-yr overload with Al added to drinking water resulted in no changes in glucose utilization in 25 discrete regions of the rat brain (94). Others have found 20–27% inhibition of hexokinase and glucose-6-phosphate dehydrogenase activities in brains of rats drinking 100 mM Al solution for 1 yr (95). 4.2. Effect on Acetyl-CoA Metabolism and Distribution in Subcellular Compartments Intramitochondrial oxidative decarboxylation of pyruvate derived from glycolytic metabolism of glucose is a main source of acetyl-CoA in mammalian brain. Several recent reports reveal that lactate/pyruvate produced in glial cells may, under some conditions, serve as an important energy substrate (96,97). The bulk of acetyl-CoA is utilized for energy production in tricarboxylic acid cycle and only a small fraction is


223

transported to the cytoplasm to serve as a substrate for lipid synthesis. Cholinergic neurons utilize additional amounts of acetyl-CoA for acetylcholine (ACh) synthesis in their cytoplasm. Thus, intracellular/intraneuronal acetyl-CoA forms two pools regulated by independent mechanisms (89). In rat brain synaptosomes depolarized with 30 mM KCl, extracellular AlPO4(OH-) caused increase of acetyl-CoA content in their mitochondria and decrease in cytoplasmic compartment (18). This phenomenon could be due to the inhibition of verapamil-sensitive Ca-channels by AlPO4(OH-) (Fig. 1). Inhibition of Ca entry would result in inhibition of direct acetyl-CoA transfer from mitochondria to cytoplasm through large conductance channels, known as permeability-transition pores, that are activated by increase of cytoplasmic [Ca2+] (18,20,23). Such a process is likely to impair cytoplasmic pathways of structural lipids in all brain cells and ACh synthesis in cholinergic neurons (see Subheading 5.). Hence, the reversal by extracellular Al increase of synaptoplasmic acetyl-CoA-evoked Ca-dependent NO excitotoxicity could be explained by its Ca channel-blocking properties (98) (Fig. 1). On the other hand, intracellular Al due to its elevating effect on cytoplasmic and intramitochondrial [Ca2+] would exert opposite effect, leading to increased permeability of the mitochondrial membranes and depletion of acetyl-CoA in mitochondria (20,23). It could generate energy shortages in Al-overloaded cells (66,98) (Fig. 1). Susceptibility of cholinergic neuroblastoma SN56 to Al was found to depend on their phenotypic differentiation. Namely, in SN56 with high expression of ACh metabolism, Al alone and with NO caused much greater suppression of acetyl-CoA levels than in undifferentiated cells (24). It is possible that differing in regional susceptibility of cholinergic neurons to Al neurotoxicity may depend on intrinsic relationships between rates of acetyl-CoA and ACh metabolism (Fig. 1). These data are compatible with findings showing that Al alone or in combination with amyloid-β prevented morphological differentiation of neuroblastoma cells by cAMP/retinoic acid or other compounds (24,39). 5. INTERFERENCE OF AL WITH CHOLINERGIC TRANSMISSION 5.1. In Vivo Effect of Al on ACh Metabolism Preferential loss of cholinergic neurons is a characteristic feature of several degenerative brain diseases including DE, AD, and some cases of PD, in which an excessive Al accumulation in the brain may take place (9,61,100). Deficiency of cholinergic function causes a wide range of psychoneurologic symptoms including progressive impairment of cognitive and emotional functions, psychic and motor hyper- or hypoexcitability, muscular weakness, and sensory disturbances. Brain autopsies of DE and AD victims revealed prominent decrease of cholinergic markers such as choline acetyltransferase (ChAT), high-affinity choline-uptake system activities, or muscarinic presynaptic M2 receptors in hippocampus, different areas of the cerebral cortex and olfactory bulbs. The degree of loss correlated with the gravity of dementia assessed immediately before death (85,101,102). In contrast, no change in M1 and increase in M4 specific ligand-binding sites were seen in AD brains (101). However, one has to bear in mind that excess Al may be only one of several cholinotoxic signals (Ca2+, NO, hypoxia, amyloid-β, glutamate) that are claimed to contribute to different forms of brain degeneration.

224

Szutowicz

Decrease of ChAT activity in encephalopathic brain is a basic indicator of loss of cholinergic neurons. Al neurotoxicity experiments provide contradictory results on this subject. There is similar number of reports showing either suppression or no effect of Al on enzyme activity in brains of Al-susceptible rabbits and Al-resistant rats (9,104,105). Moreover, in some cases Al-evoked neurofibrillary degeneration developed without changes in ChAT activity. Small decrease of ChAT activity was observed after 2-wk intraperitoneal injections of Al into nephrectomized, uremic rats, whereas nephrectomy itself had no effect on enzyme activity (106). On the other hand, there are concordant reports showing that high-affinity uptake system activity, as well as ACh levels were always decreased by Al treatment (105,107,108). One can expect that pre- and postsynaptic muscarinic receptors may differentially respond to primary and secondary effects of Al in brain. However, no detailed studies on this subject are available. It has been shown that chronic intraperitoneal Al application caused no significant change in density of nonspecificic [3H]quinuclidynyl benzilate binding sites in different brain regions (105,109). On the other hand, significant decrease in number of M1 receptor [3H]pirenzepine-binding sites was observed (109). Acetylcholinesterase is not a specific marker for cholinergic neurons since this enzyme is localized on the postsynaptic side of cholinergic synapses. This may explain the lack of Al effect on its activity in vivo (see Subheading 9 for review). On the other hand, recent studies indicate that prolonged 1–2 mo exposure to Al significantly decreased, whereas 4-d treatment increased acetylcholinesterase activity (105,110). Kinetic in vitro studies, revealed that Al activation of acetylchinesterase is due to a direct structural modifications of the enzyme (111). Thus, biphasic, time-dependent effects of Al on brain acetylcholinesterase in vivo may be explained by its primary direct activation followed by secondary neuron degeneration-evoked loss of the enzyme (105). These results suggest that an excess of Al may depress cholinergic transmission in affected brains. Accordingly, several behavioral, cognitive, neurophysiologic deficits were found in Al-treated animals. Observations on rats included decrease of active and passive avoidance tests (104,105), increased mobility and decreased efficiency in learning tasks (112) as well as impairment of hippocampal long-term potentiation in vivo and in vitro (113). Other reports revealed no effect of chronic Al administration to young and old rats on their performance in open field activities and passive avoidance tests (114). In turn, treatment of pregnant rats with subcutaneous injections of AlCl3 diminished performance and lengthened latency in avoidance-responding tasks in their pups (115). On the other hand, in rabbits all studies reported Al to cause impairment of behavioral and motor functions that correlated well with the appearance of neurofibrillary degeneration. Young animals appeared to be less susceptible to Al toxicity than the adult ones (see ref. 116 for review). 5.2. Short- and Long-Term Effects of Aluminum on Transmission at the Neuromuscular Junction Al effects on the peripheral nervous system (PNS) were studied using frog neuromuscular junction. Relatively high (0.3–8.0 mM) Al concentrations instantly aug-


225

mented spontaneous and stimulation-induced quantal ACh release from motor nerve terminals incubated in low Ca media. However no changes in postsynaptic response were recorded (9). The mechanism of this Al effect is unknown. One may suppose that Al-evoked increase of ACh release results from a rise of intraterminal [Ca2+]. On the other hand, long-term 3.5 h exposure of neuromuscular junction to Al brought about transient 20-fold increase of the transmitter release followed by irreversible cessation of synaptic activity. Such sequence of events was found to be a general feature of degenerating nerve terminals (9). Impairment of peripheral cholinergic transmission may explain mechanism of motor disturbances seen in Al encephalopathies in humans and experimental animals. 5.3. In Vitro Effects of Al on Acetylcholine Metabolism Interferencece of Al with calcium, acetyl-CoA metabolism, and intracellular signaltransduction systems indicates potential sites of its influence on ACh metabolism. Inhibitory effect of Al on choline uptake by high-affinity system of rat brain synaptosomes may explain depression of cholinergic transmission as caused by shortage of one of the substrates for ACh synthesis (107). In Ca-free medium, Al was found to increase resting/spontaneous ACh release from isolated brain nerve terminals, presumably because of its calcium-like properties (9,18,20,98). On the other hand, 0.25 mM Al resulted in over 60% inhibition of Ca-dependent transmitter release and synthesis in rat brain synaptosomes (18). In addition, Al partially reversed inhibition of Ca-dependent ACh release by verapamil. It suggests that extracellular Al inhibits Ca-dependent transmitter release from cholinergic nerve terminals by binding with verapamil site on voltage-dependent Ca channels. Omission of phosphate from the incubation medium abolished inhibitory influence of Al on ACh release. It indicates that the active form of this metal in biological systems may be Al(PO4)OH—anion (18). Depression of ACh release and synthesis correlates well with decreased acetyl-CoA content in cytoplasm of Al-treated nerve terminals. It indicates that Al may suppress cholinergic transmission by a combination of several mechanisms including decreases [Ca2+] and precursor substrates choline and acetyl-CoA in cytoplasmic compartment of nerve terminals (18,20,107). This finding rises the supposition that excessive accumulation of Al in some encephalopathic brains might by these mechanisms aggravate symptoms of already existing cholinergic loss by functional depression of surviving cholinergic neurons. Postsynaptic responses to ACh were also affected by Al. In concentrations from 0.01–1 mM Al partially reversed carbachol-evoked rise of inositol phosphates in hippocampal and cortical slices from rat brain (117). The mechanism of this action may involve both direct inhibition of phospholipase C activity and decrease of intracellular [Ca2+], independent of extracellular Ca level (76,118). No direct effect of Al on PKC was found (118). However, there are reports showing inhibition of PKC by this metal (9). 5.4. Al Actions on Cholinergic Neuroblastoma Cholinergic neuroblastoma cells have not been extensively used so far as a model for studies of mechanisms of preferential Al cholinotoxicity. The main problem is the relatively low expression of ChAT and thereby low rate of ACh metabolism, in com-

226

Szutowicz

parison to normal brain cholinergic neurons. On the other hand, the expression of ChAT and other parameters of ACh metabolism may be greatly increased by various differentiating and trophic factors known to affect cholinergic neuron function and viability in vivo. Thus, it appears that Al added to growth medium failed to change ChAT, pyruvate dehydrogenase activity, and muscarinic-receptor expression in nondifferentiated mouse SN56 and human neuroblastoma IMR32 (24,119,120). In contrast, SN56 cells differentiated with cAMP and retinoic acid were found to be sensitive to Al, which brought about 30 and 40% inhibition of ChAT and pyruvate dehydrogenase activities. In differentiated SN56 cells, Al also caused much stronger short-term suppression of ACh release and acetyl-CoA levels than in undifferentiated cells. Therefore, one may postulate that in the brain highly differentiated cholinergic neurons may be more susceptible to Al toxicity than those with lower expression of cholinergic functions. This would be a result of a greater inhibitory influence of Al on energy metabolism, transmitter synthesis, and release (24). 6. Al INTERACTION WITH DIFFERENT NEUROTRANSMITTER SYSTEMS 6.1. Effect on Monoaminergic System Al was found to have inhibitory effects on different steps of monoamine metabolism. Intracerebral injections of Al decreased activity of tyrosine hydroxylase, a ratelimiting step in the synthesis of biogenic amines, in selected regions of the rabbit brain (121). In accord with this finding remain data on decrease in dopamine and norepinephrine contents in brains of Al-treated rats and rabbits (108,122). In vitro experiments have shown inhibition of monoamine oxidase A in rat brain mitochondria (123). Al also affected postsynaptic effects of norepinephrine by inhibiting adrenergic receptor-G protein mediated inositol phosphate accumulation by inhibition of phospholipase C activity (117). This mechanism appears to be similar for effects on ACh and glutamate receptor-mediated rise of inositol phosphate level (see Subheadings 5.3. and 6.2.). Formation of Al-dopamine (catechol) complexes was found to be responsible for the inhibition of dopamine-evoked rise of nerve growth-factor synthesis by cultured mouse brain astroglial cells (124). Thus, suggesting that Al-induced cognitive impairment by prenatal exposure to Al in rats may be due to inappropriate development of the cholinergic system resulting from nerve growth-factor deficits (115). 6.2. Effect on Amino Acid Transmitters Signaling and Metabolism Early works demonstrated that Al inhibited uptake of glutamate, gammaaminobutyric acid (GABA), and glycine by rat and rabbit brain nerve terminals (9). This transport is driven by sodium gradient maintained by NaK ATP-ase, which was inhibited by millimolar concentrations of Al(65). However, concentrations of Al necessary for inhibition of amino acid transport were two orders of magnitude lower than those required for suppression of ATP-ase activity, suggesting the existence of other mechanisms for Al-evoked inhibition of amino acid uptake. The glutamate deficit in the brain may be caused by Al stimulation of glutamine synthetase activity in astrocytes accompanied by inhibition of glutaminase activity (125). Insufficient glutamatergic transmission may be the cause of disturbances seen in dialysis encephal-


227

opathy victims. Al may also affect postsynaptic glutamate signaling. It has been shown that Al decreased metabotropic glutamate (quisqualate) receptor-mediated increase of inositol phosphate accumulation in hippocampal and cortical slices from rat brain due to inhibition of phopholipase C activity (116). It also suppressed NMDA receptormediated signals, presumably by combination of its inhibitory effects on Ca entry and guanylate cyclase activity (19,81). Formation of a stable Al-glutamate complex, which prevents its conversion to glutamine, was claimed to be responsible for impairment of ammonia detoxification and development of ammonia-like encephalopathy (126). 7. BRAIN Al IN HUMAN ENCEPHALOPATHIES 7.1. Al in AD and Related Disorders: Clinical and Experimental Evidences The neurotoxicity of Al applied directly to the brain was demonstrated by its ability to develop seizures, cognitive deficits, and neurofibrillary degeneration in rabbits (see ref. 9 for review). The similarity of this model to clinical and autopsy observations in AD, DE, ALS, and Parkinson’s dementia of Guam triggered extensive studies on the role of Al as a putative etiological factor in these encephalopathies. Early studies of Crapper et al. (126,127) using atomic absorption spectrometry showed two to fourfold higher levels of Al in homogenates from different brain regions of AD in comparison to brains of unaffected subjects with normal Al concentrations approx 1–2 µg/g of tissue. Among 13 studies performed in brains of AD victims, 10 revealed increased Al content by quantitative or qualitative methods. The latter ones demonstrated increased Al content in tangle-bearing neurons or amyloid-plaque cores (see ref. 9 for review). Other reports have shown increase of Al in the aging brain as well as in brains of ALS/ PD victims (9). Laser-microprobe mass-analysis studies have shown the presence of Al within neurofibrillary tangles as well as in plaque cores and nuclei of affected neurons. Its level in ALS/PD was considerably higher than in respective samples of AD brains (129,130). Studies of parietal lobe, hippocampus, superior, and middle temporal gyri of AD brains show modest increases in Al (highest value 8 µg/g) in comparison to control samples (131). Treatment of AD brain slices with desferrioxamine, a trivalent cation chelator, enhanced their immunoreactivity with antibodies against phosphorylated epitopes of paired helical filament tau protein, and indicates their partial masking with Al (49). Another study demonstrated increased Al accumulation in neurofibrillary tangles of dementia pugilistica brains higher than those in AD patients. Moreover, excessive Al accumulation was found in nuclei of neurofibrillary tangles (132). Autopsy specimens from 10 brain regions of patients who died of PD were tested by neutron-activation analysis for Al contents. This study revealed 30–110% increase of Al in hippocampus, caudate nucleus, globus pallidus, and substantia nigra, accompanied by significant decrease in Mg and no change in Ca content in all 10 regions tested (133). This is consistent with studies showing complex interactions of Al with divalent cations-binding sites in extra- and intracellular compartments (6,18, 20). There are also studies that showed statistically insignificant increases of Al concentrations in AD brains. This finding was explained by great variation of Al content in AD brains. It indicates that AD may develop both at normal and highly increased brain Al content (134,135). A recent study, using the new technique of nuclear microscopy found Al to be absent in neurofibrillary tangles of AD (136). Also Al concentrations in the frontal

228

Szutowicz

and temporal cortex were not found to be elevated in AD (137). Thus the existence of a causal link between Al excess in AD remains unproven. On the other hand, one can assume that secondary Al accumulation in AD brain, may exacerbate existing cognitive and neurological deficits caused by a primary pathogenic factor. The contribution of Al to AD pathology is indirectly supported by the positive outcome of treatment of AD patients with intramuscular injections of Al/Fe chelating agent desferrioxamine (138). This claim is supported by experiments showing the ability of desferrioxamine to reverse Al-evoked neurofibrillary degeneration in rabbits (139). 7.2. Dialysis and Other Iatrogenic Encephalopathies Increased Al level in brains of uremic/dialyzed patients results from increased metal absorption from dialysis fluids, intake of Al-containing acid binders, and decreased excretion in urine. Clinical signs of Al intoxication included speech disorders, dementia, convulsions, and coma leading to death. The incidence of DE is about 4% among dialysis patients receiving this treatment for 3–7 yr. (140). Brain Al levels were found to increase two to fourfold in dialyzed patients without encephalopathy and 6–12-fold in those with encephalopathy. Increased levels of Al in the brain was apparently from its elevation in the serum, which is blamed for damage of BBB. Normal Al serum concentration were found to be 6.5 (2–14) µg/L and was reported to rise to the values 10–15 times higher in the course of dialyses (116,141). Introduction of Al-free dialysis fluids (Al < 10 µg/L) and new dialysis methods markedly reduced incidence of these complications with fatal outcomes. Nevertheless autopsy studies of dialysis patients showed that density of Al-rich argyrophilic inclusions in the brain correlated with global intake of Al-containing, hyperphosphatemia-controlling drugs (142). Also some patients with chronic uremia who did not undergo dialysis were reported to develop encephalopathy due to excessive Al intake and its impaired excretion (143). Desferrioxamine and products of its metabolism are known to facilitate Al removal by formation of ultrafilterable complexes (141). Transient increase of total and dialyzable Al level in serum was observed due to its extraction from the tissues (140,141). Nevertheless, depositions of proteinaceous Al complexes in the brain appeared to be irreversible. They remained unchanged up to 10 yr after renal transplantation, termination of dialyses, and intake of Al-containing drugs (142). Some standard intravenous feeding solutions contain significant amounts of Al. Preterm infants receiving standard intravenous feeding solutions with daily intake of 45 µg of Al/kg/d displayed deficits in mental and neurologic development as compared to infants fed with Al-depleted solutions (4–5 µg of Al/kg/d) (100). Also 10 of 12 children on long-term (7 mo to 4 yr) parenteral nutrition program with daily Al intake from 7–11 µg/kg/d were found to contain high Al concentrations in their sera, varying from 22–52 µg/L (144). These findings justify careful monitoring of Al level in parenteral nutrition products. Refractory status epilepticus with deadly outcome was observed in two patients after vestibular neurectomy during which Al-Ca fluorosilicate cement was used to bridge bone defects. The level of Al in their CSF was found to be 112 and 63 µg/L (145). These observations clearly indicate that Al may be neurotoxic to humans, although no clear correlation was found between its concentration in tissues and serum and severity of evoked encephalopathy.


229

8. CONCLUSIONS Al may affect several intra- and extraneuronal signaling and metabolic pathways in the brain. Many of these toxic effects are caused by Al interference with Ca and Mg binding sites on regulatory proteins, and low molecular-weight signaling compounds. Interference with glucose-energy metabolism may decrease neuronal function. Although the causative role of Al in dialysis encephalopathy is strong, its role in AD has yet to be proven. Nevertheless, high concentrations of Al present in brains affected by different degenerative diseases are likely to aggravate existing mental and neurologic deficits. Thus, despite these uncertainties, the overall experimental and clinical data indicate that the accumulation of Al in human brain may have neuropathological implications. REFERENCES 1. Driscoll, C. T. and Schecher W. D. (1990) The chemistry of aluminum in the environment. Envir. Geochem. Health. 12, 28–50. 2. Yokel R. A. (1997) The metabolism and toxicokinetics of aluminum relevant to neurotoxicity, in Mineral and Metal Neurotoxicology (Yasui, M., Strong, M. J., Ota, K., and Verity, M. A. eds.), CRC Press, Boca Raton, FL, pp. 81–89. 3. Taylor, G. A., Morre, P. B., Ferrier, I. N., Tyrer, S. P., and Edwardson, J. A. (1998) Gastrointestinal absorption of aluminium and citrate in man. J. Inorg. Biochem. 69, 165–16. 4. Eastham R. D.(ed.) (1985) Biochemical Values in Clinical Medicine. Wright, Bristol, UK, pp. 14–15. 5. Harris, W. R. (1992) Equilibrium model for speciation of aluminum in serum. Clin. Chem. 39, 1809–1818. 6. Martin R. B. (1992) Aluminium speciation in biology, in Aluminium in Biology and Medicine (Martin, R. B., ed.) J. Wiley and Sons, Chichester, UK, pp. 5–25. 7. Yasui, M. and Ota, K. (1998) Aluminum decreases the magnesium concentration of spinal cord and trabecular bone in rats fed a low calcium, high aluminum diet. J. Neurol. Sci. 157, 37–41. 8. Roskams, A. J. and Connor, J. R. (1990) Aluminum access to the brain: a role for transferrin and its receptor. Proc. Natl. Acad. Sci. USA 87, 9024–9027. 9. Meiri, H., Banin, E., Roll, M., and Rousseau, A. (1992) Toxic effects of aluminium on nerve cells and synaptic transmission. Prog. Neurobiol. 40, 89–121. 10. Deloncle, R. and Pages, N. (1997) Aluminum: on both sides of the blood-brain barrier, in Mineral and Metal Neurotoxicology (Yasui, M., Strong, M. J., Ota, K., and Verity, M. A. eds.), CRC Press, Boca Raton, FL, pp. 91–97. 11. Vorbrodt, A. W., Trowbridge, R. S., and Dobrogowska, D. H. (1994) Cytochemical study of the effect of aluminium on cultured brain microvascular endothelial cells. Histochem. J. 26, 119–126. 12. Flatent, T. P. and Garruto, R. M. (1992) Polynuclear ions in aluminum toxicity. J. Theor. Biol. 156, 129–132. 13. Ackley, D. C. and Yokel, R. A. (1997) Aluminum citrate is transported from brain into blood via the monocarboxylic acid transporter located at the blood-brain barrier. Toxicology 120, 89–97 14. Shi, B. and Haug, A. (1990) Aluminum uptake by neuroblastoma cells. J. Neurochem. 55, 551–558. 15. Golub, M. S., Han, B., and Keen, C. L. (1996) Aluminum alters iron and manganese uptake and regulation of surface transferrin receptors in primary rat oligodendrocyte cultures. Brain Res. 719, 72–77.

230

Szutowicz

16. Hoyer, S., Nitsch, R., and Oesterreich, K. (1990) Ammonia is endogenously generated in the brain in the presences of presumed and verified dementia of Alzheimer type. Neurosci. Lett. 17, 358–362. 17. Tjalve, H. and Henriksson, J. (1999) Uptake of metals in the brain via olfactory pathways. Neurotoxicology 20, 181–195. 18. Bielarczyk, H., Tomaszewicz, M., and Szutowicz, A. (1998) Effect of aluminum on acetylCoA and acetylcholine metabolism in nerve terminals. J. Neurochem. 70, 1175–1181. 19. Koenig, M. L. and Jope, R. S. (1987) Aluminum inhibits the fast phase of voltage-dependent calcium influx into synaptosomes. J. Neurochem. 49, 316–320. 20. Szutowicz, A., Bielarczyk, H., Kisielewski, Y., Jankowska, A., Madziar, B., and Tomaszewicz, M. (1998) Effects of aluminum and calcium on acetyl-CoA metabolism in rat brain mitochondria. J. Neurochem. 71, 2447–2453. 21. Mundy, W. R., Kodavanti, P. R., Dulchinos, V. F., and Tilson, H. A. (1994) Aluminum alters calcium transport in plasma membrane and endoplasmic reticulum from rat brain. J. Biochem. Toxicol. 9, 17–23. 22. Julka, D. and Gill, K. D. (1996) Altered calcium homeostasis: a possible mechanisms of aluminium-induced neurotoxicity. Biochim. Biophys. Acta. 1315, 47–54. 23. Green, D. R. and Reed, J. C. (1998) Mitochondria and apoptosis. Science 281, 1309–1312. 24. Szutowicz, A., Tomaszewicz, M., Jankowska, A., Madziar, B., and Bielarczyk, H. (2000) Acetyl-CoA metabolism in cholinergic neurons and their susceptibility to neurotoxic inputs. Met. Brain Dis. 15, 29–44. 25. Brenner, S. R. and Kong-Woo Y. (1994) Aluminum toxicity in rat hippocampal neurons. Neurosci. Lett. 178, 260–262. 26. Yang, M. S., Wong, H. F., and Yung, K. L. (1998) Determination of endogenous trace metal contents in various mouse brain regions after prolonged oral administration of aluminum chloride. J. Toxicol. Environ. Health. 55, 445–453. 27. Bondy, S. C., Ali, S. F., and Guo-Ross, S. (1998) Aluminum but not iron treatment induces pro-oxidant events in the rat brain. Mol. Chem. Neuropathol. 34, 219–232. 28. Deloncle, R., Huguet, F., Babin, P., Fernandez, B., Quellard, N., and Guillard, O. (1999) Chronic administration of aluminium L-glutamate in young mature rats: effects on iron levels and lipid peroxidation in selected brain areas. Toxicol. Lett. 104, 65–73. 29. Xie, C. X. and Yokel, R. A. (1996) Aluminum facilitation of iron-mediated lipid peroxidation is dependent on substrate, pH, and aluminum and iron concentrations. Arch. Biochem. Biophys. 327, 222–226. 30. Verstraeten, S. V., Golub, M. S., Keen, C. L., and Oteiza, P. I. (1997) Myelin is a preferential target of aluminum-mediated oxidative damage. Arch. Biochem. Biophys. 344, 289–294. 31. Xie, C. X., Pattson, M. P., Lovell, M., and Yokel, R. A. (1996) Intraneuronal aluminum potentiates iron-induced oxidative stress in cultured rat hippocampal neurons. Brain Res. 743, 271–277. 32. Abreo, K., Abreo, F., Sella, M. L., and Jain, S. (1999) Aluminum enhances iron uptake and expression of neurofibrillary tangle protein in neuroblastoma cells. J. Neurochem. 72, 2059–2064. 33. Abd-el-Fattah, A. A., al.-Yousef, H. M., al.-Bekairi, A. M. and al.-Sawaf-H. A. (1998) Vitamin E protects the brain against oxidative injury stimulated by excessive aluminum intake. Biochem. Mol. Biol. Int. 46, 1175–1180. 34. Atienzar, F., Desor, D., Burnel. D., Keller, J. M., Lehr, P., and Vasseur, P. (1998) Effect of aluminum on superoxide dismutase activity in the adult rat brain. Biol. Trace. Elem. Res. 65, 19–30. 35. Bondy, S. C., Liu, D., and Guo-Ross, S. (1998) Aluminum treatment induces nitric oxide synthase in the rat brain. Neurochem. Int. 33, 51–54.


231

36. Yang, E. Y., Guo-Ross, S. X., and Bondy, S. C. (1999) The stabilization of ferrous iron by a tocix beta-amyloid fragment and by an aluminum salt. Brain Res. 839, 221–226. 37. Ueda, K., Shinohara, S., Yagami, T., Asakura, K., and Kawasaki, K. (1997) Amyloid β protein potentiates Ca2+ influx through L-type voltage-sensitive Ca2+ channels: a possible involvement of free radicals. J. Neurochem. 68, 265–271. 38. Wis´niewski, H., Karczewski, W., and Wissniewska, K. (1966) Neurofibrillary degeneration of nerve cells after intracerebral injection of aluminium cream. Acta Neuropathol. 6, 211–219. 39. Selkoe, D. J., Liem, R. K. H., Yen, S. H., and Shelanski, M. L. (1979) Biochemical and immunological characteriation of neurofilaments in experimental neurofibrillary degeneration induced by aluminium. Brain Res. 163, 235–252. 40. Troncoso, J. C., Sternberger, N. H., Sternberger, L. A., Hoffman, P. N., and Price, D. L. (1986) Immunocytochemical studies of neurofilament antigens in the neurofibrillary pathology induced by aluminum. Brain Res. 364, 295–300. 41. Muma, N. A. and Singer M. (1996) Aluminum-induced neuropathology: transient changes in microtubule-associated proteins. Neurotoxicol Teratol. 18, 679–690. 42. Savory, J., Huang, Y., Herman, M. H., Reyes, M. R., and Wills, M. R. (1995) Tau immunoreactivity associated with aluminum maltolate-induced neurofibrillary degeneration in rabbits. Brain Res. 669, 325–329. 43. Johnson, G. V. and Jope, R. S. (1988) Phosphorylation of rat brain cytoskeletal proteins is increased after orally administered aluminum. Brain Res. 456, 95–103. 44. Shea, T. B., Clarke, J. F., Wheelock, T. R., Paskevich, P. A., and Nixon, R. A. (1989) Aluminum salts induce the accumulation of neurofilaments in perikarya of NB2a/dl neuroblastoma. Brain Res., 492, 53–64. 45. Shea, T. B., Beermann, M. L., and Nixon, R. A. (1995) Aluminum treatment of intact neuroblastoma cells alters neurofilament subunit phosphorylation, solubility and proteolysis. Mol. Chem. Neuropathol. 26, 1–11. 46. Shin R. W., Lee, V. M., and Trojanowski, J. Q. (1994) Aluminum modifies the properties of Alzheimer’s disease PHF tau proteins in vivo and in vitro. J. Neurosci. 14, 7221–7233. 47. Shea, T. B., Wheeler, E., and Jung, C. (1997) Aluminum inhibits neurofilament assembly, cytoskeletal incorporation, and axonal transport. Dynamic nature of aluminum-induced perikaryal neurofilament accumulations as revealed by subunit turnover. Mol. Chem. Neuropathol. 32, 17–39. 48. MacDonald, T. L., Humphreys, W. G., and Martin, R. B. (1987) Promotion of tubulin assembly by aluminum ion in vitro. Science 236, 183–186. 49. Murayama, H., Shin, R. W., Higuchi, J., Shibuya, S., Muramoto, T., and Kitamoto, T. (1999) Interaction of aluminum with PHFtau in Alzheimer’s disease neurofibrillary degeneration evidenced by desferrioxamine-assisted chelating autoclave method. Am. J. Pathol. 155, 877–885. 50. Scott C. W., Fieles, A., Sygowski L. A., and Caputo C. B. (1993) Aggregation of tau protein by aluminum. Brain Res. 628, 77–84. 51. Kruck, T. P. A., McLachlan, D. R. C., Bergeron, C., and Lukiw, W. J. (1993). Aluminum in neocortical nuclei. Removal by shuttle chelation and relevance to Alzheimer’s disease pharmacotheraphy, in Advances in the Biosciences. Alzheimer’s Disease and Related Disorders, vol. 87 (Nicolini, N., Zatta, P. F., and Corain, B., eds.), Pergamon Press, Oxford, pp. 255–256. 52. Dobson, C. B., Templar, J., Pay, J. P., and Itzhaki, R. F. (1993) Aluminium and Alzheimer’s disease: intracellular distribution of aluminium as revealed by Al-binding. in Advances in the Biosciences. Alzheimer’s Disease and Related Disorders, vol. 87 (Nicolini, N., Zatta, P. F., and Corain, B., eds.), Pergamon Press, Oxford, pp. 251–252. 53. Walker, P. R., LeBlanc, J., and Sikorska, M. (1989) Effects of aluminum and other cations on the structure of brain and liver chromatin. Biochemistry 28, 3911–3915.

232

Szutowicz

54. Lukiw, W. J., Kruck, T. P., and McLachlan, D. R. (1989) Linker histone-DNA complexes: enhanced stability in the presence of aluminum lactate and implications for Alzheimer’s disease. FEBS Lett. 253, 59–62. 55. Pearson, E. C., Bates, D. L., Prospero, T. D., and Thomas, J. O. (1984) Neuronal nuclei and nuclei from mammalian cerebral cortex nucleosome repeat lengths DNA contents and H1 contents. Eur. J. Biochem. 144, 353–360. 56. Parhad, I. M., Krekoski, C. A., Mathew, A., and Tran, P. M. (1989) Neuronal gene expression in aluminum myleopathy. Cell. Mol. Neurobiol. 9, 123–138. 57. Chambers, C. B. and Muma, N. A. (1997) Neuronal gene expression in aluminum-induced neurofibrillary pathology: an in situ hybridization study. Neurotoxicology, 18, 77–88. 58. Crapper McLachlan, D. R., Lukiw, W. J., Wong, L., Bergeron, C., and Bech-Hansen, N. T. (1988) Selective messenger RNA reduction in Alzheimer’s disease. Mol. Brain Res. 3, 255–262. 59. Bergeron, C., Beric-Maskarel, K., Muntasser, S., Weyer, L., Somerville, M. J., and Percy, M. E. (1994) Neurofilament light and polyadenylated mRNA levels are decreased in amyotrophic lateral sclerosis motorneurons. J. Neuropathol. Exp. Neurol. 53, 221–230. 60. Hill, W. D., Arai, M., Cohen, J. A., and Trojanowski, J. Q. (1993) Neurofilament mRNA is reduced in Parkinson’s disease substantia nigra pars compacta neurons. J. Comp. Neurol. 329, 328–336. 61. Perry T. L., Yong V. W., Kish S. J., Ito, M., Foulks, J. G., Godolphin, W. J., and Sweeney V. P. (1985) Neurochemical abnormalities in brains of renal failure patients treated by repeated hemodialysis. J. Neurochem. 45, 1043–1048. 62. Wen, G. Y. and Wisniewski, H. M. (1985) Histochemical localization of aluminium in the rabbit CNS. Acta Neuropathol. 68, 175–184. 63. Miller, C. A. and Levine, E. M. (1974) Effects of aluminium salts on cultured neuroblastoma cells. J. Neurochem. 22, 751–758. 64. Nicholls, D. M., Speares, G. M., Miller, A. C., Math, J., and Del-Bianco, G. (1991) Brain protein synthesis in rabbits following low level aluminium exposure. Int. J. Biochem. 23, 737–741. 65. Caspers, M. L., Dow, M. J., Fu, M. J., Jacques, P. S., and Kwaiser, T. M. (1994) Aluminum-induced alterations in [3H] ouabain binding and ATP hydrolysis catalyzed by the rat brain synaptosomal (Na+ + K+)-ATPase. Mol. Chem. Neuropathol. 22, 43–55. 66. Johnson, G. V. W. and Jope R. S. (1986) Aluminum increases cyclic AMP in rat cerebral cortex in vivo. Life Sci. 39, 1301–1305. 67. Johnson, G. V. W., Cogdill, K. W., and Jope, R. S. (1990) Oral aluminum alters in vitro protein phosphorylation and kinase activities in rat brain. Neurobiol. Aging 11, 209–216. 68. Johnson, G. V. W., Li, X. H., and Jope, R. S. (1989) Aluminum increases agonist-stimulated cyclic AMP production in rat cerebral cortical slices. J. Neurochem. 53, 258–263. 69. Nakamura, Y., Takeda, M., Angelides, K.J., Tanaka, T., Tada, K., and Nishimura, T. (1990) Effect of phosphorylation on 68 Kda neurofilament subunit protein assembly by the cyclic AMP dependent protein kinase in vitro. Biochem. Biophys. Res. Commun. 169, 744–750. 70. Yamamoto, H., Saitoh, Y., Yasugawa, S., and Miaymoto, E. (1990) Dephosphorylation of tau factor by protein phosphatase 2A in synaptosomal cytosol fractions and inhibition by aluminum. J. Neurochem. 55, 683–690. 71. Richardt, G., Frederolf, G., and Haberman, E. (1985) The interaction of aluminium and other metal ions with calcium-calmodulin dependent phosphodiesterase. Arch. Toxic. 57, 257–259. 72. Levi, R., Wolf, T., Fleminger, G., and Solomon, B. (1998) Immuno-detection of aluminium and aluminium induced conformational changes in calmodulin-implications in Alzheimer’s disease. Mol. Cell. Biochem. 189, 41–46. 73. You, G. F. and Nelson, D. J. (1991) Al3+ versus Ca2+ ion binding to methionine and tyrosine spin-labeled bovine brain calmodulin. J. Inorg. Biochem. 41, 283–291.


233

74. Fisher, S. K., Heacock, A. M., and Agranoff, B. W. (1992) Inositol lipids and signal transduction in the nervous system: an uptake. J. Neurochem. 58, 18–38. 75. Gilman, A. G. (1987) G proteins: transducers of receptor-generated signals. Ann. Rev. Biochem. 56, 615–649. 76. Mundy, W. R., Freudenrich, T., Shafer, T. J., and Nostrandt, A. C. (1995) In vitro aluminum inhibition of brain phosphoinositide metabolism: comparison of neonatal and adult rats. Neurotoxicology 16, 35–44. 77. Shafer, T. J., Nostrandt, A. C., Tilson, H. A., and Mundy, W. R. (1994) Mechanisms underlying AlCl3 inhibition of agonist-stimulated inositol phosphate accumulation. Role of calcium, G-proteins, phospholipase C and protein kinase C. Biochem. Pharmacol. 47, 1417–1425. 78. Cochran, M., Elliott, D. C., Brennan, P., and Chawtur, V. (1990) Inhibition of protein kinase C activation by low concentrations of aluminium. Clin. Chim. Acta. 194, 167–172. 79. Wood, P. C., Wojcikiewicz, R. J. H., Burgess, J., Castleden, C. M., and Nahorski, S. R. (1994) Aluminium inhibits muscarinic agonist-induced insoitol 1,4,5-trisphosphate production and calcium mobilization in permeabilized SH-SY5Y human neuroblastoma cells. J. Neurochem. 62, 2219–2223. 80. Nostrandt, A. C., Shafer, T. J., Mundy, W. R., and Padilla, S. (1996) Inhibition of rat brain phosphatidylinositol-specific phospholipase C by aluminum: regional differences, interactions with aluminum salts, and mechanisms. Toxicol. Appl. Pharmacol. 136, 118–125. 81. Hermenegildo, C., Saez, R., Minoia, C., Manzo, L., and Felipo, V. (1999) Chronic exposure to aluminium impairs the glutamate-nitric oxide-cyclic GMP pathway in the rat in vivo. Neurochem. Int. 34, 245–253. 82. Cucarella, C., Montoliu, C., Hermenegildo, C., Saez, R., Manzo, L., Minana, M. D., and Felipo, V. (1998) Chronic exposure to aluminium impairs neuronal glutamate-nitric oxidecyclic GMP pathway. J. Neurochem. 70, 1609–1614. 83. Llansola, M., Minana, M. D., Montoliu, C., Saez, R., Corbalan, R., Manzo, L., and Felipo, V. (1999) Prenatal exposure to aluminum reduces expression of neuronal nitric oxide synthase and soluble guanylate cyclase and impairs glutamatergic neurotransmission in rat cerebellum. J. Neurochem. 73, 712–718. 84. Bondy, S. C., Liu. D., and Guo-Ross. S. (1998) Aluminum treatment induces nitric oxide synthase in the rat brain. Neurochem. Int. 33, 51–54. 85. Bierer, L. M., Haroutunian, V., Gabriel, S., Knott, P. J., Carlin, L. S., Purohit, D. P., et al. (1995) Neurochemical correlates of dementia severity in Alzheimer’s disease: relative importance of the cholinergic deficits. J. Neurochem. 64, 749–760. 86. Muller, J. P. and Bruinink, A. (1994) Neurotoxic effects of aluminium on embryonic chick brain cultures. Acta Neuropathol. Berl. 88, 359–366. 87. Murphy, E. J., Roberts, E., and Horrocks, L. A. (1993). Aluminum silicate toxicity in cell cultures. Neuroscience 55, 597–605. 88. Roll, M., Banin, E., and Meiri, H. (1989) Differentiated neuroblastoma cells are more susceptible to aluminium toxicity than developing cells. Arch. Toxicol. 63, 231–237. 89. Szutowicz, A., Tomaszewicz, M., and Bielarczyk, H. (1996) Disturbances of acetyl-CoA, energy and acetylcholine metabolism in some encephalopathies. Acta Neurobiol. Exp. 56, 323–339. 90. Johnson, G. V. and Jope, R. S. (1986) Aluminum impairs glucose utilization and cholinergic activity in rat brain in vitro. Toxicology 40, 93–102. 91. Lai, J. C. K. and Blass, J. P. (1984) Inhibition of brain glycolysis by aluminium. J. Neurochem. 42, 438–446. 92. Lai, J. C. K., DiLorenzo, J. C., and Sheu, K. F. R. (1988) Pyruvate dehydrogenase complex is inhibited in calcium-loaded cerebro-cortical mitochondria. Neurochem. Res. 13, 1043–1048.

234

Szutowicz

93. Lipman J. J. and Tolchard, S. (1989) Comparison of the effects of central and peripheral aluminum administration on regional 2-deoxy-d-glucose incorporation in the rat brain. Life Sci. 45, 1977–1987. 94. Clauberg, M., Smith, C. B., Dang, T., Sokoloff, L., and Joshi, J. G. (1994) Effects of chronic dietary aluminum on local cerebral glucose utilization in rats. Neurobiol. Aging 15, 657–661. 95. Cho, S. and Joshi, J. G. (1988) Effect of long-term feeding of aluminium chloride on hexokinase and glucose–6-phosphate dehydrogenase in the brain. Toxicology 48, 61–69. 96. Schurr, A., Payne, R. S., Miller, J. J., and Rigor B. M. (1997) Brain lactate, not glucose, fuels the recovery of synaptic function from hypoxia upon reoxygenation: an in vitro study. Brain Res. 744, 105–111. 97. Szutowicz, A. and Bielarczyk, H. (1991) Relationships between pyruvate oxidation, acetyl-CoA compartmentalization and acetylcholine synthesis in nerve terminals, in Biochemistry and Physiology of Thiamin Diphosphate Enzymes (Bisswanger, H. and Ullrich, J. eds.) VCH Verlagesellshaft, Weinheim, pp. 367–374. 98. Szutowicz A., Tomaszewicz, M., Bielarczyk, H., and Jankowska, A. (1998) Putative significance of shifts in acetyl-CoA compartmentalization in nerve terminals for disturbances of cholinergic transmission in brain. Dev. Neurosci. 20, 485–492. 99. Uemura, E., Lartius, R. K., and Martens, C. (1993) Aluminum induces neurite elongation and sprouting in cultured hippocampal neurons. J. Geriatr. Psychiatry Neurol. 6, 239–244. 100. Bishop, N. J., Morley, R., Chir, B., Day, J. P., and Lucas, A. (1997) Aluminum neurotoxicity in preterm infants receiving intravenous-feeding solutions. N. Engl. J. Med. 336, 1557–1561. 101. Flynn, D. D., Ferrari-DiLeo, G., Levey, A. I., and Mash D. C. (1995) Differential alterations in muscarinic receptor subtypes in Alzheimer’s disease: implications for cholinergic-based therapies. Life Sci. 56, 869–876. 102. Pascual, J., Fontan, A., Zarranz, J. J., Berciano, J., Florez, J., and Pazos, A. (1991) Highaffinity choline uptake carrier in Alzheimer’s disease: implications for the cholinergic hypotheis of dementia. Brain Res. 552, 170–174. 103. Guela, C. and Mesulam, M. M. (1988) Cholinergic system and related neuropathological predilection patterns in Alzheimer’s disease, in Alzheimer’s Disease (Terry, R. D., Katzman, R., and Bick K. L. eds.), Raven Press, NY, pp. 263–291. 104. Wu, Y. H., Zhou, Z. M., Xiong, Y. L., Wang, Y. L., Sun, J. H., Liao, H. B., and Luo, X. D. (1998) Effects of aluminum potassium sulfate on learning, memory and cholinergic system in mice. Chung Kuo Yao Li Hsueh Pao 19, 509–512. 105. Julka, D., Sandhir, R., and Gill, K. D. (1995) Altered cholinergic metabolism in rat CNS following aluminum exposure: implications on learning performance. J. Neurochem. 65, 2157–2164. 106. Cherroret, G., Desor, D., Hutin, M. F., Burnel, D., Capolaghi, B., and Lehr, P. R. (1996) Effects of aluminum chloride on normal and uremic adult male rats. Tissue distribution, brain choline acetyltransferase activity, and some biological variables. Biol. Trace. Elem. Res. 54, 43–53. 107. Lai, J. C. K., Guest, J. F., Leung, T. K. C., Lim, L., and Davison, A. N. (1979) The effects of cadmium, manganese and aluminium on sodium-potassium-activated and magnesiumactivated adenosine triphosphatase activity and choline uptake in rat brain synaptosomes. Biochem. Pharmacol. 29, 141–146. 108. Zubenko, G. Z. and Hanin, I. (1989) Cholinergic and noradrenergic toxicity of intraventricular aluminium chloride in the rat hippocampus. Brain Res. 489, 381–384. 109. Harkany, T., Lengyel, Z., Kasa, P., and Gulya, K. (1996) Chronic aluminum treatment results in aluminum deposits and affects M1 muscarinic receptors in rat brain. Neurobiology Bp. 4, 35–43.


235

110. Kumar, S. (1998) Biphasic effect of aluminium on cholinergic enzyme of rat brain. Neurosci. Lett. 248, 121–123. 111. Zatta, P., Zambenedetti, P., Bruna, V., and Filippi, B. (1994) Activation of acetylcholinesterase by aluminium (III): the relevance of the metal species. Neuroreport 5, 1777–1780. 112. Jung, P. H., Lehr, P. R., Toniolo, A. M., Bertelle, D., and Desor, D. (1996) Effects of aluminium intoxication on various aspects of learning in the rat, in Metals Ions in Biology and Medicine, vol. 4 (Collery, P., Corbella, J., Domingo, J. L., Etienne, J. C., and Llobet J. M. eds.), John Libbey Eurotext, Paris, pp. 451–453. 113. Platt, B., Carpenter, D. O., Busselberg, D., Reymann, K. G., and Riedel, G. (1995) Aluminum impairs hippocampal long-term potentiation in rats in vitro and in vivo. Exp. Neurol. 134, 73–86. 114. Domingo, J. L., Llorens, J., Sanchez, D. J., Gomez, M., Llobet, J. M., and Corbella, J. (1996) Age-related effects of aluminum ingestion on brain aluminum accumulation and behavior in rats. Life Sci. 58, 1387–1395. 115. Gonda, Z., Lehotzky, K., and Miklosi, A. (1996) Neurotoxicity induced by prenatal aluminum exposure in rats. Neurotoxicology 17, 459–470. 116. Yokel, R. A. (1994) Aluminum exposure produces learning and memory deficits. A model of Alzheimer’s disease, in Toxin-Induced Models of Neurological Disorders (Woodruff, M. J. and Nonneman, A. J. eds.), Plenum Press, NY, pp. 301–318. 117. Shafer, T. J., Mundy, W. R., and Tilson H. A. (1993) Aluminum decreases muscarinic, adrenergic, and metabotropic receptor-stimulated phosphoinositide hydrolysis in hippocampal and cortical slices from rat brain. Brain Res. 629, 133–140. 118. Shafer, T. J., Nostrandt, A. C., Tilson, H. A., and Mundy, W. R. (1994) Mechanisms underlying AlCl3 inhibition of agonist-stimulated inositol phosphate accumulation: role of calcium, G-proteins, phospholipase C and protein kinase C. Biochem. Pharmacol. 47, 1417–1425. 119. Dobson, C. B., Graham, J., and Itzhaki R. F. (1998) Mechanisms of uptake of gallium by human neuroblastoma cells and effects of gallium and aluminum on cell growth, lysosomal protease, and choline acetyltransferase activity. Exp. Neurol. 153, 342–350. 120. Gotti, C., Cabrini, D., Sher, E., and Clementi, F. (1987) Effects of long-term in vitro exposure to aluminum, cadmium or lead on differentiation and cholinergic receptor expression in a human neuroblastoma cell line. Cell. Biol. Toxicol. 3, 431–440. 121. Hofstetter, J. R., Vincent, I., Bugiani, O., Ghetti, B., and Richter, J. A. (1987) Aluminuminduced decreases in choline acetyeltransferase, tyrosine hydroxylase and glutamate decarboxylase in selected regions of rabbit brain. Neurochem. Pathol. 6, 177–193. 122. Beal, M. F., Mazurek, M. F., Ellison, D. W., Kowall, N. W., Solomon, P. R., and Pendelbury, W. W. (1989) Neurochemical characteristics of aluminium-induced neurofibrillary degeneration in rabbits. Neuroscience 29, 339–346. 123. Rao, K. S. and Rao, G. V. (1994) Effect of aluminium on brain mitochondrial monoamine oxidase-A activity: an in vitro kinetic study. Mol. Cell. Biochem. 137, 57–60. 124. Kobayashi, D., Murase, K., Odani, A., Hayashi, K., and Yamauchi, O. (1996) Effects of aluminum (III) on catechol-stimulated nerve growth factor biosynthesis by cultured mouse brain astroglial cells. Bioch. Mol. Biol. Int. 38, 675–684. 125. Zielke, H. R., Jackson, M. J., Tildon, J. T., and Max, S. R. (1993) A glutamatergic mechanism for aluminum toxicity in astrocytes. Mol. Chem. Neuropath. 19, 219–233. 126. Deloncle, R. and Guillard, O. (1990) Mechanism of Alzheimer’s disease: arguments for a neurotransmitter-aluminium complex implication. Neurochem. Res. 15, 1230–1245. 127. Crapper, D. R., Krishnan, S. S., and Quittkat, S. (1976) Aluminium, neurofibrillary degeneration and Alzheimer’s disease. Brain 99, 67–80. 128. Crapper, D. R., Krishnan, S. S., and Dalton, A. J. (1973) Brain aluminium distribution in Alzheimer’s disease and experimental neurofibrillary degeneration. Science 180, 511–513.

236

Szutowicz

129. Perl, D. P. and Good, P. F. (1992) Aluminium and the neurofibrillary tangle: results of tissue microprobe studies. Ciba Found Symp. 169, 217–227. 130. Wisniewski, H. M. and Wen, G. Y. (1992) Aluminium and Alzheimer’s disease. Ciba Found. Symp. 169, 142–154. 131. Xu, N., Majidi. V., Markesbery, W. R., and Ehmann, W. D. (1992) Brain aluminum in Alzheimer’s disease using an improved GFAAS method. Neurotoxicology 13, 735–743. 132. Bouras, C., Giannakopoulos, P., Good, P. F., Hsu, A., Hof, P. R., and Perl, D. P. (1997) A laser microprobe mass analysis of brain aluminum and iron in dementia pugilistica: comparison with Alzheimer’s disease. Eur. Neurol. 38, 53–58. 133. Yasui, M., Kihira, T., and Ota, K. (1992) Calcium, magnesium and aluminum concentrations in Parkinson’s disease. Neurotoxicology 13, 593–600. 134. Corrigan, F. M., Crichton, J. S., Van Rhijn, A. G., Skinner, E. R., and Ward, N. I. (1992) Transferrin, cholesterol and aluminium in Alzheimer’s disease. Clin. Chim. Acta 211, 121–123. 135. Kienzl, E., Jellinger, K., Wruss, W., Horhager, S., and Puchinger, L. (1996) Distribution of individual elements in Alzheimer disease brain tissue, in Metal Ions in Biology and Medicine, vol. 4 (Colery, P., Corbella, J., Domingo L. J., Etienne, J. C., and Llobet, J. M., eds.), John Libbey Eurotext, Paris, pp. 617–619. 136. Makjanic, J., McDonald, B., Li-Hsian Chen, C. P., and Watt, F. (1998) Absence of aluminium in neurofibrillary tangles in Alzheimer’s disease. Neurosci. Lett. 240, 123–126. 137. Bjertness, E., Candy, J. M., Torvik, A., Ince, P., McArthur, F., Taylor, G. A., et al. (1996) Content of brain aluminum is not eleveted in Alzheimer disease. Alzheimer Dis. Assoc. Disord. 10, 171–174. 138. Crapper-McLachlan, D. R., Dalton, A. J., Kruck, T. P. A., Bell, M. Y. B., Smith, W. L., Kalow, W., and Andrews, D. F. Intramuscular desferrioxamine in patients with Alzheimer’s disease. Lancet 337, 1304–1308. 139. Savory, J., Herman, M. M., Erasmus, R. T., Boyd, J. C., and Wills, M. R. (1994) Partial reversal of aluminium-induced neurofibrillary degeneration by desferrioxamine in adult male rabbits. Neuropathol. Appl. Neurobiol. 20, 31–37. 140. Starkey, B. J. (1987) Aluminium in renal disease: current knowledge and future developments. Ann. Clin. Biochem. 24, 337–344. 141. Menendez-Fraga, P., Fernandez-Martin, J. L., Blanco-Gonzalez, E., and Cannata-Andia, J. B. (1998) Low percentage of aluminoxamine and ferrioxamine in uremic serum after desferrioxamine administration. Clin. Chem. 44, 1262–1268. 142. Reusche, E., Koch, V., Friedrich, H. J., Nunninghoff, D., Stein, P., and Rob, P. M. (1996) Correlation of drug-related aluminum intake and dialysis treatment with deposition of argyrophilic aluminum-containing inclusions in CNS and in organ systems of patients with dialysis-associated encephalopathy. Clin. Neuropathol. 15, 342–347. 143. Russo, L. S., Beale, G., Sandroni, S., and Ballinger, W. E. (1992) Aluminium intoxication in undialysed adults with chronic renal failure. J. Neurol. Neurosurg. Psychiatry 55, 697–700. 144. Popinska, K., Kierkus, J., Lyszkowska, M., Socha, J., Pietraszek, E., Kmiotek, W., and Ksiazyk, J. (1998). Concentration of serum aluminum in children on long-term parenteral nutrition, in Metal Ions in Biology and Medicine, vol. 5 (Earl, P., Colery, P., Braner, V., Negrett de Bratter, L., and Khassanova, J. C., eds.), John Libbey Eurotext, Paris, pp. 617–619. 145. Hantson, P., Mahieu, P., Gersdorff, M., Sindic, C., and Lauwerys, R. (1995) Fatal encephalopathy after otoneurosurgery procedure with an aluminum-containing biomaterial. J. Toxicol. Clin. Toxicol. 33, 645–648.

Ecology of Microbial Neurotoxins

237

III Natural Toxins of Microbial Origin

238

Llewellyn


239

13 Ecology of Microbial Neurotoxins Lyndon E. Llewellyn

1. INTRODUCTION Biosynthesis of toxins uses precious cellular energy and it would seem unlikely that evolution would be forgiving enough to tolerate wasted metabolism. Although there is much debate about this proposition in the guise of secondary vs primary metabolites (1,2), it is a reasonable hypothesis that toxins of all kinds should play some beneficial role. This return is apparent when one considers toxins used for prey capture or selfdefense as occurs with venoms. For microbial neurotoxins, however, the identity of this biological profit remains a mystery. This is especially so when one considers that the microbes that manufacture these toxins, and those microorganisms that surround them, do not possess nerves nor many of the molecular systems that characterize nerves. The few exceptions to this rule are those microbial toxins that attack generic cellular systems common to many cell types, including nerve cells. An example is the dinoflagellate toxin okadaic acid, an inhibitor of certain serine-threonine protein phosphatases, enzymes that occur widely in different cell types in animals and plants as well as microorganisms (3–5). In fact, an okadaic acid sensitive form of this enzyme has been isolated from the dinoflagellate, Prorocentrum lima, an established producer of okadaic acid (6). One can hypothesize then that okadaic acid may act as a physiological regulator of this enzyme within the dinoflagellate itself or upon unrelated microbes in its vicinity. The larger mystery lies with microbial neurotoxins that attack cellular and molecular neural processes not present in microbes. Aquatic microbes live in a unique situation in that they are surrounded by a medium that may easily transmit the toxins and the organisms comparatively large distances. In terrestrial microorganisms, toxins would have to be volatile to be able to mediate their effect far from the toxin progenitor. The effects of these terrestrial neurotoxins arise then from either infection of other organisms or by ingestion of intoxicated food items (7,8). The distribution of these toxins in the environment is therefore artificial, governed mostly by hygiene and medical measures. This does not apply to neurotoxic microbes found in marine and freshwater environments as they are not infective and management measures are reactive, rather than preemptive, to simply quantify their potential threat to public health (9). This governance by mainly natural conditions and From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

239

240

Llewellyn

the flow and distribution of the toxins throughout the environment makes them a good model with which to examine the question posed above, namely what benefit, if any, is conferred upon microbes that produce these noxious chemicals. 2. AQUATIC NEUROTOXINS AND THE MICROBES THAT PRODUCE THEM 2.1. Their Taxonomic Affinities Harking back to the days of Leeuwenhoek when he looked down the first microscope, all the organisms that were no longer invisible were by definition microscopic, and there are representatives of each of the five Kingdoms that can be seen only with the aid of magnification. Neurotoxic microbes however are found only in two of these kingdoms (Table 1; 10). Within the Protoctist Kingdom, neurotoxins are elaborated by diatoms and dinoflagellates. Protoctists are a group of eukaryotic organisms, which are not always microscopic, distinguished from other eukaryotes because they do not develop from blastula (animals), embryos (plants), or spores (fungi). Diatoms and dinoflagellates have little else in common with diatoms being characterized by a bipartite siliceous test, whereas dinoflagellates possess a cellulosic test. These two groups can be further distinguished by their mode of locomotion with dinoflagellates having two flagella while diatoms are either nonmotile or move by gliding upon secretions. Specifically, neurotoxic diatoms are pennate diatoms, a division that is bilaterally symmetrical and motile. Of all of the dinoflagellate species, only seven are presently known to elaborate neurotoxins (11,12). Similarly, only one diatom genus, Pseudo-nitzschia, contains neurotoxic species (11). Cyanobacteria are members of the kingdom Monera, which contains all the prokaryotic microorganisms, and rely upon photosynthesis for energy production. Evidence is gathering that other Monerans, such as some species of Vibrio, Pseudomonas, Alteromonas, and Moraxella, produce tetrodotoxin and possibly saxitoxin (22–25), but it is a subject of some debate (26). Like the protoctists, neurotoxin production is taxonomically restricted within the cyanobacteria with only two genera, Anabaena and Aphanizomenon, being neurotoxic (27). What we can draw from this is that neurotoxin production by microbes is only known to occur in several genera of the many thousands that make up these two large Kingdoms. Further, not every species within these genera have been found to elaborate such toxins. Finally, the many genera between these neurotoxic microbes on the phylogenetic tree are not known to be toxic and this is especially interesting when one considers the taxonomic divergence between microorganisms that elaborate the same toxins. For instance, dinoflagellates that are sources of saxitoxin and tetrodotoxin are phylogenetically very distant from the cyanobacteria and other bacteria that produce these toxins. 2.2. Diversity in Neurotoxin Structure All of the known neurotoxins from marine and freshwater microbes are small organic molecules (Table 1; Fig. 1), and are biologically active without modification. This is in contrast to the proteinaceous microbial neurotoxins such as botulinum and tetanus toxin, which are at least 150 kDa, and require proteolytic cleavage before they are fully

Parent toxin

Example producers

Taxonomic group

Molecular action

Marine (M) or freshwater (F) Reference

Brevetoxin

Gymnodinium breve

Dinoflagellate

Convert Na channel to hyperexcitable state

M

(13)

Ciguatoxin

Gambierdiscus toxicus

Dinoflagellate

Convert Na channel to hyperexcitable state

M

(12)

Saxitoxin

Anabaena circinalis Alexandrium catenella Moraxella sp.

Cyanobacteria Dinoflagellate Bacteria

Block sodium flow through Na channel

F M M

(14) (15) (16)

Tetrodotoxin

Alexandrium tamarense Vibrio spp.

Dinoflagellate Bacteria

Block sodium flow through Na channel

M M

(17) (18)

Anatoxin-a

Anabaena flos-aquae

Cyanobacteria

Stimulates nicotinic acetylcholine receptors

F

(19)

Anatoxin -a(s) Anabaena flos-aquae

Cyanobacteria

Acetylcholinesterase inhibition

F

(20)

Excitation of kainic IgluR

M

(21)

Domoic acid

Pseudo-nitzschia pungens Diatom forma multiseries


Table 1 Microbial Neurotoxins, Their Origin, and Mode of Action

241

242

Llewellyn

Fig. 1. Structures of a typical member of toxin families described here; (A) domoic acid (30); (B) saxitoxin (31); (C) tetrodotoxin (32–34); (D) anatoxin-a (35); (E) the brevetoxin PbTx-1 (37); (F) anatoxin-a(s) (36); and (G) ciguatoxin (38).

active and able to induce their neurotoxic effect (8). These organic toxins can be highly lethal to mammals and pose a public-health risk to both humans (11,27) and farming livestock (28). For example, the most toxic of these organic neurotoxins is saxitoxin, for which only several micrograms per kilogram are required to kill mice when they are injected intraperitoneally (29). Many analogs of these toxins can be manufactured by the same microbe, and in some cases, produced simultaneously. One example is Alexandrium minutum, which can produce six different saxitoxin analogues in culture at one time (39). In total, there are over 20 natural analogs of saxitoxin (40). This phenomena of high natural structural diversity is not restricted to saxitoxin with there being nine variants of brevetoxin (13) produced by the dinoflagellate Ptychodiscus breve. There is some evidence that the suite of organic toxin analogs produced by these microbes is constant and they can be used as biochemical markers (41), indicating that the mechanism for their manufacture is innate and relatively resistant to nature’s vagaries such as variations in the availability of substrates for the enzymes. This potential hardwiring of the composition of a suite of toxins is rather remarkable given that evolution of metabolite complexity would seem highly intricate. Organic neurotoxins are one step removed from the genetic code, the source of most phenotypic


243

variation, being generated by metabolizing enzymes. Stable structural diversity of organic neurotoxins requires mutations of the responsible enzymes that does not adversely affect enzyme function and must also alter the metabolic process in such a way that it results in a chemical modification of the metabolite. Further, multipleenzyme steps are sometimes required to generate a metabolite, complicating even more the changes in the machinery necessary to create analogs of organic neurotoxins. On the face of it, then, evolution of such structural diversity of these toxins would seem convoluted and improbable. It has been documented in some bacteria, however, that simple genetic mutations can generate new metabolic phenotypes with very divergent substrate needs (42), which would lead to very different enzyme products being generated. This demonstrates that minor changes in DNA sequence can indeed result in major changes in their catabolizing enzymes and therefore metabolism and metabolite production. 2.3. Nerve Mechanisms Targeted by Microbial Neurotoxins A nerve comprises a series of specialized cells that transmit an electrical impulse along its length by sequential localized electrical events across the cell membrane. The events comprise progressive elimination and then reestablishment of a transmembrane potential difference created by unequal cation distribution between the intra- and extracellular compartments. Depolarizing the cell membrane triggers conformational changes in structurally distinct and cation-specific channels that then allow them to selectively pass calcium, sodium, and potassium. Transfer of this action potential between nerve cells is achieved through ligand-gated ion channels that open in response to neurotransmitters released in vesicles from the nerve terminus preceding the gap junction. Once open, these ligand-gated ion channels pass ions to once again trigger cellular depolarization and continue the electrical impulse. In some cases the next cell after a nerve terminus may not be a nerve but another excitable cell such as a muscle or secretory cell. The mechanisms in this pathway upon which microbial neurotoxins act are listed in Table 1 and elaborated upon below. 2.3.1. Voltage Dependent Sodium Channels Far more sodium is found outside a nerve cell than within it. During a nerve’s electrical impulse, sodium selective channels react to altered cell-membrane voltages by opening and allowing these ions to enter the cell, thus removing the potential difference across the membrane. This voltage-dependent sodium (Na) channel is blocked by saxitoxin (Fig. 1B) and its relatives (43–45). Although structurally dissimilar to saxitoxin, tetrodotoxin (Fig. 1C) competes for the same binding site on the Na channel attacked by saxitoxin (43,44). This inability to transport sodium ions renders nerves nonfunctional. The Na channel is also targeted by brevetoxin (Fig. 1E; 46) and ciguatoxin (Fig. 1G; 47), binding to a different region of the Na channel than via that which saxitoxin and tetrodotoxin act, and eliciting a very different effect. They bring the Na channel nearer the threshold where the change in potential difference across the cell membrane triggers the conformational changes to open and allow passage of Na+ ions. In essence, this makes the nerve hyperexcitable and thereby renders it dysfunctional (48–50).

244

Llewellyn

2.3.2. Ligand-Gated Ion Channels 2.3.2.1. GLUTAMATE RECEPTORS

Glutamate binds to a variety of receptors but one particular class responds to its binding by opening and allowing cations to enter the nerve cell. This glutamate receptor class is referred to as ionotropic glutamate receptors (IgluRs). Several forms of IgluRs are recognized based on their sensitivity to three compounds: namely, N-methylD-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate. Not surprisingly, domoic acid (Fig. 1A), which is structurally related to kainic acid, acts upon the kainate IgluR and is a neuroexcitant (51). IgluRs initiate many intracellular processes either directly by their introduction of cations or indirectly by the second messengers produced by these cascades such as cyclic AMP, or reactive oxygen species (ROS) (52,53). 2.3.2.2. NICTONIC ACETYLCHOLINE RECEPTOR Where nerve meets nerve or other effector tissues such as muscle, one finds acetylcholine receptors in the cell membrane after the gap between the abutting cells. Binding of acetylcholine opens the pore of the channel, nonselectively allowing many cations from the extracellular milieu to enter the nerve (54). The increase of intracellular cations leads to depolarization of the nerve terminus, thereby triggering a cascade of events that continue the action potential. Anatoxin-a (Fig. 1D) from the cyanobacterium Anabaena circinalis is a powerful agonist of this particular class of acetylcholine receptors, defined by their sensitivity to nicotine (55), being many times better than acetylcholine itself in stimulating the receptor to open (56). Not only does anatoxin-a therefore outcompete and displace acetylcholine, in the process it overstimulates this receptor, thereby eliminating its ability to properly function and by extension, incapacitating the nerve. 2.3.3. Acetylcholinesterase

Within the junction between nerves, or between nerves and other excitable cells like muscle, if acetylcholine were allowed to linger it would continually stimulate the acetylcholine receptor. This would lead to uncoordinated firing of nerves and negation of the cellular potential difference necessary for any excitable cell to function. The enzyme acetylcholinesterase degrades the acetylcholine into its constituents that are reabsorbed into the presynaptic nerve terminal for acetylcholine regeneration to be released at the next nerve impulse. This acetylcholinesterase is potently inhibited by a second anatoxin, anatoxin-a(s) (Fig. 1F; 57), which leads to a build-up of acetylcholine and therefore hyperactivity of excitable cells. 2.4. Phylogenetic Distribution of Microbial Neurotoxin Targets Since none of the molecular targets of these microbial neurotoxins are present in microorganisms, and bearing in mind that these systems must have evolved from some ancestral processes, how closely related are these systems to these microbes in an evolutionary sense? The Na channel has only been detected in multicellular animals, including all vertebrates and in most invertebrate phyla including molluscs (58), jellyfish (59), and flatworms (60). It is closely related in both amino acid sequence, and presumably structure,


245

to the other major voltage-dependent cation channels, the calcium (Ca) and potassium (K) channels (61). Both the Na and Ca channels contain a quadruplicate amino acid sequence repeat believed to form four homologous domains around the central ionconducting pore. These two channels are so closely related in fact, that site-directed mutations can transfer properties between the two channels such as their cation selectivity (62). Unlike these two channel families, the K channel has no replication within its amino acid sequence but is homologous to the single domains of the Na and Ca channels. Unlike the Na and Ca channels then, which simply need to fold themselves into the correct formation within the cell membrane to function, the K channel must fold correctly and then gather together with three other similar subunits around the ionconducting pore to become a functional ion channel. As stated earlier, the Na channel has not been found very far back in evolutionary history. Genes, or fragments thereof, encoding Ca and K channels, have been found in various worms (63,64) and protozoal parasites (65), respectively. A diverse family of neuronal acetylcholine (nACh) receptors have been cloned from the nematode Caenorhabditis elegans (66). Even though this is the most primitive organism from which these receptors have been obtained, this complexity would indicate that the nematode nACh receptor is not yet the most primitive form of this channel, unless they arose de novo in this group of animals, which is rather unlikely. It is also true for IgluRs, which have been found in C. elegans (67). Curiously, gene fragments homologous to IgluRs have been isolated from the plant Arabidopsis thaliana (Genbank accession number AF170494), but no functional evidence has yet been presented that these sequences are indeed IgluRs. It must be borne in mind that the known phylogenetic distribution of these systems may only reflect search effort. What it highlights at this point is the question as to why microbes have evolved such exquisite pharmacological regulators of molecular processes apparently distant from them in evolutionary history. 3. ABIOTIC FACTORS THAT MAY INFLUENCE TOXIN PRODUCTION The aquatic environments in which these microbes live are not constant and they are exposed to many different conditions in the course of a single day. Their ability to transport themselves to different microenvironments, especially when they migrate vertically, would expose them to physical gradients (68) such as increasing pressure and decreased light penetration. Those microbes that require photosynthesis, namely dinoflagellates and cyanobacteria, require sunlight for survival and its availability would of course affect microbial growth (69). This limits also the realm in which they can survive and flourish. Some of these microbes, in particular dinoflagellates, can enter a dormant encysted stage that can be found in benthic environments and remain toxic (70,71). Sunlight may not only affect microbes dependent on photosynthesis for their nutrition, but may also have photolytic impacts on toxins released into environment. For example, anatoxin-a degrades within an hour or two when exposed to natural levels of ultra-violet (UV) light whereas it is stable for days when stored in the dark (72). Stratification can occur in oceanic and freshwater systems in terms of nutrient availability and temperature to name but two variables. Further still, these factors may vary with time as well as spatially. For instance, water temperature fluctuates during the

246

Llewellyn

course of a single day especially in the transition from day to night, as well as between annual seasons. On a gross level, temperature is an important factor controlling microbial growth in culture, and presumably in nature, and toxin production tends to be associated with various stages of microbial growth (73,74), if produced at all. For example, gametes, zygotes, and the first few stages of vegetatively growing cells of the domoic acid producing diatom Pseudo-nitzschia pungens f. multiseries do not produce domoic acid (74). Apart from this gross effect, there are many examples of enzymes whose metabolic activity changes markedly with temperature (75). Although the enzymes that metabolize these neurotoxins have yet to be identified, like most enzymes, they too would have optimal temperatures for their function. Metabolite production including toxin generation would be affected then if the microbe were in suboptimal thermal conditions. Nutrient availability will also affect toxin production. For example, nitrogen is a critical element of all of these toxins. It can be patchily distributed throughout the water column (76) and it has been shown that microbes, including dinoflagellates, will migrate to artificially stratified high areas of nitrogen and this access to nitrogen correlates with the levels of their toxin production (68). 4. MICROBIAL EFFECTS OF NEUROTOXINS There are two routes by which a microbial toxin may affect itself or its conspecifics in nature. First, while resident in the cell, it would be exposed to all of its own cellular processes unless it is compartmentalized or conjugated to another molecule. Alternatively, the toxins may act after being released from the cell. This release could be passive after degradation of the cell after death or be actively released by secretion or excretion. Such release of these neurotoxins has been reported (77,78) but whether this release is active or passive has not been determined. So then, while these toxins reside inside the cell do they play any regulatory role? And if released into the environment, does it then affect cell-surface receptors and impact upon the behavior and biology of the toxin progenitor or neighboring microbes? 4.1. Intracellular Actions As described earlier, the receptors and enzymes known to bind or react with these toxins have not been found in any representative of the groups to which these microbes belong. This does not exclude, however, that other receptors or molecular targets, not yet discovered, may be affected by these compounds and a search for any so-called “orphan” receptors has not yet been reported. Localization of these toxins within the microbes may provide a clue as to whether these molecules are bound in any form within the producing microbe. The only reported study tackling this question for these neurotoxins used antibodies generated to saxitoxin, detecting the toxin within the nucleus as well as on the periphery of small granules thought to be starch grains. Labeling occurred on or close to the permanently condensed chromosomes in the nucleus (79). In this case, it seems that the neurotoxin is not randomly distributed throughout the nucleus or cytoplasm, giving a hint that there may indeed be intracellular molecules to which saxitoxin may bind.


247

4.2. Extracellular Activity Incorporation of these neurotoxins in the media used to culture these organisms has never been reported. Discerning whether such a manipulation would have any impact upon microbial growth or behavior would be difficult, however, because one must presume that some of the toxin is always in the media and this has been shown for some (77,78). Metabolically inert agents, such as toxin antibodies, that chelate these toxins would remove the toxin from the media and eliminate this complication. Exposure in culture is exacerbated by the fact that it is an enclosed system. In nature, the watery medium would provide an immediate dilution and any impacts the released toxin may have would most likely be limited to themselves or their near neighbors, where the concentration of the toxin may remain high enough to be effective. Many of these microbes participate in a phenomenon, the so-called harmful algal blooms, where their populations undergo sudden explosions in their numbers attaining high population densities (80). Also, these blooms are not often monospecific and individual microbes would come into close proximity not only to many of their cohorts but with many other microbial species. This dense microbial accumulation of these microbes may provide an opportunity for the toxins to act upon other microbes, but such an action might be secondary as these phenomena are occasional events and somewhat haphazard in their occurrence. 5. NONMICROBIAL ORGANISMS EXPOSURE TO MICROBIAL NEUROTOXINS All of these neurotoxins, apart from the anatoxins, are accumulated into other organisms (11,12), which must be able to resist the toxin’s action at least to some degree. They may also exert their toxicity and seriously harm, if not kill, their consumers. 5.1. Plankton Planktonic life stages of many macroscopic organisms, as well as those organisms that spend their entire life cycle within the planktonic niche, may encounter these neurotoxic organisms either as potential food items or the toxins after their release after cell death. Grazing rates upon toxic microbes by micropredators such as planktonic crustacea are sometimes dramatically reduced compared to nontoxic microbes (81–83). This reduced feeding activity appears not to be a simple physiological response to the neurotoxin’s acting upon their target in these organisms (83), but may be a chemosensory effect. In fact, some microcrustacea can sense toxic vs nontoxic strains of neurotoxic microbes (81) and selectively feed upon nontoxic microbes. But this repulsion of potential predators is not universal and some plankton readily feed upon toxic microorganisms (82). These toxins therefore do not prevent consumption of the microbe and the toxin may flow into the food chain. 5.2. Molluscs Only the anatoxins and ciguatoxin have not yet been found to be bioaccumulated by certain species of molluscs. More often than not, these molluscs are bivalves that filter organic matter from the water column for digestion. As well as nutrients, they may also assimilate the neurotoxins from the digested microbes (11). By definition, shellfish

248

Llewellyn

that accumulate these toxins must be resistant to their actions, as the compounds often reside in the animal’s tissue for long periods of time (84). Shellfish do possess the receptors inhibited by some of these toxins. The mollusc Aplysia has kainate-sensitive glutamate receptors (85) and therefore they may also be sensitive to domoic acid, and a Na channel that is sensitive to tetrodotoxin (86). Resistance may arise from these animals having insensitive isoforms of the enzymes and receptors targeted by the toxins. In fact, neurons from shellfish known to accumulate saxitoxin are resistant to the toxin (87,88). Site directed-mutagenesis studies have shown that simple mutations of a saxitoxin-sensitive isoform of the Na channel can result in much reduced toxin sensitivity (89) and a naturally occurring mutation may underlie the apparent resistance of shellfish nerves to saxitoxin or its analogs. An alternative means by which resistance may be conferred is physical separation of the toxin away from its active site. Tetrodotoxin and saxitoxin need to act upon the extracellular opening of the Na channel, and if the toxins are retained intracellularly, then they will never be able to exert their pharmacological action. Some of these toxins possess chemical attributes that make them poor candidates for uptake across digestive epithelia. For instance, domoic acid is polar and therefore hydrophilic and its ability to penetrate into shellfish through digestive epithelia would be expected to be poor. This is the case, and it has been found to be very poorly accumulated by shellfish (90). By contrast, lipophilic compounds like the brevetoxins may be expected to be accumulated effectively and retained by shellfish for lengthy periods after intoxication. This has been shown to be true, with detectible levels of brevetoxin remaining in shellfish some months after the toxic dinoflagellate, Ptychodiscus brevis, had disappeared (84). But curiously, there is one example where a microbial neurotoxin is regularly found in a nonfilter-feeding mollusc. Tetrodotoxin occurs in the venom of the blue-ringed octopus (91) and can also be found in the octopus’ eggs (92). Tetrodotoxin does not affect the nerves of the blue-ringed octopus (93), which is reminiscent of bivalve shellfish that may accumulate saxitoxin. If this toxin does still originate from a microorganism within this mollusc (94), then the octopus has tamed it, for lack of a better word, so that the toxin may be used for the octopus’ own needs. 5.3. Crustacea A number of crustacea are known to harbor these microbial neurotoxins. In fact, the filter-feeding sand crab Emerita analoga was one of the first animals from which saxitoxin was discovered prior to it being realized that it came from a dinoflagellate (95). But another group of crustacea, primarily but not solely those that belong to the family Xanthidae, regularly carry saxitoxin and tetrodotoxin in their tissue (96,97). These crabs are not filter-feeders and are unlikely to become toxic in the same manner as bivalve shellfish. Bacteria from the intestines of these crabs have been suggested as producing tetrodotoxin (18). This along with evidence for saxitoxin coming from bacteria as well as dinoflagellates (23,25) has been seen by some as circumstantial evidence that the intestinal bacterial flora produces these neurotoxins for accumulation by the crabs. Like the bivalve molluscs described earlier, some of these crabs possess nerves resistant to these particular toxins (98) allowing them to tolerate the toxins within their


249

tissues. These particular species are those that are often found to harbor the toxins, whereas those that do not accumulate the toxin have nerves that cannot tolerate the toxins. Members of this family of crabs also have two distinct proteins in their haemolymph that bind tetrodotoxin and saxitoxin, respectively (99,100). In the case of the saxitoxin-binding haemolymph protein, its taxonomic occurrence was inversely related to known neural-toxin sensitivity with regularly toxic crabs not having the protein, while those not known to carry the toxins having the protein in their haemolymph (100). The haemolymph would be the first line of defense against a digested toxin and it was postulated that its removal allowed the toxin to enter the crab’s tissues where it was not biologically active because the nerves were no longer sensitive. If this postulate is correct, what then was the selective advantage to the crab to evolve a second toxin-resistance mechanism so that it could tolerate the toxin within its tissues, while eliminating an utterly different means from the haemolymph of negating the toxin before it even entered the crab’s tissues. 5.4. Fish Ciguatoxin, tetrodotoxin, and domoic acid are the only microbial neurotoxins that are accumulated by fish. Tetrodotoxin is most famous for its occurrence in puffer fish, which are the staple of the famous fugu tradition in Japan. Evidence has been presented that bacteria from the gut and skin of these fish produce tetrodotoxin (101,102). Release of this toxin appears to act as a sexual pheromone (103) and its release into the water by the fish can also be induced by artificial stress (104). Ciguatoxin can be detected in the flesh of predatory fish (12) after it has passed along the food chain. Anchovies have been found to accumulate domoic acid, which then passed the toxin higher up the food chain to kill seabirds (105) and sea lions (106). Mackerel also have been found to carry saxitoxin in their viscera and transfer this toxin to higher level consumers (107). But many other fish are vulnerable to the actions of these microbial neurotoxins and some may cause massive fish kills such as those observed in the Gulf of Mexico due to brevetoxin (108). 5.5. Mammals As these toxins travel up the food chain, they eventually reach humans. Because of the social and economic costs of human poisonings, much research effort has been directed towards the effects of microbial neurotoxins upon humans and detailed descriptions can be found elsewhere (11). In essence, though, some of these toxins lead to human fatalities, e.g., saxitoxin, tetrodotoxin, after the ingestion of shellfish or pufferfish, which, as described earlier, may bear these toxins in their flesh. Other mammals are also susceptible to these neurotoxins, with many sea lion deaths occurring due to domoic acid (106), manatees dying from brevetoxin intoxication (109), whales from saxitoxin (107), and sheep being killed after drinking from standing freshwater contaminated by a cyanobacterial bloom (28). Accumulation of the toxins within mammalian tissues is unknown. 6. SYNTHESIS From all of this, are there any common themes from which we can develop hypotheses that will allow us to unravel the biological role these toxins play for the microbes?

250

Llewellyn

First, although microbes are not nerves, they are excitable cells with a potential difference across their cell membrane (110,111). Depolarization events underlie some responses of microbes such as mechanical and chemical sensing. Other important electrical events in microbes are ciliary and flagellate movements (112–114), the means by which many microbes and indeed some of the neurotoxic microbes propel themselves. Does the action of these compounds upon excitatory processes in other animals reflect a residual propensity to bind to structures necessary for controlling transmembrane potential differences. For example, are there any common structural motifs between microbial ion channels and those we know to be attacked by these toxins that are necessary for function? An effect upon these systems need not result in the disabling of the microbial systems. To the microbes, they may not be toxic and may simply act as regulators and have subtle physiological effects. The most obvious hypothetical role for these toxins is as a chemical defense for the microorganism. These microbes and their toxins are known to have dramatic effects on the biota around them, causing fish kills and deaths of higher-level consumers. But this effect is usually at their own expense as they generally must be eaten to allow the toxin to act; thus the toxin provides little, if any, protection to the individual microorganism. Also, a number of animals tolerate the toxins, allowing them to consume the toxic microbes with impunity. It is unlikely then that on the individual level these compounds provide any defense against predation, and this avenue is unlikely to provide an answer as to why these compounds exist. Part of the mystery of some of these microbial neurotoxins is the sporadic occurrence of some of the toxins taxonomically, with some occurring in unrelated taxonomic groups. Our attention is usually drawn to these organisms when we see the results of their toxins, such as deaths of other organisms. Are the microbes that have attracted our attention merely greater producers of these toxins and happen to live in circumstances that result in our attention or curiosity being piqued? Are these toxins much more widely occurring throughout the various Kingdoms than we realize? At present, we magnify the amount of toxin being produced by laboratory culture or they are accumulated in a higher level consumer, elevating these toxins to levels within the sensitivity range of our present day analytical methods. Broader surveys for these toxins and development of highly sensitive methods for toxin detection will reveal whether these toxins are indeed unique to certain species or are more fundamental chemicals present in many organisms not yet discovered. Neurotoxins from marine and freshwater microorganisms can elicit many effects upon other organisms, either directly or after biomagnification through the food web. As can be seen, the function fulfilled by these neurotoxins for the microbe is not readily apparent. Are they merely byproducts of other metabolic processes that happen to have biological activity and not necessarily play any role for the progenitors? If this were the case, why then would an organism pursue conditions that result in increased production of these metabolites (68)? Or is this merely a consequence of elevated metabolism in the presence of more nutrients, and all metabolite levels, including secondary compounds, are increased? Many gaps remain in our knowledge of what these neurotoxins achieve for their microbial manufacturers. Attention to understanding these toxins from the microorganism’s perspective and evolution of techniques that will enable us to examine these toxins at the cellular level will greatly advance our knowledge.


251

ACKNOWLEDGMENTS This is publication number 1052 of the Australian Institute of Marine Science. REFERENCES 1. Vining, L. C. (1992) Secondary metabolism, inventive evolution and biochemical diversity: a review. Gene 115, 135–140. 2. Stone, M. J. and Williams, D. H. (1992) On the evolution of functional secondary metabolites (natural products). Mol. Microbiol. 6, 29–34 3. Takai, A., Murata, M., Torigoe, K., Isobe, M., Mieskes, G., and Yasumoto, T. (1992) Inhibitory effect of okadaic acid derivatives on protein phosphatases. Biochem. J. 284, 539–544. 4. Haystead, T. A., Sim, A. T., Carling, D., Honnor, R. C., Tsukitani, Y., Cohen, P., and Hardie, D. G. (1989) Effects of the tumour promoter okadaic acid on intracellular protein phosphorylation and metabolism. Nature 337, 78–81. 5. Solow, B., Young, J. C., and Kennelly, P. J. (1997) Gene cloning and expression and characterization of a toxin-sensitive protein phosphatase from the methanogenic archaeon Methanosarcina thermophila TM–1. J. Bacteriol. 179, 5072–5075. 6. Boland, M. P., Taylor, M. F., and Holmes, C. F. (1993) Identification and characterisation of a type–1 protein phosphatase from the okadaic acid-producing marine dinoflagellate Prorocentrum lima. FEBS Lett. 334, 13–17. 7. Sugiyama, H. (1980) Clostridium botulinum neurotoxin. Microbiol. Rev. 44, 419–448. 8. Pellizzari, R., Rossetto, O., Schiavo, G., and Montecucco, C. (1999) Tetanus and botulinum neurotoxins: mechanism of action and therapeutic uses. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 354, 259–268. 9. Park, D. L. (1995) Surveillance programmes for managing risks from naturally occurring toxicants. Food Addit. Contam. 12, 361–371. 10. Margulis, L. and Schwartz, K. V. (1998) Five Kingdoms: An Illustrated Guide to the Phyla of Life of Earth. W. H. Freeman and Co., San Francisco, CA. 11. Llewellyn, L. E. (2000) Shellfish chemical poisoning, in Foodborne Diseases Handbook, vol. 4, Seafood and Environment Toxins (Hui, Y. H., Kitts, D. D., and Stanfield, P. S., eds.), Marcel Dekker, New York, NY, pp. 77–108. 12. Lewis, R. J. and Holmes, M. J. (1993) Origin and transfer of toxins involved in ciguatera. Comp. Biochem. Physiol. C. 106, 615–628. 13. Baden, D. G. (1989) Brevetoxins: unique polyether dinoflagellate toxins. FASEB J. 3, 1807–1817. 14. Humpage, A. R., Rositano, J., Bretag, A. H., Brown, R., Baker, P. D., Nicholson, B. C., and Steffensen, D. A. (1994) Paralytic shellfish poisons from Australian cyanobacterial blooms. Aust. J. Mar. Freshwat. Res. 45, 761–771. 15. Schantz, E. J., Lynch, J. M., Vayvada, G., Matsumoto, K., and Rapoport, H. (1966) The purification and characterization of the poison produced by Gonyaulax catenella in axenic culture. Biochemistry 5, 1191–1195. 16. Kodama, M., Ogata, T., Sakamoto, S., Sato, S., Honda, T., and Miwatani, T. (1990) Production of paralytic shellfish toxins by a bacterium Moraxella sp. isolated from Protogonyaulax tamarensis. Toxicon 28, 707–714. 17. Kodama, M., Sato, S., Sakamoto, S., and Ogata, T. (1996) Occurrence of tetrodotoxin in Alexandrium tamerense, a causative dinoflagellate of paralytic shellfish poisoning. Toxicon 34, 1101–1105. 18. Noguchi, T., Jeon, J. K., Arakawa, O., Sugita, H., Deguchi, Y., Shida, Y., and Hashimoto, K. (1986) Occurrence of tetrodotoxin and anhydrotetrodotoxin in Vibrio sp. isolated from the intestines of a Xanthid crab, Atergatis floridus. J. Biochem. 99, 311–314.

252

Llewellyn

19. Devlin, J. P., Edwards, O. E., Gorham, P. R., Hunter, N. R., Pike, R. K., and Stravic, B. (1977) Anatoxin-a, a toxic alkaloid from Anabaena flos-aquae NRC–44h. Can. J. Chem. 55, 1367–1371. 20. Matsunaga, S., Moore, R. E., Niemczura, W. P., and Carmichael, W. W. (1989) Anatoxin-a(s), a potent anticholinesterase from Anabaena flos-aquae. J. Am. Chem. Soc. 111, 8021–8023. 21. Bates, S. S., Bird, J. C., De Freitas, A. S. W., Foxall, R., Gilgan, M., Hanic, L. A., et al. (1989) Pennate diatom Nitzschia pungens as the primary source of domoic acid, a toxin in shellfish from eastern Prince Edward Island, Canada. Can. J. Fish. Aquat. Sci. 46, 1203–1221. 22. Do, H. K., Kogure, K., and Simidu, U. (1990) Identification of deep-sea-sediment bacteria which produce tetrodotoxin. Appl. Environ. Microbiol. 56, 1162–1163. 23. Kodama, M., Ogata, T., Sato, S., and Sakamoto, S. (1990) Possible association of marine bacteria with paralytic shellfish toxicity of bivalves. Mar. Ecol. Prog. Ser. 61, 203–206. 24. Matsui, T., Taketsugu, S., Kodama, K., Ishii, A., Yamamori, K., and Shimizu, C. (1989) Production of tetrodotoxin by the intestinal bacteria of a puffer fish Takifugu niphobles. Bull. Jap. Soc. Sci. Fish 55, 2199–2203. 25. Kodama, M., Ogata, T., and Sato, S. (1988) Bacterial production of saxitoxin. Agric. Biol. Chem. 52, 1075–1077. 26. Matsumura, K. (1995) Reexamination of tetrodotoxin production by bacteria. Appl. Environ. Microbiol. 61, 3468–3470. 27. Falconer, I. R. (1998) Algal toxins and human health, in Handbook of Environmental Chemistry, vol. 5, Part C, Quality and Treatment of Drinking Water II (Hrubec, J., ed.), Springer Verlag, Berlin, Heidelberg, pp. 53–82. 28. Negri, A. P., Jones, G. J., and Hindmarsh, M. (1995) Sheep mortality associated with paralytic shellfish poisons from the cyanobacterium Anabaena circinalis. Toxicon 33, 1321–1329. 29. Sommer, H. and Meyer, K. F. (1937) Paralytic shellfish poisoning. Arch. Pathol. 24, 560–598. 30. Wright, J. L. C., Falk, M., McInnes, A. G., and Walter, J. A. (1990). Identification of isodomoic acid D and two new geometrical isomers of domoic acid in toxic mussels. Can. J. Chem. 68, 22–25. 31. Schantz, E. J., Ghazarossian, V. E., Schnoes, H. K., Strong, F. M., Springer, J. P., Pezzanite, J. O., and Clardy, J. (1975) The structure of saxitoxin. J. Am. Chem. Soc. 97, 1238–1239. 32. Tsuda, K., Ikuma, S., Kawamura, M., Tachikawa, R., and Sakai, K. (1964) Tetrodotoxin. VII. On the structure of tetrodotoxin and its derivatives. Chem. Pharm. Bull. 12, 1357–1374. 33. Goto, T., Kishi, Y., Takahashi, S., and Hirata, Y. (1965) Tetrodotoxin. Tetrahedron 21, 2059–2088. 34. Woodward, R. B. (1964) The structure of tetrodotoxin. Pure Appl. Chem. 9, 49–74. 35. Koskinen, A. M. P. and Rapaport, H. (1985) Synthetic and conformational studies on anatoxin-a: a potent acetylcholine agonist. J. Med. Chem. 28, 1301–1309. 36. Matsunaga, S., Moore, R. E., Niemczara, W. P., and Carmichael, W. W. (1989) Anatoxin-a(s), a potent anticholinesterase from Anabaena flos-aquae. J. Am. Chem. Soc. 111, 8021–8023. 37. Lin, Y., Risk, M., Ray, S. M., van Engen, D., Clardy, J., Golik, J., James, J. C., and Nakanishi, K. (1981) Isolation and structure of brevetoxin B from “red tide” dinoflagellate Ptychodiscus brevis (Gymnodinium breve). J. Am. Chem. Soc. 103, 6773–6776. 38. Murata, M., Legrand, A. M., Ishibashi, Y., and Yasumoto, T. (1989) Structures of ciguatoxin and its congener. J. Am. Chem. Soc. 111, 8929–8931.


253

39. Chang, F. H., Anderson, D. M., Kulis, D. M., and Till, D. G. (1997) Toxin production of Alexandrium minutum (Dinophyceae) from the Bay of Plenty, New Zealand. Toxicon 35, 393–409. 40. Oshima, Y. (1995) Postcolumn derivatization liquid chromatographic method for paralytic shellfish toxins. J. AOAC Int. 78, 528–532. 41. Oshima, Y., Blackburn, S. I., and Hallegraeff, G. M. (1993) Comparative study on paralytic shellfish toxin profiles of the dinoflagellate Gymnodinium catenatum from three different countries. Mar. Biol. 116, 471–476. 42. Betz, J. L., Brown, P. R., Smyth, M. J., and Clarke, P. H. (1974) Evolution in action. Nature 247, 261–264. 43. Satin, J., Limberis, J. T., Kyle, J. W., Rogart, R. B., and Fozzard, H. A. (1994) The saxitoxin/tetrodotoxin binding site on cloned rat brain IIa Na channels is in the transmembrane electric field. Biophys. J. 67, 1007–1014. 44. Lipkind, G. M and Fozzard, H. A. (1994) A structural model of the tetrodotoxin and saxitoxin binding site of the Na+ channel. Biophys. J. 66, 1–13. 45. Strichartz, G. (1984) Structural determinants of the affinity of saxitoxin for neuronal sodium channels. Electrophysiological studies on frog peripheral nerve. J. Gen. Physiol. 84, 281–305. 46. Atchison, W. D., Luke, V. S., Narahashi, T., and Vogel, S. M. (1986) Nerve membrane sodium channels as the target site of brevetoxins at neuromuscular junctions. Br. J. Pharmacol. 189, 731–738. 47. Lombet, A., Bidard, J. N., and Lazdunski, M. (1987) Ciguatoxin and brevetoxins share a common receptor site on the neuronal voltage-dependent Na+ channel. FEBS Lett. 219, 355–359. 48. Jeglitsch, G., Rein, K., Baden, D. G., and Adams, D. J. (1998) Brevetoxin–3 (PbTx–3) and its derivatives modulate single tetrodotoxin-sensitive sodium channels in rat sensory neurons. J. Pharmacol. Exp. Ther. 284, 516–525. 49. Huang, J. M., Wu, C. H., and Baden, D. G. (1984) Depolarizing action of a red-tide dinoflagellate brevetoxin on axonal membranes. J. Pharmacol. Exp. Ther. 229, 615–621. 50. Sheridan, R. E. and Adler, M. (1989) The actions of a red tide toxin from Ptychodiscus brevis on single sodium channels in mammalian neuroblastoma cells. FEBS Lett. 247, 448–452. 51. Biscoe, T. J., Evans, R. H., Headley, P. M., Martin, M., and Watkins, J. C. (1975) Domoic and quisqualic acids as potent amino acid excitants of frog and rat spinal neurones. Nature 255, 166–167. 52. Michaelis, E. K. (1998) Molecular biology of glutamate receptors in the central nervous system and their role in excitotoxicity, oxidative stress and aging. Prog. Neurobiol. 54, 369–415. 53. Ozawa, S., Kamiya, H., and Tsuzunki, K. (1998) Glutamate receptors in the mammalian central nervous sytem. Prog. Neurobiol. 54, 581–618. 54. Adams, D. J., Dwyer, T. M., and Hille, B. (1980) The permeability of endplate channels to monovalent and divalent metal cations. J. Gen. Physiol. 75, 493–510. 55. Sargent, P. B. (1993) The diversity of neuronal nicotinic acetylcholine receptors. Annu. Rev. Neurosci. 16, 403–443. 56. Swanson, K. L., Allen, C. N., Aronstam, R. S., Rapoport, H., and Albuquerque, E. X. (1986) Molecular mechanisms of the potent and stereospecific nicotinic receptor agonist (+)-anatoxin-a. Mol. Pharmacol. 29, 250–257. 57. Hyde, E. G. and Carmichael, W. W. (1991) Anatoxin-a(s), a naturally occurring organophosphate, is an irreversible active site-directed inhibitor of acetylcholinesterase (EC 3.1.1.7). J. Biochem. Toxicol. 6, 195–201. 58. Dyer, J. R., Johnston, W. L., Castellucci, V. F., and Dunn, R. J. (1997) Cloning and tissue

254

59.

60.

61. 62.

63.

64.

65.

66.

67. 68.

69.

70.

71. 72. 73.

74.

75. 76.

Llewellyn distribution of the Aplysia Na+ channel alpha-subunit cDNA. DNA Cell. Biol. 16, 347–356. Spafford, J. D., Spencer, A. N., and Gallin, W. J. (1998) A putative voltage-gated sodium channel alpha subunit (PpSCN1) from the hydrozoan jellyfish, Polyorchis penicillatus: structural comparisons and evolutionary considerations. Biochem. Biophys. Res. Commun. 244, 772–780. Jeziorski, M. C., Greenberg, R. M., and Anderson, P. A. (1997) Cloning of a putative voltage-gated sodium channel from the turbellarian flatworm Bdelloura candida. Parasitology 115, 289–296. Guy, H. R. and Durell, S. R. (1995) Structural models of Na+, Ca2+, and K+ channels. Soc. Gen. Physiol. Ser. 50, 1–16. Heinemann, S. H., Terlau, H., Stuhmer, W., Imoto, K., and Numa, S. (1992) Calcium channel characteristics conferred on the sodium channel by single mutations. Nature 356, 441–443. Lee, R. Y., Lobel, L., Hengartner, M., Horvitz, H. R., and Avery, L. (1997) Mutations in the alpha1 subunit of an L-type voltage-activated Ca2+ channel cause myotonia in Caenorhabditis elegans. EMBO J. 16, 6066–6076. Le Paslier, M. C., Pierce, R. J., Merlin, F., Hirai, H., Wu, W., Williams, D. L., et al. (2000) Construction and characterization of a Schistosoma mansoni bacterial artificial chromosome library. Genomics 65, 87–94. Myler, P. J., Audleman, L., deVos, T., Hixson, G., Kiser, P., Lemley, C., et al. (1999) Leishmania major Friedlin chromosome 1 has an unusual distribution of protein-coding genes. Proc. Natl. Acad. Sci. USA 96, 2902–2906. Mongan, N. P., Baylis, H. A., Adcock, C., Smith, G. R., Sansom, M. S., and Sattelle, D. B. (1998) An extensive and diverse gene family of nicotinic acetylcholine receptor alpha subunits in Caenorhabditis elegans. Receptors Channels 6, 213–228. Maricq, A. V., Peckol, E., Driscoll, M., and Bargmann, C. I. (1995) Mechanosensory signalling in C. elegans mediated by the GLR–1 glutamate receptor Nature 378, 78–81. MacIntyre, J. G., Cullen, J. J., and Cembella, A. D. (1997) Vertical migration, nutrition and toxicity in the dinoflagellate Alexandrium tamarense Mar. Ecol. Prog, Ser. 148, 201–216. Ogata, T., Ishimura, T., and Kodama, M. (1987) Effect of water temperature and light intensity on growth rate and toxicity change in Protogonyaulax tamarensis. Mar. Biol. 95, 217–220. Selvin, R. C., Lewis, C. M., Yentsch, C. M., and Hurst, J. W. (1984) Seasonal persistence of resting cyst toxicity in the dinoflagellate Gonyaulax tamarensis var. excavata. Toxicon 22, 817–820. Oshima, Y., Bolch, C. J., and Hallegraeff, G. M. (1992) Toxin composition of resting cysts of Alexandrium tamarense (Dinophyceae). Toxicon 30, 1539–1544. Stevens, D. K. and Krieger, R. I. (1991) Stability studies on the cyanobacterial nicotinic alkaloid anatoxin-A. Toxicon 29, 167–179. Boczar, B. A., Beitler, M. K., Liston, J., Sullivan, J. J., and Cottolico, R. A. (1988) Paralytic shellfish toxins in Protogonyaulax tamarensis and Protogonyaulax catanella in axenic culture. Plant Physiol. 88, 1285–1290. Subba Rao, D. V., De Freitas, A. S. W., Quilliam, M. A., Pocklington, R., and Bates, S. S. (1990) Rates of production of domoic acid, a neurotoxin amino acid in the pennate marine diatom Nitzschia pungens, in Toxic Marine Phytoplankton (Graneli, E., Sundstrom, B., Edler, L., and Anderson, D. M. eds.), Elsevier, New York, NY, pp. 413–417. Fersht, A. (1985) Enzyme Structure and Mechanism. W.H. Freeman, New York. Herbert, R. A. (1999) Nitrogen cycling in coastal marine ecosystems FEMS Microbiol. Rev. 23, 563–590.


255

77. Rapala, J., Sivonen, K., Luukkainen, R., and Niemala, S. I. (1993) Anatoxin-a concentration in Anabaena and Aphanizomenon under different environmental conditions and comparison of growth by toxic and non-toxic Anabaena-strains: a laboratory study. J. Appl. Phycol. 5, 581–591. 78. Bumke-Vogt, C., Mailahn, W., Rotard, W., and Chorus, I. (1996) A highly sensitive analytical method for the neurotoxin anatoxin-a, using GC-ECD, and first application to laboratory cultures. Phycologia 35(Suppl. 6), 51–56. 79. Anderson, D. M. and Cheng, T. P. (1988) Intracellular localization of saxitoxins in the dinoflagellate Gonyaulax tamarensis. J. Phycol. 24, 17–22. 80. Hallegraeff, G. M. (1993) A review of harmful algal blooms and their apparent global increase. Phycologia 32, 79–99. 81. Teegarden, G. J. (1999) Copepod grazing selection and particle discrimination on the basis of PSP toxin content. Mar. Ecol. Prog. Ser. 181, 163–176. 82. Teegarden, G. J. and Cembella, A. D. (1996) Grazing of toxic dinoflagellates, Alexandrium spp., by adult copepods of coastal Maine: implications for the fate of paralytic shellfish toxins in marine food webs. J. Exp. Mar. Biol. Ecol. 196, 145–176. 83. Haney, J. F., Sasner, J. J., and Ikawa, M. (1995) Effects of products released by Aphanizomenon flos-aquae and purified saxitoxin on the movements of Daphnia carinata feeding appendages. Limnol. Oceanogr. 40, 263–272. 84. Ishida, H., Muramatsu, N., Kosuge, T., and Tsuji, K. (1996) Study of neurotoxic shellfish poisoning involving New Zealand shellfish, Crassostrea gigas, in Harmful and Toxic Algal Blooms (Yasumoto, T., Oshima, Y., and Fukuyo, Y., eds.). UNESCO, Paris, pp. 491–494. 85. Trudeau, L. E. and Castellucci, V. F. (1993) Excitatory amino acid neurotransmission at sensory-motor and interneuronal synapses of Aplysia californica. J. Neurophysiol. 70, 1221–1230. 86. Fieber, L. A. (1998) Characterization of Na+ and Ca2+ currents in bag cells of sexually immature Aplysia californica. J. Exp. Biol. 201, 745–754. 87. Twarog, B. M., Hidaka, T., and Yamaguchi, H. (1972) Resistance to tetrodotoxin and saxitoxin in nerves of bivalve molluscs. A possible correlation with paralytic shellfish poisoning. Toxicon 10, 273–278. 88. Kvitek, R. G. and Beitler, R. M. (1991) Relative insensitivity of butter clam neurons to saxitoxin: a pre-adaptation for sequestering paralytic shellfish poisoning toxins as a chemical defense. Mar. Ecol. Prog. Ser. 69, 47–54. 89. Satin, J., Kyle, J. W., Chen, M., Bell, P., Cribbs, L. L., Fozzard, H. A., and Rogart, R. B. (1992) A mutant of TTX-resistant cardiac sodium channels with TTX-sensitive properties. Science 256, 1202–1205. 90. Wohlgeschaffen, G. D., Mann, K. H., Subba Rao, D. V., and Pocklington, R. (1992) Dynamics of the phycotoxin domoic acid: accumulation and excretion in two commercially important bivalves. J. Appl. Phycol. 4, 297–310. 91. Sheumack, D. D., Howden, M. E., Spence, I., and Quinn, R. J. (1978) Maculotoxin: a neurotoxin from the venom glands of the octopus Hapalochlaena maculosa identified as tetrodotoxin. Science 199, 188–189. 92. Sheumack, D. D., Howden, M. E., and Spence, I. (1984) Occurrence of a tetrodotoxinlike compound in the eggs of the venomous blue-ringed octopus (Hapalochlaena maculosa). Toxicon 22, 811–812. 93. Flachsenberger, W. and Kerr, D. I. (1985) Lack of effect of tetrodotoxin and of an extract from the posterior salivary gland of the blue-ringed octopus following injection into the octopus and following application to its brachial nerve. Toxicon 23, 997–999. 94. Hwang, D. F., Arakawa, O., Saito, T., Noguchi, T., Simidu, U., Tsukamoto, K.,et al. (1989) Tetrodotoxin-producing bacteria from the blue-ringed octopus Octopus maculosus. Mar. Biol. 100, 327–332.

256

Llewellyn

95. Sommer, H. (1932) The occurrence of the paralytic shellfish poison in the common sand crab. Science 76, 574–575. 96. Noguchi, T., Konosu, S., and Hashimoto, Y. (1969) Identity of the crab toxin with saxitoxin. Toxicon 7, 325–326. 97. Noguchi, T., Uzu, A., Koyama, K., Maruyama, J., Nagashima, Y., and Hashimoto, K. (1983) Occurrence of tetrodotoxin as the major toxin in a Xanthid crab Atergatis floridus. Bull. Jpn. Soc. Sci. Fish. 49, 1887–1892. 98. Daigo, K., Noguchi, T., Miwa, A., Kawai, N., and Hashimoto, K. (1988) Resistance of nerves from certain toxic crabs to paralytic shellfish poison and tetrodotoxin. Toxicon 26, 485–490. 99. Yamamori, K., Yamaguchi, S., Maehara, E., and Matsui, T. (1992) Tolerance of shore crabs to tetrodotoxin and saxitoxin and antagonistic effect of their body fluid against the toxins. Bull. Jap. Soc. Sci. Fish. 58, 1157–1162. 100. Llewellyn, L. E. (1997) Haemolymph protein in xanthid crabs: its selective binding of saxitoxin and possible role in toxin bioaccumulation. Mar. Biol. 128, 599–606. 101. Noguchi, T., Hwang, D. F., Arakawa, O., Sugita, H., Deguchi, Y., Shida, Y., and Hashimoto, K. (1987) Vibrio alginolyticus, a tetrodotoxin-producing bacterium, in the intestines of the fish Fugu vermicularis vermicularis. Mar. Biol. 94, 625–630. 102. Yotsu, M., Yamazaki, T., Meguro, Y., Endo, A., Murata, M., Naoki, H., and Yasumoto, T. (1987) Production of tetrodotoxin and its derivatives by Pseudomonas sp. isolated from the skin of a pufferfish. Toxicon 25, 225–228. 103. Matsumura K. (1995) Tetrodotoxin as a pheromone. Nature 378, 563–564. 104. Kodama, M., Sato, S., Ogata, T., Suzuki, Y., Kaneko, T., and Aida, K. (1986) Tetrodotoxin secreting glands in the skin of puffer fishes. Toxicon 24, 819–829. 105. Fritz, L., Quilliam, M. A., and Wright, J. L. C. (1992) An outbreak of domoic acid poisoning attributed to the pennate diatom Pseudonitzschia australis. J. Phycol. 28, 439–442. 106. Scholin, C. A., Gulland, F., Doucette, G. J., Benson, S., Busman, M., Chavez, F. P., et al. (2000) Mortality of sea lions along the central California coast linked to a toxic diatom bloom. Nature 403, 80–84. 107. Geraci, J. R., Anderson, D. M. Timperi, R. J., St. Aubin, D. J., Early, G. A., Prescott, J. H., and Mayo, C. A. (1989) Humpback whales (Megaptera novaeangliae) fatally poisoned by dinoflagellate toxin. Can. J. Fish. Aquat. Sci. 46, 1895–1898. 108. Davis, C. G. (1947) Gymnodinium breve, n. Sp. A cause of discolored water and animal mortality in the Gulf of Mexico. Bot. Gaz. 109, 358–360. 109. Bossart, G. D., Baden, D. G., Ewing, R. Y., Roberts, B., and Wright, S. D. (1998) Brevetoxicosis in manatees (Trichechus manatus latirostris) from the 1996 epizootic: gross, histologic, and immunohistochemical features. Toxicol. Pathol. 26, 276–282. 110. Saimi, Y., Martinac, B., Delcour, A. H., Minorsky, P. V., Gustin, M. C., Culbertson, M. R., et al. (1992) Patch clamp studies of microbial ion channels. Methods Enzymol. 207, 681–691. 111. Oami, K., Naitoh, Y., and Sibaoka, T. (1995) Voltage-gated ion conductances correspoding to regenerative positive and negative spikes in the dinoflagellate Noctiluca miliaris. J. Comp. Physiol. A 176, 625–633. 112. Andrivon, C. (1988) Membrane control of ciliary movement in ciliates. Biol. Cell 63, 133–142. 113. Liu, J. Z., Dapice, M., and Khan, S. (1990) Ion selectivity of the Vibrio alginolyticus flagellar motor. J. Bacteriol. 172, 5236–5244. 114. Hirota, N. and Imae, Y. (1983) Na+-driven flagellar motors of an alkalophilic Bacillus strain YN–1. J. Biol. Chem. 258, 10,577–10,581.

Biosynthesis of Important Marine Toxins

257

14 Biosynthesis of Important Marine Toxins of Microorganism Origins Yuzuru Shimizu 1. INTRODUCTION There are a great number of toxins derived from microorganisms: bacteria, fungi, microalgae, and so on. They include a wide range of different biosynthetic types such as acetogenins, peptides, and terpenes, and their biosyntheses have been studied extensively and reviewed by many authors. Therefore, in this chapter, the author will focus only on the biosyntheses of pharmacologically and environmentally important marine toxins produced by marine microorganisms. Marine metabolites are unique in many respects. Many of them have very specific modes of action and serve as important probes for pharmacological studies. They are also recognized as culprits of seafood poisonings, and their biosynthetic origins are of particular interest and importance. Regarding their biosynthetic origins, they are very different from the terrestrial counterparts. More often than not, the organisms in which the toxins are found are not the primary producers of the toxins, and microorganisms involved in the food chain, association, or symbiosis are recognized as the real sources of the compounds. 1.1. Biosynthesis of Saxitoxin Derivatives The saxitoxin group toxins are known as paralytic shellfish poisons (PSP) and responsible for the major shellfish intoxication. Saxitoxin (STX) is the first toxin discovered and best known in this group, but such toxins as neosaxitoxin and gonyautoxins occur more abundantly in shellfish (1). The identities of the primary producers of the toxins have been established as several marine dinoflagellates, but some freshwater blue-green algae also produce toxins in the same group. The representative structures of natural STX derivatives are shown in Fig. 1. As soon as saxitoxin’s very unusual structure had been elucidated by X-ray crystallography, the biosynthesis of SXT became a subject of great interest. At a first glimpse, it seemed to be a purine metabolite condensed with a C3 unit, e.g., acrylic acid for the additional five-membered ring, and a C1 unit for C-13 extra carbon (2). Acrylic acid and its precursors are, in fact, abundant in microalgae. Subsequent studies, however, proved that the biosynthesis follows a completely different pathway. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

257

258

Shimizu

Fig. 1. Structures of representative paralytic shellfish toxins. They are divided into two major groups: saxitoxin (STX) and neosaxitoxin (neoSTX) groups (1).

The earlier isotope trace studies carried out using the US East coast red-tide organism Alexandrium tamarense (Gonyaulax tamarensis) demonstrated that two guanidium carbons atoms in the nucleus and the side-chain carbamate carbon come from arginine (2). The finding clearly ruled out the normal purine pathway for the skeleton. In that experiment, however, no incorporation of 14C was observed from [1- 14C]-arginine, suggesting that either arginine serves only as a guanidine donor in the citrulline-ornithine cycle or the carboxyl group is lost in a biosynthetic step. On the assumption that the latter is the case, the fate of arginine was further studied with a toxic strain of Aphanizomenon flos-aquae, which produces neosaxitoxin and saxitoxin (3). 2-13C-2-15N-Labeled ornithine, which is the immediate precursor of arginine, was synthesized and fed to the organism. The 13C-NMR analysis of the isolated neosaxitoxin showed an enriched C-4 signal as a doublet, split by coupling with 15N, suggesting that the 13C-15N connectivity in the arginine was incorporated intact into the toxin molecule. Again, in this experiment, the feeding of 1-13C-arginine did not result in the retention of 13C, indicating that the carboxyl carbon was lost in the process. Initially, serine was the most likely candidate for the rest of the carbon skeleton, C-5, C-6, and C-13. In fact, the feeding of 3-13C-serine effectively labeled C-13 of neosaxitoxin (4). However, as later experiments showed, this was extremely misleading, and the incorporation was actually a result of a C1 transfer from serine. Likewise, 2-13C-labeled glycine was also incorporated into C-13 of neosaxitoxin very effectively. The origin of the two carbons, C-5 and C-6, were eventually found to be acetate (4). The feeding of single- and double-13C-labeled acetate was incorporated intact into C-5 and C-6, and 1-13C-acetate and 2-13C-acetate were incorporated into C-5 and C-6, respectively (3). The acetate was also incorporated, although at much lesser rates, into the arginine-derived portion in the pattern expected from the biosynthetic pathway of arginine. The origin of the side-chain carbon, C-13, was established by a feeding experiment with methyl-13C-methionine (3,4). As mentioned earlier, other C1 sources such as


259

1-13C-glycine and 2-13C-serine were also incorporated into C-13, but methionine was the best precursor, suggesting methionine is the direct source of the C1 (Fig. 2). The starting step of the biosynthesis must be the Claisen-type condensation of acetyl Co-A on the α-carbon of arginine followed by decarboxylation (4,5). Claisen condensation on the amino-bearing carbon is not so common in biosynthetic pathways, but examples include such an important reaction as the formation of aminolevulinic acid, the first step of porphirin biosynthesis. The C1 introduction from methionine is normally the electrophilic addition of a methyl group from S-adenosylmethionine (SAM) on a double bond. The detail of the C-13 introduction was investigated using double isotopte labeling and stepwise feeding techniques (6). Feeding of methionine-methyl-d3-methyl-13C resulted in the retention of only one deuterium atom at C-13 of neosaxitoxin, whose 13C nuclear magnetic resonance (NMR) signal appeared as a triplet split by spin-coupling with one deuterium atom. Very interestingly, the feeding of [1,2-13C2]-acetate-d3 resulted in retention of one deuterium at C-5, the angular carbon derived from acetate carboxyl group, which originally did not bear an hydrogen atom. The origin of the hydrogen at C-6 was investigated by the stepwise feeding of 1, 13 2- C-actate and methyl-D3-methionine, and it was found that there was no migration of deuterium from the methyl-derived group to the adjacent C-6. Based on these experimental results, the pathway and mechanisms summarized in Fig. 3 were proposed. Several important steps remain unresolved. One of them is the introduction of 12-oxygen seen in all PSP toxins and 11-hydroxyl group seen in many toxins, i.e., gonyautoxins. It is an interesting question why and how the methylene carbon derived from arginine is oxidized so exclusively, because, to date, no toxin or intermediate metabolite lacking an oxygen function at C-12 has been discovered. Stimulated largely by the well-publicized bacterial origin theory of tetrodotoxin (TTX) (vide infra), several groups have proposed bacteria as the true producer of saxitoxin derivatives (7–9). However, genetic evidence presented by Ishida’s (9) and Oshima’s groups (11) suggests that the toxigenicity is proper to the algae. Still, minute amounts of the toxins could be produced by symbiotic bacteria, which received a gene transfer, but it may have only peripheral significance in the major toxin production. It was also pointed out that there are a number of compounds in the bacterial cells that give the same chromatographic behavior as the PSP toxins (12,13). 1.2. Biosynthesis of Brevetoxins Brevetoxins are the toxins of neurotoxic shellfish poisoning (NSP) produced by the dinoflagellate Gymndinium breve (14). The toxins have three basic skeletons: brevetoxin A (BTX-A), brevetoxin-B (BTX-B), and hemibrevetoxin-B (HBT-B) (Fig. 4). Their spectacular structures of neatly arranged polycyclic ethers have stimulated many scientists’ imagination about their biosynthesis. It was suggested that such all-syn-trans structures could be attained by the concomitant opening of stereochemically uniform all-trans-epoxides of long-chain polyenes (15,16) (Fig. 5). Such polyenes are usually polyketides formed from acetate. However, brevetoxins turned out to be not ordinary polyketides. Feeding experiments carried out with 1-13C-, 2-13C-, and 1,2-13C2-acetate revealed that acetate units were not incorporated in a pattern predicted by Birch’s polyketide’s

260

Shimizu

Fig. 2. Building blocks or precursors of STX derivatives established by feeding isotopelabeled putative precursors.

Fig. 3. Proposed biosynthetic pathway and mechanisms of paralytic shellfish toxins.

rule (17,18) (Fig. 6). Another surprising finding in these experiments was that methyl branches seemed to have two completely different origins: methionine and acetate methyl group. Nomally, methyl branches on polyketide side chains come from methionine or propionic acid inserted instead of acetate. Methylation with acetate are seen with a limited number of bacterial polyketides. This apparent dual mechanism for introduction of methyl groups on the same side chain was unprecedented and met with


261

Fig. 4. Structure types of brevetoxins, the toxic principles of Gymnodinium breve, the causative organism of neurotoxic shellfish poisoning (NSP).

considerable skepticism. In fact, it was later found out that another prominent marine toxin of dinoflagellate origins, okadaic acid, has all its methyl groups derived from acetate. In repeated feeding experiments (19), however, it was confirmed that four methyl groups of BTX-B are, indeed, of methionine origin, whereas three methyl and one terminal methylene are from acetate methyl group.The reported acetate incorporation pattern has been also confirmed in the repeated experiments, and an alternative explanation given to the anomalous incorporation of acetate. The irregular acetate incorporation pattern was very puzzling and unprecedented. In the 1987 report (18), it was suggested that the mixing of dicarboxylic acids, e.g., succinate and α-ketoglutarate, as building blocks in the chain formation could explain the anomalous pattern and carbon connectivity (Fig. 7). In the TCA cycle, one acetatederived carbon will be lost by decarboxylation of α-ketoglutarate on every turn of the cycle. Also, the insertion of another dicarboxylic acid, 3-hydroxyl-3-methylglutarate would explain the acetate derived methyl branches. In fact, the secondary 13C-13C connectivities seen in the NMR spectrum of BTX-B from 2-13C-labeled acetate feeding were in good agreement with those predicted from the TCA cycle. Lee et al. (20) also resorted to the same explanation.

262

Shimizu

Fig. 5. Proposed mechanism of brevetoxin biosynthesis: the concomitant opening of transepoxides can lead too all syn-trans cyclic structures (14,15).

The detour through the TCA cycle is, however, expected to result in the differential incorporation of acetate units into those irregular portions of carbon chain. We should also see randomization in the concerned carbon units, as the TCA cycle goes around more than one cycle. In the new experiments, however, it was observed that the incorporation rates of acetate into all acetate-derived carbons in the chain were rather uniform. The observation seems to rule out the possible involvement of the TCA cycle, unless the entire TCA cycle is a part of the reading frame in the polyketide gene cluster, which is highly unlikely. As an alternative explanation, a new mechanism, which involves carbon losses by decarboxylation of glycidic acid-type intermediates during the chain elongation, was proposed (19) (Fig. 8). The glycidic acid-type intermediate can be formed by epoxidation of a crotonic acid-type trans-unsaturated acid, a normal intermediate in fatty acid and polyketide biosynthesis. The new mechanism concurs the apparent intense involvement of epoxide formation in the entire biosynthesis of brevetoxins and other dinoflagellate polyketides. One eye-catching aspect of brevetoxin structures of all three major types—BTX-As, BTX-Bs, and HBTXs—is the presence of an aldehyde function at the end the molecules. It may now be interpreted as a remnant of the decarboxylation of glycidic acids. The anomalous polyketides chains are seen with other dinoflagellate metabolites. Of them, the biosynthesis of okadaic acid derivatives, important marine toxins produced by dinoflagellates, has been investigated. Norte et al. (21) reported that the headto-tail acetate sequence is faulted at two locations in the carbon chain. A Canadian


263

Fig. 6. Anomalous acetate incorporation pattern in brevetoxin B (BTX-B) biosynthesis as was confirmed by repetitive experiments. Of the seven methyl groups and one terminal methylene, four come from acetate methyl, and four from a nonacetate origin, methionine methyl group.

Fig. 7. Unconventional building blocks were proposed to explain the unusual acetate incorporation pattern observed with brevetoxin B (17). The acetate pathway in the TCA cycle can explain the odd methyl-methyl connectivity observed with brevetoxin B derived from the [1-13C]-acetate feeding. S: succinate or equivalent; HMG: hydroxymethylglutamate; P: propionate or equivalent A; acetate

group made the same observation and determined glycolic acid as the starter unit (22) (Fig. 9). Explaining the carbon-chain shortening, the Canadian group suggested a mechanism involving Favorsky-type rearrangement, in which a cyclopropane ring intermediate formed from an α-diketone undergoes oxidative opening followed by decarboxylation (23). In fact, examples of Favorsky rearrangement (or benzil-benzilic acid rearrangement) are seen in the biosyntheses of some other compounds. A difficulty with this hypothesis, however, may be the absence of carboxyl branches in any of the known dinoflagellate polyketides. The Canadian group proved that the okadaic

264

Shimizu

Fig. 8. An alternate explanation for the carbon shortening observed with brevetoxins and other dinoflagellate polyketides. Insertion of an epoxidation step in the polyketide synthetase can cause the carbon loss or chain shortening (18).

Fig. 9. Acetate incorporation pattern of okadaic acid biosynthesis. The starter unit is glycolic acid (19,20).

acid ester derivatives found in the dinoflagellates are formed from a single carbon chain by Baeyer-Villiger reaction (23). This remarkable finding seems to be another indication of peroxidation involvement in dinoflagellate polyketides. 1.3. Biosynthesis of Tetrodotoxin TTX is probably the most prominent of all marine toxins. It was first isolated from puffers, but has been found in many marine and terrestial animals in different phyla. This occurrence of the unique molecule in a wide variety of organisms has made most reserchers suspect the presence of a common vector as the toxin source and accumulation mechanisms in the animals through the food chain or symbiosis. In fact, puffer fish raised in captivity are known to lack TTX or have greatly diminished toxicity. Japanese groups have reported the production of TTX by many strains of bacteria (24). Thus, the bacterial origin theory of TTX is now widely accepted. This author, however, feels that the issue is far from settled. First, the reported bacterial productivity of TTX by bacteria seems to be too minute to account for large amounts of TTX found in the animals. All the reports were dependent on high-performance liquid chromatography (HPLC) and mass spectroscopic identification of TTX or its derivative, and there was no report of isolation of a substantial amount of TTX from culture enough to carry out the conventional identification such


265

as NMR measurement. Our efforts to isolate a tangible amount of TTX from a large bacterial culture failed (unpublished result). Danger of total dependence on the ultramicroanalytical identification is well-demonstrated in a case of a neosaxitoxin imposter in the bacterial culture (12). In fact, there is a report that an immunological test using a monoclonal TTX antibody, which should be the most selective and sensitive, failed to detect TTX in the bacterial culture (25). In the absence of the knowledge of the real or good producer, its biosynthetic study is in jumble. The bacterial culture fed with 14C-acetate as a universal carbon source failed to label TTX, which is added to the culture in a dilution experiment. After thorough purification, no radioactivity was observed with TTX, which may cast another doubt on the bacterial theory (Shimizu et al., unpublished results). Injection of 14C precursors such as acetate and ureido-labeled arginine into the newt, Taricha granulosa, labeled all common metabolites such as cholesterol and amino acids effectively, but not TTX (26). The result indicates that the animal does not produce TTX, at least in captivity. The only clues about the possible biosynthetic pathway of TTX come from the chemical structures of various congeners found along with TTX. Yamashita and Yasumoto isolated deoxyTTX from a puffer and a newt (27,28). The author’s group carefully examined the guanidinium compounds in Taricha granulosa, and isolated 1N-hydroxyl-5, 11-deoxyTTX, which is the most ring-deoxygenated TTX derivatives (29) (Fig. 10). There was a long-time speculation that the highly oxygenated skeleton of TTX might be a sugar derivative. As a matter of fact, condensation of arginine and a branched sugar like apiose, which is known to exist in marine environments, could conveniently explain the highly oxygenated skeleton. The discovery of the deoxyTTX derivatives, however, seems to suggest the otherwise that TTX is formed from an aklyl precursor, e.g., an isoprenoid, by progressive oxygenation (Fig. 11), since the reverse pathway, i.e., the stepwise deoxygenation is rather difficult and rare in nonaromatic metabolites. At any rate, the final proof for the biosynthetic pathway of TTX has to wait for finding an appropriate organism or system for successful feeding. 1.4. Biosynthesis of Amnesic Shellfish Poison: Domoic Acid In 1987, a Canadian group discovered that domoic acid, a known anthelmintic amino acid causes a serious shellfish poisoning, amnesic shellfish poisonin (NSP) (30). The source of domoic acid found in the mussles in the first incident was traced to the diatom, Pseudonitzschia sp. The toxin was later found in several other related species. Domoic acid is, parmacologically and chemically, a close relative of kainic acid, a well-known anthelmintic glutamate agonist. It is found in the macro red alga Chondria armata (31). Its occurrence in the distantly related species raised a question about its real producer. As in the case of TTX, the associated bacteria became a primary suspect. In fact, culture with diminished number of bacteria was found to produce a less amount of domoic acid. However, the axenic culture also produced the toxin, though in a much less quantity (32). Therefore, the role of bacteria in domoic acid production is likely that of a metabolic helper and/or a promoter, which is not unusual. The biosynthetic precursors of domoic acid are easily visualized as glutamate and prenyl group: one unit for kainic acid and two for domoic acid. Thus, Wright et al. (33) demonstrated that 13C-acetate is incorporated into the glutamate and prenyl portion of the toxin molecule in the expected from the TCA cycle (32) (Fig. 12). In the experi-

266

Shimizu

Fig. 10. Structures of tetrodotoxin (TTX) and its deoxygenated congeners discovered in puffer fish and newts.

Fig. 11. Plausible biosynthetic pathways of tetrodotoxin (TTX) (B). The branched sugar pathway (A) seems less likely.

ment, however, the acetate was not incorporated well into the isoprenoid portion of the molecule. It was interpreted that the TCA cycle and prenyl biosynthesis rely on different acetate pools, but it also leaves the possibility that the organisms is utilizing the nonacetic acid pathway of isoprenoid biosynthesis.


267

Fig. 12. Biosynthetic building blocks of domoic acid, the amnesic shellfish poisoning toxin (ASP), as supported by a labeled acetate feeding experiment (32).

REFERENCES 1. Shimizu, Y. (2000) Paralytic shellfish poisoning (PSP), chemistry and mechanism of action, in Seafood and Freshwater Toxins: Pharmacology, Physiology and Detection (Botana, L. M., ed.), Marcel Dekker, New York, NY, pp. 151–172. 2. Shimizu, Y., Kobayashi, M., Genenah, A., and Ichihara, N. (1984) Biosynthesis of paralytic shellfish poisoning, in Seafood Toxins (Ragelis, E. P., ed.), American Chemical Society, Washington, DC, pp. 151–160. 3. Shimizu, Y., Norte, M., Hori, A., Genenah, A., and Kobayashi, M. (1984) Biosynthesis of saxitoxin analogues: the unexpected pathway. J. Am. Chem. Soc. 106, 6433–6434. 4. Shimizu, Y. (1986) Toxigenesis and biosynthesis of saxitoxin analogues. Pure Appl. Chem. 58, 257–262. 5. Shimizu, Y., Gupta, S., and Prasad, A. V. K. (1989) Biosynthesis of dinoflagellate toxins, in Toxic Marine Phytoplankton (Granéli, E., Sudström, S., Edler, L., and Anderson, D. M., eds.), Elservier, New York, NY, pp. 62–73. 6. Gupta, S., Norte, M., and Shimizu, Y. (1989) Biosynthesis of saxitoxin anlogues: the origin and intruduction mechanism of the side-chain carbon. J. Chem. Soc. Chem. Commun. 1421–1424. 7. Kodama, M., Ogata, T., and Sato, S. (1988) Bacterial production of saxitoxin. Agric. Biol. Chem. 52, 1075–1097. 8. Doucette, G. J. and Trick, C. G. (1995) Characterization of bacteria associated with different isolates of Alexandrium tamarense, in Harmful Algal Blooms (Lassus, P., Arzul, G., Erard, E., Gentien, P., and Marcaillou, C., eds.), Lavoisier, Paris, pp. 33–38. 9. Gallacher, S., Flynn, K. J., Franco, J. M., Brueggemann, E. E., and Hines, H. B. (1997) Evidence for production of paralytic shellfish toxins by bacteria associated with Alexandrium spp. (Dinophyta) in culture. Appl. Environ. Bicrobiol. 63, 239–245. 10. Ishida, Y., Kim, C. H., Sako, Y., Hirooka, N., and Uchida, A. (1993) PSP production is chromosome dependent in Alexandrium spp., in Toxic Phytoplankton Blooms in the Sea (Smayda, T. J. and Shimizu, Y., eds.), Elsevier, New York, NY, pp. 881–887. 11. Oshima, Y., Itakura, H., Lee, K. C., Yasumoto, T., Blackburn, S., and Hallegraeff, G. (1993) Toxin production by the dinoflagellate Gymnodinium catenatum, in Toxic Phytoplankton Blooms in the Sea (Smayda, T. J. and Shimizu, Y., eds.), Elsevier, New York, NY, pp. 907–912. 12. Sato, S. and Shimizu, Y. (1998) Purification of a fluoresccent product from the bacterium, Moraxella: a neosaxitoxin impostor, in Harmful Algae (Reguera, B., Blanco, J., Fernandez, M. L., and Wyatt, T., eds.), UNESCO, Paris, pp. 465–467.

268

Shimizu

13. Onodera, H., Oshima, Y., Watanabe, M. F., Watanabe, M., Bolch, C. J., Blackburn, S., and Yasumoto, Y. (1996) Screening of paralytic shellfish toxins in freshwater cyanobacteria and chemical confirmation of the toxins in cultured Anabaena circinalis from Australia, in Harmful and Toxic Algal Blooms (Yasumoto, Y., Oshima, Y., and Fukuyo, Y., eds.), UNESCO, Paris, pp. 563–566. 14. Baden, D. G. (1983) Marine food-borne dinoflagellate toxins. Int. Rev. Cytol. 82, 99–150. 15. Nakanishi, K. (1985) The chemistry of brevetoxins: a review. Toxicon 23, 473–479. 16. Shimizu, Y. (1986) Biosynthesis and biotransformation of marine invertebrate toxins, in Natural Toxins (Harris, J. B., ed.), Clarendon Press, Oxford, pp. 115–125. 17. Lee, M. S., Repeta, D. S., Nakanishi, K., and Zagorski, M. G. (1986) Biosynthetic origins and assignments of carbon-13 NMR peaks of brevetoxin B. J. Am. Chem. Soc. 108, 7855–7856. 18. Chou, H. N. and Shimizu, Y. (1987) Biosynthesis of brevetoxins. Evidence for the mixed origin of the backbone carbon chain and the possible involvement of dicarboxylic acids. J. Am. Chem. Soc. 109, 2184–2185. 19. Shimizu, Y., Sano, T., and Shen, J. (2000) Mystery of polyether toxin biosynthesis in dinoflagellates. Proceeding of IUPAC Symposium on Mycotoxins and Phycotoxins, Brazil 2000, in press. 20. Lee, M. S., Quin, G. W., Nakanishi, K., and Zagorski, M. G. (1989) Biosynthetic studies of brevetoxins, potent neurotoxins produced by the dinoflagellate Gymnodinium breve. J. Am. Chem. Soc. 11, 6234–6241. 21. Norte, M., Padilla, A., and Fernadez, J. J. (1994) Studies on the biosynthesis of the polyether marine toxin dinophysistoxin-1 (DTX-1). Tetrahedron Lett. 35, 1441–1444. 22. Needham, J., McLachlan, J. L., Walter, J. A., and Wright, J. L. C. (1994) Biosynthetic origin of C-37 and C-38 in the polyether toxins, okadaic acid and dinohysistoxin–1. J. Chem. Soc. Chem. Commun. 2599–2600. 23. Wright, J. L. C., Hu, T., McLachlan, J. L., Needham, J., and Walter, J. A. (1996) Biosynthesis of DTX-4: confirmation of a polyketide pathway, proof of a Baeyer-villiger oxidation step, and evidence for an unusual carbon deletion process. J. Am. Chem. Soc. 118, 8757–8758. 24. Simidu, U., Kita-Tsukamoto, K., Yasumoto, T., and Yotsu, M. (1990) Taxonomy of four marine bacterial strains that produce tetrodotoxin. Int. J. Syst. Bacteriol. 40, 331–336. 25. Matsumura, K. (1995) Reexamination of tetrodotoxin production by bacteria. Appl. Environ. Microbiol. 61, 3468–3470. 26. Shimizu, Y. and Kobayashi, M. (1983) Apparent lack of tetrodotoxin biosynthesis in cultured Taricha torosa and Taricha granulosa. Chem. Pharm Bull. 31, 3625–3631. 27. Yasumoto, T., Yotsu, M., Murata, M., and Naoki, H. (1988) New tetrodotoxin analogues from the newt Cynops ensicauda. J. Am. Chem. Soc. 110, 2344–2345. 28. Yotsu-Yamashita, M., Yamagishi, Y., and Yasumoto, T. (1995). 5,6,11-Trihydroxytetrodotoxin from the puffer fish, Fugu poecilonotus. Tetrahedron Lett. 51, 9329–9332. 29. Kotaki, Y. and Shimizu, Y. (1993) 1-Hydroxy-5,11-dideoxytetrodotoxin, the first N-hydroxy and ring-deoxy derivative of tetrodotoxin found in the newt Tarich granuylosa. J. Am. Chem. Soc. 115, 827–830. 30. Wright, J. L. C., Boyd, R. K., de Freitas, A. S. W., Falk, M., Foxall, R. A., Jamieson, W. D., et al. (1989) Identification of domoic acid, a neuroexictatory amino acid in toxic mussels from eastern Prince Edward Island. Can. J. Chem. 67, 481–490. 31. Daigo, K. (1959) Constituents of Chondria armata. II. Isolation of an anthelmintic constituent. Yakugaku Zasshi 79, 353–356. 32. Douglas, D. J., Bates, S. S., Bourque, L. A., and Selvin, R. (1993) Domoic acid production by axenic and non-axenic cultures of the pennate diatom Nitzshia pungens f. multiseries, in Toxic Phytoplankton Blooms in the Sea (Smayda, T. J. and Shimizu, Y., eds.), Elsevier, Amsterdam, pp. 595–600. 33. Douglas, D. J., Ramsey, U. P., Walter, J. A., and Wright, J. L. C. (1992) Biosynthesis of the neurotoxin domoic acid by the marine diatom Nitzschia pungens forma multiseries, determined with [13C]-labelled precursors and nuclear magnetic resonance. J. Chem. Soc. Chem. Commun. 714–716.

Marine “Shellfish” Toxins

269

15 Biological Assay and Detection Methods for Marine “Shellfish” Toxins Neale R. Towers and Ian Garthwaite

1. INTRODUCTION Marine organisms produce a wide range of bioactive compounds. There is an equally wide array of methods for their detection and quantification. The toxins range from low molecular-weight compounds produced by unicellular cyanobacteria, diatoms, and micro-algae to large polypeptide and protein toxins produced by higher organisms such as sea anemones. This review will focus on the low molecular-weight toxins, which are of particular interest because of their harmful effects in humans following consumption of contaminated shellfish or fish. Most are neurotoxins, and in many instances the common names given to the classes of toxin based on the observed symptoms (paralytic shellfish poisons [PSP], neurotoxic shellfish poisons [NSP], and amnesic shellfish poisons [ASP]) point to an interaction between the toxin and the nervous system. In other cases, toxins with in vitro cytotoxic or hemolytic activity can cause equally dramatic effects, as instanced by the discomfort that diarrhetic shellfish poisons (DSP) cause consumers and the economic losses that result from large fish kills. Although often referred to as “shellfish toxins,” these poisons are produced by freeliving micro-algae, upon which the shellfish and fish feed. They are more accurately described as phycotoxins. There are approx 20 species of dinoflagellates, and a smaller number of diatoms, that are currently known to produce phycotoxins. Most are freeliving organisms found in the water column, but some are benthic organisms that can form “lawns” on which shellfish and herbivorous reef fish may graze. Furthermore, some blue-green algae that are found in fresh or brackish waters produce toxins, including the PSP toxins that are also found in marine microalgae. In most circumstances, the presence of phycotoxin-producing algae in the waters is not a problem. Occasionally, however, toxic algae may “bloom,” becoming the predominant species available to filter-feeding shellfish. The latter then concentrate the toxins in their flesh or digestive gland (hepatopancreas) and act as a vector, transferring the toxins further up the food chain. During an algal bloom, toxin concentrations can rapidly increase within healthy shellfish to a level that is harmful to consumers. In order that such shellfish may be avoided, sensitive analytical techniques that are able to From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

269

270

Towers and Garthwaite

detect the toxins at levels well below the toxic threshold are required. These techniques should be able to deal with complicated sample matrices, and ideally should be able to detect all toxins within a given toxin group, and differentiate these from related nontoxic compounds and from members of different toxin groups. 1.1. Phycotoxins Shellfish toxins are nonproteinacious compounds of low molecular weight (Mr 250–1500 Dalton). They have very different chemical structures (Figs. 1 and 2), solubilities, and modes of action, and produce quite different toxic effects. The toxins may be polar or lipophilic and may, or may not, be thermally labile; some are sensitive to pH, oxygen, and light while others are highly stable. Five major classes of shellfish poisons have been identified. These occur worldwide, and are distinguished by symptoms of the intoxication and their solubility in organic solvents. 1.1.1. Neurotoxic Shellfish Poisons The condition known as NSP manifests with incoordination, paralysis, and convulsions, and is caused by the lipid-soluble brevetoxins (1) (10 currently identified) (Fig. 1). These polyether toxins are produced by the dinoflagellate Gymnodinium breve, but may subsequently undergo chemical modification by shellfish (2). Seaspray contaminated with the organism irritates the eyes and nasal passages, leading to coughing and asthma-like symptoms (3). G. breve blooms occur almost annually in the Gulf of Florida, where they are responsible for widespread fish kills, and have caused the death of manatees (4). No human deaths have been recorded from NSP, although the condition is severely debilitating (5). 1.1.2. Diarrhetic Shellfish Poisons DSP is caused by okadaic acid (Fig. 1) and analogs, including the dinophysis toxins (DTX-1, 2, and 3) (Fig. 2), which are also lipid-soluble. The major symptom is diarrhea, and the illness is self-limiting. Unlike bacterial diarrhea, symptoms usually begin within a few hours of eating contaminated shellfish. Okadaic acid and the DTX-toxins have been shown to cause injury to the intestinal mucosa, and are implicated in tumor promotion (6). The structurally unrelated pectenotoxins (Fig. 2) and yessotoxins (Fig. 1) are currently included in the DSP group for food safety-regulatory purposes, although these substances do not cause diarrhea and have different biological activities (7). 1.1.3. Paralytic Shellfish Poisons The saxitoxins (Fig. 1), a group of more than 20 water-soluble compounds of varying toxicity, are responsible for PSP, a serious, life-threatening syndrome. Symptoms are neurological and include numbness, tingling, and burning of the lips and skin; giddiness, ataxia, and fever. Severe poisoning may lead to general muscular incoordination, respiratory distress, and death from paralysis of the respiratory muscles (8). No antidote is available and treatment is symptomatic. It is therefore important to prevent intoxication by monitoring shellfish for the presence of these toxins. 1.1.4. Amnesic Shellfish Poisons

ASP is caused by another water-soluble toxin, domoic acid (Fig. 1). Symptoms include loss of balance, nausea, headache, disorientation, and vomiting. The condition


271

Fig. 1. Structures of marine biotoxins: brevetoxin B (NSP); okadaic acid (DSP); saxitoxin and neo-saxitoxin (PSP); tetrodotoxin; domoic acid (ASP); yessotoxin (DSP); ciguatoxin (CFP).

272


Fig. 2. Structures of marine biotoxins: dinophysistoxin-2 (DTX-2) (DSP); microcystin-LR; pectenotoxin-1 (PTX-1) (DSP); azaspiracid; gymnodimine; palytoxin.


273

results in the permanent loss of short-term memory. Action limits for this toxin were established following an outbreak of poisoning in Canada, when more than 150 people became ill after eating shellfish. Four people died and toxin levels over 900 ppm were detected in shellfish (9,10). ASP affects the whole food chain, and there have been reports of bird deaths (11) and more recently deaths in Californian sealions that had eaten anchovies contaminated with the diatom Pseudo-nitzschia pungens (12). 1.1.5. Ciguatera Fish Poisons

Ciguatera fish poisoning (CFP), as a result of phycotoxins produced by the benthic alga Gambierdiscus toxicus, is generally contracted from eating subtropical and tropical fin-fish, clams, and marine snails contaminated with the toxins. The lipophilic ciguatoxins (CTX) (Fig. 1) are highly potent, and produce symptoms similar to the NSP toxins, although the symptoms of CFP are generally more severe, and include vomiting and diarrhea. The condition can be fatal, but where it is not symptoms can persist for many months. CFP is the most frequently reported phycotoxin problem, with an estimated 10–50,000 cases/year. The high toxicity, and therefore extremely low detection limit required, poses a challenge to the analyst (13,14). 1.1.6. “Fast Acting” Toxins

More recently, other algal toxins, including azaspiracid (15) (Fig. 2) and the “fast acting toxins” (gymnodimine [16] [Fig. 2.], the spirolides [17], and pinnatoxins [18]), have been added to the list. Azaspiracid causes symptoms reminiscent of DSP in humans, although in animals the effects are distinctly different from DSP (19). There is additional concern that azaspiracid is a potent hepatotoxin. The fast-acting toxins are so called because they have an “all or nothing” action, causing death rapidly in mice injected intraperitoneally with more than a threshold quantity of toxin, but little or no effect at lower doses. Death is preceded by neurological symptoms including convulsions. There are no known cases of human intoxication by gymnodimine but the pinnatoxins are known to cause human health problems in China and Japan (18). Many countries have established shellfish screening programs, and regulatory maximum permitted levels (MPL) for human exposure have been set worldwide (20–23). For further information on the toxicity of the phycotoxins, see refs. 23–28. 2. BIOTOXIN ASSAYS The available assays fall largely into two categories. The first, known as “effectbased” assays detect a biological action. The second type, known as “structure-based” assays, employs parameters determined by chemical structure (chromatographic mobility, charge, ultraviolet [UV] absorbance, mass, shape). 2.1. Effect-Based Assays Effect-based assays include in vivo bioassays using invertebrates and mammals, the many cytotoxicity assays and many of the receptor-based binding assays. These assays will detect any bioactive compound, irrespective of chemical structure, that is able to elicit the biological response being monitored. Such assays have an inherently low specificity and high potential for interference from other bioactive compounds that may be present in the sample. With these assays, specificity is largely conferred by the

274


extraction solvent and cleanup steps employed in preparing the sample extracts for analysis. For example, several of the marine biotoxins exert their effects by binding to the voltage-gated sodium channel. The pufferfish poison, tetrodotoxin (Fig. 1), and the saxitoxin group (29) bind to site 1 on the sodium-channel complex, “locking” it shut, whereas the ciguatoxins (30) and brevetoxins (31) bind to site 5 opening the channel, exerting an opposing effect. Thus, an assay such as the neuroblastoma cytotoxicity bioassay (32), based on this effect is unsuitable for use with extracts made with “universal” solvents, such as methanol or ethanol, which may contain mixtures of these toxins, unless specific clean-up steps are employed to separate the two toxin groups. Quantitation in the effect-based assays is by comparison to response curves prepared using a “standard” or reference toxin, and the results are in “toxin equivalents,” i.e., the observed activity is equivalent to that produced by a particular amount of the standard toxin under the same conditions. The observed result is the summation of the activity of all the compounds in the extract that can elicit the effect. The low specificity and the “summing” of bioactivity can be considered an advantage when the assays are used in food safety monitoring programs because, even though they give a high rate of false-positives, they are most likely to detect the presence of novel toxins. However, the “toxicity” measured in these bioassays may bear little relationship to the risk of intoxication following oral ingestion of contaminated shellfish, where other factors such as toxin stability in acid and the rate at which the toxin is absorbed may influence the severity of toxic effects. 2.2. Structure-Based Assays 2.2.1. Chromatographic Analysis At the other end of the scale to the effect-based assays the instrumental analytical methods (high-performance liquid chromatography [HPLC], capillary electrophoresis, liquid chromatography-mass spectrometry [LC-MS]) have high specificities, and are able to quantitate those toxins for which reference standards are available. The increased availability of analytical standards has played a key role in the recent development of HPLC and LC-MS methods, although there is still a great need for further certified reference standards. Chromatographic procedures with UV or fluorescence detection are now available for the majority of the shellfish toxins (Table 1). These methods will not be considered in detail in this review. For a review of LC-MS, see ref. 33. Chromatographic methods have generally been developed and optimized for the detection of a specific target toxin, and, because many toxins have no or only weakly absorbing chromophores, often involve a derivatization step to introduce a fluorescent or UV absorbing chromophore to enable toxin detection or to improve detection limits. The very specificity of the instrumental analytical methods, coupled with the fact that the analysis may be preceded by quite extensive clean-up of the sample extract using techniques such as liquid:liquid partition or solid-phase extraction, can often lead to situations where other toxic compounds and analogs of the target toxin are either not detected or are discarded in the clean-up steps. This poses a risk of underestimating the toxicity of shellfish. However, when suitable detection systems are used, chromatographic methods may detect a series of closely-related toxins. For example, HPLC using a diode-array detector allows for the detection of materials with an absorption spectrum similar to that of


275

Table 1 Chromatographic Detection Procedures for Shellfish Toxins Toxin class

Detection method

ASP

UV 242 nm

PSP

Fluorescence LC-MS Fluorescence

Fluorescence

Derivatization required None SAX cleanup FMOC Precolumn oxidation

Post-column oxidation

CE-MS (ESI)

DSP

NSP CTX

Detection limit in shellfish(a) 400 ng/g 20–30 ng/g (15 pg/mL) 10 ng/g (1–2 pg/injection)

(5–30 pg/injection) (30 pg/injection)

UV 205–215 nm

None

(10 µg/mL)

Fluorescence

ADAM

10–20 ng/gb 100 ng/gc

LC-MS (APCI) UV 215 nm LC-MS (ESI) Fluorescence

LC-MS

None

8 ng/g (500 ng/injection) (10 pg/injection)

P-CTX-1 0.04 ng/gd C-CTX-2 0.2 ng/gd

References and comments (134,135) (136) (33) (137) SPE cleanupcorrelated with mouse at 40 µg/100 g (138) Limit in flesh not given (139) C-toxin fragmentation is a problem (140) Extensive sample cleanup required (140,141) Matrix problems:b Hepatopancreasc whole shellfish (33,142) (5) (143) Insufficient sensitivity for clinical relevance (144)

aDetection limit of instrumental method is given (pg/injection) where limit in shellfish following extraction is not given by author and unknown. bHepatopancreas only. cWhole shellfish. dFish flesh.

the known toxin, while LC-MS allows for the detection of compounds with masses derived from the parent toxin by the addition or removal of such common moieties as hydroxy, methyl, and carboxy groups. The absence of certified reference standards for most of the marine biotoxins makes it difficult to quantify the total toxicity of an algal or shellfish extract by chromatographic methods since quantitation of the “unknown” or “unreferenced” peaks can only be achieved by extrapolation from peak heights for toxins for which standards are available.

276


2.2.2. Immunoassays and Receptor-Ligand Binding Assays Intermediate between the extremes of effect-based and structure-based assays are the immunoassays and receptor-ligand binding assays. Immunoassays have greater specificity than in vivo bioassays, cytotoxicity assays, and most receptor-binding assays as the antibody will only bind specifically to the toxin against which it was raised and to other compounds that contain the same, or very similar, epitopes. (The epitope is the region to which the antibody binds; good binding requires a high level of similarity in space filling structure and electron charge distribution.) Antibody specificity provides both advantages and disadvantages with respect to other analytical methods. Antibody specificity means that it is quite possible to determine concentrations of one toxin in the presence of many others. For example, saxitoxin or brevetoxin may be determined in the presence of other sodium channel-binding toxins, which is not possible with cytotoxicity or receptor-binding assays. But antibodies rarely have such narrow specificity that they bind a single compound or such broad specificity that they bind equally well to all members of a toxin group. The problems arising from this will be discussed later (Subheading 2.7.). Receptor-ligand binding-assay specificity depends on the particular receptor and the assay format, but as the receptor was not “designed” specifically for the toxin, there is always a risk of interference from the natural ligand(s) or other toxins that can bind to the same receptor. Immunoassays and receptor-ligand binding assays are quantitated by comparison to a standard curve prepared using a reference toxin and reported in “toxin equivalents.”

2.3. In Vivo Bioassays 2.3.1. Mouse Bioassays The classical method for detecting shellfish toxins has been the mouse bioassay, and most of the known phycotoxins have been isolated using mouse bioassay-directed chromatography. The mouse bioassay procedures date back to incidents of PSP and NSP toxicity in the United States more than 63 years ago (34) and although the methods (35,36) have since undergone extensive standardization and validation, the basic procedure has changed little. Water-soluble PSP and ASP toxins are simply extracted in boiling acid (0.1 M HCl) and the neutralized extract injected intraperitoneally into mice, with quantitation of the toxin depending on careful observation of the time of death and the use of conversion tables relating mouse liveweight and death time to toxin content of the extract (36). The procedures for lipophilic toxins are more complex, and although a standardized procedure using diethyl ether as the extracting solvent has been developed for the NSP toxins (35), it is not suited to processing large numbers of samples. Hannah et al. (37) developed an extraction procedure using acetone followed by partitioning against dichloromethane that allowed for a more rapid separation of the aqueous and organic phases. This procedure is very effective in extracting lipid-soluble toxins from shellfish, but has the disadvantage of coextracting other toxins including the DSP toxins and a number of fast-acting toxins. It is often difficult, therefore, to interpret the bioassay result. The mouse assay has also been widely used for the detection of DSP toxins, although there is no universally accepted procedure, and there are wide variations in the protocols employed in different countries (38–41).


277

Thus, the mouse bioassay, although widely used, and in many cases the only accepted assay for regulatory purposes, suffers from a number of disadvantages. It has relatively low sensitivity compared to other assay methods (by up to five orders of magnitude), is very nonspecific, and is labor-intensive. It suffers from interferences from substances such as polyunsaturated fatty acids (42,43) and zinc (44) which are toxic, and mineral salts such as Na+, which decrease the toxicity of shellfish extracts (45). Furthermore, mouse sex, strain, and liveweight affect bioassay sensitivity, and quantitation is therefore dependent on careful, extensive calibration to establish the conversion factors relating death time to toxin concentration. Animal to animal variation, coupled with ethical and practical restrictions on the number of mice used per test, means that the assay precision is low. For a more detailed discussion on the use of the mouse assay, see ref. 46. 2.3.2. Other Mammalian Bioassays In vivo bioassays have been developed for the DSP and CTXs that are more directly related to their biological effects. A suckling mouse bioassay (47), which measures fluid accumulation in the intestine following DSP administration, detects only the diarrhetic members of the DSP group, as does the rat bioassay (48), which relies on assessment of the softness (water content) of feces passed by rats fed 10 g of DSPcontaining shellfish hepatopancreas. Neither assay detects the yessotoxins or pectenotoxins, which are included in the DSP group because they co-extract with okadaic acid and the dinophysis toxins. These assays are thus more specific than the mouse bioassay, but because they are less sensitive and are only semi-quantitative, they are not widely used. Bioassays for CTX involving oral administration of fish or fish liver to chickens, cats, or mongooses have been developed (49–51). These bioassays avoid the problems of incomplete or selective toxin extraction but require larger sample sizes. While they may be less sensitive than other detection methods, the fact that they involve oral dosing of suspect material means that they provide a better model for human health-risk assessment than assessments relying on intraperitoneal injection of shellfish extracts or purified toxins. 2.3.3. Invertebrate Bioassays Since poisoning by contaminated shellfish often occurs in third-world countries, there has been an interest in developing cheaper alternatives to the conventional assays. Bioassays using invertebrates have been developed specifically for marine biotoxins, and there is further potential to adapt some of the many bioassays developed for fungal and plant toxins to the detection of algal toxins. Bioassays using invertebrates include a blowfly larva (Calliphora vomitoria) bioassay for DSP toxins and for hemolytic ichthyotoxins (52), mosquito bioassays for ciguatoxin (53) and saxitoxin (and a wide range of other toxins) (54), Daphnia magna bioassays for DSP toxins (55), and a desert locust (Schistocerca gregaria) assay for PSP toxins (56). In general, the minimum toxin concentrations detected by these bioassays are higher than for receptor-ligand binding assays or immunoassays.

278


2.3.4. Ichthyotoxicity: Fish Kill Bioassays

Fish deaths are often a feature of algal-bloom events and bioassays for ichthyotoxicity have been widely used in identifying potentially toxic organisms and in guiding the isolation of the toxins responsible. The bioassays generally involve exposing small aquarium fish to water containing shellfish extract (57,58), or to algal cultures (59). The assay is at best semiquantitative. 2.4. Cell-Based Bioassays 2.4.1. Cytotoxicity Bioassays Cell-based bioassays in their simplest form depend on monitoring the effects of the toxin or toxin-containing extract on cell survival. Most are cytotoxicity bioassays in which cell death is the ultimate response, and many of these have been established empirically because the mode of action of the toxin is not fully understood. Examples include bioassays for DSP toxins using either primary cultures of hepatocytes (60) or various cultured cell lines including several intestinal epithelial cell lines (61), KB cells (62), fibroblasts (63), or Buffalo green monkey kidney cell lines (64). The effects of the toxin may be observed by microscopic examination for morphological changes (60,65), but this is less convenient than the assessment of cytotoxicity by measuring the metabolic conversion of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) to its corresponding formazan (61,62). This procedure can be readily adapted to a 96-well microplate format using an enzyme-linked immunosorbent assay (ELISA) plate reader. The sensitivity of these cytotoxicity assays can be high and is dependent on both toxin concentration and the duration of exposure to the toxin. Tubaro et al.(62) report a sensitivity as low as 50 ng okadaic acid/g of mussel hapatopancreas. 2.4.2. Neuroblastoma Sodium Channel Bioassays

The most widely used cytotoxicity assay is the neuroblastoma bioassay for sodium channel-binding toxins (32) in which a monolayer of mouse neuroblastoma cells (Neuro-2a; ATCC CCL131) is treated with a fixed concentration of the Na+ channel activator, veratridine, in the presence of oubain, an inhibitor of Na+/K+ ATPase. Together, these compounds increase Na+ influx into the cells, causing them to swell and eventually lyse. However, in the presence of Na+ channel-blocking toxins such as the PSP toxins or tetrodotoxin, the action of veratridine is inhibited and the cells remain morphologically normal and viable. The assay can also be modified for assay of ciguatoxins and the brevetoxins (66), which increase the veratradine/oubain-induced Na+ influx, thus increasing the rate of cell lysis. Cell death has been monitored visually (32), by vital-dye exclusion (67,68), and by the MTT technique (69,70) using a 96-well plate reader, and quantitation is by comparison with a response curve prepared with a reference toxin. These tissue culture-based procedures, while useful for research, require specialized technical staff and tissue-culture facilities, and are difficult to maintain in an analytical laboratory.


279

2.4.3. Hemolysis and Hemolysis Neutralization Assays

In many instances, the binding of toxins to membrane ion channels results in a concentration dependent hemolysis of red blood cells that can readily be measured spectrographically (57,58). However, since hemolysis of red cells is a common property of many bioactive compounds, hemolysis bioassays have a very low specificity. In the case of palytoxin (Fig. 2), Bignami (71) overcame this problem by introducing a hemolysis neutralization assay in which the observed hemolysis was shown to be specifically due to palytoxin by demonstrating that it was prevented by addition of excess palytoxin-binding antibody. The hemolysis neutralization assay has a lower limit of detection (1 pg/mL) than the ELISA based on the same monoclonal antibody (MAb) (1 ng/mL) (72). 2.5. Receptor-Ligand Binding Assays Many of the marine biotoxins exert their biological activity at the cellular level by binding to, and disrupting the function of, specific ion-conducting channels in nerve and muscle membranes. The ASP toxins disrupt binding of glutamate to the AMDA glutamine receptors of the brain; CTX and NSP toxins bind to site 5 on the Na+ channel; PSP toxins and tetrodotoxin bind to site 1 of the Na+ channel, while maitotoxin binds to receptors within the Ca++ channels (73). This binding to natural receptors, with binding affinities similar to that obtained with antibody-antigen binding, is the basis for receptor-ligand binding assays. These assays can have high sensitivities, with quantitation limits of the order of 1–4 ng toxin/mL of extract, but like other bioassays can be subject to nonspecific interference from co-extracting compounds (74). They are also liable to interference from other compounds that bind to the same receptor; e.g., kainic acid, glutamate, and gamma-aminobutyric acid (GABA) could all interfere with the determination of domoic acid. Radio-ligand displacement assays have been established using synaptosome preparations from frog brain for domoic acid (75), and mouse brain for the PSP toxins (29,76) and the NSP toxins (77). These assays depend on measuring the competition between the variable amount of toxin in the sample extract and a fixed amount of radio-labeled toxin for the binding receptor. The radio-labeled ligand can be the toxin itself, or any other bioactive compound that binds reversibly to the same receptor site with a similar binding affinity. Thus, 3H-kainic acid can be used in the receptor assay for domoic acid (75), 3H-tetrodotoxin in the binding assay for saxitoxin (78), and 3H-brevetoxin in the assay for ciguatoxins (75). These assays have been formatted for screening relatively large numbers of samples using 96-well plate formats (75,79). A further development intermediate to receptor-ligand binding assays and the immunoassays has been the use of saxiphilin, a saxitoxin-binding protein found in some invertebrate species, as an alternative to the Na+ channel receptor in a radio-ligand binding assay for the PSP toxins (80). Saxiphilin has the advantage of binding most members of the PSP group but not binding tetrodotoxin and thus being less subject to this potential interference. Recent assay developments have focussed on the use of reconstituted Na+ ion-channel preparations (81) (rather than crude membrane preparations), the development of

280


recombinant cell-based assays (82), and the use of an in vitro rat hippocampal-slice preparation for the detection and quantitation of saxitoxin, brevetoxin, and domoic acid (83). While the latter directly records the effects of the toxin at a potential site of action, it requires skills and equipment not found in most analytical laboratories. The assays have been used for the analysis of algal and shellfish extracts, for monitoring toxic blooms, and for confirming the cause of intoxications in man (5,84) and wildlife (12). A major drawback of the radio-ligand binding assays is the limited availability of radio-labeled standard toxins and the often high regulatory (and personal) barriers to the use and disposal of radio-isotopes. Furthermore, international agreements on the distribution of materials considered to have potential as chemical weapons has severely curtailed the availability of saxitoxin and 3H-saxitoxin (85). 2.6. Enzyme Inhibition Assays The only established enzyme inhibition assays for algal biotoxins are those for the DSP toxins (okadaic acid and the dinophysistoxins), and for the microcystins (Fig. 2) and nodularin toxins These substances inhibit protein phosphatase type 1 (PP1) and type 2A (PP2A) enzymes. Originally formatted using a 32P-labeled substrate (86), the assay has been developed to employ chromogenic (87) or fluorescent (88,89) substrates and kinetic or end-point determination. The assay has also been developed as a radioligand binding displacement assay in which 125I-labeled microcystin competes with toxins that bind to the catalytic site of PP2A. In this format, the detection limit was below 50 pM for nodularin and microcystin-LR and below 200 pM for okadaic acid (90). 2.7. Immunoassays Immunoassays for low molecular-weight toxins are, in general, a form of competitive binding assay similar in concept to the receptor-binding assays, except that the “receptor” is an antibody raised specifically against the toxin. The low molecular mass marine biotoxins are not immunogenic in their own right and do not elicit an immune response unless they are conjugated to an immunogenic macromolecule, normally a protein or polypeptide. This places additional barriers to the development of an immunoassay for a toxin, for not only must a supply of the toxin be available but it must also be amenable to chemical conjugation, often via a bifunctional “linker molecule,” to a carrier protein such as bovine serum albumin (BSA), ovalbumin, or keyhole limpet hemocyanin. Polyclonal antibodies (antisera) are raised in animals (rabbits, sheep, goats) given a series of injections of the antigen, together with adjuvants that promote an immune response, over a period of several months. The production of antibodies in large animals, such as sheep or goats, is relatively simple and can yield large volumes of antisera, sufficient for many millions of assays, although there can be some variation in antibody titer and affinity for the toxins in polyclonal antibodies prepared from individual blood collections. Monoclonal antibodies, produced in cultures of hybridomas formed by fusing splenocytes from immunized animals (normally mice) with cultured myeloma cells avoid this variability, but are more difficult to produce in quantity. However, MAb production provides a greater opportunity to select antibodies that have the desired toxin binding specificity.


281

Although immunoassays were first developed as radio immunoassays in which radioisotope-labeled toxin competes with the toxin in a sample extract for the binding sites on the antibody, the most common procedure used now is the ELISA, in which a toxinprotein conjugate competes with the toxin the sample for antibody-binding sites. In ELISA the detecting label is an enzyme, normally horseradish peroxidase, and binding of the labeled toxin is determined by measuring bound enzymic activity with chromogenic or fluorescent substrates. This provides an amplification system, which increases the assay sensitivity. ELISAs are relatively cheap and quick, and are therefore suited to handling large numbers of samples. They do not require sophisticated and expensive facilities, and can be easily automated. They also have the potential to be further developed for accurate quantitation of toxin concentration. ELISA are normally one to two orders of magnitude more sensitive than HPLC detection methods. Although antibodies binding many of the better-known marine and freshwater algal biotoxins have now been raised and used in ELISA, this is not the case for most of the more recently discovered toxins, such as azaspiracid, pectenotoxin, the spirolides, and gymnodimine. Antibodies binding saxitoxin, the major PSP toxin, were reported in 1964 (91) and since then several groups have developed ELISA based on antisaxitoxin antibodies (92–94). Despite very low limits of detection and quantitation for saxitoxin, all these assays have a major disadvantage when used to detect potentially toxic shellfish or algal cultures. The antisaxitoxin antibodies have a low crossreactivity to other common toxic members of the PSP group, particularly to neo-saxitoxin (95), and the assay can markedly underestimate the total PSP content in samples where neo-saxitoxin is a major component (96). Similarly antibodies raised against neo-saxitoxin have a low cross-reactivity to saxitoxin (97). Chu et al. (98) have suggested that this problem could be overcome by summing the results of ELISA using antisaxitoxin and antineosaxitoxin antibodies, and these authors demonstrated a good correlation between the summation and toxicity estimated using the mouse bioassay. Antibodies binding the ASP toxin, domoic acid, have been developed by several groups for use in ELISA (99–103). In this case, antibody cross-reactivity is not a problem as domoic acid is the predominant toxin. Smith and Kitts (101) reported a high correlation between the results obtained by ELISA and by HPLC when analyzing shellfish samples, although ELISA results were approx 9% higher than those from HPLC analyses, probably because the ELISA detected the presence of domoic acid isomers not detected by HPLC. More recently, antibodies against novel domoic acid haptens, designed to eliminate the possibility of cross-reactivity to analogous compounds such as kainic acid, have been used in ELISA (103). The ELISA has a limit of quantitation of 0.15 ng/mL, which is approx 500-fold lower than the regulatory limit. The high sensitivity of the assay has allowed the reclassification of several Pseudo-nitzschia species from “nontoxic” to low-level toxin producers (104). Antibodies and ELISA for members of the brevetoxin B group (105–109); okadaic acid and DTX-1 from the DSP group (110–112); yessotoxin (113); ciguatoxin (114); palytoxin (72); microcystins and nodularins (115,116); have all been reported and a number of commercially produced test kits are now available. However, while most of these “first-generation” ELISA have been shown to provide sensitive assays yielding

282


concentration estimates that correlate well with those from other quantitative methods, such as HPLC, many have shortcomings when used to estimate the total toxicity of shellfish samples (117–119). The ELISA (and HPLC) often underestimate the total toxicity when compared to mouse bioassay results. In the case of ELISA, the underestimation generally arises from the limited affinity of the antibody for some members of the toxin group, while HPLC analysis generally measures only the major known toxin analogues. This problem is being addressed by designing better antigens in which the toxin is conjugated to the carrier protein so as to present the most common structural features to the immune system (by conjugating the “linker” to functional groups showing greatest variation between congeners) or by using as the antigen a synthesized fragment of the toxin structure, which is common to all toxins in the group (120,121). Antibodies produced in this way are expected to have a broader cross-reactivity spectrum and thus to provide a better estimate of the total toxin content. Despite these problems, the greater simplicity of the ELISA, and the potential to develop commercial ELISA test kits, and simple “dipstick” test kits (114,122) or biosensors (123–125) for use “dockside,” provides a greater impetus for continued development of immunoassays compared to the receptor-based assays. 2.8. Combined Bioassay and Chromatographic Techniques Combining bioassays with chromatographic analysis can greatly extend the utility of both techniques. Most of the known toxins were isolated using bioassay-directed chromatography, in which bioassays were used to identify the chromatographic fractions containing the toxic activity. Recently, two new techniques have become increasingly important. 2.8.1. Immunoaffinity Chromatography

A growing use for toxin-binding antibodies is their incorporation into immunoaffinity chromatography (IAC) columns used to provide a simple “one-step” clean-up system for concentrating the toxin, and/or its metabolites, from water samples (126), algal extracts (127), or tissue extracts (128). The toxin-binding antibodies are bonded to chromatographic matrices and placed in a small chromatographic column through which the water sample or tissue extract is passed. The column is then washed to remove contaminants before the bound toxins are released and eluted from the column, normally with buffer containing a high methanol content. The eluent can then be analyzed by any of the analytical methods available (ELISA, HPLC, LC-MS). The removal of the majority of compounds that interfere with the determination of UV absorbance, or with the derivatization of the target toxin(s) to introduce fluorescent chromophores, leads to cleaner HPLC and LC-MS chromatograms and lower detection limits. Development of IAC methods has been limited by the availability of suitable antibodies, but methods for PSP toxins (129), DSP (127,128,130), and microcystin/ nodularin (126) have been reported. 2.8.2. Chromatogram Peak Identification

The very high sensitivities of most immunoassays and receptor-binding assays means that they provide a powerful tool for the identification of “unknown” peaks on HPLC chromatograms, even when these are so small as to be hardly above background


283

and/or are not suitable for providing absorbance spectra. If HPLC eluent fractions containing the unknown peak are analyzed by immunoassay or receptor-binding assay or the PP-2A enzyme-inhibition assay, much can be inferred with regard to the nature of the unknown compounds. Positive results in the ELISA or radioimmunoassay indicate the presence of a structure with a similar epitope as the known toxin (5,115) while a positive result in a receptor-ligand binding assay (5) or the PP-2A inhibitor assay (131) confirms the presence of a substance that binds to the same site and that may have a similar biological activity. This provides a powerful tool in investigations aimed at detecting toxins and their metabolites. 2.9. Comparing Results from Different Assays In addition to problems in comparing the total “toxicity” of shellfish determined by mouse bioassay and ELISA or HPLC discussed in Subheading 2.7., major discrepancies may be seen between the results obtained from different bioassays and/or chromatographic procedures, especially when different solvents and extraction procedures are used prior to the analysis. These discrepancies may arise from differences in the extraction efficiencies of the solvents employed (132,133), losses during clean-up steps and differences in the relative sensitivities of the assays to the toxin mixtures present in the extracts. For example, Dickey et al. (133) found a 93-fold difference in the apparent brevetoxin-3 equivalent toxicity in shellfish, when measured using the mouse bioassay after diethyl ether extraction compared with the neuroblastoma Na+ channel cytotoxicity assay after extracting the shellfish with methanol. 3. SUMMARY Numerous in vitro and in vivo bioassays for the low molecular-weight algal toxins have been developed, enabling detection of the toxins in water, algal cells, contaminated fish and shellfish, and in the body tissues and fluids of animals and people intoxicated by contaminated shellfish or fish. The effects-based bioassays generally have a low specificity and are therefore subject to interference from other compounds with a similar biological action unless specific sample extraction and clean-up steps are employed. Structure-based assays, such as the immunoassays, are more specificity but may not respond equally to all members of a toxin family. In selecting an analytical method, the advantages and disadvantages of each method must be carefully assessed against the aims of the analytical program. REFERENCES 1. Baden, D. G. (1989) Brevetoxins: unique polyether dinoflagellate toxins. FASEB J. 3, 1807–1817. 2. Ishida, H., Nozawa, A., Totoribe, K., Muramatsu, N., Nukaya, H., Tsuji, K., et al. (1995) Brevetoxin B1, a new polyether marine toxin from the New Zealand shellfish, Austrovenus stutchburyi. Tetrahedron Lett. 36, 725–728. 3. Baden, D. G., Mende, T. J., Bikhazi, G., and Leung, I. (1982) Bronchoconstriction caused by Florida red tide toxins. Toxicon 20, 929–932. 4. Bossart, G. D., Baden, D. G., Ewing, R. Y., Roberts, B., and Wright, S. D. (1998) Brevetoxicosis in Manatees (Trichechus manatus latirostris) from the 1996 Epizootic: gross, histologic, and immunohistochemical features. Toxicol. Pathol. 26, 276–282.

284


5. Poli, M. A., Musser, S. M., Dickey, R. W., Eilers, P. P., and Hall, S. (2000). Neurotoxic shellfish poisoning and brevetoxin metabolites: a case study from Florida. Toxicon 38, 981–993. 6. Fujiki, H., Suganuma, M., Suguri, H., Yoshizawa, S., Takagi, K., Uda, N., et al. (1988) Diarrhetic shellfish toxin, dinophysistoxin-1, is a potent tumor promoter on mouse skin. Jpn J. Cancer Res. (Gann) 79, 1089–1093. 7. Aune, T., Stabell, O. B., Nordstoga, K., and Tjotta, K. (1998) Oral toxicity in mice of algal toxins from the diarrheic shellfish toxin (DST) complex and associated toxins. Nat. Toxins 7, 141–158. 8. Rodrigue, D. C., Etzel, R. A., Hall, S., de Porras, E., Velasquez, O. H., Tauxe, R. V., et al. (1990) Lethal paralytic shellfish poisoning in Guatemala. Am. J. Trop. Med. Hyg. 42, 267–271. 9. Wright, J. L. C., Boyd, R. K., de Freitas, A. S. W., Falk, M., Foxall, R. A., Jamieson, W., et al. (1989) Identification of domoic acid, a neuroexcitatory amino acid, in toxic mussels from eastern Prince Edward Island. Can. J. Chem. 67, 481–490. 10. Iverson, F. and Truelove, J. F. (1994) Toxicology and seafood toxins: domoic acid. Nat. Toxins 2, 334–339. 11. Work, T. M., Barr, B., Beale, A. M., Fritz, L., Quilliam, M. A., and Wright, J. L. C. (1993) Epidemiology of domoic acid poisoning in brown pelicans (Pelecanus occidentalis) and Brandt’s cormorants (Phalacrocorax penicillatus) in California. J. Zoo Wildlife Med. 24, 54–62. 12. Scholin, C. A., Gulland, F., Doucette, G. J., Benson, S., Busman, M., Chavez, F. P., et al. (2000) Mortality of sea lions along the central California coast linked to a toxic diatom bloom. Nature 403, 80–84. 13. Hokama, Y. (1993) Recent methods for detection of seafood toxins: recent immunological methods for ciguatoxin and related polyethers. Food Addit. Contam. 10, 71–82. 14. Park, D. L. (1994) Evolution of methods for assessing ciguatera toxins in fish, in Rev. Environ. Contam. Toxicol. 136, 1–21. 15. Satake, M., Ofuji, K., Naoki, H., James, K. J., Furey, A., McMahon, T., et al. (1998) Azaspiracid, a new marine toxin having unique spiro ring assemblies, isolated from Irish mussels, Mytilus edulis. J. Am. Chem. Soc. 120, 9967–9968. 16. Seki, T., Satake, M., Mackenzie, L., Kaspar, H. F., and Yasumoto, T. (1995) Gymnodimine, a new marine toxin of unprecedented structure isolated from New Zealand oysters and the dinoflagellate Gymnodinium sp. Tetrahedron Lett. 36, 7093–7096. 17. Hu, T., Curtis, J. M., Oshima, Y., Quilliam, M. A., Walter, J. A., Watson-Wright, W. M., and Wright, J. L. C. (1995) Spirolides B and D, two novel macrocycles isolated from the digestive glands of shellfish. J. Chem. Soc. Chem. Commun. 2159–2161. 18. Uemura, D., Chou, T., Haino, T., Nagatsu, A., Fukuzawa, S., Zheng, S.-Z., and Chen, H.-S. (1995) Pinnatoxin A: a toxic amphoteric macrocycle from the Okinawan bivalve Pinna muricata. J. Am. Chem. Soc. 117, 1155–1156. 19. Ito, E., Satake, M., Ofuji, K., Kurita, N., McMahon, T., James, K., and Yasumoto, T. (2000) Multiple organ damage caused by a new toxin azaspiracid, isolated from mussels produced in Ireland. Toxicon 38, 917–930. 20. van Egmond, H. P. (1993) Rationale for regulatory programmes for mycotoxins in human foods and animal feeds. Food Addit. Contam. 10, 29–36. 21. Park, D. L. (1995) Surveillance programmes for managing risks from naturally occurring toxicants. Food Addit. Contam. 12, 361–371. 22. Park, D. L., Guzman-Perez, S. E., and Lopez-Garcia, R. (1999) Aquatic biotoxins: design and implementation of seafood safety monitoring programs. Rev. Environ. Contam. Toxicol. 161, 157–200. 23. Quilliam, M. A. (1999) Phycotoxins. J. AOAC Int. 82, 773–781.


285

24. Morse, E. V. (1977) Paralytic shellfish poisoning: a review. J. Am. Vet. Med. Assoc. 171, 1178–1180. 25. Lagos, N. (1998) Microalgal blooms: a global issue with negative impact in Chile. Biol. Res. 31, 375–386. 26. Clark, R. F., Williams, S. R., Nordt, S. P., and Manoguerra, A. S. (1999) A review of selected seafood poisonings. Undersea Hyperbaric Med. 26, 175–184. 27. Van Dolah, F. M. (2000) Marine algal toxins: origins, health effects, and their increased occurrence. Environ. Health Perspec. 108 (Suppl. 1), 133–141. 28. Shimizu, Y. (1996) Microalgal metabolites: a new perspective. Annu. Rev. Microbiol. 50, 431–465. 29. Davio, S. R. and Fontelo, P. A. (1984) A competitive displacement assay to detect saxitoxin and tetrodotoxin. Anal. Biochem. 141, 199–204. 30. Lombet, A., Bidard, J. N., and Lazdunski, M. (1987) Ciguatoxin and brevetoxins share a common receptor site on the neuronal voltage-dependent Na+ channel. FEBS Lett. 219, 355–359. 31. Catterall, W. A. and Gainer, M. (1985) Interaction of brevetoxin A with a new receptor site on the sodium channel. Toxicon 23, 497–504. 32. Kogure, K., Tamplin, M. L., Simidu, U., and Colwell, R. R. (1988) A tissue culture assay for tetrodotoxin, saxitoxin and related toxins. Toxicon 26, 191–197. 33. Quilliam, M. A. (1998) Liquid chromatography-mass spectrometry: a universal method for the analysis of toxins? in Harmful Algae (Reguera, B., Blanco, J., Fernandez, M. L., and Wyatt, T., eds), Xunta de Galicia and IOC/UNESCO, Vigo, pp. 509–514. 34. Sommer, H. and Meyer, K. F. (1927) Paralytic shellfish poisoning. Arch. Pathol. 24, 560–598. 35. Irwin, N. (1970) Method for the bioassay of Gymnodinium breve toxin(s) in shellfish, in Recommended Procedures for the Examination of Sea Water and Shellfish, 4th ed., American Public Health Association Inc., New York, pp. 61–65. 36. AOAC (1990) Paralytic shellfish poison, in Official Methods of Analysis (Hellrich, K., ed.), Association of Official Analytical Chemists, Arlington, Virginia, pp. 881–882. 37. Hannah, D. J., Till, D. G., Deverall, T., Jones, P. D., and Fry, J. M. (1995) Extraction of lipid-soluble marine biotoxins. J. AOAC Int. 78, 480–483. 38. Yasumoto, T., Oshima, Y., and Yamaguchi, M. (1978) Occurrence of a new type of shellfish poisoning in the Tohoku district. Bull. Jap. Soc. Scient. Fish. 44, 1249–1255. 39. Yasumoto, T., Murata, M., Oshima, Y., Matsumoto, K., and Clardy, J. (1984) Diarrhetic shellfish poisoning, in Seafood Toxins (Ragelis, E. P., ed.), ACS Symp. Ser. vol. 262, American Chemical Society, Washington DC, pp. 207–214. 40. Lee, J. S., Yanagi, T., Kenma, R., and Yasumoto, T. (1987) Fluorometric determination of diarrhetic shellfish toxins by high-performance liquid chromatography. Agric. Biol. Chem. 51, 877–881. 41. Dahl, E., Rogstad, A., Aune, T., Hormazabal, V., and Underdal, B. (1995) Toxicity of mussels related to occurrence of Dinophysis species, in Harmful Marine Algal Blooms (Lassus, P., Arzul, G., Erard-Le Denn, E., Gentien, P., and Marcaillou-Le Baut, C., eds.), Lavoisier Science Publishers, Paris, pp. 783–788. 42. Suzuki, T., Yoshizawa, R., Kawamura, T., and Yamasaki, M. (1996) Interference of free fatty acids from the hepatopancreas of mussels with the mouse bioassay for shellfish toxins. Lipids 31, 641–645. 43. Lawrence, J. F., Chadha, R . K., Ratnayake, N. W. M., and Truelove, J. F. (1994) An incident of elevated levels of unsaturated free fatty acids in mussels from Nova Scotia and their toxic effect in mice after intraperitoneal injection. Nat. Toxins 2, 318–321. 44. McCulloch, A., Boyd, R., de Freitas, A., Foxall, R., Jamieson, W., Laycock, M., et al.

286

45. 46.

47. 48. 49. 50. 51. 52. 53.

54. 55.

56.

57. 58. 59. 60.

61. 62.

63.

64.

65.

Towers and Garthwaite (1989) Zinc from oyster tissue as causative factor in mouse deaths in official bioassay for paralytic shellfish poison. J. Assoc. Off. Anal. Chem. 72, 384–386. Shantz, E. J., McFarren, E. F., Schaffer, M. L., and Lewie, K. H. (1958) Purified poison for bioassay standardization. J. Assoc. Off. Anal. Chem. 41, 160–168. Fernandez, M. and Cembella, A. D. (1995) Mammalian Bioassays, in Manual on Harmful Marine Microalgae (Hallegraeff, G. M., Anderson, D. M., and Cembella, A. D., eds.), IOC Manuals and Guides No 33. UNESCO, Paris, pp. 213–228. Hamano, Y. (1991) Studies on diarrhetic shellfish toxins, particularly okadaic acid. Osaka Daigaku Igaku Zasshi 43, 67–87. Kat, M. 1983. Diarrhetic mussel poisoning in the Netherlands related to the dinoflagellate Dinophysis acuminata. Antonie van Leeuwenhoek 49, 417–427. Banner, A. H., Scheuer, P. J., Sasaki, S., Helfrich, P., and Alender, C. B. (1960) Observations on ciguateria-type toxin in fish. Ann. NY Acad. Sci. 90, 770–787. Lewis, R. J. (1987) Ciguatera (fish poisoning) with special reference to cats. Veterin. Clin. Toxicol. 103, 59–61. Vernoux, J. P., Lahlou, N., Magras, L., and Greaux, J.B. (1985) Chick feeding test: a simple system to detect ciguatoxin. Acta Trop. 42, 235–240. Stabell, O. B. and Kildalsen, H. (1995) Blowflies (Calliphora vomitoria) as an in vivo bioassay for diarrhetic and hemolytic algae toxins. Toxicol. Methods 5, 213–223 Bagnis, R., Chanteau, E., Chungue, E., Drollet, J. H., Lechat, I., and Legrand, A. M., (1987) The use of the mosquito bioassay for determining the toxicity to man of ciguateric fish. Biol. Bull. 72, 137–143. Turell, M. J. and Middlebrook, J. L. (1988) Mosquito inoculation: an alternative bioassay for toxins. Toxicon 26, 1089–1094. Vernoux, J. P., Le Baut, C., Masselin, P., Marais, C., Baron, B., Choumiloff, R., et al. (1993) The use of Daphnia magna for detection of okadaic acid in mussel extracts. Food Addit. Contam. 10, 603–608. McElhiney, J., Lawton, L. A., Edwards, C., and Gallacher, S. (1998) Development of a bioassay employing the desert locust (Schistocerca gregaria) for the detection of saxitoxin and related compounds in cyanobacteria and shellfish. Toxicon 36, 417–420. Nakajima, I., Oshima, Y., and Yasumoto, T. (1981) Toxicity of benthic dinoflagellates in Okinawa. Bull. Jap. Soc. Scient. Fish. 47, 1029–1033. Igarashi, T., Aritake, S., and Yasumoto, T. (1998) Biological activities of prymnesin–2 isolated from a red tide alga Prymnesium parvum. Nat. Toxins 6, 35–41. Khan, S., Arakawa, O., and Onoue, Y. (1996) A toxicological study of the marine phytoflagellate, Chattonella antiqua (Raphidophyceae). Phycologia 35, 239–244. Aune, T., Yasumoto, T., and Engeland, E. (1991) Light and scanning electron microscopic studies on effects of marine algal toxins toward freshly prepared hepatocytes. J. Toxicol. Environ. Health 34, 1–9. Blay, J. and Poon, A. S. L. (1995) Use of cultured permanent lines of intestinal epithelial cells for the assay of okadaic acid in mussel homogenates. Toxicon 33, 739–746. Tubaro, A., Florio, C., Luxich, E., Vertua, R., Loggia, R. D., and Yasumoto, T. (1996) Suitability of the MTT-based cytotoxicity assay to detect okadaic acid contamination of mussels. Toxicon 34, 965–974. Fessard, V., Diogene, G., Dubreuil, A., Quod, J. P., Durand-Clement, M., Legay, C., and Puiseux-Dao, S. (1994) Selection of cytotoxic responses to maitotoxin and okadaic acid and evaluation of toxicity of dinoflagellate extracts. Nat. Toxins 2, 322–328. Croci, L., Cozzi, L., Stacchini, A., de Medici, D., and Toti, L. (1997) A rapid tissue culture assay for the detection of okadaic acid and related compounds in mussels. Toxicon 35, 223–230. Amzil, Z., Pouchus, Y. F., Le Boterff, J., Roussakis, C., Verbist, J. F., Marcaillou-Lebaut,


66.

67.

68.

69.

70.

71. 72.

73. 74.

75.

76.

77.

78.

79.

80.

81. 82.

287

C., and Masselin, P. (1992) Short-time cytotoxicity of mussel extracts: a new bioassay for okadaic acid detection. Toxicon 30, 1419–1425. Manger, R. L., Leja, L. S., Lee, S. Y., Hungerford, J. M., Hokama, Y., Dickey, R. W., et al. (1995) Detection of sodium channel toxins: directed cytotoxicity assays of purified ciguatoxins, brevetoxins, saxitoxins, and seafood toxins. J. AOAC Int. 78, 521–527. Jellett, J. F., Marks, L. J., Stewart, J. E., Dorey, M. L., Watson-Wright, W., and Lawrence, J. F. (1992) Paralytic shellfish poison (saxitoxin family) bioassays: automated endpoint determination and standardization of the in vitro tissue culture bioassay, and comparison with the standard mouse bioassay. Toxicon 30, 1143–1156. Gallacher, S. and Birkbeck, T. H. (1992) A tissue culture assay for direct detection of sodium channel blocking toxins in bacterial culture supernates. FEMS Microbiol. Letters 92, 101–108. Garthwaite, I., Ross, K., Poli, M., and Towers, N. R. (1996) Comparison of immunoassay, cellular, and classical mouse bioassay methods for detection of neurotoxic shellfish toxins, in Immunoassays for Residue Analysis (Beier, R. C. and Stanker, L. H., eds.), ACS Symp. Ser., vol. 621, American Chemical Society, Washington, DC, pp. 404–412. Manger, R. L., Leja, L. S., Lee, S. Y., Hungerford, J. M. and Wekell, M. M. (1993) Tetrazolium-based cell bioassay for neurotoxins active on voltage-sensitive sodium channels: semiautomated assay for saxitoxins, brevetoxins, and ciguatoxins. Anal. Biochem. 214, 190–194. Bignami, G. S. (1993) A rapid and sensitive hemolysis neutralization assay for palytoxin. Toxicon 31, 817–820. Bignami, G. S., Raybould, T. J. G., Sachinvala, N. D., Grothaus, P. G., Simpson, S .B., Lazo, C. B., et al. (1992) Monoclonal antibody-based enzyme-linked immunoassays for the measurement of palytoxin in biological samples. Toxicon 30, 687–700. Takahashi, M., Ohizumi, Y., and Yasumoto, T. (1982) Maitotoxin, a Ca2+ channel activator candidate. J. Biol. Chem. 257, 7287–7289. Whitney, P. L., Delgado, J. A., and Baden, D. G. (1997) Complex behavior of marine animal tissue extracts in the competitive binding assay of brevetoxins with rat brain synaptosomes. Nat. Toxins 5, 193–200. Van Dolah, F. M., Finley, E. L., Haynes, B. L., Doucette, G. J., Moeller, P. D., and Ramsdell, J. S. (1994) Development of rapid and sensitive high throughput pharmacologic assays for marine phycotoxins. Nat. Toxins 2, 189–196. Vieytes, M. R., Cabado, A. G., Alfonso, A., Louzao, M. C., Botana, A. M., and Botana, L. M. (1993) Solid-phase radioreceptor assay for paralytic shellfish toxins. Anal. Biochem. 211, 87–93. Baden, D. G., Mende, T. J., Szmant, A. M., Trainer, V. L., Edwards, R. A., and Roszell, L. E. (1988) Brevetoxin binding: molecular pharmacology versus immunoassay. Toxicon 26, 97–103. Doucette, G., Powell, C., Do, E., Byon, C., Cleves, F., and McClain, S. (2000) Evaluation of 11-[3H]-tetrodotoxin use in a heterologous receptor binding assay for PSP toxins. Toxicon 38, 1465–1474. Doucette, G. J., Logan, M. M., Ramsdell, J. S., and Van Dolah, F. M. (1997) Development and preliminary validation of a microtiter plate-based receptor binding assay for paralytic shellfish poisoning toxins. Toxicon 35, 625–636. Negri, A. and Llewellyn, L. (1998) Comparative analyses by HPLC and the sodium channel and saxiphilin 3H-saxitoxin receptor assays for paralytic shellfish toxins in crustaceans and molluscs from tropical North West Australia. Toxicon 36, 283–298. Trainer, V. L., Baden, D. G., and Catterall, W. A. (1995) Detection of marine toxins using reconstituted sodium channels. J. AOAC Int. 78, 570–573. Fairey, E. R., Edmunds, J. S., and Ramsdell, J. S. (1997) A cell-based assay for

288

83. 84.

85.

86.

87.

88.

89. 90. 91.

92. 93.

94.

95.

96. 97. 98.

99.

100.

Towers and Garthwaite brevetoxins, saxitoxins, and ciguatoxins using a stably expressed c-fos-luciferase reporter gene. Anal. Biochem. 251, 129–132. Kerr, D. S., Briggs, D. M., and Saba, H. I. (1999) A neurophysiological method of rapid detection and analysis of marine algal toxins. Toxicon 37, 1803–1825. Poli, M., Lewis, R., Dickey, R., Musser, S., Buckner, C., and Carpenter, L. (1997) Identification of Caribbean ciguatoxins as the cause of an outbreak of fish poisoning among US soldiers in Haiti. Toxicon 35, 733–741. United Nations (1994) Convention on the prohibition of the development, production, stockpiling and use of chemical weapons and on their destruction. Depositary Notification C.N.246–1994 TREATIES–5, United Nations Organisation, Paris. Holmes, C. F. B. (1991) Liquid chromatography-linked protein phosphatase bioassay: a highly sensitive marine bioscreen for okadaic acid and related diarrhetic shellfish toxins. Toxicon 29, 469–477. Tubaro, A., Florio, C., Luxich, E., Sosa, S., della Loggia, R., and Yasumoto, T. (1996) A protein phosphatase 2A inhibition assay for a fast and sensitive assessment of okadaic acid contamination in mussels. Toxicon 34, 743–752. Mountford, D. O., Kennedey, G., Garthwaite, I., Quilliam, M., Trueman, P., and Hannah, D. J. (1999) Evaluation of the fluorometric protein phosphatase inhibition assay in the determination of okadaic acid in mussels. Toxicon 37, 909–922. Vieytes, M. R., Fontal, O. I., Leira, F., Baptista de Sousa, J. M., and Botana, L. M. (1997) A fluorescent microplate assay for diarrheic shellfish toxins. Anal. Biochem. 248, 258–264. Serres, M., Fladmark, K., and Doskeland, S. (2000) An ultrasensitive competitive binding assay for the detection of toxins affecting protein phosphatases. Toxicon 38, 347–360. Johnson, H. M., Frey, P. A., Angelotti, R., Campbell, J. E., and Lewis, K. H. (1964) Haptenic properties of paralytic shellfish poison conjugated to proteins by formaldehyde treatment. Proc. Soc. Exp. Biol. Med. 117, 425–430. Chu, F. S. and Fan, T. S. L., (1985) Indirect enzyme-linked immunosorbent assay for saxitoxin in shellfish. J. Assoc. Off. Anal. Chem. 68, 13–16. Usleber, E., Schneider, E., and Terplan, G. (1991) Direct enzyme immunoassay in microtitration plate and test strip format for the detection of saxitoxin in shellfish. Lett. Appl. Microbiol. 13, 275–277. Cembella, A. D. and Lamoureux, G. (1993) A competitive inhibition enzyme-linked immunoassay for the detection of paralytic shellfish toxins in marine phytoplankton. Dev. Mar. Biol. 3, 857–862. Usleber, E., Dietrich, R., Martlbauer, E. P., and Terplan, G. (1994) Effect of heterologous paralytic shellfish poisoning toxin-enzyme conjugates on the cross-reactivity of a saxitoxin enzyme immunoassay. Lett. Appl. Microbiol. 18, 337–339. Usleber, E., Dietrich, R., Burke, C., Schneider, E., and Maetlbauer, E. (2001) Immunoassay methods for paralytic shellfish poisoning (PSP) toxins. J. AOAC Int., in press. Chu, F. S. and Huang, X. (1992) Production and characterization of antibodies against neosaxitoxin. J. AOAC Int. 75, 341–345. Chu, F. S., Hsu, K.-H., Huang, X., Barrett, R., and Allison, C. (1996) Screening of paralytic shellfish poisoning toxins in naturally occurring samples with three different direct competitive enzyme-linked immunosorbent assays. J. Agric. Food Chem. 44, 4043–4047. Newsome, H., Truelove, J., Hierlihy, L., and Collins, P. (1991) Determination of domoic acid in serum and urine by immunochemical analysis. Bull. Environ. Contam. Toxicol. 47, 329–334. Osada, M., Marks, L. J., and Stewart, J. E. (1995) Determination of domoic acid by two different versions of a competitive enzyme-linked immunosorbent assay (ELISA). Bull. Environ. Contam. Toxicol. 54, 797–804.


289

101. Smith, D. S. and Kitts, D. D. (1995) Enzyme immunoassay for the determination of domoic acid in mussel extracts. J. Agric. Food Chem. 43, 367–371. 102. Kawatsu, K., Hamano, Y., and Noguchi, T. (1999) Production and characterization of a monoclonal antibody against domoic acid and its application to enzyme immunoassay. Toxicon 37, 1579–1589. 103. Garthwaite, I., Ross, K. M., Miles, C. O., Hansen, R. P., Foster, D., Wilkins, A. L., and Towers, N. R. (1998) Polyclonal antibodies to domoic acid, and their use in immunoassays for domoic acid in sea water and shellfish. Nat. Toxins 6, 93–104. 104. Rhodes, L., Scholin, C., and Garthwaite, I. (1998) Pseudo-nitzschia in New Zealand and the role of DNA probes and Immunoassays in refining marine biotoxin monitoring programmes. Nat. Toxins 6, 105–111. 105. Trainer, V. L. and Baden, D. G. (1991) An enzyme immunoassay for the detection of Florida red tide brevetoxins. Toxicon 29, 1387–1394. 106. Levine, L. and Shimizu, Y. (1992) Antibodies to brevetoxin B: serologic differentiation of brevetoxin B and brevetoxin A. Toxicon 30, 411–418. 107. Poli, M. A., Rein, K. S., and Baden, D. G. (1995) Radioimmunoassay for PbTx–2-type brevetoxins: epitope specificity of two anti-PbTx sera. J. AOAC Int. 78, 538–542. 108. Baden, D. G., Melinek, R., Sechet, V., Trainer, V. L., Schultz, D. R., Rein, K. S., et al. (1995) Modified immunoassays for polyether toxins: implications of biological matrixes, metabolic states, and epitope recognition. J. AOAC Int. 78, 499–508. 109. Garthwaite, I., Ross, K. M., and Haywood, A. (1998) Detection of NSP (brevetoxins) in algal cultures and seawater by ELISA, in Proceedings of the 10th Biotoxin Science Workshop, New Zealand Ministry of Agriculture and Forestry, Wellington, New Zealand, pp. 51–54. 110. Usagawa, T., Nishimura, M., Itoh, Y., Uda, T., and Yasumoto, T. (1989) Preparation of monoclonal antibodies against okadaic acid prepared from the sponge Halichondria okadai. Toxicon 27, 1323–1330. 111. Matsuura, S., Hamano, Y., Kita, H., and Takagaki, Y. (1994) An ELISA for okadaic acid and its analogs among the diarrhetic shellfish toxins using mouse monoclonal anti-okadaic acid antibodies which are resistant to organic solvents. Jpn. J. Toxicol. Environ. Health 40, 365–373. 112. Chin, J. D., Quilliam, M. A., Fremy, J. M., Mohapatra, S. K., and Sikorska, H. M. (1995) Screening for okadaic acid by immunoassay. J. AOAC Int. 78, 508–513. 113. Briggs, L., Towers, N. R., Miles, C. O., Garthwaite, I., and Ross, K. M. (1999) Progress towards ELISAs for okadaic acid and yessotoxin, in Proceedings of the 11th Biotoxin Science Workshop, New Zealand Ministry of Agriculture and Forestry, Wellington, New Zealand, pp. 98–102. 114. Hokama, Y., Honda, S. A. A., Kobayashi, M. N., Nakagawa, L. K., Asahina, A. Y., and Miyahara, J. T. (1989) Monoclonal antibody (MAb) in detection of ciguatoxin (CTX) and related polyethers by the stick-enzyme immunoassay (S-EIA) in fish tissues associated with ciguatera poisoning, in Mycotoxins and Phycotoxins ’88 (Natori, S., Hashimoto, K., and Ueno, Y., eds), Elsevier Science, Amsterdam, pp. 303–310. 115. Chu, F. S., Huang, X., and Wei, R. D. 1990. Enzyme-linked immunosorbent assay for microcystins in blue-green algal blooms. J. AOAC Int. 73, 451–456. 116. Baier, W., Loleit, M., Fischer, B., Jung, G., Neumann, U., Weiss, M., et al. (2000). Generation of antibodies directed against the low-immunogenic peptide—toxins microcystinLR/RR and nodularin. Int. J. Immunopharmacol. 22, 339–353. 117. Tubaro, A., Sosa, S., Bruno, M., Gucci, P. M. B., Volterra, L., and Della Loggia, R. (1992) Diarrheic shellfish toxins in Adriatic Sea mussels evaluated by an ELISA method. Toxicon 30, 673–676.

290


118. Draisci, R., Croci, L., Giannetti, L., Cozzi, L., Lucentini, L., De Medici, D., and Stacchini, A. (1994) Comparison of mouse bioassay, HPLC and enzyme immunoassay methods for determining diarrhetic shellfish poisoning toxins in mussels. Toxicon 32, 1379–1384. 119. Carmody, E. P., James, K. J., and Kelly, S. S. (1995) Diarrhetic shellfish poisoning: evaluation of enzyme-linked immunosorbent assay methods for determination of dinophysistoxin–2. J. AOAC Int. 78, 1403–1408. 120. Pauillac, S., Sasaki, M., Inoue, M., Naar, J., Branaa, P., Chinain, M., et al. (2000) Characterization of mice antisera elicited with a ciguatoxin tetracyclic synthetic ring fragment (JKLM) conjugated to carrier proteins. Toxicon 38, 669–685. 121. Dietrich, D., Fischer, W., Garthwaite, I., Miles, C. O., Towers, N. R., Chamberlain, R., and Aggen, J. B. (2000) Congener independent detection of microcystin and nodularin. European Patent Application No. PCT/EP00/08711. 122. Schneider, E., Usleber, E., and Terplan, G. (1991) Test strip enzyme immunoassay for the detection of saxitoxin. Food Agric. Immunol. 3, 103–104. 123. Carter, R. M., Poli, M. A., Pesavento, M., Sibley, D. E. T., Lubrano, G. J., and Guilbault, G. G. (1993) Immunoelectrochemical biosensors for detection of saxitoxin and brevetoxin. Immunol. Methods 3, 128–133. 124. Marquette, C. A., Coulet, P. R., and Blum, L. J. (1999) Semi-automated membrane based chemiluminescent immunosensor for flow injection analysis of okadaic acid in mussels. Anal. Chim. Acta 398, 173–182. 125. Kreuzer, M. P., O’Sullivan, C. K., and Guilbault, G .G. (1999) Development of an ultrasensitive immunoassay for rapid measurement of okadaic acid and its isomers. Anal. Chem. 71, 4198–4202. 126. Kondo, F., Matsumoto, H., Yamada, S., Tsuji, K., Ueno, Y., and Harada, K. (2000) Immunoaffinity purification method for detection and quantification of microcystins in lake water. Toxicon 38, 813–823. 127. Ten-Hage, L., Delaunay, N., Pichon, V., Coute, A., Puiseux-Dao, S., and Turquet, J. (2000) Okadaic acid production from the marine benthic dinoflagellate Prorocentrum arenarium Faust (Dinophyceae) isolated from Europa Island coral reef ecosystem (SW Indian Ocean). Toxicon 38, 1043–1054. 128. Puech, L., Dragacci, S., Gleizes, E., and Fremy, J.-M. (1999) Use of immunoaffinity columns for clean-up of diarrhetic toxins (okadaic acid and dinophysistoxins) extracts from shellfish prior to their analysis by HPLC/fluorimetry. Food Addit. Contam. 16, 239–251. 129. Dietrich, R., Usleber, E., Buerk, C., and Maertlbauer, E. (1996) Immunochemical approaches to the analysis of paralytic shellfish poisoning toxins, in Immunoassays for Residue Analysis (Beier, R. C. and Stanker, L. H., eds), ACS Symp. Ser., vol. 621, American Chemical Society, Washington, DC, pp. 395–403. 130. Delaunay, N., Pichon, V., Caer, J. P. L., and Hennion, M. C. (2000) Immunoaffinity extraction as a new approach for an improved liquid chromatography-mass spectrometric or fluorimetric determination of okadaic acid in shellfish and algae. Anal. Chim. Acta 407, 173–186. 131. Holmes, C. F. (1991) Liquid chromatography-linked protein phosphatase bioassay; a highly sensitive marine bioscreen for okadaic acid and related diarrhetic shellfish toxins. Toxicon 29, 469–477. 132. Aase, B. and Rogstad, A. (1997) Optimization of sample cleanup procedure for determination of diarrhoeic shellfish poisoning toxins by use of experimental design. J. Chromatogr. A 764, 223–231. 133. Dickey, R., Jester, E., Granade, R., Mowdy, D., Moncreiff, C., Rebarchik, D., et al. (1999) Monitoring brevetoxins during a Gymnodinium breve red tide: comparison of sodium


134.

135.

136.

137.

138.

139.

140.

141.

142.

143. 144.

291

channel specific cytotoxicity assay and mouse bioassay for determination of neurotoxic shellfish toxins in shellfish extracts. Nat. Toxins 7, 157–165. Quilliam, M. A., Sim, P. G., McCulloch, A. W., and McInnes, A. G. (1989) High-performance liquid chromatography of domoic acid, a marine neurotoxin, with application to shellfish and plankton. Int. J. Environ. Anal. Chem. 36, 139–154. Lawrence, J. F., Charbonneau, C. F., Ménard, C., Quilliam, M. A., and Sim, P. G. (1989) Liquid chromatographic determination of domoic acid in shellfish products using the paralytic shellfish poison extraction procedure of the Association of Official Analytical Chemists. J. Chromatogr. 462, 349–356. Pocklington, R., Milley, J. E., Bates, S. S., Bird, C. J., de Freitas, A. S. W., and Quilliam, M. A. (1990) Trace determination of domoic acid in seawater and phytoplankton by highperformance liquid chromatography of the fluorenylmethoxycarbonyl (FMOC) derivative. Int. J. Environ. Anal. Chem. 38, 351–368. Lawrence, J. F., Ménard, C., and Cleroux, C. (1995) Evaluation of prechromatographic oxidation for liquid chromatographic determination of paralytic shellfish poisons in shellfish. J. AOAC Int. 78, 514–520. Oshima, Y. (1995) Post-column derivatization HPLC methods for paralytic shellfish poisons, in Manual on Harmful Marine Microalgae (Hallegraeff, G. M., Anderson, D. M., and Cembella, A. D., eds.), IOC Manuals and Guides No. 33., UNESCO, pp. 81–94. Locke, S. J. and Thibault, P. (1994) Improvement in detection limits for the determination of paralytic shellfish poisoning toxins in shellfish tissues using capillary electrophoresis/electrospray mass spectrometry and discontinuous buffer systems. Anal. Chem. 66, 3436–3446. Quilliam, M. A. and Wright, J. L. C. (1995) Methods for diarrhetic shellfish poisons, in Manual on Harmful Marine Microalgae (Hallegraeff, G. M., Anderson, D. M., and Cembella, A. D., eds.), IOC Manuals and Guides No. 33., UNESCO, pp. 95–112. Quilliam, M. A. (1995) Analysis of diarrhetic shellfish poisoning toxins in shellfish tissue by liquid chromatography with fluorometric and mass spectrometric detection. J. AOAC Int. 78, 555–570. James, K. J., Bishop, A. G., Gillman, M., Kelly, S. S., Roden, C., Draisci, R., et al. (1997) Liquid chromatography with fluorimetric, mass spectrometric and tandem mass spectrometric detection for the investigation of the seafood-toxin-producing phytoplankton, Dinophysis acuta. J. Chromatogr. A 777, 213–221. Hua, Y. S. and Cole, R. B. (2000) Electrospray ionization tandem mass spectrometry for structural elucidation of protonated brevetoxins in red tide algae. Anal. Chem. 72, 376–383. Lewis, R. J., Jones, A., Vernoux, J.-P., and Marquis, M. (1998) Sensitive detection of multiple ciguatoxins by HPLC/MS/MS, in Harmful Algae (Reguera, B., Blanco, J., Fernandez, M. L., and Wyatt, T., eds.), Xunta de Galicia and IOC/UNESCO, Vigo, pp. 523–524.

292


Clostridial Neurotoxins

293

16 An Overview of Clostridial Neurotoxins Mark A. Poli and Frank J. Lebeda

1. INTRODUCTION Neurotoxins produced by the anaerobic bacteria Clostridium botulinum and C. tetani are some of the most potent naturally occurring compounds known. Their exquisite toxicity coupled with their highly specific mechanism of action render them both highly dangerous but yet quite useful to medical science. Tetanus toxin (TeNT) is usually encountered as a wound contaminant and is a significant health problem in developing countries. Botulinum neurotoxins (BoNTs) are typically encountered in food poisoning, although they also occur as a result of wound infection (wound botulism) or as a colonizing infection in the neonatal intestinal tract (infant botulism). Tetanus intoxication, known for thousands of years, is effectively controlled in developed countries via childhood vaccination. In contrast, botulism became a common public health threat only after the advent of food preservation in the 19th century. Modern food-preparation practices have rendered botulism a rare occurrence from commercially prepared foods, although a small but significant number of cases occur annually from eating home-canned foods. Because the incidence of botulism is low, the general populace has not been vaccinated. During World War II, the extremely high potency of BoNTs induced both the Allied and Axis powers to evaluate them as potential biological warfare agents (1). Although this work formally ended with the signing of the 1972 Biological Weapons Convention, recent events in the Middle East have confirmed the weaponization of these toxins by the Iraqi military before and during the Gulf War (2). However, food poisoning and weapons of mass destruction are not the only legacy of human interaction with these toxins. Today, BoNT is used to treat strabismus, blepharospasm, and other medical conditions that require a relatively specific and long-lasting block of muscle contractions (3,4). In addition, these toxins have become valuable tools in understanding the natural processes underlying neural function (5,6). *The views, opinions, and/or findings contained herein are those of the authors and should not be construed as an official Department of the Army position, policy, or decision unless so designated by other documentation. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

293

294

Poli and Lebeda

Seven serologically distinct serotypes of BoNT, denoted serotypes A–G, are currently recognized. They are polypeptides with a molecular weight of 150–180 kD. Although synthesized as single polypeptide chains, cleavage by intra- or extracellular proteases coverts them into dimers consisting of a heavy (H) chain of approx 100 kD coupled to a light (L) chain of approx 50 kD by one or more disulfide bridges. Each serotype is produced as the primary toxin by a specific strain of bacteria and, although they share a high degree of homology, serotypes differ in their toxicity and molecular site of action. In bacterial culture, they are produced in association with nontoxic proteins that vary according to serotype. Although some have hemagglutinin activity, the function of these associated proteins is not completely understood. They are believed to stabilize the toxin complex against proteolytic cleavage in the gut (7). Only one serotype of TeNT is produced by all toxic strains of C. tetani. It is a paradox that, while similar in size, structure, and mechanism of action to the BoNTs, TeNT induces dramatically different clinical manifestations. Tetanus neurotoxin is not typically associated with food poisoning, probably because it is not complexed with nontoxic proteins (8). The intoxication process involves a retrograde transport of toxin from the peripheral-nerve terminals to the central nervous system (CNS) where inhibitory neurons are blocked, and a spastic paralysis results. The BoNTs remain at peripheral neuromuscular junctions, blocking the release of acetylcholine and inducing a flaccid paralysis (5). We will provide here a brief overview of the clinical picture of botulism at the organism level, the mechanism of action of the neurotoxins at the cellular level, and some of the implications of structural components on the function of these toxins at the molecular level. 2. CLINICAL ASPECTS OF BOTULISM AND TETANUS 2.1. Botulism Botulism is a very serious and potentially fatal intoxication. Food-borne botulism results from ingesting pre-formed toxin in foodstuffs contaminated with C. botulinum. However, wound botulism is caused by toxic strains of C. botulinum growing and producing toxin in an infected wound, and infant botulism results from colonization and subsequent toxin production in the intestinal tract of infants. 2.1.1. Food-Borne Botulism

In humans, food-borne botulism results from eating food containing toxins from the contaminant bacteria. In animals, food-borne botulism is typically vectored by carrion; it results from either the direct ingestion of carrion, ingestion of feed contaminated by carrion, or in the case of birds, the ingestion of insect larvae feeding on carrion (9). According to Hatheway (9), in recent years an average of 449 human outbreaks involving 930 cases per year occur worldwide. The majority of these outbreaks (72%) occurred in Poland. More than 90% of food-borne botulism is caused by eating insufficiently cooked home-canned foods. In most of the United States, botulism is caused primarily by BoNT-A or, less commonly, BoNT-B. In Alaska, where botulism is almost exclusively linked to home-prepared fish or marine mammals prepared in native fash-


295

ion, the vast majority of cases are attributable to BoNT-E. Only one case of BoNT-F intoxication was reported in the United States during 1975–1992. Serotypes C and D are typically linked to animal cases. Botulism in animals and birds is reviewed by Smith and Sugiyama (10). Victims of food-borne botulism present 12–36 h after exposure with symptoms of cranial-nerve paralysis, including blurred vision, double vision, photophobia, and dry mouth. Patients are usually afebrile, and gastrointestinal symptoms are typically mild or absent, except with BoNT-E. Victims of BoNT-E poisoning are unique in that nausea and vomiting occurs in all cases (11). Weakness of the neck muscles and difficulty in swallowing follow, and if left untreated, a symmetric descending paralysis will ultimately involve the skeletal and respiratory muscles. Patients are conscious and usually retain normal senses (10), although paresthesias may occur in a small number of cases (12). Paralysis of the respiratory musculature can prove fatal without medical intervention. 2.1.2. Wound Botulism Wound botulism is the rarest form of botulism; only one to two cases are reported in the United States annually (9). As with food-borne botulism, wound botulism is usually associated with BoNT-A. It is linked primarily to traumatic injury such as fractures of long bones or crush injuries of the extremities, but has also been reported in association with surgery or illicit drug use. Because symptoms stem from wound infection and subsequent in vivo toxin production rather than ingestion of preformed toxin, the incubation time from injury to symptoms is much longer, typically several days to 2 wk (10). However, the symptomotology is similar to that of food-borne botulism. 2.1.3. Infant Botulism Infant botulism usually occurs in patients less than 6 mo of age, and usually in those being breast-fed (13). Breast-feeding may result in a more favorable intestinal environment for spore formation (11). It is the most common form of botulism in the United States; from 1976 through 1992, 1134 cases were reported to the Centers for Disease Control (CDC), of which nearly half occurred in California (9). Unlike food-borne botulism, where the incidence of BoNT-A intoxication exceeds BoNT-B intoxication by threefold, more than 50% of infant botulism cases are linked to BoNT-B (9). The primary source of C. botulinum is the environment; the only implicated food vector is honey (14). Because infant botulism is a colonizing infection with subsequent toxin formation, incubation periods may vary from several days to 2 wk. Patients present first with autonomic dysfunction manifested as constipation, hypoactive bowel sounds, dry mouth, dilated pupils, and neurogenic bladder (11). This is followed by a descending paralysis beginning with the cranial nerves, including loss of head control, poor feeding, weak cry, decreased eye movement, facial diplegia, and finally diminished movement of the extremities and loss of deep tendon reflex (15,16). Supportive care is the mainstay of treatment and usually results in complete recovery after several weeks (17). 2.1.4. Persistence of Action

In nonfatal cases of poisoning and in clinical therapeutic usage, a hallmark feature of the BoNT paralysis, especially that produced by serotype A, is the relatively long

296

Poli and Lebeda

duration, ranging from a few weeks to several months. The persistence of paralysis by BoNT-A has been associated with a maintained intracellular enzyme activity (18) and with a long-lasting depletion of the vesicle-fusion protein, SNAP-25 (19). Both of these results must be considered when developing therapies for poisonings and when using these neurotoxins to treat neurologic disorders. 2.2. Tetanus Tetanus is a severe neurological illness characterized by intense spasms of the skeletal musculature. It was well-known even to the ancient Greek and Egyptian physicians who recognized the relationship between traumatic injury and development of fatal muscle spasms. Although the causative toxin was isolated from C. tetani culture fluids in 1884 (20), and an effective toxoid vaccine was described in 1890 (21), tetanus remains a significant public-health problem. Worldwide incidence is estimated to be about 1 million cases annually, primarily in developing countries; the United States reports about 70 cases annually (22). The mortality rate is 20–30%. Where childhood vaccination is common, most reported cases occur in the elderly, suggesting waning immunity with age. In contrast, neonatal tetanus accounts for approx half of the mortality in developing nations. The tragedy of tetanus is that it is an easily preventable disease; properly administered vaccination programs could effectively eliminate tetanus within a few years. In the century since the first tetanus vaccine, however, this potential has yet to be realized. In developed countries, punctures and lacerations are the primary portals of entry for C. tetani spores from the environment. The incubation period can vary from several days to several weeks, with shorter incubation times linked to poorer outcomes. Localized tetanus involves rigidity of the muscles around the injury site. Weakness and diminished muscle tone can persist for weeks. The localized form can resolve spontaneously, especially in persons with partial immunity, or it can progress to the generalized form of the disease (22). Generalized tetanus is the most easily recognized form, and often begins with rigidity of the masseter muscles (trismus, or “lockjaw”) and rigidity or tightening of the muscles of facial expression (risus sardonicus). Generalized muscle spasms follow, characterized by opisthotonic posturing with flexion of the arms and extension of the legs. The patient remains conscious and each spasm is extremely painful (22). If the diaphragm becomes spastic or the upper airway is obstructed, respiration can be compromised and death result. Even with managed respiratory function, however, fatal autonomic dysfunction can occur (23). The disease can progress for up to 2 wk before recovery begins. Recovery may require several weeks, but is usually complete in the absence of complications (22). Neonatal tetanus usually results from infection of the umbilical stump due to cultural practices (24) or failure of aseptic technique in nonimmune mothers (25). The patient typically presents with generalized weakness and failure to nurse; typical rigidity and spasms follow. The mortality rate exceeds 90% (26).


297

3. MECHANISM OF ACTION OF BOTULINUM AND TETANUS TOXINS Because the cellular effects of botulinum and tetanus toxins are dealt with in detail in Chapter 17 by Molgo et al. in this volume, only a brief overview will be presented here. The mechanism of action of TeNT and BoNTs are similar in many respects. Both bind tightly (KD ≈ 1 nM) to receptors on peripheral-nerve cells (27). The toxin/receptor complex is then internalized in a temperature-dependent manner within an endocytotic compartment (28). Upon acidification of the endocytic compartment by a vesicular proton pump, conformational changes in the protein result in oligomerization and ionchannel formation (29,30). Biophysics suggest that more than one toxin molecule is required for ion channel formation (31). It is generally believed that disulfide bridges holding the dimers are reduced, the H- and L-chains separate, and the L-chain (containing the enzymatic activity) is translocated into the neutral pH environment of the cytosol. At this point, the mechanisms of action of the two toxins diverge. The BoNTs remain in the presynaptic region of peripheral nerves, cleaving specific proteins critical for synaptic vesicle fusion. All BoNTs are zinc-dependent metalloproteases, but different serotypes attack different specific targets. BoNT-A and E cleave SNAP-25, BoNT-C cleaves syntaxin and SNAP-25, and the remaining toxins cleave VAMP (5,32,33). Cleavage of these proteins prevents Ca++-dependent docking and fusion of the acetylcholine-containing vesicles to the presynaptic membrane, inhibits neurotransmitter release, and results in the clinical findings of descending flaccid paralysis. While TeNT can inhibit neurotransmitter release and cause flaccid paralysis in peripheral nerves reminiscent of BoNT-B (34), these effects are typically overridden by its major effect on the CNS. Once taken up by peripheral neurons (motor, sensory, or adrenergic) and internalized by receptor-mediated endocytosis, retrograde axonal transport carries the majority of the toxin to the ventral spinal cord and brain stem (35) where blockade of inhibitory neurotransmitter release causes central disinhibition and spastic paralysis (36). 4. STRUCTURE/FUNCTION CONSIDERATIONS Within the last few years, critical new data have been presented on the three-dimensional crystal structures of TeNT and BoNT-A (37–39). With this new information, it is now possible to view the previous biochemical and physiological data in a structural context and ask new questions about the relationship between structure and function in this toxin family. Clostridial neurotoxins are expressed by the bacteria as single polypeptide chains of about 1300 amino acids. Subsequent proteolytic “nicking” produces two chains linked by disulfide bridges. The L chain comprises the first ~450 residues at the N-terminus, and the H chain comprises the remaining amino acids. A three-dimensional model of BoNT-A is presented in Fig. 1. 4.1. The L-Chain Structurally, the L-chain is comprised of both helical- and extended-strand components (Fig. 1, purple). Over a decade ago, it was observed that the L-chain of TeNT contains the subsequence HELIH , which corresponds to the zinc-binding region of the

298

Poli and Lebeda

Fig. 1. A representation of the three-dimensional structural domains of BoNT type A and their functional descriptions. Structures and locations of the domains in the whole neurotoxin are, from right to left: L-chain (purple), HN (yellow-green), belt (red), N-domain (teal) and the C-domain of HC (magenta). The catalytic L-chain is partially surrounded by a belt, a portion of the HN, the translocation/channel-forming domain. The binding domain, HC, contains highly exposed loops that are believed to be involved, at least in part, in the binding of extracellular receptors at the neuromuscular junction and in the binding to neutralizing antibodies. Atomic coordinates from the PDB file 3BTA (38,39). Illustrations made with VMD (58), SURF (59), and POV-Ray™ (60).

metalloprotease thermolysin (40). This link between primary structure and enzymatic function was later confirmed experimentally for TeNT and the BoNTs as well (29,41). In BoNT-A, the active site is a cavity containing a zinc ion (39). This cavity can accommodate at least 16 amino acid residues (38), which agrees with the minimal size required for cleavage of a BoNT-A substrate in vitro (42). 4.2. The H-Chain 4.2.1. The HN Region In contrast to the L-chain, the N-terminal portion of the H chain (HN) is mostly helical with a ~100 Å-long coiled structure (39). This type of structure is associated with membrane-fusion activities in certain viral proteins (43) and in heterotrimeric GTP-binding proteins in plants and animals (44).


299

Apart from this coiled coil structure, another region of the H chain is involved with the formation of ion-selective, voltage-dependent transmembrane channels. A series of predictions based on sequence algorithms (30,45–50) implicated clusters of amphipathic regions in the HN. These predictions were verified in studies with artificial membranes and a 23-mer peptide that formed ion-conducting channels (50) whose biophysical properties were similar to those produced by the parent neurotoxin (51). This fragment corresponded to an unstructured region located within the HN of BoNT-A (Fig. 1, yellow-green). In solution, different ratios of coil, helix, and strand configuration can be demonstrated under different solvent conditions (52). Although biophysics suggest that neurotoxin-induced ion channels can be formed by oligomerization, it is unknown whether they are also involved in translocation of the toxic moiety into the cytoplasm. Understanding the translocation of this moiety out of the endocytic vesicle in response to a reduction in pH remains a challenging problem. The three-dimensional crystallography data for BoNT-A also provided evidence for an unusual and unexpected structure intimately associated with the L-chain (38). The initial portion of the HN forms a ~55-residue belt-like structure that winds around the mid-portion of the L-chain and then forms an extensive ~50-residue loop next to the coiled-coil structure before re-joining the H-chain (Fig. 1, red). The short initial and final portions of this belt are in extended conformations that, together with a strand in the C-terminus of the L-chain, form a three-stranded sheet (not shown) that may act as a ligature to help anchor the L-chain in position. An even shorter two-stranded sheet (not shown) located near the top of the extensive loop may play a role in low pHdependent conformational changes because one of these strands is part of the channelforming, 23-residue region mentioned earlier. The portion of the belt that wraps around the L-chain also appears to occlude access to the active-site cleft in BoNT-A. A similar belt has been described for BoNT-B, but it appears to be displaced from the active site rather than covering it (53). A central question is whether movement of the belt away from the active site is required for functional binding of the substrate to the L-chain. From its position over the active site in BoNT-A, the belt may interfere with the binding of SNAP-25. If the belt in BoNT-B fulfills similar functions as in BoNT-A, one might expect this, and perhaps belts of the other serotypes, to move about when in an aqueous solution. In this case, the position of the belts in the BoNT-A and-B crystal structures may differ only due to differences in experimental conditions. From the standpoint of developing therapeutic agents, the presence of an occluding belt raises the concern that active-site inhibitors might be prevented from binding to the internalized neurotoxic component. Peptide and peptidomimetic inhibitors are effective in cell-free experimental systems in the presence of nicked and reduced neurotoxins (42,54,55). Nonreduced toxin, however, inefficiently cleaves substrate. In the presence of nontoxic neurotoxin-associated proteins (NAPs), protease activity in unreduced BoNT-A is ≈ 20-fold higher (56). The NAPs may constrain the L-chain and Hchain conformations such that the belt does not obscure the active site. If the belt is capable of sufficient movement, inhibitory ligands may be effective before internalization. On the other hand, if the active site is protected from pharmacological inhibitors by the belt, then these reagents may need to target the L-chain itself after internalization.

300

Poli and Lebeda

Another possible role for the belt is that it may help anchor together the L- and H-chains by providing additional noncovalent interactions. The belt interfaces with the L-chain over a larger buried surface area than the two chains alone. Examination of the three-dimensional structure suggests that the belt would have to move away from the active-site cleft for the chains to separate. Whether reduction of disulfide bonds, exposure to the low-pH environment, and movement of the belt are interdependent events required for chain separation is presently unclear. 4.3.2. The HC Region of the H-Chain

In contrast to both the L-chain and the HN, the C-terminal portion of the H-chain (HC) is a mostly beta-protein composed of two visually distinct structural domains. This region is thought to be involved in specific binding to neural receptors residing on the external surface of peripheral cholinergic neurons. The N-domain of the HC is in a sandwich fold in which the strands within a pair of sheets form a jelly-roll architecture related to that of the S-lectins, a carbohydrate-binding family of proteins (Fig. 1, teal). The C-domain is in a pseudo threefold trefoil conformation that is structurally similar to the sequentially unrelated interleukins-1α and 1β, as well as fibroblast growth factors (FGFs) (Fig. 1, magenta). These mostly β-proteins are involved in protein–protein interactions, a characteristic consistent with the hypothesis that clostridial neurotoxins bind to receptors comprised of functionally coupled protein and carbohydrate components (5,57). The identity of these neurotoxin receptors has not been established and remains a fertile area for further research. 5. SUMMARY The neurotoxins from C. botulinum and C. tetani have numerous similarities. Both are protein toxins produced by Gram-positive anaerobic bacilli that are widespread in distribution and pose significant public health problems. These toxins are heterodimeric proteins consisting of a ≈ 100 kD H-chain, which mediates binding to extracellular receptors and uptake into the cell, and a 50 kD L-chain that contains the Zn-dependent metalloprotease activity. Finally, these neurotoxins are specific for neural cells and interfere with the release of neurotransmitters by blocking vesicle fusion to the presynaptic membranes. In spite of these great similarities, however, these toxins produce strikingly different clinical manifestations. While BoNTs evoke a flaccid paralysis through blockade of acetylcholine release at peripheral neuromuscular junctions, TeNT produces a spastic paralysis via blockade of the release of inhibitory neurotransmitters within the CNS. Over the past decades, biochemical and biophysical experiments have provided insight into the binding, internalization, and mechanism of action of these toxins at the cellular level. In the past few years, three-dimensional crystal structures have presented a more detailed look at the biochemical and biophysical data and allow new questions to be asked about the mechanism of action of these toxins at the molecular level. The enzymatic active site on the L-chain has been identified. A region on the C-terminal portion of the H-chain has been found that is postulated to be involved in the binding of the toxin to its extracellular receptor. Similarly, a region on the N-terminal portion of the H-chain may be involved in the formation of voltage-gated channels and transloca-


301

tion of the toxic moiety into the neuroplasm. While these recent data give us a new level of understanding of how these toxins work, they also stimulate in-depth questions that will occupy scientists in the years to come. Among these are: 1. What is the nature of the specific neural receptor(s) for these toxins? How do the toxins interact with these receptors? 2. How does the oligomerization process and formation of ion channels relate to translocation of the toxic moiety into the neuroplasm? Does the toxic moiety consist of the L-chain alone, or is a portion of the H-chain translocated as well? If so, what is the role of that Hchain fragment? 3. Can the active-site geometry be used in the design of pharmacological inhibitors? Does the belt sufficiently occlude the active site to be a barrier to pharmacological intervention before internalization of the toxin? 4. What happens to the toxic moiety after translocation? Does it bind intracellularly to other cellular structures or remain free in the cytoplasm? If intracellularly bound, can differences in the avidity of this binding correlate to persistence of effects among serotypes? 5. What factors, or combination of factors, differentiate BoNTs from TeNT and underlie the difference in intracellular trafficking and subsequent differences in pharmacology?

REFERENCES 1. Middlebrook, J. L. and Franz, D. R. (1997) Botulinum toxins, in Textbook of Military Medicine, Part 1: Medical Aspects of Chemical and Biological Warfare (Sidell, F. R, Takafuji, E. T., and Franz, D. R., eds.), Office of the Surgeon General, US Army, Falls Church, VA, pp. 643–654. 2. Smart, J. K. (1997) History of chemical and biological warfare: an American perspective, in Textbook of Military Medicine, Part 1: Medical Aspects of Chemical And Biological Warfare (Sidell, F. R., Takafuji, E. T., and Franz, D.R., eds.), Office of the Surgeon General, US Army, Falls Church, VA, pp. 9–86. 3. Jankovic, J. (1994) Botulinum toxin in movement disorders. Curr. Opin. Neurol. 7, 358–366. 4. Dutton, J. J. (1996) Botulinum-A toxin in the treatment of craniocervical musclespasms: short- and long-term, local, and systemic effects. Surv. Opthamol. 41, 51–65. 5. Schiavo, G., Rosseto, O., and Montecucco, C. (1994) Clostridial neurotoxins as tools to investigate the molecular events of neurotransmitter release. Sem. Cell Biol. 5, 221–229. 6. Pellizari, R., Rossetto, O., Schiavo, G., and Montecucco, C. (1999) Tetanus and botulinum neurotoxins: mechanism of action and therapeutic uses. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 354, 259–268. 7. Kessler, K. R. and Benecke, R. (1997) Botulinum toxin: from poison to remedy. Neurotoxicology 18, 761–770. 8. Singh, B. R., Li, B., and Read, D. (1995) Botulinum versus tetanus neurotoxins: why botulinum neurotoxin but not tetanus neurotoxin a food poison? Toxicon 33, 1541–1547. 9. Hatheway, C. L. (1995) Botulism: the present status of the disease. Curr. Top. Microbiol. Immunol. 195, 55–75. 10. Smith, L. D. and Sugiyama, H. (1988) Botulism: The Organism, Its Toxins, The Disease, 2nd ed. Thomas, Springfield, IL. 11. Ferrari, N. D. and Weisse, M. E. (1995) Botulism, in Advances in Pediatric Infectious Diseases, vol. 10, Mosby Year Book, Inc., NY, pp. 81–91.

302

Poli and Lebeda

12. Hughes, J. M., Blumenthal, J. R., Merson, M. H., Lombard, G. L., Dowell, V. R., and Gangarosa, E. J. (1981) Clinical features of type A and B food-borne botulism. Ann. Intern. Med. 95, 442–445. 13. Thompson, J. A., Glascow, L. A., and Warpinski, J. R. (1980) Infant botulism: clinical spectrum and epidemiology. Pediatrics 66, 936–942. 14. Arnon, S. S. (1992) Infant botulism, in Textbook of Pediatric Infectious Diseases, 3rd ed. (Feigen, R. D and Cherry, J. D., eds.), W.B. Saunders, Philadelphia, PA. 15. Brown, L. (1984) Infant botulism. Ped. Ann. 13, 139–148. 16. Schreiner, M. S., Field, E., and Ruddy, R. (1991) Infant botulism: a review of 12 years experience at the Children’s Hospital of Philadelphia. Pediatrics 66, 159–165. 17. Bleck, T.P. (1995) Clostridium botulinum, in Principles and Practice of Infectious Disease (Mandell, G. L., Bennett, J. E., and Dolin, R., eds.), Churchill Livingstone, NY, pp. 2178–2182. 18. Keller, J. E., Neale, E. A., Oyler, G., and Adler, M. (1999) Persistence of botulinum neurotoxin action in cultured spinal cord cells. FEBS Lett. 456, 137–142. 19. O’Sullivan, G. A., Mohammed, N., Foran, P. G., Lawrence, G. W., and Dolly, J. O. (1999) Rescue of exocytosis in botulinum toxin A-poisoned chromaffin cells by expression of cleavage-resistant SNAP–25. J. Biol. Chem. 274, 36,897–36,904. 20. Nicolaier, A. (1884) Über infectiösen tetanus. Dtsch. Med. Wochenschr. 10, 842–844. 21. Behring, E. and Kitasato, S. (1890) Über das Zustandekommen der Diptherie-Immunität und der Tetanus-Immunität bei Thieren. Dtsch. Med. Wochenschr. 16, 1113–1114. 22. Bleck, T. P. (1995) Clostridium botulinum, in Principles and Practice of Infectious Disease (Mandell, G. L., Bennett, J. E., and Dolin, R. eds.), Churchill Livingstone, NY, pp. 2173–2178. 23. Edmonton, R. S. and Flowers, M. M. W. (1979) Intensive care in tetanus: management, complications, and mortality in 100 patients. BMJ 1979, 1401–1404. 24. Traverso, H., Bennett, J.V ., and Kahn, A. J. (1989) Ghee application to the umbilical cord: a risk factor for neonatal tetanus. Lancet 1, 486–488. 25. Bleck, T. P. (1989) Clinical aspects of tetanus, in Botulinum Neurotoxin and Tetanus Toxin (Simpson, L. L., ed.), Academic Press, San Diego, CA, pp. 379–398. 26. Anlar, B., Yalaz, K., and Dizman, R. (1989) Long-term prognosis after neonatal tetanus. Dev. Med. Child. Neurol. 31, 76–80. 27. Daniels-Holgate, P. U. and Dolly, J. O. (1996) Productive and nonproductive binding of botulinum neurotoxin A to motor nerve endings are distinguished by its heavy chain. J. Neurosci. Res. 44, 263–271. 28. Atassi, M. Z. and Oshima, M. (1999) Structure, activity, and immune (T and B cell) recognition of botulinum neurotoxins. Crit. Rev. Immunol. 19, 219–260. 29. Montecucco, C. and Schiavo, G. (1994) Mechanism of action of tetanus and botulinum neurotoxins. Mol. Microbiol. 13, 1–8. 30. Lebeda, F. J. and Singh, B. R. (1999) Membrane channel activity and translocation of tetanus and botulinum neurotoxins. J. Toxicol. Tox. Rev. 18(1), 45–76. 31. Donovan, J. J. and Middlebrook, J. L. (1986) Ion-conducting channels produced by botulinum toxin in planar lipid membranes. Biochemistry 5, 2872–2876. 32. Williamson, L. C., Halpern, J. L., Montecucco, C., Brown, J. E., and Neale, E. A. (1996) Clostridial neurotoxins and substrate proteolysis in intact neurons. J. Biol. Chem. 271, 7694–7699. 33. Foran, P., Lawrence, G. W., Shone, C. C., Foster, K. A., and Dolly, J. O. (1996) Botulinum neurotoxin C1 cleaves both syntaxin and SNAP–25 in intact and permeabilized chromaffin cells: correlation with its blockade of catecholamine release. Biochemistry 35, 2630–2636. 34. Dreyer, F. (1989) Peripheral actions of tetanus toxin, in Botulinum Neurotoxin and Tetanus Toxin (Simpson, L. L., ed.), Academic Press, San Diego, CA, pp. 179–203.


303

35. Bizzini, B. (1989) Axoplasmic transport and transsynaptic movement of tetanus toxin, in Botulinum Neurotoxin and Tetanus Toxin (Simpson, L. L., ed.) Academic Press, San Diego, CA, pp. 204–231. 36. Collingridge, G. L. and Herron, C. E. (1985) Effects of tetanus toxin on GABA synapses in the mammalian central nervous system, in Seventh International Conference on Tetanus (Nistico, G., Mastroeni, P., and Pitzurra, M., eds.), Gangemi, Rome, pp. 127–142. 37. Umland, T. C., Wingert, L. M., Swaminathan, S., Furey, W. F., Schmidt, J. J., and Sax, M. (1997) The structure of the receptor binding fragment HC of tetanus neurotoxin. Nature Struct. Biol. 4, 788–792. 38. Lacey, D. B., Tepp, W., Cohen, A. C., DasGupta, B. R., and Stevens, R. C. (1998) Crystal structure of botulinum neurotoxin type A and implications for toxicity. Nature Struct. Biol. 5, 898–902. 39. Lacey, D. B. and Stevens, R. C. (1999) Sequence homology and structural analysis of the clostridial neurotoxins. J. Mol. Biol. 291, 1091–1104. 40. Jongeneel, C. V., Bouvier, J., and Bairoch, A. (1989) A unique signature identifies a family of zinc-dependent metallopeptidases. FEBS Lett. 242, 211–214. 41. Schiavo, G., Poulain, B., Rossetto, O., Benfenati, F., Tauc, L., and Montecucco, C. (1992) Tetanus toxin is a zinc protein and its inhibition of neurotransmitter release and protease activity depend on zinc. EMBO J. 11, 3577–3583. 42. Schmidt, J. J. and Bostian, K. A. (1995) Proteolysis of synthetic peptides by type A botulinum neurotoxin. J. Prot. Chem. 14, 703–708. 43. Lacey, D. B. and Stevens, R. C. (1998) Unraveling the structures and modes of action of bacterial toxins Curr. Opin. Struc. Biol. 8, 778–784. 44. Obrdlik, P., Neuhaus, G., and Merkle, T. (2000) Plant heterotrimeric G protein subunit is associated with membranes via protein interactions involving coiled coil formation. FEBS Lett. 476, 208–212. 45. Eisel, U., Jarausch, W., Foretzki, K., Henschen, A., Engels, J., Weller, U., et al. (1986) Tetanus toxin: primary structure, expression in E. coli, and homology with botulinum toxins. EMBO J. 5, 2495–2501. 46. Singh, B. R. and Be, X. (1992) Use of sequence hydrophobic moment to analyze membrane interacting domains of botulinum, tetanus, and other toxins, in Techniques in Protein Chemistry III (Angeletti, R. H., ed.), Academic Press, Orlando, FL, pp. 373–383. 47. Lebeda, F. J. and Olsen, M. A. (1994) Secondary structural predictions for the clostridial neurotoxins. Proteins Struc. Func. Gen. 20, 293–300. 48. Lebeda, F. J. and Olsen, M. A. (1995) Structural predictions of the channel-forming region of the botulinum neurotoxin heavy chain. Toxicon 35, 559–567. 49. Montal, M. S., Blewitt, R., Tomich, J. M., and Montal, M. (1992) Identification of an ion channel-forming motif in the primary structure of tetanus and botulinum neurotoxins. FEBS Lett. 313, 12–18. 50. Oblatt-Montal, M., Yamazaki, M., Nelson, R., and Montal, M. (1995) Formation of ion channels in lipid bilayers by a peptide with the predicted transmembrane sequence of botulinum neurotoxin A. Prot. Sci. 4, 1490–1497. 51. Hoch, D. H., Romero-Mira, M., Ehrlich, B. E., Finkelstein, A., DasGupta, B. R., and Simpson, L. L. (1985) Channels formed by botulinum, tetanus, and diptheria toxins in planar lipid bilayers: relevance to translocation of proteins across membranes. Proc. Nat. Acad. Sci. USA 82, 1692–1696. 52. Byrne, M. P., Montal, M., Canaves, J., and Lebeda, F. J. (1998) Conformational changes of a channel-forming peptide from the translocation domain of botulinum neurotoxin as directed by circular dichroism. Prot. Sci. 7(Suppl. 1), 361–T. 53. Swaminathan, S. and Eswaramoothy, S. (1999) Crystal structure of botulinum neurotoxin type B. Abstract presented at The Movement Disorder Society’s International Conference 1999: Basic Therapeutic Aspects of Botulinum and Tetanus Toxins, Orlando, FL.

304

Poli and Lebeda

54. Martin, L., Cornille, F., Turcaud, S., Muedal, H., Roques, B. P., and Fournie-Zaluski, M-C. (1999) Metallopeptidase inhibitors of tetanus toxin: a combinatorial approach. J. Med. Chem. 42, 515–525. 55. Christine, A., Turcaud, S., Cornille, F., Fournie-Zaluski, M. C., and Roques, B. P. (1999) Structure-based design of the first highly-potent inhibitors of botulinum B neurotoxin. Abstract presented at The Movement Disorder Society’s International Conference 1999: Basic and Therapeutic Aspects of Botulinum and Tetanus Toxins, Orlando, FL. 56. Singh, B. R., Cai, S., LI, L., Sharma, S. K., Shukla, H. D., Lomneth, R., and Sarkar, H. K. (1999) Molecular basis of the unique endopeptidase activity of botulinum neurotoxin: role of dynamic protein structure. Abstract presented at The Movement Disorder Society’s International Conference 1999: Basic and Therapeutic Aspects of Botulinum and Tetanus Toxins, Orlando, FL. 57. Halpern, J. L. and Neale, E. A. (1995) Neurospecific binding, internalization, and retrograde axonal transport. Curr. Top. Microbiol. Immunol 195, 221–241. 58. Humphrey, W., Dalke, A., and Schulten, K. (1996) VMD: visual molecular dynamics. J. Mol. Graphics 14, 1091–1104. 59. Varshney, A., Brooks, F. P., and Wright, W. V. (1994) Linearly scalable computation of smooth molecular surfaces. IEEE Comp. Graphics Applic. 14, 19. 60. Persistence of Vision Ray TracerTM (POV-RayTM) (http://www.povray.org).

Botulinal Neurotoxins

305

17 Molecular Mechanism of Action of Botulinal Neurotoxins and the Synaptic Remodeling They Induce In Vivo at the Skeletal Neuromuscular Junction Frédéric A. Meunier, Judit Herreros, Giampietro Schiavo, Bernard Poulain, and Jordi Molgó

1. INTRODUCTION Botulinal neurotoxins (BoNTs) have long been known to have potent and specific paralytic effects at the vertebrate neuromuscular junction (NMJ). Although they are the most toxic substances known, the serotype A is now being used for therapeutic purposes, mainly to treat involuntary muscle contractions, but also for a number of other medical applications (reviewed in ref. 1). During the last decade, the most significant milestone discoveries have paved the way in this field starting with the discovery of their metalloprotease activity targeting key components of the exocytotic machinery (reviewed in ref. 2), and they culminated with the elucidation of the crystal structure of BoNT/A and /B (3–5). BoNTs, when injected in the vicinity of a muscle, reach the motor-nerve terminal cytosol by a sequence of several steps (reviewed in refs. 6–8). The toxins’ metalloproteolytic activity blocks acetylcholine (ACh) release and produces a profound but transient skeletalmuscle paralysis in vivo (reviewed in refs. 2,9,10). In addition, BoNTs trigger a pronounced outgrowth along intramuscular axons and nerve terminals and a remodeling of the NMJ, which contributes to the eventual functional recovery of neuromuscular transmission. These phenomena provide an excellent model for analyzing the cellular and molecular interactions involved in the plasticity of synaptic contacts. This review highlights recent progress in our understanding of the structural organization of BoNTs sheltered in their molecular complexes and as single neurotoxins. Also, the review focuses on the BoNTs’ functional domains required for binding, internalization, translocation into the nerve terminal cytosol, and intracellular protease activity responsible for the blockade of neurotransmitter release. Finally, we will discuss how, after the initial blockade of quantal ACh release leading to a relentless paralysis, the neuromuscular system through pre- and postsynaptic cellular and molecular interactions initiates NMJ remodeling, which culminates in full recovery of efficient neuromuscular transmission.


305

306

Meunier et al.

Fig. 1. BoNTs are produced as multimeric progenitor toxins. The neurotoxin moiety (S, M r ≈ 150 kDa, coefficient of sedimentation 7S) associates with a large, nontoxic, nonhaemagglutinin protein (NTNH). This dimeric complex (Mr ≈ 300 kDa, 12S) serves as a scaffold for the assembly of larger complexes formed by the incorporation of three types of additional proteins. These nontoxic components have haemagglutinin (HA) activity and have Mr of 17 kDa (HA17), 35 kDa (HA35), and 70 kDa (HA70) (14,15). These additional forms are the large “L” (16S, ≈ 500 kDa) and the extra-large “LL” (19S, ≈ 900 kDa) progenitor toxins. HA70 is present in the progenitor toxins in a cleaved state (19 and 52 kDa). Additional proteolytic cleavage of the NTNH occurs only after it is assembled in the M complex (yielding 13 and 103 kDa fragments). The molar ratios of the various components are indicated.

2. STRUCTURAL ORGANIZATION AND FUNCTIONAL DOMAINS OF BONTs The primary structure of the seven distinct serotypes of BoNTs, as well as that of the closely related tetanus neurotoxin (TeNT), lacks any leader sequence, a fact that determines the retention of the neurotoxins in the bacterial cytosol. Therefore, their release into the culture medium follows bacterial lysis. The secreted form differs among the various clostridial neurotoxins (CNTs). No accessory protein is associated with TeNT; however, the BoNTs form multimeric complexes, termed progenitor toxins, with nontoxic proteins encoded by genes present in the neurotoxin locus (11–13). A common component of these complexes is a large, nontoxic, nonhaemagglutinin protein of 119 kDa (NTNH) (Fig. 1), which is encoded by a gene upstream to the BoNT locus (13). NTNHs produced by different neurotoxigenic strains of clostridia are more conserved than the corresponding BoNTs. Interestingly, the N-terminal region of NTNH is homologous to the corresponding portion of BoNT (11). Although the significance of this


307

finding is presently unclear, it is tempting to speculate that this homology is at the basis of the intrinsic tendency of BoNTs and NTNHs to dimerize, leading to the formation of a core BoNT-NTNH complex of 300 kDa (M, 12S) (Fig. 1). This dimeric complex serves as a scaffold for the assembly of larger complexes formed by the incorporation of three proteins possessing haemagglutinin (HA) activity: HA of 17 kDa (HA17), HA of 35 kDa (HA34), and HA of 70 kDa (HA70) (14,15). These additional forms are the large L (16S, ≈ 500 kDa) and the extra-large LL (19S, ≈ 900 kDa) progenitor toxins (Fig. 1). The molar ratio between the different components of the L and LL complexes is identical in the two progenitor toxins with the remarkable exception of HA34, which is 4 in the L form and 8 in the LL form. This indicates that the 19S complex (LL) is a dimer of the 16S (L) complex cross-linked by additional HA35 components (Fig. 1) (12). Although high-resolution, three-dimensional structures of the progenitor toxins are not yet available, two-dimensional electron-density analysis of the 19S complex of BoNT/A revealed a triangularly-shaped protein core containing six lobes from which two additional smaller structures protrude, forming an elongated particle of 220 Α (16). Several remarkable features are becoming apparent from the genetic analysis of Clostridium strains bearing the neurotoxin locus (11–13). One of the most striking is that the neurotoxin genes are characterized by high mobility. Accordingly, nontoxigenic strains cocultivated with toxigenic strains can become toxigenic by gene transfer mediated by serotype-specific mobile elements, such as bacteriophages, plasmids, or transposons (13,17). As a consequence of such genetic mobility, C. botulinum may harbor more than one toxin gene (17,18). In multiple toxin-bearing strains, the neurotoxins are expressed in different proportions, with one serotype largely prevailing over the others (17). In some cases, one of the toxin genes may contain mutations and, therefore be silent (17). Strains producing mosaic BoNTs with type C and type D mixed elements have also been recently characterized (19,20). Progenitor toxins are more stable than the isolated neurotoxins to proteolysis and physical and chemical denaturation (21,22), and this stability could serve in protecting the corresponding BoNTs from the harsh conditions of the stomach. Once reaching the intestines, progenitor toxins are dissociated by the slightly alkaline pH, which releases the neurotoxin (S complex, 150 kDa) (Fig. 1). BoNT is then transcytosed to the mucosal side of the intestinal epithelium (23), from which it distributes systemically to NMJs. Haemagglutinins are not required for intestinal absorption, since pure neurotoxin can reach the general circulation (23,24). Interestingly, neurotoxin binding and transcytosis in a human gut epithelial cell line was limited to serotypes A and B, the main serotypes responsible for oral botulism in humans (23). The length of the polypeptide chains of CNTs varies from the 1251 amino acid residues of C. butyricum BoNT/E to the 1315 residues of TeNT (11,13,25). They are synthesized as inactive single chains of 150 kDa, which are activated by proteolysis within a surface-exposed loop (Fig. 2). Several bacterial and tissue proteinases are able to generate the active di-chain neurotoxin (26,27). The heavy (H) chain (100 kDa) and the light (L) chain (50 kDa) remain associated via noncovalent protein-protein interactions and via a conserved interchain

308

Meunier et al.

Fig. 2. Mechanism of activation of CNTs. BoNTs are synthesised by the bacteria as inactive single polypeptides having Mr of ≈ 150 kDa, and they are activated by proteolysis of an exposed loop (scissors). The cleavage generates an active di-chain neurotoxin composed of the L- (50 kDa) and the H- (100 kDa) chains held together by a single disulphide bond and noncovalent forces. In this form, the catalytic site is hindered by an extension of the H-chain (belt), which wraps around the L-chain. Reduction of the disulphide bond sets free the L-chain by exposing the zinc-endopeptidase active site of the L-chain.

S-S bond whose integrity is essential for neurotoxicity (28,29). The exact length of the L- and H-chains depends on the site of proteolytic cleavage within the exposed loop, and they range in size from 419–449 residues for the L chains and from 829–857 residues for the H-chain. Another site for preferential proteolysis is located in the middle of the H-chain. Treatment of the neurotoxins with papain generates a carboxy-terminal fragment (HC) and a nontoxic heterodimer composed of the L-chain and the aminoterminal portion (HN) of the H-chain (30). As more sequences of BoNTs are determined, it appears that their subdivision into seven immunologically distinct types is


309

not adequate to describe their diversity. Very relevant sequence variations are present within the same BoNT serotype (13,31), and hybrid toxins have been reported (19,20). H- and L-chains contain homologous segments separated by regions of little or no similarity. The most conserved portions of the L-chains are located at the N-terminus and in the central part of the molecule (BoNT/A, residues 216–235). The latter region contains the consensus motif of zinc-endopeptidases (HExxH) (32–35), and this observation led to the demonstration that CNTs are zinc-containing proteins. Chelators remove bound zinc and generate inactive apo-neurotoxins, but the active site metal atom can be reacquired by incubation in zinc-containing buffers to reform the active holotoxin. With the same procedure, the active-site zinc atom can be exchanged with other divalent transition-metal ions forming active metal-substituted toxins (36; reviewed in ref. 2). The crystallographic structures of BoNT/A and the carboxy-terminal TeNT domain (HC) have been determined at 3.3 Å and 1.5 Å resolution, respectively (3,37–40). BoNTs consist of three ≈50 kDa domains: an amino-terminal domain endowed with zinc-endopeptidase activity, a membrane-translocation domain characterized by long pairs of kinked α-helices, and a carboxy-terminal binding portion similar to that in legume lectins and the Kunitz-type trypsin inhibitors (3,5,40). Such structural organization mirrors the four-step mechanism followed by CNTs to intoxicate neurons; i.e., (1) binding, (2) internalization, (3) membrane translocation, and (4) target modification (6,41,42). The L-chain is responsible for the intracellular catalytic activity (43–48), the amino-terminal 50 kDa domain of the H chain (HN) is implicated in membrane-translocation (49–57), and the carboxy-terminal part (HC) is mainly responsible for the neurospecific binding (58–60). The HC domains of TeNT and BoNT/A are structurally very similar and consist of two distinct subdomains, the amino-terminal half (HCN) and the carboxy-terminal half (HCC). HCN is enriched in ß-strands arranged in a jelly-roll motif closely similar to that of legume lectins, which are carbohydrate-binding proteins. The HCC has a modified ß-trefoil fold, which has been found in several proteins involved in recognition and binding functions such as fibroblast growth factor (FGF) and Kunitz-type trypsin inhibitors. Recombinant HCC and HCN interact in solution and form heterodimers (61). In contrast to that of the HCN domain, the sequence of HCC is poorly conserved among different CNTs. Removal of HCN does not alter the neurospecific binding of HC (61), whereas the carboxy-terminus plays an active role in CNT binding (62, J. Herreros, unpublished results). The importance of the last 34 residues of HCC (and, in particular, of H1293) for polysialoganglioside binding was demonstrated in TeNT by photoaffinity labeling (63). Recently, the functional analysis of this region has been extended to BoNT/E using a monoclonal antibody (MAb) directed against a carboxy-terminal epitope containing a homologous histidine (H1227). This antibody neutralizes completely BoNT/E toxicity (64). In BoNT/A, ganglioside binding to the same region has been inferred by fluorescence-quenching analysis of W1265, the only solvent-accessible tryptophan in this serotype (40,65). The structure of this segment greatly differs between TeNT and BoNT/A. While in BoNT/A it creates a large positively-charged cleft between the two subdomains, the loop in TeNT folds in, generating a more neutral, shallow space (40). These differences are likely to be at the basis of the distinct

310

Meunier et al.

surface-recognition characterizing the binding of TeNT and BoNTs to neuronal cells. Very recently, the structure of crystals of the HC fragment of TeNT soaked with different carbohydrates has been determined (39). Its analysis revealed multiple oligosaccharide-binding sites not restricted to the extreme carboxy-terminal portion, but all within the HCC domain, which suggests that this domain could be involved in multiple interactions involving glycosylated lipid and protein receptors on the neuronal surface. Accordingly, the HCC portion of TeNT has been recently reported to be necessary and sufficient for binding to both polysialogangliosides and a putative glycoprotein receptor on nerve-growth factor-differentiated PC12 cells. The HN domains are highly homologous among the various CNTs, with the exception of a 50 amino acid region, which wraps around the catalytic domain and is called the “wrapping belt.” This high similarity suggests that the core domain of different HNs may have a closely similar three-dimensional structure (11,67), a hypothesis which has been recently confirmed by the crystallization of BoNT/B (5). The membrane translocation region of BoNT/A contains a pair of twisted 105 Å-long helices reminiscent of the hairpin observed in colicin (68), and this domain forms ion channels in artificial lipid bilayers and in cell membranes (49–57,69). Several studies have dealt with the identification of short amphipathic sequences capable of spanning the membrane and generating ion channels in vitro (55,67). Interestingly, the identified segments lie just outside the two long helices observed in the translocation domain, and they have an extended-loop conformation (40). The mechanism by which a drop in pH triggers the conformational change responsible for membrane insertion and the generation of ion channels is not known, but the overall structure of HN resembles that of some viral proteins undergoing an acid-driven conformational change (70). The catalytic domain presents little similarity with related enzymes of known structure, apart from the α-helix including the HExxH consensus sequence typical of several zinc proteases (3). The catalytic zinc ion of BoNT/A is coordinated by two histidines of the motif, a water molecule bound to the adjacent glutamic acid and by E261. Another possible interaction is provided by Y365, which points in the direction of the metal atom, but remains about 5 Å away from it. This type of zinc coordination is unique among zinc-endopeptidases, as predicted on the basis of sequence differences and on the unique properties of metal-substituted TeNT (36), and it accounts for the mutagenesis studies performed on the L-chains of TeNT and BoNTs (71–73). The presence of a tyrosine residue around the active site zinc atom was anticipated on the basis of a multiple-scattering analysis of the X-ray absorption spectra of TeNT (74,75). The active site of BoNT/A is buried deep within the L-chain and is accessible to the substrate via an anionic channel via three flexible loops (40). The active site is not accessible in the intact molecule of BoNT/A because it is shielded by HN and its wrapping belt (3) and this accounts for the lack of enzymatic activity of di-chain CNTs. Reduction of the inter-chain disulphide bridge in vitro or in vivo releases the belt, which appears to loosely interact with the L-chain (40), allowing the entry of the substrate into the active pocket.


311

3. BINDING TO NERVE TERMINALS Selective binding largely determines the exquisite neuronal specificity of CNTs. BoNTs bind to presynaptic cholinergic nerve terminals (7,76–79). This specificity could be a consequence of the unrestricted accessibility of the NMJ, but not of other regions of the neuron (for example, the axonal membrane or cell bodies within the central nervous system [CNS]), to the general circulation. In fact, binding to the whole neuronal surface of primary spinal cultures (31,66,80) and to nerve terminals within different areas of the CNS (81) can be achieved in vitro. BoNTs interact in vitro and in vivo with polysialogangliosides (reviewed in refs. 25,76,82), in particular to members of the G1b series (GD1b, GT1b, and GQ1b). Upon preincubation with the neurotoxin, these lipids protect the NMJ from BoNT-dependent inhibition of neurotransmitter release. In addition, the sensitivity of cultured chromaffin cells to BoNT/A (83) is increased by pretreatment with exogenous polysialogangliosides, whereas the removal of sialic acid residues from the membrane with neuraminidase decreases BoNTs binding (30,84). The use of ganglioside knockout mice (85,86) further emphasizes the importance of these glycolipids as CNT-binding agents. In fact, the toxicity of TeNT, BoNT/A and BoNT/B in these animals is reduced (87). Accordingly, spinal-cord neurons treated with fumonisin B1, an inhibitor of sphingolipid and ganglioside synthesis, displayed null TeNT binding and were totally protected from the neurotoxin’s intracellular activity (80). These findings clearly demonstrate a primary role of polysialogangliosides in CNTs binding, possibly as a low-affinity, high-capacity concentration mechanism, regardless of the additional role of a yet-to-be-identified protein receptor(s) (see below). It is unlikely, however, that binding to polysialogangliosides totally accounts for the absolute neurospecificity of these neurotoxins. High-affinity (subnanomolar), trypsinsensitive, BoNT-binding sites were found in isolated synaptosomes (88–90), suggesting the existence of protein receptors for BoNTs. Since lectins with affinity for sialic acid antagonize the binding of BoNTs to the NMJ (91,92), their protein receptors might be glycoproteins. Taken together, these observations led to a dual lipid-protein receptor model that extends to TeNT (82). BoNTs block neuroexocytosis at peripheral-nerve terminals, whereas TeNT is retrogradely transported to the CNS and acts on inhibitory synapses (Fig. 3). Thus, the presence of different protein receptors for BoNTs and TeNT could explain their differential intracellular sorting and fates despite their mutual binding to polysialogangliosides. Specific receptors would be responsible for internalization and sorting of TeNT to an endocytic compartment undergoing retrograde axonal transport. In contrast, receptors for BoNTs would direct them to acidic vesicles, a condition that would allow the translocation of the catalytic chain into the cytosol of the NMJ. In agreement with the double lipid-protein receptor model (82), BoNT/B interacts with the intravesicular domain of the synaptic-vesicle protein synaptotagmin in the presence of gangliosides (93–95). These findings were more recently extended to BoNT/A and /E (96), thus pointing to members of the synaptotagmin family as possible candidate protein receptors for BoNTs. However, these results remain controversial since anti-synaptotagmin antibodies do not inhibit binding or antagonize the toxin’s activity at the NMJ (97). In addition, competition experiments demonstrated that dif-

312

Meunier et al.

Fig. 3. Transport and sorting of TeNT and BoNTs. After entry at motor-nerve terminals, TeNT (right) is retrogradely transported to the cell body of the motoneuron, where it is transynaptically transferred and internalized by inhibitory-nerve terminals. There, the L-chain is translocated to the cytosol and cleaves the synaptic-vesicle protein VAMP/synaptobrevin. In contrast, BoNTs (left) are sorted to endocytic compartments that stay at the periphery, allowing translocation of the L-chain and the presentation of their proteolytic activity within the NMJ.

ferent BoNT serotypes do not seem to share the same receptors (8,89). The involvement of distinct receptors (not common to other serotypes) for different BoNTs are confirmed by studies done with the NMJ of Rana pipiens, which is resistant to BoNT/B binding but binds other BoNTs (92). Toxins generally choose as receptors molecules essential for the cell’s viability, ensuring the right positioning to display their actions. Despite the proven difficulties in identifying the receptors for BoNTs, the importance of research in that area is indisputable. They should surely provide us with key molecules for the ligand-dependent endocytosis at the cholinergic synapse.


313

4. INTERNALIZATION AND TRANSLOCATION INTO THE NEURONAL CYTOSOL The L-chains of CNTs have to reach the cytosol to display their intracellular proteolytic activity. After binding to cell-surface receptors, BoNTs are internalized by receptor-mediated endocytosis. This process is temperature- and energy-dependent (78) and is enhanced by nerve stimulation (77,98). Internalization is mediated by the H-chain, and HC fragments of BoNTs and TeNT are taken up by neuronal cells (31). However, a role of H N in increasing the efficiency of CNT internalization is possible (99). Upon internalization in the mouse NMJ, BoNT/A was found in clathrin-coated and later in uncoated vesicles (77). In the case of TeNT, this matter is still controversial. Although an ultrastructural study of TeNT uptake in an intact NMJ is not available, TeNT was found in coated pits in spinal-cord cells (100). Others, however, found it in uncoated pits and vesicles at early stages (101,102) and subsequently in multivesicular bodies and lysosomes. These discrepancies may reflect differences in the cell system, toxin concentration, or time-point analyzed. In a concealing and simplified view, protein receptor-mediated endocytosis may correlate with a clathrin-coated pathway, whereas uncoated-vesicles may be involved in ganglioside-dependent endocytosis. However, different forms of clathrin-independent endocytosis seem to exist, both dynamin-dependent and independent (103,104), and it is not clear whether coated and uncoated pathways may converge (105). In addition, several studies performed with different bacterial-protein toxins suggest that binding to distinct sphingolipids can drive internalization either via the coated or uncoated pathway (106–108). These observations raise the question as to whether or not the enrolment in a given internalization route is determined by the partition of the lipid receptor to particular subdomains of the plasma membrane. In hippocampal neurons, TeNT binding and internalization occurred only after membrane depolarization. TeNT was found to colocalize with synaptic-vesicle markers (109), suggesting its internalization via synaptic-vesicle endocytosis. This result, together with the possible role of the synaptic-vesicle protein synaptotagmin as the receptor for BoNTs, would explain the early finding that synaptic activity enhances the intoxication process (98). However, binding and uptake of CNT HC fragments appears to be independent of membrane depolarization in spinal-cord cultures (31,80) and in a neuronal cell line (66). Furthermore, TeNT gets internalized in NMJ preintoxicated with BoNT/A, where synaptic-vesicle exocytosis is completely blocked (110). Thus, a more precise analysis of the recycling pathways followed by synaptotagmin and other components of the synaptic terminal is required in order to elucidate the molecular mechanism for BoNT uptake. Regardless of the nature of the internalization pathway followed by the neurotoxins, their L-chains must cross the hydrophobic barrier of the vesicle membrane in order to reach the cytosol where they display their activity. The different trafficking of TeNT and BoNT at the NMJ clearly indicates that internalization is not necessarily followed by membrane translocation into the cytosol. Therefore, internalization and membrane translocation are clearly distinct steps in the process of cell intoxication, as is the case for most bacterial toxins (111,112). Compelling evidence indicates that CNTs have to

314

Meunier et al.

be exposed to low pH for nerve intoxication to occur. Accordingly, drugs altering the acidification of the endosomal lumen block toxin action (77,109,113–115). Acidic pH does not directly activate the neurotoxin, since the introduction of a nonacid treated L-chain into the cytosol is sufficient to block exocytosis (43–48). Hence, low pH is instrumental in the process of membrane translocation of the L-chain from the vesicle lumen into the cytosol. Low pH induces BoNTs to undergo a conformational change determining the exposure of hydrophobic segments. This enables the penetration of both the H- and L-chains into the lipid membrane (reviewed in refs. 2 and 112). CNTs form cation-selective channels in planar lipid bilayers, which are formed by the oligomerization of the HN domain (50,52,69,116). There is a general consensus that these toxin channels are related to the process of translocation of the L domain, but no details are presently available regarding their high-resolution structure and the sequential steps followed by the L-chain to transfer between the two sides of the membrane. However, the translocation is effective both in neuronal and non-neuronal cells, as demonstrated by the use of a BoNT HN-L chimera, which binds to a variety of cell types (117). 5. INTRACELLULAR ACTION AND MOLECULAR TARGETS Essential clues to the mechanism of action of CNTs arose from the observation that one of the most conserved regions of the L-chain contains the putative HExxH metalloprotease zinc-binding motif, and by the demonstration of a zinc requirement for TeNT activity (34). These findings identified CNTs as a new class of zinc-endopeptidases specific for three previously identified proteins of the synapse. TeNT and BoNT/B, /D, /F and /G cleave VAMP/synaptobrevin, but each at different sites; BoNT/A and /E cleave SNAP-25 at two different positions within the carboxy-terminus, and BoNT/C cleaves both syntaxin and SNAP-25 (see ref. 2 and references therein; 118). Strikingly, TeNT and BoNT/B cleave VAMP at the same peptide bond; however, when injected into the animal, they cause the opposite symptoms of tetanus and botulism, respectively (119). This finding clearly demonstrated that the different symptoms derive from distinct sites of action rather than from a different intracellular mechanism. CNTs are phosphorylated inside the neuron and this modification enhances the proteolytic activity of the toxins as well as their cellular lifetime (120). Altogether, these findings have been exploited to develop in vitro assays of the protease activity of CNTs. These assays are essential for better standardization of the BoNT preparations used in human therapy and of the TeNT preparations used as starting material for tetanus toxoid preparation (121–126). A continuous assay based on the use of fluorescent substrates has been reported (127). Moreover, CNT activity can be probed in cells and tissues with antibodies specific for epitopes present in the intact substrates. Thus, highly sensitive assays can be performed by following the progressive loss of SNARE staining (109,128–130) and monitoring, in parallel, the block of synaptic vesicles exo-endocytosis consequent to SNARE proteolysis. Two groups of zinc-endopeptidase inhibitors are known: (1) zinc chelators and (2) active site ligands. While chelators are very effective against the CNTs, none of the inhibitors active against the other classes of zinc-endopeptidases inhibit CNTs at low concentrations (34,131). Recently, a fluorescent coumarin derivative and several


315

aminothiol derivatives of tripeptides next to the VAMP cleavage site were found to inhibit BoNT/B at high micromolar concentrations (132,133). The combinatorial chemistry approach is being applied to these derivatives (134), and novel powerful inhibitors will be designed based on the crystallographic structure of BoNT/A and BoNT/B. The goal of having new and improved therapeutic agents effective against botulism and tetanus may therefore soon be achieved. The targets of the zinc-endopeptidase activity of CNTs are proteins located on the neuronal plasma membrane (syntaxin and SNAP-25) and on synaptic vesicles (VAMP/ synaptobrevin). These proteins are characterized by an extensive polymorphism with several distinct isoforms and form a large family, called SNARE (135–138). These proteins are largely unstructured in solution, but form a ternary complex, termed the SNARE complex, which is distinguished by a high structural stability to a variety of agents, including the denaturing detergent sodium dodecyl sulfate (SDS) (139–141). Recent structural studies have shown that the SNARE complex consists of four tightly packed α-helices forming a left-handed helical bundle (142,143). This bundle retains the membrane-anchoring sequences at one end of the rod and adopts a conformation similar to that seen in several membrane-fusion segments of viral glycoproteins (144). This quadruple helical bundle derives from the association of most of the cytoplasmic domain of synaptobrevin (residues 30–96), the carboxy-terminal portion of the cytoplasmic domain of syntaxin (residues 180–262), and the amino- and carboxy-terminal segments of SNAP-25 (residues 1–83 and 120–206). SNAP-25 contributes two parallel α-helices linked by a long extended segment, which includes a quartet of palmitoylated cysteines mediating the anchoring of SNAP-25 to the synaptic membrane. This finding suggests that the SNARE complex lies parallel to the membrane surface. The SNARE-interaction regions were previously highlighted in deletion studies of individual SNARE proteins, with the exception of SNAP-25 in which only the amino-terminal portion appeared to be required (140,145,146). The N-termini of VAMP (residues 1–27) and syntaxin (residues 1–120) do not take part in the formation of the four helix bundle, and they constitute two cytoplasmic extensions of the SNARE complex. The proline-rich amino-terminal segment of VAMP is clearly implicated in exocytosis since its removal inhibits the process (147) and amino-terminal peptides inhibit neurotransmitter release (148). The SNARE complex can recruit, under suitable conditions other cytosolic proteins, such as NSF (N-ethylmaleimide-sensitive factor) and its soluble adaptor SNAPs (soluble NSF accessory proteins). NSF and SNAPs are both recognized as essential proteins for a large number of vesicular transport steps within the cell (149–151). The 20S synaptic SNARE complex is stable in the presence of nonhydrolyzable ATP analogs, but it is rapidly disassembled by NSF in the presence of ATP and Mg2+ (139,152). The low-resolution structure of this particle has been determined with electron microscopic and rotary-shadowing techniques (153,154). In the 20S particle, NSF and SNAPs occupy one end of the rod, which constitutes the α-helical core of the SNARE, and they disappear when the complex is incubated in the presence of Mg2+-ATP (153,154). The structure and properties of the SNARE complex and of its components suggest that its formation brings the membranes anchoring the SNAREs in close proximity. The free energy released during this process may be at least partially used to promote membrane-bilayer fusion (155,156). Such a model is

316

Meunier et al.

supported by the finding that synthetic liposomes with reconstituted VAMP can interact and fuse in vitro with vesicles containing SNAP-25 and syntaxin (156,158–160). The effects of CNTs on assembly and disassembly of the SNARE complex support the idea that a cycle of assembly and disassembly is a key process in exocytosis. In fact, cleavage of individual SNAREs by CNTs does not prevent SNARE-complex formation, but this association either has reduced stability (140,161) or loses its functional connection to the membrane, resulting in an impairment of neuroexocytosis. In contrast, SNAREs are resistant to CNTs when assembled in the SNARE complex (140,162). This protection is consistent with the gain in secondary structure experienced by the SNAREs during complex formation (142,143). Compelling evidence supports the idea that CNT-induced proteolysis of VAMP, SNAP-25, or syntaxin is responsible for the inhibition of neurotransmitter release (reviewed in refs. 2,9,10 and 118). Mutations in the cleavage site generate SNARE proteins that are resistant to CNT cleavage and support regulated secretion (163). Surprisingly, full inhibition of neurotransmitter release is not accompanied by the complete proteolysis of the SNARE proteins at the nerve terminal (128–130,164,165). This finding suggests the existence within the nerve terminal of different pools of SNARE proteins characterized by different CNT sensitivity and with different functional roles in neuroexocytosis. Accordingly, SNARE molecules actively engaged in the release process are substrates for CNTs and it is their cleavage that likely determines the blockade of neurotransmitter release (166). The elucidation of the CNTs’ cleavage sites on the three SNARE proteins reveals no conserved patterns accounting for the absolute specificity of these proteases (reviewed in refs. 10 and 118). Hence, each CNT must differ in the spatial organization of the active site, in order to catalyze the hydrolysis of different peptide bonds. Interestingly, CNTs are able to cleave only very long peptides derived from the SNARE proteins, whereas short peptides are inhibitory (34,119,167,168). In addition, the minimal portion acting as substrate differs from serotype to serotype (124,131,146,148,167,169– 172), even in the case of CNTs acting on the same peptide bond, such as BoNT-A and TeNT. These findings indicate that CNTs recognize other determinants on their substrates in addition to the cleavage site. Such a structural feature has been identified by Montecucco and colleagues, in a nine residue-long motif designated the SNARE motif (173). This motif is characterized by the consensus sequence xh--xh-xhp (x, any amino acid; h, hydrophobic residue; -, acidic residue; p, polar residue) and multiple copies are present in VAMP, syntaxin, and SNAP-25. The SNARE motifs are included in the four-helix bundle of the synaptic SNARE complex, with their hydrophobic residues in the interior and their acidic residues in the exterior, or, in one case, in a three-helix bundle at the amino-terminus of syntaxin (142). Multiple experimental evidence supports the role of the SNARE motif in CNT recognition. Remarkably, the presence of the SNARE motif is an absolute requirement for the cleavage of peptide substrates by CNTs (124,131,167,170,171), and selected mutations of this motif alter in vitro and in vivo resistance to CNT-induced proteolysis (169,174–178). Moreover, antibodies against the SNARE motif inhibit the proteolytic activity of the neurotoxins (175). Although additional biological activities of CNTs have been reported (179,182), overwhelming evidence indicates that CNTs are zinc-endopeptidases specific for syn-


317

aptic SNARE proteins or their homologues, and that this is the primary mechanism of toxicity. One of the most convincing pieces of evidence is that active-site mutants of CNTs, devoid of metalloproteinase activity, are unable to inhibit ACh release at the rat NMJ (71,73). However, it is impossible to rule out the possibility of other substrates, or other activities of the toxins. Such alternative targets or activities are unlikely to make a major contribution to the blockade of neurotransmitter release, but under certain circumstances they might produce measurable effects on cell physiology. 6. BLOCKADE OF QUANTAL ACETYLCHOLINE RELEASE In this section we briefly discuss the effects of BoNTs on quantal transmitter release and neuromuscular transmission, and we attempt to correlate those actions with the toxins’ effects on their targeted synaptic proteins. In vivo and in vitro exposure to BoNTs causes skeletal-muscle paralysis (183) by reducing nerve impulse-evoked quantal ACh release, at levels insufficient to trigger endplate potentials (EPPs) of enough amplitude to reach the threshold for action potential generation in the muscle fiber (reviewed in refs. 2,6–10,184,185). Also, spontaneous quantal ACh release recorded as miniature endplate potentials (MEPPs) is almost completely abolished after exposure to BoNT/A (7–10,118,184–186) or to nicked-BoNT/E (187,188). The few MEPPs that persist after in vitro exposure to BoNT-A usually have smaller amplitudes than the spontaneous MEPPs recorded in untreated nerve terminals (189–192) and the decrease in the MEPPs’ mean amplitude usually causes a leftward shift in their amplitude distribution (but see ref. 193). However, this effect apparently is unrelated to alteration of the neurotransmitter content of synaptic vesicles because BoNTs apparently do not perturb ACh synthesis or ACh uptake and storage in synaptic vesicles (194). About 5 d after the onset of muscle paralysis induced by BoNT/A in vivo, spontaneous MEPPs having either small amplitudes or large amplitudes and prolonged time-topeaks are recorded from the majority of the junctions. These quantal events have been termed “Giant-MEPPs” (G-MEPPs) or slow MEPPs, and their frequency increases with time after poisoning. Giant MEPPs are produced by ACh released from motor-nerve terminals and in contrast to MEPPs, their frequency is not affected by nerve-terminal depolarization or by changes in extracellular Ca2+ (195). Moreover, G-MEPPs do not enter into the composition of EPPs evoked by nerve stimulation (196). Therefore, it has been suggested that the ACh release mechanism(s) giving rise to G-MEPPs is different from that responsible for the generation of MEPPs (195,197–201), and that “regulated” neurotransmitter release is not involved in generating G-MEPPs (201). G-MEPP frequency is prominent at a time when nerve-terminal sprouting and synaptic remodeling is prominent (184–186). Therefore, it is likely that G-MEPPs result from fusion events due to “constitutive” ACh secretion from nerve-terminal sprouts at sites where there is significant remodeling of postsynaptic nicotinic ACh receptors and acetylcholinesterases (see Subheading 7.). Interestingly, after the effect of BoNT/A subsides and functional recovery occurs, G-MEPP frequency decreases and normal MEPPs reappear (195). Blockade of evoked neurotransmitter release by BoNT/A occurs without changes in the propagation of the action potential to the nerve terminal and the ensuing influx of Ca2+ that triggers quantal ACh release (202–205). Moreover, neurotransmitter release

318

Meunier et al.

Fig. 4. Molecular targets and the exocytosis steps blocked by CNTs. Neurotransmitter exocytosis follows a sequence of steps during which free vesicles undergo docking at active zones, functional maturation or “priming” and fusion with the plasma membrane in response to Ca2+ influx (arrow). During this sequence of events, VAMP/synaptobrevin, syntaxin, and SNAP-25 assemble in complexes organized in a fusogenic ring between the vesicle and plasma membrane. When the CNTs’ targets (VAMP/synaptobrevin, syntaxin, and SNAP-25) are under an open configuration, they can be cleaved by the neurotoxins. After cleavage, the targets can assemble but the complexes formed cannot transit in subsequent steps (broken arrows), thereby preventing fusion.

continues to occur in a dispersed and random manner along the nerve-terminal arborization (206,207). These observations indicate that the number of release sites is not diminished by BoNT/A. In addition, detailed analysis of nerve-evoked transmitter release from motor-nerve endings indicates that BoNT/A and BoNT/E reduce the mean quantal content of EPPs, so that a few quanta are released in response to nerve stimulation (130,189,191,209). Increasing the release probability of BoNT/A-poisoned motor endings, by (1) raising extracellular Ca2+, (2) tetanic-nerve stimulation (187,189,191,205,207), or (3) enhancing phasic Ca2+ influx via blockade of fast K+ currents in nerve terminals with aminopyridines (4-aminopyridine [4-AP] and 3,4-diaminopyridine [3,4-DAP]) (214), strongly reduces the number of failures of release and increases the number of quanta released per nerve impulse (187,205,207,208,212,215–219). Transmitter quanta are released synchronously in the presence of aminopyridines at BoNT/A-treated motor nerve terminals and they can sum up, generating EPPs with amplitudes sufficient to depolarize the membrane over the threshold for action potential generation in the muscle fiber, which allows muscle twitching. However, recovery of neuromuscular transmission induced by aminopyridines can occur, provided that transmitter release has not been completely abolished. Although both BoNT/A and BoNT/E cleave SNAP-25, their blocking actions display several differences. For example, treatments aimed at increasing the release probability (e.g., exposure to 4-AP or 3-4 DAP) are relatively ineffective in producing recovery of neuromuscular transmission at BoNT/E-poisoned mouse (188,220) and rat NMJs (187,212).


319

After SNAP-25 proteolysis by BoNT/A, the transit of the ternary SNARE complex in the SDS-resistant state is diminished but can still occur (140). Thus, BoNT/A does not abolish irreversibly fusion per se (see Fig. 4). This may explain why the effects of BoNT/A on quantal transmitter release can be reversed by treatments that increase the release probability. Accordingly, kinetic analysis of exocytosis in chromaffin cells indicates that BoNT/A acts on a secretory-pathway step that is earlier than the step affected by BoNT/E; i.e., before fusion (166,221). Also, the observation that synaptic vesicles accumulate at active zones raises the question about which step after docking but before fusion is affected by BoNT/A. Truncation of SNAP-25 by BoNT/E destabilizes the four-helical bundle of the SNARE complex in such a way that it is not tight enough to cause fusion of the synaptic-vesicle membrane and the plasma membrane. This model is consistent with the observation that several treatments used to enhance release probability do not promote exocytosis at BoNT/E-poisoned motor-nerve terminals. Other suggestions for the step blocked by BoNT/A come from the observation that BoNT/Ainduced blockade can be reversed by treatments increasing Ca2+-influx. In this regard, at least two possibilities may be considered. First, BoNT/A may alter the interaction between truncated SNAP-25 and synaptotagmin (222), the main Ca2+ sensor of exocytosis. Although SNAP-25 truncated by BoNT/A or BoNT/E still binds to synaptotagmin (223), the possibility of allosteric modification of synaptotagmin-triggering functions can not be excluded (see also discussion in ref. 166). Second, BoNT/A may alter Snapin, a protein that binds to SNAP-25 and, thereby, could regulate the association of synaptotagmin with the SNARE complex (224). Among the BoNTs that cleave VAMP/synaptobrevin, BoNT/B produces an intense block of nerve-evoked quantal ACh release at rat (212,225) and mouse NMJs (219), and its intracellular injection into the presynaptic motor nerve axon abolishes neurotransmitter release at crayfish NMJs (226). BoNT/D also blocks evoked quantal ACh release at frog (130,227), rat, and mouse NMJs (205,228), and BoNT/F abolishes ACh release at mouse (212,218) and frog NMJs (229). However, BoNT/D is inactive at human NMJs (230). Although extracellular application of TeNT is far less potent than BoNTs in inhibiting ACh release at the NMJ (by a factor 100–1000), sublethal doses block neuromuscular transmission (208,209,219,231–233). A similar blockade occurs when TeNT is delivered into presynaptic terminals or nerve cells by intraneuronal injection (48,226,224,235), and the blockade is as efficient as that produced by BoNT/B (236). CNTs that cleave VAMP-synaptobrevin usually reduce MEPP frequency (205,208,209,218,219,225,226,231,232,237,238) with an efficiency that depends on the VAMP isoforms present in the various animal species. Blockade induced by TeNT and by BoNT/B, /D, and /F exhibits several features that make it distinct from that induced by BoNT/A. For example, quantal transmitter release evoked by nerve impulses at NMJs poisoned with TeNT or BoNTs cleaving VAMP/synaptobrevin is characterized by a temporal dispersion of the quanta released by presynaptic depolarization (205,208,209,220,225,237). The asynchrony of quantal ACh release prevents EPPs from building up over the muscle action potential threshold and to restore neuromuscular transmission (reviewed in refs. 185 and 212). Moreover, spontaneous quantal transmitter release from motor-nerve terminals poisoned with TeNT, BoNT/B, /D, or /F is relatively insensitive to α-latrotoxin present in black widow

320

Meunier et al.

spider venom, in contrast to BoNT-A-treated terminals (219,232). A similar insensitivity to the effect of α-latrotoxin has been observed in cultured hippocampal neurons poisoned with TeNT (239). In vitro studies indicate that cleavage of VAMP-2 by TeNT, BoNT/B, /D, /F, or /G does not prevent assembly of VAMP-2, SNAP-25, and syntaxin in ternary SNARE complexes (140). However, the blockade induced by these toxins seems to have two possible scenarios. First, when VAMP is cleaved by TeNT, BoNT/B or /G, the VAMP portion (~20 amino acid residues in length), which remains on the synaptic-vesicle membrane, does not contain interaction sites for the other SNAREs. Therefore, the membrane of the synaptic vesicle is no longer linked to the SNARE complex, and fusion with the plasma membrane cannot occur. Second, when VAMP is cleaved by BoNT/D or /F, the C-terminus VAMP fragment remaining in the vesicle membrane is long enough to anchor the synaptic vesicle to the SNARE complex, but fusion cannot occur because the latter cannot transit in the thermally stable, SDS-resistant, fusogenic state. Hence, despite the appearance of docked synaptic vesicles, treatments that increase the release probability are usually unsuccessful in reversing the toxins’ action. However, this schematic view probably needs to be amended because of several recent studies suggesting that fusion of the synaptic vesicle with the plasma membrane requires a ring of SDS-resistant SNARE complexes (reviewed in ref. 10). Thus, cleavage of some of the VAMP decorating the synaptic vesicle probably can abolish exocytosis simply by disabling some of the complexes in the fusogenic ring. It is possible that the few desynchronized quanta elicited by increasing the release probability at TeNT, BoNT/B, /D or /F-treated NMJs may represent a very small fraction of the synaptic vesicles for which the incomplete SNARE ring is still fusogenic. In this regard, some of the docked vesicles observed after VAMP-cleavage might correspond to synaptic vesicles anchored to the plasma membrane by an incomplete ring of SNARE complexes (see ref. 10), as well as to vesicles tethered to the plasma membrane by synaptotagmin (223) or other proteins. BoNT/C produces very intense blockade of neurotransmitter release evoked by nerve stimulation (92,213,239–242) and it dramatically decreases spontaneous exocytosis (92,213,239). These observations raise the question of whether the observed blockade of neurotransmitter release is due to cleavage of syntaxin or SNAP-25, or both. In vitro, cleavage of SNAP-25 by BoNT/C occurs with a much lower efficiency (~1000 fold difference) than SNAP-25 cleavage by BoNT/A or /E (146,164). Also, inhibition of the exocytotic process by BoNT/C at the squid giant synapse results only from syntaxin cleavage, because squid SNAP-25 can not be cleaved by the neurotoxin (241,242). In addition, the secretory blockade produced in chromaffin cells by BoNT/C is very distinct from that produced by BoNT/A (166). Furthermore, the action of BoNT/C is very similar to the blockade of both evoked and spontaneous quantal transmitter release that characterizes Drosophila mutants lacking syntaxin (243). Thus, the secretory blockade is likely to be due to syntaxin cleavage. A possible molecular model for BoNT/C-induced inhibition of exocytosis envisages the formation of a SNARE complex that, despite its SDS resistance and its tethering to the plasma membrane via SNAP-25, has not enough potential energy to be fusogenic. On the other hand, several studies suggest that blockade of neurotransmitter release by BoNT/C is caused by SNAP-25 inactivation. For example, BoNT/C is very efficient in removing nearly all


321

SNAP-25 immunoreactivity in cultured hippocampal slices or cultured spinal neurons (129,239). This second way of blocking exocytosis (i.e., cleavage of SNAP-25) would be similar to that produced by BoNT/A (see Fig. 4). Indeed, the frequency of spontaneous quantal transmitter release after BoNT/C-treatment can be increased by high external Ca2+ and by applying an ionophore (ionomycin) or α-latrotoxin (239). 7. BONTs-INDUCED SYNAPTIC REMODELING OF THE SKELETAL NEUROMUSCULAR JUNCTION The neuromuscular system after paralysis with BoNTs provides an excellent model for analyzing synaptogenesis, synaptic maintenance, and sprouting of new nerve processes. Time-lapse imaging of mouse NMJs has revealed that most endplates are very stable during much of an animal’s life, even if junctions enlarge as the animal grows (244,245). A remarkable example of plasticity in the adult animal is the ability of intramuscular axons and nerve terminals to sprout new processes and form additional synapses in response to muscle inactivity induced by BoNTs. The selective blockade of quantal ACh release by BoNTs provides an interesting approach since paralysis occurs without any microscopic damage of the poisoned motor-nerve terminal and, advantageously, avoids physical removal of the nerve endings. Among the seven distinct serotypes of BoNTs known, BoNT/A has been the most widely used for ex vivo, in vivo, and in vitro studies of trophic interactions between the three cellular components of the NMJ (motor nerve terminals, perisynaptic Schwann cells, and skeletal-muscle fibers). Local injection of a sublethal dose of botulinum typeA toxin complex (see Subheading 2.) into the extensor or flexor musculature of the hind limbs of rodents (246), or into the immediate vicinity of the Levator auris longus (LAL) muscle of mice (247), does not cause generalized intoxication but blocks quantal ACh release and results in muscle paralysis confined to the site of toxin administration (reviewed in ref. 248). The neuromuscular block may last for several weeks and depends on both the dose and the BoNT serotype injected. These models have been used to characterize, mainly ex vivo, synaptic plasticity changes triggered by BoNTs at the adult NMJ. Neuromuscular paralysis produced by BoNT/A triggers a marked in vivo outgrowth of intramuscular axons at the nodes of Ranvier (nodal sprouting) and at motor-nerve terminals (terminal and ultraterminal sprouting) (249–262). The ability of motor nerves to sprout in rat muscles is much greater in younger (16–31 d old) than in older, adult rats (252), varies with time after injection, and markedly depends upon the muscle examined. In general, the ability of axons to grow or sprout in vivo after BoNT/A exposure is more pronounced in slow-contracting muscles than in fast-contracting muscles. Thus, more nerve-terminal sprouting is detected in the soleus muscle than in the extensor digitorum longus (EDL) muscle. Moreover, when the relation between axon length and the abundance of outgrowth from motor-nerve terminals is examined, sprouting is usually more abundant in proximal (e.g., rhomboid and paraspinous) than in distal (e.g., EDL) muscles (253). This inverse correlation between nerve length and the abundance of sprouting from nerve terminals indicates that short axons have a greater ability or potential to sprout than long axons. Nerve-terminal sprouts have been detected 2 d after BoNT/A injection into mammalian soleus muscle and, based on the time required for the neurotoxin to act, terminals

322

Meunier et al.

were calculated to sprout within 24 h of muscle inactivity. Sprouts also were seen arising from myelinated intramuscular axons at the nodes of Ranvier 2 d after BoNT/A injection. However, unlike the terminal sprouts that elongated over time, the nodal sprouts remained short and were confined by the basal lamina overlying the nodal region (254). The pattern of innervation of the LAL muscle is markedly modified after BoNT/A injection (184,255,256,258,260–262). During the first 15 d after BoNT/A injection, nerve-terminal sprouts appeared as thin, poorly branched, unmyelinated filaments usually oriented parallel to the longitudinal axis of the muscle and extending beyond the original endplate area (256). The sprouts increased in number, length, and complexity for about 50 d after the BoNT/A injection, as revealed by morphometric analysis of the nerve-terminal arborization (258). One striking finding was that sprouts continued to grow even when phasic-muscle contraction elicited by nerve stimulation resumed. These results indicate that despite recovery of normal muscular activity, there is no immediate repression of the sprout process, a finding confirmed by time-lapse imaging of identified BoNT/A-treated nerve terminals in the mouse sternomastoid muscle (260, see below). Nerve-terminal sprouting has also been detected in the LAL muscle after BoNT/D (263) and BoNT/C1 injection (264). Characteristic nodal, terminal, and ultraterminal sprouting has been reported to occur in the human orbicularis muscle, and the persistence of nerve-terminal sprouts has been documented after repeated injections of BoNT-A (265). Perisynaptic Schwann-cell processes capping motor-nerve terminals at the NMJ insulate the terminals from the environment and probably provide them with trophic sustenance (266). Alteration of the nerve terminal-Schwann cell relationship in mouse leg muscles, by X-ray irradiation prior to the injection of BoNT/A, did not prevent sprouting but prevented the maturation of newly formed terminals and the differentiation of new endplates, and markedly delayed the recovery of neuromuscular transmission (267). Also, paralysis induced by BoNT/A causes both nerve terminal and perisynaptic Schwann-cell sprouting (259). The extended Schwann-cell processes were associated with nerve sprouts and, in some cases, they were longer than the sprouts growing next to them (259). Thus, Schwann cells seem to promote extension and guidance of nerve terminals in muscles during NMJ remodeling induced by BoNT/A. The return of measurable, nerve-induced muscle contraction after botulinization, at a time when extensive sprouts have developed, strongly supports their involvement in establishing new functional synapses. According to the SNARE hypothesis, synaptic vesicles are targeted to the plasma membrane through specific interactions between the vesicle v-SNARE (VAMP/synaptobrevin) and the plasma membrane t-SNAREs, which include SNAP-25 and syntaxin. In a recent study, excised sternomastoid muscles from control and BoNT/A-injected mice were probed for VAMP and SNAP-25 using classical immunocytochemical technique, and they were imaged by laser-scanning confocal microscopy (260). In addition, the postsynaptic nicotinic acetylcholine receptors (nAChRs) were also visualized with rhodaminated α-bungarotoxin, so as to allow receptor clustering to be correlated with the emergence of sprouts. Imaging of nonintoxicated fibers double-stained for either of the two SNAREs and for nAChRs revealed apparent co-localization and labeling of SNAP-25 and VAMP in the nerve terminals was present largely within the area occupied by the postsynaptic nAChRs.


323

Some differences in the distribution of the SNAREs were also observed; e.g., VAMP staining was concentrated at the terminals, whereas SNAP-25 labeling occurred within the nerve endings and their axons. Similar double immunostaining experiments, at various time-points after BoNT/A treatment, detected SNAP-25 and VAMP in the original terminals and in the sprouts. Twenty-eight days after BoNT/A administration, at the time effective neurotransmission was restored, both SNAP-25 and VAMP were observed along the entire length of extended sprouts. Within these processes, the staining patterns of the individual SNAREs were similar to those seen a few days postintoxication. Syntaxin immunostaining was also investigated; however, the presynaptic pattern observed before injection of BoNT/A was complicated by the presence of a strong reactivity of the perisynaptic Schwann cells (Meunier, F. A., unpublished results). Although the significance of such labeling is unclear, Schwann-cell processes play an important role during synaptic remodeling, by paving the way for the extending sprouts (259,266). Elongation of such glial processes is likely to be based on a SNARE-dependent constitutive exocytosis. Detection of nAChRs in the sternomastoid muscle revealed a clear reorganization of the postsynaptic apparatus, with distinguishable patches of α-bungarotoxin staining abutting the sprouts, particularly towards their extremities, but also, to a lesser extent, along the axis of the outgrowths. During the growth and maturation of sprouts induced by BoNT/A in vivo, neurofilaments, tubulin, and the synaptic-vesicle proteins synaptophysin and synaptotagmin-II, have also been detected in motor axons and their sprouts, especially toward points of synaptic contact with muscle fibers (258,270,271). The clear-cut detection of these important proteins within the newly formed sprouts, at the time of recovery from BoNT/A-induced paralysis, establishes that the outgrowths had acquired both presynaptic key components for vesicle-mediated neurotransmitter release and postsynaptic nAChR clusters, and indicates their capacity to form effective synapses with the muscle. In this regard, electrophysiological recordings of the membranes of pre-existing motor endings and newly formed sprouts have revealed: (1) active propagation of action potentials over most of the length of the nerve-terminal arborization, (2) the presence of a Ca2+ influx upon active depolarization, and (3) Ca2+-dependent K+ currents in the terminal sprout membrane (256). Thus, these findings indicate that nerve-terminal sprouts have the molecular machinery for ACh release, and they support previous suggestions, that terminal sprouts play a role in the recovery of neuromuscular transmission after BoNT/A treatment (250,251,256,258,272). Recently, de Paiva et al. (260) found, using time-lapse imaging of the same identified NMJs during BoNT/A-induced paralysis and subsequent recovery of neuromuscular transmission, that sprouts could establish functional synaptic contacts mediating appropriate exo-endocytosis in vivo. Their study used a protocol combining: (1) the use of the fluorescent probe FM1-43, which has been shown to be a suitable marker for exo-endocytosis in nerve endings (273), and (2) repeated viewing of the same synaptic region (244,274), which allowed to follow the blockade of exocytosis in the original terminals after BoNT/A injection and to trace the in vivo appearance of synaptic vesicle recycling in the outgrowth (260) (see Fig. 5). Nerve endings imaged immediately prior to toxin injection (d 0) revealed colocalization of the neuronal vital marker 4-Di-2ASP (274) and activity-dependent staining of FM1-43. Two days after BoNT/A injec-

324 Meunier et al.

Fig. 5. Scheme highlighting the major steps occurring before and after BoNT/A treatment at motor nerve terminals. The activity-dependent uptake of FM1-43 within nerve terminals has allowed the visualisation of exo-endoytosis sites during the remodeling events and the demonstration of the major, albeit transitory, role of the sprouts in establishing new synapses responsible for early functional recovery. Rehabilitation of synaptic activity in the originally poisoned terminal signals the elimination of the sprouts.


325

tion, dye uptake was inhibited in the same identified motor-nerve terminal. However, although FM1-43 uptake at the original terminals was greatly reduced by BoNT/Atreatment, the level of staining in the sprouts dramatically increased. By d 28, at which time twitching of the sternomastoid muscle could be elicited by electrical stimulation of the nerve, the sprouts (particularly towards their endings) exhibited long expanses of activity-dependent uptake of FM1-43 (Fig. 5). The hypothesis that the sprouts are required in the recovery process is clearly supported by the finding that sprouts are the sole synaptic structures capable of undergoing exo-endocytosis when nerve-induced muscle twitching occurs and the parent terminal is devoid of activity (260). Interestingly, growth cones can release neurotransmitter in response to electrical stimulation even before they make contact with muscle fibers (275,276). Furthermore, quantal-like-stimulated ACh release from the neuronal processes of cultured motoneurons has been elegantly shown by whole-cell patch-clamp recordings from myocytes brought into contact at various positions along the processes (277). This observation suggests that axonal processes (sprouts) are capable of both evoked and spontaneous ACh release. It further indicates that mature axons lose the ability to support exocytosis after establishment of functional and definitive endplates, since no activity-dependent uptake of FM1-43 was ever detected pre-terminally. It is now very important to determine whether synaptic vesicles transported along the axon only reach their ultimate maturation stage in the nerve ending or in the growing sprouts. Recent studies have begun to explore this question (278,279) but the molecular bases of such locationdependent maturation are presently unclear. At the present time, sprouts are thought to be functional only until exo-endocytosis resumes at the original nerve terminal. Why are the original nerve terminals unable to undergo neurotransmitter release for such a long period of time (about 2 mo after the BoNT/A injection)? Since BoNT/A inhibits release by cleaving SNAP-25, the resolution of this question requires determining whether the turnover of the toxin and/or its target SNAP-25 modulates the recovery process. Thus, at least three scenarios are possible: (1) active BoNT/A remains within the original terminal and any newly synthesized SNAP-25 is cleaved and rendered not functional, (2) SNAP-25 turnover is locally impaired and newly synthesized SNAP-25 is not able to reach the plasma membrane at release sites, and (3) the two latter proposals can be reconciled by a model where fast turnover of SNAP-25 provides enough newly synthesized SNAP-25 to overcome the proteolytic activity of BoNT/A. An interesting strategy used to gain a better understanding of this issue was to coinject BoNT/A and /E (280). Using this approach, the authors found that the paralysis time-course for type-A intoxication was greatly shortened, which seems to preclude persistence of adequate BoNT/A activity within the original endplate. Indeed, if BoNT/A activity was to survive that of BoNT/E within the nerve terminals, a much longer paralysis should have been observed. In sharp contrast, the lifetime of BoNT/A significantly exceeded that of BoNT/E in spinal-cord neuronal cultures challenged with large amounts of both toxins (281). The discrepancy between the two studies suggests that the third proposed scenario may prevail; however, further experiments involving sequential injection of the two toxins are needed to rule out any uptake inhibition of BoNT/A by BoNT/E. Thus, the persistence of BoNT/A-truncated SNAP-25 (SNAP-251-197) may be one of the key factors responsible for the duration of

326

Meunier et al.

neurotransmitter-release inhibition. This remarkable characteristic has already been observed in neurosecretory cells (282) and endocrine cells (283), where overexpression of SNAP-251-197 is a potent inhibitor of stimulated secretion without addition of toxin, and the inhibition probably results from competition between SNAP-251-197 and fulllength SNAP-25 for transmitter-release sites. In that regard, it is intriguing that SNAP251-180 elicits a much shorter recovery, and it is obvious that the two types of cleaved SNAP-25 are dealt with differently. Do SNAP-251-197 and SNAP-251-180 have different turnover rates, as previously proposed (280)? In order to answer this question, it is important to keep in mind that normal turnover of full-length SNAP-25 is rapid in the optic tract and superior colliculi (284) and in cultured PC12 cells (285). Therefore, the fast recovery observed after BoNT-E intoxication probably reflects the normally rapid SNAP-251-180 replacement by the full-length molecule. This idea suggests that BoNT/Atruncated SNAP-25 not only competes with full-length SNAP-25 for a discrete number of release sites but, more importantly, that its replacement is drastically impaired once in place. Removal of 9 C-terminal residues from SNAP-25 does not affect its binary interaction with syntaxin but slightly reduces its interaction with VAMP; i.e., the formation of the ternary SNARE complex is reduced by 50% at equilibrium (140,141,286– 288). In contrast, removal of 17 residues by cleavage with BoNT/E totally inhibits the formation of the ternary complex (141). Under physiological conditions, full-length SNAP-25 operates in conjunction with the other SNAREs during regulated exocytosis. Since SNAP-251-180 cannot assemble into the ternary SNARE complex, it is probably rapidly retrieved from the plasma membrane by constitutive endocytosis coupled with retrograde transport, and is replaced by newly synthesized SNAP-25. In contrast, SNAP-251-197 can enter the ternary SNARE complex, albeit in a nonproductive manner, and competes with full-length SNAP-25 for release sites. Thus, a plausible alternative is that SNAP-251-197 may inhibit passage from the regulated exo-endocytotic pathway to the constitutive retrieval and degradation route. Monitoring the same identified living nerve terminals during BoNT/A poisoning demonstrated, for the first time, that the endplate remaining after elimination of the superfluous sprouts is at the same location and displays the same morphology as the original one (260). It was previously thought that sprouting only stops when nerveinduced muscle twitching recovers; this deduction was based on the observation that BoNT/A-induced sprouting could be prevented by direct and chronic electrical stimulation of skeletal muscle (252,289,290). However, several studies have failed to support this dogma, since sprout elongation continues well after the onset of nerve-muscle twitch recovery (184,256,258). Recently, the aforementioned approach (260) yielded convincing evidence that the trigger for sprout elimination is not the onset of nervestimulated muscle twitch recovery but the rehabilitation of the originally poisoned motor-nerve terminals. Since sprouts are not as efficient in mediating exo-endocytosis as are the parent terminals (260), it was suggested that, after reaching a threshold level of activity at the original terminal, the muscle responds either by signaling sprout elimination or by turning-off a sprouting signal. The use of other BoNT serotypes exhibiting various time-courses of paralysis should provide valuable data to determine whether the onset of sprout elimination truly correlates with endplate rehabilitation.


327

The main characteristic of muscles injected with BoNT/A is their marked reduction in mean fiber diameter, as compared to controls (246). It is interesting that local injection of BoNT/A does not appear to affect all fibers in a given muscle. Thus, muscle atrophy is usually rather heterogeneous, since some of the muscle fibers (which probably are unpoisoned) maintain their normal diameters. This phenomenon is well-documented by distribution histograms of muscle-fiber diameter of toxin-treated muscles, where the majority of fibers are atrophic, the main peak of the histogram is shifted to the left, and a small peak of normal fiber diameter persists (Faille, L., Angaut-Petit, D. and Molgó, J., unpublished results). It is striking that, despite the fact that the NMJ is structurally intact (no degenerative changes are detected in the intramuscular nerve branches or in the motor-nerve terminals), muscle fibers undergo atrophic changes indistinguishable from those observed after denervation. Other denervation-like changes observed in muscle fibers treated with BoNT/A include: the appearance of fibrillation potentials and tetrodotoxin- (TTX) resistant action potentials, a decrease in the resting membrane potential, the disappearance of extrajunctional acetylcholinesterase, and the susceptibility to innervation by foreign nerves (reviewed in refs. 291,292). Coincident with an increased sensitivity to ACh in BoNT/A-poisoned muscles, action potentials become partially resistant to the action of TTX (293), which normally blocks voltage-gated Na+ channels in muscle. Such insensitivity to TTX results from an overexpression of mRNA encoding TTX-insensitive, voltage-dependent Na+ channels (SkM2) in adult skeletal muscle, which declines when functional recovery of neuromuscular transmission occurs (294). As with denervated muscle, BoNT/A-paralyzed muscles express a high endocytotic activity restricted to the endplate region (295). Interestingly, BoNT/A-blockade delays and prevents the retraction of polyneuronal innervation and motorneuron death during development (292). These observations support the idea that the paralyzed muscle secretes factors essential for growth and survival of motoneurons. BoNT/A-induced sprouting is also associated with changes in the pattern of cholinesterase staining and in the distribution of nAChRs (reviewed in refs. 291 and 292). In BoNT/A-paralyzed muscle fibers, the density of nAChRs increases dramatically, as determined by specific binding with radiolabeled [125I]-α-bungarotoxin (296) or by rhodaminated-α-bungarotoxin staining (297). The increase in 125I-α-bungarotoxin binding sites induced by BoNT/A occurs to a lesser extent in neonatal muscles than in adult muscles (298). Each fiber in mammalian skeletal muscles contains several hundred myonuclei. Of these, a few (usually 3–8) occur in tight clusters and are constantly associated with synaptic sites (299). BoNT/A-treated, atrophic muscle fibers contain a conspicuously large number of myonuclei, which are frequently distributed in continuous chains located in the center of the muscle fiber and are, most of the time, close to the nerveterminal sprouts. These synaptic nuclei are larger and rounder than extrasynaptic nuclei, and they are transcriptionally specialized, since they express genes encoding several synaptic proteins (including subunits of the nAChR) at levels far higher than those of extrasynaptic nuclei in the same cytoplasm (300–302). As a result, mRNAs for synaptic proteins are concentrated in synaptic areas, thus allowing local synthesis of synaptic constituents. During paralysis induced by BoNT/A, the levels and spatial distributions

328

Meunier et al.

of the different subunit-specific mRNAs encoding nAChRs change. Thus, mRNA levels for the α-subunit of the nAChR increase (303) and the changes in the levels and distribution of γ- and ε-subunit-specific mRNAs (304) correlate well with the spatial appearance of functional fetal and adult nAChR channel subtypes along the muscle fibers (304,305). The formation of the postsynaptic apparatus, including the accumulation and specialization of synaptic nuclei, is controlled by the nerve. The nerve-derived signal proteoglycan agrin is required for several aspects of postsynaptic differentiation, including transcriptional specialization of synaptic nuclei (306–308). Its receptor, the muscle-specific tyrosine kinase (MuSK) is concentrated in the postsynaptic membrane (309). Postsynaptic differentiation occurs after agrin activates MuSK; however, little is known about how activation of MuSK leads to postsynaptic differentiation or how agrin interacts with other signals such as neuregulins, which have been implicated in the induction of nAChR gene expression in synaptic nuclei (310,311). The molecular mechanisms that regulate synaptic plasticity in adult NMJs treated with BoNTs are poorly understood and involve many proteins that mediate intercellular interactions during the formation, maturation, and maintenance of the NMJ. Thus, identifying the intrinsic NMJ components that control plasticity and remodeling is of prime importance. The observation that direct electric stimulation of BoNT/A-paralyzed muscle prevents nerve-terminal sprouting (252,289) strongly suggests the involvement of musclederived signaling factors. Likely candidates for the muscle-released factors are the insulin-like growth factors IGF-1 and IGF-2, since their expression is upregulated by muscle inactivity induced by either denervation or BoNT/A (312,313). Furthermore, IGF-binding proteins (IGFbp4 or IGFbp5) delivered locally to BoNT/A-paralyzed muscle prevent nerve sprouting. Also, during paralysis induced by BoNT/A, there is an increase in muscle plasminogen activator (257), a serine protease that activates plasminogen to plasmin. This protease may be responsible for the degradation of some components of the junctional basal lamina that could have a role in neuromuscular formation and plasticity. The neural growth-associated protein (GAP-43) has been involved in axonal elongation, synaptogenesis, and nerve sprouting during development and in the adult NMJ (314,315). However, the levels of GAP-43 mRNA in mouse motoneurons were little affected during sprouting triggered by BoNT-A treatment (316), and sprouting induced by BoNT-A was not impaired at NMJs of adult, GAP-43 knockout mice (317). The cytoskeleton-associated and calmodulin-binding protein CAP-23 is functionally related to GAP-43 and plays a critical role in regulating nerve sprouting and the actin cytoskeleton. CAP-23 knockout mice exhibit little or no sprouting in soleus muscle injected with BoNT-A; however, the sprouting deficit can be rescued by transgenic overexpression of either CAP-23 or GAP-43 in adult motoneurons (317). Thus, GAP43 can functionally substitute for CAP-23 in vivo, probably by promoting subplasmalemmal actin cytoskeleton accumulation. A prominent modification of nerve terminals undergoing BoNT/A-induced remodeling is the appearance of immunoreactivity for the calcitonin gene-related peptide (CGRP) (318–321), which is associated with an increased CGRP content in motoneu-


329

rons (319,321). CGRP is packaged in large dense-core vesicles and transported to motor-nerve terminals where it is released upon nerve stimulation (reviewed in ref. 9). CGRP immunoreactivity has also been detected in motor endplates after nerve-terminal sprouting has been induced by exogenous application of ciliary neurotrophic factor (CNTF) (322). These results indicate that CGRP is upregulated when motor-nerve outgrowth is induced, even in the absence of muscle paralysis or a nerve lesion. CGRP stimulates synthesis of nAChRs when applied to cultured myotubes, and it has been hypothesized that the release of the neuropeptide provides a localized signal to stimulate nAChR gene expression by subsynaptic nuclei, which results in localized nAChRs synthesis (reviewed in ref. 323). However, recent studies of mutant mice lacking α-CGRP mRNA do not appear to support this view, because synaptogenesis at their NMJs is not impaired (324,325). The neural-cell adhesion molecule (N-CAM), an abundant cell adhesion molecule expressed in adult motoneurons is also believed to be involved in the development and plasticity of the NMJ. In rats, antibodies against N-CAM decrease the extent of nerveterminal sprouting after paralysis of the muscle with BoNT/A (326). Expression of N-CAM at a time when endogenous mouse N-CAM is absent from the myofiber (e.g., postnatally), results in a significant number of NMJs (about 20%) displaying extensive intraterminal sprouting, despite the absence of alterations in neuromuscular transmission (327). Also, sprouting in response to paralysis induced by BoNT/A is enhanced in the transgenic animals. In addition, N-CAM and laminin-1 (composed of α1, ß1, and γ1 chains) immunoreactivities are associated with nerve sprouts induced by BoNT/A (261). Taken together, these observations support the view that N-CAM regulates nerve-muscle interaction. Tenascin-C, a glycoprotein component of the extracellular matrix (ECM) (328), mediates neuron-glial interactions and regulates neurite extension and retraction during development (329). In the adult, tenascin-C is found in myelinating Schwann cells, at the nodes of Ranvier, and in NMJ-associated, perisynaptic, nonmyelinating Schwann cells (330). Also, BoNT/A induces a significantly smaller and delayed sprouting response in mutant mice deficient in tenascin-C than it does in normal animals (331,332). Thus, it is likely that tenascin-C is involved in both stabilization and plasticity of the NMJ during the action of BoNT-A. 8. CONCLUSIONS The detailed analysis of BoNTs’ mechanisms of action has markedly improved our understanding of the neurotransmitter release processes, in particular neuroexocytosis, and has greatly promoted their use as tools in neurobiology. Several synaptic mechanisms are now being explored with the help of these neurotoxins, and these studies will certainly yield new insights into synaptic-vesicle trafficking within synapses. The elucidation of the tridimensional structure of BoNT/A has focused interest on the versatility of the molecule, which intrinsically retains cholinergic specificity for binding, translocation, and delivery of its metalloproteolytic activity into motor-nerve terminals. Recent studies have begun to compare the structures of other BoNT serotypes, and they will lead to the elucidation of the molecular mechanisms underlying their selectivity of action. More interestingly, these studies could pave the way for a

330

Meunier et al.

novel type of neuropharmacologic studies using the refined structural features of CNTs as neurospecific carriers for the central and/or peripheral nervous systems. Considerable evidence indicates that BoNTs promote synaptic remodeling at the NMJ. However, the molecular mediators of synaptic-plasticity changes and the mechanisms involved remain intriguingly elusive. Their identification remains an important challenge for future research in this field and a potential pharmacological target for diseases affecting the integrity of NMJs. ACKNOWLEDGMENTS We are indebted to Dr. Arnold S. Kreger (University of Maryland School of Medicine) for his assistance with the copyediting of the manuscript. The authors’ studies of botulinal neurotoxins were supported by research grants from The European Commission Biotechnology Program (grant BC104CT965119 to F. A. M.), Human Frontier Science Program (J. H.), Imperial Cancer Research Fund (G. S.), Association Française de Lutte contre les Myopathies (B. P. and J. M.), and DSP (grant 01 34 029 to J. M.). REFERENCES 1. Johnson, E. A. (1999) Clostridial toxins as therapeutic agents: benefits of nature’s most toxic proteins. Annu. Rev. Microbiol. 53, 551–575. 2. Schiavo, G., Matteoli, M., and Montecucco, C. (2000) Neurotoxins affecting neuroexocytosis. Physiol. Rev. 80, 717–766. 3. Lacy, D. B., Tepp, W., Cohen, A. C., DasGupta, B. R., and Stevens, R. C. (1998) Crystal structure of botulinum neurotoxin type A and implications for toxicity. Nat. Struct. Biol. 5, 898–902. 4. Hanson, M. A. and Stevens, R. C. (2000) Cocrystal structure of synaptobrevin-II bound to botulinum neurotoxin type B at 2.0 A resolution. Nat. Struct. Biol. 7, 687–692. 5. Swaminathan, S. and Eswaramoorthy, S. (2000) Structural analysis of the catalytic and binding sites of Clostridium botulinum neurotoxin B. Nat. Struct. Biol. 7, 693–699. 6. Simpson, L. L. (1986) Molecular pharmacology of botulinum toxin and tetanus toxin. Annu. Rev. Pharmacol. Toxicol. 26, 427–453. 7. Simpson, L. L. (ed.) (1989) Botulinum Neurotoxin and Tetanus Toxin. Academic Press, San Diego. 8. Habermann, E. and Dreyer, F. (1986) Clostridial neurotoxins: handling and action at the cellular and molecular level. Curr. Top. Microbiol. Immunol. 129, 93–179. 9. Van der Kloot, W. and Molgó, J. (1994) Quantal acetylcholine release at the vertebrate neuromuscular junction. Physiol. Rev. 74, 915–926. 10. Humeau, Y., Doussau, F., Grant, N. J., and Poulain, B. (2000) How botulinum and tetanus neurotoxins block neurotransmitter release. Biochimie 82, 427–446. 11. Minton, N. P. (1995) Molecular genetics of clostridial neurotoxins. Curr. Top. Microbiol. Immunol. 195, 161–194. 12. Inoue, K., Fujinaga, Y., Watanabe, T., et al. (1996) Molecular composition of Clostridium botulinum type A progenitor toxins. Infect. Immun. 64, 1589–1594. 13. Popoff, M. R. and Marvaud, J.-C. (1999) Structural and genomic features of clostridial neurotoxins, in The Comprehensive Sourcebook of Bacterial Protein Toxins (Alouf, J. E. and Freer, J. H., eds.), Academic Press, London, pp. 174–201. 14. Fujita, R., Fujinaga, Y., Inoue, K., Nakajima, H., Kumon, H., and Oguma, K. (1995) Molecular characterization of two forms of nontoxic-nonhemagglutinin components of Clostridium botulinum type A progenitor toxins. FEBS Lett. 376, 41–44. 15. Henderson, I., Whelan, S. M., Davis, T. O., and Minton, N. P. (1996) Genetic


16.

17.

18.

19.

20.

21. 22.

23. 24.

25.

26.

27.

28. 29.

30.

31.

32.

331

characterisation of the botulinum toxin complex of Clostridium botulinum strain NCTC 2916. FEMS Microbiol. Lett. 140, 151–158. Burkard, F., Chen, F., Kuziemko, G. M., and Stevens, R. C. (1997) Electron-density projection map of the botulinum neurotoxin 900-kilodalton complex by electron crystallography. J. Struct. Biol. 120, 78–84. Henderson, I., Davis, T., Elmore, M., and Minton, N. (1997) The genetic basis of toxin production in Clostridium botulinum and Clostridium tetani, in The Clostridia: Molecular Biology and Pathogenesis (Rood, I., ed.), Academic Press, New York, pp. 261–294. Hutson, R. A., Zhou, Y., Collins, M. D., Johnson, E. A., Hatheway, C. L., and Sugiyama, H. (1996) Genetic characterization of Clostridium botulinum type A containing silent type B neurotoxin gene sequences. J. Biol. Chem. 271, 10,786–10,792. Moriishi, K., Koura, M., Fujii, N., et al. (1996) Molecular cloning of the gene encoding the mosaic neurotoxin, composed of parts of botulinum neurotoxin types C1 and D, and PCR detection of this gene from Clostridium botulinum type C organisms. Appl. Environ. Microbiol. 62, 662–667. Moriishi, K., Koura, M., Abe, N., et al. (1996) Mosaic structures of neurotoxins produced from Clostridium botulinum types C and D organisms. Biochim. Biophys. Acta 1307, 123–126. Sakaguchi, G. (1983) Clostridium botulinum toxins. Pharmac. Ther. 19, 165–194. Chen, F., Kuziemko, G. M., and Stevens, R. C. (1998) Biophysical characterization of the stability of the 150-kilodalton botulinum toxin, the nontoxic component, and the 900kilodalton botulinum toxin complex species. Infect. Immun. 66, 2420–2425. Maksymowych, A. B. and Simpson, L. L. (1998) Binding and transcytosis of botulinum neurotoxin by polarized human colon-carcinoma cells. J. Biol. Chem. 273, 21,950–21,957. Maksymowych, A. B., Reinhard, M., Malizio, C. J., Goodnough, M. C., Johnson, E. A., and Simpson, L. L. (1999) Pure botulinum neurotoxin is absorbed from the stomach and small intestine and produces peripheral neuromuscular blockade. Infect. Immun. 67, 4708–4712. Niemann, H. (1991) Molecular biology of clostridial neurotoxins, in A Sourcebook of Bacterial Protein Toxins (Alouf, J. E. and Freer, J. H., eds.) Academic Press, London, pp. 303–348. DasGupta, B. R. (1994) Structures of botulinum neurotoxin, its functional domains, and perspectives on the crystalline type A toxin, in Therapy with Botulinum Toxin (Jankovic, J. and Hallett, M., eds.) Marcel Dekker, New York, pp. 15–39. Krieglstein, K. G., Henschen, A. H., Weller, U., and Habermann, E. (1991) Limited proteolysis of tetanus toxin. Relation to activity and identification of cleavage sites. Eur. J. Biochem. 202, 41–51. Schiavo, G., Papini, E., Genna, G., and Montecucco, C. (1990) An intact interchain disulfide bond is required for the neurotoxicity of tetanus toxin. Infect. Immun. 58, 4136–4141. de Paiva, A., Poulain, B., Lawrence, G. W., Shone, C. C., Tauc, L., and Dolly, J. O. (1993) A role for the interchain disulfide or its participating thiols in the internalization of botulinum neurotoxin A revealed by a toxin derivative that binds to ecto-acceptors and inhibits transmitter release intracellularly. J. Biol. Chem. 268, 20,838–20,844. Kozaki, S., Miki, A., Kamata, Y., Ogasawara, J., and Sakaguchi, G. (1989) Immunological characterization of papain-induced fragments of Clostridium botulinum type A neurotoxin and interaction of the fragments with brain synaptosomes. Infect. Immun. 57, 2634–2639. Lalli, G., Herreros, J., Osborne, S. L., Montecucco, C., Rossetto, O., and Schiavo, G. (1999) Functional characterisation of tetanus and botulinum neurotoxins binding domains. J. Cell Sci. 112, 2715–2724. Kurazono, H., Mochida, S., Binz, T., et al. (1992) Minimal essential domains specifying toxicity of the light chains of tetanus toxin and botulinum neurotoxin type A. J. Biol. Chem. 267, 14,721–14,729.

332

Meunier et al.

33. Jiang, W. and Bond, J. S. (1992) Families of metalloendopeptidases and their relationships. FEBS Lett. 312, 110–114. 34. Schiavo, G., Poulain, B., Rossetto, O., Benfenati, F., Tauc, L., and Montecucco, C. (1992) Tetanus toxin is a zinc protein and its inhibition of neurotransmitter release and protease activity depends on zinc. EMBO J. 11, 3577–3583. 35. Wright, J. F., Pernollet, M., Reboul, A., Aude, C., and Colomb, M. G. (1992) Identification and partial characterization of a low affinity metal-binding site in the light chain of tetanus toxin. J. Biol. Chem. 267, 9053–9058. 36. Tonello, F., Schiavo, G., and Montecucco, C. (1997) Metal substitution of tetanus neurotoxin. Biochem. J. 322, 507–510. 37. Umland, T. C., Wingert, L. M., Swaminathan, S., Furey, W. F., Schmidt, J. J., and Sax, M. (1997) Structure of the receptor binding fragment Hc of tetanus toxin. Nat. Struct. Biol. 4, 788–792. 38. Knapp, M., Segelke, B., and Rupp, B. (1998) The 1.61 Angstrom structure of the tetanus toxin. Ganglioside binding region: solved by MAD and MIR phase combination. Am. Cryst. Assoc., Abstract. Papers 25, 90. 39. Emsley, P., Fotinou, C., Black, I., et al. (2000) The structures of the H-C fragment of tetanus toxin with carbohydrate subunit complexes provide insight into ganglioside binding. J. Biol. Chem. 275, 8889–8894. 40. Lacy, D. B. and Stevens, R. C. (1999) Sequence homology and structural analysis of the clostridial neurotoxins. J. Mol. Biol. 291, 1091–1104. 41. Schmitt, A., Dreyer, F., and John, C. (1981) At least three sequential steps are involved in the tetanus toxin-induced block of neuromuscular transmission. Naunyn Schmiedebergs Arch. Pharmacol. 317, 326–330. 42. Montecucco, C. and Schiavo, G. (1995) Structure and function of tetanus and botulinum neurotoxins. Q. Rev. Biophys. 28, 423–472. 43. Penner, R., Neher, E., and Dreyer, F. (1986) Intracellularly injected tetanus toxin inhibits exocytosis in bovine adrenal chromaffin cells. Nature 324, 76–78. 44. Poulain, B., Tauc, L., Maisey, E. A., Wadsworth, J. D., Mohan, P. M., and Dolly, J. O. (1988) Neurotransmitter release is blocked intracellularly by botulinum neurotoxin, and this requires uptake of both toxin polypeptides by a process mediated by the larger chain. Proc. Natl. Acad. Sci. USA 85, 4090–4094. 45. Ahnert-Hilger, G., Weller, U., Dauzenroth, M. E., Habermann, E., and Gratzl, M. (1989) The tetanus toxin light chain inhibits exocytosis. FEBS Lett. 242, 245–248. 46. Bittner, M. A., Habig, W. H., and Holz, R. W. (1989) Isolated light chain of tetanus toxin inhibits exocytosis: studies in digitonin-permeabilized cells. J. Neurochem. 53, 966–968. 47. Bittner, M. A., DasGupta, B. R., and Holz, R. W. (1989) Isolated light chains of botulinum neurotoxins inhibit exocytosis. Studies in digitonin-permeabilized chromaffin cells. J. Biol. Chem. 264, 10,354–10,360. 48. Mochida, S., Poulain, B., Weller, U., Habermann, E., and Tauc, L. (1989) Light chain of tetanus toxin intracellularly inhibits acetylcholine release at neuro-neuronal synapses, and its internalization is mediated by heavy chain. FEBS Lett. 253, 47–51. 49. Hoch, D. H., Romero-Mira, M., Ehrlich, B. E., Finkelstein, A., DasGupta, B. R., and Simpson, L. L. (1985) Channels formed by botulinum, tetanus, and diphtheria toxins in planar lipid bilayers: relevance to translocation of proteins across membranes. Proc. Natl. Acad. Sci. USA 82, 1692–1696. 50. Donovan, J. J. and Middlebrook, J. L. (1986) Ion-conducting channels produced by botulinum toxin in planar lipid membranes. Biochemistry 25, 2872–2876. 51. Blaustein, R. O., Germann, W. J., Finkelstein, A., and DasGupta, B. R. (1987) The N-terminal half of the heavy chain of botulinum type A neurotoxin forms channels in planar phospholipid bilayers. FEBS Lett. 226, 115–120.


333

52. Shone, C. C., Hambleton, P., and Melling, J. (1987) A 50-kDa fragment from the NH2terminus of the heavy subunit of Clostridium botulinum type A neurotoxin forms channels in lipid vesicles. Eur. J. Biochem. 167, 175–180. 53. Gambale, F. and Montal, M. (1988) Characterization of the channel properties of tetanus toxin in planar lipid bilayers. Biophys. J. 53, 771–783. 54. Montal, M. S., Blewitt, R., Tomich, J. M., and Montal, M. (1992) Identification of an ion channel-forming motif in the primary structure of tetanus and botulinum neurotoxins. FEBS Lett. 313, 12–18. 55. Oblatt-Montal, M., Yamazaki, M., Nelson, R., and Montal, M. (1995) Formation of ion channels in lipid bilayers by a peptide with the predicted transmembrane sequence of botulinum neurotoxin A. Protein Sci. 4, 1490–1497. 56. Sheridan, R. E., Deshpande, S. S., Nicholson, J. D., and Adler, M. (1997) Structural features of aminoquinolines necessary for antagonist activity against botulinum neurotoxin. Toxicon 35, 1439–1451. 57. Fu, F. N. and Singh, B. R. (1999) Calcein permeability of liposomes mediated by type A botulinum neurotoxin and its light and heavy chains. J. Prot. Chem. 18, 701–707. 58. Bizzini, B., Stoeckel, K., and Schwab, M. (1977) An antigenic polypeptide fragment isolated from tetanus toxin: chemical characterization, binding to gangliosides and retrograde axonal transport in various neuron systems. J. Neurochem. 28, 529–542. 59. Morris, N. P., Consiglio, E., Kohn, L. D., Habig, W. H., Hardegree, M. C., and Helting, T. B. (1980) Interaction of fragment B and C of tetanus toxin with neural and thyroid membranes and with gangliosides. J. Biol. Chem. 255, 6071–6076. 60. Weller, U., Taylor, C. F., and Habermann, E. (1986) Quantitative comparison between tetanus toxin, some fragments and toxoid for binding and axonal transport in the rat. Toxicon 24, 1055–1063. 61. Herreros, J., Lalli, G., and Schiavo, G. (2000) C-terminal half of tetanus toxin fragment C is sufficient for neuronal binding and interaction with a putative protein receptor. Biochem. J. 347, 199–204. 62. Halpern, J. L. and Loftus, A. (1993) Characterization of the receptor-binding domain of tetanus toxin. J. Biol. Chem. 268, 11,188–11,192. 63. Shapiro, R. E., Specht, C. D., Collins, B. E., Woods, A. S., Cotter, R. J., and Schnaar, R. L. (1997) Identification of a ganglioside recognition domain of tetanus toxin using a novel ganglioside photoaffinity ligand. J. Biol. Chem. 272, 30,380–30,386. 64. Kubota, T., Watanabe, T., Yokosawa, N., Tsuzuki, K., Indoh, T., Moriishi, K., et al. (1997) Epitope regions in the heavy chain in Clostridium botulinum type E neurotoxin recognized by monoclonal antibodies. Appl. Environ. Microbiol. 63, 1214–1218. 65. Kamata, Y., Yoshimoto, M., and Kozaki, S. (1997) Interaction between botulinum neurotoxin type-A and ganglioside-ganglioside inactivates the neurotoxin and quenches its tryptophan fluorescence. Toxicon 35, 1337–1340. 66. Herreros, J., Lalli, G., Montecucco, C., and Schiavo, G. (2000) Tetanus toxin fragment C binds to a protein present in neuronal cell lines and motoneurons. J. Neurochem. 74, 1941–1950. 67. Lebeda, F. J. and Olson, M. A. (1995) Structural predictions of the channel-forming region of botulinum neurotoxin heavy chain. Toxicon 33, 559–567. 68. Wiener, M., Freymann, D., Ghosh, P., and Stroud, R. M. (1997) Crystal structure of colicin Ia. Nature 385, 461–464. 69. Menestrina, G., Forti, S., and Gambale, F. (1989) Interaction of tetanus toxin with lipid vesicles. Effects of pH, surface charge, and transmembrane potential on the kinetics of channel formation. Biophys. J. 55, 393–405. 70. Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Atomic structure of the ectodomain from HIV-1 gp41. Nature 387, 426–430.

334

Meunier et al.

71. Li, Y., Foran, P., Fairweather, N. F., et al. (1994) A single mutation in the recombinant light chain of tetanus toxin abolishes its proteolytic activity and removes the toxicity seen after reconstitution with native heavy chain. Biochemistry 33, 7014–7020. 72. Yamasaki, S., Hu, Y., Binz, T., et al. (1994) Synaptobrevin/vesicle-associated membrane protein (VAMP) of Aplysia californica: structure and proteolysis by tetanus toxin and botulinal neurotoxins type D and F. Proc. Natl. Acad. Sci. USA 91, 4688–4692. 73. Zhou, L., de Paiva, A., Liu, D., Aoki, R., and Dolly, J. O. (1995) Expression and purification of the light chain of botulinum neurotoxin A: a single mutation abolishes its cleavage of SNAP-25 and neurotoxicity after reconstitution with the heavy chain. Biochemistry 34, 15,175–15,181. 74. Morante, S., Furenlid, L., Schiavo, G., Tonello, F., Zwilling, R., and Montecucco, C. (1996) X-ray absorption spectroscopy study of zinc coordination in tetanus neurotoxin, astacin, alkaline protease and thermolysin. Eur. J. Biochem. 235, 606–612. 75. Meneghini, C. and Morante, S. (1998) The active site structure of tetanus neurotoxin resolved by multiple scattering analysis in X-ray absorption spectroscopy. Biophys. J. 75, 1953–1963. 76. Halpern, J. L. and Neale, E. A. (1995) Neurospecific binding, internalization, and retrograde axonal transport. Curr. Top. Microbiol. Immunol. 195, 221–241. 77. Black, J. D. and Dolly, J. O. (1986) Interaction of 125I-labeled botulinum neurotoxins with nerve terminals. II. Autoradiographic evidence for its uptake into motor nerves by acceptor-mediated endocytosis. J. Cell Biol. 103, 535–544. 78. Dolly, J. O., Black, J., Williams, R. S., and Melling, J. (1984) Acceptors for botulinum neurotoxin reside on motor nerve terminals and mediate its internalization. Nature 307, 457–460. 79. Hirokawa, N. and Kitamura, M. (1979) Binding of Clostridium botulinum neurotoxin to the presynaptic membrane in the central nervous system. J. Cell Biol. 81, 43–49. 80. Williamson, L. C., Bateman, K. E., Clifford, J. C. M., and Neale, E. A. (1999) Neuronal sensitivity to tetanus toxin requires gangliosides. J. Biol. Chem. 274, 25,173–25,180. 81. Herreros, J., Marti, E., Ruiz-Montasell, B., Casanova, A., Niemann, H., and Blasi, J. (1997) Localisation of putative receptors for tetanus toxin and botulinum neurotoxin type A in rat central nervous system. Eur. J. Neurosci. 9, 2677–2686. 82. Montecucco, C. (1986) How do tetanus and botulinum toxins bind to neuronal membranes? Trends Biochem. Sci. 11, 315–317. 83. Marxen, P., Fuhrmann, U., and Bigalke, H. (1989) Gangliosides mediate inhibitory effects of tetanus and botulinum A neurotoxins on exocytosis in chromaffin cells. Toxicon 27, 849–859. 84. Bigalke, H., Muller, H., and Dreyer, F. (1986) Botulinum A neurotoxin unlike tetanus toxin acts via a neuraminidase sensitive structure. Toxicon 24, 1065–1074. 85. Sheikh, K. A., Sun, J., Liu, Y., Kawai, H., Crawford, T. O., Proia, R. L., et al. (1999) Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc. Natl. Acad. Sci. USA 96, 7532–7537. 86. Takamiya, K., Yamamoto, A., Furukawa, K., Yamashiro, S., Shin, M., Okada, M., et al. (1996) Mice with disrupted GM2/GD2 synthase gene lack complex gangliosides but exhibit only subtle defects in their nervous system. Proc. Natl. Acad. Sci. USA 93, 10,662-10,667. 87. Kitamura, M., Takamiya, K., Aizawa, S., and Furukawa, K. (1999) Gangliosides are the receptor for C. botulinum neurotoxin in mice. J. Neurochem. 73, S64–S64. 88. Williams, R. S., Tse, C. K., Dolly, J. O., Hambleton, P., and Melling, J. (1983) Radioiodination of botulinum neurotoxin type A with retention of biological activity and its binding to brain synaptosomes. Eur. J. Biochem. 131, 437–445. 89. Evans, D. M., Williams, R. S., Shone, C. C., Hambleton, P., Melling, J., and Dolly, J. O. (1986) Botulinum neurotoxin type B. Its purification, radioiodination and interaction with rat-brain synaptosomal membranes. Eur. J. Biochem. 154, 409–416.


335

90. Agui, T., Syuto, B., Oguma, K., Iida, H., and Kubo, S. (1983) Binding of Clostridium botulinum type C neurotoxin to rat brain synaptosomes. J. Biochem. 94, 521–527. 91. Bakry, N., Kamata, Y., Sorensen, R., and Simpson, L. L. (1991) Tetanus toxin and neuronal membranes: the relationship between binding and toxicity. J. Pharmacol. Exp. Ther. 258, 613–619. 92. Coffield, J. A., Bakry, N. M., Maksymowych, A. B., and Simpson, L. L. (1999) Characterization of a vertebrate neuromuscular junction that demonstrates selective resistance to botulinum toxin. J. Pharmacol. Exp. Ther. 289, 1509–1516. 93. Nishiki, T., Kamata, Y., Nemoto, Y., Omori, A., Ito, T., Takahashi, M., and Kozaki S. (1994) Identification of protein receptor for Clostridium botulinum type B neurotoxin in rat brain synaptosomes. J. Biol. Chem. 269, 10,498–10,503. 94. Nishiki, T., Tokuyama, Y., Kamata, Y., Nemoto, Y., Yoshida, A., Sekiguchi, M., et al. (1996) Binding of botulinum type B neurotoxin to Chinese hamster ovary cells transfected with rat synaptotagmin II cDNA. Neurosci. Lett. 208, 105–108. 95. Nishiki, T., Tokuyama, Y., Kamata, Y., Nemoto, Y., Yoshida, A., Sato, K., et al. (1996) The high-affinity binding of Clostridium botulinum type B neurotoxin to synaptotagmin II associated with gangliosides GT1b/GD1a. FEBS Lett. 378, 253–257. 96. Li, L. and Singh, B. R. (1998) Isolation of synaptotagmin as a receptor for type A and type E botulinum neurotoxin and analysis of their comparative binding using a new microtiter plate assay. Nat. Toxins 7, 215–226. 97. Bakry, N. M., Kamata, Y., and Simpson, L. L. (1997) Expression of botulinum toxin binding sites in Xenopus oocytes. Infect. Immun. 65, 2225–2232. 98. Hughes, R. and Whaler, B. C. (1962) Influence of nerve-endings activity and of drugs on the rate of paralysis of rat diaphragm preparations by Clostridium botulinum type A toxin. J. Physiol. (Lond.) 160, 221–233. 99. Fishman, P. S., Parks, D. A., Patwardhan, A. J., and Matthews, C. C. (1999) Neuronal binding of tetanus toxin compared to its ganglioside binding fragment (H-c). Nat. Toxins 7, 151–156. 100. Parton, R. G., Ockleford, C. D., and Critchley, D. R. (1987) A study of the mechanism of internalisation of tetanus toxin by primary mouse spinal cord cultures. J. Neurochem. 49, 1057–1068. 101. Schwab, M. E. and Thoenen, H. (1978) Selective binding, uptake, and retrograde transport of tetanus toxin by nerve terminals in the rat iris. An electron microscope study using colloidal gold as a tracer. J. Cell Biol. 77, 1–13. 102. Montesano, R., Roth, J., Robert, A., and Orci, L. (1982) Non-coated membrane invaginations are involved in binding and internalization of cholera and tetanus toxin. Nature 296, 651–653. 103. Damke, H., Baba, T., van der Bliek, A. M., and Schmid, S. L. (1995) Clathrin-independent pinocytosis is induced in cells overexpressing a temperature-sensitive mutant of dynamin. J. Cell Biol. 131, 69–80. 104. Henley, J. R., Krueger, W. A., Oswald, B. J., and McNiven, M. A. (1998) Dynaminmediated internalization of caveolae. J. Cell Biol. 141, 85–99. 105. Turek, J. J., Leamon, C. P., and Low, P. S. (1993) Endocytosis of folate-protein conjugates: ultrastructural localization in KB cells. J. Cell Sci. 106, 423–430. 106. Parton, R. G., Ockleford, C. D., and Critchley, D. R. (1988) Tetanus toxin binding to mouse spinal cord cells: an evaluation of the role of gangliosides in toxin internalization. Brain Res. 475, 118–127. 107. Sandvig, K., Olsnes, S., Petersen, O. W., and van Deurs, B. (1989) Endocytosis from coated pits of Shiga toxins: a glycolipid-binding protein from Shigella dysenteriae. J. Cell Biol. 108, 1331–1343. 108. Orlandi, P. A. and Fishman, P. H. (1998) Filipin-dependent inhibition of cholera-toxin evidence for toxin internalization and activation through caveolae-like domains. J. Cell Biol. 141, 905–915.

336

Meunier et al.

109. Matteoli, M., Verderio, C., Rossetto, O., Iezzi N., Coco, S., Schiavo, G., and Montecucco, C. (1996) Synaptic vesicle endocytosis mediates the entry of tetanus neurotoxin into hippocampal neurons. Proc. Natl. Acad. Sci. USA 93, 13,310–13,315. 110. Habermann, E. and Erdmann, G. (1978) Pharmacokinetic and histoautoradiographic evidence for the intraaxonal movement of toxin in the pathogenesis of tetanus. Toxicon 16, 611–623. 111. Menestrina, G., Schiavo, G., and Montecucco, C. (1994) Molecular mechanisms of action of bacterial protein toxins. Mol. Aspects Med. 15, 79–193. 112. Montecucco, C., Papini, E., and Schiavo, G. (1994) Bacterial protein toxins penetrate cells via a four-step mechanism. FEBS Lett. 346, 92–98. 113. Simpson, L. L. (1982) The interaction between aminoquinolines and presynaptically acting neurotoxins. J. Pharmacol. Exp. Ther. 222, 43–48. 114. Simpson, L. L. (1983) Ammonium chloride and methylamine hydrochloride antagonize clostridial neurotoxins. J. Pharmacol. Exp. Ther. 225, 546–552. 115. Williamson, L. C. and Neale, E. A. (1994) Bafilomycin A1 inhibits the action of tetanus toxin in spinal cord neurons in cell culture. J. Neurochem. 63, 2342–2345. 116. Schmid, M. F., Robinson, J. P., and DasGupta, B. R. (1993) Direct visualization of botulinum neurotoxin-induced channels in phospholipid vesicles. Nature 364, 827–830. 117. Chaddock, J. A., Purkiss, J. R., Friis, L. M., et al. (2000) Inhibition of vesicular secretion in both neuronal and nonneuronal cells by a retargeted endopeptidase derivative of Clostridium botulinum neurotoxin type A. Infect. Immun. 68, 2587–2593. 118. Herreros, J., Lalli, G., Montecucco, C., and Schiavo, G. (1999) Pathophysiological properties of clostridial neurotoxins, in The Comprehensive Sourcebook of Bacterial Protein Toxins (Freer, J. H. and Alouf, J. E., eds.) Academic Press, London, pp. 202–228. 119. Schiavo, G., Benfenati, F., Poulain, B., et al. (1992) Tetanus and botulinum B neurotoxins block neurotransmitter release by proteolytic cleavage of synaptobrevin. Nature 359, 832–835. 120. Ferrer Montiel, A. V., Canaves, J. M., DasGupta, B. R., Wilson, M. C., and Montal, M. (1996) Tyrosine phosphorylation modulates the activity of clostridial neurotoxins. J. Biol. Chem. 271, 18,322–18,325. 121. Schiavo, G. and Montecucco, C. (1995) Tetanus and botulism neurotoxins: isolation and assay. Methods Enzymol. 248, 643–652. 122. Ekong, T. A., McLellan, K., and Sesardic, D. (1995) Immunological detection of Clostridium botulinum toxin type A in therapeutic preparations. J. Immunol. Methods 180, 181–191. 123. Hallis, B., James, B. A., and Shone, C. C. (1996) Development of novel assays for botulinum type A and B neurotoxins based on their endopeptidase activities. J. Clin. Microbiol. 34, 1934–1938. 124. Soleilhac, J. M., Cornille, F., Martin, L., Lenoir, C., Fournie-Zaluski, M. C., and Roques, B. P. (1996) A sensitive and rapid fluorescence-based assay for determination of tetanus toxin peptidase activity. Anal. Biochem. 241, 120–127. 125. Ekong, T. A., McLellan, K., and Sesardic, D. (1996) Recombinant SNAP-25 is an effective substrate for Clostridium botulinum type A toxin endopeptidase activity in vitro. Microbiology 143, 3337–3347. 126. Wictome, M., Newton, K., Jameson, K., et al. (1999) Development of an in vitro bioassay for Clostridium botulinum type B neurotoxin in foods that is more sensitive than the mouse bioassay. Appl. Environ. Microbiol. 65, 3787–3792. 127. Knight, C. G. (1995) Fluorimetric assay of proteolytic enzymes. Methods Enzymol. 248, 18–34. 128. Osen Sand, A., Staple, J. K., Naldi, E., et al. (1996) Common and distinct fusion proteins in axonal growth and transmitter release. J. Comp. Neurol. 367, 222–234. 129. Williamson, L. C., Halpern, J. L., Montecucco, C., Brown, J. E. and Neale, E. A. (1996) Clostridial neurotoxins and substrate proteolysis in intact neurons: botulinum neurotoxin C acts on synaptosomal-associated protein of 25 kDa. J. Biol. Chem. 271, 7694–7699.


337

130. Raciborska, D. A., Trimble, W. S., and Charlton, M. P. (1998) Presynaptic protein interactions in vivo. Evidence from botulinum A, botulinum C, botulinum D and botulinum E action at frog neuromuscular junction. Eur. J. Neurosci. 10, 2617–2628. 131. Cornille, F., Goudreau, N., Ficheux, D., Niemann, H., and Roques, B. P. (1994) Solidphase synthesis, conformational analysis and in vitro cleavage of synthetic human synaptobrevin II 1-93 by tetanus toxin L chain. Eur. J. Biochem. 222, 173–181. 132. Adler, M., Nicholson, J. D., and Hackley, B. E. (1998) Efficacy of a novel metalloprotease inhibitor on botulinum neurotoxin B activity. FEBS Lett. 429, 234–238. 133. Martin, L., Cornille, F., Coric, P., Roques, B. P., and Fournie-Zaluski, M. C. (1998) Betaamino-thiols inhibit the zinc metallopeptidase activity of tetanus toxin light chain. J. Med. Chem. 41, 3450–3460. 134. Martin, L., Cornille, F., Turcaud, S., Meudal, H., Roques, B. P., and Fournie-Zaluski, M. C. (1999) Metallopeptidase inhibitors of tetanus toxin: A combinatorial approach. J. Med. Chem. 42, 515–525. 135. Söllner, T. (1995) SNAREs and targeted membrane fusion. FEBS Lett. 369, 80–83. 136. Robinson, L. J. and Martin, T. F. (1998) Docking and fusion in neurosecretion. Curr. Opin. Cell Biol. 10, 483–492. 137. Schiavo, G. and Stenbeck, G. (1998) Molecular analysis of neurotransmitter release. Essays Biochem. 33, 29–41. 138. Mochida, S. (2000) Protein-protein interactions in neurotransmitter release. Neurosci. Res. 36, 175–182. 139. Söllner, T., Whiteheart, S. W., Brunner, M., et al. (1993) SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318–324. 140. Hayashi, T., McMahon, H., Yamasaki, S., et al. (1994) Synaptic vesicle membrane fusion complex: action of clostridial neurotoxins on assembly. EMBO J. 13, 5051–5061. 141. Hayashi, T., Yamasaki, S., Nauenburg, S., Binz, T., and Niemann, H. (1995) Disassembly of the reconstituted synaptic vesicle membrane fusion complex in vitro. EMBO J. 14, 2317–2325. 142. Sutton, R. B., Fasshauer, D., Jahn, R., and Brunger, A. T. (1998) Crystal-structure of a SNARE complex involved in synaptic exocytosis at 2.4 angstrom resolution. Nature 395, 347–353. 143. Poirier, M. A., Xiao, W. Z., MacOsko, J. C., Chan, C., Shin, Y. K., and Bennett, M. K. (1998) The synaptic SNARE complex is a parallel 4-stranded helical bundle. Nat. Struct. Biol. 5, 765–769. 144. Skehel, J. J. and Wiley, D. C. (1998) Coiled coils in both intracellular vesicle and viral membrane fusion. Cell 95, 871–874. 145. Kee, Y., Lin, R. C., Hsu, S. C., and Scheller, R. H. (1995) Distinct domains of syntaxin are required for synaptic vesicle fusion complex formation and dissociation. Neuron 14, 991–998. 146. Vaidyanathan, V. V., Yoshino, K., Jahnz, M., et al. (1999) Proteolysis of SNAP-25 isoforms by botulinum neurotoxin types A, C, and E: domains and amino acid residues controlling the formation of enzyme-substrate complexes and cleavage. J. Neurochem. 72, 327–337. 147. Binscheck, T., Bartels, F., Bergel, H., et al. (1995) IgA protease from Neisseria gonorrhoeae inhibits exocytosis in bovine chromaffin cells like tetanus toxin. J. Biol. Chem. 270, 1770–1774. 148. Cornille, F., Deloye, F., Fournie-Zaluski, M. C., Roques, B. P., and Poulain, B. (1995) Inhibition of neurotransmitter release by synthetic proline-rich peptides shows that the N-terminal domain of vesicle-associated membrane protein/synaptobrevin is critical for neuroexocytosis. J. Biol. Chem. 270, 16,826–16,832. 149. Woodman, P. G. (1997) The roles of NSF, SNAPs and SNAREs during membrane fusion. Biochim. Biophys. Acta 1357, 155–172.

338

Meunier et al.

150. Haas, A. (1998) NSF: fusion and beyond. Trends Cell Biol. 8, 471–473. 151. Owen, D. J. and Schiavo, G. (1999) A handle on NSF. Nat. Cell Biol. 1, E127–128. 152. Barnard, R. J. O., Morgan, A., and Burgoyne, R. D. (1997) Stimulation of NSF ATPase activity by alpha-SNAP is required for SNARE complex disassembly and exocytosis. J. Cell Biol. 139, 875–883. 153. Hanson, P. I., Roth, R., Morisaki, H., Jahn, R., and Heuser, J. E. (1997) Structure and conformational changes in NSF and its membrane receptor complexes visualized by quick-freeze/deep-etch electron microscopy. Cell 90, 523–535. 154. Hohl, T. M., Parlati, F., Wimmer, C., Rothman, J. E., Sollner, T. H., and Engelhardt, H. (1998) Arrangement of subunits in 20S particles consisting of NSF, SNAPs, and SNARE complexes. Mol. Cell 2, 539–548. 155. Rizo, J. and Südhof, T. C. (1998) Mechanics of membrane fusion. Nat. Struct. Biol. 5, 839–842. 156. Weber, T., Zemelman, B. V., McNew, J. A., et al. (1998) SNAREpins: minimal machinery for membrane fusion. Cell 92, 759–772. 157. Nickel, W., Weber, T., McNew, J. A., Parlati, F., Sollner, T. H., and Rothman, J. E. (1999) Content mixing and membrane integrity during membrane fusion driven by pairing of isolated v-SNAREs and t-SNAREs. Proc. Natl. Acad. Sci. USA 96, 12,571–12,576. 158. Weber, T., Parlati, F., McNew, J. A., et al. (2000) SNAREpins are functionally resistant to disruption by NSF and alpha SNAP. J. Cell Biol. 149, 1063–1072. 159. Parlati, F., McNew, J. A., Fukuda, R., Miller, R., Sollner, T. H., and Rothman, J. E. (2000) Topological restriction of SNARE-dependent membrane fusion. Nature 407, 194–198. 160. Fukuda, R., McNew, J. A., Weber, T., Parlati, F., Engel, T., Nickel, W., et al. (2000) Functional architecture of an intracellular membrane t-SNARE. Nature 407, 198–202. 161. Pellegrini, L. L., O’Connor, V., Lottspeich, F., and Betz, H. (1995) Clostridial neurotoxins compromise the stability of a low energy SNARE complex mediating NSF activation of synaptic vesicle fusion. EMBO J. 14, 4705–4713. 162. Pellegrini, L. L., O’Connor, V., and Betz, H. (1994) Fusion complex formation protects synaptobrevin against proteolysis by tetanus toxin light chain. FEBS Lett. 353, 319–323. 163. Washbourne, P., Bortoletto, N., Graham, M. E., Wilson, M. C., Burgoyne, R. D., and Montecucco, C. (1999) Botulinum neurotoxin E-insensitive mutants of SNAP-25 fail to bind VAMP but support exocytosis. J. Neurochem. 73, 2424–2433. 164. Foran, P., Lawrence, G. W., Shone, C. C., Foster, K. A., and Dolly, J. O. (1996) Botulinum neurotoxin C1 cleaves both syntaxin and SNAP-25 in intact and permeabilized chromaffin cells: correlation with its blockade of catecholamine release. Biochemistry 35, 2630–2636. 165. Bruns, D., Engers, S., Yang, C., Ossig, R., Jeromin, A., and Jahn, R. (1997) Inhibition of transmitter release correlates with the proteolytic activity of tetanus toxin and botulinus toxin A in individual cultured synapses of Hirudo medicinalis. J. Neurosci. 17, 1898–1910. 166. Xu, T., Binz, T., Niemann, H., and Neher, E. (1998) Multiple kinetic components of exocytosis distinguished by neurotoxin sensitivity. Nature Neurosci. 1, 192–200. 167. Shone, C. C., Quinn, C. P., Wait, R., Hallis, B., Fooks, S. G., and Hambleton, P. (1993) Proteolytic cleavage of synthetic fragments of vesicle-associated membrane protein, isoform-2 by botulinum type B neurotoxin. Eur. J. Biochem. 217, 965–971. 168. Dayanithi, G., Stecher, B., Höhne-Zell, B., et al. (1994) Exploring the functional domain and the target of the tetanus toxin light chain in neurohypophysial terminals. Neuroscience 58, 423–431. 169. Shone, C. C. and Roberts, A. K. (1994) Peptide substrate specificity and properties of the zinc-endopeptidase activity of botulinum type B neurotoxin. Eur. J. Biochem. 225, 263–270.


339

170. Cornille, F., Martin, L., Lenoir, C., Cussac, D., Roques, B. P., and Fourniezaluski, M. C. (1997) Cooperative exosite-dependent cleavage of synaptobrevin by tetanus toxin light chain. J. Biol. Chem. 272, 3459–3464. 171. Foran, P., Shone, C. C., and Dolly, J. O. (1994) Differences in the protease activities of tetanus and botulinum B toxins revealed by the cleavage of vesicle-associated membrane protein and various sized fragments. Biochemistry 33, 15,365–15,374. 172. Yamasaki, S., Baumeister, A., Binz, T., et al. (1994) Cleavage of members of the synaptobrevin/VAMP family by types D and F botulinal neurotoxins and tetanus toxin. J. Biol. Chem. 269, 12,764–12,772. 173. Rossetto, O., Schiavo, G., Montecucco, C., et al. (1994) SNARE motif and neurotoxins. Nature 372, 415–416. 174. Sweeney, S. T., Broadie, K., Keane, J., Niemann, H., and O’Kane, C. J. (1995) Targeted expression of tetanus toxin light chain in Drosophila specifically eliminates synaptic transmission and causes behavioral defects. Neuron 14, 341–351. 175. Pellizzari, R., Rossetto, O., Lozzi, L., et al. (1996) Structural determinants of the specificity for synaptic vesicle-associated membrane protein/synaptobrevin of tetanus and botulinum type B and G neurotoxins. J. Biol. Chem. 271, 20,353–20,358. 176. Wictome, M., Rossetto, O., Montecucco, C., and Shone, C. C. (1996) Substrate residues N-terminal to the cleavage site of botulinum type B neurotoxin play a role in determining the specificity of its endopeptidase activity. FEBS Lett. 386, 133–136. 177. Pellizzari, R., Mason, S., Shone, C. C., and Montecucco, C. (1997) The interaction of synaptic vesicle-associated membrane protein/synaptobrevin with botulinum neurotoxins D and F. FEBS Lett. 409, 339–342. 178. Washbourne, P., Pellizzari, R., Baldini, G., Wilson, M. C., and Montecucco, C. (1997) Botulinum neurotoxin type A and type E require the SNARE motif in SNAP-25 for proteolysis. FEBS Lett. 418, 1–5. 179. Facchiano, F. and Luini, A. (1992) Tetanus toxin potently stimulates tissue transglutaminase. A possible mechanism of neurotoxicity. J. Biol. Chem. 267, 13,267– 13,271. 180. Facchiano, F., Benfenati, F., Valtorta, F., and Luini, A. (1993) Covalent modification of synapsin I by a tetanus toxin-activated transglutaminase. J. Biol. Chem. 268, 4588–4591. 181. Ashton, A. C., Li, Y., Doussau, F., et al. (1995) Tetanus toxin inhibits neuroexocytosis even when its Zn2+-dependent protease activity is removed. J. Biol. Chem. 270, 31,386– 31,390. 182. Ray, P., Berman, J. D., Middleton, W., and Brendle, J. (1993) Botulinum toxin inhibits arachidonic acid release associated with acetylcholine release from PC12 cells. J. Biol. Chem. 268, 11,057–11,064. 183. Burgen, A. S. V., Dickens, F., and Zatman, L. T. (1949) The action of botulinum toxin on the neuromuscular junction. J. Physiol. (Lond.) 109, 10–24. 184. Molgó, J., Comella, J. X., Angaut-Petit, D., Pecot-Dechavassine, M., Tabti, N., Faille, L., et al. (1990) Presynaptic actions of botulinal neurotoxins at vertebrate neuromuscular junctions. J. Physiol. (Paris) 84, 152–166. 185. Poulain, B., Molgó, J., and Thesleff, S. (1995) Quantal neurotransmitter release and the clostridial neurotoxins’ targets. Curr. Top. Microbiol. Immunol. 195, 237–249. 186. Molgó, J., Meunier, F. A., and Sellin, L. C. (1997) Quantal transmitter release at botulinum-treated vertebrate neuromuscular junctions, in Neurochemistry: Cellular, Molecular, and Clinical Aspects (Teelken, A. W. and Korf, J., eds.), Plenum Press, New York, pp. 713–717. 187. Molgó, J., Dasgupta, B. R., and Thesleff, S. (1989) Characterization of the actions of botulinum neurotoxin type E at the rat neuromuscular junction. Acta. Physiol. Scand. 137, 497–501.

340

Meunier et al.

188. Wieszt, L. and Dreyer, F. (1991) Mode of action of botulinum toxin E on the transmitter release process at the mouse neuromuscular junction. Naunyn Schmiedebergs Arch. Pharmacol. 344, R74. 189. Boroff, D. A., del Castillo, J., Evoy, W. H., and Steinhardt, R. A. (1974) Observations on the action of type A botulinum toxin on frog neuromuscular junctions. J. Physiol. (Lond.) 240, 227–253. 190. Kriebel, M. E., Llados, F., and Matteson, D. R. (1976) Spontaneous subminature endplate potentials in mouse diaphragm muscle: evidence for synchronous release. J. Physiol. (Lond.) 262, 553–581. 191. Cull-Candy, S. G., Lundh, H., and Thesleff, S. (1976) Effects of botulinum toxin on neuromuscular transmission in the rat. J. Physiol. (Lond.) 260, 177–203. 192. Tse, C. K., Wray, D., Melling, J., and Dolly, J. O. (1986) Actions of beta-bungarotoxin on spontaneous release of transmitter at muscle end-plates treated with botulinum toxin. Toxicon 24, 123–130. 193. Dolly, J. O., Lande, S., and Wray, D. W. (1987) The effects of in vitro application of purified botulinum neurotoxin at mouse motor nerve terminals. J. Physiol. (Lond.) 386, 475–484. 194. Gundersen, C. B. (1980) The effects of botulinum toxin on the synthesis, storage and release of acetylcholine. Prog. Neurobiol. 14, 99–119. 195. Thesleff, S. and Molgó, J. (1983) A new type of transmitter release at the neuromuscular junction. Neuroscience 9, 1–8. 196. Molgó, J. and Thesleff, S. (1982) 4-aminoquinoline induced “giant” miniature end-plate potentials at mammalian neuromuscular junctions. Proc. R. Soc. Lond. B. Biol. Sci. 214, 229–247. 197. Colméus, C., Gomez, S., Molgó, J., and Thesleff, S. (1982) Discrepancies between spontaneous and evoked synaptic potentials at normal, regenerating and botulinum toxin poisoned mammalian neuromuscular junctions. Proc. R. Soc. Lond. B. Biol. Sci. 215, 63–74. 198. Thesleff, S., Molgó, J., and Lundh, H. (1983) Botulinum toxin and 4-aminoquinoline induce a similar abnormal type of spontaneous quantal transmitter release at the rat neuromuscular junction. Brain Res. 264, 89–97. 199. Kim, Y. I., Lomo, T., Lupa, M. T., and Thesleff, S. (1984) Miniature end-plate potentials in rat skeletal muscle poisoned with botulinum toxin. J. Physiol. (Lond.) 356, 587–599. 200. Vautrin, J. (1992) Miniature endplate potentials induced by ammonium chloride, hypertonic shock, and botulinum toxin. J. Neurosci. Res. 31, 318–326. 201. Sellin, L. C., Molgó, J., Isberg, P.-E., Törnquist, K., Hansson, B., and Thesleff, S. (1996) On the possible origin of Giant or slow rising miniature end-plate potentials at the neuromuscular junctions. Pflügers Arch. 431, 325–334. 202. Gundersen, C. B., Katz, B., and Miledi, R. (1982). The antagonism between botulinum toxin and calcium in motor nerve terminals. Proc. R. Soc. Lond. B. Biol. Sci. 216, 369–376. 203. Dreyer, F., Mallart, A., and Brigant, J. L. (1983). Botulinum A toxin and tetanus toxin do not affect presynaptic membrane currents in mammalian motor nerve endings. Brain Res. 270, 373–375. 204. Mallart, A., Molgó, J., Angaut-Petit, D., and Thesleff, S. (1989) Is the internal calcium regulation altered in type A botulinum toxin-poisoned motor endings? Brain Res. 479, 167–171. 205. Molgó, J., Siegel, L. S., Tabti, N., and Thesleff, S. (1989) A study of synchronization of quantal transmitter release from mammalian motor endings by the use of botulinal toxins type A and D. J. Physiol. (Lond.) 411, 195–205. 206. Gundersen, C. B., Katz, B., and Miledi, R. (1981) The reduction of endplate responses by botulinum toxin. Proc. R. Soc. Lond. B. Biol. Sci. 213, 489–493.


341

207. Molgó, J., Lemeignan, M., and Thesleff, S. (1987) Aminoglycosides and 3,4diaminopyridine on neuromuscular block caused by botulinum type A toxin. Muscle Nerve 10, 464–470. 208. Dreyer, F. and Schmitt, A. (1981) Different effects of botulinum A toxin and tetanus toxin on the transmitter releasing process at the mammalian neuromuscular junction. Neurosci. Lett. 26, 307–311. 209. Dreyer, F. and Schmitt, A. (1983) Transmitter release in tetanus and botulinum A toxinpoisoned mammalian motor endplates and its dependence on nerve stimulation and temperature. Pflügers. Arch. 399, 228–234. 210. Simpson, L. L. and Dasgupta, B. R. (1983) Botulinum neurotoxin type E: studies on mechanism of action and on structure-activity relationships. J. Pharmacol. Exp. Ther. 224, 135–140. 211. Lomneth, R., Suszkiw, J. B., and DasGupta, B. R. (1990) Response of the chick ciliary ganglion-iris neuromuscular preparation to botulinum neurotoxin. Neurosci. Lett. 113, 211–216. 212. Adler, M., Macdonald, D. A., Sellin, L. C., and Parker, G. W. (1996) Effect of 3,4diaminopyridine on rat extensor digitorum longus muscle paralyzed by local injection of botulinum neurotoxin. Toxicon 34, 237–249. 213. Coffield, J. A., Bakry, N., Zhang, R. D., Carlson, J., Comella, L. G., and Simpson, L. L. (1997) In vitro characterization of botulinum toxin types A, C and D action on human tissues: combined electrophysiologic, pharmacologic and molecular biologic approaches. J. Pharmacol. Exp. Ther. 280, 1489–1498. 214. Molgó, J. (1982) Effects of aminopyridines on neuromuscular transmission, in Aminopyridines and Similarly Acting Drugs, Advances in the Biosciences, vol. 35 (Lechat, P., Thesleff, S., and Bowman, W. C., eds.), Pergamon Press, Oxford, pp. 95–116. 215. Lundh, H., Leander, S., and Thesleff, S. (1977) Antagonism of the paralysis produced by botulinum toxin in the rat. The effects of tetraethylammonium, guanidine and 4aminopyridine. J. Neurol. Sci. 32, 29–43. 216. Molgó, J., Lundh, H., and Thesleff, S. (1980) Potency of 3,4-diaminopyridine and 4aminopyridine on mammalian neuromuscular transmission and the effects of pH changes. Eur. J. Pharmacol. 61, 25–34. 217. Sellin, L. C., Kauffman, J. A., and Dasgupta, B. R. (1983) Comparison of the effects of botulinum neurotoxin types A and E at the rat neuromuscular junction. Med. Biol. 61, 120–125. 218. Kauffman, J. A., Way, J. F. Jr, Siegel, L. S., and Sellin, L. C. (1985). Comparison of the action of types A and F botulinum toxin at the rat neuromuscular junction. Toxicol. Appl. Pharmacol. 79, 211–217. 219. Gansel, M., Penner, R., and Dreyer, F. (1987) Distinct sites of action of clostridial neurotoxins revealed by double-poisoning of mouse motor nerve terminals. Pflügers Arch. 409, 533–539. 220. Simpson, L. L. (1986) A preclinical evaluation of aminopyridines as putative therapeutic agents in the treatment of botulism. Infect. Immun. 52, 858–862. 221. Lawrence, G. W., Foran, P., Mohammed, N., DasGupta, B. R., and Dolly, J. O. (1997) Importance of two adjacent C-terminal sequences of SNAP-25 in exocytosis from intact and permeabilized chromaffin cells revealed by inhibition with botulinum neurotoxins A and E. Biochemistry 36, 3061–3067. 222. Gerona, R. R., Larsen, E. C., Kowalchyk, J. A., and Martin, T. F. (2000) The C terminus of SNAP25 is essential for Ca2+ -dependent binding of synaptotagmin to SNARE complexes. J. Biol. Chem. 275, 6328–6336. 223. Schiavo, G., Stenbeck, G., Rothman, J. E., and Sollner, T. H. (1997) Binding of the synaptic vesicle v-SNARE, synaptotagmin, to the plasma membrane t-SNARE, SNAP-25,

342

224. 225.

226.

227. 228.

229.

230.

231. 232.

233.

234.

235.

236.

237. 238.

239. 240.

Meunier et al. can explain docked vesicles at neurotoxin-treated synapses. Proc. Natl. Acad. Sci. USA 94, 997–1001. Ilardi, J. M., Mochida, S., and Sheng, Z. H. (1999) Snapin: a SNARE-associated protein implicated in synaptic transmission. Nat. Neurosci. 2, 119–124. Sellin, L. C., Thesleff, S., and Dasgupta, B. R. (1983) Different effects of types A and B botulinum toxin on transmitter release at the rat neuromuscular junction. Acta. Physiol. Scand. 119, 127–133. Hua, S. Y., Raciborska, D. A., Trimble, W. S., and Charlton, M. P. (1998) Different VAMP/synaptobrevin complexes for spontaneous and evoked transmitter release at the crayfish neuromuscular junction. J. Neurophysiol. 80, 3233–3246. Harris, A. J. and Miledi, R. (1971) The effect of type D botulinum toxin on frog neuromuscular junctions. J. Physiol. (Lond.) 217, 497–515. Bray, J. J. and Harris, A. J. (1975) Dissociation between nerve-muscle transmission and nerve trophic effects on rat diaphragm using type D botulinum toxin. J. Physiol. (Lond.) 253, 53–77. Molgó, J., Meunier, F. A., and Poulain, B. (1996) Effects of 3,4-diaminopyridine on quantal acetylcholine release from neuromuscular junctions paralysed in vivo with botulinum type-F toxin. Toxicon 34, 1092. Coffield, J. A., Bakry, N., Zhang, R. D., Carlson, J., Gomella, L. G., and Simpson, L. L. (1997) In vitro characterization of botulinum toxin types A, C and D action on human tissues: combined electrophysiologic, pharmacologic and molecular biologic approaches. J. Pharmacol. Exp. Ther. 280, 1489–1498. Mellanby, J. and Thompson, P. A. (1972) The effect of tetanus toxin at the neuromuscular junction in the goldfish. J. Physiol. (Lond.) 224, 407–419. Duchen, L. W. and Tonge, D. A. (1973) The effects of tetanus toxin on neuromuscular transmission and on the morphology of motor end-plates in slow and fast skeletal muscle of the mouse. J. Physiol. (Lond.) 228, 157–172. Dreyer, F., Rosenberg, F., Becker, C., Bigalke, H., and Penner, R. (1987) Differential effects of various secretagogues on quantal transmitter release from mouse motor nerve terminals treated with botulinum A and tetanus toxin. Naunyn Schmiedebergs Arch. Pharmacol. 335, 1–7. Hunt, J. M., Bommert, K., Charlton, M. P., Kistner, A., Habermann, E., Augustine, G. J., and Betz, H. (1994). A post-docking role for synaptobrevin in synaptic vesicle fusion. Neuron 12, 1269–1279. Llinas, R., Sugimori, M., Chu, D., Morita, M., Blasi, J., Herreros, J., et al. (1994) Transmission at the squid giant synapse was blocked by tetanus toxin by affecting synaptobrevin, a vesicle-bound protein. J. Physiol. (Lond.) 477, 129–133. Poulain, B., de Paiva, A., Deloye, F., Doussau, F., Tauc, L., Weller, U., and Dolly, J. O. (1996). Differences in the multiple step process of inhibition of neurotransmitter release induced by tetanus toxin and botulinum neurotoxins type A and B at Aplysia synapses. Neuroscience 70, 567–576. Bevan, S. and Wendon, L. M. (1984) A study of the action of tetanus toxin at rat soleus neuromuscular junctions. J. Physiol. (Lond.) 348, 1–17. Herreros, J., Miralles, F. X., Solsona, C., Bizzini, B., Blasi, J., and Marsal, J. (1995) Tetanus toxin inhibits spontaneous quantal release and cleaves VAMP/synaptobrevin. Brain Res. 699, 165–170. Capogna, M., McKinney, R. A., O’Connor, V., Gahwiler, B. H., and Thompson, S. M. (1997) Ca2+ or Sr2+ partially rescues synaptic transmission in hippocampal cultures treated with botulinum toxin A and C, but not tetanus toxin. J. Neurosci. 17, 7190–7202. Mochida, S., Saisu, H., Kobayashi, H., and Abe, T. (1995) Impairment of syntaxin by botulinum neurotoxin C1 or antibodies inhibits acetylcholine release but not Ca2+ channel activity. Neuroscience 65, 905–915.


343

241. O’Connor, V., Heuss, C., De Bello, W. M., Dresbach, T., Charlton, M. P., Hunt, J. H., et al. (1997) Disruption of syntaxin-mediated protein interactions blocks neurotransmitter secretion. Proc. Natl. Acad. Sci. USA 94, 12,186–12,191. 242. Marsal, J., Ruiz-Montasell, B., Blasi, J., Moreira, J. E., Contreras, D., Sugimori, M., and Llinas, R. (1997) Block of transmitter release by botulinum C1 action on syntaxin at the squid giant synapse. Proc. Natl. Acad. Sci. USA 94, 14,871–14,876. 243. Broadie, K., Prokop, A., Bellen, H. J., O’Kane, C. J., Schulze, K. L., and Sweeney, S. T. (1995) Syntaxin and synaptobrevin function downstream of vesicle docking in Drosophila. Neuron 15, 663–673. 244. Lichtman, J. W., Magrassi, L., and Purves, D. (1987) Visualization of neuromuscular junctions over periods of several months in living mice. J. Neurosci. 7, 1215–1222. 245. Balice-Gordon, R. J. and Lichtman, J. W. (1990) In vivo visualization of the growth of pre- and postsynaptic elements of neuromuscular junctions in the mouse. J. Neurosci. 10, 894–908. 246. Jirmanova, I., Sobotroka, M., Thesleff, S., and Zelena, J. (1964) Atrophy in skeletal muscles poisoned with botulinum toxin. Physiol. Bohemoslov. 13, 467–472. 247. Angaut-Petit, D., Molgó, J., Connold, A., and Faille, L. (1987) The levator auris longus muscle of the mouse: a convenient preparation for studies of short-and long-term presynaptic effects of drugs or toxins. Neurosci. Lett. 82, 83–88. 248. Poulain, B., Bader, M. F., and Molgó, J. (2000) In vitro physiological studies on clostridial neurotoxins. Biological models and procedures for extracellular and intracellular application of toxins. Methods Mol. Biol. 145, 259–286. 249. Duchen, L. W. and Strich, S. J. (1968) The effects of botulinum toxin on the pattern of innervation of skeletal muscle in the mouse. Q. J. Exp. Physiol. 53, 84–89. 250. Duchen, L. W. (1970) Changes in motor innervation and cholinesterase localization induced by botulinum toxin in skeletal muscle of the mouse: differences between fast and slow muscles. J. Neurol. Neurosurg. Psychiat. 33, 40–54. 251. Duchen, L. W. (1971) An electron microscopic study of the changes induced by botulinum toxin in the motor end-plates of slow and fast skeletal muscle fibres of the mouse. J. Neurol. Sci. 14, 47–60. 252. Brown, M. C., Holland, R. L., and Hopkins, W. G. (1981) Motor nerve sprouting. Annu. Rev. Neurosci. 4, 17–42. 253. Pestronk, A. and Drachman, D. B. (1988) Motor nerve outgrowth: reduced capacity for sprouting in the terminals of longer axons. Brain Res. 463, 218–222. 254. Pamphlett, R. (1989) Early terminal and nodal sprouting of motor axons after botulinum toxin. J. Neurol. Sci. 92, 181–192. 255. Angaut-Petit, D. and Molgó, J. (1989) Presynaptic effects of in vivo injection of type A botulinum toxin in the Levator auris longus muscle of the mouse, in Neuromuscular Junction (Sellin, L. C., Libelius, R., and Thesleff, S., eds,), Elsevier Science, Amsterdam, p. 577. 256. Angaut-Petit, D., Molgó, J., Comella, J. X., Faille, L., and Tabti, N. (1990) Terminal sprouting in mouse neuromuscular junctions poisoned with botulinum type A toxin: morphological and electrophysiological features. Neuroscience 37, 799–808. 257. Tian, W. H., Festoff, B. W., Blot, S., Diaz, J., and Hantaï, D. (1995) Synaptic transmission blockade increases plasminogen activator activity in mouse skeletal muscle poisoned with botulinum toxin type A. Synapse 20, 24–32. 258. Juzans, P., Comella, J. X., Molgó, J., Faille, L., and Angaut-Petit, D. (1996) Nerve terminal sprouting in botulinum type-A treated mouse Levator auris longus muscle. Neuromusc. Disord. 6, 177–185. 259. Son, Y. J. and Thompson, W. J. (1995) Nerve sprouting in muscle is induced and guided by processes extended by Schwann cells. Neuron 14, 133–141. 260. de Paiva, A., Meunier, F. A., Molgó, J., Aoki, K. R., and Dolly, J. O. (1999) Functional repair of motor endplates after botulinum neurotoxin A poisoning: bi-phasic switch of

344

261. 262. 263. 264.

265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277.

278.

279.

Meunier et al. synaptic activity between nerve sprouts and their parent terminals. Proc. Natl. Acad. Sci. USA 96, 3200–3205. Lee, R. E., Tartell, P. B., Karmody, C. S., and Hunter, D. D. (1999) Association of adhesive macromolecules with terminal sprouts at the neuromuscular junction after botulinum treatment. Otolaryngol. Head Neck Surg. 120, 255–261. Santafe, M. M., Urbano, F. J., Lanuza, M. A., and Uchitel, O. D. (2000) Multiple types of calcium channels mediate transmitter release during functional recovery of botulinum toxin type A-poisoned mouse motor nerve terminals. Neuroscience 95, 227–234. Comella, J. X., Molgó, J., and Faille, L. (1993) Sprouting of mammalian motor nerve terminals induced by in vivo injection of botulinum type D toxin and the functional recovery of paralysed neuromuscular junctions. Neurosci. Lett. 153, 61–64. Molgó, J., Meunier, F. A., Faille, L., Cifuentes-Diaz, C., Comella, J. X., Popoff, M. R., and Poulain, B. (1999) Bourgeonnement des terminaisons nerveuses motrices déclenché par différents sérotypes de neurotoxines botuliques, in Dystonie, Neurone et Plasticité (Christen, Y., Nieoullon, A., and Rascol, O., eds.), Solal éditeur, Marseille, pp. 77–91. Holds, J. B., Alderson, K., Fogg, S. G., and Anderson, R. L. (1990) Motor nerve sprouting in human orbicularis muscle after botulinum A injection. Invest. Ophthalmol. Vis. Sci. 31, 964–967. Son, Y. J., Trachtenberg, J. T., and Thompson, W. J. (1996) Schwann cells induce and guide sprouting and reinnervation of neuromuscular junctions. Trends Neurosci. 19, 280–285. Gomez, S., Duchen, L. W., and Hornsey, S. (1982) Effects of x-irradiation on axonal sprouting induced by botulinum toxin. Neuroscience 7, 1023–1036. Sollner, T. and Rothman, J. E. (1994) Neurotransmission: harnessing fusion machinery at the synapse. Trends Neurosci. 17, 344–348. Sollner, T., Whiteheart, S. W., Brunner, M., Erdjument-Bromage, H., Geromanos, S., Tempst, P., and Rothman, J. E. (1993) SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318–324. Alderson, K., Yee, W. C., and Pestronk, A. (1989) Reorganization of intrinsic components in the distal motor axon during outgrowth. J. Neurocytol. 18, 541–552. Juzans, P., Molgó, J., Faille, L., and Angaut-Petit, D. (1996) Synaptotagmin II immunoreactivity in normal and botulinum type-A treated mouse motor nerve terminals. Pflügers Arch. 431(Suppl.), R283–R284. Tonge, D. A. (1974) Chronic effects of botulinum toxin on neuromuscular transmission and sensitivity to acetylcholine in slow and fast skeletal muscle of the mouse. J. Physiol. (Lond.) 241, 127–139. Betz, W. J., Mao, F., and Bewick, G. S. (1992) Activity-dependent fluorescent staining and destaining of living vertebrate motor nerve terminals. J. Neurosci. 12, 363–375. Magrassi, L., Purves, D., and Lichtman, J. W. (1987) Fluorescent probes that stain living nerve terminals. J. Neurosci. 7, 1207–1214. Young, S. H. and Poo, M. M. (1983) Spontaneous release of transmitter from growth cones of embryonic neurones. Nature 305, 634–637. Hume, R. I., Role, L. W., and Fischbach, G. D. (1983) Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes. Nature 305, 632–634. Zakharenko, S., Chang, S., O’Donoghue, M., and Popov, S. V. (1999) Neurotransmitter secretion along growing nerve processes: comparison with synaptic vesicle exocytosis. J. Cell Biol. 144, 507–518. Becher, A., Drenckhahn, A., Pahner, I., Margittai, M., Jahn, R., and Ahnert-Hilger, G. (1999) The synaptophysin-synaptobrevin complex: a hallmark of synaptic vesicle maturation. J. Neurosci. 19, 1922–1931. Verderio, C., Coco, S., Bacci, A., Rossetto, O., De Camilli, P., Montecucco, C., and


280.

281. 282.

283.

284.

285. 286.

287.

288.

289.

290.

291. 292. 293. 294.

295.

296. 297.

298.

345

Matteoli, M. (1999) Tetanus toxin blocks the exocytosis of synaptic vesicles clustered at synapses but not of synaptic vesicles in isolated axons. J. Neurosci. 19, 6723–6732. Eleopra, R., Tugnoli, V., Rossetto, O., De Grandis, D., and Montecucco, C. (1998). Different time courses of recovery after poisoning with botulinum neurotoxin serotypes A and E in humans. Neurosci. Lett. 256, 135–138. Keller, J. E., Neale, E. A., Oyler, G., and Adler, M. (1999) Persistence of botulinum neurotoxin action in cultured spinal cord cells. FEBS Lett. 456, 137–142. Criado, M., Gil, A., Viniegra, S., and Gutiérrez, L. M. (1999) A single amino acid near the C terminus of the synaptosome associated protein of 25 kDa (SNAP-25) is essential for exocytosis in chromaffin cells. Proc. Natl. Acad. Sci. USA 96, 7256–7261. Huang, X. H., Wheeler, M. B., Kang, Y. H., Sheu, L., Lukacs, G. L., Trimble, W. S., and Gaisano, H. Y. (1998) Truncated SNAP-25 (1-197), like botulinum neurotoxin A, can inhibit insulin secretion from HIT-T15 insulinoma cells. Mol. Endocrinol. 12, 1060–1070. Loewy, A., Liu, W.-S., Baitinger, C., and Willard, M. B. (1991) The major 35S-methionine-labeled rapidly transported protein (superprotein) is identical to SNAP-25, a protein of synaptic terminals. J. Neurosci. 11, 3412–3421. Lane, S. R. and Liu, Y. C. (1997) Characterization of the palmitoylation domain of SNAP25. J. Neurochem. 69, 1864–1869. Canaves, J. M. and Montal, M. (1998) Assembly of a ternary complex by the predicted minimal coiled-coil-forming domains of syntaxin, SNAP-25, and synaptobrevin. A circular dichroism study. J. Biol. Chem. 273, 34,214–34,221. Chapman, E., An, S., Barton, N., and Jahn, R. (1994) SNAP-25, a t-SNARE which binds to both syntaxin and synaptobrevin via domains that may form coiled coils. J. Biol. Chem. 269, 27,427–27,432. Raciborska, D. and Charlton, M. (1999) Retention of cleaved synaptosome-associated protein of 25 kDa (SNAP-25) in neuromuscular junctions: a new hypothesis to explain persistence of botulinum A poisoning. Can. J. Physiol. Pharmacol. 77, 679–688. Brown, M. C., Goodwin, G. M., and Ironton, R. (1977) Prevention of motor nerve sprouting in botulinum toxin poisoned mouse soleus muscles by direct stimulation of the muscle. J. Physiol. (Lond.) 267, 42P–43P. Lomo, T. (1976). The role of activity in the control of membranes and contractile properties of skeletal muscle, in Motor Innervation of Muscle (Thesleff, S., ed.), Academic Press, New York, pp. 289–312. Thesleff, S. (1989) Botulinal neurotoxins as tools in studies of synaptic mechanisms. Q. J. Exp. Physiol. 74, 1003–1017. Thesleff, S., Molgó, J., and Tägerud, S. (1990). Trophic interrelations at the neuromuscular junction as revealed by the use of botulinal neurotoxins. J. Physiol. (Paris) 84, 167–173. Mathers, D. A. and Thesleff, S. (1978) Studies on neurotrophic regulation of murine skeletal muscle. J. Physiol. (Lond.) 282, 105–114. Yang, J. S., Sladky, J. T., Kallen, R. G., and Barchi, R. L. (1991) TTX-sensitive and TTX-insensitive sodium channel mRNA transcripts are independently regulated in adult skeletal muscle after denervation. Neuron 7, 421–427. Tägerud, S., Libelius, R., and Thesleff, S. (1986) Effects of botulinum toxin induced muscle paralysis on endocytosis and lysosomal enzyme activities in mouse skeletal muscle. Pflügers Arch. 407, 275–278. Bambrick, L. and Gordon, T. (1987) Acetylcholine receptors and sodium channels in denervated and botulinum-toxin-treated adult rat muscle. J. Physiol. (Lond.) 382, 69–86. Yee, W. C. and Pestronk, A. (1987) Mechanisms of postsynaptic plasticity remodeling of the junctional acetylcholine receptor cluster induced by motor nerve terminal outgrowth. J. Neurosci. 7, 2019–2024. Bambrick, L. and Gordon, T. (1992) Neural regulation of acetylcholine receptors in rat neonatal muscle. J. Physiol. (Lond.) 449, 479–492.

346

Meunier et al.

299. Couteaux, R. (1978) Recherches morphologiques et cytochimiques sur l’organisation des tissues excitables. Robin et Mareuge, Paris. pp. 51–77. 300. Merlie, J. P. and Sanes J. R. (1985) Concentration of acetylcholine receptor mRNA in synaptic regions of adult muscle fibres. Nature 317, 66–68. 301. Burden, S. J. (1993) Synapse-specific gene expression. Trends Genet. 9, 12–16. 302. Moscoso, L. M., Chu, G. C., Gautam, M., Noakes, P. G., Merlie, J. P., and Sanes, J. R. (1995) Synapse-associated expression of an acetylcholine receptor-inducing protein, ARIA/heregulin, and its putative receptors, ErbB2 and ErbB3, in developing mammalian muscle. Dev. Biol. 172, 158–169. 303. Lipsky, N. G., Drachman, D. B., Pestronk, A., and Shih, P. J. (1989) Neural regulation of mRNA for the alpha-subunit of acetylcholine receptors: role of neuromuscular transmission. Exp. Neurol. 105, 171–176. 304. Witzemann, V., Brenner, H. R., and Sakmann, B. (1991) Neural factors regulate AChR subunit mRNAs at rat neuromuscular synapses. J. Cell Biol. 114, 125–141. 305. Koltgen, D., Ceballos-Baumann, A. O., and Franke, C. (1994) Botulinum toxin converts muscle acetylcholine receptors from adult to embryonic type. Muscle Nerve 17, 779–784. 306. McMahan, U. J. (1990) The agrin hypothesis. Cold Spring Harb. Symp. Quant. Biol. 55, 407–418. 307. Gautam, M., Noakes, P. G., Moscoso, L., Rupp, F., Scheller, R. H., Merlie, J. P., and Sanes, J. R. (1996) Defective neuromuscular synaptogenesis in agrin-deficient mutant mice. Cell 85, 525–535. 308. Burgess, R. W., Nguyen, Q. T., Son, Y. J., Lichtman, J. W., and Sanes, J. R. (1999) Alternatively spliced isoforms of nerve- and muscle-derived agrin: their roles at the neuromuscular junction. Neuron 23, 33–44. 309. Valenzuela, D. M., Stitt, T. N., DiStefano, P. S., Rojas, E., Mattsson, K., Compton, D. L., et al. (1995) Receptor tyrosine kinase specific for the skeletal muscle lineage: expression in embryonic muscle, at the neuromuscular junction, and after injury. Neuron 15, 573–584. 310. Fischbach, G. D. and Rosen, K. M. (1997) ARIA: a neuromuscular junction neuregulin. Annu. Rev. Neurosci. 20, 429–458. 311. Sanes, J. R. and Lichtman, J. W. (1999) Development of the vertebrate neuromuscular junction. Annu. Rev. Neurosci. 22, 389–442. 312. Ishii, D. N. (1989) Relationship of insulin-like growth factor II gene expression in muscle to synaptogenesis. Proc. Natl. Acad. Sci. USA 86, 2898–2902. 313. Caroni, P., Schneider, C., Kiefer, Mc., and Zapf, J. (1994) Role of muscle insulin-like growth factors in nerve sprouting: suppression of terminal sprouting in paralyzed muscle by IGF-binding protein 4. J. Cell Biol. 125, 893–902. 314. Skene, J. H. (1989) Axonal growth-associated proteins. Annu. Rev. Neurosci. 12, 127–156. 315. Benowitz, L. I. and Routtenberg, A. (1997) GAP-43: an intrinsic determinant of neuronal development and plasticity. Trends Neurosci. 20, 84–91. 316. Bisby, M. A., Tetzlaff, W., and Brown, M. C. (1996) GAP-43 mRNA in mouse motoneurons undergoing axonal sprouting in response to muscle paralysis of partial denervation. Eur. J. Neurosci. 8, 1240–1248. 317. Frey, D., Laux, T., Xu, L., Schneider, C., and Caroni, P. (2000) Shared and unique roles of CAP23 and GAP43 in actin regulation, neurite outgrowth, and anatomical plasticity. J. Cell Biol. 149, 1443–1454. 318. Hassan, S. M., Jennekens, F. G. I., Wieneke, G., and Veldman, H. (1994) Calcitonin gene-related peptide like immunoreactivity, in botulinum toxin-paralysed rat muscles. Neuromusc. Disord. 4, 489–496. 319. Sala, C., Andreose, J. S., Fumagalli, G., and Lomo, T. (1995) Calcitonin gene-related peptide: posible role in formation and maintenance of neuromuscular junctions. J. Neurosci. 15, 520–528.


347

320. Meunier, F. A., Colasante, C., Faille, L., Gastard, M., and Molgó, J. (1996) Upregulation of calcitonin gene-related peptide at mouse motor nerve terminals poisoned with botulinum type-A toxin. Pflügers Arch. 431(Suppl.), R297–R298. 321. Tarabal, O., Calderó, J., Rivera, J., Sorribas, A., Lopez, R., Molgó, J., and Esquerda, J. E. (1996) Regulation of motoneural calcitonin gene-related peptide (CGRP) during axonal growth and neuromuscular synaptic plasticity induced by botulinum toxin in rats. Eur. J. Neurosci. 8, 829–836. 322. Tarabal, O., Calderó, J., and Esquerda, J. E. (1996) Intramuscular nerve sprouting induced by CNTF is associated with increases in CGRP content in mouse motor nerve terminals. Neurosci. Lett. 219, 60–64. 323. Changeux, J. P., Duclert, A., and Sekine, S. (1992) Calcitonin gene-related peptides and neuromuscular interactions. Ann. NY Acad. Sci. 657, 361–378. 324. Sanes, J. R., Appel, E. D., Burgess, R. W., Emerson, R. B., Feng, G., Gautam, M., et al. (1998) Development of the neuromuscular junction: genetic analysis in mice. J. Physiol. (Paris) 92, 167–172. 325. Salmon, A. M., Damaj, I., Sekine, S., Picciotto, M. R., Marubio, L., and Changeux, J. P. (1999) Modulation of morphine analgesia in alphaCGRP mutant mice. Neuroreport 10, 849–854. 326. Booth, C. M., Kemplay, S. K., and Brown, M. C.(1990) An antibody to neural cell adhesion molecule impairs motor nerve terminal sprouting in a mouse muscle locally paralysed with botulinum toxin. Neuroscience 35, 85–91. 327. Walsh, F. S., Hobbs, C., Wells, D. J., Slater, C. R., and Fazeli, S. (2000) Ectopic expression of NCAM in skeletal muscle of transgenic mice results in terminal sprouting at the neuromuscular junction and altered structure but not function. Mol. Cell. Neurosci. 15, 244–261. 328. Chiquet-Ehrismann, R. (1995) Tenascins, a growing family of extracellular matrix proteins. Experientia 51, 853–862. 329. Werle-Haller, B. and Chiquet, M. (1993) Dual function of tenascin: simultaneous promotion of neurite growth and inhibition of glial migration. J. Cell Sci. 106, 597–610. 330. Daniloff, J. K., Crossin, K. L., Pinçon-Raymond, M., Murawsky, M., Rieger, F., and Edelman, G. M. (1989) Expression of cytotactin in the normal and regenerating neuromuscular system. J. Cell Biol. 108, 625–635. 331. Cifuentes-Diaz, C., Velasco, E., Meunier, F. A., Goudou, D., Belkadi, L., Faille, L., et al. (1998) The peripheral nerve and the neuromuscular junction are affected in the tenascinC-deficient mouse. Cell. Mol. Biol. 44, 357–379. 332. Cifuentes-Diaz, C., Meunier, F. A., Velasco, E., Faille, L., Goudou, D., Belkadi, L., et al. (1998) Morphological alterations of motor nerve terminals after botulinum type-A poisoning or reinnervation of skeletal muscle in the tenascin-C deficient mouse. J. Physiol. (Paris) 92, 421–422.

348

Meunier et al.

Sentinels of Environmental Biotoxins

349

18 Marine Mammals as Sentinels of Environmental Biotoxins Vera L. Trainer

1. INTRODUCTION Harmful algal blooms (HABs), commonly known as “red tides,” are documented in almost every coastal region of the world. Over the past decade, these toxic blooms appear to have increased in number, magnitude, and seasonal duration, potentially due to the spread of toxic algal species to new areas, eutrophication caused by human development of coastal areas, and global climate change. Some of the microscopic, singlecelled algae that constitute the base of the marine food chain produce the potent toxins found in these harmful blooms. These toxins can accumulate in fish, shellfish, and other marine organisms, and move through the food chain, at times affecting the highest consumers, including marine mammals and humans. Humans are usually well-protected by federal and state monitoring programs that detect the toxins at an early stage and restrict the harvest or sale of the toxic seafood. However, marine mammal health is not protected by the same routine monitoring programs. Recent evidence has indicated that mass mortalities and strandings of whales, dolphins, sea lions, manatees, and sea otters, in some cases, may be caused by their exposure to marine biotoxins. 2. ALGAL TOXINS The possibility that marine mammal mortalities are caused by toxic algal blooms has been indicated in the scientific record since the late 1800s (1). In a mass mortality of animals off the Gulf coast of Florida between November 1946 and August 1947 (1), it was documented that “the mass death of marine organisms was associated with a flowering of the dinoflagellate, Gymnodinium breve (recently placed into the new genus Karenia as Karenia brevis, but the previous nomenclature is used in this chapter). In some places, G. breve reproduced so abundantly that patches of the water became saffron yellow in color and noticeably viscous. Schools of fishes entering this water died immediately.” Lethal and sublethal effects of algal toxins on turtles, fish, invertebrates, and sea grasses have also been documented (see, for example, refs. 2–7). However, only re-


349

350

Table 1 Algal Toxins and Their Effects on Marine Mammals Major toxin structure(s)a

Symptoms of toxin exposure in mammalsb

Toxin

Toxic syndrome

Site of action

Mode of action

Saxitoxin

Paralytic shellfish poisoning (PSP)

Sodium channel site 1

Inhibits ion conductance

Mild case: dizziness, nausea, vomiting, diarrhea. Severe case: muscular paraly sis, respiratory difficulty, choking sensation, death through respiratory paralysis.

Sodium channel site 5

Repetitive firing, shift voltage dependence of activation

Incoordination, inability to maintain a righting reflex, muscle fasciculations. Marine mammal would likely stop eating and lose buoyancy.

Brevetoxin Neurotoxic shellfish poisoning (NSP)

Domoic acid

Amnesic shellfish Kainate-sensitive Receptor-induced poisoning (ASP) or glutamate receptor depolarization domoic acid and excitation poisoning

Decreased reaction to severe pain, hallucinations, confusion, memory loss, seizures, opisthotonus, death may occur.

aThe major backbone structure of a common toxin isoform is shown here. Asterisks indication the position of 3H in commercially available radioactive derivatives

Trainer

(used in receptor-binding assays). Detailed structures of all toxin isoforms can be found in a number of references including Hall et al. (9) and Baden and Trainer (10). bThis list is by no means comprehensive, but rather details those symptoms that either have been observed in marine mammals or are believed to possibly occur in marine mammals due to observations of toxin exposure in terrestrial mammals.


351

cently have sufficient clinical, biochemical, and environmental data been collected to characterize thoroughly a marine mammal poisoning event by measuring the transfer of toxin from algae to a vector organism and finally to the stranded mammal (8). A number of different toxins are now implicated in the stranding and death of marine mammals (Table 1). These include saxitoxin, produced by dinoflagellates of the genera Gonyaulax, Alexandrium, and Pyrodinium; brevetoxin, produced by the dinoflagellate, G. breve; and domoic acid, produced by diatoms of the genus Pseudo-nitzschia. The structure, chemical characteristics, and mode of action of these toxins help to determine the concentration, efficacy, and route of exposure that cause a specific physiological response in the affected organism. For example, the water-soluble toxins, saxitoxin and domoic acid, concentrate in the metabolically active, physiologically sensitive tissues such as the heart and brain. The lipophilic toxin, brevetoxin, collects in lipid-rich tissues such as blubber, allowing for toxin retention in the organism for long periods of time. 3. RECENT US STRANDING EVENTS Stranding events caused by HABs in US coastal waters have recently been documented in detail, largely due to organized volunteer efforts, coordinated scientific responses, and improved analytical techniques. In some cases, a large amount of circumstantial evidence has indicated the strong possibility that algal toxins were the cause of death; in other cases, histological, biochemical, and environmental data have led to firm conclusions. Some toxins are quickly metabolized and excreted; therefore, toxin analysis of mammal tissues and body fluids may, at times, give a negative result. However, this does not necessarily indicate that the animal did not die from toxin exposure; merely that toxin is no longer detectable within the organism. All coastal areas of the United States have now documented marine mortalities believed to be a result of algal toxin exposure. Worldwide, many incidences of endangered and protected species mortalities are suspected to be due to toxins from HABs. For example, over 100 monk seals (Monachus monachus) in the Mediterranean (11), as well as sea otter (Enhydra lutris) and Stellar sea lion (Eumetopias jubatus) deaths in Alaska (12; Plumley, G., personal communication) are thought to have been caused by paralytic shellfish poisoning (PSP). The most thoroughly investigated marine mammal toxification events in US coastal waters will be described here; however, this is by no means a complete coverage of marine mammal mortalities due to algal toxins worldwide. On the Gulf coast of Florida, G. breve has been implicated in mortalities of the endangered Florida manatee, Trichechus manatus latirostris, in both 1963 (13,14), and 1996 (15,16). In 1996, at least 149 manatees died and the presence of brevetoxin was immunohistochemically demonstrated in several manatees (16). Several lines of evidence implicated a neurotoxin as the cause of death. Severe lesions of the upper respiratory tract and catarrhal rhinitis were likely due to inhalation of aerosolized brevetoxin (16). Four manatees that were rescued exhibited neurological signs that included muscle fasciculations, incoordination, and the inability to maintain a righting reflex (16). Since brevetoxins are aerosolized (17,18,19,20), a possible route of manatee exposure was through inhalation; however, the stomach contents of three manatees contained tunicates, a possible filter-feeding toxin vector (15).

352

Trainer

Brevetoxin-associated mortality was also postulated in bottlenose dolphins off southwestern Florida from 1946–1947 (1) and along the US Atlantic coast from 1987–1988 (21). In the event beginning in June 1987, over 740 bottlenose dolphins stranded during an 11-mo period. This was estimated to be 50% or more of the coastal migratory stock between Florida and New Jersey (22). The key pieces of evidence that pointed to brevetoxicosis as a cause of death were the presence of brevetoxin in the viscera of menhaden taken from the stomach of one dolphin, and in the livers of three nursing calves and several adults. Menhaden and thread herring collected from the area in which dolphins stranded also tested positive for brevetoxin in their livers. Although direct observations of intoxication of marine mammals by brevetoxin are difficult to make, symptoms of poisoning in humans are known to include nausea, vomiting, diarrhea, reversal of temperature sensation, ataxia, and numbness and tingling of extremities (23). A dolphin affected by brevetoxin would likely stop eating, eventually exhausting its blubber reserve, resulting in loss of its buoyancy and thermal shield. The stress associated with these problems would set the stage for infection by the ubiquitous opportunistic organisms that were also isolated in the affected dolphins (21). Deaths of humpback whales (Megaptera novaeangliae) off the coast of Cape Cod, Massachusetts were also reported in 1987 but were believed to be due to toxins produced by other dinoflagellate species (24). These whales died suddenly at sea and many still had fish in their stomachs, evidence of recent feeding. One whale observed close to the beach appeared to behave normally but was dead 90 min later, possibly indicative of sudden paralysis (24). These animals had considerable blubber reserves, indicating that they were in good health prior to death. Analysis of mackerel that the whales had been eating revealed the presence of saxitoxin in liver, kidney, and viscera of this fish. Extracts of whale kidney, liver, and viscera also tested positive for saxitoxin or its derivatives. Most recently, beginning in the spring of 1998, over 70 California sea lions (Zalophus californicus) and one northern fur seal (Collorhinus ursinus) stranded along the central California coast (8,25). Adult sea lions and northern fur seals suffering from neurological dysfunction were also documented in 1978, 1986, and 1992 (25). However, during the 1998 stranding event, a collaborative effort among a number of scientists allowed a detailed documentation of the algal bloom and the transfer of toxin to the mammals. These animals had neurological signs, including seizures, which often became increasingly frequent, resulting in opisthotonus (arching of the back), then death. Some sea lions were observed wandering in confusion along the roads that border Monterey Bay (Scholin, C., personal communication). Most of these animals died despite treatment. Several lines of evidence indicated their poisoning was due to ingestion of domoic acid. The predominant histological lesion in affected animals was neuronal necrosis that was most severe in the dentate gyrus and CA3 and CA4 zones of the hippocampus, damage characteristic of domoic acid poisoning. Domoic acid was detected in serum, urine, and feces of several of the stranded animals by three separate analytical methods: the receptor-binding assay, high-performance liquid chromatography (HPLC), and mass spectroscopy (see Subheading 5. for a description of these techniques). Several blooms of Pseudo-nitzschia australis were observed off the central California coast in the locations where sea lions stranded. Seawater samples containing


353

this phytoplankton species were shown to contain domoic acid by receptor binding assay. The stomachs of anchovies collected from Monterey Bay, where several of the sea lions had stranded, contained high amounts of domoic acid and frustules of P. australis. Fecal samples from two sea lions in which domoic acid was detected also contained P. australis frustules (8,26). This combination of clinical signs, histopathology, epidemiology, and oceanography led to the diagnosis of domoic acid toxicity in the mammals. A similar domoic acid poisoning of sea lions occurred again beginning in June 2000, involving over 90 animals (Gulland, F., personal communication), indicating that these events may be more common than previously believed. 4. ROUTES OF EXPOSURE The routes of exposure of marine mammals to algal toxins can vary. The transfer of toxins to these animals, however, is often poorly understood. Only by studying the clinical signs during a poisoning event, fully understanding predator-prey relationships, knowing the animal’s preferred habitat and typical feeding behavior, and completing rigorous studies of toxin potencies and bioavailability, can toxin transfer mechanisms be known. Possible routes of exposure include ingestion of prey or seawater containing toxic algae, and inhalation, in the case of brevetoxin. Because shellfish are the only organisms typically monitored for the prevention of human poisoning by algal toxins, a number of other possible vector species important in the transfer of toxins to marine mammals have been overlooked. For example, recent evidence indicates that spiny mole crabs (Blepharipoda occidentalis), small crustaceans that feed on detritus and algae, likely transferred toxins to sea otters in central California, about 1 mo after the sea lion deaths in 1998 (27; Kvitek, R., personal communication). It is possible that decaying, less buoyant, but still toxic Pseudo-nitzschia blooms sink into the benthic layer where they affect another habitat in the marine food chain. A survey of the ability of marine mammal prey to concentrate and sequester toxins is important in the complete characterization of marine mammal intoxications. An unexpected route of exposure of marine mammals to toxins may also occur through mother’s milk to newborn calves. In fact, brevetoxin was found in the livers of three nursing dolphin calves during the 1987 mortality event. It is believed that the toxin was delivered to calves via the mother’s milk (21). Laboratory studies in rats have shown that neonates are particularly sensitive to domoic acid (28), therefore the transfer of toxin via breast milk is of particular concern. This indicates the need to fully study the mode of toxin transfer at all levels of the food web and in all life stages of an exposed organism. 5. METHODS OF DETECTION As we continue into the 21st century, methods of toxin and toxic cell detection will become more sensitive, highly automated, and accurate. Currently, comprehensive diagnoses of marine mammal poisoning events are possible due to the large number of complementary analytical techniques that are available. A summary of these methods, including the characteristics and limitations of each, is listed in Table 2. Some techniques are used specifically as screening tools, in order to quickly detect toxins or toxic cells in a given sample. Examples of these screening tools are the enzyme-linked immunosorbent assay (ELISA), cell identification using light microscopy, cell-based

354

Table 2 Methods of Toxin and Toxigenic Organism Detection Assay

Performance

1. Light microscopy

Visual inspection of phytoplankton cells using a microscope.

2. ELISA

Recognition of toxin structure using antibodies. Measure dose-dependent cell mortality caused by a specific toxin action. Displacement of radiolabeled toxin from receptor by toxin in sample. Results compared to a standard curve. Recognition of specific genetic sequence by a complementary molecular probe. Separation of toxin by migration through a packed column. Comparison of sample migration to a toxin standard.

3. Cell toxicity

4. Receptor binding

5. Molecular probes 6. HPLC

7. Mouse bioassay 8. SEM

9. Mass spectrometry

Injection of sample into a mouse and observation of time of death. Electron microscopy of plankton sample to determine fine structure and to identify to the species level. Separation and identification based directly on molecular weight.

Field or lab Lab. Can be done in the field, but field microscopes are less effective. Lab. Future modification for field. Lab

Disadvantage

Advantage

Time-consuming

Standard method of phytoplankton identification.

Often does not recognize all toxin structures. Requires specialized equipment. Some false-negatives.

Results can be read by eye. High throughput analysis is possible. Fast results that can be read by eye.

Lab

Requires use of radioisotopes

High through-put sample analysis is possible.

Lab. Currently being tested for use on buoys.

Probes must be fine-tuned for each geographical region. Time-consuming. Toxin retention times are not always unique. Uses expensive and hazardous solvents. Uses live animals. Total cost is high. Requires facilities. Highly time-consuming. Difficult to enumerate cells.

Tests can be automated (sandwich assay, buoy formats). High throughput analysis is possible. Standard method of toxin detection recognized by regulatory agencies. Can be automated for processing large numbers of samples.

Time-consuming, requires expensive equipment, trained operator.

Combined with HPLC, isomers of toxins can be identified and structural information can be deduced. Powerful confirmation technique.

Lab

Lab Lab

Lab

Standard method recognized by regulatory agencies. Sometimes the only method for identification of species.

Trainer


355

toxicity assays, receptor-binding assays, and molecular-based toxic species determination. Other techniques are used for confirmation of the presence of toxins, including HPLC and liquid chromatography-tandem mass spectroscopy (LC-MS/MS). The diagnosis of toxic effects at the cellular level of the affected organisms relies on methods such as immunocytochemistry and histopathology. Scanning electron microscopy (SEM), capable of magnification at 10,000-fold or greater, is used to confirm the species of algae present in a seawater sample or even in the viscera of the poisoned animal. This is an intensive, time-consuming method that, in some cases, is the only means of knowing whether a potentially toxic algal species is present in a given water mass. The key to rapid diagnosis of marine mammal poisoning by algal toxins in the future will be automation of analyses, including real-time detection of HABs in the areas where mammals are affected. The ecological significance of the sea lion mortalities in central California in the spring of 1998 became a subject of study only after reports of marine mammal deaths were made. In this event, marine mammals were the sentinels of a large-scale toxicbloom event. In the future, more appropriate sentinels of biotoxin threats to coastal ecosystems must be used, allowing for early warning of impending blooms. This HAB “forecasting” capability will be possible with buoys that can provide real-time data at offshore locations where toxic blooms are known to originate. The detection of toxic cells using molecular probes on automated buoys is currently being tested for effectiveness and accuracy (29). Finally, on cloudless days, satellite remote sensing can be used to track coastal water masses and chlorophyll signatures in which toxic algae may reside. Only when a suite of real-time, automated methods is developed for use synergistically will early warning and accurate, rapid detection of HABs be possible. 6. MARINE MAMMAL SUSCEPTIBILITY TO TOXINS The sequence of events in toxic episodes during HABs begins with the entry of toxin into the animal. After ingestion or inhalation, a toxin must reach its target, a specific site of action on nerve membranes. A recent study has documented the high-affinity, specificity, and reversible binding of both saxitoxin and brevetoxin to marine mammal brain tissue (30). Similar research that documents the affinity of domoic acid binding to marine mammal brain tissue in vitro is in progress. The affinities of brevetoxin and saxitoxin for marine mammal nerve receptors are similar to those that have been demonstrated for terrestrial mammals, known to have a strong biological response to toxin exposure (Table 3). In fact, because of its high affinity for neurotoxins, the mouse is currently used as the primary diagnostic organism for detection of toxins in seafood by regulatory agencies worldwide (31). It is suspected that other physiological factors may increase the effect of a given dose of toxin, especially in cetaceans. Upon toxin exposure, a whale may lose control of its vital peripheral heat-conserving mechanism (32) or become unable to return to the surface to breathe because of peripheral nerve impairment (33). During a cetacean dive, toxin will be concentrated in the heart and brain because of increased blood flow to those organs (34), thereby restricting the elimination of toxin by the liver and kidney. Water-soluble saxitoxin would tend to concentrate in these physiologically-sensitive, metabolically-active areas of the organism (30).

356

Trainer

Table 3 Brevetoxin and Saxitoxin Binding Affinity in Excitable Tissue Organism Terrestrial Cat Rat Chick Locust Drosophila Aquatic Squid Eel Tilapia Gambusia Trout Manatee Gray whale Sea lion Humpback whale

KD Brevetoxin 2.9

Saxitoxin 0.6 2.0a 0.2b 0.3 0.5 1.9 4.3 6.0

6.1 10.0 7.5

3.8 3.0 2.7 1.9 4.9

KD values are expressed in nM. Amphibians are not included. Table reprinted with permission from ref. (30). aMeasurement made at 36°C. bMeasurement made at 4°C.

Marine mammals that already have compromised health, such as a Gram-negative bacterial infection that was prominent in some manatees in the 1996 Florida incident (16), may be particularly sensitive to further health stressors such as toxins from HABs. For example, a Vibrio species was present at the same time that brevetoxins were identified in tissues of dolphin that had stranded off the US East coast. Other physical factors, such as pregnancy, may also play a role as stressors. In the 1998 sea lion stranding event, over 50% of the poisoned animals were pregnant females (25). Alternatively, the health of animals could initially be compromised by exposure to toxins, thereby making them more susceptible to subsequent opportunistic infections. Agerelated differences in susceptibilities to marine toxins have also been demonstrated. In the case of the domoic acid poisoning due to the ingestion of toxic mussels by humans in eastern Canada in 1987, older patients were more likely to have memory loss and to require hospitalization (35). A similar age-dependent sensitivity to biotoxins could be expected in marine mammals. 7. ENVIRONMENTAL CONSIDERATIONS Environmental factors play an important role in determining the dose of toxin to which the marine mammal is exposed. Dilution effects will clearly be present in an aqueous system, thereby requiring a physical means of toxin concentration for lethal exposure to result. For example, at the times of manatee deaths that occurred in conjunction with G. breve blooms in 1963, 1982, and 1996, large numbers of manatees congregated in canals where upstream power plants warm the water (15). G. breve


357

Fig. 1. Food chain transfer of algal toxins. (1) Toxin production by phytoplankton is influenced by environmental conditions that must be favorable for HAB development. The HAB can be concentrated by physical trapping of cells (e.g., in canals) or by layering of cells in the water column. HABs are diluted by physical methods such as currents or storms. (2) Toxin is transferred to filter feeders or planktivorous fish, resulting in bioaccumulation. Subsequent toxin transfer can be short-term (high-dose) or long-term (low-dose). (3) The toxin action is initiated by recognition of a specific binding site in the affected organism. Particular organs are targeted and affinity of the toxin for the receptor will be directly proportional to the dose required for a biological response to occur. Toxin can be removed from the system by excretion. (4) The toxin action at a specific receptor will result in a physiological response, including illness and in extreme cases, death. Other mechanisms of toxin transfer are possible but not depicted here, such as by inhalation of aerosolized toxin or ingestion of zooplankton.

blooms persisted in the canals for several days, resulting in the sustained exposure of manatees to brevetoxin. These blooms do not usually penetrate the upper reaches of bays and estuaries, however, in early 1982, drought and water-management practices reduced the amount of freshwater input into those areas (14), thereby limiting flushing rates. Environmental factors may act in conjunction with prey availability to provide optimal conditions for toxin transfer. For example, the 1998 Pseudo-nitzschia bloom on the central California coast occurred at the end of a large El Niño event. It is known that food availability is greatly reduced in El Niño years, evidenced in the spring of 1998 by the many emaciated sea lion pups that were found dead on the same beaches where adults with neurological symptoms stranded (25). It is possible that the schools of domoic acid-containing anchovies and sardines were one of the few food sources available in key locations along central California coast that season. Recurring blooms of harmful algae have been documented in many of the areas where marine-mammal mortalities are observed. The acute effects of blooms on mammals have received the most attention because of the immediate impact on populations, the most serious of which is death. It has been suggested that in the manatee stranding

358

Trainer

event attributed to brevetoxin in 1996, mortality may not necessarily have been acute, but could have occurred after days or perhaps even weeks after inhalation or ingestion of subacute concentrations of toxin (16). Manatees are exposed to brevetoxin repeatedly throughout their lives in southwest Florida, and the concentration of brevetoxin to which these manatees are chronically exposed is unknown (14). It is possible that HABs adversely affect mammal populations by having sublethal, chronic impacts (e.g., reduced fecundity, growth, and behavior). For example, okadaic acid, a toxin produced by benthic dinoflagellates of the genus Prorocentrum, is a known tumor promoter that is suspected to cause chronic problems in aquatic animals (7,36). These chronic effects, which may have long-term consequences on population health and vigor, have received little attention. A schematic of the key environmental and physiological factors involved in toxin transfer is shown in Fig. 1. Finally, marine toxins have been shown to act synergistically with several pyrethroids (37), a group of synthetic insecticidal compounds. It is therefore possible that the effects of biotoxins may be magnified in polluted coastal areas. The additive effects of the suite of natural and anthropogenic toxins must be carefully studied in order to assess the potential damage of these agents in nearshore marine ecosystems. Likewise, the potential for sublethal impacts of HABs on marine mammal populations may be modeled in controlled laboratory studies. Better diagnostic tools for marine mammal exposure will aid in determining low-level doses of toxins to which these animals are perhaps routinely exposed. 8. FUTURE RESEARCH As this chapter is being written, there are reports of sea otters with seizures and ear scratching behavior stranding on the central Washington coast, signs that may indicate domoic acid poisoning. These otters are usually found along the Northern coast, in rocky coastal areas, where they typically feed on sea urchins. Perhaps population pressures have forced them into more Southern habitats, where their feeding behaviors are changed. The likely food sources for these otters in the more southern Washington coastal areas are razor clams and Dungeness crabs, organisms that are known to concentrate domoic acid (38). How can scientists quickly determine whether or not algal toxins are the cause of the death of these sea otters? Currently, time-consuming laboratory analyses can be performed on urine, blood, and fecal samples, but few laboratories have the expertise to run these tests. Samples must be frozen or kept cold prior to analysis. Antibody-based assays show great promise for modification into “test kits” that can be used immediately in the field, especially the ELISA for domoic acid (39). A quick diagnostic test for the detection of brevetoxin in blood is also in the developmental stage (40). This test involves placing a small spot of blood on a collection card that can be mailed at ambient temperature to a lab for analysis. This type of assay opens up the possibility of biomonitoring for toxins in larger numbers of both live and freshly stranded animals. Nasal and throat swabs for brevetoxin detection also show promise as a biomarker of exposure that can be performed using live subjects (41). Effective diagnostics will give scientists insights into the potential impacts of HABs, but can these toxic blooms ultimately be controlled or even prevented? Although HABs are natural phenomena, they may be stimulated by human activities such as pollution


359

or habitat alteration. Because of their position near the top of the food chain, mammals are sentinels of environmental health and population well-being. In order to assist in preservation of marine mammal populations, the environmental influences on toxin production and the means of toxin transfer through the food chain to marine mammals should be rigorously studied. Altered habitat quality, changed nutrient supplies, and global warming are human impacts on coastal marine environments that potentially affect the health and welfare of marine mammal populations. Although these are difficult problems to address, ensuring the welfare of our marine ecosystems will ultimately protect the complex interdependence of global biological systems. REFERENCES 1. Gunter, G., Williams, R. H., Davis, C. C., and Smith F. G. W. (1948) Catastrophic mass mortality of marine animals and coincident phytoplankton bloom on the west coast of Florida, November 1946 to August 1947. Ecolog. Monogr. 18, 309–324. 2. Tracey, G. A. (1998) Feeding reduction, reproductive failure, and mortality in Mytilus edulis during the 1985 “brown tide” in Narragansett Bay, Rhode Island. Mar. Ecol. Prog. Ser. 50, 73–81. 3. Dennison, W. C., Marshal, G. J., and Wigand, C. (1989) Effect of ‘brown tide’ shading on eelgrass (Zostera marine L.) distributions, in Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Coastal and Estuarine Studies, vol. 35 (Cosper, E. M., Bricelj, V. M., and Carpenter, E. J., eds.), Springer Verlag, Berlin, pp. 675–692. 4. Burkholder, J. M., Noga, E. J., Hobbs, C. H., and Glasgow, Jr., H. B. (1992) New “phantom” dinoflagellate is the causative agent of major estuarine fish kills. Nature 358, 407–410. 5. Taylor, F. J. R. (1993) Current problems with harmful phytoplankton blooms in British Columbia waters, in Toxic Phytoplankton Blooms in the Sea (Smayda, T. J. and Shimizu Y., eds.), Elsevier Science, Amsterdam, pp. 699–703. 6. Landsberg, J. H. (1995) Tropical reef-fish disease outbreaks and mass mortalities in Florida, USA: what is the role of dietary biological toxins? Dis. Aquat. Org. 22, 83–100. 7. Landsberg, J. H., Balazs, G. H., Steidinger, K. A., Baden, D. G., Work, T. M., and Russell, D. J. (1999) The potential role of natural tumor promoters in marine turtle fibropapillomatosis. J. Aq. Animal Health 11, 199–210. 8. Scholin, C.A., Gulland, F., Doucette, G. J., Benson, S., Busman, M., Chavez, et al. (2000) Mortality of sea lions along the central California coast linked to a toxic diatom bloom. Nature 403, 80–84. 9. Hall, S., Strichartz, G. R., Moczydlowski, E., Ravindran, A., and Reichardt, P. B. (1990) The saxitoxins: sources, chemistry, and pharmacology, in Marine Toxins (Hall, S. and Strichartz, G. R., eds.), ACS Symposium Series 418. American Chemical Society Washington, DC, pp. 29–65. 10. Baden, D. G. and Trainer, V. L. (1993) Mode of action of toxins of seafood poisoning, in Algal Toxins in Seafood and Drinking Water (Falconer, I. R., ed.), Academic Press, New York, pp. 49–74. 11. Hernandez, M., Robinson, I., Aguilar, A., Gonzalez, L. M., Lopez-Jurado, L. F., Reyero, M. I., et al. (1998) Did algal toxins cause monk seal mortality? Nature 393, 28–29. 12. DeGange, A. R. and Vacca, M. M. (1989) Sea otter mortality at Kodiak Island, Alaska, during summer 1987. J. Mammol. 70(4), 836–838. 13. Layne, J. N. (1965) Observations on marine mammals in Florida waters. Bull. Fl. State Mus. 9, 131–181.

360

Trainer

14. O’Shea, T. J., Rathbun, G. B., Bonde, R. K., Buergelt, C. D., and Odell, D. K. (1991) An epizootic of Florida manatees associated with a dinoflagellate bloom. Mar. Mamm. Sci. 7(2), 165–179. 15. Landsberg, J. H. and Steidinger, K. A. (1998) A historical review of Gymnodinium breve red tides implicated in mass mortalities of the manatee (Trichechus manatus latirostris) in Florida, USA, in Harmful Algae (Reguera, B., Blanco, J., Fernandez, M. L., and Wyatt, T., eds.), UNESCO, Paris, pp. 97–100. 16. Bossart, G. D., Baden, D. G., Ewing, R. Y., Roberts, B., and Wright, S. D. (1998) Brevetoxicosis in manatees (Trichechus manatus latirostris) from the 1996 epizootic: gross, histologic, and immunohistochemical features. Toxicol. Path. 26, 276–282. 17. Steidinger, K. A., Burklew, M. A., and Ingle, R. M. (1973) The effects of Gymnodinium breve toxin on estuarine animals, in Marine Pharmacognosy: Action of Marine Toxins at the Cellular Level (Martin, D. F. and Padilla, G. M., eds.), Academic Press, New York, pp. 179–202. 18. Steidinger, K. A. and Haddad, K. (1981) Biologic and hydrographic aspects of red tides. Bioscience 31, 814–819. 19. Baden, D. G. and Mende, T. J. (1982) Toxicity of two toxins from the Florida red tide marine dinoflagellate, Gymnodinium breve. Toxicon 20, 457–461. 20. Pierce, R. H. (1986) Red tide (Gymnodinium breve) toxin aerosols: a review. Toxicon 24, 955–965. 21. Geraci, J. R. (1989) Clinical investigation of the 1987–88 mass mortality of bottlenose dolphins along the U.S. central and south Atlantic coast. Final report to National Marine Fisheries Service, U.S. Navy, Office of Naval Research, and Marine Mammal Commission, pp. 63. 22. Scott, G. P., Burn, D. M., and Hansen, L. J. (1988) The dolphin dieoff: long-term effects and recovery of the population. Proceedings Oceans ’88, Baltimore, MD. pp, 819–823. 23. Baden, D. G. (1983) Marine food-borne dinoflagellate toxins. Int. Rev. Cytol. 82, 99–150. 24. Geraci, J. R., Anderson, D. M., Timperi, R. J., St. Aubin, D. J., Early, G. A., Prescott, J. H., and Mayo, C. A. (1989) Humpback whales (Megaptera novaeangliae) fatally poisoned by dinoflagellate toxin. Can. J. Fish. Aquat. Sci. 46, 1895–1898. 25. Gulland, F., et al. (2000) Domoic acid toxicity in California sea lions (Zalophus californicus) stranded along the central California coast, May–October 1998. Report to the National Marine Fisheries Service Working Group on Unusual Marine Mortality Events. US Dept. of Commerce, NOAA Tech. Memo. NMFS-OPR-17, 45 p. 26. Lefebvre, K. A., Powell, C. L., Doucette, G. J., Silver, J. B., Miller, P. E., Hughes, M. P., et al. (1999) Detection of domoic acid in northern anchovies and California sea lions associated with an unusual mortality event. Nat. Toxins 7, 85–92. 27. Trainer, V. L., Adams, N. G., Bill, B. D., Stehr, C. M., Wekell, J. C., Moeller, P., et al. (2000) Domoic acid production near California coastal upwelling zones, June 1998. Limnol. Oceanogr. 45(8), 401–440. 28. Xi, D., Peng, Y. G., and Ramsdell, J. S. (1997) Domoic acid is a potent neurotoxin to neonatal rats. Nat. Toxins 5(2), 74–79. 29. Scholin, C., Massion, G., Mellinger, E., Brown, M., Wright, D., and Cline, D. (1999) The development and application of molecular probes and novel instrumentation for detection of harmful algae. Ocean Community Conference 98 Proceedings, Marine Technology Society, Washington, DC, vol. 1, pp. 367–370. 30. Trainer, V. L. and Baden, D. G. (1999) High affinity binding of red tide neurotoxins to marine mammal brain. Aquat. Toxicol. 46, 139–148. 31. AOAC (1984). Official Methods of Analysis of the Association of Official Analytical Chemists (Williams, S., ed.), 14th ed. Association of Official Analytical Chemists, Secs 18.086-18. 092, AOAC Inc., Arlington, VA, pp. 344–345.


361

32. Ridgway, S. H., McCormick, J. G., and Wever, E. G. (1974) Surgical approach to the dolphin’s ear. J. Exp. Zool. 188(3), 265–276. 33. McFarren, E. F., Schafer, M. L., Campbell, J. E., Lewis, K. H., Jensen, E. T., and Schantz, E. J. (1960) Public health significance of paralytic shellfish poison. Adv. Food Res. 10, 135–179. 34. Ridgway, S. H. (1972) Homeostasis in the aquatic environment, in Mammals of the Sea: Biology and Medicine (Ridgeway, S. H., ed.), Thomas, Springfield, IL, pp. 590–603. 35. Perl, T. M., Bedard, L., Kosatsky, T., Hocking, J. C., Todd, E. C. and Remis, R. C. (1990) An outbreak of toxic encephalopathy caused by eating mussels contaminated with domoic acid. N. Engl. J. Med. 322, 1775–1780. 36. Landsberg, J. H. (1996) Neoplasia and biotoxins in bivalves: is there a connection? J. Shellfish Res. 15, 203–230. 37. Trainer, V. L., McPhee, J. C., Boutelet-Bochan, H., Baker, C., Scheuer, T., Babin, D., et al. (1997) High affinity binding of pyrethroids to the alpha subunit of brain sodium channels. Mol. Pharmacol. 51, 651–657. 38. Wekell, J. C., Gauglitz, Jr., E. J., Barnett, H. J., Hatfield, C. L., and Eklund, M. (1994) The occurrence of domoic acid in razor clams (Siliqua patula), Dungeness crab (Cancer magister), and anchovies (Engraulis mordax). J. Shellfish Res. 13, 587–593. 39. Garthwaite, I., Ross, K. M., Miles, C. O., Hansen, R. P., Foster, D., Wilkins, A. L., and Towers, N. R. (1998) Polyclonal antibodies to domoic acid, and their use in immunoassays for domoic acid in sea water and shellfish. Nat. Toxins 6, 93–104. 40. Fairey, E., Stuart, N., Busman, M., Kimm-Brinson, D., Moeller, P., and Ramsdell, J. (2001) Biomonitoring brevetoxin exposure in mammals using blood spot cards, in Harmful Algal Blooms (Hallegraeff, G., ed.), 9th International Conference Proceedings. 41. Baden, D. G., Abraham, W., Fleming, L., Bossart, G., Benson, J. and Cheng, Y. S. (2001) What is the respiratory irritant in the air during Florida red tides? in Harmful Algal Blooms (Hallegraeff, G., ed.), 9th International Conference Proceedings.

362

Trainer

Human Illnesses and HABs

363

19 The Epidemiology of Human Illnesses Associated with Harmful Algal Blooms Lora E. Fleming, Lorraine Backer, and Alan Rowan

1. INTRODUCTION Harmful algal blooms (HABs) are harmful to people predominantly through their elaboration of a wide variety of very potent natural toxins that can accumulate in water or food. The toxins can be acutely lethal, and can cause a wide range of both acute and chronic health effects, including neurologic, dermatologic, pulmonotoxic, hepatotoxic, and immunotoxic illnesses and cancer in humans and other species. Many of these toxins are tasteless, odorless, and heat- and acid-stable; thus normal food preparation methods and conventional water treatment processes will not prevent intoxication if the food or water is contaminated (1–7). Previous chapters in this book have discussed the definition, ecology, formation, and geographic distribution of HABs, as well as the biological and biochemical activities of the toxins produced by the phytoplankton comprising HABs. Clearly, there is considerable scientific information concerning the identification and characterization of many of the HAB organisms as well as the molecular basis for the activities of the toxins they produce. However, beyond anecdotal reports and minimal published literature, there is very little information on the epidemiology of human health effects caused by exposure to HAB organisms or any associated toxins. In this chapter, we present an overview of what is known about the epidemiology of human illnesses caused by, or associated with, exposure to HABs. Specifically, we will review general epidemiologic principles, possible routes of human exposure to the etiologic agents, known and suspected human health effects, specific populations at risk, and possible exposure and disease prevention strategies. 2. OVERVIEW OF EPIDEMIOLOGY Epidemiology is the science of examining the occurrence of disease in populations. Disease occurrence is measured and related to the different characteristics of individuals and of their environments (8). Laboratory scientists such as toxicologists can estab*The opinions expressed in this chapter are those of the authors, not their employers or funding agencies. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

363

364

Fleming et al.

lish causation by manipulating the laboratory environment. Although this same principle is used in one form of epidemiologic study (human clinical trials), epidemiologists typically attempt to establish causation by observing and assessing naturally occurring patterns of disease in populations. One epidemiologic activity that may provide the first evidence of an unusual disease pattern, such as a cluster (an unusual increase in morbidity or mortality over a particular time and space), is surveillance. The Centers for Disease Control and Prevention (CDC) define public-health surveillance as the ongoing, systematic collection, analysis, and interpretation of health data that is needed for the planning, implementation, and evaluation of public health practice (9). Once there is evidence that there is an association between a particular exposure and an adverse health outcome, epidemiologists can then design studies to examine more closely examine a possible causal relationship. The data from epidemiologic studies are evaluated in terms of the risk of developing a disease or health outcome as the result of a given exposure. When investigating causality, epidemiologists interpret their data by examining: • The temporal sequence from exposure to subsequent development of the disease; • The strength of the association (the magnitude of the increased risk for the outcome that is associated with the particular exposure); • The evidence of a dose-response relationship (the probability of illness increases as the exposure increases); • The biologic plausibility of the hypothesized association; • Whether the results of current studies are consistent with the results of previous work; and • Whether there is any laboratory evidence that supports the causal relationship (10–17).

One example of how these criteria can be applied would be during the investigation of a seafood-poisoning outbreak. As with many food-borne diseases, seafood-borne poisonings often appear as disease clusters because seafood is often shared among families and friends and because it may be given to several customers dining at a single restaurant (12,13). To examine the source of a seafood-poisoning outbreak and thus identify a cause, the investigators must demonstrate that the persons who became ill were exposed to the suspected organism or toxin and that they had consumed the appropriate seafood vehicle or transvector before they became ill. Investigators must also show that the attack rate (prevalence of ill persons among all persons exposed) was higher among those who consumed the seafood than among those who abstained, and that it was highest among those who ate the most seafood. Finally, it must be documented that the seafood vehicle was contaminated with the pathogen or toxin of concern. One type of epidemiologic study used to investigate the association between exposure and subsequent disease occurrence is an outbreak investigation. Specific activities conducted during an outbreak investigation include: (1) identifying the nature of the outbreak; (2) identifying the target population (the group of people to whom the study results will apply) and a study population (a subset of the target population); (3) formulating a case definition; (4) measuring the exposures of interest; (5) comparing exposures and/or outcomes within the study groups; (6) formulating recommendations for further study; and (7) defining and conducting the relevant public-health follow-up measures (see Box 1 for the steps involved in assessing a HAB-related disease outbreak).


365

Box 1 Specific Activities Conducted During an Investigation of a Potential HAB-Related Disease Outbreak (3,12–15,23,67,68) I.

Identify the existence and nature of the outbreak a. Report of index or sentinel case(s) b. Initial public health response c. Confirm the existence of an epidemic II. Identify the target population and the study population a. Search for additional cases i. Active surveillance activities (e.g., house-to-house canvassing) ii. Passive surveillance activities (e.g., set up a hot-line for the public) III. Formulate a case definition a. Subjective clinical definition i. Route of exposure ii. Incubation time (minutes, hours, or days) iii. Acute symptoms iv. Duration v. Chronic symptoms b. Objective laboratory definition i. Exposure assessment 1. Environmental sample analysis 2. Food sample analysis 3. Biological fluid or tissue analysis ii. Patient evaluation 1. Biological fluid or tissue analysis 2. Other clinical tests (e.g., pulmonary function, neuropsychological testing) c. Medical evaluation i. Exclude other possible etiologies ii. Evaluate response to treatment (as a rule-out for other etiologies) IV. Measure exposures of interest a. Environmental samples b. Samples of food c. Biological samples V. Compare exposures and/or outcomes within the study population a. Apply case definition b. Calculate measures of association (risk ratio, odds ratio) VI. Further study a. Surveillance for future outbreaks i. Determine prevalence of human exposure b. Investigate possible clusters c. Conduct environmental monitoring activities VII. Define and conduct public health follow-up measures a. Institute control measures if appropriate (e.g., close shellfish harvest) b. Establish or continue surveillance activities c. Prepare risk communication materials i. Public ii. Health professionals iii. Trade groups iv. Media

366

Fleming et al.

Box 2 Case Study of Inhalation and Dermatologic Exposure: Harmful Algal Blooms Screening Pilot Study in Florida (71) The proliferation of harmful algal blooms (HABs) has reportedly intensified worldwide, causing deleterious effects to plants, animals, and humans (28,72,73). In 1997, fish kills in North Carolina estuaries were attributed to a new organism, Pfiesteria piscicida. That year, the CDC and six eastern seaboard states collaborated to develop the public health response to the presence of this and similar (Pfiesteria-like) organisms. For public health-surveillance purposes, an initial set of exposure and symptom criteria were developed for a general condition that might be associated with exposure to estuarine fish kills. Although Pfiesteria had not been found in Florida, a cryptoperidiniopsoid dinoflagellate (a Pfiesteria-like organism) had been found in Florida in association with fish with lesions. A cross-sectional epidemiologic study was conducted to determine whether there was an association between occupational exposure to estuarine HABs and human health effects. Florida environmental workers regularly exposed to estuarine waters as part of their work activities were recruited to participate in the study. There were three exposure groups: (1) workers exposed to waters containing cryptoperidiniopsoids, (2) workers exposed to fish kills and/or fish with lesions, but at times when no cryptoperidiniopsoids were found in the water, and (3) workers not exposed to the organism, fish kills, or fish with lesions (control group). Phone interviews were conducted to inquire about reported exposure and symptoms. The study population was a homogeneous group of 51 workers. Forty-one percent of the participants had at some time been exposed to a fish kill or to fish with lesions, and 13 were specifically exposed to water in which cryptoperidiniopsoids were found. None of the participants met the published criteria for human health outcomes associated with exposure to estuarine water containing Pfiesteria or Pfiesteria-like organisms (69). The subgroup (n = 28) exposed only to fish kills/fish lesions reported more health effects than both of the other groups (p = 0.08). The symptoms reported by this group may have been associated with a red-tide event that occurred simultaneously with the cryptoperidiniopsoid bloom. These occupationally exposed workers represent an excellent population to follow for future investigations of occupational HAB exposure.

Ideally, the target population is at high risk for both exposure and disease, and thus has a high incidence of disease. For example, if a new HAB were identified in an estuarine environment, a possible target population would be persons who are occupationally exposed to estuarine waters (see Box 2 for an example). In order to measure the amount of disease caused by a specific exposure in the target population, a case definition is required. The case definition may be based on historical consensus or developed for the purposes of a particular investigation. To the extent possible, the case definition should incorporate some objective measurements (such as biomarkers) to reduce misdiagnosis. The case definition also should have clinical relevance so that it can be used in the diagnosis and reporting of disease. When investigating newly emerging environmental health issues, epidemiologists may not be able to establish a firm case definition. For example, in response to reports of human illness associated with exposure to Pfiesteria piscicida, the CDC in collabo-


367

Box 3 PEAS Surveillance Criteria In March 2000, Centers for Disease Control and Prevention (CDC), in collaboration with investigators from 6 eastern seaboard states, published the surveillance criteria for possible estuary-associated syndrome (PEAS). For purposes of surveillance, people are considered to have PEAS if: (1) they report developing symptoms within 2 wk after confirmed exposure to estuarine water, (2) they report memory loss or confusion of any duration and/or they report three or more specific symptoms either of any duration (skin rash at the site of water contact or sensation of burning skin) or that persist for at least 2 wk after exposure (headache, eye irritation, upper respiratory irritation, muscle cramps, and gastrointestinal symptoms), and (3) a health-care provider reports that they cannot identify another cause for the symptoms. It is difficult to rapidly and accurately identify Pfiesteria piscicida in estuarine water samples. In addition, the toxin(s) associated with this and similar organisms has not been identified or characterized, therefore, there is no biological marker with which to confirm exposure. Finally, as P. piscicida blooms are ephemeral and unpredictable, it is unclear when and where to collect appropriate environmental samples when human exposure is suspected. Without a toxin with which to monitor exposure, and without any specific human health symptoms comprising PEAS criteria, it is extremely difficult to associate human exposure to estuarine waters that may contain P. piscicida with a specific illness.

ration with investigators from six eastern seaboard states published the surveillance criteria for possible estuary-associated syndrome (PEAS) (15). Because there was no method to confirm exposure, and because the reported symptoms were nonspecific, a series of criteria, rather than a case definition, were developed to identify individuals for participation in epidemiologic studies of PEAS (see Box 3). An exposure is any agent (chemical, biological, etc.) or characteristic (a human behavior, a particular genotype) suspected of causing or increasing the risk of disease. For HAB-related illnesses, the exposure of interest is exposure to natural toxins produced by the phytoplankton. Measures of these exposures, their surrogates, and subsequent physiologic effects are known as biomarkers (16–19). Biomarkers of exposure are the levels of the toxin or its metabolites in tissues, such as the levels of brevetoxin in the tissues of manatees who died after exposure to a red tide (20). Biomarkers of effect are indicators of subclinical physiologic change, such as conduction changes in the peripheral nerves of persons suffering from the ciguatera fish poisoning (21,22). In addition to biomarkers, epidemiologists commonly use other measurements of exposure such as self-reported information about food consumption collected through questionnaires (23). HAB-related illnesses are a good example of how important it is to be able to verify exposure. Human illness caused by exposure to brevetoxin from contaminated seafood has a well-documented clinical description based in part on our understanding of the toxin’s biochemical activity. Ongoing work on aerosolized brevetoxin and its possible acute and chronic effects on the human respiratory system is being aided by newly developed biomarkers of effect and exposure. By contrast, the definition of PEAS remains vague because, despite reports of human health effects associated with expo-

368

Fleming et al.

Box 4 Laboratory Advances and the Epidemiology of HAB-Associated Diseases Advances in marine-toxin laboratory science form the scientific basis for understanding the biology and epidemiology of the human illnesses caused by these toxins. The lack of such advances for particular HABs has affected our understanding of the epidemiology of human illnesses caused by exposure to two dinoflagellate organisms, Gymnodinium breve and Pfiesteria piscicida. G. breve is a marine dinoflagellate that produces polyether toxins, called brevetoxins. Fish kills associated with blooms of G. breve, called red tides, have been reported since 1844. Food poisonings (neurotoxic shellfish poisoning [NSP]) from consuming shellfish from Florida have been reported since the early 1900s, but the causative organism was not identified until the 1960s (after the discovery of G. breve in 1947). Work conducted in the 1980s identified and characterized the brevetoxins, allowing researchers to determine their biochemical activity (induction of channel-medicated Na+ ion influx), pharmacology, biological activity (depolarization of muscle fiber membranes), and physiological effect (neurotoxicity). Brevetoxins can be detected and quantified in shellfish as well as in water and air samples and human exposure to brevetoxins occurs either from consuming contaminated shellfish or from inhaling sea spray containing fragments of the cells or the toxin. Therefore, because we have considerable information about routes of exposure, biochemical activity, and human symptoms, it is possible to conduct the appropriate epidemiologic studies to evaluate the public-health importance of G. breve-associated diseases in human populations. By contrast, despite diligent activity by many researchers over a number of years, the toxin(s) associated with P. piscicida and similar organisms, as well as their specific biochemical, pharmacological, biological, and physiologic activities, remain uncharacterized. In addition, these organisms are difficult to identify, appear to form fleeting blooms, and are resistant to culturing. We currently cannot quantify environmental exposure to the organism or to any toxin(s) it may produce. New molecular biology techniques may allow at least a rapid identification and enumeration of cells during a bloom, which could provide a relative estimate of exposure. However, without knowing the biochemical activity of the toxin(s) or an accurate assessment of exposure, we cannot develop a case definition for human illness. Because we do not have the critical information about exposure and human symptoms needed to conduct epidemiologic studies, the current studies of human health effects from exposure to P. piscicida will produce limited information in evaluating the public-health importance of this organism.

sure, no toxin has been isolated from P. piscicida, and no biological indicator (i.e., biomarker) of either human exposure or biological effect has been identified (see Box 4). The final step in an epidemiologic investigation is calculating a measure of effect that indicates the magnitude of the health risk associated with a specific exposure. These measures usually are the risk ratio (the change in risk for a specific health outcome given a particular exposure) or the odds ratio (the odds that people with the health outcome are more likely to have been exposed than people without the health outcome were) (8,13,15).


369

Table 1 Routes of Exposure and Human Health Outcomes Associated with Microalgae Microbial Toxins Route of exposure Food

Source of human exposure to the toxin Shellfish

Fish

Water: inhaled aerosol Water: skin contact

Water: drinking water

Human health outcome Paralytic shellfish poisoning (PSP) Neurotoxic shellfish poisoning (NSP) Diarrheic shellfish poisoning (DSP) Amnesiac shellfish poisoning (ASP) Ciguatera Tetrodotoxin poisoning or Pufferfish poisoning (Fugu) ? Amnesiac shellfish poisoning (ASP)

Red tide organisms exacerbation Cyanobacteria (blue-green algae)

Respiratory irritation, asthma

Cyanobacteria (blue-green algae)

Skin irritation

Pfiesteria

? Skin irritation, skin lesions, possible estuary-associated syndrome (PEAS)

Cyanobacteria (blue-green algae)

Gastroenteritis; liver, kidney toxicity

Respiratory irritation

? Theoretical, but not epidemiologically established. Adapted from ref. 81.

3. ROUTES OF EXPOSURE There are a variety of routes through which people can be exposed to HABs and the toxins associated with them (Table 1). The two major routes of exposure are through food (many of these toxins become concentrated in animal tissues via the food web) and water. Unfortunately, the marine organisms that humans prefer as food tend to concentrate environmental chemicals, including the toxins produced by HAB organisms, through processes that represent both the lowest and upper-most levels in the food web. For example, molluscan shellfish are filter-feeders that concentrate natural and chemical toxins indiscriminately (24–26). In addition, many of the fish that humans eat are at the top of the marine food chain. Toxins in the prey can accumulate in the tissues of the predator fish, especially if the toxins are fat-soluble (i.e., lipophilic), and can reach concentrations that cause illness in people who consume them. The largest and oldest fish of each species accumulate the greatest body burden of toxin, and also tend to be the most desired by seafood consumers. Furthermore, the toxins that marine

370

Fleming et al.

organisms accumulate may have no obvious health effects on the fish and shellfish, and they appear healthy and appetizing to the consumer. As the number of people living in coastal areas increases and the international seafood trade expands, there are increasing opportunities for human exposure to HABs and their toxins (27–29). From 1988–1992 in the United States, seafood was responsible for at least 1 in 6 food poisoning outbreaks with a known etiology, as well as for 15% of the deaths associated with these outbreaks (30). This represents a 50% increase over the previous decade in the reported number of food-borne illnesses associated with seafood (24,30,31). In other parts of the world, the impact of seafood poisoning is even greater. In the period from 1971 to 1990, seafood was the single most important vehicle in food-poisoning outbreaks in Korea (32%) and Japan (22%). Further, seafood poisoning was responsible for 43 and 62%, respectively, of outbreak-related fatalities (32,33). The other major route of exposure to HABs and their toxins is through water (Box 2). For example, humans can be exposed to the blue green algae (cyanobacteria) toxins through direct skin contact, by drinking contaminated water, or through dialysis (5,34). Another possible water-related route of exposure is the inhalation of aerosols carrying cellular debris or significant levels of toxins. Recreational water users (e.g., people at the ocean beach) as well as people occupationally exposed to aerosols (e.g., from irrigating with contaminated water, fishing, or sampling water for monitoring purposes) may receive significant inhalation exposures to toxins during HAB events (35,36). 4. POPULATIONS Study populations for epidemiologic studies include people who are most likely to be exposed to, and thus affected by, HAB toxins. These populations include those occupationally involved in seafood harvesting, shipping, and processing; seafood consumers (including those consuming fish they caught or fish served in a restaurant); environmental workers (especially those collecting water samples); individuals who work and play on or near the water; and coastal communities, especially indigenous peoples who rely on seafood for a substantial proportion of their diet. Among seafood consumers, the risk for seafood poisoning varies considerably. Different ethnic subpopulations also have different seafood consumption and preparation patterns. For example, many Southeast Asian and South American groups traditionally eat raw fish, while northern Europeans consume fish preserved by pickling. In addition to the public health issues associated with risk communication, some ethnic subpopulations believe that the consumption of raw seafood or particular species of seafood is healthier and/or has aphrodisiac properties (e.g., raw oysters or pufferfish) (37–39). These cultural differences have important implications for the prevention and control of seafood poisonings. Health warnings or consumption advisories that are not culturally targeted and in the appropriate language may not reach these ethnic subpopulations (39). The importance of restaurants and other commercial establishments in the epidemiology of seafood poisoning is unclear. Nevertheless, people who consume seafood at a restaurant represent another population at risk for developing a seafood-related illness. At least 20% of the seafood consumed in the developed nations and an even greater percentage in the developing nations is derived from recreational and/or subsistence


371

fishing, both of which are not well-regulated by public-health controls. Outbreaks associated with recreational or subsistence fishing are much less likely to be reported than outbreaks originating in restaurants, making the contribution of this food source to the overall incidence of seafood poisonings unknown (24). In addition to persons at risk through food consumption, it is evident that HABs may pose a significant risk to persons through direct and indirect marine, fresh, and estuarine water exposure. Therefore, as noted earlier, occupational groups such as fishermen, watermen, water treatment workers, and scientists, as well as recreational boaters and swimmers, must be considered possible HAB target populations with water as the primary exposure vehicle (34,40,41). Because HAB toxins are so highly toxic and because they can induce damage to specific organ systems, children and people with underlying respiratory disease, liver disease, or neurologic illnesses may be particularly at risk from exposure to these toxins (42). 5. DISEASES 5.1. Disease Incidence and Prevalence Under-diagnosis and under-reporting, especially in endemic areas, make it difficult to know the true worldwide incidence of human illness associated with HABs. For example, it is believed that ciguatera affects at least 50,000–100,000 people per year who live in or visit tropical and subtropical areas. However, ciguatera is difficult to diagnose, particularly in areas where it is uncommon. Except in endemic areas such as Florida and Hawaii, physicians are not required to report the cases they identify. Clearly, there is significant under-reporting of this relatively common marine-toxin disease even in endemic areas (1,13,43). In addition to significant under-reporting of acute disease, almost no research has been performed on the possible chronic health effects of exposure to the HAB toxins, either after an acute exposure event or after long-term, low-level exposure (3,13,28,29). Thus, we also have little information about the prevalence or incidence of these illnesses. 5.2. HAB-Associated Human Illnesses The primary target of many HAB toxins is the neurologic system, although affected individuals usually present with a wide range of symptoms, resulting in a confusing clinical picture. Gastrointestinal symptoms begin minutes to hours after eating contaminated seafood. In the case of paralytic shellfish poisoning (PSP), pufferfish poisoning, and ciguatera, accompanying acute respiratory distress may be fatal within hours. Ciguatera and amnesic shellfish poisoning (ASP) may also produce debilitating chronic neurologic symptoms lasting from months to years. Other toxins, for example those from the blue-green algae, can cause dermatotoxicity, hepatotoxicity, immunotoxicity, and respiratory toxicity, and they can be carcinogenic (1). Again, there is very little information on the long-term health effects from chronic low level or from acute high-level exposures to these toxins. Chronic disease (neurologic, immunologic, etc) associated with the HAB toxins is an area of active scientific research. For example, Landsberg and others (44,45) have reported an increasing incidence of gonadal tumors in shellfish exposed to HAB toxins

372

Fleming et al.

with carcinogenic properties. The long-term health of the human populations who consume these potentially carcinogenic organisms should be considered an important priority in terms of their potential public-health impact. 5.3. Toxin Transmission The HAB toxin diseases associated with seafood consumption can generally be categorized into two groups based on their primary transvectors (Table 2). Shellfish harbor the toxins that produce paralytic shellfish poisoning (PSP), neurotoxic shellfish poisoning (NSP), diarrheic shellfish poisoning (DSP), and amnesiac shellfish poisoning (ASP). Fish carry the toxins responsible for ciguatera poisoning and tetrodotoxin (fugu or pufferfish) poisoning. The shellfish-associated diseases generally occur in association with algal blooms or “red tides,” which may be characterized by patches of discolored water and dead or dying fish. The fish-associated diseases are more localized to specific reef areas (ciguatera poisoning) and specific fish (pufferfish poisoning). This categorization method is somewhat limited as domoic acid (associated with ASP) has also caused disease in sea lions that consumed contaminated fish as opposed to contaminated shellfish (46). The water-associated diseases include those caused by exposure to blue-green algae, and aerosolized red-tide toxins (Table 2). The blue-green algal toxins are more commonly associated with freshwater exposure, although marine and estuarine waters can be contaminated as well. Estuarine and beach-water exposure are associated with the other HABs (5,47,48). 5.4. Financial Impact of Disease As noted earlier, in the past many of these illnesses have been highly localized to island and coastal communities as endemic diseases. With increasing worldwide seafood consumption and trade, as well as international tourism, these diseases are expanding beyond their traditional geographic boundaries. One side effect has been the high costs of diagnosis and treatment of disease in traditionally nonendemic areas. For example, in Canada, with an estimated 1000 cases per year related to tourism and food importation, the average medical cost per case of ciguatera fish poisoning was $2470 in 1990 (49,50). Accurate estimates of the human costs of these diseases necessitate adequate knowledge of their prevalence and incidence, as well as understanding of their acute and chronic human health effects. 6. PREVENTION From the public-health perspective, disease prevention comprises primary, secondary, and tertiary prevention activities. Primary prevention, or the decrease in disease incidence, would be ideal (10,12–15). However, incomplete scientific information, inadequate tools with which to assess exposure and disease, and limited resources can make primary prevention unachievable. The prevention of the HAB-related diseases in human populations is based on monitoring and preventing exposures, as well as surveillance of the diseases in human populations (13,31,51–53). For example, in the case of the diseases such as PSP, primary prevention is comprised of shellfish-bed monitoring. When saxitoxin levels in the shellfish become dangerously high, the beds are closed to harvesting. Continued monitoring


373

allows reopening of the shellfish beds when the toxin levels drop to levels that make the shellfish safe for human consumption. Secondary prevention activities are developed to decrease the prevalence of disease by reducing the duration of clinical illness and by instituting early detection. Secondary prevention of the HAB-related illnesses could involve the surveillance of biomarkers of both exposure and subclinical effects in target populations. Additional activities would include the development of educational and monitoring systems (including environmental and human health monitoring) for future primary prevention. The goal of tertiary prevention is to reduce the complications resulting from actual HAB-related illness and disease. This would involve the early treatment of clinical disease, such as the use of mannitol in the treatment of acute ciguatera to prevent chronic sequelae (54). All forms of public-health prevention require the existence of an infrastructure including knowledgeable public-health personnel (environmental health and epidemiology) who are able to work in the field collecting surveillance information and exposure samples. Also important is the education of healthcare and public health personnel concerning the diagnosis, treatment, and reporting of HAB-related illnesses. At-risk populations should be informed about how to recognize HAB-related illnesses and about measures to prevent exposure to HAB toxins (such as not consuming shellfish during times of red tides and/or fish kills or avoiding drinking from fresh water with toxic blue green blooms). Finally, the cooperation of the seafood and marine recreational industry and drinking-water utilities will be required to implement a successful prevention program (10–15). 6.1. Exposure Monitoring and Public Recreational Health Response Monitoring techniques for the presence of organisms in the environment, as well as for the toxins in both water and in seafood, exist for most of the known HAB organisms that produce characterized toxins. In areas where HABs are a regular and expected event, these monitoring techniques can prevent disease in human populations. For example, the Florida Department of Environmental Protection monitoring program leads to the closure of shellfish harvesting when increased levels of the organism (Gynodinium breve) and/or toxin (brevetoxin) associated with Florida Red Tide are found. This monitoring program has virtually eliminated human cases of NSP associated with red tide in Florida. However, such a monitoring program does not work when there are organisms and/or toxins that have not previously been identified as toxinproducing (e.g., the first report that a Canadian red-tide diatom produced domoic acid, which is associated with ASP) or that appear for the first time in a new geographic area (e.g., the first outbreak of PSP in Guatemala, Box 5) (54,56). Monitoring programs designed to assess environmental samples or seafood (as is now required in the United States by the Food Safety Act [57]) are expensive and labor-intensive and often require access to sophisticated analytical instruments. In addition, with diseases such as ciguatera-fish poisoning, every individual reef fish must be tested in order to prevent human exposure to the toxin; and monitoring programs simply cannot eliminate the problem. Increases in the international trade of seafood further complicate seafood monitoring. The tracing and possible recall of seafood is more dif-

374

Table 2 Human Intoxication Syndromes Caused by Marine Microbial Toxinsa

Ciguatoxin (10+) maitotoxin, scaritoxin

Anatoxins (3+), saxitoxins (2+), microcystins, nodularins, cylindrospermopsins

Brevetoxin (10+)

Okadaic acid (4)

Domoic acid (3)

Disease(s)

Paralytic Shellfish Poisoninga (PSP)

Neurotoxic Shellfish Poisoning (NSP)

Diarrheic Shellfish Poisoning (DSP)

Amnesiac Shellfish Poisoning (ASP)

CiguateraFish Poisoning

Respiratory illness, asthma exacerbations

Skin irritation, gastrointestinal (liver) disease

Causative organism(s)

Red-tide dinoflagellate



Red-tide diatom

Epibenthic dinoflagellate


Blue-green algae, Cyanobacteria

Route of exposure(s)

Ingestion

Ingestion

Ingestion

Ingestion

Ingestion

Inhalation, ?skin

Ingestion, ?inhalation, ?skin

Major transvector(s)

Shellfish

Shellfish

Shellfish

Shellfish, ?Fish

Fish

Water

Water

Moleculary mechanism(s)

Na+ channel blocker

Na+ channel activator

Phosphorylase phosphatase inhibitor

Glutamatereceptor agonist

Na+, Ca++ channel activators

Na+ channel activator

Acetylcholinesterase inhibitor, Na+ channel blocker, phosphorylase phosphatase inhibitor, proteinsynthesis inhibitor

Brevetoxin (10+)

Fleming et al.

Saxitoxin (18+)

Toxin(s) (number)

Toxin(s) (number)

Saxitoxin (18+)

Brevetoxin (10+)

Okadaic acid (4)

Domoic acid (3)

Ciguatoxin (10+) maitotoxin, scaritoxin

Brevetoxin (10+)

Anatoxins (3+), saxitoxins (2+), microcystins, nodularins, cylindrospermopsins

Incubation time

5–30 min

30 min–24 h

α4ß2 (Kd = 10–7 to 10–6 M) >> α3ß4, α2ß2, α4ß4 (Kd >> 10–6 M) (129, 181).

4. Nonconventional Neurotoxins Long-chain toxins

From kraits, coral snakes, and cobras. Long toxins with 4 conserved disulfide bonds and a fifth disulfide bond in loop I, with low affinity for muscular-type AChR (9a,9b).

This review will focus on the progress that has been achieved in the course of the last decade in understanding the structural and biological molecular properties of α-neurotoxins. 2. SOURCES OF TOXINS 2.1. Isolation from Venoms Since the early discovery of the first neurotoxins (16,17), the methods for isolating new toxins have continuously improved. It is beyond the scope of this review to describe such improvements and their exploitation in the domain of neurotoxin isolation. The interested reader may consult previous reviews (18) and/or some recent studies of these questions (19,20). It is worth mentioning that neurotoxins can be nicely

Snake Neurotoxins

387

identified in crude snake venoms using on-line liquid chromatography-electrospraymass spectrometry (LC-ES/MS)(21). 2.2. Production, Purification, and Characterization of Recombinant Neurotoxins Readers interested in a detailed description of the artificial procedures used to overproduce three-fingered toxins may consult a recent review (22). The different systems that have been used to produce neurotoxins are summarized in Table 2. In the early studies, the bacterial systems were particularly appealing because neurotoxins usually do not undergo post-translational modification. Initially, a fused toxin was produced in the periplasm of Escherichia coli (23), where the formation of disulfide bonds is specifically facilitated by specific redox machinery (24). For the fused toxin to be addressed to the periplasm of E. coli, a pRIT5A vector was used, under the control of Protein A promotor (25). The vector was later transformed into pEZZ 18, which included two IgG-binding domains of protein A, called ZZ, from Streptococcus aureus (26). Thus, a gene encoding the protein A sequence signal was followed by the gene coding for the ZZ domain, which in turn was linked to the toxin gene. This vector even permitted direct secretion of the synthesized ZZ-toxin hybrid into the growth medium of E. coli, HB101 strains often having been used for transformation. This general procedure was successfully used for the production of fused recombinant forms of Erabutoxin a from Laticauda semifasciata (Ea) (27), NmmI toxin from Naja mossambica mossambica (28), and κ-bungarotoxin from Bungarus multicinctus (29) (Table 2). In these procedures, the toxins were directly and appropriately folded. The yield of purified unfused toxin was rarely higher than 0.5 mg/L. A higher yield was sometimes observed with cytoplasmic expression systems. In this case, the sequence of the ZZ, without a signal sequence, was cloned into the pET3a vector downstream of the T7 promotor (30), leading to a pCP vector (31). Thus, the cDNA encoding Ea, as well as a synthetic gene encoding toxin α or α-cobratoxin, isolated from Naja nigricollis and Naja kaouthia, respectively, were inserted in this modified vector and used to transform a BL21(DE3) E. coli strain (27,31,32) (Table 2). This procedure yielded purified toxin in the range of milligrams per liter of culture. Using this procedure, a neurotoxin was uniformly labeled with stable isotopes (31). The recombinant fused toxin was efficiently refolded by passing it slowly through a column containing both immobilized IgG and an appropriate redox medium. Evidently, for both periplasmic and cytoplasmic systems an appropriate cleavage site also needs to be introduced in the fusion protein. Since neurotoxins often lack methionine, this residue was purposely introduced at position –1 of the toxin sequence, allowing the toxin to be adequately released by a simple treatment with cyanogen bromide. This approach led to recombinant toxins which were always indistinguishable from those isolated from venom. Rosenthal et al. described the cytoplasmic production of α-Bgtx as a fusion protein appended to the COOH-terminal portion of the T7 gene 9 coat protein. In this case, the formation of insoluble inclusion bodies was avoided but the recombinant toxin contains 10 additional residues linked to the N-terminal sequence due to an inappropriate cleavage by Factor Xa (33). Also using a pET vector, recombinant cobrotoxin was

388

Table 2 Recombinant Expressions and Mutations of Snake Toxins Acting on AChRs

Toxins

Host cells

Expression system

Production yields in purified toxin (mg/L of culture)

Mutants or modified toxins

References

α-Neurotoxins (short-chain toxins) Erabutoxin a E. coli Protein A fusion (Laticauda (HB 101, BL21) pEZZ 18 or pET3a vectors for semifasciata) periplasmic or cytoplasmic expressions

0.05–0.5

NmmI (Naja mossambica mossambica)

F4A; N5V; H6A; Q7L; S8G,T (23,27,31,67,73) S9G; Q10A; P11N; Q12A; T13V T14A;K15A; T16A; DS18; E21A; S22A;Y25F;K27E; Q28A; W29H,F S30A; D31H,N;F32L; R33E,K,Q; G34S; T35A; I36R;E38Q,K,L; G42A; P44V; T45A; V46A;K47A,E; P48Q; G49V; I50Q; L52A;S53A;S57N; V59A; N62A

Protein A fusion pEZZ 18 vector periplasmic expression

0.5–1.5

S8T; E10A,R; K27E; R33E; R36E; K47A,E; K48E;

(28,126,127,171)

Toxin α (Naja nigricollis)

E. coli BL21(DE3)LysS

Protein A fusion pET3a vector for cytoplasmic expression

2.2

Uniformly labeled 15N-toxin α

(31)

Cobrotoxin (Naja naja atra)

E. coli BL21(DE3)

pET20b(+) vector inclusion bodies

E38Q, K47Q,E

(34)

Servent and Ménez

E. coli (HB 101)

α/κ-Neurotoxins α-Bungarotoxin E. coli (Bungarus BL21(DE3) multicinctus)

Pichia pastoris (GS115) α-Cobratoxin E. coli (Naja kaouthia) BL21(DE3)

κ-Neurotoxins κ-Bungarotoxin (Bungarus multicinctus)

Gene 9 fusion protein; pSR9 vector cytoplasmic expression with 10 additional residues in N-terminal

0.4

D30A

(33)

Gene 9 fusion protein with His tag pPIC9 vector

0.5–1.5

K26A, R36A, ∆68-74

(128)

K38P-L42Q

(36)

F4A; I5A; T6A; P7A; D8A,R; I9A; T10A; S11A; K12E; D13A; Y21F; K23E; W25A,F,H; D27N,R; A28G,R; F29A,L,W; C26S-C30S; S31A; I32A; R33E; K35A; R36A; D38L; T47A; K49E; D53K; F65A; P66A; ∆67-71

(32)

Q26W; P36A,K; R34A

(29,129,181)

I20A; Q48A; F49A; DR54; R54A; L57A; T60A; C3A/C21A; C14A/C42A; C27A/C31A; C46A/C58A; C59A/C64A

(35,90,130)

Protein A fusion pET3a vector for cytoplasmic expression

Protein A fusion, pEZZ 18 vector periplasmic expression

Pichia pastoris (GS115)

pPIC9 vector

HEK 293

Prolactin fusion, pAdlox vector

0.1

(182)

389

E. coli

0.5–1.5

Snake Neurotoxins

Table 2 Continued

390

Servent and Ménez

produced in the cytoplasmic space of E. coli, but this time as inclusion bodies. The recombinant toxin was then subjected to refolding (34) (Table 2). Recently, two groups described the use of a Pichia pastoris system to produce recombinant κ- and α-bungarotoxin (35,36) (Table 2). This yeast expression system allows the direct secretion of the recombinant toxin into the medium without in vitro proteolytic and refolding reactions. It might appear from these descriptions that production of a recombinant toxin is easy. However, it must be stressed that the efficiency may vary from one toxin to another and it often requires much effort to adjust the expression system correctly and thereby achieve appropriate and efficient production. 2.3. Chemical Synthesis Toxin α from Naja nigricollis, a 61-amino-acid protein containing four disulfides, and some variants of a long-chain toxin, have been successfully synthesized chemically, despite their numerous disulfide bonds (37). This was done using the general Fmoc strategy and an in vitro redox glutathione system. About 10 mg of the synthetic peptide was produced with an overall yield of 40% (37,38). The chemical approach leads to large quantities of toxin, permits the introduction of non-natural amino acids, and enables site-directed introduction of 13C- and 15N-labeled amino acids. Furthermore, solid-phase synthesis appears to be an appropriate approach not only to modifying natural toxins but also to inserting new bioactive functions, as shown recently (39,40). 3. RNAS AND DNAS ENCODING NEUROTOXINS 3.1. cDNAs and Genes that Encode Three-Fingered α-Neurotoxins The first cDNA encoding a neurotoxin was reported in 1985 (41). Its constitution indicated most clearly that the toxin followed the classical route of synthesis of secreted proteins (42). Since then, cDNAs encoding various other neurotoxins, like those of Aipysurus laevis (43), long-chain neurotoxins from Laticauda semifasciata (44), κ-neurotoxin (45), and α-bungarotoxin (46) from Bungarus multicinctus have been isolated using experimental procedures that have been described in detail in various reviews (47,48). Inspection of the cDNAs encoding three-finger toxins from snake venoms indicated that they contain an open reading frame of approx 280 nucleotides encoding 81–86 residues, 21 of which consistently form a signal peptide (49). Clearly, the 5' and 3' UTRs are much more conserved than are the protein-coding regions. For example, when considering cDNAs encoding five different toxins, the average nucleotide identities are 96, 91.5, and 80.8% for the 5' UTR, the 3' UTR and the protein-coding region, respectively. Moreover, the signal peptide-coding region is 95.5% conserved, implying that the region coding for the mature toxins is characterized by only 66.3% conservation. Clearly, a higher rate of mutation occurs within the region encoding the mature toxins. A similar phenomenon was initially observed for phospholipases A2 from snake venoms (49). These observations suggest that the three-finger neurotoxins might have evolved via accelerated evolution and gene duplication. However, the biochemical process, if any,

Snake Neurotoxins

391

that might be associated with this phenomenon is not known. Interestingly, a comparative study of the cDNAs of 53 conopeptide precursors from nine conus species revealed that inside the most variable regions there is a pronounced position-specific codon conservation for the conserved cysteine residues (50). This observation was even extended to other families of proteins rich in disulfide bonds. In all cases, it seems that the cysteines would be “protected” from mutations by a kind of stencil-like system. The authors even proposed that specific “protecting” molecules may bind to cysteine codons in hypervariable regions of the toxins. This interesting idea now requires experimental verification. As initially shown in the pioneering work by Fuse et al., the gene encoding erabutoxin c (Ec) is 1.2 kb long with three exons separated by two introns (51). The first intron is located within the sequence encoding the signal peptide, whereas the second is located in the region encoding the central loop of the mature toxin. The intron sequences share a high degree of sequence similarity. Also conserved is the sequence of exon 1, which codes for signal peptides. In contrast, substantial differences have been observed for exons 2 and 3. More recently, two genomic DNAs of approx 2.8 kb encoding the precursors of the long neurotoxins, α-bgtx (A31) and α-bgtx (V31), have also been isolated (46). Both toxin genes possess the same organization with three exons separated by two introns, their nucleotide sequences sharing approx 98% identity. The intron/exon organization of α-bgtx genes is globally similar to that of short neurotoxins. This observation suggests that the long and short neurotoxins have a common evolutionary origin (46,52). Comparison of neurotoxin genes revealed that the coding regions of the exons are more variable than those coding for introns, the nucleotide sequences encoding signal-peptide regions being highly conserved (44,46). These findings again agree with the hypothesis that only the gene region encoding the mature part of three-fingered toxins undergoes an accelerated rate of mutation, perhaps indicating an accelerated evolution in this particular region of the toxin genes. 4. STRUCTURAL ASPECTS OF THREE-FINGERED TOXINS We have known how snake neurotoxins fold since 1976, when two groups independently reported the three-dimensional (3D) crystal structure of two neurotoxins isolated from venom of the sea snake Laticauda semifasciata and that later appeared to be identical. Thus, the structure of erabutoxin b (Eb) was solved at 2.75 Å resolution (53) and the other toxin, at that time named neurotoxin b, was solved at 2.2 Å resolution (54). These studies revealed that neurotoxins are rather flat molecules with three adjacent loops forming a large ß-pleated sheet with five antiparallel ß-strands. These loops can be seen as three fingers emerging from a small globular core where four disulfide bonds are embedded. The base of the three loops seems to be sitting on the small Cterminal loop, giving the impression that the back of a three-fingered hand is lying on a small plate. This fold is shown in Fig. 1, using erabutoxin a, α-cobratoxin, κ-neurotoxin, and bucandin as prototypes of short-chain α-neurotoxins, α/κ-neurotoxins, κ-neuronal toxins, and nonconventional neurotoxins, respectively. Many structural studies have been performed on snake curaremimetic toxins by X-ray diffraction or nuclear magnetic resonance (NMR) and all the structures available on the Protein Data Bank (PDB) are shown in Table 3.

392

Servent and Ménez

Fig. 1. Three-dimensional structurees of snake toxins. The β-strands are visualized by arrows and the disulfide bonds by dotted lines

4.1. Three-Dimensional Structures of α-Neurotoxins 4.1.1. The Erabutoxins The crystal structure of erabutoxin b was analyzed in detail at 2.5 Å resolution (55), and then at 1.4 Å resolution (56–58). A dimeric crystal form of Eb was obtained later using thiocyanate solution as a crystallizing agent, and its crystal structure was solved at a resolution of 1.7 Å (59). The authors suggested that this form resulted from the presence of the thiocyanate. However, this is unlikely since a dimeric form was also observed for the very similar erabutoxin a (Ea) (60) and erabutoxin c (61), without the use of this particular agent. For both Eb and Ea, the dimer involves an antiparallel association between residues 52 and 56 (59,60). This association, however, has little effect on the toxin architecture, the differences between the backbones of the two molecules forming the dimers in Eb being characterized by a root mean square deviation (rmsd) not greater than 0.5 Å. Evidently, the differences are distinctly spread along the

Snake Neurotoxins

393

Table 3 Three-Finger Fold Structures of Snake Toxins Blocking AChRs

Toxin names and snake species α-Neurotoxins (short-chain toxins) Erabutoxin b (Laticauda semifasciata)

Method resolution (X-ray) and average backbone rmsd and number of structures (NMR)

PDB code

References

1FRA

NMR (0.66 Å, 14) X-ray diffraction (1.4 Å) X-ray diffraction (1.38 Å) X-ray diffraction (1.4 Å) X-ray diffraction (1.70 Å)

1NXB 3EBX 6EBX

(62) (53,56,57) (54) (58) (59)

Erabutoxin a (Laticauda semifasciata)

X-ray diffraction (2 Å) X-ray diffraction (1.49 Å)

5EBX 1QXD

(64) (60)

S8G mutant S8T mutant

X-ray diffraction (1.8 Å) X-ray diffraction (1.7 Å)

2ERA 3ERA

(66) (66)

Toxin α (Naja nigricollis)

NMR (0.51 Å, 8) X-ray diffraction (1.8 Å)

1NEA 1IQ9

(70)

Cobrotoxin (Naja naja atra)

NMR (1.67 Å, 23)

1COD

(71)

Dpp α (Dendroaspis polylepis polylepis)

NMR (0.66 Å, 20)

1NTX

(69)

NnoII (Naja naja oxiana)

NMR (0.53 Å, 19)

1NOR

(72)

α/κ-Neurotoxins NnoI (Naja naja oxiana)

X-ray diffraction (1.9 Å)

1NTN

(85)

α-Cobratoxin (Naja kaouthia, Naja siamensis)

X-ray diffraction (2.8 Å) X-ray diffraction (2.4 Å) NMR (0.5 Å, 10)

1CTX 2CTX

(75) (76) (77)

α-Bungarotoxin (Bungarus multicinctus) Complex Bgtx / peptide Complex Bgtx / peptide

X-ray diffraction (2.5 Å)

2ABX

(80)

NMR (2.6 Å, 4) NMR (1.35 Å, 20)

1ABT 2BTX

(82) (83)

LS III NMR (0.82 Å, 23) (Laticauda semifasciata) Toxin b NMR (0.15 Å, 21) (Ophiophagus hannah)

1LSI

(84)

1TXA

(86)

NMR (1.3 Å, 10) X-ray diffraction (2.3 Å)

2NBT 1KBA

(88) (89)

Nonconventional neurotoxins Bucandin (Bungarus candidus) X-ray diffraction (0.97Å)

1F94

(90a)

κ-Neurotoxins κ-Bungarotoxin (Bungarus multicinctus)

394

Servent and Ménez

polypeptide chain, the largest being seen at the very tip of the central finger, which illustrates the flexibility of the chain in this region (59). The NMR studies of Eb then confirmed the folding pattern of the toxin in solution (62). However, a number of differences have been observed between X-ray and solution structures, especially within the central and third loops, a finding that suggests that these regions are more flexible. Erabutoxin a (Ea) was also submitted to structural studies. This toxin differs from Eb by a single mutation, position 26 being a histidine and asparagine in Eb and Ea, respectively. The crystal structure of monomeric Ea was initially solved at 2.5 Å resolution (63) and then at 2.0 Å (64). The structure of the recombinant form of this toxin was also elucidated at 2.0 Å resolution (65). More recently, the structure of monomeric and dimeric Ea was solved at 1.5 Å resolution (60). Comparison of the backbone conformation of this high-resolution monomeric structure with those of the other monomeric structures of Eb or Ea revealed an rmsd of no more than about 0.22 Å (60). Clearly, the mutation His/Asn at position 26 has little effect on the polypeptide-chain structure of erabutoxins. This conclusion can probably be extended to many other mutations, as illustrated by the crystal structures solved at 1.7 Å and 1.8 Å resolution for two Ea mutants where serine at position 8 was respectively replaced by a threonine and a glycine (66). Also, many individual mutations have been introduced in Ea without causing any substantial conformational change, as indicated at least by circular dichroism measurements (27,67). 4.1.2. Other Short-Chain α-Neurotoxins Toxin α from the spitting cobra Naja nigricollis was the first neurotoxin to be isolated and sequenced (17) but its crystal structure was solved only recently (R. Ménez personal communication). The high resolution of this structure (1.8 Å) shows that despite its 17 different residues, toxin α adopts a folding that is very similar to that of erabutoxins, all the toxins having similar affinities for the Torpedo AChR (27,67,68). However, a number of slight structural differences, mostly located around the tips of the fingers, were noted between the structures of the two types of toxins. Due to their small size and high stability, snake neurotoxins have been rapidly perceived as interesting protein tools for NMR investigations and indeed several short toxins have been studied by means of NMR. Until the early 1990s, NMR mostly allowed identification of both secondary structure and local conformational details (see ref. 1 for review). Three-dimensional structures of various short neurotoxins in solution were subsequently elucidated from NMR data and molecular modeling. The highquality structure of the 60-residue α-neurotoxin from the black mamba (Dendroaspis polylepis polylepis) (69) was elucidated from 656 NOE distance constraints and 143 dihedral angle constraints. Although this toxin and Eb have only 39 residues in common, the two toxins adopt quite similar 3D structures, with five homologous antiparallel ß-strands forming a large ß-sheet whose backbone atoms N, Cα, and C' were characterized by an rmsd no greater than 0.62 Å. A number of local differences were observed, however, especially around the tips of the first and second fingers, in the external part of loop III and around the turn region where the Eb possesses two additional residues at positions 18 and 19. Almost simultaneously, the solution structure of toxin α from Naja nigricollis was elucidated from 409 distance constraints and 73

Snake Neurotoxins

395

dihedral angle constraints (70). The same overall three-finger structure was observed, again accompanied by a number of deviations around the tips of the loops. The solution structures of cobrotoxin from venom of the Taiwan cobra Naja naja atra (71), neurotoxin II from Naja naja oxiana (72) and Eb (62), were reported. The 3D structures of these toxins clearly share five homologous ß-strands, which are welldefined, with low rmsd relative to the crystal structure of Eb. In general, ill-defined regions with greater mobility are seen in the vicinity of the tips of the fingers, especially around the first and second fingers. It has often been speculated that these mobile regions may be involved in binding to receptor. As we shall see later, these regions are indeed likely to play a functional role, although less mobile regions also appear to be functionally important (27,67,73). 4.1.3. Waglerins

Although waglerin-1 from the Viperidae venom displays high affinity for muscular acetylcholine receptor (3), its 3D structure is unrelated to that of Elapidae α-neurotoxins, as inferred from 2D-NMR and molecular dynamic simulation studies (74). Using 247 inter-proton distance constraints derived from NOE measurements, 19 structures were calculated by molecular modeling. Strikingly, the toxin has no special secondary structure, the central ring region being well defined but the N- and C-termini being more disordered (74). 4.2. Three-Dimensional Structure of α/κ-Neurotoxins 4.2.1. α-Cobratoxin The 3D structure of a long-chain neurotoxin was reported for the first time in 1980. The 3D structure of α-cobratoxin (Fig. 1) from the Asian snake Naja naja siamensis (now called Naja kaouthia) was solved at 2.8 Å resolution (75) and then refined at 2.4 Å (76). The solution structure of this toxin was also solved by NMR and modeling (77). Clearly, this long toxin adopts an overall fold that is reminiscent of that of Elapidae α-neurotoxins, with three fingers emerging from a small globular core and four invariant disulfide bonds. Perhaps a major difference resides in the presence of only three antiparallel ß-strands on loops II (residues 19–25 and 36–41) and III (residues 52–57). The other most important structural deviations are a result of large insertions or deletions that occur in the sequence of the long-chain toxin. Thus, the tip of the central finger displays a small loop cyclized by an additional disulfide, the first loop of the long toxin is smaller in size, and its C-terminal tail is markedly longer. Strikingly, these deviations are well-integrated into the fold. It could be argued that the crystals of the toxin were obtained at the unconventional pH of about 2.8, but this condition is unlikely to have affected the overall toxin conformation, as judged from the high stability of the toxin (1). Moreover, the solution structure of the toxin at acidic pH (77) was not greatly different from that observed at neutral pH, although the latter was somewhat more compact (78). In acidic solution, however, a short helix (residues 29–34) was seen at the tip of the central finger. This helical structure was neither observed in the crystal structure (76) nor in solution at neutral pH (78).

396

Servent and Ménez

4.2.2. Other α/κ-Neurotoxins In 1986, the crystal structure of α-bungarotoxin from the krait Bungarus multicinctus was solved at 3.5 Å (79) and then at 2.5 Å resolution (80). The structure was globally comparable to those of other known snake neurotoxins. However, an unexpected difference was seen around the invariant tryptophan 28. Like in other neurotoxins this residue belongs to a ß-strand of the central finger but its orientation is clearly opposite to that adopted by the homologous tryptophan in other neurotoxins, with respect to the plane defined by the triple-stranded ß-sheet. The invariant indole side chain points towards the concave face in all other neurotoxin structures, whereas it points toward the convex face for α-bungarotoxin, this face interacting with the C-terminal loop of the neurotoxins. Structural NMR studies have examined α-bungarotoxin, either free (81) or complexed with a peptide (82,83). These data clearly show that the toxin also adopts a three-finger fold in solution, with two antiparallel ß-strands in loop I (residues 2–5 and 12–16) and three antiparallel ß-strands spread on loop II (residues 22–27 and 40–45) and loop III (residues 57–60). In the solution structure, however, the indole of Trp28 adopts the conventional location on the concave face, suggesting that upon crystallization, the indole undergoes a reorientation from the concave to the convex face. This is a most surprising case of a major difference between solution and crystal structures in a neurotoxin. It might be interesting to investigate whether the unconventional orientation of the tryptophan 28 in the crystal structure of α-bungarotoxin is maintained when different conditions of crystallization are used. LSIII, a neurotoxin from venom of Laticauda semifasciata, shares 53% sequence similarity with α-cobratoxin and, not surprisingly, the solution structure revealed that its fold is globally similar to that of other neurotoxins (84). LSIII also shows a triplestranded ß-sheet on loops II (residues 19–25 and 37–41) and III (residues 51–58), though in contrast to α-cobratoxin, it does not display helicity at the tip of the central loop (77), this region, however, being highly mobile in both toxins. Neurotoxin I from Naja naja oxiana, whose crystal structure was solved at a resolution of 1.9 Å, possesses five ß-strands like α-bungarotoxin (85). A triple ß-stranded sheet is composed of two strands on the central finger (residues 20–26 and 37–43) and of one on the third finger (residues 54–58), all of them being homologous to those observed in other long-chain neurotoxins. In addition, a double antiparallel stranded ß-sheet (residues 3–5 and 10–12) is reminiscent of what is also seen in short neurotoxins. A short helix turn is seen at the tip of the central finger (residues 31–34), which is homologous to what is seen with α-cobratoxin (77). The solution structure of toxin b from venom of the King cobra Ophiophagus hannah also revealed the presence of five ß-sheet strands (86). A very short double-stranded ß-sheet is seen in the first loop (residues 3–4 and 12–13) and a triple-stranded ß-sheet encompasses the second loop (residues 20–25 and 38–43) and the third loop (residues 54–59). However, toxin b does not display any short helix at the tip of the central loop. Therefore, the three-finger fold adopted by the α- and ακ-neurotoxins displays a variety of structural deviations, sometimes with three or five ß-strands, a short helix at the tip of the central loop and many other differential features like the size or the twist of the loops. The three-finger fold, therefore, is highly permissive to structural variations. A related question is whether or not these differences reflect any functional features?

Snake Neurotoxins

397

4.3. Three-Dimensional Structure of κ-Neurotoxins A 1991 H1 NMR study showed that a neuronal bungarotoxin (87) adopts a threefinger structure with a central triple-stranded anti-parallel ß-sheet. The toxin forms a dimer in solution, the two monomers having the same average conformation. The interface between the two monomers is located along one edge of the ß-sheet, thus generating a six-stranded intermolecular anti-parallel ß-sheet. By combining these H1 NMR spectroscopy findings with homologous model building, together with dynamic simulated annealing and molecular refinements, a structural description of the neuronal toxin dimer was proposed (88). Using 582 experimentally determined NOE constraints, together with 27 ϕ-angle constraints, the authors calculated 14 different structures for the dimeric molecule. They observed a rather well-defined six-stranded ß-sheet, whereas the N-terminal (Arg1 to Ile20) and C-terminal (Thr60 to His66) regions were poorly defined. They observed, for example, that Phe49 has an important role in dimer formation. The crystal structure of κ-bungarotoxin was described 2 yr later (89). The X-ray structure was solved with a resolution of 2.3 Å. The data confirmed the solution structure and added more precise information about a number of regions which appeared rather flexible in solution. An extended six-stranded, anti-parallel ß-sheet was observed by virtue of an approximate twofold symmetry of the dimer. The interactions at the dimer interface involved six main-chain/hydrogen bonds and three hydrogen-bonding interactions in which some side chains were implicated. Phe49 and Leu57 were observed to form van der Waals interactions across the dimer interface. More recently, mutational studies confirmed the critical role of Phe49 and showed the involvement of Ile20 at the dimer interface of κ-bungarotoxin (90). Most strikingly, the structures of the two subunits of the dimer are not identical, a major difference occurring at the tip of the central loop (between Cys27 and Pro36), a region that is believed to be functionally important. The rmsd between the Cα-atoms of two subunits of the κ-bungarotoxin homodimer is 1.2 Å, a value that drops to 0.7 Å when the residues at the tip of loop 2 are excluded from the superposition calculation. The authors suggested that the experimentally determined dimensions of the toxin dimer are consistent with the possibility that the two monomers of the dimer interact simultaneously with two α-subunits of an AChR molecule (88). They also suggested that such a crosslink might inhibit the opening of the ion-channel of the receptor. The structure of a dimeric neuronal toxin is shown in Fig. 1. 4.4. Three-Dimensional Structure of Nonconventional Neurotoxins Recently, the X-ray crystallographic structure of bucandin, a novel toxin isolated from the Malayan krait Bungarus candidus was solved at 0.97 Å resolution (90b). This toxin possesses a fifth disulfide in the first loop. It was observed that the first loop of this toxin is characterized by a 90° turn relative to the rest of the molecule. This feature is the major structural difference as compared to monomeric three-fingered neurotoxins. The authors claimed that this toxin could act presynaptically by enhancing release of acetylcholine. However, this hypothesis has not yet received any experimental support. It may be interesting to investigate whether this toxin, like weak neurotoxins, block muscular AChRs.

398 Servent and Ménez

Fig. 2. Amino acid sequences of various snake toxins. Bold residues and disulfide bonds showed in (Ea) are conserved in all toxins with few exceptions in nonconventional neurotoxins. The proposed numbering largely used in this review is that of Ea without considering the spaces used for alignment.

Snake Neurotoxins

399

4.5. The Functional Significance of Structural Deviations in Three-Finger Toxins Do structural deviations between two toxins with similar folds reflect functional differences? A first answer to this question emerged from the observation that α/κ neurotoxins have a high affinity toward both α7 neuronal (Kd around 10–9 M) and muscular (Kd around 10–10 M) acetylcholine receptors, whereas α-neurotoxins bind with high affinity (Kd around 10–10 M) to muscular receptors only (7). Comparison of primary and tertiary structures of the two toxin categories suggested that this functional difference may be because of the presence in the α/κ neurotoxins of an additional cyclic loop at the tip of the central finger (see Figs. 1 and 2). Not only mutational data support this proposal (90b), but also grafting of this particular small loop on a αneurotoxins caused an increase in affinity constant (Kd) for the neuronal receptor from 2000 nM to 100 nM, the affinity for the muscular receptor remaining unchanged (37). Therefore, the additional cyclic loop uniquely found in long neurotoxins clearly contributes at least in part to the specific recognition of the neuronal receptor. Another striking example that structural differences may be associated with differential functionality was recently reported (91,92). The α-neurotoxin from Naja nigricollis shares much structural similarity with fasciculin, a toxin isolated from the green mamba Dendroaspis angusticeps (93). Despite their structural analogies, the two toxins have no detectable cross biological activities. The neurotoxin blocks muscular AChRs (see above), whereas fasciculin selectively blocks acetylcholinesterase. Superimposition of the 3D crystal structures of the two toxins reveals that the first two fingers display remarkable deviations from one toxin to another. Finger I, half of finger II and a C-terminal residue of fasciculin 2 were “transferred” into toxin α. Not only did the chimeric toxin display a high affinity for acetycholinesterase, only 15-fold lower than that of fasciculin 2 (91), but it also retained the general characteristics of the threefinger toxins and, more strikingly perhaps, the structural features associated with the transferred function were recovered in the chimeric toxin (92). From the aforementioned studies, it is tempting to suggest that structural deviations between two three-finger toxins, and perhaps more generally between two proteins adopting the same fold, may reflect their functional differences. However, the pairwise structural analyses performed hitherto concerned three-finger toxin structures (α- and α/κ neurotoxins; a short-chain α-neurotoxin and a short-chain fasciculin) that could be readily superimposed. This may not be equally simple for any pair of three-finger toxins and more generally for any pair of three-finger proteins. This fold is widespread among living organisms, including animals and plants, where it has various unrelated biological functions, including nontoxic ones (49). For example, it is adopted by a protein from the skin of Xenopus (10), wheat germ agglutinin (11), an inhibitor of activation of complement (12), and a domain of the receptor of the urokinase-type plasminogen activator (13). From available structural data, it is clear that a pair threefinger proteins can display considerable different structural deviations. We anticipate that pairwise comparison of two similarly folded proteins may be functionally informative provided the compared structures are not too different. In any case, it would be of interest to perform a computational pairwise search for progressive structural deviations between all toxins of the three-fingered family. If successful, such an analysis is

400

Servent and Ménez

expected to greatly help identify the different functional regions displayed by the fold, thus enhancing understanding of the molecular evolution of a variation of functions around a structurally conserved theme. 4.6. Refolding of Three-Fingered Neurotoxins About two decades ago, the refolding of neurotoxins was initially studied using a number of natural variants (94). Upon reduction of its four disulfide bonds, a threefinger neurotoxin unfolds, but removal of the reducing agent followed by an incubation in an aqueous medium, at physiological pH, sufficed for neurotoxins to refold and to recover virtually 100% of their physicochemical and biological properties. However, not all the neurotoxins refolded at the same rate and it was observed that neurotoxins with 62 residues refold substantially more slowly than those with 61 or 60 residues (94). It was observed that the slow toxins have a longer turn 2, which joins loop I to loop II, between the invariant cysteine residues 17 and 24 (using the numbering of erabutoxins; see Fig. 2). In the slow toxins (like Eb), the turn comprises 6 residues, whereas it possesses 5 (like N. nigricollis toxin α) or 4 residues for the rapid toxins. Therefore, it was suggested, for the first time, that a turn might control the rate of refolding of a protein. However, as the toxins differed from each other by many other substitutions, the proposed correlation required additional experimental support. This has been recently achieved in a study describing the comparative refolding pathways of toxin α from Naja nigricollis (61 residues), Eb and two synthetic variants (α60 and α62) of toxin α (95). The two mutants differ from toxin α by a single residue that was either inserted at position 18 (α62) or deleted at position 21 (α60) in the turn supposed to be critical for refolding. The refolding was performed in the presence of oxidized (0.3 mM) and reduced (3 mM) glutathione and followed by electrospray mass spectrometry, which allowed identification and quantification of the population of intermediates that occur during the refolding processes. As shown in the paper by Ruoppolo et al., the shorter the turn 2, the faster the refolding (95). This turn does indeed control the folding rate of three-finger neurotoxins. It was shown that the natural and synthetic toxins follow the same overall refolding pathway. This process eventually leads to a unique native-like species with four disulfides that contains only native toxin. Identification of the disulfide bonds that progressively appear during the refolding process may shed light on how a single turn controls the overall folding rate of the three-finger toxin. 4.7. Identification of Binding Topographies on the α-Neurotoxin Surface by Structural Studies How neurotoxins bind to their targets (receptors, ion channels, enzymes, antibodies, etc.) is obviously an important question, which will be reviewed in detail later (see below). Here we shall only present the results of structural studies. A number of previous studies showed that small peptides corresponding to the region 173–204 of the αsubunit of the muscular AChR bind to toxins, although usually with low affinities (96–98). A 1993 paper reported the solution structure of a complex between αbungarotoxin and a receptor peptide (82) that encompasses the sequence 185–196 (KHWVYYTCCPDT) of the α-subunit of the muscular receptor. The two partners bind with an apparent Kd of about 1.6 µM. The peptide residues predominantly interact with

Snake Neurotoxins

401

toxin residues in the N-terminal loop I, in the middle loop II and His 68 in the Cterminal tail. Thus, His186 in the peptide interacts with toxin residues Thr5, Thr6, Ala7, Tyr24, and Gly43. The peptide Trp187 interacts with toxin Thr6, whereas Val188 makes contact with toxin Val39 and Val40. The peptide Tyr189 interacts with toxin residues Thr6, Ile11, and His68. Tyr190 in the peptide interacts with both His68 and Asp30 in the toxin. A similar structural study was reported with α-bungarotoxin bound to a library-derived 13-residue peptide (83). The amino acid sequence of this peptide is MRYYESSLKSYPD. It shares a number of similarities with the receptor peptide used previously, including two adjacent tyrosine residues that seem homologous to Tyr189 and Tyr190 in the receptor peptide 185–196. Four intermolecular interactions involving both tyrosines were common in the two complexes (82,83), the first tyrosine interacting with Thr6, Ile11, and His68 of the toxin and the second with Asp30. However, the library-derived peptide, including its two tyrosine residues, establishes more interactions with the toxin than the receptor peptide (65 interactions vs 25), which may account for the higher affinity of the library-derived peptide for the receptor. Very recently, the same two groups have reported the refined structures of complexes between α-bungarotoxin and peptides derived from the α-subunit of AChR (83a,83b). In the latter study, the peptide 182-202 of αAChR was observed to adopt a β-hairpin conformation when bound to the toxin, forming an intermolecular β-sheet with residues Lys38-Val40 of the second finger of the toxin. It may be argued that the structure of a complex between a toxin and a receptor peptide does not necessarily reflect how the toxin binds to the native receptor. More structural work is now needed to clarify this important question. Proton 2D-NMR studies have also been used to map epitopes at the surface of the short-chain toxin α from Naja nigricollis. The principle is simple. Exchange behavior can be observed by NMR for a substantial proportion of individual amides and labile side-chain hydrogens and antibody binding is expected to affect this exchange profile, especially within the site where the complex is formed. Comparison of the exchange rates in the free toxin and in the antibody-complexed toxin should provide information on the protons that are protected against exchange upon antibody binding and hence on the toxin region that is recognized by the antibody. Two toxin-α-specific monoclonal antibodies (MAbs), Mα1 and Mα2-3, were previously produced and both were shown to neutralize the toxin and to recognize different topographical epitopes (99,100). These epitopes were therefore mapped by 2D-NMR (101). The antigenic region recognized by Mα1 is mostly composed of side chains centered on finger I and in the C-terminal part in contact with loop I. This group comprises the side chains of Asn5, Thr13, Lys15, which are all oriented like Lys15, the side chains of Glu2, Thr4, Thr16, Pro18, and Lys58, which are all oriented like Pro18, and the NH2 terminus (see sequences in Fig. 2). These side chains form a homogeneous surface of about 750 Å2, which is consistent with the surface of epitopes defined by crystallographic means (102). This description is nicely supported by a number of chemical modifications of the toxin performed at a single position (99). Mα1 is essentially toxin α-specific and does not recognize most other neurotoxins, like Eb, for instance. A close inspection of the 3D structures of the epitope in toxin α and of the homologous region in Eb, suggested that the latter toxin may not be recognized by the antibody because of the presence of an additional serine at position 18 in the large turn 2 that joins loops I and II. A mutant of

402

Servent and Ménez

Eb was therefore synthesized by genetic engineering to delete the additional serine (103). Strikingly, this simple deletion sufficed to confer on Eb the full capacity to be recognized by Mα1. Clearly, the additional serine at position 18 not only affects the folding rate of the toxin (94,95) but also largely contributes to the antigenic properties of the toxin. However, the deletion of this residue has no effect on the capacity of the toxin to bind to AChR (103). In sharp contrast to Mα1, the other toxin-α-specific MAb, Mα2-3, binds to all short neurotoxins, including Eb, to which it binds with an even higher affinity, suggesting that the epitope is shared by the α-neurotoxins (100). Proton 2D-NMR was used to identify the surface recognized by this antibody (101). This epitope involves a group of side chains that are located in the central area of finger II and parts of fingers I and III, and that are mostly oriented toward the concave face of the toxin. Using the numbering of Eb (see Fig. 2), this group of side chains includes Tyr25, Lys27, Trp29, Ile36, and Glu38 in the central part of loop II, the side chain of Gln7 in loop I, which interacts with Glu38; the side chain of Ile50 in loop III, which interacts with the side chain of Trp28; and the side chain of Lys47, which interacts with Ile50. Since this study was published, an extensive mutational study of Eb has confirmed and even refined the localization of the epitope. It clearly encompasses the three fingers on the concave face of the toxin. And as expected, the epitope is composed of residues that are largely conserved in neurotoxins, most epitope residues in fact also being major actors in toxin recognition of AChR (73,104). Even more recently, a low-resolution model of the structure of the toxin-Mα2-3 complex was reported, based on combined mutational and modeling studies. More precisely, a model of the structure of the variable fragment of Mα2-3 was first established by a homology modeling-based protocol (105). Then, the residues forming the paratope of the antibody were identified on the basis of mutational analyses of its Complementarity Determining Region (CDR) residues (104). Using all available data, the structure of an Mα2-3/toxin complex was studied by directly docking the toxin epitope onto the antibody paratope (106). However, a large number of possible solutions were found, the toxin never adopting the same orientation. To introduce some experimental constraints into the docking procedure, a double-mutation cycle procedure was used to identify pairs of interacting residues between the toxin and the antibody (107). Thus, evidence was obtained which indicated that Arg33 in the toxin is close to, and perhaps interacts with, Asp31 in CDR1H. The use of this pair of proximate residues during the selection procedure yielded three models based on docking calculations. The selected models predicted the proximity of Lys47 in the toxin to Tyr49 and/or Tyr50 in the antibody. This was experimentally confirmed using another round of double-mutation cycles. The two models finally selected were submitted to energy minimization in a CHARMM22 force field, and were characterized by an rmsd of 7.0 ± 2.9 Å. Both models display most features of antibody-antigen structures. First, the buried surfaces are about 2000 Å2 in area, in agreement with the 1200–2300 Å2 range seen in crystallographic protein-Ab complexes (102). Second, a shape complementarity of surfaces was observed with the prominent CDR3H of the paratope which fits with the concave shape of the epitope. Third, the prediction of one salt-bridge agrees with the observation that most Ab/Ag complexes contain between 0 and 3 of these linkages (108,109). Fourth, the interacting residues in the antibody involve five tyrosines, in agreement with the observation that an Ab paratope in Ag/Ab

Snake Neurotoxins

403

complexes usually comprises between 3 and 8 aromatic residues (110). Fifth, as observed in most other Ab/Ag complexes (102), the contact area comprises a large hydrophobic core that includes residues Ala54, Met100, Gly101, Ala102, Thr103, and Leu106 in the CDR3H. Although the models proposed in the study of the toxin/antibody structure suffer from some weaknesses, they offer an interesting molecular basis to the understanding of the fine specificity of recognition of the antibody and therefore the mechanism by which this antibody neutralizes the toxin. Also, since Mα2-3 also partially mimics some binding properties of AChR, these structural features may also further clarify how toxins bind to AChR. 4.8. Dynamic Properties of Neurotoxins In most studies aiming at identifying how a toxin binds to its target or to antibodies, the available technologies essentially identify the elements that are in contact, without giving much information about the importance of the dynamic properties in the binding process. To appreciate the importance of the dynamic parameters in the formation of a toxin/target complex, it is necessary to identify them for the two partners both in the free and bound states. We still are rather far from this knowledge, although a recent study described the dynamics of a snake neurotoxin on a picosecond to hour time scale, suggesting some correlation with its toxic and antigenic properties (111). The study was achieved using a large amount (14 mg) of toxin α from Naja nigricollis, fully labeled with –15N. Movements of the toxin backbone and/or side chain have been determined on a time scale of picoseconds to nanoseconds and microseconds to milliseconds using 15N relaxation measurements, microseconds to milliseconds by means of off-resonance ROESY, and minutes to hours following H-D exchange kinetics. Clearly, the central part of the loop I (His4, Lys15), loop II (residues Tyr25 to Trp29 and Ile37 to Gly40), and the C-terminal region (Cys55, Lys59) have a rigid backbone on all these time scales. Around this rigid core, the toxin displays variable movements on the different time scales. It is unknown, however, whether this particular mobility pattern is similar in other three-finger proteins. This knowledge clearly provided an interesting new basis on which to shed light on some specific properties of three-finger neurotoxins. Thus, turn 2, which is associated with the rate of toxin refolding (see above), has 5 residues in toxin α, and is highly mobile on all time scales, the backbone being particularly subject to substantial movements on the picosecond to nanosecond time scale. Is this mobility altered when the loop shortens to five residues, as in 60-residue toxins, or when it lengthens to 6 residues as in Eb? More precisely, does the decrease in size of the turn from 6 to 5 and 4 residues between cysteines 17 and 24 cause a change (a decrease?) in the local mobility, which in turn would favor the involvement of these two cysteines in the disulfide bonds most appropriate for rapid refolding? Identification of the disulfide bonds that appear progressively during the folding process should provide an answer. Assuming that toxin α has a similar AchR binding site as compared to Eb (see above and below), it was observed that this site is formed around a rigid core (Lys27, Trp29, and Glu38), surrounded by residues from the two most mobile regions of the molecule on the picosecond to nanosecond time scale, Asp31 and Arg33 at the tip of the central loop and Lys47 on the third finger. This observation is all the more interesting as these

404

Servent and Ménez

residues are among the most important in terms of binding contributions to the acetylcholine receptor (104). With mobility on the microsecond to millisecond time scale, the functionally important residues Gln7, Ser8, and Gln10 are located on the tip of the first finger. The side chains of Gln7 and Gln10 are even mobile on the second time scale. Therefore, the “toxic” site of a neurotoxin appears as a small rigid core from which radiate residues with variable degrees of mobility on various time scales. It is unknown whether or not this constitution is important for the high affinity with which the toxin binds to the receptor, and whether this pattern of mobility is modified upon formation of the toxin-receptor complex. 5. THE SITES BY WHICH SNAKE TOXINS INTERACT WITH NICOTINIC ACETYLCHOLINE RECEPTORS 5.1. The Nicotinic Acetylcholine Receptors (AChRs) AChRs control the transformation of a chemical message into an electrical signal, the binding of the neurotransmitter, acetylcholine, generating postsynaptic depolarization. AChRs belong to the same ligand-gated ionic channel family (LGIC) as 5HT3, GABAA, and glycine receptors, which are all composed of five homologous transmembrane subunits that surround a central channel (112). Each subunit comprises a large extracellular N-terminal domain (about 210 residues), four transmembrane regions (M1 to M4), a cytoplasmic part between the M3 and M4 domains and a short extracellular C-terminal tail (Fig. 3A) (reviewed in refs. 112,113). Both the subunit composition and stoichiometry of AChRs vary with tissue localization and are associated with different pharmacological and electrophysiological properties (114). At the neuromuscular junction or in the electric organs of Torpedo and electric eel, the receptor possesses four different subunits arranged around the channel in the order αγαδß or αεαδß in the fetal and adult muscle, respectively. In the peripheral and central nervous systems (PNS/CNS), a great variety of subunit arrangements is also observed, some receptors are αnßm with various combinations of α (from α2 to α6) and ß (from ß2 to ß4) subunits, whereas others are homopentameric with five identical subunits (α7, α8, and α9) (reviewed in ref. 114). The extracellular domain of the receptor possesses two binding sites for agonists (acetylcholine, nicotine) and competitive antagonists (neurotoxins, d-tubocurarine). This domain is linked allosterically to the receptor channel that is mainly constituted by the second transmembrane region of each subunit, the affinities of the ligands depending on the different receptor states (resting or desensitized). Electron microscopy studies have indicated that AChR is a pentamer 115 Å long, approx half of which is extracellular. The external diameter of the receptor is about 80 Å, whereas the diameter of the pore is around 30 Å in the synaptic part and about 7 Å in the channel (Fig. 3) (115,116). While preparing this review, the 3D structure of a pentameric protein sharing a number of similarities with the extracellular domain of AChRs was published at atomic resolution (116a). This protein was isolated from the neurones of the mollusc Lymnaea stagnalis, where it plays a role in the regulation of the cholinergic transmission (116b). The protein (AChBP) possesses the major characteristics of the extracellular domains of the members of the Ligand Gated Ion Channel superfamily. It adopts an IgG-topology (Fig. 3B). The ligand binding sites are localized at the interfaces between two monomers, a region which also includes the residues that interact with snake toxins.

Snake Neurotoxins

405

Fig. 3. (A) Topological organization of the nicotinic acetylcholine receptor and model of the transmembrane organization of the AChR subunits. (B) The pentameric structure of AChBP. Top view of the 5 subunits (A to E), with the 5 ligand-binding sites at each interface. The toxins are supposed to interact at each promoter interface (indicated by arrows). Adapted from ref. 116a with Nature’s copyright permission.

406

Servent and Ménez

This major finding opens the possibility to investigate the 3D structure of toxin-AChR complexes either by docking toxins to models derived from the X-ray structure of AChBP, and/or by proceeding to crystallographic analyses of toxin-AChBP complexes. This is a most exciting phase of our understanding as to how toxins block AChRs, which may stimulate many efforts. Then, it will be of particular interest to compare these data with those describing the various 3D structures in solution of complexes between toxin and receptor-derived peptides. 5.2. Receptor Binding Sites on α-Neurotoxins Since their discovery more than 30 yr ago (16,17), α-neurotoxins from venom of Elapidae have been subjected to many studies designed to understand the molecular basis underlying their specific and tight (equilibrium dissociation constants 10–9 to 10–11 M) interaction with AChRs (1,117,118). 5.2.1. The Initial Approaches

Amino acid sequence comparisons and chemical modifications were pioneering approaches in identifying functionally important residues of α-neurotoxins (reviewed in ref. 1). This work revealed the functional importance of positively charged (Lys, Arg) residues (118–122), aromatic residues (Trp, Tyr) (118,120,123), and the structural role of disulfide bonds (118,121). Also, the C-terminal tail of the long-chain toxins was proposed to be functionally important (119,124). However, despite considerable efforts involving a sort of “mutational” change by chemical means, many toxin residues remained refractory to modifications (1). 5.2.2. The Recombinant Approaches

To have access to a complete analysis of an α-neurotoxin sequence by site-directed mutagenesis, the cDNA encoding Ea was isolated (41), fused to two domains of Protein A from S. aureus, and expressed as a fusion protein in the periplasmic space of E. coli (23). The toxin and Protein A domains being separated from each other by a methionine, the fusion protein was cleaved efficiently by cyanogen bromide, leading to a toxin that was biologically and chemically indistinguishable from venom Ea (23,125). Even its 3D structure at 1.8 Å resolution was identical, within experimental error, to that of the venom toxin (65). Since then, various expression systems have been developed to produce different recombinant neurotoxins, as described in Subheading 2.2. and summarized in Table 2. 5.2.2.1. FUNCTIONAL SITES BY WHICH α-NEUROTOXINS FROM ELAPIDAE INTERACT WITH MUSCULAR ACHR Initial attempts to explore the functional residues through which α-neurotoxins interact with the muscular-type receptor were made with Ea (27,67,73). Mutations of 41 of the 62 toxin residues, allowed identification of most, if not all, of the residues that constitute the site by which this short toxins interact with the Torpedo receptor. This functional site includes conserved and variant residues, located at the tip of loop I (Gln7, Ser8, Gln10), on the central loop (Lys27, Trp29, Asp31, Arg33, Ile36, Glu38) and in loop III (Lys47) (Fig. 4). These results have confirmed the initial chemical studies. They also concord with mutational investigations of NmmI from Naja naja mossambica, which demonstrate the crucial role of the positively charged residues at positions 27, 33, 36 and to a lesser extent 47 (28,126,127). Furthermore, these authors

Snake Neurotoxins

407

Fig. 4. Functional sites by which Ea and Cbtx interact on Torpedo and α7 nicotinic receptors. Residues whose mutations caused affinity decrease for AchRs less than fivefold are in green, between five- and 10-fold in yellow, 10- and 100-fold in orange, and up to 100-fold in red. The only residue for which mutation induced an affinity increase is in blue.

408

Servent and Ménez

observed a differential effect of these mutations at the two NmmI binding sites and identified the high- and low-affinity sites on the mouse muscular receptor at the α/δ and α/γ interfaces, respectively (28). Similarly, a differential binding effect was also recently reported for a number of Ea mutants (32). 5.2.2.2. FUNCTIONAL SITES BY WHICH α/κ-NEUROTOXINS INTERACT WITH MUSCULAR ACHR Two long toxins, α-Bgtx and α-Cbtx, have also been submitted to site-directed mutagenesis (Table 1). Of the four mutations made on α-Bgtx, those made at Arg36 (which is homologous to Arg33 in short toxins) indicated that this residue is the most critical in toxin binding to muscular-type AChR (128). This study also indicated the importance of Lys26. In contrast, truncation of the last seven residues had little effect on toxin affinity, the mutation D30A being of no consequence (33,128). A recent extensive mutational analysis of α-Cbtx revealed that mutations at only 8 of the 29 explored positions caused significant decreases in affinity toward the Torpedo receptor (32). These important positions mostly belong to the toxin loop II (Lys23, Trp25, Asp27, Phe29, Arg33, and Arg36), with the exception of Lys 49, which belongs to the third loop, and Phe65, which is in the C-terminal tail (Fig. 4). The crucial role of Arg33 (homologous to Arg36 in α-Bgtx), as well as the functional importance of Lys23 (Lys 26 in α-Bgtx), were confirmed, suggesting that these two residues play comparable functional roles in long-chain toxins. It was noted that the quantitative effects of mutations introduced at the homologous positions in the two toxins were not identical, although this may well reflect the differential nature of the introduced mutations (alanine substitution in α-Bgtx and charged inversion in α-Cbtx). In addition, it was observed that mutations K23E, K49E, and perhaps R33E caused differential effects at the two toxin-binding sites, suggesting a distinct interaction of these residues at the α/γ and α/δ interfaces, in agreement with data on NmmI toxin (28). In summary, the available data indicate that a number of structurally equivalent residues are commonly involved in the binding of α- and α/κ-neurotoxins to musculartype AChRs. Thus, using respective numberings as in α-Cbtx and Ea (Fig. 2), these residues seem to systematically include the highly conserved Lys23/27, Trp25/29, Asp27/31, Phe29/32, Arg33/33, and Lys49/47. Beside these similarities, the functional sites of long and short toxins seem to share differences, some of their regions or residues being selectively important for one toxin type only. Thus, Glu 38 and the tip of loop I are important for Ea but not for α-Cbtx, whereas a small part of the C-terminal tail (Phe65) is slightly important for the long toxin only (Fig. 4). Therefore, the α- and α/κ-neurotoxins from Elapidae venoms do not form a homogeneous family wherein all members interact identically with muscular-type AChR. This conclusion therefore supports a separate classification of the two subfamilies as shown in Table 1. 5.2.2.3. FUNCTIONAL SITES BY WHICH α/κ-NEUROTOXINS BIND TO NEURONAL α7 ACHR The numerous mutants prepared for α-Cbtx have also been exploited in an attempt to understand the molecular basis of this toxin’s ability to block not only muscular AChR but also the neuronal α7 receptor. The mutational experiments showed that Trp25, Asp27, and Arg33 are involved in the binding to both receptors and that their energetic contributions to bindings are comparable for the two receptors. They also showed that Phe29, Arg36, and Phe65 are involved in binding to both receptors, but their contribution varies substantially from one receptor to the other. In addition, Lys23

Snake Neurotoxins

409

and Lys 49 are selectively involved in binding to the Torpedo receptor, whereas Cys26Cys30 and Lys35 are uniquely implicated in binding to the α7 subtype (Fig. 4) (90b). Therefore, α-Cbtx interacts with two different AChRs using both a common core of positively charged and aromatic residues and few additional amino acids specific to each receptor subtype. That the same residues are involved in the two binding sites suggests that the toxin binds to comparable complementary determinants on the two receptor subtypes. These determinants possibly occupy homologous positions in the two receptors. However, since the two binding sites also involve subtype-specific residues, the two homologous binding determinants most likely display marked differences. 5.2.2.4. FUNCTIONAL SITES BY WHICH κ-NEUROTOXINS INTERACT WITH NEURONAL ACHR Mutational experiments have been used to understand how the dimeric κ-Bgtx binds to the neuronal receptor from chick ciliary ganglia (29,129). The authors focused on residues highly conserved in neuronal toxins. They observed that mutations Q26W and P36A caused no affinity decrease, suggesting that Gln26 and Pro36 were not critical for the selective interaction of the toxin in the κ-Bgtx with the neuronal receptor. In contrast, P32K caused a 13-fold affinity decrease, suggesting its functional importance (29,129). The drastic loss of affinity observed upon mutation of Arg34 (homologous to Arg33 in Ea or α-Cbtx) into alanine, shows that this residue plays a major role. Removal of the C27-C31 disulfide bond, at the tip of the loop II, causes a 50-fold affinity decrease, indicating a potentially specific role of this disulfide in the selective interaction with neuronal receptor (130). Other mutations have indicated that two toxin residues, Ile20 and Phe49, are required for the formation of an active dimeric κ-Bgtx (90). The systematic removal of the disulfide bonds of the toxin by site-specific mutagenesis indicates that the mutations C3A-C21A and C14A-C42A have no effect on the toxin activity, in sharp contrast to C46A-C58A and C59A-C64A, which completely abolished the toxin’s folding and thus its biological function (130). 5.2.2.5. CONCLUSIONS The body of currently available data suggests some general rules regarding how a snake toxin binds to an AChR subtype. First, Arg33 at the tip of the central loop appears to be a most critical toxin residue in all kinds of toxin-receptor interactions. Second, some conserved positively charged residues (Lys27, Arg33, and Lys47 using Ea numbering) interact differentially at the two binding sites of the muscular-type AChR, suggesting a nonidentical positioning of the toxins on these two sites. Third, α- and α/κ-neurotoxins bind to muscular-type receptor using conserved functional amino acids but also specific residues in each toxin family. Fourth, an α/κ neurotoxin can interact with muscular-type and α7 receptors using a common core of residues assisted by subtype-specific residues. Finally, the fifth disulfide bond at the tip of loop II might play a specific role in the capacity of α/κ- and κ-neurotoxins to interact with neuronal receptors, as observed with α7 and α3ß2. 5.2.3. The Chemical Approaches The small size of waglerins (22–24 residues) and the presence of only one disulfide bond in their sequences was shown to be compatible with chemical synthesis. Thus, structure-function studies involving this approach proved appropriate to identify amino

410

Servent and Ménez

acids involved in the high- and low-affinity interaction site at the α/ε and α/δ interfaces of the muscular AChR (131,132). The single disulfide bond of waglerin I is critical for its bioactivity, as judged from the large decrease in toxicity that results from replacement of the two half cystines by alanine or serine (2,3). Also, part of the active site of waglerin I is close to the disulfide bond, including His10 (133). The C-terminal region of waglerin I is also functionally important, since deletion by chymotrypsin of its 7 last residues caused inactivation (3). Also, some of the first five residues of waglerin I may be functionally important as indicated by proteolytic treatment. The negative charges seem unimportant since substitution D5N and amidation of the carboxy-terminus has no effect on the toxicity of waglerin I (2). Progressive deletions at the N- or C-terminal regions of waglerin I suggested that Asp5 and Leu6 and Ile16, Pro17, and Arg18 are involved in the high-affinity binding at the α/ε interface (132), whereas individual substitutions close to the disulfide bond, as at histidines 10 and 14, selectively altered affinity at the α/δ binding site (132). Waglerin I, therefore, seems to interact differentially at the two binding sites of the muscular receptor. 5.3. Neurotoxin Binding Sites on AChRs 5.3.1. Affinity and Photoaffinity Labeling Experiments

The topographical mapping of residues involved in the binding of various ligands to AChRs was initially investigated with specific labeling probes such as MBTA (134), DDF (135,136), nicotine (137), d-tubocurarine (138) and lophotoxin (139). On the basis of these studies, a three-loop model of the agonist and small competitive antagonist binding site on the α subunit of the muscular receptor was proposed (reviewed in ref. 140,141). As indicated in Fig. 5, loop A is centered on Tyr93, loop B around Trp149, and the critical loop C included Tyr190, the Cys doublet 192–193 and Tyr198. More recently, labeling of the non-α subunit was reported, showing involvement of residues from the γ and δ subunits in the constitution of the ligand-binding sites (138,142). These results have been confirmed and extended by site-directed mutagenesis experiments on the various subunits of muscular and neuronal receptors. In particular, the functional role of residues in the 180–200 region was investigated in detail in the α1 and α7 subunits (143-148) and the role of residues of the γ or δ subunits in the interaction of agonist/competitive antagonist was reinforced, with identification of four complementary functional loops around residues 34, 55, 110, and 170 (138,149–152) (reviewed in ref. 153). If the site by which small ligands (ACh, nicotine, d-tubocurarine, etc.) bind to various AChRs is now well documented, this was not the case for neurotoxins until recently. Early attempts to localize the α-neurotoxin-binding sites on AChRs were based on electronic microscopy analyses and fluorescence measurements, which indicated that α-Bgtx binds to two sites on the receptor located at the outside face (154) or on the top of the receptor (155–157). Photoaffinity labeling experiments using neurotoxin derivatives reported coupling to α and non-α subunits to different extents, depending on the position and type of derivatives introduced in the toxin (122,158–162). From these

Snake Neurotoxins

411

Fig. 5. Toxin-binding sites on the different subunits of the muscular and α7 receptors. (The sequences of the α,γ,δ, and ε subunits are those of the mouse muscular receptor, whereas the rat α7 sequence is presented. The sequence alignment was made using ClustalX. The residues in bold are included in the different funtional loops involved in the ligand interaction on the α: loops A, B, C, and γ or δ subunits: loop I, II, III, and IV)

studies, structural models of the toxin-binding sites were proposed, suggesting a location at the top of the receptor (162) or in a deep pocket near the membrane at the subunit interface (163,164). Direct-affinity labeling, with different cysteine derivatives of toxin α from Naja nigricollis and mildly reduced Torpedo receptor, suggested a model in which the tips of loops I and II of the toxin are located close to the disulfide bond 192–193 of the receptor α subunit (165,166).

412

Servent and Ménez

5.3.2. Chemical Synthesis of Receptor Peptides

Another way of identifying regions or residues of the receptor involved in toxin recognition is to use synthetic peptides from the α subunit sequences of muscular or neuronal receptors and to test their ability to bind to α- or κ-Bgtx, respectively. Although this approach may not reflect the genuine role of these sequences in the intact receptor (see Subheading 5.5.), it has been largely exploited to probe the putative functional role of the 180–200 region of the α-subunit (review in ref. 96–98). As shown by an NMR study, α-Bgtx undergoes a conformational change upon binding to the α-185–196 peptide fragment (82). To understand why AChRs from snakes and mongoose fail to interact efficiently with α-neurotoxins, various peptides encompassing the 180–200 region of their α-subunits were synthesized. A number of residues substituted in this receptor fragment were proposed to account for toxin resistance (167,168). Glycosylation at position 187 or 189 in the snake and mongoose AChRs was identified as responsible for toxin resistance (169,170). 5.3.3. Site-Directed Mutagenesis Experiments

Figure 5 shows a sequence alignment of different AChR subunits and indicates the different residues that were identified as being involved in interactions with various neurotoxins as a result of receptor mutations. Taylor’s group has nicely analyzed how mutations introduced into the α, γ, δ, and ε subunits of the recombinant murine muscular receptor affect the binding of the short-chain NmmI toxin (28,127,171,172). They observed that although acetylcholine binds to a site that encompasses the three loops A, B, and C, only mutations at 4 positions in loop C caused substantial affinity decreases (Fig. 5), which were differentially expressed at the α/γ and α/δ interfaces. Using double-mutant cycle analyses, specific pairwise interactions were then identified between 3 toxin residues (Lys27, Arg33, Lys47) and 4 residues on the receptor α subunit (Val188, Tyr190, Pro197, Asp200). The most important energy coupling (∆∆G = 2.6 kcal/mol) was observed between the toxin Arg33 and αVal188 (171). Furthermore, the construction of chimeras between the γ and ε subunits allowed this laboratory to show that the residues γPro175 and γGlu176 confer high affinity for the toxin on the α/γ interface. Finally, studies with various toxin-receptor mutant combinations have indicated strong pairwise interactions between γGlu176 and Lys27, γLeu119 and Arg33, and γTrp55 and Lys27 or Arg33 (127,172). The cysteine-scanning mutation of residues in the loop III of the γ, δ, and ε subunits (Fig. 5), associated with the use of sulfhydryl-specific reagents, showed that Leu119 is a potential site of interaction for α-Bgtx (173). An additional study based on the same methodology suggested that Trp187, Val188, Phe189, Tyr190, and Pro194 in the loop C of the α-subunit contribute to the α-Bgtx binding site (174). Finally, introduction into toxin-insensitive subunits of fragments of toxin-sensitive subunits shed more light on subunit regions involved in toxin binding. Thus, insertion of the sequence 184–191 of the highly sensitive Torpedo α1 or rat α7 subunits into the insensitive α3 subunits sufficed to provide the α3 receptor with a nanomolar affinity for α-Bgtx (175). Also, the simple substitution of the Lys189 from the α3 sequence by the homologous Tyr of the α1 subunit sufficed to make this neuronal receptor significantly sensitive to αBgtx, thus demonstrating the critical binding role of this aromatic residue (175). There-

Snake Neurotoxins

413

fore, as previously observed for agonists, small antagonists and short-toxin NmmI, the loop C and particularly the aromatic residues in this region are highly critical in long αBgtx toxin’s interaction with AChRs. As previously described, waglerins are small toxins that interact selectively with mature mouse muscular receptor subtypes. Structure-activity relationship studies have shown that Asp59, His61, Tyr115, and Asp173 of the ε subunit are involved in the determinant recognized by this toxin (Fig. 5) (132,176). More precisely, the length of the side chain at position 59 and the aromaticity at position 115 were shown to be highly critical for the nanomolar interaction of waglerin-1 at the α/ε interface of the adult mouse muscular AChR. The functional residues of muscular and α7 receptor subunits involved in the interaction with conotoxins GI, MI or ImI are shown in Fig. 5, illustrating the multipoint attachment of these toxins to AChRs. Clearly, residues from the α and non-α subunits are involved in these interactions. They will be described in more detail in chapter 9 of this book. Using both individual mutations and constructions of chimeric receptors, it was shown that when κ-Bgtx binds to the neuronal α3ß2 receptor, (1) α3Thr147 is involved as part of a consensus glycosylation crucial to toxin sensitivity; (2) some residues in the region 185–205 of the α3 subunit, such as Lys185, Ile188 and Gln198, are involved in toxin binding; and (3) Thr59 in the ß2 subunit is critical in determining κ-Bgtx sensitivity (177–179). 6. CONCLUSION Molecular studies carried out over the past decade have greatly improved our understanding of how an α-neurotoxin fulfills its biological function. The sites by which three-fingered α- and α/κ-neurotoxins bind to muscular-type AChR have now been delineated. The dynamic properties of an α-neurotoxin, including the region that is responsible for its AChR-blocking activity, have been identified on time scales ranging from picosecond to hours. The site by which a α/κ-neurotoxin blocks a neuronal AChR (α7) has also been identified. The available data have revealed a remarkable plasticity in the number and nature of functional residues that surround a few consistently important residues located at the tip of the central loop of the three-finger architecture. These findings may provide a base on which to build a possible scenario for the molecular evolution of AChR-blocking sites in toxins adopting a three-fingered architecture. This in turn might explain how the three-finger fold can exhibit so many different toxic and nontoxic functions. Explanation of how a toxin and AChR form a stable complex still requires much work. Mutational analyses and fine photoaffinity labeling have proven very helpful in identifying pairs of interacting residues between toxins and receptors. Accumulation of such data are of considerable value in constraining novel structural models of the receptor (153,180). The recently solved X-ray structure of an analog of the extracellular domain of AChR (116a) should rapidly provide an answer as to how toxins block AChRs. Finally, the three-fingered neurotoxins have proven particularly useful in investigations of more general questions associated with biochemistry of proteins. Thus, preliminary studies have shed some light on a possible pathway of refolding of a

414

Servent and Ménez

three-fingered neurotoxin. Also, the mechanisms by which a protein can be neutralized by its antibodies are now better understood. In particular, a three-dimensional model of a complex between a toxin and a MAb covering the “toxic” site has been proposed on the basis of combined mutational analyses, modeling of the antibody, and docking of both partners. This knowledge may be of value for the design of new approaches aimed at protecting human beings from envenomation. 7. ACKNOWLEDGMENTS We wish to thank Drs. F. Legall for his help and Drs F. Ducancel and P. Drevet for critical reading of the manuscript and fruitful discussions. REFERENCES 1. Endo, T. and Tamiya, N. (1991) Structure-function relationships of postsynaptic neurotoxins from snake venoms, in Snake Toxins (Harvey, A. L., ed.), Pergamon Press, New York, pp. 165–222. 2. Schmidt, J. J. and Weinstein, S. A. (1995) Structure-function studies of waglerin-I, a lethal peptide from the venom of Wagler’s pit viper, Trimeresurus wagleri. Toxicon 33, 1043–1049. 3. Schmidt, J. J., Weinstein, S. A., and Smith, L. A. (1992) Molecular properties and structure-function relationships of lethal peptides from venom of Wagler’s pit viper. Toxicon 30, 1027–1037. 4. Couturier, S., Bertrand, D., Matter, J. M., Hernandez, M. C., Bertrand, S., Millar, N., et al. (1990) A neuronal nicotinic acetylcholine receptor subunit (α7) is developmentally regulated and forms a homo-oligomeric channel blocked by α-Btx. Neuron 5, 847–856. 5. Gerzanich, V., Anand, R., and Lindstrom, J. (1994) Homomers of α8 and α7 subunits of nicotinic receptors exhibit similar channel but contrasting binding site properties. Mol. Pharmacol. 45, 212–220. 6. Elgoyhen, A. B., Johnson, D. S., Boulter, J., Vetter, D. E., and Heinemann, S. (1994) Alpha 9: an acetylcholine receptor with novel pharmacological properties expressed in rat cochlear hair cells. Cell 79, 705–715. 7. Servent, D., Winckler-Dietrich, V., Hu, H. Y., Kessler, P., Drevet, P., Bertrand, D., and Ménez, A. (1997) Only snake curaremimetic toxins with a fifth disulfide bond have high affinity for the neuronal alpha 7 nicotinic receptor. J. Biol. Chem. 272, 24,279–24,286. 8. Chiappinelli, V. A., Wolf, K. M., DeBin, J. A., and Holt, I. L. (1987) Kappa-flavitoxin: isolation of a new neuronal nicotinic receptor antagonist that is structurally related to kappa-bungarotoxin. Brain Res. 402, 21–29. 9. Chiappinelli, V. A. (1991) κ-neurotoxins and α-neurotoxins: effects on neuronal acetylcholine receptors, in Snake Toxin (Harvey, A. L., ed.), Pergamon Press, New York, pp. 223–258. 9a. Chang, L. Lin, S., Wang, J., Hu, W-P., Wu, B., and Huang, H. (2000) Structure-function studies on Taiwan cobra long neurotoxin homolog. Biochim. Biophys. Acta 1480, 293–301. 9b. Utkin, Y. N., Kukhtina, V. V., Maslennikov, I. V., Eletsky, A. V., Starkov, V. G., Weise, C., Franke, P., Hucho, P., and Tsetlin, V. I. (2001) First, tryptophan-containing weak neurotoxins from cobra venom. Toxicon 39, 921–927. 9c. Aird, S. D., Womble, G. C., Yates, J. R., and Griffin, P. R. (1999) Primary structure of γbungarotoxin, a new postsynaptic neurotoxin from venom of Bungarus multicinctus. Toxicon 37, 609–625. 9d. Nirthanan, S., Gopalakrishnakone, P., Gwee, M. C. E., Khoo, H. E., Cheah, L. S., and Kini, M. R. (2000) Candoxin, a three-finger toxin isolated from Bungarus candidus snake

Snake Neurotoxins

10.

11.

12.

13.

14. 15. 16.

17. 18.

19.

20.

21.

22.

23. 24. 25. 26.

415

venom, is a reversible post-synaptic neuromuscular blocker of nicotinic acteylcholine receptors. Abstract L108 of the XIIIth world Congress on the International Society on Toxicology. Kolbe, H. V. J., Huber, A., Cordier, P., Rasmussen, U. B., Bouchon, B., Jaquinot, M., et al. (1993) Xenoxins, a family of peptides from dorsal gland secretion of Xenopus laevis related to snake venom cytotoxins and neurotoxins. J. Biol. Chem. 268, 16,458–16,464. Drenth, J., Low, B., Richardson, J. S., and Wright, J. S. (1980) The toxin-agglutinin fold. A new group of small protein strucutres organized around a four disulfide core. J. Biol. Chem. 255, 2652–2655. Kieffer, B., Driscoll, P. C., Campbell, I. D., Willis, A. C., Van der Merve, A. P., and Davis, S. J. (1994) Three-dimensional solution structure of the extracellular region of the complement regulatory protein CD59, a new cell-surface protein domain related to snake venom neurotoxins. Biochemistry 33, 4471–4482. Ploug, M. and Ellis, V. (1994) Stucture-function relationships in the receptor for urokinase-type plasminogen activator. Comparison to other members of the Ly-6 family and snake venom α-neurotoxins. FEBS Lett. 349, 163–168. Ménez, A., Bontemps, F., Roumestand, C., Gilquin, B., and Toma, F. (1992) Structural basis for functional diversity of animal toxins. Proc. R. Soc. Edinburgh 99B, 83–103. Tsetlin, V. (1999) Snake venom alpha-neurotoxins and other “three-finger” proteins. Eur. J. Biochem. 264, 281–286. Chang, C. C. and Lee, C. Y. (1963) Isolation of neurotoxins from the venom of Bungarus multicinctus and on their modes of neuromuscular blocking action. Arch. Int. Pharmacodyn. Ther. 144, 241–257. Eaker, D. and Porath, J. (1967) The amino acid sequence of neurotoxin from Naja nigricollis venom. Jpn. J. Microbiol. 11, 353–355. Hider, R. C., Karlsson, E., and Namiranian, S. (1991) Separation and purification of toxins from snake venoms, in Snake Toxins (Harvey, A. L., ed.), Pergamon Press, New York, pp. 1–34. Chang, C. C., Huang, T. Y., Kuo, K. W., Chen, S. W., Huang, K. F., and Chiou, S. H. (1993) Sequence characterization of a novel alpha-neurotoxin from the king cobra (Ophiophagus hannah) venom. Biochem. Biophys. Res. Commun. 191, 214–223. Lin, S. R., Leu, L. F., Chang, L. S.. and Chang, C. C. (1997) Amino acid sequence and chemical modification of a novel α-neurotoxin (Oh-5) from king cobra (Ophiophagus hannah) venom. J. Biochem. 121, 690–695. Stöcklin, R., Mebs, D., Boulain, J. C., Panchaud, P.-A., Virelizier, H., and Gillard-Factor, C. (2000) Identification of snake species by toxin mass fingerprinting of their venoms, in Methods in Molecular Biology, vol. 146 (Chapman, J. R., ed.), Humana Press Inc., Totowa, NJ, pp. 1–19. Boulain, J. C., Ducancel, F., Mourier, G., Drevet, P., and Ménez, A. (1999) “Three-fingered” toxins from hydrophid and elapid snakes: artificial procedures to overproduce wildtype and mutated euraremimetic toxins in Animal Toxins, vol. 15 (Rochat, H. and Martin-Eauclaire, M. F., eds.), Birkhäuser Verlag, Basel, pp. 229–245. Ducancel, F., Boulain, J. C., Trémeau, O., and Ménez, A. (1989) Direct expression in E. coli of a functionally active protein A-snake toxin fusion protein. Prot. Eng. 3, 139–143. Missiakas, D. and Raina, S. (1997) Protein folding in the bacterial periplasm. J. Bact. 12, 2465–2471. Nilsson, B., Abrahmsen, L., and Uhlen, M. (1985) Immobilization and purification of enzymes with Staphylococcal protein A gene fusion vectors. EMBO J. 4, 1075–1080. Nilsson, B., Moks, T., Jansson, B., Abrahmsen, L., Emblad, A., Henrichson, C., et al. (1987) A synthetic IgG-binding domain based on Staphylococcal protein A. Prot. Eng. 1, 107–113.

416

Servent and Ménez

27. Trémeau, O., Lemaire, C., Drevet, P., Pinkasfeld, S., Ducancel, F., Boulain, J. C., and Ménez, A. (1995) Genetic engineering of snake toxins. The functional site of Erabutoxin a, as delineated by site-directed mutagenesis, includes variant residues. J. Biol. Chem. 270, 9362–9369. 28. Ackermann, E. J. and Taylor, P. (1997) Nonidentity of the alpha-neurotoxin binding sites on the nicotinic acetylcholine receptor revealed by modification in alpha-neurotoxin and receptor structures. Biochemistry 36, 12,836–12,844. 29. Fiordalisi, J. J., Al-Rabiee, R., Chiappinelli, V. A., and Grant, G. A. (1994) Affinity of native kappa-bungarotoxin and site-directed mutants for the muscle nicotinic acetylcholine receptor. Biochemistry 33, 12,962–12,967. 30. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990) Use of T7 polymerase to direct expression of cloned genes. Methods Enzymol. 185, 60–89. 31. Drevet, P., Lemaire, C., Gasparini, S., ZinnJustin, S., Lajeunesse, E., Ducancel, F., et al. (1997) High-level production and isotope labeling of snake neurotoxins, disulfide-rich proteins. Protein Express. Purif. 10, 293–300. 32. Antil, S., Servent, D., and Ménez, A. (1999) Variability among the sites by which curaremimetic toxins bind to Torpedo acetylcholine receptor, as revealed by identification of the functional residues of α-Cobratoxin. J. Biol. Chem. 274, 34,851–34,858. 33. Rosenthal, J. A., Hsu, S. H., Schneider, D., Gentile, L. N., Messier, N. J., Vaslet, C. A., and Hawrot, E. (1994) Functional expression and site-directed mutagenesis of a synthetic gene for α-bungarotoxin. J. Biol. Chem. 269, 11,178–11,185. 34. Chang, L. S., Chen, K. C., Wu, B. N., Lin, S. K., Wu, P. F., Hong, Y. R., and Yang, C. C. (1999) Expression and mutagenesis studies of cobrotoxin from Taiwan cobra. Biochem. Biophys. Res. Commun. 263, 652–656. 35. Fiordalisi, J. J., James, P. L., Zhang, Y., and Grant, G. A. (1996) Facile production of native-like kappa-bungarotoxin in yeast: an enhanced system for the production of a neuronal nicotinic acetylcholine receptor probe. Toxicon 34, 213–224. 36. Levandoski, M. M., Caffery, P. M., Rogowski, R. S., Lin, Y., Shi, Q. L., and Hawrot, E. (2000) Recombinant expression of alpha-bungarotoxin in Pichia pastoris facilitates identification of mutant toxins engineered to recognize neuronal nicotinic acetylcholine receptors. J. Neurochem. 74, 1279–1289. 37. Mourier, G., Servent, D., Zinn-Justin, S., and Ménez, A. (2000) Chemical engineering of a three-fingered toxin with anti-α7 neuronal acetylcholine receptor activity. Prot. Eng. 13, 217–225. 38. Mourier, G., Zinn-Justin, S., Trémeau, O., and Ménez, A. (1996) Chemical synthesis and in vitro folding of the curarimetic toxin α from Naja nigricollis: a three finger fold protein. Proceedings of the 24th European Peptides Symposium, Edinburgh, 659–660. 39. Vita, C., Roumestand, C., Toma, F., and Ménez, A. (1995) Scorpion toxins as natural scaffolds for protein engineering. Proc. Natl. Acad. Sci. USA 92, 6404–6408. 40. Vita, C., Drakopoulou, E., Vizzavona, J., Rochette, S., Martin, L., Ménez, A., et al. (1999) Rational engineering of a miniprotein that reproduces the core of the CD4 site interacting with HIV-1 envelope glycoprotein. Proc. Natl. Acad. Sci. USA 96, 13,091–13,096. 41. Tamiya, T., Lamouroux, A., Julien, J. F., Grima, B., Mallet, J., and Ménez, A. (1985) Cloning and sequence analysis of the cDNA encoding a snake neurotoxin precursor. Biochimie 67, 185–189. 42. Ducancel, F., Bouchier, C., Tamiya, T., Boulain, J. C., and Ménez, A. (1991) Cloning and expression of cDNAs encoding snake toxins, in Snake Toxins, vol. 10 (Harvey, A. L., ed.), Pergamon Press, New York, pp.385–414. 43. Ducancel, F., Guignery-Frelat, G., Boulain, J.-C., and Ménez, A. (1990) Nucleotide sequence and structure analysis of cDNAs encoding short-chain neurotoxins from venom glands of a sea snake (Aipysurus laevis). Toxicon 28, 119–123.

Snake Neurotoxins

417

44. Tamiya, T., Ohno, S., Nishimura, E., Fujimi, T. J., and Tsuchiya, T. (1999) Complete nucleotide sequences of cDNAs encoding long chain alpha-neurotoxins from sea krait, Laticauda semifasciata. Toxicon 37, 181–185. 45. Danse, J. M. and Garnier, J. M. (1990) cDNA deduced amino-acid sequences of two novel Kappa-neurotoxins from Bungarus multicinctus. Nucleic Acids Res. 18, 4610–4616. 46. Chang, L., Lin, S., Huang, H., and Hsiao, M. (1999) Genetic organization of alphabungarotoxins from Bungarus multicinctus (Taiwan banded krait): evidence showing that the production of alpha-bungarotoxin isotoxins is not derived from edited mRNAs. Nucleic Acids Res. 27, 3970–3975. 47. Lajeunesse, E., Ducancel, F., Gillet, D., Trémeau, O., Drevet, P., Boulain, J. C., et al. (1994) Molecular biology of sea-snake toxins: cloning, expression of cDNAss and analysis of gene organization, in Sea Snake Toxinology (Gopalakrishnakone, P., ed.), Singapore University Press, Singapore, pp. 66–92. 48. Smith, L. A. (1990) Cloning, characterization, and expression of animal toxin genes for vaccine development. J. Toxicol. Toxin Rev. 9, 243–283. 49. Ohno, M., Ménez, R., Ogawa, T., Danse, J. M., Shimohigashi, Y., Fromen, C., et al. (1998) Molecular evolution of snake toxins: Is the functional diversity of snake toxins associated with a mechanism of accelerated evolution. Prog. Nucleic Acid Res. Mol. Biol. 59, 307–364. 50. Contecillo, S. G., Pilpel, Y., Glusman, G., and Fainzilber, M. (2000) Position-specific codon conservation in hypervariable gene family. Trends Genet. 16, 57–59. 51. Fuse, N., Tsuchiya, T., Nonomura, Y., Ménez, A., and Tamiya, T. (1990) Structure of the snake short-chain neurotoxin, erabutoxin c, precursor gene. Eur. J. Biochem. 193, 629–633. 52. Afifiyan, F., Armugam, A., Tan, C. H., Gopalakrishnakone, P., and Jeyaseelan, K. (1999) Postsynaptic alpha-neurotoxin gene of the spitting cobra, Naja naja sputatrix: structure, organization, and phylogenetic analysis. Genome Res. 9, 259–266. 53. Low, B. W., Preston, H. S., Sato, A., Rosen, L. S., Searl, J. E., Rudko, A. D., and Richardson, J. S. (1976) Three-dimensional structure of erabutoxin b neurotoxic protein: inhibitor of acetylcholine receptor. Proc. Natl. Acad. Sci. USA 78, 2991–2994. 54. Tsernoglou, D. and Petsko, G. A. (1976) The crystal structure of a post-synaptic neurotoxin from sea snake at 2.2 Å resolution. FEBS Lett. 68, 1–4. 55. Kimball, M. R., Sato, A., Richardson, J. S., Rosen, L. S., and Low, B. W. (1979) Molecular conformation of erabutoxin b: atomic coordinates at 2.5 Å resolution. Biochem. Biophys. Res. Comm. 88, 950–959. 56. Bourne, P. E., Sato, A., Corfield, P. W. R., Rosen, L. S., Birken, S., and Low, B. W. (1985) Erabutoxin b. Initial protein refinement and sequence analysis at 0.140-nm resolution. Eur. J. Biochem. 153, 521–527. 57. Low, B. W. and Corfield, P. W. R. (1986) Erabutoxin b: Structure/Function relationships following initial protein refinement at 0.140-nm resolution. Eur. J. Biochem. 161, 579–587. 58. Smith, J. L., Corfield, P. W., Hendrickson, W. A., and Low, B. W. (1988) Refinement at 1.4 Å resolution of a model of erabutoxin b: treatment of ordered solvent and discrete disorder. Acta Cryst. A 44, 357–368. 59. Saludjian, P., Prangé, T., Navaza, J., Ménez, R., Guilloteau, J. P., Riès-Kautt, M., and Ducruix, A. (1992) Structure determination of a dimeric form of erabutoxin-b, crystallized from a thiocyanate sodium. Acta Cryst. B48, 520–531. 60. Nastopoulos, V., Kanellopoulos, P. N., and Tsernoglou, D. (1998) Structure of Dimeric and Monomeric erabutoxin a refined at 1.5 Angstrom resolution. Acta Crystallogr. D. Biol. Cryst. 54, 964–974. 61. Preston, H. S., Kay, J., Sato, A., Low, B. W., and Tamiya, N. (1975) Crystalline erabutoxin c. Toxicon 13, 273–275.

418

Servent and Ménez

62. Hatanaka, H., Oka, M., Kohda, D., Tate, S.-I., Suda, A., Tamiya, N., and Inagaki, F. (1994) Tertiary structure of erabutoxin b in aqueous solution as elucidated by two-dimensional nuclear magnetic resonance. J. Mol. Biol. 240, 155–166. 63. Tsernoglou, D. and Petsko, G. A. (1977) Three-dimensional structure of neurotoxin a from venom of the Philippines sea snake. Proc. Natl. Acad. Sci. USA 74, 971–974. 64. Corfield, P. W. R., Lee, T.-J., and Low, B. W. (1989) The crystal structure of erabutoxin a at 2.0-Å resolution. J. Biol. Chem. 264, 9239–9242. 65. Arnoux, B., Ménez, R., Drevet, P., Boulain, J. C., Ducruix, A., and Ménez, A. (1994) Three-dimensional crystal structure of recombinant erabutoxin a at 2.0 Å resolution. FEBS Lett. 342, 12–14. 66. Gaucher, J. F., Ménez, R., Arnoux, B., Pusset, J., and Ducruix, A. (2000) High resolution X-ray analysis of two mutants of a curaremimetic snake toxin. Eur. J. Biochem. 267, 1323–1329. 67. Pillet, L., Trémeau, O., Ducancel, F., Drevet, P., Zinn Justin, S., Pinkasfeld, S., et al. (1993) Genetic engineering of snake toxins. Role of invariant residues in the structural and functional properties of a curaremimetic toxin, as probed by site-directed mutagenesis. J. Biol. Chem. 268, 909–916. 68. Ishikawa, Y., Ménez, A., Hori, H., Yoshida, Y., and Tamiya, N. (1977) Structure of snake toxins and their affinity to the acetylcholine receptor of fish electric organ. Toxicon 15, 477–488. 69. Brown, L. R. and Wüthrich, K. (1992) Nuclear magnetic resonance solution structure of the α-neurotoxin from the black mamba (Dendroaspis polylepis polylepis). J. Mol. Biol. 227, 1118–1135. 70. Zinn Justin, S., Roumestand, C., Gilquin, B., Bontems, F., Ménez, A., and Toma, F. (1992) Three-dimensional solution structure of a curaremimetic toxin from Naja nigricollis venom: a proton NMR and molecular modeling study. Biochemistry 31, 11,335–11,347. 71. Yu, C., Bhaskaran, R., Chuang, L. C., and Yang, C. C. (1993) Solution conformation of cobrotoxin: a nuclear magnetic resonance and hybrid distance geometry-dynamical simulated annealing study. Biochemistry 32, 2131–2136. 72. Golovanov, A. P., Lomize, A. L., Arseniev, A. S., Utkin, Y. N., and Tsetlin, V. I. (1993) Two-dimensional 1H-NMR study of the spatial structure of neurotoxin II from Naja naja oxiana. Eur. J. Biochem. 213, 1213–1223. 73. Ducancel, F., Mérienne, K., Fromen-Romano, C., Trémeau, O., Pillet, L., Drevet, P., et al. (1996) Mimicry between receptors and antibodies. J. Biol. Chem. 271, 31,345–31,353. 74. Chuang, L. C., Yu, H. M., Chen, C., Huang, T. H., Wu, S. H., and Wang, K. T. (1996) Determination of three-dimensional solution structure of waglerin-1, a toxin from Trimeresurus wagleri, using 2D-NMR and molecular dynamics simulation. Biochim. Biophys. Acta. 1292, 145–155. 75. Walkinshaw, M. D., Saenger, W., and Maelicke, A. (1980) Three-dimensional structure of the “long” neurotoxin from cobra venom. Proc. Natl. Acad. Sci. USA 77, 2400–2404. 76. Betzel, C., Lange, G., Pal, G. P., Wilson, K. S., Maelicke, A., and Saenger, W. (1991) The refined crystal structure of α-cobratoxin from Naja naja siamensis at 2.4 Å resolution. J. Biol. Chem. 266, 21,530–21,536. 77. Le Goas, R., Laplante, S. R., Mikou, A., Delsuc, M. A., Guittet, E., Robin, M., et al. (1992) α-cobratoxin: proton NMR assignments and solution structure. Biochemistry 31, 4867–4875. 78. Leroy, E., Mikou, A., Yang, Y., and Guittet, E. (1994) The three-dimensional NMR solution structure of α-cobratoxin at pH 7.5 and conformational differences with the NMR solution structure at pH 3.2. J. Biomol. Struct. Dynamics 12, 1–17. 79. Agard, D. A. and Stroud, R. M. (1982) α-bungarotoxin structure revealed by a rapid method for averaging electron density of non-crystallographically translationally related molecules. Acta Cryst. A38, 186–194.

Snake Neurotoxins

419

80. Love, R. A. and Stroud, R. M. (1986) The crystal structure of α-bungarotoxin at 2.5 Å resolution: relation to solution structure and binding to acetylcholine receptor. Prot. Eng. 1, 37–46. 81. Basus, V. J., Billeter, M., Love, R. A., Stroud, R. M., and Kuntz, I. D. (1988) Structural Studies of α-Bungarotoxin. 1. Sequence-Specific 1H NMR Resonance Assignments. Biochemistry 27, 2763–2771. 82. Basus, V. J., Song, G., and Hawrot, E. (1993) NMR solution structure of an alphabungarotoxin/nicotinic receptor peptide complex. Biochemistry 32, 12,290–12,298. 83. Scherf, T., Balass, M., Fuchs, S., KatchalskiKatzir, E., and Anglister, J. (1997) Threedimensional solution structure of the complex of alpha-bungarotoxin with a libraryderived peptide. Proc. Natl. Acad. Sci. USA 94, 6059–6064. 83a. Zeng, H., Moise, L., Grant M. A. and Hawrot, E. (2001) The solution structure of the complex formed between alpha-bungarotoxin and an 18-mer cognate peptide derived from the alpha 1 subunit of the nicotinic acetylcholine receptor from Torpedo californica. J. Biol. Chem. 276 22930–22940. 83b. Samson, A. O., Chill, J. H., Rodriguez, E., Scherf, T., and Anglister, J. (2001) NMR mapping and secondary structure determination of the major acetylcholine receptor alpha-subunit determinant interacting with alpha-bungarotoxin. Biochemistry 40, 5464–5473. 84. Connolly, P. J., Stern, A. S., and Hoch, J. C. (1996) Solution structure of LSIII, a long neurotoxin from the venom of Laticauda semifasciata. Biochemistry 35, 418–426. 85. Nickitenko, A. V., Michailov, A. M., Betzel, C., and Wilson, K. S. (1993) Three-dimensional structure of neurotoxin-1 from Naja naja oxiana venom at 1.9 Å resolution. FEBS Lett. 320, 111–117. 86. Peng, S. S., Kumar, T. K. S., Jayaraman, G., Chang, C. C., and Yu, C. (1997) Solution structure of toxin b, a long neurotoxin from the venom of the king cobra (Ophiophagus hannah). J. Biol. Chem. 272, 7817–7823. 87. Oswald, R. E., Sutcliffe, M. J., Bamberger, M., Loring, R., Braswell, E., and Dobson, C. M. (1991) Solution structure of neuronal bungarotoxin determined by two-dimensional NMR spectroscopy: sequence-specific asignments, secondary structure and dimer formation. Biochemistry 30, 4901–4909. 88. Sutcliffe, M. J., Dobson, C. M., and Oswald, R. E. (1992) Solution structure of neuronal bungarotoxin determined by two-dimensional NMR spectroscopy: calculation of tertiary structure using systematic homologous model building, dynamical simulated annealing and restrained molecular dynamics. Biochemistry 31, 2962–2970. 89. Dewan, J. C., Grant, G. A., and Sacchettini, J. C. (1994) Crystal structure of κbungarotoxin at 2.3 Å resolution. Biochemistry 33, 13,147–13,154. 90. Grant, G. A., Al-Rabiee, R., Xu, X. L., and Zhang, Y. P. (1997) Critical interactions at the dimer interface of kappa-bungarotoxin, a neuronal nicotinic acetylcholine receptor antagonist. Biochemistry 36, 3353–3358. 90a. Kuhn, P., Deacon, A. M., Comoso, S., Rajaseger, G., Kini, R. M., Uson, I., and Kolatkar, P. R. (2000) The atomic resolution structure of bucandin, a novel toxin isolated from the Malayan krait, determined by direct methods. Acta Cryst. D56, 1401–1407. 90b. Antil-Delbeke, C., Gaillard, T., Tamiya, P. J., Carninger, J. P., Changeux, D., Servent, and Ménez, A. (2000) Molecular determinants by which a long chain toxin from snake venom interacts with the neuronal α7 nicotinic acetylcholine receptor. J. Biol. Chem. 275, 29,594–29,601. 91. Ricciardi, A., Le Du, M. H., Khayati, M., Dajas, F., Boulain, J. C., Ménez, A., and Ducancel, F. (2000) Do structural deviations between toxins adopting the same fold reflect functional differences? J. Biol. Chem. 275, 18,302–18,310. 92. Le Du, M. H., Ricciardi, A., Khayati, M., Ménez, R., Boulain, J. C., Ménez, A., and Ducancel, F. (2000) Stability of a structural scaffold upon activity transfer. X-Ray structure of a three fingers chimeric protein. J. Mol. Biol. 296, 1017–1026.

420

Servent and Ménez

93. Le Du, M. H. (1992) 1.9 Å resolution structure of fasciculin 1, an anti-acetylcholinesterase toxin from green mamba snake venom. J. Biol. Chem. 267, 22,122–22,130. 94. Ménez, A., Bouet, F., Guschlbauer, W., and Fromageot, P. (1980) Refolding of reduced short neurotoxins: circular dichroism analysis. Biochemistry 19, 4166–4172. 95. Ruoppolo, M., Moutiez, M., Mazzeo, M. F., Pucci, P., Ménez, A., Marino, G., and Quémeneur, E. (1998) The length of a single turn controls the overall folding rate of “three-fingered” snake toxins. Biochemistry 37, 16,060–16,068. 96. Lentz, T. L. and Wilson, P. T. (1988) Neurotoxin binding site on the acetylcholine receptor. Int. Rev. Neurobiol. 29, 117–159. 97. Conti-Fine, B. M., Maelicke, A., Reinhardt-Maelicke, S., Chiapinelli, V., and McLane, K. E. (1995) Binding sites for neurotoxins and cholinergic ligands in peripheral and neuronal receptors. Ann. NY Acad. Sci. 757, 133–152. 98. McLane, K. E., Dunn, S. J. M., Manfredi, A. A., Conti-Tronconi, B. M., and Raftery, M. A. (1996) The nicotinic acetylcholine receptor as a model for a superfamily of ligandgated ion channel proteins. Prot. Eng. Design 10, 289–352. 99. Boulain, J. C., Ménez, A., Couderc, J., Faure, G., Liacopoulos, P., and Fromageot, P. (1982) Neutralizing monoclonal antibody specific for Naja nigricollis toxin α: preparation, characterization and localization of the antigenic binding site. Biochemistry 21, 2910–2915. 100. Trémeau, O., Boulain, J. C., Couderc, J., Fromageot, P., and Ménez, A. (1986) A monoclonal antibody which recognized the functional site of snake neurotoxins and which neutralizes all short-chain variants. FEBS Lett. 208, 236–240. 101. Zinn Justin, S., Roumestand, C., Drevet, P., Ménez, A., and Toma, F. (1993) Mapping of two “neutralizing” epitopes of a snake curaremimetic toxin by proton nuclear magnetic resonance spectroscopy. Biochemistry 32, 6884–6891. 102. Lo Conte, L., Chotia, C., and Janin, J. (1999) The atomic structure of protein-protein recognition sites. J. Mol. Biol. 285, 2177–2198. 103. Fromen-Romano, C., Maillère, B., Drevet, P., Lajeunesse, E., Ducancel, F., Boulain, J. C., and Ménez, A. (1997) Transformation of a non-enzymatic toxin into a toxoid by genetic engineering. Prot. Eng. 10, 1213–1220. 104. Mérienne, K., Germain, N., Zinn-Justin, S., Boulain, J. C., Ducancel, F., and Ménez, A. (1997) The functional architecture of an acetylcholine receptor-mimicking antibody. J. Biol. Chem. 272, 23,775–23,783. 105. Tenette, C., Ducancel, F., and Smith, J. (1996) Structural model of the anti-snake toxin antibody Mα2-3. Prot. Struct. Funct. Genet. 26, 9–31. 106. Tenette-Souaille, C. and Smith, J. (1998) Structural modeling of the complex between an acetylcholine receptor-mimicking antibody and its snake toxin antigen. Prot. Struct. Funct. Genet. 30, 249–263. 107. Germain, N., Mérienne, K., Zinn-Justin, S., Boulain, J. C., Ducancel, F., and Ménez, A. (2000) Molecular and structural basis of the specificity of a neutralizing acetylcholine receptor-mimicking antibody, using combined mutational and molecular modelling analyses. J. Biol. Chem. 275, 21,578–21,586. 108. Amzel, L. M. and Poljak, R. J. (1979) Three-dimensional structure of immunoglobulins. Annu. Rev. Biochem. 48, 961–997. 109. Pruett, P. S. and Air, G. M. (1998) Critical interactions in binding antibody NC41 to influenza N9 neuraminidase: amino acid contacts on the antibody heavy chain. Biochemistry 37, 10,660–10,670. 110. Davies, D. R. and Cohen, G. H. (1996) Interactions of protein antigens with antibodies. Proc. Natl. Acad. Sci. USA 93, 7–12. 111. Guenneuges, M., Drevet, P., Pinkasfeld, S., Gilquin, B., Ménez, A., and Zinn-Justin, S. (1997) Picosecond to hour time scale dynamics of a “three finger” toxin: correlation with its toxic and antigenic properties. Biochemistry 36, 16,097–16,108.

Snake Neurotoxins

421

112. Devillers-Thiéry, A., Galzi, J. L., Eiselé, J. L., Bertrand, S., Bertrand, D., and Changeux, J. P. (1993) Functional architecture of the nicotinic acetylcholine receptor: a prototype of ligand-gated ion channels. J. Membr. Biol. 136, 97–112. 113. Galzi, J. L. and Changeux, J. P. (1992) The nicotinic acetylcholine receptor: a model of ligand-gated ion channels, in Membrane Proteins: Structures, Interactions and Models (Pullman, A., ed.), Kluwer Academic, The Netherlands, pp. 127–146. 114. Lindstrom, J. (1999) Purification and cloning of nicotinic acetylcholine receptors, in Neuronal Nicotinic Receptors. Pharmacology and Therapeutic Opportunities (Arneric, S. P. and Brioni, J. D., eds.), Wiley-Liss, New York. 115. Unwin, N. (1993) Nicotinic acetylcholine receptor at 9 Å resolution. J Mol. Biol. 229, 1101–1124. 116. Miyazawa, A., Fujiyoshi, Y., Stowell, M., and Unwin, N. (1999) Nicotinic acetylcholine receptor at 4.6 Angstrom resolution: transverse tunnels in the channel wall. J. Mol. Biol. 288, 765–786. 116a. Brejc, K., van Dijk, W. J., Klaassen, R. V., Schuurmans, M., van der Oost, J., Smit, A. B. and Sixma, T. K. (2001) Crystal structure of an ACh-binding protein reveals the ligandbinding domain of nicotinic receptors. Nature 411, 269–278. 116b. Smit, A. B., Syed, N. I., Schaap, D., van Minnen, J., Klumperman, J., Kits, K. S., et al. (2001) A glia-derived acteylcholine binding protein that modulates synaptic transmission. Nature 411, 261–268. 117. Weber, M. and Changeux, J. P. (1974) Binding of Naja nigricollis [3H] α-toxin to membrane fragments from Electrophorus and Torpedo electric organs. Mol. Pharmacol. 10, 15–34. 118. Martin, B. M., Chibber, B. A., and Maelicke, A. (1983) The sites of neurotoxicity in αcobratoxin. J. Biol. Chem. 258, 8714–8722. 119. Lin, S., R, and Chang, C., C. (1991) Studies on the status of amino groups in αbungarotoxin. Toxicon 29, 937–950. 120. Faure, G., Boulain, J. C., Bouet, F., Montenay-Garestier, T., Fromageot, P., and Ménez, A. (1983) Role of indole and amino groups in the structure and function of Naja nigricollis toxin α. Biochemistry 22, 2068–2076. 121. Chicheportiche, R., Vincent, J. P., Kopeyan, C., Schweitz, H., and Lazdunski, M. (1975) Structure-function relationship in the binding of snake neurotoxins to the Torpedo membrane receptor. Biochemistry 14, 2081–2091. 122. Hervé, M., Pillet, L., Humbert, P., Tremeau, O., Ducancel, F., Hirth, C., and Ménez, A. (1992) Role and environment of the conserved Lys27 of snake curaremimetic toxins as probed by chemical modifications, site-directed mutagenesis and photolabelling experiments. Eur. J. Biochem. 208, 125–131. 123. Chang, C.-C., Kawata, Y., Sakayama, F., and Hayashi, K. (1990) The role of an invariant tryptophan residue in α-bungarotoxin and cobrotoxin. Eur. J. Biochem. 193, 567–572. 124. Endo, T., Oya, M., Tamiya, N., and Hayashi, K. (1987) Role of C-terminal tail of long neurotoxin from snake venoms in molecular conformation and acetylcholine receptor binding: Protein nuclear magnetic resonance and competition binding studies. Biochemistry 26, 4592–4598. 125. Boyot, P., Pillet, L., Ducancel, F., and Ménez, A. (1990) A recombinant snake neurotoxin generated by a chemical cleavage of a hybrid protein recovers full biological properties. FEBS Let. 266, 87–90. 126. Malany, S., Ackermann, E., Osaka, H., and Taylor, P. (1998) Complementary binding studies between α-neurotoxin and the nicotinic acetylcholine receptor. J. Physiol. (Paris) 92, 462-463. 127. Osaka, H., Malany, S., Kanter, J. R., Sine, S. M., and Taylor, P. (1999) Subunit interface selectivity of the alpha-neurotoxins for the nicotinic acetylcholine receptor. J. Biol. Chem. 274, 9581–9586.

422

Servent and Ménez

128. Rosenthal, J. A., Levandoski, M. M., Chang, B., Potts, J. F., Shi, Q.-L., and Hawrot, E. (1999) The functional role of positively charged amino acid side chains in α-bungarotoxin revealed by site-directed mutagenesis of a His-tagged recombinant α-bungarotoxin. Biochemistry 38, 7847–7855. 129. Fiordalisi, J. J., Al-Rabiee, R., Chiapinelli, V. A., and Grant, G. A. (1994) Site-directed mutagenesis of κ-bungarotoxin: implications for neuronal receptor specificity. Biochemistry 33, 3872–3877. 130. Grant, G. A., Luetje, C. W., Summers, R., and Xu, X. L. (1998) Differential roles for disulfide bonds in the structural integrity and biological activity of kappa-bungarotoxin, a neuronal nicotinic acetylcholine receptor antagonist. Biochemistry 37, 12,166–12,171. 131. McArdle, J. J., Lentz, T. L., Witzemann, V., Schwarz, H., Weinstein, S. A., and Schmidt, J. J. (1999) Waglerin-1 selectively blocks the epsilon form of the muscle nicotinic acetylcholine receptor. J. Pharm. Exp. Ther. 289, 543–550. 132. Molles, B. E., Kline, E. F., Sine, S. M., McArdle, J. J., and Taylor, P. (1998) Probing the structure of the ligand binding site on the muscle nicotinic receptor with Waglerin peptides. J. Physiol. (Paris) 92, 470–471. 133. Hsio, Y. M., Chuang, C. C., Chuang, L. C., Yu, H. M., Wang, K. T., Chiou, S. H., and Wu, S. H. (1996) protein engineering of venom toxins by synthetic approach and NMR dynamic simulation: status of basic amino acid residues in waglerin I. Biochem. Biophys. Res. Commun. 227, 59–63. 134. Kao, P. N. and Karlin, A. (1986) Acetylcholine receptor binding site contains a disulfide cross-link between adjacent half-cystinyl residues. J. Biol. Chem. 261, 8085–8088. 135. Dennis, M., Giraudat, J., Kotzyba-Hibert, F., Goeldner, M., Hirth, C., Chang, J. Y., et al. (1988) Amino acids of the Torpedo marmorata acetylcholine receptor labeled by a photoaffinity ligand for the acetylcholine binding site. Biochemistry 27, 2346–2357. 136. Galzi, J. L., Revah, F., Black, D., Goeldner, M., Hirth, C., and Changeux, J. P. (1990) Identification of a novel amino acid alpha-Tyr93 within the active site of the acetylcholine receptor by photoaffinity labeling: additional evidence for a three loop model of the acetylcholine binding site. J. Biol. Chem. 265, 10,430–10,437. 137. Middleton, R. E. and Cohen, J. B. (1991) Mapping of the acetylcholine binding site of the acetylcholine receptor: [3H]nicotine as an agonist photoaffinity label. Biochemistry 30, 6987–6997. 138. Chiara, D. C. and Cohen, J. B. (1997) Identification of amino acids contributing to high and low affinity d-tubocurarine sites in the Torpedo nicotinic acetylcholine receptor. J. Biol. Chem. 272, 32,940–32,950. 139. Abramson, S. N., Culver, P., Klines, T., Li, Y., P., G., Gutman, L., and Taylor, P. (1988) Lophotoxin and related coral toxins covalently label tha α-subunit of the nicotinic acetylcholine receptor. J. Biol. Chem. 263, 18,568–18,573. 140. Galzi, J. L., Revah, F., Bessis, A., and Changeux, J. P. (1991) Functional architecture of the nicotinic acetylcholine receptor: From electric organ to brain. Annu. Rev. Pharmacol. Toxicol. 31, 37–72. 141. Galzi, J. L. and Changeux, J. P. (1994) Neurotransmitter-gated ion channels as unconventional allosteric proteins. Curr. Opin. Struct. Biol. 4, 554–565. 142. Chiara, D. C., Middleton, R. E., and Cohen, J. B. (1998) Identification of tryptophan 55 as the primary site of [H-3]nicotine photoincorporation in the gamma-subunit of the Torpedo nicotinic acetylcholine receptor. FEBS Lett. 423, 223–226. 143. Tomaselli, G. F., McLaughlin, J. T., Jurman, M. E., Hawrot, E.. and Yellen, G. (1991) Mutations affecting agonist sensitivity of the nicotinic acetylcholine receptor. Biophys. J. 60, 721–727. 144. Sine, S. M., Quiram, P., Papanikolaou, F., Kreienkamp, H. J., and Taylor, P. (1994) Conserved tyrosines in the alpha subunit of the nicotinic acetylcholine receptor stabilize qua-

Snake Neurotoxins

145. 146.

147.

148.

149.

150. 151.

152.

153.

154. 155.

156. 157. 158.

159.

160.

161.

423

ternary ammonium groups of agonists and curariform antagonists. J. Biol. Chem. 269, 8808–8816. O’Leary, M. E. and White, M. M. (1992) Mutational analysis of ligand-induced activation of the Torpedo acetylcholine receptor. J. Biol. Chem. 267, 8360–8365. Aylwin, M. L. and White, M. M. (1994) Ligand-receptor interactions in the nicotinic acetylcholine receptor probed using multiple substitutions at conserved tyrosines on the alpha subunit. FEBS Lett. 349, 99–103. Galzi, J. L. (1991) Functional significance of aromatic amino acids from three peptide loops of the α7 neuronal nicotinic receptor site investigated by site-directed mutagenesis. FEBS Leters 294, 198-202. Corringer, P. J., Bertrand, S., Bohler, S., Edelstein, S. J., Changeux, J. P., and Bertrand, D. (1998) Critical elements determining diversity in agonist binding and desensitization of neuronal nicotinic acetylcholine receptors. J. Neurosci. 18, 648–657. Czajkowski, C. and Karlin, A. (1995) Structure of the nicotinic receptor acetylcholinebinding site. Identification of acidic residues in the delta subunit within 0.9 nm of the 5 alpha subunit-binding. J. Biol. Chem. 270, 3160–3164. Prince, R. J. and Sine, S. M. (1996) Molecular dissection of subunit interfaces in the acetylcholine receptor. J. Biol. Chem. 271, 25,770–25,777. Sine, S. M., Kreienkamp, H. J., Bren, N., Maeda, R., and Taylor, P. (1995) Molecular dissection of subunit interfaces in the acetylcholine receptor: identification of determinants of alpha-conotoxin M1 selectivity. Neuron 15, 205–211. Chiara, D. C., Xie, Y., and Cohen, J. B. (1999) Structure of the agonist-binding sites of the Torpedo nicotinic acetylcholine receptor: Affinity-labeling and mutational analyses identify gamma Tyr-111/delta Arg-113 as antagonist affinity determinants. Biochemistry 38, 6689–6698. Tsigelny, I., Sugiyama, N., Sine, S. M., and Taylor, P. (1997) A model of the nicotinic receptor extracellular domain based on sequence identity and residue location. Biophys. J. 73, 52–66. Holtzman, E., Wise, D., Wall, J., and Karlin, A. (1982) Electron microscopy of complexes of isolated acetylcholine receptor, biotynyl-toxin and avidin. PNAS 79, 310–314. Zingsheim, H. P., Barrantes, F. J., Frank, J., Hänike, W., and Neugebauer, D. C. (1982) Direct structural localization of two toxin-recognition sites on ACh receptor protein. Nature 299, 81–84. Kistler, J. and Stroud, R. M. (1982) Structure and function of an acetylcholine receptor. Biophys. J. 37, 371–383. Johnson, D. A., Cushman, R., and Malekzadeh, R. (1990) Orientation of cobra α-toxin on the nicotinic acetylcholine receptor. J. Biol. Chem. 265, 7360–7368. Kreienkamp, H. J., Utkin, Y. N., Weise, C., Machold, J., Tsetlin, V. I., and Hucho, F. (1992) Investigation of ligand-binding sites of the acetylcholine receptor using photoactivatable derivatives of neurotoxin II from Naja naja oxiana. Biochemistry 31, 8239–8244. Utkin, Y. N., Krivoshein, A. V., Davydov, V. L., Kasheverov, I. E., Franke, P., Maslennikov, I. V., et al. (1998) Labeling of Torpedo californica nicotinic acetylcholine receptor subunits by cobratoxin derivatives with photoactivatable groups of different chemical nature at Lys23. Eur. J. Biochem. 253, 229–235. Utkin, Y. N., Hatanaka, Y., Franke, P., Machold, J., Hucho, F., and Tsetlin, V. I. (1995) Synthesis of nitrodiazirinyl derivatives of neurotoxin II from Naja naja oxiana and their interaction with the Torpedo californica nicotinic acetylcholine receptor. J. Prot. Chem. 14, 197–203. Machold, J., Weise, C., Utkin, Y. N., Franke, P., Tsetlin, V. I., and Hucho, F. (1995) A new class of photoactivatable and cleavable derivatives of neurotoxin II from Naja naja

424

162.

163.

164. 165. 166.

167.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

Servent and Ménez oxiana. Synthesis, characterisation, and application for affinity labelling of the nicotinic acetylcholine receptor from Torpedo californica. Eur. J. Biochem. 228, 947–954. Chatrenet, B., Kotzba Hibert, F., Mulle, C., Changeux, J. P., Goeldner, M. P., and Hirth, C. (1992) Photoactivatable agonist of the nicotinic acetylcholine receptor: potential probe to characterize the structural transitions of the acetylcholine binding site in different states of the receptor. Mol. Pharmacol. 41, 1100–1106. Machold, J., Utkin, Y., Kirsch, D., Kaufmann, R., Tsetlin, V., and Hucho, F. (1995) Photolabeling reveals the proximity of the alpha-neurotoxin binding site to the M2 helix of the ion channel in the nicotinic acetylcholine receptor. Proc. Natl. Acad. Sci. USA 92, 7282–7286. Hucho, F., Tsetlin, V. I., and Machold, J. (1996) The emerging three-dimensional structure of a receptor: the nicotinic acetylcholine receptor. Eur. J. Biochem. 239, 539–557. Kessler, P., Maurin, S., and Ménez, A. (1998) nAChR and α-neurotoxins: new tools for old acquaintances. J. Physiol. (Paris) 92, 447. Michalet, S., Teixeira, F., Gilquin, B., Mourier, G., Servent, D., Drevet, P., et al. (2000) Relative spatial position of a snake neurotoxin and the reduced disulfide bond α(Cys192cys193) at the αγ interface of the nicotinic acetylcholine receptor. J. Biol. Chem. 275, 25,608–25,615. Barchan, D., Kachalsky, S., Neumann, D., Vogel, Z., Ovadia, M., Kochva, E. and Fuchs, S. (1992) How the mongoose can fight the snake: the binding site of the mongoose acetylcholine receptor. Proc. Natl. Acad. Sci. USA 89, 7717–7721. Chaturvedi, V., Donnelly-Roberts, D. L., and Lentz, T. L. (1992) Substitutions of Torpedo acetylcholine receptor α1 subunit residues with snake α1 and rat nerve α3 subunit residues in recombinant fusion proteins: effect on α-bungarotoxin binding. Biochemistry 31, 1370–1375. Keller, S. H., Kreienkamp, H. J., Kawanishi, C., and Taylor, P. (1995) Molecular determinants conferring alpha-toxin resistance in recombinant DNA-derived acetylcholine receptors. J. Biol. Chem. 270, 4165–4171. Kreienkamp, H. J., Sine, S. M., Maeda, R. K., and Taylor, P. (1994) Glycosylation sites selectively interfere with alpha-toxin binding to the nicotinic acetylcholine receptor. J. Biol. Chem. 269, 8108–8114. Ackermann, E. J., Ang, E. T. H., Kanter, J. R., Tsigelny, I., and Taylor, P. (1998) Identification of pairwise interactions in the alpha-neurotoxin-nicotinic acetylcholine receptor complex through double mutant cycles. J. Biol. Chem. 273, 10,958–10,964. Osaka, H., Malany, S., Molles, B. E., Sine, S. M., and Taylor, P. (2000) Pairwise electrostatic interactions between α-neurotoxins and γ, δ, and ε subunits of the nicotinic acetylcholine receptor. J. Biol. Chem. 275, 5478–5484. Sine, S. M. (1997) Identification of equivalent residues in the gamma, delta, and epsilon subunits of the nicotinic receptor that contribute to alpha-bungarotoxin binding. J. Biol. Chem. 272, 23,521–23,527. Spura, A., Russin, T. S., Freedman, N. D., Grant, M., McLaughlin, J. Y., and Hawrot, E. (1999) Probing the agonist domain of the nicotinic acetylcholine receptor by cysteine scanning mutagenesis reveals residues in proximity to the α-bungarotoxin binding site. Biochemistry 38, 4912–4921. Levandoski, M. M., Lin, Y. X., Moise, L., McLaughlin, J. T., Cooper, E., and Hawrot, E. (1999) Chimeric analysis of a neuronal nicotinic acetylcholine receptor reveals amino acids conferring sensitivity to alpha-bungarotoxin. J. Biol. Chem. 274, 26,113–26,119. Taylor, P., Osaka, H., Molles, B. E., Sugiyama, N., Marchot, P., Ackermann, et al. (1998) Toxins selective for subunit interfaces as probes of nicotinic acetylcholine receptor structure. J. Physiol. Paris 92, 79–83. Luetje, C. W., Piattoni, M., and Patrick, J. (1993) Mapping of ligand binding sites of

Snake Neurotoxins

178.

179. 180.

181.

182.

183.

184.

185.

186.

425

neuronal nicotinic acetylcholine receptors using chimeric alpha subunits. Mol. Pharmacol. 44, 657–666. Luetje, C. W., Maddox, F. N., and Harvey, S. C. (1998) Glycosylation within the cysteine loop and six residues near conserved Cys192/Cys193 are determinants of neuronal bungarotoxin sensitivity on the neuronal nicotinic receptor alpha 3 subunit. Mol. Pharmacol. 53, 1112–1119. Harvey, S. C. and Luetje, C. W. (1996) Determinants of competitive antagonist sensitivity on neuronal nicotinic receptor beta subunits. J. Neurosci. 16, 3798–3806. Le Novere, N., Corringer, P. J., and Changeux, J. P. (1999) Improved secondary structure predictions for a nicotinic receptor subunit: incorporation of solvent accessibility and experimental data into a two-dimensional representation. Biophys. J. 76, 2329–2345. Chiappinelli, V. A., Weaver, W. R., McLane, K. E., ContiFine, B. M., Fiordalisi, J. J., and Grant, G. A. (1996) Binding of native kappa-neurotoxins and site-directed mutants to nicotinic acetylcholine receptors. Toxicon 34, 1243–1256. Gorman, S., Viseshakul, N., Cohen, B., Hardy, S., Grant, G. A., Yost, C. S., and Forsayeth, J. R. (1997) A recombinant adenovirus that directs secretion of biologically active k-bungarotoxin from mammalian cells. Mol. Brain Res. 44, 143–146. Sugiyama, N., Marchot, P., Kawanishi, C., Osaka, H., Molles, B., Sine, S. M., and Taylor, P. (1998) Residues at the subunit interfaces of the nicotinic acetylcholine receptor that contribute to alpha-conotoxin M1 binding. Mol. Pharmacol. 53, 787–794. Quiram, P. A. and Sine, S. M. (1998) Identification of residues in the neuronal alpha(7) acetylcholine receptor that confer selectivity for conotoxin ImI. J. Biol. Chem. 273, 11,001–11,006. Quiram, P. A., Jones, J. J., and Sine, S. M. (1999) Pairwise interactions between neuronal alpha(7) acetylcholine receptors and alpha-conotoxin ImI. J. Biol. Chem. 274, 19,51719,524. Malany, S., Osaka, H., Sine, S. M. and Taylor, P. (2000) Orientation of α-neurotoxin at the subunit interfaces of the nicotinic acetylcholine receptor. Biochemistry 39, 15,388– 15,398.

426

Servent and Ménez

A2 Neurotoxins

427

21 Presynaptic Phospholipase A2 Neurotoxins from Snake Venoms John B. Harris

1. INTRODUCTION Neurotoxicity has been defined as “a structural change or a functionally adverse response of the nervous system to a chemical, biological or physical agent” (1). The nervous system is particularly susceptible to neurotoxic chemical, biological, or physical agents for a number of reasons: neurons cannot divide because they are postmitotic and so the death of a neuron is an irreversible event; metabolic activity is very high in all parts of the nervous system; the nervous system develops relatively slowly in the higher vertebrates, and adverse events during development can have significant longterm behavioral, functional, and physiological consequences (1,2). The effects of a given neurotoxic agent (neurotoxin) will depend on its access to the various parts of the nervous system and on its subcellular target. Those targets will be determined by the properties of the toxin. If the neurotoxin is taken up by nerve terminals, its target may be intracellular and may also be remote from its point of entry into the nervous system; widespread entry into various parts of the nervous system may be facilitated if vascular endothelium is damaged; neurotoxins may specifically target particular ion channels, receptors, or uptake processes; changes in membrane structure may secondarily influence the behavior of a large range of membrane associated ionchannels, receptors, and enzymes. For example, clostridial toxins are endocytosed by motor-nerve terminals in the peripheral motor system. Botulinum toxins act at the point of entry by the proteolytic inactivation of synaptic proteins involved in transmitter release; tetanus toxins are transferred to the spinal cord where they act at the Renshaw cell to reduce the release of the inhibitory transmitter glycine. Domoic acid (the toxin responsible for amnesic shellfish poisoning) enters the central nervous system (CNS) and causes the long-term excitation of glutamate receptors. Most parts of the nervous system are protected from exogenous toxins. The bloodbrain barrier (BBB) protects most parts of the CNS and the blood-axon barrier protects the peripheral nervous system (PNS). Unprotected areas include the choroid plexus, the neurohypophysis, area postrema, the pineal gland, and locus caeruleus in the CNS and the neuromuscular junction, dorsal-root ganglia and autonomic ganglia in the PNS. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

427

428

Harris

Broadly, these regions act as potential sites of entry into the wider nervous system, but unless specific uptake processes exist, polar compounds, like charged polypeptides, are unlikely to enter because they are not lipid-soluble. 1.1. Neurotoxic Symptoms Following Snake Bite The accurate reporting of the symptoms of snake bite is not straightforward. Most victims of snake bite, whether by venomous or nonvenomous snakes, report first to local or traditional healers. This is a simple reflection of the fact that most snake bites occur in rural Africa, South and Central America, and Southeast Asia—regions where access to medical care is difficult, costly, and often unhelpful. By the time a victim of an envenoming bite is referred to a physician, the direct consequences of the bite may have been complicated by the treatments of the local healer which, typically, include multiple incisions at the bite site, the application of overtight tourniquets, and by lengthy lapses of time between the administration of local treatment and the seeking of qualified medical advice. The clinical examination should look for evidence that the condition of the patient cannot be adequately explained by natural events or causes, or by any interventions made by a third party and there should be reasonable physical and/or anecdotal evidence that a bite has been inflicted. It is rarely possible to rely on the patient’s identification of the offending animal. Confusion is common when numerous biting and stinging animals are present in a given location and local knowledge of dangerous fauna is invariably limited (3). Fear, anxiety, mental confusion, and incoherence are not definitive signs of neurotoxicity (see ref. 4, for example) and neither is pain (though this is undoubtedly a neurological problem). The definitive signs of neurotoxicity are those associated with neuromuscular weaknesses: ptosis, exophthalmoplegia, low grip strength, inability to protrude the tongue, difficulties in speaking and swallowing, and an inability to open or close fully the mouth. In severe cases of neurotoxic envenoming, respiration is grossly (and sometimes fatally) impaired. Central manifestations of neurotoxic damage are rarely acute. The best known central problems associated with snake bite are the secondary result of hemorrhage into the pituitary (5) or into the CNS more generally (6). Even those central structures unprotected by the BBB are not known targets for the various toxic components of snake venoms. It is reasonable to generalize that the neurotoxicity associated with envenoming snake bites is peripheral rather than central, and that neurotoxic components of snake venoms attack peripheral targets to cause neuromuscular weakness. 1.2. Neurotoxins and the PNS The potential causes of significant toxin-induced neuromuscular weakness are numerous and are summarized in Fig. 1. It is not always easy to determine precisely the target of a neurotoxin without reference to multidisciplinary studies. In the human victim of a “neurotoxic” envenoming bite, this might include acute and longer-term studies of electrophysiological function of PNS and CNS. Neuropsychometric testing for potential changes in cognition, memory and mood, and biopsy (especially of the nervemuscle junction) for the study of muscle fibers and synaptic structures would be of great value. Clearly, studies such as these require an array of skills rarely available and they have not, with one or two notable exceptions, been made. Most current understanding of the biological basis of neurotoxic snake bite in humans has been made on

A2 Neurotoxins

429

Fig. 1. Possible causes of neuromuscular paralysis.

the basis of the clinical examination, intuition, and subjective interpretation rather than on objective evidence. To a large degree this process depends on a good knowledge of the toxic constituents of the venoms of the relevant dangerous snakes and of the pharmacological properties of those venoms and toxins. 1.3. Neurotoxins in Snake Venoms Neurotoxins, defined as toxins whose activity primarily affects the CNS or either of the two major branches of the PNS, the somatic PNS and the autonomic PNS, respectively, can be isolated from the venoms of all elapid snakes; from the venoms of their close relatives, the sea snakes; and from the venoms of some viperid, crotalid and colubrid snakes. The toxins fall into a number of specific classes: Postsynaptically active neurotoxins. a. Toxins binding selectively to nicotinic acetylcholine receptors (AChR). b. Toxins binding selectively to muscarinic AChR. c. Fasciculins.

Presynaptically active toxins. a. Dendrotoxins. b. Neurotoxic phospholipases A2.

Myotoxic phospholipases A2. The inclusion of myotoxic phospholipases A2 in a list of neurotoxic constituents of snake venoms may appear odd, but is defended on the grounds that many toxic phospholipases A2 are both neurotoxic and myotoxic and that the clinician interpreting the signs and symptoms expressed by a victim of an envenoming snake bite must be able to identify and differentiate between neuromuscular weakness caused by transmission failure and that caused by rhabdomyolysis. 1.3.1. Postsynaptically Active Neurotoxins 1.3.1.1. TOXINS BINDING SELECTIVELY TO NICOTINIC AChR The majority of toxins binding to nicotinic AChR bind selectively to peripheral receptors at the neuromuscular junction. These toxins are conveniently known as α-neurotoxins. A small class of toxins that bind to nicotinic AChR mainly in the CNS

430

Harris

has been isolated from the venoms of kraits (Bungarus species). These are identified as κ-bungarotoxins. The α-toxins have been extremely well-characterized and details of their origin, their molecular and functional biochemistry, and their molecular and cellular pharmacology are described in depth by Servent and Ménez in Chapter 20. Envenoming bites by snakes whose venoms are rich in α-neurotoxins cause a neuromuscular weakness with very rapid onset and of potentially fatal depth. Reversal of paralysis, even in severely envenomed victims, occurs approx 24 h after the bite, provided respiration is maintained, and recovery can be accelerated by the administration of antivenom or anticholinesterase (7). The venoms of all elapid and hydrophiid snakes contain high levels of such toxins but the venoms of viperid and crotalid snakes are devoid of such toxins. Non-elapid land snakes of the genus Boiga may elaborate a venom that contains toxins capable of binding to nicotinic AChR (8) but clinically important bites by these snakes are rare. The postsynaptic toxins that bind to nicotinic AChR cause no structural damage to any part of the nervous system or its target cells and any long-term sequelae to the bite must be considered due to the presence of other toxic fractions in the crude venom. 1.3.1.2. TOXINS BINDING TO MUSCARINIC AChR This group of toxins comprises a number of closely related polypeptides isolated from the venom of the mamba Dendroaspis augusticeps (9). They are small polypeptides of 64–66 amino acid residues crosslinked by four disulphide bonds. The toxins have become valuable tools in the study of muscarinic receptor subtypes in the mammalian body. The widespread distribution of muscarinic receptors and the involvement of ACh and mAChR in the cognitive functions of the CNS suggest that these potentially important neurotoxins might play a significant role in the syndrome that characterizes envenoming bites by mambas. Interest is growing in the effects of envenoming by snakes on autonomic function (10). 1.3.1.3. FASCICULINS Fasciculins comprise a small group of toxins isolated from the venoms of Dendroaspis species that block AChE activity. Anticholinesterase activity is not uncommon in snake venoms, but although the fasciculins can induce tremors and fasciculation in animals, they probably contribute little to clinical signs of neurotoxicity following bites by relevant snakes. For a detailed review of the structure and function of fasciculins, see ref. 11. 1.3.2. Presynaptically Active Toxins 1.3.2.1. DENDROTOXINS Dendrotoxins are small polypeptides of 57–60 amino acid residues attached in a single chain and cross-linked by three disulphide bonds. They are found exclusively in the venoms of mambas (Dendroaspis sp.). They block a number of voltage-dependent Κ+ channels in nerve terminals (12). As a result they cause the repetitive firing of action potentials and the enhancement of evoked transmitter release at the motor nerve terminal. The toxins are used extensively as biological tools but they are not particularly toxic (LD50 approx 25 µg g–1 i.v. mouse) and are unlikely to contribute significantly to the signs of neurotoxic poisoning.

A2 Neurotoxins

431

1.3.2.2. TOXIC PHOSPHOLIPASES A2 Phospholipases A2 are common constituents of snake venoms. The enzymes are typical of other phospholipases A 2 in that they hydrolyze the 2-acyl bond of 3-sn phosphoglycerides, liberating a fatty acid and leaving the relevant lysophosphatide as the reaction product. The hydrolytic activity is strictly Ca2+-dependent; other divalent actions are inhibitory even though they may form structural complexes with the enzymes (13,14). Snake venom phospholipases A2 (E.C.3.1.1.4; phosphatide acylhydrolase) are members of a larger group of secretory phospholipases A2, all of which are small polypeptides of 13–14 kDa. They are highly resistant to denaturation, retain hydrolytic activity over wide ranges of both temperature and pH, and are active at lipid-aqueous interfaces, hydrolyzing micelles and structured membranes with equal facility (13). Snake-venom phospholipases A2 fall into two major classes (15). Class I phospholipases A2 may be isolated from the venoms of elapid snakes and their close relatives the sea snakes; class II phospholipases may be isolated from the venoms of viperid and crotalid snakes (16). Class I phospholipases A2 are homologous with pancreatic phospholipases A2; class II phospholipases A2 are homologous with the phospholipases A2 of platelets, neutrophils, and other cell types. Despite their differences in origin, and a rather low sequence homology (the between-class homology is approx 20% compared to a within-class homology of more than 60%), the two classes are very similar in secondary and tertiary structure (17,18). The class II enzymes have a slightly extended C-terminal region, have lost the disulphide link between Cys 11 and Cys 77, which is characteristic of class I phospholipases A2; and gained a disulphide link between Cys 50 and Cys 133. The Ca2+ binding domain is similar for both classes of phospholipase A2 (18) and involves, in each, conserved His-48 and Asp-99 residues. The venom glands of snakes are derived from salivary glands, and so it is not unreasonable to suppose that the primary purpose of venom phospholipases is digestive. Snakes cannot chew their prey. Prey items are swallowed whole and so the ability to inoculate into a prey item a highly active phospholipase provides a means whereby the digestive process can begin very early. But, although the venom phospholipases A2 may originally have been purely digestive in function, many possess a wide range of other activities. They may, for example, be anticoagulant, cytotoxic, haemolytic, neurotoxic or myotoxic. It is the neurotoxic, and myotoxic activities of the toxic phospholipases A2 with which this review is primarily concerned. The neurotoxic phospholipases A2 are commonly referred to as ß-neurotoxins to distinguish them from the α-neurotoxins, which bind to junctional nicotinic AChR at the vertebrate neuromuscular junction. The ß-neurotoxins are presynaptically active, inducing neuromuscular weakness by attacking the motor-nerve terminal. They are highly variable with respect to quaternary structure; some are monomers of approx 120 amino acid residues, some are dimers, and others are multimers of up to 5 polypeptides. In a few cases among the dimers, the two chains are covalently linked. Among the more complex multimers, one chain is often glycosylated. In every case at least one chain is hydrolytically active and toxic. There is general agreement that among the noncovalently linked di- and multimeric toxins, the toxicity of the hydrolytically active, toxic subunit is enhanced by one or more of its “partner” subunits and that its hydrolytic activity is often reduced (19–21). The diversity of quaternary structures of the

432

Harris

various ß-neurotoxins makes it difficult to discuss them as a single coherent group of toxins, and it is becoming more common to treat them as distinct groups of toxins (22). In this review, I shall consider the monomeric, dimeric, and multimeric toxins as separate groups because a consideration of the relationship between structure and function raises questions that specifically relate to quaternary structure. 1.3.2.2.1. Monomeric ß-Neurotoxins. Monomeric neurotoxins may be class I phospholipases A2 (e.g., notexin from the venom of Notechis scutatus and pseudexin from that of Pseudechis porphyriacus) or class II phospholipases A2 (e.g., agkistrodotoxin from the venom of Agkistrodon halys; caudoxin from Bitis caudalis and the ammodytoxins from the venom of Vipera ammodytes). It is of considerable interest that multiple isoforms of these major toxins may be isolated from a single venom sample. For example, six isoforms of notexin can be isolated from a single sample of venom from Notechis scutatus, two of which are nontoxic. Three isoforms of pseudexin have been identified, one of which has rather low toxicity (23). It is accepted that an organism will only maintain, transcribe, and express useful protein products of multiple copies of a gene under intense selective pressure because of the high metabolic demand. It is difficult to understand, therefore, why the nontoxic homologs are produced. Perhaps their toxicity is highly selective and their targets have not yet been identified. Kinetic studies of a number of monomeric phospholipases A2 suggested that under optimal conditions the phospholipases form homodimers (24). This has encouraged the view that dimers might be the active form of monomeric phospholipases A2, especially where multiple isoforms exist. But current thinking is that dimer formation might actually impede access to the substrate and thus inhibit hydrolytic activity (25), and it has been suggested (18) that dimerization might act as a way of storing the toxins in inactive form in the venom gland. There is no reason to suspect that the biological activity of the neurotoxic, monomeric phospholipases A2 requires the formation of multimers. 1.3.2.2.2. Dimeric ß-Neurotoxins. Two major groups of presynaptically active neurotoxins exist as stable dimers. The ß-bungarotoxins, isolated exclusively from the venom of kraits (genus Bungarus) consist of two covalently bound polypeptides. Chain A is 120 amino acids long. It is homologous with monomeric class I phospholipases A2 but does not possess the disulphide bond between residues Cys 11 and 77; it possesses instead an extra half cysteine at position 15. This half cysteine is involved in the formation of a covalent link between Chain A and its partner in the dimeric toxin, Chain B. Chain B is 60 amino acids long and is homologous with Kunitz-type inhibitors and the dendrotoxins. Chain B is linked to chain A via a disulphide bridge spanning Cys 15 on Chain A and Cys 55 on Chain B. The presence of homologues of a Kunitz-type inhibitor in intact ß-bungarotoxins does not confer proteinase inhibition on the intact toxin. Reduction of the linking disulphide bond of the ß-bungarotoxins leads to the separation of Chains A and B and their complete inactivation. Recombination does not restore activity. Numerous isoforms of both Chain A and Chain B have been identified and there are correspondingly numerous combinations of chains A and B. Although the toxicity of the different combinations is variable (26,27) mechanisms of toxicity (see below) are probably identical.

A2 Neurotoxins

433

The second major group of dimeric toxins is the crotoxin-like toxins. Crotoxin, the first toxic phospholipase A2 to be isolated, is from the venom of the South American rattlesnake Crotalus durissus terrificus. Crotoxin consists of two noncovalently linked polypeptides: CB (or crotactin) a basic, neurotoxic class II phospholipase A2 of 13.5 kDa and CA, an acidic complex of three small polypeptides produced by the posttranslational modification of a precursor phospholipase A2 homolog (28). In the absence of CA, CB is only weakly toxic. In combination, at a molar ratio of 1, the hydrolytic activity of CB is reduced but its toxicity is greatly enhanced. Similar toxins to crotoxin have been isolated from the venoms of several other species of Crotalus and from Viperid snakes of the genus Pseudocerastes, Vipera, and Trimeresurus. As with monomeric phospholipases A 2 and with the two subunits of ß-bungarotoxins, there are numerous isoforms of both subunits of intact crotoxin-like toxins, some arising as the result of the transcription, translation, and expression of multiple genes and some arising as a result of modifications introduced during posttranslational processing (28–31). 1.3.2.2.3. Multimeric Neurotoxic Phospholipases A2. A small number of neurotoxic phospholipases A2 comprise three or more subunits. Taipoxin, for example, the principle neurotoxin of the venom of the taipan Oxyuranus scutellatus has three subunits: α, ß, and γ. The α and ß subunits are typical class I homologs of 118 and 120 amino acid residues respectively; γ-taipoxin is 135 amino acid residues long and resembles a pancreatic phospholipase A2 with a pro-enzyme sequence intact. It is glycosylated and acidic (19). The α-subunit is pharmacologically active and its activity is enhanced when combined in a molar ratio of 1 with the γ-subunit. The activity does not appear to be affected by the presence or absence of the ß-subunit (32). Textilotoxin, isolated from the venom of another Australian snake, Pseudonaja textilis, comprises five subunits. Two of the subunits (A and D) are obligatory for the expression of neurotoxicity, but altogether only three of the five subunits are required for full toxicity to be expressed (33,34). The very short discussion of the quaternary structures of major neurotoxic phospholipases A2 raises numerous questions to which answers are only slowly emerging. For example, why do the monomeric toxins require no co-subunit? Do they aggregate to form homo-dimers or multimers during normal activity? What is the role of the nonhydrolytic subunits such as Chain B of the ß-bungarotoxins or CA of the crotoxins? Are they chaperones, allowing the precise targeting of the active subunit? If so, what does the concept mean? Do the di- and multimeric toxins retain their quaternary structure once inoculated into the victim or prey? If so, how are they so stable if they are not covalently limited? If they disaggregate, how can one subunit act as a chaperone to the other? A discussion of these topics is not appropriate in this review, but the reader is referred to a collection of recent reviews of these important concepts and problems (35). 1.3.2.3. CLINICAL ASPECTS OF NEUROTOXIC POISONING BY PHOSPHOLIPASES A2 The presence of neurotoxic phospholipases A2 in a venom does not necessarily mean that the human victim of an envenoming bite will express any signs or symptoms of neurotoxicity. This is simply because snake venoms are complex mixtures of toxic and nontoxic components. The prevailing signs and symptoms will therefore reflect the

434

Harris

nature and respective quantities of the toxins in the venom. With particular respect to the neurotoxic phospholipases A2, there is an additional significant problem: with the sole exception of the ß-bungarotoxins from the venoms of kraits (Bungarus species), the toxins are also potent myotoxins (see below). Thus, a patient with a severe envenoming bite by a snake whose venom contains significant quanitites of presynaptically active phospholipase A2 toxins may express rhabdomyolysis as well as a neurotoxin-induced failure of neuromuscular transmission. An equally important confounding factor is the fact that the venoms of elapid and hydrophiid snakes that contain neurotoxic phospholipases A2 also contain postsynaptically active neurotoxins, and so neuromuscular weakness and paralysis may be an expression of either postor presynaptic neurotoxicity or myotoxicity or a combination of all three. Determining the respective roles of these toxins in an envenoming incident is not straightforward, and can cause difficulties for physicians treating victims of envenoming bites. Very rapid identification of the envenoming snake (ideally based on enzyme-linked immunosorbent assay [ELISA] or similar techniques for the recognition of venom antigen in a victim’s circulation) combined with a good knowledge of venom constituents could allow much better management of patients (see refs. 5 and 36 for some comments on the epidemiology of snakebite and its implications for management). For the neurotoxicologist, the signs and symptoms following envenoming bites by snakes inoculating very complex venoms do not contribute much to an understanding of the cellular and molecular basis of the toxicity of the individual neurotoxic phospholipases A2. Rather, the clinical observations pose the questions while work on purified toxins provides most of the answers. 2. CELLULAR AND MOLECULAR TOXICOLOGY OF PURIFIED NEUROTOXIC PHOSPHOLIPASES A2 The neurotoxic phospholipases A2 all target the neuromuscular junction of skeletal muscle. In the mammal, a single muscle fiber is innervated by a single axonal branch from a single motor neurone. That single motor neurone may innervate a large number of muscle fibers (ranging from 10 to several hundreds). The motor neurone, its axon, and the muscle fibers it innervates is the motor unit. The muscle fibers comprising a single motor unit are usually dispersed widely throughout a muscle and are rarely contiguous. The metabolic profile of the muscle fiber is determined by the physiological properties (principally the firing rate) of the motor neurone, and is matched to the role of the motor unit. Motor units involved in posture are fatigue-resistant, fire action potentials at a resting (background) rate of c. 10 Hz and utilize oxidative metabolic pathways. The muscle fibers possess many mitochondria and high levels of myoglobin. Motor units involved in explosive activity fatigue rapidly, tend to be electrically silent at rest, but are capable of responding to short periods of very high firing rates. The muscle fibers are glycolytic, and possess few mitochondria and lower levels of myoglobin. Oxidative fibers contract slowly and express an adult “slow” myosin isoform. Glycolytic fibers contract rapidly and express an adult “fast” myosin isoform. As the axonal branch of the motor axon approaches the muscle fiber, it loses its myelin sheath and expands into a small spray of nerve terminals nestled within a synaptic cleft on the surface of the muscle fiber. The postsynaptic membrane is deeply

A2 Neurotoxins

435

invaginated and highly specialized, richly endowed with ACh receptors at the top of the postsynaptic folds and with voltage-gated Na+ channels in the deep regions of the folds. The synaptic cleft accommodates the nerve terminal, leaving a gap between nerve terminal and muscle fiber of approx 50 nm. The cleft and the postsynaptic folds are lined with basal lamina, which is continuous with the muscle fiber basal lamina and which covers the terminal Schwann cell of the motor axon. The nerve terminal contains a large population of clear synaptic vesicles 50–60 nm in diameter. The vesicles contain the motor transmitter ACh and are clustered around “active zones” in the motornerve terminal. These zones are rich in voltage-gated Ca2+ channels and are the localization sites for a number of transmembrane proteins involved in synaptic vesicle binding, fusion, and recycling, such as SNAP 25, synaptobrevin, and syntaxin. The active zones are localized directly opposite the peaks of postsynaptic folds, and act as the release sites at which synaptic vesicles discharge, by exocytosis, their contents into the synaptic cleft. Released ACh diffuses across the synaptic cleft to bind with the AChR. The process of binding induces a conformational change in the AChR and the opening of its ionophore. The movement of Na+, K+, and Ca2+ along their concentration gradients depolarizes the postsynaptic membrane. If this depolarization is sufficiently large, the postsynaptic voltage-gated Na + channels open to generate a propagated action potential that eventually initiates excitation-contraction coupling in the muscle fibers. The release of ACh is triggered by the invasion of an action potential into the motornerve terminal. The action potential causes the opening of voltage-gated Ca2+ channels and the entry of Ca2+ from the extracellular phase (where [Ca2+]o = 10–3 M) into the nerve terminal (resting [Ca2+]i = 10–7 M). The local internal concentration of Ca2+ at the active zone increases very rapidly to approx 10–4 M. Synaptic vesicles in the nerve terminal are distributed between a number of different pools. The immediately available pool comprises vesicles already docked at the active zones. A second, reserve pool, is bound to actin filaments by a group of vesicle-associated proteins, of which synapsin is the best known. Phosphorylation of synapsin releases the vesicles from their bound state, allowing them to dock at the active zones that become vacant following exocytosis. The docking of the synaptic vesicles relies on the formation of complexes of vesiclespecific proteins (for example, synaptophysin and synaptotagmin) with proteins specific to the release site of the nerve terminal plasma membrane (for example, SNAP-25 and syntaxin). Phosphorylation of these proteins is responsible for the fusion of the vesicle with the nerve-terminal plasma membrane and the exocytosis of its contents. The synaptic vesicle is then retrieved by a process of endocytosis. This process, which also appears to involve synaptotagmin and Ca2+-dependent protein phosphorylation, is integral to the recycling of the synaptic vesicles. The importance of understanding the molecular biology of transmitter release and synaptic-vesicle recycling is that the relevant proteins are frequently significant targets for natural toxins. For example, botulinum-induced paralysis results from the cleavage of syntaxin and SNAP-25 (37); and α-latrotoxin, which targets neurexin (another protein component of the presynaptic nerve terminal membrane) (38), induces massive transmitter release that is Ca2+-independent.

436

Harris

3. FAILURE OF NEUROMUSCULAR TRANSMISSION CAUSED BY NEUROTOXIC PHOSPHOLIPASES Detailed studies on the cellular biology of transmission failure caused by the neurotoxic phospholipases A2 have been made using only a small number of toxins, particularly ß-bungarotoxin, crotoxin, notexin, and taipoxin. Of these toxins, only ß-bungarotoxin has a specific neurotoxic action. Crotoxin, notexin, and taipoxin are all potent myotoxins and thus data on neurotoxic activity are sometimes difficult to interpret. For this reason, this review will concentrate on findings generated using ß-bungarotoxin. 3.1. Site of Action of Neurotoxic Phospholipases A2 Studies on ß-bungarotoxin have shown beyond doubt that the target for the neurotoxic phospholipases A2 is the nerve terminal rather than the axon of the motor neurone. The definitive evidence for this is that, at the cessation of spontaneous respiration following poisoning by ß-bungarotoxin, electrical discharges can still be recorded from the phrenic nerve (39). This simple observation demonstrates that fatal poisoning involves neither damage to the lower motor neurone nor the impairment of axonal conduction of the action potential. There is also extensive evidence that the presynaptically active neurotoxic phospholipases A2 do not block AChR on the postsynaptic surface of the neuromuscular junction (40–42), although it should be noted that crotoxin may bind to junctional AChR. This toxin does not influence the binding of AChR ligands but instead stabilizes the AChR in its desensitized form (43). 3.2. Are the Toxins Active on the Nerve-Terminal Membrane or Are They Internalized? This extremely important question has not yet been satisfactorily resolved. The simplest explanation for the highly selective action of ß-bungarotoxin and related toxins would be that it binds to an “acceptor” on the nerve-terminal membrane and is then suitably orientated so that it can attack the plasma-membrane phospholipids. The model would then suggest that the hydrolysis of the nerve-terminal phospholipids would result in the loss of ion homeostasis, depolarization, and the entry of Ca2+ into the intracellular compartment. This would initiate transmitter release and nerve-terminal degeneration. A less economical possibility is that the toxins bind to the acceptor and that binding is followed by the internalization of either the toxin or the toxin/acceptor complex. The intracellular targets could then include mitochondria, synaptic vesicles, or any of the proteins involved with the synthesis and packaging of transmitter, with vesicle mobilization, docking, or fusion or with exocytosis. There is general agreement that there is a rate limiting activation step. In isolated nerve-muscle preparations, for example, there is a latent period of up to 20 min between exposure to toxin and the onset of neuromuscular failure (4). This latent period is only marginally affected by toxin concentration. The binding of toxin to its acceptor is rapid, however, and its rapidity is demonstrated by the fact that the toxin cannot be removed by washing if exposure has lasted for more than 5 min, even though the latent period lasts for approx 20 min (Fig. 2). Similarly, the effects of the toxins can be inhibited in vitro by neutralizing antibodies only during the earliest stages of the latent period (44).

A2 Neurotoxins

437

Fig. 2. Failure of the indirectly-elicited muscle twitch in rat hemidiaphragm following exposure to ß-bungarotoxin (3 µg mL–1). The toxin was added at time 0 and the preparation was washed × 3 changes of bathing fluid 1–20 min after the addition of toxin. The onset of paralysis can only be prevented if the exposure period lasts for 5 min or less. (Figure courtesy of S. Prasarnpun.)

The data suggest that the binding of toxin to acceptor is very tight and that after binding the toxin is either internalized, or that it needs time to begin to destroy the nerve-terminal plasma membrane, or that it intitiates a series of secondary events that lead to transmission failure. There is only very limited evidence that ß-bungarotoxin is internalized. Strong et al. (45) reported the binding of horse radish peroxidase (HRP)-conjugated ß-bungarotoxin to nerve-terminal plasma membranes and mitochondria in mouse diaphragm, and also suggested localization to synaptic vesicles. Esquerda et al. (46) used Torpedo synaptosomes and showed binding of radiolabeled ß-bungarotoxin to the plasma membranes. Dixon and Harris (unpublished data) used immunohistochemistry with gold-conjugated notexin, immunolabeling with gold-conjugated antibodies to notexin and gold-labeled ß-bungarotoxin in an attempt to visualize the binding site at the level of electron-microscopy without success, despite its routine use in the study of toxin labeling in skeletal muscle (47,48). The difficulty of studies such as those of Strong et al., Esquerda et al., and Dixon and Harris is that controlling for false-positives and false-negatives is very difficult, and that the risk of sampling errors is very high. Howard and Wu (49) conjugated ß-bungarotoxin to polystyrene beads that were too large to be internalized into synaptosomes. The effects of ß-bungarotoxin were not impaired, implying that internalization is not essential for the expression of pharmacological activity. Neither is hydrolytic activity essential for binding; there is good evidence that ß-bungarotoxin inactivated by conjugation with paraphenacyl bromide and Sr2+ will still bind to nerve terminals (50). 3.3. Neuronal Binding Sites Despite the difficulty of determining whether the presynaptically active neurotoxic phospholipases A2 are internalized, numerous studies have identified putative accep-

438

Harris

tors or binding sites on synaptosomes. The data on binding are confusing and contentious and suggest that rather than there being a single binding site there are numerous “acceptors” or binding sites, each site binding a different, but possibly overlapping, group of neurotoxic phospholipases A2. Most data have been derived from studies on ß-bungarotoxin, crotoxin, a less well-characterized toxin from the venom of the taipan, 0S2, and the class II neurotoxin ammodytoxin. The binding studies made in vitro typically use synaptosomal membranes or membrane fragments. The binding sites, though, may be irrelevant to the expression of neurotoxicity, representing instead binding sites for physiological ligands of different kind. These and other concerns on the relevance of the studies on the binding to central synaptosomes of toxins that act peripherally should ensure that data are interpreted cautiously. 3.3.1. Neuronal Binding Sites for ß-Bungarotoxins Rehm and Betz (51–53) have made a series of detailed studies of the binding of ß-bungarotoxin to synaptosomes prepared from rat and chicken brain. They used [125I] ß-bungarotoxin and affinity labeling of chicken brain to show that the principal high affinity (Kd = 1 nM) binding site was present at a density of 50–150 f moles mg–1 protein, that it was destroyed by proteolysis and that it had a molecular mass of about 430 kDa. The binding of the toxin was not inhibited by other representatives of either class I or class II neurotoxic phospholipases A2 (notexin, taipoxin, crotoxin). A minor binding site of 95 kDa was also identified by Rehm and Betz (51) but its status has not been determined. It is of interest that dendrotoxin inhibited the binding of ß-bungarotoxin to its specific binding sites and that the ß-bungarotoxin inhibited the binding of dendrotoxin. Dendrotoxin binds to two sites in rat and chicken synaptosomes, a high-affinity binding site (Kd = 0.5 nM; Bmax = 90 pmoles mg–1 protein) and a lower-affinity binding site (Kd = 15 nM; Bmax = 400 pmoles mg–1 protein). The dendrotoxin receptor is probably a neuronal voltage-gated K+ channel (54,55); it is of interest, therefore, that there is good physiological evidence that the presynaptically active neurotoxic phospholipases A2 bind to voltage-gated K+ channels in the motor-nerve terminal (56,57) and that preincubation with dendrotoxin inhibits neuromuscular failure caused by ß-bungarotoxin (58). It is also of interest that the B-chain of ß-bungarotoxin is homologous with dendrotoxin. Whether it is the B-chain of ß-bungarotoxin that acts as the binding region of the whole toxin is not known. 3.3.2. Neuronal Binding Sites for Crotoxin Two major binding sites for crotoxin have been identified on brain synaptosomes. One has a molecular mass of 85 kDa, the other of about 45 kDa with Kd’s of 4 nM and 87 nM, respectively (59–61). The binding of crotoxin was inhibited by the functionally related neurotoxic phospholipases A2, and by porcine pancreatic phospholipase A2. More recent studies (62) have identified two populations of binding sites for crotoxin, differentiated by binding affinity (Kd = 10 nM and 40 pM, respectively). The high-affinity binding site was sensitive to heat treatment and to exposure to proteases and so is probably a protein. The low-affinity binding site was sensitive to neither and this binding site probably corresponds to a membrane lipid. Ammodytoxin, a closely related class II neurotoxic phospholipase A2, did not inhibit the binding of crotoxin; neither did crotoxin inhibit the binding of ammodytoxin (see below).

A2 Neurotoxins

439

Bon and his colleagues have concentrated on the binding of the two chains of crotoxin (CB and CA) to putative receptors. They have shown that crotoxin binds with a Kd of 1 µM to basic phospholipid vesicles. It is the CB subunit that binds, the CA subunit being liberated in the course of binding. Crotoxin does not bind with high affinity to neutral phospholipid vesicles (63–65). The functional relationship between CB and CA has been a matter of discussion for many years (see ref. 66 for example) but has still not been resolved. It is similarly difficult to reconcile into a coherent functional model the respective roles of phospholipid and protein binding sites in the expression of pharmacological activity of neurotoxic phospholipases A2 like crotoxin. 3.3.3. Neuronal Binding Sites for OS2

OS2 is a class I phospholipase A2 isolated (67) from the venom of the taipan Oxyuranus scutellutus. It is neurotoxic, a definition based on the physiological effects of inoculation into cerebral ventricles rather than at the neuromuscular junction (67,68). OS2 also inhibits ACh release at cholinergic neuro-neuronal synapses of Aplysia (68). Using [125I] -OS2 Lambeau et al. (67) demonstrated two high-affinity binding sites in rat brain synaptosomal membranes: one site had a Kd = 1.5 pM; Bmax = 1 p mole mg–1 protein, the other a Kd = 45 pM; Bmax = 3 p moles mg–1 protein. The two classes of binding sites had molecular masses 36–51 kDa and 85–88 kDa, respectively. These binding sites also act as receptors for the high-affinity binding of many other neurotoxic phospholipases A2, including taipoxin, crotoxin, ammodytoxin, notexin, and textilotoxin, but neither ß-bungarotoxin nor non-neurotoxic phospholipases A2 bound with high affinity. The binding of OS2 to the receptors was Ca2+-dependent but this should not be interpreted as indicating that binding depends on hydrolytic activity because Zn2+, which is a potent inhibitor of hydrolytic activity, also supported binding. Although the high-affinity neuronal binding site for the neurotoxic phospholipases was termed by Lazdunski’s group the N-type receptor (to distinguish it from a muscletype receptor, the M-type receptor), the receptor has been identified in a wide range of non-neuronal tissues such as kidney, liver, and lung. This suggests that the N-type receptor may not be a specific receptor for presynaptically active phospholipases A2 but may be a receptor for a number of physiological ligands (endogenous phospholipases for example) to which the neurotoxins, perhaps fortuitously, also bind. Neither the function nor the detailed structure of the N-type receptor is known. 3.3.4. Neuronal Binding Sites for Ammodytoxins

The ammodytoxins (ammodytoxin A, B, and C) are class II neurotoxic phospholipases A2 isolated from the venom of Vipera ammodytes (69–71). Ammodytoxin A binds with high affinity (Kd = 4 nM) to a binding site with a Bmax = 6.7 p moles mg–1 protein on bovine cortical synaptic membranes. The binding site is a protein with a molecular mass of 53–56 kDa. The binding of ammodytoxin A was inhibited by other ammodytoxins and by crotoxin, but not by non-neurotoxic phospholipases A2 or by ß-bungarotoxin. A nonspecific low affinity but high-density binding site, unaffected by heating or treatment with proteases, was thought to be lipid. Binding to the highaffinity binding site was blocked by the CB subunit of crotoxin but was unaffected by CA and ß-bungarotoxin. Ca2+ was an essential co-factor for binding. Binding to Torpedo synaptosomal membranes was also characterized by both highand low-affinity kinetics. Binding to the low affinity site was blocked by other neuro-

440

Harris

toxic phospholipases (but not by ß-bungarotoxin); binding to the high-affinity binding site was not Ca2+-dependant and was blocked to a varying degree by other neurotoxic phospholipases A2 (73). In summary, it is clear from these various studies that high-affinity, selective putative receptors for the neurotoxic phospholipases A2 fall into two broad classes, one with a molecular mass of between 35 and 55 kDa and one with a molecular mass of between 70 and 95 kDa. The receptors are almost certainly membrane-bound proteins, but the lipid environment in which the proteins are embedded may play a significant role in toxin-receptor binding. No studies have localized the receptors to specific regions in the target cells, and little is known of either the function of the putative receptor proteins or of their natural ligands. Although most interest with respect to the function of the binding sites relates to the possibility that they correspond to K+ channels (74), it is reasonable to ask whether the binding studies shed any light on the question of whether the toxins act exclusively on the plasma membrane or whether they are internalized. The most detailed studies of any of the binding sites have been made on the N-type receptors identified originally as binding sites for OS2, but molecular and functional information on the N-receptor is sparse (75). Most recently, neuronal pentaxin, a secreted neuronal protein of molecular mass 47 kDa, has been found to bind taipoxin, with high-affinity (76). Pentaxin is an homolog of a number of “acutephase proteins” that may be involved in the uptake of foreign agents such as bacteria and cytotoxins (77). A second, intracellular Ca2+-binding protein (TCBP), a 49 kDa component of the reticular system in neurons and glia, has also been identified as a putative binding protein for taipoxin. It has been suggested that the two proteins might be involved in the uptake and intracellular activation of toxic phospholipases (76,77). Note, however, that recloning pentaxin and TCBP into various cell lines did not result in the binding of OS2 or taipoxin. It seems unlikely that either pentaxin nor TCBP are natural ligands for secretory phospholipases. 3.4. Neuromuscular Transmission and Neurotoxic Phospholipases A2: Transmitter Accumulation, Synthesis, and Release The presynaptically active phospholipases all cause neuromuscular paralysis. In isolated nerve-muscle preparations, the onset of paralysis is typically preceded by an initial inhibition of neuromuscular transmission (phase I), followed by a facilitation of transmission (phase II), and then a third phase of progressive and unrelenting decline in effective transmission. The appearance of phase I and phase II is variable (see refs. 42,66 for example) and seems to depend on both the toxin under examination and the species from which the nerve-muscle preparation has been made. Phase III is invariable. Electrophysiological recordings have shown consistently that the inhibitory phase is reflected in a reduction in both spontaneous transmitter release, measured as a reduced frequency of spontaneous miniature end-plate potentials (MEPPs), and in a reduced quantal content of stimulus-evoked end-plate potentials (EPPs). The secondary facilitation is reflected in an increased in mepp frequency and in quantal content of evoked epps. The phase of facilitation may be associated with the appearance of giant mepps and spontaneous epps (78–80; Fig. 3). The final phase of unrelenting decline in transmission is associated with a steady decline in the frequency (but not the amplitude) of

A2 Neurotoxins

441

Fig. 3. MEPP frequency and quantal content of EPPs at a frog neuromuscular junction exposed to crotoxin 250 pM from time 0. Inset, MEPPs and a giant MEPP recorded 58 min after addition of crotoxin. Giant MEPPs (GMEPPs appeared at arrow). (Figure courtesy of B. Hawgood and C. Bon.)

mepps and in the quantal content of epps (81). As might be expected, mepps can be recorded for some time after the failure of effective (nerve-stimulation-evoked) neuromuscular transmission because the quantal content of epps falls below that necessary to initiate an action potential in the postsynaptic membrane, even though the nerve terminal still contains synaptic vesicles full of transmitter (82). The biological basis of the three phases described previously is not well-understood. The early inhibition appears to be independent of hydrolytic activity because neither the substitution of Sr2+ for Ca2+ nor acylation of the toxins (procedures that abolish phospholipase A2 activity) appears to be prevent the initial inhibition (79,83–86), although there are exceptions to this general observation (50). The inhibition of transmitter release under these circumstances is difficult to understand. Harris (87) suggested that binding of the toxins is sufficient to block, transiently, the movement of Ca2+ or to hyperpolarize the nerve terminal membrane, but there is no evidence for this at all. There is a similar lack of understanding of the biological basis of the facilitation of transmitter release. There is a general perception that hydrolytic activity is essential for the expression of phase II (85–88) but that does not seem to be an invariable rule (89). In accord with the general perception, it has been suggested that the facilitation results

442

Harris

Fig. 4. Effects from presynaptically active phospholipase A2 neurotoxins on perineural wave forms recorded from mouse neuromuscular junctions after 60-min exposure to the toxins. Note the selective reduction of component b, generally considered to represent current flow through a slowly activating K+ channel. (Figure courtesy of A. Harvey.)

from the hydrolysis of nerve-terminal lipids. The concept is that during early stages of hydrolysis, the increase in membrane fluidity would result in depolarization and a rise in [Ca2+]i, the rise in [Ca2+]i stimulating an increase in both the spontaneous and evoked release of transmitter. There is a tantalizing alternative explanation. Peterson et al. (90) reported that ß-bungarotoxin partially blocked a noninactivating K+-current in isolated dorsal-root ganglion cells. Subsequently, Dreyer and Penner (56) and Rowan and Harvey (57) obtained data that suggested that a slowly activating K+ current associated with the arrival of an invading axonal action potential at the nerve terminal was reduced by a number of neurotoxic phospholipases A2 (Fig. 4). Such an action could result in a prolonged depolarization of the nerve terminal, an enhanced entry of Ca2+ through voltage-gated Ca2+ channels, and an enhancement of both spontaneous and evoked transmitter release. The putative role of binding to a K+ channel is not only of interest because it offers an explanation for phase II, but because it suggests that the mutual competition between dendrotoxins and neurotoxic phospholipases A2 at binding sites on a number of neuronal preparations reflects the presence of mutually acceptable K+ channels on the neuronal membranes. Rowan and Harvey (57) pointed out that the toxins did not affect putative K+ currents in amphibian preparations, and that although notexin did block mammalian K+ currents, it did not facilitate transmission. These anomalies may simply reflect the presence of varying types of K+ channels on neuronal membranes, and varying affinities of the neurotoxins for these putative K+ channels. The difficulty in obtaining definitive data on the mechanism of action (rather than a description of events) of the presynaptically active neurotoxic phospholipases A2 has prompted numerous studies on the supposedly simpler preparations of synaptosomes. Most have worked on synaptosomes prepared from mammalian brain or from the electric organ of Torpedo. Whatever the source, synaptosomes are depolarized following

A2 Neurotoxins

443

exposure to the neurotoxic phospholipases A2 (91–95). The depolarization is Ca2+-dependant, but is neither dependant on [Na+]o nor mediated by Na+/K+ ATPase activity (91,93,94). The depolarization is probably caused by the nonspecific loss of ion gradients as a result of hydrolysis of membrane phospholipids, the release of fatty acids, and the formation of lysophosphatides. This interpretation is strengthened by the evidence that depolarization is a nonspecific response to phospholipases A2 in general (96) and is not specific to neurotoxic phospholipases A2. Preincubation of synaptosomes with neurotoxic phospholipases A2 results in the inhibiton of uptake of a wide range of transmitters and transmitter precursors (92, 97–102 for example). All the uptake processes are energy-dependent and are driven by the electrical gradient across the plasma membrane of the nerve terminal. A significant depolarization inhibits the uptake process. Thus, early suggestions that the inhibition of uptake of choline is specific and leads to reduced transmitter synthesis and, eventually, to the inhibition of neuromuscular transmission may be inaccurate in the sense that the inhibition of uptake of choline is nonspecific and probably represents the indirect consequence of physical damage to the synaptosome and depolarization. Enhanced uptake of choline has been reported when labeled choline is incubated with synpatosomes in the presence of the neurotoxic phospholipases A2 (103–105). This is not inconsistent with the more general view that uptake is inhibited. During early stages of incubation, a small depolarization may enhance the uptake process; the activity of the choline carrier will be inactivated as the membrane potential continues to fall (106). The data relating to the effects of the neurotoxic phospholipases A2 on transmitter synthesis are as confusing as those relating to uptake processes. The preincubation of synaptosomes with neurotoxic phospholipases A2 (conditions that cause the inhibition of choline uptake) results in a reduction of ACh synthesis (107,108); if the synaptosomes are incubated with choline and neurotoxic phospholipases together (conditions that enhance choline uptake) there is an enhancement of transmitter synthesis (109). Since the neurotoxic phospholipases have no direct effect on the activity of choline acetyltransferase (ChAT) or AChE (101–107; Harris, unpublished) the effects on synthesis probably reflect the availability of choline. Exposure of synaptosomes to neurotoxic phospholipases A2 for longer periods results in the release of transmitters together with lactate dehydrogenase and deoxyglucose (94,97,107,110–112). The biological data overwhelmingly suggest that the release is associated with membrane damage, and consistent with this interpretation are the regular findings that non-neurotoxic as well as neurotoxic phospholipases A2 can initiate release from synaptosomes (102,113). It is of interest, however, that neurotoxic phospholipases A2 tend to be more potent potentiators of transmitter release than nontoxic phospholipases A2, and that ß-bungarotoxin, at least, can initiate the release of transmitter under conditions when there is no evidence of any physical or structural damage to the synaptosomes (114). Perhaps the neurotoxic phospholipases A2 are capable of initiating release at lower concentrations and with greater selectively because they act as ligands for specific synaptosomal binding sites with high affinity and thus gain immediate access to target phospholipids. There is also much evidence that many of the pharmacological effects of phospholipases A2 on synaptosomes are

444

Harris

caused by fatty acids released as the result of hydrolysis of phospholipids (102,108,115). 3.5. Protein Phosphorylation and Transmitter Release Given the importance of phosphorylation and dephosphorylation of proteins involved in synaptic vesicle mobilization and exocytosis (see p. 435), it is surprising that very little work has been done on the possibility that the neurotoxic phospholipases A2 might modulate protein phosphorylation in nerve terminals. Only Rosenberg and his colleagues (discussed at length in ref. 116) have explored these possibilities. They have reported that in synaptosomal preparations ß-bungarotoxin and a number of other neurotoxic phospholipases A2 inhibit protein phosphorylation. This inhibition includes that of synapsin. Since the phosphorylation of synapsin is an integral step in the mobilization of bound stores of synaptic vesicles, inhibition of phosphorylation would be expected to lead to a block of synaptic-vesicle mobilization and, eventually, the cessation of transmitter release. If vesicle mobilization were to be blocked by the phosphorylation of synapsin, one would expect nerve terminals to be still filled with synaptic vesicles and those vesicles would not be located at release sites but deeeper within the nerve terminal (117). 4. NEUROMUSCULAR PATHOLOGY The majority of studies on the effects of neurotoxic phospholipases A2 on neural tisue have concentrated on the acute responses of isolated nerve-muscle preparations or synaptosomes. But, for the clinician dealing with victims of envenoming bites by snakes whose venoms are particularly rich in such neurotoxic agents, the problems are not simply the acute neuromuscular paralysis, but the treatment and management of envenomed victims during what can be a prolonged post-acute phase. The most detailed clinical studies of victims of such bites have been made by Trevett et al. (118–121). Trevett and his colleagues showed that bites by the Papua/New Guinea taipan caused an immediate, acute severe neuromuscular weakness, a second phase of profound paralysis and a third phase of slow recovery of normal function lasting up to 3 wk. The patients were extremely difficult to treat, being unresponsive to antivenom, anticholinesterases, and agents that enhance transmitter output such as diaminopyridines. Physiological studies in isolated tissues may explain the onset of weakness as a “blockade of transmitter release,” but only detailed neuropathological studies combined with physiology are likely to provide a real explanation of the syndrome. A few studies of the morphology of neuromuscular junctions have been made in preparations exposed in vitro to presynaptically active neurotoxins. Most reports have suggested that at the time of acute paralysis motor nerve terminals are depleted of synaptic vesicles and exhibit “Ω” profiles on the presynaptic plasma membrane and swollen, disrupted mitochondria (45,122). It has been claimed that synaptic vesicle depletion is a transient event and that at the time of complete blockade of transmission, the vesicle population has returned to normal (123), but the observations that underpinned this suggestion have never been replicated and were probably the result of a sampling error. The toxins also seem to activate terminal Schwann cells. These cells enwrap nerve terminals, isolating nerve terminals from the postsynaptic membrane, and then phagocytose degen-

A2 Neurotoxins

445

erating nerve terminals (45,122). It has been claimed that the morphological features described above are rather nonspecific and reflect “extensive damage observed after a long period of exposure, or after exposure to high concentrations, to the enzymatic activity of these toxins” (116). The observations raise a more important question: is the basis of paralysis the blockade of transmitter release or the depletion of synaptic vesicles, and if the latter, how does depletion occur? Prasarnpun and Harris (in progress) have made detailed studies of the acute responses of the rat-isolated hemidiaphragm muscle exposed to ß-bungarotoxin. They have shown that at the time of complete block of neuromuscular transmission all identifiable motor-nerve terminals in the rat hemidiaphragm muscle were seriously depleted or completely devoid of synaptic vesicles. The nerve-terminal membranes were intact and smooth, and nerve-terminal perimeter was unchanged compared with that in preparations not exposed to ß-bungarotoxin. This would suggest that depletion was not caused by the inhibition of either the mobilization or the recycling of synaptic vesicles by the general destruction (by lipid hydrolysis) of the terminal plasma membrane. Instead the data suggest that ß-bungarotoxin caused the mobilization of all pools of synaptic vesicles, including the tightly bound reserve pools (117). Quantitative fluorescence microscopy showed that there was no loss of intensity of labeling of AChR or of two proteins involved in vesicle docking and exocytosis, SNAP 25 and syntaxin, thus supporting the conclusion that the release process was probably intact. Synaptophysin, an integral protein of synaptic vesicles, was reduced by only 30% and very few neuromuscular junctions were identified at which synaptophysin labeling was lost entirely. What then was the fate of synaptic vesicles: how were they mobilized, even from the most tightly regulated reserve pools, and where did they go if they were not irreversibly fused to the nerve terminal plasma membrane? And where was the synaptophysin that could be so clearly visualized? There are no specific answers to these questions as yet, but given the role of protein kinases in vesicle mobilization (117), it seems possible that the highly effective mobilization results from the activation of a number of secondary intracellular signaling systems, the most likely candidates of which are the protein kinases A and C and, possibly, Ca2+/calmodulin (124). It is more difficult to speculate on the fate of the synaptic vesicles. The evidence that they do not remain fused with the nerve terminal membrane is indirect but strong. Clathrin-coated synaptic vesicles could be found in those terminals in which depletion was substantial but incomplete, and there would have been an increase in nerve-terminal area had the vesicle been incorporated into the nerve-terminal membrane and not recycled. It seems unlikely that the vesicles could have been retrogradely transported out of the nerve terminal, and the only obvious conclusion is that they were destroyed internally, perhaps by phospholipase activity stimulated internally as a result of the secondary signaling processes initiated by the binding of the toxins to the nerve terminal. In this context, it is of interest that intracellular structures in a number of systems, including mitochondria, sarcoplasmic reticulum, and synaptic vesicles, are very sensitive to phospholipase activity (125,126). It might be objected that spontaneous transmitter release (measured in the form of MEPPs) can be recorded from nerve-muscle preparations that have been paralyzed by exposure to the presynaptically active neurotoxic phospholipases A2, because this

446

Harris

would imply that synaptic vesicles were still available for docking and release (82). This observation is not inconsistent with the general thesis. It would support the contention that mobilization and release is unimpaired and it is exactly what would be expected during the latter stages of transmission failure. Experiments in vivo have confirmed that vesicle depletion is an early response (1–3 h postinoculation) of somatic motor nerve terminals exposed to the neurotoxic phospholipases, but the in vivo studies have also shown that by 24 h nerve-terminal degeneration (including the terminal axon) is an almost invariable response (127,128). The recovery of transmission is rapid. In the rat the earliest signs of regeneration appear 3 d post inoculation and neuromuscular function is complete by 7–21 d postinoculation (129,130). There appears to be no loss of the number of functioning motor units and so it is possible to conclude that the ability of the neurotoxic phospholipases to destroy immature motor neurones in vivo (131–135) is not retained in the adult. There is, however, extensive axonal sprouting (terminal and nodal) in the reinnervated muscles (127,128) the causes and implications of which have not been explored. These data offer a sufficient explanation for the clinical observations. The rapid onset of neuromuscular paralysis would represent the early depletion of transmitter. The longer period of profound weakness would represent the period of nerveterminal degeneration and the slow recovery of neuromuscular function would represent the regeneration and maturation of the neuromuscular junction. The first two phases—depletion of transmitter and the degeneration of the nerve terminal— would naturally be insensitive to treatment. 5. MYOTOXIC PHOSPHOLIPASES A2 Three groups of toxins are responsible for the necrosis of skeletal muscle: the small (42–45 amino acid residues) polypeptides such as crotamine, myotoxin a, peptide c, toxin c, and myotoxin l found in the venoms of crotalid snakes of the Americas, the 60– 62 amino acid residues long cardiotoxins (also known as cytotoxins, direct lytic factors, and membrane toxins) found in the venoms of the cobras, the ringhals and (possibly) the mambas, and the myotoxic phospholipases A2 (136). Of these toxins, the most important are the myotoxic phospholipases A2 because they are found in numerous venoms (e.g., the elapid snakes of Australasia, the South American rattlesnakes, Russell’s viper, Western sand viper) and because with the solitary exception of ß-bungarotoxin all presynaptically active neurotoxic phospholipases A2 are also potent myotoxins. Thus an envenoming bite by a snake whose venom is rich in neurotoxic phospholipases A2 may precipitate neuromuscular weakness by the combined effect of destruction of the motor-nerve terminal and of the muscle fiber it innervates (128). The pathological features of skeletal-muscle degeneration initiated by the myotoxic phospholipases A2 have been described often, and detailed studies on the primary site of action and the relative rates of degeneration of muscle-specific proteins and on the regeneration of the damaged muscle are slowly emerging. The interested reader should approach some of the recent reviews for an overview of myotoxicity (86,136–138) and source references (47,48,139,140) for an introduction to the cellular basis of myotoxicity and myotoxic phospholipases A2. Note added in proof: Shortly after the submission of this article two imporant reviews appeared: Schiavo, G., Matteoli, M., and Montecucco, C. (2000) Neurotoxins

A2 Neurotoxins

447

affecting neuroexocytosis. Physiol. Rev. 80(2), 717–766, and Montecucco, C. and Rossetto, O. (2000) How do presynaptic PLA2 neurotoxins block nerve terminals? Trends Biochem. Sci. 25, 266–270. REFERENCES 1. Harry, G. J. (1999) Developmental neurotoxicity, in Medical Neurotoxicology (Blain, P. G. and Harris, J. B., eds.), Arnold, London, pp.13–30. 2. Bates, D. (1999) Diagnosis of neurotoxic syndromes, in Medical Neurotoxicology (Blain, P. G. and Harris, J. B., eds.), Arnold, London, pp. 3–12. 3. Mirtschin, P. and Davis, R. (1987) Clinical diagnosis, in Dangerous Snakes of Australia, rev. ed. Ure Smith Press, Sydney, pp. 164–167. 4. Maretic, Z. and Russell, F. E. (1979) An unusual non venomous snake bite. Toxicon 17, 425–427. 5. Warrell, D. A. (1986) Tropical snake bite: clinical studies in south east Asia, in Natural Toxins (Harris, J. B., ed.), Oxford University Press, Oxford, pp. 25–45. 6. Kouyoumdjian, J. A., Polizella, C., Lobo, S. M. A., and Guimares, S. M. (1991) Fatal extradural haematoma after snake bite (Bothrops moojeni). Trans. R. Soc. Trop. Med. Hyg. 85, 552. 7. Tin-Myint, Rai-Mua, Maung-Chit, Tun-Pe, and Warrell, D. A. (1991) Bites by king cobra in Myanmar: successful treatment of severe neurotoxic envenoming. Q. J. Med. N Series 80, 751–762. 8. Weinstein, S. A., Stiles, B. G., McCoid, M. J., Smith, L. A., and Kardong, K. V. (1993) Variation of lethal potencies and acetylcholine receptor binding activity of Duvenoy’s secretions from the brown tree snake, Boiga irregularis Merrem. J. Nat. Toxins 2, 187–198. 9. Adem, A., Asblom, A., Johansson, G., Mbugua, P. M., and Karlsson, E. (1988) Toxins from the venom of the green mamba Dendroaspis augusticeps that inhibit the binding of quinuclidinyl benzilate to muscarinic acetylcholine receptors. Biochim. Biophys. Acta 968, 340–345. 10. Laothong, C. and Sitprija, V. (2001) Decreased parasympathetic activities in Malayan krait (Bungaris candidus) envenoming. Toxicon. 39, 1353–1357. 11. Cervenansky, C., Dajas, F., Harvey, A. L., and Karlsson, E. (1991) Fasciculins, anticholinesterase toxins from mamba venoms: biochemistry and pharmacology, in Snake Toxins (Harvey, A. L., ed.), International Encyclopedia of Pharmacology and Therapeutics, Pergamon Press, New York, pp. 303–321. 12. Harvey, A. L. (1997) Dendrotoxins (Dendroaspis species), in Protein Toxins and Their Use in Cell Biology (Rappuoli, R. and Montecucco, C., eds.), Oxford University Press, Oxford, UK, pp. 159–161. 13. Harris, J. B. and Macdonell, C. A. (1981) Phospholipase A2 activity of notexin and its role in muscle damage. Toxicon 19, 419–430. 14. Scott, D. L. and Sigler, P. B. (1994) The structural and functional roles of Calcium ions in secretory phospholipases A2, in Advances in Inorganic Biochemistry, vol. 10 (Eichorn, G. L. and Marzilli, L. G., eds.), pp. 139–155. 15. Harris, J. B. (1997) Toxic phospholipases in snake venom: an introductory review, in Venomous Snakes: Ecology, Evolution and Snakebite (Thorpe, R. S., Wüster, W., and Mulhotra, A., eds.), Oxford University Press, Oxford, UK, pp. 235–250. 16. Heinrikson, R. L. (1982) Structure-function relationships in phospholipases in proteins in biology and medicine, in Biology and Medicine (Bradshaw, R. A., R. L. H. R, Tang, J., Liang, C., Tsao, T. C., and Tsou, C. L., eds.), Academic Press, New York, pp. 132. 17. Danse, J. M., Gasparini, S., and Menez, A. (1997) Molecular biology of snake venom phospholipases A2, in Venom Phospholipase A2 Enzymes (Kini, R. M., ed.), Wiley, Chichester, UK, pp. 29–71.

448

Harris

18. Scott, D. L. (1997) Phospholipases A2 structure and catalytic properties, in Venom Phospholipase A2 Enzymes (Kini, R. M., ed.), Wiley, Chichester, UK, pp. 97–128. 19. Fohlman, J., Eaker, D., Karlsson, E., and Thesleff, S. (1976) Taipoxin, an extremely potent presynaptic neurotoxin from the venom of the Australian taipan (Oxyuranus scutellatus). Isolation, characterization, quaternary structure and pharmacological properties. Eur. J. Biochem. 68, 457–468. 20. Marlas, G. and Bon, C. (1982) Relationship between the pharmacological action of crotoxin and its phospholipase activity. Eur. J. Biochem. 68, 157–165. 21. Kouyoumdjian, J. A., Harris, J. B., and Johnson, M. A. (1986) Muscle necrosis caused by the sub-units of crotoxin. Toxicon 24, 575–583. 22. Kini, R. M. (1997) Phospholipase A2 - a complex multifunctional protein puzzle, in Venom Phospholipase A2 Enzymes (Kini, R. M., ed.), Wiley, Chichester, UK, pp. 1–28. 23. Schmidt, J. J. and Middlebrook, J. L. (1989) Purification, sequencing and characterization of pseudexin isoenzymes. Toxicon 27, 805–818. 24. Deems, R. A. and Dennis, E. A. (1975) Characterization and physical properties of the major form of a phospholipase A2 from cobra venom (Naja naja naja) that has a molecular weight of 11,000. J. Biol. Chem. 250, 9008–9012. 25. Brunie, S., Bolin, J., Gewith, D., and Sigler, P. B. (1985) The refined crystal structure of dimeric phospholipase A2 at 2.5 Å. J. Biol. Chem. 260, 9742–9749. 26. Kondo, K., Toda, H., Narita, K., and Lee, C. Y. (1982) Amino acid sequence of B2-bungarotoxin from Bungarus multicinctus venom. The amino acid substitutions in the B-chains. J. Biochem. 91, 1519–1530. 27. Kondo, K., Toda, H., Narita, K., and Lee, C. Y. (1982) Amino acid sequences of three Beta-bungarotoxins (ß3-, ß4-, and ß5-bungarotoxins from Bungarus multicinctus venom). Animo acid substitutions in the A-chains. J. Biochem. 91, 1531–1548. 28. Faure, G., Guillaume, J. L., Camoin, L., and Bon, C. (1991) Multiplicity of acidic subunit isoforms of crotoxin, the phospholipase A2 neurotoxin from Crotalus durissus terrificus venom, results from post translational modifications. Biochemistry 30, 8074–8083. 29. Faure, G. and Bon, C. (1988) Crotoxin, a phospholipase A2 neurotoxin from the South American rattlesnake Crotalus durissus terrificus: purification of several isoforms and comparison of their modular structure and of their biological activities. Biochemistry 27, 730–738. 30. Faure, G. and Bon, C. (1987) Several isofoms of crotoxin are present in individual venoms from the South American rattlesnake Crotalus durissus terrificus. Toxicon 25, 229–234. 31. Faure, G., Choumet, V., Bouchier, C., Camoin, L., Guillaume, J. L., and Monegier, B. (1994) The origin of the diversity of crotoxin isoforms in the venom of Crotalus durissus terrificus. Eur. J. Biochem. 223, 161–164. 32. Harris, J. B. and Maltin, C. A. 1982. Myotoxic activity of the crude venom and the principal neurotoxin taipoxin, of the Australian taipan (Oxyuranus scutellatus). Br. J. Pharmacol. 76, 61–75. 33. Tyler, M. I., Barnett, D., Nicholson, P., Spence, I., and Howden, M. E. H. (1987) Studies on the subunit structure of textilotoxin, a potent neurotoxin from the venom of the Australian common brown snake (Pseudonaja textilis). Biochim. Biophys. Acta 915, 210–216. 34. Pearson, J. A., Tyler, M. D., Retson, K. V., and Howden, M. E. H. (1993) Studies on the subunit structure of textilotoxin, a potent presynaptic neurotoxin from the venom of the Australian common brown snake (Pseudonaja textilis). 3. The complete amino-acid sequences of all the subunits. Biochim. Biophys. Acta 1161, 223–229. 35. Kini, R. M. (ed.) (1972) Venom Phospholipase A2 Ezymes. Wiley, Chichester, UK. 36. Theakston, R. D. G. (1991) Immunological aspects of snake venom research in Handbook of Natural Toxins, vol. 5 (Tu, A. T., ed.), Marcel Dekker, New York, pp. 495–527. 37. Schiavo, G., Santucci, A., Gupta, B. R. D., Metha, P. P., and Jontes, J. (1993) Botulinum

A2 Neurotoxins

38.

39. 40. 41.

42.

43. 44.

45.

46.

47.

48.

49. 50.

51. 52. 53. 54.

55.

449

neurotoxins serotype A and E cleave SNAP 25 at distinct COOH-terminal peptide bonds. FEBS Lett. 336, 99–103. Davletov, B. A., Shamotienko, O. G., Lelianova, V. G., Grishin, E. V., and Ushkaoyov, V. A. (1996) Isolation and biochemical characterization of a Ca2+-independent alphalatratoxin-binding protein. J. Biol. Chem. 271, 23,239–23,245. Vital-Brazil, O. (1972) Venoms, their inhibitory action on neuromuscular transmission. Intl. Encylop. Pharmacol. 14, 145–167. Howard, B. D. and Gunderson, C. B. (1980) Effects and mechanisms of polypeptide neurotoxins that act presynaptically. Ann. Rev. Pharmacol. Toxicol. 20, 307–336. Strong, P. N. 1987. Presynaptic phospholipase A2 neurotoxins: relationships between biochemical and electrophysiological approaches to the mechanism of toxin action, in Cellular and Molecular Basis of Cholinergic Function (Dowdall, M. J. and Hawthorne, J. N., eds.), Ellis-Horwood, Chichester, UK, pp. 534–549. Hawgood, B. and Bon, C. (1991) Snake venom presynaptic toxins in reptile venoms and toxins, in Handbook of Natural Toxins, vol. 5 (Tu, A. T, Ed.), Marcel Dekker, New York, pp. 3–52. Bon, C., Changeux, J. B., Jeng, T. W., and Fraenkel-Conrat, H. (1979) Postsynaptic effects of crotoxin and of its isolated subunits. Eur. J. Biochem. 99, 471–481. Simpson, L. L., Lautenslager, G. T., Kaiser, I. I., and Middlebrook, J. L. (1993) Identification of the site at which the phospholipase A2 neurotoxins act to produce their neuromuscular blocking effects. Toxicon 31, 13–26. Strong, P. N., Heuser, J. E., and Kelly, R. B. (1977) Selective enzyme hydrolysis of nerve terminal phospholipids by ß-bungarotoxin: biochemical and morphological studies, in Cellular Neurobiology (Hall, Z., Kelly, R. B., and Fox, C. F., eds.), Alan Liss, New York, pp. 227–249. Esquerda, J. E., Solsona, C. W., and Marsal, J. (1982) Binding of Beta-bungarotoxin to Torpedo electric organ synaptosomes. A high resolution autoradiographic study. Neuroscience 7, 751–758. Brenes, F., Gutierezz, J. M., and Lomonte, L. (1997) Immunohistochemical demonstration of the binding of Bothrops asper myotoxin to skeletal muscle sarcolemma. Toxicon 25, 574–577. Dixon, R. W. and Harrris, J. B. (1996) Myotoxic activity of the toxic phospholipase, notexin, from the venom of the Australian tiger snake. J. Neuropath. Exp. Neurol. 55, 1230–1237. Howard, B. D. and Wu, W. C. S. (1976) Evidence that ß-bungarotoxin acts at the exterior of nerve terminals. Brain Res. 103, 190–192. Hawgood, B. J. and Smith, I. C. H. (1989) The importance of phospholipase A2 in the early induction by crotoxin of biphasic changes in the end plate potentials at the frog neuromuscular-junction. Toxicon 27, 272–276. Rehm, H. and Betz, H. (1982) Binding of beta-bungarotoxin to synaptic membrane fractions of chick brain. J. Biol. Chem. 257, 10,015–10,022. Rehm, H. and Betz, H. (1983) Identification by cross-linking of a ß-bungarotoxin binding polypeptide in chick brain membranes. EMBO J. 2, 119–122. Rehm, H. and Betz, H. (1984) Solubilization and characterization of the ß-bungarotoxinbinding protein of chick brain membranes. J. Biol. Chem. 259, 6865–6869. Rehm, H. and Lazdunski, M. (1988) Existence of different populations of dendrotoxin 1 binding protein associated with neuronal K+ channels. Biochem. Biophys. Res. Commun. 153, 231–240. Rehm, H. and Lazdunski, M. (1988) Purification and subunit structure of a putative K+ channel protein identified by its binding properties for dendrotoxin. Proc. Natl. Acad. Sci. USA 85, 4919–4923.

450

Harris

56. Dreyer, F. and Penner, R. (1987) The actions of presynaptic snake toxins on membrane currents of mouse motor nerve terminals. J. Physiol. Lond. 386, 455–463. 57. Rowan, E. G. and Harvey, A. L. (1988) Potassium channel blocking action of ß-bungarotoxin and related toxins on mouse and frog motor nerve terminals. Br. J. Pharmacol. 94, 839847. 58. Harvey, A. L. and Karlsson, E. (1982) Protease inhibitor homologues from mamba venoms: Facilitation of acetylcholine release and interaction with prejuntional blocking toxins. Br. J. Pharmacol. 77, 153–161. 59. Tzeng, M.-C., Hseu, M. J., Yang, J. H., and Guillory, R. J. (1986) Specific binding of three neurotoxins with phospholipase A2 activity to a synaptosomal membrane preparation from the guinea pig brain. J. Prot. Chem. 5, 221–228. 60. Hseu, M. J., Guillory, R. J., and Tzeng, M.-C. (1990) Identification of a crotoxin binding protein in membranes from guinea pig brain by photoaffinity labelling. J. Bioenerg. Biomembr. 22, 39–50. 61. Yen, C.-H. and Tzeng, M.-C. (1991) Identification of a new binding protein on synaptic membranes of the brain for crotoxin and other neurotoxic phospholipase A2. Biochemistry 30, 11,473–11,477. 62. Krizaj, I., Faure, G., Gubensek, F., and Bon, C. (1996) Re-examination of crotoxin-membrane interactions. Toxicon 34, 1003–1009. 63. Radvanyi, F., Saliou, B., Bon, C., and Strong, P. N. (1987) The interaction between the presynaptic phospholipase neurotoxins Beta-bungarotoxin and crotoxin and mixed detergent-phosphatidylcholine micelles. A comparison with non neurotoxic snake venom phospholipases A2. J. Biol. Chem. 262, 8966–8974. 64. Radvanyi, F., Saliou, B., Lembezat, M. P., and Bon, C. (1989) Negatively charged phospholipids, a possible target for crotoxin, a presynaptic phospholipases A2 neurotoxin. J. Neurochem. 53, 1252–1260. 65. Delot, E. and Bon, C. (1993) Model for the interaction of crotoxin, a phospholipase A2 neurotoxin, with presynaptic membranes. Biochemistry 32, 10,708–10,713. 66. Harris, J. B. (1984) Polypeptides from snake venoms which act on nerve and muscle, in Progress in Medicinal Chemistry, vol. 21 (Ellis, G. P. and West, G. B., Eds.), Elsevier, Holland, pp. 63–110. 67. Lambeau, G., Barhanin, J., Schweitz, H., Qar, J., and Lazdunski, M. (1989) Identification and properties of a very high affinity brain membrane binding site for a neurotoxic phospholipase from the taipan venom. J. Biol. Chem. 264, 11,503–11,510. 68. Grandollo, G., Lambeau, G., Lazdunski, M., and Gottesman, C. (1996) Effects on behaviour and EEG of single chain phospholipases A2 from snake and bee venoms injected into rat brain: search for a functional antagonism. Pharmacol. Toxicol. 78, 341–347. 69. Fossier, P., Lambeau, G., Lazdunski, M., and Baux, G. (1995) Inhibition of ACh release at an Aplysia synapse by neurotoxic phospholipases A2: specific receptors and mechanisms of action. J. Physiol. Lond. 489, 29–40. 70. Ritonja, A. and Gubensek, F. 1985. Ammodytoxin A, a highly lethal phospholipase A2 from Vipera ammodytes venom. Biochim. Biophys. Acta 828, 306–312. 71. Ritonja, A., Machleidt, W., Turk, V., and Gubensek, F. (1986) Amino acid sequence of ammodytoxin B partially reveals the location of the site of toxicity of ammodytoxins. Biol. Chem. Hoppe-Seyler 367, 919–923. 72. Pungecar, J., Kordis, D., Strukelj, B., Liang, N. S., and Gubensek, F. (1991) Cloning and nucleotide seqence of a cDNA encoding ammodytoxin A, the most toxic phospholipase A2 from the venom of the long-nosed viper (Vipera ammodytes). Toxicon 29, 269–273. 73. Gubensek, F., and Krizaj, I. 1997. Unpublished data cited in Gubensek, F. and Krizaj, I. (1997). Ammodytoxins (Vipera ammodytes ammodytes), in Guidebook to Protein Toxins and Their Use in Cell Biology (Rappuoli, R. and Montecucco, C., eds.), Oxford University Press, Oxford, UK, pp. 224–226.

A2 Neurotoxins

451

74. Harvey, A. L., Anderson, A. J., and Rowan, E. G. (1992) Potassium channel-blocking toxins from snake venoms and neuromuscular transmission, in Methods in Neurosciences, vol. 8, Neurotoxins (Conn, P. N., ed.), Academic Press, San Diego, CA, pp. 396–407. 75. Lambeau, G., Cupillard, L., and Lazdunski, M. (1997) Membrane receptors for venom phospholipases A2, in Venom Phospholipase A2 Enzymes: Structure, Function and Mechanism (Kini, R. M., ed.), Wiley, Chichester, UK, pp. 389–412. 76. Schlimgen, A. K., Helms, J. A., Vogel, H., and Perin, M. S. (1995) Neuronal pentaxin, a secreted protein with homology to acute phase proteins of the immune system. Neuron 14, 1–20. 77. Dodds, D., Schlimgen, A. K., Lu, S. Y., and Perin, M. S. (1995) Novel reticular calcium binding protein is purified on taipoxin columns. J. Neurochem. 64, 2339–2344. 78. Kamenskaya, M. A. and Thesleff, S. (1974) Neuromuscular blocking action of an isolated toxin from the elapid (Oxyuranus scutellatus). Acta Physiol. Scand. 90, 716–724. 79. Abe, T., Alema, S., and Miledi, R. (1977) Isolation and characterization of pre-synaptically acting neurotoxins from the venom of Bungarus snakes. Eur. J. Biochem. 80, 1–12. 80. Hawgood, B. J. and Santana de Sa, S. (1979) Changes in spontaneous and evoked release of transmitter induced by the crotoxin complex and its component phospholipase A2 at the frog neuromuscular junction. Neuroscience 4, 293–303. 81. Chang, C. C., Chen, T. F., and Lee, C. Y. (1973) Studies on the presynaptic effect of ßbungarotoxin on neuromuscular transmission. J. Pharmacol. Exp. Ther. 184, 339–345. 82. Rowan, E. G., Pemberton, K. E., and Harvey, A. L. (1990) On the blockade of acetylcholine release at mouse motor nerve terminals by ß-bungarotoxin and crotoxin. Br. J. Pharmacol. 100, 301–304. 83. Chang, C. C., Su, M. J., Lee, J. D., and Eaker, D. (1977) Effects of Sr2+ and Mg2+ on the phospholipase A and presynaptic neuromuscular blocking action of ß-bungarotoxin, crotoxin and taipoxin. Naunyn Schmiedebergs Arch. Pharmacol. 299, 155–161. 84. Abe, T. and Miledi, R. (1978) Inhibition of ß-bungarotoxin action by bee venom phospholipase A2. Proc. R. Soc. Lond. B. 200, 225–230. 85. Caratsch, C. G., Maranda, B., Miledi, R., and Strong, P. N. (1981) A further study of the phospholipase-independent action of ß-bungarotoxin at frog endplates. J. Physiol. Lond. 319, 179–191. 86. Caratsch, C. G., Miledi, R., and Strong, P. N. (1985) Influence of divalent cations on the phospholipase-independant action of ß-bungarotoxin at frog neuromuscular junctions. J. Physiol. Lond. 363, 169–179. 87. Harris, J. B. (1991) Phospholipases in snake venoms and their effects on nerve and muscle in snake toxins, in International Encyclopedia of Pharmacology and Therapeutics, Pharmacology and Therapeutics (Harvey, A. L., ed.), Pergamon Press, New York, pp. 91–129. 88. Hawgood, B. R., Smith, I. C. H., and Strong, P. N. (1988) Early induction by crotoxin of biphasic frequency changes and giant miniature end-plate potentials in frog muscle. Br. J. Pharmacol. 94, 765–772. 89. Allerdice, M. T. and Volle, R. L. (1981) The increase in spontaneous transmitter release produced by ß-bungarotoxin and its modification by inorganic ions. J. Pharm. Exp. Therap. 205, 58–68. 90. Petersen, M., Penner, R., Pieneau, F. K., and Dreyer, F. (1986) Beta-bungarotoxin inhibits a non-inactivating potassium current in guinea-pig dorsal root ganglion neurons. Neurosci Lett. 68, 141–145. 91. Bieber, A. L., Mill, J. P., Ziolkowski, C., and Harris, J. (1990) Rattlesnake neurotoxins: their biochemical and biological effects. J. Toxicol. Toxin. Rev. 9, 285–306. 92. Sen, I. and Cooper, J. R. (1978) Similarities of ß-bungarotoxin and phospholipase A2 and their mechanism of action. J. Neurochem. 30, 1369–1375. 93. Nichols, D., Snelling, R., and Dolly, J. O. (1985) Bioenergetic actions of ß-bungarotoxin, dendrotoxin and bee-venom phospholipase A2 on guinea-pig synaptosomes. Biochem. J. 229, 653–662.

452

Harris

94. Rugulo, M., Dolly, J. O., and Nicholls, D. (1986) The mechanism of action of betabungarotoxin at the presynaptic plasma membrane. Biochem. J. 233, 519–523. 95. Delot, E. and Bon, C. (1992) Differential effects of presynaptic phospholipase A neurotoxins on Torpedo synaptosomes. J. Neurochem. 58, 311–319. 96. Yates, S. L., Burns, M., Condrea, E., Ghassemi, A., Shina, R., and Rosenberg, P. (1990) Phospholipid hydrolysis and loss of membrane integrity following treatment of rat brain synaptosomes with ß-bungarotoxin, notexin and Naja naja atra and Naja nigricollis phospholipase A2. Toxicon 28, 939–951. 97. Wernicke, J. F., Oberjat, T., and Howard, B. D. (1974) ß-neurotoxin reduces neurotransmitter storage in brain synapses. J. Neurochem. 22, 781–788. 98. Dowdall, M. J., Fohlman, J. P., and Eaker, D. (1977) Inhibition of high-affinity choline transport in peripheral cholinergic endings by presynaptic snake venom neurotoxins. Nature 269, 700–702. 99. Tse, C. K., Dolly, J. O., and Diniz, C. R. (1980) Effects of ß-bungarotoxin and tityus toxin on accumulation of putative amino acid neurotransmitters by rat cortex synaptosomes. Neuroscience 5, 135–143. 100. Ng, R. H. and Howard, B. D. (1981) De-energisation of nerve terminals by ß-bungarotoxin. Biochemistry 17, 4978–4986. 101. Mollier, P., Brochier, G., and Talarmain, Y. M. G. (1990) The action of notexin from tiger snake venom (Notechis scutatus scutatus) on acetylcholine release and compartmentation in synaptosomes from electric organ of Torpedo marmorata. Toxicon 28, 1039–1052. 102. Fletcher, J. E., Storella, R. J., and Jiang, M. S. (1995) Bovine serum albumin does not completely block synaptosomal cholinergic activities of presynaptically acting snake venom phospholipase A2 enzymes. Toxicon 33, 1051–1060. 103. Dowdall, M. J., Fohlman, J. P., and Watts, A. (1979) Presynaptic action of snake venom neurotoxins on cholinergic systems, in Neurotoxins: Tools in Neurobiology (Ceccarelli, B. and Clementi, F., eds.), Raven, New York, pp. 63–76. 104. Gundersen, C. B., Newton, M. W., and Jenden, D. J. (1980) Beta-bungarotoxin elevates acetylcholine levels in rat diaphragm. Brain Res. 182, 486–490. 105. Gundersen, C. B., Jenden, D. J., and Newton, M. W. (1981) ß-bungarotoxin stimulates the synthesis and accumulation of acetyl choline in rat phrenic nerve diaphragm preparations. J. Physiol. Lond. 310, 13–35. 106. Tucek, S. (1984) Problems in the organisation and control of aceylcholine synthesis in brain neurones. Prog. Biophys. Mol. Biol. 4, 1–46. 107. Sen, I., Baba, A., Schulz, R. A., and Cooper, J. R. (1978) Effects of Beta-bungarotoxin and Naja naja citra snake venom phospholipase A2 on acetyl choline release and choline uptake in synaptosomes. Toxicon 24, 91–99. 108. Fletcher, J. E. and Middlebrook, J. L. (1986) Effects of beta-bungarotoxin and Naja naja citra snake venom phospholipase A2 on acetylcholine release and choline uptake in synaptosomes. Toxicon 24, 91–99. 109. Gundersen, C. B. and Jenden, D. J. (1981) Notexin preferentially inhibits the release of newly synthesized acetylcholine from rat brain synaptosomal fractions. J. Neursci. 1, 1113–1116. 110. Sen, I., Grantham, P. A., and Cooper, J. R. (1976) Mechanisms of action of ß-bungarotoxin on synaptosomal preparations. Proc. Natl. Acad. Sci. USA 73, 2664–2668. 111. Smith, C. C. T., Bradford, H. F., Thompson, E. J., and MacDermot, J. (1980) Actions of ß-bungarotoxin on amino acid transmitter release. J. Neurochem. 34. 487–494. 112. Jiang, M. S., Haggblad, J., Heilbronn, E., Rydgvist, B., and Eaker, D. (19871) Some biochemical characteristics and cell membrane actions of a toxic phospholipase A2 isolated from the venom of the pit viper Agkistrodon halys (Pallas). Toxicon 25, 785–792.

A2 Neurotoxins

453

113. Fletcher, J. E. and Jiang, M. S. (1995) Presynaptically acting snake venom phospholipase A2 enzymes attack unique substrates. Toxicon 33, 1565–1576. 114. Chapell, R. and Rosenberg, P. (1992) Specificity of action of ß-bungarotoxin on acetylcholine release from synaptosomes. Toxicon 30, 621633. 115. Yates, S. L. and Rosenberg, P. (1991) Comparative effects of phospholipase A2 neurotoxins and enzymes on membrane potential and Na+/K+ ATPase activity of rat brain synpatosomes. Toxicol. Appl. Pharmacol. 109, 207–218. 116. Fletcher, J. E. and Rosenberg, P. (1997) The cellular effects and mechanisms of action of presynapticaly acting phospholipase A2 toxins, in Venom Phospholipases A2 Enzymes: Structure, Function and Mechanism (Kini, R. M, ed.), Wiley, Chichester, UK, pp. 413– 454. 117. Augustine, G. J., Burns, M. E., Bello, W. M. D., Hilfiker, S., Morgan, J. R., Schweizer, F. E., et al. (1999) Proteins involved in synaptic vesicle trafficking. J. Physiol. Lond. 520, 33–41. 118. Trevett, A. J., Lalloo, D. G., Nwokolo, N. C., Naraqi, S., Kevan, I. H., Theakston, R. D. G., and Warrell, D. A. (1995) Failure of 3-4 diamonipyridine and edrophonium to produce significant clinical benefit in neurotoxicity following the bite of the Papua taipan (Oxyuranus scutellatus canni). Trans. R. Soc. Trop. Med. Hyg. 89, 444–446. 119. Trevett, A. J., Lalloo, D. G., Nwokolo, N. C., Naraqi, S., Kevan, I. H., Theakson, R. D. G., and Warell, D. A. (1995) The efficacy of antivenom in the treatment of bites by the Papuan taipan (Oxyuranus scutellatus canni). Trans. R. Soc. Trop. Med. Hyg. 89, 322–325. 120. Trevett, A. J., Lalloo, D. G., Nwokolo, N. C., Naraqi, S., Kevan, I. H., Theakston, R. D. G., and Warrell, D. A. (1995) Electrophysiological findings in patients envenomed following the bite of a Papuan taipan (Oxyuranus scutellatus canni). Trans. R. Soc. Trop. Med. Hyg. 89, 415–417. 121. Connolly, S., Trevett, A. J., Nwokolo, N. C., Lalloo, D. G., Naraqi, S., Mantle, D., et al. (1995) Neuromuscular effects of Papuan Taipan snake venom. Ann. Neurol. 88, 916–920. 122. Abe, T., Limbrick, A. R., and Miledi, R. (1976) Acute muscle denervation induced by Beta-bungarotoxin. Proc. R. Soc. B. 194, 545–553. 123. Landon, D. N., Westgaard, R. H., MacDermot, J., and Thompson, E. J. (1980) The morphology of rat soleus meuromuscular junctions treated in vitro with purified Betabungarotoxin. Brain Res. 202, 1–20. 124. Hille, B., Billiard, J., Babcock, D. F., Nguyen, T., and Koh, D. S. (1999) Stimulation of exocytosis without a Ca2+ signal. J. Physiol. Lond. 520, 23–31. 125. Ng, R. H. and Howard, B. D. (1980) Mitochondria and sarcoplasmic reticulum as model targets for neurotoxic and myotoxic phospholipases A2. Proc. Natl. Acad. Sci. USA 77, 1346–1350. 126. Noremberg, K. and Parsons, S. M. (1986) Selectivity and regulation in phospholipase A2 mediated attack on cholinergic vesicles by ß-bungarotoxin. J. Neurochem. 47, 1312–1317. 127. Dixon, R. W. and Harris, J. B. (1999) Nerve terminal damage by ß-bungarotoxin - its clinical significance. Am. J. Pathol. 154, 447–455. 128. Harris, J. B., Grubb, B. D., Maltin, C. A., and Dixon, R. (2000) The neurotoxicity of the venom phospholipases A2, notexin and taipoxin. Exp. Neurol. 161, 517–526. 129. Grubb, B. D., Harris, J. B., and Schoffield, I. S. 1991. Neuromuscular transmission at newly formed neuromuscular junctions in the regenerating soleus muscle of the rat. J. Physiol. Lond. 441, 405–421. 130. Davis, C. E., Harris, J. B., and Nicholson, L. V. B. (1991) Myosin isoform transitions and physiological properties of regenerated and re-innervated soleus muscles of the rat. Neuromusc. Disord. 1, 411–421. 131. Hirokawa, N. (1978) Characterizationof various nervous tissues of the chick embryo through responses to chronic app.lication and immunocytochemistry of Betabungarotoxin. J. Comp. Neurol. 180, 449–466.

454

Harris

132. Pittman, R., Oppenheim, R. W., and Wang, I. C. (1978) ß-bungarotoxin induced neuronal degeneration in the chick embryo spinal cord. Brain Res. 153, 199–204. 133. Henley, M. R. and Emson, P. C. (1979) Neuronal degeneration induced by stereotaxic injection of ß-bungarotoxin into rat brain. Neurosci. Lett. 11, 143–148. 134. Olek, A. J. (1980) Effects of α- and ß-bungarotoxin on motor neuron loss in Zenopus larvae. Neuroscience 5, 1557–1563. 135. Rehm, H., Schafer, T., and Betz, H. (1982) ß-bungarotoxin-induced cell-death in neurons in chick retina. Brain Res. 250, 309–319. 136. Harris, J. B. and Cullen, M. J. (1990) Muscle necrosis caused by snake venoms and toxins. Electron Microsc. Rev. 3, 183–211. 137. Mebs, D. and Ownby, C. L. (1990) Myotoxic components of snake venoms: their biochemical and biological activities. Pharmacol. Ther. 48, 223–235. 138. Fletcher, J. E., Arranjo, H. S. S. D., and Ownby, C. L. (1997) Molecular events in the myotoxic action of phospholipases, in Venom Phospholipase A2 Enzymes: Structure, Function and Mechanisms (Kini, R. M., Ed.), Wiley, Chichester, UK, pp. 455–497. 139. Vater, R., Cullen, M. J., and Harris, J. B. (1992) The fate of dystrophin during the degeneration and regeneration of the soleus muscle of the rat. Acta Neuropathol. 83, 140–148. 140. Vater, R., Cullen, M. J., and Harris, J. B. (1992) The fate of desmin and titin during the degeneration and regeneration of the soleus muscle of the rat. Acta Neuropathol. 84, 278–288.

Dendrotoxins

455

22 Dendrotoxins from Mamba Snakes J. Oliver Dolly and Giacinto Bagetta 1. INTRODUCTION Dendrotoxins (DTXs) are a family of homologs, single-chain, basic polypeptides from mamba snake venoms that exibit potent convulsant activity in experimental animals. Since the discovery that α-DTX and its homolog toxin I could attenuate an A-like K+ current in hippocampal pyramidal neurones (1,2), these neurotoxins have been shown to be even more potent in blocking slowly inactivating variants (3,4) in a variety of neuronal preparations (reviewed in refs. 5–7). Application of these and other members of the DTX family as tools in neurobiology has been instrumental for the molecular identification, localization, purification, and characterization of one subfamily (Kv) of K+ channel proteins. Moreover, the complementary approach of exploiting the IA channel gene of the Drosophila Shaker mutant for cross-hybridization, cloning, and expression (8,9), in conjunction with electrophysiological studies, provided further detailed information on the structural/functional relations of the channels’ α subunits (reviewed in refs. 10,11). More recently, dendrotoxin homologues have been successfully employed in neurodegenerative studies carried out in intact animals where, again, their differential inhibition of subsets of K+ currents has aided the elucidation of different pathophysiological mechanisms underlying neuronal death. These studies have opened new avenues of research towards the development of novel therapeutic strategies for neuroprotection. 2. STRUCTURAL CHARACTERISTICS A number of neurotoxins have been purified from the venoms of the green (Dendroaspis angusticeps) and black (D. polylepis) mamba snakes (see ref. 11) that are known to facilitate transmitter release at numerous peripheral (12–15) and central (16,17; reviewed in refs. 5,18) synapses. These single-chain, basic polypeptides with Mr ~ 7 kDa (57–60 residues) contain 6 cysteines that form 3 disulphide bonds. All the toxins show significant sequence homology to Kunitz protease inhibitors (e.g., bovine pancreatic trypsin inhibitor [BPTI]; Table 1), though the most potent neurotoxins


455

456

Table 1 Amino Acid Sequences of α-DTX and Homologs

Dolly and Bagetta

α-,δ-DTX, from Dendroaspis angusticeps; toxin I, K, B, and D from D. polylepis; β1-BuTX, B-chain of β1-bungarotoxin, from Bungarus multicintus; BPTI, bovine pancreatic trypsin inhibitor. Z denotes L-pyrrolidone glutamic acid. Residues are numbered according to the sequence of α-DTX, the asterisk indicates the reactive site residues for protease inhibition. Disulphide bonds link cysteine residues at positions 7–57, 16–40, and 32–53. See ref. 80 for references. Adapted with Permission from ref. 80.

Dendrotoxins

457

exhibit negligible inhibitory activity. Conversely, the nontoxic homolog E does inhibit serine proteases, whereas B lies between the two extremes, being devoid of both activities (19). Note that the C-terminal portion of all the neurotoxic homologs (α- and δ-DTX, toxin I and DTX-K) is virtually identical with dissimilarities existing in their N-terminal regions (Table l); this also applies to other congeners from D. angusticeps (β- and γ-DTX; 20) and D. viridis (D.v-14; see ref. 21). Interestingly, noticeable homology (including the positioning of the cysteines) exists between the mamba neurotoxins and the smaller B-chain of β-bungarotoxin (β1-BuTX); the latter occurs disulphide-linked to a phospholipase A2 chain. Notably, toxin I has been shown to totally inhibit the binding of [3H] β-BuTX to its neuronal acceptors in brain cortical synaptosomal preparations of rat (22); in addition, this di-chain protein from Bungarus multicinctus has been shown to inhibit certain voltage-activated K+ currents in peripheral neurones (23–24). Although the three-dimensional structure of BPTI has been known for some time (25), the high-resolution structure of α-DTX (26) and toxin I and DTX-K (27,28) was solved more recently by X-ray diffraction of protein crystals (α-DTX) and by nuclear magnetic resonance (NMR) spectroscopy (DTX-K and I), respectively. The schematic of the backbone secondary structure is similar for all these toxins (see ref. 11) and consists of an N-terminal region either with little structure (α-DTX) or containing a short 310-helix (toxin I and DTX-K), a loop responsible (in BPTI) for antiprotease activity, a twisted, double-stranded antiparallel ß sheet, and, at the C-terminus, a short α-helix. Much interest has focused upon the ß turn between the two ß strands, since this element contains a region conserved between the toxins that has two (DTX-K, δ-DTX) or three (α-DTX, toxin I) basic lysine residues and may be responsible for their K+ channel-blocking activity (reviewed in ref. 29). Cloning and functional expression of DTX-K cDNA (30) have allowed site-directed mutagenesis experiments to be conducted in conjunction with binding studies. These have demonstrated that positively charged residues in the ß turn and those (K6 and K3) in the 310-helix are functionally important for the high-affinity binding of DTX-K to neuronal K+ channels from the brain cortex of rat (31). On the other hand, mutation of two positively charged (R52A and R53A) residues in the α-helix region at the C-terminus had little influence on the binding properties of DTX-K (31). A report (32) in which the three lysines were substituted in a recombinant α-DTX appears to contradict the notion that these residues may be responsible for K+ channel-blocking activity (reviewed in ref. 33). Site-directed mutagenesis experiments have indicated that the functional site for α-DTX includes six major binding residues, all located in its N-terminal region, with Lys5 and Leu9 being the most important (34). Quantitative differences in binding affinity to K+ channels yielded by mutation of positively charged residues in the ß turn of α-DTX (32) and DTX-K (31) probably reflect their subtle structural dissimilarity (26,27) which also may account, in part, for their distinct specificities for different K+-channel subtypes (see below). More recently, the 310-helical N-terminal region of DTX-K was found to be responsible for recognition of Kv 1.1 (component of the Shaker-related, Kv 1.x, subfamily) channels because mutation of K3A led to 1246-fold reduction in the inhibitory potency for [125I]DTX-K binding and a large decrease in its ability to block the Kv 1.1 current; the effect of this substitution on the affinity of DTX-K for Kv 1.2-possessing oligomers was much less dramatic (16-fold) (35). A schematic representation of a

458

Dolly and Bagetta

Fig. 1. A speculative model to explain the experimentally observed patterns of DTX-K and α-DTX to neuronal K+ channels postulated to contain different contents of Kv 1.1 and 1.2 α-subunits. (upper panel) Analysis of numerous mutants suggest that [125I]DTX-K binds strongly (or exerts antagonism in the case of unlabeled DTX-K) via identified residues (e.g., K26, and K3) to channels prposed to contain at least two Kv 1.1 subunits (group A and B) could represent Kv 1.1 or Kv 1.4 in A and Kv 1.1, 1.2 and 1.6 in B according to the data on separated K+ channels (64); unlabeled DTX-K antagonizes this binding with high affinity. It is hypothesized that [125I]αDTX binds with high afinity to all Kv 1.2 containing channels, which are divided in B and C according to their nominal Kv 1.1 content. Our binding data are most readily explained by postulating that DTX-K binds via K3 and K26 to the two copies of Kv 1.1 in group B channels with high affinity (KiA), as reflected by its potent antagonism of [125I]αDTX binding to the Kv 1.2 constituent. In the case of group C oligomers, DTX-K interacts through K26 with the only copy of Kv 1.1 subunit, leaving K3 to attach weakly to Kv 1.2 or othe subunits (another copy of Kv 1.2 or 1.6) yielding overall affinity competition of [125I]αDTX binding (KiB). (lower panel) Residues of the rat Kv 1.1 known to form part of a DTX acceptor site (see ref. 35), whose location seems to coincide with the “turret” described in the X-ray structure of a bacterial K+ channel (see ref. 35). All of these amino acids are different in Kv 1.2 (as well as Kv 1.3, 1.4, and 1.6, which do not bind DTX-K) possibly explaining the toxin’s decreasing affinity for group C channels observed experimentally. Adapted with permission from ref. 35.

hypothetical mode of interaction of α-DTX and DTX-K with neuronal K+ channels is shown in Fig. 1. In the structure of α-DTX, tyrosine 17 sticks out of the top of the molecule; its exposed position makes this the likely site of the toxin’s radioiodination. In this case, tyrosine 17 must not be essential for biological activity because the 125I-labeled α-DTX retains toxicity and ability to bind to K+ channels (36); consistently, in two neurotoxic homologs (δ-DTX and DTX-K) lysine occupies this position. 3. DENDROTOXIN HOMOLOGS ARE SELECTIVE BLOCKERS OF SUBSETS OF FAST-ACTIVATING, VOLTAGE-GATED K+ CHANNELS The first K+ channel identified as a target for α-DTX and toxin I was that giving rise to a fast-activating, rapidly inactivating “A” current of rat hippocampal neurones (1,2). Neurons in rodent superior cervical or nodose (C-cells) ganglia possess transient IA

Dendrotoxins

459

currents that are not susceptible to α-DTX, whereas nodose A cells and dorsal-root ganglion neurones display slowly inactivating, α-DTX sensitive K+ current variants (3,4,37). In nodes of Ranvier of frog sciatic nerves, α-DTX, like toxin I, preferentially inhibits a slowly inactivating component of the K+ current (IKFI) though with a potency that is about 25-fold less (IC50 of 0.4 nM; 38,39). Thus, α-DTX preferentially blocks a family of neuronal K+ currents with slow inactivation rates but relatively fast activation kinetics. In agreement with this concept, low concentrations [1 nM] of α-DTX have been shown to block preferentially the slowly inactivating component of a voltage dependent K+ current recorded from cultured sensory neurones from neonatal rats, using the whole-cell voltage-clamp technique (40). Under these experimental conditions, δ-DTX inhibited more selectively the noninactivating component of the native K+ current because even at [10 nM] inhibition by δ-DTX was mostly restricted to the noninactivating component, whereas in the case of α-DTX the two K+ current components were inhibited equally (Fig. 2) (40). Cloning and stable expression of human and rat Kv1.1 channels have allowed the relative potency of various dendrotoxin homologs as blockers of the corresponding underlying currents to be established. The results of these studies demonstrate that the potency of α-DTX homologs, DTX-K, and δ-DTX depends on the species from which the Kv 1.x subunit is cloned. Thus, δ-DTX (and, similarly, DTX-K) is much more potent than α-DTX and toxin I in blocking human Kv.1.1 expressed in mammalian cells (IC50 = 3 pM), whereas the rat form is considerably less susceptible (see ref. 11). α-DTX preferentially blocks Kv 1.2, 1.1, and 1.6 of mammalian-cloned K+ channel subunits expressed in Xenopus oocytes and vertebrate cells. 4. DENDROTOXINS AS TOOLS TO LOCALIZE K+ CHANNELS Biologically active, radioiodinated α-DTX binds to an apparently homogeneous set of sites in cerebrocortical synaptosomes from rat (36), bovine (41), and guinea-pig (42) with an affinity of 0.1–0.5 nM. In accord with their selective blockade of at least two fast-activating, voltage-gated + K currents in peripheral sensory neurons, 125I-labeled α- and δ-DTX were found to bind with high affinity to K+ channels in rodent synaptosomal membranes (Table 2); notice that the site content is lower for δ-DTX. Although it has not been possible to obtain direct evidence for α-DTX being able to discriminate subtypes of K+ channels, δ-DTX competition of 125I-α-DTX binding to membranes from guinea-pig cerebral cortex revealed heterogeneity in these channels. Using the Grafit program to analyze the data, the competition curve was best fitted to two sites, with δ-DTX showing KI values of 0.47 (± 0.35) and 45 (± 30) nM. The binding constant of δ-DTX for the higher-affinity site approximates to the KD determined for 125I-δ-DTX (Table 2). Direct detection of the second low-affinity site using 125I-δ-DTX was precluded owing to its high level of nonsaturable binding relative to that observed with 125I-α-DTX. Further evidence for multiple K+ channel forms has been obtained by demonstrating their characteristic distributions in rat brain using the same toxin probes (43). As first reported for 125I-labeled β-BuTX (44) and α-DTX (45), sheet-film autoradiographic analysis of brain cryostat sections labeled with 125I-δ-DTX demonstrated a widespread occurrence of saturable acceptors (Fig. 3). Notably, the sites for all three toxins showed distinct

460

Dolly and Bagetta

Fig. 2. α- and δ-DTX block selective components of total outward K+ current in sensory neurons. Total voltage-activated, K+ current (i) was evoked with a voltage step from a conditioning prepulse of –100 mV to +60 mV (as above). Controls are indicated by C throughout. (A) (i) 1 nM α-DTX blocked a proportion of the total current. K+ current activated from –40 mV (see [iii]) to +60 mV was subtracted from the total outward current (i) to reveal a slowly inactivating component of voltage-activated current (ii). This component was sensitive to inhibition by 1 nM α-DTX. In (iii), current evoked from –40 mV to +60 mV was largely insensitive to block by 1 nM α-DTX. (B) (i) 1 nM δ-DTX blocked the total outward current in a characteristic time-dependent manner. in contast to α-DTX, the current activated from –40 mV (iii) was reduced by δ-DTX. Note that the apparent increase in the slowly inactivating current (ii) in the presence of δ-DTX is a result of the time-dependence of the block, and the subtraction procedure. Adapted with permission from ref. 80.

location patterns, though variable extents of colocalization were apparent. Consistent with the toxins’ site contents, high-affinity sites for 125I-β-BuTX were most restricted, being very prominent in the dentate gyrus and molecular layer of the hippocampus (Fig. 3E). Occurrence of acceptors for the other two toxins was more extensive, with overlapping distributions in most areas (Fig. 3A,C). However, notable exceptions

Dendrotoxins

461

Fig. 3. Characteristic distribution patterns of sites in rat hippocampus labeled with radioiodinated δ-DTX, α-DTX and β-BuTX. Cryostat sections were incubated with the 125Ilabeled δ-DTX (A), α-DTX (C), or β-BuTx (E) alone with the inclusion of 1 µM toxin I (B) or δ-DTX (D,F), respectively. Sections were washed and processed for 3H-Ultrofilm autoradiography. CC, corpus callosum; G, Granule cell layer; H, hilus; LM, stratum lacunosum moleculare; M, statum moleculare; Or, statum oriens; P, pyramidal cell layer; R, stratum radiatum; CA1, CA2, and CA3 define the subfield of hippocampus, whereas DG identifies the denate gyrus. Scale bar (in A) = 1 mm. Adapted with permission from ref. 80.

462

Dolly and Bagetta

Table 2 Parameter Estimates for α-DTX, βBuTX, and δ-DTX Binding in Rat and Guinea-Pig Cerebrocortical Synaptosomal Membranes

Rat Guinea-pig

Toxin

Bmax (pmol/mg protein)

KD (nM)

α-DTX ß-BuTX α-DTX δ-DTX

1.4–1.7 0.4 0.26 + 0.02 0.23 + 0.04 (0.16 + 0.05)

0.3 0.4 0.23 + 0.02 0.30 + 0.09 (0.17 + 0.09)

The binding of varying concentrations of 125I-α-DTX and 125I-β-BuTX to rat synaptosomal membranes was determined in Krebs/phosphate buffer and of 125I-α- and δ-DTX to guinea-pig synaptosomal membranes in TES-buffer (see ref. 80) All values for Bmax and KD were obtained by graphical analysis of Skatchard plots of specific binding, except for values in parentheses (for δ-DTX) which were derived from the weighted, least-squared, nonlinear regression curve-fitting program (see ref. 80). The latter data were best fitted to a one-site model. Adapted with permission from ref. 80.

include the lacunosum molecular layer of the hippocampus and Purkinje-cell layer of the cerebellar cortex (43), which were devoid of δ-DTX sites, though labeled clearly with α-DTX and β-BuTX (Fig. 3A–C). The absence of high-affinity δ-DTX sites in this particular hippocampal area was confirmed by the inability of δ-DTX to block the binding therein of 125I-α-DTX or 125I-β-BuTX, though extensive competition was seen, to variable extents, in all other regions of the sections (Fig. 3D, F). One possible interpretation of these interesting results is that δ-DTX at low concentrations (e.g., 0.5– 1.0 nM used for autoradiographic and electrophysiological experiments) labels predominantly one subtype of α-DTX acceptor; hence, the noted extensive codistribution and mutual antagonism. Additionally, δ-DTX displays weak affinity for a second population of α-DTX (and β-BuTX) acceptors present in the hippocampal layers specified earlier, thereby explaining the inability of δ-DTX to block these sites (except at very high doses > 1 µM). Such a proposal could be reconcilable with the preferential blockade by 1 nM δ-DTX of a sustained K+ current in sensory neurons whereas much higher doses are required to abolish the slowly inactivating K+ current component. α-DTX has the opposite preference though less clear-cut, as expected from its affinity for both K+ channel subtypes being quite similar (thus, the two acceptors can only be detected by blockade of α-DTX binding with δ-DTX). 5. DENDROTOXINS AS TOOLS TO PROBE K+ CHANNELS SUBUNIT AND OLIGOMERIC STRUCTURES IN THE BRAIN α-DTX and the closely related toxin I have been most useful in probing the structures of the K+ channel proteins. The first identification of a component of native K+ channels was achieved by covalent crosslinking of 125I-α-DTX to its binding protein in rat brain synaptosomes, with subsequent separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (46). This labeled a component of about 65,000 Mr, after allowance for the size contribution of one bound toxin molecule. Bands of similar size were later observed in rat brain synaptosomes using ß-BuTX (47) and mast cell degranulating peptide (MCPD) (48) as well as a much smaller one (28,000

Dendrotoxins

463

Mr) which was labelled only by ß-BuTX. In chick brain, two proteins of Mr 75,000 and 69,000 were crosslinked to 125I-α-DTX (36); labeling of both could be prevented by cold α-DTX and ß-BuTX. Similarly, two components (Mr = 74,000 and 97,000) were also identified in guinea-pig brain synaptosomes using 125I-δ-DTX as the probe; although labeling of both bands could be inhibited by α-DTX, the latter appeared to bind to the smaller component (35,42). Further studies using toxin I and α-DTX have allowed the purification of K+ channel proteins from the bovine (41) and rat (49) brain, respectively. Analysis of purified, radiolabeled acceptors on sucrose density gradients showed that most contain both α- and ß-subunits. However, additionally, a minor species was observed that possessed only α-subunits. The detergent-corrected Mr of this minor form was found to be 250 kDa (50), which can most likely be explained by an α4 stoichiometry. Furthermore, the difference in size between the smaller and the larger (Mr 400 kDa) oligomeric form is consistent with the presence of 4 ß-subunits. Thus, it appears that the vast majority of purified α-DTX K+ channels have an α4/β4 oligomeric structure. The successful expression of individual cloned α subunits yielding voltage-sensitive K+ currents was presumed to result from formation of tetramers, by analogy with Ca2+ and Na+ channels (reviewed in ref. 51) and from the elegant mixing measurements by MacKinnon (10,52). However, co-expression of two different α genes gave K+ channels with distinct characteristics (53–55). Considering the evidence emerging from biochemical studies on the α-DTX sensitive K+ channel proteins (outlined below), it seems likely that hetero-multimers are responsible for the K+ currents recorded in neuronal membranes, an idea (56) that is now gaining support from interesting results on “hybrid” α-subunit K+ channels. Furthermore, the protein chemistry experiments showed that not only are K+ channel subtypes composed of a combination of different α isoforms, but most oligomeric forms also contain multiple copies of β subunit. Indeed, X-ray crystallography at 2.8 Å resolution of the conserved core of the mammalian ß subunits has established that these form a fourfold symmetric structure where each subunit is an oxyreductase complete with a nicotinamide cofactor in its active site, suggesting that this may interact with the K+ channel’s voltage sensor (57). The (α)4 (ß)4 stoichiometry observed for the naturally occurring channels is consistent with indirect studies of K+ channels expressed in Xenopus oocytes after injection of cRNAs encoding α subunits only of either Drosophila Shaker or mammalian Kv1.1 (Shaker-related) proteins (see ref. 11). Co-injection of cRNAs encoding different α-subunit leads to the generation of K+ currents with properties intermediate between those produced by each α subunit cRNA when injected alone, indicating the formation of hetero-oligomeric K+ channel tetramers (see ref. 11 for details). The primary determinants for the specificity of assembly of K+ channel α subunits appear to lie in the cytosolic N-terminal region (see ref. 11). Evidence for the formation of such hetero-oligomers in the brain was sought using antibodies specific for each of the α-subunits (58). A number of α-subunit variants are detected in the purified preparation of α-DTX-sensitive K+ channels (Fig. 4A). The presence of Kv 1.1, 1.2 (also detected by direct N-terminal sequencing), and 1.6 was not surprising since α-DTX-sensitive K+ currents are obtained when their cRNA is injected in Xenopus oocytes. Importantly, however, Kv 1.4, which gave rise to α-DTX

464

Dolly and Bagetta

Fig. 4. Immunoprecipitation and Western Blotting of α-DTX sensitive K+ channels with α-subunit specific antibodies: demonstration of oligomeric subtypes. (A) K+ channels purified by toxin affinity chromatography (see ref. 11) were separated by SDS-PAGE, electroblotted onto PVDF membrane, probed with α-subunit specific antibodies, and visualized after addition anti-species antibody coupled to alkaline phosphatase. Lane 1, anti-Kv 1.1 antibody; 2, anti-Kv 1.6; 3, anti-Kv 1.4; 4, anti-Kv 1.2 (rabbit polyclonal); 5, anti-Kv 1.2 (mouse monoclonal antibody). The positions of molecular mass markers are also shown. (B) α-DTX sensitive K+ channels were labeled in crude detergent extracts by incubation with 125I-labeled α-DTX. Anti-α subunit antibodies were added, followed by anti-IgG-sepharose. After centrifugation and washing to remove unbound 125I-α-DTX, pellets were γ-counted to assess the amount of K+ channeltoxin complex precipitated. Data are expressed relative to the 125I-α-DTX binding in crude extract. Adapted with permission from ref. 11.

insensitive K+ currents when expressed, was also observed. This could only have been adsorbed and eluted from the toxin I column if it was in association with α-DTXsensitive α-subunit in a hetero-oligomer. Further analysis of quantitative immunoprecipitation (Fig. 4B) and by immunoprecipitation followed with Western blotting (58) showed that the pure α-DTX K+ channels contain a number of oligomeric subtypes. All the oligomers contained Kv 1.2, confirming N-terminal sequencing where it gave the most abundant signal. Approximately 50% of these oligomers also possess Kv 1.1, while minor populations have Kv 1.6 (approx 20%) or Kv 1.4 (approx 10%). These data appear to agree well with demonstrated overlapping distributions of Kv 1.2 and Kv 1.1 in the juxta-paranodal regions of nodes of ranvier in mouse brain stem and in the terminal fields of the basket cells in mouse cerebellum (59,60) and of Kv 1.2 and 1.4 in the middle third of the dentate gyrus, molecular layer, and stratum lacunosum moleculare of rat hippocampus (61), which may indicate hetero-multimeric complexes. Recently, sequential immunoprecipitation with specific anti-Kv1 antibodies identified conclusively one fully defined tetramer Kv 1.2/1.3/1.4/1.6, and several other pos-

Dendrotoxins

465

sible multimeric channels in synaptic membranes from bovine cerebral cortex (62). From the aforementioned studies, it has emerged that only a small fraction of the possible oligomeric subtypes is actually present. In particular, all but one (Kv 1.4 putative homo-oligomer) of the oligomers identified contained at least one copy of Kv 1.2, the α subunit found to be the prominent costituent of K+ channels purified from bovine brain using toxin I (63), and ß2 subunit (62); interestingly, Kv 1.2 has also been shown to occur in the absence of these other subunits as a putative homo-oligomer (62). More recent solubilization and immunoprecipitation experiments of channel subunits covalently bound to [125I]DTX-K allowed, by affinity chromatography on immobilized DTX-K and toxin I, the purification of K+ channel subtypes. In particular, these experiments unveiled two types of Kv 1.1-containing channels: Kv1.1/1.4 and Kv1.1/1.2 coassembled with or without Kv 1.6 (64). 6. EXPLOITATION OF DENDROTOXINS IN NEURODEGENERATIVE STUDIES It is well documented that dendrotoxins are potent convulsants in experimental animals, both for systemic and intracerebral administration (see ref. 46). The epileptic syndrome elicited by systemic administration of α-DTX is characterized by complex behavioral symptoms and electrographic seizure activity in limbic structures of the brain (65), resembling amygdaloid epilepsy elicited by kainic acid in rat (see ref. 66). Focal injection into one dorsal hipocampus of minute quantities of α-DTX into rats has been shown to produce motor and electrocortical (ECoG) seizures accompanied at 24 h by multifocal damage to the hippocampal (CA1, CA3, and CA4 areas) formation (67). Antagonism studies, in which selective N-methyl-D-aspartate (NMDA) (e.g., MK801 and CGP37849) and non-NMDA (e.g., NBQX and GYKI52466) receptor blockers have been used, demonstrate that excitatory, glutamate-mediated neurotransmission does not play a mayor role in the mechanisms that trigger α-DTX-induced seizures and hippocampal damage (67,68). In agreement with the aforementioned deduction, it has been reported that NMDA and non-NMDA receptor antagonists are ineffective against seizures induced by α-DTX in mice (69). It is well known that the hippocampus represents a three-synaptic feed-forward excitatory loop (70,71) with the peculiar characteristic of rapid neuronal recruitment and synchronization, firing bursts of action potentials (72) and exhibiting high susceptibility to neurodegenerative insults (see ref. 73). These may, in part, be a result of intrinsic characteristics of the membrane of hippocampal neurones and to the robust, glutamatemediated, excitatory innervation that the CA1 pyramids receive from the enthorinal cortex, through the perforant and Schaffer collateral pathways (70,72). Therefore, to establish more conclusively the role of excitatory transmission in the mechanism of seizures and damage induced to the hippocampus by α-DTX, this has been studied in rats bearing a monolateral surgical lesion of the Schaffer collaterals, in order to eliminate the possible influence of glutamate released by the toxin from the Schaffer collateral terminals. Under these experimental conditions, treatment with α-DTX induced seizures and damage to the CA1 (site of α-DTX injection) and CA4 (area distant to the site of injection) hippocampal areas leaving unaffected the CA3 region (anatomically and functionally isolated by the lesion). Silver staining of lesioned hippocampal sec-

466

Dolly and Bagetta

Fig. 5. Photomicrographs of rat hippocampal sectors showing the neurodegenerative effects of DTX-K administered into the CA1 region. Rats were injected stereotaxically with either 105 pmol of BSA (A) or 35 pmol of DTX-K (B) into one side (T) of hippocampal formation; 24 h later, 10-µm coronal sections were cut and stained with Cresyl Fast Violet, as detailed previously (see ref. 68). Note the neuronal loss that occurred in the CA1 pyramidal cell layer injected with toxin (T) and the contralateral side (C) whereas the samples treated with BSA (A; T and C) appeared normal. Adapted with permission from ref. 68.

tions (74) and [125I]-α-DTX binding analysis (75) have shown that motor and ECoG seizures and CA1 neuronal loss induced by α-DTX occurred in the absence of CA3 excitatory, glutamatergic afferents to the dorsal hippocampus. Together, the antagonism and lesion studies suggest that K+ channel blockade by α-DTX increases the excitability of CA1 pyramids and this is responsible for triggering seizures and CA1 pyramidal cell death. Epileptogenic and neurodegenerative effects similar to those described above have been recently reported in rats injected with DTX-K (68), though neuronal damage was bilateral (Fig. 5) rather than restricted to the side of injection as seen with α-DTX (67); another notable difference is the more extensive cell loss that resulted from injection of DTX-K with respect to the effect of α-DTX (68). More importantly, at variance with the findings of the α-DTX study, systemic administration of NMDA and non-NMDA receptor antagonists prevented seizures (Fig. 6) and CA1 hippocampal damage (67,76); similarly, seizures and hippocampal damage (Fig. 7) induced by DTX-K, but not α-DTX, have been reduced by systemic administration of U-74389G (76), a free radical scavenger of the 21–aminosteroid group (see ref. 77). Taken together, these data support an excitotoxic, glutamate-mediated mechanism (see ref. 78) of neurotoxicity for DTX-K. Collectively, the earlier data indicate that important differences exist in the mechanisms of seizures and hippocampal damage caused by these two DTX homologs. In particular, it is reasonable to hypothesize that DTX-K-evoked neurotoxicity results from a predominant presynatic blockade of K+ channels leading to glutamate release, which in turn induces seizures and CA1 pyramidal cell death (68). In contrast, α-DTX seems to act mainly at postsynaptic level because any enhancement of glutamate release elicited by this toxin was found not to be involded in triggering seizures and CA1

Dendrotoxins Fig. 6. Electrocortical records of epileptogenic spikes induced by intrahippocampal injection of DTX-K into rats and their prevention by NMDA and non-NMDA receptor antagonists. Elecrocortical activity, monitored in both fronto-parietal cortices as detailed in Experimental Procedures (see ref. 68), is shown before (control) and at indicated times after stereotaxic injection of DTX-K (35 pmol) into the CA1 area of one dorsal hippocampus of rats that had (B) or had not (A) been given 15 min previously an intraperitoneal injection of LY 274614 (5 mg/kg) or (C) GYKI 52466 (10 mg/kg), NMDA and non-NMDA receptor antagonists, respectively. Adapted with permission from ref. 68.

467

468

Dolly and Bagetta

Fig. 7. Light microscopy pictures showing DTX-K-induced CA1 pyramidal cell loss and its prevention by U-74389G. Brain tissue coronal sections (10 µm) from a control rat, injected into one dorsal hippocampus with bovine serum albumin (BSA; 300 ng), and those injected with DTX-K (35 pmol) alone or receiving additionally, 30 min previously, an intraperitoneal injection of the 21-aminosteriod U-74389G (5 mg/kg). Note the loss of neurones in the pyramidal cell layer of the CA1 hippocampal area (between arrows) induced by DTX-K when administered alone (see panel DTX-K) and the protection afforded by U-74389G (see panel U-74389G + DTX-K). By contrast, the 21-aminosteroid failed to prevent the CA1 neuronal loss (see panel U-74389G + α-DTX) typically evoked by α-DTX (35 pmol; see panel α-DTX), a DTX-K homolog. Glial cells infiltrating the pyramidal cell layer are also evident in some instance (see asterisks). Adapted with permission from ref. 76.

Dendrotoxins

469

pyramidal cell death (67,68,75). In agreement with a differential, pre- vs postsynaptic mechanism of neurotoxicity for DTX-K and α-DTX, respectively, is the recent evidence that, in comparison to the effect of α-DTX, intra-hippocampal injection of DTX-K causes a longer lasting and statistically significant increase in extracellular glutamate in the dorsal hippocampus of rat (79). Furthermore, the aforementioned hypothesis can be reconciled with the reported selectivity displayed by these two DTX homologs with respect to K+ channel blockade (see Subheading 3). In conclusion, the data yielded so far demonstrate that DTX-K (likewise δ-DTX [Bagetta, G. and Dolly, J. O., unpublished observations]) and α-DTX produce seizures and hippocampal damage via different mechanisms and are useful neurobiological tools to study subsets of experimental epilepsies and neuronal cell death. ACKNOWLEDGMENT This work is financed by MRC (to J.O.D.) and by MURST ex quota 60% (to G.B.). REFERENCES 1. Dolly, J. O., Halliwell, J. V., Black, J. D., Williams, R. S., Pelchen-Matthews, A., Breeze, A. L., et al. (1984) Botulinum neurotoxin and dendrotoxin as probes for studies on transmitter release. J. Physiol. (Paris) 79, 280–303. 2. Halliwell, J. V., Othman, I. B., Pelchen-Matthews, A., and Dolly, J. O. (1986) Central action of dendrotoxin: selective reduction of a transient K+ conductance in hippocampus and binding to localized acceptors. Proc. Natl. Acad. Sci. USA 83, 493–497. 3. Stansfeld, C. E., Marsh, S. J., Halliwell, J. V., and Brown, D. A. (1986) 4-Aminopyridine and dendrotoxin induce repetitive firing in rat visceral sensory neurones by blocking a slowly inactivating outward current. Neurosci. Lett. 64, 299–304. 4. Stansfeld, C. E., Marsh, S., Parcej, D. N., Dolly, J. O., and Brown, D. A. (1987) Mast cell degranulating peptide and dendrotoxin selectively inhibit a fast-activating potassium current and bind to common neuronal proteins. Neuroscience 23, 893–902. 5. Dreyer, F. (1990) Peptide toxins and potassium channels. Rev. Physiol. Biochem. Pharmacol. 115, 93–116. 6. Dolly, J. O. (1991) Components involved in neurotransmission probed with toxins: voltage-dependent K+ channels, in Probes for Neurochemical Target Sites. Royal Irish Academy Press (Tipton, K. F. and Iversen, L. L., eds.), Dublin, pp. 127–140. 7. Meves, H. (1992) Potassium channel toxins, in Handbook of Experimental Pharmacology, vol. 102, (Herken, H. and Hucho, F., eds.), Springer-Verlag, Berlin, pp. 739–774. 8. Stuhmer, W., Ruppersberg, J. P., Schroter, K. H., Sakmann, B., Stocker, M., Giese, K. P., et al. (1989) Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J. 8, 3235–3244. 9. Grupe, A., Schróter, K. H., Ruppersberg, J. P., Stocker, M., Drewes, T., Beckh, S., and Pongs, O. (1990) Cloning and expression of a human voltage-gated potassium channel. A novel member of the RCK potassium channel family. EMBO J. 9, 1749–1756. 10. MacKinnon, R. (199l) New insights into the structure and function of potassiurn channels. Curr. Opin. Neurobiol. 1, 14–19. 11. Dolly, J. O. and Parcej, D. N. (1996) Molecular properties of voltage-gated K+ channels. J. Bioenerg. Biomembr. 28, 231–253. 12. Harvey, A. L. and Karlsson, E. (1980) Dendrotoxin from venom of the green mamba, Dendroaspis angusticeps, a neurotoxin that enhances acetylcholine release at neuromuscular junctions. Naunyn-Schmiedeberg’s Arch. Pharmacol. 312, 1–6.

470

Dolly and Bagetta

13. Harvey, A. L. and Karlsson, E. (1982) Protease inhibitor homologues from mamba venoms: facilitation of acetylcholine release and interactions with prejunctional blocking toxins. Br. J. Pharmacol. 77, 153–161. 14. Harvey, A. L., Anderson, A. J., Marshall, D. L., Pemberton, K. E., and Rowan, E. G. (1990) Facilitatory neurotoxins and transmitter release. J. Toxicol. Toxin Rev. 9, 225–242. 15. Hu, P. S., Benishin, C., and Fredholm, B. B. (1991) Comparison of the effects of four dendrotoxin peptides, 4-aminopyridine and tetraethylammonium on the electrically evoked [3H]noradrenaline release from rat hippocampus. Eur. J. Pharmacol. 209, 87–93. 16. Dolly, J. O., Black, J. D., Black, A. R., Pelchen-Matthews, A., and Halliwell, J. V. (1986) Novel roles of neural acceptors for inhibitory and facilitatory toxins, in Natural Toxins: Animal, Plant and Microbial (Harris, J. B., ed.), Oxford University Press, pp. 237–264. 17. Tibbs, G. R., Dolly, J. O., and Nicholls, D. G. (1989) Dendrotoxin, 4-aminopyridine and ß-bungarotoxin act at common loci but by two distinct mechanisms to induce Ca2+-dependent release of glutamate from guinea-pig cerebro-cortical synaptosomes. J. Neurochem. 52, 201–206. 18. Dolly, J. O. (1992) Peptide toxins that alter neurotransmitter release, in Handbook of Experimental Pharmacology, vol. 102 (Herken, H. and Hucho, F., eds.), Springer-Verlag, Berlin, pp. 681–717. 19. Dufton, M. J. (1985) Proteinase inhibitors and dendrotoxins. Sequence classification, structural prediction and structure/activity. Eur. J. Biochem. 153, 647–654. 20. Benishin, C. G., Sorensen, R. G., Brown, W. E., Krueger, B. K., and Blaustein, M. P. (1988) Four polypeptide components of green mamba venom selectively block certain potassium channels in rat brain synaptosomes. Mol. Pharmacol. 34, 152–159. 21. Mehraban, F., Haines, A., and Dolly, J. O. (1986) Monoclonal and polyclonal antibodies against dendrotoxin: their effects on its convulsive activity and interaction with neuronal acceptors. Neurochem. Int. 9, 11–22. 22. Othman, I. B., Spokes, J. W., and Dolly, J. O. (1982) Preparation of neurotoxic 3H-betabungarotoxin: demonstration of saturable binding to brain synapses and its inhibition by toxin I. Eur. J. Biochem. 128, 267–276. 23. Petersen, M., Penner, R., Pierau, Fr.-K., and Dreyer, F. (1986) ß-Bungarotoxin inhibits a non-inactivating potassium current in guinea pig dorsal root ganglion neurones. Neurosci. Lett. 68, 141–145. 24. Rowan, E. G. and Harvey, A. L. (1988) Potassium channel blocking actions of ß-bungarotoxin and related toxins on mouse and frog motor nerve terminals. Br. J. Pharmacol. 94, 839–847. 25. Deisenhofer, J. and Steigemann, W. (1975) Crystallographic refinement of the structure of bovine pancreatic trypsin inhíbitor at 1.5 A resolution. Acta Crystallogr. 23, 544–548. 26. Skarzynski, T. (1992) Crystal structure of α-dendrotoxin from the green mamba venom and its comparison with the structure of bovine pancreatic trypsin inhibitor. J. Mol. Biol. 224, 671–683. 27. Berndt, K. D., Guntert, P., and Wuthrich, K. (1993) Nuclear magnetic resonance solution structure of dendrotoxin K from the venom of Dendroaspis polylepis polylepis. J. Mol. Biol. 234, 735–750. 28. Lancelin, J. M., Foray, M. F., Poncin, M., Hollecker, M., and Marion, D. (1994) Proteinase-inhibitor homologs as potassium channel blockers. Nature Struct. Biol. 1, 246–250. 29. Harvey, A. L. (1993) Neuropharmacology of potassium ion channels. Med. Rev. Res. 13, 81–104. 30. Smith L. A., Lafaye, P. J., LaPenotiere, H. F., Spain, T., and Dolly, O. J. (1993) Cloning and functional expression of dendrotoxin K from black mamba, a K+-channel blocker. Biochemistry 32, 5692–5697. 31. Smith, L. A., Reid, P. F., Wang, F. C., Parcej, D. N., Schmidt, J. J., Olson, M. A., and

Dendrotoxins

32.

33. 34. 35.

36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

49.

471

Dolly, O. J. (1997). Site-directed mutagenesis of dendrotoxin K reveals amino acids critical for its interaction with neuronal K+ channels. Biochemistry 36, 7690–7696. Danse, J. M., Rowan, E. G., Gasparini, S., Ducancel, F., Vatanpour, H., Young, L. C., et al. (1994) On the site of which α-dendrotoxin binds to voltage-dependent potassium channels: site-directed mutagenesis reveals that the lysine triplet 28–30 is not essential for binding. FEBS Lett. 356, 153–158. Harvey, A. L. (1997) Recent studies on dendrotoxins and potassium ion channels. Gen. Pharmacol. 28, 7–12. Gasparini, S., Danse, J.M., Lecoq, A., Pinkasfeld, S., Zinn-Justin, S., Young, C.L., et al. (1998) Delineation of the functional site of α-dendrotoxin. J. Biol. Chem. 273, 25,393– 25,403. Wang, F. C., Bell, N., Reid, P., Smith, L., McIntosh, P., Robertson, B., and Dolly, J. O. (1999) Identification of residues in dendrotoxin K responsible for its discrimination between neuronal K+ channels containing Kv 1.1 and 1.2 α subunits. Eur. J. Biochem. 263, 222–229. Black, A. R., Breeze, A. L., Othman, I. B., and Dolly, J. O. (1986) Involvement of neuronal acceptors for dendrotoxin in its convulsive action in rat brain. Biochem. J. 237, 397–404. Penner, R., Petersen, M., Pierau, F.-K., and Dreyer F. (1986) Dendrotoxin: a selective blocker of a non-inactivating potassium current in guinea-pig dorsal root ganglion neurones. Pflügers Arch. 407, 365–369. Benoit, E. and Dubois, J. M. (1986) Toxin I from the snake Dendroaspis polylepis polylepis: a high specific blocker of one type of potassium channel in mielynated nerve fibre. Brain Res. 37, 374–377. Brau, M. E., Dreyer, F., Jonas, P., Repp, H., and Vogel, W. (1990) A K+ channel in Xenopus nerve fibres selectively blocked by bee and snake toxins: binding and voltage-clamp experiments. J. Physiol. (Lond.) 420, 365–385. Hall, A. C., Stow, J., Sorensen, R., Dolly, J. O., and Owen, D. G. (1994) Blockade by dendrotoxin homologues of voltage-dependent K+ currents in cultured sensory neurones from neonatal rats. Br. J. Pharmacol. 113, 959–967. Parcej, D. N. and Dolly, J. O. (1989) Dendrotoxin acceptor from bovine synaptic plasma membranes: binding properties, purification and subunit composition of a putative constituent of certain voltage-activated K+ channels. Biochem. J. 257, 899–903. Muniz, Z. M., Diniz, C. R., and Dolly, J. O. (1990) Characterization of binding sites for δ-dendrotoxin in guinea-pig synaptosomes: relationship to acceptors for the K+ channel probe, α-dendrotoxin. J. Neurochem. 54, 343–346. Awan, K. A. and Dolly, J. O. (1991) K+ channel sub-types in rat brain: characteristic locations revealed using ß-bungarotoxin, α- and δ-dendrotoxins. Neuroscience 40, 29–39. Pelchen-Matthews, A. and Dolly, J. O. (1988) Distribution of acceptors for ß-bungarotoxin in the central nervous system of the rat. Brain Res. 441, 127–138. Pelchen-Matthews, A. and Dolly, J. O. (1989) Distribution in the rat central nervous system of acceptor sub-types for dendrotoxin, a K+ channel probe. Neuroscience 29, 347–361. Mehraban, F., Breeze, A. L., and Dolly, J. O. (1984) Identification by cross-linking of a neuronal acceptor protein for dendrotoxin, a convulsant polypeptide. FEBS Lett. 174, 116–122. Schmidt, R. R. and Betz, H. (1989) Cross-linking of beta-bungarotoxin to chick brain membranes. Identification of subunits of a putative voltage-gated K+ channel. Biochemistry 28, 8346–8350. Rehm, H., Bidard, J. N., Schweitz, H., and Lazdunski, M. (1988) The receptor site for the bee venom mast cell degranulating peptide. Affinity labeling and evidence for a common molecular target for mast cell degranulating peptide and dendrotoxin I, a snake toxin active on K+ channels. Biochemistry 27, 1827–1832. Rehm, H. and Lazdunski, M. (1988) Purification and subunit structure of a putative K+

472

50.

51. 52. 53.

54.

55. 56. 57. 58.

59.

60.

61. 62.

63.

64.

65.

66.

67.

Dolly and Bagetta channel protein identified by its binding properties for dendrotoxin I. Proc. Natl. Acad. Sci. USA 85, 4919–4923. Parcej, D. N., Scott, V. E. S. and Dolly, J. O. (1992) Oligomeric properties of α-dendrotoxin-sensitive K+ channels purified from bovine brain. Biochemistry 31, 11,084– 11,088. Catterall, W. A. (1988) Structure and function of voltage-sensitive ion channels. Science 242, 50–61. MacKinnon, R. (1991) Determination of the subunit stoichiometry of a voltage-activated potassiurn channel. Nature 350, 232–235. Ruppersberg, J. P., Schroter, K. H., Sakinann, B., Stocker, M., Sewing, S., and Pongs, O. (1990) Heteromultimeric channels formed by rat brain potassium-channel proteins. Nature 345, 535–537. Christie, M. J., North, R. A., Osborne, P. B., Douglass, J., and Adelman, J. P. (1990) Heteropolymeric potassium channels expressed in Xenopus oocytes from cloned subunits. Neuron 2, 405–411. Isacoff, E. Y., Jan, Y. N., and Jan, L. Y. (1990) Evidence for the formation of heteromultimeric potassium channels in Xenopus oocytes. Nature 345, 530–534. Dolly, J. O. (1988) Potassium channels: what can the protein chemistry contribute? Trends Neurosci. 11, 186–188. Gulbis, J. M., Mann, S., and MacKinnon, R. (1999) Structure of a voltage-dependent K+ channel beta subunit. Cell 97, 943–952. Scott, V. E. S., Muniz, Z., Sewing, S., Lichtinghagen, R., Parcej, D. N., Pongs, O., and Dolly, J. O. (1994) Antibodies specific for distinct Kv subunits unveil a heterooligomeric basis for subtypes of α-dendrotoxin-sensitive K+ channels in bovine brain. Biochemistry 33, 1617–1623. Wang, H., Kunkel, D. D., Martin, T. M., Schwartzkroin, P. A., and Tempel, B. L. (1993) Heteromultimeric K+ channels in terminal and juxtaparanodal regions of neurons. Nature 365, 75–79. Wang, H., Kunkel, D. D., Schwartzkroin, P. A., and Tempel, B. L. (1994) Localization of Kv1.1 and Kv1.2, two K-channel proteins, to synaptic terminals, somata, and dendrites in the mouse brain. J. Neurosci. 14, 4588–4599. Sheng, M., Liao, Y. J., Jan, Y. N., and Yan, L. Y. (1993) Presynaptic A-current based on heteromultimeric K+ channels detected in vivo. Nature 365, 72–75. Shamotienko, O. G., Parcey, D. N., and Dolly, J. O. (1997) Subunit combinations defined for K+ channel Kv1 subtypes in synaptic membranes from bovine brain. Biochemistry 36, 8195–8201. Scott, V. E. S., Parcej, D. N., Keen, J. N., Findlay, J. B. C., and Dolly, J. O. (1990) α-dendrotoxin acceptor from bovine brain is a K+ channel protein: evidence from N-terminal sequence of its larger subunit. J. Biol. Chem. 265, 20,094–20,097. Wang, F. C., Parcej, D. N., and Dolly, J. O. (1999) α Subunit composition of Kv 1.1containing K+ channel subtypes fractionated from rat brain using dendrotoxins. Eur. J. Biochem. 263, 230–237. Velluti J. C., Caputi A., and Macadar O. (1987). Limbic epilepsy induced in the rat by dendrotoxin, a polypeptide isolated from the green mamba (Dendroaspis angusticeps) venom. Toxicon 25, 649–657. Ben-Ari, Y., Tremblay, E., and Ottersen, O. P. (1980) Injection of kainic acid into the amygdala complex of the rat: and electrographic, clinical and histological study in relation to the pathology of epilepsy. Neuroscience 5, 515. Bagetta, G., Nisticò, G., and Dolly J. O. (1992) Production of seizures and brain damage in rats by α-dendrotoxin, a selective K+ channel blocker. Neurosci. Lett. 139, 34–40.

Dendrotoxins

473

68. Bagetta, G., Iannone, M., Palma, E., Nisticò, G., and Dolly, J. O. (1996) N-methyl-Daspartate and non-N-methyl-D-aspartate receptors mediate seizures and CA1 hippocampal damage induced by dendrotoxin-K in rats. Neuroscience 71, 613–624. 69. Coleman, M. H., Yamaguchi S., and Rogawski M. A. (1992) Protection against dendrotoxin-induced clonic seizures in mice by anticonvulsant drugs. Brain Res. 575, 138–142. 70. Amaral, D. G. and Witter, M. P. (1989) The three-dimensional organization of the hippocampal formation: a review of anatomical data. Neuroscience 31, 571–591. 71. Andersen, P. (1975) Organization of hippocampal neurones and their interconnections, in The Hippocampus (Isaacson, R. L. and Pribram, K. H., eds.), Plenum Press, New York, pp. 155–175. 72. Parè, D., deCurtis, M., and Llinàs, R. (1992) Role of hippocampal-enthorinal loop in temporal lobe epilepsy: extra- and intracellular study in the isolated guinea-pig brain in vitro. J. Neurosci. 12, 1867–1881. 73. Olney, J. W., Collins, R. C., and Sloviter, R. S. (1986) Excitotoxic mechanisms of epileptic brain damage, in Adv. Neurol., vol. 44 (Delgado-Escueta, A. V., Ward, A .A. Jr., Woodbury, D. M., and Porter, R. J., eds.), Raven Press, pp. 857–877. 74. Bagetta, G., Nisticò, G., and Bowery, N.G. (1991) Characteristics of tetanus toxin and its exploitation in neurodegenerative studies. Trends Pharmacol. Sci. 12, 285–289. 75. Bagetta, G., Nair, S., Nisticò, G., and Dolly J. O. (1994) Hippocampal damage produced in rats by α-dendrotoxin, a selective K+ channel blocker, involves non-NMDA receptor activation. Neurochem. Int. 24, 81–90. 76. Bagetta, G., Palma, E., Piccirilli, S., Nisticò, G., and Dolly, J. O. (1997) Seizures and hippocampal damage produced by dendrotoxin-K in rats is prevented by the 21aminosteroid U-74389G. Exp. Neurol. 147, 204–210. 77. Bagetta, G., Iannone, M., Vecchio, I., Rispoli, V., Rotiroti, D., and Nisticò, G. (1994) Neurodegeneration produced by intrahippocampal injection of paraquat is reduced by systemic administration of the 21-aminosteroid U-74389F in rats. Free Rad. Res. 21, 85–93. 78. Choi, D. W. (1988) Glutamate neurotoxicity and diseases of the nervous system. Neuron 1, 623–634. 79. Richards, D. A. and Morrone, L. A. (1999) Extracellular glutamate levels following intrahippocampal infusion of α-dendrotoxin and dendrotoxin-K in rat. Br. J. Pharmacol. 126, 253P. 80. Dolly, J. O., Muniz, Z. M., Parcej, D. N., Hall, A. C., Scott, V. E. S., Awan, K. A., and Owen, D.G. (1994) Subtypes of fast activating, voltage-gated K+ channels in the nervous system: location and molecular properties unveiled with α-dendrotoxin and homologues, in Neurotoxins in Neurobiology (Tipton, K. F. and Dajas, F., eds.), Ellis Horwood, Chichester, pp. 103–124.

474

Dolly and Bagetta

Neurotoxins from Spider Venom

475

23 Neurotoxins from Spider Venoms Alfonso Grasso and Stefano Rufini

1. INTRODUCTION In his excellent book “Arachnida,” rich in learned citations, T. Savory (1) writes: “The spider is the dominant arachnid; it surpasses all others in the number and variety of its species, in the complexity of its habits, and in the breadth of its range across the world. So well it is advertised by the beauty of the orb-web that all men know the spider and for many it represents the whole class of Arachnids. The spider may be encountered in mythology, in history, in art, in literature; its reputation is not unspotted and its merits are seldom recognized.”

It is not difficult to share this synthetic and passionate opinion on spiders, in a book having the purpose of promoting the role of Arachnids. However, in an article dealing with spider venoms this incipit sounds a little bit exaggerated. It is necessary to remember here that spiders have also developed diverse biological and behavioral specializations that are essentially used for predation purposes. The majority of spiders use venoms for food collection (immobilizing and killing the prey: generally an insect) and for self-defense. Thus, spiders carry a well-developed venom apparatus where a set of different interesting molecules with diverse functions is synthesized. Considering the physiology of the spiders’ digestive functions, it is likely that venom glands, besides toxic molecules, it also produce enzymes having proteolytic, glycolytic, nucleolytic, collagenolytic, and so forth, activity. There seems to be a largely accepted empirical rule that spiders living in webs generally have venoms characterized by a neurotoxic action, whereas the free-ranging species have cytotoxic venom (2). Although reports on spider-venom composition suggest that they contain a large number of active toxins: (1) neurotoxins, (2) necrotic toxins, (3) toxins with hemolytic activity, and (4) toxins affecting ion channels, the previously mentioned rule can be accepted as synthetically distinctive as well as descriptive of spider-venom composition. The data reported here will be primarily limited to the neurotoxic molecules. Historically, the study of the molecular composition of the venom of spiders, mites, insects and so on was carried out according to a traditional biochemical approach. This From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

475

476

Grasso and Rufini

was quite difficult to accomplish because a large number of venom glands were necessary from animals which are normally small, and difficult to collect and to rear. The use of molecular biology in the study of spider venom has certainly been advantageous. In fact, cloning and expression of cDNAs has offered an alternative to secure sufficient material to detect molecules having pharmacological activities (polypeptides). As a consequence of this last approach, the sequences of a number of toxic molecules have been described and characterized, in recent years. The natural prey of spiders are insects, and as it was expected, spider venom contains toxins directed against animals belonging to this class. However, the toxins found in the spiders’ venom is often very toxic also for vertebrate, although the latter are not a source of food nor are they known to be predators of spiders. Thus, we have to suggest a purely fortuitous toxic action for molecules actually having a different function in the venom. If this proposition is true, it may even be possible that the venom of a greater number of spiders is toxic to humans, but that the consequent pathological syndrome in humans is unknown because of the inadequacy of their piercing apparatus. Indeed spider venoms contain a variety of toxic components. According to a recent review article (3) the polypeptide toxins of spiders can be conveniently subdivided into low and high molecular-mass types. Small polypeptide toxins interacting with cation channels, display spatial-structure homology. They can affect the functioning of calcium, sodium, or potassium channels. In contrast, a family of high molecular-mass toxic proteins (latrotoxins) was found in the venom of the spider genus Latrodectus and these cause a massive transmitter release from a diversity of nerve endings. This article intends to cover the most recent studies on neuroactive spider toxins only and in reviewing the structural and functional properties of spider polypeptide toxins we can share this naive subdivision and proceed accordingly: (small peptides = channels modifiers/large polypeptides = modifiers of presynaptic functions). In addition, the article is confined within limits of those toxins having a well-described target, so that the perspectives of their use in neurobiology, beside being promising, are also well-suited for the understanding of the mechanisms involved. 2. TOXINS ACTING ON IONIC-CHANNELS The ionic fluxes of the excitable cells occur through three main different classes of channels: (1) passive channels; (2) ligand-gated-channels; (3) voltage-gated-channels. These three classes derive probably from a common ancestor gene and, after a complex evolutionary pathway, originate a great number of channels showing peculiar characteristics of ion specificity, molecular regulation, and conductivity. Basically, spiders venom contains two chemical classes of ionic-chanel modulators: Blockers of ligand-gated channels. Typically, these low molecular weight molecules are formed by a phenolic or indolic moiety linked to the N-terminus of a polycationic part composed of long-chain polyamines. Modifiers of voltage-activated channels open state. Toxins belonging to this class are peptides that, despite their low overall amino acidic sequence homology and their diversity in molecular target, show a conserved structural motif consisting of a cystine knot and a triple-stranded ß-sheet.


477

2.1. Toxins Acting on Voltage-Gated Ca2+ Channels Multiple types of voltage-operated calcium channels (VOCCs) that possess a characteristic molecular structure, a different pharmacology, a different localization and possibly, a different physiological role have been described both in the nervous system and in peripheral tissues. VOCCs consist of an α1 subunit that possesses a voltagesensitive gate coupled with a calcium-conductive pore and auxiliary subunits (ß, γ, and α2δ) involved in the modulation and in trafficking of the α1 subunits. The mRNAs for the ion-conducting α1 subunits of VOCCs have been categorized into five structural classes A–E (4), whereas studies of native calcium channels have defined a variety of currents with distinctive electrophysiological properties: L, N, P, Q, R, and T (for a review, see ref. 5). The availability of natural toxins that bind to specific calcium channel subtypes has greatly helped in calcium channels classification (3). Polypeptide antagonists of calcium channels were first purified from the venom of sea snail of the genus Conus (for a review, see ref. 6). More recently the venom of various species of spiders have also proven to be sources of peptides with specificity toward diverse calcium channels subtypes (see Table 1). The original observation (7) that the venom of the funnel-web spider Agelenopsis aperta irreversibly blocked synaptic transmission in chick cochlear nucleus neurons without affecting the neurotransmitter-induced cell depolarization, opened the way for the discovery of a class of polypeptide spider toxins that bind and block VOCCs at nanomolar concentrations (8). A complete purification of A. aperta venom, revealed the presence of at least four classes of toxins acting as blockers on calcium channels, named ω-Aga-I, II, III, and IV. Both type I (A, B, C) and type II (A, B) strongly inhibit insect neuromuscular transmission (9,10). This block is additive when the toxins are applied jointly, suggesting that they bind to distinct sites (11). The mature ω-Aga-IA comprises two polypeptide chains, a larger fragment of 66 amino acids that is disulfide-bond to a small tripeptide with sequence Ser-Pro-Cys. Mass spectroscopy and sequencing data are consistent with a model where mature toxins are derived from the precursor by excision of an internal, glutamate-rich, heptapeptide (10). The mature precursor region of the ω-Aga-IA does not immediately follow the signal sequence, but it is preceded by a relatively short intervening region, enriched with glutamate residues, which acts as an intramolecular chaperone that organizes the correct folding of the polypeptide, as demonstrated in other proteins (12). The presence of a glutamate-rich sequence at the C-terminus preceding the major chain polypeptide was also described in the cDNA precursor encoding the neurotoxin Tx1 from Phoneutria nigriventer (13) and there are reasons to believe this to be a common gene encoding feature for spider toxins. A more detailed study on the specificity of the different types of agatoxins was carried out by competition binding to synaptosomal membranes of different fractions of A. aperta venom with [125I]ω-conotoxin GVIA (14). Through this approach, it was demonstrated that both type II and type III ω-agatoxins compete for the binding of [125I]ω-conotoxin GVIA to synaptosomal membrane prepared from chick or rat brain, whereas type I agatoxins do not (14). In the same way the K+-induced calcium entry into chick brain synaptosomes is blocked only by type II and III ω-agatoxins, but not

478

Grasso and Rufini

Table. 1 Spider Toxins Acting on Voltage-Activated Ion Channels Spider Phoneutria nigriventer

Agelenopsis aperta

Toxin PhTx2-A PhTx3-1 PhTx3-2 PhTx4 ω-Aga-I ω-Aga-II

ω-Aga-III ω-Aga-IV (A;B) µ-Aga-(I-IV) FTx Agelenopsis opulenta Agelenine Plectreurys tristis PLTx-II Hadronyche versuta Versutoxin (δ-atracotoxin) Atrax robustus Robustotoxin Grammostola spatulata Hanatoxin (HaTx-I -II) ω-grammotoxin (SIA) Filistata hibernalis DW13.3 Segestria florentina SNX-325 Hysterocrates gigas SNX-482 Heteropoda venatoria Heteropodatoxin (HpTx 1,2,3) Selenocosmia huwena huwentoxin-II (HWTX-II) Eurypelma californicum ESTX Hololena curta Curtatoxin (I-III)

AA

Target

76 48 36 38 37 44 42

Ca2+-channel (N > L > P/Q) K+-channel Ca2+-channel (L) Inhibitor of glutamate uptake Ca2+-channel (insect neuronal type) Na+- and Ca2+-channels (insect neuronal type) Ca2+-channel (N > L) Ca2+-channel (P/Q) Na+-channel (insect neuronal type) Ca2+-channel (P/Q > T) Na+-channel (insect neuronal type) Ca2+-channel (insect neuronal type) TTx sensitive Na+-channel

42 35

TTx sensible Na+-channel K+-channel (Kv2.1 > Kv4.2)

36

Ca2+-channel (N > P > Q)

74 49 42 30/33

Ca2+-channel Ca2+-channel (N) Ca2+-channel (R) K+-channel (Kv4.2-Shal subfamily)

37

Block of NM transmission

? 36

? Na+-channel (insect neuronal type)

76 76 76 ? 66 90/95

Ttx, tetrodotoxin; NM, neuromuscular.

by type I. Ertel and coworkers (15), reported a difference in specificity between the two main components of Type III ω-agatoxins, namely ω-Aga-IIIA and ω-Aga-IIIB. The two toxins share high sequence (the 76 amino acid long peptide possesses 66 identical residues) and structural homology, but different pharmacological properties. Both toxins recognize a binding site in N type, but ω-Aga-IIIA blocks L and N-type VOCCs with an equally high potency (IC50 = 1 nM) whereas ω-Aga-IIIB possesses 100-fold more affinity for L-type vs N-type calcium channel (15). Notably, the calcium blockers of type II and type III totally inhibit the depolarization-induced calcium entry in chick brain synaptosomes, but it does so only partially in rat brain synaptosomes. With the aim of isolating a mammalian VOCs antagonist, fractions from A. aperta venom were systematically tested for antagonism of depolarization-stimulated 45Ca2+ entry in rat


479

Fig. 1. Spider and snail toxins share a similar tertiary structure and “cystine motif.” Molscript representation of tertiary structure of: ω-agatoxin IVA (A); µ-agatoxin (B), and ω-conotoxin GVIA (C). The structural topology of antiparallel ß-sheet is shown. Arrows and rectangles represent the location of ß-strands in the solution structures of the peptide toxins. The structure coordinates of the toxins were extracted from the Research Collaboratory for Structural Bioinformatic, PDB.

brain synaptosomes. This approach yielded the type IV channel antagonists ω-AgaIVA (16) and ω-aga-IVB (17), which have amino acid sequences unrelated to types I-III ω-agatoxins. Both ω-Aga-IVA and ω-Aga-IVB inhibit the high-voltage-activated calcium current in cerebellar Purkinje neurons (8,16,17) by blocking the “P-type” channel (8) and, with lower potency, a distinct “Q-type” channel (18,19), but do not show any effect on L-type and N-type calcium channels (19–21). The channels sensitive to ω-Aga-IVA contribute a major part of the calcium entry underlying synaptic transmission at a variety of central (22,23) and peripheral synapses (24); thus the protein is a very powerful blocker of voltage-activated vesicles exocytosis (25,26). The ω-Aga-IVA sensitive P-type channels seem to be involved not only in the exocytosis, but also in the endocytotic membrane retrieval to compensate for the excess surface membrane after exocytosis (27). ω-Aga-IVA seems to block the channel not by interacting with the pore, but rather by altering the voltage-dependent gating of the channels (28). In fact, after ω-Aga-IV interaction, P-type channel cannot be opened by moderate depolarization but it can be opened with the toxin still bound by a strong depolarization. Moreover, the unbinding reaction of the toxin is dramatically accelerated by repeated strong depolarization (17). The three-dimensional structure of ω-Aga-IVA was determined in aqueous solution by two dimensional 1H NMR and simulated annealing calculations (29). The overall fold for ω-Aga-IVA is similar to that of ω-CTX GVIA (30) from Conus snail, sharing an identical disulfide bonding pattern, a small three-stranded antiparallel ß-sheet, four loop regions, a solvent inaccessible core, and disordered carboxyl and amino terminal (Fig. 1).

480

Grasso and Rufini

Although mass spectroscopic data indicate the presence of four disulfide bonds, it has not been possible to obtain the disulfide pairing information using classical enzymatic and/or chemical standard techniques due to the high density of cysteine residues in the protein. The most reasonable pattern of the disulfide bonds based on NMR observation was C4-C20, C12-C25, C19-C36, and C27-C34 (31). The results obtained by NMR indicate that, together with a very rigid nucleus, regions inconsistent with a single static structure are present in loops and most notably in the C-terminal tail (29). This suggests that there are residues capable of undergoing relatively low-cost rearrangements to adopt specific geometry imposed by ligand-channel complex. Comparison of the structure of ω-Aga-IVA with the structure of other toxins having the same biological target but different activity, gives important information on the structure-activity relationship for these toxins. ω-Aga-IVA and ω-Aga-IVB are closely related peptides, in size (48 aa), cysteine residues, sequence identity (73%), and overall three dimensional structure (31). These peptides show the same affinity for P-channel (Kd for both peptides 2/3 nM [8,16,17]), as well as the unbinding reaction of ω-AgaIVA results similar to ω-Aga-IVB being dramatically accelerated by depolarization of the membrane, increasing more than one thousand fold at +120 mV compared with –80 mV (17). All these evidences lead to suppose an identical interaction of these peptides with the P-channels. NMR data suggest that both ω-Aga-IVA and ω-Aga-IVB possess a region exhibiting a large positive potential represented by a cluster of basic residues occurring between position 10 and 27. This positively charged group is clustered on the same side in both toxins, thus this region has been proposed to play a pivotal role for the specific interaction with P-type calcium channel. Combined with the suggestion that there are clusters of negatively charged residues near the pore of the Ca2+-channel, one speculative model of interaction between the toxin and the channel has emerged. The complementary electrostatic attractions between the toxin and the extracellular loops near the mouth of the pore may contribute to binding affinity, while the hydrophobic tail of the toxin (corresponding to the carboxyl terminal residues) may cause the blockade of the channel either simply by occluding the conductance pathway or by establishing additional molecular contacts with the channel proteins. The only difference in channels-blocking activity between ω-Aga-IVA and ω-Aga-IVB is that the latter develops channel-blocking activity and reverses during washing out more slowly than ω-Aga-IVA. For this reason ω-Aga-IVA results more potent than ω-AgaIVB to inhibit calcium current in vivo. A comparison of the two peptide structures reveals surfaces that contain distinct clusters of charged residues in the amino terminal portion (i.e., basic in ω-Aga-IVB and acidic in ω-Aga-IVA) that deeply influence the calculated electrostatic topographies (32). Thus, it is possible that the reason for different activity observed in ω-Aga-IVA and ω-Aga-IVB may reside in the different clusters present in the amino terminal tail of the toxin. A novel pharmacological property of ω-Aga-IVA has been described by Herrero et al. (33) who studied the inhibition of neuronal acetylcholine receptor (nAChR) expressed in Xenopus laevis oocytes, by voltage-activated channel blockers such as ω-conotoxin and ω-agatoxins. These authors report that a high ω-Aga-IVA concentration (1 µM) is able to affect nAChR ionic currents elicited by dimetylphenilpiperazidinium (33). Previously, (34,35) it was reported that huwentoxin-I (HWTX-I) from the venom of the Chinese bird spider Selenocosmia huwena, a neurotoxic polypep-


481

tide of 33 amino acids, reversibly blocks the neuromuscular transmission in isolated phrenic nerve-diaphragm preparation (35). It has been shown by binding competition experiments of HWTX-I with tubocurarine, that the toxin acts as an inhibitor of nAChR (35). Nuclear magnetic resonance (NMR) studies reveal that HWTX-I acquires a threedimensional structure in solution normally adopted by channel-modulator polypeptides such conotoxins and agatoxins (36,37). The presence of neurotoxic polypeptides with specificity for calcium channels seems to be a common feature in the spider venom. Venom fractions of Filistata hibernalis (38), Hololena curta (39,40), Plectreurys tristis (13), Grammostola spatulata (41), Segestria florentina (42), Hysterocrates gigas (43), among others, show antagonism for insect and vertebrate calcium channels (see Table 1). The complete structure of PLTXII, a presynaptic calcium blocker in Drosophila, purified from Plectreurys tristis, reveals a 44-amino acids peptide with an O-palmytoil threonine amide residue at its carboxyl terminus. Cleavage of the O-palmitoyl ester from native toxin by base treatment results in a loss of biological activity, suggesting that acylation is required for activity (44). PLTXII, which is primarily a water-soluble peptide, acquires a significant solubility in organic solvent thanks to its bond to palmitate. Thus, it is possible that the lipid moiety allows toxin to penetrate the membrane and to act at an intracellular site. Complete sequence and cDNA cloning of about 50 toxins from the venom of P. tristis, indicate that several of these toxins undergo posttranslational modification such as amidation of C-termini, to produce the active products (45,46). The amidation of the C-terminus has also been demonstrated in other spider neurotoxins. At least four of the six Na+ channel modifier µ-Agatoxins (I–VI) are amidated at the carboxyl terminus (47). Post translational modification of toxins does not seems confined to polypeptide spider toxins, but to be a common feature in many channel-modifying peptides. For example, in a recent review (48), a great number of post-translational modifications in Conus peptides are listed, some of which are very unusual, and others (such as bromination of tryptophan) are first described in Conus peptide. The first selective antagonist of R-type calcium channels is the 41 amino acids neurotoxic peptide SNX-482 purified (43) from the venom of the African tarantula Hysterocrates gigas. The target of this toxin was identified by screening with SNX482 a set of Xenopus oocytes, stably expressing different voltage-gated channels (43). The SNX-482 specificity towards the R-type channel has allowed to determine that this channels controls the release of oxytocin from neurohypophysial nerve terminals (49). ω-Grammotoxins was first isolated and pharmacologically characterized from the venom of the tarantula spider Grammostola spatulata (41). The authors report that this peptide, which shares the common cystine knot motif, inhibits neuronal N- and P-type calcium channels, without affecting mammalian vascular L-type. The effects of ω-grammotoxin on gating P-type channels are very similar to those of ω-Aga-IVA, but its binding to the P-channel is not prevented by saturating binding of ω-Aga-IVA. Thus the conclusion is that grammatoxin strongly inhibits P-type channel gating, by binding to distinct (or additional) sites of the channel compared with ω-Aga-IVA (28,50). Despite the fact that toxins blocking the voltage-activated ionic channels have a high specificity for the different channels subtypes, they all share a common secondary and tertiary structural motif. This motif, which bridges the ends of the peptide to the

482

Grasso and Rufini

center, forming a very compact structure (39), is essential for biological activity, thus, when disulfide bonds are reduced, the resulting peptide is not toxic even at very high dosage (47). Interestingly, this structural motif has been found in an increasing number of small toxic polypeptides purified from different sources such as the venom of scorpion (51), sea snail of the genus Conus (30,52), and plant (53), all acting as modulator of voltage-activated ionic-channel activity. These data arise the fundamental question on the relationship between the nature of this common structural motif and the modification of channels activity. To answer this question few hypothesis are plausible. First, highly folded native toxins are quite resistant to heat denaturation as well as to proteolytic digestion (47). Second, the fold appears to be an ideal compact globular scaffold for the presentation of a variety of functional groups. Thus, it is possible that the hypermutation of an ancestor gene reached polypeptides with a great numbers of exposed amino acids, creating a practically infinite number of different ligands and finally generate a range of polypeptides with diverse biological targets. A third hypothesis could be that this common structural motif, could posses the right molecular geometry to associate a domain that recognizes a variable binding site specific for a target with a domain that recognizes a highly conserved three-dimensional voltage-sensing region in voltage-gated ion channels. This hypothesis is in agreement with the results obtained by using two spider toxins, purified from the venom of the Chilean rose tarantula, Phrixotrichus spatulata (54). Hanatoxin and grammatoxin modify preferentially the voltage-dependent gating of voltage-gated K+ and voltage-gated Ca2+ channels, respectively. The authors, by using several mutants of voltage-gated channels that display an altered binding of both neurotoxins, suggest that the two toxins interact with a structural domain conserved both in voltage-gated Ca2+ and K+ channels. 2.2. Toxins Acting on Na+ Channel Toxins of the class µ from A. aperta venom cause repetitive firing in presynaptic motor axons of Manduca sexta thus they produce a gradual but irreversible paralysis and death of the insect after injection with an LD50 values ranging between 7 and 18 nmol/g. The first identification and characterization of µ-neurotoxins was done by Skinner et al. (45), who achieved the complete sequences of six µ-agatoxins from the venom of A. aperta. Three dimensional solution structures of both µ-Aga-I and µ-Aga-IV, obtained by NMR, confirm a common secondary and tertiary structural motif with phylogenetically diverse peptide toxins targeting a variety of channel type (55). The main structural differences between µ and ω class of Agelenopsis toxins lies in the presence of a long hydrophobic C-terminal tail in the ω class. Although the functional significance is not clear, an interesting aspect is that µ-Aga-IV exhibits two distinct and equally populated conformations in solution, arising from cis and trans peptide bonds, which involve a proline in position 15 (55). Skinner first proposed that the action mechanism of the toxins belonging to this class, appears similar to that of the scorpion toxins, which act as sodium-channel agonists and increase cell excitability (56). In fact, modification of Na+ channels by spider toxins resulted in a dramatic slowing down of INa decay, consisting in the modification of channel inactivation from the open state (45). Original studies were carried out on the effect of the venom of Atrax robustus


483

(57). A. robustus venom produces multiple end-plate potentials on mouse diaphragm in response to a single stimulus of the phrenic nerve (57), and cause the prolongation of evoked action potentials (58), but have no effect on the resting membrane potential of skeletal muscle or on the amplitude or time course of end-plate potentials (59). The major toxic component of the venom has been identified with a polypeptide, named robustotoxin (δ-Atracotoxin-Hv1 or RTX) (60), that is able to mimic the neurotoxic symptoms of whole A. robustus venom in anaesthetized monkey (61). NMR analysis reveals that in solution RTX adopts the classical inhibitor cystine-knot structure (62) found in several other voltage-activated channel-modifier toxins. More detailed studies on the molecular mechanism of action of spider toxins acting on voltage-gated Na+ channels have been carried out using versutoxin (δ-AtracotoxinAr1 or VTX), the major component of the venom of Australian Blue Mountains funnel web spider Hydronyche versuta (63). VTX shows a considerable primary sequence homology with RTX (only seven amino acidic substitutions) and both toxins produce the same neurotoxic syndrome in humans. In rat dorsal-root ganglion cells that have two types of sodium channel, one tetrodotoxin insensitive and the other tetrodotoxin sensitive, VTX shows a selective interaction only with TTx-sensitive sodium channels gating and kinetics (64). Both toxins have been show to exert their neurotoxicity by slowing or removing tetrodotoxin-sensitive sodium current inactivation in rat-dorsal root ganglion cells (65,66), an action similar to that of scorpion α-toxin and sea anemone toxin (67). Using different radiolabeled toxins, seven neurotoxin receptor sites, in the voltage-activated sodium channel, have been identified (68). Competitive binding experiments demonstrate that atracotoxins do not displace [3H]saxitoxin from site 1 on sodium channel of rat brain synaptosomes (69). Moreover electrophysiological data (i.e., VTX does not alter the selectivity of the sodium channel), exclude also that VTX acts as an agonist on site 2 like such as batrachotoxin and grayanotoxin (70). However, competitive binding experiments demonstrate that atracotoxins were able to completely displace 125I-scorpion α-toxin Lqh II in a concentration-dependent manner (71), suggesting that they bind site 3, i.e., the same site of scorpion α-toxins. Recently, the VTX solution structure determined by using NMR spectroscopy, shows homology with µ-agatoxins from A. aperta but not with sea anemone nor scorpion toxins, both modifying the opening kinetic of sodium-channel (72). Despite this lack of homology, VTX contains charged residues that are topologically related with those implicated in the binding of scorpion toxins, indicating a common binding-mechanism consistent with the idea that Atrax toxins slow down the inactivation of voltage-gated sodiumchannel by interacting with channel-recognition site 3 (72). Despite the structural analogy and the binding competition evidence, some electrophysiological data reveal several differences between the action mechanism of α-scorpion and atracotoxins, and at the moment further studies are necessary to understand the mechanism of atracotoxins action on the dynamic gating processes of voltage-gated sodium channels. It is of interest that during evolution different organisms have developed toxins aimed at modifying different processes of sodium-channel gating. Some of the toxins prevent the sodium channel from activation thus causing paralysis of the prey. Other toxins produce a persistent activation of sodium channels in the prey, causing convulsions and preventing normal motility. A possible explanation could be that the evolutionary selection of the molecular target on the sodium channel may be related to the

484

Grasso and Rufini

visual acuity of the predator (Zlotkin, personal communication in ref. 73). Predators with poor visual acuity will have difficulty in finding a still immobile prey. 2.3. Toxins Acting on K+ Channel A large number of K+ channel inhibitors have been already isolated principally from the venom of scorpion (74,75) and snake (76). A common feature of these toxins is the ability to bind and block primarily the Shaker-related subfamily of voltage-activated K+ channels (77) without affecting the other four subfamilies of the K+-channel (namely: Shab-related, Shaw-related, Shal-related, and eag-related). Sanguinetti et al. (78) purified three peptide toxins from the venom of the spider Heteropoda venatoria, named Heteropodatoxins (HpTxs), able to block transient outward K+ current in rat ventricular myocites. Using X. laevis oocytes transfected with different transient outward K+, these authors demonstrated that HpTxs specifically block the Kv4.2 channel (Shal-related subfamily), and that this block is strongly voltage-dependent. This could result either from dissociation of the toxin from its binding site in or near the pore caused by an outward flow of K+ and/or from a modulation of intrinsic channel-gating behavior. Hanatoxins (HaTxs) from the venom of a Chilean tarantula (Grammostola spatulata) share significant sequence homology with the HpTxs but show a specific blocking activity on a potassium channel, the Kv2.1, belonging to the Shal-related subfamily with an IC50 value of 42 nM (79). Furthermore, HaTxs 500 nM, are able to block 75% of the current flowing in Kv4.2 but do not show any inhibition on other K+ channel tested (79). The identification of toxins that target specifically Shal- and Shabrelated subfamily of K+ voltage-activated channels opens new perspectives in the study of the physiological role of these classes of channels in different tissues. Moreover, the mapping of spider toxin binding sites, as already done for some scorpion toxins, could reveal new structural features of the K+-channel itself. Studies in this direction have been carried out (80,81) using HaTxs as probes. The experiments indicate that HaTx inhibits the K+ channel not by occluding the ion-conductance pore but by modifying channel-gating properties. Moreover, inhibition kinetics, as well as concentration dependence for equilibrium binding, reveal that HaTx alters the energetic of channel gating by binding to the surface of the channel at four equivalent sites. These equivalent sites seem to correspond to the residues near the outer edges of S3 and S4 domains, located at least 1.5 nm from the central pore axis of the channel (81). Several groups report the identification of other protein toxins that act on K+ channels. In particular, Kushmerick et al. (82) reported that type 3 toxin purified from Phoneutria nigriventer (namely toxin Tx3-1), as 4-aminopyridine, controls Ca2+ oscillation frequency in GH3 cells through a mechanism involving the block of A-type K+ currents, but without affecting other K+ currents. A novel K+-channel blocking toxin, named SGTx1, was recently purified from the venom of Scodra griseipes (83). SGTx1 shares a high homology with hanatoxins and similarity to TxP5 toxin (32% identity) from Brachyphelma smithii (84) and to huwentoxin (26% identity) from S. huwena (36). Preliminary studies indicate that SGTx1 160 nM is able to gradually and reversibly depress both the outward K+ fast-transient current and the delayed rectifier current in rat cerebellum granular cells suggesting a mode of action reminiscent of that shown by HaTxs on Kv2.1 and Kv4.2 channels (83).


485

2.4. Toxins Affecting Ligand-Gated Channels Since evidence exists that glutamate receptors are involved in several neurological diseases as mediators of epileptogenesis, and excitotoxic cell death in brain ischemia and that block of glutamate receptors can exert a neuroprotective role in these neurodegenerative disorders, we intend to briefly describe some molecules present in spider venom having a potentially protective effect. Glutamate seems to be the common mediator in almost all invertebrate neuromuscular junctions (NMJ) as well as the principal excitatory neurotransmitter in mammalian central nervous system (CNS). Glutamate receptors are classified on the basis of their response to different agonists (AMPA, NMDA, kainate, etc.) and subdivided into metabotropic, coupled to a second messenger pathway (mainly through phosphatidylinositolphosphates hydrolysis), and ionotropic, associated to the opening of ionic channels. Low molecular weight arylamines of the venom of spider belonging to the family Araneidae, seem to be the main factor causing spider-bite induced paralysis of insect by blocking glutamate receptors at the NMJ (85–87). So far, glutamate-receptor blocking toxins have been isolated from several spider venoms and utilized for functional and structural studies of glutamate receptors. Typically these molecules consist of an aromatic moiety (generally phenyl or indole) linked to the N-terminus of a polycationic part of long-chain polyamines (88). Polyamines closely related to those found in the spider venom have also been found in the venom of solitary digger wasp Philanthus triangulum (89). The similarity of these wasp and spider neurotoxins provides a notable example of convergence in the evolution of secondary metabolites directed against a common target. After the discovery of the neurotoxin that acts postsinaptically on the ionotropic glutamate receptor (90), other toxins purified from the venom of spider belonging to the Genus Nephila, (i.e., JSTX from N. clavata and NsTx from N. maculata venom) were described as potent antagonists of glutamate synapse both in invertebrate NMJ and in vertebrate CNS (91–93). These toxins appear to be very effective in blocking glutamate-induced excitatory postsynaptic potential (EPSP) with an IC50 of 10 nM, without affecting inhibitory postsynaptic potentials (91). Studies carried out on the specificity of acylpolyamine toxins for different classes of glutamate receptor gave conflicting results. In NMJ of invertebrate, JSTX blocked quisqualate-type receptors while aspartate-induced responses were insensitive to the toxin (92). In vertebrate, JSTX suppresses the EPSPs in CA1 pyramidal neurons evoked both by stimulation of Schaffer collateral fibers and by ionthophoretically added glutamate (94,95). These studies suggest that JSTX is preferentially antagonist of the AMPA receptor-mediated component of excitatory transmission, with little effect on the NMDA receptor-mediated component (94). Using the single electrode voltage-clamp method in CA1 pyramidal neurons, it was shown (96) that JSTX affects, at different levels, both NMDA receptor-mediated currents and non-NMDA-mediated currents. Furthermore, it was reported (97) that in cerebellar granule cells in cultures, JSTX is able to stimulate specifically both AMPA and kainate receptors. These conflicting observations can probably be justified by assuming that spider venom arylpolyamines show different affinity for different subtypes of the same class of glutamate receptor. Studies carried out on the action of arylpolyamine

486

Grasso and Rufini

on recombinant AMPA and NMDA receptors expressed in Xenopus oocytes, reveal a puzzling picture of the interaction: spider-toxin-arylamine/glutamate-receptor. In fact, these toxins are very specific in blocking AMPA-receptor formed by GluR1, GluR3, GluR4 subunits but show a very weak specificity for AMPA receptor containing GluR2 subunit (98,99). The amino acid composition of the toxin binding-site in the glutamate receptor, seems to be very critical in this context, being a single amino acid substitution sufficient to markedly reduce the sensitivity of the glutamate receptor to toxins (98). Thus, specificity of the toxin for a glutamate receptor, depends on the presence of a binding site of the subunit forming the receptor and is not restricted to a pharmacological class of glutamate-receptors. Argiotoxin636, is an acylpolyamine acting specifically on NMDA-receptor in a voltage-dependent fashion, purified from the venom of Argiope lobata (100–102). Reynolds reported that argiotoxin636 inhibits the binding of [3H]MK-801 to rat brain membranes in a manner that is insensitive to glutamate, glycine, and spermidine (103). It was concluded that argiotoxin636 exerts its inhibitory effect on the NMDA receptor complex by binding to one of the Mg2+ sites located within the NMDA-gated ion channel. As the previously described toxins, also the block of NMDA receptor by argiotoxin636 seems to be dependent on the subunit present in the receptor (104). A detailed characterization of argiotoxins mechanism of action in cultured rat hippocampal neurons was done by (105) using whole-cell recording techniques. The authors conclude that arylamines spider toxins exert their blocking action on NMDA receptor by two distinct mechanisms: channel block and competition at the NMDA recognition site. Furthermore, at positive holding potential and in the presence of high NMDA concentrations the toxins can also induce a polyamine-like potentiation, an effect that in vivo is masked by the various blocking actions of the toxins. Agelenopsis aperta venom contains a significant amount of both voltage-activated calcium channel-blocking peptides and arylamine toxins (47). Low molecular-weight Agelenopsis toxins, named α-agatoxins, antagonize synaptic transmission mediated by glutamate receptors in both arthropods and vertebrates (11,106). Agatoxin-489 appears to be the most potent and most selective NMDA antagonist among the spider toxins, being at least 10-fold more effective than argiotoxin636 (EC50 value of 10–100 nM and 100–200 nM, respectively; 107,108). Kiskin et al. (108) demonstrated that agatoxin489 can block the NMDA receptor only if the channel is in the open glutamate-activated state, an observation that implies that the toxin affects primarily the channel part of the NMDA receptor-channel complex rather than the agonist recognition site; thus, after toxin binding, the channel leads to an “open but non-conducing state.” Structures of the molecule of α-agatoxins reveal that these polyamine-containing toxins lack the amino acids found in other spider arylamine toxin and contain an unprecedented hydroxylamine moiety proximal to the acylamine chain terminus (108). 3. TOXINS THAT AFFECT PRESYNAPTIC TERMINALS AND ACTIVATE EXOCYTOSIS 3.1. The Venom of Latrodectus tredecimguttatus (BWS) Latrotoxins, the neurotoxins found in the venom of spiders of genus Latrodectus are known to have presynaptic terminals affinity and to activate secretagogic effects


487

Table 2 Chronological Description of the Latrotoxins, Characterized in the Venom-Gland Extract of Black Widow Spider (L. tredecimguttatus) Denomination Latrodectus toxin LV1 LV2 LV3 Toxic to vertebrates LV1 LV2 LV3/α-Latrotoxin LV4 α-Latrotoxin α-Latroinsectotoxin ß-Latroinsectotoxin γ-Latroinsectotoxin δ-Latroinsectotoxin ε-Latroinsectotoxin α-Latrocrustatoxin Low molecular-weight protein Latrodectin

Components One Three

Description Toxic to guinea pigs Cytotoxic

References (110,111) (1120

Toxic to insects Four

Seven

One One

Toxic to insects

(114)

Toxic to vertebrate Active on SRO Cloned (ALR) Cloned (ALR)

(114,113) (115) (116)

Cloned (ALR)

(118)

Cloned (ALR) Cloned Cloned

(148) (122) (121)

SRO, stretch receptor organ; ALR, ankyrin like repeats.

(109–111). Latrotoxins are proteins of about 1000 amino acid residues and share a high level of structure identity . Because of the existence of different toxins having different animal targets, we consider essential to summarize the most relevant features of the various toxins warning that in the literature, the effects of total venom were taken as effects of individual toxins. The potent neurotoxic activity of the venom of black widow spider has been under investigation for several years, the first attempts to purify the toxin dating in the early 1960s. Most of the biochemical work on the isolation and characterization of toxins was performed on the venom gland extract of the Mediterranean female L. tredecimguttatus (112–114). The molecular biology studies were mostly taken on the Asiatic varieties of L. mactans (115). Although the potent neurotoxic activity of the black widow spider venom has been studied in both vertebrates and invertebrates for several years, the actual number of active toxic components has not yet been clearly defined. The progress reached on the composition of the venom, from a chronological point of view, is reconstructed in Table 2. In a very detailed study Frontali et al. (114) verified early observations (112), suggesting the presence of distinct components responsible for toxicity to vertebrates and insects and described four distinct toxic molecules in the venom. As a result of the molecular biology approach (116–119) seven active components in the venom of L. mactans have been characterized. While the progression in the detection of active components mirrors the improvement of the techniques employed and represents a better understanding of the venom gland composi-

488

Grasso and Rufini

tion, description of the molecular organization of Latrodectus venom has not yet been completely achieved. In summary, the high-molecular mass protein toxins (110–130 kDa) of the venom constitute a family of molecules that stimulate the release of virtually all neurotransmitters from nerve endings. Five molecules active on insects are called latroinsectotoxins (α-, ß-, γ- δ-, and ε-), one toxic to crustacean: α-latrocrustatoxin. The structure of α-latrotoxin, α- and δ- latroinsectotoxin as well as α-latrocrustatoxin has been determined (115,118–120). In addition to latrotoxins, a small molecular mass protein, named latrodectin, has been characterized, sequenced, and implicated in venom action (121,122). 3.2. α-Latrotoxin as a Prototype of Latrotoxins Family 3.2.1. Molecular Structure of α-Latrotoxin Although it may appear as a tautological consideration, α-latrotoxin is indeed an exclusive product of the venom gland. Immunocytochemical localisation studies (111,123) and Northern-blot analysis confirmed its presence in responsive cells only within the venom gland (111). Identification, purification, and characterization of the molecule toxic to vertebrates (113), was initially described as mortality on newborn mice (113). Toxicity tests were later substituted by in vitro techniques following one of the biological effects elicited by toxin (124–126). These protocols have also been utilized to establish the specific activity (ED50) of the toxins, always found in the pico to nanomolar range. The α-latrotoxin structure has been defined and can be taken as representative of the general molecular arrangement of the entire group (family) of latrotoxins. The toxin derives by a post-translational cleavage, from a precursor having a molecular mass of 157 kDa. Four structural domains have been determined: (1) the leader sequence that differs structurally from the classic signal peptide, but has a reminiscent function; (2) the N-terminal domain consisting of about 430–450 amino acid residues; (3) the central domain containing about 15–20 ankyrin-like repeats (ankyrins constitute a family of proteins coordinating the interaction between integral membrane proteins and cytoskeletal elements of the cell); (4) the C-terminal domain consisting of about 160 amino acid residues. In this later domain the proteolytic modifications, that bring to the correct molecular mass, occur. To validate this hypothesis, in the hope that a mutational analysis of α-latrotoxin could facilitate the understanding of its mechanism of action, the recombinant toxin was produced in bacteria (119). Protein folding in bacteria was incorrect, whereas protein expression of the precursor in baculovirus gave a nonfunctional and insoluble product (117). Recombinant baculovirus carrying two furin (a subtilisin-like proteolytic enzyme) sites gave a product of 131.5 kDa (119) undistinguishable from the natural toxin with respect to: molecular mass, toxicity to mice, binding to latrotoxin receptors, electro-physiological response, and reactivity to anti-latrotoxin antibodies. Thus, the recombinant toxin, possessing all the properties of natural “wild” α-latrotoxin, suggests the occurrence, in nature, of a maturation event in the spider venom glands, reminiscent of the furin sites cleavage. The oligomerization of this unit was suggested to bring the formation of an active complex (127) to explain some effects of α-latrotoxin. From the analysis of the sequence of this molecule representing the functional unit, it is clear that the structure of α-latrotoxin does not carry any peculiar motif to predict


489

its toxic action. The ankyrin-like motif that characterizes the family of molecules may be implicated in a protein-protein interaction process only (128). 3.2.2. Mechanism of Action of α-Latrotoxin

A dramatic increase in the frequency of miniature postsynaptic potentials from nerve terminals of various types of neurons challenged with the toxin indicates that it may cause massive release of different neurotransmitters (for a review, see ref. 109). A primary target is obviously found in the NMJ of the peripheral nervous system (PNS), where the toxin apparently triggers exocytosis of clear synaptic vesicles (129). But α-latrotoxin also stimulates secretion of peptides and catecholamines, stored in large dense core vesicles of sensory neurons, endocrine, and neuroendocrine cells as well as from preparations from the CNS (synaptosomes and primary cultures) although it does not normally get access there (130). All these observations, suggest that α- latrotoxin acts on a general process that is common for cells having regulated secretion. In addition, it has repeatedly been reported that toxin activates and supports secretion even in the absence of added external calcium (131). The accepted notion is that α-latrotoxin, stimulates neurotransmitter release from neuronal cells via Ca2+-dependent as well as Ca2+-independent mechanisms. This notion is based on a series of observations in which release also occurs in absence of extracellular calcium. To validate this statement, we report here few recent examples. By the patch-clamp technique in conjunction with a fluorescent Ca2+ indicator to simultaneously measure the cytosolic Ca2+ concentration and ionic current (132), the action of α- latrotoxin in rat pituitary gonadotropes secreting the peptide LH, was studied. α-Latrotoxin triggers Ca2+-dependent exocytosis via extracellular Ca2+ entry as well as intracellular Ca2+ release, but in approx 25% of the cells, α-latrotoxin could also trigger a slow exocytosis in the absence of intracellular Ca2+ elevation. Therefore, α-latrotoxin has both Ca2+-dependent and Ca2+-independent actions in gonadotropes. The notion that the toxin has a signaling function in addition to its channel-forming capability is essentially based on the findings that α-latrotox independent secretion of insulin from pancreatic ß-cells occurs in the absence of any ion fluxes (133); that the entry of Ca2+ in the presynapses of the hippocampus is not necessary for toxin dependent transmitter release (134); and that facilitation of exocytosis produced by α-latrotoxin from cromaffin cells does not need Ca2+ fluxes (135). Thus, in answering the question: what are the effects consequent to the interaction of α-latrotoxin with the target structures, we can agree with the solutions suggested (136): (1) the interaction of α-latrotoxin with the receptor may cause cytoskeletal rearrangements such as those known to trigger exocytosis; (2) α-latrotoxin in itself functions as a fusogenic agent directly stimulating synaptic vesicles exocytosis; (3) the dockingfusion apparatus that activates exocytosis, is spontaneously and independently modified from the calcium trigger, after α-latrotoxin binding with the receptor; and (4) α-latrotoxin activates agonists for the receptor triggering in turn second-messenger signals that lead to a rapid activation of exocytosis. 3.2.3. Models for Stimulation of Exocytosis by α-Latrotoxin Because of the complexities reported previously, few models of exocytosis stimulation by α-latrotoxin were proposed (136–138). These models can take into account most of the results described; yet the emerging pictures remain rather unclear, since

490

Grasso and Rufini

α-latrotoxin, as a function of its concentration and cell type, is capable of exploiting more than one secretory pathway. The models proposed so far, are the following: 3.2.3.1. CHANNEL-FORMING EFFECTS α-Latrotoxin forms channels in plasma membranes. The toxic molecules aggregate, insert themselves into the phospholipid bilayer and form a channel like pore that stays permanently open. The pore is permeable to cations and thereby the membrane gets depolarized. The increase in cytosolic calcium concentration triggers exocytosis. Since its early description, α-latrotoxin has been known to form cations channels in artificial phospholipid bilayers (139,140) and this property has been recalled to explain the activation of exocytosis in culture cells (110,141). The interest in this suggestion, has been strengthened by a very recent paper (142) in which the three-dimensional structure of α-latrotoxin has been investigated. Accordingly, α-latrotoxin is dimeric in calciumfree conditions, but forms tetramers in the presence of divalent cations (calcium or magnesium). The oligomerization process allows the formation of an amphipathic pore forming complex from the essentially hydrophilic protein (monomer) (Fig. 2). In fact, tetramerization correlates with α-latrotoxin activity. Three characteristic domains were detected in the structure and named: the wing, the body, and the head; the latter concurs to form the channel in the middle of the tetrameric structure. By single particle cryoelectron microscopy, analysis of the tetramers inserted into artificial lipid bilayer of liposomes revealed several interesting conclusions: The tetramers insert into the membrane with the base, the base fully permeates the membrane while the upper part of the tetramere remains exposed on the membrane. Thus the pore is formed and the membrane permeabilized, meanwhile the wing may interact with receptors involved in synaptic vesicle exocytosis. The requirement for divalent cations in the formation of the complex brings us to the previous question related with the ion dependence of the toxin mechanism of action. 3.2.3.2. RECEPTOR-MEDIATED EFFECTS α-Latrotoxin has been shown to be the ligand for two types of cell surface proteins: neurexins (128) and latrophilin (143,144). In the presence of external calcium α-latrotoxin binds with high affinity both to neurexin Ia (a member of the family of cell-surface proteins with single transmembrane domain and large extracellular domains reminiscent of cell-adhesion molecules) and to a calcium-independent receptor (CIRL or latrophilin). The intracellular domain of neurexins mediates exocytosis by activating unknown intracellular signaling processes (137). In “zero calcium” conditions, α-latrotoxin mainly binds latrophilin (143), the receptor having seven transmembrane domains that resemble members of the secretin family of G protein-coupled receptors, and triggers exocytosis possibly via a second messenger-mediated pathway. The members of the secretin family activate G-proteins, which regulate a still unknown molecular cascade that ultimately leads to exocytosis (145). A third intermediate condition remains the most plausible (141): The receptors mainly serve to recruit the toxin and to facilitate its insertion in the membrane. Exocytosis results from the interaction of toxin with receptor molecules, combined with the control of the conductance of toxin-dependent ionic channels (144).


491

Fig. 2. Analysis of the α-latrotoxin structure. (A) A diagram of the α-latrotoxin sequence (top) and a linear representation of the domain structure of α-latrotoxin. The numbers above the diagram mark amino acid positions of region boundaries. N- and C- terminal regions removed proteolytically during α-latrotoxin maturation are colored black. Light gray is the fragment used to affinity purify the N-terminus specific antibodies (Ab). Small boxes (some of which are numbered underneath) represent ankyrin repeats. Imperfect or incomplete repeats and nonhomologous regions are colored gray. The “head” domain is dark gray. The wavy line is a coiled-coil region connecting the “wing” and “body” domains. The α-latrotoxin molecule has been thought to consist essentially of two major domains, the N-terminal domain (1–430) and the ankyrin repeats-containing domain (430–1179). However, the three dimensional structure of α-latrotoxin reveals that it contains three distinct, well-separated domains—“wing,” “body,” and “head”—that correspond to sequence A, B, and C. These regions, two of which (B and C) contain ankyrin repeats, are interconnected by relatively short sequences lacking homology with other proteins; these regions seem to form the narrow hinges between the threedimensional domains. (B) View of the monomer illustrating the three structural domains of α-latrotoxin: wing, body and head. The dotted line delimits the position of the central pore. (C) Immunolocalization of the N-terminal domain by cryo-EM in presence of Mg2+. The position of N-terminal antibodies associated with α-latrotoxin tetramers derives from selected analysis of a large number of images. Figure courtesy of Dr. Y. A. Ushkaryov (142).

492

Grasso and Rufini

3.3. Latroinsectotoxins It is evident from the consideration reported earlier, that α-latrotoxin is the toxic molecule of Latrodectus venom that has received the most attention thus far. However, in the venom of this spider, other toxins, primarily if not exclusively toxic to insect, have been described. Latroinsectotoxins seem to affect neurotransmission at insect NMJs and their mechanism of action seems to be reminiscent of that described for vertebrate synapses. This is also true for the structure of the latter. In fact, the cDNA encoding the putative sequences of two insectotoxins have been cloned and functionally expressed (118,120). The toxins are α- and δ-latroinsectotoxin, the structure of this last has been thoroughly investigated. The δ-latroinsectotoxin precursor comprises four structural domains: a signal peptide removed upon protein maturation, the N-terminal domain (this is the region exhibiting the highest degree of identity with other latrotoxins), the central region composed of 15 ankyrin-like repeats, and the C-terminal portion where the postranslational modification event (truncation) occurs. When expressed in bacteria, the precursor was inactive, whereas the truncated form, a protein of about 100 kDa, caused a massive neurotransmitter release at the locust NMJ (120). Similarly, native α-latroinsectotoxin (146) caused an increase of the frequency of miniature postsynaptic potential at the neuromuscular junction of blow fly larvae. Although the phenomenology of the action of these toxins on insects resembles that described for α-latrotoxin, no overlapping of their action in the various preparations tested (mouse and fly) was manifested. It remains to be explained what determines the specificity for different targets in the organization of molecules producing similar effects. 3.4. Latrocrustatoxin The high toxicity of the venom for Crustacean (terrestrial Crustacean, Isopoda, Oniscoidea, are frequently found caught in the web of L. tredecimguttatus) was taken into consideration since the early description of the effects of black widow spider venom on the cray fish stretch-receptor (147). This effect was attributed to the presence of a crustacean-specific toxin named latrocrustatoxin, which has recently been cloned and sequenced (148). The toxin has about 40% sequence homology with other latrotoxins. The maximum extent in structural homology is found in the N-terminal region, the most variable region being the C-terminus and the central domain consisting in tandem ankyrin like repeats. In a recent paper (149), analyzing the effects of a commercial preparation of toxin on the NMJ of the cray-fish, it was observed that it produces burstlike spontaneous discharge of transmitter quanta, accompanied by periodic fluctuations in the levels of calcium ions within the nerve terminal. The fluctuating periods of transmitter release correlated well with the periods of increased permeability of calcium ions. Other divalent cations (Sr and Ba) can support toxin effects, whereas La and Gd block exocytosis in the same manner as with α-latrotoxin. These effects can be explained by supposing an influx of extracellular calcium into nerve terminals. In addition, the properties of α-latrocrustatoxin acceptors are strongly reminiscent of the properties described for latrophilin/CIRL (145).


493

4. CONCLUSIONS Given the information reported here, it is evident that only a tiny percentage of spider venoms has been so far studied. Nothing is known about the venom of the several thousand different spiders known today and it is plausible that extremely interesting pharmacological data can be derived from a systematic study on the properties of their venom. However, it appears reasonable to hypothesize that, given a certain uniformity of behavior in catching, paralyzing, and killing the prey, an uniformity in the action of the molecules composing the venom fluid may exist. In other words, the neurotoxins present in the venom should always be directed against functional units that characterize the neurons: ionic channels and synaptic transmission. Thus, spider toxins will interfere with the functioning of the various ionic channels of the neuron or with the functioning of the chemical transmission either by altering presynaptic neurosecretion or by blocking the ligand action at the postsynaptic target. Although the effect on mammals is probably fortuitous, we can group the latrotoxins, among the neurotoxins having a defensive function, in other words sufficiently potent to provide protection against predators. However, in doing this, toxins seem to utilize few very efficient structural features. The key structural features of the inhibitor cystine knot motif are adopted by diverse polypeptides. Cone snails and spiders appear to be rich sources of polypeptides based on this structural motif and it is likely that protein engineering and phage-display technologies will contribute to the amplification of the range of molecules having this structure (150). An interesting development of these considerations derives from the recent observation (151) that α-latrotoxin shares structural homology with the glucagon-like bioactive peptides causing secretion of insulin. In fact, two apparently unrelated toxins (α-latrotoxin and exendin-4, a venom toxin of the salivary glands of the Gila monster Heloderma suspectum), not only show similarities in structure, but also the receptors that these toxins recognize and interact with, have structural similarities. Such observations provide evidence for a convergent form of molecular evolution, by which conserved structural motifs contributing to toxin-receptor interactions are generated by natural selection. ACKNOWLEDGMENTS We are grateful to Mario Pescatori (IBC, C. N. R., Rome) for helpful discussion during the course of this work, and Maria LoPonte for assisting in the english editing of the manuscript. The work of the authors was supported primarily by grants from the Consiglio Nazionale delle Ricerche (C. N. R.).The assistance of Y. A. Ushkaryov, who provided us with the images of Fig. 2, is gratefully acknowledged. M. Falconi assisted us in the elaboration of Fig. 1. REFERENCES 1. Savory, T. (ed.) (1977) Arachnida. Academic Press, London. 2. Newlands, G. and Atkison, P. (1988) Review of southern African spiders of medical importance, with notes on signs and symptoms of envenomation. S. Afr. Med. J. 73, 235–239. 3. Grishin, E. (1999) Polypeptide neurotoxins from spider venoms. Eur. J. Biochem. 264, 276–280.

494

Grasso and Rufini

4. Snutch, T. P., Leonard, J. P., Gilbert, M. M., Lester, H. A., and Davidson, N. (1990) Rat brain expresses a heterogeneous family of calcium channels. Proc. Natl. Acad. Sci USA 87, 3391–3395. 5. Tsien, R. W., Ellinor, P. T., and Horna, W. A. (1991) Molecular diversity of voltagedependent Ca2+ channels. Trends Pharmacol. Sci. 12, 349–354. 6. Oliveira, B. M, Miljanih, G. P., Ramachandran, J., and Adams, M. E. (1994) Calcium channel diversity and neurotransmitter release: the omega-conotoxin and omega-agatoxin. Annu. Rev. Biochem. 63, 823–867. 7. Jackson, H., Urnes, M., and Parks, T. N. (1986) Presynaptic blockade of transmission by a potent, long-lasting toxin from Agelenopsis aperta spiders. Neurosci. Abstr. 12, 730. 8. Mintz, I. M., Venema, V. J., Swiderek, K. M., Lee, T. D., Bean, B. P., and Adams, M. E. (1992) P-type calcium channels in rat central and peripheral neurons. Neuron 9, 1–20. 9. Bindokas, V. P. and Adams, M. E. (1989) omega-Aga-I: a presynaptic calcium channel antagonist from venom of the funnel web spider Agelenopsis aperta. J. Neurobiol. 20, 171–188. 10. Santos, A. D., Imperial, J. S., Chaudhary, T., Beavis, R. C., Chait, B. T., Hunsperger J. P., et al. (1992) Heterodimeric structure of the spider toxin ω-agatoxin IA revealed by precursor analysis and mass spectrometry. J. Biol. Chem. 267, 20,701–20,705. 11. Adams, M. E., Bindokas, V. P., Hasegawa, L., and Venema, V. J. (1990) Omega-agatoxins: a novel calium hannel antagonists of two subtypes from funnel web spider (Agelenopsis aperta) venom. J. Biol. Chem. 265, 861–867. 12. Shinde, U. P., Liu, J. J., and Inouye, M. (1997) Protein memory through altered folding mediated by intramolecular chaperones. Nature 389, 520–522. 13. Diniz, M. R. V., Paine, M. J. I., Diniz, C. R., Theakston, R. D. G., and Crampton, J. M. (1993) Sequence of the cDNA coding for the lethal neurotoxin TX1 from the brazilian “armed” spider Phoneutria nigriventer predicts the synthesis and processing of a preprotoxin. J. Biol. Chem. 268, 15,340–15,342. 14. Venema, V. J., Swiderek, K. M., Lee, T. D., Hathaway, G. M., and Adams, M. E. (1992) Antagonism of synaptosomal calcium channels by subtypes of omega-agatoxins. J. Biol. Chem. 267, 2610–2615. 15. Ertel, E. A., Warren V. A., Adams, M. E., Griffin, P. R., Cohen, C. J., and McHardy, M. S. (1994) Type III ω-agatoxins: a family of probes for similar binding sites on L- and N-type calcium channels. Biochemistry 33, 5098–5108. 16. Mintz, I. M., Venema, V. J., Swiderek, K. M., Lee, T. D., Bean, B. P., and Adams, M. E. (1992) P-type calcium channels blocked by the spider toxin ω-Aga-IVA. Nature 355, 827–829. 17. Adams, M. E., Mintz, I. M., Reily, M. D., Venkataraman, T., and Bean B. P. (1993) Structure and properties of ω-Agatoxin-IVB, a new antagonist of P-type calcium channels. Mol. Pharmacol. 44, 681–688. 18. Sather, W. A., Tanabe, T., Zhang, J.-F., Mori, Y., Adams, M. E., and Tsien, R. W. (1993) Distinctive biophysical properties of class A (BI) calcium channel alpha 1 subunit. Neuron 11, 291–303. 19. Randall, A. and Tsien, R. W. (1995) Pharmacological dissection of multiple types of Ca2+ channel currents in rat cerebellar granule neurons. J. Neurosci. 15, 2995–3012. 20. Wu, L.-G. and Saggau, 1994 Pharmacological identification of two types of presinaptic voltage-dependent calcium channels at CA3-CA1 synapses of the hippocampus. J. Neurosci. 14, 5613–5622. 21. Bargas, J., Howe, A., Eberwine, J., Cao, Y., and Surmeier, D. J. (1994) Cellular and molecular characterization of Ca2+ currents in acutely isolated, adult rat neostriatal neurons. J. Neurosci. 14, 6667–6686. 22. Turner, T. J., Adams, M. E., and Dunlap, K. (1992) Calcium channels coupled to glutamate release identified by ω-Aga-IVA. Science 258, 310–313.


495

23. Wheeler, D. B., Randall, A., and Tsien, R. W. (1994) Roles of N-type and Q-type Ca2+ channels in supporting hippocampal synaptic transmission. Science 264, 107–111. 24. Villaroya, M., Olivares, R., Ruiz, A., Cano-Abad, M. F., de Pascual, R., Lomax, R. B., et al. (1999) J. Physiol. (Lond.) 516, 421–432. 25. Pocock, J. M., Cousin, M. A., and Nicholls, D. G. (1993) The calcium channel coupled to the exocytosis of L-glutamate from cerebellar granule cells is inhibited by the spider toxin AGA-GI. Neuropharmacology 32, 1185–1194. 26. Regehr, W. G. and Mintz, I. M. (1994) Partecipation of multiple calcium channel types in transmission at single climbing fiber to Purkinje cell synapses. Neuron 12, 605–613. 27. Vogel, S. S., Smith, R. M., Baibakov, B., Ikebuchi, Y., and Lambert, N. A. (1999) Calcium influx is required for endocytotic membrane retrieval. Proc. Natl. Acad. Sci. USA 96, 5019–5024. 28. McDonough, S. I., Mintz, I. M., and Bean, B. P. (1997) Alteration of P-type calcium channel gating by spider toxin ω-Aga-IVA. Biophys. J. 72, 2117–2128. 29. Kim, J. I., Konishi, S., Iwai, H., Kohno, T., Gouda, H., Shimada, I., et al. (1995) Threedimensional solution structure of the calcium channel antagonist ω-Agatoxin IVB: consensus molecular folding of calcium channel blockers. J. Mol. Biol. 250, 659–671. 30. Davis, J. H., Brandley, E. K., Miljanich, G. P., Nadasdi, L., Ramachandran, J., and Basus, V. J. (1993) Solution structure of ω-conotoxin GVIA using 2-D NMR spectroscopy and relaxation matrix analysis. Biochemistry 32, 7396–7405. 31. Yu, H., Rosen, M. K., Saccomano, N. A., Philips, D., Volkmann, R., and Schreiber, S. L. (1993) Sequential assignment and structure determination of spider toxin ω-Aga-IVB. Biochemistry 32, 13,123–13,129. 32. Reily, M. D., Holub, K. E., Gray, W. R., Norris, T. M., and Adams, M. E. (1994) Structureactivity relationship for P-type calcium channel-selective ω-agatoxins. Struct. Biol. 1, 853–856 33. Herrero, C. J., Garcia-Palomero, E., Pintado, A. J., Garcia, A. G., and Montiel, C. (1999) Differential blockade of rat alpha3beta4 and alpha7 neuronal nicotinic receptors by omegaconotoxin MVIIC, omega-conotoxin GVIA and diltiazem. Br. J. Pharmacol. 127, 1375–1387. 34. Liang, S. P., Zhang, D. Y., Pan, X., Chen, Q.,and Zhou, P. A. (1993) Properties and aminoacid sequence of huwentoxin-I, a neurotoxin purified from the venom of the Chinese bird spider Selenocosmia huwena. Toxicon 31, 969–978. 35. Zhou, P. A., Xie, X. J., Li, M., Yang, D. M., Xie, Z. P., Zong, X., and Liang, S. P. (1997) Blockade of neuromuscular transmission by huwentoxin-I, purified from the venom of the Chinese bird spider Selenocosmia huwena. Toxicon 35, 39–45. 36. Qu, Y., Liang, S., Ding, J., Ma, L., Zhang, R., and Gu, Y. (1995) Proton nuclear magnetic resonance studies on Huwetnoxin-I from the venom of the spider Selenocosmia huwena: 1. Sequence-specific 1H-NMR assignements. J. Prot. Chem. 14, 549–557. 37. Qu, Y., Liang, S. P., Ding, J., Liu, J., Zhang, R., and Gu, X. (1997) Proton nuclear magnetic resonance studies on Huwentoxin-I from the venom of the spider Selenocosmia huwena: 2. Three-dimensional structure in solution. J. Prot. Chem. 16, 565–574. 38. Sutton, K. G., Siok, C., Stea, A., Zamponi, G. W., Heck, S. D., Volkmann, R. A, et al. (1998) Inhibition of neuronal calcium channels by a novel peptide spider toxin, DW13.3. Mol. Pharmacol. 54, 407–418. 39. Strapleton, A., Blankenship, D. T., Ackermann, B. L., Chen, T.-M., Gorder, G. W., Manley, G.D., et al. (1990) Curtatoxins. Neurotoxic insecticidal polypeptides isolated from the funnel-web spider Hololena curta. J. Biol. Chem. 265, 2054–2059. 40. Lundy, P. M., Hong, A., and Few, R. (1992) Inhibition of a dihydropyridine, omegaconotoxin insensitive Ca2+ channel in rat synaptosomes by venom of the spider Hololena curta. Eur. J. Pharmacol. 225, 51–56.

496

Grasso and Rufini

41. Lampe, R. A., Defeo, P. A., Davison, M. D., Young, J., Herman, J. L., Spreen, R. C., et al. (1993) Isolation and pharmacological characterization of ω-grammotoxin-SIA, a novel peptide inhibitor of neural voltage-sensitive calcium channel responses. Mol. Pharmacol. 44, 451–460. 42. Newcomb, R., Palma, A., Fox, J., Gaur, S., Lau, K., Chung, D., et al. (1995) SNX-325, a novel calcium antagonist from the spider Segestria florentina. Biochemistry 34, 8341–8347. 43. Newcomb, R., Szoke, B., Palma, A., Wang, G., Chen, X.-H., Hopkins, W., et al. (1998) Selective peptide agonist of the class E calcium channel from the venom of the tarantula Hysterocrates gigas. Biochemistry 37, 15,353–15,362. 44. Branton, W. D., Rudnick, M. S., Zhou, Y., Eccleston, E. D., Fields, G. B., and Bowers, L. D. (1993) Fatty acylated toxin structure. Nature 365, 496–49 45. Skinner, W. S., Adams, M. E., Quistad, G. B., Kataoka, H., Cesarin, B. J., Enderlin, F. E., and Schooley, D. A. (1989) Purification and characterization of two classes of neurotoxins from the funnel web spider, Agelenopsis aperta. J. Biol. Chem. 264, 2150-2155 46. Leisy, D. J., Mattson, J.D., Quistad, G. B., Kramer, S. J., Van Beek, N., Tsai, L. W., et al. (1996) Molecular cloning and sequencing of cDNAs encoding insecticidial peptides from the primitive hunting spider Plectreurys tristis. Insect. Biochem. Mol. Biol. 26, 411–441. 47. Quistad, G. B. and Skinner, W. S. (1994) Isolation and sequencing of insecticidial peptides from the primitive hunting spider, Plectreurys tristis (Simon). J. Biol. Chem. 269, 11,098– 11,101. 48. Craig, A. G., Bandyopadhyay, P., and Olivera, B. M. (1999) Post-translationally modified neuropeptides from Conus venom. Eur. J. Biochem. 264, 271–275. 49. Wang, G., Dayanithi, G., Newcomb, R., and Lemos, J. R. (1999) An R-type Ca2+ current in neurohypophysial terminals preferentially regulates oxytocin secretion. J. Neurosci. 19, 9235–9241. 50. McDonough, S. I., Lampe, R. A., Keith, R. A., and Bean, B. P. (1997) Voltage-dependent inhibition of N- and P-type calcium channels by the peptide toxin ω-grammotoxin-SIA. Mol. Pharmacol. 52, 1095–1104. 51. Fazal, A., Beg, O. U., Shafqat, J., Zaidi, Z. H., and Jornvall, H. (1989) Characterization of two different peptides from the venom of the scorpion Buthus indicus. FEBS Lett. 257, 260–262. 52. Pallaghy, P. K., Duggan, B. M., Pennington, M. W., and Norton, R. S. (1993) Threedimensional structure in solution of the calcium channel blocker ω-conotoxin. J. Mol. Biol. 234, 405–420. 53. Cammue, B. P., De Bolle, M. F., Terras, F. R., Proost, P., Van Damme, J., Rees, S. B., et al. (1992) Isolation and characterization of a novel class of plant antimicrobial peptides from Mirabilis jalapa L. seeds. J. Biol. Chem. 267, 2228–2232. 54. Li-Smerin, Y. and Swartz, K. J. (1998) Gating modifier toxins reveal a conserved structural motif in voltage-gated Ca2+ and K+ channels. Proc. Natl. Acad. Sci. USA 95, 8585–8589. 55. Omecinsky, D. O., Holub, K. E., Adams, M. E., and Reily, M. D. (1996) Three-dimensional structure analysis of µ-agatoxins: further evidence for common motifs among neurotoxins with diverse ion specificities. Biochemistry 35, 2836–2844. 56. Romey, G., Abita, J. B., Chicheportiche, R., Rochat, H., and Lazdunski, M. (1976) Scorpion neurotoxins. Mode of action on neuromuscular junctions and synaptosomes. Biochem. Biophys. Acta 448, 607–619. 57. Gage, P. W. and Spence, I. (1977) The origin of the muscle fasciculation caused by funnelweb spider venom. Aust. J. Exp. Biol. Med. Sci. 55, 453–461. 58. Spence, I. (1979) Electrophysiological studies of the venoms of the funnel-web spider Atrax robustus and the tick Ixodes holocyclus, in Neurotoxins: Fundamental and Clinical Advances (Chubb, I. W. and Geffen, L. B., eds.), Adelaide University Union Press, Adelaide, pp. 161–173.


497

59. Spence, I., Adams, D. J., and Gage, P. W. (1977) Funnel web spider venom produces spontaneus action potentials in nerve. Life Sci. 20, 243–250. 60. Sheumack, D. D., Claassens, R., Whiteley, N. M., and Howden, M. E. H. (1985) Complete amino acid sequence of a new type of lethal neurotoxin from the venom of the funnel-web spider Atrax robustus. FEBS Lett. 181, 154–156. 61. Mylecharane, E. J., Spence, I., Sheumack, D. D., Claassens, R., and Howden, M. E. H. (1989) Actions of robustotoxin a neurotoxic polypeptide from the venom of the male funnel-web spider (Atrax robustus), in anaesthetized monkeys. Toxicon 27, 481–492. 62. Temple, H. D., Hinds, M. G., Sheumack, D. D., Howden, M. E., and Norton, R. S. (1999) 1H NMR study of robustotoxin, the lethal neurotoxin from the funnel web spider Atrax robustus. Toxicon 37, 485–506. 63. Brown, M. R., Sheumack, D. D., Tyler, M. I., and Howden, M. E. H. (1988) Amino acid sequence of versutoxin, a lethal neurotoxin from the venom of the funnel-web spider Atrax versutus. Biochem. J. 250, 401–405. 64. Nicholson, G. M., Willow, M., Howden, M. E. H., and Narahashi T. (1994) Modification od sodium channel gating and kinetics by versutoxin from the Australian funnel-web spider Hydronyche versuta. Pflugers Arch. 428, 400–440. 65. Nicholson, G. M., Little M. J., Tyler, M. I., and Narahashi T. (1996) Selective alteration of sodium channel gating by Australian funnel-web spider toxins. Toxicon 34, 14–43. 66. Nicholson, G. M., Walsh, R., Little M. J., and Tyler, M. I. (1998) Characterisation of the effects of robustotoxin, the lethal neurotoxin from the Sydney funnel-web spider Atrax robustus, on sodium channel activation and inactivation. Pflugers Arch. 436, 117–126. 67. Catteral, W. A. (1995) Structure and function of voltage-gated ion channels. Annu. Rev. Biochem. 64, 493–531. 68. Gordon, D. (1997) A new approach to insect-pest control-combination of neurotoxins interacting with voltage sensitive sodium channels to increase selectivity and specificity. Invert. Neurosci. 3, 103–116. 69. Little, M. J., Zappia, C., Gilles, N., Connor, M., Tyler, M. I., Martin-Eauclair, M. F., et al. (1998) Delta-Atracotoxins from Australian funnel-web spiders compete with scorpion alpha-toxin binding but differentially modulate alkaloid toxin activation of voltage-gated sodium channels. J. Biol. Chem. 273, 27,076–27,083. 70. Khodorov, B. I. and Revenko, S. V. (1979) Further analysis of the mechanism of action of batrachotoxin on the membrane of myelinated nerve. Neuroscience 4, 1315–1330. 71. Nicholson, G. M., Walsh, R., Little M. J., and Tyler, M. I. (1998) Characterisation of the effects of robustotoxin, the lethal neurotoxin from the Sydney funnel-web spider Atrax robustus, on sodium channel activation and inactivation. Pflugers Arch. 436, 117–126 72. Fletcher, J. I., Chapman, B. E., Mackay, J. P., Howden, M. E., and King, G. F. (1997) The structure of versutoxin (delta-atracotoxin-Hv1) provides insights into the binding of site 3 neurotoxins to the voltage-gated sodium channel. Structure 15, 1525–1535. 73. Meir, A., Ginsburg, S., Butkevich, A., Kachalsky, S. G., Kaiserman, I., Ahdut, R., et al. (1999) Ion channels in presynaptic nerve terminals and control of transmitter release. Physiol. Rev. 79,1019–1088. 74. Miller, C., Moczydlowski, E., Latorre, R., and Philips, M. (1985) Charybdotoxin, a protein inhibitor of single Ca2+ -activated K+ channels from mammalian skeletal muscle. Nature 313, 316–318. 75. Possani, L. D., Martin, B. M., and Svendsen, I. B. (1982) The primary structure of noxiustoxin: a K+ channel blocking peptide, purified from the venom of the scorpion Centruroides noxius hoffmani. Carlsberg Res. Commun. 47, 285–289. 76. Penner, R., Petersen, M., Pierau, F. K., and Dreyer, F. (1986) Dendrotoxin: a selective blocker of a non-inactivating potasium current in guinea-pig dorsal root ganglion neurones. Pflugers Arch. 407, 365–369.

498

Grasso and Rufini

77. Grissmer, S., Nguyen, A. N., Aiyar, J., Hanson, D. C., Mather, R. J., Gutman, G. A., et al. (1994) Pharmacological characterization of five cloned voltage-gated K+ channels, types 1.1, 1.2, 1.3, 1.5, 3.1 stably expressed in mammalian cell lines. Mol. Pharmacol. 45, 1227–1234. 78. Sanguinetti, M. C., Johnson, J. H., Hammerland, L. G., Kelbaugh, P. R., Volkmann, R. A., Saccomanno, N. A., and Mueller A. L. (1997) Heteropodatoxins: peptides isolated from spider venom that block Kv4.2 potassium channels. Mol. Pharmacol. 51, 491–498. 79. Swartz, K. J. and MacKinnon, R. (1995) An inhibitor of the Kv2.1 potassium channel isolated from the venom of a Chilean Tarantula. Neuron 15, 941–949. 80. Swartz, K. J. and MacKinnon, R. (1997) Hantoxin modifies the gating of a voltage-dependent K+ channel through multiple binding sites. Neuron 18, 665–673. 81. Swartz, K. J. and MacKinnon, R. (1997) Mapping the receptor site for hanatoxin a gating modifier of voltage-dependent K+ channels. Neuron 18, 675–682. 82. Kushmerick, C., Kalapothakis, E., Beirao, P. S. L., Penaforte, C. L., Prado, V. F., Cruz, J. S., et al. (1999) Phoneutria nigriventer toxin Tx3-1 blocks A-type K+ currents controlling Ca2+ oscillation frequency in GH3 cells. J. Neurochem. 72,1472–1481. 83. Marvin, L., De, E., Cosette, P., Gagnon, J., Molle, G., and Lange, C. (1999) Isolation, amino acid sequence and functional assay of SGTx1. The first toxin purified from the venom of the spider Scodra griseipes. Eur. J. Biochem. 265, 572–579. 84. Kaiser, I. I., Griffin, P. R., Aird, S. D., Hudiburg, S., Shabanowitz, J., Francis, B., Jhon, T. R., et al. (1994) Primary structures of two proteins from the venom of the mexican red knee tarantula (Brachypelma smithii) Toxicon 32, 1083–1093. 85. Bateman, A., Boden, P., Dell, A., Duce, I. R., Quicke, D. L. J., and Usherwood, P. N. R. (1985) Postsynaptic block of glutamatergic synapse by low molecular weight fractions of spider venom. Brain Res. 339, 237–244. 86. Kerry, C. J., Ramsey, R. L., Sansom, M. S. P., and Usherwood, P. N. R. (1988) Single channel studies on noncompetitive antagonism of a quisqualate-sensitive glutamate receptor by argiotoxin-636- a fraction isolated from orb-weaver spider venom. Brain Res. 459, 312–327. 87. Kawai, N., Miwa, A., Saito, M., Pan-Hou, H. S., and Yoshioka, M. (1984) Spider toxin (JSTX) on the glutamate synapse. J. Physiol. (Paris) 79, 228–231. 88. Quistad, G.B., Suwanrumpha, S., Jarema, M.A., Shapiro, M.J., Skinner, W.S., Jamieson, G.C., et al. (1990) Structures of paralytic acylpolyamines from the spider Agelenopsis aperta. Biochem. Biophys. Res. Commun. 169, 51–56. 89. Eldefrawi, A. T., Eldefrawi M. E., Konno, K., Mansour, N. A., Nakanishi, K., Oltz, E., and Usherwood, P. (1988) Structure and synthesis of a potent glutamate receptor antagonist in wasp venom. Proc. Natl. Acad. Sci. USA 85, 4910–4913. 90. Kawai, N., Miwa, A., and Abe, T. (1982) Spider venom contains specific receptor blocker of glutaminergic synapses. Brain Res. 247, 169–171. 91. Kawai, N., Yamagishi, S., Saito, M., and Furuya, K. (1983) Blockade of synaptic transmission in the squid giant synapse by a spider toxin (JSTX). Brain Res. 278, 346–349. 92. Abe, T., Kawai, N., and Miwa, A. (1983) Effects of a spider toxin on the glutaminergic synapse of lobster muscle. J. Physiol. 339, 243–252. 93. Jackson, H. and Usherwood, P. N. R. (1988) Spider toxins as a tool for dissecting elements of excitatory amino acid transmission. Trends Neurosci. 10, 278–283. 94. Akaike, N., Kawai, N., Kiskin, N., Kljuchko, E. M., Krishtal, O. A., and Tsyndrenko, A. Y. (1987) Spider toxin blocks excitatory amino acid responses in isolated hippocampal pyramidal neurons. Neurosci. Lett. 79, 326–330. 95. Sayto, M., Sahara, Y., Miwa, A., Shimazaki, K., Nakajima, T., and Kawai, N. (1989) Effect of a spider toxin (JSTX) on hippocampal CA1 neurons in vitro. Brain Res. 481, 16–24. 96. Sahara, Y., Robinson, H. P. C., Miwa, A., and Kawai, N. (1991) The effect of spider toxin (JSTX-3) and zinc on excitatory postsynaptic currents in CA1 pyramidal cells of guineapig hippocampal slice. Neurosci. Res. 10, 200–210.


499

97. Savidge, J. R. and Bristow, D. R. (1998) Ca2+ permeability and joro spider toxin sensitivity of AMPA and kainate receptors on cerebellar granule cells. Eur. J. Pharmacol. 351, 131–138. 98. Blaschke, M., Keller, B. U., Rivosecchi, R., Hollmann, M., Heinemann, S., and Konnerth, A. (1993) A single amino acid determines the subunit-specific spider toxin block of a-amino-3-hydroxy-5-methylisoxalone-4-propionate/kainate receptor channels. Proc. Natl. Acad. Sci. USA 90, 6528–6532. 99. Herlitze, S., Raditsch, M., Ruppersberg, J. P., Jahn, W., Monyer, H., Schoepfer, R., and Witzemann, V. (1993) Argiotoxin detects molecular differences in AMPA receptor channels. Neuron 10, 1131–1140. 100. Mueller, A., Albensi, R. C., Ganong, A. H., Reynols, L. S., and Jackson, J. (1992) Arylamine spider toxins antagonize NMDA receptor-mediated synaptic transmission in rat hippocampal slices. Synapse 9, 244–250. 101. Draguhn, A., Jahn, W., and Witzemann, V. (1991) Argiotoxin 636 inhibits NMDA-activated ion channels expressed in Xenopus oocytes. Neurosci. Lett. 132, 187–190. 102. Priestley, T., Woodruff, G. N., and Kemp J. A. (1989) Antagonism of responses to excitatory aminoacids on rat cortical neurones by the spider toxin, argiotoxin636. Br. J. Pharmacol. 97, 1315–1323. 103. Reynolds I. J. (1991) The spider toxin argiotoxin636, binds to a Mg2+ site on the N-methyl-D-aspartate receptor complex. Br. J. Pharmacol. 103, 1373–1376 104. Raditsch, M., Ruppersberg, J. P., Kuner, T., Gunther, W., Schoepfer, R., Seeburg, P.H., et al. (1993). Subunit-specific block of cloned NMDA receptros by argiotoxin636. Fedn. Eur. Biochem. Socs. Lett. 324, 63–66 105. Donevan, S. D. and Rogawski, M. A. (1996) Multiple actions of arylalkylamine arthropod toxins on the N-methyl-D-aspartate receptor. Neuroscience 70, 361–375. 106. Jackson, H. and Parks, T. N. (1989) Spider Toxins: recent application in neurobiology. Annu. Rev. Neurosci. 12, 405–414. 107. Kiskin, N. I., Chizhmakov, I. V., Tsyndrenko, A. Y., Mueller, A. L., Jackson, H., and Khrishtal (1992) A hyghly potent and selective N-methyl-D-aspartate receptor antagonist from the venom of the Agelenopsis aperta spider. Neuroscience 51, 11–18. 108. Parks, T. N., Mueller, A. L., Artman, L. D., Albensi, B. C., Nemeth, E. F., Jackson, H., et al. (1991) Arylamine toxins from funnel-web spider (Agelenopsis aperta) venom antagonize N-methyl-D-aspartate receptor functions in mammalian brain. J. Biol. Chem. 266, 21,523–21,529. 109. Rosenthal, L. and Meldolesi, J. (1989) α-Latrotoxin and related toxins. Pharmac. Ther. 42, 115–134. 110. Grasso, A. (1988) α-Latrotoxin as a tool for studying ionic channels and transmitter release process, in Neurotoxins in Neurochemistry (Dolly, O., ed.), J. Wiley and Sons, Chichester, New York, pp. 67–78. 111. Grasso A., Kawai, S., Kobayashi, M., and.Pescatori, M. (1998). Purification, function and selectivity in α-latrotoxin, in Secretory Systems and Toxins (Lazarovici, P., Grasso, A., and Linial, M., eds.), Harwood Academic, Amsterdam, pp. 333–355. 112. Frontali, N. and Grasso, A. (1964) Separation of three toxicologically different protein components from the venom of the spider Latrodectus tredecimguttatus. Arch. Biochem. Biophys. 106, 213–218. 113. Grasso, A. (1976) Preparation and properties of a neurotoxin purified from the venom of black widow spider (Latrodectus mactans tredecimguttatus). Biochim. Biophys. Acta 439, 409–412. 114. Frontali, N., Ceccarelli, B,. Gorio, A. Mauro, A., Tzeng, M. C., and Hurlbut., W. P (1976) Purification from black widow spider venom of a protein factor causing the depletion of synaptic vesicles at neuromuscular junctions. J. Cell Biol. 68, 462–479.

500

Grasso and Rufini

115. Kiyatkin, N. I., Dulubova, I. E., Chekhovskaya, I. A., and Grishin, E. V. (1990) Cloning and structure of cDNA encoding α-latrotoxin from black widow spider venom. FEBS Lett. 270, 127–131. 116. Krasnoperov, V. G., Shamotienko, O. G., and Grishin, E. V. (1992) Isolation and properties of insect and crustacean specific neurotoxins from the venom of the black widow spider (Latrodectus mactans tredecimguttatus). J. Nat. Toxins 1, 17–23. 117. Kiyatkin, N., Kulikovskaya, I. M., Grishin, E., Beadle, D. J., and King, L. (1995) Functional characterization of black widow spider neurotoxins synthesised in insect cells. Eur. J. Biochem. 230, 854–859. 118. Dulubova, I.E., Krasnoperov, V.G., Khvotchev, M.V., Pluzhnikov K.A., Volkova, T.M., Grishin, E.V., et al. (1996) Cloning and structure of δ-latroinsectotoxin, a novel insectspecific member of the latrotoxin family. J. Biol. Chem. 271, 7535–7543. 119. Volynski, K. E., Nosyreva, E. D., Ushkaryov, Y. A., and Grishin, E. V. (1999) Functional expression of a-latrotoxin in baculovirus system. FEBS Lett. 442, 25–28. 120. Kiyatkin, N., Dulubova, I., and Grishin, E. (1993) Cloning and structural analysis of alpha-latroinsectotoxin cDNA. Abundance of ankyrin-like repeats. Eur. J. Biochem. 213, 121–127. 121. Pescatori, M., Bradbury, A., Bouet, F., Gargano, N., Mastrogiacomo, A., and Grasso, A. (1995) The cloning of a cDNA encoding a protein (Latrodectin) which co-purifies with the α-latrotoxin from the black widow spider Latrodectus tredecimguttatus (Theridiidae). Eur. J. Biochem. 230, 322–328. 122. Gasparini, S., Kiyatkin, N., Drevet, P., Boulain, J. C., Tecnet, F., Ripoche, P., et al. (1994) The low molecular weight protein which purifies with α-latrotoxin is structurally related to crustacean hyperglycemic hormones. J. Biol. Chem. 269, 19,803–19,809. 123. Cavalieri, M., Corvaja, N., and Grasso, A.(1990) Immunocytological localization by monoclonal antibodies of α-latrotoxin in the venom gland of the spider Latrodectus tredecimguttatus. Toxicon 28, 341–346. 124. Grasso, A., Rufini, S., and Senni, M. (1978) Concanavalin A blocks black widow spider toxin stimulation of transmitter release from synaptosomes. FEBS Lett. 85, 241–244. 125. Grasso, A. and Senni, M. I. (1979) A toxin purified from the venom of black widow spider affects the uptake and release of radioactive γ-amino butyrate and N-epinephrine from rat brain synaptosomes. Eur. J. Biochem. 102, 337–344. 126. Grasso, A. and Mastrogiacomo, A. (1992). α-Latrotoxin: preparation and effects on calcium fluxes. FEMS Microbiol. Immunol. 105, 131–138. 127. Lunev, A. V., Demin, V. V., Zaitsev, O. I., Spadar, S. I., and Grishin, E.V. (1991) Electronmicroscopy of α-latrotoxin from the venom of black widow spider Latrodectus mactans tredecimguttatus. Bioorg. Khim. 17, 1021–1026. 128. Geppert, M., Ushkaryov, Y. A., Hata, Y., Davletov, B., Petrenko, A. G., and Sudhof, T. C. (1992) Neurexins, Cold Spring Harb. Symp. Quant. Biol. 57, 483–490. 129. Matteoli, M., Haimann, C., Torri-Tarelli, F., Polak, J. M., Ceccarelli, B., and De Camilli, P. (1988) Differential effect of alpha-latrotoxin on exocytosis from small synaptic vesicles and from large dense-core vesicles containing calcitonin gene-related peptide at the frog neuromuscular junction. Proc. Natl. Acad. Sci. USA 85, 7366–7370. 130. Pothos, E. N., Davila, V., and Sulzer, D. (1998) Presynaptic recording of quanta from midbrain dopamine neurons and modulation of the quantal size. J. Neurosci. 18, 4106–4118. 131. Misler, S. and Falke, L. C. (1987) Dependence on multivalent cations of quantal release of transmitter induced by black widow spider venom. Am. J. Physiol. 253, 469–476. 132. Tse, F. W. and Tse, A. (1999) Alpha-latrotoxin stimulates inward current, rise in cytosolic calcium concentration, and exocytosis at pituitary gonadotropes. Endocrinology 140, 3025–3033. 133. Lang, J., Ushkaryov, Y., Grasso, A., and Wollheim, C. B. (1998). Ca2+ independent insulin exocytosis induced by α-Latrotoxin requires latrophilin, a G Protein-coupled receptor. EMBO J. 17, 648–657.


501

134. Capogna, M., Gahwiler, B. H., and Thompson, S. M. (1996) Calcium-independent actions of alpha-latrotoxin on spontaneous and evoked synaptic transmission in the hippocampus. J. Neurophysiol. 76, 3149–3158. 135. Michelena, P., dela Fuente, M. T., Vega, T., Lara, B., Lopez, M. G., Gandia, L., and Garcia, A. G. (1997) Drastic facilitation by α-latrotoxin of bovine chromaffin cell exocytosis without measurable enhancement of Ca2+ entry or [Ca2+ ]i. J. Physiol. 502, 481–496. 136. Petrenko, A. G. and Krasnoperov, V. G. (1998) α-Latrotoxin receptors, in Secretory Systems and Toxins (Linial, M., Grasso, A., and Lazarovici, P., eds.), Harwood Academic, Amsterdam, pp. 185–212. 137. Petrenko, A. G. (1993) α-Latrotoxin receptor. Implications in nerve terminal function. FEBS Lett. 325, 81–85. 138. Henkel, A. and Sankaranarayanan, S. (1999) Mechanisms of α-latrotoxin action. Cell Tissue Res. 296, 229–233. 139. Finkelstein, A., Rubin, L. L., and Tzeng, M. C. (1976) Black widow spider venom: effect of purified toxin on lipid bilayer membranes. Science 193, 1009–1011. 140. Robello, M., Rolandi, R., Alemà, S., and Grasso A. (1984) trans-Bilayer orientation and voltage dependence of α-latrotoxin induced channels. Proc. R. Soc. Lond. B. 220, 474–487 141. Grasso, A., Alemà, S., Rufini, S., and Senni, M. I. (1980) Black widow spider toxininduced calcium fluxes and transmitter release in a neurosecretory cell line. Nature 283, 774–776. 142. Orlova, E. V., Rahaman, A. M., Gowen, B., Volynski, K. E., Ashton, A. C., Manser, C., et al. (2000) Structure of α-latrotoxin oligomers reveals that divalent cation-dependent tetramers form membrane pores. Nature Struct. Biol. 7, 48–53. 143. Davletov, B. A., Shamotienko, O. G., Lelianova, V. G., Grishin E. V., and Ushkaryov, Y. A. (1996) Isolation and biochemical characterization of a Ca2+-independent α-latrotoxinbinding protein. J. Biol. Chem. 271, 23,239–23,245. 144. Hlubek, M. D., Stuenkel, E. L., Krasnoperov, V. G., Petrenko, A. G., and Holz, R. W. (2000) Calcium-independent receptor for α-latrotoxin and neurexin 1α facilitate toxininduced channel formation: evidence that channel formation results from tethering of toxin to membrane. Mol. Pharmacol. 57, 519–528. 145. Lelianova, V. G., Davletov, B. A., Sterling, A., Rahman, M. A., Grishin, E., Totty, N. F., and Ushkaryov, Y. (1997) α-Latrotoxin receptor, latrophilin, is a novel member of the secretin family of G protein-coupled receptors. J. Biol. Chem. 272, 21,504–21,508. 146. Magazanik, L. J., Fedorova, I. M., Kovalevskaya, G. I., Pashkov, V. N., Bulgakov, O. V., and Grishin, E. V. (1992) Selective presynaptic insectotoxin (α-Latroinsectotoxin) isolated from black widow spider venom. Neuroscience 46, 188–192. 147. Grasso, A. and Paggi, P. (1967). Effect of Latrodectus mactans tredecimguttatus venom on the crayfish stretch receptor neurone. Toxicon 5, 1–4. 148. Volynskii, K. E., Volkova, T. M., Galkina, T. G., Krasnoperov, V. G., Pluzhnikov, K. A., Khvoshchev, M. V., and Grishin, E. V. (1999) Cloning and sequencing of a fragment of the cDNA for α-latrocrustotoxin from black widow spider venom. Bioorg. Kim. 25, 25–30. 149. Elrick, D. B. and Charlton, M. P. (1999) α-Latrocrustatoxin increases neurotransmitter release by activating a calcium influx pathway at crayfish neuromuscular junction. J. Neurophysiol. 82, 3550–3562. 150. Norton, R. S. and Pallaghy, P. K. (1998) The cystine knot structure of ion channel toxins and related polypeptides. Toxicon 36, 1573–1583. 151. Holz, G. G. and Habener, J. F. (1998) Black widow spider alpha-latrotoxin: a presynaptic neurotoxin that shares structural homology with the glucagon-like peptide-1 family of insulin secretagogic hormones. Comp. Biochem. Physiol. B. Biochem. Mol. Biol. 121, 177–184.

502

Grasso and Rufini

Neurotoxins from Scorpion Venoms

503

24 Neurotoxins from Scorpion Venoms Marie-France Martin-Eauclaire

1. INTRODUCTION Scorpions have been classified into six families: Bothriuridae, Scorpionidae, Buthidae, Vejovidae, Chlaerilidae, and Chactidae. Only the venoms of Buthidae scorpions have been described as potentially dangerous for humans. This Buthidae family is divided into subfamilies on the basis of geographic and morphological criteria: Isometrinae, which is of minor importance; Buthinae, from Africa and Asia; Centrurinae, living in North or Central America; and Tityinae, from South America (1). Scorpions use a cocktail of toxins to immobilize their prey and to protect themselves (2). Their venoms are thus composed of a complex mixture of homologous proteins with similar physicochemical properties but different pharmacological activities. Ionic channels are the targets of these toxic polypeptides. The most studied toxins, responsible for the high toxicity of scorpion venom to mammals and insects, modify the gating mechanism of voltage-sensitive Na+-channels from excitable cells (3–11). Others block K+-channels in various tissues with high affinity (12–15). A peptide blocking a Cl– -channel has also been purified (16), as recently have two other peptides, able to inhibit (17) or increase (18) [3H] ryanodine binding to the ryanodine-sensitive cardiac and skeletal sarcoplasmic reticulum Ca2+ channel. Toxins that exclusively act on voltage-dependent sodium channels can be broadly classified according to their animal group (phyletic) specificity, as toxins active against vertebrates (mammal) or invertebrates (arthropods). They can be further classified into different categories according to the physiological effect they elicit. 2. TOXINS ACTIVE AGAINST VOLTAGE-SENSITIVE NA+ CHANNELS 2.1. Toxins Active Against the Vertebrate Voltage-Sensitive Na+ Channel The rapid depolarization phase of the action potential of nerve, muscle, and heart cells is due to the voltage-sensitive Na+-channels (7). Scorpion toxins increase the depolarization of the membrane and the release of neurotransmitters by affecting the activation or inactivation of these channels. Studies of electrophysiological behaviors,


503

504

Martin-Eauclaire

differences in binding sites, and ion-flux have shown that scorpion toxins active against vertebrate Na+-channels are of two types: (1) α-type toxins, slowing down the inactivation of the channel, prolonging the action potential (4,9,19,20); and (2) ß-type toxins, shifting the voltage-dependent activation to more negative membrane potentials and leading to repetitive firing in both muscles and nerves (10,21–23). In both cases, the toxins exert their effect from the extracellular side of the membrane and the Na+ influx of the cells increases. In spite of their different mode of action, α- and ß-toxins are homologous small basic proteins: their single chain is 61–70 amino acid residues long and is highly crosslinked by four disulfide bridges folded similarly. α- and ß-toxins can be aligned using these 1/2 Cys residues (Fig. 1). Clearly, some amino acid residues are strictly conserved or only substituted in a conservative manner. These residues are believed to play an important structural role. The three-dimensional (3D) structures of several scorpion toxins targeting the vertebrate sodium channel have been determined by X-ray crystallography or 1H-nuclear magnetic resonance (NMR) (for review, see ref. 24). The protein scaffold is highly conserved between α- and ß-toxins, in spite of their different biological activities. Their backbone is composed of a short α-helix (located on one face of the molecule) and a three-stranded, anti-parallel ß-sheet (located on the opposite face), stabilized by two of the disulfide bridges. The N- and C-terminal regions are stabilized by a disulfide bridge and the three-stranded ß-sheet by a conserved aromatic cluster. Superposition of the backbone structures of α- and ß-toxins reveals some structural differences, due to insertions or deletions of polypeptide segments. α-Toxins contain a ß-hairpin elongated by up to five amino acids residues, and longer N- and C-termini. However, the structural elements responsible for the various pharmacological activities have not been clearly identified. An hydrophobic surface and some of the basic residues have been described as being involved in the high-affinity binding of scorpion toxins to their pharmacological targets (25–28). 2.1.1. α-Toxins

α-type toxins were the first scorpion toxins to be purified and chemically characterized from the scorpion Androctonus australis, a Buthinae from North Africa (2). They are also found in the venoms of Androctonus mauretanicus, Leiurus quinquestriatus, and various Buthus subspecies from Africa and Asia (29). These toxins are principally responsible for venom toxicity in human beings and other mammals (for example, it was shown that 90% of the toxicity of the Androctonus australis venom is due to only four toxins, AaH I to IV, accounting for only 2–3% of the weight of the crude venom obtained by electric stimulation) (30). They prolong the action potential in nerve and muscle by slowing the inactivation of sodium channels (19,20). They specifically bind to site 3 of the voltage-sensitive sodium channels of excitable cells, and binding is dependent on the membrane potential (4,9). Work with cells transfected with the α-subunit of the type IIA Na+ channel from rat brain suggests that site 3 is constituted of extracellular loops between transmembrane segments S5 and S6 of domain I (31) and domain IV (32). The residue Glu 1613 on the S3-S4 extracellular segment of domain IV of the α-subunit (33) is required for α-toxin binding (Fig. 2). Only small amounts of α-toxins are found in American scorpion venoms (those from Tityus serrulatus and Centruroides sculpturatus). Moreover, these α-toxins show


505

Fig. 1. Amino acids sequences of toxins active against the vertebrate voltage-sensitive Na+ channels. AaH, Androctonus australis Hector; Amm, Androctonus mauretanicus mauretanicus; Lqh, Leiurus quinquestriatus hebraeus; Lqq, Leiurus quinquestriatus quinquestriatus; Bot, Buthus occitanus tunetanus; Bom, Buthus occitanus mardochei; Bm, Buthus martenzi; Be, Buthus eupeus; Ts, Tityus serrulatus; Os, Orthochirus scrobiculosus; Cll, Centruroides limpidus limpidus; Clt, Centruroides limpidus tecomanus; Cn, Centruroides noxius; Css, Centruroides suffusus suffusus; Cs, Centruroides sculpturatus Ewing (eM1, toxin active against mammals; v, variant). Four groups (from the top) are α and α-like (group 3) toxins; the two last groups (to the bottom) are ß toxins. *C-terminal amidated. The list is not exhaustive (for references, see Subheading 2.1.).

a much lower affinity for the site 3 than those purified from North African species (34–38). Recent comparative studies have separated the α-scorpion toxins affecting the inactivation of sodium current into two groups according to their activity against mammals and insects (39). Some α-toxins cause a contractile paralysis of the insect, with a time

506

Martin-Eauclaire

Fig. 2. Proposition of an interaction model between the α-toxins and their receptor site (site 3) on the voltage-sensitive Na+ channels. In this model, the AaH II X-ray structure is used (26). Domain I and IV of the sodium channel are represented; IFM, inactivation particule; P, phosphorylation site; ψ, glycosylation sites. Adapted from ref. 33.

lag after exposure. The paralysis is a result of extreme prolongation of the action potential by slowing the inactivation of sodium currents. These toxins are found mostly in Leiurus quinquestriatus subspecies hebraeus and quinquestriatus, and Buthus tunetanus (40–42). They are called α-like toxins and bind to similar but not identical receptor sites in rat brain and insect sodium channels. They are highly toxic to mammals. However, they do not compete for α-toxin binding to site 3 of the rat brain Na+channels and no specific binding of 125I-α-like toxins could be detected on rat brain synaptosomes. The difficulty in measuring a specific binding on this preparation (but not on insect synaptosomes) was attributed to low-affinity binding due to the acidic properties of the α-like toxins (43). This is evidence that α-like toxins bind to a different receptor site on sodium channels (closely related to, but different from, receptor site 3). New investigations indicated that one of these α-like toxins displays a subtype specificity and strongly inhibits sodium current inactivation of rat CA1 pyramidal neurons in acute hippocampal slices, wheareas α-toxins have only weak or no effects (44). Some of these α-like toxins are so highly active on insects that they are at the end of the animal-selectivity scale and can be considered as being specifically active on insects (24). They are called α-insect-toxins. One such α-insect-toxin, Lqh αIT, has been produced as recombinant toxin in insect cells and lepidopteran larvae using baculovirus (45) and in Escherichia coli (46). Kurtoxin, purified from the venom of the South African Parabuthus transvaalicus, was described as a potent inhibitor of the T-type voltage-gated calcium channel (47). The toxin is able to discriminate between the different α subunits of voltage-gated calcium channels and binds to α1G and α1H, but not to α1B (N-type), α1A (P/Qtype), α1C (L-type), and α1E. However, its amino acid sequence is very similar to


507

those of α-toxins (up to 60%). Application of kurtoxin to the rat brain IIA sodium channel α subunit slows down the inactivation process. The toxin effect is strictly similar to that observed for other α-toxins. However, although all the molecules described in this paragraph are able to induce the same electrophysiological response (i.e., an inhibition of the sodium current inactivation), and are referred to as α-type toxins, it is important to keep in mind that the initial definition of an α-toxin is its ability to compete for binding to site 3 with 125IAaH II, the most active and prototype α-toxin (9). 2.1.2. ß-Toxins ß-type toxins are present in the venoms of scorpions from North and South America. They consist of a single polypeptide chain of 60–66 residues crosslinked by four disulfide bridges (48–54). Variant 3 from Centruroides sculpturatus Ewing, which has only weak activity but an amino acid sequence very similar to the most potent ß-toxins, was the first scorpion toxin crystallized. Its structure has been determined at high resolution by X-ray crystallography (25). ß-toxins are those that behave like the reference toxin Css II, purified from Centruroides suffusus suffusus (22). They bind to site 4 of the voltage-sensitive sodium channels of rat brain synaptosomes. This binding does not depend on membrane potential and there are no signs of allosteric effects due to toxins binding to other sites on the Na+ channel. They affect the activation process and cause repetitive firing in voltage-clamped frog myelinated nerve (21,22). By using Na+ channel chimeras, it was shown that another ß-toxin from Centruroides suffusus suffusus, Css IV, binds to a receptor site that requires Gly 845 in the S3-S4 loop, at the extracellular end of the S4 segment in domain II of the α-subunit (55). Ts γ (49,56), also called as Ts VII (57–59), is the main toxic compound purified from the venom of the Brazilian scorpion Tityus serrulatus. Its X-ray structure has been determined at 1.7 Α resolution (60). Its effects have been investigated using voltage-gated channels from human heart (hH1) and rat skeletal muscle (rSkM1) expressed in Xenopus oocytes (61). Ts γ has little effect on hH1. Its action on chimeric channels depends on the origin of the domain II: If the domain II is from rSkM1, the toxin affects activation. This is in agreement with the results found for the binding of Css IV to the rat brain sodium channel type II. Ts γ also binds with high affinity to insect sodium channels (62–64). This may be because of the extreme flexibility of the molecule allowing a good fit at various receptor sites (65). Using chemical modifications, it was demonstrated that Lys12, Trp 39, and Trp54 are important for the activity of the molecule, both on insect and mammalian sodium channels (28). Crystallographic data revealed that the positively charged group at position 12 is determinant for the specificity of ß-toxins, and that a number of basic amino acid residues located on the face of the molecule opposite to the binding surface may account for the high toxicity of Ts γ (60). Two other toxins should be mentioned here, because they share a broader specificity for binding to sodium channel and are able to compete with radiolabeled ß-type toxins bound to their receptor site on rat brain synaptosomes. The first is AaH IT4, purified from the venom of the Tunisian scorpion Androctonus australis. Its sequence has little in common with those of ß-type toxins, but is recognized by anti-Css II-antibodies. It competes with the α-toxin, AaH II, the ß-toxin, Css II, and an insect toxin, AaH IT, for

508

Martin-Eauclaire

Fig. 3. Amino acids sequences of toxins active against the invertebrate voltage-sensitive Na+ channels. AaH, Androctonus australis Hector; Lqh, Leiurus quinquestriatus hebraeus; Lqq, Leiurus quinquestriatus quinquestriatus; Bot, Buthus occitanus tunetanus; Be Buthus eupeus; Bj, Buthotus judaicus. Toxins of the second group (from the top) are “depressant anti-insect toxins”; toxins of the last group (to the bottom) are “excitatory anti-insect toxins”; the sequences of some “α-anti-insect toxins” are found in Fig. 1 (group 3 of the “α-like toxins.” The list is not exhaustive (for references, see Subheading 2.2.).

their respective sites on mammalian and insect sodium channels. Thus, AaH IT4 has been suggested to be an ancestral scorpion toxin. The second toxin is active against Crustaceans and was isolated from the venom of Centruroides noxius. At very high doses (100 µM), it competes with the ß-type toxin Cn2 (Centruroides noxius 2) for binding. Its amino acid sequence is very similar to those of ß-type toxins, but the differences in sequence are concentrated in two stretches forming a continuous surface region, which may be involved in species specificity (66). The ß-toxins can also be classified according to their activity against mammals and insects as reported for Ts γ and AaH IT4. However, ß-like toxins that are specifically active against arthropods have also been described (67). 2.2. Toxins Active Against Invertebrate Voltage-Sensitive Na+-Channels The high toxicity of scorpion venoms to insects is mainly due to the actions of various polypeptides (Fig. 3) that have specific effects on the voltage-sensitive sodium channels of one or several insect families. These toxins include: 2.2.1. The Excitatory Anti-Insect Toxins AaH IT from Androctonus australis was the first toxin to be isolated as being toxic for insects. It causes repetitive firing in the motor nerve of insects, resulting in immediate contractile paralysis of the animal (68). Using a radioiodinated derivative of AaH


509

IT, it was shown that the toxin binds specifically only to insect neuronal membranes (locust and cockroach) and does not interact with the mammalian sodium channel. The binding is independent of voltage (64,69,70). AaH IT has the highest specificity of all the toxins characterized as anti-insect toxins, (71) and one of the highest affinities for the insect sodium channel. So, AaH IT can be used as a reference model for the design of new biopesticides (see Subheading 6.3.). Recent studies showed that Musca kdr and super-kdr flies, with a single- or a double-point mutation, respectively, in domain II of the para gene encoding the voltage-gated sodium channel, are about 9- and 14-fold more susceptible than the wild-type to AaH IT (72). Similar toxins have been obtained from the venoms of other Old Word Buthinae (73,74). AaH IT consists of 70 amino acids and its disulfide-bridge pattern is different from that of the α- and ß-type toxins (75). This particular feature may be responsible for its high specificity for the insect sodium channel (65). Its structure in solution, determined by NMR, suggests that the difference in specificity between toxins able to discriminate between mammalian and insect voltage-gated Na+-channels, may be due to the position of the C-terminal peptide relative to a hydrophobic surface common to all scorpion toxins (76). Another excitatory anti-insect toxin with a very distinctive feature was recently described (77): Its crystal structure, determined at 2.1 Å resolution, showed an additional short α-helix at the C-terminus. 2.2.2. The Depressant Anti-Insect Toxins These toxins cause the slow progressive paralysis (described as flaccid paralysis) of insects. This paralysis results from the blocking of the evoked action potentials. Like the excitatory anti-insect toxins, these toxins do not bind to rat brain neuronal membranes (74). They are found in Old Word Buthinae venoms including those of Leiurus (LqhIT2, LqqIT2), Buthotus (Bj IT2), Buthus (Bot IT4 and Bot IT5), and Buthacus (Ba IT2) (78–83). They are only 61 amino acid residues long. Their disulfide bridges are in the same positions as those in other toxins active against Na+ channels, except for the contracturant-insect-toxins. No X-ray or NMR structure studies have been reported. An anti-insect toxin, Aa IT5, more potent than AaH IT and than the depressantinsect toxins against tobacco budworms, has been purified from Androctonus australis (84). Its amino acid sequence shows 38% similarity to the sequence of depressantinsect-toxins. There have been no reported in vitro competition experiments with radiolabeled excitatory anti-insect-toxins and depressant-insect-toxins to characterize the receptor sites of these toxins. Also from the Androctonus australis venom, Aah VI was the first glycosylated scorpion toxin to be described (85). Its activity in Blatella germanica is very weak. Aah VI is heterogeneously N-glycosylated on Asn9 (AsnGly-Thr sequence) and its glycan core structure is Asn-GlcNAc(α1-6Fuc)(α13Fuc)(ß1-4GlcNAc). Bot IT2 from Buthus tunetanus differs from the other toxins of this group: its overall effects in current-clamp conditions are relatively similar to those of contracturantinsect-toxins on cockroach axon preparations (42,82,86). It also has a repetitive activity similar to the contractions caused by the contracturant insect toxins. However, its sequence is more similar to those of depressant anti-insect toxins (67%). Competitive binding experiments with other known sodium-channel neurotoxins suggested that Bot IT2 binds to a receptor site located close to, but not identical to, the binding site of the excitatory anti-insect toxins (87).

510

Martin-Eauclaire

3. SCORPION TOXINS ACTIVE ON K+ CHANNELS K+ channels have been found in virtually all eukaryotic cells of the animal and plant kingdom, and even in prokaryotes cells. They perform a multitude of functions, the most basic being the setting of resting potentials (88). They can be activated by membrane potential (Kv), by an increase in intracellular calcium (KCa), and by the binding of ligands (KATP) (for reviews, see refs. 89,90). Some scorpion toxins are able to block diverse types of K+ channels with high affinity. These toxins are often present at low concentrations in venom (from 0.01–1% by weight). They are only weakly toxic when injected subcutaneously but are very active when injected intracerebroventricularly. They are powerful tools for the study of the physiological role and the architecture of the pore-region of these channels. They are short polypeptides of about 30–40 amino acid residues, cross-linked by three or four disulfide bridges. They display a high degree of primary sequence similarity with each other (for review, see refs. 14,15,91). The resolution of their three-dimenstional (3D) structure by proton NMR in solution is relatively straightforward and the structures of many have been determined. They have a common structural motif: an α-helix of 8–11 residues and a ß-sheet composed of 2 or 3 anti-parallel ß-strands. The conformation is stabilized by two disulfide bridges between the ß-sheet and the α-helix. This structural arrangement is common to both long and short scorpion toxins and also insect defensins, a group of inducible antibacterial peptides (92). These scorpion toxins are difficult to classify according to their activity against K+ channels, because of their lack of specificity. They have, however, been classified into families on the basis of primary sequence similarities (15,91,93). A family of toxins (very short peptides of 29–31 amino acid residues), shows a very specific activity against the so-called “apamin-sensitive,” small-conductance K+ channels activated by calcium (type SKCa). The other families (containing peptides of up to 39 residues long) exhibit various affinities for slowly or rapidly inactivated voltage-dependent (Kv), and calcium-activated big (BKCa) or intermediate (IKCa) conductance K+ channels. Only the Kv1 channels (in particular Kv1.1, Kv1.2, and Kv1.3) are highly sensitive to these toxins. The mechanism of inhibition of Kv and BKCa channels has been described in detail (14,94,95). Here, some of the characteristics common to these molecules (12 families and 50 molecules have been identified to date) will be summarized, but they will not be described in detail. Recent reviews (15,91,93,96,97) focus only on scorpion toxins active on K+ channels. Classification is based on primary structures, rather than pharmacological activities, which are often nonselective or even not known. 3.1. Scorpion Toxins Specific for Voltage-Sensitive K+ Channels Voltage-dependent K+ channels are further subdivided into: (1) A-channels, which activate and inactivate quickly upon membrane depolarization; (2) delayed rectifier channels, which activate slowly and do not or only slowly inactivate; and (3) the inward rectifier channels, which open upon membrane hyperpolarization. Structural information about the voltage-gated K+ channels family are available because of the successful cloning and functional expression of the Drosophila Shaker gene (for recent reviews, see refs. 98,99). This has led to an exponential growth of K+ channel research with the effect that at least nine distinct voltage-gated K+ channels families have now been


511

cloned and characterized in various species as different as Aplysia and human. The primary sequences deduced from cDNAs indicate that voltage-gated K+ channels are members of a superfamily and have several features in common: the N- and C-terminal ends are hydrophobic and have different lengths; the core region comprises five hydrophobic segments (S1-S3, S5, and S6) and a positively charged amphipathic segment (S4). Segments S1 to S6 traverse the membrane. Between S5 and S6 is the region H5 (or P region), which is highly conserved among the different K+ channels and determines the ion selectivity of the channels, lines the pore, and constitutes the receptor site of peptide blockers. Site-directed mutagenesis experiments and the study of channel mutants have made large contributions toward understanding structure and function of these ion channels. The most described scorpion toxins specific for voltage-sensitive K+ channels are discussed below. Noxiustoxin (NTX) from the Mexican scorpion Centruroides noxius Hoffman, was the first scorpion toxin shown to have a direct effect on potassium channels. It reversibly blocks the delayed rectifier potassium current in the squid giant axon. It is a 39 amino acid polypeptide (MW = 4200 Da), reticulated by 3 disulfide bridges (15). It blocks voltage-dependent K+-channels of T lymphocytes with high affinity (Kd = 0.2 nM) and is a potent inhibitor of radioiodinated charybdotoxin (125I-ChTX) binding to brain membranes but does not affect the binding of the same toxin to bovine aortic sarcolemmal membrane vesicles (97). Margatoxin belongs to the same family as NTX. Purified from the venom of the scorpion Centruroides margaritus, this toxin is the most potent blocker described for voltage-dependent K+-channels. It is an extraordinary high-affinity selective ligand (Kd < 0.1 pM) for Kv1.2 and Kv1.3 channels (100,101). The Kaliotoxin subfamily has eight highly similar members (primary structure more than 70% identical): kaliotoxin (KTX); kaliotoxin 2 and 3 (KTX2 and KTX3); BmKTX; agitoxins 1, 2 and 3; and OsK-1. OsK-1, a toxin rescently purified from the scorpion Orthochirus scrobiculosus, exhibits a type of pharmacology different from that described for the other seven toxins. KTX was purified from the venom of the Moroccan scorpion, Androctonus mauretanicus mauretanicus, and was initially described as a specific inhibitor of IKCa channels of the invertebrate, Helix pomatia, but is inactive against Kv currents of snail neurons. Tests of competitive binding of radioiodinated KTX and of Dendrotoxin from the mamba Dendroaspis angusticep (DTX) to rat brain synaptosomes have shown that KTX has high affinity for the Kv channels of mammals. Structure and function relationship studies using synthetic short peptides including KTX(27–37), KTX(25–32), and KTX(1–11) have been conducted (102). KTX(27–37) and KTX(25–32), which contain the C-terminal region conserved in all short toxins, compete with radiolabeled KTX for its receptor on rat brain synaptosomes. Although these peptides cannot themselves block IKCa currents, they act as antagonists of KTX. This demonstrates that the C-terminal region, particularly the ß-sheet, is involved in the interaction with the receptor (Fig. 4). This interaction has also been demonstrated by studies of numerous ChTX mutants and experiments with the double mutant KTX/Kv1-3 mutant (95). KTX is a fairly specific ligand for the Kv1.3 channel, which it blocks with an high affinity (103,104). It is thus a useful probe for studying the topology of the vestible of the lymphocyte Kv1.3 channel. The side

512

Martin-Eauclaire

Fig. 4. Three-dimensional structures of two toxins blockers of K+ channels. The structures were obtained by NMR: Charybdotoxin (ChTX) from Leiurus quinquestriatus hebraeus on the left side (92); Kaliotoxin (KTX) from Androctonus mauretanicus mauretanicus (95) on the right side. The amino acid residues of the ß-sheet important for the interaction between the toxins and their receptor site are materialized

chain of the lysine 27 residue enters the pore and interacts with the Asp402 residue of each channel subunit. The Phe25 residue of KTX is located between 2.4 and 6 Å from the His404 residue of the channel. The Arg24 residue of the toxin interacts electrostatically with the Asp 386 residue of the channel and residues Leu15 and Arg31 are brought close to the Gly380 residue of the channel (95,105). These results and structural analysis of KTX have led to a model of the topology of the pore vestible, in which the receptor is flat, 28–34 Å across with edges 4–8 Å high. This model is slightly different to that produced for the Shaker channel using ChTX as a probe. 3.2. Scorpion Toxins Specific for the Calcium-Activated K+ Channels Ca2+-activated K+ channels have been subdivided according to their conductance, which falls in the range from a few to several hundred picosiemens (pS): (1) highconductance Ca2+-activated K+ channels (100–300 pS): their open probability increases with increasing intracellular Ca2+ concentration (0.1–10 µM) and with membrane depolarization at constant intracellular Ca2+ concentration; (2) intermediateconductance Ca2+-activated K+ channels (18–50 pS) are activated by internal Ca2+ concentration, and can be voltage-sensitive, e.g., those present in molluscan neurons, or voltage-insensitive, e.g., those found in red blood cells; (3) small-conductance Ca2+activated K+ channels (6–14 pS) are also called apamin-sensitive Ca2+-activated K+ channels because they are blocked by pM concentrations of the bee venom peptide apamin. Their sensitivity to the intracellular Ca2+ concentration is greater than that of the intermediate-conductance Ca2+-activated K+ channels at negative membrane potentials; and they show little or no voltage-dependence. 3.2.1. Scorpion Toxins Specific for the Large Conductance Calcium-Activated K+ Channel (BKCa) Charybdotoxin (ChTX), the most extensively studied toxin active on K+ channels, was first described in the venom of Leiurus quinquestriatus hebraeus as a potent blocker of the high-conductance Ca2+-activated K+ channel of the mammalian skeletal


513

muscle (at the external side). It was shown that ChTX is also able to block other types of K+ channels, like those of the Kv1 family. It is a 37 amino acid peptide reticulated by three disulfide bridges and its N-terminal amino group is blocked (pyroglutamine). For some years, it was believed that the ChTX was selective for Ca2+-activated dependent K+ channels of large conductance. However, ChTX also blocks Ca2+ activated K+ channels of smaller conductance: 127 pS channels in cultured kidney cells, 35 pS channels in Aplysia neurons, 25 pS channels in human erythrocytes, and various types of Ca2+ activated K+ channels from rat brain plasma-membrane vesicles in planar lipid bilayers. Indeed, this toxin has been used to study the pharmacological properties of highconductance Ca2+ -activated K+ channels in vascular smooth muscle, portal vein, and trachea (for review, see ref. 91). ChTX blocks some voltage-dependent K+ channels not activated by intracellular Ca2+ ions, including the delay-rectifier in human and murine T-lymphocytes, inhibiting proliferation and interleukin-2 (IL-2) production. It also inhibits voltage-gated K+ channels in human platelets.125I-ChTX binds to sarcolemmal membrane vesicles from either bovine aortic or tracheal smooth muscle with a Kd of 0.1 pM. The main targets of ChTX in the brain appear to be dendrotoxinsensitive voltage-dependent K+ channels. ChTX physically plugs the pore of the high-conductance Ca2+ activated channel by binding to its external mouth. Interestingly, K+ and Rb+ ions relieve the toxin block when added to the opposite (i.e., internal) side of the membrane, due to an increase in the ChTX dissociation rate. By analyzing point mutants of ChTX expressed in E. coli assayed with single Ca2+ activated channels reconstituted into planar lipid bilayers, was shown that a single positively charged residue of the peptide (Lys27), wholly mediated the interaction of K+ with ChTX. These results argue that ChTX bound to the channel physically close to a K+-specific binding site at the external end of the conduction pathway and that a K+ ion occupying this site destabilizes ChTX via direct electrostatic repulsion from the ε-amino group of Lys27 (106,107). The 3D structure of natural charybdotoxin in aqueous solution has been established by 1H-NMR: ChTX is composed of a small triple-stranded anti-parallel ß-sheet linked to an α-helix by two disulfide bridges. This structure is common to all known scorpion toxins, irrespective of their size, sequence, and function. Strikingly, the same 3D organization is found in all members of another small protein family found in insect hemolymph, called insect defensins (92). Iberiotoxin (IbTX) was isolated from the venom of the scorpion Buthus tamulus and is a potent and the only specific blocker of the high conductance Ca2+-activated K+ channel (108). IbTX is a 37 amino acid polypeptide with 68% sequence identity with ChTX. Like ChTX, the N-terminus is blocked in the form of a pyroglutamic acid residue. The most interesting feature of IbTX is its high selectivity for the high conductance Ca2+-activated K+ channel. It does not compete with 125I-ChTX binding to either brain synaptic-membrane vesicles or human T lymphocytes, making IbTX a unique tool for investigating the physiological role of the high-conductance Ca2+-activated K+ channel. The interaction of IbTX with this channel has been examined by measuring singlechannel currents from high-conductance Ca2+-activated K+ channels of bovine aortic smooth muscle incorporated into planar lipid bilayers. Raising the internal potassium

514

Martin-Eauclaire

concentration increases the dissociation rate constant. The effects of increasing the external concentration of Na+ and K+ suggested that the association of IbTX with the channel was enhanced by a Coulombic attraction between the positively charged toxin and the negatively charged channel mouth. The mechanism proposed for IbTX block was the same as that identified for the ChTX block of the same channel from skeletal muscle and has the following characteristics: (1) a 1:1 stochiometry for toxin blocking of the channel; (2) trans-enhanced dissociation of toxin by internal K+ ions; (3) involvement of surface electrostatic charges in toxin association with the channel. Both toxins bind near the external mouth of the pore and occlude ion flux through the channel. The solution structure of chemically synthesized IbTX (109,110) has been determined by 2D 1H-NMR spectroscopy. At least three areas of differences between IbTX and ChTX could explain the differences in channel specificity. These are differences in net charge, electrostatic asymmetry, and specific residue–residue interactions between the toxins and the channel. To identify regions of IbTX responsible for the selectivity for high conductance, Ca2+-activated K+ channels or voltage-gated K+ channels, chimeric toxins have been constructed from ChTX and IbTX (111). The C-terminal domain of the ChTX homologs defines the toxin-channel interaction that distinguishes between the high-conductance Ca2+-activated K+ channels and voltage-gated K+ channels. Analysis of the primary amino acid sequences of the two toxins suggests that only three residues determine the high-affinity interaction with the high-conductance Ca2+activated K+ channels (Glycine 22, Aspartic acid 24, and Glycine 30, all on the ß-sheet face). 3.2.2. Scorpion Toxins Specific for the Small-Conductance K+ Channel Sensitive to Apamin (SKCa)

This family contains several known toxins, including leiurotoxin I or LTX1 (Scyllatoxin) from the venom of the scorpion Leiurus quinquestriatus Hebraeus; P05 and P01 from Androctonus mauretanicus; and Ts κ from Tityus serrulatus (112–116). They display high affinity (8–300 pM) and high specificity for the apamin-sensitive SKCa channel. α-Amidation of their C-terminal residue has a significant effect on their biological activity. With the exception of Ts κ, they are all structurally similar, with only two anti-parallel ß-strands. Structure-function studies have shown that a positively charged region at the face of the α-helix exposed to the solvent is involved in binding to the receptor. 3.3. Other Less Well-Characterized Toxins Acting on K+ Channels Tityustoxin Kß (TsTX Kß) in the venom of Tityus serrulatus was described as blocker of a voltage-sensitive delayed-rectifyer. It was the first “long” toxin active on K+ channels to be described (117). Later, cDNA encoding this toxin was isolated from a Tityus serrulatus cDNA library (93). The encoded toxin was 60 amino acids long, crosslinked by only three disulfide bridges in positions similar to those of the other scorpion toxins active on K+ channels. A cDNA encoding an analogous peptide (64 amino acids) was amplified by polymerase chain reaction (PCR) from the cDNA library of Androctonus australis. These toxins are related to the scorpion defensins (42% of sequence similarity).


515

Ergtoxin (ErgTX) isolated from Centruroides noxius specifically inhibits (IC50 = 16 nM) the ether-a-go-go-related channels (ERG). These channels are crucial in neurons and are impaired in the human long-QT syndrome. ErgTX has an unusual structure that differs from those previously described. The particularity of this molecule (42 residues, 3 disulfide bridges) is that the the 1/2 Cys residues are in different positions to those in the other K+ channel blockers from scorpion venoms. ErgTX blocks neuronal, cardiac, and endocrine ERG channels, but does not affect other varieties of classical inwardly rectifying currents (118). Continuing comutagenesis studies with both the K+ channel and selected toxins should further elucidate which residues in each molecule are involved in the binding interaction. 4. SCORPION TOXINS ACTIVE ON CL– CHANNELS This kind of molecule was first purified from Buthus epeus venom (119). BeI1 and Be 15 are small peptides (35 amino acid residues, 4 disulfide bridges), devoid of activity on mammals but at high doses able to kill insects. They were therefore called insectotoxins. P2 was found in Androctonus mauretanicus venom obtained by electric stimulation of the animals (120), but not in the physiological secretion obtained by manual stimulation. The authors explained the weak anti-insect activity by a crosscontamination (1/5000) with a potent excitatory anti-insect toxin. The NMR structure of these peptides showed for the first time that long and short scorpion toxin share a common scaffold (an α-helix and a three-stranded antiparallel ß-sheet) (121). Subsequent electrophysiological experiments allowed the characterization of a new toxin, chlorotoxin, from Leiurus quinquestriatus venom. The toxin reversibly inhibits reconstituted small-conductance Cl– channels from colonic epithelial cells (122). Its amino acid sequence shows considerable similarity with those of insectotoxins and its toxicity to arthropods has been demonstrated (16). Its structure, elucidated by NMR, adopts the same folding as described for the insectotoxins (123). A voltage-dependent outwardly rectifying Cl– current, sensitive to application of chlorotoxin, was identified in human astrocytomas (124). Its block resulted in altered cell proliferation. Radiolabeled chlorotoxin binds specifically to glioma cells on high- (Kd = 4.2 nM) or low- (Kd = 660 nM) affinity binding sites, but not to normal tissus (125). The data suggested that chlorotoxin could be used as glioma-specific marker and is a valuable diagnostic tool. 5. SCORPION TOXINS ACTIVE ON CA2+ CHANNELS Ryanodine receptors (Ryr) play a major role in excitation-contraction coupling of cardiac and skeletal muscle. They release Ca2+ from intracellular Ca2+ pools of the sarcoplasmic reticulum in response to diverse triggering signals. Two modulators of the Ryr activity have been purified from the venom of Pandinus imperator: Imperatoxin I (IpTxi) is an inhibitor, and Imperatoxin A (IpTxa) is an activator (17,18). IpTxi is a 15 kDa heterodimeric protein constituted of a high molecular weight subunit (104 amino acid residues), exhibiting a phospholipase A2 (PLA2) activity, covalently linked by a disulfide bridge to a small subunit (27 amino acids). Its complete amino acid sequence has been determined as has the nucleotide sequence of its gene. A single cDNA clone encodes the two subunits, connected by a basic pentapep-

516

Martin-Eauclaire

tide that is presumably removed during the maturation phase. The large subunit, with lipolytic activity, is closely related to PLA2 from bee and lizard venoms, i.e., group III PLA2. It was thus the first PLA2 described in a scorpion venom. The small subunit showed no similarity with any previously described peptide. The treatment of IpTXi with the PLA2 inhibitor p-bromophenacylbromide decreased IpTXi’s inhibitory activity. This suggests that a lipid product was involved in the inhibition. Further experiments supported this view: the kinetics of Ryr inhibition were mimicked by addition of linoleic acid (or other long-chain, unsaturated fatty acids) to the cytoplasmic side of the channel (17). IpTxa is a 33 amino-acid residue peptide with three disulfide bridges. Its sequence shows no similarity with that of any other scorpion toxin (18). From its biological activity, IpTxa seems to selectively activate the skeletal-type Ryr isoforms (sRyr). Its binding site appears to be distinct from caffeine- and adenine nucleotide-regulatory sites. It has been shown (126) that IpTxa mimics a cytoplasmic loop between repeats II and III of the α1 subunit of the skeletal dihydropyridine receptor (the II-III loop). This loop is an important determinant of skeletal-type exitation-contraction coupling. Synthetic truncated peptides were used to localize its minimum essential unit for in situ triggering and the Arg681-Leu690 region was thereby identified (127). A structural motif composed of a cluster of basic residues followed by Ser or Thr was found, both on the II-III loop and on IpTxa. Mutations in this cluster (on the II-III loop or on the toxin) dramatically reduce or completely abolish the activation of sRyr (126). In addition to these toxins from Pandinus imperator, able to modulate the Ryr, the kurtoxin from Parabuthus transvaalicus was described as a potent inhibitor of the T-type voltage-gated calcium channel (47). However, this toxin is also able to delay the voltage-gated Na+ channel inactivation (see Subheading 2.1.1.). 6. MOLECULAR BIOLOGY OF SCORPION TOXINS Molecular biology techniques have been introduced over the last decade into the study of scorpion toxins (128). These techniques have elucidated the structure of toxin precursors, and explained their post-translational processing and the organization of their genes. Such studies should also be helpful in the study of the evolution of these toxins. 6.1. Precursors of Scorpion Toxins Scorpion cDNA libraries have been made from mRNA extracted and purified from the venom gland. There are currently six scorpion cDNA libraries described in the literature: Androctonus australis (128), Leiurus quinquestriatus hebraeus (79), Tityus serrulatus (59), Centruroides noxius (129), Buthus occitanus tunetanus (130), and Buthus martensi Karsch (131). 6.1.1. Structure of Precursors of Long Toxins Active Against Voltage-Dependent Na+ Channels cDNAs encoding scorpion toxins active against voltage-gated Na+ channels of mammals and insects have been isolated by screening cDNA libraries with radiolabeled oligonucleotide probes and DNA amplification by PCR. The complete cDNAs isolated consist of 300–370 bp encoding toxin precursors of 82–89 amino acid residues. They


517

have a 5' noncoding region (less than 40 bp), a sequence encoding a signal peptide of 18–21 residues, a sequence encoding the mature peptide and a 3' noncoding sequence that often contains a polyadenylation signal (AATAAA), situated 50–65 bp downstream from the stop codon and 4–10 bp upstream from the poly-A tail. 6.1.2. Structure of the Precursors of Short Toxins Active Against K+ Channels

Using oligonucleotides encoding the highly conserved C-terminal motif implicated in the interaction of these toxins with the receptor (i.e., the anti-parallel strands of the ß-sheet and the ß-turn), the first nucleic acid sequences encoding precursors of toxins active against K+ channels (132,133), have been determined. The previously described structural similarity between the insect defensins and the K+ channels blockers from scorpion venoms does not reflect similarity in the organization of their precursors. The precursors of toxins active against K+ channels have a long signal peptide (more than 22 amino acid residues) followed by the mature peptide, whereas the precursors of insect defensins are organized as prepropeptides. cDNAs of more recently characterized short neurotoxins, some of them being active on SKCa channels, (134,135), and cDNAs encoding the toxins Kß, novel long-chain toxins active on a nonidentified K+ channel (93,136), were subsequently amplified and sequenced. 6.1.3. Structure of the Precursors of Short Toxins Active Against Cl+ Channels

Two cDNA clones, encoding peptides resembling the reported chlorotoxin, were isolated from the Leiurus quinquestriatus hebraeus library (135). The deduced precursors were the same length and had the same organization as toxins active against K+ channels. 6.1.4. The Signal Peptides of Scorpion Toxin Precursors

The signal peptides of both long and short scorpion toxins precursors have characteristics typical of signal peptides in general. They are organized into three regions: usually a positively charged region at the N-terminus, a central hydrophobic α-helix and a polar cleavage region at the C-terminal end. There is substantial sequence similarity between the signal peptides of the α-toxin precursors of Androctonus australis, but no such similarity between those of toxin precursors purified from the venoms of scorpions of different species. Recent sequence analysis suggested that the longest signal peptides (e.g., those of the Kß toxin), could in fact be composed of a signal peptide of about 20 residues and a short propeptide of seven to eight residues (136). Froy et al. (135) also suggested that the codons downstream from the intron (and perhaps the split codon itself), do not belong to the leader sequence. Such stretches could become structural entities corresponding to a propeptide. However, no traditional consensus cleavage site was found to support this hypothesis, which remains to be verified. 6.1.5. C-Terminal Post-Translational Processing of the Precursors

There are extra amino acid residues C-terminal to the end of the mature peptide in some toxins, indicating that these proteins are post-translationally processed. Precursors known to give α-amidated toxins have a C-terminal di- or tripeptide, Gly-Lys-Lys or Gly-Arg. The precursor is attacked by exopeptidases (carboxypeptidase B), which remove the extra, basic residues. The glycine residue may be processed by a peptidylglycine α-amidating monooxygenase (PAM). This α-amidation of the C-ter-

518

Martin-Eauclaire

minus was shown important for the activity of some of the peptides, including P05 and LTX1. 6.2. Structure of Genes Encoding Scorpion Toxins The genes encoding scorpion toxins active against Na+ channels were the first to be described. The most complete study involved the isolation of a gene encoding the long AaH α-toxin I' from an Androctonus australis DNA library (137). The transcriptional unit of this gene is 793 bp long and contains two exons separated by an intron of 425 bp. Upstream from the experimentally determined transcription initiation site AACAA, the promoter region was defined by two consensus sequences: the TATA box (binding site for RNA polymerase II) and the CAAT box (transcription-factor binding site). Mobility-shift assays show that an intrinsic bend in the DNA close to the transcription initiation site may be functionally important for a high rate of transcription. Other studies analyzed the structure of other scorpion toxin genes including those encoding ß-toxins from Tityus, α- and depressant insect toxins (129,135). Gene fragments have been obtained by PCR using primers with sequences based on those flanking the cDNAs already isolated. The fragments contain introns between 350 and about 600 bp long. The genes of the short toxins active against K+ or Cl- channels have similar structures to those of the long toxins (134,135,138). Their introns are relatively small (about 90 bp). Thus, the scorpion toxin genes contain intron regions of different sizes but in the same relative positions (the intron interrupts the sequence encoding the signal peptide), and with a high proportion of A and T residues (more than 80%) (Fig. 3). All introns depicted have a consensus GT/AG splice junction. They contain a putative branch point, 5'-TAAT-3', 47–61 bp upstream from the 3' splice site (135). They all split a codon near the end of the signal sequence, at position –4 (Gly or Val) or –6 (Glu, Val, or Ala). These studies show that the polymorphism of the long and short toxins could result from a large number of different genes rather than from alternative splicing. This suggests that either a common ancestral gene diverged to form two lines a very long time ago, or there has been convergent evolution to produce this structural motif. 6.3. Production of Recombinant Scorpion Toxins Chemical modifications and peptide synthesis have been used successfully to investigate the relationships between structure and function of scorpion toxins. However, there are still technical limitations to experiments involving changing the chemical functions of the side chains of amino acid residues exposed to the solvent (size of the reagent, denaturation of the protein, etc.). Also, only short scorpion toxins and their analogs have been chemically synthesized with success. Thus, the use of molecular biology techniques is invaluable. The production of recombinant molecules, using material that can be genetically modified as required, would facilitate studies of the structure-function relationships of these proteins. Several scorpion toxins active against K+ channels have been produced in E. coli: From synthetic genes, this is the case of ChTX, MgTX, and AgTX2 (100,107,139,140). The same pET-type expression vector, pCSP105, was used in each case. The sequence of interest is expressed, under control of the inducible phage T7 promoter, to produce a fusion protein with the product of the gene IX in the cytoplasm of the bacterium. Disul-


519

fide bridges can be formed in vitro. In each case, a synthetic gene encoding the toxin to be produced was constructed. The system was first developed with ChTX (139). The fusion protein is purified, then cleaved by factor Xa to produce the recombinant ChTX, which is then heat treated to convert the C/N terminal Gln residue into a pyroglutamate residue. About 100 mutants have been obtained (107). The natural cDNA encoding KTX2 was also produced in the periplasmic space of E. coli as a fusion protein with the maltose binding protein followed by the recognition site for enterokinase preceding the first amino acid residue of the toxin (141). The production of recombinant long toxins active on Na+ channels remains the hardest task. The potent α-toxin AaH II was the first toxin successfully expressed. However, its transient production in monkey kidney COS-7 cells was very low (128). The strict selectivity of AaH IT (71) has led various groups to use its gene (128) in the design of biopesticides. Synthesis of the toxin in the hemolymph of tobacco budworms (Heliothis virescens) by a recombinant baculovirus increases the speed of the animal’s death and thus restricts feeding damage to the plant (142–144). Highly insecticidal recombinant toxins (LqhαIT and Bj-xtrIT) have been produced in E. coli and their solution structure determined (46,145,146). This approach revealed the importance of some of the aromatic and basic residues (Lys8 in particular) to LqhαIT activity. The potent bioreactive role of the C-terminal region was also highlighted: the substitution of the C-terminal residue Arg64 of LqhαIT by a His increased the toxicity to insect by a factor 3.2, whereas its replacement with an Asp reduced the toxicity to 1/20th. The solution structure of the mutant demonstrated the importance of the spacial orientation of the side chain of the residue at position 64 (C-terminus residue), and suggested that this region may be involved in recognition of the receptor site. The properties and a series of peptide with various deletions of the structurally unique C-terminal region of Bj-xtrIT are consistent with the C-terminus being determinant for activity. 7. CONCLUSIONS AND PERSPECTIVES The combination of molecular biology and biochemical studies has led to the purification and characterization of numerous scorpion toxins. These toxins, which are potent and specific modulators of ion channels, are as diverse as their targets. Structure–function relationships of these families of toxins provide substantial information about the mechanisms of toxin–target interaction, channel structure, and mechanisms of ion permeation. The amino acid residues involved at the binding interface are beginning to be identified. These studies may lead to the design of new therapeutic agents, able to mimic the interactive surface of the toxins and to specifically modulate any channel involved in a particular disease. REFERENCES 1. Bucherl, W. (1971) Classification and biology of scorpions, in Venomous Animals and Their Venoms, vol. III (Bücherl, W. and Buckley, E., eds.), Academic Press, New York, pp. 339–342. 2. Miranda, F., Kopeyan, C., Rochat, C., Rochat, H., and Lissitzky, S. (1970) Purification of animal neurotoxins. Isolation and charac terization of eleven neurotoxins from the venom of the scorpion Androctonus australis Hector, Buthus occitanus tunetanus and Leiurus quinquestriatus quinquestriatus. Eur. J. Biochem. 16, 514–523.

520

Martin-Eauclaire

3. Catterall, W. A. (1975) Cooperative activation of action potential sodium ionophore by neurotoxins. Proc. Natl. Acad. Sci. USA 72, 1782–1786. 4. Catterall, W. A. (1979) Binding of scorpion toxin to receptor sites associated with sodium channels in frog muscle. J. Gen. Physiol. 74, 375–391. 5. Catterall, W. A. (1980) Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Ann. Rev. Pharmacol. Toxicol. 20, 15–43. 6. Catterall, W. A. (1986) Molecular properties of voltage-sensitive sodium channels. Ann. Rev. Biochem. 55, 953–985. 7. Catterall, W. A., Trainer, V., and Baden, D. G. (1992) Molecular properties of the sodium channels: a receptor formultiple neurotoxins. Bull. Soc. Pathol. Exot. 85(5 pt2), 481–485. 8. Catterall, W. A. (1995) Structure and function of voltage-gated ion channels. Annu. Rev. Biochem. 64, 493–531. 9. Jover, E., Martin-Moutot, N., Couraud, F., and Rochat, H. (1980a) Binding of scorpion toxins to rat brain synaptosomal fractions. Effects of membrane potential ions and other neurotoxins. Biochemistry 19, 463–467. 10. Jover, E., Couraud, F., and Rochat, H. (1980b) Two types of scorpion neurotoxins characterized by their binding to two separate receptor sites on rat brain synaptosomes. Biochem. Biophys. Res. Commun. 95, 1607–1614. 11. Strichartz, G., Rando, T., and Wang, G. K. (1987) An integrated view of the molecular toxinology of sodium channel gating in excitable cells. Ann. Rev. Neurosci. 10, 237–269. 12. Carbone, E., Wanke, E., Prestipino, G., Possani, L. D., and Haelicke, A. (1982) Selective blockage of voltage-dependent K+ channels by a novel scorpion toxin. Nature 296, 90–91. 13. Strong, P. N. (1990) Potassium channel toxins. Pharmacol. J. 46, 137–162. 14. Miller, C. (1995) The charybdotoxin family of K+ channel-blocking peptides. Neuron. 15, 5–10. 15. Tytgat, J., Chandy, K. G., Garcia, M. L., Gutman, G. A., Martin-Eauclaire, M-F., Van der Walt, J. J., and Possani, L. D. (1999) A unified nomenclature for short-chain peptides isolated from scorpion venoms: α-KTx molecular subfamilies. TIPS, 20, 444–447. 16. DeBin, J. A., Maggio, J. E., and Strichartz, G. R. (1993) Purification and characterization of chlorotoxin, a chloride channel ligand from the venom of the scorpion Leiurus quinquestriatus hebraeus. Am. J. Physiol. 264, 361–369. 17. Zamudio, F. Z., Conde, R., Arévalo, C., Becerril, B., Martin, B. M., Valdivia, H. H., and Possani, L. D. (1997a) The mechanism of inhibition of ryanodine receptor channels by imperatoxin I, a heterodimeric protein from the scorpion Pandinus imperator. J. Biol. Chem. 272, 11,886–11,894. 18. Zamudio, F. Z., Gurrola, G. B., Arévalo, C., Sreekumar, R., Walker, J. W., Valdivia, H. H. and Possani, L. D. (1997b) Primary structure and synthesis of Imperatoxin A (IpTXa), a peptide activator of Ca2+ release channels/ryanodine receptors. FEBS Lett. 405, 385–389. 19. Naharashi, T. (1974) Chemicals as tools in the study of excitable membranes. Physiol. Rev. 54, 813–889. 20. Romey, G., Chicheportiche, R., Lazdunski, M., Rochat, H., Miranda, F., and Lissitzky, S. (1975) Scorpion neurotoxin: a presynaptic toxin which affects both Na+ and K+ channels in axons. Biochem. Biophys. Res. Commun. 64, 115–121. 21. Cahalan, M. D. (1975) Modification of sodium channel gating in frog myelinated nerve fibres by Centruroides sculpturatus scorpion venom. J. Physiol. 244, 511–534. 22. Couraud, F., Jover, E., Dubois, J. M., and Rochat, H. (1982) Two types of scorpion toxin receptor sites, one related to the activation, the other to the inactivation of the action potential sodium channel. Toxicon 40, 9–16. 23. Yatani, A., Kirsch, G. E., Possani, L. D., and Brown, A. M. (1988) Effects of New World scorpion toxins on single-channel and whole cell cardiac sodium currents. Am. J. Physiol. 254, H 443, H 451.


521

24. Gordon, D., Savarin, P., Gurevitz, M., and Zinn-Justin, S. (1998) Functional anatomy of scorpion toxins affecting sodium channels. J. Toxicol. Toxin Rev. 17, 131–159. 25. Fontecilla-Camps, J. C., Almassy, R. J, Suddath, F. L., Watt, D. D., and Bugg, C. E. (1980) Three-dimensional structure of a protein from scorpion venom: a new structural class of neurotoxins. Proc. Natl. Acad. Sci. USA 85, 7443–7447. 26. Fontecilla-Camps, J. C., Habersetzer-Rochat, C., and Rochat, F. (1988) Orthorombic crystals and three-dimensional structure of the potent toxin II from the scorpion Androctonus australis Hector. Proc. Natl. Acad. Sci. USA 85, 7443–7447. 27. Kharrat, R., Darbon, H., Rochat, H., and Granier, C. (1989) Structure-activity relationships of scorpion α-toxins. Multiple residues contribute to the interaction with receptors. Eur. J. Biochem. 181, 381–390. 28. Hassani, O., Mansuelle, P., Cestèle, S., Bourdeaux, M., Rochat, H., and Sampieri, F. (1999a) Role of lysine and tryptophan residues in the biological activity of toxin VII (Ts γ) from the scorpion Tityus serrulatus. Eur. J. Biochem. 260, 76–86. 29. Martin-Eauclaire, M-F. and Couraud, F. (1995) Scorpion neurotoxins: effects and mechanisms, in Handbook of Neurotoxicology (Chang, L. W., Dyer, R. S., eds.), Basel, HongKong, pp. 683–716. 30. Martin-Eauclaire, M-F. and Rochat, H. (1986) Large scale purification of toxins from the venom of the scorpion Androctonus australis Hector. Toxicon 24, 1131–1139. 31. Tejedor, F. J. and Catterall, W. A. (1988) Site of covalent attachment of α-scorpion toxin derivatives in domain I of the sodium channel α-subunit. Proc. Natl. Acad. Sci. USA 85, 8742–8746. 32. Thomsen, W. J. and Catterall, W. A. (1989) Localization of the receptor site for α-scorpion toxins by antibody mapping: implications for sodium channel topology. Proc. Natl. Acad. Sci. USA 86, 10,161–10,165. 33. Rogers, J. C., Qu, Y., Tanada, T. N., Scheuer, T., and Catterall, W. A. (1996) Molecular determinants of high affinity binding of α-scorpion toxin and sea anemone toxin in S3-S4 extracellular loop in domain IV of the Na+ channel α subunit. J. Biol. Chem. 271, 15,950– 15,962. 34. Diniz, C. R. (1978) Chemical and pharmacological aspects of Tityniae venoms, in Arthropod Venoms, Handbook of Experimental Pharmacology (Bettini, S. ed.), Springer-Verlag, Berlin, 379–394. 35. Barhanin, J., Giglio, J. R., Leopold, P., Schmid, A., Sampaio, S. V., and Lazdunski, M. (1982) Tityus serrulatus venom contains two classes of toxins. J. Biol. Chem. 257, 12,553– 12,558. 36. Vijverberg, H. P. M., Pauron, O., and Lazdunski, M. (1984) The effects of Tityus serrulatus scorpion toxin γ on Na channels in neuroblastoma cells. Pflugers Arch. 401, 297–303. 37. De Lima, M. E., Martin-Eauclaire, M. F., Chavez-Olortegui, C., Diniz, C. R., and Granier, C. (1992) Tityus serrulatus venom toxins display a complex pattern of antigenic reactivity. Toxicon 31, 223–227. 38. Martin-Eauclaire, M. F., Céard, B., Ribeiro, A. M., Diniz, C. R., Rochat, H., and Bougis, P. E. (1994) Biochemical, pharmacological and genomic characterization of Ts IV, an αtoxin from the venom of the South American scorpion Tityus serrulatus. FEBS Lett. 342, 181–184. 39. Gordon, D., Martin-Eauclaire, M-F., Cestele, S., Kopeyan, C., Carlier, E., Ben Khalifa, R., et al. (1996) Scorpion toxins affecting sodium current inactivation bind to distinct homologous receptor sites on rat brain and insect sodium channels. J. Biol. Chem. 271, 8034– 8045. 40. Eitan, M., Fowler, E., Herrman, R., Duval, A., Pelhate, M., and Zlotkin, E. (1990) A scorpion venom neurotoxin paralytic to insects that affects sodium current inactivation: purification, primary structure, and mode of action. Biochemistry 29, 5941–5947.

522

Martin-Eauclaire

41. Kopeyan, C., Mansuelle, P., Martin-Eauclaire, M-F., Rochat, H., and Miranda, F. (1993) Characterization of toxin III of the scorpion Leiurus quinquestriatus quinquestriatus: a new type of alpha-toxin highly toxic both to mammals and insects. Natural Toxins 1, 308–312. 42. Borchani, L., Stankiewicz, M., Kopeyan, C., Mansuelle, P., Kharrat, R., Cestele, S., et al. (1997a) Purification, structure and activity of three insect toxins from Buthus occitanus tunetanus venom. Toxicon 35, 365–382. 43. Cestéle, S., Stankiewicz, M., Mansuelle, P., De Waard, M., Dargent, B., Gilles, N., et al. (1999) Scorpion α-like toxins, toxic to both mammals and insects, differentially interact with receptor site 3 on voltage-gated sodium channels in mammals and insects. Eur. J. Neurosci. 11, 975–985. 44. Gilles, N., Blanchet, C., Shichor, I., Zaninetti, M., Lotan, I., Bertrand, D. and Gordon, D. (1999) A scorpion α-like toxin that is active on insects and mammals reveals an unexpected specificity and distribution of sodium channel subtypes in rat brain neurons. J. Neurosci. 19, 8730–8739. 45. Chejanovsky, N., Zilberberg, N., Rivkin, H., Zlotkin, E., and Gurevitz, M. (1995) Functional expression of an alpha anti-insect scorpion neurotoxin in insect cells and lepidopterous larvae. FEBS Lett. 376, 181–184. 46. Zilberberg, N., Froy, O., Loret, E., Cestele, S., Arad, D., Gordon, D., and Gurevitz, M. (1997) Identification of structural elements of a scorpion α-neurotoxin important for receptor site recognition. J. Biol. Chem. 272, 14,810–14,816. 47. Chuang, R. S-I., Jaffe, H., Cribbs, L., Perez-Reyes, E., and Swartz, K. J. (1998) Inhibition of T-type voltage-gated calcium channels by a new scorpion toxin. Nature Neurosci. 1, 668–674. 48. Babin, D. R., Watt, D. D. Goos, S. M. and Mlejnek, R. V. (1974) Amino acid sequence of neurotoxin protein variants from the venom of Centruroides sculpturatus Ewing. Arch. Biochem. Biophys. 164, 694–706. 49. Possani, L. D. Martin, B. M., Mochca-Morales, J., and Svendsen, I. (1981a) Purification and chemical characterization of the major toxins from the venom of the Brazilian scorpion Tityus serrulatus Lutz, and Mello, Carlsberg Res. Comm. 46, 195–205. 50. Possani, L. D., Steinmetz, W. E., Dent, M. A. R., Alagon, A., C. and Wüthrich, K. (1981b) Preliminary spectroscopic characterization of six toxins from latin american scorpions. Biochem. Biophys. Acta. 669, 183–192. 51. Possani, L. D., Martin, B. M., and Svendsen, I. (1982) The primary structure of noxius toxin: a K+ channel blocking peptide, purified from the the venom of the scorpion Centruroides noxius Hoffman Carlsberg. Res. Comm. 47, 285–289. 52. Possani, L. D., Martin, B. M., Svendsen, I., Rode, G. S. and Erickson, B. W. (1985) Scorpion toxins from Centruroides noxius and Tityus serrulatus. Biochem. J. 229, 739–750. 53. Possani, L. D., Martin, B. M., Fletcher, M. D., and Fletcher, P.L. (1991) Discharge effect on pancreatic exocrine secretion produced by toxins purified from Tityus serrulatus scorpion venom. J. Biol. Chem. 266, 3178–3185. 54. Martin, M. F., Garcia y Perez, L. G., El Ayeb, M., Kopeyan, C., Bechis, G., Jover, E., and Rochat, H. (1987a) Purification and chemical and biological characterization of seven toxins from the Mexican scorpion toxin: a neural property. J. Biol. Chem. 262, 4452–4459. 55. Cestèle, S., Qu, Y., Rogers, J., Rochat, H., Scheuer, T., and Catterall, W. (1998) Voltage sensor-trapping: enhanced activation of sodium channels by ß-scorpion toxin bound to the S3-S4 loop in domain II. Neuron 21, 919–931. 56. Barhanin, J., Pauron, D., Lombet, A., Norman, R. I., Vijverberg, P. M., Giglio, J. R., and Lazdunski, M. (1983a) Electrophysiological characterization, solubilization and purification of the Tityus gamma toxin receptor associated with the gating of the sodium channel from rat brain. EMBO J. 2, 915–920.


523

57. Bechis, G., Sampieri, F., Yuan, P. M., Brando, T., Martin, M. F., Diniz, C. R., and Rochat, H. (1984) Amino acid sequence of toxin VII, a ß toxin from the venom of the scorpion Tityus serrulatus. Biochem. Biophys. Res. Comm. 122, 1146–1153. 58. Céard, B., De Lima, M. E., Bougis, P. E., and Martin-Eauclaire, M. F. (1992) Purification of the main ß-toxin from Tityus serrulatus venom using HPLC. Toxicon 30, 105–110. 59. Martin-Eauclaire, M. F., Céard, B., Ribeiro, A. M., Diniz, C. R., Rochat, H., and Bougis, P. E. (1992) Molecular cloning and nucleotide sequence analysis of a cDNA encoding the main ß-neurotoxin from the venom of the South American scorpion Tityus serrulatus. FEBS Lett. 302, 220–222. 60. Polikarpov, I., Sanches Matilde Junior, M., Marangoni, S., Toyoma, M. H. and Teplyakov, A. (1999) Crystal structure of neurotoxin Ts1 from Tityus serrulatus provides insights into the specificity and toxicity of scorpion toxins. J. Mol. Biol. 290, 175–184. 61. Marcotte, P., Chen, L. Q., Kallen, R. G., and Chahine, M. (1997) Effects of Tityus serrulatus scorpion toxin gamma on voltage-gated Na+ channels. Circ. Res. 80, 363–369. 62. Pauron, D., Barhanin, J., and Lazdunski, M. (1985) The voltage-dependent sodium channel of insect nervous system identified by receptor sites for tetrodotoxin and scorpion and sea anemone toxins. Biophys. Res. Commun. 131, 1226–1233. 63. De Lima, M. E., Martin, M. F., Diniz, C. R., and Rochat, H. (1986) Tityus serrulatus bears pharmacological properties of both ß-toxin and insect toxin from scorpion venoms. Biochem. Biophys. Res. Comm. 139, 296–302. 64. De Lima, M. E., Martin-Eauclaire, M. F., Hue, B., Loret, E., Diniz, C. R., and Rochat, H.(1989) On the binding of two scorpion toxins to the central nervous system of the cockroach Periplaneta americana. Insect Biochem. 19, 413–422. 65. Loret, E., Sampieri, F., Roussel, A., Granier, C., and Rochat, H. (1990) Conformational flexibility of a scorpion active on mammals and insects: a circular dichroism study. Proteins 8, 164–172. 66. Garcia, C., Becerril, B., Selisko, B., Delepierre, M., and Possani, L. D. (1997) Isolation, characterization and comparison of a novel crustacean toxin with a mammalian toxin from the venom of the scorpion Centruroides noxius Hoffmann. Comp. Biochem. Physiol. 116B(3), 315–322. 67. Selisko, B., Garcia, C., Becerril, B., Delepierre, M., and Possani, L. D. (1996) An insectspecific toxin from Centruroides noxius Hoffmann: cDNA, primary structure, three dimensional model and electrostatic surface potentials in comparison with other toxin variants. Eur. J. Biochem. 242, 235–241. 68. Zlotkin, E., Miranda, F., and Rochat, H. (1972) Proteins in scorpion venoms toxic on mammals and insects. Toxicon 10, 211–216. 69. Gordon, D., Jover, E., Couraud, F., and Zlotkin, E. (1984) The binding site of the insect selective neurotoxin (AaIT) from scorpion venom locust synaptosomal membrane. Biochem. Biophys. Acta. 778, 349–358. 70. Gordon, D., Moskowitz, H., Warmer, C., Catterall, W. A., and Zlotkin, E. (1992) Localization of the receptor sites for insect-selective toxins on sodium channels by site-directed antibodies. Biochemistry 31, 7622–7628. 71. De Dianous, S., Hoaro, F., and Rochat, H. (1987) Re-examination of the specificity of the scorpion Androctonus australis Hector insect toxin towards arthropods. Toxicon 25, 411– 417. 72. Zlotkin, E., Devonshire, A. L., and Warmke, J. W. (1999) The pharmacological flexibility of the insect voltage gated sodium: toxicity of AaIT to knockdown resistant (kdr) flies. Insect Biochem. Mol. Biol. 29, 849–853. 73. Lazarovici, P., Yanai, P., Pelhate, H. G., and Zlotkin, E. (1982) Insect toxic components from the venom of a chactoid scorpion, Scorpio maurus palmatus (scorpionidae). J. Biol. Chem. 257, 8397–8404.

524

Martin-Eauclaire

74. Zlotkin, E., Kadouri, D., Gordon, D., Pelhate, M., Martin, M. F., and Rochat, H. (1985) An excitatory and a depressant insect toxin from scorpion venom both affect sodium conductance and possess a common binding site. Arch. Biochem. Biophys. 240, 877–887. 75. Darbon, H., Zlotkin, E., Kopeyan, C., Van Rietschoten, J., and Rochat, H. (1982) Covalent structure of the insect toxin of the North African scorpion Androctonus australis Hector. Int. J. Pept. Prot. Res. 20, 320–330. 76. Darbon, H., Weber, C., and Brawn, W. (1991) Covalent structure of the insect toxin of the north african scorpion Androctonus australis Hector. Biochemistry 30, 1836–1844. 77. Oren, D. A., Froy, O., Amit, E., Kleinberger-Doron, N., Gurevitz, M., and Shaanan, B. (1998) An excitatory scorpion toxin with a distinctive feature: an additional α-helix at the C-terminus and its implications for interaction with insect sodium channels. Structure 6, 1095–1103. 78. Kopeyan, C., Mansuelle, P., Sampieri, F., Brando, T., Bahraoui, E. M., Rochat, H., and Granier, C. (1990) Primary structure of scorpion anti-insect toxins isolated from the venom of Leiurus quinquestriatus quinquestriatus. FEBS Lett. 261, 423–426. 79. Zilberberg, N., Zlotkin, E., and Gurevitz, M. (1991) The cDNA sequences of a depressant insect selective neurotoxin from the scorpion Buthatus judaicus. Toxicon 29, 1155–1158. 80. Zlotkin, E., Eitan, M., Bindokas, V. P., Adams, M. E., Moyer, M., Burkhart, W., and Fowler, E. (1991) Functional duality and structural uniqueness of depressant insect-selective neurotoxins. Biochemistry 30, 4814–4821. 81. Zlotkin, E., Gurevitz, M., Fowler, E., and Adams M. E. (1993) Depressant insect selective neurotoxins from scorpion venom: chemistry, action and gene cloning. Arch. Insect. Biochem. Physiol. 22, 55–73. 82. Borchani, L., Mansuelle, P., Stankiewicz, M., Grolleau, F., Cestele, S., Karoui, H., Lapied, B., et al. (1997b) A new scorpion venom toxin paralytic to insects that affects Na+ channel activation. Purification, structure, antigenicity and mode of action. Eur. J. Biochem. 24, 525–532. 83. Cestele, S., Kopeyan, C., Oughideni, R., Mansuelle, P., Granier, C., and Rochat, H. (1997a) Biochemical and pharmacological characterization of a depressant insect toxin from the venom of the scorpion Buthacus arenicola. Eur. J. Biochem. 243, 93–99. 84. Nakagawa, Y., Lee, Y. M., Lehmberg, E., Herrmann, R., Herrmann, R., Moskowitz, H., Jones, A. D., and Hammock, B. D. (1997) Anti-insect toxin (AaIT5) from Androctonus australis. Eur. J. Biochem. 240, 496–501. 85. Hassani, O., Loew, D., Van Dorsselaer, A., Papandreou, M. J., Sorokine, O., Rochat, H., et al. (1999b) Aah VI, a novel, N-glycosylated anti-insect toxin from Androctonus australis hector scorpion venom: isolation, characterization, and glycan structure determination. FEBS Lett. 443, 175–180. 86. Stankiewicz, M., Grolleau, B., Lapied, B., Borchani, M., El Ayeb, M., and Pelhate, M. (1996) Bot IT2, a toxin paralytic to insects from the Buthus occitanus tunetanus venom modifying the activity of insect sodium channels. J. Insect Physiol. 42, 397–405. 87. Cestèle, S., Borchani, L., El Ayeb, M., and Rochat, H. (1997b) Bot IT2: a new scorpion toxin to study receptor site on insect sodium channels. FEBS Lett. 405, 77–80. 88. Hille, B. (1992) Ionic Channels of Excitable Membranes, 2nd ed., Sinauer Associates, Sunderland, MA. 89. Wei, A., Jegla, T., and Salkoff, L. (1996) Eight potassium channel families revealed by the C. elegans genome project. Neuropharmacology 35, 805–829. 90. Kaczorowski, G. J. and Garcia, M. L. (1999) Pharmacology of voltage-gated and calciumactivated potassium channels. Curr. Opin. Chem. Biol. 3, 448–458. 91. Garcia, M. L., Haner, M., Knaus, H. G., Slaughter, R., and Kaczorowski, G. J. (1999) Scorpion toxins as tools for studing potassium channels. Methods Enzymol. 294, 624–649. 92. Bontems, F., Roumestand, C., Boyot, P., Gilquin, B., Doljansky, Y., Menez, A., and Toma, F. (1991) Three-dimensional structure of natural charybdotoxin in aquous solution by 1HNMR. Charybdotoxin possesses a structural motif found in other scorpion toxins. Eur. J. Biochem. 196, 19–28.


525

93. Legros, C., Bougis, P. E., and Martin-Eauclaire, M-F. (1998) Evidence for a new class of scorpion toxins active against K+ channels. FEBS Lett. 431, 375–380. 94. Goldstein, S. A. N, Pheasant, D. J., and Miller, C. (1994) The charybdotoxin receptor of a Shaker K+ channel: peptide and channel residues mediating recognition. Neuron 12, 1377–1388. 95. Aiyar, J., Withka, J. M., Rizzi, J. P., Singleton, D. H., Andrews, G. C., Simon, M., et al. (1995) Toplogy of the pore-region of a K+ channel revealed by the NMR-derived structures of scorpion toxins. Neuron 15, 1169–1181. 96. Darbon, H., Blanc, E., and Sabatier, J.-M. (1999) Three-dimentional structure of scorpion toxins: towards a new model of interaction with potassium channels, in Animals Toxins and Potassium Channels (Darbon, H. and Sabatier, J,-M., eds.), Perspectives in Drug Discovery and Design, vol. 15/16, pp. 41–60. 97. Possani, L. D., Selisko, B., and Gurrola, G. B. (1999) Structure and function of scorpion toxins affecting K+ channels, in Animals Toxins and Potassium Channels (Darbon, H. and Sabatier, J.-M., eds.), Perspectives in Drug Discovery and Design, vol. 15/16, pp. 15-40. 98. Pongs, O. (1999a) Voltage-gated potassium channels: from hyperexcitability to excitement. FEBS Lett. 452, 31–35. 99. Pongs, O. (1999b) Pharmacology of voltage-gated potassium channels, in Animals Toxins and Potassium Channels (Darbon, H. and Sabatier, J.-M., eds.), Perspectives in Drug Discovery and Design, vol. 15/16, pp. 215–225. 100. Garcia-Calvo, M., Leonard, R. J., Novick, J., Stevens, S. P., Schmalhofer, W., Kaczorowski, G. J., and Garcia, M. L. (1993) Purification, characterization, and biosynthesis of margatoxin, a component of Centruroides margaritus venom that selectively inhibits voltage-dependent potassium channels. J. Biol. Chem. 268, 18,866–18,874. 101. Knaus, H. G., Koch, R O. A., Eberhart, A., Kaczorwski, G. J., Garcia, M. L., and Slaughter, R. S. (1995) 125I-Margatoxin, an extraordinarily high affinity ligand for voltage-gated potassium channels in mammalian brain. Biochemistry 34, 13,627–13,634. 102. Romi, R., Crest, M., Gola, M., Sampieri, F., Jacquet, G., Zerrouk, H., Mansuelle, P., Sorokine, O., Van Dersaller, A., Rochat,, H., et al. (1993) Synthesis and characterization of Kaliotoxin. Is the 26-32 sequence essential for potassium channel recognition? J. Biol. Chem. 268, 26,302–26,309. 103. Grissmer, S., Nguyen, A. N., Aiyar, J., Hanson, D. C., Mather, R. J., Gutman, G. A., et al. (1994) Pharmacological characterization of five cloned voltage-gated K+ channels, type Kv1.1, 1.2, 1.3, 1.5, and 3.1, stably expressed in mammalian cell lines. Mol. Pharmacol. 45, 1227–1234. 104. Mourre, C., Chernova, M. N., Martin-Eauclaire, M-F., Bessone, R., Jacquet, G., Gola, M., et al. (1999) Distribution in rat brain of binding sites of Kaliotoxin, a blocker of Kv1.1 and Kv1.3 α-subunits. J. Pharmacol. Exp. Ther. 291, 943–952. 105. Lipkind, G. M. and Fozzard, H. A. (1999) Molecular modeling of scorpion toxin binding to voltage-gated K+ channels, in Animals Toxins and Potassium Channels (Darbon, H. and Sabatier, J.-M., eds.), Perspectives in Drug Discovery and Design, vol. 15/16, pp. 245–255. 106. Park, C. S. and Miller, C. (1992a) Interaction of Charybdotoxin with permeant ions inside the pore of a K+ channel. Neuron 9, 307–313. 107. Park, C. S. and Miller, C. (1992b) Mapping function to structure in a channel-blocking peptide: electrostatic mutants of charybdotoxin. Biochemistry 32, 7749–7755. 108. Galvez, M. L., Gimenez-Gallego, G., Reuben, J. P., Roy-Constancin, L., Feigenbaum, P., Kaczorowski, G. J., and Garcia, M. L. (1990) Purification and characterization of a unique, potent, peptidyl probe for the high conductance calcium-activated potassium channel from the venom of the scorpion Buthus tamulus. J. Biol. Chem. 265, 11,083– 11,090. 109. Sugg, E. E., Garcia, M. L., Reuben, J. P., Patchett, A. A., and Kaczorowski, G. J. (1990)

526

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121. 122. 123.

124.

125.

Martin-Eauclaire Synthesis and structural characterization of charybdotoxin, a potent peptidyl inhibitor of the high conductance Ca2+ activated K+ channel. J. Biol. Chem. 265, 18,745–18,748. Johnson, B. A. and Sugg, E. E. (1992) Determination of the three-dimensional structure of Iberiotoxin in solution by 1H nuclear resonance spectroscopy. Biochemistry 31, 8151– 8159. Giangiacomo, K. M., Sugg, E. E., Garcia-Calvo, M., Leonard, R. J., McManus, O. B., Kaczorowski, G. J., and Garcia, M. L. (1993) Synthetic charybdotoxin-iberiotoxin chimeric peptides define toxin binding sites on calcium-activated and voltage-dependent potassium channels. Biochemistry 32, 2363–2370. Chicci, G. G., Gimenez-Gallego, G., Ber, E., Garcia, M. L., Winquist, R., and Cascieri, M. A. (1988) Purification and characterization of a unique potent inhibitor of apamin binding from Leiurus quinquestriatus hebraeus. J. Biol. Chem. 263, 10,192–10,197. Auguste, P., Hugues, M., Gravés, B., Gesquiers, J. C., Maes, P., Tartar A., et al. (1990) Leiurotoxin I (Scyllatoxin), a peptide ligand for Ca2+ activated K+ channels: structurefunction relationships and brain localization of the bin ding sites. J. Biol. Chem. 265, 4753–4759. Sabatier, J. M., Zerrouk, H., Darbon, H., Mabrouk, K., Benslimane, A., Rochat, H., et al. (1993) P05, a new Leiurotoxin I-like scorpion toxin: synthesis and structure-activity relationships of the α-amidated analog, a ligand of Ca2+ activated K+ channels with increased affinity. Biochemistry 32, 2763–2770. Romi-Lebrun, R., Martin-Eauclaire, M-F., Escoubas, P., Wu, F. Q., Lebrun, B., Hisada, M., and Nakajima, T. (1997) Characterization of four toxins from Buthus martenzi scorpion venom, which act on apamin-sensitive Ca2+ activated K+ channels. Eur. J. Biochem. 245, 457–464. Lecomte, K., Ferrat, G., Fajloun, Z., Van Rietschoten, J., Rochat, H., Martin-Eauclaire, M-F., et al. (1999) Chemical synthesis and structure-activity relationships of Ts κ, a novel scorpion toxin acting on apamin-sensitive SK channel. J. Peptide. Res. 54, 369–376. Rogowski, R. S., Krueger, B. K., Collins, J. H., and Blaustein, M. P. (1994) Tityustoxin Kα blocks voltage-gated noninactivating K+ channels and unblocks inactivating K+ channels blocked by α-dendrotoxin in synaptosomes. Proc. Natl. Acad. Sci. USA 91, 1475–1479. Gurrola, G., Rosati, B., Rocchetti, M., Pimienta, G., Zaza, A., Arcangeli, A., et al. (1999a) A toxin to nervous, cardiac, and endocrine ERG K+ channels isolated from Centruroides noxius scorpion venom. FASEB J. 13, 953–962. Grishin, E. V., Volkova, T. M., and Soldotova, L. N. (1982) A study of the toxic component of the venom of the caucasian subspecies of the scorpion Buthus eupeus. Bioorganika Khim. 8, 155–164. Rosso, J. P. and Rochat, H. (1985) Characterization of 10 proteins from the venom of the Moroccan scorpion Androctonus mauretanicus mauretanicus, six of which are toxic to the mouse. Toxicon 23, 113–125. Arseniev A. S., Kondakov U. N., Maiorov, V. N., and Bystrov V. F. (1984) NMR solution spatial structure of “short” scorpion insectotoxin I5A. FEBS Lett. 165, 57–62 DeBin, J. A. and Strichartz G. R. (1991) Chloride channel inhibition by the venom of the scorpion Leiurus quinquestriatus. Toxicon 29, 1403–1408. Lippens, G., Najib, J., Wodak, S. J., and Tartar, A. (1995) NMR sequential assignments and solution structure of chlorotoxin, a small scorpion toxin that blocks chloride channels. Biochemistry 34, 13–21. Ullrich, N. and Sontheimer, H. (1996) Biophysical and pharmacological characterization of chloride currents in human astrocytoma cells. Am. J. Physiol. Cell Physiol. 270, C1511–C1521. Soroceaunu, L., Gillespie, Y., Khazaeli, M. B., and Sontheimer H. (1998) Use of chlorotoxin for targeting of primary brain tumors. Neurol. Res. 58, 4871–4879.


527

126. Gurrola, G., Arevalo, C., Sreekumar, R., Lokuta, A. J., Walker, J. W., and Valdivia, H. H. (1999b) Activation of ryanodine receptors by imperatoxin A and a peptide segment of the II-III loop of the dihydropyridine receptor. J. Biol. Chem. 274, 7879–7886. 127. El Hayek, R. and Ikemoto. N. (1998) Identification of the minimum essential region in the II-III loop of the dihydropyridine receptor α1 subunit required for activation of skeletal muscle-type excitation-contraction coupling. Biochemistry 37, 7015–7020. 128. Bougis, P., Rochat, H., and Smith, L. A. (1989) Precursors of Androctonus australis scorpion neurotoxin: structure of precursors, precessing outcomes and expression of a functional recombinant toxin II. J. Biol. Chem. 264, 19,259–19,265. 129. Becerril, B., Corona, M., Garcia., M., Bolivar, F., and Possani, L. D. (1995) Cloning of genes encoding scorpion toxins: an interpretative review. J. Toxicol. Toxins Rev. 14, 339–357. 130. Bouhaouada-Zahar, B., Ducancel, F., Zenouaki, I., Ben Khalifa, R., Borchani, L., Pelhate, M., et al. (1996) A recombinant insect-specific alpha-toxin of Buthus occitanus tunetanus scorpion confers protection against homologous mammal toxins. Eur. J. Biochem. 238, 653–660. 131. Xiong, Y-M., Ling, M-H., Wang, D-C., and Chi, C-W. (1997) The cDNA and genomic DNA sequences of a mammalian neurotoxin from the scorpion Buthus martensii Karsch. Toxicon 35, 1025–1031. 132. Laraba-Djebari, F., Legros, C., Crest, M., Céard, B., Romi, R., Mansuelle, P., et al. (1994) The kaliotoxin family enlarged: purification, characterization and precursor nucleotide sequence of KTX2 from Androctonus australis venom. J. Biol. Chem. 269, 32,835– 32,843. 133. Legros, C., Oughideni, R., Darbon, H., Rochat, H., Bougis, P. E., and Martin-Eauclaire, M-F. (1996) Characterization of a new peptide from Tityus serrulatus scorpion venom which is a ligand of the apamin-binding site. FEBS Lett. 390, 81–84. 134. Wu, J-J., Dai, L., Lan, Z-D., and Chi, C-W. (1999) Genomic organization of three neurotoxins active on small conductance Ca2+ activated potassium channels from the scorpion Buthus martensi Karch. FEBS Lett. 452, 360–364. 135. Froy, O., Sagiv., T., Poreh, M., Urbach., D., Zilberberg, N., and and Gurevitz, M. (1999a) Dynamic diversification from a putative common ancestor of scorpion toxins affecting sodium, potassium, and chloride channels. J. Mol. Evol. 48, 187–196. 136. Zhu, S., Li, W., Zeng, X., Jiang, D., Mao, X., and Liu, H. (1999) Molecular cloning and sequencing of two “short chain” and two “long” chain K+ channel blocking peptides from the Chinese scorpion Buthus martensii Karsch. FEBS Lett. 457, 509–514. 137. Delabre, M-L., Pasero, P., Marilley, M., and Bougis, P. E. (1995) Promoter structure and intron-exon organization of a scorpion α-toxin gene. Biochemistry 34, 6729–6736. 138. Legros, C., Bougis, P. E., and Martin-Eauclaire, M-F. (1997a) Genomic organization of the KTX2 gene, encoding a «short» scorpion toxin active on K+ channel. FEBS Lett. 402, 45. 139. Park, C. S., Hausdorff, S. F., and Miller, C. (1991) Design, synthesis and functional expression of a gene for charybdotoxin, a peptide blocker of K+ channels. Proc. Natl. Acad. Sci. USA 88, 2046–2050. 140. Garcia, M. L., Garcia-Calvo, M., Hidalgo, P., Lee, A., and McKinnon, R. (1994) Purification and characterization of three inhibitors of voltage-dependent K+ channels from Leiurus quinquestriatus var. hebraeus venom. Biochemistry 33, 6834–6839. 141. Legros, C., Feyfant, E., Sampieri, F., Rochat, H., Bougis, P. E., and Martin-Eauclaire, MF. (1997b) Influence of a NH2-terminal extension on the activity of KTX2, a K+ channel blocker purified from Androctonus australis scorpion venom. FEBS Lett. 417, 123. 142. Stewart, L. M. D., Hirst, M., Ferber, M. G., Merryweather, A. T., Cayley, P. J., and Possee, R. D. (1991) Construction of an improved baculovirus insecticide containing an insect-specific toxin gene. Nature 352, 85–88. 143. Maeda, S., Volrath, S. L., Hanzlik, T. N., Harpers, S. A., Majuma, K., Maddox, D. W., et

528

Martin-Eauclaire

al. (1991) Insecticidal effects of an insect-specific neurotoxin expressed by a recombinant baculovirus. Virology 164, 777–780. 144. McCutchen, B. F., Choudary, P. V., Crenshaw, R., Maddox, D., Kamita, S. G., Palekar, N., et al. (1991) Development of a recombinant baculovirus expressing an insect selective neurotoxin: potential for pest control. Bio/technology (NY) 9, 848–852. 145. Tugarinov, V., Kustanovich, I., Zilberberg, N., Gurevitz, M., and Anglisher, J. (1997) Solution structures of a highly insecticidal recombinant recombinant scorpion α-toxin and a mutant with increased activity. Biochemistry 36, 2414–2424. 146. Froy, O., Zilberberg, N., Gordon, D., Turkov, M., Gilles, N., Stankiewicz, M., et al. (1999b) Th putative bioactive surface of insect-selective scorpion excitatory neurotoxins. J. Biol. Chem. 274, 5769–5776.

Anthozoan Neurotoxins

529

25 Anthozoan Neurotoxins William R. Kem

1. INTRODUCTION 1.1. Toxins as Research Tools and Models for Drug Design Naturally occurring toxins can be exquisitely useful chemical tools in neurobiological research. For instance, using an Indian arrow poison extract (curare), the French physiologist Claude Bernard obtained the initial evidence that cells communicate with each other in the nervous system by chemical signals. Also, two potent and highly selective toxins, α-bungarotoxin and tetrodotoxin, served as important chemical tools in first isolating two ion channel membrane proteins, the muscle nicotinic receptor and sodium channel, respectively. A toxin that selectively blocks a particular ion channel can be used to study the physiological and behavioral processes regulated by the channel. Radioisotopically-labeled or fluorescent toxin derivatives can be used to map the distribution of the ion channel, even in different regions within a single cell. In the past decade, these labeled toxins have been extremely valuable probes for identifying new drug leads during high throughput screening of chemical libraries. In addition, some toxins may become excellent molecular models for designing simpler compounds to be used as drugs or pesticides. Nature rarely provides a compound already possessing ideal therapeutic characteristics, such as high efficacy and bioavailability as well as low toxicity. Thus, most natural products with novel activities become “lead” compounds. The process of drug development generally requires synthesis and testing of many structural analogs of the lead compound to optimize therapeutic and minimize toxic properties. It is as important today, as in the past, to find compounds with novel mechanisms and sites of action. Because of their high potency and selectivity of action, toxins are one of the most important classes of natural products providing new leads for drug design. Because of their high affinity and receptor selectivity, radiolabeled toxins are very useful in screening large numbers of extracts for new lead compounds. Because of remarkable improvements in techniques for isolating, chemically characterizing, and synthesizing larger molecules, even relatively large molecules such as these peptides may become attractive models for drug design. The investigative journey from the initial discovery of a novel toxin to the development of a new drug based on it follows an unpredictable route that may require From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

529

530

Kem

many years to complete. Bernard’s curare of 1859 only became a medically useful muscle relaxant in 1941. Fortunately, this journey requires far less time today. 1.2. A Major Anthozoan Neurotoxin Target: Voltage-Gated Ion Channels Because the normal functioning of voltage-gated ion channels is required for body movement, they are natural targets for venom toxins, which must act quickly to avoid loss of prey or to prevent further atttack by a predator. This is particularly important for the sedentary anthozoans. There are three major types of cation-selective channels that are activated by alteration of membrane potential, namely those passing potassium, sodium, or calcium ions. Perhaps because K channels are almost universally involved in the control of cell resting-membrane potential as well as generating electrical signals, they display the greatest diversity of expression (>60 distinct K channel-subunit genes have been found in the genome of the nematode, Caenorhabdites elegans). Whereas the pore regions of K channels are generally composed of four homologous α subunits, the pore regions of sodium and calcium channels are composed of a single large polypeptide α subunit of ~260,000 Daltons molecular size, which contains four homologous (but nonidentical) domains, each of which is structurally and functionally equivalent to one of the four potassium channel α subunits. Generally, these voltagegated ion channels also contain one or more associated ß subunits, which primarily subserve regulatory functions, such as anchoring the channel to cytoskeletal proteins expressed in particular membrane regions of the cell. Each K channel α subunit or Naor Ca-channel α subunit domain contains six membrane-spanning helical segments (S1-S6). A general molecular model for these channels, first proposed by Guy and Conti (1), predicted that each of the four domains contains two short segments, SSl and SS2, located between the S5 and S6 segments, and these segments together form the central pore. Interaction of the SS1 and SS2 segments of all four domains may allow formation of a ß-barrel type pore during the activated state. The postulated S4 amphipathic helix segments that contain Arg or Lys side chains at regular intervals serve as voltage sensors, which move outwards upon membrane depolarization to convert the channel into its open state. Unfortunately, the folded structures of the various extracellular and intracellular loops connecting the membrane-spanning segments, which are likely receptor sites for many drugs and toxins, are still undefined. While the essential aspects of the Guy-Conti model have received abundant confirmation during the past few years, the secondary structure of the pore-lining region is still uncertain. At least eight distinct toxin binding sites have now been identified on brain sodium channels. Polypeptide toxins, because of their large size, bind at the external surface of the ion channel. One of these sites, called Site 1, binds mu-type conotoxin peptides as well as the alkaloids tetrodotoxin and saxitoxin; all of these toxins block this channel by occluding the external vestibule of this channel. The Site 2 steroidal toxins batrachotoxin and veratridine bind at a more interior, lipophilic site; these toxins cause a persistent sodium-channel activation, even at a normal membrane resting potential. Site 3 peptide toxins, including scorpion α-toxins as well as the sea anemone sodium channel toxins discussed in this chapter, seem to bind at the edge of the external vestibule of this channel; they delay channel inactivation (see Fig. 1). Site 4 peptide toxins (scorpion ß-toxins) bind to an as yet unidentified site on the external side of the chan-


531

Fig. 1. A simplified model of the major conformational states of the voltage-gated sodium channel. The closed (C) state predominates at resting membrane potentials. The open (O) state allows selective passage of sodium ions through the membrane by a process of diffusion. The inactivated (I) state is the stable nonconducting state of the channel under depolarizing conditions, which develops from the open state. Another inactivated state, IC, is reached from the closed state, without prior channel opening. The sea anemone toxins inhibit the development of the inactivated state from the open state, without significantly affecting inactivation from the closed state.

nel; they primarily inhibit channel deactivation (the transition from open state back to the resting state, which is enhanced by membrane repolarization). The Site 5 (ciguatoxin) and Site 6 (pyrethroid insecticides) toxins both enhance channel opening, but probably accomplish this by different mechanisms. Two other external sites, respectively, bind coral (Goniopora) and gastropod (Conus) peptide toxins; both toxins primarily inhibit channel inactivation (2). Molecular biological studies over the past decade have revealed the presence of many different ion-channel variants that are usually products of separate genes (3). Thus, one cannot expect that a toxin’s action will be the same on the different variants of an ion channel even when expressed within a given species. At this time, data for some of the aforementioned toxin-binding sites has been obtained almost exclusively from experiments using the dominant mammalian brain sodium channel variant IIA. Similarly, large pharmacological differences may exist for a given toxin binding site depending upon the animal species. Often, toxin binding is very sensitive to the slightest change (for instance, a single amino acid side chain) in its binding site. Differences between vertebrate and insect channels are best known, as development of selective insecticides depends on such pharmacological differences. Some examples of selective toxicity will be presented later in this article. 1.3. Biology of Anthozoan Venoms The Anthozoa are a subphylum of the Cnidaria (formerly called Coelenterata), a large invertebrate phylum characterized by radial symmetry (Fig. 2A), possession of two well-defined body layers (ectoderm and endoderm) separated by a jelly-like layer known as the mesoglea, a simple nervous system composed of a nerve net, a gastric cavity with a single opening and, most importantly for this chapter, universal occurrence of a unique venom cell, the cnidocyte, in the body column and tentacular integument. The most commonly observed anthozoans are sea anemones, gorgonians, hard

532

Kem


533

corals, sea fans, soft corals, and zoanthids. The other Cnidarian subphylum contains the hydrozoans (Examples: Hydra, Physalia) and scyphozoans (jellyfish); these animals have a mobile medusoid life stage as well as a sedentary polypoid stage reminiscent of the anthozoan body plan. The Cnidarian phylum is rather unique in that practically all of its members are toxic. In other phyla, toxic species may be very numerous within a small number of taxonomic classes, but these species usually represent only a small fraction of the total number of species within the phylum. This high incidence of toxic species makes the Cnidarian phylum particularly attractive for future toxinological investigations. Of the approx 10,000 living Cnidarian species that have been described, a surprisingly large number (about 6500) are anthozoans, and these are mostly sea anemones and corals (4). While our major focus in this chapter will be the neurotoxic constituents of the cnidocyte’s venom, one cannot but marvel at the cellular versatility of the cnidocyte, which possesses not only the capacity to sense efficiently the presence of another animal, but is also able to inject a venom subdermally. Thus, in contrast to most venomous animals that possess well-defined multicellular venom glands, cnidarians possess a cellular apparatus widely distributed over the entire body integument that acts as a chemical defense against predators. Even the lining of the gastric cavity contains some of these stinging cells, which probably to aid in prey immobilization and digestion. The cnidocyte contains a large specialized capsule, called either a cnidocyst or nematocyst (older name), which when discharged shoots a small proteinaceous harpoon-like tubule into the victim that acts as a conduit for passage of the poisonous venom. Once the venom is injected subdermally into the victim, the peptide toxins can quickly reach their sites of action. A considerable amount of kinetic energy is required to drive the everting cnidocyte tubule through the victim’s skin, and this seems to be provided by a high osmotic pressure generated by proteins dissolved within the interior fluid of the cnidocyte (5). Since the wall of the cnidocyst is only permeable to molecules less than 800 Daltons, it it is not so surprising that the enclosed toxins and enzymes are larger peptides; otherwise, they would escape into the cynidocyte cytoplasmic space and possibly damage the cell. It is now well-accepted that cnidocytes are fully capable of acting as independent effectors, sensing as well as activating cnidocyst discharge toward the intended victim. Each cell contains a “cnidocil” ciliary sensory apparatus at its surface, which can be stimulated chemically as well as mechanically; usually both types of stimuli are required for discharge (6). The mechanism by which the little trapdoor at the surface of the cnidocyst is opened, allowing rapid tubular eversion and venom discharge, is still unknown. Fig. 2. Diagram of the basic anatomy of an acontiate sea anemone (A) (shown in crosssection) and its common cnidocysts (B) and (C). Note the presence of acontial threads in the gastric cavity that contain stinging cnidocytes. These threads are passively ejected towards potential predators when the sea anemone rapidly contracts, at the same time ejecting sea water from its coelenteron and gastric cavity. The “cnidome” of a particular anthozoan species is a list of the different morphological types of cnidocytes present. In sea anemones the most common cnidocyst types are mastigaphores (B), which can penetrate the skin of the prey and presumeably inject venom, and spirocysts (C), whose threads stick to the surface of the prey and mechanically prevent its escape. The cnidocyst drawings are modified from Cutress 7.

534

Kem

Each cnidarian possesses several morphologically distinguishable types of cnidocysts; this “cnidome” is very useful in species identification (7,8). While most cnidocysts are designed to penetrate the victim’s skin and deliver a venom, the spirocyst type (Fig. 2B) instead ejects a sticky thread used to adhere to the victim’s skin; many such adhering threads can preventing the victim from escaping. The anthozoan then uses its tentacles to pull the victim slowly towards its mouth. Besides cnidocytes, the anthozoan integument contains several types of glandular cells that together produce a mucous secretion that may also be toxic, although this has not yet been established. The most common known toxic constituents of anthozoans such as sea anemones are small proteins. The two majors types of toxin are (1) cytolytic proteins, which permeabilize cell membranes (9); and (2) neurotoxic peptides, which alter ion channels in nerve- and muscle-cell membranes (10). In addition, at least some sea anemone venoms also contain A-type phospholipases, which catalyze the production and membrane accumulation of lysolecithin (11). This detergent-like substance can also disrupt membranous barriers, including the gill and gastrointestinal mucosal membranes of potential predators. In many venoms, cytolysins and phospholipases work in a synergistic manner, each enhancing the other’s activity (12). Because this volume is only concerned with neurotoxins, these interesting toxins will have to be considered elsewhere. 1.4. Methods of Toxin Purification and Recognition Because cnidocytes are distributed widely over the entire anthozoan surface, it is difficult to extract venom efficiently by irritating the intact living organism, particularly since many anthozoans can rapidly contract their bodies when touched, creating a slimy little ball a mere fraction of the normal resting size. Consequently, most labs, including ours, have resorted to extracting the toxins from whole animals. If possible, this should be done without homogenizing the animals, which renders purification of the relatively small quantitites of peptide toxins even more difficult. Two generally useful methods for initially extracting the toxins from the tissues will now be described briefly. The first method relies on the toxins being soluble and stable to alcohol exposure. Live or frozen animals are soaked in ethanol in the cold room for several days, after which the resulting toxic alcoholic phase is recovered by centrifugation and then concentrated with a rotary evaporator prior to dialysis and chromatography. This method works well for isolating Anthopleura and Anemonia toxins, which are alcohol-soluble (13). Alternately, frozen whole animals can be soaked for a short period of time (2 h) in chilled water. Afterwards, the adhering toxic mucus exudate is stripped from the bodies with gloved hands, and then separated from them by passage through cheese-cloth. After the bodies are then subjected to another freeze-thaw cycle, the pooled watery exudate is again collected as before and centrifuged to remove insoluble materials before the supernatant proteins (including toxins) are precipitated with sufficient solid ammonium sulfate to reach 90% of saturation (about 600 g/L). A third whole body method, especially useful for freeze-dried whole animal extracts, is to homogenize this material in five parts (v/w) cold water, centrifuge, collect the supernatant, and precipitate the proteins in 80% acetone at least as cold as –10°C; this method removes considerable low molecular-weight contaminants such as salts and lipids, which otherwise tend to clog chromatography columns. An appropriate method for


535

extracting the toxins of one species may not be successful for other species. For instance, whereas Stichodactyla helianthus toxins, including the larger cytolytic proteins (14) were efficiently extracted by the aqueous freeze-thaw procedure, the Anthopleura toxins could not be harvested in high yield using this method (Kem, W. R., unpublished results). To obtain only small amounts of toxins, some labs have applied electrical stimulation to whole animals with some success, but the yields have not been reported (15). This method would provide a very large “biological” purification step that avoids killing the anemone as well as having to purify the venom toxins from all the other body proteins. Using relatively small amounts of tissue, McKay and Anderson (16) isolated undischarged cnidocysts from tentacles exposed to papain to liberate these cells and then separated the heavier cnidocytes from tissue debris by centrifugation in the presence of Percoll. These methods provide an alternative to the large-scale methods most labs have been using. Generally sea anemone K-channel toxins occur in much smaller amounts than the sodium-channel toxins and cytolysins. Fortunately, because of their relatively small size and high basicity, these toxins can still be isolated in pure form from whole animal extracts using a combination of conventional and high-pressure column chromatographic methods. The short K-channel toxins also possess high thermal stability, which permits selective denaturation of other proteins including proteases, by heating the extract at 90°C for several minutes and susequently removing them by centrifugation. All of the toxins described in this article were isolated from crude extracts by gel and ion-exchange chromatography followed by reversed-phase high-pressure liquid chromatography (RP-HPLC), usually with C18 bonded phases. A simple, practical bioassay is needed for identification and quantitation of toxic fractions at each step. Very small amounts of most sea anemone sodium-channel toxins potently paralyze crustaceans (crabs and freshwater crayfish). Many labs including ours have relied upon this simple bioassay method, not only because it does not require expensive equipment and reagents, but also because whole animal assays are more likely to respond to a wide variety of toxins acting on different receptors. Recognition of the K-channel toxins has been more problematic, since peripherally injected crustaceans and mice are generally rather insensitive to these toxins (Kem, W. R., unpublished results). However, when injected directly into the ventricular space of the mouse brain, these toxins act in very minute amounts. Karlsson’s lab and subsequently others have utilized a rat brain membrane-binding assay that measures displacement of a radioiodinated mamba snake toxin (either dendrotoxin I [DTX] from Dendroaspis polylepis or α-dendrotoxin from D. augusticeps) to detect the toxins in chromatographic fractions (17,18). Since dendrotoxin binding is very sensitive to the presence of salts, especially sodium and potassium, which are particularly abundant in tissue extracts, one must measure dendrotoxin binding under the appropriate conditions (less than 200 mM monovalent salts). A distinct limitation of most receptor assays is that they will only detect toxins that bind to the same receptor site as the radioligand, and will thus fail to detect new toxins. Physiological assays are very useful for discovering new types of ion-channel toxins but are generally not as practical as the other types. In view of the aforementioned difficulties in isolating some of these toxins from crude extracts, an attractive alternative approach for detecting homologous toxins

536

Kem

would be to use reverse transcriptase-polymerase chain reaction (RT-PCR) methods to obtain toxin cDNAs from anemone mRNA extracts. Spagnuolo et al. (19) cloned the cDNAs for two Calliactis toxins. A Singapore lab cloned the genomic DNA sequence of the Pacific stichodactylid toxin HmK (20). The gene for this toxin contains two exons, one for the signal sequence and a second exon providing the sequence of the active toxin. This DNA sequence data should be useful for the preparation of primers for RT-PCR experiments aimed at isolating homologous toxin cDNAs from other species of anemones. Once the amino acid sequence of a toxin is established, the most efficient means of obtaining milligram quantities for pharmacological studies will be solid-phase synthesis or recombinant DNA-protein expression. Both ShK and BgK have been produced in excellent yield using solid-phase coupling methods (21,22). The deprotected peptides fold readily without special conditions and elute as sharp, baseline-resolved peaks during HPLC purification. A cDNA for Heteractis macrodactylus (HmK) toxin was expressed; the product possessed equivalent activity to the natural product (23). Similarly, Kelso and Blumenthal (24) obtained the cDNA sequences (Table 1) of several minor variants of sodium-channel toxins from Anthopleura xanthogrammica, then expressed them to allow pharmacological comparisons. Both strategies for producing peptide toxins have been successful for the anthozoan toxins discussed in this chapter. A distinct advantage of the solid-phase approach for structure-activity studies is that a wide variety of unnatural amino acids can be introduced. A disadvantage is that the overall yield decreases precipitously with increasing molecular size. Thus, the yield for a 50-mer will be much less than for a 35-mer, sometimes being insufficient for nuclear magnetic resonance (NMR) analyses of tertiary structure. Clearly, the recombinant approach is the only feasible means of making toxins of larger molecular size. 2. SEA ANEMONE SODIUM-CHANNEL TOXINS 2.1. Structures Most sea anemone sodium channel toxins have been found to contain 46–49 residues, although three shorter (27–31 residues) toxins have also been found. Thus far, at least 34 homologous “long” toxins have been completely sequenced, either as peptides or as cDNAs (Table 1). The identification of the three conserved disulfide bonds was reported by Wunderer (25). Most toxins, formerly called type 1 (14), were obtained from species belonging to two families: the Actiniidae and the Stichodactylidae. The former contains more species than any other sea anemone family, while the latter is a small family of largely Pacific species (only one, Stichodactyla helianthus, occurs in the Atlantic). Because toxin sequences from other sea anemone families are now known to be similarly distinctive, it seems preferable to classify these toxins according to their family of origin rather than to further elaborate an arbitrary numbering system. At the present time only a few sequences are available for homologous toxins from other less common families, but this is likely to change in the next few years as more sequences are reported. An incomplete sequence has been reported for Bolocera tuediae toxin II, representing the family Boloceroidae (26). The peptide sequences of Calliactis parasitica toxins I and II (19,27) are presently the only toxin representatives for a large family (Hormathiidae) of acontiate anemones. Sequencing the cDNA of the first toxin

Long Toxins Actiniidae: AequI Aery 1 Aele 1–1 Aele 1–2 Aele 2–1 Afus I Afus II Asul Ia Asul Ib Asul II Asul V Axan I Axan II Axan 1–2 Axan 2–1 Axan 2–5 Axan 2–10 Axan 3–3 Axan 3–6 Axan 3–7 Bcai III Bgra II Bgra III Cpas II

GIPCLC VSDGPSTRGNKLSGTIWMTGGYGGNG GAPCLC ANSGPNTRGNDLNGI VWVFG GIACLC DSDGPSVRGNTLSGT YWLAG GVPCLC DSDGPNVRGNTLSGT YWLAG GVPCLC DSDGPSVRGNTLSGI LWLAG GVACLC DSDGPNVRGNTLSGT IWLAG GGVPCLC DSDGPSVRGNTLSGI IWLAG GAACLC KSDGPNTRGNSMSGTTIWVFG GAPCLC KSDGPNTRGNSMSGTTIWVFG GVPCLC DSDGPSVRGNTLSGII WLAG GVPCLC DSDGPSVRGNTLSGIL WLAG GVSCLC DSDGPSVRGNTLSGTL WLYPSG GVPCLC DSDGPRPRGNTLSGIL WFYPSG GVPCLC DSDGPNVRGNTLSGTY WLAG GAACFC DSDGPSVSGNTLSGIL WLAG GVACFC DSDGPSVSGNTLSGIL WLAG GVPCLC DSDGPSVRGNSLSGII WLFG GVSCLC DSDGPSVRGNTLSGTL WLYPSG GVSCLC DSDGPSVSGNTLSGII WLAG GVSCLC DSDGPSVRGNTLSGIL WFYPSG GVACRC DSDGPTSRGNTLTGTL WLLTGG GASCRC KSDGPTSRGNTLTGTL WLIGR GASCRC KSDGPTSRGDTLTGTL WLIGR GVHCRC DSDGPSVHGNTLSGTV WVGS

CPKG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CPSG CASG

WHFCGKSRGLLSDCCKQ WHDCHGRAMV GYCCQED WHNCKSSGQLIGACCKQ WHNCKSSGPLIGACCKQ WHNCKAHGPTIGWCCKQ WHNCKAHGPTIGWCCKQ WHNCKAHGPTIGWCCKQ WNNCEGRA IIGTCCKQ WNNCEGRA IIGTCCKQ WHNCKKHGPTIGWCCKQ WHNCKKHKPTIGWCCK WHNCKAHGPTIGWCCKQ WHNCKAHGPNIGWCCKK WHNCKSSGPNIGWCCKK WHNCKAHGPTIGWCCKK WHNCKAHGPTIGWCCKK WHNCRDHGPTIGWCCKK WHNCKAHGPTIGWCCKQ WHNCKAHGPNIGWCCKK WHNCKAHGPTIGWCCKQ WHNCRGSGPFIGYCCKK WHNCRGSGPFIGYCCKQ WHNCRGSGPFIGYCCKQ WHKCNDEYNIAYECCKE


Table 1 Sea Anemone “Long” and “Short” Sodium Channel Neurotoxin Sequencesa

537

538

Table 1 Continued Stichodactylidae: Hmac I Hmac II Hmac III Hmac IV Hmac V Hmag II Hmag III Shel I Hormathidae: Cpar I Cpar II Halicuridae: Hsp?I Boloceroidae: Btue I

ASCKC DDDGPDVRSATFTGTVDFAY GTCKC DDDGPDVRTATFTGSIEFAN GNCKC DDEGPYVRTAPLTGYVDLGY GNCKC DDEGPNVRTAPLTGYVDLGY GNCKC DDEGPNVRTAPLTGYVDLGY ASCKC DDDGPDVRSATFTGTVDFWN GNCKC DDEGPNVRTAPLTGYVDLGY AACKC DDEGPDIRTAPLTGTVDLGS ECKC EGDAPDDLSHMTGTVYFS ECKC KGDAPDDLSHMTGTVYFS

CNAG CNES CNEG CNEG CNEG CNEG CNEG CNAG

WEKCLAVYTPVASCCRKKK WEKCLAVYTPVASCCRKKK WEKCASYYSPIAECCRKKK WEKCASYYSPIAECCRKKK WDKCASYYSPIAECCRKK WEKCTAVYTPVASCCRKKK WEKCASYYSPIAECCRKKK WEKCAS YYTIIDCCRKKK

CKGGDGS WSKCNT YTAVADCCHQA CKGGDGS WSKCNT YTAVADCCHQA

VACRC ESDGPDVRSATFTGTVDLWN GIPCWC GDPRQTELDGTTFFRD

CNTG

WHKCIATYTAVASCCKKD

CNSYTGGKWKSTKWINAISDK—Short Toxins

Actiniidae: Asul III Asul IV Equa I Equa I

RSCCP RSCCP AGGKSTCCP AGGKSTCCP

CYWG CYWG CAM CAM

GCPWGQN GCPWGQN CKYAG CKYAG

CYPE CYPE CPWG QCAHHC CPWG QCAHHC

GCSGPKV GCSGP GCS GCS

aThe

Kem

sequences were aligned manually to account for the different overall sequence and loop lengths of the toxins. The toxins are designated by genus (capital letter) and the first three letters of the species name (now required to eliminate ambiguity of names formerly referred to with a two-letter designation). Only sequence papers that were not cited in Norton’s review (37) are cited here: Actinia equina, Lin et al. (31); Anemonia erythraea, Shiomi et al. (32); Anthopleura elegentissima, Beress et al. (13); Malpezzi et al. (15); Bunodosoma granulifera, Loret et al. (33); Calliactis parasitica, modified by Spagnuolo et al. (19); Condylactis passifloria, Shiomi et al. (29); Halicurias sp., Ishida et al. (28).


539

detected an error in the original protein sequence, which has been corrected in Table 1. The sequence of Halicura toxin (Hsp I) shown at the bottom of Table 1, from yet another anemone family, seems intermediate between those of the actiniid and stichodactylid toxins (28). A general caveat must be mentioned for data based totally on cDNA sequences: post-translational modifications can occur in sea anemone toxins. Hydroxylation at certain proline residues is particularly common (29,30). Several major differences in sequence are apparent between the actiniid and stichodactylid toxins. First, the stichodactylid toxins possess only two residues at the N-terminus before the first half-cystine, in contrast with the actiniid toxins, which possess three residues before the first half-cystine. Second, the stichodactylid toxins also have an acidic triad of amino acids occurring at positions 6–8. Thirdly, they possess only one invariant aromatic residue, at position 30. Fourth, their C-termini generally possess four consecutive basic amino acids. Only about 25% of the sequence is identical in the two types of sea anemone toxin, when the six common half-cystines are included in the comparison (14). Within each group of toxins there is a much higher degree of sequence homology. In a few instances, isotoxins from the same species may differ by only a single residue, which permits assessment of its possible influence upon toxicity. For example, two H. macrodactylus toxin variants differ only at position 11 (Tyr in toxin variant III, Asn in IV); since their mammalian toxicities are very similar, one can infer that these two residues are roughly equivalent in their contribution to toxin activity, or that this position contributes minimally to the toxin-receptor interaction. Usually there are several differences between toxin variants, so one cannot unequivocally assess the influence of a single side chain upon activity by comparison. Odinokov et al. (35) reported the hydrophobicity profiles of many “long” sea anemone toxins known at that time. A major difference between the actiniid and stichodactylid types of toxins occurs within the 18–24 residue region, which is much less polar in the stichodactylid toxins. This stretch of hydrophobic sequence in the stichodacylid toxins may provide a hydrophobic interior that compensates for the absence of a tryptophan residue, which is always present at position 23 in the actiniid toxins (Table 1). Norton’s group has reported a detailed tertiary structure of anthopleurin-A, which indicates that Trp23 and Trp30 side chains together form a hydrophobic core in this type 1 toxin (36). Like the scorpion α-toxins, the sea anemone long toxins possess a scaffold composed of four anti-parallel B-pleated sheets (see Norton’s review, ref. 37). This endows the molecules with considerable structural stability which is further enhanced by the presence of the three disulfide bridges. None of the toxins contain α-helical segments. Solution structures have also been published for several long toxins (36,38–41). The structure with the highest resolution was the stichodactylid toxin Shel I. This peptide, unlike AP-A and Asul II, displayed conformational homogeneity during two dimensional NMR spectroscopy. Actually, the tertiary structure of Shel I was found to more similar to that of the actiniid toxin Asul Ia than to the Heteractis (stichodactylid) toxins whose sequences are more similar to that of Shel I. Refinement of the ShI structure revealed that acidic residues 6–8 constitute a type 1 B-turn. (Fig. 3) In proteins, most peptide bonds where Pro contributes the amino function are of the trans-configuration. However, Scanlon and Norton (42) investigated the conformational status of proline residues in AP-A, and found that the Gly 40-Pro 41 peptide bond is

540

Kem

Fig. 3. NMR-derived folded structure of Stichodactyla helianthus toxin. Only the peptide “back-bone” is shown. The top and bottom halves of the figures show the toxin from different orientations. The eight best computer-derived structures are superimposed on each other. While the lower portion of each composite figure shows good superimposition, the top right-hand side of each structure shows poor superimposition, because this loop of the toxin is flexible. Adapted with permission from ref. 40.

predominantly in the cis conformation, the trans-conformer only being about half as abundant. The free-energy change associated with the cis-trans isomerization in this peptide bond is quite large (>78 kJ/mol). Two additional conformations of this toxin, though minor, were also detected. It seems quite possible that the predominant cis and trans forms of AP-A may have different affinities and efficacies for their sodium channel target. Pallaghy et al. (36) reported the most detailed tertiary structure for AP-A. The Asp 7 carboxylate lies close to the Lys 37 amino group. The side chains of these two residues, plus those of nearby residues Asp 9 and His 39, were suggested to constitute part of the AP-A pharmacophore. Only three “short” toxins have been reported (Table 1). The NMR structure of Anemonia sulcata toxin III failed to reveal any regular helical or B-pleated sheet secondary structures, but B-turns were found (43). These toxins bear no clear sequence homology with the long toxins. Since immunological comparisons of toxins can pro-


541

Fig. 4. Effect of Stichodactyla helianthus toxin I on sodium currents in the voltage-clamped crayfish giant axon. (A) Sodium currents at different test-pulse potentials. (B) Peak Na currentvoltage relations for the same axon. (C) Steady-state Na channel inactivation. The graph shows normalized peak current for test pulses to –10 mV following 30 ms conditioning pulses to the indicated potentials, with a 150 µs recovery interval between the conditioning and test pulses. Adapted with permission from ref. 49.

vide a measure of relatedness, it is of interest that polyclonal antibodies (PAbs) for Asul III crossreacted with the long toxin Asul I (44). This suggests that this short toxin shares some common surface structure, even though its sequence is very different. The short toxins have not been extensively investigated, probably because their affinities for vertebrate and insect sodium channels are quite weak (45). Schweitz et al. (46) reported that Anemonia and Heteractis toxins were immunologically distinct. Using the Ochterlony radial-diffusion technique, polyclonal rabbit antibodies prepared with As II as antigen failed to precipitate Heteractis magnifica toxins. We have confirmed this immunological distinctiveness of the actiniid and stichodactylid toxins using polyclonal antibodies prepared with different toxins. Rabbit PAbs directed towards ShI failed to react with Asul I, Asul II, or Btue II. However, Btue II did react with the AP-A antibody preparation, which suggests that this toxin is more closely related to the actiniid rather than stichodactylid toxins (Kem, W. R., unpublished results).

542

Kem

Fig. 5. Effect of holding potential (Vhold) on the rate of action of Stichodactyla helianthus toxin I on sodium channel inactivation. The same ShK concentration was applied to each of three crayfish (Procambarus clarkii) voltage-clamped median giant axons under identical conditions except that the membrane was clamped at different holding potentials for the initial 10-min period after the toxin was added. Slowing of the inactivation rate was measured as the amplitude of the toxin-induced, slow-inactivating current. The unclamped resting potential was –90 mV. The toxin acted quickly when the holding potential was hyperpolarizing (–110 mV), less rapidly when slightly depolarizing (–75 mV) and most slowly when a strong depolarizing potential (–65 mV) was used. The maximum effect was not greatly affected by holding potential because the toxin effect was essentially irreversible. Adapted with permission from ref. 49.

2.2. Physiological Actions Initial studies of the mechanisms of action of these toxins on neurons utilized voltage-clamp methods. Condylactis toxin (actually a crude extract containing several isotoxins) action was studied with crayfish axons (47), while the effect of Asul II was measured on frog-myelinated neurons (47). It was shown that the major action of these toxins, regardless of axon preparation, was to drastically slow the rate of inactivation of the sodium current, a process that is essential for normal recovery of membrane excitability. Some increase in peak sodium current was also observed, which resulted secondarily from a lesser state of inactivation when the peak current was attained (Fig. 4). The rate at which inactivation was affected strongly depended on membraneholding potential (Fig. 5). Since maintained depolarization greatly slows the onset of action, this indicates that the resting state of the channel displays a much higher affinity for the sea anemone toxin than does the inactivated state (Fig. 1). Some decrease in the slope of the steady-state inactivation (Hodkin-Huxley h-infinity parameter, now usually referred to as sodium channel “availability”)-membrane potential relationship, and even some increase in the steady-state h parameter at very positive internal potentials has been observed, as with the scorpion α-toxins. Current-clamped axons devel-


543

oped trains of action potentials resulting from early after-potentials which develop as the membrane begins to repolarize. Considerable macroscopic electrophysiological data on long toxins isolated from different sea anemone families indicates that the sea anemone long toxins from different families have essentially identical actions (49). Single-channel analyses of actiniid toxin actions indicate that the toxins only inhibit channel inactivation from open-channel states (50,51); inactivation from resting or closed states (Fig. 1) is not significantly affected. The peak sodium channel conductance of sea anemone-treated cells decreases when voltage-clamping pulses are applied at frequencies that are too high to allow the inactivated sodium channels to be recycled back to their resting state before the next clamping pulse (52). This predicts that excitable cells with sodium channels may become relatively quiescent after a peak effect of the sea anemone toxin, as a result of this frequency-dependent block of many sodium channels. One of the most interesting recent discoveries has been identification of an external region of the sodium channel that influences the fast inactivation process, in addition to the well-known intracellular loop between domains 3 and 4 (53). Actually, the sea anemone toxin Asul II was an extremely useful tool in demonstrating how steady-state activation of only a small number (1 µM. On the basis of this data, it was concluded that the stichodactylid toxins bind to the same Site 3 as the scorpion α-toxin, while the actiniid toxins bind to a separate site on the sodium channel (46). Since binding of a ß-scorpion toxin was not affected by the H. magnifica toxins, it was inferred that this other binding site must be located somewhere else on the sodium channel other than at Site 4 (46). We investigated the binding of toxins from the stichodactylid and actiniid sea anemone families to rat brain and crustacean peripheral-nerve preparations, using direct as well as indirect radioligand-binding assays. We found that both types of toxin (Asul II and Hmac III) displaying high mammalian toxicity were able to displace the specific binding of iodinated α-scorpion (Aa I) toxin in rat brain membranes (62), as shown in Fig. 6. The largely crustacean-specific stichodactylid toxin Shel I did not bind to the


545

rat brain sodium channels except at concentrations higher than are shown in Fig. 6, but it did bind avidly to crab sodium channels. However, the specific binding of labeled Asul II to the crab nerve channels was not displaced by Sh I (Kem, W. R., and Pennington, M. W., unpublished results). Thus, the receptor-binding sites for these two sea anemone toxins are apparently not identical, although they may partially overlap each other. It is difficult to compare the affinities of varioius toxins when they have been measured using different techniques. A given method may preferentially measure affinity when the channels are in a particular state. For instance, the electrophysiologically measured affinities of anthopleurins A and B for the cardiac sodium channel, recently reported to be about 3 nM and 0.1 nM, respectively, were much lower than the respective estimates of 14 and 9 nM, based on 22Na influx assays. Apparently, the closed form of the channel, the most likely state to be measured electrophysiologically, has a much higher affinity for the toxins than the open state, which would be measured in the flux assay (63). 2.4. Structure-Activity Studies: Identification of the Toxin Molecular Surface That Interacts with the Sodium Channel Chemical modification studies in the late 1970s provided some tantalizing clues as to which residues in the actiniid toxins were important for activity, but were limited by an inability to direct the modifications to particular residues containing a reactive side chain (64). Chemical-modification studies on sea anemone toxins were carried out before the diversity of sodium channels became apparent. Thus, from a present perspective, much of the chemical-modification data based on toxicity estimates obtained with whole organisms containing numerous types of sodium channels is now of only limited utility. Nevertheless, this approach still offers a unique opportunity to obtain properly disulfide-paired toxin analogs when solid-phase synthesis and recombinant methods fail to provide a suitably folded peptide. Also, chemical modification provides a means of inserting a considerable variety of unnatural side-chain substituents into a peptide. Therefore its potential should not be ignored when planning structure-activity studies. During the past decade, solid-phase chemical synthesis techniques have succeeded in providing a variety of natural and unnatural sea anemone toxin sequences possessing full biological activity. Pennington et al. (65) first reported a successful solid-phase synthesis of Shel I. Six monosubstituted analogs of the toxin were then synthesized and tested (66). When the Asp and Glu residues constituting the acidic triad at positions 6–8 were individually substituted with Asn and Gln, respectively, the resulting toxin analogs were found to be several orders of magnitude less active than the native toxin. Replacing Asp at position 11 with Asn also reduced crab toxicity and nerve binding affinity significantly, as did acetylation of the amino group at position 4. All of these analogs displayed sharp HPLC peaks and circular dichroism (CD) spectra that were practically identical with that of the native toxin, with the exception of Asn 6-Sh I. The CD method has a relatively limited utility for assessing correct folding of sea anemone sodium-channel toxins, since the spectra of the native toxins display only a small ellipticity shoulder around 217 nM, which is diagnostic of B-pleated sheet residues. NMR analysis is clearly the method of choice for determining whether a synthetic toxin has folded properly, but in these initial synthetic studies rather low yields were experi-

546

Kem

enced when using a Boc-amino acid coupling strategy, so insufficient quantities of each analog were available for NMR analysis. With this caveat in mind, five of the six analogs seemed by CD to fold in a manner similar to the native toxin. The toxicity and binding data thus led us to conclude that acidic residues 6–8 are important for activity, and may also affect the folding of the large loop. The lesser reductions in activity associated with the replacement of the ionized side chains at positions 4 and 11 are also consistent with this inference. The Elyakov group in Vladivostok had previously inferred from carboxyl modification of H. macrodactylus III that these side chains are important for activity (67). Mutants of AP-B showed that mammalian activity was reduced but not eliminated by substituting Asn or Ala for Asp 9 of this toxin; it was concluded that a hydrogen-bonding rather than negatively charged group was essential at this position for both folding and interaction with the sodium channel. An AP-B mutant containing an Asn instead of Asp at position 7 was not expressed, suggesting that this sequence would not fold properly (68). Synthesis of Axan I (AP-A) using a Boc-amino acid coupling strategy turned out to be a major challenge because of the difficulties in folding the product. Pennington et al. (69) succeeded in synthesizing a few milligrams of the native toxin that were indistinguishable from the natural toxin by many criteria, including two-dimensional proton magnetic-resonance spectroscopy and cardiac inotropic bioassays. Several analogs were also synthesized, but in very low yields, which did not allow further characterization. Blumenthal and coworkers have prepared and tested a variety of mono-substituted sea anemone toxin analogs, utilizing a synthetic gene for AP-B. After glutathionecontrolled oxidative folding of the fusion protein, the toxic portion was proteolytically cleaved from bacteriophage protein with staphylococcal protease. Generally, about 1 mg of toxin was obtained from a liter of E. coli fermentation broth after proteolytic processing and a single HPLC purification step (70). The native AP-B product was properly folded according to its CD spectrum and displayed the same activity as the natural toxin. An additional analog containing a Gly-Arg extension at the N-terminus was also obtained from the fermentation broth and found to have CD and sodiumchannel stimulatory properties indistinguishable from the native toxin. Their conclusion that the bulk of the N-terminus is not critical for mammalian-nerve sodium channel activity is consistent with previously published data for chemically modified Asul II. The positively charged side-chain residues Arg 12, Arg 14, Lys 48, and Lys 49 of AP-B were individually replaced with other natural amino acid residues to assess their importance. When Arg 12 was replaced with either a Ser or Lys residue, no significant diminution occurred in the responsiveness of cultured cells expressing either neuronalor myocardial-type sodium channels. In contrast, toxin activity was greatly reduced when an Ala replaced Arg 12. The authors concluded that a cationic residue is not necessary at position 12, just a polar one. Replacement of Arg 14 with a Gln or Lys residue only reduced activity about twofold, while replacement with Ala diminished activity by about fourfold. Thus, neither Arg 12 nor Arg 14 are absolutely required, when tested individually (71–73). Before discounting any significant contribution of the large-loop Arg residues to the sodium channel-receptor interaction, one must test whether a mutant lacking both of these residues possesses activity in order to eliminate the distinct possibility that only a


547

single cationic Arg in the large loop is important for toxin activity. Loret et al. (33) have suggested that in both sea anemone and scorpion toxins, Arg side chains may be dominant contributors to the surface electrostatic potential. Thus, it is possible that the position of the Arg is not so critical as long as it occupies the same general region of the toxin surface. In the large loop of Bolocera toxin II (Table 1), which also possesses strong myocardial potency, there are two Arg residues at positions 10 and 21. Again, the exact position of the Arg in this flexible loop may not be so critical. In at least two crustacean-active toxins, Condylactis gigantea II (30) and Calliactis parasitica I (27), the Arg in this loop is replaced by His, which may be nonionized at physiological pH; thus, toxin interaction with arthropod neuronal sodium channels seems less dependent upon the presence of Arg in the large loop. Substitution of Pro 13 in the large loop also was detrimental to activity (74). Clearly other toxin analogs with substitutions in this loop need to be prepared and investigated in order to obtain a better understanding of the contribution of the large loop to the toxin’s interaction with sodium channels. For instance, substitution of the hydrophobic residue of Leu 18 in AP-B greatly reduces activity (75). Perhaps it would also be possible to prepare loop-constrained analogs by crosslinking of reactive side chains; this may shed some light on the possible importance of loop flexibility to toxin action. Some loops in other proteins have a characteristic shape similar to the Greek letter, omega (76). In some cases these loop structures actually move laterally in a hinge-like fashion when the protein is activated by some ligand. For the sake of stimulating experimentation, it can be suggested that the sea anemone long toxin loop may also move in a hinge-like fashion during receptor binding, providing extra stabilization of a noninactivated state of the channel. Benzinger et al. (77) have provided new insights into the pharmacological differences between AP-A and AP-B. Using what might be called the “mutual mutant” method, evidence for the close interaction between a single residue (Lys 37) in the toxin and a side chain (Asp 1612) in the fourth domain of the cardiac sodium channel α subunit was obtained. This approach is clearly becoming indispensable for demonstrating the occurrence of specific interactions between side chains on ligand and receptor, and will be discussed again when we consider the binding of potassium channel toxins to their receptor site. Efforts at truncating the size of a long toxin, AP-A, to obtain smaller, simpler active peptides have only met with marginal success, because these simpler sequences probably do not adopt a similar folded structure (78). The various naturally occurring toxin variants are mostly examples of multiple substitution. Quantitative comparison of the activities of these various sea anemone toxins using multivariate statistical methods could potentially provide further insight into the structure–activity relationship, and also help to identify the interactive surface or pharmacophore on the toxin (30). Loret et al. (33) reported modeling Bunodosoma granulifera toxin II (Bgra II) by surperimposing the toxin backbone upon the published Asul II backbone coordinates. Computation of the surface electrostatic potential of Bgra II revealed two regions of high electropositivity, which were postulated to be important for interaction of this toxin with Site 3. The surface of Bgra II was also compared with that of the most extensively studied scorpion α-toxin, Androctonus australis toxin II. Certain positively charged side chains on both toxins were postulated to interact with common sites on

548

Kem

Fig. 7. Relationship between cardiac Na channel actions of sea anemone toxins and their ability to displace α-scorpion toxin (125I-Leiurus quinquestriatus quinquestriatus toxin V) binding to rat brain neurosomes. Electrophysiological data (increase of sodium current integral, which largely reflects delay of inactivation) were obtained from Xenopus oocytes expressing guinea pig myocardial sodium channels. Displacement of radioiodinated Lqq-V toxin (0.2 nM) by 1 µM sea anemone polypeptide was measured under conditions where the neurosomes possess a resting potential. Nonspecific binding was measured in the presence of 2 µM unlabeled scorpion toxin. Adapted with permission from ref. 94.

the sodium channel. Further comparisons of scorpion α-toxin tertiary structures with those of sea anemones long toxins may reveal other possible surface similarities important for binding to sodium channels, since they bind to overlapping sites. One of the most interesting facets of certain sea anemone toxins is their relatively high affinity for cardiac vs skeletal muscle sodium channels (Fig. 7). This property was originally thought to be largely confined to the actiniid-type toxins (79). However, in a limited study of nine sea anemone toxins comparing their relative abilities to inhibit the binding of Leiurus scorpion toxin to rat brain membranes and to prolong the inactivation of guinea pig myocardial sodium channels, it was found that two of the stichodactylid toxins showed an even higher selectivity for the myocardial type sodium channel (62). Clearly more data on a greater variety of natural as well as synthetic sea anemone toxins is desirable to better define the structural features that confer selective affinity for myocardial vs neuronal channels (81). 2.5. Applications The primary utility of the sea anemone sodium-channel toxins will probably be as research tools for investigating the contribution of the external surface of the channel in activation–inactivation coupling. With the renewed interest in the peripheral sodium channels (3) as novel targets for developing analgesic drugs to treat neuropathic pain disorders, it is also possible that some sea anemone toxin variants will be useful ligands for investigating these sodium channels. The greater potencies of some sea anemone long toxins for the myocardial sodium channel allows augmentation of the force of contraction of the mammalian heart at concentrations that do not seem to affect myocardial and other motor neurons


549

(8,59,80,81). This selectivity has not been reported for the scorpion toxins. While the actual mechanism of the increased inotropy is not completely settled, it is highly likely that the enhanced sodium-ion influx due to delayed inactivation causes an influx of calcium ions which, further stimulates contractile proteins. It has been demonstrated that the therapeutic index of AP-A is better than that of digitalis-type inotropes. It has been shown that these peptides actually enhance performance of the failing heart even at its most advanced stages of failure. A major problem with the use of sea anemone peptides as inotropes is their high propensity to cause delayed after-potentials, which lead to life-threatening arrhythmias. In the past decade, it has become increasingly clear that sodium and consequently calcium loading of the myocardium is life-threatening, and increased contractility should be attained by other mechanisms, perhaps by making the contractile machinery more sensitive to intracellular calcium. It seems that the only possible way in which the sea anemone peptides might again be realistic models for inotropic-drug development would be to modify (greatly reduce) their propensity to severely inhibit sodium inactivation. Theoretically, a substance that enhances sodium influx during the initial phase of the myocardial action potential, without prolonging the time course of this influx, would be expected to exert an inotropic action with less arrhythmogenic activity. The observations of the Catterall and Blumenthal labs that some sea anemone toxin variants have a lower efficacy of action on sodium channels in vitro suggests that it may be possible to design peptide analogs that display a lower propensity for causing arrhythmias. 3. CORAL SODIUM-CHANNEL NEUROTOXINS 3.1. Isolation of Goniopora Toxin (GTX) Goniopora toxin (GTX) was the first polypeptide toxin to be isolated from the hexacorals, which are reef-building anthozoans. This large group has received much less attention from toxinologists than the octacorals (soft corals), which are well known for their diterpene chemical-defense compounds. Species belonging to the genus Goniopora possess relatively long (2–6 cm) polyps that are often active during the daylight hours (unlike most corals) and can thus be readily cut free of the calcareous base of the coral for toxin extraction. GTX was discovered while screening numerous coral species for lipid-soluble ciguatoxins and palytoxins. Its purification has been reported (82). The toxin consists of a polypeptide chain of 88 amino acid residues, including 10 half-cystines, whose sequence apparently does not resemble those of other known polypeptide toxins affecting Na channels (83). Unfortunately, the structure of this interesting toxin, alluded to in the Ashida et al. abstract, still has not been published. 3.2. Physiological Actions Fujiwara et al. (84) initially investigated the action of this polypeptide upon isolated rabbit atrial myocardial strips. At 3 nM, GTX produced a significant increase in myocardial contractility, which was related to an increased action potential duration. While these two actions were irreversible, TTX was able to reversibly counteract these actions. At concentrations above 30 nM, GTX generated arrhythmias. Thus, GTX inhibits Nachannel inactivation in a manner that is similar to the Site 3 sea anemone and scorpion

550

Kem

α-toxins. Since the toxin’s effects upon myocardial contractility and electrogenesis were not inhibited by 1 µM propranolol, it must act directly upon the myocardial membrane rather than upon the autonomic nerves innervating the heart. The actions of the toxin upon were also investigated. Prolongation of the action potential and enhancement of heart contractility was closely related (85). While these studies provided considerable insight regarding the influence of action potential duration upon GTX enhanced contractility, two voltage-clamp investigations of GTX action unequivocally demonstrated that prolongation of the action potential is due to an effect of the toxin specifically upon Na-channel inactivation (86,87). In addition to this action, it was also found that the toxin depolarized the resting crayfish axon membrane in a manner unlikely to be related to its Na-channel effect, because the resting depolarization was unaffected by TTX or replacement of external Na+ with Tris ion. Concomitant with this depolarizing action, an increase in resting (leak) conductance of the membrane was observed. Neither the AP nor the resting potential effects of the toxin could be reversed by prolonged washing of the preparation with saline. In the cardiac studies, no irreversible effects upon the resting membrane conductance were mentioned. 3.3. Pharmacology and Receptor Binding An investigation (88) of GTX action, using cultured neuroblastoma cells, revealed three additional properties of this toxin that are noteworthy. First, the ability of the toxin to affect Na channel inactivation displayed a voltage-dependence similar to some other Site 3 toxins. However, this voltage-dependence was less steep than for Leiurus toxin binding. Second, GTX required extracellular Na or some other alkali metal cation for its action. The cation dependence of GTX action was quite remarkable. The toxin was not active in the absence of (Na+)o or similar cations, which is a relatively unique phenomenon for polypeptide toxins, although TTX binding shows a similar displacement with monovalent cations. Since the order of cation dependence was not the same as the selectivity sequence for channel conductance, this site must be separate from the ionic selectivity or filter site for the channel. It would be interesting to determine if GTX binding displays a cation dependence identical with that for its action. Third, GTX action apparently was mediated through some site other than Site 3, since much higher (at least 10×) concentrations of the toxin were required to inhibit [125I]Leiurus toxin V binding than were necessary to affect Na-channel inactivation. 3.4. A Goniopora Species Calcium Channel Toxin Qar et al. (89) isolated a polypeptide toxin that acts as a Ca-channel activator from a Goniopora species collected from the Red Sea. Its molecular size was estimated as 19,000 Daltons. These initial studies on Goniopora suggest that further investigations of coral toxins may provide other toxins as well. Since neither Goniopora species was identified with certainty, this seems like an appropriate place in this review article to stress the importance of proper animal identification (and collection of type specimens for future examination) in toxin research. Otherwise, investigators have difficulty in obtaining further samples of the reported toxins.


551

Fig. 8. Ribbon diagram of ShK and Charybdotoxin. ShK lacks ß-sheet secondary structure, while charybdotoxin and other related scorpion toxins have ß-sheet at both ends of the molecule. The molecular scaffolds for the K channel-binding surfaces of each type of toxin are different: α-helix for ShK and ß-sheet for ChTx. Adapted with permission from ref. 94.

4. POTASSIUM-CHANNEL TOXINS FROM SEA ANEMONES 4.1. Structures The first two of these peptide toxins were isolated from the Caribbean anemones, Bunodosoma granulifera (17) and Stichodactyla helianthus (18) by Cuban scientists working with Evert Karlsson at the University of Uppsala in Sweden. Later, chemical synthesis of these toxins permitted extensive investigations of their chemical and pharmacological properties (21,22). Subsequently, some homologous peptides have also been detected and isolated from other sea anemone species (90,91). The amino acid sequences of the so-called short K channel toxins (Table 2) revealed some conserved residues in addition to the six half-cystines involved in disulfide bond formation. Using the residue numbering for ShK, these included Asp 5, Ser 20, Lys 22, Tyr 23, Lys 30, and Thr 31. These residues were considered likely to be important for proper folding and/or receptor binding, and the effects of their replacement with other amino acids will be discussed below. The sequences of the two stichodactylid toxins, ShK and HmK, were particularly similar to each other, as were the sequences of the two actiniid toxins, AsK and BgK. Thus, we again find significant differences between toxins from these two sea anemone families, especially in the presence of a larger loop in the actiniid K toxins. However, the six half-cystines that provide the three disulfide bridges are absolutely conserved. The disulfide linkages in three of the short toxins were assigned and shown to be equivalent (22,23,92). The disulfide pairing pattern of the anemone toxins differs from that of the scorpion toxins. Comparison of ShK and BgK with charybdotoxin indicated that no homology exists between anemone toxins and scorpion K-channel toxins (93,94). However, as will be discussed later, the scorpion and sea anemone toxins seem to bind in a very similar fashion within the outer vestibules of certain K channels. The first intimation that the folded structures of the anemone and scorpion K-toxin peptides were very different (Fig. 8) emanated from a combined circular dichroism and laser Raman spectroscopic investigation (93). The CD spectra of both ShK and BgK

552

Kem

Table 2 Sequences of Sea Anemone Neurotoxins That Block K Channelsa Short toxins Actiniidae: 1 BgraK AsulK AequK

10 20 30 VCRDWFKETACRHAKSLGNCRTSQKYRAN ACKDNFAAATCKHVKENKNC GSQKYATN GCKDNFSANTCKHVKAN NC GSQKYATN

Stichodactylidae: 1 10 HmagK RTCKDLIPVSEC ShelK RSCIDTIPKSRC

20 TDIRC TAFQC

37 CAKTCELC CAKTCGHC CAKTCGKC

30 35 RTSMKYRLNLCRKTCGSC KHSMKYRLSFCRKTCGTC

Intermediate-Size Toxins 1 BdS*I BdS*II

10 20 30 AAPCFCSGKPGRGDLWILRGTCPGGYGYTSNCYKWPNICCYPH AAPCFCPGKPDRGDLWILRGTCPGGYGYTSNCYKWPNICCYPH

40

Long Toxins 1

AsKC1 AsKC2 AsKC3 DTX

10 20 30 40 50 INKDCLLPMDVGRCRASHPRYYYNSSSKRCEKFIYGGCRGNANNFHTLEECEKVCGVR INKDCLLPMDVGRCRARHPRYYYNSSSKRCEKFIYGGCRGNANNFITKKECEKVCGVR INGDCELPKCCGRCRARFPRYYYNLSSRRCEKFIYGGCGGNANNFHTLEECEKVCGVRS QPLRKLCILHRNPGRCYQKIPAFYYNQKKKQCEGFTWSGCGGNSNRFKTIEECRRTCIRK

aThe toxins are identified by species as in Table 1, except for the two BdS toxins. These toxins, isolated from Anemonia sulcata by the Beress laboratory, are referred to by these initials because they are blood pressure-depressing substances. The papers from which the sequences were obtained are as follows: Actinia equina, Minagawa et al. (90); Anemonia sulcata, Schweitz et al. (97); Bunodosoma granulifera, Aneiros et al. (17), Cotton et al. (22); Heteractis magnifica, Gendeh et al. (23); Stichodactyla helianthus, Castaneda et al. (18).

(94) possess double minima at 213 and 220 nM typical of a predominantly helical peptide. A computer analysis of the ShK CD spectrum predicted 31% α-helix and 18% ß-sheet secondary structures, while analysis of the vibrational spectra predicted 34% helix and 30% sheet. In contrast, the scorpion K-channel toxin CD spectra display CD spectral peak near 217 nM, which is characteristic of peptides containing mostly ßsheet structure. Two-dimensional 1H NMR analysis of ShK unequivocally demonstrated that the toxin contained approx one-third α-helix but no ß-sheet (94). The family of ShK structures calculated from the NMR data was well defined, except at the N- and C-termini, which had few NOE restraints. Two α-helical segments were present, encompassing residues 14–19 and 21–24, the first helix being stabilized by a capping box (94a). The N-terminal region consisting of residues 1–8 had an extended conformation that was followed by a pair of interlocking B-turns, which could also be regarded as a 310 helix. Near the C-terminus the peptide backbone displayed several chain reversals, including a type 1 turn at residues 28–31. The homologous sea anemone toxin BgK was also


553

reported to contain two longer helices, involving residues 9–16 and 24–31 (95). Its overall topology was similar to that of ShK toxin. NMR analysis also corroborated the assignments of the three disulfide bonds of ShK toxin. While the 3–35 bridge was fully exposed on the surface of the molecule, the 17–32 bond was partially buried and the 12–28 bond almost fully buried within the core of the molecule. The relative exposures of the three disulfide bridges are in excellent agreement with the relative susceptibility to reduction and alkylation of the equivalent disulfides in the closely related HmK toxin (23). A search of the Protein Data Bank found no structural folds similar to ShK toxin. The solution structure of ShK showed that all of the positively charged groups were located on its surface, with Arg 1 and Arg 11 being the most exposed of the four arginines and Lys 9 and Lys 18 being the most exposed lysines. The epsilon-amino group of Lys 30 lies close to the carboxylate of Asp 5, consistent with the possibility of a salt bridge between these two groups. The effects of pH, temperature, and polypeptide concentration on the solution structure and side-chain interactions of ShK toxin were investigated (96). The toxin’s structure was stable even at a high temperature, showing little change even at 80°C. This stability allowed backbone amide temperature coefficients and solvent-exchange rates to be measured and correlated with hydrogen bonds observed in the toxin. Asp 5 (pKa 2.8) displayed an electrostatic interaction with Lys30, which may be partially responsible for the importance of these side chains in the folding of synthetic toxin. The phenolic pKa of Tyr23 in the native toxin was only 8.7, as a result of interactions with the positively charged sidechains of Arg 11 and to a lesser extent Lys 22. Indeed, several hydrogen bonds between the pharmacologically important Arg 11 guanidino and Tyr 23 phenolic groups were found in the solution structure. As will be described later, these three residues are implicated in the tight binding of ShK toxin to the T-lymphocyte Kv1.3 and IK(Ca) channels. Thus, their close intramolecular interactions should be taken into account in models of binding of this toxin to the pore and vestibule of these potassium channels. Schweitz et al. (97), using dendrotoxin as radioligand, found not only an ShK homolog in Anemonia sulcata, but also two longer K-channel toxins (Table 2) homologous with both dendrotoxin and bovine pancreatic trypsin inhibitor. Dendrotoxin had previously been shown to be homologous with the pancreatic trypsin inhibitor. This homology between K channel toxins from different phyla is almost unique for animal toxins, which usually occur only within a small group of closely related species. Apparently, the long sea anemone K channel toxins evolved from a duplicated gene for a serinetype protease inhbitor. More recently, it was demonstrated that two blood pressure-depressing substances (BdS) with amino acid sequence homology to the sea anemone sodium-channel long toxins are selective blockers of Kv3.4 type voltage-gated K channels (98). These peptides are intermediate in size between the short toxins and long toxins described above (Table 2). The BdS peptides were previously shown to inhibit the contractility of the isolated guinea pig atrium and to displace the specific binding of radiolabeled AP-A to rat brain sodium channels (99). It is not yet clear whether the sodium and/or potassium channel altering activities of these interesting peptides are responsible for their hypotensive effects.

554

Kem

Fig. 9. Inhibition of Charybdotoxin binding to Jurkat T-lymphocytes by ShK toxin. Percent inhibition of specifically bound [125I]-charybdotoxin is plotted as a function of the test concentration of Shel K toxin. Data are presented as means ± SEM (n = 4). A similar inhibition of whole cell potassium currents was observed in these cells. Adapted with permission from ref. 100.

4.2. Physiological Effects of the Anemone K-Channel Toxins ShK displayed a surprisingly low toxicity when administered intravenously, the median paralytic dose for Swiss-Webster mice being approx 25 mg/kg (101). Commonly observed effects were tremors, muscular fasciculations, and motor paralysis. The intra-cerebroventricular toxicity of ShK has not yet been assessed, but is predicted to be in the µg/kg range because the intracerebroventricular lethal dose for BgK was estimated as approx 4.5 µg/kg (22). Also, dendrotoxin acts at very low doses when administered within the brain, relative to when it is administered peripherally. A K-channel toxin would be expected to enhance excitability by causing a massive release of peripheral neurotransmitters secondary to increasing the duration of peripheral nerve terminal action potentials. Indeed, relatively high concentrations (>100 nM) of ShK have been found to prolong the repolarization phase of neuronal-action potentials recorded at chick neuromuscular terminals (17). Later, ShK (Fig. 9) was found to block the Kv1.3 channel of Jurkat cells at much lower (10–50 picomolar) concentrations (100). Binding of the toxin within the channel outer vestibule occurred at normal resting membrane potentials and thus does not seem to greatly depend on channel activation or inactivation (Fig. 10). Most electrophysiological studies of sea anemone K-channel toxins have utilized Xenopus oocytes or mammalian cells expressing homomeric K channels. Dauplais et al. (95) studied the channel-blocking actions of BgK upon mammalian cells expressing Kv1.1, 1.2, 1.3, or 3.1 channels. The toxin selectively inhibited the three Shaker-type Kv1 delayed-rectifier channels at similar concentrations (IC50 range, 6–15 nM). The rates of association and dissociation of BgK toxin to the different Kv1 channels as measured electrophysiologically were in agreement with the Kd values measured from the steady-state concentration-response curves. In contrast with the relatively similar


555

Fig. 10. ShK blockade of Kv1.3 current in Jurkat T-lymphocytes is not state-dependent. The voltage protocol is illustrated above the current traces. Holding potential was –80 mV and test potential was +30 mV. Current traces are shown immediately prior to exposure to ShK toxin (100 pM) and following 6 min after exposure with no pulsing (closed-channel block). Open-channel block was assessed by applying an additional pulse 1 min later. Adapted with permission from ref. 100.

affinities for the three Kv channels, Kalman et al. (101) found that ShK potently inhibited Kv1.1 and 1.3 homomeric channels at 20–100 pM concentrations, while Kv1.2 homomers were only inhibited at nanomolar concentrations. ShK toxin displayed intermediate IC50s (100–400 pM) for inhibition of homomeric Kv1.4 and 1.6 channels, relative to its nanomolar IC50 for blocking Kv1.2 homomers. Even at 100 nM, ShK failed to affect the Shaw-type Kv3.1 channel (Table 3). Anemone short K toxins affect at least some K(Ca) channels. Brugnara et al. (102) reported that ShK blocks the human erythrocyte intermediate conductance K(Ca) channel. ShK (103) also blocks the T-lymphocyte IK(Ca) channel at relatively high concentrations (IC50 approx 30 nM) relative to its Kv1.3 apparent affinity. Some differences may exist between the expressed lymphocyte and red-cell channels, though they apparently have the same gene (104). Since the IK(Ca) channel is abundantly expressed in the human T-lymphocyte, inhibition of this channel as well as the Kv1.3 homomer may be therapeutically advantageous to more completely suppress the Ca stimulus for lymphocyte proliferation. Cotton et al. (22) tested BgK on rat skeletal muscle maxi-conductance K(Ca) channels reconstituted in lipid bilayers, but failed to observe an effect. The maxi-K(Ca) channel that occurs in cultured bovine chromaffin cells was also insensitive to block by ShK (C. Lingle, personal communication). Many, but not all, scorpion toxins affect maxi-K(Ca) channels, which are widely distributed throughout the cardiovascular, gastrointestinal, and nervous systems. Further studies of the actions of sea anemone toxins upon K(Ca) channels may identify toxins that influence other subtypes of these channels. It should also be mentioned that evidence for a maxi-K(Ca) channel blocking sea anemone toxin has been reported (105). 4.3. Pharmacology: Identification of the K Channel-Binding Surface One of the most useful approaches for identifying the interactive surface of polypeptide ligand has been the “alanine scan.” In this approach, each amino acid residue

556

Kem

Table 3 Relative Potency of ShK Toxin and its DAP 22 Analog for Several Kv Channels the Intermediate Conductance hKCa4 Channel and Heterologously Expressed in Mammalian Cellsa IC50 (pM) K channel Kv1.1 Kv1.2 Kv1.3 Kv1.4 Kv1.5 Kv1.6 Kv1.7 Kv3.1 hKCa4

ShK

DAP 22-ShK

16 9000 11 312 >100,000 165 13,000 >100,000 28,000

3000 39,000 23 37,000 >100,000 10,500 ——>100,000 >100,000

aThe

IC50 is the concentration required to inhibit half of the channels. Adapted with permission from ref. 101.

(excluding key structural residues such as Cys that are required to form disulfide bonds) is replaced individually with an Ala residue. Alanine was selected the replacement because its side chain is nonionized and can be accomodated within most secondary structures (95,106). The scan utilizes a panel of monosubstituted analogs that are identical to each other except at one site. The folded structures of the analogs can be initially assessed by CD spectroscopy to determine if their folded structure is at least superficially similar to the wild-type peptide or protein. When combined with a sensitive binding or functional assay, the effect of each substitution can be measured by comparison of analog affinity or activity with that of the wild-type polypeptide. Using this approach, two conserved residues of ShK—Lys 22 and Tyr 23—were identified as being especially important for interaction with Kv channels (Fig. 11). Ala substitutions at several other residues also significantly affected binding to rat brain channels, but to a lesser extent than observed for Ala substitutions in the Lys-Tyr diad. These substitutions were at Ile 7, Arg 11, Ser 20, and Phe 27 (96,97). Arg11 was a more important determinant for specific high affinity binding to the Kv1.3 channel subtype than for brain heteromeric type K channels. These results with Arg 11 and Lys 22 first demonstrated that it was possible to manipulate ShK to obtain even higher selectivity for Kv1.3 channels (100). The channel-interactive surface of BgK has also been probed using the same Ala scan approach by the Menez laboratory (95). Although BgK is slightly longer than ShK, its potassium channel-interactive surface is very similar to that of ShK. Thus, the “hot-spot” for BgK is the Lys25-Tyr26 diad, while several other peripheral residues, Phe 6, His 13, Ser 23, and Thr 33, also contribute to the binding (94). Alessandri-Haber et al. (107) succeeded in preparing interesting BgK analogs that were much more K channel-selective than was BgK itself (Table 3). Their paper demonstrates how sea anemone toxin selectivity for Kv1.1 and 1.2 channels may be enhanced. Also, it is interesting, and not yet understood, why substitution of the orni-


557

Fig. 11. “Alanine scan” of ShK blockade of DTX binding to brain potassium channels. The relative free energy of binding relative to the wild-type ShK sequence is shown. A positive value indicates that the Ala-substituted toxin analog binds less avidly to the rat brain membrane K channels, which are predominantly heteromers of Kv1.1 and 1.2 subunits. Adapted with permission from ref. 106.

thine group for BgK’s essential Lys group did not produce the same effects as we observed for ShK. In both ShK and BgK, most residues identified by the Ala scan as being important for binding are clustered together on the toxin surface. The elliptical “patch” or pharmacophore formed by these residues positions the critical residues Lys and Tyr approx 6.6 Å apart measuring from the amino group of the Lys to the middle of the Tyr aromatic ring (95). A noncontiguous Lys and Phe form a similar diad in the scorpion K-channel toxins (Fig. 12). Thus, through what may be referred to as “convergent” molecular evolution, scorpions and sea anemones have developed very similar K channel pharmacophores, each possessing a Lys/Tyr “hot spot” (95,106). 4.4. Structure-Activity Relationships In an effort to identify and better clarify the interactive surface of ShK, we have synthesized and tested a large number (>100) of analogs. Many different substitutions were made at certain key residues, particularly Lys 22 and Tyr 23 (100,106). Later, to further assess that the predicted Kv docking orientation was correct, several multisubstituted ShK analogs were prepared that substituted Ala for three relatively bulky sites. Binding was enhanced when steric bulk was reduced at these positions, perhaps by allowing a better fit of the toxin into the outer vestibule. Several Lys 22 analogs were made to determine the effect of shortening the distance that the amino group could extend into the pore. Substitution of Orn for Lys, representing loss of one methylene group and making the side chain at position 22 approx 1.3 Α shorter, had little

558

Kem

Fig. 12. Space filling molecular structure of ShK, showing residues important for brain and lymphocyte K channel-blocking activities. Collectively these important residues constitute what is called the “pharmacophore” or surface actively involved in binding to the receptor. ShK is approximately oriented as if it were docking with the center of the K channel outer vestibule, where the inner pore is located. The structure shown is the closest to the average over the family of solution structures of Tudor et al. (94). Adapted with permission from ref. 108.

effect on binding to either rat brain or Kv1.3 channels. However, removal of three methylene units (by substitution with diaminopropionic acid, Dap) practically destroyed rat brain potency while high Kv1.3 potency was retained. This toxin analog, ShK-Dap 22, displayed drastically reduced (>100-fold) affinities for every other homomeric Kv1 channel type tested, from 1.1–1.7 (Table 4). The Dap 22 ammonium group interacts most strongly with side chains of three channel-pore residues (His 404, Asp 386, and Asp 402) instead of the deeper Tyr400, which interacts with Lys 22 in the wild-type toxin. This apparently permits a slightly more favorable interaction of some less critical residues such as Arg11 and, to a lesser extent, Phe27 with the Kv1.3 vestibule. At this time it is unclear how many low affinity Kv1 subunits can be added to Kv1.3 channels before potency is greatly diminished. Fortunately, ShK does not seem to interact potently with heteromeric Kv1 channels found in brain neurons, or myocardial, vascular, and gastrointestinal muscle cells. In brains of the two mammalian species that have been examined, neuronal channels containing both Kv1.1 and 1.2 subunits are most abundant. Some of these heteromers also possess one or more Kv1.3, 1.4, or 1.6 subunits. Kv1.2 and 1.5 subunits seem to predominate in gastrointestinal smooth muscle channels, which are mostly heteromeric.


559

Table 4 Comparative Toxicity of Three Sea Anemone Long Toxins to Crabs, Mice, and Cockroachesa Median lethal dose (µg/kg) Neurotoxin

Crab

Cockroach

Mouse

Condylactis gigantea II Calliactis parasitica I Stichodactyla helianthus I

0.2 201b 0.3

780 14,000 3,600

>50,000 >15,000 >15,000

aAll animals were injected peripherally: the arthropods were injected into the hemocoele because the mice were injected intra-peritoneally. bData from ref. 27. Table adapted with permission from ref. 49.

4.5. Possible Therapeutic Application: Immunosuppression by Blockade of Lymphocyte K Channels The rather unique expression of Kv1.3 channels on lymphocytes makes them a novel target for the design of immunosuppressant drugs. A sustained elevation of intracellular calcium is the key stimulus for activating T lymphocytes (Fig. 13). When the cell responds to antigen or mitogens, calcium is liberated from intracellular stores through the phospho-inositide signaling pathway, but this is an insufficient stimulus alone. Calcium must also enter through a voltage-independent cell membrane channel as well. It is this calcium influx that is reduced by K-channel blockers, due to their depolarization of the resting-membrane potential, which reduces the driving force for calcium entry. While blockade of this calcium channel would seem to be the optimal method for inhibiting lymphocyte activation and proliferation, this so-called calcium-release activated channel (CRAC) is rather widely distributed throughout the body, and thus its direct blockade might cause many adverse effects. Many pharmaceutical firms are actively searching for inhibitors of the two K channels that control membrane potential of the human lymphocyte, namely Kv1.3 and the intermediate conductance type IK(Ca) channel (108). We have successfully modified the structure of ShK so that its whole animal toxicity is insignificant when applied peripherally. This seems one of the main advantages of Dap 22-ShK over the scorpion peptide kaliotoxin, which is rather toxic to mammals. Since ShK is also able to inhibit the IK(Ca) channel, albeit at concentrations which are roughly 1000-fold higher than inhibit Kv1.3, it seems possible that the structure of ShK could be further modified to enhance its IK(Ca) channel-blocking potency. The success of the Crest-Menez laboratories in enhancing the potency of BgK is a quite remarkable precedent. IK(Ca) channels are not so important for the initial activation process as for the maintainence of lymphocyte proliferation, since they only appear in large numbers several days after the initial activation. Thus, Dap 22-ShK might be useful in decreasing organ rejection during the initial recovery period after organ transplantation. A fundamental barrier to this drug-development approach is that a satisfactory “proof of principle” test has not yet been made, where organ-transplant animals are chronically treated with one of these peptides. Cyclosporin is currently the most widely

560

Kem

Fig. 13. Diagram showing the influence of K channels upon lymphocyte proliferation stimulated by antigen presentation to the T-cell receptor. M refers to a hypothetical chemical mediator for activating ICrac, a Ca current stimulated by depletion of intracellular endoplasmic reticulum Ca stores. Interleukin-2 transcription is activated by Ca stimulation of calcineurin, a phosphatase. The inhibitory effects of K channel blockers, including scorpion and anemone peptide toxins, can be alleviated by addition of interleukin-2. Adapted with permission from ref. 108.

used drug to reduce graft rejection, but it has many toxic effects that limit its use, especially kidney failure. Thus, the search for safer drugs to replace or complement cyclosporin and related drugs will intensify as the demand for organ transplants increases. Ultimately, a small molecule inhibitor of these lymphocyte K channels would be most desirable, expecially for long-term treatment of auto-immune disorders like rheumatoid arthritis (RA), multiple sclerosis (MS), and type 1 diabetes. While design of a peptide-mimetic of ShK is our ultimate goal, we have initially attempted to reduce the size of this peptide. Initial experiments provided some evidence that it should be possible to reduce the size of ShK without destroying its selective Kv1.3 blocking activity. It was possible to eliminate some of the N- and C-terminal regions of the ShK sequence without eliminating activity (109). Elimination of the 3-35 disulfide linkage was tolerated, but removal of either of the other two disulfides largely eliminates K-channel activity (110). Considering that our present molecular model of ShK interaction with the Kv1.3 channel indicates that the natural toxin interacts with several, if not all four Kv subunits, it may be a considerable challenge to retain Kv1.3 selectivity as the molecule is reduced in size. Most side effects of Kv1.3 targeted drugs would be predicted to occur as a result of blockade of other Kv channels, which are generally heteromers. The inherent symmetry of the homomeric Kv1.3 channel might be exploited further in drug design by adding groups that enhance binding to other identical subunits in the channel but reduce binding to nonidentical groups in other heteromeric K channels. For instance, Arg 11 is though to interact with His 404 in one subunit of Kv1.3, whereas another subunit His404 may primarily interact with some other residue on the surface of the toxin. Substitution of Gln at position 11 significantly reduced Kv1.3 binding


561

without greatly affecting binding to brain-membrane K channels. In this case the Arg11containing natural toxin seems well adapted for high-affinity binding to Kv1.3 relative to brain heteromeric channels, which are probably composed mainly of Kv1.1and 1.2 subunits. At this time it is difficult to predict the affinities of Kv1.3 selective peptides for heteromeric Kv1 complexes such as occur in the brain and gastrointestinal systems. Clearly, subunits possessing positively charged groups in their outer vestibule or poreforming regions may repel the approach of a like charge on the sea anemone toxin. The only study that has systematically analyzed the influence of sequentially adding a new subunit to a Kv1 channel is that of Tytgat et al. (111). They measured the free energy of binding of α dendrotoxin to heteromeric complexes containing decreasing numbers (4 to 1) of Kv1.1 and increasing numbers (1–4) of a Kv1.1 pore mutant displaying negligible affinity in the homomeric state. Increasing numbers of the mutant subunits progressively diminished the free-energy decrease in an essentially additive fashion. Whether or not this relationship is applicable to the smaller scorpion and sea anemone peptides, which may bind more tightly to the vestibular region immediately adjacent to the outer pore, is not yet known. 5. CONCLUDING COMMENTS 5.1. Toxins Provide New Insights into Ion-Channel Structure and Function Peptide toxins will continue to be of most use in biomedical research as chemical probes of receptors, including ion channels. For instance, the nanomolar affinities of some sea anemone toxins for cardiac sodium channels makes them attractive radioligands for measuring the concentration of myocardial sodium channels. However, their remarkable cardiac-inotropic activity, even upon the failing heart, suggests that they might still be useful molecular models for designing a new type of inotropic drug (see below). It is highly likely that sea anemone toxins with selective activity upon other Kchannel subtypes will be found. The studies of BgK by Alessandri-Haber et al. (107) and of ShK by our group indicate that these short toxins can be chemically manipulated in the laboratory to enhance selective toxicity. They actually found that one analog possessed 300-fold greater affinity for the Kv1.3 homomeric channel than natural BgK. This suggests that searching for natural toxin homologs possessing other channel specificities may not be the only promising strategy of obtaining new chemical tools for studying K channels. 5.2. Peptide Toxins and Drug Design Although technological advances in the investigation of peptides have been impressive, pharmaceutical companies are generally uninterested in developing drugs from exogenous peptides, in spite of the fact that some of the most powerful drugs are exogenous peptides—cyclosporin being one obvious example. The often-stated reasons for this bias might not actually apply to all foreign peptides. Such reasons include: poor bioavailability, high antigenicity, and short duration of action. There are so many new ways to administer drugs now that the first justification seems no longer to be an unsurmountable problem. Antigenicity is dependent on the number of antigenic determinants, which is often quite limited for small peptides. Janin and Chothia (112)

562

Kem

reported that protein antigen-antibody interaction typically involves a surface area of 300–400 square Α. Novotny and Haber (113) analyzed the antigenicity of a scorpion α-toxin containing 65 residues and found only four distinguishable antigenic sites; the sea anemone toxins discussed here would be expected to have even fewer epitopes. Also, since peptides like Dap 22-ShK are inhibitors of lypmphocyte proliferation, development of an immune response should be much less than otherwise. Thus, it is possible that many peptides, including K short-toxin analogs like Dap 22-ShK, will have minimal antigenic potency. If anemone and scorpion peptide analogs are to be used as intravenous immunosuppressants in the immediate post-transplantation period, they will not only have to be free of adverse effects upon nontargeted K channels, but will also need to possess favorable pharmacokinetic properties such as a relatively long plasma half-life. Efforts are under way in the author’s lab to identify the mechanisms of removal of ShK from the plasma space and to enhance plasma half-life of Dap22-ShK. It can be predicted that rapid progress will be made in the next few years in understanding the intimate interactions between peptide toxins and the outer-vestibular receptor. Several groups are actively working on the structural analysis of two-dimensional crystals of homomeric K channels, including Kv1.3. Doyle et al. (114) have recently reported the crystal structure of a much simpler but homologous P region structure expressed in a bacterial protein, which forms functional K channels and can even bind certain scorpion toxins with moderate affinity (115). This simplified system is very attractive for high-resolution structural analysis by NMR and X-ray crystallography. A high resolution structure of the toxin-channel complex would provide a much firmer basis for rational drug design, as long as this synthetic channel contains all of the determinants involved in complex formation with the peptide toxin. It was demonstrated previously that the P-region of a Shaker type channel determines scorpion toxin binding. The derivation of peptide-mimetic drugs from a peptide pharmacophore is a difficult challenge. Success will depend on the selection of appropriate molecular models resembling the peptide pharmacophore, and will undoubtedly require synthesis and pharmacological testing of many compounds. As the structure of a peptide is “minimized” one can expect that potency and selectivity will often be reduced in most, but not all products. In terms of selecting an overall strategy for peptidomimetic drug design, it will be of considerable interest whether rational design approaches like ours, or the combinatorial high-throughput screening approaches, are most successful in designing selective Kv1.3 channel immunosuppressants. 5.3. Chemical Biology of Anthozoan Toxins Until recently, investigation of cnidarian venoms and toxins has been relatively slight in comparison to toxins from terrestrial organisms such as snakes and scorpions. As knowledge of the components of cnidarian venoms becomes advanced, it is possible that some interesting insights regarding the co-evolution of the toxins and their cnidocyst means of administration will be forthcoming. Some questions for which we still have no answers will now be considered. How do the different venom constituents act together under natural conditions? It is possible that the three known types of sea anemone toxins (cytolysins, sodium-channel


563

toxins, and potassium-channel toxins) act in a synergistic fashion on electrically excitable cells. The cytolytic actinoporins (9,10) would initially depolarize the electrically excitable cell, causing activation of the sodium channels. These would remain open for an abnormally long period of time in the presence of the sodium channel toxins that delay inactivation. To ensure that the excitable cell remains depolarized, the sea anemone also injects toxins that block the K channels and also assist in the membrane repolarization! The ultimate effect of these toxins would be to cause maintained depolarization of presynaptic nerve endings, and thus an abnormally large release of neurotransmitters. It has to be admitted that no one has yet shown that the resulting convulsive paralysis of a crab or fish results from such toxin synergism, but it at least seems plausible. It certainly makes sense to couple enhancement of sodium inactivation with potassium-channel blockade. Perhaps this may also be the reason why only minute amounts of the short K-channel toxins have been found in sea anemones. How do some commensal animals avoid being poisoned by sea anemones? It is known that certain symbiotic crustaceans live among sea anemones bearing potent toxins. These include “cleaner” shrimps and the anemone fishes (116,117), which can touch the tentacles of a poisonous anemone without being paralyzed. Various theories have been presented and considered. The most tenable ones, at least for the more completely studied anemone fish-anemone symbiosis, involve some means of preventing cnidocyst discharge, either by eliminating the characteristic chemical stimuli that usually trigger the discharge, or producing inhibitors of the discharge. Anemone (Amphiprion sp.) fish which have had some of their body mucus removed are readily stung. This would be consistent with either of the aforementioned hypotheses as well as with a “thick mucus” hypothesis, which claims that anemone fish are not readily stung because of a thicker mucus layer that protects their skin. These hypotheses were tested before there was much knowledge of sea anemone venoms and toxins. The availability of purified toxins might permit other experiments that would assess whether these commensal organisms also display an innate or acquired resistance to the toxins. It is hoped that the next decade will see some application of our laboratory knowledge of sea anemone toxin properties back to the natural situation, so that we can better understand how these substances confer biological advantages to the anthozoans that make them. Why are the ion channels of some organisms extremely sensitive to sea anemone toxins while those of other groups are quite resistant? Amongst the sea anemone long sodium-channel toxins most are extremely potent toxins on crustaceans, whereas the ones which are potent toxins on vertebrates are often less potent on crustaceans (30,64). It has already been shown that a single amino acid difference in a toxin’s receptorbinding site can often change toxin affinity >1000-fold. The more interesting question is, why do the sea anemone toxins rather consistently show high crustacean toxicity rather than vertebrate toxicity? One is tempted to speculate that, in the course of evolution, it was initially crustaceans that were the important targets of sea anemone toxins, because they not only represented a good food source for the sea anemone, but large crustaceans were also probably dominant predators on sea anemones. The vertebrates appeared much later than the crustaceans, and may generally be less important targets for most sea anemones, but probably with exceptions. It should be pointed out that while insects also are affected by these sea anemone long toxins, they are also more

564

Kem

resistant than are the crustaceans (Table 4). It is also interesting that sea anemones with highly potent cytolysins tend to have fewer ion channel toxins, as if this would be redundant. Do different types of nematocyst contain different toxins and other constituents like phospholipases and protease inhibitors? Answering this question has not been easy, because of the relative difficulties encountered in separating the different kinds of sea anemone cnidocysts. Furthermore, application of immunohistochemical techniques is difficult, probably for two reasons. First, nematocysts seem to bind immunoglobins nonspecifically, requiring use of affinity-purified antibodies and extensive control incubations (118,119). Second, fixed cnidocytes are not readily penetrated by immunoglobulins (Kem, W. R., unpublished results). Despite these experimental difficulties, it should be possible to directly visualize the toxins and other venom constituents using immunohistochemical methods. It should also be possible to shed further light on a perennial controversy in this field, namely whether some of the purified toxins (and other peptides such as phospholipases and protease inhibitors) are packaged in glands rather than cnidocysts. It will be of interest to observe whether laboratory investigations of anthozoan toxins can provide insights into what may be termed the “chemical ecology” of these organisms. Few scientists are currently investigating the biological roles of toxins, since it is relatively difficult to obtain funds to carry out such research. Nevertheless, it is hoped that biochemical and pharmacological studies like those described in this chapter may ultimately provide a strong scientific foundation for investigating how organisms in the wild use venoms to their own evolutionary advantage. ACKNOWLEDGMENTS The author gratefully acknowledges the longterm collaboration of Dr. Mike Pennington (Bachem Biosciences, Inc., King of Prussia, Pennsylvania), who synthesized all of the sea anemone toxin analogs we have investigated, and Dr. Ray Norton (NMR Laboratory, Biomolecular Research Institute, Melbourne, Australia) whose lab carried out NMR structural determinations on the Sh toxins. The author’s research would have been much more limited without their respective contributions and enthusiasm for this research. In addition, Dr. George Chandy (University of CaliforniaIrvine) and Dr. Doug Krafte (now at ICagen, Research Triangle Park, North Carolina) were important contributors to the ShK immunosuppressant research project. Barbara Seymour provided the sea anemone illustrations. The author’s research on sea anemone toxins was supported by NIH grants RO1 GM- and GM-54221. REFERENCES 1. Guy, H. R. and Conti, F. (1990) Pursuing the structure and function of voltage-gated channels. Trends Neurosci. 13, 201–206. 2. Catterall, W. A. (1980) Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Ann. Rev. Pharmacol. Toxicol. 20, 15–43. 3. Goldin, S. (1999) Diversity of mammalian voltage-gated sodium channels. Ann. NY Acad. 868, 38–50. 4. George, J. D. and George, J. J. (1979) Marine Life: An Illustrated Encyclopedia of Invertebrates in the Sea. John Wiley & Sons, NY, pp. 288.


565

5. Tardent, P. (1995) The cnidarian cnidocyte, a high-tech cellular weaponry. BioEssays 17, 351–362. 6. Watson, G. M. and Hessinger, D. A. (1988) Localization of a purported chemoreceptor involved in triggering cnida discharge in sea anemones, in The Biology of Nematocysts (Hessinger, D. A. and Lenhoff, H. M., eds.), Academic Press, NY, pp. 255–272. 7. Cutress, C. E. (1955) An interpretation of the structure and distribution of cnidae in Anthozoa. Syst. Zool. 4, 120–137. 8. Fautin, D. G. (1988) Importance of nematocysts to Actiniian taxonomy, in Biology of Nematocysts (Hessinger, D. A. and Lenhoff, H. M., eds.), Academic Press, NY, pp. 487–500. 9. Kem, W. R. (1988b) Sea anemone toxins: structure and action, in The Biology of Nematocysts (Hessinger, D. A. and Lenhofff, H. M., eds), Academic Press, NY, pp. 375–405. 10. Kem, W. R. (1988a) Peptide chain toxins of marine animals, in Biomedical Importance of Marine Organisms, vol. 13 (Fautin, D., ed.), Calif. Acad. Sci., San Francisco, CA pp. 69–83. 11. Hessinger, D. A., Lenhoff, H. M., and Kahan, L. B. (1973) Haemolytic, phospholipase A and nerve-affecting activities of sea anemone nematocyst venom. Nat. New Biol. 241, 125–127. 12. Grotendorst, G. R. and Hessinger, D. A. (1999) Purification and partial characterization of the phospholipase A2 and co-lytic factor from sea anemone (Aiptasia pallida) nematocyst venom. Toxicon 37, 1779–1796. 13. Beress, L., Bruhn, T., Sanchez-Rodriquez, Wachter, E., and Schweitz, H. (2000) Sea anemone toxins acting on Na+-channels and K+-channels: isolation and investigation, in Animal Toxins: Facts and Protocols (Rochat, H. and Martin-Euclaire, M.-F., eds.), Birkhauser, Basel, pp. 31–56. 14. Kem, W. R., Parten, B., Pennington, M. W., Dunn, B. M., and Price, D. (1989) Isolation, characterization, and amino acid sequence of a polypeptide neurotoxin occurring in the sea anemone Stichodactyla helianthus. Biochemistry 28, 3483–3489. 15. Malpezzi, E. L. A., Freitas, J. C., Muramoto, K., and Kamiya, H. (1993) Characterization of peptides in sea anemone venom collected by a novel procedure. Toxicon 31, 853–864. 16. McKay, M. C. and Anderson, P. A. V. (1988) On the preparation and properties of isolated cnidocytes and cnidae, in The Biology of Nematocysts (Hessinger, D. A. and Lenhoff, H. M., eds.), Academic Press, NY, pp. 273–294, 17. Aneiros, A., Garcia, I., Martinez, J. R., Harvey, A. L., Anderson, A. J., Marshall, D. L., et al. (1993) A potassium channel toxin from the secretion of the sea anemone Bundosoma granulifera. Biochim. Biophys. Acta 1157, 86–92. 18. Castaneda, O., Sotolongo, V., Amor, A. M., Stocklin, R., Anderson, A. J., Harvey, A. L., et al. (1995) Characterization of a potassium channel toxin from the Caribbean sea anemone Stichodactyla helianthus. Toxicon 33, 603–613. 19. Spagnuolo, A., Zanetti, L., Cariello, L., and Piccoli, R. (1994) Isolation and characterization of two genes encoding calitoxins, neurotoxic peptides from Calliactis parasitica (Cnidaria). Gene 138, 187–191. 20. Gendeh, G. S., Chung, M. C. M., and Jeyaseelan, K. (1997a) Genomic structure of a potassium channel toxin from Heteractis magnifica. FEBS Lett. 418, 183–188. 21. Pennington, M. W., Byrnes, M. E., Zaydenberg, I., Khaytin, I., de Chastonay, J., Krafte, D., et al. (1995) Chemical synthesis and characterization of ShK toxin: a potent potassium channel inhibitor from a sea anemone. Int. J. Pept. Prot. Res. 46, 354–358. 22. Cotton, J., Crest, M., Bouet, F., Alessandri, N., Gola, M., Forest, E., et al. (1997) A potassium-channel toxin from the sea anemone Bunodosoma granulifera, an inhibitor for Kv1 channels. Revision of the amino acid sequence, disulfide-bridge assignment, chemical synthesis, and biological activity. Eur. J. Biochem. 244, 192–202. 23. Gendeh, G. S., Young, L. C., de Medeiros, C. L. C., Jeyaseelan, K., Harvey, A. L., and Chung, M. C. M. (1997) A new potassium channel toxin from the sea anemone

566

24.

25. 26.

27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37. 38.

39.

40.

41.

Kem Heteractis magnifica: isolation, cDNA cloning, and functional expression. Biochemistry 36, 11,461–11,471. Kelso, G. J. and Blumenthal, K. M. (1998) Identification and characterization of novel sodium channel toxins from the sea anemone Anthopleura xanthogrammica. Toxicon 36, 41–51. Wunderer, G. (1978): Die Disulfidbrucken von Toxin II aus Anemonia sulcata. HoppeSeyler’s Z. Physiol Chem. 359, 1193–1201. Beress, L. (1988) Sea anemone toxins as tools for physiological, pharmacological and biophysical research, in Poisonous and Venomous Marine Animals of the World, 2nd ed. (Halstead, B. W., eds.), Darwin Press, Princeton, pp. 150–161. Cariello, L., de Santis, A., Fiore, F., Piccoli, R., Spagnuolo, A., Zanetti, L., and Parente, A. (1989) Calitoxin, a neurotoxic peptide from the sea anemone Calliactis parasitica: amino acid sequence and electrophysiological properties. Biochemistry 28, 2484–2489. Ishida, M., Yokoyama, A., Shimakura, K., Nagashima, Y., and Shiomi, K. (1997) Halcurin, a polypeptide from the sea anemone Halicurias sp., with a structural resemblance to type 1 and type 2 toxins. Toxicon 35, 537–544. Shiomi, K., Lin, X.-Y., Nagashima, Y., and Ishida, M. (1995) Isolation and amino acid sequence of polypeptide toxins in the sea anemone Condylactis passiflora. Fish. Sci. 61, 1016–1021. Hellberg, S. and Kem, W. R. (1990) Quantitative structure-activity relationships for sea anemone polypeptide toxins. Int. J. Peptide Prot. Res. 36, 440–444. Lin, X.-Y., Ishida, M., Nagashima, Y., and Shiomi, K. (1996) A polypeptide toxin in the sea anemone Actinia equina homologous with other sea anemone sodium channel toxins: isolation and amino acid sequence. Toxicon 34, 57–65. Shiomi, K., Qian, W.-H., Lin, X.-Y., Shimakura, Nagashima, Y., and Ishida, M. (1997) Novel polypeptide toxins with crab lethality from the sea anemone Anemonia erythraea. Biochim. Biophys. Acta. Loret, E. P., de Valle, R. M., Mansuelle, P., Sampieri, F., and Rochat, H. (1994) Positively charged amino acid residues located similarly in sea anemone and scorpion toxins. J. Biol. Chem. 269, 16,785–16,788. Nishida, S., Fujita, S., Warashina, A., Satake, M., and Tamiya, N. (1985) Amino acid sequence of a sea anemone toxin from Parasicyonis actinostoloides. Eur. J. Biochem. 150, 171–173. Odinokov, S. E., Nabiullin, A. A., Kozlovskaya, E. P., and Elyakov, G. B. (1989) Structure-function relationship of polypeptide toxins: modifying gating mechanism of sodium channel. Pure Appl. Chem. 61, 497–500. Pallaghy, P. K., Scanlon, M. J., Monks, S. A., and Norton, R. S. (1995) Three-dimensional structure in solution of the polypeptide cardiac stimulant anthopleurin-A. Biochemistry 34, 3782–3794. Norton, R. S. (1991) Structure and structure-function relationships of sea anemone proteins that interact with the sodium channel. Toxicon 29, 1051–1084. Monks, S. A., Pallaghy, P. K., Scanlon, M. J., and Norton, R. S. (1995) Solution structure of the cardiostimulant polypeptide anthopleurin-B and comparison with anthpleurin-A. Structure 3, 791–803. Hinds, M. G. and Norton, R. S. (1993) Sequential H-NMR assingments of neurotoxin III from the sea anemone Heteractis macrodactylus and structural comparison with related toxins. J. Protein Chem. 12, 371–378. Fogh, E., Kem, W. R., and Norton, R. S. (1990) Solution structure of neurotoxin I from the sea anemone Stichodactylus helianthus. A nuclear magnetic resonance, distance geometry, and restrained molecular dynamics study. J. Biol. Chem. 265, 13,016–13,028. Wilcox, G. R., Fogh, R. H., and Norton, R. S. (1993) Refined structure in solution of the sea anemone neurotoxin ShI. J. Biol. Chem. 268, 24,707–24,719.


567

42. Scanlon, M. J. and Norton, R. S. (1994) Multiple conformations of the sea anemone polypeptide anthopleurin-A in solution. Prot. Sci. 3, 1121–1124. 43. Norton, R. S., Cross, K., Braach-Maksvytis, V., and Wachter, E. (1993) 1H-n.m.r. study of the solution properties and secondary structure of neurotoxin III from the sea anemone Anemonia sulcata. Biochem. J. 293, 545–551. 44. Bahraoui, E. M., El Ayab, M., Granier, C., Beress, L., and Rochat, H. (1989) Specificity of antibodies to sea anemone toxin III and immunogenicity of the pharmacological site of anemone and scorpion toxins. Eur. J. Biochem. 180, 55–60. 45. Pauron, D., Barhanin, J., and Lazdunski, M. (1985) The voltage-dependent Na+ channel of insect nervous system identified by receptor sites for tetrodotoxin, and scorpion and sea anemone toxins. Biochem. Biophys. Res. Comm. 131, 1226–1233. 46. Schweitz, H., Bidard, J. N., Frelin, C., Pauron, D., Vijverberg, H. P. M., Mahasneh, D. M., and Lazdunski, M. (1985) Purification, sequence, and pharmacological properties of sea anemone toxins from Radianthus paumotensis. A new class of sea anemone toxins acting on the sodium channel. Biochemistry 24, 3554–3561. 47. Murayama, K. N., Abbott, N. J., Narahashi, T., and Shapiro, B. (1972) Effect of allethrin and Condylactis toxin on the kinetics of sodium conductance of crayfish axon membranes. Comp. Gen. Pharmacol. 3, 391–400. 48. Bergman, C., DuBois, J. M., Rojas, E., and Rathmayer, W. (1976) Decreased rate of sodium conductance inactivation in the node of Ranvier induced by a polypeptide toxin from sea anemone. Biochim. Biophys. Acta 455, 173–184. 49. Salgado, V. L. and Kem, W. R. (1992) Actions of three structurally distinct sea anemone toxins on crustacean and insect sodium channels. Toxicon 30, 1365–1381. 50. El-Sherif, N., Fozzard, H. A., and Hanck, D. A. (1992) Dose-dependent modulation of the cardiac sodium channel by sea anemone toxin ATX II. Circ. Res. 70, 285–301. 51. Hanck, D. A. and Sheets, M. F. (1995) Modification of inactivation in cardiac sodium channels: ionic current studies with anthopleurin-A toxin. J. Gen. Physiol. 106, 601–616. 52. Wasserstrom, J. A., Kelly, J. E., and Liberty, K. N. (1993) Modification of cardiac Na+ channels by anthopleurin-A: effects on gating and kinetics. Pflügers Arch. 424, 15–24. 53. Cannon, S. C. (1996) Sodium channel defects in myotonia and periodic paralysis. Ann. Rev. Neurosci. 19, 141–164. 54. Cannon, S. C. and Corey, D. P. (1993) Loss of Na+ channel inactivation by anemone toxin (ATX II) mimics the myotonic state in hyperkalaemic periodic paralysis. J. Physiol. 466, 501–520. 55. Rogers, J. C., Qu, Y., Tanada, T. N., Scheuer, T. and Catterall, W. A. (1996) Molecular determinants of high affinity binding of alpha-scorpion toxin and sea anemone toxin in the S3-S4 extracellular loop in domain IV of the Na+ channel a subunit. J. Biol. Chem. 271, 15,950–15,962. 56. Chen, L.-Q., Santarelli, V., Horn, R., and Kallen, R. G. (1996) A unique role for the S4 segment of domain 4 in the inactivation of sodium channels. J. Gen. Physiol. 108, 549– 556. 57. Sheets, M. F. and Hanck, D. A. (1995) Voltage-dependent open-state inactivation of cardiac sodium channels: gating current studies with anthopleurin-A toxin. J. Gen. Physiol. 106, 617–640. 58. Lawrence, J. C. and Catterall, W. A. (1981) Tetrodotoxin-insensitive sodium channels. Binding of polypeptide neurotoxins in primary cultures of rat muscle cells. J. Biol Chem. 256, 6223–6229. 59. Scriabine, A., Van Arman, G., Morgan, G., Morris, A. A., Bennett, C. D., and Bohidar, N. R. (1979) Cardiotonic effects of anthopleurin-A, a polypeptide from a sea anemone. J. Cardiovasc. Pharmacol. 1, 571–583. 60. Catterall, W. A. and Beress, L. (1978) Sea anemone toxin and scorpion toxin share a com-

568

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73. 74. 75. 76. 77.

Kem mon receptor site associated with the action potential sodium iontophore. J. Biol. Chem. 253, 7393–7396. Couraud, F., Rochat, H., and Lissitzky, S. (1978) Binding of scorpion and sea anemone neurotoxins to a common site related to the action potential Na+ ionophore in neuroblastoma cells. Biochem. Biophys. Res. Commun. 83, 1525–1530. Kem, W. R., Pennington, M. W., Krafte, D. S., and Hill, R. J. (1996a) Sea anemone toxins affecting sodium channels: are the similarities greater than the differences? in Biochemical Aspects of Marine Pharmacology (Lazarovici, P., Spira, M., and Zlotkin, E., eds.), Alaken Press, Ft. Collins, pp. 98–120. Benzinger, G. R., Kyle, J. W., Blumenthal, K. M., and Hanck, D. A. (1998) A specific interaction between the cardiac sodium channel and site-3 toxin anthopleurin B. J. Biol. Chem. 273, 80–84. Kem, W. R., Pennington, M. W., and Dunn, B. M. (1990) Sea anemone polytpeptide toxins affecting sodium channels. Initial structure-activity investigations, in Marine Toxins: Origin, Structure, and Molecular Pharmacology, vol. 418 (Hall, S., and Strichartz, F., eds.) American Chemical Soc., Washington, DC, pp. 279–289. Pennington, M. W., Kem, W. R., Norton, R. S., and Dunn, B. M. (1990a) Chemical synthesis of a neurotoxic polypeptide from the sea anemone Stichodactyla helianthus. Intern. J. Pept. Protein Res. 36, 335–343. Pennington, M. W., Kem, W. R., and Dunn, B. M. (1990) Synthesis and biological activity of six monosubstituted analogs of a sea anemone (Stichodactyla helianthus) type 2 polypeptide toxin. Peptide Res. 3, 1–5. Mahnir, V. M., Kozlovskaya, E. P., and Elyakov, G. B. (1990) Modification of carboxyl groups in sea anemone toxin RTX-III from Radianthus macrodactylus. Toxicon 28, 1255–1263. Khera, P. K. and Blumenthal, K. M. (1996) Importance of highly conserved anionic residues and electrostatic interactions in the activity and structure of the cardiotonic polypeptide anthopleurin B. Biochemistry 35, 3503–3507. Pennington, M. W., Zadenberg, I., Byrnes, M. E., Norton, R. S., and Kem, W. R. (1994) Synthesis of the cardiac inotropic polypeptide anthopleurin-A. J. Pept. Prot. Res. 43, 463–470. Gallagher, M. J. and Blumenthal, K. M. (1992) Cloning and expression of wild-type and mutant forms of the cardiotonic polypeptide anthopleurin B. J. Biol. Chem. 267, 13,958– 13,963. Gallagher, M. J. and Blumenthal, K. M. (1994) Importance of the unique cationic residues arginine 12 and lysine 49 in the activity of the cardiotonic polypeptide anthopleurin B. J. Biol. Chem. 269, 254–259. Khera, P. K. and Blumenthal, K. M. (1994) Role of the cationic residues arginine 14 and lysine 48 in the function of the cardiotonic polypeptide anthopleurin B. J. Biol. Chem. 269, 921–925. Khera, P. K., Benzinger, F. R., Lipkind, G., Drum, C. L., Hanck, D. A., and Blumenthal, K. M. (1995) Multiple cationic residues of anthopleurin B that determine high affinity and channel isoform discrimination. Biochemistry 34, 8533–8541. Kelso, G. J., Drum, C. L., Hanck, D. A., and Blumenthal, K. M. (1996) Role for Pro-13 in directing high-affinity binding of anthopleurin B to the voltage-sensitive sodium channel. Biochemistry 35, 14,157–14,164. Dias-Kadambi, B. L., Drum, C. L., Hanck, D. A., and Blumenthal, K. M. (1996) Leucine 18, a hydrophobic residue essential for high affinity binding of anthopleurin B to the voltage-sensitive sodium channel. J. Biol. Chem. 271, 9422–9428. Fetrow, J. S. (1995) Omega loops: nonregular secondary structures significant in protein function and stability. FASEB J. 9, 708–717. Benzinger, G. R, Drum, C. L., Chen, L.-Q., Kallen, R. G., and Hanck, D. A. (1997) Differ-


569

ences in the binding sites of two site-3 sodium channel toxins. Pflügers Arch. Eur. J. Physiol. 434, 742–749. 78. Gould, A. R., Mabbutt, B. C., Llewellyn, L. E., Goss, N. H., and Norton, R. S. (1992) Linear and cyclic peptide analogues of the polypeptide cardiac stimulant, anthopleurin-A. lHNMR and biological activity studies. Eur. J. Biochem. 206, 641–651. 79. Renaud, J. F., Fosset, M., Schweitz, H., and Lazdunski, M. (1986) The interaction of polypeptide neurotoxins with tetrodotoxin-resistant Na+ channels in mammalian cardiac cells. Correlation with inotropic and arrhythmic effects. Eur. J. Pharmacol. 120, 161–170. 80. Blair, R. W., Peterson, D. F., and Bishop, V. S. (1978) The effect of anthopleurin-A on cardiac dynamics in conscious dogs. J. Pharm. Exp. Ther. 207, 271–276. 81. Norton, R. S. (1997) Polypeptide modulators of sodium channel function as a basis for the development of novel cardiac stimulants, in Structure Based Drug Design (Veerapandian, P., ed.), Marcel Dekker, NY, pp. 295–319. 82. Hashimoto, Y. and Ashida, K. (1987) Screening of toxic corals and isolation of a toxic polypeptide from Goniopora spp. Publ. Seto. Mar. Biol. Lab. 20, 703–711. 83. Ashida, K., Toda, H., Fujiwara, M., and Sakiyama, F. (1987) Amino acid sequence of Goniopora toxin. Jpn. J. Pharmacol. 43(Suppl.)33, (abstract P-33). 84. Fujiwara, M., Muramatsu, I., Hidaka, H., Ikushima, S., and Ashida, K. (1979) Effects of Goniopora toxin, a polypeptide isolated from coral, on electromechanical properties of rabbit myocardium. J. Pharm. Exp. Ther. 210, 153–157. 85. Noda, M., Muramatsu, I., and Fujiwara, M. (1984) Effects of Goniopora toxin on the membrane currents of fullfrog atrial muscle. N. S. Arch. Pharmacol. 327, 75–80. 86. Ikushima, S., Muramatsu, I., Fujiwara, M., and Ashida, K. (1981) Relationship between the effects of Goniopora toxin on action potential and on contractile force in guinea-pig papillary muscle. Jpn. J. Pharmacol. 31, 1051–1060. 87. Muramatsu, I., Fujiwara, M., Miura, A., and Narahashi, T. (1985) Effects of Goniopora toxin on crayfish giant axons. J. Pharm. Exp. Ther. 234, 307–315. 88. Gonoi, T., Ashida, K., Feller, D., Schmidt, J., Fujiwara, M., and Catterall, W. A. (1986) Mechanism of action of a polypeptide neurotoxin from the coral Goniopora on sodium channels in mouse neuroblastoma cells. Mol. Pharmacol. 29, 347–354. 89. Qar, J., Schweitz, H., Schmid, A., and Lazdunski, M. (1986) A polypeptide toxin from the coral Goniopora. Purification and action on Ca2+ channels. FEBS Lett. 202, 331–336. 90. Minagawa, S., Isida, I., Nagashima, Y., and Shiomi, K. (1998) Primary structure of a potassium channel toxin from the sea anemone Actinia equina. FEBS Lett. 427, 149–151. 91. Harvey, A. L., Rowan, E. G., Vatanpour, H., Young, L. C., Castaneda, O., Mebs, D., et al. (1996) Potassium channel neurotoxins from sea anemones, in Biochemical aspects of Marine Pharmacology (Lazarovici, P., Spira, M., and Zlotkin, E., eds.), Alaken Press, Ft. Collins, pp. 121–131. 92. Pohl, J., Hubalek, F., Byrnes, M. E., Nielsen, K. R., Woods, A., and Pennington, M. W. (1995) Assignment of the three disulfide bonds in ShK toxin: a potent potassium channel inhibitor from the sea anemone Stichodactyla helianthus. Lett. Peptide Sci. 1, 291–297. 93. Kem, W. R., Sanyal, G., Williams, R. W., and Pennington, M. W. (1996b) Secondary structure of ShK toxin, a potassium channel-blocking peptide. Lett. Peptide Sci. 3, 69–72. 94. Tudor, J. E., Pallaghy, P. K., Pennington, M. W., and Norton, R. S. (1996) Solution structure of ShK toxin, a novel potassium channel inhibitor from a sea anemone. Nature Str. Biol. 3, 317–320. 94a.Lanigan, M. D., Tudor, J. E., Pennington, M. W., and Norton, R. S. (2001) A helical capping motif in ShK toxin and its role in helix stabilization. Biopolymers 58, 422–436. 95. Dauplais, M., Lecoq, A., Song, J., Cotton, J., Jamin, N., Gilquin, B., et al. (1997) On the convergent evolution of animal toxins. conservation of a diad of functional residues in potassium channel-blocking toxins with unrelated structures. J. Biol. Chem. 272, 4302–4309.

570

Kem

96. Tudor, J. E., Pennington, M. W., and Norton, R. S. (1998) Ionization behaviour and solution properties of the potassium-channel blocker ShK toxin. Eur. J. Biochem. 251, 133–141. 97. Schweitz, H., Bruhn, T., Guillemare, M. D, Lancelin, J.-M., Beress, L., and Lazdunshi, M. (1995) Kalicludines and Kaliseptine: Two different classes of sea anemone toxins for voltage-sensitive K+ channels. J. Biol. Chem. 270, 25,121–25,126. 98. Diochot, S., Schweitz, H., Beress, L., and Lazdunski, M. (1998) Sea anemone peptides with a specific blocking activity against fast inactivating potassium channel Kv3.4. J. Biol. Chem. 73, 6744–6749. 99. Llewellyn, L. E. and Norton, R. S. (1991) Binding of the sea anemone polypeptide BdS II to the voltage-gated sodium channel. Biochem. Intern. 24, 937–946. 100. Pennington, M. W., Mahnir, V. M., Krafte, D. S., Zadenberg, I., Byrnes, M. E., Khaytin, I., et al. (1996) Identification of three separate binding sites on ShK toxin, a potent inhibitor of voltage-dependent potassium channels in human T-lymphocytes and rat brain. Biochem. Biophys. Res. Commun. 219, 696–701. 101. Kalman, K., Pennington, M., Nguyen, A., Mahnir, V. M., Kem, W R., Grissmer, S., et al. (1998) ShK-K22DAP: A potent Kv1.3-specific immunosuppressive peptide. J. Biol. Chem. 273, 32,697–32,707. 102. Brugnara, C., Armsby, C. C., De Franceschi, L., Crest, M., Martin Euclaire, M. F., and Alper, S. L. (1995) Ca2+-activated K+ channels of human and rabbit erythrocytes display distinctive patterns of inhibition by venom peptide toxins. J. Membr. Biol. 147, 71–82. 103. Rauer, H., Pennington, M. W., Cahalan, M. D., and Chandy, K. G. (1999) Structural conservation of the pores of calcium-activated and voltage-gated potassium channels determined by a sea anemone toxin. J. Biol. Chem. 274, 21,885–21,892. 104. Aiyar, J. (1999) Potassium channels in leukocytes and toxins that block them: structure, function and therapeutic implications. Perspect Drug Disc. Design 15/16, 257–280. 105. Araque, A., Urbano, F. J., Cervenansky, C., Gandia, L., and Buno, W. (1995) Selective block of Ca2+-dependent K+ current in crayfish neuromuscular system and chromaffin cells by sea anemone Bunodosoma cangicum venom. J. Neurosci. Res. 42, 539–546. 106. Pennington, M. W., Mahnir, V. M., Khaytin, I., Zaydenberg, I, Byrnes, M. E., and Kem, W. R. (1996): An essential binding surface for ShK toxin interaction with rat brain potassium channels. Biochemistry 35, 16,407-16,411. 107. Alessandri-Haber, N., Lecoq, A., Gasparin, S., Grangier-Macmath, G., Jacquet, G., Harvey, A. L., et al. (1999) Mapping the functional anatomy of BgK on Kv1.1, Kv1.2, and Kv1.3. Clues to design analogs with enhanced selectivity. J. Biol. Chem. 274, 35,653– 35,661. 108. Kem, W. R., Pennington, M. W., and Norton, R. S. (1999) Sea anemone toxins as templates for the design of immunosuppressant drugs. Perspec. Drug Disc. Design 15/16, 111–129. 109. Pennington, M. W., Mahnir, V. M., Baur, P., McVaugh, C. T., Behm, D., and Kem, W. R. (1997b) The effect of truncation on ShK toxin: elimination of the amino-carboxyl terminal (3-35) disulfide linkage stabilizing the amino and carboxyl terminal segments. Prot. Peptide Lett. 4, 237–242. 110. Pennington, M. W., Lanigan, M. D., Kalman, K., Mahnir, V. M., Rauer, H., McVaugh, C. T., et al. (1999) Role of disulfide bonds in the structure and potassium channel blocking activity of ShK toxin. Biochemistry 38, 14,549–14,558. 111. Tytgat, J., Debont, T., Carmeliet, E., and Daenens, P. (1995) The alpha-dendrotoxin footprint on a mammalian potassium channel. J. Biol. Chem. 270, 24,776–24,781. 112. Janin, J. and Chothia, C. (1990) The structure of protein-protein recognition sites. J. Biol. Chem. 265, 16,027–16,030. 113. Novotny, J. and Haber, E. (1986) Static accessibility model of protein antigenicity: the case of scorpion neurotoxin. Biochemistry 25, 6748–6754.


571

114. Doyle, D. A., Cabral, J. M., Pfuetzner, R. A., Kuo, A., Gulbis, J. M., Cohen, S. L., et al. (1998) The structure of the potassium channel: molecular basis of K+ conduction and selectivity. Science 280, 69–77. 115. MacKinnon, R., Cohen, S. L., Kuo, A., Lee, A., and Chait, B. T. (1998) Structural conservation in prokaryotic and eukaryotic potassium channels. Science 280, 106–109. 116. Mariscal, R. N. (1970), The nature of the symbiosis between Indo-Pacific anemone fishes and sea anemones. Mar. Biol. 6, 58–65. 117. Fautin, D. G. (1991), The anemone-fish symbiosis: what is known and what is not. Symbiosis 10, 23–46. 118. de Couet, H. G. (1982), Coelenterate nematocysts bind immunoglobulins. Experientia 38, 353–354. 119. Maier, L. and Rathmayer, W. (1982), Lokalisierung von Anemonentoxin in den Tentakeln der Wachsrose Anemonia sulcata (Coelenterata) mit Hilfe spezifischer Antikorper. Verh. Dtsch. Zool. Ges. 281 (Abstr.).

572

Kem

Nemertine Toxins

573

26 Nemertine Toxins William R. Kem

1. INTRODUCTION In 1936 a Belgian pharmacologist reported the serendipitous discovery of at least two different toxins in nemertines, a relatively small phylum of marine worms. Bacq demonstrated that an aqueous homogenate of the hoplonemertine Amphiporus lactifloreus potently contracted isolated frog skeletal muscle and stimulated the cat cervical autonomic ganglion in a manner similar to the neurotransmitter acetylcholine (ACh). However, since the this activity was stable in highly alkaline solution, it could not be due to ACh (1,2). In other nemertine extracts Bacq also found a neurotoxic activity lacking nicotinic-receptor effects, which he referred to as “nemertine.” Both “amphiporine” and “nemertine” extracts caused convulsions, paralysis, and death when injected into crabs. In contrast with “amphiporine” activity, “nemertine” activity only slowly traversed a dialysis membrane. Harold King, an organic chemist who had previously determined the structures of a variety of plant natural products, including the arrow poison d-tubocurarine, attempted crystallization of the active constituent from an extract of 1000 worms. Although this was unsuccessful, the solubility of “amphiporine” activity in chloroform under basic but not acidic conditions indicated that it was a weakly basic compound (3). This was also consistent with Bacq’s inference that “amphiporine” was an alkaloid similar to nicotine. Thirty years elapsed before nemertine toxins were investigated again. During the intervening decades, many new isolation and analytical techniques had been introduced. These included chromatographic methods permitting isolation of even minute amounts of natural products, nuclear magnetic resonance (NMR) and mass spectroscopic techniques, and for peptides and proteins, sensitive amino acid analysis, and Edman sequencing methods. The author, as a graduate student, isolated the hoplonemertine alkaloid anabaseine, a nicotinoid compound possessing a biological and chemical profile similar to Bacq’s “amphiporine.” Related compounds were found in other hoplonemertines (4–8). In contrast, anoplan (physically unarmed) nemertines were found to contain peptide neurotoxins resembling the “nemertine” activity profile, thus explaining Bacq’s observation of a slow rate of “nemertine” activity dialysis (9). While almost 900 species of nemertines have already been described in the biological literature, it is almost certain that this relatively inconspicuous animal phylum conFrom: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

573

574

Kem

tains many more, as yet undescribed species, perhaps several times this number. While no nemertine fossils have been reported, this group of marine animals is thought to have evolved from the flatworms (Phylum Platyhelminthes) back in Precambrian times (>500 million years ago). Nemertines possess several evolutionary innovations relative to flatworms, including separation of sexes (they are dioecious rather than hermaphroditic), a closed circulatory system composed of pulsating blood vessels, and a unidirectional gastrointestinal system allowing digested food to be eliminated through an anus rather than by mouth (10). They also possess another important structure, namely a long mobile proboscis, which is primarily used to capture prey (Fig. 1A). The phylum is divided into two large systematic classes, based on whether the proboscis is armed (Fig. 1B) with a skin-puncturing stylet or is “unarmed.” The armed nemertines are called hoplonemertines, whereas the unarmed species are either paleonemertines (a group thought to represent a more primitive stage in nemertine evolution) or heteronemertines. While the hoplonemertines paralyze their prey (usually other worms or cructaceans, depending on the species), paleonemertine and heteronemertine toxins are likely to be used for defensive purposes only. Since the integuments of hoplonemertines as well as paleonemertines and heteronemertines contain toxins to repel predators, this suggests that they originated for defensive purposes, but in the case of hoplonemertines also became offensive toxins for prey capture. This chapter will mainly focus on the chemical and pharmacological properties of the few toxins that have been isolated and characterized until now. Their mechanisms of action, insofar as they are known, will then be described. Finally, we will consider the potential utility of the toxins as neurobiological research tools and models for drug design. 2. HOPLONEMERTINE TOXINS A plethora of alkaloids have been discovered in hoplonemertines, but only a small number of these compounds have as yet been isolated and studied (4–9). The major problem is one of collecting and identifying satisfactory amounts of biomass from which the compounds can be extracted in sufficient quantity for structural identification. In this section we will only describe compounds whose structures have been previously reported. 2.1. Anabaseine 2.1.1. Chemistry Anabaseine (Fig. 2) was first isolated from the Peregrine Hoplonemertine Paranemertes peregrina (4,5). This moderately large (length > 15 cm) species wanders over exposed surfaces at low tide, searching for its annelid prey in full view of potential predators. Anabaseine was later found in certain ant venoms (11). While the structure of anabaseine chemically resembles anabasine, a tobacco alkaloid, it differs from the latter compound in one important chemical bond: there is a double bond between the nitrogen atom in the otherwise saturated ring and the carbon atom, which also is connected to the pyridyl ring. An imine-enamine tautomerism makes the tetrahydropyridyl ring beta-carbon lie within the same plane as this alpha-carbon and the imine N. This system, in turn, is conjugated with the pi electrons of the pyridyl ring. There-

Nemertine Toxins

575

Fig. 1. (A) Hoplonemertines use proboscis toxins for prey capture as well as defense against potential predators. (A) The Chevron hoplonemertine, Amphiporus angulatus, attacking its prey, an amphipod crustacean. This hoplonemertine (maximum length about 10–15 cm) occurs along the Pacific and Atlantic coasts of North America. (B) A general diagram of the hoplonemertine median proboscis stylet apparatus. The mineralized stylet of this apparatus is used to puncture the skin of the prey, thus allowing pyridyl alkaloid toxins produced in the glandular epitheliuim of the anterior proboscis and stored in the posterior chambers to readily enter the crustacean. The actual mechanism by which the venom exits the posterior chamber and enters the victim is not yet clear. The stylet is often lost during prey capture, but is readily replaced with another stylet kept in one of the two stylet accessory pouches. The integument covering the rest of the worm is continuous with the secretory epithelium of the anterior proboscis and also produces and secretes toxins used as a chemical defence against predators. Another group of hoplonemertines, the Polystyliferans, possesses multiple stylets for use in attacking prey (18a).

fore, the two rings of anabaseine are approximately coplanar. This contrasts with nicotine and anabasine, whose two rings are approximately at right angles with respect to each other in aqueous solution. Anabaseine was first obtained as a intermediate in the synthesis of anabasine by two Austrian tobacco chemists (12). This classical method generally provided anabaseine in a very low yield (5,13). Subseqently, several modifications were made to provide a more efficient synthesis and isolation (14,15); these papers should be consulted by those wishing to prepare the compound, which is not commercially available. The protection of the piperidone nitrogen can be accomplished with a variety of chemical groups. Direct crystallization of the ammonium-ketone open-chain form as a

576

Kem

Fig. 2. Covalent structures of nicotine, anabaseine (Paranemertes toxin), 3-(2,4dimethoxybenzylidene)-anabaseine (also called GTS-21 or DMXBA), and DMAC-anabaseine. While both nicotine and anabaseine stimulate most nicotinic receptors, GTS-21 and DMACanabaseine only stimulates the α7-type nicotinic receptors occurring in the mammalian brain. GTS-21 is currently undergoing clinical tests for possible use in treating neurodegenerative diseases (26,28).

dihydrochloride salt from the final reaction was found to be much more efficient than the older workup methods. Synthetic anabaseine dihydrochloride obtained in this manner exists as the ammonium-ketone form. While stable as the dried salt, to avoid any decomposition aqueous solutions of the toxin should be refrigerated when not in use and replaced after 1–2 wk. The cationic forms of anabaseine are quite soluble in protic solvents such as water, methanol, and ethanol, but the less hydrophilic free base is best dissolved in nonaqueous solvents such as alcohols, acetone, or ethyl acetate. Anabaseine occurs in several different forms in the presence of water (4). An NMR investigation demonstrated that at neutral pH there are three main forms present in roughly equal concentrations (16). These are the free base (cyclic imine), the monocationic cyclic iminium, and the monocationic ammonium-ketone. This multiplicity of forms complicated our initial attempts at identifying the pharmacologically active form that interacts with nicotinic receptors. Thus, stable analogs of these three forms were prepared so that their individual pharmacological properties could be examined. The fully aromatized free-base analog 2,3'-bipyridyl can be expected to possess a chemical conformation similar to the cyclic imine form of anabaseine, while 2-(3,4,5,6-tetrahydropyrimidinyl)-3-pyridine (called PTHP) was selected as an appropriate analog for the cyclic iminium form. To obtain stable open-chain forms of anabaseine, the open-chain nitrogen of anabasine was di- or tri-methylated. Since only PTHP displayed an ability to contract skeletal muscle and to bind to brain nicotinic receptors, we concluded that the mono-protonated cyclic iminium species is the only form of anabaseine that possesses significant affinty for the nicotinic receptor (17). 2.1.2. Pharmacology

Anabaseine stimulates a variety of vertebrate nicotinic receptors, like nicotine (18). However, it preferentially stimulates nicotinic receptors, namely skeletal muscle and brain α7 subtypes, which display high affinities for the snake toxin α-bungarotoxin (Table 1). Nicotine preferentially stimulates other neuronal nicotinic receptors that are

Nemertine Toxins

577

Table 1 Comparison of the Relative Efficacies of Anabaseine, Nicotine, and GTS-21 on Several Vertebrate Nicotinic Receptors Receptor type Central α7 (Rat) α4-ß2 (Rat) Peripheral Sympathetic (Rat PC12) Skeletal muscle (Frog)

Anabaseine

Nicotine

GTS-21

Full agonist

Weak partial agonist

Partial agonist

Weak partial agonist

Strong partial agonist

Antagonist

Full agonist

Full agonist

Weak antagonist

Full agonist

Full agonist

Weak antagonist

Table data summarized from results from refs. 18 and 28.

involved in its euphoric action in the brain. Anabaseine is one of the most potent neurotoxin nicotinic agonists; only epibatidine, anatoxin, and leptodactyline are more potent, when the ionized forms of these compounds are compared (Table 2). It was previously established that the monocationic form of nicotine stimulates the neuromuscular receptor (19). Patch-clamp analysis of anabaseine action on BC3H cell neuromuscular nicotinic receptors showed that anabaseine’s efficacy is comparable with that of ACh; thus it may be considered a full agonist on the neuromuscular type receptor (18). Analysis of the single channel openings provided evidence that at relatively high concentrations, anabaseine also is a channel-blocker (Fig. 3). While nicotine with high potency stimulates central neuronal receptors containing ß2 subunits, anabaseine displays a relatively low potency for stimulating these central receptors that have been implicated in tobacco addiction as well as cognitive function. Anabaseine was only a weak partial agonist at the rat α4-ß2 subtype of receptor, but a full agonist at the brain α7 subunit containing receptor (Fig. 4). The prolonged timecourses of the ionic currents generated by anabaseine or nicotine, relative to that of ACh, suggests that these nicotinoid compounds also act as channel-blockers as well as agonists at this receptor subtype. Nicotine was a partial agonist (relative to the natural agonist ACh) at both of the major brain nicotinic receptors, but its maximum effect on the α4-ß2 subtype was much greater than upon the α7 receptor. Since nicotine also binds to α4-ß2 receptors at much lower (about 100-fold) concentrations than at α7 receptors, its in vivo effects at smoking concentrations seem to be mediated primarily through the ß2 subunit-containing receptors. The whole animal (mouse) toxicity of anabaseine is very similar to that of nicotine (6). Because of its lack of receptor selectivity, few in vivo studies have been carried out with anabaseine. Meyer et al. (20) found that anabaseine improved passive avoidance in nucleus basalis-lesioned rats. We found that, when injected into the lateral ventricle

578

Kem

Table 2 Relative Potencies of Nicotinic Agonists on the Frog Rectus Abdominis Musclea Compound

EC50 (µM)

pKa

EC50,I (nM)

Epibatidine (+)-Anatoxin-a Leptodactyline Anabaseine Acetylcholine (S)-Nicotine Cytisine Carbamylcholine (S)-Anabasine

0.018a 0.067 0.12 0.74 0.53 1.96 6.70 7.38 7.05

9.3b 9.3 None NA None 7.9 7.9g None 8.7

0.018 0.066 0.12 0.25 0.53 1.63 5.56 7.38 6.83

Relative potency (EC50Carb/EC50,I) 410 112 62 30 14 4.5 1.3 1.0 0.93

In the column on the far right, the potencies are calculated assuming that the mono-cationic (I) form of each compound is solely active. The median effective concentration (EC50) values are calculated for a pH of 7.2. Adapted with permission from ref. 18.

Fig. 3. Nicotinic agonist activity of anabaseine and ACh on BC3H-1 cells. Data were recorded using the cell-attached voltage-clamp method. Groupings of openings activated by either ACh (A) or anabaseine (B) are shown. Groups for analysis were selected as defined in the Methods. Qualitatively, closed intervals within groups of openings become shorter with increases in agonist concentration. With anabaseine, open intervals become shorter with increases in agonist concentration as a result of channel block by anabaseine, and an increase in frequency of a short duration gap is apparent. Adapted with permission from ref. 18.

Nemertine Toxins

579

Fig. 4. Agonist actions of anabaseine and anabasine on rat brain nicotinic receptors expressed in Xenopus oocytes. (A) Responsiveness of the α7 receptor. Note that nicotine has a relatively low efficacy for stimulating this receptor subtype. (B) Responsiveness of the α4-ß2 receptor. On this receptor nicotine is a much more potent and efficaceous agonist than anabaseine or anabasine. Agonist responses were normalized to the individual oocyte’s response to a control ACh (500 µM) application made 5 min before the compound application. Each point represents the average response (± S.E.) of at least four oocytes. Adapted with permission from ref. 18.

580

Kem

of the rat brain, anabaseine elicited the same prostration behavior as many other nicotinic agonists (18). The significantly lower potency of anabaseine, relative to nicotine, in causing prostration is consistent with the notion that ß2 subunit-containing nicotinic receptors primarily mediate this behavior. In the rat fronto-parietal cortex, anabaseine elevated both norepinephrine and acetylcholine levels without affecting serotonin and dopamine (21). The noncompetitive nicotinic antagonist mecamylamine inhibited the anabaseine elevation of these two neurotransmitters (Fig. 5). These central anabaseine actions were most likely mediated through high-affinity nicotinic receptors containing beta2 subunits, whose channels are much more sensitive to mecamylamine blockade than are those of the α7 receptors. Anabaseine also affects a variety of invertebrate nicotinic receptors. Marine annelids, the usual prey of Paranemertes, are paralyzed, as are crustaceans and insects; it is assumed that these responses result from stimulatory actions upon nicotinic receptors. In these organisms nicotinic cholinergic receptors primarily reside on central neurons without readily recognizable soma, which makes experimental analysis of toxin effects more difficult. 2,3'-bipyridyl, a largely nonionized analog of anabaseine, is even more active than anabaseine in paralyzing crustaceans (6). While it does not cause paralysis, nemertelline (a tetrapyridyl found in Amphiporus angulatus), in common with anabaseine and 2,3'-bipyridyl, stimulates an unusual receptor in the stomatogastric muscle of the crayfish, which is apparently a chloride channel (22). At present this is the only known action of this complex alkaloid, which is the most abundant pyridine in this species of Amphiporus. Nemertine body-wall muscles (including those of the heteronemertine Cerebratulus) also contain nicotinic receptors, but they only respond to extremely high concentrations of anabaseine. Thus a natural resistance to this toxin may be advantageous to hoplonemertines that produce anabaseine or related compounds (23). Anabaseine also affects molluscan ganglionic nicotinic receptors, some of which are chloride channels. The ganglionic receptors are of three major types: chloride channels that either rapidly or slowly desensitize and cation channels that desensitize rather slowly. While anabaseine primarily blocks the very fast desensitizing receptor-chloride channel, it activates the sustained chloride channel response as well as the cation channel. In contrast, the anabaseine derivative DMXB anabaseine transiently activates and subsequently blocks the rapidly desensitizing chloride channel, at concentrations that do not affect the slowly inactivating chloride and cation channels. This pattern is consistent with its selective agonistic action on α7 nicotinic receptors in vertebrates. A variety of pyridine compounds including anabaseine and 2,3'-bipyridyl (see Subheading 2.3.) stimulate chemoreceptor neurons in crayfish and spiny lobster walking legs (25; Hatt, Ache, and Kem, unpublished results). Observations of feeding behavior in marine aquaria indicated that spiny lobsters attack but subsequently reject living Amphiporus angulatus. Anabaseine and 2,3'-bipyridyl were found to be two of the most active compounds in stimulating similar pyridine receptors on spiny lobster antennule nerves (22). We suspect that nemertine alkaloids, by acting upon these chemoreceptors, may act as repellants against certain predators. 2.2. DMXB-Anabaseine (GTS-21): A Synthetic Anabaseine Drug Candidate While anabaseine is a broad spectrum nicotinic agonist, a variety of 3-substituted anabaseines have been found to possess greater nicotinic receptor selectivity (21; Kem,

Nemertine Toxins

581

Fig. 5. Microdialysis study of anabaseine effects upon neurotransmitter levels in the rat frontoparietal cortex. Anabaseine was administered alone (left side) or 90 min after pretreatment (4.9 µmol/kg or 1.0 mg/kg, i.p.) with the noncompetitive antagonist mecamylamine (right side).The s.c. dose was 3.6 µmol/kg for each compound (0.90 mg/kg for anabaseine. Data are expressed as a percentage of the pre-injection control levels (average of the six samples prior to injection = 100%); mean ± S.E.M., n = 6, *p < 0.05 by paired Students t-test analysis. Adapted with permission from ref. 21.

582

Kem

W. R., et al., in preparation). Here we shall only consider 3-(2,4-dimethoxybenzylidene) -anabaseine, whose pharmaceutical code name is GTS-21 (26–28). This compound is of special interest because it has been shown to be a neuroprotective agent in stroke and amino acid neurocytotoxicity models as well as a “cognition enhancer” in aged and brain lesioned animals. Initial (Phase I) tests in humans demonstrated the lack of toxicity of the compound and also indicate improved cognitive function in healthy young adults (29). 2.2.1. Chemistry GTS-21 (Fig. 2) is readily prepared by reaction of 2,4-dimethoxybenzaldehyde with anabaseine in acidic alcohol at elevated temperature, in a manner similar to the preparation of 3-(4-dimethylaminobenzylidene)-anabaseine (13,30). The resulting product can be precipitated and recrystallized using less polar solvents. The 3-arylideneanabaseines do not hydrolyze to open-chain forms at physiological pH like anabaseine, but they are moderately photolabile and must thus be protected from strong light. In principle, the benzylidene ring of such a compound can adopt two possible conformations with respect to the tetrahydropylridyl ringt, namely E (entegegen) or Z (zusamenfassung). By NMR we have shown that the E form is preferred in aqueous solution (16). Only in the presence of intense light does the E to Z conversion become significant. The Z-form does not display significant affinity for the α4-ß2 receptor (Kem et al., unpublished results). While the two rings in anabaseine are essentially co-planar, in the 3-benzylideneanabaseines these two rings as well as the benzylidene ring are predicted to lie in different planes (30). Perhaps the lack of agonist activity of GTS-21 upon certain nicotinic receptor subtypes that are quite sensitive to anabaseine, such as α4-ß2 and muscle 2.2.2. Pharmacology In contrast to anabaseine, DMXB-anabaseine is only agonistic upon one known nicotinic receptor subtype, the neuronal homo-oligomeric α7 receptor, which is primarily found in the brain. It also acts as an antagonist at the brain α4-ß2 nicotinic receptor. Only at much higher concentrations does it act as a weak antagonist at other peripherally located nicotinic receptors. What makes this compound of considerable interest is its selective stimulation of a central nervous system (CNS) receptor whose physiological function has been very difficult to investigate in the past. Initially the α7 receptor could only be recognized by its ability to bind α-bungarotoxin (BTX). Only later, after cloning and expression in cultured cells, was it found to be physiologically active as a ligand-gated ion channel with high permeability for calcium ions. The compound is a partial agonist at this receptor, since it displays an efficacy approximately half that of ACh for stimulating this receptor when expressed in the Xenopus oocyte (Fig. 6). Another cinnamylidene-anabaseine derivative, DMAC-anabaseine, is about as efficaceous as ACh (27). 2.2.3. Behavioral Effects It has now been over a decade since several laboratories reported large (as much as 50%) decreases in nicotinic cholinergic receptors in Alzheimer’s patients (31). This stimulated considerable academic and pharmaceutical interest in the development of nicotinic agonists that could selectively stimulate the remaining brain nicotinic receptors involved in cognitive and other critical mental functions. Its effects upon cognitive

Nemertine Toxins

583

Fig. 6. Comparison of the effects of DMXB-anabaseine and DMAC-anabaseine with ACh action on the rat α7 nicotinic receptor. Responses were normalized to the responsiveness of the oocyte to 500 µM ACh applied 5 min earlier. The curves were obtained with Kaleidograph sotfware. Each point represents the mean ± 1SE of the response from four oocytes (anabaseine compounds) or three oocytes (Ach). Adapted with permission from ref. 27.

behavior have been investigated by four different laboratories using a variety of mammals. Initially it was observed that the compound enhanced passive avoidance performance in rats (20), active avoidance (Fig. 7) in aged rats (32), and acquisition of conditioned eye-blink reflex (Fig. 8) in aging rabbits (33). More intricate learning tasks such as water- and radial-maze performance by rats (34) and delayed matching by monkeys (35) were also enhanced, which suggests that the compound may also be able to enhance cognition in aging humans, particularly Alzheimer’s patients. The compound is currently in clinical trials. DMXB-anabaseine, like nicotine, has been demonstrated to enhance auditory gating in mice (36). Since this action of both compounds is prevented by prior administration of BTX, α7 receptors probably mediate this action. It is also a moderately potent antagonist at mouse 5-HT3 receptors, which display a high degree of homology with α7 nicotinic receptors (37). 2.2.4. Neuroprotection DMXB-anabaseine displays neuroprotective actions upon differentiated pheochromocytoma (PC12) cells stressed by nerve growth-factor depletion (38), and neocortical primary cultures exposed to glutamate (39,40) or ß-amyloid (41). This action seems to be mediated through α7 nicotinic receptors since it can be inhibited by α-bungarotoxin or mecamylamine. Neuroprotection by DMXB-anabaseine has also been observed in rats (42,43). Thus, this drug candidate could delay the process of neurodegeneration as well as ameliorate some of the cognitive deficits in neurodegenerative diseases like Alzheimer’s.

584

Kem

Fig. 7. Enhancement of one-way active avoidance acquisition in aged Sprague-Dawley rats pretreated with GTS-21 (1 mg/kg), nicotine (NIC; 0.2 mg/kg) or saline vehicle 15 min prior to daily testing for 12 d. The number of animals in each group is indicated in parentheses. Asterisks at individual 3-d time blocks indicate a significantly greater percentage of conditioned avoidance responses (CARs) for that group compared to the % CAR’s exhibited by aged control rats for that block (p < 0.05 or higher level of significance). Adapted with permission from ref. 32.

2.3. Bipyridyl and Other Hoplonemertine Alkaloids The Chevron Nemertine (Amphiporus angulatus) is relatively common in the cold coastal waters of both the Atlantic and Pacific coasts of North America. It produces a wide variety of pyridyl alkaloids (6,7,22). Anabaseine is only a relatively minor constituent in this species, which feeds upon crustaceans, in contrast with the anabaseinerich, vermiferous Paranemertes peregrina. The major alkaloids are 2,3'-bipyridyl and a tetrapyridyl alkaloid named nemertelline (Fig. 9). While the crustacean paralyzing activity of the former compound is even greater than that of anabaseine, nemertelline lacks an obvious toxic activity upon mammals (mice) as well as crustaceans. In addition to these compounds we have recently identified a new dihydroisoquinoline alkaloid and an isomer of anabaseine, 2-(3-pyridyl)-1,2,5,6-tetrahydropyridine (Soti and Kem, in preparation). Many other compounds are present in smaller amounts. Since some of these compounds may be intermediates in the biosynthesis of the most abundant compounds, knowledge of their structures may assist in constructing potential biosynthetic pathways for these alkaloids. The crustacean paralyzing activity of 2,3'-bipyridyl is particularly interesting because this molecule apparently acts in its nonionized form in this group of invertebrates. Since its most basic nitrogen pKa is 4.4, only about 1 in 10,000 molecules will exist as monocations at physiological pH (44). This leads one to ask whether this substance acts upon nicotinic receptors, since all known forms of this receptor so far studied, invertebrate as well as vertebrate, seem to require that a stimulatory molecule contains a cationic group. Because crustacean nicotinic receptors are centrally located and radioligand receptor-binding investigations have yet to be done on these receptors, it is still only an assumption that this bipyridyl is acting on such receptors. Since 2,3'bipyridyl also acts as a feeding repellent for spiny lobsters when incorporated into agar

Nemertine Toxins

585

Fig. 8. Enhancement of eyeblink conditioning in aged rabbits and lack of nicotinic receptor upregulation by DMXBA. (A) on the left indicates the number of paired stimulus trials required to learn the eyeblink reflex at three different doses (subcutaneous) of GTS-21, relative to a saline control group (n = 8 animals for each group). (B) on the right shows measurements of the concentration of high nicotine affinity receptors in cerebral-cortex samples from the same rabbits. No statistically significant difference in receptor concentrations were observed between any of the GTS-21 treated animal groups and the control group. Adapted with permission from ref. 33.

gel blocks containing a feeding attractant, it may serve as a feeding deterrent as well as a defensive toxin. It is highly likely that nemertelline also has some action on potential predators or prey. One possible site of its action would be the skeletal muscles composing the gastric-mill apparatus in the crustacean stomach. These skeletal muscles possess chloride ion-permeable ion channels that are activated by most commonly used nicotinic cholinergic agonists. Prolonged activation of these depolarizing chloride currents by 2,3'bipyridyl and nemertelline would be expected to block the initial grinding process that breaks up ingested organisms and other food materials. Further studies of nemertelline are planned, now that a method of laboratory synthesis has been worked out (45). 3. HETERONEMERTINE NEUROTOXINS Many heteronemertines possess peptide neurotoxins (8,9). However, the only neurotoxins that have been isolated to date belong to a large (>1 m, 20 g) Atlantic coast species, Cerebratulus lacteus. The other group of anoplans, paleonemertines, thus far have been found to contain only protein cytolysins (Kem, unpublished results). 3.1. Cerebratulus Neurotoxins 3.1.1. Chemistry The so-called Cerebratulus B neurotoxins have molecular sizes of approx 6000, similar to the scorpion neurotoxins, and are crosslinked by three disulfide bonds (46). The sequences (Fig. 10) of the two most abundant and active isotoxins, B-II and B-IV, were reported some time ago (47,48). Both are very basic peptides and contain a single

586

Kem

Fig. 9. Molecular structures of the major pyridyl alkaloids of the hoplonemertine Amphiporus angulatus. 2,3'-Bipyridyl (left side) is the major neurotoxin of this nemertine, while the most abundant alkaloid is nemertelline (right side). Nemertelline affects crustacean stomach muscle nicotinic receptors but is not toxic (Kem and Soti, submitted). The structure of nemertelline was recently revised (45).

residue of hydroxyproline at position 10. Unlike the scorpion and sea anemone peptide toxins whose secondary structures are largely composed of anti-parallel B-strands, the B toxins are devoid of B-sheet structure but rich in α-helix (49). The secondary and tertiary structures of B-IV were determined by NMR (50,51). It can be observed that there are two long stretches of helix, represented by positions 11-23 and 34-49. The two helices are connected by a loop consisting of two inverse γ-turns and a ß-turn. The entire sequence, 11-49, thus constitutes a rather unique helical hairpin structure (Fig. 11). 3.1.2. Pharmacology The two B toxins are potent toxins when injected into crustaceans, especially crayfish. Initially the animal displays tremors (including flipping of the tail), but then convulses in a massive contracture of the limbs and tail. In a few minutes the contractural paralysis is replaced by flaccid paralysis and eventually death. This toxicity of the Cerebratulus B toxins (and unpurified Lineus toxins) seems confined to the crustacean nervous system. The limb contracture can also be observed in the perfused crayfish cheliped. The nerve terminals seem to be the site of action, since tetrodotoxin effectively blocks the action of toxin B-IV, even when the peptide is applied directly on the muscle. At relatively high (0.1–1.0 µM) concentrations toxin B-IV also affects the evoked compound action potential recorded from isolated crab-leg nerves, causing some repetitive spiking, which also probably contributes to the prolongation of the compound action potential (Kem, W. R., unpublished results). The current conception of how the toxins act is that they activate a small population of sodium channels, which leads to neuronal repetititive spiking, which causes a massive release of excitatory (and perhaps inhibitory) neurotransmitters at neuromuscular synapes. In arthropods the skeletal muscles are generally electrically inexcitable, but are innervated along their entire length by excitatory glutamatergic and inhibitory GABAergic synapes. A more intensive analysis of their actions using intracellular or patch clamp recording techniques would certainly be desirable. B-IV was inactive when tested on a variety of isolated axons and neurons from various noncrustacean groups including vertebrates and molluscs (46).

Nemertine Toxins

587

Fig. 10. Amino acid sequences of the two Cerebratulus lacteus neurotoxins B-II and B-IV. Identical sequence regions are enclosed within boxes. Adapted with permission from refs. 47 and 48.

3.1.3. Receptor Binding Studies Lieberman and Blumenthal (52) measured the binding of iodinated toxin B-IV to membranes prepared from Maine lobster (Homarus vulgaris) muscle. High-affinity binding was observed. The specific binding could not be displaced by scorpion α-toxin. Further experiments are required to identify the membrane receptor involved. While they suceeded in chemically crosslinking the toxin to the membranes, the solubilized membrane proteins labeled by the iodinated toxin possessed a smaller molecular size (about 40,000) than would be predicted for the sodium channel α-subunit, which contains the binding site for scorpion and sea anemone peptide toxins. This result could be explained in several ways. Membrane proteins can be degraded by endogenous proteases unless a cocktail of inhibitors was added. This seems unlikely to be the reason here, since several inhibitors were added to the incubation saline. Other alternative explanations are that the toxin either binds predominantly to a B subunit of the sodium channel or that it interacts with some other membrane protein, perhaps another ion channel. 3.1.4. Structure-Action Relationships Considerable data is available implicating some amino acid sidechains in the toxic action of toxin B-IV. The initial studies utilized a chemical modification approach and focused on the few aromatic residues (2 tyrosyls, 2 tryptophanyls). By manipulating the conditions of the reactions, it was possible to differentially label the two Tyr and two Trp residues, the former by nitration (53) and the latter by alkylation (54). By bioassaying the toxin samples at different degrees of modification, it was deduced that Tyr9 and Trp30 are probably involved in receptor binding, since their modification did not affect secondary structure as measured by CD spectroscopy. During the past decade Blumenthal’s lab has utilized molecular biological methods to express B-IV in Escherichia coli and to obtain mutants for structure-activity relationship (SAR) studies (55,56). Initial experiments showed that replacement of the

588

Kem

Fig. 11. An NMR-derived solution backbone structure of Cerebratulus neurotoxin B-IV. Practically the entire peptide consists of a helical hairpin structure crosslinked by four disulfide bonds. Adapted with permission from ref. 51.

hydroxyproline at position 10 with Pro did not affect crayfish paralytic activity, nor did replacement of Ala residues at either position 3 or 8 with serine. In fact, the toxicity of B-IV was enhanced by simultaneous substitution of serine at positions 3 and 8. Arg17 has been implicated in receptor binding as toxicity was undetectable but the CD spectrum unchanged when glutamine, Ala or Lys was substituted. Replacement of Arg25 with Lys reduced crayfish toxicity 400-fold (57). Some of the side chains implicated in toxicity of B-IV include the guanidinyl side chains of Args 17, 25, and 34, and the aromatic side chains of Trp 30 and Tyr 9. It was rather surprising that the implicated residues are found along the entire length of one surface of the toxin. This implies that the toxin binds to an extensive portion of receptor surface. Alternately, modification of certain residues such as Tyr9 and Trp30 may have altered the folded structure sufficiently to deleteriously affect activity without affecting the CD spectrum. Experiments with other toxin mutants, coupled with more intensive NMR structural analyses of tertiary structure should provide further insights regarding the binding surface of this toxin. Since a scorpion α-toxin did not inhibit the Cerebratulus toxin binding, the latter probably binds to another site (52). 3.2. Lineus Neurotoxins All species of this large heteronemertine genus so far examined have been found to be quite toxic (8,58,59). These toxins are also low molecular weight peptides of 3000–

Nemertine Toxins

589

6000 Daltons (4). While they paralyze crustaceans in a manner indistinguishable from the Cerebratulus toxins, the Lineus toxins primarily prolong action potential duration in crustacean neurons, whereas Cerebratulus toxin B-IV causes repetitive spiking. Isolation and characterization of the Lineus toxins is currently being attempted in the author’s laboratory. 4. HETERONEMERTINE AND PALEONEMERTINE CYTOLYSINS Lytic toxins are practically ubiquitous among all living organisms, including bacteria, plants, and animals. Almost all animal venoms include some lytic substances that enhance the penetration of the other toxins into the circulation of the affected organism and possibly potentiate the actions of certain toxins. So far, cytolytic proteins have only been found in the anoplan nemertines. Cerebratulus contains at least four homologous protein lysins called A toxins (60). The most abundant isotoxin, A-III, was sequenced and shown to contain three disulfide bonds (61). CD and Raman spectroscopic analyses of this toxin revealed the presence of approx 60% α-helix and 10% B-sheet (Kem, W. R., et al., in preparation). The Cterminal portion that is not cross-linked by a disulfide bond is thought to exist as a helical hairpin structure. Because of its amphipathic nature, this region of toxin sequence may interact with membrane lipids and possibly participate in the formation of a pore. At sublytic concentrations Cerebratulus toxin A-III blocks the squid axon sodium channel and affects the kinetics of opening and closing of voltage-gated potassium channels. Then it increases the so-called leakage conductance, which probably reflects a dimunition of membrane integrity (58). These differential effects on sodium vs potassium channels are interesting and deserve further study. Perhaps due to its ability to disrupt lipid bilayers, toxin A-III also inhibits brain phospholipid sensitive Ca2+dependent protein kinase in vitro, as do other peptide toxins that possess a detergentlike action (61). Also, these proteins may be neurotoxic at lower, sublytic concentrations and potentiate the actions of the smaller neurotoxic B peptide toxins. 5. CONCLUDING COMMENTS It is predicted that systematic investigations of other nemertine species will reveal a plethora of molecules used as toxins by this relatively unstudied phylum. Besides providing unique tools for biomedical research, some of the toxins (particularly alkaloids like anabaseine) may become molecular models for drug design (63,64). The toxins will also provide insights into chemical mechanisms and structures. For instance, the NMR investigation of anabaseine has provided an improved understanding of the stability of tetrahydropyridine compounds under aqueous conditions. Norton’s NMR solution structural analysis of Cerebraulus toxin B-IV has revealed a paired helical structure that is so far unique among the known neurotoxins. Many invertebrate toxins are primarily directed towards other invertebrates; that is, they have evolved over time to deal with potential invertebrate predators or prey. Study of their actions upon invertebrate nervous system receptors may ultimately lead to the design of more selective pesticides and anti-parasitic drugs. Ultimately it is expected that homologous toxins will be found that also target homologous receptors in the mammalian nervous system. For

590

Kem

instance, although the narrow phylogenetic activity of the Cerebratulus B neurotoxins presently limits their utility as molecular probes to crustacean nervous systems, it seems likely that natural or synthetic variants of these toxins will eventually be found that act on homologous ion channels of mammals and other vertebrates. ACKNOWLEDGMENTS The author is grateful to many students, colleagues, and technical personnel for their contributions to this research, which has spanned the past three decades. Barbara Seymour prepared the drawings in Fig. 1. Many of the author’s investigations of nemertine neurotoxins were supported by NSF, NIH, or Taiho Pharmaceutical Company grants. REFERENCES 1. Bacq, Z. M. (1936) Les poisons des Nemertiens. Bull. Acad. R. Belg. Cl. Sci. (Ser. 5) 22: 1072–1079. 2. Bacq, Z. M. (1937) L”amphiporine” et la “nemertine,” poisons des vers nemertiens. Arch. Int. Physiol. 44, 190–204. 3. King, H. (1939) Amphiporine, an active base from the marine worm Amphiporus lactifloreus. J. Chem. Soc. 1365. 4. Kem, W. R. (1969) A chemical investigation of nemertine toxins. Ph.D. dissertation, University of Illinois, Urbana. 5. Kem, W. R., Abbott, B. C., and Coates, R. M. (1971) Isolation and structure of a hoplonemertine toxin. Toxicon 9, 15–22. 6. Kem, W. R., Scott, K. N., and Duncan, J. H. (1976) Hoplonemertine worms: a new source of pyridine neurotoxins. Experientia 32, 684–686. 7. Kem, W. R. (1988) Pyridine alkaloid distribution in the hoplonemertinea. Hydrobiologia 156, 145–153. 8. Kem, W. R. (1971) A study of the occurrence of anabaseine in Paranemertes and other nemertines. Toxicon 9, 23–32. 9. Kem, W. R. (1988) Worm toxins, in Handbook of Natural Toxins, vol. 4, Marine Toxins and Venoms (Tu, ed.), Marcel Dekker, New York, pp. 253–378. 10. Gibson, R. (1972) Nemerteans. Hutchinson University Library, London, pp. 224. 11. Wheeler, J. W., Olubajo, O., Storm, C. B., and Duffield, R. M. (1981) Anabaseine: venom alkaloid of Aphaenogaster ants. Science 211, 1051–1052. 12. Spath, E. and Mamoli, L. (1936) Eine Neue Synthese Des D,L-Anabasins. Chem. Ber. 69, 1082–1085. 13. Kem, W. R. (1973) Biochemistry of nemertine toxins, in Marine Pharmacognosy: Marine Biotoxins as Probes of Cellular Function (Martin, D. F. and Padilla, G. M., eds.), Monographs on Cell Biology Series, Academic Press, NY, pp. 37–84. 14. Bloom, L. B. (1990) Influence of solvent on the ring-chain hydrolysis equilibrium of anabaseine and synthesis of anabaseine and nicotine analogues. Ph.D. dissertation, University of Florida, Department of Chemistry. 15. Zoltewicz, J. A. and Cruskie, M. P., Jr. (1995a) A superior synthesis of cholinergic anabaseine. OPPI Briefs 27, 510–513. 16. Zoltewicz, J. A., Bloom, L. B., and Kem, W. R. (1989) Quantitative determination of the ring-chain hydrolysis equilibrium constant for anabaseine and related tobacco alkaloids. J. Org. Chem. 54, 4462–4468. 17. Kem, W. R., Mahnir, V. M., Bloom, L. B., and Gabrielson, B. J. (1994) The active form of the nicotinic receptor agonist anabaseine is the cyclic iminium cation. 11th World Congress on Animal, Plant, and Microbial Toxins, Tel Aviv.

Nemertine Toxins

591

18. Kem, W. R., Mahnir, V. M., Papke, R., and Lingle, C. (1997) Anabaseine is a potent agonist upon muscle and neuronal alpha-bungarotoxin sensitive nicotinic receptors. J. Pharmacol. Exper. Therap. 283, 979–992. 18a.Gibson, R. (1972) Nemerteans. Hutchinson University Library, London, pp. 224. 19. Barlow, R. B. and Hamilton, J. T. (1962) The effects of pH on the activity of nicotine and nicotine methiodide on the rat diaphragm preparation. Br. J. Pharmacol. 18, 543–549. 20. Meyer, E. M., deFiebre, C. M., Hunter, B. E., Simpkins, C. E., and deFiebre N. E. (1994) Effects of anabaseine related analogs on rat brain nicotinic receptor binding and on avoidance behavior. Drug Dev. Res. 31, 135–141. 21. Summers, K., Kem, W. R., and Giacobini, E. (1997) Nicotinic agonist modulation of neurotransmitter levels in the rat frontoparietal cortex. Jpn. J. Pharmacol. 74, 139–146. 22. Kem, W. R. and Soti, F. (2001) Amphiporus alkaloid multiplicity implies functional diversity: Initial studies on crustacean pyridyl receptors. Hydrobiolog. In press. 23. Kem, W. R. (1997b) Nemertine body wall and proboscis longitudinal muscles possess unique nicotinic receptors. Fifth Intern. Conf. Invertebrate Neurochem. Neurophysiol. Mtg, Eilat, Israel (Abstr.). 24. Kehoe, J. S. and Kem, W. R. (2001) Anabaseine and GTS-21 act differentially on two alpha-bungarotoxin-sensitive Aplysia nicotinic receptors. Toxicon, in press. 25. Hatt, H. and Schmiedel-Jacob, I. (1984) Electrophysiological studies of pyridine-sensitive units on the crayfish walking leg. I. Characteristics of stimulatory molecules. J. Comp. Physiol. 154A, 855–863. 26. Kem, W. R. (1998) Alzheimer’s drug design based upon an invertebrate toxin (anabaseine) which is a potent nicotinic receptor antagonist. Invertebr. Neurosci. 3, 251–259. 27. De Fiebre, C. M., Meyer, E. M., Henry, J. C., Muraskin, S. I., Kem, W. R., and Papke, R. L. (1995) Characterization of a series of anabaseine-derived compounds reveals that the 3-(4)-Dimethylaminocinnamylidine derivative (DMAC) is a selective agonist at neuronal nicotinic alpha 7/[125K] alpha-bungarotoxin receptor subtypes. Mol. Pharmacol. 47, 164–171. 28. Kem, W. R. (2000) The brain alpha7 nicotinic receptor may be an important therapeutic target for the treatment of Alzheimer’s disease: studies with DMXBA (GTS-21). Behav. Brain Res. 113, 169–183. 29. Kitagawa, H., Moriyama, A. A., Takenouchi, T., Wesnes, K., Kramer, W., and Clody, D. R. (1998) Phase I studies of GTS-21 to assess the safety, tolerability, PK and effects on measures of cognitive function in normal volunteers. Neurobiol. Aging 19, S182 (Abstr.). 30. Zoltewicz, J. A., Prokai-Tatrai, K., Bloom, L. B., and Kem, W. R. (1993) Long range transmission of polar effects of cholinergic 3-arylideneanabaseines. Conformations calculated by molecular modelling. Heterocycles 35, 171–179. 31. Whitehouse, R. J., Price, D. L., Clark, A. W., Coyle, J. T., and DeLong, M. R. (1986) Nicotinic acetylcholine binding in Alzheimer’s disease. Brain Res. 371, 146–151. 32. Arendash, G. W., Sengstock, G. J., Sanberg, R., and Kem, W. R. (1995) Improved learning and memory in aged rats with chronic administration of the nicotinic receptor agonist GTS21. Brain Res. 674, 252–259. 33. Woodruff-Pak, D. S., Li, Y.-T., and Kem, W. R. (1994) A nicotinic receptor agonist (GTS21), eyeblink classical conditioning, and nicotinic receptor binding in rabbit brain. Brain Res. 645, 309–317. 34. Bjugstad, K. B., Mahnir, V. M., Kem, W. R., and Arendash, G. W. (1996) Long-term treatment with GTS-21 or nicotine enhances water maze performance in aged rats without affecting the density of nicotinic receptor subtypes in neocortex. Drug Dev. Res. 39, 19–28. 35. Briggs, C. A., Anderson, D. J., Brioni, J. D., Buccafusco, J. J., Buckley, M. J., Campbell, J. E., et al. (1997) Functional characterization of the novel neuronal nicotinic acetylcholine receptor ligand GTS-21 in vitro and in vivo. Pharmacol. Biochem. Behav. 57, 231–241.

592

Kem

36. Stevens, K. E., Kem, W. R., Mahnir, V. M., and Freedman, R. (1998) Selective alpha7nicotinic agonists normalize inhibition of auditory response in DBA mice. Psychopharmacology 136, 320–327. 37. Machu, T. K., Strahlendorf, J., and Kem, W. R. (1996) Nicotinic receptor ligands antagonize 5-HT3 receptors expressed in Xenopus oocytes. J. Neurosci. 22, 1780 (Abstr.). 38. Martin, E. J., Panickar, K. S., King, M. A., Deyrup, M., Hunter, B. E., Wang, G., and Meyer, E. M. (1994) Cytoprotective actions of 2,4-dimethoxybenzylidene anabaseine in differentiated PC12 cells and septal cholinergic neurons. Drug Dev. Res. 31, 135–141. 39. Akaike, A., Tamura, Y., Yokota, T., Shimohama, S., and Kimura, J. (1994) Nicotine-induced protection of cultured cortical neurons against N-methyl-D-aspartate recetpor-mediated glutamate cytotoxicity. Brain Res. 644, 181–187. 40. Shimohama, S., Greenwald, D. L., Shafron, D. H., Akaika, A., Maeda, T., Kaneko, S., et al. (1998) Nicotinic alpha7 receptors protect against glutamate neurotoxicity and neuronal ischemic damage. Brain Res. 779, 359–363. 41. Kihara, T., Shmohama, S., Sawada, H., Kimura, J., Kume, T., Kochiyama, H., et al. (1997) Nicotinic receptor stimulation protects neurons against B-amyloid toxicity. Ann. Neurol. 42, 159–163. 42. Nanri, M., Yamamoto, J., Miyake, H., and Watanabe, H. (1998) Protective effect of GTS21, a novel nicotinic receptor agonist, on delayed neuronal death induced by ischemia in gerbils. Jpn. J. Pharmacol. 76, 23–29. 43. Meyer, E. M., King, M. A., and Meyers, C. (1998) Neuroprotective effects of 2,4dimethoxybenzylidene anabaseine (DMXB) and tetrahydroaminoacridine (THA) in nerocortices of nucleus basalis lesioned rats. Brain Res. 786, 252–254. 44. Yamamoto, I., Kamimura, H., Yamamoto, R., Skai, S., and Goda, M. (1962) Studies on nicotinoids as insecticides. I. Relation of structure to toxicity. Agr. Biol. Chem. (Tokyo) 26, 709–716. 45. Zoltewicz, J. A. and Cruskie, M. P., Jr. (1995b) Strategies for the synthesis of unsymmetrical quaterpyridines using palladium-catalyzed cross-coupling reactions. Tetrahedron 51, 11,393–11,400. 46. Kem, W. R. (1976) Purification and characterization of a new family of polypeptide neurotoxins from the heteronemertine Cerebratulus lacteus (Leidy). J. Biol. Chem. 251, 4184–4192. 47. Blumenthal, K. M. and Kem, W. R. (1976) Primary structure of Cerebratulus lacteus toxin B-IV. J. Biol. Chem. 251, 6025–6029. 48. Blumenthal, K. M., Keim, P. S., Heinrikson, R. L., and Kem, W. R. (1981) Structure and action of heteronemertine polypeptide toxins. Amino acid sequence of Cerebratulus lacteus toxin B-II and revised structure of toxin B-IV. J. Biol. Chem. 256, 9063–9067. 49. Kem, W. R., Tu, C.-K., Williams, R. W., Toumadje, A., and Johnson, W. C., Jr. (1990) Circular dichroism and laser Raman spectroscopic analysis of the secondary structure of Cerebratulus lacteus toxin B-IV. J. Prot. Chem. 9, 433–443. 50. Hansen, P. E., Kem, W. R., Bieber, A. L., and Norton, R. S. (1992) 1H-NMR study of neurotoxin B-IV from the marine worm Cerebratulus lacteus. Solution properties, sequence-specific resonance assignments, secondary structure and global fold. Eur. J. Biochem. 210, 231–240. 51. Barnham, K. J., Dyke, T. R., Kem, W. R., and Norton, R. S. (1997) Structure of neurotoxin B-IV from the marine worm Cerebratulus lacteus: a helical hairpin cross-linked by disulphide bonding. J. Mol. Biol. 268, 886–902. 52. Lieberman, D. L. and Blumenthal, K. M. (1986) Structure and action of heteronemertine polypeptide toxins. Specific cross-linking of Cerebratulus lacteus toxin B-IV to lobster axon memebrane vesicles. Biochim. Biophys. Acta 855, 1–48. 53. Blumenthal, K. M. and Kem, W. R. (1980) Inactivation of Cerebratulus lacteus toxin B-IV by tyrosine nitration. Arch. Biochem. Biophys. 203,816–203,821.

Nemertine Toxins

593

54. Blumenthal, K. M. (1980) Inactivation of Cerebratulus lacteus toxin B-IV concomitant with tryptophan alkylation. Arch. Biochem. Biophys. 203, 822–826. 55. Howell, M. L. and Blumenthal, K. M. (1989) Cloning and expression of a synthetic gene for Cerebratulus lacteus neurotoxin B-IV. J. Biol. Chem. 264, 15,268–15,273. 56. Howell, M. L. and Blumenthal, K. M. (1991) Mutagenesis of Cerebratulus lacteus neurotoxin B-IV identifies NH2-terminal sequences important for biological activity. J. Biol. Chem. 266, 12,884–12,888. 57. Wen, P. H. and Blumenthal, K. M. (1996) Role of electrostatic interactions in defining the potency of neurotoxin B-IV from Cerebratulus lacteus. J. Biol. Chem. 271, 29,752–29,758. 58. Kem, W. R. (1994) Structure and membrane actions of a marine worm cytolysin, Cerebratulus toxin A-III. Toxicol 87, 189–203. 59. Kem, W. R. (1985) Structure and action of nemertine toxins. Am. Zool. 2, 99–111. 60. Kem, W. R. and Blumenthal, K. M. (1978) Purification and characterization of the cytolytic Cerebratulus A toxins. J. Biol. Chem. 253, 5752–5757. 61. Blumenthal, K. M. and Kem, W. R. (1980) Primary structure of Cerebratulus lacteus toxin A-III. J. Biol. Chem. 255, 8266–8272. 62. Kuo, J. F., Raynor, R. L., Mazzei, G. J., Schatzman, R. C., Turner, R. S., and Kem, W. R. (1983) Cobra polypeptide cytotoxin I and marine worm polypeptide cytotoxin A-IV are potent and selective inhibitors of phospholipid sensitive Ca2+-dependent protein kinase. FEBS Lett. 153, 183–186. 63. Kem, W. R. (2000), Nemertine neurotoxins, in Neurotoxicology Handbook: Natural Toxins of Animal Origin (Harvey, A., ed.), Humana Press, Totowa, NJ.

594

Kem

Trachynilysin

595

27 Secretagogue Activity of Trachynilysin, a Neurotoxic Protein Isolated from Stonefish (Synanceia trachynis) Venom Frédéric A. Meunier, Gilles Ouanounou, Cesar Mattei, Pascal Chameau, Cesare Colasante, Yuri A. Ushkaryov, J. Oliver Dolly, Arnold S. Kreger, and Jordi Molgó

1. INTRODUCTION Approximately 400 to 500 species of marine fish may be poisonous to humans after ingestion. Most poisonous fish are nonmigratory reef fish and can be either herbivores or carnivores. Some of them have tissues that are toxic at all times, others are poisonous during certain periods of the year or in certain geographical areas, and still others have only specific organs that are toxic, and their toxicity may vary with time, location, and habitat (reviewed in ref. 1). More than 200 species of marine fish, including stingrays, scorpionfish, weevers, and stargazers, possess some form of venom apparatus capable of inflicting serious and occasionally fatal wounds to humans. Most venomous fish are nonmigratory, shallowwater reef or inshore fish, and they use their venom apparatus as a defensive weapon. The Indian and Pacific Oceans and the Red Sea contain numerous genera and species of venomous fish belonging to the family Scorpaenidae (2,3). Members of this family are commonly called scorpionfish, and they include three stonefish (Synanceia) species, which possess numerous pairs of well-developed venom glands and cause severe disease in humans (2–14). The results of clinical and pharmacological studies indicate that stonefish venoms cause: (1) intense pain in humans; (2) extensive local tissue edema and necrosis of skin tissue; (3) a marked increase in the spontaneous release of neurotransmitters; (4) irreversible damage to and depolarization of muscle cells; (5) muscle twitches, incoordination, and paralysis (neuromuscular blockade); (6) hypotension; and (7) respiratory and cardiac failure (5–10,15–24). Although the venomous properties of certain fish have been recognized for centuries, during the last decade there has been a breakthrough in our knowledge of the chemical and pharmacological properties of protein toxins present in some of the venoms (reviewed in refs. 21,25,26). For example, lethal, membrane-perturbing, neurotoxic and myotoxic proteins have been isolated from the venoms of all three stonefish From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

595

596

Meunier et al.

species (i.e., stonustoxin from S. horrida [23,27–29], verrucotoxin from S. verrucosa [22,30], and trachynilysin from S. trachynis [31–33]). It is not possible within the confines of this review to consider in detail the properties of all of the bioactive agents and toxins present in fish venoms and, in particular, those present in the venoms of stonefish (the most dangerous of the known venomous fish), for which we apologize. This brief report reviews our studies of the pharmacologic activity of trachynilysin (TLY), a membrane-perturbing neurotoxic protein isolated from S. trachynis venom, which has the ability to increase neurotransmitter release. 2. PURIFICATION AND PARTIAL CHEMICAL CHARACTERIZATION OF TLY The availability of rapid, specific, and accurate assays is essential for the successful purification of the biologically active constituents of venoms. In that regard, TLY, albeit a neurotoxin, is most rapidly and accurately quantified by virtue of its in vitro, membrane-perturbing, and lytic activity against washed rabbit erythrocytes (31,33). The neurotoxins in the venoms of other stonefish species (i.e., stonustoxin [34] and verrucotoxin [22]) also possess hemolytic activity, and they are commonly detected and quantitated by a similar in vitro assay. Lyophilized S. trachynis venom, obtained as previously described (31), was mixed gently with Tris-buffered saline (TBS; pH 7.4) containing glycerol (TBS-G), and insoluble matter was removed by centrifugation followed by membrane filtration. The solution of reconstituted venom was fractionated by anion-exchange fast protein liquid chromatography (FPLC) using a Resource Q column equilibrated at room temperature with TBS-G. After application of the venom, the gel was washed sequentially with TBS-G, TBS-G supplemented with 0.1 M ammonium sulfate, and TBS-G supplemented with 0.8 M ammonium sulfate. The effluent was continuously monitored for absorbance at 280 nm, and the fractions were assayed for cytolytic activity against rabbit erythrocytes (Fig. 1). The pool (ca. 2 mL) of cytolytic peak fractions obtained by anion-exchange FPLC was concentrated four to fivefold with a Centricon-10 microconcentrator, and aliquots (0.2 mL) of the concentrated pool were fractionated by size-exclusion FPLC using a Superose 12 HR 10/30 column equilibrated with TBS-G at room temperature. The gel was washed with TBS-G, the effluent was continuously monitored for absorbance at 280 nm, and fractions were collected and assayed for cytolytic activity (Fig. 2). The cytolytic peak fractions (usually 5 or 6 fractions) were pooled, concentrated with a Centricon-10 microconcentrator to yield a preparation containing 1–2 mg protein/mL, and stored at –60°C. Fractionation of S. trachynis venom (ca. 60 mg dry weight containing 44–48 mg of protein and 1.8–3.0 × 106 HU), by the two-step purification procedure described earlier, yielded highly purified TLY preparations containing 3–4 mg of protein and having a specific activity of 3–4 × 105 HU/mg protein (33). Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Fig. 3) demonstrated the homogeneity of the final preparation obtained by the purification procedure, and indicated that the toxin is composed of subunits having molecular masses of ca. 76 kDa (α-subunit) and 83 kDa (ß-subunit).

Trachynilysin

597

Fig. 1. Anion-exchange FPLC of reconstituted S. trachynis venom. Fractions obtained during elution of the reconstituted venom with the indicated buffers were continuously monitored for absorbance at 280 nm and were assayed for cytolytic activity against washed rabbit erythrocytes. One hemolytic unit (HU) is the amount of toxin needed to release the hemoglobin from 50% of the erythrocytes.

Quantitative amino acid analysis confirmed the protein nature of TLY, and the sequence of the first 37 N-terminal amino acid residues in the native TLY preparation and in the ß-subunit of the denatured, reduced toxin was found to be: Pro-Ser-Asp-IleLeu-Val-Val-Ala-Ala-Leu-Gly-Arg-Pro-Phe-Thr-Leu-Gly-Met-Leu-Tyr-Asp-AlaArg-Asn-Asp-Lys-Leu-Ile-Pro-Gly-Phe-Thr-Leu-Val-Glu-Asp-Glu (33). The N-terminal amino acid sequence of the α-subunit could not be determined. The apparent molecular mass of native TLY (determined by size-exclusion FPLC), assuming that the toxin has a globular shape and does not interact with Superose 12 gel, is ca. 158,000 Daltons and the toxin’s isoelectric point (determined by liquid density gradient electrofocusing) is ca. 5.7 (31).

598

Meunier et al.

Fig. 2. Size-exclusion FPLC of TLY preparation obtained by anion-exchange FPLC. Fractions obtained during elution with TBS-G were continuously monitored for absorbance at 280 nm and were assayed for cytolytic activity against washed rabbit erythrocytes. The column void volume was 8.2 mL.

3. TLY INCREASES SPONTANEOUS AND EVOKED QUANTAL ACETYLCHOLINE RELEASE FROM VERTEBRATE MOTOR-NERVE TERMINALS Low concentrations of S. trachynis venom have been reported (18) to enhance spontaneous quantal acetylcholine (ACh) release, detected as an increase in miniature endplate potentials (MEPPs) or currents at mouse and frog neuromuscular junctions, and to transiently augment nerve-evoked endplate currents or potentials. These results prompted us to purify TLY, the venom’s membrane-perturbing, hemolytic, protein toxin (see Subheading 1.), in order to determine whether it was the venom constituent responsible for the previously observed presynaptic stimulation of the neurotransmitter release process. Nanomolar concentrations of TLY caused a profound increase in the frequency of spontaneous MEPPs (Fig. 4) of isolated frog neuromuscular preparations equilibrated in standard frog saline (33). The increased MEPP frequency developed after a delay of a few minutes and was characterized by oscillations in the mean frequency of the spontaneous quantal events, in which high-intensity bursts of MEPP discharges of variable

Trachynilysin

599

Fig. 3. SDS-PAGE of S. trachynis venom and TLY preparations. The patterns shown were obtained with Coomassie brilliant blue R-250-stained gels composed of stacking gels containing 4% acrylamide and separating gels containing 12% acrylamide (A) and 7.5% acrylamide (B). Lane 1, protein standards (Novex); molecular masses of the standards are indicated in kDa, to the left of lane 1. Lane 2, reconstituted venom. Lane 3, TLY isolated from venom by anionexchange FPLC. Lane 4, TLY isolated from venom by sequential anion-exchange FPLC and size-exclusion FPLC. The molecular masses of the TLY subunits were estimated from their mobilities in separating gels containing 7.5% acrylamide (lane 4, B). Reproduced with permission from Eur. J. Neurosci. (33).

duration alternated with periods in which the frequency returned to near control levels (as before TLY exposure). During MEPP bursts, the mean frequency was elevated from a control level of about 0.4 events/s (Hz) to frequencies around 100–150 Hz. The TLY-induced appearance of MEPP bursts was not affected by tetrodotoxin (0.5–1 µM) blockade of axonal- and nerve-terminal voltage-dependent sodium channels. Thus, it is unlikely that the bursts of spontaneous MEPP discharges arise because of spontaneous firing of motor axons or nerve terminals. During periods of enhanced MEPP frequency induced by 63 nM TLY, there was a simultaneous increase in the amplitude of endplate potentials, suggesting that a common mechanism might be responsible for both effects. However, exposure to TLY for more than 60 min caused an irreversible failure of neuromuscular transmission. The increased MEPP frequency induced by TLY was irreversible and was sustained for about 3–4 h, after which time it declined in all junctions (possibly due to exhaustion of transmitter stores). In normal frog saline, the range of

600

Meunier et al.

Fig. 4. Effect of TLY on miniature endplate potential frequency. Intracellular recordings were obtained with the same frog neuromuscular junction before (A) and 10 min after addition of 63 nM TLY (B) to the standard physiological solution supplemented with 1 µM tetrodotoxin. The resting-membrane potential of the muscle fiber during the recordings was –86 mV.

TLY concentrations that markedly increased the mean MEPP frequency did not affect their amplitude distribution or the resting membrane potential of the muscle fibers. Interestingly, exposure to TLY for 1 h failed to increase MEPP frequency in preparations preequilibrated in frog saline without Ca2+ and containing the Ca2+-chelator EGTA (1 mM). However, extensive washing to remove unbound TLY, and replacing the Ca2+-free and EGTA supplemented medium with standard frog saline containing Ca2+ (1.8 mM), caused an immediate increase in MEPP frequency. Also, removing Ca2+ from the extracellular medium, by replacing the standard medium with a Ca2+free medium supplemented with EGTA, completely suppressed the increase in quantal transmitter release. These observations suggest that: (1) extracellular Ca2+ is not required for TLY binding to its membrane receptors, (2) there is very little dissociation of TLY from its binding site, and (3) the TLY-elicited response involves a presynaptic mechanism requiring the influx of Ca2+ for enhancing quantal transmitter release from motor nerve endings. TLY-mediated responses similar to those herein described with frog neuromuscular preparations also have been observed with TLY-treated mouse (J. Molgó, unpublished observations) and fish (Torpedo marmorata) (36) skeletal neuromuscular preparations. In order to determine whether Ca2+ entry via nerve-terminal voltage-dependent Ca2+ channels is involved in the increase of MEPP frequency caused by TLY, the selective N-type Ca2+ channel blocker ω-conotoxin GVIA was used specifically to block Ca2+ entry into frog motor-nerve terminals (36,37). However, application of ω-conotoxin

Trachynilysin

601

GVIA (2–3.5 µM) to frog neuromuscular preparations did not prevent the TLY-elicited increase in MEPP frequency. Thus, it appears that the Ca2+ influx triggered by TLY does not occur through voltage-sensitive Ca2+ channels, but must occur via some other Ca2+ entry pathway (see Subheadings 7. and 9.). The lectin concanavalin-A (500 µg/mL), from the plant Canavalia ensiformis, prevented the TLY-induced increase in spontaneous quantal ACh release. The idea that the effect of the lectin probably involves inhibition of TLY binding to its receptor is supported by our observation that concanavalin-A applied after treatment with TLY failed to affect TLY-increased MEPP frequency. 4. MONITORING OF TLY-INDUCED, Ca2+-DEPENDENT EXOCYTOSIS AND IMPAIRMENT OF THE SYNAPTIC-VESICLE RECYCLING PROCESS IN LIVING MOTOR-NERVE TERMINALS WITH FM1-43 The fluorescent styryl dye FM1-43 has been widely used since 1992 as a tracer allowing real-time observation of exo-endocytosis in a large range of neuronal and non-neuronal cells (38–43). Stimulation of neurotransmitter release at the neuromuscular junction, in the presence of FM1-43, produces an alignment of brightly stained clusters of recycled synaptic vesicles along presynaptic nerve-terminal branches (38,39). The staining is stable during resting conditions, but stimulation of exocytosis results in a reversible loss of the fluorescent spots, which is believed to result from an activity-dependent release of FM1-43. This unique property has been exploited to gain novel insights into various parameters of vesicular recycling and to optically measure synaptic-vesicle turn-over during endo- and/or exocytosis in living nerve terminals (38,41,42). Thus, we investigated the effect of TLY on living FM1-43- prestained nerve terminals (Fig. 5). In the absence of external Ca2+, TLY did not alter the pattern of staining (Fig. 5A, B), but the subsequent addition of Ca2+ modified the staining. Swelling of the nerve-terminal branches was observed together with detectable shrinkage, dimming, and segregation of the fluorescent spots, which later coalesced into brighter and larger spots (Fig. 5C–F). The overall staining intensity was markedly reduced, as expected from the known secretion of FM1-43 during TLY-induced bursts of asynchronous quantal release. The TLY-induced alteration of FM1-43 staining is indicative of an impairment of the synaptic recycling process, and the very bright coalescent spots could correspond to the incorrectly recycled inner membranes previously observed by ultrastructural analysis (33). This result contrasts with the effect of α-latrotoxin (α-LTX) on the same preparation, where recycling is only blocked when Ca2+ is removed from the external medium (39,44). 5. TLY DEPLETES CLEAR SYNAPTIC VESICLES BUT DOES NOT AFFECT LARGE DENSE-CORE VESICLES IN MOTOR-NERVE TERMINALS The effect of TLY on nerve-terminal ultrastructure was examined to determine whether the toxin causes structural changes related to the observed massive quantal transmitter release (see Subheading 3.). Frog neuromuscular junctions exposed for 3 h to TLY, a condition that greatly increased quantal ACh release leading to exhaustion of the neurotransmitter release process, showed a marked depletion of small clear synap-

602

Meunier et al.

Fig. 5. TLY-induced destaining and swelling of frog motor-nerve terminals loaded with FM1-43 dye in the presence of Ca2+. (A) Frog nerve endings stained by stimulating for 10 min transmitter release with an isotonic 60 mM K+ medium, in the presence of 2 µM FM1-43, and then washed for 30 min in a standard Ca2+-free solution containing 2 mM EGTA. Each fluorescent spot marks clusters of synaptic vesicles. Addition of 63 nM TLY to the Ca2+-free medium did not trigger destaining (B), but extensive washing of the preparation with a solution containing 2 mM Ca2+ for 15 (C), 30 (D), 60 (E), and 120 min (F) destained and swelled the nerve terminal branches (C–F).

tic vesicles and swelling of motor-nerve terminals (33). Thus, in addition to eliciting the release of ACh quanta, TLY seems to impair the retrieval of the membranes of exocytosed small clear synaptic vesicles and their incorporation into new vesicles. Small clear synaptic vesicles are known to store and release ACh by repeated cycles of exo-endocytosis (37,44–47). Synaptophysin immunolabeling of nerve terminals was performed to determine the fate of synaptic vesicles after intense quantal transmitter release induced by TLY. Synaptophysin is a well known integral membrane protein of synaptic vesicles that can only be detected in resting motor-nerve terminals after detergent treatment. Under this condition, synaptophysin immunolabeling exhibits a segmented pattern that corresponds to clusters of synaptic vesicles. However, after 3 h exposure to TLY, nonpermeabilized swollen nerve terminals exhibited an irregularly distributed synaptophysin immunoreactivity, i.e., the regular bands were no longer observed and the immunofluorescence observed appeared to be due to synaptophysin incorporation into the nerve-terminal axolemma as a consequence of permanent synaptic-vesicle fusion. Thus, these results support the view that TLY impairs the small clear synaptic vesicle-recycling process (33). In marked contrast, TLY did not induce a parallel depletion of the small population of large dense-core vesicles (80–150 nm diameter) scattered within motor-nerve endings. These vesicles are known to store and secrete neuropeptides (48,49), among which the calcitonin gene-related peptide

Trachynilysin

603

(CGRP) and substance-P are the best characterized at motor nerve endings (50,51). TLY-exposed nerve terminals were shown to contain CGRP, as revealed by colloidal gold immunostaining, and they exhibited CGRP-like immunofluorescence similar to that of untreated terminals (33). Interestingly, the only toxin previously reported to cause differential depletion of the two types of secretory vesicles present in motor nerve endings is α-LTX (51) purified from black widow spider venom (52). However, α-LTX only induces depletion of small clear synaptic vesicles in the absence of extracellular Ca2+ (53), whereas TLY’s action is dependent on the presence of external Ca2+. It remains to be determined whether TLY interacts with and activates some molecular constituents of clear synaptic vesicles that are absent in large dense-core vesicles. However, whatever the mechanism may be for TLY’s stimulation of neurotransmitter release from small clear synaptic vesicles, the toxin may be a useful tool for discriminating between neurotransmitter and neuropeptide release at motor-nerve terminals. 6. TLY IS A SECRETAGOGUE FOR BOVINE CHROMAFFIN CELLS Recent studies have demonstrated that TLY triggers the release of catecholamine from large dense-core vesicles of adrenal chromaffin cells (54) (Fig. 6A). Ca2+ was required in the external medium in order to induce secretion during the initial phase (20 min post-treatment) of release, and intoxication of the chromaffin cells with various serotypes of botulinum neurotoxins abolished this early stimulatory effect (54). Botulinum neurotoxin serotypes A–G selectively cleave presynaptic proteins that are members of a tripartite complex (Soluble N-ethylmaleimide-sensitive fusion protein Attachment Protein [SNAP] receptor) involved in the last stage of exocytosis, thereby resulting in impairment of the latter process in motor-nerve terminals and secretory cells (reviewed in ref. 55). Also, treatment of chromaffin cells with botulinum toxins types A and B cleaves the vast majority of SNAP-25 and vesicle-associated protein (VAMP), respectively (56,57). Thus, the use of botulinum toxin-treated cells showed that the release is SNARE-dependent. However, a later phase (60 min postintoxication) of TLY-induced release is Ca2+- and SNARE (SNAP-receptor)-independent and is believed to result from an alteration of membrane permeability, because concomitant “leakage” of cytoplasmic lactate dehydrogenase was detected in the external medium. The binding of TLY did not require the presence of Ca2+ but was partially inhibited by pretreatment of the cells with the lectin concanavalin A or its derivative succinyl concanavalin A (Fig. 6B), which suggests that the toxin interacts with a glycoprotein situated on the chromaffin-cell surface. 7. TLY ELICITS A TRANSIENT Ca2+ INCREASE IN HIPPOCAMPAL NEURONS AND CHROMAFFIN CELLS The fact that Ca2+ was required for TLY’s stimulatory effect on both nerve terminals and neurosecretory cells prompted us to investigate the possible effect of TLY on intracellular Ca2+ levels in hippocampal neurons and chromaffin cells. Ca2+ imaging experiments, using the dye fluo-3 and cultured mouse hippocampal neurons, revealed that TLY causes a transient increase in intracellular Ca2+ levels that is dependent on external Ca2+ (58) (Fig. 7A). The voltage-dependent Ca2+ channels expressed in these neurons (59) were not responsible for the Ca2+ influx (Fig. 7B). Additional Ca2+ imaging experiments showed that the binding of the toxin is not dependent on the presence of

604

Meunier et al.

Fig. 6. TLY-induced catecholamine secretion from chromaffin cells and the effects of concanavalin-A, succinyl concanavalin-A, and botulinum neurotoxins type A and B. (A) Primary cultures of adrenal chromaffin cells bathed in standard physiological solution containing 2 mM Ca2+ were exposed to various concentrations of TLY for 20 min before removing of an aliquot and assaying of its catecholamine content by fluorometry. The amounts of catecholamine released (± S. E. M., n = 4) are plotted as a % of the total content. (B) Chromaffin cells were preincubated for 15 min with concanavalin-A (CA) or succinyl concanavalin-A (SCA) (500 µg/ mL) before incubating with TLY (60 nM) for 20 min and determining the amount of secreted catecholamines. (C) Intact chromaffin cells poisoned by prolonged exposure to BoNT-A, -B (methods detailed in ref. 54), and control cells treated only with vehicle, were exposed for 20 min to TLY (60 nM) in the presence of Ca2+ (2 mM), or for 60 min in the absence of Ca2+, before measuring the extent of catecholamine release.

external Ca2+ (Fig. 7C) and is abolished by concanavalin-A (Fig. 7D). These observations suggest that TLY probably triggers Ca2+ influx via a Ca2+ channel-independent pathway. None of the Ca2+ channels reported in chromaffin cells (60–62) seems to be implicated in TLY-induced catecholamine release (54). Investigating the possible effect of TLY on Ca2+ levels in fluo-3-loaded chromaffin cells, with high-resolution confocal microscopy, revealed increased cytoplasmic fluorescence and a particularly strong signal in a restricted area of the optical section of the cell (Fig. 8A). Such polarized Ca2+ signals are typical of chromaffin cells’ response to agonists that mobilize Ca2+ from internal stores (reviewed in ref. 63). Also, the precise colocalization of intracellular Ca2+ signals evoked by TLY and caffeine (Fig. 8B) strongly suggests a role for the

Trachynilysin

605

Fig. 7. TLY-induced Ca2+ transients in fluo-3-loaded cultured mouse hippocampal neurons. Ca2+ transients elicited by bath application of TLY (60 nM) in the presence and absence of external Ca2+ (A) and in the presence of external Ca2+ and the Ca2+ channel blockers nifedipine (10 µM), ω-conotoxin GVIA (3 µM), and ω-conotoxin MVIIC (1 µM) (B). After bath application of TLY (60 nM) in the absence of external Ca2+, the superfusion of TLY-free medium containing Ca2+ induced an increase in intracellular Ca2+ (C). The Ca2+ signal was completely abolished by pretreating the neurons with 10 µM concanavalin A (D).

mobilization of caffeine-sensitive Ca2+ stores by the toxin. Since caffeine is a poor secretagogue (63), we hypothesize that TLY stimulates Ca2+ influx across the plasma membrane, thereby triggering Ca2+-induced Ca2+ release from the endoplasmic reticulum stores and raising cytosolic Ca2+ to a level capable of supporting exocytosis. Intracellular, caffeine-sensitive Ca2+-stores could therefore be recruited by TLY as an extra source of Ca2+ available for prolonging the triggering of exocytosis. 8. TLY-BINDING SITES ARE SEGREGATED ON THE PLASMA MEMBRANE OF CHROMAFFIN CELLS The observation that TLY triggers a polarized increase in Ca2+ prompted us to investigate the possibility of restricted localization of TLY-binding sites to a single pole of chromaffin cells. Probing of Western blots of purified TLY with rabbit antiTLY antibodies detected both of the toxin’s subunits, although labeling of the 76 kDa α-subunit was stronger than that of the 83 kDa ß-subunit (Fig. 9). In order to determine the precise location of the TLY-binding sites on chromaffin cells, immunocytochemistry was performed with control and TLY-treated preparations. Anti-TLY labeling was always restricted to an area representing 15–30% of the total surface of the intoxicated cells (Fig. 9B), whereas nonintoxicated chromaffin cells exhibited very faint, nonspecific, anti-TLY immunostaining (ca. 4% of that shown by intoxicated cells; p < 0.001; Fig. 9B). Moreover, the TLY signal from the contaminating fibroblasts was negligible (ca. 1% of that exhibited by intoxicated cells; p < 0.001; Fig. 9), which demonstrated

606

Meunier et al.

Fig. 8. Confocal microscopy-obtained evidence of the colocalization of caffeine- and TLYinduced intracellular Ca2+ signals in fluo-3-loaded chromaffin cells. Confocal images (optical slice 1 µm), acquired every 5 s, of a single fluo-3-loaded, living chromaffin cell (A) sequentially treated with 5 mM caffeine (B–D), washed out of caffeine for 10 min (E), and treated with 60 nM TLY (F–J) in the presence of external Ca2+. Note from the images (B–D) and their intensity profiles (K) that the intracellular Ca2+ transient evoked by caffeine is restricted to a specific region of the chromaffin cell, and that subsequent treatment with 60 nM TLY (G–J, L) produces a localized increase in Ca2+ fluorescence signal of higher amplitude but at the same area, thus suggesting the involvement of intracellular Ca2+ stores.

that TLY-binding sites are highly specific for chromaffin cells. Thus, binding by and activation of an as-yet-to-be discovered receptor for TLY seems likely, since toxinbinding sites are clustered within a restricted area of the chromaffin cell. Interaction of TLY with the putative receptor would result in Ca2+-influx, which in turn would activate Ca2+-induced Ca2+ release from caffeine-sensitive stores and, ultimately, large dense-core vesicle exocytosis. 9. NEURONAL MEMBRANE EFFECTS AND PORE-FORMING ACTIVITY OF TLY Since Ca2+ influx is required for TLY’s action on nerve terminals and neuroendocrine cells, and the L, N, and P/Q types of voltage-sensitive Ca2+ channels are not accountable for TLY-induced neurotransmitter secretion or intracellular Ca2+ increases, it was of interest to investigate the possibility that TLY forms pores or channels in neuronal cells. The ability of TLY to form membrane channels was investigated with

Trachynilysin

607

Fig. 9. Immunocytochemical localization of TLY binding sites by confocal microscopy reveals their restriction to a single pole of the chromaffin cell. (A) Purified TLY was subjected to SDS-PAGE (6% polyacrylamide gel) followed by Western blotting with rabbit polyclonal IgG raised against TLY. The toxin’s 76 kDa α-subunit reacted more strongly with the antibody preparation than did the 83 kDa ß-subunit (using ECL system staining). (B) Unexposed cells and cells exposed to 60 nM TLY for 30 min were washed before fixation (4% paraformaldehyde) and probing for TLY binding with anti-TLY IgG, and the location of the TLY-IgG complexes was revealed by immunostaining with FITC-conjugated, anti-rabbit IgG. Rhodaminated-phalloidin was used to differentiate between chromaffin cells (showing a ringlike cortical actin network) and contaminating fibroblasts (displaying a star-like actin network). Note that only TLY-treated chromaffin cells (+ TLY) display anti-TLY immunostaining, which is restricted to one area of the cell, and that control, untreated (– TLY) chromaffin cells do not exhibit such staining.

cultured neuroblastoma × glioma NG108-15 hybrid cells (64) previously differentiated with dibutyryl-cyclic adenosine-monophosphate. Because of the irreversibility of TLYbinding, TLY (12 nM) could be applied for just 1 min to NG108-15 cells, using a fast perfusion system followed by washing with a TLY-free medium. Under these conditions, using the whole-cell configuration of the patch-clamp technique (65), it was evident that TLY induced a persistent inward current at a holding potential of –60 mV (Fig. 10A), which fluctuated in abrupt steps that were multiples of the same value and was due to changes in membrane permeability. All physiological cations contributed to this current, and the nonselective cationic current reversed at –3 mV in standard physiological solution (66). La3+ and Ni2+, which are known to block nonselective cation channels in a dose dependent-manner (67), reversibly blocked the TLY-induced current. The kinetics of the current induced by TLY, with the development of abrupt current steps (Fig. 10A), has also been observed in the presence of other pore-forming toxins, such as α-LTX (68,69) and equinatoxin-II, a cytolysin from the sea anemone Actinia equina L. (70–72). Furthermore, the TLY-induced conductance change was higher at positive holding potentials than at negative holding potentials (Fig. 10B).

608

Meunier et al.

Fig. 10. Membrane effects and pore-forming activity of TLY in neuroblastoma cells. (A) Whole-cell patch-clamp recording of the TLY-induced transmembrane current of a neuroblastoma (NG108-15) cell bathed in physiological medium at a holding potential of –60 mV. Note that the inward current develops with abrupt steps, and the changes between preferential levels. (B) Current-voltage relationship of TLY-induced current under steady-state conditions. The current-voltage relationship exhibited outward rectification and had a reversal potential of –3 mV in physiological medium. Each point is the mean ± S. E. M. of 8 measurements in different cells. (C) Inside-out patch-clamp recording in which TLY was added to the pipet medium, at a pipet potential of +60 mV (membrane potential of –60 mV). The figure shows the activity of a 78 pS pore, which was found only at positive pipet potential; c and o indicate the “closed” and “open” state of the pore, respectively.

Elementary currents recorded in membrane patches of differentiated NG108-15 cells, using the inside-out configuration of the patch-clamp technique, revealed that the addition of TLY to the pipet-medium evoked single-channel activity having various conductance values and kinetics. The pore-forming activity of TLY was more distinct for positive pipet potentials (negative membrane potential) than for negative pipet potentials, but the number of induced pores was higher at negative pipet potentials (Ouanounou, et al., manuscript in preparation). These data can be interpreted as indicating a “carpet-like” mechanism for pore formation (73), which has been proposed for stonustoxin, a hemolytic, pore-forming toxin in S. horrida venom (34). According to this model, the basic residues of the toxin monomers initiate interaction with the negative cell surface. The peptides then bind parallel to the surface in an amphiphilic α-helix structure covering the membrane without penetrating deeply into the hydrophobic core. When a threshold concentration of the peptide is reached, the monomers may penetrate more deeply into the membrane core and disorganize the membrane structure. At this stage, a fraction of the peptides may polymerize and reorient the

Trachynilysin

609

hydrophobic face of their amphiphilic α-helices toward the hydrophobic membrane core, thus forming the transmembrane pores. The observation of the TLY-induced macroscopic current, recorded with the whole-cell configuration, suggests that macroscopic steps of current may result from the formation of single “macropores.” However, because only multiples of 26 pS were found for the elementary conductances in the inside-out configuration of the patch-clamp technique, some pore sizes seemed more stable than others, which suggests that some oligomerized TLY structures are more stable than others. 10. DIFFERENT RECEPTORS BUT POSSIBLE COMMON MECHANISM OF ACTION FOR TLY AND α-LTX? Both purified TLY and α-LTX induce a massive increase in the frequency of MEPPs of frog neuromuscular junctions in a Ca2+-containing physiological solution (see Subheading 3.). Therefore, we were interested in determining whether both toxins target a common neuroexocytosis pathway and have a common receptor. Since α-LTX has recently been shown to be a potent secretagogue in the “user-friendly” primary culture of bovine chromaffin cells and in rat cerebrocortical synaptosomes, and two of its receptors have been cloned and well characterized (74–79), we decided to use cerebrocortical synaptosomes and chromaffin cells in order to examine the possibility that the two toxins have a common receptor and mechanism of action. Insights into the mechanism of action of α-LTX were recently obtained because of the purification, cloning, and sequencing of latrophilin, a neuronal protein also known as CIRL for Ca2+-Independent Receptor for α-Latrotoxin (76,79), which binds α-LTX with high affinity in the absence of Ca2+ (75). The amino acid sequence of latrophilin shows a high-degree of homology with the G-protein-coupled members of the secretin superfamily (79–81). Moreover, some evidence suggests that α-LTX activates a phospholipase C via a signal-transduction mechanism (82), although the relevance of this activation to stimulation of exocytosis by the toxin is unclear (83). However, despite the similar secretagogue effect of TLY and α-LTX on cholinergic nerve terminals and neuroendocrine cells, TLY did not compete with α-LTX for binding to cortical synaptosomal membranes in the absence of external Ca2+ (Fig. 11A). Moreover, chromaffin cells transiently transfected so as to express latrophilin did not exhibit significantly higher sensitivity to TLY than did nontransfected control cells (Fig. 11B). Taken together, these data do not support the hypothesis that latrophilin is a functional receptor for TLY. However, additional experiments are needed to evaluate the possibility that neurexin, the other α-LTX receptor (74), is a receptor for TLY. Striking similarities in the secretagogue activity of α-LTX (81,82) and TLY (54) have emerged from recent studies implicating the role of internal Ca2+ stores in the toxins’ modes of action. Thus, even if α-LTX and TLY do not have the same receptor, they may have a similar mechanism of action for synaptic-vesicle exocytosis. A model for this action is presented in Fig. 12B. We propose that the mode of action of these stimulatory toxins have two important features in common: The toxins (1) bind specifically to the membrane of presynaptic nerve terminals and neuroendocrine cells, and (2) form pores or channels after binding to the membranes. These actions would enable a very small amount of toxin specifically and efficiently to reach the critical concentration enabling it to insert into the membrane and form Ca2+-permeable pores. The resul-

610

Meunier et al.

Fig. 11. TLY does not bind to the α-LTX-receptor latrophilin. (A) Binding of [I125] α-LTX (5 nM) to rat cerebrocortical synaptosomes was measured in the presence of increasing concentrations of TLY (closed circles) or a 100-fold excess of unlabeled α-LTX (open circles). Only α-LTX displaced the radiolabeled α-LTX from its receptors, and TLY stimulated secretion of [H3]norephinephrine (NE) from the synaptosomes (not shown). (B) Chromaffin cells were cotransfected (method detailed in ref. 79) with the hGH plasmid and with either a control plasmid (pcDNA3.1) or a plasmid encoding for latrophilin (LPH2A). After 5 d, both preparations were treated with 20 pM α-LTX or 1 nM TLY (concentrations just below their respective EC50) in Ca2+-containing medium. The small % of cotransfected cells containing hGH and LPH2A plasmids were sensitized to α-LTX but not to the low dose of TLY, since hGH release remained very similar to the catecholamine basal level.

ting Ca2+ influx would then activate internal Ca2+ stores and maintain a sustained rise in intracellular free Ca2+ at levels capable of supporting exocytosis (Fig. 12B). Recent studies of the structure of α-LTX (84,85) have yielded novel perspectives for a plausible mode of action for the toxin (Fig. 12A) and suggest further studies needed to elucidate TLY’s mechanism of action. It has long been known that α-LTX has pore-forming activity in artificial membrane bilayers and neuroblastoma cells (68,69). However, the toxin does not create FITC-permeant pores in naive Cos-7 cells

Trachynilysin

611

Fig. 12. Proposed model for α-LTX mode of action and schematic diagram of the model for TLY-induced, Ca2+-dependent exocytosis. (A) Cryo-electronmicroscopy reconstruction of the α-LTX tetramer; (a) top view and (b) side view (c) model of the α-LTX tetramer (cut open to reveal the channel) in the membrane bilayer. Reproduced with permission from ref. 85. Copyright Nature Structural Biology. (B) Schematic view of TLY action: TLY binding to and activation of its unknown receptor results in Ca2+ influx via pore formation, which in turn triggers Ca2+-induced Ca2+ release from Ca2+ stores and, finally, exocytosis of synaptic vesicles (SV) (see text for details).

unless the cells are transfected to express either of the toxin’s receptors (86). Also, it has recently been suggested that α-LTX uses latrophilin and neurexin in order to be recruited and concentrated on motor nerve terminals selectively expressing both receptors (82,86), and that α-LTX pores are formed by the insertion of divalent cation-dependent α-LTX tetramers into the plasma membrane (84,85). However, it is not yet clear how the topology of the two receptors elicits oligomerization and/or insertion of the toxin in the plasmalemma bilayer. Whatever the mechanism may be, however, α-LTX creates pores at crucial sites, thereby causing Ca2+ influx into the vicinity of the presynaptic and transmitter release sites, which triggers maximal stimulation of neurotransmitter release (86). Thus, TLY, which is also a pore-forming toxin (see Subheading 9.), may use another “docking anchor” protein selectively expressed on motor-nerve terminals in order to insert into the plasma membrane. This hypothesis needs to be evaluated experimentally because it may lead to the discovery of a novel presynaptic protein.

612

Meunier et al.

ACKNOWLEDGMENT This work was supported in part by a research grant from DSP (98/1051 to J.M) and by Public Health Service grant GM-43728 from the National Institutes of Health, USA (to A.S.K.). We thank Dr. B. Rouzaire-Dubois for providing the NG108-15 cells used in our studies and David McColm (University of Queensland, Brisbane, Australia) for his assistance in helping us to obtain the S. trachynis venom used as the source of TLY needed in our studies. F.A. M. was supported by a fellowship from the European Commission Biotechnology Program (grant B104CT965119), G.O. and P.C. were supported by fellowships from the Ministère de l’Education Nationale, de la Recherche Scientifique et de la Technologie, and C.M. was supported by a fellowship from the Direction des Recherches Etudes et Techniques. REFERENCES 1. Molgó, J., Benoit, E., Legrand, A. M., and Kreger, A. S. (1999) Bioactive agents involved in fish poisoning: an overview, in Proceedings of the 5th Indo-Pacific Fish Conference (Noumea, 1997), (Séret, B. and Sire, J. Y., eds.), Soc. Fr. Ichtyol., Paris, pp. 721–738. 2. Russell, F. E. (1965) Marine toxins and venomous and poisonous marine animals, in Advances in Marine Biology, vol. 3 (Russell, F. E., ed.), Academic Press, London, pp. 255–385. 3. Halstead, B. W. (1988) Scorpionfishes, in Poisonous and Venomous Marine Animals of the World, 2nd rev. ed., The Darwin Press, Princeton, pp. 839–906. 4. Sutherland, S. K. (1983) Genus Synanceia (Linnaeus), stonefishes: S. verrucosa (Bloch & Schneider) & S. trachynis (Richardson), in Australian Animal Toxins, the Creatures, Their Venoms and Care of the Poisoned Patient, Oxford University Press, Melbourne, pp. 400–410. 5. Wiener, S. (1958) Stonefish sting and its treatment. Med. J. Aust. 2, 218–222. 6. Wiener, S. (1959) Observations on the venom of the stonefish (Synanceja trachynis). Med. J. Aust. 1, 620–627. 7. Saunders, P. R. (1959) Venom of the stonefish Synanceja horrida (Linnaeus). Arch. Int. Pharmacodyn. Thér. 123, 195–205. 8. Austin, L., Cairncross, K. D., and McCallum, I. A. N. (1961) Some pharmacological actions of the venom of the stone fish (Synanceja horrida). Arch. Int. Pharmacodyn. 131, 339–347. 9. Austin, L., Gillis, R. G., and Youatt, G. (1965) Stonefish venom: some biochemical and chemical observations. Aust. J. Exp. Biol. Med. Sci. 43, 79–90. 10. Saunders, P. R. (1959) Venom of stonefish Synanceja verrucosa. Science 129, 272–274. 11. Bagnis, R. (1968) A propos de 51 cas de piqûres venimeuses par la “Rascasse” tropicale Synanceia verrucosa. Méd. Trop. 28, 612–620. 12. Deakins, D. E. and Saunders, P. R. (1967) Purification of the lethal fraction of the venom of the stonefish Synanceja horrida (Linnaeus). Toxicon 4, 257–262. 13. Lagraulet, J., Tapu, J., Cuzon, G., Fabre-Teste, R., and Toudic, A. (1972) Recent data on stings of poisonous fishes of the Synanceja genus and disk electrophoresis study of their venom. Bull. Soc. Pathol. Exot. 65, 605–621. 14. Scharnagl E., Smola, M., and Pierer, G. (1987) Steinfisch-stichverletzung an der hand. Handchirurgie 19, 46–48. 15. Low K. S., Gwee, M. C. E., and Yuen, R. (1990) Neuromuscular effects of the venom of the stonefish Synanceja horrida. Eur. J. Pharmacol. 183, 574. 16. Low, K. S., Gwee, M. C., Yuen, R., Gopalakrishnakone, P., and Khoo, H. E. (1993) Stonustoxin: a highly potent endothelium-dependent vasorelaxant in the rat. Toxicon 31, 1471–1478.

Trachynilysin

613

17. Khoo, H. E., Yuen, R., Poh, C. H., and Tan, C. H. (1992) Biological activities of Synanceja horrida (stonefish) venom. Nat. Toxins 1, 54–60. 18. Kreger, A. S., Molgó, J., Comella, J. X., Hansson, B., and Thesleff, S. (1993). Effects of stonefish (Synanceia trachynis) venom on murine and frog neuromuscular junctions. Toxicon 31, 307–317. 19. Shiomi, K., Hosaka, M., and Kikuchi, T. (1993). Properties of a lethal factor in stonefish Synanceia verrucosa venom. Nippon Suis. Gakk. 59, 1099. 20. Hopkins, B. J, Hodgson, W. C., and Sutherland S. K. (1994) Pharmacological studies of stonefish (Synanceja trachynis) venom. Toxicon 32, 1197–1210. 21. Gwee, M. C., Gopalakrishnakone, P., Yuen, R., Khoo, H. E., and Low, K. S. (1994) A review of stonefish venoms and toxins. Pharmacol. Ther. 64, 509–528. 22. Garnier, P., Goudey-Perrière, F., Breton, P., Dewulf, C., Petek, F., and Perrière, C. (1995) Enzymatic properties of the stonefish (Synanceia verrucosa Bloch and Schneider, 1801) venom and purification of a lethal, hypotensive and cytolytic factor. Toxicon 33, 143–155. 23. Poh, C. H., Yuen, R., Khoo, H. E., Chung, M., Gwee, M., and Gopalakrishnakone, P. (1991) Purification and partial characterization of stonustoxin (lethal factor) from Synanceia horrida venom. Comp. Biochem. Physiol. 99B, 793–798. 24. Low, K. S., Gwee, M. C., Yuen, R., Gopalakrishnakone, P., and Khoo, H. E. (1994) Stonustoxin: effects on neuromuscular function in vitro and in vivo. Toxicon 32, 573–581. 25. Goudey-Perriere, F. and Perriere, C. (1998) Pharmacological properties of fish venoms. C. R. Seances Soc. Biol. Fil. 192, 503–548. 26. Hodgson, W. C. (1997). Pharmacological action of Australian animal venoms. Clin. Exp. Pharmacol. Physiol. 24, 10–17. 27. Khoo, H. E, Chen, D., and Yuen, R. (1998) Role of free thiol groups in the biological activities of stonustoxin, a lethal factor from stonefish (Synanceja horrida) venom. Toxicon 36, 469–476. 28. Khoo, H. E., Chen, D., and Yuen, R. (1998) The role of cationic amino acid residues in the lethal activity of stonustoxin from stonefish (Synanceja horrida) venom. Biochem. Mol. Biol. Int. 44, 643–646. 29. Ghadessy, F. J., Chen, D., Kini, R. M., Chung, M. C. M., Jeyaseelan, K., Khoo, H. E., and Yuen, R. (1996) Stonustoxin is a novel lethal factor from stonefish (Synanceja horrida) venom. cDNA cloning and characterization. J. Biol. Chem. 271, 25,575–25,581. 30. Garnier, P., Ducancel, F., Ogawa, T., Boulain, J. C., Goudey-Perrière, F., Perrière, C., and Ménez, A. (1997) Complete amino-acid sequence of the beta-subunit of VTX from venom of the stonefish (Synanceia verrucosa) as identified from cDNA cloning experiments. Biochem. Biophys. Acta 1337, 1–5. 31. Kreger, A. S. (1991) Detection of a cytolytic toxin in the venom of the stonefish (Synanceia trachynis). Toxicon 29, 733–743. 32. Hopkins, B. J. and Hodgson, W. C. (1998) Enzyme and biochemical studies of stonefish (Synanceja trachynis) and soldierfish (Gymnapistes marmoratus) venoms. Toxicon 36, 791–793. 33. Colasante, C., Meunier, F. A., Kreger, A. S., and Molgó, J. (1996) Selective depletion of clear synaptic vesicles and enhanced quantal transmitter release at frog motor nerve endings produced by trachynilysin, a protein toxin isolated from stonefish (Synanceia trachynis) venom. Eur. J. Neurosci. 8, 2149–2156. 34. Chen, D., Kini, R. M., Yuen, R., and Khoo, H. E. (1997) Haemolytic activity of stonustoxin from stonefish (Synanceja horrida) venom: pore formation and the role of cationic amino acid residues. Biochem. J. 325, 685–691. 35. Ouanounou, G., Mattei, C., Meunier, F. A., Kreger, A. S., and Molgó, J. (2000) Trachynilysin, a protein neurotoxin isolated from Synanceia trachynis venom, increases spontaneous quantal acetylcholine release from Torpedo marmorata neuromuscular junctions. Cybium 24, (Suppl), 149–156.

614

Meunier et al.

36. Kerr, L. M. and Yosikami, D. (1984) A venom peptide with a novel presynaptic blocking action. Nature 308, 282–284. 37. Van der Kloot, W., and Molgó, J. (1994) Quantal acetylcholine release at the vertebrate neuromuscular junction. Physiol. Rev. 74, 899–991. 38. Betz, W. J., Mao, F., and Bewic, G. S. (1992) Activity-dependent fluorescent staining and destaining of living vertebrate motor nerve terminals. J. Neurosci. 12, 363–375. 39. Henkel, A.W. and Betz, W.J. (1995. Monitoring of black widow spider venom (BWSV) induced exo- and endocytosis in living frog motor nerve terminals with FM1-43. Neuropharmacology 34, 1397–1406 40. Giovannucci, D. R., Yule, D. I., and Stuenkel, E. L. (1998) Optical measurement of stimulus-evoked membrane dynamics in single pancreatic acinar cells. Am. J. Physiol. 275, C732–739. 41. De Paiva, A., Meunier, F. A., Molgó, J., Aoki, K. R., and Dolly, J. O. (1999) Functional repair of motor endplates after botulinum neurotoxin type A poisoning: biphasic switch of synaptic activity between nerve sprouts and their parent terminals. Proc. Natl. Acad. Sci. USA 96, 3200–3205. 42. Cochilla, A. J., Angleson, J. K., and Betz, W. J. (1999) Monitoring secretory membrane with FM1-43 fluorescence. Ann. Rev. Neurosci. 22, 1–10. 43. Kavalali, E. T., Klingauf, J., and Tsien, R. W. (1999) Properties of fast endocytosis at hippocampal synapses. Phil. Trans. R. Soc. Lond. Series B 354, 337–346. 44. Hurlbut, W. P., Iezzi, N., Fesce, R., and Ceccarelli, B. (1988) Correlation between quantal secretion and vesicle loss at the frog neuromuscular junction. J. Physiol. (Lond.) 425, 501–526. 45. Heuser, J. E. (1989) Review of electron microscopic evidence favouring vesicle exocytosis as the structural basis for quantal release during synaptic transmission. Q. J. Exp. Physiol. 74, 1051–1069. 46. Kelly, R. (1993) Storage and release of neurotransmitters. Neuron 10, 43–53. 47. Betz, W. J. and Angleson, J. K. (1998) The synaptic vesicle cycle. Annu. Rev. Physiol. 60, 347–363. 48. Thureson-Klein, A. K., Klein R. L., Zhu P. C., and Kong, J. Y. (1988) Differential release of transmitters and neuropeptides co-stored in central and peripheral neurons, in Cellular and Molecular Basis of Synaptic Transmission (Zimmermann, H., ed.), Springer Verlag, Berlin, pp. 137–151. 49. De Camilli, P. and Jahn, R. (1990) Pathways to regulated exocytosis in neurons. Annu. Rev. Physiol. 52, 625–645. 50. Matteoli, M., Haimann, C., Torri-Tarelli, F., Polak, J. M., Ceccarelli, B., and De Camilli P. (1988) Differential effect of alpha-latrotoxin on exocytosis from small synaptic vesicles and from large dense-core vesicles containing calcitonin gene-related peptide at the frog neuromuscular junction. Proc. Natl. Acad. Sci. USA 85, 7366–7370. 51. Matteoli, M., Haimann C., and De Camilli, P. (1990) Substance P-like immunoreactivity at the frog neuromuscular junction. Neuroscience 37, 271–275. 52. Rosenthal, L. and Meldolesi, J. (1989) Alpha-latrotoxin and related toxins. Pharmacol. Ther. 42, 115–134. 53. Torri-Tarelli, F., Villa, A., Valtorta, F., De Camilli, P., Greengard, P., and Ceccarelli, B. (1990) Redistribution of synaptophysin and synapsin I during α-latrotoxin-induced release of neurotransmitter at the neuromuscular junction. J. Cell Biol. 110, 449–459. 54. Meunier, F. A., Mattei, C., Chameau, P., Lawrence, G., Colasante, C., Kreger, A.S., et al. (2000) Trachynilysin mediates SNARE-dependent release of catecholamines from chromaffin cells via external and stored Ca2+. J. Cell Sci. 113, 1119–1125. 55. Herreros, J., Lalli, G., Montecucco, C., and Schiavo, G. (1999) Pathophysiological properties of clostridial neurotoxins, in The Comprehensive Sourcebook of Bacterial Protein Toxins (Freer, J. H. and Alouf, J. E., eds.), Academic Press, London, pp. 202–228. 56. Lawrence, G. W., Foran, P., and Dolly, J. O. (1996). Distinct exocytotic responses of intact

Trachynilysin

57.

58.

59.

60.

61.

62. 63. 64. 65.

66.

67. 68. 69. 70.

71.

72.

73. 74.

75.

615

and permeabilised chromaffin cells after cleavage of the 25-kDa synaptosomal-associated protein (SNAP-25) or synaptobrevin by botulinum toxin A or B. Eur. J. Biochem. 236, 877–886. Foran, P., Lawrence, G. W., Shone, C. C., Foster, K. A., and Dolly, J. O. (1996) Botulinum neurotoxin C1 cleaves both syntaxin and SNAP-25 in intact and permeabilized chromaffin cells - correlation with its blockade of catecholamine release. Biochemistry 35, 2630–2636. Chameau, C., Meunier, F. A., Kreger, A. S., Shimahara, T., and Molgó, J. (1997) Increase of intracellular Ca2+ in cultured mouse hippocampal neurons by trachynilysin, a protein toxin extracted from stonefish (Synanceia trachynis) venom. Toxicon 35, 1662. Chameau, P., Lucas, P., Melliti, K., Bournaud, R., and Shimahara, T. (1999) Development of multiple calcium channel types in cultured mouse hippocampal neurons. Neuroscience 90, 383–388. Lopez, M. G., Albillos, A., de la Fuente, M. T., Borges, R., Gandia, L., Carbone, E., et al. (1994) Localized L-type calcium channels control exocytosis in cat chromaffin cells. Pflügers Arch. Eur. J. Physiol. 427, 348–354. Albillos, A., Garcia, A. G., Olivera, B., and Gandia, L. (1996) Re-evaluation of the P/Q Ca2+ channel components of Ba2+ currents in bovine chromaffin cells superfused with solutions containing low and high Ba2+ concentrations. Pflügers Arch. Eur. J. Physiol. 432, 1030–1038. Artalejo, C. R., Adams, M. E., and Fox, A. P. (1994) Three types of Ca2+ channels trigger secretion with different efficacies in chromaffin cells. Nature 367, 72–76. Burgoyne, R. D. (1991) Control of exocytosis in adrenal chromaffin cells. Biochim. Biophys. Acta 1071, 174–202. Hamprecht, B. (1977) Structural, electrophysiological, biochemical, and pharmacological properties of neuroblastoma × glioma hybrids in cell culture. Int. Rev. Cytol. 49, 99–170. Hamill, O. P., Marty, A., Neher, E., Sakmann, B., and Sigworth, F. J. (1981) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflügers Arch. 391, 85–100. Ouanounou, G., Malo, M., Kreger, A. S., Prado de Carvalho, L., and Molgo, J. (1999) Changes in ionic permeability induced by trachynilysin in differentiated NG108-15 neuroblastoma cells. Toxicon 37, 1234. Hille, B. (1992) Ionic Channels of Excitable Membranes. Sinauer Associates, Sunderland, MA, p. 607. Finkelstein, A., Rubin, L. L., and Tzeng, M. C. (1976) Black widow spider venom: effect of purified toxin on lipid bilayer membranes. Science 193, 1009–1011. Hurlbut, W. P., Chieregatti, E., Valtorta, F., and Haimann, C. (1994) Alpha-latrotoxin channels in neuroblastoma cells. J. Membr. Biol. 138, 91–102. Zorec, R., Tester, M., Macek, P., and Mason, W. T. (1990) Cytotoxicity of equinatoxin-II from the sea anemone Actinia equina involves ion channel formation and an increase in intracellular calcium activity. J. Membr. Biol. 118, 243–249. Belmonte, G., Pederzolli, C., Macek, P., and Menestrina, G. (1993) Pore formation by the sea anemone cytolysin equinatoxin II in red blood cells and model lipid membranes. J. Membr. Biol. 131, 11–22. Macek, P., Belmonte, G., Pederzolli, C., and Menestrina, G. (1994) Mechanism of action of equinatoxin-II, a cytolysin from the sea anemone Actinia equina L. belonging to the family of actinoporins. Toxicology 87, 205–227. Shai, Y. (1995). Molecular recognition between membrane-spanning polypeptides. Trends Biochem. Sci. 20, 460–464. Ushkaryov, Y. A., Petrenko, A. G., Geppert, M., and Südhof, T. C. (1992) Neurexins: synaptic cell surface proteins related to the alpha-latrotoxin receptor and laminin. Science 257, 50–56. Davletov, B. A., Shamotienko, O. G., Lelianova, V. G., Grishin, E. V., and Ushkaryov, Y.

616

76.

77.

78.

79.

80.

81.

82.

83.

84.

85. 86.

Meunier et al. A. (1996) Isolation and biochemical characterization of a Ca2+-independent alphalatrotoxin-binding protein. J. Biol. Chem. 271, 23,239–23,245 Lelianova, V. G., Davletov, B. A., Sterling, A., Rahman, M. A., Grishin, E. V., Totty, N. F., and Ushkaryov, Y. A. (1997). Alpha-latrotoxin receptor, latrophilin, is a novel member of the secretin family of G protein-coupled receptors. J. Biol. Chem. 272, 21,504–21,508. Krasnoperov, V. G., Beavis, R., Chepurny, O. G., Little, A. R., Plotnikov, A. N., and Petrenko, A. G. (1996) The calcium-independent receptor of alpha-latrotoxin is not a neurexin. Biochem. Biophy. Res. Com. 227, 868–875. Bittner, M. A., Krasnoperov, V. G., Stuenkel, E. L., Petrenko, A. G., and Holz, R. W. (1998) A Ca2+-independent receptor for alpha-latrotoxin, CIRL, mediates effects on secretion via multiple mechanisms. J. Neurosci. 18, 2914–2922. Krasnoperov, V. G., Bittner, M. A., Beavis, R., Kuang, Y., Salnikow, K. V., Chepurny, O. G., et al. (1997) Alpha-latrotoxin stimulates exocytosis by the interaction with a neuronal G-protein-coupled receptor. Neuron 18, 925–937. Sugita, S., Ichtchenko, K., Khvotchev, M., and Südhof, T. C. (1998) Alpha-latrotoxin receptor CIRL/latrophilin 1 (CL1) defines an unusual family of ubiquitous G-protein-linked receptors. G-protein coupling not required for triggering exocytosis. J. Biol. Chem. 273, 32,715–32,724. Davletov, B. A., Meunier, F. A., Ashton, A. C., Matsushita, H., Hirst, W. D., Lelianova, V. G., et al. (1998) Vesicle exocytosis stimulated by alpha-latrotoxin is mediated by latrophilin and requires both external and stored Ca2+. EMBO J. 17, 3909–3920. Rahman, M. A., Ashton, A. C., Meunier, F. A., Davletov, B. A., Dolly, J. O., and Ushkaryov, Y. A. (1999). Norepinephrine exocytosis stimulated by alpha-latrotoxin requires both external and stored Ca2+ and is mediated by latrophilin, G proteins and phospholipase C. Phil. Trans. R. Soc. Lond. Series B 354, 379–386. Ichtchenko, K., Khvotchev, M., Kiyatkin, N., Simpson, L., Sugita, S., and Südhof, T. (1998) alpha-latrotoxin action probed with recombinant toxin: receptors recruit alphalatrotoxin but do not transduce an exocytotic signal. EMBO J. 17, 6188–6199. Orlova, E. V., Rahman, M. A., Gowen, B., Volynski, K. E., Ashton, A. C., Manser, C., et al. (2000) Structure of alpha-latrotoxin oligomers reveals that divalent cation-dependent tetramers form membrane pores. Nat. Struct. Biol. 7, 48–53. Saibil, H. R. (2000) The black widow’s versatile venom. Nat. Struct. Biol. 7, 3–4. Volynski, K. E., Meunier, F.A., Lelianova, V. G., Dudina, E. E., Volkova, T. M., Rahman, M. A., et al. (2000) Latrophilin, neurexin, and their signaling-deficient mutants facilitate alpha-latrotoxin insertion into membranes but are not involved in pore formation. J. Biol. Chem. 275, 41,175–41,183.

Neurotoxins of Cone Snail Venoms

617

28 Neurotoxins of Cone Snail Venoms* Robert Newcomb and George Miljanich

1. INTRODUCTION Cone snails are predatory marine mollusks that rely on their venom components to immobilize and capture fish, worms, or other mollusks. Cones employ a number of prey-hunting strategies, ultimately involving the injection of venom through a hollow, harpoon-like, modified-tooth structure (1). Like most animal venoms, those of the cones can contain proteins and small molecules, as well as a variety of peptides that are most often conformationally constrained by internal disulfide bridges. When envenomated, a cone’s piscine, molluscan, or vermicular prey is rapidly subdued by the concerted, high-affinity binding of the venom’s protein and peptide toxins to voltage- and ligandgated ion channels essential for the proper function of the prey’s nervous and muscular systems. Envenomations of species outside the cones’ usual fare are an occasional occurrence, and the consequences are instructive. While rare, dozens of stingings of humans by cones have been reported and about half of these, most often by the large fish hunter, Conus geographus, have proven fatal for the victims (2,3). These incidents serve as some of the earliest evidence suggesting that ion channel structural motifs have been evolutionarily conserved and that, therefore, cone toxins can bind and block mammalian ion channels homologous to those in fish and invertebrates. Several of these “crossreacting” toxins, especially those from the venoms of the fish hunters, are powerful tools for probing structural and functional aspects of these homologous mammalian channels. In addition, this cross-reactivity allows the consideration of individual venom components for medical use. Most notable among these is the calcium-channel antagonist, ω-MVIIA (also known as SNX-111 or ziconotide), which has completed latestage clinical trials as an analgesic agent (4,5). The purpose of this review is to: (1) discuss the overall nature of cone venoms, (2) cite analyses of the venom of the molluscivore, Conus textile, as a case study, and (3) summarize published work on the components of cone-snail venoms, with a focus on * This chapter is dedicated to the memory of Dr. Robert W. Newcomb. His ultimely death has robbed the scientific community of a creative and prolific contributor to our understanding of the structure and function of neuroactive agents, including the toxins derived from animal venoms. A kind and passionate friend and colleague to all who worked with him, we miss him very much. From: Handbook of Neurotoxicology, vol. 1 Edited by: E. J. Massaro © Humana Press Inc., Totowa, NJ

617

618

Newcomb and Miljanich

Table 1 Proteins and Enzymes in Cone Snail Venoms Species C. magus C. distans

Approx size (kDa) 13.6 kDa 25 kDa

C. striatus

25 kDa

C. striatus

25 kDa

C. tessulatus 55 kDa (2 chains)

C. eburneus

28 kDa

Activity (approx IC50) Phospholipase A2 Inhibition of norepinepherine release (1 nM) Inhibition of sodium channel inactivation (

Handbook of Neurotoxicology, Volume 1

Handbook of Neurotoxicology, Volume 2

Handbook of Neurotoxicology (Neurological Disease and Therapy)

Handbook of Algebra : Volume 1

Handbook of Econometrics Volume 1

Handbook of Magnetic Materials, Volume Volume 1

Credit Derivatives Handbook - Volume 1

Handbook of Income Distribution, Volume 1

Handbook of Extractive Metallurgy Volume 1

Handbook of Mathematical Fluid Dynamics, Volume 1

Handbook of Management Accounting Research, Volume 1

Handbook of Philosophical Logic Volume 1

Handbook of Offshore Engineering Volume 1

The Control Handbook. Volume 1

Handbook of Shock Waves, Volume 1

Handbook of experimental economics results, Volume 1

Handbook of Perceptual Dialectology ~ Volume 1

Handbook of Monetary Economics. Volume 1

Handbook of herbs and spices: Volume 1

Human Developmental Neurotoxicology

Handbook of the Economics of Innovation, Volume 1, Volume 1 (Handbooks in Economics)

Gallium Arsenide IC Applications Handbook (Volume 1)

Neurotoxicology: Approaches and Methods

Handbook of the Geometry of Banach Spaces, Volume 1

Handbook of Psychology, Volume 1: History of Psychology

Handbook of the Economics of Education, Volume 1

Polymer Blends Handbook, Volume 1 and Volume 2

Handbook of Philosophical Logic, Second Edition, Volume 1

Handbook of Automation Computation and Control, Volume 1: Control Fundamentals

Handbook of Software Engineering and Knowledge Engineering, Volume 1

Handbook of Differential Equations: Ordinary Differential Equations, Volume 1

Handbook of Neurotoxicology, Volume 1

Handbook of Neurotoxicology, Volume 2

Handbook of Neurotoxicology (Neurological Disease and Therapy)

Handbook of Algebra : Volume 1

Handbook of Econometrics Volume 1

Handbook of Magnetic Materials, Volume Volume 1

Credit Derivatives Handbook - Volume 1

Handbook of Income Distribution, Volume 1

Handbook of Extractive Metallurgy Volume 1

Handbook of Mathematical Fluid Dynamics, Volume 1

Handbook of Management Accounting Research, Volume 1

Handbook of Philosophical Logic Volume 1

Handbook of Offshore Engineering Volume 1

The Control Handbook. Volume 1

Handbook of Shock Waves, Volume 1

Handbook of experimental economics results, Volume 1

Handbook of Perceptual Dialectology ~ Volume 1

Handbook of Monetary Economics. Volume 1

Handbook of herbs and spices: Volume 1

Human Developmental Neurotoxicology

Handbook of the Economics of Innovation, Volume 1, Volume 1 (Handbooks in Economics)

Gallium Arsenide IC Applications Handbook (Volume 1)

Neurotoxicology: Approaches and Methods

Handbook of the Geometry of Banach Spaces, Volume 1

Handbook of Psychology, Volume 1: History of Psychology

Handbook of the Economics of Education, Volume 1

Polymer Blends Handbook, Volume 1 and Volume 2

Handbook of Philosophical Logic, Second Edition, Volume 1

Handbook of Automation Computation and Control, Volume 1: Control Fundamentals

Handbook of Software Engineering and Knowledge Engineering, Volume 1

Handbook of Differential Equations: Ordinary Differential Equations, Volume 1

Recommend Documents