The Eicosanoids
The Eicosanoids Edited by
Peter Curtis-Prior Cambridge Research Institute, Cambridge and Anglia Polytechnic University, UK
Copyright u 2004
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England Telephone (+44) 1243 779777
E-mail (for orders and customer service enquiries):
[email protected] Visit our Home Page on www.wileyeurope.com or www.wiley.com All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission in writing of the Publisher. Requests to the Publisher should be addressed to the Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, or emailed to
[email protected], or faxed to (+44) 1243 770620. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered. It is sold on the understanding that the Publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.
Other Wiley Editorial Offices John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA Wiley-VCH Verlag GmbH, Boschstr. 12, D-69469 Weinheim, Germany John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1 Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.
British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0 471 48984 0 Typeset in 9/10pt Times by Dobbie Typesetting Ltd, Tavistock, Devon Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire This book is printed on acid-free paper responsibly manufactured from sustainable forestry in which at least two trees are planted for each one used for paper production.
Contents List of Contributors vii Preface xi Preface from Prostaglandins Acknowledgements xvii Foreword xix
12 Time-resolved Fluoroimmunoassay in Eicosanoid Analysis 123 Werner Schlegel and Harald John
xiii
SECTION THREE BIOCHEMICAL AND MOLECULAR PHARMACOLOGY 129
SECTION ONE BIOSYNTHESIS AND METABOLISM 1 1
2 3 4 5 6 7
Perspectives on the Biosynthesis and Metabolism of Eicosanoids 3 Robert C. Murphy, Rebecca C. Bowers, Jennifer Dickinson and Karin Zemski Berry Control of Eicosanoid Production by Cellular and Secreted Phospholipase A2 17 Marise Andreani, Jean-Luc Olivier and Gilbert Be´re´ziat Mechanisms of PGH Synthase-1 (COX-1) Activity and Role of Radical States 29 Carol Deby and Ginette Deby-Dupont Regulation and Function of Prostaglandin Synthase-2/ Cyclooxygenase II 43 Harvey R. Herschman Mammalian Lipoxygenases 53 Shozo Yamamoto, Hiroshi Suzuki, Natsuo Ueda, Yoshitaka Takahashi and Tanihiro Yoshimoto Biosynthesis and Biological Effects of 5-oxo-ETE and Other Oxoeicosatetraenoic Acids 61 William S. Powell Synthetic Eicosanoids 69 Norrie. H. Wilson
SECTION TWO
ANALYTICAL METHODS
95
8 Perspectives of Analytical Methods for Eicosanoids 97 Jay Y. Westcott and Angelo Sala 9 Enzyme Immunoassays of Metabolites and Enzymes Using Acetylcholinesterase as Label 103 Christophe Cre´minon and The Late Jacques Maclouf 10 Bioassay of Eicosanoids 111 Robert L. Jones 11 Gas Chromatography and Mass Spectrometry in Eicosanoid Analysis 117 Michinao Mizugaki, Takanori Hishinuma, Naoto Suzuki and Junichi Goto
13 Perspectives and Clinical Significance of the Biochemical and Molecular Pharmacology of Eicosanoids 131 Subhash P. Khanapure and L. Gordon Letts 14 Eicosanoid Antagonists 163 Kiyoshi Yasui and Akinori Arimura 15 Biosynthesis and Degradation of Anandamide, an Endogenous Ligand of Cannabinoid Receptors 179 Natsuo Ueda and Dale G. Deutsch 16 Inhibitors of Eicosanoids 189 K. D. Rainsford 17 Biology and Chemistry of Products of the Isoprostane Pathway 211 L. Jackson Roberts II and Jason D. Morrow 18 Insight into Prostanoid Functions: Lessons from Receptor-knockout Mice 219 Yukihiko Sugimoto, Shuh Narumiya and Atsushi Ichikawa
SECTION FOUR IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION 227 19 Perspectives and Clinical Significance of Eicosanoids in Immunology, Endocrinology and Metabolic Regulation 229 Milan R. Henzl 20 Prostaglandins and the Immune Response 237 Diana C. Fleming and Rodney W. Kelly 21 Leukotrienes in Aspirin-intolerant Asthma 247 Anthony P. Sampson and Stephen T. Holgate 22 Essential Fatty Acids 257 Yoeju Min and Michael A. Crawford 23 Endothelial Secretory Function and Atherothrombosis 267 Stefan Chlopicki and Richard J. Gryglewski
vi
CONTENTS
24 Molecular Regulation of Pancreatic Islet Prostaglandin Synthesis and its Relevance to Diabetes Mellitus 277 R. Paul Robertson 25 Prostaglandins, Leukotrienes and Bone 289 Carol C. Pilbeam and Lawrence G. Raisz 26 Ageing and Prostaglandins 299 A. Hornych SECTION FIVE
INFLAMMATION
319
27 Perspectives and Clinical Significance of Eicosanoids in Pain and Inflammation 321 Burkhard Hinz and Kay Brune 28 Antiinflammatory Steroids 327 Franc¸oise Russo-Marie 29 Eicosanoids and Algesia in Inflammation 333 Lynne Murray, Henry Sarau and Kristen E. Belmonte 30 Cyclooxygenase-2 in Cancer 341 Ovidiu C. Trifan and Jaime L. Masferrer 31 Cytokines and Eicosanoids in Arthritis 347 K. D. Rainsford SECTION SIX
CIRCULATORY SYSTEM
359
32 Perspectives and Clinical Significance of Eicosanoids in the Circulatory System 361 The Late James B. Lee 33 Aspirin and Activated Platelets 373 Artur-Aron Weber 34 Generation of Vasoactive Prostanoids by the Cyclooxygenase-2 Pathway in the Cardiovascular System of the Rat 387 P. J. Kadowitz, S. R. Baber, M. M. Mazim, M. Keebler, H. C. Champion, T. J. Bivalacqua, D. B. McNamara and A. L. Hyman 35 Eicosanoid Generation and Effects in Cardiac Muscle and Coronary Vessels 393 Karsten Schro¨r SECTION SEVEN
DIGESTIVE SYSTEM
405
36 Perspectives and Clinical Significance of Eicosanoids in the Digestive System 407 Chi Hin Cho, Joshua Ka Shun Ko and Marcel Wing Leung Koo 37 Eicosanoids and Liver Regeneration 415 David A. Rudnick and Louis J. Muglia 38 Eicosanoids and the Intestine 423 Klaus Bukhave and Jørgen Rask-Madsen 39 Eicosanoids and Stomach Physiology 431 Brigitta M. Peskar SECTION EIGHT
NERVOUS SYSTEM
445
40 Perspectives and Clinical Significance of Arachidonic Acid Release, Action and Metabolism in the Nervous System 447 Christopher D. Breder
41 Eicosanoid Pathways in the Ageing of the Central Nervous System 457 Hari Manev and Tolga Uz 42 Arachidonate Metabolites in the Neurophysiological System: the Fever Pathway 463 Ji Zhang and Serge Rivest 43 Prostanoids in Pain 473 Tony L. Yaksh, Patrick W. Mantyh and Camilla I. Svensson 44 Eicosanoids: Roles in the Pathophysiology of Cerebral Ischaemia 481 Robert W. Hickey and Steven H. Graham 45 NSAIDs in the Treatment of Alzheimer’s Disease 487 Paul S. Aisen 46 Prostaglandins and Eicosanoids in Mental Illness 493 A.I.M. Glen and B.M. Ross 47 Essential Fatty Acids: Eicosanoid Precursors in the Treatment of Huntington’s Disease 499 Krishna Vaddadi
SECTION NINE
REPRODUCTIVE SYSTEM
507
48 Perspectives and Clinical Significance of Eicosanoids in Obstetric and Gynaecological Practice 509 I. Z. MacKenzie 49 Prostaglandins and Male Reproductive Physiology 517 Rodney W. Kelly 50 Prostaglandin F2a: the Luteolytic Hormone 525 John A. McCracken 51 Prostaglandins in Implantation 547 Norman L. Poyser 52 Parturition and the Clinical Interruption of Pregnancy 559 S. Cowan and A.A. Calder 53 Foetal and Neonatal Ductus Arteriosus 569 Kazuo Momma
SECTION TEN CONCLUSIONS AND CORRELATIONS 583 54 Biochemical Interactions of Platelet-activating Factor with Eicosanoids 585 Joseph T. O’Flaherty and Robert L. Wykle 55 Eicosanoid Precursors as Pharmaceuticals 593 The Late David F. Horrobin 56 Pharmaceutical Exploitation: Cyclooxygenase and Lipoxygenase Inhibitors 599 Paola Patrignani and Maria G. Sciulli 57 Pharmaceutical Exploitation: Eicosanoids and their Analogues 613 David F. Woodward and June Chen Epilogue Index
617
619
List of Contributors Paul S. Aisen Department of Neurology, Georgetown University Medical Center, 1 Bles Building, 3800 Reservoir Road NW, Washington, DC 20007, USA Marise Andreani Laboratoire de Physiologie et de Physiopathologie, University Pierre and Marie Curie, 9 Quai Saint Bernard, 75252 Paris Cedex 05, France Akinori Arimura Discovery Research Laboratories, Shionogi & Co. Ltd, 3-1-1 Futaba-cho, Toyonaka, Osaka 561-0825, Japan S. R. Baber Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA Kristen E. Belmonte Respiratory and Inflammation, Centre for Excellence in Drug Discovery, GlaxoSmithKline, UW2532, 709 Swedeland Road, King of Prussia, PA 19406, USA Gilbert Be´re´ziat University Pierre and Marie Curie, Case 256 Batiment a` 5 Etage, 7 Quai Saint Bernard, 75252 Paris Cedex 05, France T. J. Bivalacqua Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA Rebecca C. Bowers Division of Cell Biology, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA Christopher D. Breder Division of Neuroscience, Clinical Design and Evaluation, Bristol-Myers Squibb Co., Wallingford, CT, USA Kay Brune Department of Experimental and Clinical Pharmacology and Toxicology, Friedrich Alexander University Erlangen-Nu¨rnberg, Fahrstrasse 17, 91054 Erlangen, Germany Klaus Bukhave The Royal Veterinary and Agricultural University, Copenhagen, Denmark
June Chen Department of Biological Sciences, Allergan Inc., 2525 Dupont Drive, Irvine, CA 92612, USA Stefan Chlopicki Department of Pharmacology, Jagiellonian University Medical College, Grzego´rzecka 16, 31531 Krako´w, Poland Chi Hin Cho Department of Pharmacology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China S. Cowan Department of Obstetrics and Gynaecology, University of Edinburgh, UK Michael A. Crawford Institute of Brain Chemistry and Human Nutrition, London Metropolitan University, 166–220 Holloway Road, London N7 8DB, UK Christophe Cre´minon CEA, Service de Pharmacologie et d’Immunologie, De´partement de Recherche Me´dicale, CEA-Saclay, 91191 Gif-sur-Yvette Cedex, France Carol Deby Centre for Oxygen Research and Development, Institut de Chimie, B6a, Universite´ de Lie`ge, 4000 Lie`ge, Belgium Ginette Deby-Dupont Centre for Oxygen Research and Development, Institut de Chimie, B6a, Universite´ de Lie`ge, 4000 Lie`ge, Belgium Dale G. Deutsch Department of Biochemistry and Cell Biology, State University of New York at Stony Brook, Stony Brook, NY 11794-5215, USA Jennifer Dickinson Division of Cell Biology, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA Diana C. Fleming Medical Research Council Human Reproductive Science Unit, University of Edinburgh Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9ET, UK
A. A. Calder Department of Obstetrics and Gynaecology, University of Edinburgh, UK
Junichi Goto Department of Pharmaceutical Sciences, Tohoku University Hospital, 1-1 Seiryo-machi Aoba-ku, Sendai 980-8574, Japan
H. C. Champion Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA
A. I. M. Glen Ness Foundation, UHI Millennium Institute, Ness House, Dochfour Business Centre, Inverness IV3 8GY, UK
viii
CONTRIBUTORS
Steven H. Graham Department of Neurology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA Richard J. Gryglewski Department of Pharmacology, Jagiellonian University Medical College, Grzego´rzecka 16, 31-531 Krako´w, Poland Milan R. Henzl Stanford University Medical School, 4210 Ynigo Way, Palo Alto, CA 94306, USA Harvey R. Herschman Department of Biological Chemistry, David Geffen School of Medicine at UCLA, 341 Boyer Hall, 611 Charles E. Young Drive East, Los Angeles, CA 90095-1570, USA Robert W. Hickey Department of Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA Burkhard Hinz Department of Experimental and Clinical Pharmacology and Toxicology, Friedrich Alexander University Erlangen-Nu¨rnberg, Fahrstrasse 17, 91054 Erlangen, Germany Takanori Hishinuma Department of Pharmaceutical Sciences, Tohoku University Hospital, 1-1 Seiryomachi, Aoba-ku, Sendai 980-8574, Japan Stephen T. Holgate Respiratory Cell and Molecular Biology, Division of Infection, Inflammation and Repair, Southampton University School of Medicine, Southampton, UK
Philip J. Kadowitz Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA M. Keebler Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA Rodney W. Kelly Medical Research Council Human Reproductive Science Unit, University of Edinburgh Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9ET, UK Joshua Ka Shun Ko Department of Pharmacology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China Subhash P. Khanapure Nitro Med Inc, 12 Oak Park Drive, Bedford, MA 01730, USA The Late James B. Lee* School of Medicine and Biomedical Sciences, University at Buffalo, USA Gordon Letts NitroMed Inc., 12 Oak Park Drive, Bedford, Massachusetts 01730, USA Marcel Wing Leung Koo Department of Pharmacology, Faculty of Medicine, Hong Kong Baptist University, Hong Kong, China I. Z. MacKenzie Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK
A. Hornych Department of Nephrology, European Hospital George Pompidou, 20 Rue Leblanc, 75908 Paris Cedex 15, France
The Late Jacques Maclouf U348 INSERM I.F.R. Vaisseaux-Lariboisie`re, Hoˆpital Lariboisie`re, 8 Rue Guy Patin, 75475 Paris Cedex 10, France
The Late David F. Horrobin{ Chairman, Laxdale Ltd, Kings Park House, Laurelhill Business Park, Stirling FK7 9JQ, UK
Hari Manev The Psychiatric Institute, Department of Psychiatry, University of Illinois at Chicago, 1601 West Taylor Street, Chicago, IL 60612, USA
A. L. Hyman Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA
Patrick W. Mantyh Department of Preventive Sciences, University of Minnesota, Minneapolis, MN 55455, USA
Atsushi Ichikawa Department of Physiological Chemistry, Graduate School of Pharmaceutical Sciences and Department of Pharmacology, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto 6068501, Japan
Jaime L. Masferrer Oncology Department, AA5C, Pharmacia Corporation, 700 Chesterfield Parkway North, Chesterfield, MO 63198, USA M. M. Mazim Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA
Harald John IPF Pharmaceuticals GmbH, 30625 Hannover, Germany
John A. McCracken Department of Animal Science, University of Connecticut, Storrs, CT 06269-4040, USA
Robert L. Jones Department of Pharmacology, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong
D. B. McNamara Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA 70112, USA
{ Correspondence to Dr Crispin Bennett, Research Information Manager (
[email protected])
*Correspondence to Professor Peter Curtis-Prior, Editor (
[email protected]
CONTRIBUTORS
ix
Yoeju Min Institute of Brain Chemistry and Human Nutrition, London Metropolitan University, 166–220 Holloway Road, London N7 8DB, UK
K. D. Rainsford Biomedical Research Centre, School of Science and Mathematics, Sheffield Hallam University, Howard Street, Sheffield S1 1WB, UK
Michinao Mizugaki Department of Pharmaceutical Sciences, Tohoku University Hospital, 4-4-1 Komatsushima, Aoba-ku, Sendai, 981-8558 wcl, Japan
Lawrence G. Raisz University of Connecticut Health Center, 63 Farmington Avenue, Farmington, CT 06030-2806, USA
Kazuo Momma Tokyo Women’s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162-8666, Japan
Jørgen Rask-Madsen Department of Medical Gastroenterology, Herlev Hospital, University of Copenhagen, Denmark
Jason D. Morrow Departments of Pharmacology and Medicine, Vanderbilt University, Nashville, TN 37232, USA Louis J. Muglia Department of Pediatrics, Washington University School of Medicine, St Louis, MO 63110, USA Robert C. Murphy Division of Cell Biology, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA
Serge Rivest Laboratory of Molecular Endocrinology, CHUL Research Center and Department of Anatomy and Physiology, Laval University, 2705, Boul. Laurier, Que´bec, Canada G1V 4G2 L. Jackson Roberts II Departments of Pharmacology and Medicine, Vanderbilt University, Nashville, TN 37232, USA R. Paul Robertson Pacific Northwest Research Institute, Seattle, WA 98122, USA
Lynne Murray Imperial College School of Medicine, SAF Building, Exhibition Road, Leukocyte Biology, BMS Division, South Kensington, London SW7 2AZ, UK
B. M. Ross Ness Foundation, UHI Millennium Institute, Ness House, Dochfour Business Centre, Inverness IV3 8GY, UK
Shuh Narumiya Department of Physiological Chemistry, Graduate School of Pharmaceutical Sciences and Department of Pharmacology, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto 6068501, Japan
David A. Rudnick Department of Pediatrics, Washington University School of Medicine, St Louis, MO 63110, USA
Joseph T. O’Flaherty Department of Internal Medicine, Section on Infectious Disease, Wake Forest University School of Medicine, Winston Salem, NC, USA
Angelo Sala Department of Pharmacological Sciences, Via Balzaretti, 20133 Milan, Italy
Jean-Luc Olivier Laboratory of Medical Biology, Faculty of Medicine of Nancy, France Paola Patrignani Department of Medicine and Aging, Division of Pharmacology, ‘‘G. D’Annunzio’’ University of Chieti School of Medicine, 66013 Chieti, Italy Brigitta M. Peskar Department of Experimental Clinical Medicine, Ruhr-University of Bochum, Bochum, Germany Carol C. Pilbeam University of Connecticut Health Center, 63 Farmington Avenue, Farmington, CT 06030-2806, USA William S. Powell Meakins-Christie Laboratories, Department of Medicine, McGill University, 3626 St Urbain Street, Montreal, Quebec, Canada H2X 2P2 Norman L. Poyser Division of Biomedical and Clinical Laboratory Sciences, University of Edinburgh, Hugh Robson Building, George Square, Edinburgh EH8 9XD, UK
Franc¸oise Russo-Marie BIONEXIS, CEA Saclay Baˆtiment 520, 91191 Gif-sur-Yvette Cedex, France
Anthony P. Sampson Respiratory Cell and Molecular Biology, Division of Infection, Inflammation and Repair, Southampton University School of Medicine, Southampton, UK Henry Sarau Respiratory and Inflammation, Centre for Excellence in Drug Discovery, GlaxoSmithKline, UW2532, 709 Swedeland Road, King of Prussia, PA 19406, USA Werner Schlegel Universita¨tsklinikum Mu¨nster, Klinik und Poliklinik fu¨r Frauenheilkunde und Geburtshilfe, Albert-Schweitzer-Strasse 33, D-48149 Mu¨nster, Germany Karsten Schro¨r Institut fu¨r Pharmakologie und Klinische Pharmakologie, Universita¨tsklinikum Du¨sseldorf, Moorenstrasse 5, D-40225 Du¨sseldorf, Germany Maria G. Sciulli Department of Medicine and Aging, Division of Pharmacology, ‘‘G. D’Annunzio’’ University of Chieti School of Medicine, 66013 Chieti, Italy Yukihiko Sugimoto Department of Physiological Chemistry, Graduate School of Pharmaceutical Sciences and
x
CONTRIBUTORS
Department of Pharmacology, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto 6068501, Japan Hiroshi Suzuki Department of Biochemistry, Tokushima University, School of Medicine, Kuramoto-cho, Tokushima 770-8503, Japan Naoto Suzuki Department of Pharmaceutical Sciences, Tohoku University Hospital, 1-1 Seiryo-machi, Aobaku, Sendai 980-8574, Japan Camilla I. Svensson Department of Anesthesiology, University of California, San Diego, La Jolla, CA 92093, USA Yoshitaka Takahashi Department of Molecular Pharmacology, Kanazawa University Graduate School of Medicine, 13-1 Takara-machi, Kanazawa 920-8640, Japan Ovidiu C. Trifan Oncology Department, AA5C, Pharmacia Corporation, 700 Chesterfield Parkway North, Chesterfield, MO 63198, USA Natsuo Ueda Department of Biochemistry, Kagawa University School of Medicine, Miki-cho, Kita-gun, Kagawa 761-0793, Japan Tolga Uz The Psychiatric Institute, Department of Psychiatry, University of Illinois at Chicago, 1601 West Taylor Street, Chicago, IL 60612, USA Krishna Vaddadi Department of Psychological Medicine Monash Medical Centre, (Monash University), 246 Clayton Road, Clayton 3186, Victoria, Australia Artur-Aron Weber Institut fu¨r Pharmakologie und Klinische Pharmakologie, Universita¨tsklinikum Du¨sseldorf, Moorenstrasse 5, D-40225 Du¨sseldorf, Germany
Jay Y. Westcott National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA Norrie H. Wilson Division of Biomedical Sciences, University of Edinburgh, 1–3 George Square, Edinburgh EH8 9XD, UK David F. Woodward Department of Biological Sciences, Allergan Inc., 2525 Dupont Drive, Irvine, CA 92612, USA Robert L. Wykle Department of Biochemistry, Wake Forest University School of Medicine, Winston Salem, NC, USA Tony Yaksh Department of Anesthesiology, University of California, San Diego, La Jolla, CA 92093, USA Shozo Yamamoto Department of Food and Nutrition, Faculty of Home Economics, Kyoto Women’s University, Imakumano, Higashiyama-ku, Kyoto 605-8501, Japan Kiyoshi Yasui Discovery Research Laboratories, Shionogi & Co. Ltd, 3-1-1 Futaba-cho, Toyonaka, Osaka 561-0825, Japan Tanihiro Yoshimoto Department of Molecular Pharmacology, Kanazawa University Graduate School of Medicine, 13-1 Takara-machi, Kanazawa 920-8640, Japan Karin Zemski Berry Division of Cell Biology, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA Ji Zhang Laboratory of Molecular Endocrinology, CHUL Research Center and Department of Anatomy and Physiology, Laval University, 2705 Boul. Laurier, Que´bec, Canada G1V 4G2
Preface As the year 2000 approached, there developed in people around the world a sense of anticipation, positive for some but negative for others. However, in both groups of people there was a deep desire to mark such a rare moment in history. It was at this time that it was first suggested to me as the ideal moment for a new, comprehensive volume on the biomedical significance of polyunsaturated acids and their derivatives, bringing together the important discoveries and developments which had taken place in this area since the publication of a previous book with which I had been involved (Curtis-Prior 1988). It was that prompting which has led to this latest book, The Eicosanoids (see also Acknowledgements). So, although what might be considered as ‘‘. . . important discoveries and developments . . .’’ is to some extent a subjective exercise and certainly every chapter in this book reveals developments of fascinating new knowledge in its field of specialism, there has been one landmark advance during these fifteen years which, among much else, has reconciled the seeming paradox that prostaglandins may be both harmful and beneficial. This was, of course, the identification and characterization of an alternative cyclooxygenase enzyme (Fu et al 1990; Masferrer et al 1990; Herschman et al 1993), i.e. the proposal of the coexistence of the known enzyme responsible for basal, constitutive prostaglandin synthesis (COX-1) together with a new enzyme (designated COX-2) implicated in various inflammatory and ‘‘induced’’ settings (Funk 2001; see also Chapter 4, this volume, Regulation and Function of Prostaglandin Synthase-2/Cyclooxygenase II). This revelation provoked a whole new wave of discovery research activity in acadaemia and in the pharmaceutical industry. There followed a race to develop a new class of COX-2 selective inhibitors which inhibited prostaglandin production in inflammatory cells without the, hitherto, inevitable undesirable interference on the production of COX-1-generated prostaglandins fulfilling their physiological regulatory roles in the gastrointestinal tract (Warner et al 1999) in renal function and in the reproductive and haemostatic systems as is presented in superb detail elsewhere in this book (see for example Chapters 13, this volume, Perspectives and Clinical Significance of the Biochemical and Molecular Pharmacology of Eicosanoids and Chapter 16 Inhibitors of Eicosanoids). Some of the early fruits of this burst of scientific creativity were presented at specialized conferences in Bangkok and Boston and recorded in a state of the art ‘‘book-of-the-meetings’’ edited by Vane and Botting (1998), and this was later succeeded by their comprehensive monograph on the subject, entitled Therapeutic Roles of Selective COX-2 Inhibitors (Vane and Botting 2001). The Pharmacia drug Celebrex (celecoxib) was the first of this new generation of COX-2 selective inhibitors to receive marketing authorization (by the US FDA) for the treatment of inflammatory conditions (Penning et al 1997) and the second was Merck Sharp & Dohme’s Vioxx (rofecoxib) (Chan et al 1999; Prasit et al 1999). In a pivotal trial, when celecoxib was compared with ibuprofen and diclofenac in a 6 months-long arthritis safety study (CLASS), it was reported (Silverstein et al 1999) that the COX-2 inhibitor was associated with a lower incidence of symptomatic ulcers and
ulcer complications than the traditional COX-1 non-steroidal antiinflammatory drugs, as had been the result anticipated. However, it appears that there was some incongruity (Okie 2001; Berg et al 2001; Wright et al 2001) between this account from Silverstein and his co-workers and ‘‘. . . the complete information available to the United States Food and Drug Administration . . .’’ (Juni et al 2002) in what has been described as a ‘‘. . . failure in the therapeutic chain as a cause of drug ineffectiveness . . .’’ (Figueras and Laporte 2003). However, it should be noted that a subsequent major study (designated VIGOR) to compare the incidence of upper gastrointestinal events provoked by MSD’s selective COX-2 inhibitor rofecoxib and the non-selective NSAID naproxen, in patients with rheumatoid arthritis, demonstrated significant clinical advantages for rofecoxib (Bombardier et al 2000). Whatever might be the eventual outcome of this particular and complex situation, where we well know that the worldwide clinic is the ultimate challenge of any therapeutic product, the drug target remains fascinatingly attractive, especially for industry, in the light of a global market for analgesics estimated at US $10 billion.* Interestingly, there have also been reports of a herbal preparation called NexrutineTM (Lavelle 2003). This is a new patent-pending dietary supplement containing an extract of the plant Phellodendron amurense and has been reported to be useful in the treatment of inflammatory diseases. It possesses COX-2inhibitory qualities, but also protects the gastrointestinal tract against ulceration. In clinical studies, it was shown that Nexrutine inhibits COX-2 without interfering with the activity of COX-1. In animal studies, Nexrutine has proved to be as effective as naproxen in reducing pain and inflammation. However, its mechanism of action has been reported to be different from that of what might be termed true COX-2 inhibitors, since it does not act directly on the cyclooxygenase, but inhibits the gene responsible for the production of COX-2, as well as other inflammatory mediators. Other novel developments treated in this book include: endogenous ligands of cannabinoid receptors, referred to as endocannabinoids (Mechoulam et al 1998; see also Chapter 15, this volume, Biosynthesis and Degradation of Anandamide, an Endogenous Ligand of Cannabinoid Receptors); identification of the noxious roles of cysteinyl leukotrienes in aspirin-intolerant asthma (Samson and Holgate 1999; see also Chapter 21, this volume, Leukotrienes in Aspirin-intolerant Asthma); prostaglandin receptor knock-out mice (Negishi and Katoh 2003; see also Chapter 18, this volume, Insight into Prostanoid Functions: Lessons from Receptor-knockout Mice); the biosynthesis and roles of isoprostanes (Morrow et al, 1990; see also Chapter 17, this volume, Biology and Chemistry of Products of the Isoprostane Pathway) and their utility in the assessment of oxidative stress status in vivo, in the clinic. Finally there is in this volume (see also below) a wealth of new knowledge shared over a broad range of the actions and *Current nomenclature has been adopted, taking 1 billion (bn) as 109 rather than the traditional value of 1012.
xii
PREFACE
interactions of eicosanoids in biomedicine and physiology, from immunology to central nervous system disorders. The book is divided into thematic sections (see Contents) each of which begins with a chapter designated ‘‘Perspectives and clinical significance . . .’’ The intention is for these ‘‘Perspectives’’ chapters to provide an integrated synopsis of the subject area covered in the respective Section, indicating how the different specialist chapters (strands of the theme) relate to one another in the overall context of the theme. In this way, it is hoped that THE EICOSANOIDS may interest, encourage and inspire all who are fascinated by the biosciences in seeking further to understand more about the myriad roles of essential fatty acids to the benefit of human wellbeing. Notwithstanding the unpredictability of the peaks and troughs in the evolution and development of any discipline, it appears valid, still, to look forward to even more fascinating discoveries ahead through the immortal words from Robert Browning, cited on a previous occasion (Curtis-Prior 1983): Grow old along with me The best is yet to be . . .
Cambridge 2004 Peter Curtis-Prior
[email protected] REFERENCES Berg Hrach J and Mora M (2001) Reporting of 6-month vs. 12-month data in a clinical trial of celecoxib. J Am Med Assoc, 286, 2398. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Eng J Med, 343, 1520–1528. Chan CC, Bouce S, Brideau C et al (1999) Rofecoxib [Vioxx, MK-0966; 4(4’-methylsulfonylphenyl)-3phenyl-2-(5H)-furanone]: a potent and orally active cyclooxygenase-2 inhibitor. Pharmacological and biochemical profiles. J Pharmacol Exp Therapeut, 290, 551–560. Curtis-Prior PB (ed.) (1988) Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids. Edinburgh: Churchill Livingstone. Curtis-Prior PB (1983) Trends in prostaglandin studies [Orientations des e´tudes actuelles sur les prostaglandines]. Curr Clin Concepts, 1, 9–19. Di Marzo V (1998) Endocannabinoids and other fatty acid derivatives with cannabimimetic properties: biochemistry and possible physiopathological relevance. Biochim Biophys Acta, 1392, 153–175. Figueras A and Laporte J-R (2003) Failures of the therapeutic chain as a cause of drug ineffectiveness. Br Med J, 326, 895–896. Fu JY, Masferrer JL, Seibert K et al (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740. Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science, 294, 1871–1875.
Herschman HR, Fletcher BS and Kujubu DA (1993) TIS10, a mitogeninducible glucocorticoid-inhibited gene that encodes a second prostaglandin synthase/cyclooxygenase enzyme. J Lipid Mediat, 6, 89–99. Juni P, Rutjes AWS and Dieppe PA (2002) Are selective COX-2 inhibitors superior to traditional non-steroidal antiinflammatory drugs? Adequate analysis of the CLASS trial indicates that this may not be the case. Br Med J, 324, 1287–1288. Lavelle JB (ed.) (2003) The COX-2 Connection: Natural Breakthrough Treatments for Arthritis, Alzheimer’s and Cancer. Rochester VT: Healing Arts Press. Masferrer JL, Zweifel BS, Seibert K et al (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. Mechoulam R, Fride E and Di Marzo V (1998) Endocannabinoids. Eur J Pharmacol, 359, 1–18. Morrow JD, Hill KE, Burk RF et al (1990) A series of prostaglandin F2like compounds are produced in vivo in humans by a noncyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci USA, 87, 9383–9387. Negishi M and Katoh H (2002) Cyclopentanone prostaglandin receptors. Prostagland Lipid Mediat, Special Issue: Molecular Biology of the Arachidonate Cascade, Guest Eds S Yamamoto and WL Smith, 68–69, 611–617. Okie S (2001) Missing data on Celebrex. Full study altered picture of drug. Washington Post, 5 (Aug), A11. Penning TD, Talley JJ, Bertenshaw SR et al (1997) Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase2 inhibitors: identification of 4-[5-(4-methyl phenyl)-3-(trifluoromethyl)1H-pyrazol-1-yl]benzenesulfonamide (SC-58635, Celecoxib). J Med Chem, 40, 1347–1365. Prasit P, Wang Z, Brideau C et al (1999) The discovery of rofecoxib [MK966, Vioxx, 4-(methylsulfonyl phenyl)3-phenyl-2(5H)-furanone], an orally active cyclooxygenase-2 inhibitor. Bioorg Med Chem Lett, 9, 563–567. Sampson AP and Holgate ST (1999) Leukotriene Modifiers in Asthma Treatment. London: Martin Dunitz. Silverstein FE, Faich G, Goldstein JL et al (2000) Gastrointestinal toxicity with celecoxib vs. non-steroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. J Am Med Assoc, 284, 1247–1255. Vane JR and Botting RM (1998) Clinical Significance and Potential of COX-2 Inhibitors. London: William Harvey Press. Vane JR and Botting RM (2000) Therapeutic Roles of Selective COX-2 Inhibitors. London: William Harvey Press. Warner TD, Giuliano F, Vojnovic I et al (1999) Non-steroid drug selectivities for cyclooxygenase-1 rather than cyclooxygenase-2 are associated with human gastrointestinal toxicity: a full in vitro analysis. Proc Natl Acad Sci USA, 96, 7563–7568. Wright JM, Perry TL, Bassett KL and Chambers KG (2001) Reporting of 6-month vs. 12-month data in a clinical trial of celecoxib. J Am Med Assoc, 286, 2398.
Preface from Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids, edited by P. B. Curtis-Prior (1988) My purpose in preparing this volume was to provide, in a single source, a comprehensive collection of information on prostaglandins and related eicosanoids.* The 57 chapters presented here are divided into ten sections, according to subject, and are the work of a total of almost 80 authors from over a dozen countries, distributed among four of the world’s continents. Although the chapters have been edited I have tried, as far as possible, to retain the original style of each author, or group of authors, in order to present a richer, international, intercultural final product. The book has been over 5 years in the making and, in spite of the more recent possibility for authors to carry out updating procedures on their contributions, must remain less than comprehensive owing, in part, to the constant influences of man’s innate curiosity and creativity providing new knowledge more quickly even than the old can be recorded. However, it is perhaps germane at this time to recall the maxim below, suggested by Dr John A. McCracken (The Worcester Foundation, Shrewsbury, Massachusetts), with which many of the ‘‘initiated’’ in this field will, no doubt, readily identify: The one who places the last stone and steps across to the terra firma of accomplished discovery gets all the credit. Only the initiated know and honor those whose patient integrity and devotion to exact observation have made the last step possible. Hans Zinsser
The demonstration of the essentiality of dietary linoleic acid, and the coining of the term ‘‘essential fatty acid (EFA)’’ by Burr and Burr (1930), the subsequent dawning of the real significance of this observation, plus the preliminary clinical studies of Kurzrok and Lieb (1930) showing the uterine muscle-stimulating properties of human semen, followed shortly afterwards by the observations of hypotensive actions of semen and acidic lipid extracts thereof by Goldblatt (1935) and von Euler (1935), really set the scene for the story recounted in this volume, a story which is certainly not yet completed. Since the 1930s there has been an enormous growth and development of interest and investigation which, however, has not been a continuum. Indeed, following the initial spate of papers of the ‘past historic’ phase of attention to prostaglandins (CurtisPrior 1985), there was a clear arrest of devotion during the years of the Second World War, prior to a resurgence—including the first isolation of prostaglandins (Bergstrom and Sorjvall 1957) in the postwar years, since when there has developed a rate of issue of 4.5–5.06103 publications each year, i.e. approaching 100 per
*‘‘Eicosanoids’’ is a useful term coined by Corey EJ, Nuira H, Falek JR, Mioskowski C, Arai Y, Marfat A (1980) Recent studies on the chemical synthesis of eicosanoids. Advances in Prostaglandin and Thromboxane Research, 6: 19–25 . . . ‘‘to describe the broad group of compounds derived from C20 fatty acids . . .’’
week.{ The origin of this present and continuing interest in prostanoids is not entirely clear, but probably stems to a large degree from the independent and virtually simultaneous findings relating to the actions of non-steroidal anti-inflammatory drugs on prostaglandin biosynthesis in France and in Britain, especially of the effects of the most widely-used drug substance the world has yet known: aspirin, as was surmised previously by Collier (1971). At L’Institut Pasteur in Paris, Vargaftig and co-workers had spent several years examining phospholipases A and C, and showed subsequently that aspirin could block the actions of vasoactive substances normally formed by phospholipases from guinea-pig lung (Vargaftig and Dao 1971). Meanwhile, Willis and Smith at the Royal College of Surgeons of England in London had observed that the production of prostaglandins by blood platelets was presented in the presence of aspirin. The significance of these observations was quickly recognized, with legendary vision, by Vane (also at the Royal College), who corroborated them using guinea-pig lung homogenate; this led to the famous trilogy of papers published in Nature that year (Ferreira et al 1971; Smith and Willis 1971; Vane 1971). Subsequently, the work of Morley and Williams (1973) demonstrated an enhancement of paw oedema by prostaglandins which entirely corroborated the association of prostaglandins and inflammation. Although the greatest interest and the major proportion of published data relate to arachidonic acid and its metabolites, there is increasing attention being paid also to eicosanoids of the 1-series and of the 3-series. Considerable public interest has been provoked latterly in health food preparations of seed oil of the evening primrose, an oleaginous (willow herb) plant of the genus Oenothera misnamed because of the resemblance of its bright yellow flowers to those of real primroses. This plant’s seed is rich in linoleic acid (72%) and contains also g-linolenic acid (c.9%). It is the latter component which is proposed as the active principle through conversion to prostaglandins of the 1-series (mainly E1) in remarkably wideranging therapeutic claims (Graham 1984). This is compatible with the wide-held belief that our main source of arachidonic acid is probably the meat of our diet (i.e. other animals’ bodies) and, since on the one hand the 5-desaturase enzyme catalysing the further transformation of dihomo-g-linolenic acid to arachidonic acid is of minor importance in man, and on the other hand that activity of the 6-desaturase responsible for the formation of glinolenic acid from linoleic acid is much reduced by our modern lifestyle, as well as the pathologies of severe hepatic insufficiency, undernutrition, insulin-dependent diabetes and senescence (Paccalin et al 1984), it is tacitly assumed that additional dietary glinolenic acid is converted mainly to prostaglandins of the 1-series. { These data were kindly supplied by the Information Services (Corporate Technical Library), Upjohn Company, Kalamazoo, USA and calculated under the revised system, 1983.
xiv
PREFACE FROM PROSTAGLANDINS
However, the identification of other angiosperm plant seeds containing increased amounts of g-linolenic acid, like the blackcurrant (Ribes nigrum) and borage (Borago officianalis), and the recent discovery, in several parts of the world, of the very important nutritional potential of the blue-green alga of the genus Spirulina (Chan et al 1981; Ciferri 1983; Devi and Venkataraman 1983) containing much higher concentrations of g-linolenic acid synthesized by direct desaturation of linoleic acid, suggest they may shortly eclipse the present fashionable interest in ‘‘primrose oil’’. Indeed, recent advances in biotechnology could replace agricultural seed oil production altogether as a source of glinolenic and other polyunsaturated fatty acids of therapeutic potential (Ratledge 1987). The origins of man’s cognisance of the potential therapeutic importance of prostaglandins of the 3-series and their precursor polyunsaturated fatty acids has been indicated recently by Sinclair (1985).* About 3500 years ago, the Israelites were bored by making bricks without straw, and Moses (or perhaps Aaron) hit the sea with a rod—a poor tool for that purpose but the Lord was on his side. All the fish died. Shortly thereafter Pharoah had an infarct followed by an epidemic of hardening of the heart amongst the Egyptians. This is the earliest indication of the importance of fish in preventing myocardial infarction.
More recently, the extremely low incidence of myocardial infarction and a tendency to bleed attributed to an Eskimo diet has emphasized the apparent protective aspects of fish oils (Dyerberg et al 1975), although in rats fed cod liver oil PGI3 was not detectable (Hornstra and Nugteren 1981). While this observation has thrown doubt on a thromboregulatory role for 3series prostaglandins, it cannot dispute the observed non-genetic low cardiovascular morbidity and mortality of Eskimo populations and may indicate the identity of the protective principle as the polyunsaturated fatty acid (potential prostaglandin precursor) eicosapentanoic acid (Dyerberg et al 1978). However, it may be significant that furanoid or urofuran acid compounds which have been shown to be minor components of fish oils (Glass et al 1975) may also be credited with a possible cardiovascular protective role in view of their potent hypolipidaemic properties in animal models in reducing both blood cholesterol and triglycerides (Hall 1985). The recently-discovered arachidonic acid metabolites lipoxins, formed by human leukocytes stimulated with calcium ionophore (Serhan et al 1984), appear to be involved in inflammation and their possible interactions and/or associations with other eicosanoids are discussed in several contexts in this volume. Apart from the arachidonic acid metabolites, phospholipase A2 may give rise also to another lipid mediator, platelet activating factor (PAF, PAF-acether or acetyl-glyceryl-ether-phosphorylcholine) (Benveniste et al 1972), which was first recognized by its degranulating activity of rabbit platelets (Siragnanian and Osler 1971), implicated as a mediator of anaphylaxis (Henson and Pinckard 1977) and subsequently implicated in most aspects of cellular inflammation (Benveniste and Arnoux 1983). Because of the unceasing evidence of the intimate associations of eicosanoids with PAF in pathophysiological processes, a sign-post chapter (Ch. 55) has been incorporated in the final part of this volume. The recognition of the cytokine peptide interleukin-1 (IL-1) as an inflammatory mediator inducing fever, the release of acute phase proteins and cartilage degradation led naturally to the
*This is quoted with permission from a pre-publication script kindly supplied by Professor Hugh Sinclair of his pre-dinner presentation entitled: ‘‘History of EFA and their prostanoids: some personal reminiscences’’, on the occasion of the Second International Congress on Essential Fatty Acids and their Eicosanoids.
supposition of prostaglandin involvement in these pathophysiological processes. However, it is only recently that IL-1 has been clearly shown to activate cellular phospholipase A2 and thus suggests the whole gamut of arachidonic acid metabolites and PAF as its potential secondary mediators (Chang et al 1986). Superimposed on this are the complicating preliminary observations that PAF itself may IL-1 production from human macrophages (Barrett et al 1987). The early implication of eicosanoids in inflammation, referred to above, is evidently the visible tip of a complex metaphorical iceberg wherein the indicated interrelationship of prostanoids/ eicosanoids with PAF, lipoxins and cytokines and kinins are only now beginning to be unravelled. These exciting developments in inflammation and the body homeostatic defence mechanisms, those related to cardiovascular disease and diet, recent implications of PAF in nidation and a potential further means of regulating fertility (O’Neil 1986), and the most recent suggestion of a role for eicosanoids as second messengers in the nervous system (Piorrelli et al 1987) are but a few of the potentially exciting avenues awaiting further elucidation in the future of ‘‘prostaglandins’’, and may explain why often ‘‘the true innovatory researcher cannot define the eventual length or cost of a research programme’’ (Celsus 1985). P.B.C-P. Cambridge, 1988 REFERENCES Barrett ML, Lewis GP, Ward S and Westwick J (1987) Platelet activating factor induces interleukin-1 production from human adherent macrophages. Br J Pharmacol, 90, 113P. Benveniste J and Arnoux B (1983) Platelet-activating Factor and Structurally-related Ether Lipids. Amsterdam: Elsevier. Benveniste J, Henson PM, Cochrane CG 1972 Leukocyte-dependent histamine release from rabbit platelets. The role of IgE, basophils and a platelet-activating factor. J Exp Med, 136, 1356–1376. Bergstrom S and Sjovall J (1957) The isolation of prostaglandin. Acta Chem Scand, 11, 1086. Burr GO, Burr MM (1930) On the nature and the role of fatty acids essential in nutrition. J Biol Chem, 86, 587–621. Chang J, Gilman SC and Lewis AJ (1986) Interleukin 1 activates phospholipase A2 in rabbit chondrocytes: a possible signal for IL 1 action. J Immunol, 136, 1283–1287. Celsus (1985) Primroses and prostaglandin. Pharm J, 235, 607–608. Chen L-C, Chen J-S and Tung T-C (1981) Effects of Spirulina on serum lipoproteins and its hypocholesterolemic effect. J Formosan Med Assoc, 80, 934–942. Ciferri O (1983) Spirulina, the edible microorganism. Microbiol Rev, 47, 551–578. Collier HOJ (1971) Prostaglandins and aspirin. Nature New Biol, 231, 17– 19. Curtis-Prior PB (1985) Trends in prostaglandin studies. Curr Clin Concepts, 1, 9–19. Dyerberg J, Bang HO and Hjorne N (1975) Fatty acid composition of the plasma lipids in Greenland eskimos. Am J Clin Nutrit, 28, 958–966. Dyerberg J, Bang HO, Stofferson E et al (1978) Eicosapentanoic acid and prevention of thrombosis and atherosclerosis. Lancet, ii, 117–119. Devi MA and Venkaturaman LV (1983) Hypocholesterolemic effect of blue green algae Spirulina platensis in albino rats. Nutrit Rep Int, 28, 519–530. von Euler US (1935) A depressor substance in the vesicular gland. J Physiol, 84, 21P. Ferreira SM, Moncada S and Vane JR (1971) Indomethacin and aspirin abolish prostaglandin release from the spleen. Nature, 231, 237–239. Glass, RL, Krick TP, Sand DM et al (1975) Furanoid fatty acids from fish lipids. Lipids, 10, 695–702. Goldblatt MW (1935) A depressor substance in seminal plasma. J Physiol, 84, 201–218. Graham J (1984) Evening Primrose Oil. Wellingborough, UK: Thorsons.
PREFACE FROM PROSTAGLANDINS Hall H (1985) The hypolipidaemic activity of furanoic acid and furyl acrylic acid derivatives in rodents. Pharm Res, 5, 233–238. Henson PM and Pinckard RN (1977) Basophil derived platelet activating factor (PAF) as an in vivo mediator of acute allergic reactions. Demonstration of specific desensitization of platelets to PAF during IgE-induced anaphylaxis in the rabbit. J Immunol, 119, 1279–1286. Hornstra G and Nugteren D (1981) Fish oil feeding does not result in the endogenous formation of PGI3 in rats. Progr Lipid Res, 20, 911. Kurzrok R and Lieb CC (1930) Biochemical studies of human semen. The action of semen on the human uterus. Proc Soc Exp Biol Med, 28, 268– 272. O’Neill C (1986) PAF in early pregnancy. Conference papers: an International Seminar on the Therapeutic and Commercial Potential of Drugs Affecting Platelet Activating Factor (London). Paccalin J, Dabadie H, Bernard M et al (1984) Interet d’une nouvelle plante oleagineuse: l’onagre (Oenothera biennis ou larmarkiania) apport en acide g-linolenique et troubles de la desaturation en pathologie. Med Nutr, xxi, 132–136. Piorrelli D, Volterra A, Dale N et al (1987) Lipoxygenase metabolites of arachidonic acid as second messengers for presynaptic inhibition of Aplysia sensory cells. Nature, 328, 38–43.
xv
Ratledge C (1987) Oleaginous yeasts and moulds. Conference papers: International Conference on Biotransformations (Cambridge). Serhan CN, Hamberg M and Samuelsson B (1984) Trihydroxytetraenes: novel series of oxygenated derivatives formed from arachidonic acid in human leukocytes. Biochem Biophys Res Commun, 118, 943–949. Siraganian RP and Osler AG (1971) Destruction of rabbit platelets in the allergic response of sensitized leukocytes. 1. Demonstration of a fluid phase intermediate. J Immunol, 106, 1244–1251. Smith JB and Willis AL (1971) Aspirin selectively inhibits prostaglandin production in human platelets. Nature, 231, 235–237. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature, 231, 232–235. Vargaftig BB and Dao NN (1971) Release of vaso-active substances from guinea-pig lungs by ‘‘slow-reacting substance C’’ and arachidonic acid. Its blockade by non-steroidal antiinflammatory agents. Pharmacology, 6, 99–108. Williams TJ and Morley J (1973) Prostaglandins as potentiators of increased vascular permeability in inflammation. Nature, 246, 215–217.
Acknowledgements Editing this substantial volume has required, above all, a great commitment of that most valuable commodity . . . time. Time taken out of a life is not easily replaced, so I thank my wife for her enabling me to devote so much time to this exciting editorial adventure. The stimulus for this book, The Eicosanoids, grew out of suggestions from several colleagues and especially Professor Kim Rainsford (Biomedical Research Centre, Sheffield Hallam University, UK). It was suggested that a new book in this area would be a fitting and appropriate marking of the new millennium and enable recent scientific discoveries and developments to be drawn together into a work of reference, which a number of other colleagues believed would be timely. It was, thus, through these encouragements that the idea finally took on form and became a reality. The Editor wishes to express his appreciation to the publishers, John Wiley & Sons, for their patience and to the indefatigable and complementary Charlotte Brabants and Layla Paggetti from the London office, who have been such absolute stalwarts during these several years and to Monica Twine and her team at the Chichester Office for wisdom and support during the production phase. It is also my great pleasure to acknowledge the consistently reliable support from staff of the Medical Library in Cambridge University, Addenbrooke’s Hospital in Cambridge. I am most grateful to Nobel Laureate Professor Bengt Samuelsson (Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden) for agreeing to contribute a Foreword to this new volume. His pioneering work, together with that of his co-workers, in their pivotal role in elucidating eicosanoid structures, truly laid the foundations for the subsequent development in our understanding of the biochemical pathways involving polyunsaturated fatty acids. A number of the eminent scientists who were involved in a previous multi-author book on prostaglandins (see above) kindly agreed to contribute also to this new volume. To them and to the newer, distinguished contributors (particularly those who so helpfully stepped in at the proverbial ‘‘eleventh hour’’ as replacements to fill a totally unanticipated vacancy) I wish to express my thanks and warm appreciation for the vital part they have played in the production of this exciting book. Also, I should like to make a special point of thanking Professor Elisabeth Granstrom for her help, wisdom and in sharing her extensive knowledge in the eicosanoid field. I am indebted to many colleagues who kindly gave assistance, advice or made recommendations of potential contributors and in this regard wish to thank, particularly: Prof Shozo Yamamoto, Dr Christopher D. Breder, Dr Ignatius Tavares, Prof Alan R. Brash, Dr David Horrobin, Prof Brigitta Peskar, Dr Keith Crowshaw, Dr Bob
Coleman, Dr Howard R. Knapp, Dr Stephen Cunnane, Prof Michael Crawford, Dr A. Kirchenbaum, Dr Andrew J. Dannenberg, Prof Alastair Watson, Dr Joel K. Elmquist, Dr Denis McCarthy, Dr Martin L. Ogletree, Dr Frank Fitzpatrick, Prof Simonetta Nicosia, Prof Colin C. Funk, Dr R. Michael Garavito, Dr Krister Green, Prof Constantino Iadecola, Prof Richard J. Kulmacz, Prof William E. M. Lands, Prof W. L. Smith, Dr John A. McCracken, Prof Stephen Smith, Prof John C. McGiff, Prof Daniele Piomelli, Prof Rodolpho Paoletti, Prof Carlo Patrono, Dr Malcolm Peet, Prof Brendan J. Whittle, Dr Norman L. Poyser, Prof Ingvar Bjarnason, Prof Clifford B. Saper, Prof Flavio Cocceani, Prof K. Schroer, Prof Jorge Capdevila, Prof A. Barry Kay, Prof Tak H. Lee, Dr Jonathan P. Arm, Dr Crispin Bennett, Dr Patrick Loll, Dr Marie Therese Droy-Lefaix, Dr John Wallace and Professor K. C. Nicolaou. Sadly, I have to report that during the gestation period of this book from inception to receipt of the manuscripts several wouldbe contributors, who were research scientists of international repute . . . teachers . . . industrialists . . . colleagues and friends to many who will read this book, have died. Some of them had completed their contribution to this book, but all their names appear below, as a mark of respect for their different, but very significant contributions to the current pool of knowledge in the field of eicosanoids:
IN MEMORIAM Docteur Jacques Maclouf INSERM, Hoˆpital Lariboisie`re, Paris, France Doctor Robin Hoult Division of Pharmacology and Therapeutics, GKT School of Biomedical Sciences, London, UK Professor Emeritus Alan Bennett, DSc Academic Department of Surgery, GKT School of Medicine, London, UK Professor Simonetta Nicosia Department of Pharmacological Sciences, University of Milan, Milan, Italy Professor of Medicine Emeritus James B. Lee, MD University at Buffalo, New York, USA Doctor David Horrobin Chairman, Laxdale Ltd, Kings Park House, Stirling, UK
Foreword Professor Bengt Samuelsson, Nobel Laureate Department of Medical Biochemistry and Biophysics, Karolinska Institute, S-171 77 Stockholm, Sweden
This book edited by Dr Curtis-Prior is entitled The Eicosanoids, which term, introduced by one of the major contributors to the field, Dr E. J. Corey, includes all derivatives of 20-carbon-atom fatty acids. The emphasis in this book is of course on arachidonic acid and products formed by oxidation and further transformation. Structure and function have always played important roles in the eicosanoid field, as well as in many other areas. It was the knowledge of the structures of prostaglandins E1, E2 and E3 (Bergstro¨m et al 1963a; Samuelsson et al 1963a) that paved the way for the discovery of the precursor product relationship between essential fatty acids like arachidonic acid and prostaglandins (Bergstro¨m et al 1964b; van Dorp et al 1964). And this finding made it possible to discover the inhibition of the transformation of arachidonic acid into prostaglandins by aspirin and other NSAIDs in 1971 (Vane 1971). At that time cyclooxygenase had not been discovered. This term was introduced in 1974 (Hamberg et al 1974a) after the endoperoxide structures PGH2 and PGG2 had been discovered. Since cyclooxygenase and COX-1/2 have become household words, I am citing from the 1974 paper (Hamberg et al 1974a): ‘‘. . . It seems likely that PGG2 is the first stable compound formed from arachidonic acid by the ‘prostaglandin synthetase’. By the isolation of PGG2 it was demonstrated for the first time that introduction of the oxygen function at C-15 of the prostaglandin occurs by a dioxygenase reaction. We propose the name fatty acid cyclooxygenase for the enzyme that catalyzes the conversion of arachidonic acid into PGG2 by oxygenation at C-11 and C-15 . . .’’
The availability of the endoperoxides PGG2 and PGH2 was a prerequisite in the discovery of the transformation product thromboxane A2 with pro-aggregatory and vasoconstrictor effects (Hamberg et al 1975b) and subsequently prostacyclin with opposite effects (Moncada et al 1976). These findings extended the cyclooxygenase pathway products from three (PGE2, PGF2a and PGD2) to five. The molecular basis of the multiple biological effects of single prostaglandins has been elucidated by the discovery of subclasses of prostaglandin receptors (Narumiya 2001). The prostaglandins and their analogues are being used as therapeutic agents for induction of labour and termination of pregnancy and also more recently in the treatment of glaucoma. The identification of the receptors will probably stimulate the development of specific prostaglandin agonists and antagonists as therapeutic agents. The isolation and cloning of cyclooxygenase constitutes another important milestone in the eicosanoid field (DeWitt et al 1988; Merlie et al 1988). Although it had been found that the cyclooxygenase can be induced (Masferrer et al 1990), it was through the advent of molecular biology that it was discovered that the induced cycloxygenase activity is due to a different chemical entity, the isozyme COX-2 (Kujubu et al 1991; Xie et al
1991; O’Banion et al 1991). This finding momentarily gave the signal to the pharmaceutical industry to develop new and specific inhibitors of COX-2. Celebra from Searle and Pharmacia and Vioxx from Merck were the first drugs of this category and achieved multibillion dollar sales in their first year. Structure and function also played an important role in the identification of the elusive ‘‘slow-reacting substance of anaphylaxis’’ (SRS-A). In 1976 Borgeat, Hamberg and Samuelsson discovered the 5-lipoxygenase pathway for transformation of arachidonic acid (Borgeat et al 1976). Work on the mechanism of formation of a dihydroxy acid (LTB4) led to the identification of an unstable intermediate (LTA4). This finding played a crucial role in determining the structures of the cysteinyl leukotrienes, LTC4, LTD4 and LTE4 (Murphy et al 1979; Samuelsson et al 1987b). The potent actions of the leukotrienes in inflammation and allergy stimulated the pharmaceutical industry to develop new drugs. Singulair and Accolate from Merck and Zeneca (now Astra-Zeneca), respectively, are now in use in the treatment of bronchial asthma. I have touched on some of the key findings in the eicosanoid field and especially those related to structure and function. The area is still developing rapidly. This development occurs in many disciplines, such as molecular biology, physiology, pathophysiology, pharmacology and clinical applications. New compounds formed by the transformation of arachidonic acid by other lipoxygenases or a combination of lipoxygenases known as the lipoxins (Samuelsson et al 1987b; Serhan et al 1999) are still being discovered. This book edited by Curtis-Prior provides an important update in many different fields of eicosanoid research. It is indeed a very useful summary of the state of the art in a rapidly developing area of research.
REFERENCES Bergstro¨m S, Ryhage R, Samuelsson B and Sjo¨vall J (1963a) The structures of prostaglandin E1, F1a and F1b. J Biol Chem, 238, 3555– 3564. Bergstro¨m S, Danielsson H and Samuelsson B (1964b) The enzymatic formation of prostaglandin E2 from arachidonic acid. Biochim Biophys Acta, 90, 207–210. Borgeat P, Hamberg M and Samuelsson B (1976) Transformation of arachidonic acid and homo-g-linolenic acid by rabbit polymorphonuclear leukocytes. J Biol Chem, 251, 7816–7820. DeWitt DL and Smith WL (1988) Primary structure of prostglandin G/H synthase from sheep vesicular gland determined from the complementary DNA sequence. Proc Natl Acad Sci USA, 85, 1412– 1416. Hamberg M, Svensson J, Wakabayashi T and Samuelsson B (1974a) Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Natl Acad Sci USA 71, 345–349.
xx
FOREWORD
Hamberg M, Svensson J and Samuelsson B (1975b) Thromboxanes. A new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Natl Acad Sci USA, 72, 2994– 2998. Kujubu DA, Fletcher BS, Varnum BC et al (1991) TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J Biol Chem, 266, 12866–12872. Masferrer JL, Zweifel BS, Seibert K and Needleman P (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. Merlie JP, Fagan D, Mudd J and Needleman P (1988) Isolation and characterization of the complementary DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase). J Biol Chem, 263, 3550–3553. Moncada S, Gryglewski R, Bunting S and Vane R (1976) An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature, 263, 663– 665. Murphy RC, Hammarstro¨m S and Samuelsson B (1979) Leukotriene C: a slow reacting substance (SRS) from murine mastocytoma cells. Proc Natl Acad Sci USA, 76, 4275–4279.
Narumiya S (2001) Physiological and pathophysiological roles of prostanoids; lessons from receptor-knockout mice for clinical applications. In Advances in Prostaglandin and Leukotriene Research: Basic Science and New Clinical Applications. Dordrecht: Kluwer Academic, 115–120. O’Banion MK, Sadowski HB, Winn V and Young DA (1991) A serumand glucocorticoid-regulated 4-kilobase mRNA encodes a cyclooxygenase-related protein. J Biol Chem, 266, 23261–23267. Samuelsson B (1963a) Prostaglandins and related factors: the structure of prostaglandin E3. J Am Chem Soc, 85, 1878–1879. Samuelsson B, Dahle´n S-E, Lindgren JA˚ et al (1987b) Leukotrienes and lipoxins: structures, biosynthesis and biological effects. Science, 237, 1171–1176. Serhan CN, Takano T, Gronert K et al (1999) Lipoxin and aspirintriggered 15-epi-lipoxin cellular interactions anti-inflammatory lipid mediators. Clin Chem Lab Med, 37, 299–309. Van Dorp DA, Beerthuis RK, Nugteren DH and Vonkeman H (1964) The biosynthesis of prostaglandins. Biochim Biophys Acta, 90, 204–206. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature, 231, 232–235. Xie W, Chipman JG, Robertson DL et al (1991) Expression of a mitogenresponsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA, 88, 2692–2696.
Section One Biosynthesis and Metabolism
1 Perspectives on the Biosynthesis and Metabolism of Eicosanoids Robert C. Murphy, Rebecca C. Bowers, Jennifer Dickinson and Karin Zemski Berry National Jewish Medical and Research Center, Denver, CO, USA
INTRODUCTION The regulated oxygenation of arachidonic acid leads to a large family of metabolites which have been termed eicosanoids. During investigation of the biochemical events leading to the synthesis of this family of oxygenated lipid substances, a number of quite surprising biochemical features were revealed, some unique to the eicosanoid pathway. Other features illustrated the fundamental biochemical events taking place with polyunsaturated fatty acids within the living cell. The eicosanoids represent a class of molecules that have been termed ‘‘lipid mediators’’ because they are chemical messengers that carry information of cell activation from one cell to another. These cellular messenger molecules have a diverse set of important physiological and pathophysiological roles and coordinate events between cells so that proper tissue function can result. Furthermore, these molecules play a central role in mounting important host defence reactions to protect tissues from adverse conditions, including bacterial infections, that may affect cell viability. While we stand at the threshold of understanding a great deal about these molecules, there is still much that is yet to be learned about the biosynthesis and metabolism of these molecules that further challenges investigators in the field. The first step in the synthesis of eicosanoids, common to all the pathways including prostaglandin and leukotriene biosynthesis, is the release of free arachidonic acid from its storage site in membrane phospholipids. This seemingly simple event is itself a complex biochemical process. For several decades, we have known that arachidonic acid is conserved by the cell in a manner that sequesters this polyunsaturated fatty acid in glycerophospholipids as well as triglycerides (Irvine 1982). Since the cell very tightly regulates the quantity of arachidonic acid through activation of CoA synthase and CoA transferase, as well as phospholipid remodelling (Chilton and Murphy 1986; Walsh et al 1981), there is typically little substrate available for the initial enzyme that oxygenates arachidonic acid. This being the case, it is clear that a great deal of attention has been placed on those enzymes, in particular phospholipase A2, which release arachidonic acid from phospholipid stores (Dennis 2000). Arachidonic acid is highly concentrated at the sn-2 position of glycerophospholipids and the pioneering work of Leslie and co-workers (Alonso et al 1986; Gijon and Leslie 1999) demonstrated that a high molecular weight cytosolic phospholipase A2 (cPLA2) had a unique substrate specificity in releasing arachidonic acid from glycerophosphoethanolamine and glycerophosphocholine lipids. As described by Andreani et al (Chapter 2, this volume), the The Eicosanoids. Edited by Peter Curtis-Prior &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
details of cytosolic phospholipase A2, its phosphorylation, translocation and role in eicosanoid biosynthesis continue to emerge. However, other phospholipases are also present in the cell and, interestingly, they also participate in the regulation of arachidonic acid release, particularly release for eicosanoid biosynthesis (Tougui and Alaoui-El-Axher 2001). The interaction of these phospholipases, including the secreted phospholipases (Type V) and cPLA2 (Type IV), continues to be an area of active investigation and a clear picture of the role of these enzymes and auxiliary enzymes, such as transacylases (Surette and Chilton 1998), is only beginning to emerge. Nonetheless, these enzymes remain interesting therapeutic targets to control eicosanoid biosynthesis within target cells.
PGH SYNTHASE The oxygenation of arachidonic acid in the prostaglandin pathway (Figure 1.1) is mediated by the enzyme PGH synthase for which two genes exist, coding for similar yet significantly different proteins, PGH synthase-1 and PGH synthase-2, commonly referred to as the cyclooxygenases, COX-1 and COX-2 (Smith et al 2000). PGH synthase was the first enzyme studied in the metabolism of arachidonic acid and yet new information continues to emerge from studies of these proteins. Chapter 3 by Deby and Deby-Dupont (this volume) describes many of the mechanistic details of PGH synthase-1, its regulation and mechanism by which the cyclic endoperoxide PGH2 emerges. The regulation and function of PGH synthase-2 is a subject of Chapter 4 (this volume) by H. R. Herschman. Fundamental questions remain concerning these two enzymes from a more chemical point of view. The three-dimensional X-ray crystallographic structures of both PGH synthase-1 and PGH synthase-2 have now been determined and have revealed important insights into the chemical mechanism involved in the cyclooxygenase reaction, including the feature that this enzyme works as a dimer (Kiefer et al 2000; Malkowski et al 2000). These enzymes are membrane-bound and located on the endoplasmic reticulum. Arachidonic acid is drawn from the phospholipid membrane into the active site through a channel in the membrane-associated region of the PGH synthase dimer. The detailed mechanisms by which cyclooxygenase becomes activated through hydroperoxide tone are also emerging. One of the interesting suggestions is that the dimeric enzyme is a critical feature, coupling the peroxidase active site with the endoperoxide active site (Malkowski et al 2000). A detailed understanding of the chemical transformation of
4
BIOSYNTHESIS AND METABOLISM
Figure 1.1 Biochemical pathway involved in the conversion of arachidonic acid into the chemically reactive endoperoxide intermediate prostaglandin H2, mediated by both PGH synthase-1 and PGH synthase-2. The subsequent conversion of PGH2 to the active prostaglandins (PGE2, TXA2 and prostacyclin) is mediated by auxiliary synthases that convert PGH2 into the biologically active prostanoid
arachidonic acid into the cyclic endoperoxide and the critical role played by the tyrosine radical intermediate in the active site have been important advances in this field in understanding the free radical events taking place during the cyclization reaction (Goodwin et al 2000; Shi et al 2000). Within the past several years, it has been discovered that arachidonic acid is not the only preferred substrate for PGH synthase-2. In fact, an ester and amide-linked arachidonic acid can be efficiently converted into prostaglandin-like molecules with the carboxylic acid group linked to ethanolamine (anandamide) or to 2-arachidonoyl glycerol (Yu et al 1997). The structures of the 2-arachidonoyl glycerol products are illustrated in Figure 1.2 and represent an unexpected family of molecules which could themselves play important roles in vivo. In part they have unique physical chemical properties and thus would be distributed quite differently from the normal prostaglandin molecules. For example, these carboxyl-linked prostanoids could be stored within membranes for release of the biochemically active prostaglandin following activation of a lipase or amidase, or they may merely represent an epiphenomenon in that they are accidental substrates for PGH synthase-2. Nevertheless, the ability of PGH synthase-2, rather than PGH synthase-1, to form these molecules is remarkable and further investigation into the role that these carboxylic acid-linked prostaglandins play within cellular biochemistry and pharmacology remains to be elucidated.
Another aspect of the endoperoxide biosynthetic pathway is illustrated by two themes common to eicosanoid biosynthesis. The first theme is that chemically reactive intermediates formed by radical reactions are the primary products during the oxygenation mechanism of arachidonic acid. For PGH synthase, this intermediate is the chemically reactive PGH2. This cyclic endoperoxide has a relatively short half-life of approximately 3 min at physiological pH (Hamberg and Samuelsson 1973; Maclouf et al 1980), yet it is transformed to a host of diverse structures, including prostaglandins D2, E2, F2a, thromboxane A2 and prostacyclin (PGI2). Each of these secondary products of PGH2 are biologically active prostanoids that have substantially less chemical reactivity (except for TXA2 and PGI2) and result from the action of secondary enzymes which process PGH2 into the corresponding prostanoid. This chemical reactivity is a feature that has escaped the attention of many investigators. Prostanoids are highly lipid-soluble and in some cases have very poor water solubility, particularly PGH2. Yet PGH2, which is made within close proximity of the lipophilic membrane of the endoplasmic reticulum, must find auxiliary enzymes that in many cases are in the cytosol. Perhaps most surprising is the fact that there is clear evidence that PGH2 can escape the cell in which it is synthesized, be transferred into another cell, and be metabolized into a unique prostanoid by a secondary cell. This process has been known for approximately two decades and was first recognized when
PERSPECTIVES
5
Figure 1.2 Biochemical conversion of 2-arachidonoyl glycerol by PGH synthase-2 into glyceryl-prostaglandins
mixtures of platelets and endothelial cells were found to synthesize prostacyclin (a unique endothelial cell product), in spite of the fact that the stimulus leading to the release of arachidonic acid and formation of PGH2 took place within the platelet, a cell which only expresses thromboxane synthase and not prostacyclin synthase (Marcus et al 1980). Some of the current active areas of investigation include probing the mechanism by which these chemically reactive intermediates can traverse membranes within and between cells, yet remain intact. Related to this feature is a growing body of evidence that a family of unique proteins that act as transporters are present within cells and move lipophilic prostanoids from the extracellular milieu into a cell where they can either exert unique biological actions or can be the substrate for metabolic processing (Schuster 1998; Pucci et al 1999). We are only at the elementary level in understanding the facile uptake of these lipophilic molecules, their transport through membranes into various compartments within the cell, and transport into peroxisomes and mitochondria where metabolic processing can take place. A second unique theme central to prostaglandin biosynthesis is the suicide inactivation of PGH synthase (Song et al 2001) as well as some of the PGH2-processing enzymes, including prostacyclin synthase (Wade et al 1995) and thromboxane synthase (Jones and Fitzpatrick 1991). Suicide inactivation of PGH synthase was
observed and reported almost 30 years ago (Smith and Lands 1972). While the suicide inactivation of an enzyme by its substrate is certainly known in biochemistry, there are perhaps no other examples of a single biochemical pathway where so many enzymes involved in the production of the biologically active products are susceptible to suicide inactivation. The exact events and detailed chemical mechanisms of suicide inactivation are also not understood. Major questions remain as to whether or not this process is a way by which cells can regulate the full extent of prostanoid production, or whether it is merely a curious artifact of PGH synthase. Yet, it is clear that the suicide inactivation of PGH synthase can limit the extent of prostaglandin biosynthesis within cells. LEUKOTRIENE BIOSYNTHESIS The second major pathway of arachidonate metabolism in terms of production of a family of biologically active products was first described in 1979 as the leukotriene pathway (Figure 1.3). The enzyme 5-lipoxygenase (5-LO) is the first critical protein involved in the oxidation of arachidonic acid and formation of leukotrienes (Samuelsson et al 1979). Following the discovery of the leukotriene pathway and elucidation of the biological activities
6
BIOSYNTHESIS AND METABOLISM
Figure 1.3 Leukotriene biosynthetic pathway involving the conversion of arachidonic acid into 5-hydroperoxyeicosatetraenoic acid (5-HPETE) and LTA4 by 5-lipoxygenase. Subsequent enzymatic processing of the chemically reactive LTA4 into LTB4 and LTC4 occurs through LTA4 hydrolase and LTC4 synthase. The series of cysteinyl leukotrienes are derived from sequential amino acid removal from the tripeptide of LTC4
PERSPECTIVES of leukotriene B4 (Ford-Hutchinson et al 1980) and leukotriene C4 (Barnes et al 1984), detailed studies of the biochemical events taking place during the 5-LO reaction began to emerge. The most significant advances were obtained once 5-LO was sequenced, cloned and expressed so that reasonable quantities were available for detailed biochemical investigation (Rouzer et al 1986). One of the first important discoveries was that 5-LO carried out two different enzymatic steps, both involving the removal of a bisallylic hydrogen atom, initially leading to the formation of 5hydroperoxyeicosatetraenoic acid (5-HPETE) and secondly converting 5-HPETE into leukotriene A4 (LTA4), the conjugated triene epoxide (Shimizu et al 1984). The first step involved removal of a hydrogen atom from carbon atom 7 of arachidonic acid, with concomitant reduction of the non-haeme iron in 5-LO from FeIII to FeII in a radical reaction mechanism. An unexpected observation which resulted from the development of a 5-LO inhibitor, MK886 (Dixon et al 1990), was the discovery of a second membrane-bound protein that was required for biosynthesis of LTA4 in the intact cell, but not necessarily required for biosynthesis in broken cell preparations or with isolated 5-LO (Shimizu et al 1984). This protein, termed five lipoxygenase activating protein (FLAP), was similar in primary sequence to only one other protein, that being LTC4-synthase (Penrose et al 1997). Furthermore, FLAP was found to be localized in or near the nuclear envelope. This observation was consistent with studies that demonstrated translocation of 5-LO to the nucleus during activation, where it likely encountered FLAP (Rouzer et al 1986). In some cells, 5-LO is localized not in the cytosol but in the nucleus (Coffey et al 1992) and activation of such cells resulted in the translocation of the nuclear-localized 5-LO to the perinuclear membrane. An understanding of the importance of the nuclear localization of 5-LO as well as the factors which control or target 5-LO to the nuclear region are still areas of active investigation. Recent evidence has suggested that unique structural features in 5-LO may target 5-LO to this region of the cell (Kulkarni et al 2002; Jones et al 2002). Fundamental questions also remain as to the nature of FLAP and its precise biochemical role. The first stable product of 5-LO is 5-HPETE. Another family of biologically active eicosanoids (Figure 1.4) is directly derived from 5-HPETE and in some cells, e.g. the neutrophil, from the peroxidase reduction product 5-hydroxyeicosatetraenoic acid (5HETE; Skoog et al 1988). Powell and co-workers (1994) described the oxidation of 5-HETE to 5-oxo-eicosatetraenoic acid (5-oxoETE) with a substantial alteration in biological activity, most likely through the engagement of specific receptors that recognize 5-oxo-ETE (Hosoi et al in press). 5-oxo-ETE and some of the related molecules resulting from oxidation at C-15 (Powell et al 1994) or conjugation with glutathione at C-7 (Bowers et al 2000) are highly chemotactic for both eosinophils and neutrophils. Dr William Powell presents an overview of the biosynthesis and pharmacological properties of these oxo-eicosanoids, which represent a more recently discovered family of 5-LO products that have substantial biochemical relevance (Powell, Chapter 6, this volume). LTA4 hydrolase carries out the stereospecific hydrolysis of the conjugated triene epoxide, LTA4, to 5(S),12(R)-dihydroxy-6,14cis-8,10-trans-eicosatetraenoic acid, LTB4. A crystal structure of LTA4 hydrolase has now been obtained (Thunnissen et al 2001) and a more detailed understanding of the mechanism involved in directing water to C-12 of LTA4 is emerging (Rudberg et al 2002). Interestingly, LTA4 hydrolase has a fair degree of homology to zinc-containing metalloproteases and in fact can carry out this enzymatic activity (Medina et al 1991; Muskardin et al 1994). Furthermore, inhibitors of certain metalloproteases, such as bestatin, have been found to inhibit LTA4 hydrolase (Muskardin et al 1994). LTA4 hydrolase is also a suicide inactivated enzyme
7
(Orning et al 1992) and it is clear that LTA4 and LTA3 can become covalently bound to LTA4 hydrolase in the active site (Mueller et al 1995, 1996). In fact, using mass spectrometric techniques it was possible to isolate a peptide derived from intact LTA4 hydrolase to which LTA3 attached to tyrosine-383 in the sequence D371-K385 (Mancini et al 1998). The conjugation of LTA4 with glutathione results in the formation of the myotropic leukotriene, LTC4, which is the initial member of the cysteinyl leukotriene series (Figure 1.3). This enzymatic step is catalysed by a unique glutathione-S-transferase (GST), termed LTC4 synthase. This 18 kDa membrane-bound protein has sequence homology only to FLAP, as mentioned above, rather than to the ubiquitous GSTs found in the cytosol of most cells. Molecular recognition of the cysteinyl leukotrienes is mediated by at least two G-protein-linked membraneassociated receptors (Christie and Henderson 2002), termed cysLT1 and cysLT2, that have the highest affinity for the cysteinyl leukotriene formed by the action of g-glutamyl transpeptidase on LTC4 (Figure 1.3). This product, leukotriene D4, and the subsequent peptide hydrolysis product, leukotriene E4, complete the family of cysteinyl leukotrienes that are potent activators of the cysLT receptors. The transport of these unique peptidolipids has been studied by Keppler et al (1997), who have described involvement of the multi-drug resistance receptor in this important process of export of LTC4 from the synthetic cell. The biosynthesis of leukotrienes within tissues is surprisingly more complex than that revealed by the initial biochemical studies of events taking place within the cell or near the nucleus of the cell. Specifically, approximately 60–70% of the LTA4 made by the neutrophil (Sala et al 1996) is not processed within this cell, but is transferred to auxiliary cells in the process described as transcellular biosynthesis (Murphy et al 1989). Thus, when mixtures of platelets and neutrophils are prepared and neutrophils stimulated, the leukotriene produced in largest amount is LTC4 rather than LTB4 (Maclouf and Murphy 1988). Since the platelet has LTC4 synthase, but no 5-LO or LTA4 hydrolase, this clearly implicates a fundamental role of transcellular biosynthesis during cell–cell interaction. However, the exact mechanisms involved are less than clear. One of the most important steps in transcellular biosynthesis is the transfer of LTA4 from the nuclear site of biosynthesis through the cellular membranes and into a recipient cell, e.g. the platelet. LTA4 must traverse multiple membranes as well as the aqueous cytosol. Since LTA4 has a chemical half-life of less than 3 s at pH 7.4 (Fitzpatrick et al 1982), there is clearly a factor present within cells which protects LTA4 during its movement within the cell and during export from the cell. Even less is known about how LTA4 is taken up by a cell and transported intact to the site where enzymatic processing occurs. Clearly, cells are very efficient in this process, since it has been shown that eosinophils treated with exogenous LTA4 at 08C have intact LTA4 present within the eosinophil for over 1 h, so long as the temperature remains low (Lam et al 1989). Virtually nothing is known about the transport process. It is likely that binding proteins are involved but what role such proteins play in total eicosanoid, and specifically leukotriene, biosynthesis is unclear. Another interesting aspect of LTA4 biochemistry has emerged following the discovery that LTA4 can covalently bind to nucleosides, oligonucleotides (Reiber and Murphy 2000) and, specifically, DNA and RNA (Hankin and Murphy unpublished). Since LTA4 biosynthesis takes place close to the nucleus where high concentrations of DNA and RNA exist, covalent binding of LTA4 to nucleic acids raises important questions as to a potential role that such covalent modification of nucleic acids could play within cells. The concept of covalent binding of LTA4 is beginning to emerge and much remains to be done.
8
BIOSYNTHESIS AND METABOLISM
Figure 1.4 Biosynthesis of 5-oxo-eicosatetraenoic acid (5-oxo-ETE), either directly, as is observed in macrophages, or through the intermediate formation of 5-hydroxyeicosatetraenoic acid (5-HETE) and subsequent oxidation catalysed by the NADP+-dependent dehydrogenase present in neutrophils. The conjugation of 5-oxo-ETE by LTC4 synthase and glutathione leads to the formation of 5-oxo-7-glutathionyl-eicosatrienoic acid (FOG7)
12-/15-LIPOXYGENASE The metabolic oxidation of arachidonic acid by 12-lipoxygenase and 15-lipoxygenase (Figure 1.5) is known to take place within several cell types (Yamamoto et al, Chapter 5, this volume). Within the platelet the major eicosanoid formed by platelet 12lipoxygenase is 12-HPETE following activation of phospholipase A2 (Yamamoto et al 1997). The exact biochemical role of 12lipoxygenase products, however, remains to be established. A similar story exists for 15-lipoxygenase (Kuhn and Borngraber 1999). Important insight into the mechanism of lipoxygenases and a better understanding of the catalytic domain that contains the histidine-coordinated iron molecules was obtained through the crystal structure of 15-lipoxygenase (Gillmor et al 1997). 15Lipoxygenase is the only lipoxygenase that can actually metabolize intact phospholipids to hydroperoxides, and a role has been
suggested for 15-lipoxygenase in cardiovascular disease, particularly atherosclerosis (Funk and Cyrus 2001). CYTOCHROME P-450 AND EPOXYGENASE PATHWAY A third major arm of arachidonate oxygenation involves a series of specific cytochrome P-450 isozymes, which include CYP1A, CYP2B, CYP2C, CYP2D, CYP2G, CYP2J, CYP2N and CYP4A (Zeldin 2001). These enzymes catalyse the formation of epoxyeicosatrienoic acid (EET) regioisomers (Figure 1.6). Other pathways have also been described, including o-oxidation of arachidonic acid, which result in formation of 20-hydroxyand 20-carboxyarachidonic acid, as well as lipoxygenase-like products, with the opposite stereochemistry formed by the specific
PERSPECTIVES
9
Figure 1.5 Metabolism of arachidonic acid catalysed by the 12-lipoxygenase and 15-lipoxygenase into the stereo- and regiospecific monohydroperoxyeicosatetraenoic acids
lipoxygenase (Bylund et al 1998). The pharmacological properties of these cytochrome P-450 metabolites of arachidonic acid have been widely studied, but perhaps the clearest evidence for a role of these compounds in vivo involves the effects of EETs and 20HETE in the kidney (Roman et al 2000). It is likely that a unique role for this pathway of arachidonate metabolism may include modulation of ion channels, mediating sodium potassium ATPase and inhibiting the effects of arginine vasopressin in the kidney (Zeldin 2001). EICOSANOID METABOLISM The role that eicosanoids, specifically prostaglandins and leukotrienes, play within tissues is critically linked to the effective extracellular concentration of these molecules as they present themselves to specific G-protein-coupled receptors. The extracellular concentration is a combination of not only the biosynthesis and biosynthetic mechanisms leading to the eicosanoid product, but also the mechanisms involved in metabolic transformation of these lipid mediators into inactive products. Therefore, an understanding of the catabolic events that occur within specific cells and tissues is an important facet of the entire picture of eicosanoid lipid mediators. Several major pathways exist for processing these endogenous polyunsaturated and
oxygenated fatty acid compounds, with ultimate conversion of the constitutive elements to acetate and water. The metabolism of eicosanoids involves events taking place within the cytosol, the mitochondrion, peroxisome and endoplasmic reticulum (cytochrome P-450). Finally, for most eicosanoids unique pathways of metabolism exist which are specific to each covalent structure. o-OXIDATION A major pathway of prostaglandin metabolism involves o-oxidation, in which a cytochrome P-450 isozyme specifically converts the methyl-terminus of the prostanoid into a primary carbinol (Figure 1.7), a much more polar moiety (Okita and Okita 1996). A specific example is the metabolism of PGE2 in the hepatocyte, where the major metabolites are 20-hydroxy- and 20-carboxy-PGE2. In addition, oxidation of the penultimate carbon atom also occurs, with formation of 19-hydroxy-PGE2, one of the important metabolic products of PGE2 in the isolated hepatocyte (Hankin et al 1997). The metabolism of LTB4 (Figure 1.8), as well as LTC4, is mediated by specific cytochrome P-450s of the CYP4F series. There have now been 10 different enzymes identified in various tissues and animal species that can carry out the o-oxidation of leukotrienes (Kikuta et al 1999). For example, LTB4 is converted by human CYP4F2 (neutrophil) into 20-hydroxy-LTB4 (Kikuta et al 1998;
10
BIOSYNTHESIS AND METABOLISM
Figure 1.6 Summary of the biosynthetic products obtained following the action of various cytochrome P-450 isozymes, using arachidonic acid as substrate with formation of four different regioisomers of epoxyeicosatrienoic acid (EET), the lipoxygenase-like reaction exemplified by formation of 12(R)-HETE and o-oxidation exemplified by 20-HETE
Shak and Goldstein 1985). In the neutrophil itself, it appears that this cytochrome P-450 further oxidizes 20-hydroxy-LTB4 into 20oxo-LTB4 and then an aldehyde dehydrogenase oxidizes this further into 20-carboxy-LTB4 (Sutyak et al 1989). However, in the hepatocyte, 20-hydroxy-LTB4 is a substrate for alcohol dehydrogenase to form 20-oxo-LTB4, which is a substrate for aldehyde dehydrogenase to form 20-carboxy-LTB4 (Baumert et al 1989). b-OXIDATION Recovery of the carbon atoms of endogenous lipids as well as the generation of ATP result during the process of b-oxidation (Diczfalusy 1994). b-Oxidation of lipids takes place not only in mitochondria, but also in the peroxisome. The peroxisomal boxidation pathway is particularly important in eicosanoid metabolism. The process of b-oxidation for a prostaglandin involves initial formation of the CoA ester of the carboxylic acid followed by removal of two and four carbon atoms, leading to the dinor and tetranor prostaglandin metabolites, as shown in Figure 1.7 for PGE2. Typically, chain-shortened metabolites are some of the more abundant metabolites of prostaglandins in urine. However, it should be noted that much of the actual quantity of prostaglandin synthesized within the intact organism is never
excreted as a urinary metabolite such as the tetranor-PGE1 species. Detailed studies of the excretion rate of TXA2 in human subjects have revealed that only 1–2% of the entire biosynthetic pool of TXA2 can be found in any specific excreted product (Patrono et al 1986). Leukotrienes are metabolized by the peroxisomal b-oxidation pathway and LTB4 (Figure 1.8) also by the mitochondrial b-oxidation pathway (Jedlitschky et al 1991). However, the b-oxidation pathway most often occurs after the initial oxidation of the o-terminus into 20-carboxy-LTB4 (Figure 1.8) or 20-carboxy-LTE4. b-Oxidation then proceeds from the o-terminus, leading to the dinor and tetranor metabolites. Under some circumstances, b-oxidation from the carboxyl-terminus has been observed (Jedlitschky et al 1991), but the evidence to date suggests that most b-oxidation is subsequent to o-oxidation of the leukotriene. CONJUGATION Only recently has evidence emerged that conjugation of prostaglandins and leukotrienes by either glucuronidation or sulphation can result in major metabolic products within certain cells. In large part, this was a consequence of the inability to analyse such molecules as intact species; however, the emergence of powerful
PERSPECTIVES
11
Figure 1.7 Metabolism of prostaglandins by o-oxidation and b-oxidation, as exemplified by formation of specific metabolites of PGE2
techniques, specifically electrospray combined liquid chromatography–mass spectrometry (LC–MS and LC–MS–MS) have enabled direct analysis of these molecules. Conjugation is clearly one of the major metabolic pathways which leads to a large number of excreted products. For example, in Figure 1.9, some of the potential metabolites of LTB4 by the glucuronidation pathway are indicated. Thus, it is not surprising that a complex mixture of glucuronides has been observed when techniques that can specifically detect these compounds are employed (Wheelan et al 1999). The extent of conjugation by the sulphate pathway or the glucuronidation pathway remains an important unexplored area of eicosanoid biochemistry. Another important conjugation reaction of some eicosanoids is covalent binding of the tripeptide glutathione (GSH, g-glu–cys– gly) with various electrophilic intermediates formed by oxidation
of arachidonic acid. As discussed above, LTA4 is conjugated with GSH to yield LTC4, a reaction catalysed by LTC4 synthase. This same enzyme has also been found to catalyse formation of another biologically active GSH adduct, FOG7 from 5-oxo-ETE (Hevko and Murphy 2002). The epoxyeicosatrienoic acids are substrates for the cytosolic glutathione-S-transferases and form glutathione adducts as common metabolites (Spearman et al 1985). For all glutathione conjugates, sequential removal of the gglutamic acid and glycine amino acids by g-glutamyl-transpeptidase and dipeptidases is a common metabolic pathway for glutathione adducts to form cysteinyl adducts (Figure 1.3). The cysteinyl adduct is often acetylated to an N-acetyl cysteinyl adduct in what is termed the mercapturic acid pathway and exemplified by formation of N-acetyl-LTE4 (Stene and Murphy 1988; Huber et al 1990).
12
BIOSYNTHESIS AND METABOLISM
Figure 1.8 Metabolism of leukotrienes as exemplified by o-oxidation and subsequent b-oxidation from the o-terminus of LTB4. The conversion of 20hydroxy-LTB4 to 20-carboxy-LTB4 involves a two-step enzymatic oxidation catalysed by alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (AldDH). These metabolites were observed in isolated hepatocytes (Shirley and Murphy 1990. Reproduced with permission of The American Society for Biochemistry & Molecular Biology)
SPECIFIC PATHWAYS Prostaglandins, leukotrienes, and the cytochrome P-450-derived eicosanoids have specific enzymes that can chemically alter the structure of the primary eicosanoid. One of the interesting metabolic enzymes for the prostaglandins is 15-hydroxyprostaglandin dehydrogenase, which is highly expressed in the lung, particularly in the pregnant woman (Bergholte et al 1987). This enzyme converts the 15-hydroxy group into the 15-oxo group, as well as terminating the activity of the prostaglandin, since these oxidized metabolites are typically not recognized by the prostanoid receptor (Figure 1.10). Once the 15-oxo-moiety is present, the a,b-unsaturated keto moiety is recognized by prostaglandin 13,14-reductase, which reduces the 13,14 double bond. This 13,14trans double bond with the allylic hydroxyl substituent is the same structural unit recognized by two enzymes that metabolize leukotrienes (insert, Figure 1.10). The enzyme
12-hydroxyeicosanoid dehydrogenase metabolizes LTB4 into the 12-oxo form of LTB4, which can then be a substrate for the 10,11reductase, leading to a family of 10,11-dihydro-leukotrienes which have substantially diminished biological activity. The 12-hydroxyeicosanoid dehydrogenase has recently been cloned and expressed (Yokomizo et al 1996). The 6-trans-LTB4 non-enzymatic hydrolysis products of LTA4 are also metabolized by reductase in a similar manner, first forming the initial 5-oxo product followed by the dehydrogenase reaction leading to the 6,7-dihydro-5-oxo product (Powell and Gravelle 1988; Wheelan and Murphy 1995). The epoxyeicosatrienoic acids are substrates for epoxide hydrolases. These cytosolic enzymes add water to the epoxide group, leading to vicinal dihydroxyeicosatrienoic acids. Various epoxide hydrolases, both in cytosolic and microsomal locations, can catalyse this hydration reaction, but considerable regioselectivity is observed, e.g. with (14R,15S)-EET (Zeldin et al 1993). Some EETs, specifically 8,9-EET and 5,6-EET, are substrates for
PERSPECTIVES
13
Figure 1.9 Possible glucuronide conjugates of LTB4 at various positions, including the acyl glucuronide which undergoes ester migration to other hydroxyl groups on the glucuronic acid moiety to form several isomers
PGH synthase (Carrol et al 1993) and can be converted to unique prostanoids, including 5-hydroxy-PGI1 and 5,6-epoxy-PGE1 (Oliw 1985). CONCLUSION The biosynthesis and metabolism of eicosanoids is complex, but nonetheless of considerable interest. Numerous challenges remain in developing a full understanding of the role of each of the biosynthetic steps involved in prostaglandin, leukotriene and other eicosanoid production. Much remains to be understood in terms of the different biological roles of PGH synthase-1 and PGH synthase-2. The view that the PGH synthase-1 is the normally expressed ‘‘housekeeping’’ enzyme and that PGH synthase-2 is an enzyme expressed during cellular activation, such as during inflammation, is now known to be overly simplistic and probably does not reveal the true roles for these two enzymes. The mechanism of 5-LO is still only partially understood,
including the events occurring within the active site and the overall regulation of enzymatic activity. The role of FLAP, as well as the nuclear site of LTA4 biosynthesis, remain as enigmas. Finally, the trafficking of these highly lipophilic mediators within a cell or even between cells is only now becoming a serious target of investigation. However, these events are clearly critical to the fundamental process by which eicosanoids act as lipid mediators and a full understanding of the detailed events taking place in the biosynthesis and release of these compounds will very likely reveal interesting new insights into the role of these molecules in health and disease.
ACKNOWLEDGEMENTS This work was supported, in part, by grants from the National Institutes of Health (HL25785 and HL34303).
14
BIOSYNTHESIS AND METABOLISM
Figure 1.10 Metabolism of PGE2 by 15-hydroxyprostaglandin dehydrogenase, leading to formation of 15-oxo-PGE2 and subsequent reduction of the 13,14 double bond. The inset shows the structural unit common to other eicosanoids which undergoes oxidation of an allylic hydroxyl followed by reduction of the adjacent double bond
REFERENCES Alonso F, Henson PM and Leslie CC (1986) A cytosolic phospholipase in human neutrophils that hydrolyzes arachidonoyl-containing phosphatidylcholine. Biochim Biophys Acta, 878, 273–280. Barnes NC, Piper PJ and Costello JF (1984) Comparative effects of inhaled leukotriene C4, leukotriene D4, and histamine in normal human subjects. Thorax, 39, 500–504. Baumert T, Huber M, Mayer D and Keppler D (1989) Ethanol-induced inhibition of leukotriene degradation by o-oxidation. Eur J Biochem, 182, 223–229. Bergholte JM, Soberman RJ, Hayes R et al (1987) Oxidation of 15hydroxyeicosatetraenoic acid and other hydroxy fatty acids by lung prostaglandin dehydrogenase. Arch Biochem Biophys, 257, 444–450. Bowers RC, Hevko J, Henson PM and Murphy RC (2000) A novel glutathione containing eicosanoid (FOG7) chemotactic for human granulocytes. J Biol Chem, 275, 29931–29934. Bylund J, Kunz T, Valmsen K and Oliw EH (1998) Cytochromes P450 with bisallylic hydroxylation activity on arachidonic and linoleic acids studied with human recombinant enzymes and with human and rat liver microsomes. J Pharmacol Exp Ther, 284, 51–60. Carroll MA, Balazy M, Margiotta P et al (1993) Renal vasodilator activity of 5,6-epoxyeicosatrienoic acid depends upon conversion by
cyclooxygenase and release of prostaglandins. J Biol Chem, 268, 12260–12266. Chilton FH and Murphy RC (1986) Remodeling of arachidonatecontaining phosphoglycerides within the human neutrophil. J Biol Chem, 261, 7771–7777. Christie PE and Henderson WR Jr (2002) Lipid inflammatory mediators: leukotrienes, prostaglandins, platelet-activating factor. Clin Allergy Immunol, 16, 233–254. Coffey M, Peters-Golden M, Fantone JC and Sporn PH (1992) Membrane association of active 5-lipoxygenase in resting cells. Evidence for novel regulation of the enzyme in the rat alveolar macrophage. J Biol Chem, 267, 570–576. Dennis EA (2000) Phospholipase A2 in eicosanoid generation. Am J Respir Crit Care Med, 161, S32–S35. Diczfalusy U (1994) b-oxidation of eicosanoids. Prog Lipid Res, 33, 403– 428. Dixon RA, Diehl RE, Opas E et al (1990) Requirement of a 5lipoxygenase-activating protein for leukotriene synthesis. Nature, 343, 282–284. Fitzpatrick FA, Morton DR and Wynalda MA (1982) Albumin stabilizes leukotriene A4. J Biol Chem, 257, 4680–4683. Ford-Hutchinson AW, Bray MA, Doig MV et al (1980) Leukotriene B, a potent chemokinetic and aggregating substance released from polymorphonuclear leukocytes. Nature, 286, 264–265.
PERSPECTIVES Funk CD and Cyrus T (2001) 12/15-lipoxygenase, oxidative modification of LDL and atherogenesis. Trends Cardiovasc Med, 11, 116–124. Gijon MA and Leslie CC (1999) Regulation of arachidonic acid release and cytosolic phospholipase A2 activation. J Leuk Biol, 65, 330–336. Gillmor SA, Villasenor A, Fletterick R et al (1997) The structure of mammalian 15-lipoxygenase reveals similarity to the lipases and the determinants of substrate specificity. Nature Struct Biol, 4, 1003–1009. Goodwin DCF, Rowlinson SWF and Marnett LJ (2000) Substitution of tyrosine for the proximal histidine ligand to the heme of prostaglandin endoperoxide synthase 2: implications for the mechanism of cyclooxygenase activation and catalysis. Biochemistry, 39, 5422–5432. Hamberg M and Samuelsson B (1973) Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Natl Acad Sci USA, 70, 899–903. Hankin JA, Wheelan P and Murphy RC (1997) Identification of novel metabolites of prostaglandin E2 formed by isolated rat hepatocytes. Arch Biochem Biophys, 340, 317–330. Hevko JM and Murphy RC (2002) Formation of murine macrophage derived FOG7 is catalyzed by LTC4 synthase. J Biol Chem, 277, 7037– 7043. Hosoi T, Koguchi Y, Sugikawa E et al (2002) Identification of a novel human eicosanoid receptor coupled to Gi/o. J Biol Chem, 277, 31459– 31465. Huber M, Muller J, Leier I et al (1990) Metabolism of cysteinyl leukotrienes in monkey and man. Eur J Biochem, 194, 309–315. Irvine RF (1982) How is the level of free arachidonic acid controlled in mammalian cells? Biochem J, 204, 3–16. Jedlitschky G, Huber M, Volkl A et al (1991) Peroxisomal degradation of leukotrienes by b-oxidation from the o-end. J Biol Chem, 266, 24763– 24772. Jones DA and Fitzpatrick FA (1991) Thromboxane A2 synthase. Modification during ‘‘suicide’’ inactivation. J Biol Chem, 266, 23510– 23514. Jones SM, Luo M, Healy AM et al (2002) Structural and functional criteria reveal a new nuclear import sequence on the 5-lipoxygenase protein. J Biol Chem, 277, 38550–38556. Keppler D, Leier I and Jedlitschky G (1997) Transport of glutathione conjugates and glucuronides by the multidrug resistance proteins MRP1 and MRP2. Biol Chem, 378, 787–791. Kiefer JR, Pawlitz JL, Moreland KT et al (2000) Structural insights into the stereochemistry of the cyclooxygenase reaction. Nature, 405, 97– 101. Kikuta Y, Kusunose E, Ito M and Kusunose M (1999) Purification and characterization of recombinant rat hepatic CYP4F1. Arch Biochem Biophys, 369, 193–196. Kikuta Y, Kusunose E, Sumimoto H et al (1998) Purification and characterization of recombinant human neutrophil leukotriene B4 ohydroxylase (cytochrome P450 4F3). Arch Biochem Biophys, 355, 201– 205. Kuhn H and Borngraber S (1999) Mammalian 15-lipoxygenases. Enzymatic properties and biological implications. Adv Exp Med Biol, 447, 5–28. Kulkarni S, Das S, Funk CD et al (2002) Molecular basis of the specific subcellular localization of the C2-like domain of 5-lipoxygenase. J Biol Chem, 277, 13167–13174. Lam BK, Owen WFJ, Austen KF and Soberman RJ (1989) The identification of a distinct export step following the biosynthesis of leukotriene C4 by human eosinophils. J Biol Chem 264, 12885–12889. Maclouf J, Kindahl H, Granstrom E and Samuelsson B (1980) Interactions of prostaglandin H2 and thromboxane A2 with human serum albumin. Eur J Biochem, 109, 561–666. Maclouf J and Murphy RC (1988) Transcellular metabolism of neutrophil-derived leukotriene A4 by human platelets. J Biol Chem, 263, 174–181. Malkowski MG, Ginell SL, Smith WL and Garavito RM (2000) The productive conformation of arachidonic acid bound to prostaglandin synthase. Science, 289, 1933–1937. Mancini JA, Waugh RJ, Thompson JA et al (1998) Structural characterization of the covalent attachment of leukotriene A3 to leukotriene A4 hydrolase. Arch Biochem Biophys, 354, 117–124. Marcus AJ, Weksler BB, Jaffe EA and Broekman MJ (1980) Synthesis of prostacyclin from platelet-derived endoperoxides by cultured human endothelial cells. J Clin Invest, 66, 979–986.
15
Medina JF, Wetterholm A, Radmark O et al (1991) Leukotriene A4 hydrolase: determination of the three zinc-binding ligands by sitedirected mutagenesis and zinc analysis. Proc Natl Acad Sci USA, 88, 7620–7624. Mueller MJ, Blomster M, Oppermann UC et al (1996) Leukotriene A4 hydrolase: protection from mechanism-based inactivation by mutation of tyrosine-378. Proc Natl Acad Sci USA, 93, 5931–5935. Mueller MJ, Wetterholm A, Blomster M et al (1995) Leukotriene A4 hydrolase: mapping of a henicosapeptide involved in mechanism-based inactivation. Proc Natl Acad Sci USA, 92, 8383–8387. Murphy RC, Fitzpatrick FA and Maclouf J (1989) Transcellular biosynthesis of eicosanoids and inflammation. In The Handbook of Inflammation: Mediators of the Inflammatory Process, Henson PM, Murphy RC (eds). Elsevier Science, Amsterdam, 147–154. Muskardin DT, Voelkel NF and Fitzpatrick FA (1994) Modulation of pulmonary leukotriene formation and perfusion pressure by bestatin, an inhibitor of leukotriene A4 hydrolase. Biochem Pharmacol, 48, 131– 137. Okita RT and Okita JR (1996) Prostaglandin-metabolizing enzymes during pregnancy: characterization of NAD(+)-dependent prostaglandin dehydrogenase, carbonyl reductase and cytochrome P450-dependent prostaglandin omega-hydroxylase. Crit Rev Biochem Mol Biol, 31, 101–126. Oliw EH (1985) On the metabolism of epoxyeicosatrienoic acids by ram seminal vesicles: isolation of 5(6)epoxy-prostaglandin F1 a. Biochem Biophys Res Commun, 126, 1090–1096. Orning L, Gierse J, Duffin K et al (1992) Mechanism-based inactivation of leukotriene A4 hydrolase/aminopeptidase by leukotriene A4. Mass spectrometric and kinetic characterization. J Biol Chem, 267, 22733– 22739. Patrono C, Ciabattoni G, Pugliese, F et al (1986) Estimated rate of thromboxane secretion into the circulation of normal humans. J Clin Invest, 77, 590–594. Penrose JF, Baldasaro MH, Webster M et al (1997) Molecular cloning of the gene for mouse leukotriene C4 synthase. Eur J Biochem, 248, 807– 813. Powell WS and Gravelle F (1988) Metabolism of 6-trans isomers of leukotriene B4 to dihydro products by human polymorphonuclear leukocytes. J Biol Chem, 263, 2170–2177. Powell WS, Gravelle F and Gravel S (1994) Phorbol myristate acetate stimulates the formation of 5-oxo-6,8,11,14-eicosatetraenoic acid by human neutrophils by activating NADPH oxidase. J Biol Chem, 269, 25373–25380. Pucci ML, Bao Y, Chan B et al (1999) Cloning of mouse prostaglandin transporter PGT cDNA: species-specific substrate affinities. Am J Physiol, 277, R734–R741. Reiber DC and Murphy RC (2000) Covalent binding of LTA4 to nucleosides and nucleotides. Arch Biochem Biophys, 379, 119– 126. Roman RJ, Maier KG, Sun CW et al (2000) Renal and cardiovascular actions of 20-hydroxyeicosatetraenoic acid and epoxyeicosatrienoic acids. Clin Exp Pharmacol Physiol, 27, 855–865. Rouzer CA, Matsumoto T and Samuelsson B (1986) Single protein from human leukocytes possesses 5-lipoxygenase and leukotriene A4 synthase activities. Proc Natl Acad Sci USA, 83, 857–861. Rudberg PC, Tholander F, Thunnissen MM et al (2002) Leukotriene A4 hydrolase: selective abrogation of leukotriene B4 formation by mutation of aspartic acid 375. Proc Natl Acad Sci USA, 99, 4215–4220. Sala A, Bolla M, Zarini S et al (1996) Release of leukotriene A4 versus leukotriene B4 from human polymorphonuclear leukocytes. J Biol Chem, 271, 17944–17948. Samuelsson B, Borgeat P, Hammarstrom S and Murphy RC (1979) Introduction of a nomenclature: leukotrienes. Prostaglandins, 17, 785– 787. Schuster VL (1998) Molecular mechanisms of prostaglandin transport. Annu Rev Physiol, 60, 221–242. Shak S and Goldstein IM (1985) Leukotriene B4 omega-hydroxylase in human polymorphonuclear leukocytes. Partial purification and identification as a cytochrome P-450. J Clin Invest, 76, 1218–1228. Shi WF, Hoganson CWF, Espe MF et al (2000) Electron paramagnetic resonance and electron nuclear double resonance spectroscopic identification and characterization of the tyrosyl radicals in prostaglandin H synthase 1. Biochemistry, 39, 4112–4121.
16
BIOSYNTHESIS AND METABOLISM
Shimizu T, Radmark O and Samuelsson B (1984) Enzyme with dual lipoxygenase activities catalyzes leukotriene A4 synthesis from arachidonic acid. Proc Natl Acad Sci USA, 81, 689–693. Shirley MA and Murphy RC (1990) Metabolism of leukotriene B4 in isolated rat hepatocytes: Involvement of 2,4-dienoyl CoA reductase in leukotriene B4 metabolism. J Biol Chem, 265, 16288–16295. Skoog MT, Nichols JS, Harrison BL and Wiseman JS (1988) Specificity of an HPETE peroxidase from rat PMN. Prostaglandins, 36, 373–384. Smith WL, DeWitt DL and Garavito RM (2000) Cyclooxygenases: structural, cellular and molecular biology. Annu Rev Biochem, 69, 145– 182. Smith WL and Lands WE (1972) Oxygenation of polyunsaturated fatty acids during prostaglandin biosynthesis by sheep vesicular gland. Biochemistry, 11, 3276–3285. Spearman ME, Prough RA, Estabrook RW et al (1985) Novel glutathione conjugates formed from epoxyeicosatrienoic acids (EETs). Arch Biochem Biophys, 242, 225–230. Song I, Ball TM and Smith WL (2001) Different suicide inactivation processes for the peroxidase and cyclooxygenase activities of prostaglandin endoperoxide H synthase-1. Biochem Biophys Res Commun, 289, 869–875. Stene DO and Murphy RC (1988) Metabolism of leukotriene E4 in isolated rat hepatocytes. Identification of b-oxidation products of sulfidopeptide leukotrienes. J Biol Chem, 263, 2773–2778. Surette ME and Chilton FH (1998) The distribution and metabolism of arachidonate-containing phospholipids in cellular nuclei. Biochem J, 330, 915–921. Sutyak J, Austen KF and Soberman RJ (1989) Identification of an aldehyde dehydrogenase in the microsomes of human polymorphonuclear leukocytes that metabolizes 20-aldehyde leukotriene B4. J Biol Chem, 264, 14818–14823. Thunnissen MM, Nordlund P and Haeggstrom JZ (2001) Crystal structure of human leukotriene A4 hydrolase, a bifunctional enzyme in inflammation. Nature Struct Biol, 8, 131–135.
Tougui L and Alaoui-El-Axher M (2001) Mammalian secreted phospholipases A2 and their pathophysiological significance in inflammatory diseases. Curr Mol Med, 1, 739–754. Wade ML, Voelkel NF and Fitzpatrick FA (1995) ‘‘Suicide’’ inactivation of prostaglandin I2 synthase: characterization of mechanism-based inactivation with isolated enzyme and endothelial cells. Arch Biochem Biophys, 321, 453–458. Walsh CE, Waite BM, Thomas MJ and DeChatelet LR (1981) Release and metabolism of arachidonic acid in human neutrophils. J Biol Chem, 256, 7228–7234. Wheelan P, Hankin JA, Bilir B, Guenette D and Murphy RC (1999) Metabolic transformations of leukotriene B4 in primary cultures of human hepatocytes. J Pharmacol Exp Ther, 288, 326–334. Wheelan P and Murphy RC (1995) Metabolism of 6-trans-isomers of leukotriene B4 in cultured hepatoma cells and in human polymorphonuclear leukocytes. Identification of a D6-reductase metabolic pathway. J Biol Chem, 270, 19845–19852. Yamamoto S, Kishimoto K, Arakawa T et al (1997) Arachidonate 12lipoxygenases. Catalytic properties and regulation of the enzyme gene. Adv Exp Med Biol, 407, 191–196. Yokomizo T, Ogawa Y, Uozumi N et al (1996) cDNA cloning, expression and mutagenesis study of leukotriene B4 12-hydroxydehydrogenase. J Biol Chem, 271, 2844–2850. Yu M, Ives D and Ramesha CS (1997) Synthesis of prostaglandin E2 ethanolamide from anandamide by cyclooxygenase-2. J Biol Chem, 272, 21181–21186. Zeldin DC (2001) Epoxygenase pathways of arachidonic acid metabolism. J Biol Chem, 276, 36059–36062. Zeldin DC, Kobayashi J, Falck JR et al (1993) Regio- and enantiofacial selectivity of epoxyeicosatrienoic acid hydration by cytosolic epoxide hydrolase. J Biol Chem, 268, 6402–6407.
2 Control of Eicosanoid Production by Cellular and Secreted Phospholipase A2 Marise Andreani1, Jean-Luc Olivier2 and Gilbert Be´re´ziat1 1University
Pierre and Marie Curie, Paris; and 2Faculty of Medicine, Nancy, France
Phospholipases A2 (PLA2) are a structurally diverse group of enzymes that hydrolyse the fatty acid which esterifies the sn-2 position of glycerophospholipids (Figure 2.1), leading to the formation of a free fatty acid and a lysoderivative. The presence of various PLA2 enzymes in mammalian cells provides multiple differentially regulated pathways for the important process of fatty acid turnover and lipid mediator synthesis. VARIOUS CLASSES OF PHOSPHOLIPASES A2 The mammalian PLA2 enzymes consist of three unrelated families of enzymes: secretory PLA2s (sPLA2s), cellular PLA2s and PLA2s which hydrolyse the platelet-activating factor (PAF) or oxidized phospholipids. Cellular Phospholipases A2 Cellular PLA2s comprise at least five categories of enzymes which differ considerably from each other. The best characterized are cytosolic PLA2s (cPLA2s) (Figure 2.2). Three different enzymes have been cloned to date: cPLA2a cPLA2b and cPLA2g. cPLA2a has a molecular weight of 85 kDa (749 amino acids), is widely distributed and its membrane translocation is achieved by a calcium-binding domain (CalB) (Leslie 1997). This enzyme possesses multiple enzymatic activities exhibiting, in addition to phospholipase A2 activity, intense lysophospholipase activity and weak transacylase activity. It can also hydrolyse fatty esters of 7hydroxycoumarin. The lysophospholipase activity provides an efficient mechanism to prevent an untimely increase of deleterious cytotoxic lysophospholipids. Mutagenesis studies have demonstrated two important amino acid residues for the catalytic process. Serine 228, which is located in a region of sequence homology with the phospholipase B from Penicillum notatum that aligns the lipase consensus sequence (Gly–Leu–Ser–Gly–Gly), is involved in the esterase process in the same way as the serine of the catalytic site of trypsin. This suggests that serine 228 [S228] covalently reacts with the substrate to form an acyl-enzyme intermediate. In addition, cysteine 331 [C331] is the residue that is targeted by sulphydryl-modifying reagents leading to a loss of PLA2 activity, which suggests that it is located near the active site where its integrity is required for correct hydrolysis. However, when [S228] is replaced by an alanine [S228A], low phospholipase activity persists, suggesting that [C331] can replace serine as the active nucleophil The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
site, albeit with reduced catalytic activity. Arginine 200 [R200] and aspartic acid 549 [D549] are also essential for cPLA2 activity, but the network of electron/proton charges which allows [S228] and/or [C331] activation is not understood. Several serine residues are phosphorylated in cPLA2: 437, 454, 505 and 727. While most studies have emphasized the major role of serine 505 [S505] phosphorylation for enzyme activation, other studies also suggest the participation of serine 727 [S727]. At the present time, cPLA2a is the only characterized cellular phospholipase clearly involved in agonist-induced release of lipid mediators. cPLA2b and cPLA2g have recently been cloned (Underwood et al 1998) but only cPLA2g has been characterized. This enzyme has a molecular weight of 60 kDa (541 amino acids) and is calciumindependent, since it lacks the N-terminal CalB domain. cPLA2g is closely related to cPLA2a and contains the potential critical amino acids for the active site with the exception of cysteine [C331], but lacks the key elements that regulate cPLA2a activity (serine phosphorylation sites and calcium ion-induced membrane translocation). Its membrane anchoring is obtained by a prenylation motif at the C-terminus and a myristoylation site at the Nterminus (Figure 2.2). This expression is restricted to heart and skeletal muscles and it is suggested that this enzyme participates in ischaemia-induced injury of heart muscle. Another calcium-independent PLA2 (iPLA2) has been cloned in CHO cells and macrophages (Balboa et al 1997; Tang et al 1997) (Figure 2.3). The CHO enzyme is an 85 kDa protein present in the form of a multimeric complex of 270–350 kDa. The full-length cDNA encodes a 752 amino acid cytoplasmic protein with one lipase motif (GXS465XG) and eight ankyrin repeats, which are required for enzymatic activity. It lacks extensive homologies with other phospholipases and was named iPLA2. The macrophage enzyme has marked homology with CHO iPLA2. These enzymes hydrolyse choline phospholipids at the same rate as ethanolamine phospholipids, but at a higher rate than phospholipids with negatively charged head groups (PI and PA). They exhibit a rather broad specificity towards phospholipid substrates and an additional intrinsic lysophospholipase activity. No known consensus sequences for post-translational modification, such as phosphorylation sites, are present. This indicates that this enzyme is most probably involved in the housekeeping action of remodelling phospholipids. Phospholipases A2 specific for plasmalogens have been described. One has been isolated from bovine brain. It is localized in cytosol and does not require Ca2+ for its activity. It has a molecular weight of 39 kDa and is strongly inhibited by
18
BIOSYNTHESIS AND METABOLISM
Figure 2.1 Various sites of phospholipid cleavage by phospholipases
Figure 2.2 Protein domains of the two cytosolic phospholipases A2 indicating the important amino acid residues involved in the catalytic process (R, arginine; D, aspartic acid; C, cysteine; S, serine), the calcium binding domain (CalB), which is involved in calcium-mediated membrane translocation of the enzyme, and the sites for prenylation and myristoylation that might be involved in enzyme activation. Serine505 is the main residue in which phosphorylation by Map-kinase induces enzyme activation
Figure 2.3 Calcium-insensitive cellular phospholipases A2 showing the serine of the active site and the ankyrin domain
PLA2 CONTROL OF EICOSANOID PRODUCTION glycosaminoglycans, gangliosides and sialoglycoproteins. The interactions between plasmalogen-selective PLA2 and glycoconjugates may be involved in the regulation of enzymatic activity. Plasmalogen rather than monocyl phospholipids is also the preferred substrate for the cardiac PLA2 isoform activated during ischaemia. The diacyl metabolite, lysophosphatidylcholine, is arrhythmogenic, but the effects of the plasmalogen metabolite, lysoplasmenylcholine (LPLC), are essentially unknown (Caldwell and Baumgarten 1998).
Secreted Phospholipases A2 The mammalian sPLA2s constitute a homogeneous family of enzymes closely related to snake venom sPLA2s (Figure 2.4). They share an ancestral precursor and have retained several very stable characteristics: a low molecular weight (13–16 kDa); several disulphide bridges (4–8); the need for millimolar concentrations of calcium ions and a strongly conserved active site. This site comprises the presence of a nearly invariable calcium binding loop, whose role is to orientate the ester bond of the phospholipid substrate at the active site, and a network of electronic/protonic charges (Figure 2.5) that activates the histidine residue, which is always present at the active centre. This network involves one aspartic acid and two tyrosine residues which have been highly conserved through evolution.
Figure 2.4 Phylogeny of the secreted phospholipases A2
19
The pancreatic enzymes belong to type IB sPLA2 and contain seven disulphide bridges and the characteristic ‘‘elapidae loop’’. Enzymes secreted during the inflammatory processes belong mostly to type IIA and type V sPLA2. Type IIA sPLA2 enzymes also possess seven disulphide bridges, but one is in a different position than in type I sPLA2. It lacks the ‘‘elapidae loop’’ and presents a characteristic extended C-terminus. It is the most important of the inflammatory enzymes (Kramer et al 1989; Seilhamer et al 1989). Type V sPLA2 possesses the six disulphide bridges common to type IB sPLA2 and type IIA sPLA2 but it lacks the ‘‘elapidae loop’’ and the C-terminal extension. This enzyme is mainly expressed and secreted in the heart and in macrophages in response to inflammatory cytokines (Tischfield 1997). Type X sPLA2 possesses eight disulphide bridges and differs from other sPLA2s by the presence of a propeptide and more acidic properties (Cupillard et al 1997).
Platelet-activating Factor Acetylhydrolases Platelet-activating factor acetylhydrolases are structurally diverse isoenzymes that catalyse the hydrolysis of the acyl group at the second position of glycerol in unusual phospholipids (Stafforini et al 1997). They have a marked specificity for phospholipids with a short acyl chain at the sn-2 position. Their main substrate is platelet-activating factor (PAF) and they are consequently
20
BIOSYNTHESIS AND METABOLISM
Figure 2.5 Catalytic site of secreted phospholipase A2 showing the calcium ion which coordinates the carbonyl of the fatty acid esterifying the sn-1 position of the substrate, the phosphate group of the substrate, and two amino acid residues of the enzyme. Therefore, the calcium ion orientates the substrate in order to present the sn-2 ester bond to the reactive histidine residue, which is stimulated by an ionic network involving two tyrosines and one aspartic acid
considered to be signal terminators, but they are also able to hydrolyse phospholipids with sn-2 acyl chains up to nine methylene groups, especially when they carry a carbonyl group at the o-end of the acyl chain. These enzymes were first studied as secreted proteins, found in mammalian plasma. Further activities were identified in the cytosolic fraction of various mammalian tissues and in human blood cells. The plasma (pPAH) and intracellular (iPAH) forms of PAF acetylhydrolases are encoded by individual genes. The first 17 codons of the cDNA of the secreted isoform pPAH encode for a hydrophobic peptide that targets the protein to the secretory pathway; this sequence is followed by a propeptide sequence of 23 residues which is not present in the enzyme purified from plasma with a molecular weight of 44 kDa. The active site contains a catalytic triad composed of serine, histine and aspartic acid; the serine residue is a component of a small region of homology domain with other lipases: glycine–X–serine273–X– glycine (GXSXG). Two intracellular isoforms have been characterized. They possess the minimal lipase motif GXSXG (or valine). The first isoform, iPAH-I, shares a high degree of identity of the amino acid sequence (41%) with the plasma form. It is present mainly in liver and kidneys. The second intracellular isoform, iPAH-II, is composed of three subunits with molecular weights of 45, 30 and 29 kDa (a, b and g). Only the a and b subunits possess lipolytic activity; the g subunit is considered to be a regulatory component. This isoform is totally specific for PAF hydrolysis and its function is to precisely modulate cellular PAF levels.
PHOSPHOLIPASES A2 AND CELL PHYSIOLOGY Phospholipases A2 play a role in a very wide range of physiological functions. PLA2 enzymes participate in the digestion of dietary lipids and in phospholipid acyl turnover. They are involved in several pathophysiological situations, such as bacterial infection, sepsis and other inflammatory disorders. They are also
involved in major cellular functions, such as cell signalling, cell growth, apoptosis, endocytosis/exocytosis, cell migration and adhesion and the regulation of gene expression. These enzymes are involved in three major types of functions at the molecular level: (a) they participate in the metabolic fate of phospholipids; (b) they have a housekeeping role for lipoprotein and membrane repair; and (c) they are key enzymes for the control of bioactive mediator synthesis.
Phospholipases A2 and the Metabolic Fate of Phospholipids (Figure 2.6) The neosynthesis of all phospholipids shares a common pathway from glycerol-3-phosphate and acyl-CoA to phosphatidic acid (PA). Thereafter, two different pathways lead to aminophospholipids [phosphatidylcholine (PC), phosphatidylethanolamine (PE) and phosphatidylserine (PS)] and other phopholipids [phosphatidylinositol (PI), phosphatidylinositoldiphosphate (PIP2), phosphatidylglycerol (PG) and diphosphatidylglycerol (DPG or cardiolipid)] (Figure 2.6). This explains why the fatty acid profile of neosynthesized phospholipids is broadly the same and reflects the repertoire of cellular acyl-CoAs, which is itself greatly influenced by dietary conditions. The only specificity achieved by de novo synthesis is the presence of a saturated or monounsaturated chain in the sn-1 position of the glycerol backbone and the presence of a mono-unsaturated and poly-unsaturated chain in the sn-2 position. When considering the actual composition of cellular phospholipids, a vast number of studies have emphasized the specific fatty acid profile of each class of phospholipids. Furthermore, most mammals exhibit subcellular pools of phospholipids which remained rich in poly-unsaturated fatty acids when animals were submitted to drastic nutritional conditions of essential poly-unsaturated fatty acid deficiency. This, clearly, cannot be achieved by de novo synthesis, but requires several remodelling pathways, including either the polar head group or the fatty acids. Polar head groups are interchanged between phospholipids by several enzymes: base-exchange
PLA2 CONTROL OF EICOSANOID PRODUCTION
21
Figure 2.6 De novo pathway of phospholipid synthesis (top) and fatty acid remodelling (bottom) by deacylation/reacylation using various transacylating enzymes
enzymes, PS decarboxylase, PE transmethylases, etc. (Figure 2.6). Fatty acid composition is modified by several pathways. A deacylation step catalysed by either PLA1 or PLA2 produced lysoderivatives. These lysoderivatives are then reacylated by (a) acyl-CoA:lysophospholipidacyl-transferases which exhibit lysophospholipid and acyl-CoA specificity and differential expression in various cell types, and by (b) phospholipid:lysophospholipid acyl-transferases, enzymes very specific for poly-unsaturated fatty acid transfer, which are mainly found in platelets, muscle cells and cells involved in the immuno-inflammatory process (Yamashita et
al 1997). Acyl-CoAs are either produced by acyl-CoA synthases, which are specific for a set of fatty acids (saturated, mono- or diunsaturated fatty acids and poly-unsaturated fatty acids) or by acyl-CoA:lysophospholipid acyl-transferase acting in reverse. In the lung, a particulate acyl-transferase synthesizes the characteristic dipalmitoyl-lecithin of surfactant by a CoA-independent transfer of palmitic acid from one sn-1-palmitoyl-lysolecithin to another. It has been reported that iPLA2 contributes to the remodelling of membrane phospholipids. In support of this proposal, iPLA2 overexpression leads to an increase in basal fatty
22
BIOSYNTHESIS AND METABOLISM
acid release. Of particular interest was the finding that arachidonic acid released by iPLA2 was not metabolized to prostaglandin E2 (PGE2) (Balsinde et al 1998). Phospholipases A2 Housekeeping Enzymes for Lipoprotein and Membrane Repair Oxidative stress appears to be a major pathophysiological disturbance at work during the early atherosclerotic process (Livrea et al 1998). The phospholipids in LDL particles are known to undergo oxidation to produce oxidized phospholipids, which become substrates for PAF acetylhydrolases. These oxidized phospholipids could be detrimental in two ways; they mimic PAF to support sustained inflammation but could also covalently modify the apolipoprotein, resulting in unregulated cholesterolester accumulation by macrophages. In contrast, the products of oxidized phospholipid hydrolysis, lysoPC and fragments of fatty acids, are water-soluble and can be further metabolized into nontoxic products. Consequently, LDL-associated PAF acetylhydrolase can prevent modification of LDL into atherogenic particles. This protective effect might also be at work intracellularly, e.g. in erythrocytes and other cells subject to oxidative stress to ensure membrane phospholipid integrity, which is highly disturbed by the presence of oxidized phospholipids. Phospholipases A2 and Control of Lipid Mediator Production When activated by extracellular stimuli, cPLA2 triggers rapid hydrolysis of membrane phospholipids, leading to the formation of free fatty acids, mainly arachidonic acid, linoleic acid or oleic acid. This is the rate-limiting step for eicosanoid production (Figure 2.7). Arachidonic acid is the precursor of the vast family of eicosanoids which comprises two subfamilies, the cyclooxygenase family (prostaglandins and thromboxanes) and the lipoxygenase family [leukotrienes and hydroxyeicosatetranoic acids (HETEs)]. Linoleic acid gives rise to hydroxyoctadecanoic acids (HODEs) via the lipoxygenase pathway. The other cPLA2 product, lysophospholipid, might be a lipid mediator (lysophosphatidic acid or lysophosphatidylcholine) or the precusor of PAF. The production of these various mediators is tissue- and stimulation-dependent and this is achieved through the differential expression of the various enzymes involved in lipid mediator synthesis, and through the preferential coupling between cPLA2 and lipoxygenase or cyclo-oxygenase and the subcellular localization of these enzymes. Role of cPLA2 in Eicosanoid Synthesis Activation of cPLA2 is dependent on at least two mechanisms: phosphorylation of cPLA2 by various kinases and its translocation to membrane phospholipids by the calcium-binding domain (CalB) following an increase of intracellular Ca2+, which allows access of this cytosolic enzyme to its membrane substrate (Leslie 1997). The pathway by which cPLA2 is activated by extracellular stimuli, or whether this activation occurs at specific cellular sites, has not been elucidated. To date, various intracellular sites of translocation have been described for cPLA2 according to the cell type and stimuli used. They include nuclear and endoplasmic reticulum membranes, where they would co-localize with cyclooxygenases or lipoxygenases. Plasma membrane has also been proposed as a possible site for translocation of cPLA2, but early studies proposing a direct interaction of cPLA2 with heterotrimeric G proteins driven by serpentine membrane receptors have not been confirmed. Recent data reporting an interaction of
cPLA2 with receptor-activated Jak-kinases, or its activation by phosphatidyl inositol 4,5-biphosphate, have confirmed the possibility that translocation to the plasma membrane might indeed constitute the actual activation process. Using chimeric cPLA2 bearing the plasma membrane targeting signal of the Lck protein kinase, we have shown that increased expression of this chimeric cPLA2 in CHO cells gave a stronger response to epinephrine, Ca2+ ionophore or both than increased expression of the wildtype enzyme (Klapisz et al 1999). We also recently demonstrated that, in epithelial cells bearing the DF508 cystic fibrosis mutation, alteration of the cystic fibrosis transmembrane regulator (CFTR) molecule alters the functional coupling between certain receptors and cPLA2, leading to increased eicosanoid production. This might increase the inflammatory state at least in some circumstances (Berguerand et al 1997). cPLA2 has one consensus site for Map-kinase and phosphorylation of this serine (S505) by Map-kinases occurs in vivo and in vitro (Figure 2.2). This phosphorylation, which induces a characteristic electrophoretic mobility shift, is essential for receptor-mediated cPLA2 activation, since the S505A mutant can no longer be activated. Other kinases, such as protein kinases C (PKC) and A, can phosphorylate cPLA2 in vitro, but this does not result in a significant increase in enzyme activity. However, PKC might indirectly activate cPLA2 in vivo by activating other kinases. In some cell systems (neutrophils, platelets), TNFa or thrombin activate cPLA2 independently of Map-kinase activation, but via a proline-directed kinase which also phosphorylates S505. Although phosphorylation is important for cPLA2 activation in certain cells, it is not sufficient for complete activation. We recently demonstrated, in CHO cells overexpressing adrenergic receptors, that Ca2+ also participates in the activating process itself. This dual activation of cPLA2 by adrenergic receptors was also evidenced in rat cardiomyocytes and has been implicated in the contractile effect of b2-adrenergic receptors (Pavoine et al 1999). However, there is evidence to suggest alternative pathways without increasing Ca2+, e.g. in macrophages and neutrophils stimulated by the phorbol ester PMA or the phosphatase inhibitor okadaic acid. In this case, phosphorylation of S727 in addition to S505 phosphorylation might dispense with the Ca2+ increase. The actual protein kinase involved is not known but might be a basotrophic kinase. It is now clear that cPLA2 is a major element of the signalling cascade. However, this enzyme might not be as ubiquitous as may appear at first glance, e.g. its expression is confined in murine brain grey matter to astrocytes at the pial surface of the brain, and was found in white matter in addition to fibrous astrocytes, in interfascicular oligodendrocytes (Lautens et al 1998). In contrast, mRNA cPLA2-immunoreactivity was undetectable in neurons of murine brain. A direct role of cPLA2 in muscarinic signalling must be ruled out, but an indirect regulation of neurotransmission may be provided via the activity of cPLA2 in glial cells. Role of sPLA2 in Eicosanoid Synthesis Type II sPLA2 has been previously shown to hydrolyse phospholipids in the membrane of Escherichia coli, after the action of bacterial permeability-increasing protein (BPI), suggesting that it might participate in antimicrobial defence. Only indirect evidence has been published to indicate that type II sPLA2 can hydrolyse mammalian cell phospholipids in vivo to generate lipid mediators. In several cell systems, treatment by inflammatory cytokines strongly increases prostaglandin and type II sPLA2 synthesis (vascular smooth muscle cells, articular chondrocytes, mesangial cells, macrophages). This enzyme accumulates in the Golgi apparatus associated with proteoglycans. Part of the
Figure 2.7 Biosynthetic pathways of oxygenated fatty acids. The cyclooxygenase pathways (right) and the various lipoxygenase pathways (left)
PLA2 CONTROL OF EICOSANOID PRODUCTION 23
24
BIOSYNTHESIS AND METABOLISM
enzyme may remain intracellular, but the majority is secreted and found associated with proteoglycans on the external cell surface, where phospholipid hydrolysis can occur. Some studies have shown that treatment of cell cultures with neutralizing anti-type II sPLA2 or anti-type V sPLA2 antibodies or antisense cDNA corresponding to the mRNA of these enzymes decreased prostaglandin synthesis. As type II sPLA2 and type V sPLA2 are not synthesized as inactive proenzymes, mechanisms must exist in vivo to protect membranes of the secretory system from untimely hydrolysis. Furthermore, numerous studies have shown that secreted enzymes cannot hydrolyse phospholipids of the plasma membrane of intact cells. It is thought that activation of the enzyme might be the result of membrane ‘‘scrambling’’ or synergistic activation of sphingomyelinase. We have also demonstrated that sphingomyelin, which is located in the external leaflet of plasma membranes, inhibits phospholipid hydrolysis by type II sPLA2 and that cholesterol, which strongly complexes with sphingomyelin, relieves this inhibitory effect (Koumanov et al 1997). Furthermore, when cells are activated by inflammatory cytokines, membrane sphingomyelin is degraded, and phospholipid asymmetry and lipid packing are known to be widely modified under these circumstances. This may have a critical role in modulating the activity of type II sPLA2 acting on biological membranes. Type II sPLA2 activity is also regulated by polypeptide factors such as phospholipase activating peptides (PLAPs) or annexins. Annexins are known to inhibit type II sPLA2 activity via so-called substrate depletion inhibition (Comera et al 1990). Annexin is strongly bound to acidic phospholipids; large amounts of annexin are supposed to mask phospholipid substrates from the action of type II sPLA2. However, we demonstrated that its action is more complex, since at low concentrations, which more probably mimic the intracellular situation, annexins increased enzyme activity. It is known that annexin binding to the cell surface induces disorders on both sides of the membrane. In addition, annexin VI participates in severing of the stalk during the budding process of endocytosis. It is therefore tempting to implicate type II sPLA2, which is present in secretory vesicles or at the surface of the plasma membrane, in events leading to exocytosis, endocytosis and to the intracellular production of lipid mediators, which often occurs during these processes. In the kidney, uteroglobin was shown to inhibit the type II sPLA2 enzyme secreted by mesangial cells and consequently decreases the level of lysophosphatidic acid (Mukherjee et al 1998). Specific receptors for secreted type I and type II enzymes have been characterized and cloned (Lambeau et al 1995). These receptors belong to the family of mannose receptors which are involved in the clearance of plasma glycoproteins. It has been suggested that its function is to eliminate the overproduction of secreted PLA2. However, some data suggest that the binding of type I sPLA2 might induce intracellular signals (Figure 2.8). For example, in a human astrocytoma cell line, type II PLA2 was found able to activate the cascade of Map-kinases and increase cPLA2 activity (Hernandez et al 1998). But other data have challenged this possibility and conflicting results have been reported concerning the actual involvement of PLA2-receptors in this process. REGULATION OF PHOSPHOLIPASE A2 SYNTHESIS The mechanisms controlling cPLA2 activity are mainly posttranslational. This occurs by phosphorylating events and Ca2+dependent translocation and allows rapid activation of cPLA2 activity in response to a variety of physiological stimuli. The mechanisms that control cPLA2 synthesis are less clearly understood. Although there is evidence that cPLA2 mRNA levels may
be subject to modest stimulation in some cells by mediators such as TNFa, glucocorticoids, macrophage colony stimulatory factor, epidermal growth factors, interferon-g, leukaemia inhibitory factor or interleukin-13, cPLA2 function is so important for a variety of cellular processes that it is likely that transcription of this gene is closely controlled. The TNFa effect was demonstrated to be mediated by the sphingomyelinase–ceramide pathway (Hayakawa et al 1996). Conversely, IL-4 inhibits cPLA2 gene expression in LPS-stimulated macrophages and rheumatoid synovial cells (Kuroda et al 1997). It has been demonstrated that a specific element (717,78) of the cPLA2 promoter is responsible for the low-level constitutive expression of the gene (Miyashita et al 1995). Furthermore, the presence of three ATTTA motifs in the 3’ untranslated region (3’UTR) suggests that the gene might be regulated at a post-transcriptional level by modulation of mRNA stability (Tay et al 1994). Inflammatory cytokines induce the synthesis and secretion of type II sPLA2 in various cell types. In several cell systems, particularly in rat mesangial cells, the cytokine effect is reinforced by the stimulation of receptors or drugs that increase cytoplasmic cAMP (Walker et al 1995). In contrast, peptide growth factors (TGFb, PDGF, EGF and bFGF) markedly inhibited the extracellular release of type II sPLA2 (Pruzanski et al 1998). The precise signalling events leading to type II sPLA2 gene regulation are not fully understood. In most cells, interleukin-1b and TNFa trigger their action by at least two main pathways involving specific receptors. They activate acid sphingomyelinase to initiate the ceramide cascade (Mathias et al 1993). IL-1b and TNFa induce signalling cascades which activate members of the NFkB family. In most cells, NFkB is retained in the cytoplasm by inhibitory proteins called IkBs (Beg and Baldwin 1993). IL-1b and TNFa induce IkB phosphorylation and promote its degradation by the proteosome pathway and the subsequent nuclear translocation and activation of NFkB family members (Baldwin 1996). The actual pathway by which they induce phosphorylation of IkBs is still poorly understood. Recently, cPLA2 has been implicated in IL-1b-mediated type II sPLA2 gene induction in rat fibroblasts (Kuwata et al 1998). This raises the possibility of a control of the type II sPLA2 gene by free fatty acids or their derivatives via peroxisome proliferator activated receptors (PPARs). PPARs are key players in lipid metabolism and are members of the nuclear receptor superfamily of transcription factors that regulate the pattern of gene expression in response to the binding of small molecular weight ligands (Mangelsdorf et al 1995). Three different subtypes have been described, PPARa, PPARb(d) and PPARg. Although the identities of the ligands that regulate activity in vivo remain to be established with certainty, clofibric acid, leukotriene B4 and oleic acid have been demonstrated to stimulate PPARa and 15-deoxyD12,14-prostaglandin J2 (15DPGJ2) (Forman et al 1995), 12- or 9hydroxyoctadecanoic acid (HODE) (Nagy et al 1998), and certain polyunsaturated fatty acids (Kliewer et al 1997) have been demonstrated to stimulate PPARg-dependent transcription. We demonstrate that, in rat vascular smooth muscle cells, Il-1b induces the type II sPLA2 gene via the NFkB pathway and PLA2/ PPARg pathway (Figure 2.9), but not via the ceramide pathway or the cPLA2–PPARa pathway. Concomitant activation of both the NFkB pathway and the cPLA2–PPARg pathway is necessary to achieve the transcriptional process, suggesting that they cooperate at the level of transcriptional machinery (Couturier et al 1999). In rabbit articular chondrocytes, we have demonstrated that large amounts of type II sPLA2 are secreted in response to IL-1b and that insulin-like growth factor-1 (IGF-1), which counteracts cartilage degradation in arthritis, inhibits this effect. We demonstrated that the IL-1b-induced synthesis and released of type II sPLA2 occurs in parallel with release of PGE2, and results in indirect stimulation of
PLA2 CONTROL OF EICOSANOID PRODUCTION
25
Figure 2.8 Putative effects of secreted phospholipase A2 via unspecific interaction with cellular membranes or via specific receptors
the transcription of type II sPLA2 gene, giving rise to a very stable type II sPLA2 mRNA. We also demonstrated that the inhibitory effect of IGF-1 markedly decreased the half-life of type II sPLA2 mRNA (Jacques et al 1997). Activation of the human counterpart of the type II sPLA2 gene is quite different. We have demonstrated that the NFkB pathway is not directly involved in this process, at least in HepG2 cells, chondrocytes or smooth muscle cells. NFkB must stimulate an intermediary gene, the product of which in turn participates in the induction of human type II sPLA2 gene. We found that the promoter of the human gene possesses a degenerated kB site, which binds C/EBP factors (Figure 2.10) (Couturier et al 2000). This binding relieves the basal silencing of the type II sPLA2 gene (Fan et al 1997). These factors were identified to bind and stimulate many acute-phase response genes. The C/EBP family has three main members, C/EBPa, C/EBPb and C/EBPd. The cell mRNA content was markedly increased after the incubation of hepatocytes with IL-6 or during an experimental systemic inflammatory process. In addition, C/EBPb and C/EBPd are also regulated by phosphorylation. The difference that we observed between humans and rodents in the regulation of the type II sPLA2 gene might be explained by differences in the promoter region of human and rat genes, which strongly diverge. Additional experiments are currently being performed in our laboratory to define the molecular basis for these differences. This might be of importance, since type II sPLA2 is thought to play an important role in human inflammatory disease and might disfavour the rat model in the search for new classes of therapeutic agents.
CONCLUSION It is now evident that phospholipases A2 are very important enzymes in cell physiology. Both cellular and secreted PLA2
Figure 2.9 IL-1-mediated signalling pathways inducing the rat type IIA sPLA2 gene in vascular smooth muscle cells. One of the pathways involves the IkB–NF-kB system, which is activated by proteosome degradation of IkB. The second pathway involves MAP-kinase-mediated phosphorylation of the cytosolic phospholipase A2 and activation of PPAR/RXR by oxygenated fatty acids. The third pathway, via Jun kinases (JNK/SAPK), remains hypothetical
26
BIOSYNTHESIS AND METABOLISM
Figure 2.10 IL-1-mediated activation of the human type IIA sPLA2 gene. The IkB–NF-kB system is replaced by activation of C/EBP dimers via phosphorylation
participate in the housekeeping of cell membrane and lipoprotein integrity. They are also involved in either the immediate control of lipid mediator production after stimulation of membrane receptors by their cognate agonists (cPLA2) or the delayed production of lipid mediators after cell stimulation (type IIA or type V sPLA2). This therefore constitutes a putative target for pharmacological agents. To date, despite tremendous work conducted by pharmaceutical companies, and with the remarkable exception of quinacrine derivatives (which, however, were not initially intended to be PLA2 inhibitors), no PLA2 inhibitors can be used in human pathology. The inhibitors that are available on the market are not very specific and must be used with caution in experiments. This is probably due to the physical properties of their substrates. One additional way to search for new drugs in the future will be to identify the signalling pathways more precisely, controlling their gene expression in order to obtain efficient inhibitors of PLA2 synthesis.
REFERENCES Balboa MA, Balsinde J, Jones SS and Dennis EA (1997) Identity between the Ca2+-independent phospholipase A2 enzymes from P388D1 macrophages and Chinese hamster ovary cells. J Biol Chem, 272, 8576–8580. Baldwin AS Jr (1996) The NF-kB and IkB proteins: new discoveries and insights. Ann Rev Immunol, 14, 649–681. Balsinde J, Balboa MA and Dennis EA (1998) Functional coupling between secretory phospholipase A2 and cyclo-oxygenase-2 and its regulation by cytosolic group IV phospholipase A2. Proc Natl Acad Sci USA, 95, 7951–7956.
Beg AA and Baldwin AS Jr (1993) The IkB proteins: multifunctional regulators of Rel/NF-kB transcription factors. Genes Dev, 7, 2064– 2070. Berguerand M, Klapisz E, Thomas G et al (1997) Differential stimulation of cytosolic phospholipase A2 by bradykinin in human cystic fibrosis cell lines. Am J Resp Cell Mol Biol, 17, 481–490. Caldwell RA and Baumgarten CM (1998) Plasmalogen-derived lysolipid induces a depolarizing cation current in rabbit ventricular myocytes. Circulat Res, 83, 533–540. Comera C, Rothhut B and Russo-Marie MF (1990) Identification and characterization of phospholipase A2 inhibitory proteins in human mononuclear cells. Eur J Biochem, 188, 139–146. Couturier C, Antonio V, Brouillet A et al (2000) Protein kinase Adependent stimulation of rat type II secreted phospholipase A2 gene transcription involves C/EBP-beta and -delta in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol, 20, 2559–2565. Couturier C, Brouillet A, Couriaud C et al (1999) Interleukin 1b induces type II secreted phospholipase A2 gene in vascular smooth muscle cells by a NFkB- and cytosolic phospholipase A2-mediated process. J Biol Chem, 274, 23085–23093. Cupillard L, Koumanov K, Mattei MG et al (1997) Cloning, chromosomal mapping and expression of a novel secretory phospholipase A2. J Biol Chem, 272, 15745–15752. Fan Q, Paradon M, Salvat C et al (1997) C/EBP factor suppression of inhibition of type II secreted phospholipase A2 promoter in HepG2 cells: possible role of single-strand binding proteins. Mol Cell Biol, 17, 4238–4248. Forman BM, Tontonoz P, Chen J et al (1995) 15-Deoxy-delta12–14prostaglandin J2 is a ligand for the adipocyte determination factor PPARg. Cell, 83, 803–812. Hayakawa M, Jayadev S, Tsujimoto M et al (1996) Role of ceramide in the stimulation of the transcription of cytosolic phospholipase A2 and cyclooxygenase 2. Biochem Biophys Res Commun, 220, 681–686. Hernandez M, Burillo SL, Crespo SM and Nieto ML (1998) Secretory phospholipase A2 activates the cascade of mitogen activated protein kinases and cytosolic phospholipase A2 in the human astrocytoma cell line 1321N1. J Biol Chem, 273, 606–612. Jacques C, Bereziat G, Humbert L et al (1997) Posttranscriptional effect of insulin-like growth factor-I on interleukin-1b-induced type II-secreted phospholipase A2 gene expression in rabbit articular chondrocytes. J Clin Invest, 99, 1864–1872. Klapisz E, Ziari M, Wendum D et al (1999) N- and C-terminal plasma membrane anchoring have opposite effects on cytosolic phospholipase A2 activation. Eur J Biochem, 265, 957–966. Kliewer SA, Sundseth SS, Jones SA et al (1997) Fatty acids and eicosanoids regulate gene expression through direct interactions with peroxisome proliferator-activated receptors alpha and gamma. Proc Natl Acad Sci USA, 94, 4318–4323. Kramer RH, Hession C, Johansen B et al (1989) Structure and properties of a human non pancreatic phospholipase A2. J Biol Chem, 264, 5768– 5775. Koumanov K, Wolf C and Bereziat G (1997) Modulation of human type II secretory phospholipase A2 by sphingomyelin and annexin VI. Biochem J, 326, 227–233. Kuroda A, Sugiyama E, Taki H et al (1997) Interleukin-4 inhibits the gene expression and biosynthesis of cytosolic phospholipase A2 in lipopolysaccharide-stimulated U937 macrophage cell line and freshly prepared adherent rheumatoid synovial cells. Biochem Biophys Res Commun, 230, 40–43. Kuwata H, Nakatani Y, Murakami M and Kudo I (1998) Cytosolic phospholipase A2 is required for cytokine-induced expression of type IIA secretory phospholipase A2 that mediates optimal cyclooxygenase2-dependent delayed prostaglandin E2 generation in rat 3Y1 fibroblasts. J Biol Chem, 273, 1733–1740. Lambeau G, Ancian P, Nicolas JP et al (1995) Structural elements of secretory phospholipase A2 involved in the binding to M-type receptors. J Biol Chem, 270, 5534–5540. Lautens LL, Chiou XG, Sharp JD et al (1998) Cytosolic phospholipase A2 (cPLA2) distribution in murine brain and functional studies indicate that cPLA2 does not participate in muscarinic receptor-mediated signaling in neurons. Brain Res, 809, 18–30. Leslie CC (1997) Properties and regulation of cytosolic phospholipase A2. J Biol Chem, 272, 16709–16712.
PLA2 CONTROL OF EICOSANOID PRODUCTION Livrea MA, Tesoriere L, Maggio A et al (1998) Oxidative modification of low-density lipoprotein and atherogenetic risk in b-thalassemia. Blood, 92, 3936–3942. Mangelsdorf DJ, Thummel C, Beato M et al (1995) The nuclear receptor superfamily: the second decade. Cell, 83, 835–839. Mathias S, Younes A, Kan CC et al (1993) Activation of the sphingomyelin signaling pathway in intact EL4 cells and in a cellfree system by IL-1b. Science, 259, 519–522. Miyashita A, Crystal RG and Hay GG (1995) Identification of a 27 bp 5’flanking region element responsible for the low level constitutive expression of the human cytosolic phospholipase A2 gene. Nucleic Acids Res, 23, 293–301. Mukherjee AB, Kundu GC, Mandal AK et al (1998) Uteroglobin: physiological role in normal glomerular function uncovered by targeted disruption of the uteroglobin gene in mice. Am J Kidney Dis, 32, 1106–1120. Nagy L, Tontonoz P, Alvarez, JG et al (1998) Oxydized LDL regulates macrophage gene expression through ligand activation of PPARg. Cell, 93, 229–240. Pavoine C, Magne S, Sauvadet A and Pecker F (1999) Evidence for a b2adrenergic/arachidonic acid pathway in ventricular cardiomyocytes. Regulation by the b1-adrenergic/cAMP pathway. J Biol Chem, 274, 628–637. Pruzanski W, Stefanski E, Vadas P et al (1998) Regulation of the cellular expression of secretory and cytosolic phospholipases A2 and cyclooxygenase-2 by peptide growth factors. Biochem Biophys Res Commun, 1403, 47–56.
27
Seilhamer JJ, Pruzanski W, Vadas P et al (1989) Cloning and recombinant expression of phospholipase A2 present in rheumatoid arthritic synovial fluid. J Biol Chem, 264, 5335–5338. Stafforini DM, McIntyre TM, Zimmerman GA and Prescott SM (1997) Platelet-activating factor acetylhydrolases. J Biol Chem, 272, 17895– 17898. Tang J, Kritz RW, Wolfman N et al (1997) A novel cytosolic calciumindependent phospholipase A2 contains eight ankyrin motifs. J Biol Chem, 272, 8567–8575. Tay A, Maxwell P, Li ZG et al (1994) Cytosolic phospholipase A2 gene expression in rat mesangial cells is regulated post-transcriptionally. Biochem J, 304, 417–422. Tischfield JA (1997) A reassessment of the low molecular weight phospholipase A2 gene family in mammals. J Biol Chem, 272, 17247– 17250. Underwood KW, Song C, Kriz RW et al (1998) A novel calciumindependent phospholipase A2; cPLA2-g that is prenylated and contains homology to cPLA2. J Biol Chem, 373, 21926–21932. Walker G, Kunz D, Pignat W et al (1995) Pyrrolidone dithiocarbamate differently affects cytokine- and cAMP-induced expression of group II phospholipase A2 in rat renal mesangial cells. FEBS Lett, 364, 218– 222. Yamashita A, Sugiura T and Waku K (1997) Acyltransferases and transacylases involved in fatty acid remodeling of phospholipids and metabolism of bioactive lipids in mammalian cells. J Biochem (Tokyo), 122, 1–16.
3 Mechanisms of PGH Synthase-1 (COX-1) Activity and Role of Radical States Carol Deby and Ginette Deby-Dupont University of Lie`ge, Lie`ge, Belgium
BRIEF OVERVIEW The first works in the research on prostaglandin endoperoxide H (PGH) synthase (PGHS) were performed simultaneously by Van Dorp et al (1964) and Bergstro¨m et al (1964), who described a sequence of reactions starting when a proteic extract of sheep seminal vesicles was incubated with arachidonic acid. Further research established that three polyunsaturated fatty acids (chain length of 20 carbon atoms; Figure 3.1) incubated with seminal vesicle extracts were oxidized by two molecules of oxygen. These three precursors became cyclic compounds, passing through free radical states, finally producing the prostaglandin PGH, the direct precursor of the various biologically active prostanoids (prostaglandins, prostacyclin, thromboxanes) and their numerous metabolites. Further data demonstrated the complexity of the peroxide and endoperoxide rearrangements, leading first to endoperoxide G and second to endoperoxide H, highly unstable and quickly metabolized by a cascade of reactions into more stable compounds, the prostanoids. At the origin of the synthesis of these cyclic compounds derived from arachidonic acid and the two other 20-carbon chain unsaturated fatty acids, there are two common root compounds, endoperoxide G (PGG) and endoperoxide H (PGH), deriving from the reduction of PGG. A single enzyme, displaying two enzymatic activities (cyclooxygenasic and peroxidasic), produces the root compounds: this enzyme is referred to as prostaglandin endoperoxide H synthase (PGHS). The main pathways of arachidonic acid peroxidation by PGHS are presented in (Figure 3.2). The first step is a double dioxygenase or cyclooxygenase reaction (bis-dioxygenation; Ogino et al 1978), binding two oxygen molecules on the polyunsaturated fatty acid precursor. A cyclic molecule is produced with both 9,11endoperoxide and 15-hydroperoxyl groups (Figure 3.2), called endoperoxide G or PGG (Samuelsson 1972; Nugteren and Hazelhof 1973). The following reaction summarizes this first step: C20:3, C20:4, C20:5+2O2!PGG1,2,3 (bis-dioxygenation) In the second step, the 15-hydroperoxide group of PGG is reduced to hydroxyl by the peroxidase activity of PGHS, forming the endoperoxide H or PGH. This reaction follows the classical scheme, using a hydrogen donor DH2: PGG+DH2!PGH+H2O+D
(D is often a free radical)
From PGG and PGH, several prostanoids can be produced; their nature is determined by the enzymes of the considered tissue The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
and according to the biochemical circumstances, e.g. from endoperoxides there is a generation of thromboxane A2 in platelets, while in endothelial cells, prostacyclin (PGI2) seems to be predominant. PGD appears to be relatively abundant in the rat spleen (Christ-Hazelhof and Nugteren 1979), as shown by in vitro experiments in the presence of albumin (Christ-Hazelhof et al 1976; Hamberg and Fredholm 1976) or glutathione-S-transferase (Christ-Hazelhof et al 1976). Until the beginning of the 1990s, it was generally admitted that only one gene existed for PGH synthase (Yokoyama and Tanabe 1989), but it was recognized that this enzyme, originally named cyclooxygenase (COX), was found in tissues in the form of two isoenzymes, COX-1 and COX-2. It is now demonstrated that these two isoforms are encoded by two different genes (Kraemer et al 1992). COX-1 is the classical constitutive (‘‘housekeeping’’) enzyme, and COX-2 is the inducible isoform. Whereas these isoenzymes differ substantially with respect to their expression and biology, they possess similar structures and express the same two catalytic activities. This chapter deals only with PGHS-1 (COX-1). A general review of COX-2 properties was published by Herschman (1996) and can be found in recent reviews with a comparison of the molecular biology (biosynthetic and signalling pathways) of the two enzymes (Smith et al 2000). The general enzymatic mechanisms described for COX-1 are applicable to COX-2.
STRUCTURE OF PGH SYNTHASE-1 (EC 1.14.99.1) PGH synthase-1 (or cyclooxygenase-1: COX-1) is a bis-dioxygenase displaying both cyclooxygenase and peroxidase activities (O’Brien and Rahimtula 1976; Miyamoto et al 1976), thus catalysing the two following reactions where AA is arachidonic acid, C20:4 (but this fatty acid can be replaced by C20:3 or C20:5): 1. AA+2O2 !PGG2 (cyclooxygenase reaction) 2. PGG2 !PGH2 (peroxidase reaction) This enzyme can be separated from the protein fraction converting PGH into PGE (Miyamoto et al 1974, 1976). PGHS-1 is a membrane-bound haemoglycoprotein containing mannose and Nacetyl-glucosamine (van der Ouderaa et al 1977). Its 576 amino acid sequence has been completely elucidated (Smith et al 1996). A dimeric structure has been proposed, with both dimers structurally and functionally similar (Roth et al 1980; van der Ouderaa et al 1977; Xiao et al 1998). The monomer is a single polypeptide chain of 72 kDa with three structural domains: an epidermal
30
BIOSYNTHESIS AND METABOLISM
Figure 3.1 The three polyunsaturated fatty acids, precursors of the prostanoids, and their corresponding derived PGHs. I, eicosatrienoic (o-6) or homo-g-linolenic acid C20-3; Ia, PGH-1; II, eicosatetraenoic (o-6) or arachidonic acid C20-4; IIa, PGH-2; III, eicosapentaenoic (o-3) acid C20-5; IIIa, PGH-3
growth factor domain at the N-terminus; a neighbouring membrane-binding domain; and a large globular catalytic domain at the C-terminus (closely similar to the catalytic domain of neutrophil myeloperoxidase) with a hydrophobic channel (Smith et al 2000). Each subunit contains a haem (van der Ouderaa et al 1977; Roth et al 1980; Nugteren et al 1981). The enzyme can be solubilized and purified in non-ionic detergents (Miyamoto et al 1976; van der Ouderaa et al 1977; Hemler and Lands 1976). During purification, the apoenzyme can be separated from its haem and becomes completely inactive. The reconstitution of the enzymic activity can be driven by haemin addition (Miyamoto et al 1974). The distribution of PGHS-1 in the cell was studied by Morita et al (1995a, 1995b), using quantitative confocal fluorescence imaging microscopy. The enzyme was found equally distributed in the endoplasmic reticulum and perinuclear membranes using immunological techniques, but the activities were mainly localized in the perinuclear region. It is not our purpose to give details on the techniques and methods of purification of PGHS; however, it must be underlined that, according to Roth et al (1980), PGHS-1 can be purified to near homogeneity in an active form containing a single subunit. The combination of classical protein chemistry, UV-visible spectroscopy (Lambeir et al 1985; Dietz et al 1988), electron spin resonance (Karthein et al 1987, 1988; DeGray et al 1988, 1992; Tsai et al 1992, 1994a) and X-ray crystallography (Picot et al 1994) permits us to imagine the PGHS as a spheroid (Marnett et al 1999) (Figure 3.3).
The active site of the enzyme is formed by a hydrophobic channel opening into the outside of the protein, accessible for dioxygen molecules and presenting specific sites that intervene in the binding and stereospecific oxidation of arachidonic acid. Three amino acid residues are especially important: Arg120, which serves as counter-ion for the carboxylic function of the substrate and is the anchoring base of the substrate (Picot and Garavito 1994; Bhattacharyya et al 1996); Tyr 385, which plays the main role in the oxidation of the fatty acid; and Ser 530, of which the acetylation blocks the dioxygenase activity (Loll et al 1995). The substrate carboxylate is positioned adjacent to Arg 120, which plays a crucial role in binding arachidonate and arylalkanoic acid-type inhibitors (Kulmacz et al 1994) and the 13-pro-S hydrogen is placed near the phenolic hydroxyl of Tyr 385. The importance of the Arg 120 residue in PGHS-1 is recognized for its interaction with arachidonic acid and non-steroidal antiinflammatory drugs (NSAIDs) containing a free carboxylic acid moiety (Mancini et al 1995). The o-end of the fatty acid is inserted into a channel at the top of the cyclooxygenase active site that eventually leads to the surface of the protein. This end of arachidonate straddles the a-helix containing Ser 530, the site that is acetylated by aspirin. Acetylation blocks the arachidonate binding to PGHS (Shimokawa and Smith 1992; Barnett et al 1994). PGHS appears to have two high-affinity binding sites for metalloporphyrins. The two sites have slightly different affinities for haem. The synthase dimer is capable of cyclooxygenase catalysis when the site with the highest affinity is occupied by the haem. The two subunits of the enzyme are thus not completely identical (Kulmacz and Lands 1984). Haem can be replaced by manganese protoporphyrin, giving a system where the enzyme only catalyses the bis-dioxygenation producing prostaglandin G1 (Ogino et al 1978; Tsai et al 1998). NSAIDs compete directly with arachidonate for binding to the cyclooxygenase site and inhibit cyclooxygenase activity, but have little effect on peroxidase activity (Capdevila et al 1995; Kulmacz and Lands 1985a). THE TWO MAIN MECHANISMS OF PGH SYNTHASE ACTIVITY Two Distinct Sites The two activities of PGHS (bis-dioxygenasic and peroxidasic) are displayed on the same enzyme molecule on two distinct sites, physically and functionally separated, that are closely localized on the protein chain (Marshall and Kulmacz 1988). Tyr 385 is located between the haem- and arachidonate-binding sites, well positioned to serve a role in coupling the two enzyme activities (Picot and Garavito 1994). The existence of these distinct sites was first demonstrated by using acetyl salicylic acid (ASA), which blocks the dioxygenase reaction but does not affect the peroxidase activity. Aspirin acetylates a single serine residue of an active site on the enxyme (Roth et al 1975, 1981; van der Ouderaa et al 1977, 1980; Hemler et al 1978), related to the cyclooxygenase function, without modifications of the peroxidase activity (Miyamoto et al 1976; van der Ouderaa et al 1977). On the other hand, replacement of the haem by manganese protoporphyrin maintains only the cyclooxygenase but not the peroxidase activity (Ogino et al 1978). After treatment with ASA, PGHS is always able to transform exogenous PGG2 into PGH2. The peroxidase reaction acts not only on PGG2 but also on other peroxides. Mechanism of PGG Formation: Cyclooxygenasic Function This function is believed to involve a free radical state, initiated by the abstraction of a hydrogen atom at carbon 13 on arachidonic
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES
31
Figure 3.2 Chemical steps in the conversion of arachidonic acid into PGG2. The enzyme removes the 13-pro-S-hydrogen (I), which generates a pentadienyl radical with maximal electron density at carbons 11 and 15 (II). The trapping of the radical at carbon 11 with O2 produces a peroxyl radical, which adds to carbon 9, generating an endoperoxide and a carbon-centred radical at carbon 8 (III). The carbon 8 radical adds to the double bond at carbon 12, generating the bicyclic peroxide and an allylic radical with maximal electron density at carbons 13 and 15 (IV). The trapping of the carbon radical at carbon 15 with O2 generates a peroxyl radical (V) which, in the presence of an electron donor (D.H), is reduced to PGG2 (VI)
acid (Hamberg and Samuelsson 1967). The reality of the radical state, carbon-centred, of arachidonic acid was evidenced by Mason et al (1980). Since the end of the 1980s, it has been demonstrated that the hydrogen atom abstraction was conducted by Tyr 385 in PGHS-1 (see below). At this state, an oxygen molecule could react with carbon atom 11 to form a peroxyl radical, which could abstract a hydrogen atom to form a hydroperoxide (Schreiber et al 1986). From the work of Dietz et al (1988), it appears clearly that the hydrogen abstraction is done by the tyrosyl radical, Tyr. 385. Until 1994, the role of Tyr. was hardly debated (Kulmacz et al 1994). But recent papers have afforded numerous arguments in favour of the importance of the Tyr. 385 step (Tsai et al 1994a, 1999). Since the regeneration of Tyr. 385 is dependent on peroxidasic activity, the cyclooxygenasic function closely depends on the presence of a sufficient hydroperoxide level. A balance must thus exist between the two activities that will regulate the global synthesis of PGH. Kulmacz et al (1994) realized a computerized model which can simulate PGHS activity. However, in in vivo conditions, outcoming hydroperoxides such as hydrogen peroxide (H2O2) are able deeply to modify the expected results (see below). Other factors can also modify the reactions, such as antioxidants and the presence of carbonates (from CO2). Thus, in vitro experiences are often very difficult to reproduce. The presence of glutathione (GSH) peroxidase
results in a strong PGHS inhibition (Hemler et al 1978) because this enzyme reduces inorganic and organic hydroperoxides. Mechanism of PGH Formation: Peroxidasic Function Analogies between PGH Synthase and the Classical Haem Peroxidases The first analogy between the PGHS and compound I (Cd I) of horseradish peroxidase (HRP), the typical model of peroxidases, was given by Nastainczyk et al (1984). They observed the classical spectral characteristics of Cd I during the kinetic study of PGHS, i.e. the formation of a ferryl-oxo complex (FeO)3+ of the haem, analogous to the formation of Cd I of horseradish peroxidase. These results were confimed by Lambeir et al (1985), who calculated that Cd I of PGH synthase is quickly converted to compound II. Confirmations were afforded by the team of Marnett (Markey et al 1987). Co-factors are strongly required as electron donors since, in their absence, arachidonic acid irreversibly inactivates PGHS (Markey et al 1987), but until recently these biological reducers were unknown (Zenser et al 1999). Many naturally occurring compounds have been assayed as electron donors, such as NADPH, NADH, glutathione, methionine, tryptophan,
32
BIOSYNTHESIS AND METABOLISM
Figure 3.3 Schematic representation of the monomer of PGH synthase-1. For details see text
epinephrine, ascorbic acid, lipoic acid and uric acid (Markey et al 1987; Zenser et al 1999). However, in some cases, some of these substances could act as free radical scavengers (see section on regulation, below). Different molecules can play the role of electron donor, each itself becoming a free radical. A good model is given by various oestrogens (Freyberger and Degen 1989). Glutathione seems to slow the rate of PGHS activity (Freyberger and Degen 1989). In the resting (native) state, the haem of PGHS is in the ferric high-spin state and the iron atom is at oxidation degree III (Karthein et al 1987). On the porphyrin macrocycle, there are numerous sequences of the carbon chain containing conjugated double bonds, a favourable structure for the resonance of a free electron. During peroxidase activity, an electron of one of these sequences can be removed at the same time as an electron of the ferric atom. This phenomenon occurs by the reaction of the haem with a hydroperoxide molecule (i.e. H2O2, cumene peroxide, lipid peroxide).
The Cycle of Oxidoreduction of the Haem in PGHS (Figure 3.4) In the presence of a peroxide (R–OOH), the haem is able to reduce the –OOH function by giving two electrons. One of these electrons originates from the porphyrin macrocycle. The departure of this electron causes the formation of a radical state, with resonance by way of the conjugated double bonds of the macrocycle. This electron delocalization is represented in Figure 3.4 by dotted lines. The electron loss is associated with gain of a positive charge (+ on Figure 3.4). The second electron comes from the haem iron, which is thus oxidized to the ferryl state (FeIV) associated with an oxygen atom. This structure is analogous to the Cd I first described for horseradish peroxidase. The radical structure on the porphyric ring and the positive charge explain the name given to Cd I, pcation radical. During the activity of PGHS, it is clear that the
peroxide used will be PGG2, but other peroxides can trigger PGHS activity. H2O2 is one of these possible triggering peroxides. We shall see further what kind of peroxides are used to form Cd I. Recently, Marnett et al (1999) underlined the role played by peroxynitrite (see below: triggering). At the end of the 1980s, Karthein et al (1987, 1988) demonstrated by ESR and crystallographic studies that Cd I is transformed into Cd II by a long-range electron transfer (LRET) from a tyrosyl residue of the protein chain (Tyr 385), an observation confirmed by Shimokawa et al (1990). This electron given by the tyrosyl residue is coupled to the free electron of the porphyrin macrocycle, which returns to a non-radical state, losing its positive charge. The radical state thus becomes localized in a stereospecific site that has been described above. Facing the tyrosyl radical is carbon 13 of arachidonic acid. A hydrogen atom is abstracted from arachidonic acid at this carbon 13, forming an arachidonyl radical, which is able successively to trap two exogenous dioxygen molecules (present in the protein channel), following a mechanism previously described (see above). Finally, Cd II is reduced to the ferric state by mechanisms generating free radicals. Effectively, two molecules of hydrogen donor are needed to reduce the oxygen atom of the ferryl structure and FeIV into FeIII. These side-reactions associated with prostaglandin synthesis are described in a separate part of this chapter. How is the radical located on Tyr 385 reduced for a next dioxygenase cycle? The final step in each round of arachidonic acid oxygenation is the reduction of the peroxyl radical precursor of PGG2 by Tyr 385, which regenerates the Tyr 385 radical for the next round of cyclooxygenase catalysis (Smith et al 1992). This leads to multiple turnovers per activation event and allows the accumulation of PGG2 (Seibert et al 1994; Xie and Herschman 1995). To be able to initiate a novel cycle, Cd II must be reduced to the resting form (FeIII). It is one of the steps where an electron donor (DH) is required. By donation of an H., DH can become a free radical D. (for the central role of Tyr 385, see Figure 3.5). The interaction of haem with PGHS is stoichiometric (Tsai et al 1997). Only one subunit needs to be bound to a haem to form a catalytically active synthase dimer, and an excess of haem is an inhibitor (Kulmacz and Lands 1984). Preincubation of the enzyme with haematin or haemoglobin results in the loss of enzyme activity. The enzyme inactivation is prevented by tryptophan or various other aromatic compounds (Ogino et al 1978). Coupling of the Two Mechanisms of PGHS Activity Two mechanisms are proposed to account for the combination of the cyclooxygenase and peroxidase activities of PGHS. Dietz et al (1988) proposed a branched-chain process, hypothesizing that the cyclooxygenase reaction propagates independently of peroxidase activity. Bakovic and Dunford (1994) proposed tightly coupled mechanisms; for these authors, the peroxidase cycle participates totally in cyclooxygenase propagation. Wei et al (1995) consider the branched-chain mechanism to be more probable than the tightly coupled mechanism. NATURE OF THE AGENTS TRIGGERING PEROXIDASE ACTIVITY Requirement for a Hydroperoxide ROOH Hemler and co-workers (Hemler et al 1979; Hemler and Lands 1980) observed that the addition of purified arachidonic acid to a preparation of cyclooxygenase did not consume oxygen during a more or less long time period that depends on the degree of purity of arachidonic acid (lag time). The addition of trace amounts of
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES peroxidized arachidonic acid suppressed this lag time. Other peroxides, such as peroxidized linoleic acid, can play the same triggering role; tert-butyl-hydroperoxide (O’Brien and Rahimtula 1976) and arachidonate peroxides (Narayanan and Harrington 1980) have a significant enhancing effect. The efficient triggering action of hydroperoxides is evidenced when GSH peroxidase is added to the enzyme system: PGHS is blocked. This triggering phenomenon is now easily explained by the need to oxidize the ferric haem into its ferryl state (see Figure 3.4). A source of hydroperoxides is cytochrome P-450BM-3 (Capdevila et al 1996). The Problem of H2O2 The triggering effect of hydroperoxides may be simply obtained by H2O2. The stimulation of PGHS by optimal H2O2 concentrations was observed at the end of the 1970s (Hemler et al 1979; Deby and Deby-Dupont 1980). However, the direct role of H2O2, as cyclooxygenase trigger, was minimized by Hemler et al (1979): the action of H2O2 would be too weak and would be prevented by catalase. Later on, a complete inactivation of PGHS by H2O2 with bleaching of the haem Soret absorbance peak was found when using 20 moles H2O2/mole PGHS haem (Tsai et al 1992). But more recently the triggering effect of H2O2 has again been contested (Hajjar et al 1995). The latest publications on the subject admit that H2O2 will play an activating role at optimal concentrations (Bakovic and Dunford 1996a). The formation of Cd I, using H2O2 as substrate, was studied as a function of pH and temperature. The protonated form of H2O2 preferentially reacts with the unprotonated form of the enzyme over the pH range of 3.5–9.1, suggesting the importance of acid–base catalysis for Cd I formation (Bakovic and Dunford 1996b). Many other observations have confirmed the role of hydrogen peroxide. Ascorbate activates cyclooxygenase through the generation of H2O2 (Polgar and Taylor 1980). GSH peroxidase reduces H2O2 as well as other hydroperoxides, and is an inhibitor of PGHS activity (Hemler et al 1978). Activated leukocytes produce H2O2, which is partially released outside these cells (Badwey and Karnovsky 1980). It was also observed that activated leukocytes stimulated cyclooxygenase activity in endothelial cells (Rampart et al 1981). It is thus logical to hypothesize that H2O2 acts as a messenger from leukocytes to neighbouring cells and platelets. The intracellular sources of H2O2 are numerous, since H2O2 is formed in mitochondria and by numerous enzymatic systems (e.g. cytosolic xanthine oxidase, NADPH-oxidase). It has recently been demonstrated that endothelial cells possess a functional leukocyte-type NADPHoxidase (Meyer et al 1999) and that this enzyme (or a parent form) is present in vascular smooth muscle cells, fibroblasts and myocytes (Griendling et al 2000). It is therefore highly probable that H2O2 effectively interacts in vivo with PGHS activity. However, if many data demonstrated that H2O2 can play an efficient role during PGHS activity and that H2O2 and O.2generating systems are stimulating on prostanoid production, it also appears that an optimal concentration of H2O2 is needed to reach PGHS stimulation, H2O2 becoming an inhibitor when used above this optimal concentration (Figure 3.6) (Deby 1988; Deby and Deby-Dupont 1981). The favourable role of an optimal concentration of peroxide has long remained not clearly understood. This is now explained by the mechanism of action of PGHS, which uses peroxide to break a covalent bond on the haemic structure of the enzyme, starting the chain reaction that will lead to arachidonic acid transformation. PGHS-1 would have a higher threshold for peroxide than PGHS-2 (Kulmacz and Wang 1995; Capdevila et al 1996) and a negative allosteric regulation by low arachidonic acid concentrations (Swinney et al 1997).
33
Role of Nitric Oxide and Peroxynitrite The Role of NO . The effect of nitric oxide (NO.) on PGHS remains controversial. For some authors, NO. is an inhibitor (Kanner et al 1992; Picot et al 1994; Curtis et al 1996), without effect (Tsai et al 1994a, 1994b) or a significant activator (Corbett et al 1993; Franchi et al 1994; Kelner and Uglik 1994; Riutta et al 1994). Direct activation of PGHS-1 by nitric oxide or NO. donors has been reported (Salvemini et al 1993, 1995a, 1995b; Hajjar et al 1995). However, the effect of NO. is still discussed, as results reporting activation of PGHS by NO. have proved to be difficult to reproduce, even with enzymes from the same sources (Capdevila et al 1995; Kulmacz and Wang 1995). The biochemical basis for direct cyclooxygenase activation by NO. has yet to be demonstrated convincingly. NO has only a very weak affinity for the ferric haem in PGHS-1, and PGHS–haem ligands generally inhibit, rather than activate, cyclooxygenase activity (Kulmacz and Lands 1985b). If NO. binds tightly to ferrous PGHS-1 haem (Capdevila et al 1995), the ferrous state is rapidly oxidized in air and is not observed during catalysis (Masferrer et al 1994). Published data of a PGHS activation by NO. or derived nitrosothiols report the use of high concentrations of NO., implausible when compared to the low micromolar levels expected in vivo. These properties make it very unlikely that NO. is a significant PGHS haem ligand at the low micromolar levels expected in vivo. Nitrosothiol formation on PGHS cysteine residues has been proposed as a mechanism for cyclooxygenase activation by NO. (Hajjar et al 1995), but the published data yield an implausibly high stoichiometry of PGHS nitrosothiol formed vs. NO. added, indicating that the issue needs to be re-examined (Kulmacz 1998). The Role of ONOO7 Taking into account a regulatory regime of PGHS activity by peroxides, it is conceivable that peroxynitrite (ONOO7), the coupling product of NO. and superoxide anion (Beckman et al 1990), is a PGHS activator (Landino et al 1996; Kulmacz 1998). The indirect cyclooxygenase activation via ONOO requires appreciable levels of O2. (Landino et al 1996), which is not produced during catalysis by purified PGHS unless particular cosubstrates, such as NADH or NADPH, are added (Kukreja et al 1986). The team of Marnett recently published that peroxynitrite (ONOO7) is an excellent oxidant of the haemic FeIII (Figure 3.7) (Gunther et al 1997; Goodwin et al 1998, 1999). GSH peroxidase is unable to inhibit this step as it does when ROOHs are used as triggers. This triggering role of ONOO7 can occur in some cells, producing simultaneously NO. and O.2 (endothelial cells, leukocytes). When ONOO7 is the triggering agent, the PGHS activation is inhibited by superoxide dismutase, which prevents the formation of ONOO7 from NO. and O.2 (Marnett et al 1999). These data provide a biochemical link between NO. biosynthesis and prostaglandin biosynthesis, and may explain the observations that NO synthase inhibitors reduce prostaglandin biosynthesis in inflammatory lesions in vivo (Marnett et al 1999). ASSOCIATED OXIDATIVE MECHANISMS Co-oxidation Marnett et al (1975) first described co-oxygenation as the oxidation of compounds by the arachidonic acid–PGH synthase
34
BIOSYNTHESIS AND METABOLISM
Figure 3.4 Mechanisms of PGHS-1 peroxidasic function. DH2, electron donor (co-factor) which can produce a radical species; grey ellipsoid, schematic view of the apo-protein, containing Tyr–OH (Tyrosine 385). ROOH is a hydroperoxide able to oxidize the haem. A, formation of the intermediate I (Cd I); B, formation of the intermediate II, after a long-range electron transfer (LRET), reducing the macrocycle radical, and forming a tyrosyl radical, at Tyr 385; C, abstraction of H. at carbon 13 of the eicosapolyenoic acid with formation of a lipidic radical and reduction of the tyrosyl radical, giving Cd II; D, reduction of Cd II to the resting enzyme by an electron donor, DH2, which becomes a radical (D.)
system. This phenomenon has been now termed co-oxidation. A wide variety of chemicals are co-oxidized during the reduction of PGG2 to PGH2. The best knowledge of the mechanisms involved in prostaglandin biosynthesis now provides a clearer understanding of the co-oxidation processes, which appear as important as those occurring in the cytochrome P-450 system. The co-oxidation occurs when the reduction of the ferryl state FeIV of the haem to the resting state, FeIII, requires the oxidation of an electron donor, DH (see Figure 3.4). A good example is given by oxidation of p-coumaric acid by PGHS and H2O2 (Bakovic and Dunford 1993): coumaric acid serves as a reducing substrate for PGHS through two one-electron oxidation steps. The electron donor DH can become a free radical, D., which initiates radical reactions, as concluded by Eling et al (1990), reviewing the results published by different research teams. For these authors, the common point in the different oxidative mechanisms intervening in co-oxidation (epoxidation, peroxidation, etc.) was the formation of free radicals. PGHS, as well as other peroxidases (horseradish peroxidase, thyroid peroxidase, myeloperoxidase), is able to oxidize various compounds and to transform some of these into toxic derivatives. Dix et al (1986) described a hydroperoxidedependent epoxidation of an aromatic molecule derived from
anthracene. Baumann et al (1983) explained the weak effect of cyclooxygenase inhibitors of several NSAIDs (phenylbutazone, aminopyrine) by a co-oxidation, transforming these molecules into less active metabolites, while other, more active NSAIDs were found to be less oxidized by PGHS activity (diclofenac, indomethacin). Electron spin resonance (ESR) techniques using spin-traps indicated that free radical intermediates are formed during the metabolism of certain drugs by PGHS. Phenytoin is oxidized into a toxic compound by PGHS, with a free radical step, as demonstrated by ESR (Kubow and Wells 1989). Samokyszyn et al (1995) and Freyaldenhoven et al (1996) published interesting results, obtained by ESR, on the generation of free radical species during the co-oxidation of all-transand 13-cis-retinoic acid.
Singlet Oxygen Generation Marnett et al (1974) early on observed that PGH synthase activity was associated with an ultra-weak red chemiluminescence, which was attributed to the formation of singlet oxygen. Cadenas et al (1983) later confirmed Marnett’s hypothesis.
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES
35
Figure 3.5 The cycles of oxido-reduction, and the central role of Tyr 385. DH, electron donor; grey ellipsoid, schematic view of the enzyme with tyrosine 385 and the haem; AA, arachidonic acid
Co-oxidation and Cancer The relationship between PGHS activity, generating free radicals and oxidation products, and carcinogenesis has been actively studied since the 1970s. The first compound to be studied was a-benzopyrene, a polycyclic aromatic hydrocarbon, which was easily oxidized into carcinogenic epoxides (Huberman et al 1976). Further studies demonstrated the carcinogenic role of arachidonic acid oxidation (Eling et al 1990). The presumably carcinogenic food antioxidant 2(3)-tert-butyl-4-hydroxyanisole (BHA) is oxidized into various metabolites, particularly 2-tert-butyl (1,4) hydroquinone (TBHQ) and its quinone, TBQ. ESR analyses showed that PGHS activity resulted in a substantially accelerated metabolism of TBHQ into TBQ, which was accompanied by the formation of superoxide anion, hydroxyl radical and H2O2. Both acetylsalicylic acid and indomethacin induced a significant decrease in TBQ excretion into urine. Co-oxidation by PGHS of the BHA metabolite TBHQ into TBQ, yielding reactive oxygen species, might therefore be responsible for the carcinogenic and toxic responses elicited by this antioxidant (Schilderman et al 1993). NSAIDs such as aspirin are considered to be anticancerous agents. Colorectal cancer prevention by aspirin is demonstrated by a recent epidemiological study (Giovannucci 1999). In studies on animals also, colorectal carcinogenesis induced by azoxymethane was inhibited by aspirin (Li et al 1999). NATURAL REGULATION OF PGHS-1 PGHS-1 has a cyclooxygenase turnover number of +3500 moles arachidonate/min/mole dimer (Kulmacz et al 1994; Barnett et al 1994), similar to that of PGHS-2 (Smith et al 2000). The formation of prostaglandins by PGHS-1 can be limited by the availability of the fatty acid substrate, by the concentration of
triggering hydroperoxides, by cellular and cytoplasmic factors and by the self-catalysed inactivation (suicide inactivation) of PGHS. Control of the Bioavailability of the Precursors PGHS-1 function requires exogenous arachidonic acid (Reddy and Herschman 1994) and seems to have a sigmoidal dependence on this free fatty acid (allosteric activation). At this level, a control can be exerted on the release of the free fatty acids and on their transport to the active site of the enzyme. The intracellular concentration of arachidonic acid available for prostanoid synthesis seems to equilibrate with the extracellular pool of the free fatty acids. The concentration of arachidonic acid is itself a regulator of PGHS, since arachidonic acid in excess, leading to increased PGHS activity, will also induce the simultaneous production of free radicals and inactivation of PGH synthase (Gonchar et al 1999). The release of free fatty acids appears to be controlled by hormonal and adrenergic stimulations, but also by inflammatory, immunological and even mechanical mechanisms. These actions may be explained by the activity of several enzymes, phospholipases A2, cholesterol esterase and triglyceride lipases. Phospholipases exist in different forms, the cytosolic (cPLA2) and secretory (sPLA2; two isoforms) phospholipases and a Ca2+independent phospholipase (iPLA2). Phospholipase A2 activity is under hormonal control and is stimulated by bradykinin, catecholamines, angiotensin II or histamine. Calcium mobilization acts on cytosolic and secretory phospholipase stimulation by a non-receptor pathway. It also appears that the cytosolic phospholipase A2 is activated by tumour necrosis factor-a (TNFa): this cytokine produces an activation of the mitogenactivated protein kinase cascade, which is followed by the phosphorylation of phospholipase A2 and the release of arachidonic
36
BIOSYNTHESIS AND METABOLISM
acid (Hernandez et al 1999). From recent work on the interrelations between the phospholipases and PGHS activities, it appears that PGHS-1 and -2 use cPLA2 and sPLA2 but that iPLA2, which is the dominant phospholipase involved in arachidonic acid release, serves in cell membrane remodelling but does not induce PGHS activity (Murakami et al 1999; Fitzpatrick and Soberman 2001). The role of glucocorticoids still remains a matter of discussion. From much reported data, it appears that glucocorticoids may initially stimulate phospholipase A2 activity, while inhibiting it at a later phase. This biphasic behaviour may be attributed to different concentrations of a PLA2-modulating protein, possibly lipocortin, that accumulates during exposure to glucocorticoids (Prajgrod and Danon 1992). The glucocorticoids thus act as inhibitors by an indirect pathway: they induce the synthesis of the proteic phospholipase inhibitors, lipocortin and its related polypeptides, macrocortin and renocortin (Douglas et al 1992). Glucocorticoids and lipocortin-1 (lipocortin peptide 2-26) may also act by inhibiting the expression of phospholipase A2 and PGHS, but this last effect has only been described for PGHS-2 (Masferrer and Seibert 1994; Masferrer et al 1994; Minghetti et al 1999; Ferreira et al 1997). Cholesterol esterase is subject to hormonal control. Its activity of free fatty acid release from cholesteryl esters may be partially controlled by the activity of lecithin cholesterol acyl transferase (LCAT), which can use the free fatty acids carried by HDL for cholesterol esterification. The activation of the triglyceride lipases that release free fatty acids from the triglyceride reserves is dependent on the activation of adenyl cyclase, a mechanism that may be important in the kidney, where the triglycerides are rich in arachidonic acid. The transport of free fatty acids in plasma is mainly regulated by albumin, with a molar ratio of free fatty acid to albumin not exceeding 4. However, free fatty acids are also carried by other plasma compounds, cyclodextrin, haptoglobin and lipoproteins. Heparin, by releasing lipoprotein lipase, induces a lysis of LDL triglycerides, increasing the plasma concentrations of free fatty acids. Heparin also weakens the bond between free fatty acids and albumin, making them more available for PGHS activity (Deby 1988).
Availability of the Triggering Agent (ROOH or ONOO7) Hydroperoxide dependence is now a firmly established fact (Chen et al 1999). At the end of the 1970s we demonstrated the enhancer role of an optimal concentration of H2O2 on seminal vesicle homogenates (Deby and Deby-Dupont 1980; see Figure 3.6). Catalase, which specifically destroys H2O2, decreases prostaglandin synthesis in activated leukocytes, while U937 cells incubated with labelled arachidonic acid show a strong enhancement of prostaglandin formation when H2O2 is added to the system (Marshall et al 1987). An important intracellular inhibitor is GSH peroxidase, acting through its peroxidasic activity, which eliminates H2O2 and other lipoperoxides necessary for the triggering of PGHS (see above). The addition of GSH-peroxidase to the PGHS–arachidonic acid solution significantly decreases the synthesis of PGH2 (Marshall et al 1987), a fact that is explained by the destruction of both lipid and hydrogen peroxides. GSH peroxidase may also be inactivated by scavenging PGG2, the hydroperoxide derived from PGHS activity that could further activate other PGHS molecules and involve a chain reaction (Marshall et al 1987). Margalit et al (1998) observed that elevated GSH levels inhibit prostanoid production in in vivo models, providing evidence for the role of GSH in the regulation of prostanoid biosynthesis.
Figure 3.6 Effects of increasing concentrations of hydrogen peroxide (H2O2) on the biosynthesis of prostanoids (PGs) by PGHS-1 (bull seminal vesicles). The rate of prostanoid formation is expressed as a percentage of the control value (absence of H2O2). Role of the addition of catalase or superoxide dismutase (SOD) on the PG biosynthesis: 1078–1077 M H2O2 (arrow I), no effect of catalase or SOD; 1077–1076 M H2O2 (arrow II), no effect of catalase, increasing effect of SOD; 1076–1075 M H2O2 (arrow III), decreasing effect of catalase, no effect or slightly inhibiting effect of SOD; 1075–1072 M H2O2 (arrow IV), increasing effect of catalase and decreasing effect of SOD
Various physiological effects that are produced by hydrogen peroxide are thus mediated by cyclooxygenase activation and are diminished or abolished by NSAIDs, e.g. tracheal smooth muscle contractions in guinea-pig (Rhoden and Barnes 1989; Gao and Vanhoutte 1993) and human airway smooth muscle tone enhancement (Rabe et al 1995). Human leukocytes produce activator hydroperoxides in quantities sufficient to enhance prostaglandin synthesis in cells (Marshall and Lands 1986). The activators appeared to be partly H2O2 and partly a lipid hydroperoxide. These effects are completely inhibited by glutathione peroxidase (0.5 U/ml) and partially by catalase (500 U/ml) (Marshall et al 1987). Low levels of H2O2 enhance arachidonic acid-induced platelet aggregation by cyclooxygenase activation, which is inhibited by aspirin (Pratico et al 1991). H2O2 alone does not induce aggregation. H2O2 thus plays a role in the activation of cyclooxygenase by providing an adequate peroxide tone (Hecker et al 1991). The role of antioxidants on PGHS activity could thus be largely explained in this perspective of a regulation role of lipoperoxides. By lowering the lipoperoxide levels, antioxidants such as atocopherol (vitamin E) could inhibit PGHS activity. But the results reported in the literature are divergent, and the role of tocopherol on PGHS remains a matter of discussion. Panganamala et al (1977) found no effects of DL-a-tocopherol, although Seeger et al (1988) observed that inhibition was obtained using Trolox, a water-soluble derivative of tocopherol. However, strong antioxidants such as a-naphthol, guaiacol, NDGA and propyl gallate are inhibitors (Panganamala et al 1977).
Regulation by Cell and Plasma Factors In vivo Experiments Intravascular injection of high doses of arachidonate is normally necessary to obtain physiological events (fall of blood pressure and
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES
37
Figure 3.7 Role of peroxynitrite (ONOO7) in the formation of the p-cation radical on PGH synthase
myocardial symptoms) that appear to be linked to the production of prostanoids. However, heparinization, short-duration hypoxia and low amounts of plasmatic free haemoglobin increase the hypotensive effect of arachidonic acid injection, that hypotensive effect being accompanied by significant rises of plasma prostanoid concentration (Deby and Deby-Dupont 1981; Deby et al 1981). These observations demonstrate that control mechanisms of PGHS are normally present in vivo. A control is also exerted at the level of the availability of the precursors of prostanoids, the free fatty acids. In man, plasma levels of unesterified arachidonic acid are normally low (a few mg/ml), but important increases of this free fatty acid were measured in pathological and stress conditions (Deby-Dupont et al 1983).
Effects of Blood Components Plasma and cells of different mammalian species have inhibitory properties on PGHS, attributed mainly to albumin (Heinsohn et al 1987) but also to other proteins such as haptoglobin and lipoproteins (LDL). Inhibitors of PGHS are present in tissue extracts (e.g. stomach, placental extracts) but are often attributed to the cells present in these tissues. The inhibitory properties of plasma proteins still remain under discussion, as it seems difficult to explain an intracellular inhibition of PGHS by intravascular proteins that cannot enter the cells. An action on cell membrane receptors would be an explanation for these inhibitory phenomena. PGHS inhibitors are present in cells such as platelets, where the inhibitory capacity would be due to b-thromboglobulin. Numerous stimulating factors of PGHS are present in blood, one of them being free haemoglobin. Another important stimulating factor of blood is heparin, which acts by releasing arachidonic acid from its albumin-binding sites and by inhibiting GSH-peroxidase and LCAT activities. A stimulating effect on PGHS is also attributed to hypoxia, by way of plasmatic uric acid increase. The stimulating activity of this purine can be explained by its free radical scavenging properties (Deby et al 1981; Deby 1988).
Self-catalysed Inactivation: the Suicidal Character of PGH Synthase The phenomenon of autoinactivation was first described during the 1970s (Smith and Lands 1972; Egan et al 1976; Ohki et al 1979). One molecule of PGHS can only run a limited number of cycles, thus transforming a limited number of substrate molecules. This number was estimated, in conditions in vitro, at 103 moles PGH produced per molecule of PGHS (Marshall et al 1987). Studies performed with human endothelial cells overexpressing PGHS-1 allowed us to quantify the kinetics of PGHS-1 turnover and indicated that approximately 30% of PGHS-1 was degraded during each catalysis-induced autoinactivation, so that the extent of prostanoid synthesis would be governed by the level of PGHS-1 mass (Sanduja et al 1994). Egan et al (1976) proposed that the inactivating factors were oxygen-centred radicals. The free radical species are thus responsible for enzyme inactivation but, as the enzymatic activity of PGHS produces free radicals at different steps of its activity, this free radical production leads to the autoinactivation of the enzyme. This autoinactivation explains the role played by antilipoperoxidants and radical scavengers on prostanoid production, which were contradictorily described as inhibitors or stimulators of PGHS. The presence of free radical scavengers such as phenols (Egan et al 1976, 1978), lipoic acid (Marnett and Wilcox 1977), uric acid (Deby et al 1981) and antioxidants (tocopherol, NDGA) prolongs enzyme activity more or less considerably according to their efficiency (affinity coefficient for radicals, concentrations, etc.). Antilipoperoxidant and radical scavenging activities are often associated. A free radical scavenger may thus protect PGHS from autoinactivation while inactivating the lipoperoxides needed to trigger the enzymatic activity. However, the active site of PGHS being hydrophobic, its access is only possible to lipophilic molecules or to hydroxyl radicals (.OH) or small alkoxyl radicals (RO.). Lipophilic antioxidants will thus protect the enzyme from inactivation by free radicals, but will also reach the active site and affect the peroxide concentration, while hydrophilic antioxidants will only inactivate free radicals in the aqueous phase without altering the lipoperoxide activity in the lipophilic phase. Substances which exert a scavenging activity only against hydroxyl
38
BIOSYNTHESIS AND METABOLISM
and oxy-ferryl radicals, without inhibition of lipoperoxidation, thus protect PGHS and significantly enhance prostanoid synthesis. A typical example is given by uric acid, which traps .OH and oxoferryl radicals but is inefficient against peroxidation cycles (Deby 1988). Uric acid was demonstrated to enhance prostaglandin synthesis in vitro (Deby et al 1981) and in vivo (Bourgain et al 1982). A new explanation of the self-inactivation of the enzyme has recently been proposed, based on kinetic considerations: there is an unbalanced process which paralyses the activity of PGHS. But why does the self-inactivated enzyme remain definitively in this state? Some authors prefer the oxidative theory, where strong oxidants, such as free radicals and singlet oxygen, would destroy the active sites. But the main impact of oxidative agents seems to be exerted on the haem. Deep haem spectral changes were observed during suicidal inactivation (Tsai et al 1992). Recently, Wu et al (1999) explained the haem spectral changes: the Cd II is first converted in a faster step (0.5–2/s) into a new compound, intermediate III, which undergoes a subsequent slower (0.01–0.05/ s) transition to a terminal species. These results were obtained by the study of the kinetics of formation of intermediate I (Cd I) and intermediate II during reaction of PGHS. The ability of compounds to protect against hydroperoxide-induced inactivation correlates directly with their ability to act as reducing substrates. Hydroquinone, an excellent reducing substrate, protected against hydroperoxide-induced inactivation when present in less than threefold molar excess over hydroperoxide (Markey et al 1987). The presence of a highly efficient hydroperoxide-reducing activity appears absolutely essential for the protection of the cyclooxygenase capacity of PGHS. We may conclude that the mechanisms of control of PGHS are complex, and remain incompletely elucidated. This control is modulated in an interdependent manner by both the peroxide level and the supply of arachidonate. Moreover, it is clear that there are differences but also resemblances in the regulatory mechanisms that apply to PGHS-1 or to PGHS-2, this latter being induced within the context of inflammatory response, resulting in an exacerbated production of prostanoids. There is thus a need to control the two PGHSs separately, which can only be performed by acting on the activity of the enzyme without affecting the bioavailability of the prostanoid precursor (Fitzpatrick and Soberman 2001).
In experiments on macrophages, Masferrer et al (1994) observed that both PGHS-1 message and PGHS-1 protein levels were unaffected either by adrenalectomy or by dexamethasone administration, unlike PGHS-2. The expression of PGHS-1 appears to be stable, and mRNA for PGHS-1 is found in most cells at relatively stable concentrations, so that PGHS-1 is considered as a constitutive (‘‘housekeeping’’) enzyme supplying the low prostanoid levels indispensable for basal functions of the cells and the organisms (Smith et al 1996; Kulmacz 1998). PGHS1 and -2 are encoded by separate genes, located on different chromosomes. In contrast to PGHS-2, PGHS-1 expression is not induced in inflammatory situations. The two PGHSs are often coexpressed in the same cells but seem to function independently, channelling prostanoids to the nucleus and to the extracellular milieu, respectively. PGHS-1, like PGHS-2, is attached to the membrane, anchored to one leaflet of the lipid bilayer through the hydrophobic surfaces of amphipathic helices, and not through transmembrane motifs. PGHS-1 is especially inhibited by aspirin, acting via the inhibition of platelet thromboxane A2 formation, lowering the relative risk for mortality from cardiovascular disease. It thus seems that PGHS-1 activity is not importantly regulated at the level of its gene expression. PGHS-1 is constitutively present, expressed at constant levels throughout the cell cycle, in most tissues but not within all cells of a tissue. However, it is an oversimplification to consider that PGHS-1 is only constitutive, as PGHS-1 levels have been demonstrated to change during development, to be downregulated in endothelial cells and upregulated in mast cells under the influence of specific factors (Smith et al 1996). The role of glucocorticoids on PGHS-1 regulation remains a matter of discussion. These hormones have an inhibitory effect on PGI2 production in endothelial and mesangial cells, this effect being generally attributed to their activity on phospholipase A2 and thus to the availability of arachidonic acid. An effect of glucocorticoids on PGHS expression has been described only for COX-2 (Masferrer et al 1994; Minghetti et al 1999; Ferreira et al 1997). However, in 1999, it was reported that the glucocorticoids downregulated PGHS-1 expression in foetal pulmonary artery endothelial cells, as demonstrated by the decrease of COX-1 mRNA expression in these cells when they are treated by dexamethasone (Jun et al 1999). Glucocorticoids may downregulate COX-1 expression by acting on the glucocorticoid receptor and on gene transcription.
BIOSYNTHESIS OF PGHS-1 The two PGHSs serve separate physiological functions, as demonstrated by studies with knockout mice for either PHGS-1 or PGHS-2 (Langenbach et al 1999), and have segregated biosynthetic pathways (Smith et al 2000). PHGS-1 is expressed constitutively and assumes the regulated production of prostanoids that are needed for normal physiological functions, such as haemostasis. The subcellular location of PGHS-1 appears similar to that of PGHS-2 although the latter appears more concentrated in the nuclear envelope (Morita et al 1995a). PGHS-1 can be detected in most cells, but is preferentially expressed at high levels in endothelial cells, monocytes, platelets, renal collecting tubules and seminal vesicles (Smith and DeWitt 1996). The gene for PGHS-1 is approximately 22 kilobase pairs and contains 11 exons (Kraemer et al 1992). Typical of developmentally regulated ‘‘housekeeping’’ genes, the PGHS-1 gene lacks a TATA box. PGHS-1 expression does not vary greatly in adult animals, so that study on its transcriptional regulation is difficult. Virtually nothing is known about the details of the regulation of PGHS-1 gene expression, contrary to the large amount of data that has accumulated for PGHS-2 expression (Smith and Dewitt 1996; Smith et al 1996, 2000).
PHARMACOLOGICAL REGULATION A major method of PGHS control is the use of NSAIDs, which act through their ability to inhibit cyclooxygenase activity, reducing the synthesis of prostaglandins, prostacyclins and thromboxane (Vane 1987). Acetyl salicylic acid (aspirin) remains the model of these NSAIDs. The mechanism of acetylation of PGHS active site has been the subject of numerous studies (DeWitt et al 1990; Shimokawa and Smith 1992; Loll et al 1995). At low doses (40 mg/day), aspirin selectively inhibits production of thromboxane A2 without affecting prostacyclin. This is explained by an irreversible acetylation of PGHS and the impossibility for platelets to regenerate PGHS, while the systemic plasma aspirin concentration is likely to be too low to affect prostacyclin synthesis. However, the inhibition activity of most of the NSAIDs does not distinguish between PGHS-1 and PGHS-2. Attempts are under way to produce NSAIDs with a specific inhibition activity on PGHS-2 (Smith et al 2000; Fitzpatrick and Soberman 2001). Callan et al (1996) presented a kinetic model for the comparison of NSAID efficiencies.
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES CONCLUSIONS Since the first description of its activity in the beginning of the 1960s, PGHS has been isolated to a high degree of purity. Its amino acid sequence has been elucidated and its gene identified. PGHS-1 could be distinguished from its parent enzyme, PGHS-2, and the specific characteristics of PGHS-1 as a ‘‘housekeeping enzyme’’, not induced by an inflammatory response, could be defined. Because of its isolation and in vitro studies of pure PGHS, the molecular mechanisms of its cyclooxygenasic and peroxidasic activities could be described, underlining its connections with haemic peroxidases. PGHS belongs to the family of ‘‘radical proteins’’ and is the source of an important free radical production, which leads to an autoinactivation of the enzyme. However, a major problem remains the pharmacological regulation of its activity. As the two PGHS enzymes are free radicalgenerating enzymes which may be responsible for radical chain reactions and potentially dangerous co-oxidations of neighbouring molecules, it would be useful to inhibit the two enzymes separately by free radical scavengers. However, this remains an unsolved problem. REFERENCES Badwey JA and Karnovsky ML (1980) Active oxygen species and the functions of phagocytic leukocytes. Annu Rev Biochem, 149, 695–726. Bakovic M and Dunford HB (1993) Kinetics of the oxidation of pcoumaric acid by prostaglandin H synthase and hydrogen peroxide. Biochemistry, 32, 833–840. Bakovic M and Dunford HB (1994) Intimate relation between cyclooxygenase and peroxidase activities of prostaglandin H synthase. Peroxidase reaction of ferulic acid and its influence on the reaction of arachidonic acid. Biochemistry, 33, 6475–6482. Bakovic M and Dunford HB (1996a) Reactions of prostaglandin endoperoxide synthase and its compound I with hydroperoxides. J Biol Chem, 271, 2048–2056. Bakovic M and Dunford HB (1996b) pH and temperature dependence of the rate of compound I formation from the reaction of prostaglandin endoperoxide synthase with hydrogen peroxide. Biochem Cell Biol, 74, 117–124. Barnett J, Chow J, Ives D et al (1994) Purification, characterization and selective inhibition of human prostaglandin G/H synthase 1 and 2 expressed in the baculovirus system. Biochim Biophys Acta, 1209, 130– 139. Baumann J, von Bruchhausen F and Wurm G (1983) Decreasing inhibitory potency of prostaglandin synthetase inhibitors during their cooxidative metabolism. Studies on amino-phenols, pyrazolon derivatives and 1,3-diphenylisobenzofuran. Pharmacology, 27, 267–280. Beckman JS, Beckman TW, Chen J et al (1990) Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci USA, 87, 1620–1624. Bergstro¨m S, Danielsson H and Samuelsson B (1964) The enzymatic formation of prostaglandin E from arachidonic acid. Biochim Biophys Acta, 90, 207–219. Bhattacharyya DK, Lecomte M, Rieke CJ et al (1996) Involvement of arginine 120, glutamate 524, and tyrosine 355 in the binding of arachidonate and 2-phenylpropionic acid inhibitors to the cyclooxygenase active site of ovine prostaglandin endoperoxide H synthase-1. J Biol Chem, 271, 2179–2184. Bourgain RH, Deby C, Deby-Dupont G and Andries R (1982) Enhancement of arterial thromboformation by uric acid, a free radical scavenger. Biochem Pharmacol, 31, 3011–3013. Cadenas E, Sies H, Nastainczyk W and Ullrich V (1983) Singlet oxygen formation detected by low-level chemiluminescence during enzymatic reduction of prostaglandin G2 to H2. Hoppe Seylers Z Physiol Chem, 364, 519–528. Callan OH, On-Yee S and Swinney DC (1996) The kinetic factors that determine the affinity and selectivity for slow binding inhibition of human prostaglandin H synthase 1 and 2 by indomethacin and flurbiprofen. J Biol Chem, 271, 3548–3554.
39
Capdevila JH, Morrow JD, Belosludtsev YY et al (1995) The catalytic outcomes of the constitutive and the mitogen-inducible isoforms of prostaglandin H2 synthase are markedly affected by glutathione and glutathione peroxidase(s). Biochemistry, 34, 3325–3337. Capdevila JH, Wei S, Helvig C et al (1996) The highly stereoselective oxidation of polyunsaturated fatty acids by cytochrome P450BM-3. J Biol Chem, 271, 22263–22271. Chen W, Pawelek TR and Kulmacz RJ (1999) Hydroperoxide dependence and cooperative cyclooxygenase kinetics in prostaglandin H synthase-1 and -2. J Biol Chem, 274, 20301–20306. Christ-Hazelhof E and Nugteren DH (1979) Purification and characterisation of prostaglandin endoperoxide D-isomerase, a cytoplasmic, glutathione-requiring enzyme. Biochim Biophys Acta, 572, 43–51. Christ-Hazelhof E, Nugteren DH and Van Dorp DA (1976) Conversions of prostaglandin endoperoxides by glutathione-S-transferases and serum albumins. Biochim Biophys Acta, 450, 450–461. Corbett JA, Kwon G, Turk J and McDaniel ML (1993 ) IL-1b induces the co-expression of both nitric oxide synthase and cyclooxygenase by islets of Langerhans: activation of cyclooxygenase by nitric oxide. Biochemistry, 32, 13767–13770. Curtis JF, Reddy NG, Mason RP et al (1996) Nitric oxide: a prostaglandin H synthase 1 and 2 reducing co-substrate that does not stimulate cyclooxygenase activity or prostaglandin H synthase expression in murine macrophages. Arch Biochem Biophys, 335, 369–376. Deby C (1988) Metabolism of polyunsaturated fatty acids, precursors of eicosanoids. In Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids, Curtis-Prior PB (ed.). Churchill Livingstone, Edinburgh, 11–36. Deby C and Deby-Dupont G (1980) Oxygen species in prostaglandin biosynthesis in vitro and in vivo. In Biological and Clinical Aspects of Superoxide and Superoxide Dismutase, Bannister WH, Bannister JV (eds). Elsevier, Amsterdam, 84–97. Deby C and Deby-Dupont G (1981) Natural factors modulating the intervention of activated oxygen in the biosynthesis of prostanoids. Clin Respir Physiol, 17, 129–139. Deby C, Deby-Dupont G, Noe¨l FX and Lavergne L (1981) In vitro and in vivo arachidonic conversions into biologically active derivatives are enhanced by uric acid. Biochem Pharmacol, 30, 2243–2249. Deby-Dupont G, Ducarne H, De Landsheere C et al (1983) Intense rises of unesterified arachidonate plasma levels in stressed humans. Biomed Pharmacother, 37, 386–391. DeGray JA, Lassmann G, Curtis JF et al (1988) Higher oxidation states of prostaglandin H synthase. Rapid electronic spectroscopy detected two spectral intermediates during the peroxidase reaction with prostaglandin G2. Eur J Biochem, 171, 321–328. DeGray JA, Lassmann G, Curtis JF et al (1992) Spectral analysis of the protein-derived tyrosyl radicals from prostaglandin H synthase. J Biol Chem, 267, 23583–23588. DeWitt DL, el-Harith EA, Kraemer SA et al (1990) The aspirin and hemebinding sites of ovine and murine prostaglandin endoperoxide synthases. J Biol Chem, 265, 5192–5198. Dietz R, Nastainczyk W and Ruf HH (1988) Higher oxidation states of prostaglandin H synthase. Rapid electronic spectroscopy detected two spectral intermediates during the peroxidase reaction with prostaglandin G2. Eur J Biochem, 171, 321–328. Dix TA, Buck JR and Marnett LJ (1986) Hydroperoxide-dependent epoxidation of 3,4-dihydroxy-3,4-dihydrobenzo-a-anthracene by ram seminal vesicle microsomes and by hematin. Biochem Biophys Res Commun, 140, 181–187. Douglas GJ, Flower RJ, Parente L and Perretti M (1992) Peptide 204-212 of lipocortin 5 inhibits the generation of a prostacyclin-like factor from rat aorta preparations in vitro. Prostaglandins, 44, 381–388. Egan RW, Paxton J and Kuehl FA Jr (1976) Mechanism for irreversible self-deactivation of prostaglandin synthetase. J Biol Chem, 251, 7329– 7335. Egan RW, Gale PH, Beveridge GC et al (1978) Radical scavenging as the mechanism for stimulation of prostaglandin cyclooxygenase and depression of inflammation by lipoic acid and sodium iodide. Prostaglandins, 16, 861–869. Eling TE, Thompson DC, Foureman GL, Curtis JF et al (1990) Prostaglandin H synthase and xenobiotic oxidation. Annu Rev Pharmacol Toxicol, 30, 1–45.
40
BIOSYNTHESIS AND METABOLISM
Ferreira SH, Cunha FQ, Lorenzetti BB et al (1997) Role of lipocortin-1 in the anti-hyperalgesic actions of dexamethasone. Br J Pharmacol, 121, 883–888. Fitzpatrick FA and Soberman R (2001) Regulated formation of eicosanoids. J Clin Invest, 107, 1347–1351. Franchi AM, Chaud M, Rettori V et al (1994) Role of nitric oxide in eicosanoid synthesis and uterine motility in estrogen-treated rat uteri. Proc Natl Acad Sci USA, 91, 539–543. Freyaldenhoven MA, Lloyd RV and Samokyszyn VM (1996) Prostaglandin H synthase-catalyzed oxidation of all-trans- and 13cis-retinoic acid to carbon-centered and peroxyl radical intermediates. Chem Res Toxicol, 9, 677–681. Freyberger A and Degen GH (1989) Studies on the stoichiometry of estrogen oxidation catalyzed by purified prostaglandin-H-synthase holoenzyme. J Steroid Biochem, 33, 473–481. Gao Y and Vanhoutte PM (1993) Products of cyclooxygenase mediate the responses of the guinea pig trachea to hydrogen peroxide. J Appl Physiol, 74, 2105–2111. Giovannucci E (1999) The prevention of colorectal cancer by aspirin use. Biomed Pharmacother, 53, 303–308. Gonchar M, Sergeeva M, Mevkh A and Varfolomeyev S (1999) Kinetics of prostanoid synthesis by macrophages is regulated by arachidonic acid sources. Eur J Biochem, 265, 779–787. Goodwin DC, Gunther MR, Hsi LC et al (1998) Nitric oxide trapping of tyrosyl radicals generated during prostaglandin endoperoxide synthase turnover. Detection of the radical derivative of tyrosine 385. J Biol Chem, 273, 8903–8909. Goodwin DC, Landino LM, and Marnett LJ (1999) Effects of nitric oxidederived species on prostaglandin endoperoxide synthase and prostaglandin biosynthesis. FASEB J, 13, 1121–1136. Griendling KK, Sorescu D and Ushio-Fukai M (2000) NAD(P)H oxidase. Role in cardiovascular biology and disease. Circ Res, 86, 494–501. Gunther MR, Hsi LC, Curtis JF et al (1997) Nitric oxide trapping of the tyrosyl radical of prostaglandin H synthase-2 leads to tyrosine iminoxyl radical and nitrotyrosine formation. J Biol Chem, 272, 17086–17090. Hajjar DP, Lander HM, Pearce SFA et al (1995) Nitric oxide enhances prostaglandin-H synthase-1 activity by a heme-independent mechanism: evidence implicating nitrosothiols. J Am Chem Soc, 117, 3340–3346. Hamberg M and Fredholm BB (1976) Isomerization of prostaglandin H2 into prostaglandin D2 in the presence of serum albumin. Biochim Biophys Acta, 431, 189–193. Hamberg M and Samuelsson B (1967) On the mechanism of the biosynthesis of prostaglandins E-1 and F-1a. J Biol Chem, 242, 5336–5343. Hecker G, Utz J, Kupferschmidt RJ and Ullrich V (1991) Low levels of hydrogen peroxide enhance platelet aggregation by cyclooxygenase activation. Eicosanoids, 4, 107–113. Heinsohn C, Polgar P, Fishman J and Taylor L (1987) The effect of bovine serum albumin on the synthesis of prostaglandin and incorporation of [3H]acetate into platelet-activating factor. Arch Biochem Biophys, 257, 251–258. Hemler M and Lands WE (1976) Purification of the cyclooxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyme. J Biol Chem, 251, 5575–5579. Hemler ME and Lands WE (1980) Evidence for a peroxide-initiated free radical mechanism of prostaglandin biosynthesis. J Biol Chem, 255, 6253–6261. Hemler ME, Graff G and Lands WE (1978) Accelerative autoactivation of prostaglandin biosynthesis by PGG2. Biochem Biophys Res Commun, 85, 1325–1331. Hemler ME, Cook HW and Lands WE (1979) Prostaglandin biosynthesis can be triggered by lipid peroxides. Arch Biochem Biophys, 193, 340– 345. Hernandez M, Bayon Y, Sanchez Crespo M and Nieto ML (1999) Signaling mechanisms involved in the activation of arachidonic acid metabolism in human astrocytoma cells by tumor necrosis factor-a: phosphorylation of cytosolic phospholipase A2 and transactivation of cyclooxygenase-2. J Neurochem, 73, 1641–1649. Herschman HR (1996) Prostaglandin synthase 2. Biochim Biophys Acta, 1299, 125–140.
Huberman E, Mager R and Sachs L (1976) Mutagenesis and transformation of normal cells by chemical carcinogens. Nature, 264, 360–361. Jun SS, Chen Z, Pace MC and Shaul PW (1999) Glucocorticoids downregulate cyclooxygenase-1 gene expression and prostacyclin synthesis in fetal pulmonary artery endothelium. Circ Res, 84, 193–200. Kanner J, Harel S and Granit R (1992) Nitric oxide, an inhibitor of lipid oxidation by lipoxygenase, cyclooxygenase and hemoglobin. Lipids, 27, 46–49. Karthein R, Nastainczyk W and Ruf HH (1987) EPR study of ferric native prostaglandin H synthase and its ferrous NO derivative. Eur J Biochem, 166, 173–180. Karthein R, Dietz R, Nastainczyk W and Ruf HH (1988) Higher oxidation states of prostaglandin H synthase. EPR study of a transient tyrosyl radical in the enzyme during the peroxidase reaction. Eur J Biochem, 171, 313–320. Kelner MJ and Uglik SF (1994) Mechanism of prostaglandin E2 release and increase in PGH2/PGE2 isomerase activity by PDGF: involvement of nitric oxide. Arch Biochem Biophys, 312, 240–243. Kraemer SA, Meade SA and DeWitt DL (1992) Prostaglandin endoperoxide synthase gene structure: identification of the transcriptional start site and 5’-flanking regulatory sequences. Arch Biochem Biophys, 293, 391–400. Kubow S and Wells PG (1989) In vitro bioactivation of phenytoin to a reactive free radical intermediate by prostaglandin synthetase, horseradish peroxidase, and thyroid peroxidase. Mol Pharmacol, 35, 504–511. Kukreja RC, Kontos HA, Hess ML and Ellis EF (1986) PGH synthase and lipoxygenase generate superoxide in the presence of NADH or NADPH. Circ Res, 59, 612–619. Kulmacz RJ (1998) Cellular regulation of prostaglandin H synthase catalysis. FEBS Lett, 430, 154–157. Kulmacz RJ and Lands WE (1984) Prostaglandin H synthase. Stoichiometry of heme co-factor. J Biol Chem, 259, 6358–6363. Kulmacz RJ and Lands WE (1985a) Stoichiometry and kinetics of the interaction of prostaglandin H synthase with anti-inflammatory agents. J Biol Chem, 260, 12572–12578. Kulmacz RJ and Lands WE (1985b) Quantitative similarities in the several actions of cyanide on prostaglandin H synthase. Prostaglandins, 29, 175–190. Kulmacz RJ and Wang LH (1995) Comparison of hydroperoxide initiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J Biol Chem, 270, 24019–24023. Kulmacz RJ, Pendleton RB and Lands WE (1994) Interaction between peroxidase and cyclooxygenase activities in prostaglandinendoperoxide synthase. Interpretation of reaction kinetics. J Biol Chem, 269, 5527–5536. Langenbach R, Loftin C, Lee C and Tiano H (1999) Cyclooxygenase knockout mice: models for elucidating isoform-specific functions. Biochem Pharmacol, 58, 1237–1246. Lambeir AM, Markey CM, Dunford HB and Marnett LJ (1985) Spectral properties of the higher oxidation states of prostaglandin H synthase. J Biol Chem, 260, 14894–14896. Landino LM, Crews BC, Timmons MD et al (1996) Peroxynitrite, the coupling product of nitric oxide and superoxide, activates prostaglandin biosynthesis. Proc Natl Acad Sci USA, 93, 15069–15074. Li H, Schut HA, Conran P et al (1999) Prevention by aspirin and its combination with a-difluoromethyl-ornithine of azoxymethaneinduced tumors, aberrant crypt foci and prostaglandin E2 levels in rat colon. Carcinogenesis, 20, 425–430. Loll PJ, Picot D and Garavito RM (1995) The structural basis of aspirin activity inferred from the crystal structure of inactivated prostaglandin H2 synthase. Nat Struct Biol, 2, 637–643. Mancini JA, Riendeau D, Falgueyret JP et al (1995) Arginine 120 of prostaglandin G/H synthase-1 is required for the inhibition by nonsteroidal antiinflammatory drugs containing a carboxylic acid moiety. J Biol Chem, 270, 29372–29377. Margalit A, Hauser SD, Zweifel BS et al (1998) Regulation of prostaglandin biosynthesis in vivo by glutathione. Am J Physiol, 274, R294–R302. Markey CM, Alward A, Weller PE and Marnett LJ (1987) Quantitative studies of hydroperoxide reduction by prostaglandin H synthase. Reducing substrate specificity and the relationship of peroxidase to cyclooxygenase activities. J Biol Chem, 262, 6266–6279.
PGH SYNTHASE-1 ACTIVITY AND RADICAL STATES Marnett LJ and Wilcox CL (1977) Stimulation of prostaglandin biosynthesis by lipoic acid. Biochim Biophys Acta, 487, 222–230. Marnett LJ, Wlodawer P and Samuelsson B (1974) Light emission during the action of prostaglandin synthetase. Biochem Biophys Res Commun, 60, 1286–1294. Marnett LJ, Wlodawer P and Samuelsson B (1975) Co-oxygenation of organic substrates by the prostaglandin synthetase of sheep vesicular gland. J Biol Chem, 250, 8510–8517. Marnett LJ, Rowlinson SW, Goodwin DC et al (1999) Arachidonic acid oxygenation by COX-1 and COX-2. Mechanisms of catalysis and inhibition. J Biol Chem, 274, 22903–22906. Marshall PJ and Kulmacz RJ (1988) Prostaglandin H synthase: distinct binding sites for cyclooxygenase and peroxidase substrates. Arch Biochem Biophys, 266, 162–170. Marshall PJ and Lands WE (1986) In vitro formation of activators for prostaglandin synthesis by neutrophils and macrophages from humans and guinea pigs. J Lab Clin Med, 108, 525–534. Marshall PJ, Kulmacz RJ and Lands WE (1987) Constraints on prostaglandin biosynthesis in tissues. J Biol Chem, 262, 3510–3517. Masferrer JL and Seibert K (1994) Regulation of prostaglandin synthesis by glucocorticoids. Receptor, 4, 25–30. Masferrer JL, Reddy ST, Zweifel BS et al (1994) In vivo glucocorticoids regulate cyclooxygenase-2 but not cyclooxygenase-1 in peritoneal macrophages. J Pharmacol Exp Ther, 270, 1340–1344. Mason RP, Kalyanaraman B, Tainer BE and Eling TE (1980) A carboncentered free radical intermediate in the prostaglandin synthetase oxidation of arachidonic acid. Spin trapping and oxygen uptake studies. J Biol Chem, 255, 5019–5022. Meyer JM, Holland JA, Ziegler LM et al (1999) Identification of a functional leukocyte-type NADPH oxidase in human endothelial cells: a potential atherogenic souce of reactive oxygen species. Endothelium, 7, 11–22. Minghetti L, Nicolini A, Polazzi E et al (1999) Downregulation of microglial cyclo-oxygenase-2 and inducible nitric oxide synthase expression by lipocortin 1. Br J Pharmacol, 126, 1307–1314. Miyamoto T, Yamamoto S and Hayaishi O (1974) Prostaglandin synthetase system resolution into oxygenase and isomerase components. Proc Natl Acad Sci USA, 71, 3645–3648. Miyamoto T, Ogino N, Yamamoto S and Hayaishi O (1976) Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J Biol Chem, 251, 2629–2636. Morita I, Schindler M, Regier MK et al (1995a) Different intracellular locations for prostaglandin endoperoxide H synthase-1 and -2. J Biol Chem, 270, 10902–10908. Morita I, Smith WL, DeWitt DL and Schindler M (1995b) Expressionactivity profiles of cells transfected with prostaglandin endoperoxide H synthase measured by quantitative fluorescence microscopy. Biochemistry, 34, 7194–7199. Murakami M, Kambe T, Shimbara S and Kudo I (1999) Functional coupling between various phospholipase A2s and cyclooxygenases in immediate and delayed prostanoid biosynthetic pathways. J Biol Chem, 274, 3103–3115. Narayanan R and Harrington MG (1980) Hydroperoxyeicosatetraenoic acid: an activator of prostaglandin synthase activity. Biochem Soc Trans, 8, 449–450. Nastainczyk W, Schuhn D and Ullrich V (1984) Spectral intermediates of prostaglandin hydroperoxidase. Eur J Biochem, 144, 381–385. Nugteren DH and Hazelhof E (1973) Isolation and properties of intermediates in prostaglandin biosynthesis. Biochim Biophys Acta, 326, 448–461. Nugteren DH, Buytenhek M, Christ-Hazelhof E et al (1981) Enzymes involved in the conversion of endoperoxides. Prog Lipid Res, 20, 169– 172. O’Brien PJ and Rahimtula A (1976) The possible involvement of a peroxidase in prostaglandin biosynthesis. Biochem Biophys Res Commun, 170, 832–838. Ogino N, Ohki S, Yamamoto S and Hayaishi O (1978) Prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. Inactivation and activation by heme and other metalloporphyrins. J Biol Chem, 253, 5061–5068. Ohki S, Ogino N, Yamamoto S and Hayaishi O (1979) Prostaglandin hydroperoxidase, an integral part of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J Biol Chem, 254, 829–836.
41
Panganamala RV, Miller JS, Gwebu ET et al (1977) Differential inhibitory effects of vitamin E and other antioxidants on prostaglandin synthetase, platelet aggregation and lipoxidase. Prostaglandins, 14, 261–271. Picot D and Garavito RM (1994) Prostaglandin H synthase: implications for membrane structure. FEBS Lett, 346, 21–25. Picot D, Loll PJ and Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature, 367, 243– 249. Polgar P and Taylor L (1980) Stimulation of prostaglandin synthesis by ascorbic acid via hydrogen peroxide formation. Prostaglandins, 19, 693–700. Prajgrod G and Danon A (1992) Biphasic regulation by dexamethasone of IL-1- and LPS-stimulated endothelial prostacyclin production. Agents Actions, 36, 70–76. Pratico D, Iuliano L, Ghiselli A et al (1991) Hydrogen peroxide as trigger of platelet aggregation. Haemostasis, 21, 169–174. Rabe KF, Dent G and Magnussen H (1995) Hydrogen peroxide contracts human airways in vitro: role of epithelium. Am J Physiol, 269, L332– L338. Rampart M, Bult H and Herman AG (1981) Effect of complement activation on the biosynthesis of prostacyclin (PGI2) by rabbit peritoneum in vitro. Arch Int Pharmacodyn Ther, 253, 327–329. Reddy SR and Herschman HR (1994) Ligand-induced prostaglandin synthesis requires expression of the TIS10/PGS-2 prostaglandin synthase gene in murine fibroblasts and macrophages. J Biol Chem, 269, 15473–15480. Rhoden KJ and Barnes PJ (1989) Effect of hydrogen peroxide on guineapig tracheal smooth muscle in vitro: role of cyclo-oxygenase and airway epithelium. Br J Pharmacol, 98, 325–330. Riutta A, Alanko J, Mucha I and Vapaatalo H (1994) Effects of Trolox C and SIN-1 on arachidonic acid metabolism and on cyclic GMP formation in leukocytes. Eur J Pharmacol, 262, 217–222. Roth GJ, Stanford N and Majerus PW (1975) Acetylation of prostaglandin synthase by aspirin. Proc Natl Acad Sci USA, 72, 3073–3076. Roth GJ, Siok CJ and Ozols J (1980) Structural characteristics of prostaglandin synthetase from sheep vesicular gland. J Biol Chem, 255, 1301–1304. Roth GJ, Machuga ET and Strittmatter P (1981) The heme-binding properties of prostaglandin synthetase from sheep vesicular gland. J Biol Chem, 256, 10018–10022. Salvemini D, Misko TP, Masferrer JL et al (1993) Nitric oxide activates cyclooxygenase enzymes. Proc Natl Acad Sci USA, 90, 7240–7244. Salvemini D, Manning PT, Zweifel BS et al (1995a) Dual inhibition of nitric oxide and prostaglandin production contributes to the antiinflammatory properties of nitric oxide synthase inhibitors. J Clin Invest, 96, 301–308. Salvemini D, Settle SL, Masferrer JL et al (1995b) Regulation of prostaglandin production by nitric oxide; an in vivo analysis. Br J Pharmacol, 114, 1171–1178. Samokyszyn VM, Chen T, Maddipati KR et al (1995) Free radical oxidation of (E)-retinoic acid by prostaglandin H synthase. Chem Res Toxicol, 8, 807–815. Samuelsson B (1972) Biosynthesis of prostaglandins. Fed Proc, 31, 1442– 1450. Sanduja SK, Tsai AL, Matijevic-Aleksic N and Wu KK (1994) Kinetics of prostacyclin synthesis in PGHS-1-overexpressed endothelial cells. Am J Physiol, 267, C1459–C1466. Schilderman PA, van Maanen JM, Smeets EJ et al (1993) Oxygen radical formation during prostaglandin H synthase-mediated biotransformation of butylated hydroxyanisole. Carcinogenesis, 14, 347–353. Schreiber J, Mason RP and Eling TE (1986) Carbon-centered free radical intermediates in the hematin- and ram seminal vesicle-catalyzed decomposition of fatty acid hydroperoxides. Arch Biochem Biophys, 251, 17–24. Seeger W, Moser U and Roka L (1988) Effects of a-tocopherol, its carboxylic acid chromane compound and two novel antioxidant isoflavanones on prostaglandin H synthase activity and autodeactivation. N S Arch Pharmacol, 338, 74–81. Seibert K, Zhang Y, Leahy K et al (1994) Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proc Natl Acad Sci USA, 91, 12013–12017.
42
BIOSYNTHESIS AND METABOLISM
Shimokawa T and Smith WL (1992) Prostaglandin endoperoxide synthase. The aspirin acetylation region. J Biol Chem, 267, 12387– 12392. Shimokawa T, Kulmacz RJ, DeWitt DL and Smith WL (1990) Tyrosine 385 of prostaglandin endoperoxide synthase is required for cyclooxygenase catalysis. J Biol Chem, 265, 20073–20076. Smith WL and Dewitt DL (1996) Prostaglandin endoperoxide H synthases-1 and -2. Adv Immunol, 62, 167–215. Smith WL and Lands WEM (1972) Oxygenation of polyunsaturated fatty acids during prostaglandin biosynthesis by sheep vesicular gland. Biochemistry, 11, 3276–3285. Smith WL, Eling TE, Kulmacz RJ et al (1992) Tyrosyl radicals and their role in hydroperoxide-dependent activation and inactivation of prostaglandin endoperoxide synthase. Biochemistry, 31, 3–7. Smith WL, Garavito RM and DeWitt DL (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem, 271, 33157–33160. Smith WL, DeWitt DL and Garavito RM (2000) Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem, 69, 145– 182. Swinney DC, Mak AY, Barnett J and Ramesha CS (1997) Differential allosteric regulation of prostaglandin H synthase 1 and 2 by arachidonic acid. J Biol Chem, 272, 12393–12398. Tsai AL, Palmer G and Kulmacz RJ (1992) Prostaglandin H synthase. Kinetics of tyrosyl radical formation and of cyclooxygenase catalysis. J Biol Chem, 267, 17753–17759. Tsai AL, Hsi LC, Kulmacz RJ et al (1994a) Characterization of the tyrosyl radicals in ovine prostaglandin H synthase-1 by isotope replacement and site-directed mutagenesis. J Biol Chem, 269, 5085–5091. Tsai AL, Wei C, Kulmacz RJ (1994b) Interaction between nitric oxide and prostaglandin H synthase. Arch Biochem Biophys, 313, 367–372. Tsai AL,Wu G and Kulmacz RJ (1997) Stoichiometry of the interaction of prostaglandin H synthase with substrates. Biochemistry, 36, 13085– 13094. Tsai AL, Palmer G, Xiao G et al (1998) Structural characterization of arachidonyl radicals formed by prostaglandin H synthase-2 and
prostaglandin H synthase-1 reconstituted with mangano protoporphyrin IX. J Biol Chem, 273, 3888–3894. Tsai AL, Wu G, Palmer G et al (1999) Rapid kinetics of tyrosyl radical formation and heme redox state changes in prostaglandin H synthase-1 and -2. J Biol Chem, 274, 21695–21700. van der Ouderaa FJ, Buytenhek M, Nugteren DH and Van Dorp DA (1977) Purification and characterization of prostaglandin endoperoxide synthetase from sheep vesicular glands. Biochim Biophys Acta, 487, 315–331. van der Ouderaa FJ, Buytenhek M and van Dorp DA (1980) Characterization of prostaglandin H2 synthetase. Adv Prostaglandin Thromboxane Res, 6, 139–144. Van Dorp DA, Beerthuis RK, Nugteren DH and Vonkeman H (1964) Enzymatic conversion of all-cis-polyunsaturated fatty acids into prostaglandins. Nature, 203, 839–841. Vane J (1987) The evolution of non-steroidal antiinflammatory drugs and their mechanisms of action. Drugs, 33 (suppl 1), 18–27. Wei C, Kulmacz RJ and Tsai AL (1995) Comparison of branched-chain and tightly coupled reaction mechanisms for prostaglandin H synthase. Biochemistry, 34, 8499–8512. Wu G, Wei C, Kulmacz RJ et al (1999) A mechanistic study of selfinactivation of the peroxidase activity in prostaglandin H synthase-1. J Biol Chem, 274, 9231–9237. Xie W and Herschman HR (1995) v-src Induces prostaglandin synthase 2 gene expression by activation of the c-Jun N-terminal kinase and the c-Jun. J Biol Chem, 270, 27622–27628. Xiao GS, Chen W and Kulmacz RJ (1998) Comparisons of structural stabilities of prostaglandin H synthase-1 and-2. J Biol Chem, 273, 6801–6811. Yokoyama C and Tanabe T (1989) Cloning of human gene encoding prostaglandin endoperoxide synthase and primary structure of the enzyme. Biochem Biophys Res Commun, 165, 888–894. Zenser TV, Lakshmi VM, Hsu FF and Davis BB (1999) Peroxygenase metabolism of N-acetylbenzidine by prostaglandin H synthase. Formation of an N-hydroxylamine. J Biol Chem, 274, 14850–14856.
4 Regulation and Function of Prostaglandin Synthase 2/Cyclooxygenase II Harvey R. Herschman David Geffen School of Medicine at UCLA, Los Angeles, CA, USA
The prostaglandins (PGs) play central roles in acute and chronic inflammatory responses. Pharmacological inhibition studies of prostanoid synthesis have provided enormous insight into the physiological and pathophysiological roles of the prostanoids. Vane (1971) reported that aspirin and other nonsteroidal antiinflammatory drugs (NSAIDs) exert their biological effects by blocking prostaglandin synthesis. This landmark experiment implicated prostanoids in thermoregulation, pain perception and inflammatory responses, because of the antipyretic, analgesic and antiinflammatory effects of NSAIDs, and in blood clotting, because of a major aspirin side effect— increased clotting time. Prostanoids and the leukotrienes are derived from C20 fatty acids, principally arachidonic acid. Following appropriate stimulation, cells activate a phospholipase which releases free arachidonic acid from membrane phosphoglycerolipid stores. Prostaglandin synthase, or cyclooxygenase, converts arachidonic acid to prostaglandin H2 (PGH2), the common precursor to all prostanoids. Prostaglandin synthase catalyses two concerted reactions. Initially, arachidonic acid is cyclized to PGG2 by a bis-oxygenation. PGG2 is converted by a hydroperoxidase reaction, at a distinct enzyme site, to PGH2. PGH2 serves as substrate for production of the PGE2, PGF2a, PGD2, PGI2, TXA2, etc. Cell-type-specific prostaglandin production is the consequence of cell-type-specific expression of distinct prostaglandin synthases—PGE2 synthase, PGD2 synthase, etc.
THE DOGMA WITH RESPECT TO INDUCED PROSTAGLANDIN SYNTHESIS Until the mid-1980s it was thought that: (a) the rate-limiting step in prostaglandin synthesis following an inflammatory insult is phospholipase activation to release arachidonic acid from membrane stores, and (b) constitutive cyclooxygenase in tissues is sufficient to convert released arachidonic acid to PGH2. Whiteley and Needleman (1984) demonstrated that PGE2 production induced with conditioned medium in human fibroblasts could be blocked by either cycloheximide or actinomycin D, suggesting that protein translation and mRNA transcription were necessary. Moreover, an increase in microsomal cyclooxygenase Vmax suggested that ligand stimulation induced an increase in cyclooxygenase activity. Similarly, it was found that platelet derived growth factor (PDGF)-induced PGE2 accumulation in murine fibroblasts was blocked by cycloheximide (Habenicht et al 1985) and epidermal growth factor (EGF)-induced PGE2 production in murine osteoblast-like MC3T3E1 cells was blocked by cycloheximide and actinomycin D (Yokota et al 1986). Moreover, if constitutive cyclooxygenase in 3T3 cells is inactivated by aspirin treatment, PDGF-induced PGE2 synthesis is hardly affected, suggesting that pre-existing cyclooxygenase does not play a significant role in mitogen-induced prostaglandin production (Habenicht et al 1985).
EVIDENCE FOR INCREASED CYCLOOXYGENASE EXPRESSION IN RESPONSE TO LIGAND STIMULATION PURIFICATION, CHARACTERIZATION AND CLONING OF SHEEP SEMINAL VESICLE CYCLOOXYGENASE Sheep seminal vesicles were the primary source for cyclooxygenase purification and biochemical characterization (Hemler et al 1976; Miyamoto et al 1976). Cyclooxygenase, a homodimer of two 70 kDa subunits, is found in the endoplasmic reticulum and nuclear membrane. Oligonucleotide probes, prepared from amino acid sequence data, were used to clone sheep seminal vesicle cyclooxygenase (DeWitt et al 1988; Merlie et al 1988; Yokayama et al 1988). Ovine cyclooxygenase encodes a 576 amino acid protein. Human and murine sheep seminal vesicle cyclooxygenase orthologues were later cloned. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
PGI2 induction in endothelial cells by interleukin-2 (IL-2) is inhibited by actinomycin D and cycloheximide (Frasier-Scott et al 1988). Using an affinity-purified antibody to sheep seminal vesicle cyclooxygenase, a substantial increase in immunoreactive material was observed in IL-2-treated cells. Similarly, ligand-stimulated MC3T3-E1 murine osteoblast-like cells demonstrated an increase in microsomal cyclooxygenase activity which could be immunoprecipitated by antisera to sheep seminal vesicle cyclooxygenase. Raz et al (1988) used antisera to sheep seminal vesicle cyclooxygenase to demonstrate that IL-1 treatment of human fibroblasts stimulated incorporation of radioactive methionine into immunoprecipitable cyclooxygenase. This immunoprecipitated cyclooxygenase was reported to be similar to sheep seminal vesicle cyclooxygenase, based on N-terminal amino acid sequencing and
44
BIOSYNTHESIS AND METABOLISM
endogycosidase H treatment. Ras et al (1989) concluded that IL1-induced PGE2 production in human dermal fibroblasts is the consequence of ligand-induced cyclooxygenase synthesis. Following luteinizing hormone (LH) treatment of rats, granulosa cells produce large amounts of immunoreactive cyclooxygenase (Wong et al 1989). Although LH-stimulated cyclooxygenase accumulation was blocked by a transcriptional inhibitor, cyclooxygenase message was not increased in LHtreated cells. Wong et al (1989) concluded that this LH-induced increase in cyclooxygenase did not result from increased transcription of the cyclooxygenase gene. Endotoxin/lipopolysaccharide (LPS) stimulation of human monocytes also led to increased microsomal cyclooxygenase activity and immunoprecipitable cyclooxygenase. LPS-induced PGE2 production, elevated microsomal cyclooxygenase activity and increased immunoprecipitable cyclooxygenase protein were all inhibited by dexamethasone, a potent inhibitor of prostaglandin production (Fu et al 1990). These authors suggested that monocytes ‘‘. . . may contain two pools of cyclooxygenase, each with a differential sensitivity to endotoxin or dexamethasone’’. Masferrer et al (1990) demonstrated that peritoneal macrophages isolated from endotoxin-treated mice had increased levels of cyclooxygenase activity and—when incubated with radioactive amino acids in culture—had increased immunoprecipitable cyclooxygenase. Both induced cyclooxygenase activity and immunoprecipitable radioactive cyclooxygenase protein were blocked by dexamethasone. Chicken embryo fibroblasts expressing the v-src oncogene also produce elevated levels of immunoprecipitable cyclooxygenase (Han et al 1990). Thus, by 1990, it was clear that mitogens (PDGF, EGF), inflammatory agents (LPS), hormones (LH), cytokines (IL-1) and oncogenes (v-src) could induce transcription- and translation-dependent accumulation of cyclooxygenase activity and immunoprecipitable cyclooxygenase. EVIDENCE FOR A SECOND CYCLOOXYGENASE GENE Rosen et al (1989) demonstrated that anti-sheep seminal vesicle cyclooxygenase antibodies could precipitate a 12-fold increase in immunoprecipitable, radioactively labelled protein from serumstimulated sheep tracheal epithelial cells in culture. They probed northern blots from serum-stimulated and control cells with a cDNA probe for the 2.8 kb cyclooxygenase message. At high stringency, they saw no increase in this message. At lower stringency, they observed a 4.0 cross-reacting RNA species present in serum-treated cells, absent in untreated cells. They suggested that the 4.0 mRNA ‘‘may be derived from a distinct, cyclooxygenase-related gene’’ and ‘‘. . . may encode for a protein with cyclooxygenase activity’’. The following year, Masferrer et al (1990) suggested that two cyclooxygenase genes may encode two cyclooxygenase proteins, one regulated by glucocorticoids. CLONING THE COX-2 cDNA Despite circumstantial evidence for an inducible cyclooxygenase in hormone-, cytokine-, growth factor-, mitogen- and oncogenetreated cells, identification of the inducible cyclooxygenase did not result from a directed search for a second cyclooxygenase in stimulated cells. Each laboratory that reported the cloning of a second, inducible cyclooxygenase in 1991 was searching for genes that modulate growth-factor or oncogene-stimulated proliferative responses. Simmons et al (1989) identified cDNAs induced in chicken embryo fibroblasts in response to v-src expression or serum
stimulation. Xie et al (1991) sequenced one of these genes, CEF147, whose open reading frame encoded a protein with 59% homology to sheep seminal vesicle cyclooxygenase. They suggested that this gene, which they named miPHSch for ‘‘mitogeninducible PGSchicken’’, might be a new cyclooxygenase and pointed out the similar size of the CEF-147 message and the 4.1 kb message described by Rosen et al (1989). However, they could not be certain of their suggestion because, at the time their paper was published, the chicken orthologue of sheep seminal vesicle cyclooxygenase had not been cloned. Consequently, they could not be certain whether miPHSch was a distinct, inducible cyclooxygenase or the chicken orthologue of the previously cloned ovine/human/murine cyclooxygenase. Our laboratory was interested in identifying ‘‘immediate early’’ or ‘‘primary response’’ genes that mediate the transition from G0arrested cells to proliferating cells in response to mitogenic stimulation. We used differential screening of a cDNA library prepared from mitogen-treated cells to identify cDNAs for mitogen-stimulated genes. The library was screened with probes prepared from mRNAs isolated from G0 cells and from tetradecanoyl phorbol acetate (TPA)-treated cells to identify TPA-induced sequences, or TIS genes (Lim et al 1987). In 1991, we reported (Kujubu et al 1991) that the sequence of the TIS10 ‘‘immediate early’’ gene had striking sequence homology to the previously cloned murine cyclooxygenase cDNA. In contrast to the studies with miPHSch, we could conclude unambiguously that the 2.8 kb murine cyclooxygenase message and the 4.0 TIS10 message encoded highly related proteins that were derived from distinct genes. The TIS10/inducible cyclooxygenase cDNA was induced by EGF, forskolin and serum, in addition to TPA, in murine fibroblasts. The constitutive cyclooxygenase gene encoding the 2.8 kb message is now called COX-1 and the inducible cyclooxygenase gene encoding the 4.0 kb message is called COX-2. COX-2 mRNA has multiple AUUUA sequences (Xie et al 1991; Kujubu et al 1991), associated with rapid mRNA degradation. O’Banion et al (1991) described a v-src- and serum-inducible cyclooxygenase immunoprecipitable with antiserum to sheep seminal vesicle cyclooxygenase in murine fibroblasts. Induction was suppressed by glucocorticoids. Low-stringency hybridization identified a glucocorticoid-sensitive 4.0 kb message that crossreacted with the 2.8 kb probe encoding ovine seminal vesicle cyclooxygenase present in the serum-treated fibroblasts. They concluded, based on ‘‘stringency analysis and preliminary sequence data’’, that this RNA ‘‘arises from a gene distinct from that transcribed into the previously cloned 2.8 kb PGHS cDNA’’ and demonstrated that the product of in vitro translation of the 4.0 kb message could be immunoprecipitated by antiserum to sheep seminal vesicle cyclooxygenase. O’Banion et al (1991) concluded that their data ‘‘suggests that two distinct and differentially regulated cyclooxygenase species exist’’. In 1992, O’Banion et al (1992) also reported the sequence of the murine COX-2 cDNA. Using sequence data from the murine COX-2 cDNA, Hla and Neilson (1992) and Jones et al (1993) cloned the human COX-2 cDNA and demonstrated its induction in various cell types. The COX-2 cDNA was later cloned from chicken, rat, dog, guinea-pig, sheep, rabbit, mink, horse, cow and trout. THE COX-2 AND COX-1 GENES COX-1 is constitutively expressed in most cells. In contrast, COX2 is inducible by a wide range of inducers, in a variety of cell types (Herschman 1991). ‘‘Immediate early genes’’ are usually small compared to other genes (Herschman 1996). As we thought might be the case, the murine COX-2 gene is indeed smaller than the COX-1 gene. The murine COX-2 gene is found in an 8 kb DNA segment (Fletcher et al 1992), even though murine COX-2 mRNA
PROSTAGLANDIN SYNTHASE 2/CYCLOOXYGENASE II is 4.0 kb and murine COX-1 mRNA, encoded from a 22 kb gene (Yokoyama and Tanabe 1989), is only 2.8 kb. The COX-2 gene has only 10 exons and 9 introns; a COX-1 exon coding an N-terminal hydrophobic leader sequence is not present in the COX-2 gene. The remaining exons of the COX-1 and COX-2 genes and the splice sites at the exon–intron junctions are similar for COX-1 and COX-2, with the exception of their final 30 /C-terminal exons. The last, longer exon of the COX-2 gene encodes the longer 30 untranslated region of the COX-2 message. The major differences between the COX-2 and COX-2 transcription units are the absence of the first exon found in COX-1 and the consistently smaller introns in the COX-2 gene, as compared to the COX-1 gene. COX-1 and COX-2 map to distinct chromosomes, both in mouse and human. The sequences upstream of the transcriptional start sites of the COX-1 and COX-2 genes bear no resemblance to one another. In contrast, the upstream regulatory regions of the murine and human COX-2 genes have substantial sequence similarity, sharing a number of putative transcription factor binding sites. THE COX-1 AND COX-2 PROTEINS The COX-1 protein is 602 amino acids; the COX-2 protein is 604 amino acids. COX-1 and COX-2 share about 80% amino acid identity/similarity, differing most extensively at their amino- and carboxyl-terminal ends. COX-1 has a 17 amino acid hydrophobic sequence at its N-terminal that is not shared by COX-2. Conversely, COX-2 has an 18 amino acid insertion, located near its carboxyl terminal, that is not found in COX-1. COX-1 and COX-2 are membrane-bound homodimers, found in the endoplasmic reticulum and nuclear membrane. COX-1 (Picot et al 1994) and COX-2 (Kurumbail et al 1996; Luong et al 1996) share remarkable similarities in tertiary structure. Both are bound to the membrane by amphipathic helices that form monotopic membrane attachment sites. The helices comprising the COX-1 and COX-2 membrane binding sites contain the greatest divergence in their sequences, with the exception of the unique N-terminal and C-terminal peptide insertions. The cyclooxygenase and peroxidase active sites are located at distinct regions of the proteins. Arachidonate gains access to the COX-1 and COX-2 cyclooxygenase sites by a hydrophobic channel adjacent to the membrane binding sites. The COX-2 active site contains a ‘‘side pocket’’ not present in the COX-1 active site. The serine 530 residue that is the site of COX-1 aspirin acetylation is conserved at serine 516 in COX-2, and is also subject to irreversible aspirin acetylation, leading to COX-2 inhibition. COX-1 tyrosine 385, essential for cyclooxygenase activity, is conserved as tyrosine 371 in COX-2, and is also essential for activity. The sequences required for haem binding (TIWLREHNRV and RGLGF) are also conserved in COX-1 and COX-2. For additional information about COX-1 and COX-2 membrane association, substrate and inhibitor accessibility to the COX-1 and COX-2 active sites, COX-1 and COX-2 kinetic parameters and mechanisms of catalysis by COX-1 and COX-2, readers are referred to reviews that emphasize COX-1 and COX-2 biochemistry, enzymology and pharmacology (e.g. Smith et al 2000). WHY ARE TRADITIONAL NSAIDs EFFECTIVE INHIBITORS OF BOTH COX-1 AND COX-2? COX inhibitors were first identified by their ability to inhibit prostaglandin production in vitro by COX-1 purified from sheep seminal vesicles. Active compounds were then tested for efficacy in
45
in vivo inflammation models. Candidates effective in inhibiting inflammation and demonstrating appropriate pharmacokinetic and pharmacodynamic properties were then brought to clinical trial. Since the animal model, in retrospect, required COX-2 inhibition, compounds that ‘‘passed’’ both the in vitro and in vivo preclinical tests were, by definition, both COX-1 and COX-2 inhibitors. Lead compounds that could preferentially inhibit COX-2 were identified only when COX-2 was used to screen chemical libraries for inhibitors that could distinguish between COX-1 and COX-2. Subsequent refinement led to development of the blockbuster drugs Celebrex and Vioxx, with additional COX-2 pharmacological agents on the horizon. THE COX-1/COX-2 PARADOX: WHY ARE THERE TWO CYCLOOXGENASES? Until the discovery of COX-2, the dogma for prostanoid synthesis suggested that the rate-limiting step in ligand-stimulated prostaglandin synthesis was the activation of phospholipase to release arachidonic acid from membrane stores. Constitutive cyclooxygenase activity in cells—now known as COX-1—was thought to be present in excess, and available to convert the stimulus-released arachidonic acid to PGH2. With the discovery of an inducible cyclooxygenase, COX-2, the question arose as to why there are two cyclooxygenases. If constitutive cyclooxygenase is available to convert arachidonate to PGH2, what is the role of an inducible cyclooxygenase? Moreover, if arachidonate released in cells following ligand stimulation is available to COX-1 and COX-2, why do COX-2-specific inhibitors such as Celebrex and Vioxx work? How can the COX-2-specific inhibitors prevent stimulusinduced prostaglandin synthesis if COX-1 is available to convert arachidonate acid to PGH2? To determine whether arachidonic acid availability is functionally restricted to distinct COX isozymes, Reddy and Herschman (1994) examined the production of prostaglandins in activated cells unable to synthesize COX-2. Antisense oligonucleotides (ASOs) directed against COX-2 mRNA were used to block COX-2 protein production in both mitogen-activated fibroblasts and endotoxin-activated macrophages. In both cases, ligandstimulated PGE2 production was blocked by ASO inhibition of COX-2 synthesis. When the fibroblasts or macrophages activated in the presence of COX-2 ASO were provided with exogenous arachidonic acid, they were able to synthesize PGE2. The data suggest: (a) that COX-1 present in fibroblasts and macrophages is active—it can convert exogenously supplied arachidonate to PGH2; and (b) that COX-1 present in fibroblasts and macrophages is unavailable to convert endogenous arachidonic acid, released following ligand stimulation, to PGH2. Endogenous arachidonic acid released from membrane stores following ligandstimulated phospholipase activation is not available to COX-1, but is available to COX-2. Using retrovirally expressed COX-1 and COX-2, Chulada et al (1996) reached fundamentally the same conclusions. When the first COX-2 inhibitor, NS-398, became available, we examined its ability to inhibit prostaglandin production from endogenously and exogenously supplied arachidonic acid, following mitogen stimulation of fibroblasts. NS-398 blocked PGE2 production from endogenous arachidonic acid following PDGF treatment, but had no effect on COX-1-dependent PGE2 synthesis when the cells were supplied with exogenous arachidonic acid (Herschman 1996). Endogenous arachidonic acid released in response to ligand stimulation can be converted to PGH2 by COX-2, but not by COX-1, under intracellular conditions. The simplest explanation for segregated access to endogenous arachidonic acid for COX-1 and COX-2 would be a difference in their intracellular location. Although initially thought to be the
46
BIOSYNTHESIS AND METABOLISM
case (Regier et al 1993), immunofluorescence analyses (Reddy and Herschman 1994; Spencer et al 1998) and immunoelectron microscopy studies (Spencer et al 1998) suggest that the subcellular localizations of COX-1 and COX-2 are indistinguishable. In most cell types (fibroblasts, endothelial cells, macrophages, epithelial cells, osteoblasts) ligand-stimulated prostaglandin synthesis is relatively slow and is dependent on COX-2 gene expression. In contrast, when mast cells are activated by IgE receptor aggregation, PGD2 expression occurs in a biphasic fashion; there is a rapid ‘‘early’’ burst of PGD2 synthesis, followed by a ‘‘delayed’’ phase of PGD2 production (Kawata et al 1995; Murakami et al 1995). Aspirin inhibition, glucocorticoid inhibition and COX-1- and COX-2-specific inhibitors have been employed to demonstrate that the ‘‘early’’ PGD2 production phase is due to pre-existing, constitutive COX-1, while ‘‘delayed’’ PGD2 production is due to induced COX-2. In this special case of prostaglandin production, the temporal separation of COX-1 and COX-2 expression contributes to distinctions in the production of PGD2. COX-1 AND COX-2 USE ARACHIDONIC ACID WITH DIFFERING EFFICIENCIES AT LOW SUBSTRATE CONCENTRATIONS COX-2 is more efficient at low arachidonic acid concentrations (below 0.5 mM), while COX-1 is more efficient at higher arachidonic acid concentrations (above 0.25 mM) (Swinney et al 1997). COX-1 shows positive cooperation with arachidonic acid as substrate, in contrast to COX-2 (Swinney et al 1997; Chen et al 1999). Swinney et al (1997) suggest that COX-1 positive cooperation plays a major role in COX-1’s inability to utilize endogenous arachidonic acid at the concentrations released intracellularly by ligand-stimulated phospholipases. The haem groups within the COX-1 and COX-2 peroxidase sites require a lipid peroxide-dependent ‘‘initiation’’ reaction. The oxidized haem at the peroxidase side then oxidizes a tyrosine residue in the COX cyclooxygenase site. This oxidized tyrosine removes a hydrogen atom from arachidonate, forming an arachidonic acid radical that reacts with molecular oxygen to form PGG2. PGG2 is, itself, a lipid peroxide. Chen et al (1999) postulate that PGG2 can participate in a feedback loop to activate the cyclooxygenase activity of latent COX molecules. The required concentration for initiator peroxide is about 10-fold lower for COX-2 than for COX-1. Kulmacz and Wang (1995) suggest that the reduced requirement for initiator can explain the ability of COX-2 to preferentially utilize low concentrations of arachidonate as substrate in cells. COUPLING OF COX-1 AND COX-2 TO UPSTREAM PHOSPHOLIPASES AND DOWNSTREAM PROSTAGLANDIN/THROMBOXANE/ PROSTACYCLIN SYNTHASES Prostanoid production requires three enzymatic reactions: activation of phospholipase to release arachidonic acid from glycerophospholipids, conversion of arachidonic acid to PGG2 and then to PGH2 by COX-1 or COX-2, and conversion of PGH2 to the various prostanoids by tissue-specific prostaglandin/ thromboxane/prostacyclin synthases. In activated mast cells, we found that a secretory phospholipase (Reddy and Herschman 1996), subsequently identified as type V sPLA2 (Reddy et al 1997), provides arachidonate to COX-1 and a cytoplasmic phospholipase provides arachidonate to COX-2. Although these results were initially controversial (Bingham et al 1996), studies with cPLA2 knockout mice confirmed the requirement of cPLA2 for the
delayed, COX-2-dependent PGD2 synthesis in mast cells (Fujishima et al 1999) and transfection studies demonstrated the role of type V sPLA2 in providing arachidonate to COX-1 for ‘‘early’’ PGD2 synthesis (Enomoto et al 2000). In the past several years, a profusion of cytoplasmic and secreted phospholipases has been described, and an extensive, confusing and contradictory literature has developed concerning the identities of phospholipases that provide arachidonic acid to COX-1 and COX-2. The upshot of these studies is that both cPLA2 and several secretory phospholipases can supply arachidonic acid to both COX-1 and COX-2. The coupling of specific phospholipases and cyclooxygenases is dependent on cell type, ligand and relative enzyme concentrations. The field is currently an area of intense research, conflicting studies and relatively tentative conclusions. Previously, the rate-limiting step in prostanoid production was thought to be phospholipase activation. With the discovery of COX-2, the model shifted to COX-2 as a subsequent limiting factor for ligand-induced prostanoid production. The conversion of PGH2 to the final prostanoid product was not thought to be subject to transient, ligand-dependent events. However, Matsumoto et al (1997) demonstrated that the qualitative nature of prostanoids produced in rat peritoneal macrophages changed, depending on the stimulating ligand. Cells activated with a calcium ionophore, to stimulate COX-1-dependent prostanoid synthesis, produce TXA2 and PGD2. In contrast, when stimulated with endotoxin, these cells produce PGE2 and PGI2. When exogenous PGH2 was provided (to bypass COX-2 synthesis), LPS-induced PGE2 synthesis required transcription- and translation-dependent production of PGE2 synthase (PGES). The data suggested that (a) PGES is induced by LPS and (b) induced PGES is coupled to induced COX-2. Similarly, unstimulated rat peritoneal macrophages produce TXA2 and PGD2 predominantly but, when stimulated with endotoxin, produce primarily PGI2 and PGE2 (Brock et al 1999). However, these authors did not observe an increase in PGES levels following endotoxin stimulation. They suggested that COX-1 and COX-2 might be able to preferentially and differentially couple with alternative prostanoid synthases in cells, thereby qualitatively changing the spectrum of the prostanoids produced. Similar results have also been observed in endothelial cells (Caughey et al 2001). Unstimulated human umbilical vein endothelial cells (HUVEC) produce TXA2. Following IL-1b treatment, thromboxane levels increase two-fold. In contrast, IL-1b-treated HUVECs produce vastly increased levels of PGI2 (54-fold) and PGE2 (84-fold) from endogenous arachidonic acid. Increased PGI2 and PGE2 production relative to TXA2 also occurs if the IL-1b-treated cells are tested for prostanoid synthesis with exogenous PGH2 as substrate, suggesting that the levels of PGES and PGI2 synthase must be elevated in IL-1b-treated HUVECs. A new chapter in our understanding of the nature of prostaglandin synthesis was opened when Jakobsson et al (1999) described the cloning of an inducible, membrane-bound PGES, mPGES, from IL-1b-treated cells. Murakami et al (2000) also cloned mPGES and demonstrated that it is inducible by inflammatory signals in macrophages and osteoblasts. In a companion paper, Tanioka et al (2000) cloned a constitutive, cytoplasmic cPGES isozyme. These papers showed that mPGES is functionally coupled to COX-2 and cPGES is functionally coupled to COX-1, when cells use endogenous arachidonic acid. However, if arachidonic acid is artificially elevated in cells, this strict coupling breaks down; mPGES can use COX-2-derived PGH2 to produce PGE2 (Murakami et al 2000). Like COX-2, mPGES induction can be inhibited by glucocorticoids (Stichtenoth et al 2001). MPGES has emerged as an interesting candidate for pharmacological inhibition, since it modulates PGE2 production in an inflammatory context.
PROSTAGLANDIN SYNTHASE 2/CYCLOOXYGENASE II Using co-transfection studies to characterize the coupling of other prostanoid synthases with COX-1 and COX-2, Ueno et al (2001) found that thromboxane synthase and PGI2 synthase, like mPGES, preferentially couple with COX-2. However, the functional linkage of COX-1 and COX-2 with the prostanoid synthases was strongly modulated by the levels of arachidonic acid available to the enzymes in cells. An even more interesting ‘‘twist’’ to the qualitative changes in prostanoid production was provided by Sun et al (2001). Unstimulated RAW264.7 macrophages produce small amounts of TXA2 and PGE2. When stimulated with endotoxin or by crosslinking of ab1 integrin, COX-2 levels and PGES levels are elevated, and both TXA2 and PGES message synthesis is substantially induced. In contrast, if the hyaluronan receptor (CD44) is cross-linked, COX-2 mRNA levels are induced but PGES mRNA levels are not elevated. Moreover, the production of TXA2 is substantially increased in these cells, reaching levels obtained by LPS treatment of ab1 integrin cross-linking, but no elevation in PGE2 production occurs. Thus, the nature of the inducer determines the qualitative distribution of prostaglandin production by differentially regulating COX-2 and terminal prostaglandin/thromboxane/prostacyline synthase gene expression. Clearly, the coupling of the myriad phospholipases, the COX-1 and COX-2 enzymes and the terminal prostanoid synthases to produce prostanoids following ligand stimulation will continue to provide fertile ground for research for some time to come.
CANCER AND CYCLOOXYGENASES We have known for many years that low-dose aspirin reduces colon cancer morbidity and mortality. A substantial literature now exists demonstrating elevation of COX-2 gene expression both in colon tumours and in epithelial cells adjacent to the sites of tumour initiation, suggesting that COX-2 misregulation plays a role in colon cancer initiation and progression. COX-2-specific inhibitors are extremely effective at inhibiting induction of colon tumours in experimental animals (Kawamori et al 1998) and reduce the frequency and size of colon polyps in familial adenomatous polyposis (FAP) patients (Steinbach et al 2000). Mice heterozygous for a mutation in the multiple intestinal neoplasia (min) gene, the orthologue of the human FAP gene, develop multiple intestinal tract neoplasms. When crossed to COX-2 null mice, the frequency of intestinal tract tumours was dramatically decreased (Oshima et al 1996). Pharmacological studies suggest that antiangiogenic effects may play a major role in the efficacy of COX-2 inhibitor-based prevention of colon and lung cancer progression (Masferrer et al 2000). These data lead to the suggestion that elevated COX-2 expression plays a role in colon tumour initiation, vascularization and progression. However, when min (+) mice were crossed to COX-1 null mice a similar decrease in intestinal tumours occurs (Chulada et al 2000), suggesting that both COX isoforms play roles in colon tumour development. Because of the interest generated in ectopic COX-2 overexpression in colon cancer, COX-2 expression has been examined in a variety of tumours. Elevated COX-2 has been described in bladder, cervix, oesophageal, gastric, glioblastoma, head and neck, hepatocellular, lung, mammary, mesothelioma, ovarian, pancreatic, phaeochromocytoma, prostate and skin (melanoma) tumours, in addition to colorectal tumours. The role of COX-2 in tumours and the efficacy of COX-2 inhibitors as adjuvant chemotherapy are now major investigative areas.
47
REGULATION OF THE COX-2 GENE Within a few years of its discovery, a variety of stimuli, including growth factors, cytokines, fluid shear, free-radical generators, hormones, inflammatory stimuli, ischaemia reperfusion, neurotransmitters, oncogenes, radiation and cell stressors, were shown to induce COX-2. COX-2 can be induced in amnion cells, astrocytes, decidual cells, dendritic cells, endometrial cells, endothelial cells, epithelial cells, fibroblasts, granulosa cells, lymphocytes, macrophages, mast cells, mesangial cells, mesothelial cells, microglia, neurons, neutrophils, osteoblasts, smooth muscle cells, synovial cells and others I have surely left out. The main issues considered in stimulus-generated transcriptional gene activation are: (a) the cytoplasmic signalling pathway that transduces the initial signalling event (e.g. receptor occupation) to the nucleus; (b) the transcription factors, activated by the signalling event, that initiate target gene transcription; and (c) the cis-acting regulatory elements of the target gene that control transcription. In addition to regulation by transcriptional activation, gene expression can be controlled by message splicing, message movement from nucleus to cytoplasm, message stability and message translational efficiency. COX-2 gene regulation appears to vary widely among cell types and among inducing agents.
Cis-Acting Sites of the COX-2 Gene Because the same COX-2 gene exists in all cells, it is easiest to first consider cis-acting COX-2 regulatory sites. Signalling pathways and transcription factors must converge on some subset(s) of cisacting COX-2 regulatory sites. Murine and human COX-2 genes share a number of putative transcription factor binding sequences, including an E-Box, a cyclic AMP response element (CRE), NFIL6/C-EBP sites, an NF-kB binding site, etc., in the first kilobase of DNA 50 of the transcription start site. Transient transfection assays, with constructs in which luciferase reporter gene transcription is driven either by the COX-2 promoter region or by promoter regions containing mutations of putative regulatory elements, demonstrated that the CRE plays the major role in v-src, serum and PGDF COX-2 gene stimulation in murine fibroblasts (Xie and Herschman 1995, 1996). The CRE of the murine and human COX-2 gene also plays a role in COX-2 induction of endotoxin-stimulated macrophages (Wadleigh et al 2000), activated mast cells (Reddy et al 2000), UV-induced keratinocytes (Gorgoni et al 2001), TNFa (Potter et al 2000; Hansen et al 2000) and IL-1 (Wang and Tai 1999) stimulated amnion cells, hormone and phorbol ester treated osteoblasts (Okada et al 2000; Wadleigh and Herschman 1999), and both IL-1 (Kirtikara et al 2000) and endotoxin (Inoue et al 1995) treated endothelial cells. Interestingly, the rat COX-2 gene has no CRE. Reporter gene experiments suggest that an E-Box plays a major role in stimulus-induced COX-2 gene expression in rat cells (Morris and Richards 1996). Sirois and Richards (1993) first suggested that the COX-2 NFIL-6 site(s) might play a role in regulating COX-2 expression. They demonstrated that an NFIL-6 mutation blocks hormonal COX-2 induction in rat granulosa cells. Mutational studies suggest that COX-2 NFIL-6 sites play a role in TNFa-stimulated osteoblasts (Yamamoto et al 1995; Wadleigh and Herschman 1999), TNFa-treated osteoblasts (Yamamoto et al 1995), amnion cells (Hansen et al 2000) and pancreatic islet cells (Sorli et al 1998), fluid shear-stressed osteoblasts (Ogasawara et al 2001), IL1b-treated chondrocytes (Thomas et al 2000), amnion (Potter et al 2000) and endothelial cells (Kirtikara et al 2000), endotoxintreated vascular endothelial cells (Inoue et al 1995) and
48
BIOSYNTHESIS AND METABOLISM
macrophages (Wadleigh et al 2000; Mestre et al 2001), and activated mast cells (Reddy et al 2000). The most controversial of the COX-2 cis-acting prospective regulatory sites is the NF-kB site. We found that mutation of the NF-kB site does not inhibit induction of COX-2 promoter/ luciferase reporter genes in hormone-stimulated osteoblasts (Wadleigh and Herschman 1999), activated mast cells (Reddy et al 2000) or endotoxin-stimulated macrophages (Wadleigh et al 2000). Mutation of the human NF-kB site does not suppress COX-2 promoter activity in response to IL-1 (Sorli et al 1998). In the rat, which has no CRE in the COX-2 gene, mutation of COX2 NF-kB sites also did not significantly reduce promoter activity (Chen et al 1999). In contrast, several laboratories have reported that mutation of the human COX-2 NF-kB site attenuated induction from the COX-2 promoter (Wang and Tai 1998; Schmedtje et al 1997; Allport et al 2000). An extensive literature suggests that activation of NF-kB transcription factors plays a role in induced COX-2 gene expression, but in many cases the affected COX-2 cis-acting site has not been examined. The role of the COX-2 NF-kB site in stimulus-induced gene expression appears to be quite context-sensitive. Transcription Factors We initially assumed that the cyclic AMP binding protein, CREB, would be the transcription factor modulating COX-2 expression at the CRE in mitogen- and oncogene-stimulated fibroblasts. Gel shift experiments demonstrated a strong interaction between the COX-2 CRE and CREB (Xie et al 1994). However, cotransfection of a plasmid expressing wild-type CREB, along with a COX-2 promoter/luciferase reporter gene, suppressed vsrc-induced luciferase expression (Xie and Herschman 1995). The simplest interpretation is that CREB acts as a ‘‘dominant negative’’, binding at the COX-2 CRE but not acting as a functional transcription factor. ‘‘Supershift’’ experiments with anti-Jun antibody demonstrated that c-Jun also binds to the COX-2 CRE (Xie and Herschman 1995). In contrast to the inhibition with overexpressed CREB, c-Jun overexpression enhanced v-src stimulated expression from the COX-2 promoter. We also replaced the CRE in the COX-2 promoter/luciferase reporter gene with the DNA binding domain of the gal4 promoter, and tested the ability of GAL4-CREB, GAL4-ATF2 and GAL4-JUN chimeric proteins to stimulate luciferase production. Only GAL4-JUN was active; GAL4-CREB and GAL-4ATF2 were inactive. We concluded that c-JUN, not CREB, plays the major role in mediating gene expression at the COX-2 CRE following v-src stimulation in fibroblasts. CREB co-expression with the COX-2 promoter/luciferase reporter gene also suppresses luciferase activity in mitogen-stimulated fibroblasts (Xie and Herschman 1996), hormone-stimulated osteoblasts (Wadleigh and Herschman 1999), activated mast cells (Reddy et al 2000) and endotoxin-treated macrophages (Wadleigh et al 2000). In contrast, c-JUN overexpression enhances COX-2 expression in all these contexts, leading us to conclude that, in most cell contexts, c-JUN is likely to be the transcription factor modulating CREmediated COX-2 gene expression in response to ligand/oncogene stimulation. Caivano and Cohen (2000) suggest that CREB is the transcription factor active at the COX-2 CRE in endotoxin-stimulated RAW264 macrophages. Their conclusion is based primarily on pharmacological inhibition of signalling kinases upstream of transcription factor activation and the observation that endotoxin-stimulated CREB phosphorylation correlates with COX-2 gene expression. However, endotoxin treatment of these same cells also stimulates transient phosphorylation of c-JUN (Wadleigh et al 2000). Moreover, despite activation of CREB
phosphorylation by forskolin in RAW264 cells, COX-2 gene expression is not enhanced (Caviano and Cohen 2000). The correlation of CREB phosphorylation and COX-2 expression and their coordinate pharmacological inhibition following stimulation also lead to the conclusion that CREB mediates lysophosphatidylcholine induction of COX-2 in vascular endothelial cells (Rikitake et al 2001) and UV-induced COX-2 expression in human keratinocytes (Tang et al 2001). Based only on increased CREB phosphorylation, Ogasawara et al (2001) concluded that CREB mediates fluid shear-stressed COX-2 expression in osteoblasts. Inoue et al (1995) were the first to report that C/EBPd overexpression enhances induced expression from the COX-2 promoter, in this case in endotoxin- and phorbol ester-treated vascular endothelial cells. Similar co-transfection experiments suggested that C/EBP proteins, binding at COX-2 NFIL6 sites, play a role in IL-1b-treated chrondrocytes (Thomas et al 2000), hormone-treated osteoblasts (Wadleigh and Herschman 1999; Harrison et al 2000), endotoxin-stimulated macrophages (Wadleigh et al 2000) and activated mast cells (Reddy et al 2000). COX-2 induction is impaired in endotoxin-treated macrophages prepared from C/EBPb7/7 mice (Gorgoni et al 2001). In contrast, COX-2 induction is not diminished in serum, TNFa, IL1b or v-src treated fibroblasts from these C/EBPb7/7 mice (Gorgoni et al 2001) or in epidermal cells of TPA-treated C/ EBPb7/7 mice (Wang et al 2001). These experiments demonstrate both the complexity and cell-type specificity of COX-2 gene regulation and the power that genetically manipulated mice bring to the analysis of gene expression. The role of NFkB transcription factors is among the most controversial aspects of COX-2 gene expression. p65 reduction by antisense in human rheumatoid synovial fibroblasts is reported to reduce IL-1b induction of COX-2 (Roshak et al 1996; Crofford et al 1997). Nucleotide decoy experiments sequestering p65 suggest that hypoxia-induced COX-2 expression in human endothelial cells also requires p65 (Schmedtje et al 1997). Co-transfection of p65 and p50 expression vectors synergistically stimulate expression from a COX-2 reporter gene (Newton et al 1997), also suggesting a role for NF-kB-mediated COX-2 gene expression. However, when NF-kB activation is blocked in murine macrophages with a dominant negative I-kB, both basal and endotoxininduced expression from a COX-2 promoter/luciferase reporter gene are elevated, while expression from an NF-kB reporter plasmid is strongly inhibited (Wadleigh et al 2000). Again, distinctions in cell type and ligand specificity for COX-2 induction are likely to account for much of the apparently contradictory literature concerning NFkB in COX-2 induction. Signalling Pathways Because of our initial observations regarding the c-JUN transcription factor in activating COX-2 in fibroblasts (Xie and Herschman 1995, 1996) and subsequent studies in osteoblasts (Wadleigh and Herschman 1999), macrophages (Wadleigh et al 2000) and mast cells (Reddy et al 2000), we have concentrated on signalling pathways that activate the RAS–MEKK1–JNKK– JNK–JUN pathway. Using primarily co-transfection strategies with dominant-negative and activated constructs of RAS and the various signalling kinases, we demonstrated that this pathway plays a role in mitogen- and oncogene-activated COX-2 expression in fibroblasts (Xie and Herschman 1995, 1996), in hormone-induced COX-2 expression in osteoblasts (Wadleigh and Herschman 1999) and in COX-2 expression in activated mast cells (Reddy et al 2000). Using similar approaches, we also find that the RAS–RAF–MAPK–ERK pathway plays a role in COX-2 induction in these cells. Since our original reports on the role of
PROSTAGLANDIN SYNTHASE 2/CYCLOOXYGENASE II RAS-dependent activation of the COX-2 gene, many studies implicating ERK 1/2, p38 and JNK have been reported, using cotransfection and pharmacological inhibitor approaches (Herschman 2002). Specific signalling pathways appear to have differing degrees of importance in COX-2 induction, depending on the cell type and inducer examined. In contrast to fibroblasts, osteoblasts and mast cells, DN-RAS co-transfection does not block endotoxin activation of a COX-2 reporter gene in macrophages (Wadleigh et al 2000). The ability of DN-JNK and DN-MEKK1 to block endotoxin-stimulated expression from the COX-2 promoter in macrophages (Wadleigh et al 2000) suggests that c-JUN is also a major COX-2 transcriptional activator in these cells. TLR4 is the major endotoxin receptor. Evolutionary conserved signalling intermediate in toll pathways (ECSIT) is an adaptor protein that connects toll/tumour necrosis factor receptor-associated factor (TRAF60) activation to MEKK1, leading to c-JUN activation (Kopp et al 1999). Co-expression of DN-ECSIT blocks endotoxin-induced reporter gene expression from the COX-2 promoter (Wadleigh et al 2000), suggesting that ECSIT couples occupation of TLR4 by endotoxin to the MEKK1–JNK–c-JUN pathway to activate COX-2 expression. A complete pathway from TLR4 receptor ligand binding to COX-2 gene activation has been proposed (Wadleigh et al 2000). Caivano et al (2001) suggest that endotoxin/LTR4 activation stimulates the classical MAPK pathway and the p38-mediated pathway to phosphorylate MSK kinases that initially activate the COX-2 gene via CREB phosphorylation in macrophages. They do not describe the pathway(s) by which TLR4 activation is coupled to MAPK– SAPK–MSK to phosphorylate CREB. COX-2 MESSAGE STABILITY Ristimaki et al (1994) suggested that COX-2 mRNA stabilization might play a role in COX-2 elevation following cell stimulation. Research on this topic has focused on two main areas: (a) the mechanisms that alter COX-2 mRNA stability in stimulated cells and (b) the sequences in the COX-2 mRNA that regulate message stability. Several laboratories have demonstrated that ligand-dependent COX-2 mRNA stabilization can be blocked with p38 kinase inhibitors (Ridley et al 1998; Dean et al 1999; Matsuura et al 1999; Jang et al 2000; Fiebich et al 2000). Lasa et al (2001) suggest that the ability of glucocorticoids to suppress COX-2 induction is, in part, the result of glucocorticoid inhibition of p38 activation, thus blocking ligand-induced, p38-dependent stabilization of COX-2 message. Lasa et al (2000) propose that the target of p38 for COX2 message stabilization is MAPKAP2, but the downstream target of this latter kinase has not been identified. Sheng et al (2001) suggest that RAS-mediated increases in COX-2 in transformed cells may be regulated by AKT/PKB-induced stabilization of COX-2 message, as well as increased COX-2 transcription. Attempts to identify the regions of COX-2 mRNA that modulate its stability in response to ligand stimulation have used chimeric reporter genes that fuse regions of COX-2 message to stable transcripts (Dixon et al 2000; Lasa et al 2000). Both groups agree that the 123 nucleotides immediately 50 to the translation termination site play a critical role in COX-2 message regulation. This region of the COX-2 30 untranslated message is rich in AUUUA sequences that confer instability to cytokine messages. While several cytokine instability regions share the stabilizing response to p38 activation with the COX-2 sequence, this property is not universal to all AUUUA-rich regions (Lasa et al 2000). Both groups identified proteins that can bind to the region of the COX-2 message associated with regulated stability, but no causal relationships have been established. The character-
49
ization of critical sequences responsible for COX-2 message stabilization, the proteins that mediate this phenomenon and the signalling pathways that translate receptor occupancy into changes in COX-2 message stability should be a fascinating chapter in the story of COX-2 regulation. ACKNOWLEDGEMENTS Thanks to all the members of my laboratory, past and present, who have contributed to the studies from our group described in this review. I also thank my many colleagues in the field of COX-2 biology who have provided advice, reagents and information. My apologies to those whose work I have not cited, due either to lack of space or to oversights by me in reviewing the literature. Supported by Grant AI50495, the UCLA Asthma and Allergic Disease Center supported by the NIAID and NIEHS and Grant CA84572 from the National Cancer Institute. REFERENCES Allport VC, Slater DM, Newton R and Bennett PR (2000) NF-kB and AP-1 are required for cyclo-oxygenase 2 gene expression in amnion epithelial cell line (WISH). Mol Hum Reprod, 6, 561–565. Bingham CO III, Murakani M, Fujishama H et al (1996) A heparinsensitive phospholipase A2 and prostaglandin endoperoxide synthase-2 are functionally linked in the delayed phase of prostaglandin D2 generation in mouse bone marrow-derived mast cells. J Biol Chem, 271, 25936–25944. Brock TG, McNish RW and Peters-Golden M (1999) Arachidonic acid is preferentially metabolized by cyclooxygenase-2 to prostacyclin and prostaglandin E2. J Biol Chem, 274, 11660–11666. Caivano M and Cohen P (2000) Role of mitogen-activated protein kinase cascades in mediating lipopolysaccharide-stimulated induction of cyclooxygenase-2 and IL-1b in RAW264 macrophages. J Immunol, 164, 3018–3025. Caivano M, Gorgoni B, Cohen P and Poli V (2001) The induction of cyclooxygenase-2 mRNA in macrophages is biphasic and requires both CCAAT enhancer-binding protein b (C/EBPb) and C/EBP D transcription factors. J Biol Chem, 276, 48693–48701. Caughey GE, Cleland LG, Penglis PS et al (2001) Roles of cyclooxygenase (COX)-1 and COX-2 in prostanoid production by human endothelial cells: selective upregulation of prostacyclin synthesis by COX-2. J Immunol, 167, 2831–2838. Chen G, Wood EG, Wang SH and Warner TD (1999) Expression of cyclooxygenase-2 in rat vascular smooth muscle cells is unrelated to nuclear factor-kB activation. Life Sci, 64, 1231–1242. Chen W, Pawelek TR and Kulmacz RJ (1999) Hydroperoxide dependence and cooperative cyclooxygenase kinetics in prostaglandin H synthase-1 and -2. J Biol Chem, 274, 20301–20306. Chulada PC, Loftin CD, Winn VD et al (1996) Relative activities of retrovirally expressed murine prostaglandin synthase-1 and -2 depend on source of arachidonic acid. Arch Biochem Biophys, 330, 301–313. Chulada PC, Thompson MB, Mahler JF et al (2000) Genetic disruption of Ptgs-1, as well as Ptgs-2, reduces intestinal tumorigenesis in min mice. Cancer Res, 60, 4705–4708. Crofford LJ, Tan B, McCarthy CJ and Hla T (1997) Involvement of nuclear factor kB in the regulation of cyclooxygenase-2 expression by interleukin-1 in rheumatoid synoviocytes. Arthrit Rheumat, 40, 226– 236. Dean JL, Brook M, Clark AR and Saklatvala J (1999) p38 mitogenactivated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. J Biol Chem, 274, 264–269. DeWitt DL and Smith WL (1988) Primary structure of prostaglandin G/H synthase from sheep vesicular gland determined from the complementary DNA sequence. Proc Natl Acad Sci USA, 85, 1412– 1416. Dixon DA, Kaplan CD, McIntyre TM et al (2000) Post-transcriptional control of cyclooxygenase-2 gene expression. J Biol Chem, 275, 11750– 11757.
50
BIOSYNTHESIS AND METABOLISM
Enomoto A, Murakami M, Valentin E et al (2000) Redundant and segregated functions of granule-associated heparin-binding group II subfamily of secretory phospholipases A2 in the regulation of degranulation and prostaglandin D2 synthesis in mast cells. J Immunol, 165, 4007–4014. Fiebich BL, Mueksch B, Boehringer M and Hull M (2000) Interleukin-1b induces cyclooxygenase-2 and prostaglandin E(2) synthesis in human neuroblastoma cells: involvement of p38 mitogen-activated protein kinase and nuclear factor-kB. J Neurochem, 75, 2020–2028. Fletcher BS, Kujubu DA, Perrin DM and Herschman HR (1992) Structure of the mitogen-inducible TIS10 gene and demonstration that the TIS10-encoded protein is a functional prostaglandin G/H synthase. J Biol Chem, 267, 4338–4344. Frasier-Scott K, Hatzakis H, Seong D et al (1988) Influence of natural and recombinant interleukin 2 on endothelial cell arachidonate metabolism. J Clin Invest, 82, 1877–1883. Fu JY, Masferrer JL, Seibert K et al (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740. Fujishima H, Sanchez Mejia RO, Bingham CO III et al (1999) Cytosolic phospholipase A2 is essential for both the immediate and the delayed phases of eicosanoid generation in mouse bone marrow-derived mast cells. Proc Natl Acad Sci USA, 96, 4803–4807. Gorgoni B, Caivano M, Arizmendi C and Poli V (2001) The transcription factor C/EBPb is essential for inducible expression of the cox-2 gene in macrophages but not in fibroblasts. J Biol Chem, 276, 40769–40777. Habenicht AJ, Goerig M, Grulich J et al (1985) Human platelet-derived growth factor stimulates prostaglandin synthesis by activation and by rapid de novo synthesis of cyclooxygenase. J Clin Invest, 75, 1381–1387. Han JW, Sadowski H, Young DA and Macara IG (1990) Persistent induction of cyclooxygenase in p60v-src-transformed 3T3 fibroblasts. Proc Natl Acad Sci USA, 87, 3373–3377. Hansen WR, Marvin KW, Potter S and Mitchell MD (2000) Tumour necrosis factor-a regulation of prostaglandin H synthase-2 transcription is not through nuclear factor-kB in amnion-derived AV-3 cells. Placenta, 21, 789–798. Harrison JR, Kelly PL and Pilbeam CC (2000) Involvement of CCAAT enhancer binding protein transcription factors in the regulation of prostaglandin G/H synthase 2 expression by interleukin-1 in osteoblastic MC3T3-E1 cells. J Bone Min Res, 15, 1138–1146. Hemler M, Lands WEM and Smith WL (1976) Purification of the cyclooxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyme. J Biol Chem, 251, 5575–5579. Herschman HR (1991) Primary response genes induced by growth factors and tumor promoters. In Annual Review of Biochemistry, Richardson CC, Abelson JN, Meister A and Walsh CT (eds), vol. 60. Annual Reviews, Palo Alto, CA, 281–319. Herschman HR (1996) Prostaglandin synthase 2. Biochim Biophys Acta, 1299, 125–140. Herschman HR (2003) Historical aspects of COX-2; cloning and characterization of the cDNA, protein and gene. In COX-2 Blockage in Cancer Prevention and Therapy, Harris R (ed.). Humana, Totowa, NJ, USA, 13–32. Hla T and Neilson K (1992) Human cyclooxygenase-2 cDNA. Proc Natl Acad Sci USA, 89, 7384–7385. Inoue H, Yokoyama C, Hara S, Tone Y and Tanabe T (1995) Transcriptional regulation of human prostaglandin-endoperoxide synthase-2 gene by lipopolysaccharide and phorbol ester in vascular endothelial cells. Involvement of both nuclear factor for interleukin-6 expression site and cAMP response element. J Biol Chem, 270, 24965– 24971. Jakobsson PJ, Thoren S, Morgenstern R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96, 7220–7225. Jang BC, Sanchez T, Schaefers HJ et al (2000) Serum withdrawal-induced post-transcriptional stabilization of cyclooxygenase-2 mRNA in MDA-MB-231 mammary carcinoma cells requires the activity of the p38 stress-activated protein kinase. J Biol Chem, 275, 39507–39515. Jones DA, Carlton DP, McIntyre TM et al (1993) Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem, 268, 9049–9054.
Kawamori T, Rao CV, Seibert K and Reddy BS (1998) Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res, 58, 409–412. Kawata R, Reddy ST, Wolner B and Herschman HR (1995) Prostaglandin synthase 1 and prostaglandin synthase 2 both participate in activationinduced PGD2 production in mast cells. J Immunol, 155, 818–825. Kirtikara K, Raghow R, Laulederkind SJ et al (2000) Transcriptional regulation of cyclooxygenase-2 in the human microvascular endothelial cell line, HMEC-1: control by the combinatorial actions of AP2, NFIL-6 and CRE elements. Mol Cell Biochem, 203, 41–51. Kopp E, Medzhitov R, Carothers J et al (1999) ECSIT is an evolutionarily conserved intermediate in the Toll/IL-1 signal transduction pathway. Genes Dev, 13, 2059–2071. Kujubu DA, Fletcher BS, Varnum BC et al (1991) TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J Biol Chem, 266, 12866–12872. Kulmacz RJ and Wang LH (1995) Comparison of hydroperoxide initiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J Biol Chem, 270, 24019–24023. Kurumbail RG, Stevens AM, Gierse JK et al (1996) Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature, 384, 644–648. Lasa M, Mahtani KR, Finch A et al (2000) Regulation of cyclooxygenase 2 mRNA stability by the mitogen-activated protein kinase p38 signaling cascade. Mol Cell Biol, 20, 4265–4274. Lasa M, Brook M, Saklatvala J and Clark AR (2001) Dexamethasone destabilizes cyclooxygenase 2 mRNA by inhibiting mitogen-activated protein kinase p38. Mol Cell Biol, 21, 771–780. Lim RW, Varnum BC and Herschman HR (1987) Cloning of tetradecanoyl phorbol ester induced ‘‘primary response’’ sequences and their expression in density-arrested Swiss 3T3 cells and a TPA nonproliferative variant. Oncogene, 1, 263–270. Luong C, Miller A, Barnett J et al (1996) Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nature Struct Biol, 3, 927–933. Masferrer JL, Zweifel BS, Seibert K and Needleman P (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. Masferrer JL, Leahy KM, Koki AT et al (2000) Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res, 60, 1306–1311. Matsumoto H, Naraba H, Murakami M et al (1997) Concordant induction of prostaglandin E2 synthase with cyclooxygenase-2 leads to preferred production of prostaglandin E2 over thromboxane and prostaglandin D2 in lipopolysaccharide-stimulated rat peritoneal macrophages. Biochem Biophys Res Commun, 230, 110–114. Matsuura H, Sakaue M, Subbaramaiah K et al (1999) Regulation of cyclooxygenase-2 by interferon gamma and transforming growth factor a in normal human epidermal keratinocytes and squamous carcinoma cells. Role of mitogen-activated protein kinases. J Biol Chem, 274, 29138–29148. Merlie JP, Fagan D, Mudd J and Needleman P (1988) Isolation and characterization of the complementary DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase). J Biol Chem, 263, 3550–3553. Mestre JR, Mackrell PJ, Rivadeneira DE et al (2001) Redundancy in the signaling pathways and promoter elements regulating cyclooxygenase2 gene expression in endotoxin-treated macrophage/monocytic cells. J Biol Chem, 276, 3977–3982. Miyamoto T, Ogino N, Yamamoto S and Hayaishi O (1976) Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J Biol Chem, 251, 2629–2636. Morris JK and Richards JS (1996) An E-box region within the prostaglandin endoperoxide synthase-2 (PGS-2) promoter is required for transcription in rat ovarian granulosa cells. J Biol Chem, 271, 16633–16643. Murakami M, Bingham CO III, Matsumoto R et al (1995) IgE-dependent activation of cytokine-primed mouse cultured mast cells induces a delayed phase of prostaglandin D2 generation via prostaglandin endoperoxide synthase-2. J Immunol, 155, 4445–4453. Murakami M, Naraba H, Tanioka T et al (2000) Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated
PROSTAGLANDIN SYNTHASE 2/CYCLOOXYGENASE II prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem, 275, 32783–32792. Newton R, Kuitert LM, Bergmann M et al (1997) Evidence for involvement of NF-kB in the transcriptional control of COX-2 gene expression by IL-1b. Biochem Biophys Res Commun, 237, 28–32. O’Banion MK, Sadowski HB, Winn V and Young DA (1991) A serumand glucocorticoid-regulated 4-kilobase mRNA encodes a cyclooxygenase-related protein. J Biol Chem, 266, 23261–23267. O’Banion MK, Winn VD and Young DA (1992) cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc Natl Acad Sci USA, 89, 4888–4892. Ogasawara A, Arakawa T, Kaneda T et al (2001) Fluid shear stressinduced cyclooxygenase-2 expression is mediated by C/EBPb, cAMPresponse element-binding protein, and AP-1 in osteoblastic MC3T3-E1 cells. J Biol Chem, 276, 7048–7054. Okada Y, Voznesensky O, Herschman H et al (2000) Identification of multiple cis-acting elements mediating the induction of prostaglandin G/H synthase-2 by phorbol ester in murine osteoblastic cells. J Cell Biochem, 78, 197–209. Oshima M, Dinchuk JE, Kargman SL et al (1996) Suppression of intestinal polyposis in Apc D716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell, 87, 803–809. Picot D, Loll PJ and Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature, 367, 243– 249. Potter S, Mitchell MD, Hansen WR and Marvin KW (2000) NF-IL6 and CRE elements principally account for both basal and interleukin-1binduced transcriptional activity of the proximal 528 bp of the PGHS-2 promoter in amnion-derived AV3 cells: evidence for involvement of C/ EBPb. Mol Hum Reprod, 6, 771–778. Raz A, Wyche A, Siegel N and Needleman P (1988) Regulation of fibroblast cyclooxygenase synthesis by interleukin-1. J Biol Chem, 263, 3022–3025. Raz A, Wyche A and Needleman P (1989) Temporal and pharmacological division of fibroblast cyclooxygenase expression into transcriptional and translational phases. Proc Natl Acad Sci USA, 86, 1657–1661. Reddy ST and Herschman HR (1994) Ligand-induced prostaglandin synthesis requires expression of the TIS10/PGS-2 prostaglandin synthase gene in murine fibroblasts and macrophages. J Biol Chem, 269, 15473–15480. Reddy ST and Herschman HR (1996) Transcellular prostaglandin production following mast cell activation is mediated by proximal secretory phospholipase A2 and distal prostaglandin synthase 1. J Biol Chem, 271, 186–191. Reddy ST, Wadleigh DJ and Herschman HR (2000) Transcriptional regulation of the cyclooxygenase-2 gene in activated mast cells. J Biol Chem, 275, 3107–3113. Reddy ST, Winstead MV, Tischfield JA and Herschman HR (1997) Analysis of the secretory phospholipase A2 that mediates prostaglandin production in mast cells. J Biol Chem, 272, 13591–13596. Regier MK, DeWitt DL, Schindler MS and Smith WL (1993) Subcellular localization of prostaglandin endoperoxide synthase-2 in murine 3T3 cells. Arch Biochem Biophys, 301, 439–444. Ridley SH, Dean JL, Sarsfield SJ et al (1998) A p38 MAP kinase inhibitor regulates stability of interleukin-1-induced cyclooxygenase-2 mRNA. FEBS Lett, 439, 75–80. Rikitake Y, Hirata K, Kawashima S et al (2001) Signaling mechanism underlying COX-2 induction by lysophosphatidylcholine. Biochem Biophys Res Commun, 281, 1291–1297. Ristimaki A, Garfinkel S, Wessendorf J et al (1994) Induction of cyclooxygenase-2 by interleukin-1a. Evidence for post-transcriptional regulation. J Biol Chem, 269, 11769–11775. Rosen GD, Birkenmeier TM, Raz A and Holtzman MJ (1989) Identification of cyclooxygenase-related gene and its potential role in prostaglandin formation. Biochem Biophys Res Commun, 164, 1358– 1365. Roshak AK, Jackson JR, McGough K et al (1996) Manipulation of distinct NFkB proteins alters interleukin-1b-induced human rheumatoid synovial fibroblast prostaglandin E2 formation. J Biol Chem, 271, 31496–31501. Schmedtje JF Jr, Ji Y-S, Liu WL et al (1997) Hypoxia induces cyclooxygenase-2 via the NFkB p65 transcription factor in human vascular endothelial cells. J Biol Chem, 272, 601–608.
51
Sheng H, Shao J and Dubois RN (2001) K-Ras-mediated increase in cyclooxygenase 2 mRNA stability involves activation of the protein kinase B1. Cancer Res, 61, 2670–2675. Simmons DL, Levy DB, Yannoni Y and Erikson RL (1989) Identification of a phorbol ester-repressible v-src-inducible gene. Proc Natl Acad Sci USA, 86, 1178–1182. Sirois J and Richards JS (1993) Transcriptional regulation of the rat prostaglandin endoperoxide synthase 2 gene in granulosa cells. Evidence for the role of a cis-acting C/EBPb promoter element. J Biol Chem, 268, 21931–21938. Smith WL, DeWitt DL and Garavito RM (2000) Cyclooxygenases: structural, cellular, and molecular biology. Ann Rev Biochem, 69, 145– 182. Sorli CH, Zhang HJ, Armstrong MB et al (1998) Basal expression of cyclooxygenase-2 and nuclear factor-interleukin 6 are dominant and coordinately regulated by interleukin 1 in the pancreatic islet. Proc Natl Acad Sci USA, 95, 1788–1793. Spencer AG, Woods JW, Arakawa T et al (1998) Subcellular localization of prostaglandin endoperoxide H synthases-1 and -2 by immunoelectron microscopy. J Biol Chem, 273, 9886–9893. Steinbach G, Lynch PM, Phillips RK et al (2000) The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med, 342, 1946–1952. Stichtenoth DO, Thoren S, Bian H et al (2001) Microsomal prostaglandin E synthase is regulated by proinflammatory cytokines and glucocorticoids in primary rheumatoid synovial cells. J Immunol, 167, 469–474. Sun LK, Wahl P, Bilic G and Wu¨thrich RP (2001) CD44-mediated cyclooxygenase-2 expression and thromboxane A2 production in RAW 264.7 macrophages. Inflamm Res, 50, 496–499. Swinney DC, Mak AY, Barnett J and Ramesha CS (1997) Differential allosteric regulation of prostaglandin H synthase 1 and 2 by arachidonic acid. J Biol Chem, 272, 12393–12398. Tang Q, Chen W, Gonzales MS et al (2001) Role of cyclic AMP responsive element in the UVB induction of cyclooxygenase-2 transcription in human keratinocytes. Oncogene, 20, 5164–5172. Tanioka T, Nakatani Y, Semmyo N et al (2000) Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem, 275, 32775–32782. Thomas B, Berenbaum F, Humbert L et al (2000) Critical role of C/EBPD and C/EBPb factors in the stimulation of the cyclooxygenase-2 gene transcription by interleukin-1b in articular chondrocytes. Eur J Biochem, 267, 6798–6809. Ueno N, Murakami M, Tanioka T et al (2001) Coupling between cyclooxygenase, terminal prostanoid synthase, and phospholipase A2. J Biol Chem, 276, 34918–34927. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231, 232–235. Wadleigh DJ and Herschman HR (1999) Transcriptional regulation of the cyclooxygenase-2 gene by diverse ligands in murine osteoblasts. Biochem Biophys Res Commun, 264, 865–870. Wadleigh DJ, Reddy ST, Kopp E et al (2000) Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. J Biol Chem, 275, 6259–6266. Wang HQ, Kim MP, Tiano HF et al (2001) Protein kinase C-a coordinately regulates cytosolic phospholipase A(2) activity and the expression of cyclooxygenase-2 through different mechanisms in mouse keratinocytes. Mol Pharmacol, 59, 860–866. Wang Z and Tai HH (1998) Interleukin-1b and dexamethasone regulate gene expression of prostaglandin H synthase-2 via the NFkB pathway in human amnion derived WISH cells. Prostagland Leukotrienes Essent Fatty Acids, 59, 63–69. Wang Z and Tai HH (1999) Cyclic AMP response element mediates dexamethasone-induced suppression of prostaglandin H synthase-2 gene expression in human amnion derived WISH cells. Prostagland, Leukotrienes Essent Fatty Acids, 60, 243–248. Whiteley PJ and Needleman P (1984) Mechanism of enhanced fibroblast arachidonic acid metabolism by mononuclear cell factor. J Clin Invest, 74, 2249–2253. Wong WYL, DeWitt DL, Smith WL and Richards JS (1989) Rapid induction of prostaglandin endoperoxide synthase in rat preovulatory follicles by luteinizing hormone and cAMP is blocked by inhibitors of transcription and translation. Mol Endocrinol, 3, 1714–1723.
52
BIOSYNTHESIS AND METABOLISM
Xie W and Herschman HR (1995) v-src Induces prostaglandin synthase 2 gene expression by activation of the c-Jun N-terminal kinase and the cJun transcription factor. J Biol Chem, 270, 27622–27628. Xie W and Herschman HR (1996) Transcriptional regulation of prostaglandin synthase 2 gene expression by platelet derived growth factor and serum. J Biol Chem, 271, 31742–31748. Xie W, Chipman JG, Robertson DL et al (1991) Expression of a mitogenresponsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA, 88, 2692–2696. Xie W, Fletcher BS, Andersen RD and Herschman HR (1994) v-src Induction of the TIS10/PGS2 prostaglandin synthase gene is mediated by an ATF/CRE transcription response element. Mol Cell Biol, 14, 6531–6539.
Yamamoto K, Arakawa T, Ueda N and Yamamoto S (1995) Transcriptional roles of nuclear factor kB and nuclear factorinterleukin-6 in the tumor necrosis factor a-dependent induction of cyclooxygenase-2 in MC3T3-E1 cells. J Biol Chem, 270, 31315–31320. Yokota K, Kusaka M, Ohshima T et al (1986) Stimulation of prostaglandin E2 synthesis in cloned osteoblastic cells of mouse (MC3T3-E1) by epidermal growth factor. J Biol Chem, 261, 15410– 15415. Yokoyama C and Tanabe T (1989) Cloning of human gene encoding prostaglandin endoperoxide synthase and primary structure of the enzyme. Biochem Biophys Res Commun, 165, 888–894. Yokoyama C, Takai T and Tanabe T (1988) Primary structure of sheep prostaglandin endoperoxide synthase deduced from cDNA sequence. FEBS Lett, 231, 347–351.
5 Mammalian Lipoxygenases Shozo Yamamoto1, Hiroshi Suzuki2, Natsuo Ueda3, Yoshitaka Takahashi4 and Tanihiro Yoshimoto4 1Kyoto
Women’s University, Kyoto; 2Tokushima University School of Medicine, Tokushima; 3Kagawa University School of Medicine, Kagawa; and 4Kanazawa University Graduate School of Medicine, Kanazawa, Japan
LIPOXYGENASE ENZYMES Lipoxygenase enzymes recognize the 1-cis,4-cis-pentadiene structure of unsaturated fatty acids such as arachidonic, linoleic and linolenic acids, and incorporate one molecule of oxygen into the polyenoic acids to produce corresponding hydroperoxy acids. The enzymes belong to the group of dioxygenases. In the 1960s the enzyme was found in soybean, and referred to as lipoxidase at that time; it was believed that there was no lipoxidase in animal tissues (Tappel 1963). Along with the discovery of bioactive thromboxane in 1974, however, an enzyme which oxygenated the position 12 of arachidonic acid was found in human platelets (Hamberg and Samuelsson 1974) and in animal platelets (Nugteren 1975). The enzyme was referred to as arachidonate 12-lipoxygenase, and it was clearly demonstrated that lipoxygenase enzymes were also present in animals. Furthermore, a protein that was isolated from rabbit reticulocytes and inhibited mitochondrial respiration was found to be a lipoxygenase (Rapoport et al 1979). The reticulocyte enzyme was later identified as 15-lipoxygenase upon reaction with arachidonic acid (Bryant et al 1982). Upon the discovery of bioactive leukotrienes, 5-lipoxygenase was found in rabbit polymorphonuclear leukocytes (Borgeat et al 1976). The 8lipoxygenase activity was detected in mouse skin (Gschwendt et al 1986). Fatty acid cyclooxygenase was also shown to be a lipoxygenase in terms of its mechanism of arachidonate oxygenation (Hamberg and Samuelsson 1967b). Figure 5.1 summarizes the oxygenation reactions catalysed by these lipoxygenases. The oxygenation site of the lipoxygenase substrate is different from enzyme to enzyme. Each enzyme is named on the basis of the oxygenation site upon reaction with arachidonic acid, which has 20 carbon atoms; thus, ‘‘arachidonate X-lipoxygenase’’ oxygenates position X of arachidonic acid as substrate, counted starting from its carboxylic carbon. For reference citation in this chapter, review articles are cited in most cases. If necessary, the readers are advised to see the original articles cited in these review articles.
REACTION MECHANISM OF SOYBEAN LIPOXYGENASE In addition to the regiospecific oxygenation by lipoxygenase enzyme, as described above, the stereospecific oxygenation initiated by stereoselective hydrogen abstraction is characteristic of lipoxygenase. This was first shown by an isotope tracer experiment of soybean lipoxygenase using [13DR-3H] and [13LSThe Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
3
H] 8,11,14-eicosatrienoic acid (Hamberg and Samuelsson 1967a). As illustrated in Figure 5.2, one of the two hydrogen atoms (13proS hydrogen) of the methylene group close to the oxygenation site is stereoselectively eliminated, followed by migration of a double bond and formation of a conjugated diene. The resultant radical is attacked by an oxygen molecule, and the final product is a stereospecific hydroperoxy acid, i.e. 15S-hydroperoxy-8-cis,11cis,13-trans-eicosatrienoic acid. Soybean lipoxygenase contains an equimolar amount of iron (Pistorius and Axelrod 1974), and three histidine residues and C-terminal isoleucine are considered to be the ligands, as demonstrated by site-directed mutagenesis (Steczko and Axelrod 1992) and X-ray crystallography (Boyington et al 1993).
ARACHIDONATE 5-LIPOXYGENASE 5-Lipoxygenase initiates the biosynthesis of leukotrienes in leukocytes, and plays important pathophysiological roles in allergy and inflammation (Ford-Hutchinson et al 1994). The enzyme abstracts proS hydrogen from the carbon 7, and oxygenates the carbon 5 of not only arachidonic acid but also 5,8,11-eicosatrienoic acid and 5,8,11,14,17-eicosapentaenoic acid, and produces 5S-hydroperoxy-6-trans,8-cis,11-cis,14-cis-eicosatetraenoic acid (5S-HPETE) from arachidonic acid. However, the enzyme is much less active with linoleic and linolenic acids (Yamamoto 1992). For leukotriene synthesis, the 5S-HPETE produced is further converted to leukotriene A4 with 5,6-epoxide by liberation of one molecule of water (Figure 5.3). Several lines of enzymological and molecular biological evidence indicate that leukotriene A4 synthesis from 5S-HPETE is attributed to the 5-lipoxygenase enzyme. This reaction is initiated by abstraction of proR hydrogen from position 10. Thus, 5-lipoxygenase is a bifunctional enzyme, with 5-oxygenase activity and leukotriene A4 synthase activity (Yamamoto 1991, 1992; Ford-Hutchinson et al 1994). 5-Lipoxygenase contains iron, which is essential for catalytic activity (Ford-Hutchinson et al 1994). Ca2+ is required for 5lipoxygenase activity, and Ca2+-dependent activity is stimulated by the addition of ATP (Yamamoto 1991; Ford-Hutchinson et al 1994; Ra˚dmark 2000). The intracellular function of 5-lipoxygenase is assisted by an 18 kDa arachidonate-binding protein, which is referred to as five-lipoxygenase activating protein (FLAP) and facilitates the supply of arachidonic acid to the enzyme (FordHutchinson et al 1994). Earlier, 5-lipoxygenase was known as a soluble cytosolic protein, and shown to become membrane-bound upon increase of intracellular calcium concentration. Recent
54
BIOSYNTHESIS AND METABOLISM
Figure 5.2 Selective hydrogen abstraction and oxygenation by soybean lipoxygenase
Figure 5.1 Various mammalian lipoxygenases. PG, prostaglandin; HPETE, hydroperoxyeicosatetranoic acid
investigations have demonstrated the occurrence of a large part of the enzyme inside the nucleus of various types of cell (Ra˚dmark 2000; Funk and Chen 2000). Furthermore, involvement of the SH3 binding motif was suggested in the translocation of 5lipoxygenase, and a two-hybrid method has demonstrated the binding of the enzyme to cytoskeletal proteins such as a-actinin and actin (Ra˚dmark 2000). 5-Lipoxygenase cDNAs of various animals have been cloned, and the primary structures of 673- or 674-amino acid enzymes have been deduced. The 5-lipoxygenase gene consists of 14 exons, and is located on human chromosome 10q 11.2 and on murine chromosome 6. The promoter region is rich in GC boxes, and has neither TATA nor CCAAT boxes (Ra˚dmark 1995; Funk 1993, 1996). The 5-lipoxygenase-deficient mice grow normally, and are resistant to the shock induced by platelet-activating factor (Funk 1996; Austin and Funk 1999; Funk and Chen 2000).
Figure 5.3 5-Lipoxygenase catalysis. hydroperoxyeicosatetraenoic acid
LT,
leukotriene;
HPETE,
ARACHIDONATE 12-LIPOXYGENASE As described above, 12-lipoxygenase was first found in platelets, and later in leukocytes and various other animal cells (Yamamoto 1992; Yoshimoto and Yamamoto 1995; Yamamoto et al 1997). According to our earlier observations, there are two types of 12lipoxygenase. The enzymes of human and bovine platelets are almost inactive with linoleic and linolenic acids, while the enzymes of porcine and bovine leukocytes are active with these C18 fatty acids as well as C20 fatty acids such as arachidonic acid (Yamamoto 1992). Thus, platelet and leukocyte 12-lipoxygenases are distinguishable in terms of substrate specificity. The 12-lipoxygenase reaction is initiated by abstraction of proS hydrogen at position 10 of arachidonic acid, followed by the migration of a double bond and the formation of 8-cis,10-trans conjugated diene, and 12S-hydroperoxy-5-cis,8-cis,10-trans,14cis-eicosatetraenoic acid (12S-HPETE) is produced (Yamamoto 1992). Unlike 5-lipoxygenase, the 12S-HPETE produced is not
subjected to conversion to an epoxide of leukotriene type. In addition to 12S-HPETE, as shown in Figure 5.4, leukocyte 12lipoxygenase produces 15S-hydroperoxy-5-cis,8-cis,11-cis,13trans-eicosatetraenoic acid (15S-HPETE) from arachidonic acid in an amount of about 10% of the total products, whereas platelet enzyme produces 15-HPETE in a quantity of only 1–2% of 12HPETE (Yamamoto et al 1997). The 15S-HPETE produced is converted to 14,15-epoxy acid with a conjugated triene of leukotriene type by leukocyte 12-lipoxygenase at a significant rate. In this reaction, proS hydrogen is removed from carbon 10. Furthermore, 15S-HPETE is subjected to 8S-oxygenation and 14R-oxygenation by the same enzyme, and dihydroperoxyeicosatetraenoic acids (diHPETE) are produced (Figure 5.4). Such multifunctional activities of 12-lipoxygenase have been reviewed previously (Yamamoto 1991, 1992; Yamamoto et al 1997).
MAMMALIAN LIPOXYGENASES
55
Figure 5.4 12-Lipoxygenase catalysis. LT, leukotriene; HX, hepoxilin; HPETE, hydroperoxyeicosatetraenoic acid; diHPETE, dihydroperoxyeicosatetraenoic acid.
cDNAs and genomic DNAs of 12-lipoxygenase have been cloned from various animal species (Funk 1993; Yoshimoto and Yamamoto 1995; Funk 1996; Ku¨hn and Thiele 1999; Brash 1999). In terms of the deduced amino acid sequence and the intron size of the enzyme gene, leukocyte 12-lipoxygenases are closer to reticulocyte 15-lipoxygenases than to platelet 12-lipoxygenases. On the basis of these findings, together with the different antigenicities and substrate specificities, we proposed a classification of two isozyme types (platelet type and leukocyte type) for 12-lipoxygenases (Yamamoto et al 1997). As shown in Table 5.1, the platelet type and the leukocyte type are also found in tissues other than platelets and leukocytes, respectively. Tissue distribution of 12-lipoxygenase isozymes has been listed in detail in the review article by Yoshimoto and Yamamoto (1995). Recently, 12lipoxygenases with lower homologies for platelet and leukocyte enzymes have been isolated from human and murine skin and referred to as the ‘‘epidermis type’’ (Ku¨hn and Thiele 1999). As for the regulation of enzyme gene expression, 12-lipoxygenase is not a rapidly inducible enzyme, in contrast to cyclooxygenase-2. The gene structures of various 12-lipoxygenases have been described and discussed in the studies by Yoshimoto
and Yamamoto (1995) and Funk (1996). The expression of human platelet enzyme is negatively regulated by NFkB (Arakawa et al 1995). The mRNA of the platelet-type enzyme in human epidermoid carcinoma cells was induced by epidermal growth factor after a 10 h lag period (Chang et al 1993), and the interaction of c-Jun and Sp1 was shown to be involved in this induction (Chen and Chang 2000). The protein level of leukocytetype enzyme in porcine aortic smooth muscle cells is increased around 6 h after the addition of platelet-derived growth factor, and remains elevated for up to 24 h (Natarajan et al 1996). It was reported that translocation of human platelet 12-lipoxygenase from cytosol to membrane was required for the activation of the enzyme (Ozeki et al 1999). A clinical case with myeloproliferative disease showed a decreased expression of platelet 12-lipoxygenase protein and mRNA (Matsuda et al 1993). The biological functions of 12-lipoxygenase remain unestablished, although many studies have reported various biological activities of 12-HPETE or 12-HETE (Yamamoto et al 1997). Mice lacking leukocyte-type 12-lipoxygenase grow normally and are fertile, with no macroscopic abnormality of the external or internal organs (Funk 1996). The gene disruption resulted in a
Table 5.1 12-Lipoxygenase isozymes Properties Substrate specificity Arachidonic acid Linoleic and linolenic acids Phospholipids Reaction products Amino acid identity with 15-lipoxygenase* Exon-intron structure compared with 15-lipoxygenase* Distribution Human Porcine Bovine Canine Rat Mouse *Rabbit reticulocyte 15-lipoxygenase.
Leukocyte type
Platelet type
Epidermis type
Active Active Active 12-HPETE415-HPETE Higher Similar
Active Almost inactive Almost inactive 12-HPETE4 415-HPETE Lower Different
Active – – – Lower Different
Adrenal Leukocyte, pituitary Leukocyte, trachea, cornea Leukocyte, brain Leukocyte, pineal gland, aorta, lung, pancreas, spleen Leukocyte, pituitary, pineal gland, kidney
Platelet, skin, uterine cervix
Skin
Platelet Platelet Platelet Platelet, skin
Skin
56
BIOSYNTHESIS AND METABOLISM
Figure 5.5 Lipoxin biosynthesis. LT, leukotriene; LX, lipoxin; HPETE, hydroperoxyeicosatetraenoic acid
reduced capacity of the macrophages to oxidize low-density lipoprotein (LDL) (Austin and Funk 1999). Furthermore, the 12lipoxygenase-lacking mice showed diminished atherosclerosis (Cyrus et al 1999). The involvement of LDL receptor-related protein (LRP) in macrophages was suggested in the oxidation process of LDL by macrophage 12-lipoxygenase (Xu et al 2001). The gene disruption of murine platelet-type 12-lipoxygenase did not enhance thromboxane formation, but the platelets were hyperresponsive to ADP-induced aggregation (Austin and Funk 1999). 12-Lipoxygenase is involved in the biosyntheses of two groups of bioactive eicosanoids. As shown in Figure 5.4, hepoxilins A3 and B3 with hydroxyl and epoxy groups are produced from 12HPETE. These compounds show various biological activities related to the release of intracellular calcium and the opening of potassium channels (Pace-Asciak 1994, 1995). Recently it was established that 12-lipoxygenase increased membrane excitability by inhibiting M-type potassium channels without affecting any other ion channels (Takahashi et al 1999). Figure 5.5 shows the biosynthesis of lipoxins A4 and B4 with a conjugated tetraene and 3 hydroxyl groups. These compounds may function as endogenous braking signals in host defence, inflammation and hypersensitivity. Platelet 12-lipoxygenase is involved in the transcellular synthesis of lipoxins from arachidonic acid via leukotriene A4 (Serhan 1994; Brady and Serhan 1996). In the 1980s the lesional skin of psoriasis patients was reported to contain an increased amount of 12R-HETE rather than 12SHETE. The 12R-HETE production was attributed presumably to cytochrome P-450 (Woollard 1986). Recently, cDNA of a new 12lipoxygenase was cloned from human skin, and the recombinant enzyme produced 12R-HPETE rather than 12S-HPETE. Thus, the enzyme was referred to as 12R-lipoxygenase (Boeglin et al
1998). cDNA of the same enzyme was also cloned from a human B cell line, CCL-156, and its gene was located on chromosome 17p13. A mouse orthologue was also cloned (Sun et al 1998). ARACHIDONATE 15-LIPOXYGENASE The 15-lipoxygenase reaction is initiated by abstraction of proS hydrogen from position 13 of arachidonic acid, followed by the formation of 11-cis,13-trans conjugated diene, and 15S-HPETE is produced. As demonstrated with the rabbit reticulocyte enzyme (Bryant et al 1985), 15-lipoxygenase is also a multi-functional enzyme. The same enzyme transforms 15S-HPETE to 14,15epoxy acid with a conjugated triene (14,15-leukotriene A4) as illustrated in Figure 5.6. In this reaction, proS hydrogen is abstracted from position 10 (Yamamoto 1992). In addition to 14,15-leukotriene A4 production, the enzyme oxygenates position 5 or 8 of 15S-HPETE, and produces 5S,15S- and 8S,15SdiHPETEs. 15-Lipoxygenase of reticulocytes has also 12-lipoxygenase activity, and 12S-HPETE is produced in one-ninth the amount of 15S-HPETE (Ku¨hn et al 1990). Linoleic and linolenic acids as substrates of the enzyme are more active than arachidonic acid (Ku¨hn et al 1990). Recently, a new 15-lipoxygenase was found and its cDNA was cloned from human hair root. The enzyme was also expressed in normal prostate epithelium, but the expression level was reduced in prostate adenocarcinoma (Shappell et al 1999). The primary structure of the 676 amino acid enzyme has 40% identity to the known 15-lipoxygenase. With the recombinant enzyme, linoleic acid as substrate was less active than arachidonic acid (Brash et al 1997). As mentioned above, 15-lipoxygenase has a broad substrate specificity in terms of the carbon chain length of substrate
MAMMALIAN LIPOXYGENASES
57
Figure 5.6 15-Lipoxygenase catalysis. LT, leukotriene; HPETE, hydroperoxyeicosatetraenoic acid; diHPETE, dihydroperoxyeicosatetraenoic acid
(Yamamoto 1992). Furthermore, the rabbit reticulocyte enzyme is also active with arachidonic acid esterified in phosphatidylcholine, and 12S- and 15S-oxygenated products are found (Murray and Brash 1988). cDNA cloning for 15-lipoxygenases of rabbit reticulocytes, human reticulocytes and human airway epithelium deduced the primary structures of 661 or 662 amino acid enzymes (Ku¨hn and Thiele 1995; Funk 1996). The gene of rabbit 15-lipoxygenase is composed of 14 exons, and its promoter is characteristic of a house-keeping enzyme (Ku¨hn and Thiele 1995). The expression of the enzyme gene is upregulated by interleukin-4 or -13, and suppressed by interferon-g (Ku¨hn and Thiele 1995; Funk 1996). Translation of the mRNA encoding rabbit reticulocyte 15lipoxygenase is inhibited by a 48 kDa protein which binds to a repeated sequence in the 30 -untranslated region (Ostareck-Lederer et al 1994). In connection with 15-lipoxygenase physiology, various biological activities have been reported for 15-HPETE and 15-HETE (FordHutchinson 1991; Ku¨hn and Thiele 1995). A currently active research project with 15-lipoxygenase is its possible role in the pathogenesis of atherosclerosis. As mentioned above, 15-lipoxygenase oxygenates esterified unsaturated fatty acids contained in membranes and lipoproteins (Murray and Brash 1988; Parthasarathy et al 1989; Ku¨hn and Brash 1990). On the basis of these findings and further experimental results, a possible role of 15-lipoxygenase is the oxygenation of LDL, which triggers the production of an atherogenic form of lipoprotein (Steinberg et al 1989; Ku¨hn et al 1994; Harats et al 1995).
ARACHIDONATE 8-LIPOXYGENASE An earlier work reported that an application of phorbol ester to mouse back skin resulted in a production of 8-hydroxy-5-cis,9trans,11-cis,14-cis-eicosatetraenoic acid (8-HETE) from arachidonic acid by a cell-free epidermal preparation (Gschwendt et al 1986; Fu¨rstenberger et al 1991). Later, the metabolite was identified as 8S-HETE (Hughes and Brash 1991). After attempts for a long time to characterize the 8-lipoxygenase, its cDNA encoding 667 amino acids was cloned from the mouse epidermal
library (Jisaka et al 1997; Krieg et al 1998). The amino acid sequence has 78% identity to the new human skin 15-lipoxygenase mentioned above and about 40% identity to mouse 5-lipoxygenase and mouse 12-lipoxygenases of three types. The recombinant enzyme oxygenates arachidonic acid exclusively to 8S-HPETE, but linoleic acid is a poor substrate and is converted to 9Shydroperoxy derivative (Jisaka et al 1997). The enzyme also converts 5S-HPETE to leukotriene A4 at a rate comparable to that of 5-lipoxygenase (Qiao et al 1999). 8S-HETE is one of the ligands for nuclear peroxisome proliferator-activated receptor-a (Brash 1999), although many other compounds, including leukotriene B4 and non-steroidal antiinflammatory drugs (NSAIDs), can activate this receptor (Lehman et al 1997; Funk 2001).
MOLECULAR EVOLUTION OF LIPOXYGENASES Since the time when 5-, 12- and 15-lipoxygenases and cyclooxygenase were subjected to computer-assisted sequence comparison by Sigal (1991) and Toh et al (1992), many versions of the lipoxygenase phylogenetic tree have been presented (Funk 1996; Ku¨hn and Thiele 1999; Brash 1999; Kikuno et al 1999) and the tree has grown bigger and bigger. According to these analyses, plant lipoxygenases are distantly related to mammalian lipoxygenases, but the C-terminal portion containing all the ligand residues for the catalytically essential iron (Boyington et al 1993) is well conserved between soybean lipoxygenase and mammalian lipoxygenases. In terms of the amino acid sequence identity, 5lipoxygenase is distinct from other mammalian lipoxygenases. As we demonstrated previously (Yamamoto 1992), the platelet type and the leukocyte type of 12-lipoxygenase are distinguishable. Moreover, in the phylogenetic tree there is a closer relationship between leukocyte 12-lipoxygenases and reticulocyte 15-lipoxygenases rather than platelet 12-lipoxygenase. This finding agrees with their similar catalytic activities, described above. Therefore, 15-lipoxygenase of the reticulocyte type and 12-lipoxygenase of the leukocyte type in combination are referred to as 12/15lipoxygenase by some investigators (Funk 1996; Ku¨hn and Thiele 1999). However, this terminology may be misleading and it should
58
BIOSYNTHESIS AND METABOLISM
be noted that the major product of leukocyte 12-lipoxygenase is 12-HPETE, and that of reticulocyte 15-lipoxygenase is 15HPETE. As mentioned above, epidermis-type 12-lipoxygenase was proposed as the third isozyme type (Ku¨hn and Thiele 1999). CONCLUSION Various mammalian lipoxygenases have been investigated extensively and intensively, and their molecular and catalytic properties are now known in detail. However, except for the leukotrieneproducing 5-lipoxygenase, their physiological and pathological roles have not yet been established. Ku¨hn and Thiele (1999) have stated: ‘‘many sequence data but little information on biological significance’’. To establish the biological significance of lipoxygenases, we must demonstrate specific biological activity associated with a lipoxygenase metabolite of arachidonic acid and other polyunsaturated fatty acids, with stereoselective structure produced in various (not one single) animal species. REFERENCES Arakawa T, Nakamura M, Yoshimoto T and Yamamoto S (1995) The transcriptional regulation of human arachidonate 12-lipoxygenase gene by NFkB/Rel. FEBS Lett, 363, 105–110. Austin SC and Funk CD (1999) Insight into prostaglandin, leukotriene, and other eicosanoid functions using mice with targeted gene disruption. Prostagland Other Lipid Mediators, 58, 231–252. Boeglin WE, Kim RB and Brash AR (1998) A 12R-lipoxygenase in human skin: mechanistic evidence, molecular cloning, and expression. Proc Natl Acad Sci USA, 95, 6744–6749. Borgeat P, Hamberg M and Samuelsson B (1976) Transformation of arachidonic acid and homo-g-linolenic acid by rabbit polymorphonuclear leukocytes. Monohydroxy acids from novel lipoxygenases. J Biol Chem, 251, 7816–7820. Boyington JC, Gaffney BJ and Amzel LM (1993) The three-dimensional structure of an arachidonic acid 15-lipoxygenase. Science, 260, 1482– 1486. Brady HR and Serhan CN (1996) Lipoxins: putative braking signals in host defense, inflammation and hypersensitivity. Curr Opin Nephrol Hypertens, 5, 20–27. Brash AR (1999) Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J Biol Chem, 274, 23679–23682. Brash AR, Boeglin WE and Chang MS (1997) Discovery of a second 15Slipoxygenase in humans. Proc Natl Acad Sci USA, 94, 6148–6152. Bryant RW, Bailey JM, Schewe T and Rapoport SM (1982) Positional specificity of a reticulocyte lipoxygenase. Conversion of arachidonic acid to 15S-hydroperoxy-eicosatetraenoic acid. J Biol Chem, 257, 6050–6055. Bryant RW, Schewe T, Rapoport SM and Bailey JM (1985) Leukotriene formation by a purified reticulocyte lipoxygenase enzyme. Conversion of arachidonic acid and 15-hydroperoxyeicosatetraenoic acid to 14,15leukotriene A4. J Biol Chem, 260, 3548–3555. Chang W-C, Liu Y-W, Ning C-C et al (1993) Induction of arachidonate 12-lipoxygenase mRNA by epidermal growth factor in A431 cells. J Biol Chem, 268, 18734–18739. Chen B-K and Chang WC (2000) Functional interaction between c-Jun and promoter factor Sp1 in epidermal growth factor-induced gene expression of human 12(S)-lipoxygenase. Proc Natl Acad Sci USA, 97, 10406–10411. Cyrus T, Witztum JL, Rader DJ et al (1999) Disruption of the 12/15lipoxygenase gene dimisishes atherosclerosis in apo E-deficient mice. J Clin Invest, 103, 1597–1604. Ford-Hutchinson AW (1991) Arachidonate 15-lipoxygenase; characteristics and potential biological significance. Eicosanoids, 4, 65–74. Ford-Hutchinson AW, Gresser M and Young RN (1994) 5-Lipoxygenase. Annu Rev Biochem, 63, 383–417. Funk CD (1993) Molecular biology in the eicosanoid field. Prog Nucleic Acid Res Mol Biol, 45, 67–98.
Funk CD (1996) The molecular biology of mammalian lipoxygenases and the quest for eicosanoid functions using lipoxygenase-deficient mice. Biochim Biophys Acta, 1304, 65–84. Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science, 294, 1871–1875. Funk CD and Chen X-S (2000) 5-Lipoxygenase and leukotrienes. Transgenic mouse and nuclear targeting studies. Am J Respir Crit Care Med, 161, S120–S124. Fu¨rstenberger G, Hagedorn H, Jacobi T et al (1991) Characterization of an 8-lipoxygenase activity induced by the phorbol ester tumor promoter 12-O-tetradecanoylphorbol-13-acetate in mouse skin in vivo. J Biol Chem, 266, 15738–15745. Gschwendt M, Fu¨rstenberger G, Kittstein W et al (1986) Generation of the arachidonic acid metabolite 8-HETE by extracts of mouse skin treated with phorbol ester in vivo; identification by 1H n.m.r. and GC– MS spectroscopy. Carcinogenesis, 7, 449–455. Hamberg M and Samuelsson B (1967a) On the specificity of the oxygenation of unsaturated fatty acids catalyzed by soybean lipoxidase. J Biol Chem, 242, 5329–5335. Hamberg M and Samuelsson B (1967b) On the mechanism of the biosynthesis of prostaglandins E1 and F1a. J Biol Chem, 242, 5336– 5343. Hamberg M and Samuelsson B (1974) Prostaglandin endoperoxides. Novel transformations of arachidonic acid in human platelets. Proc Natl Acad Sci USA, 71, 3400–3404. Harats D, Mulkins MA and Sigal E (1995) A possible role for 15lipoxygenase in atherogenesis. Trends Cardiovasc Med, 5, 29–36. Hughes MA and Brash AR (1991) Investigation of the mechanism of biosynthesis of 8-hydroxyeicosatetraenoic acid in mouse skin. Biochim Biophys Acta, 1081, 347–354. Jisaka M, Kim RB, Boeglin WE et al (1997) Molecular cloning and functional expression of a phorbol ester-inducible 8S-lipoxygenase from mouse skin. J Biol Chem, 272, 24410–24416. Kikuno R, Daiyasu H and Toh H (1999) Molecular evolution of proteins involved in the arachidonic acid cascade. In Comprehensive Natural Products Chemistry, vol 1: Polyketides and Other Secondary Metabolites Including Fatty Acids and Their Derivatives, Sankawa U (ed.) Elsevier, Amsterdam, 273–284. Krieg P, Kinzig A, Heidt M et al (1998) cDNA cloning of a 8-lipoxygenase and a novel epidermis-type lipoxygenase from phorbol ester-treated mouse skin. Biochim Biophys Acta, 1391, 7–12. Ku¨hn H and Brash AR (1990) Occurrence of lipoxygenase products in membranes of rabbit reticulocytes. Evidence for a role of the reticulocyte lipoxygenase in the maturation of red cells. J Biol Chem, 265, 1454–1458. Ku¨hn H and Thiele B-J (1995) Arachidonate 15-lipoxygenase. J Lipid Mediators Cell Signalling, 12, 157–170. Ku¨hn H and Thiele BJ (1999) The diversity of the lipoxygenase family. Many sequence data but little information on biological significance. FEBS Lett, 449, 7–11. Ku¨hn H, Sprecher H and Brash AR (1990) On singular or dual positional specificity of lipoxygenase. The number of chiral products varies with alignment of methylene groups at the active site of the enzyme. J Biol Chem, 265, 16300–16305. Ku¨hn H, Belkner J, Zaiss S et al (1994) Involvement of 15-lipoxygenase in early stages of atherogenesis. J Exp Med, 179, 1903–1911. Lehmann JM, Lenhard JM, Oliver BB et al (1997) Peroxisome proliferator-activated receptors a and g are activated by indomethacin and other non-steroidal anti-inflammatory drugs. J Biol Chem, 272, 3406–3410. Matsuda S, Murakami J, Yamamoto Y et al (1993) Decreased messenger RNA of arachidonate 12-lipoxygenase in platelets of patients with myeloproliferative disorders. Biochim Biophys Acta, 1180, 243–249. Murray JJ and Brash AR (1988) Rabbit reticulocyte lipoxygenase catalyzes specific 12(S) and 15(S) oxygenation of arachidonoylphosphatidylcholine. Arch Biochem Biophys, 265, 514–523. Natarajan R, Bai W, Rangarajan V et al (1996) Platelet-derived growth factor BB mediated regulation of 12-lipoxygenase in porcine aortic smooth muscle cells. J Cell Physiol, 169, 391–400. Nugteren DH (1975) Arachidonate lipoxygenase in blood platelets. Biochim Biophys Acta, 380, 299–307. Ostareck-Lederer A, Ostareck DH, Standart N and Thiele BJ (1994) Translation of 15-lipoxygenase mRNA is inhibited by a protein that
MAMMALIAN LIPOXYGENASES binds to a repeated sequence in the 30 untranslated region. EMBO J, 13, 1476–1481. Ozeki Y, Nagamura Y, Ito H et al (1999) An anti-platelet agent, OPC29030, inhibits translocation of 12-lipoxygenase and 12hydroxyeicosatetraenoic acid production in human platelets. Br J Pharmacol, 128, 1699–1704. Pace-Asciak CR (1994) Hepoxilins: a review on their cellular actions. Biochim Biophys Acta, 1215, 1–8. Pace-Asciak CR, Reynaud D and Demin PM (1995) Hepoxilins: a review on their enzymatic formation, metabolism and chemical synthesis. Lipids, 30, 107–114. Parthasarathy S, Wieland E and Steinberg D (1989) A role for endothelial cell lipoxygenase in the oxidative modification of low density lipoprotein. Proc Natl Acad Sci USA, 86, 1046–1050. Pistorius EK and Axelrod B (1974) Iron, an essential component of lipoxygenase. J Biol Chem, 249, 3183–3186. Qiao N, Takahashi Y, Takamatsu H and Yoshimoto T (1999) Leukotriene A synthase activity of purified mouse skin arachidonate 8-lipoxygenase expressed in Escherichia coli. Biochim Biophys Acta, 1438, 131–139. Ra˚dmark O (1995) Arachidonate 5-lipoxygenase. J Lipid Mediators Cell Signalling, 12, 171–184. Ra˚dmark O (2000) The molecular biology and regulation of 5lipoxygenase. Am J Respir Crit Care Med, 161, S11–S15. Rapoport SM, Schewe T, Wiesner R et al (1979) The lipoxygenase of reticulocytes. Purification, characterization and biological dynamics of the lipoxygenase; its identity with the respiratory inhibitors of the reticulocyte. Eur J Biochem, 96, 545–561. Serhan CN (1994) Lipoxin biosynthesis and its impact in inflammatory and vascular events. Biochim Biophys Acta, 1212, 1–25. Shappell SB, Boeglin WE, Olson SJ et al (1999) 15-Lipoxygenase-2 (15LOX-2) is expressed in benign prostatic epithelium and reduced in prostate adenocarcinoma. Am J Pathol, 155, 235–245. Sigal E (1991) The molecular biology of mammalian arachidonic acid metabolism. Am J Physiol, 260, L13–L28.
59
Steczko J and Axelrod B (1992) Identification of the iron-binding histidine residues in soybean lipoxygenase L-1. Biochem Biophys Res Commun, 186, 686–689. Steinberg D, Parthasarathy S, Carew TE et al (1989) Beyond cholesterol. Modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med, 320, 915–924. Sun D, McDonnell M, Chen X-S et al (1998) Human 12(R)-lipoxygenase and the mouse ortholog. Molecular cloning, expression, and gene chromosomal assignment. J Biol Chem, 273, 33540–33547. Takahashi Y, Kawajiri H, Yoshimoto T et al (1999) 12-Lipoxygenase overexpression in rodent NG108-15 cells enhances membrane excitability by inhibiting M-type K+ channels. J Physiol, 521.3, 567– 574. Tappel AL (1963) Lipoxidase. In The Enzymes, vol 8, Boyer PD, Lardy H and Myrba¨ck K (eds). Academic Press, New York, 275–283. Toh H, Yokoyama C, Tanabe T et al (1992) Molecular evolution of cyclooxygenase and lipoxygenase. Prostaglandins, 44, 291–315. Woollard PM (1986) Stereochemical difference between 12-hydroxy5,8,10,14-eicosatetraenoic acid in platelets and psoriatic lesions. Biochem Biophys Res Commun, 136, 169–176. Xu W, Takahashi Y, Sakashita T et al (2001) Low density lipoprotein receptor-related protein is required for macrophage-mediated oxidation of low density lipoprotein by 12/15-lipoxygenase. J Biol Chem, 276, 35454–36459. Yamamoto S (1991) ‘‘Enzymatic’’ lipid peroxidation: reactions of mammalian lipoxygenases. Free Rad Biol Med, 10, 149–159. Yamamoto S (1992) Mammalian lipoxygenases: molecular structures and functions. Biochim Biophys Acta, 1128, 117–131. Yamamoto S, Suzuki H and Ueda N (1997) Arachidonate 12lipoxygenases. Prog Lipid Res, 36, 23–41. Yoshimoto T and Yamamoto S (1995) Arachidonate 12-lipoxygenase. J Lipid Mediators Cell Signalling, 12, 195–212.
6 Biosynthesis and Biological Effects of 5-oxo-ETE and Other Oxoeicosatetraenoic Acids William S. Powell McGill University, Montreal, QC, Canada
INTRODUCTION The eicosanoids comprise a large number of 20-carbon fatty acids, many of which contain oxo groups. Oxoeicosanoids can be formed in various ways, including the PG synthase-catalysed rearrangement of PGH2 to form PGE2 and PGD2, the nonenzymatic degradation of hydroperoxyeicosanoids, which is enhanced by the presence of haem compounds, and the dehydrogenase-catalysed oxidation of hydroxyeicosanoids. This chapter will deal with the formation of oxoeicosatetraenoic acids (oxo-ETEs) from hydroxyeicosatetraenoic acids (HETEs) by specific dehydrogenases and the biological activities of the resulting products. 15-Hydroxyprostaglandin dehydrogenases, the first such eicosanoid-metabolizing enzymes to be discovered, convert prostanoids and other 15-hydroxyeicosanoids to their biologically inactive 15-oxo metabolites. 12-Hydroxyeicosanoid dehydrogenases convert LTB4, 12-HETE and similar compounds to 12-oxoeicosanoids, whereas 5-hydroxyeicosanoid dehydrogenase (5h-dh) converts 5-HETE and other 5-hydroxyeicosanoids to their 5-oxo metabolites. The latter pathways will be the major focus of this chapter, as they give rise to the potent inflammatory mediator, 5-oxo-ETE.
15-OXOEICOSANOIDS Formation of 15-Oxoeicosanoids The 15-hydroxy fatty acid precursors of 15-oxoeicosanoids can be synthesized by both cyclooxygenases and lipoxygenases. The major initial step in the metabolism of prostaglandins is oxidation of the 15-hydroxyl group by a cytosolic 15-hydroxyprostaglandin dehydrogenase (15h-PG dh), resulting, in most cases, in their biological inactivation (Okita and Okita 1996). These enzymes (referred to here collectively as 15h-PG dh) are found in a wide variety of tissues, most notably the lung (Anggard et al 1971). In addition to prostaglandins, 15h-PG dh oxidizes a variety of other eicosanoids containing a hydroxyl group in the 15-position. The 15-lipoxygenase (15-LO) products 15-HETE, 5,15-diHETE and 8,15-diHETE (Bergholte et al 1987) as well as LXA4 (Serhan et al 1993) are all excellent substrates, being converted to their 15-oxo metabolites. This enzyme also metabolizes shorter chain fatty acids with hydroxyl groups in the o6-position, including the cyclooxygenase product 12-HHTrE (Liu et al 1985) and the lipoxygenase product 13-hydroxy-9,11-octadecadienoic acid (13HODE) (Agins et al 1987) to 12-oxo-5,8,11-heptadecatrienoic The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
acid and 13-oxo-9,11-octadecadienoic acid (13-oxo-ODE), respectively. Other dehydrogenases also metabolize o6-hydroxy fatty acids. An NAD+-dependent dehydrogenase in rat liver cytosol converts 13-HODE to 13-oxo-ODE (Bronstein and Bull 1997). This enzyme appears to be relatively specific, as the only other substrates found to be metabolized to an appreciable extent were 9-HODE and 15-HETE. In contrast to both the above enzyme and 15h-PG dh, which are found in the cytosol, a microsomal 15-HETE dehydrogenase has been reported in mouse liver (Sok et al 1988). Metabolism of 15-Oxoeicosanoids As with 15-oxoprostaglandins, uncyclized 15-oxoeicosanoids are metabolized by D13-prostaglandin reductase. 5-Hydroxy-15-oxoETE is converted to 5-hydroxy-15-oxo-13,14-dihydro-ETE by a NADPH-dependent cytosolic reductase in human neutrophils (Berhane et al 1998). LXA4 is converted to an analogous dihydro metabolite by monocytes, following initial oxidation of the 15hydroxyl group (Serhan et al 1993). It is likely that 15-oxo-ETE would be metabolized in a similar fashion, but this would be accompanied by the loss of the UV-absorbing chromophore, which makes this reaction more difficult to monitor by HPLC. 12-OXOEICOSANOIDS Formation of 12-oxo eicosanoids 12-Oxoeicosanoids are formed from 12-hydroxy or 12-hydroperoxy precursors, which in turn are synthesized by 12-lipoxygenases. Several such enzymes exist, including the highly specific platelet 12-LO and the less specific leukocyte 12-LO, which both convert arachidonic acid to 12S-HETE (Brash 1999). In addition, a 12Rlipoxygenase has recently been identified in human skin (Boeglin et al 1998). 12-Oxo-ETE can be formed both enzymatically from 12S-HETE or 12R-HETE or non-enzymatically by decomposition of 12-HPETE. In contrast to 15h-PG dh, 12-hydroxyeicosanoid dehydrogenases have been characterized only relatively recently. The first clue for the enzymatic formation of 12-oxoeicosanoids came from the identification of 10,11-dihydro metabolites of 12HETE and LTB4 in incubations with leukocytes (Powell 1987; Wainwright et al 1990) and corneal microsomes (Murphy et al 1988). Porcine leukocytes convert 12S-HETE to 10,11-dihydro
62
BIOSYNTHESIS AND METABOLISM
and 10,11-dihydro-12-oxo metabolites. Initial mass spectrometric studies suggested that the 10,11-dihydro metabolite was formed from deuterium-labelled 12-HETE with partial retention of the deuterium at C12, suggesting direct reduction of the 10,11-double bond by an olefin reductase. However, subsequent experiments clearly showed that the initial step in the formation of this compound was oxidation of the 12-hydroxyl group of 12-HETE to give 12-oxo-5,8,10,14-eicosatetraenoic acid (12-oxo-ETE) (Wainwright and Powell 1991). Unlike cytosolic 15h-PG dh, 12h-dh resides in the microsomal fraction of neutrophils and requires NAD+ as a co-factor (Wainwright and Powell 1991). This enzyme also converts LTB4 to its 12-oxo metabolite. It appears to be relatively specific for a hydroxyl group in the 12-position, but may also exhibit some activity towards 15-HETE and 13-HODE (Wainwright et al 1990) Interestingly, 12h-dh may not be stereospecific, as porcine neutrophil microsomes convert both 12S-HETE and 12R-HETE to 12-oxo-ETE at similar rates. Furthermore, 12S-HETE and LTB4, which have opposite configurations at C12, are equally good substrates for this enzyme (Wainwright and Powell 1991). On the other hand, it is possible that this apparent lack of specificity could be due to the presence of multiple 12h-dhs in neutrophils. Corneal microsomes appear to contain a similar dehydrogenase that oxidizes 12R-HETE (Nishimura et al 1991). The neutrophil 12h-dh is quite distinct from LTB4 12hydroxy-dehydrogenase, which oxidizes the 12R-hydroxyl group of LTB4 and is found in the kidney (Yokomizo et al 1993). The kidney enzyme differs from the neutrophil dehydrogenase in that it is found in the cytosol, requires NADP+ as a co-factor, does not metabolize 12-HETE (Yokomizo et al 1993) and also has 15-oxoPG D13 reductase activity (Ensor et al 1998). 12-Oxo-ETE can also be formed non-enzymatically from 12HPETE by cells that contain 12-LO. This reaction has been investigated in platelets, and is thought to occur via the haemcatalysed loss of water from the hydroperoxide (Fruteau de Laclos et al 1987). 12-oxo-ETE has also been identified as a metabolite of arachidonic acid in nervous tissue from the marine mollusc Aplysia californica and, as in platelets, was thought to have arisen from 12-HPETE (Piomelli et al 1988). Metabolism of 12-oxo-ETE As with 12-oxo-LTB4, 12-oxo-ETE is metabolized by a D10reductase to 10,11-dihydro-12-oxo-ETE (i.e. 12-oxo-5,8,14-eicosatrienoic acid; 12-oxo-ETrE), which is further metabolized to 10,11-dihydro-12-HETE (i.e. 12-hydroxy-5,8,14-eicosatrienoic acid; 12-HETrE). The major product of metabolism of 12S-HETE by porcine neutrophils is 12R-HETrE. Although the 12-hydroxyl group of the product is in the R configuration because of the priority rules, it has the same stereochemistry as the 12S-hydroxyl group of the substrate. 12S-HETrE is also formed from 12S-HETE, but much more slowly than 12R-HETrE. The 12h-dh/D10-reductase pathway has also been found in other types of cells and tissues. Dihydro metabolites of 12-HETE, presumably formed by the metabolism of 12-oxo-ETE, are formed by this or a related process in corneal microsomes (Nishimura et al 1991). Using LTB4 as a substrate, several groups have demonstrated this or a related pathway to exist in a variety of cell types, including mesangial cells, T lymphocytes, macrophages (Kaever et al 1987; Scho¨nfeld et al 1988), monocytes (Fauler et al 1989) and keratinocytes (Wheelan et al 1993), as well as in the lung (Kumlin et al 1990). 12-Oxo-ETE can be directly reduced to 12-HETE by 12ketoreductases in microsomal fractions from rat liver (Falgueyret et al 1988), skin (Falgueyret et al 1990) and leukocytes (Falgueyret et al 1990). In the case of liver microsomes, a mixture of 12R- and
12S-HETE was formed, whereas the skin and leucocyte enzymes converted 12-oxo-ETE stereospecifically to 12S-HETE. It is not known whether or not the 12-ketoreductase and the 12-hydroxydehydrogenase activities are due to the same enzyme. Biological Activity of Metabolites of 12-HETE Formed by the 12h-dh Pathway 12R-HETE, which has the same configuration at C12 as LTB4, stimulates neutrophils by activating the LTB4 receptor (Cunningham and Woollard 1987). Although it induces the same maximal response as LTB4, it is about 250 times less potent in mobilizing calcium in these cells (Table 6.1). 12S-HETE is about 20-fold less potent than 12R-HETE in activating neutrophils. Both 12SHETE and 12R-HETE can be metabolized by the 12h-dh/D10reductase pathway (Wainwright and Powell 1991). However, neither the initial product, 12-oxo-ETE, nor its metabolite 10,11dihydro-12-oxo-ETE (i.e. 12-oxo-ETrE) has appreciable agonist activity in neutrophils (Powell et al 1995c). Reduction of the keto group restores biological activity, with 12S-HETrE and 12RHETrE exhibiting about one-third the potency of the corresponding 12-HETE (Table 6.1). Although the 12h-dh/D10-reductase pathway serves to biologically inactivate 12R-HETE and LTB4 with respect to neutrophils, it may have other functions in other cells. As noted above, 12-oxoETE is formed in Aplysia neural tissue and has been shown to elicit changes in membrane potential in neurons from this species (Piomelli et al 1988). The dihydro metabolite of 12-oxo-ETE, 12R-HETrE, which is formed in the cornea (Murphy et al 1988) and blood vessels, may have an important role as a proinflammatory agent in these tissues (Ramboer et al 1992). This compound has been reported to have vasodilatory effects (Murphy et al 1988) and it is a highly potent angiogenic agent, both in vivo in the rabbit eye (Masferrer et al 1991) and in vitro (Stoltz et al 1996b). This effect may be mediated by a specific receptor in endothelial cells (Stoltz and Schwartzman 1997). 12-HETrE may act through NF-kB in microvessel endothelial cells (Stoltz et al 1996a) and has recently been reported to stimulate VEGF production by these cells (Mezentsev et al 2002). 5-OXOEICOSANOIDS Formation of 5-Oxoeicosanoids Synthesis by 5h-dh As with 12-HETE and 15-HETE, 5-hydroxyeicosanoids are metabolized by a dehydrogenase pathway. Evidence for the existence of such a pathway initially came from studies showing that neutrophils convert 6-trans isomers of LTB4 to dihydro
Table 6.1 EC50 values for the effects of 12R-HETE and 12S-HETE and their metabolites formed by the 12h-dh/D10-reductase pathway on calcium mobilization in human neutrophils Compound LTB4 12R-HETE 12S-HETrE 12S-HETE 12R-HETrE 12-oxo-ETE 12-oxo-ETrE
EC50 (nM) 0.75 190 530 2000 6000 4 410 000 4 410 000
5-OXO-ETE AND OTHER OXO-ETES
Figure 6.1 Relative rates of conversion of different eicosanoids to oxometabolites by human neutrophil microsomes in the presence of NADP+ (1 mM)
Figure 6.2 Regulation of 5-oxo-ETE synthesis in human neutrophils. By stimulating NADPH oxidase, PKC increases the intracellular levels of NADP+, the co-factor required for the synthesis of 5-oxo-ETE by 5h-dh. The effects of PMA are inhibited by Ssp (staurosporine) or DPI (diphenylene iodonium), or by heating neutrophils at 468C for 9 min. NADPH oxidase can be bypassed by addition of PMS (phenazine methosulphate), which directly converts NADPH to NADP+
Figure 6.3 Metabolism of 5-oxo-ETE
63
metabolites (Powell 1986; Powell and Gravelle 1988). Deuteriumlabelling experiments indicated that the formation of these metabolites requires prior oxidation of the 5-hydroxyl group. Interestingly, this reaction is quite specific for 6-trans metabolites of LTB4, as LTB4 itself is not a substrate (Powell et al 1992). It seemed unlikely that the major substrate for 5h-dh would be 6trans isomers of LTB4, as these compounds are formed nonenzymatically and have little biological activity. The most likely candidate for a biologically relevant substrate for this enzyme was 5S-HETE, which has a 5S-hydroxyl group and a 6-trans double bond in common with 6-trans-LTB4. Examination of the substrate specificity for 5h-dh revealed that this enzyme is highly specific for eicosanoids containing a 5S-hydroxyl group followed by a 6-trans double bond, and that 5S-HETE was indeed the best substrate (Figure 6.1). In contrast to 12-hydroxyeicosanoids, which are rapidly metabolized by 12h-dh in unstimulated porcine and rat neutrophils, 5-hydroxyeicosanoids are oxidized to only a small extent by 5h-dh unactivated human neutrophils. Instead, 5-HETE undergoes rapid hydroxylation to 5,20-diHETE (O’Flaherty et al 1986; Powell et al 1994). Activation of neutrophils with PMA or opsonized zymosan results in dramatically enhanced synthesis of 5-oxo-ETE. This effect appears to be mediated by protein kinase C (PKC) stimulation of NADPH oxidase, as it can be blocked by the PKC inhibitor staurosporine and by inhibition of the oxidase, either by gentle heating or by treatment with diphenylene iodonium (DPI) (Figure 6.2) (Powell et al 1994). The stimulatory effect of NADPH oxidase on 5-oxo-ETE production is due to elevation of intracellular levels of NADP+, the co-factor required for its synthesis. This effect can be bypassed by artificially raising intracellular NADP+ levels by addition of phenazine methosulphate, which non-enzymatically converts NADPH to NADP+ (Powell et al 1994). These results suggest that the release of 5-oxoETE is highly regulated, and requires not only activation of PLA2 and 5-LO, but also the adjustment of intracellular co-factor levels to provide an adequate level of NADP+ to permit oxidation of 5-HETE. Other blood cells, including monocytes (Zhang et al 1996), lymphocytes (Zhang et al 1996), eosinophils (Powell et al 1995a) and platelets (Powell et al 1999b), also contain 5h-dh. Of these,
64
BIOSYNTHESIS AND METABOLISM
As with other oxoeicosanoids, 5-oxo-ETE can be generated nonenzymatically from 5-HPETE. 5-oxo-ETE, probably formed by this mechanism, was detected following incubation of supernatants from MC-9 cells with arachidonic acid (Bryant et al 1986). Treatment of red blood cell ghosts with t-butyl hydroperoxide led to the formation of 5-oxo-ETE esterified to phospholipids, which was detected by liquid chromatography–mass spectrometry (Hall and Murphy 1998).
that eosinophils convert arachidonic acid to a lipid chemoattractant that was not identical to LTB4 and had a UV spectrum (Morita et al 1990) similar to that of 5-oxo-15-HETE, which we isolated after incubation of 15-HETE with neutrophil microsomes in the presence of NADP+ (Powell et al 1992). Schro¨der’s group went on to show that 5-oxo-15-HETE, which they synthesized by incubation of arachidonic acid with soybean lipoxygenase, is a potent eosinophil chemoattractant, suggesting that this is the substance responsible for the eosinophil-derived lipid chemoattractant activity that they previously detected. The finding that 5-HETE is converted to 5-oxo-ETE by neutrophils can also explain earlier work by O’Flaherty’s group showing that 5-HETE is a weak neutrophil agonist that acts independently of receptors for other lipid mediators (O’Flaherty 1985; O’Flaherty et al 1988). The relatively low potency of 5-HETE suggested that it may be acting through a receptor for another unknown lipid mediator. From subsequent work (discussed below), it would appear that the biological activity of 5-HETE is due to its interaction with the receptor for the more potent 5oxo-ETE, which is a potent activator of eosinophils and neutrophils and also has effects on other cells including monocytes and epithelial cells.
Metabolism of 5-oxo-ETE
Neutrophils
5-oxo-ETE is metabolized by a variety of pathways, including o-oxidation and reduction of both olefinic and keto groups, as well as by oxidation by lipoxygenases (Figure 6.3). The major pathway for the metabolism of 5-oxo-ETE in neutrophils is ohydroxylation to its 20-hydroxy metabolite (Powell et al 1996), which is presumably catalysed by LTB4 20-hydroxylase. This enzyme is present in large amounts in neutrophils, but is absent from monocytes and other blood cells (Powell 1984). Mouse macrophages, which lack the 20-hydroxylase enzyme, convert 5-oxo-ETE to 18- and 19-hydroxylated products (Hevko et al 2001). Neutrophils also contain a cytosolic D6-reductase that converts 5-oxo-ETE to its 6,7-dihydro metabolite (i.e. 5-oxo-ETrE). This enzyme, which requires NADPH as a co-factor, is clearly distinct from PG D13-reductase because of its requirement for calcium and calmodulin, suggesting that its activity may be tightly regulated (Berhane et al 1998). 5-oxo-ETE can also be converted back to its precursor, 5S-HETE, probably by 5h-dh acting in the reverse direction (Powell et al 1992), and this is the major pathway for the metabolism of 5-oxo-ETE by unstimulated platelets (Powell et al 1999b). In contrast, stimulation of platelets with thrombin, which results in activation of 12-LO, results in the conversion of 5-oxoETE to its 12-hydroxy metabolite, 5-oxo-12-HETE. 5-oxo-ETE is also a substrate for 15-LO, which converts it to 5-oxo-15S-HETE. This may be a significant reaction in eosinophils, which possess high 15-LO activity (Turk et al 1982). 5-oxo-ETE can be converted to a glutathione conjugate, 5-oxo7-glutathionyl-8,11,14-eicosatrienoic acid (FOG7), by mouse macrophages (Bowers et al 2000). This compound, which is biologically active, appears to be formed by LTC4 synthase in these cells (Hevko and Murphy 2002). Other glutathione transferases also convert 5-oxo-ETE to similar products, but in this case they are not biologically active, suggesting that they are inactive isomers of FOG7 (Hevko and Murphy 2002).
5-oxo-ETE is nearly 100 times more potent than its precursor, 5SHETE, in stimulating neutrophil migration and calcium mobilization (Powell et al 1993). It also induces a variety of other responses in neutrophils, including actin polymerization, surface expression of the adhesion molecules CD11b and CD11c, adherence and aggregation (Norgauer et al 1996; O’Flaherty et al 1994; Powell et al 1997). Although 5-oxo-ETE by itself has only modest effects on degranulation and the oxidative burst in neutrophils, it is a strong inducer of these responses in the presence of both platelet-activating factor (PAF) and tumour necrosis factor-a (TNFa) (O’Flaherty et al 1994). Pretreatment of neutrophils with either granulocyte-macrophage colony stimulating factor (GM-CSF) or G-CSF also strongly enhances these responses (O’Flaherty et al 1994). 5-oxo-ETE has been shown to activate certain kinases involved in intracellular signalling in neutrophils. It stimulates both phosphatidylinositol 3-kinase (Norgauer et al 1996) and MAP kinases (ERK-1 and ERK-2) (O’Flaherty et al 1996a) in these cells. It also stimulates phosphorylation of cPLA2 in neutrophils and the release of arachidonic acid, the latter effect being potentiated by pretreatment with low concentrations of GMCSF (O’Flaherty et al 1996a). The 5-oxo-ETE metabolite FOG7 is also capable of activating both neutrophils and eosinophils (Bowers et al 2000). This substance is a chemoattractant for these cells, and induces actin polymerization. However, unlike 5-oxo-ETE, it does not elicit a calcium mobilization response. The nature of the receptor mediating the response to FOG7 is unclear, and may well be the cys-LT1 receptor rather than the 5-oxo-ETE receptor. The effects of 5-oxo-ETE on neutrophils are very similar to those of LTB4. However, in all of the responses tested, LTB4 is about 10–50 times more potent than 5-oxo-ETE, although the maximal responses to the two eicosanoids are similar. As LTB4 and 5-oxo-ETE are likely to be formed concomitantly and in comparable amounts, this raises the question of the physiological relevance of 5-oxo-ETE as a neutrophil agonist. In normal circumstances, it would seem likely that LTB4 is more important than 5-oxo-ETE in regulating neutrophil activity. However, it is possible that in certain pathological situations 5-oxo-ETE may also play a role. For example, the latter compound could be involved in persistent inflammation, in which case neutrophils
monocytes are the most active, producing levels of 5-oxo-ETE comparable to or greater than those of LTB4. Although platelet microsomes rapidly convert 5-HETE to 5-oxo-ETE in the presence of NADP+, intact platelets produce little 5-oxo-ETE, but instead catalyse the reverse reaction (i.e. formation of 5HETE from 5-oxo-ETE) and could thus be involved in the biological inactivation of 5-oxo-ETE. As with neutrophils, conversion of 5-HETE to 5-oxo-ETE by intact platelets is dramatically enhanced by increasing intracellular NADP+ levels by addition of phenazine methosulphate (Powell et al 1999b). Non-enzymatic Synthesis of 5-oxo-ETE
Biological Activity of 5-Oxoeicosanoids The high degree of specificity of 5h-dh suggested that the product of this reaction may have some biological significance. This concept was supported by the finding by Schro¨der’s group
5-OXO-ETE AND OTHER OXO-ETES
65
may become desensitized to LTB4 (Marleau et al 1993). Furthermore, 5-oxo-ETE appears to be metabolized more slowly than LTB4, which is very rapidly converted to o-oxidation products by LTB4 20-hydroxylase in neutrophils (Powell 1984).
C5a, LTB4 and fMLP (O’Flaherty et al 1996b). 5-oxo-ETE also induces phosphorylation of MAP kinases (ERK-1 and ERK-2) in eosinophils, and is much more potent in stimulating this activity in eosinophils than in neutrophils (O’Flaherty et al 1996b).
Eosinophils
Monocytes
In contrast to neutrophils, human eosinophils undergo a much stronger chemotactic response to 5-oxo-ETE than to LTB4, which has only weak effects on the migration of these cells (Powell et al 1995a). The maximal chemotactic response elicited by 5-oxo-ETE is considerably greater than that induced by other lipid mediators, including PAF (O’Flaherty et al 1996b; Powell et al 1995a) and PGD2 (Monneret et al 2001), and is also greater than that of the potent eosinophil-specific chemokine eotaxin (Powell et al 2001). At low concentrations (1–10 nM), 5-oxo-ETE also induces a stronger response than PAF and other lipid mediators, with the exception of PGD2, which has about the same effect on eosinophil migration in this concentration range (Monneret et al 2001). However, in this concentration range eotaxin induces a somewhat stronger response than 5-oxo-ETE (Powell et al 2001). Eotaxin and 5-oxo-ETE appear to have a synergistic effect on eosinophils, as low concentrations of this chemokine induce a leftward shift in the concentration–response curve for 5-oxo-ETE (Powell et al 2001). 5-oxo-15-HETE has a similar effect to 5-oxo-ETE on eosinophil migration (Schwenk and Schro¨der 1995) but is somewhat less potent (Powell et al 1995a). 5-oxo-ETE is also a highly potent stimulator of eosinophil migration through a Matrigel matrix, an effect that may be mediated by proteinase release (Guilbert et al 1999). 5-oxo-ETE is also active in vivo, as it induces infiltration of eosinophils into the lungs of Brown Norway rats following intratracheal administration (Stamatiou et al 1998). This effect is dependent on 5-oxo-ETE itself, as it could not be blocked by antagonists to either PAF or LTB4 and was not shared by cysteinyl leukotrienes. However, it could be blocked by administration of antibodies against either of the adhesion molecules VLA-4 (CD49d) or LFA-1 (CD11a). We have also recently shown that intradermal injection of 5-oxo-ETE induces infiltration of eosinophils into human skin. However, this effect is not specific for eosinophils, as an increase in neutrophils was also observed (Muro et al 2003). In addition to its effects on eosinophil migration, 5-oxo-ETE can also promote eosinophil survival, an effect not shared by the eosinophil chemokines eotaxin and RANTES (regulated upon activation, normal T cell expressed and secreted) (Stamatiou et al 2001). However, unlike its other effects on eosinophils, this effect is indirect, as it was nearly completely dependent on the presence of small numbers of contaminating monocytes. This study suggests that 5-oxo-ETE can stimulate monocytes to release a survival factor for eosinophils (Stamatiou et al 2001). 5-oxo-ETE also induces a variety of rapid responses in eosinophils, including calcium mobilization, actin polymerization, surface expression of CD11b, and shedding of L-selectin (Czech et al 1997; Powell et al 1999a). Interestingly, LTB4 also induces most of these responses in eosinophils, with a potency similar to that of 5-oxo-ETE, in spite of the fact that it is a very weak chemoattractant for human eosinophils. However, LTB4 has only a weak effect on actin polymerization compared to 5-oxoETE, suggesting that this response may correlate more closely than other responses, such as calcium mobilization, to cell migration (Powell et al 1999a). 5-oxo-ETE displays only very weak degranulating activity for eosinophils, but its potency is markedly increased following priming with GM-CSF. Low concentrations of 5-oxo-ETE also dramatically enhance degranulation responses to a variety of other mediators, including PAF,
5-oxo-ETE also induces a chemotactic response in human monocytes (Sozzani et al 1996). Although not as potent as fMLP or MCP-1, it strongly enhances the chemotactic response to both MCP-1 and MCP-3. 5-oxo-ETE also stimulates actin polymerization, but has no effect on intracellular calcium levels in these cells (Sozzani et al 1996). Epithelial cells Guinea-pig jejunal crypt epithelial cells undergo a reduction in cell volume following treatment with 5-oxo-ETE (MacLeod et al 1999). This effect is selective for crypt cells, as villus cells do not respond to this substance. 5-oxo-ETE is much more potent than a variety of other mediators, including LTD4, bradykinin, vasoactive intestinal peptide and 5-HETE. In contrast, LTB4 has no effect on the volumes of these cells. The effect of 5-oxo-ETE is blocked by inhibitors of chloride and potassium channels, suggesting that it is due to activation of ion channels, possibly through activation of protein kinase C. Prostate Cancer Cells The 5-lipoxygenase inhibitors MK-886 and AA861 were found to be strong inducers of apoptosis in PC3 and LNCaP prostate cancer cells (Ghosh and Myers 1997, 1998). This effect was reversed by 5-oxo-ETE and, to a somewhat lesser extent, by 5HETE. In contrast, LTB4 was inactive. These cell lines were reported to constitutively synthesize immunoreactive 5-HETE from endogenous substrate. However, the production of 5-oxoETE itself was not measured directly. Thus, 5-oxo-ETE may be an endogenous survival factor for prostate tumour cells, and inhibition of its synthesis may induce cell death (Ghosh and Myers 1998). The 5-oxo-ETE Receptor The actions of 5-oxo-ETE are mediated by a unique G proteincoupled receptor. 5-oxo-ETE is specifically bound by high affinity binding sites on neutrophils (O’Flaherty et al 1998). Structure– activity studies indicate that this receptor is highly selective for 5oxo-ETE. Biological responses to this eicosanoid are subject to homologous, but not heterologous desensitization, and are not blocked by selective antagonists of other chemoattractant receptors. Structure–Activity Relationships The structural requirements for 5-oxo-ETE-induced calcium mobilization in neutrophils have been investigated in considerable detail (Powell et al 1996). This response was found to be highly selective for 5-oxo-ETE, as minor structural modifications resulted in dramatic losses in biological activity. Interestingly, nearly all of the metabolites of 5-oxo-ETE have much lower biological activities than 5-oxo-ETE itself, with the possible exception of 5-oxo-15-HETE (Figure 6.4). Methylation of the
66
BIOSYNTHESIS AND METABOLISM responses, indicating that its actions are not mediated by the BLT1 receptor (O’Flaherty et al 1993; Powell et al 1993). Similarly, both LTB4 and PAF antagonists had no effect on 5-oxo-ETE-induced eosinophil infiltration into rat lungs, but strongly inhibited eosinophil infiltration induced by LTB4 and PAF (Stamatiou et al 1998).
Coupling to G Proteins The effects of 5-oxo-ETE on leukocytes can be blocked by the Gi protein inhibitor pertussis toxin, suggesting that it acts via a G protein-coupled receptor (Norgauer et al 1996; O’Flaherty et al 2000; Powell et al 1996). In agreement with this, 5-oxo-ETE stimulates GTP/GDP exchange in membrane fractions from neutrophils (O’Flaherty et al 2000). Pertussis toxin can also inhibit the effects of other chemoattractants, but to a lesser extent than those of 5-oxo-ETE, which are completely blocked by this substance (O’Flaherty et al 2000). This suggests that 5-oxo-ETE has an absolute requirement for Gi proteins, whereas other chemoattractants can also signal though other G proteins such as Gq and G12/13. Figure 6.4 Effects of structural modifications of 5-oxo-ETE on biological activity (calcium mobilization). The values were calculated by dividing the EC50 value for 5-oxo-ETE (4 nM) by that for the 5-oxo-ETE analogue, and multiplying by 100
Cloning and expression of the 5-oxo-ETE receptor
carboxyl group results in a 20-fold reduction in potency, whereas reduction of the oxo group to a hydroxyl group, as in its precursor, 5-HETE, results in a nearly 100-fold loss in potency. Reduction of the 6,7-double bond (Berhane et al 1998) results in an even greater loss in potency, whereas isomerization of the 8,9double bond from the cis to the trans configuration results in an approximately six-fold loss in potency (O’Flaherty et al 1994; Powell et al 1996). The metabolite of 5-oxo-ETE formed by the 12-lipoxygenase pathway, 5-oxo-12-HETE, has virtually no calcium-mobilizing activity in neutrophils (Powell et al 1999b). In contrast, the product formed by the 15-lipoxygenase pathway, 5-oxo-15-HETE, has about one-tenth the potency of 5-oxo-ETE (Powell et al 1993; Schwenk and Schro¨der 1995). 5-oxo-EPE, the product analogous to 5-oxo-ETE formed from 5,8,11,14,17eicosapentaenoic acid, has about one-tenth the potency of 5oxo-ETE (Powell et al 1995b). Finally, metabolism of 5-oxo-ETE by the o-oxidation pathway, to 20-hydroxy-5-oxo-ETE, results in an almost 100-fold loss in potency (Powell et al 1996).
CONCLUSIONS
Desensitization to 5-oxo-ETE Desensitization studies strongly support the existence of a specific 5-oxo-ETE receptor. Preincubation of neutrophils (O’Flaherty et al 1993; Powell et al 1993), eosinophils (Powell et al 1999a; Schwenk and Schro¨der 1995) or monocytes (Sozzani et al 1996) with 5-oxo-ETE resulted in complete desensitization to subsequent addition of this substance but did not block responses to other chemoattractants, including LTB4 and PAF. Similarly, initial treatment of leukocytes with other chemoattractants did not prevent 5-oxo-ETE-induced cell activation. Effects of Antagonists on Other Chemoattractant Receptors The effects of 5-oxo-ETE on neutrophils are not prevented by concentrations of LTB4 antagonists that block LTB4-induced
The receptor for 5-oxo-ETE has recently been cloned and expressed by two groups independently, each using a bioinformatics approach (Hosoi et al 2002; Jones et al 2003). In agreement with prior pharmacological data, the cloned receptor is highly selective for 5-oxo-ETE and is coupled to Gi. 5-oxo-ETE induced migration and inhibited forskolin-stimulated cAMP formation in cell transfected with this receptor. The 5-oxo-ETE receptor is very highly expressed in human eosinophils, with smaller amounts being present in neutrophils and macrophages.
In the cases of many eicosanoids, in particular prostaglandins and LTB4, oxidation of one of the hydroxyl groups to a keto group is an important step in their biological inactivation. However, this is not true for 5-HETE and 12-HETE, which can be converted by dehydrogenases to products with potent biological activities. Oxidation of 5S-HETE by 5h-dh results in the formation of 5oxo-ETE, a potent chemoattractant for human eosinophils that may be an important mediator in asthma. The formation of this substance is controlled not only by substrate levels, but also by the availability of the co-factor of the enzyme, NADP+. The biological actions of 5-oxo-ETE are mediated by a highly specific Gi protein-coupled receptor. 12-HETE can also be oxidized to biologically active products. In this case, the most potent product (12-HHTrE) is formed by the further metabolism of the initial oxidation product, 12-oxo-ETE, by a D10-reductase to give an 11,12-dihydro product that is a potent angiogenic agent. Thus, both of these HETE dehydrogenase pathways can give rise to potent biological products that may be important mediators in inflammatory diseases.
REFERENCES Agins AP, Zipkin RE and Taffer IM (1987) Metabolism of cyclooxygenase and lipoxygenase products by 15-prostaglandin dehydrogenase from human HL-60 leukemia cells. Agents Actions, 21, 397–399.
5-OXO-ETE AND OTHER OXO-ETES Anggard E, Larsson C and Samuelsson B (1971) The distribution of 15hydroxy prostaglandin dehydrogenase and prostaglandin-D13-reductase in tissues of the swine. Acta Physiol Scand, 81, 396–404. Bergholte JM, Soberman RJ, Hayes R et al (1987) Oxidation of 15hydroxyeicosatetraenoic acid and other hydroxy fatty acids by lung prostaglandin dehydrogenase. Arch Biochem Biophys, 257, 444–450. Berhane K, Ray AA, Khanapure SP et al (1998) Calcium/calmodulindependent conversion of 5-oxoeicosanoids to 6,7-dihydro metabolites by a cytosolic olefin reductase in human neutrophils. J Biol Chem, 273, 20951–20959. Boeglin WE, Kim RB and Brash AR (1998) A 12R-lipoxygenase in human skin: mechanistic evidence, molecular cloning, and expression. Proc Natl Acad Sci USA, 95, 6744–6749. Bowers RC, Hevko J, Henson PM and Murphy RC (2000) A novel glutathione containing eicosanoid (FOG7) chemotactic for human granulocytes. J Biol Chem, 275, 29931–29934. Brash AR (1999) Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J Biol Chem, 274, 23679–23682. Bronstein JC, Bull AW (1997) Substrate specificity and characterization of partially purified rat liver 13-hydroxyoctadecadienoic acid (13-HODE) dehydrogenase. Arch Biochem Biophys, 348, 219–225. Bryant RW, She HS, Ng KJ and Siegel MI (1986) Modulation of the 5-lipoxygenase activity of MC-9 mast cells: activation by hydroperoxides. Prostaglandins, 32, 615–627. Cunningham FM and Woollard PM (1987) 12(R)-hydroxy-5,8,10,14eicosatetraenoic acid is a chemoattractant for human polymorphonuclear leucocytes in vitro. Prostaglandins, 34, 71–78. Czech W, Barbisch M, Tenscher K et al (1997) Chemotactic 5-oxoeicosatetraenoic acids induce oxygen radical production, Ca2+mobilization, and actin reorganization in human eosinophils via a pertussis toxin-sensitive G-protein. J Invest Dermatol, 108, 108–112. Ensor CM, Zhang H, Tai HH (1998) Purification, cDNA cloning and expression of 15-oxoprostaglandin 13-reductase from pig lung. Biochem J, 330(Pt 1), 103–108. Falgueyret JP, Leblanc Y and Riendeau D (1990) Stereoselective carbonyl reductases from rat skin and leukocyte microsomes converting 12ketoeicosatetraenoic acid to 12(S)-HETE. FEBS Lett, 262, 197–200. Falgueyret JP, Leblanc Y, Rokach J and Riendeau D (1988) NAD(P)Hdependent reduction of 12-ketoeicosatetraenoic acid to 12(R)- and 12(S)-hydroxyeicosatetraenoic acid by rat liver microsomes. Biochem Biophys Res Commun, 156, 1083–1089. Fauler J, Marx KH, Kaever V and Frolich JC (1989) Human monocytes convert leukotriene B4 to two dihydro- leukotriene B4-metabolites. Prostagland Leukotrienes Essent Fatty Acids, 37, 193–196. Fruteau de Laclos B, Maclouf J, Poubelle P et al (1987) Conversion of arachidonic acid into 12-oxo derivatives in human platelets. A pathway possibly involving the heme-catalysed transformation of 12hydroperoxy-eicosatetraenoic acid. Prostaglandins, 33, 315–337. Ghosh J and Myers CE (1997) Arachidonic acid stimulates prostate cancer cell growth: critical role of 5-lipoxygenase. Biochem Biophys Res Commun, 235, 418–423. Ghosh J and Myers CE (1998) Inhibition of arachidonate 5-lipoxygenase triggers massive apoptosis in human prostate cancer cells. Proc Natl Acad Sci USA, 95, 13182–13187. Guilbert M, Ferland C, Bosse M et al (1999) 5-oxo-6,8,11,14Eicosatetraenoic acid induces important eosinophil transmigration through basement membrane components: comparison of normal and asthmatic eosinophils. Am J Resp Cell Mol Biol, 21, 97–104. Hall LM and Murphy RC (1998) Electrospray mass spectrometric analysis of 5-hydroperoxy and 5- hydroxyeicosatetraenoic acids generated by lipid peroxidation of red blood cell ghost phospholipids. J Am Soc Mass Spectrom, 9, 527–532. Hevko JM, Bowers RC and Murphy RC (2001) Synthesis of 5-oxo6,8,11,14-eicosatetraenoic acid and identification of novel o-oxidized metabolites in the mouse macrophage. J Pharmacol Exp Therapeut, 296, 293–305. Hevko JM and Murphy RC (2002) Formation of murine macrophagederived 5-oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG7) is catalyzed by leukotriene C4 synthase. J Biol Chem, 277, 7037–7043. Hosoi T, Koguchi Y, Sugikawa E et al (2002) Identification of a novel eicosanoid receptor coupled to Gi/o. J Biol Chem, 277, 31459–31465. Jones CE, Holden S, Tenaillon L et al (2003) Expression and characterization of a 5-oxo-6E,8Z,11Z,14Z-eicosatetraenoic acid
67
receptor highly expressed on human eosinophils and neutrophils. Mol Pharmacol, 63, 471–477. Kaever V, Martin M, Fauler J et al (1987) A novel metabolic pathway for leukotriene B4 in different cell types: primary reduction of a double bond. Biochim Biophys Acta, 922, 337–344. Kumlin M, Falck JR, Raud J et al (1990) Identification and biological activity of dihydroleukotriene B4: a prominent metabolite of leukotriene B4 in the human lung. Biochem Biophys Res Commun, 170, 23–29. Liu Y, Yoden K, Shen RF and Tai HH (1985) 12-L-hydroxy-5,8,10heptadecatrienoic acid (HHT) is an excellent substrate for NAD+dependent 15-hydroxyprostaglandin dehydrogenase. Biochem Biophys Res Commun, 129, 268–274. MacLeod RJ, Lembessis P, Hamilton JR and Powell WS (1999) 5-oxo6,8,11,14-Eicosatetraenoic acid stimulates isotonic volume reduction of guinea pig jejunal crypt epithelial cells. J Pharmacol Exp Therapeut, 291, 511–516. Marleau S, Fortin C, Poubelle PE and Borgeat P (1993) In vivo desensitization to leukotriene B4 (LTB4) in the rabbit. Inhibition of LTB4-induced neutropenia during intravenous infusion of LTB4. J Immunol, 150, 206–213. Masferrer JL, Rimarachin JA, Gerritsen ME et al (1991) 12(R)hydroxyeicosatrienoic acid, a potent chemotactic and angiogenic factor produced by the cornea. Exp Eye Res, 52, 417–424. Mezentsev A, Seta F, Dunn MW et al (2002) Eicosanoid regulation of VEGF expression and angiogenesis in microvessel endothelial cells. J Biol Chem, 277, 18670–18676. Monneret G, Gravel S, Diamond M et al (2001) Prostaglandin D2 is a potent chemoattractant for human eosinophils that acts via a novel DP receptor. Blood, 98, 1942–1948. Morita E, Schro¨der JM and Christophers E (1990) Identification of a novel and highly potent eosinophil chemotactic lipid in human eosinophils treated with arachidonic acid. J Immunol, 144, 1893–1900. Muro S, Hamid Q, Olivenstein R et al (2003) 5-oxo-6,8,11,14eicosatetraenoic acid induces the infiltration of granulocytes into human skin. J Allergy Clin Immunol, 112, 768–774. Murphy RC, Falck JR, Lumin S et al (1988) 12(R)-Hydroxyeicosatrienoic acid: a vasodilator cytochrome P-450-dependent arachidonate metabolite from the bovine corneal epithelium. J Biol Chem, 263, 17197–17202. Nishimura M, Schwartzman ML, Falck JR et al (1991) Metabolism of 12(R)-hydroxy-5,8,10,14-eicosatetraenoic acid [12(R)-HETE] in corneal tissues: formation of novel metabolites. Arch Biochem Biophys, 290, 326–335. Norgauer J, Barbisch M, Czech W et al (1996) Chemotactic 5-oxoicosatetraenoic acids activate a unique pattern of neutrophil responses—analysis of phospholipid metabolism, intracellular Ca2+ transients, actin reorganization, superoxide-anion production and receptor upregulation. Eur J Biochem, 236, 1003–1009. O’Flaherty JT (1985) Neutrophil degranulation: evidence pertaining to its mediation by the combined effects of leukotriene B4, platelet-activating factor, and 5-HETE. J Cell Physiol, 122, 229–239. O’Flaherty JT, Cordes J, Redman J and Thomas MJ (1993) 5-oxoEicosatetraenoate, a potent human neutrophil stimulus. Biochem Biophys Res Commun, 192, 129–134. O’Flaherty JT, Cordes JF, Lee SL et al (1994) Chemical and biological characterization of oxo-eicosatetraenoic acids. Biochim Biophys Acta, 1201, 505–515. O’Flaherty JT, Jacobson D and Redman J (1988) Mechanism involved in the mobilization of neutrophil calcium by 5-hydroxyeicosatetraenoate. J Immunol, 140, 4323–4328. O’Flaherty JT, Kuroki M, Nixon AB et al (1996a) 5-oxo-Eicosanoids and hematopoietic cytokines cooperate in stimulating neutrophil function and the mitogen-activated protein kinase pathway. J Biol Chem, 271, 17821–17828. O’Flaherty JT, Kuroki M, Nixon AB et al (1996b) 5-oxo-eicosatetraenoate is a broadly active, eosinophil-selective stimulus for human granulocytes. J Immunol, 157, 336–342. O’Flaherty JT, Taylor JS and Kuroki M (2000) The coupling of 5-oxoeicosanoid receptors to heterotrimeric G proteins. J Immunol, 164, 3345–3352. O’Flaherty JT, Taylor JS and Thomas MJ (1998) Receptors for the 5-oxo class of eicosanoids in neutrophils. J Biol Chem, 273, 32535–32541.
68
BIOSYNTHESIS AND METABOLISM
O’Flaherty JT, Wykle R, Redman J et al (1986) Metabolism of 5hydroxyeicosatetraenoate by human neutrophils: production of a novel o-oxidized derivative. J Immunol, 137, 3277–3283. Okita RT and Okita JR (1996) Prostaglandin-metabolizing enzymes during pregnancy: characterization of NAD(+)-dependent prostaglandin dehydrogenase, carbonyl reductase, and cytochrome P450-dependent prostaglandin o-hydroxylase. Crit Rev Biochem Mol Biol, 31, 101–126. Piomelli D, Feinmark SJ, Shapiro E and Schwartz JH (1988) Formation and biological actions of 12-ketoeicosatetraenoic acid in the nervous system of Aplysia. J Biol Chem, 263, 16591–16596. Powell WS (1984) Properties of leukotriene B4 20-hydroxylase from polymorphonuclear leukocytes. J Biol Chem, 259, 3082–3089. Powell WS (1986) Novel pathway for the metabolism of 6-transleukotriene B4 by human polymorphonuclear leukocytes. Biochem Biophys Res Commun, 136, 707–712. Powell WS (1987) Conversion of leukotriene B4 to dihydro- and 19hydroxy metabolites by rat polymorphonuclear leukocytes. Biochem Biophys Res Commun, 145, 991–998. Powell WS, Ahmed S, Gravel S and Rokach J (2001) Eotaxin and RANTES enhance 5-oxo-6,8,11,14-eicosatetraenoic acid-induced eosinophil chemotaxis. J Allergy Clin Immunol, 107, 272–278. Powell WS, Chung D and Gravel S (1995a) 5-oxo-6,8,11,14Eicosatetraenoic acid is a potent stimulator of human eosinophil migration. J Immunol, 154, 4123–4132. Powell WS, Gravel S and Gravelle F (1995b) Formation of a 5-oxo metabolite of 5,8,11,14,17-eicosapentaenoic acid and its effects on human neutrophils and eosinophils. J Lipid Res, 36, 2590–2598. Powell WS, Gravel S and Halwani F (1999a) 5-oxo-6,8,11,14Eicosatetraenoic acid is a potent stimulator of L-selectin shedding, surface expression of CD11b, actin polymerization, and calcium mobilization in human eosinophils. Am J Resp Cell Mol Biol, 20, 163–170. Powell WS, Gravel S, Halwani F et al (1997) Effects of 5-oxo-6,8,11,14eicosatetraenoic acid on expression of CD11b, actin polymerization and adherence in human neutrophils. J Immunol, 159, 2952–2959. Powell WS, Gravel S, Khanapure SP and Rokach J (1999b) Biological inactivation of 5-oxo-6,8,11,14-eicosatetraenoic acid by human platelets. Blood, 93, 1086–1096. Powell WS, Gravel S, MacLeod RJ et al (1993) Stimulation of human neutrophils by 5-oxo-6,8,11,14-eicosatetraenoic acid by a mechanism independent of the leukotriene B4 receptor. J Biol Chem, 268, 9280– 9286. Powell WS and Gravelle F (1988) Metabolism of 6-trans isomers of leukotriene B4 to dihydro products by human polymorphonuclear leukocytes. J Biol Chem, 263, 2170–2177. Powell WS, Gravelle F and Gravel S (1992) Metabolism of 5(S)-hydroxy6,8,11,14-eicosatetraenoic acid and other 5(S)-hydroxyeicosanoids by a specific dehydrogenase in human polymorphonuclear leukocytes. J Biol Chem, 267, 19233–19241. Powell WS, Gravelle F and Gravel S (1994) Phorbol myristate acetate stimulates the formation of 5-oxo-6,8,11,14-eicosatetraenoic acid by human neutrophils by activating NADPH oxidase. J Biol Chem, 269, 25373–25380. Powell WS, Hashefi M, Falck JR et al (1995c) Effects of oxo and dihydro metabolites of 12-hydroxy-5,8,10,14-eicosatetraenoic acid on chemotaxis and cytosolic calcium levels in human neutrophils. J Leukocyte Biol, 57, 257–263.
Powell WS, MacLeod RJ, Gravel S et al (1996) Metabolism and biologic effects of 5-oxoeicosanoids on human neutrophils. J Immunol, 156, 336–342. Ramboer I, Blin P, Lacape G, Daret D et al (1992) Effects of monohydroxylated fatty acids on arterial smooth muscle cell properties. Kidney Int, 41, S67–S72. Scho¨nfeld W, Schlu¨ter B, Hilger R and Ko¨nig W (1988) Leukotriene generation and metabolism in isolated human lung macrophages. Immunology, 65, 529–536. Schwenk U and Schro¨der JM (1995) 5-oxo-Eicosanoids are potent eosinophil chemotactic factors—functional characterization and structural requirements. J Biol Chem, 270, 15029–15036. Serhan CN, Fiore S, Brezinski DA and Lynch S (1993) Lipoxin A4 metabolism by differentiated HL-60 cells and human monocytes: conversion to novel 15-oxo and dihydro products. Biochemistry, 32, 6313–6319. Sok DE, Kang JB and Shin HD (1988) 15-Hydroxyeicosatetraenoic acid dehydrogenase activity in microsomal fraction of mouse liver homogenate. Biochem Biophys Res Commun, 156, 524–529. Sozzani S, Zhou D, Locati M et al (1996) Stimulating properties of 5-oxoeicosanoids for human monocytes: synergism with monocyte chemotactic protein-1 and -3. J Immunol, 157, 4664–4671. Stamatiou P, Chan CC, Monneret G et al (2001) 5-oxo-6,8,11,14Eicosatetraenoic acid (5-oxo-ETE) stimulates eosinophil survival. Am J Resp Crit Care Med, 163, A321. Stamatiou P, Hamid Q, Taha R et al (1998) 5-oxo-ETE induces pulmonary eosinophilia in an integrin-dependent manner in brown Norway rats. J Clin Invest, 102, 2165–2172. Stoltz RA, Abraham NG and Laniado-Schwartzman M (1996a) The role of NF-kB in the angiogenic response of coronary microvessel endothelial cells. Proc Natl Acad Sci USA, 93, 2832–2837. Stoltz RA, Conners MS, Gerritsen ME et al (1996b) Direct stimulation of limbal microvessel endothelial cell proliferation and capillary formation in vitro by a corneal-derived eicosanoid. Am J Pathol, 148, 129–139. Stoltz RA and Schwartzman ML (1997) High affinity binding sites for 12(R)-hydroxyeicosatrienoic acid [12(R)-HETrE] in microvessel endothelial cells. J Ocul Pharmacol Ther, 13, 191–199. Turk J, Maas RL, Brash AR et al (1982) Arachidonic acid 15-lipoxygenase products from human eosinophils. J Biol Chem, 257, 7068–7076. Wainwright S, Falck JR, Yadagiri P and Powell WS (1990) Metabolism of 12(S)-hydroxy-5,8,10,14-eicosatetraenoic acid and other hydroxylated fatty acids by the reductase pathway in porcine polymorphonuclear leukocytes. Biochemistry, 29, 10126–10135. Wainwright SL and Powell WS (1991) Mechanism for the formation of dihydro metabolites of 12-hydroxyeicosanoids. Conversion of leukotriene B4 and 12-hydroxy-5, 8,10,14-eicosatetraenoic acid to 12oxo intermediates. J Biol Chem, 266, 20899–20906. Wheelan P, Zirrolli JA, Morelli JG and Murphy RC (1993) Metabolism of leukotriene B4 by cultured human keratinocytes. Formation of glutathione conjugates and dihydro metabolites. J Biol Chem, 268, 25439–25448. Yokomizo T, Izumi T, Takahashi T et al (1993) Enzymatic inactivation of leukotriene B4 by a novel enzyme found in the porcine kidney. Purification and properties of leukotriene B4 12-hydroxydehydrogenase. J Biol Chem, 268, 18128–18135. Zhang Y, Styhler A and Powell WS (1996) Synthesis of 5-oxo-6,8,11,14eicosatetraenoic acid by human monocytes and lymphocytes. J Leukocyte Biol, 59, 847–854.
7 Synthetic Eicosanoids Norrie H. Wilson Division of Biomedical Sciences, University of Edinburgh, Edinburgh, UK
Polyunsaturated fatty acids (PUFAs) with 20 carbon atoms, present in the tissues of living organisms, can be mobilized and oxidatively metabolized to form a vast range of highly biologically active substances which mediate many important functions. Oxidative processes arise in the living organism from chemical activity due to cyclooxygenase (COX), lipoxygenase (LOX), cytochrome P450 enzyme systems and aerial/free radical-initiated oxidation. Under usual pathophysiological conditions most of these important mediators are formed in very low quantities and the ability to obtain reasonable amounts of them by synthetic organic chemistry has been crucial to advances in this research field. The biosynthetic pathways are illustrated in Figure 7.1. Virtually all eicosanoids are readily available by chemical synthesis. However, the precursor PUFAs are usually obtained from natural biological sources. Chemical synthesis of these substances is normally only employed to obtain isotopically labelled versions for tracer and analytical studies (Adlof 1999).
INTRODUCTION TO PROSTANOIDS The prostaglandins (PGs) are formed from three precursor polyunsaturated fatty acids, as shown in Figure 7.2. In mammals the 1- and 2-series PGs are most important and the latter seems to predominate in man. The 3-series precursor, timnodonic acid, in the literature often called eicosapentaenoic acid or EPA, occurs mainly in fish and marine animals. Of the three precursor fatty acids, timnodonic is regarded as the poorest substrate for COX. Usually, therefore, a diminished PG biosynthesis results from fish oil-supplemented diets. In most of this review PG synthesis is focused on the 2-series molecules, as they appear to be most important in mammals, including humans. The shorthand notation for the precursor fatty acids is shown and reflects the requirement of double bonds at 8,11,14 (n-6, n-9, n-12), since these structural features are necessary to allow prostaglandins to form. Virtually every cell type in the mammal has the capacity to produce PGs and the biology of these eicosanoids forms one of the major topics of this treatise. The salient details of the rest of the PG biosynthesis pathway are shown in Figure 7.3 for the 2-series. The 1- and 3-series precursors are transformed similarly, producing analogous PGs. The many synthetic pathways for the production of the natural eicosanoids have been modified for the creation of huge numbers of analogues. This has allowed more selective agonist or antagonist versions to be elaborated, which in turn has increased our understanding of the various roles for these substances in the organism, as well as fostering the development of useful pharmaceutical products. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
THROMBOXANE A2 AND PROSTAGLANDIN I2— PROSTANOIDS WITH OPPOSING BIOLOGICAL ACTIONS Thromboxane A2 (TXA2) is the main prostanoid metabolite of arachidonic acid in blood platelets, where it mediates aggregation and is one of the initiators of the clotting process. TXA2 also constricts vascular smooth muscle (Armstrong and Wilson 1995). These important pathophysiological actions are limited by the, pH-dependent, short half-life of TXA2 in aqueous media. The action of TXA2 is also offset by the functional antagonism of other mediators, such as nitric oxide and particularly prostaglandin I2 (PGI2; also called prostacyclin and, in the early literature before its structure was fully known, PGX). TXA2 adds water in a general acid–base-catalysed reaction to form the biologically inactive TXB2 under all pH conditions. This consideration, and clever trapping experiments where methanol or ethanol is added across the oxetane ring instead of water, led to the proposed structure shown in Figure 7.4 (Hamberg et al 1975; see also Figure 7.16 for the total synthesis of TXA2). PGI2 is formed in the endothelial cells lining the blood vessel walls, and this PG is the most powerful platelet antiaggregating agent known (Moncada et al 1976). PGI2 is unstable in acid solution, since the structure contains an enol ether group (Johnson et al 1977, 1978). It is likely that the carboxyl group, as a neighbouring acid group, can participate in the hydrolytic decay to accelerate the formation of 6-keto-PGF1a by proton delivery to the b-carbon of the enol ether via a seven-membered transition state (see Figure 7.5). PGI2, in contrast to TXA2, is stable in alkaline solution and therefore can be stored conveniently as the sodium salt. Figures 7.4 and 7.5 show the hydrolytic inactivation of TXA2 and PGI2 and also illustrate the hemiacetal and hemiketal nature of TXB2 and 6-keto-PGF1a, respectively. The ring–chain tautomerism of these species is also shown and is analogous to that found in glucose and fructose structures, respectively. In each case the open chain carbonyl form is in very low concentration, but they can both be made to react as carbonyl compounds, forming typical derivatives such as oximes and hydrazones (this ring–chain tautomerism is the reason for the undefined stereochemistry of the hydroxyls attached to the anomeric carbons, in the ring form of these molecules).
PROSTAGLANDINS E2, F2a AND D2 These primary PGs are important biologically in pain, fever, reproduction mechanisms and many other biological functions.
70
BIOSYNTHESIS AND METABOLISM
Figure 7.1 Eicosanoid biosynthetic pathways
PGE2 and PGF2a are inactivated mainly, via the circulation, by the lung 15-dehydrogenase, where the 15-hydroxyl is oxidized to a carbonyl group. Thereafter, a series of reductions of the 13,14 double bond and b-oxidations truncating the a-chain produce biologically deactivated species which are excreted in the urine. PGD2 is not a good substrate for the 15-dehydrogenase, presumably because it would lead to the formation of the fluxional 15-carbonyl species which exists in an enol form resistant to further oxidation (Jones and Wilson 1978). Instead, PGD is
reduced to the PGF ring structure, which is then metabolized (Barrow et al 1984). THE SECONDARY PROSTAGLANDINS— PROSTAGLANDINS A, B, C AND J The secondary prostaglandins, PGA, PGB (Pike et al 1969), PGC (Jones et al 1972) and PGJ (Fitzpatrick and Wynalda 1983;
Figure 7.2 The three series of prostaglandins showing the numbering system and the primary product PGGx
SYNTHETIC EICOSANOIDS
71
Figure 7.3 The derivation of the primary prostaglandins
Mahmud et al 1984; Narumiya and Fukushima 1985) are derived from the primary PGs by dehydration (Figure 7.6). These substances perhaps arise during extraction and analytical procedures and it is uncertain whether they are physiologically formed. They have varying degrees of biological activity but are in general less potent than the primary PGs. The secondary PGs can be made from the primary PGs as shown (Figure 7.6). In the case of PGC and PGJ, the conditions for the isomerization or dehydration need to be very mild, and involve the use of plasma or albumin (see references cited). However, enzymes that appear to perform these transformations have been characterized (Polet and Levine 1975). A PGA2 source is in a coral, Plexaura homomalla, where it occurs as the methyl ester, 15R-acetate. The biosynthetic origin of this PG is completely different from that in mammals
(Weinheimer and Spraggins 1969). Procedures for conversion of PGA2 to primary PGs are known (Bundy et al 1971). This route is now of lesser importance with the development of efficient total syntheses.
SYNTHETIC ROUTES TO PROSTAGLANDINS A vast number of synthetic routes to the prostanoids exist and the reader is guided to several specialist texts on the subject (Bindra and Bindra 1977; Roberts and Scheinmann 1982; Pike and Morton 1985). Only the most versatile and useful approaches are discussed in this review, with indications of how other analogues were elaborated from the basic methods.
72
BIOSYNTHESIS AND METABOLISM
Figure 7.4 The decay of TXA2 to the biologically inactive TXB2. The ring–chain tautomers of TXB2
Figure 7.5 The hydrolytic inactivation of PGI2. The ring–chain tautomers of 6-keto-PGF1a
THE COREY SYNTHESIS AND RELATED METHODS This versatile route can generate a large range of PGs, including all the primary PGs and their analogues, for research purposes and pharmaceutical development. The starting materials are available by bulk production methods. The route is illustrated in Figures 7.7 and 7.8 for PGF2a and PGE2. It was possible to perform the alkylation of cyclopentadiene 1 with benzylchloromethylether to give 2 under mild conditions (Corey et al 1971a) to avoid moving the olefinic bonds such that the correct product was obtained on adding the ketene equivalent a-chloroacrylonitrile or a-chloroacryloyl chloride (Corey et al 1971b). Hydrolysis of the
initial Diels–Alder adduct then gave ketone 3, which underwent Baeyer–Villiger oxidation to give a lactone which was hydrolysed to the hydroxy acid 4. This acid 4 was optically resolved by crystallization of the (+)-amphetamine diastereoisomeric salts (Corey et al 1971c). The resolved 4 was then relactonized to 5 by iodolactonization on the olefinic bond, and then removal of the iodo group by reduction, and adjustment of the hydroxyl protecting groups yielded the key intermediate, the Corey lactone 7 (Corey and Suggs 1975). Conversion of the Corey lactone primary alcohol to the aldehyde 8 (Figure 7.8) originally employed a chromic oxidation (Corey et al 1971c); however, even this simple step gave very low
SYNTHETIC EICOSANOIDS
73
Figure 7.6 The secondary prostaglandins
yields and subsequently the preferred method was shown to be the Pfitzner–Moffat oxidation (Bindra and Bindra 1977, p. 200; Corey and Kim 1972, 1973). The o-chain was then added by the Horner– Emmons version of the Wittig reaction, yielding the ketone 9. Reduction of this ketone was originally performed by zinc borohydride, and then the two hydroxy epimers were separated by chromatography. The yield of the naturally occurring 15Sepimer 10 can be enhanced by inversion of the 15R-isomer. Alternatively, particularly in a batch process, the unwanted epimer can be reoxidized to the ketone and added to the next batch for reduction. Later investigations using chiral borane reductants showed that the yield of 10 could be made 490% (Corey et al 1972b, 1987). The hydroxyl groups were protected as the tetrahydropyranyl (THP) ethers, and this was followed by di-isobutylaluminium hydride (DIBAL) reduction of the lactone to the lactol (hydroxyaldehyde) 11. The Wittig reaction then allowed addition of the a-chain with a very high proportion of the cis-olefin 12. Only a very small amount of trans-olefin is formed using this procedure, using dimesylsodium (formed from dissolution of sodium hydride in dimethylsulphoxide at 658C) in DMSO as the base. The removal of the THP groups gave PGF2a. This synthesis allows PGE2 to be obtained also by chromic oxidation
of the unprotected hydroxyl in 12 to carbonyl. The final step of the sequence, the removal of the remaining THP groups to produce PGE2, must be done under mild conditions to avoid dehydration to PGA2 (see above). The required synthons for the a- and o-chains are readily obtained, as shown in Figure 7.9. These reactions are readily extended to the preparation of a large range of analogues. The ‘‘Corey synthesis’’ is versatile and can be adapted to produce the 1-series PGs by reduction of the a-chain olefinic bond (Corey et al 1971b). By employing the appropriate ochain precursor, the 3-series compounds can also be prepared (Corey et al 1971d). A further modification where the a-chain is added first, in order that a large variety of o-chain 1-series analogues could be conveniently obtained, has been explored (Schaaf and Corey 1972). A huge literature exists embellishing the Corey route. For example, there are many alternatives to the ketene equivalents discussed in relation to the 4+2 cycloaddition to cyclopentadiene (Figure 7.7) (Ranganathan et al 1977). Perhaps the most convenient, and inexpensive, synthesis of the Corey lactone is that of To¨mo¨sko¨zi et al 1976, using a highly regioselective Prins reaction (Figure 7.10) on the lactone 16, which in turn can be conveniently prepared in large amounts from 2+2 cycloaddition of dichloroketene to
74
BIOSYNTHESIS AND METABOLISM
Figure 7.7 The Corey lactone synthesis: a key intermediate. Reagents: i, Base/Tl2SO4; ii, BnOCH2Cl; iii, a-chloroacrylonitrile/Cu(BF4)2; iv, KOH/ DMSO/H2O; v, MCPBA; vi, NaOH/H2O; vii, CO2; viii, (+)-amphetamine salt resolution; ix, H+/H2O; x, KI3/H2O; xi, PhBzCl/pyridine; xii, Bu3SnH; xiii, H2, Pd/C, H+
cyclopentadiene (To¨mo¨sko¨zi et al 1976). The Prins reaction is interesting in this system, as it is unusually high-yielding and regiospecific. Other reactions for selective reactions at the olefinic bond in this bicyclic system are most effective on the more tightly constrained ketone 15, which can also lead to several sophisticated PG syntheses (Newton and Roberts 1980). The lactone hydroxyaldehyde 18 can be taken on into the ochain addition or alternatively the intermediate dihydroxy compound (not shown) can be selectively protected at the secondary hydroxyl before the generation of the aldehyde group. Large-scale synthesis of vast numbers of analogues in which the two side-chains can be totally differentiated was developed by the ICI Group (Brown et al 1978), using acetoxyfulvene 19 as the starting point to control the 4+2 cycloaddition of ketene equivalent. The modified route, allowing the inverse addition of the two side-chains, is depicted in Figure 7.11 (Mallion and Walker 1975). These syntheses have some advantages, since they do not use the highly toxic thallous sulphate or benzylchloromethylether as in the Corey synthesis (Figure 7.7). The rest of the process from 25 is very similar to the Corey route. The dimethoxyacetal protecting group was removed by gentle treatment with acid and water before the o-chain was deployed.
THE CONJUGATE ADDITION APPROACH All the PG syntheses discussed so far suffer from the fairly large number of steps involved. Several groups have developed more convergent processes for PG production utilizing conjugate additions. The method involves the 1,4-addition, usually of an organocopper o-chain synthon, to a suitable cyclopentenone. There
are several procedures from which PGE1, and PGE2 were successfully prepared. A typical reaction is illustrated in Figure 7.12. The components themselves for this type of synthesis require several steps in their preparation. Nevertheless the synthesis can be made very efficient. Examples are provided by Sih et al (1973, 1975a, 1975b) and Heather et al (1973). Suitable cyclopentenone precursors such as 26 have been made by several methods described in these papers. The o-chain components, such as 27, are easily available by addition of acyl chloride to acetylene (Price and Pappalardo 1961) in very high yield (Figure 7.13). The chlorine can then be replaced by the more active iodo group for use in the conjugate addition after reduction of the ketone to the alcohol(s). The o-synthon has been enantiomerically resolved by making the phthalate half-ester of the iodovinyl alcohol and purifying the diastereoisomeric (7)-a-methylbenzylamine salts by crystallization (Kluge et al 1972). The conjugate addition of the protected resolved alcohol to the unresolved cyclopentadienone then gave rise to a pair of diastereoisomers, which were separated by chromatography, thereby achieving a chiral synthesis from that point. The reader is directed to the monographs cited for other examples. Advantagous modifications of this type of synthesis include the use of organo-aluminium (Bernady et al 1979) and organo-zirconium species (Schwartz et al 1980) o-chain synthons. The conjugate addition methods can be made even more efficient and a great deal of painstaking research has been done to find reaction conditions allowing the initially formed enolate anion generated from the addition of the o-chain to be trapped using an a-chain synthon (Figure 7.14). This means that the two side-chains can be deployed in a ‘‘one-pot’’ reaction. Initial experiments succeeded in the synthesis of 11-deoxy-PG analogues (e.g. Patterson and Fried 1974) but the method failed in the preparation of the fully functionalized PG.
SYNTHETIC EICOSANOIDS
75
Figure 7.8 Corey route to prostaglandins E2 and F2a. Reagents: i, DCC/pyridinium trifluoroacetate/DMSO; ii, NaH/DMSO/(MeO)2POCH2COC5H11; iii, chiral borane/tBuLi; iv, HOH/K2CO3/MeOH; v, DHP/H+; vi, Dibal[H]; vii, DMSO/Ph3PCHCH2CH2CH2COO7; viii, CrO3/PhH/H2O; ix, AcOH/H2O
The use of more electrophilic a-chain synthons (see 32, 33; Figure 7.15), in tandem with modification of the intermediate enolate with triphenyltin chloride and very carefully controlled reaction conditions, successfully generated the desired product which was transformed into a PG. This ‘‘one-pot’’, threecomponent coupling step gave 475% yields (Suzuki et al 1988). Removal of protecting groups completed the synthesis of PGE2. The most elaborate but successful process, devised by Noyori and his colleagues, required the modification of the organolithium ochain precursor with dimethyl zinc. This presumably stabilizes the intermediate enolate, thereby allowing time for the trapping to occur (Suzuki et al 1990). There was advantage also in the use of the stannane 35 for the salt-free generation of the organolithium 36. Synthon 33 was also used in this procedure and gives rise to 5,6-didehydro-PGs which can be cyclized to the important carbacyclins (see section on PGI2 synthesis, below). Methods for the preparation of suitable cyclopentenones, such as 29, are well documented and the enolate
trapping synthesis has been made even more efficient by the use of enantiomerically resolved versions (Harre et al 1982). The Synthesis of the Labile Prostanoids PGH, PGG and TXA These rather unstable molecules were able to be isolated from biological systems by excluding the co-factors required for the conversion to other PGs (Nugteren and Hazelhof 1973; Hamberg and Samuelsson 1973). The COX enzyme used was from ram seminal vesicles (RSV), the traditional source, and the original biological research was performed using this approach. This method is probably still a very convenient way of obtaining the endoperoxides. Nevertheless, both PGG2 and PGH2 have been chemically synthesized, providing proof of their structure and a supply of these unstable species. The pathway for PGH2 43 is illustrated in Figure 7.16, starting from PGF2a (Porter et al
76
BIOSYNTHESIS AND METABOLISM
Figure 7.9 Synthons for the a- and o-chains
Figure 7.10 Corey lactone via Kovac’s regiospecific Prins reaction. Reagents: i, CCl3COCl/Et3N; ii, Zn/AcOH; iii, H2O2/AcOH/08C; iv, Prins reaction (HCHO)n/AcOH/H2SO4/708C; v, H+/MeOH; vi, Pfitzner–Moffat [O]
1978, 1979a). PGG2 was obtained by a similar route (Porter et al 1980), in which all three labile oxygen functions were added simultaneously by displacement. Of particular note is the use of 2-chloro-3-ethyloxazolium tetrafluoroborate, the Mukaiyama reagent, in the SN2 (substitution nucleophilic 2nd order kinetics) displacement reaction to introduce the bromo groups. This lowtemperature displacement reaction avoids randomizing of the stereochemistry of the leaving groups, which was initially a problem when toluenesulphonate was employed as the leaving group. Other special features of these syntheses include the following. Enzymatic ester hydrolysis, the penultimate step, was required to avoid damage to the bromo groups. Hydrogen peroxide 98% and freshly prepared silver trifluoroacetate were also necessary to optimize the yield of PGH2 (17% in the final step). In spite of the difficulties, both chemical and COX enzyme methods are viable for the production of PG endoperoxides.
An even greater synthetic challenge is TXA2. The ring structure of this eicosanoid is attacked by water and has a half-life of about 40 s in aqueous media. In consequence, TXA2 must be generated quickly and used immediately. Most biological investigations have used TX synthase enzyme prepared from microsomes of animal blood platelets treated with PGH2. It is important to appreciate that PGH2 has a great deal of TXA-like bioactivity and, since TXA2 is normally generated from it, the detection of enhanced biological activity and the subsequent decay of the activity engendered in the solution thus obtained testified to the presence of this labile substance (Needleman et al 1976; Yoshimoto et al 1977). Indeed, experiments of this type led to the discovery and elucidation of the hitherto unknown ring structure of TXA2 (Hamberg et al 1975). It was realized that there was no reason why TXA2 could not be isolated and stored at low temperatures in a dry non-protic organic solvent. Some 10 years after the original description of TXA2, a chemical synthesis was finally devised by a
SYNTHETIC EICOSANOIDS
77
Figure 7.11 The acetoxyfulvene approach. Reagents: i, NaOEt/HCOOEt; ii, a-chloroacrylonitrile; iii, H2O/H+; iv, (MeO)3CH/H+; v, NaOH/DMSO; vi, H2O2/OH7; vii, KI3/H2O; viii, Bu3SnH; ix, DHP/H+
Figure 7.12 Conjugate addition of the o-chain
78
BIOSYNTHESIS AND METABOLISM
Figure 7.13 The o-chain synthon preparation
Figure 7.14 The three-component coupling reaction
Figure 7.15 Synthons for the three-component coupling
SYNTHETIC EICOSANOIDS
79
Figure 7.16 Synthesis of PGH2. Reagents: i, n-butylboronic acid/dimethoxypropane; ii, TBDMSCl/imidazole/DMF; iii, H2O2/H2O; iv, 2-chloro-3ethylbenzoxazolium tetrafluoroborate/Et4N+Br7/CH2Cl2/7158C; v, H+/H2O; vi, porcine lipase ester hydrolysis; vii, H2O2 98%/CF3COOAg/Et2O
group of American chemists (Bagwat et al 1985a, 1985b). The strategy depended on conversion of TXB2, for which there are several syntheses, back to TXA2. The route is shown in Figure 7.17. Interesting aspects are the use of the large ring lactone formed between the 15-hydroxyl and the carboxylic acid as a hydroxyl/carboxyl protecting group, which also prevents sidereactions at the 13,14-double bond, particularly during the free radical debromination. The formation of the oxetane ring, by a modified Mitsunobu reaction, required the presence of the bromine atom in the precursor 48. The beautiful rationale for this was the realization that in the chloral (trichloroacetaldehyde) the carbonyl reacts spontaneously with alcohols to form the acetal, and even in water the energetically favoured form of the chloral is the hydrate (gem-dihydroxy compound). This contrasts with acetaldehyde itself, where the carbonyl form is preferred. Previous attempts to make an oxetane, by dehydrative cyclization across a larger ring, have been unsuccessful and led to alkene formation. A rare, successful example of oxetane formation was that of a 1,3-dihydroxy steroid, where the fairly rigid geometry of the fused ring system did not allow alkene formation (Clayton et al 1957; see also Shibasaki et al 1980). At least one halogen atom adjacent to the putative carbonyl group was shown to be required for successful ring closure (47–49) to the oxetane. The synthetic material 49, after lactone hydrolysis in dry tetrahydrofuran with potassium trimethylsiloxide, was able to be stored dry as the solid potassium salt, which could be quickly dissolved in buffer for biological use as required. The chemical and biological properties of the synthetic TXA2 50 were identical to the biologically derived material.
STABLE MIMETIC ANALOGUES OF PGH2 AND TXA2 TXA2 suffers from instability in the aqueous media in which biological experiments are conducted. Although the aforementioned methods of obtaining TXA2 are theoretically interesting and prove the structure of this important mediator, it is extremely difficult to use the natural substance in routine investigations, where a constant concentration of pharmacologically active compound is desirable for making quantitative measurements. Since TXA2 has such potentially pathological effects, a stable mimetic was required in the search for blocking drugs for use as pharmacological tools or which could have clinical potential. The compound generally accepted as the standard TXA2 mimetic is an Upjohn analogue, U-46619 55, the so-called 9a,11a-epoxymethano-PGH2. This compound has the desired selectivity and potency on the TXA2 receptors (TP receptors) and should be more correctly regarded as a stable carba-analogue of PGH2, and named as 9a-carba-PGH2. Most analogues having this general shape, including PGH2 itself, are excellent TXA2 mimetics. However, PGH2 itself is unstable, and is converted into the large range of other PGs, thus rendering it unsuitable for biological investigations. For background and leading references to this vast and interesting field about the characterization and classification of TP and other PG receptors, see Coleman and Humphrey (1993). U46619 55 is an important research PG due to its usefulness in the discovery and testing of TXA2 agonists and antagonists (Coleman and Humphrey, 1993). The elegant synthesis of
80
BIOSYNTHESIS AND METABOLISM
Figure 7.17. Synthesis of TXA2. Reagents: i, Ph3P/dipyridyl disulphide; ii, 2-chloro-1-methylpyridinium iodide/-H2O; iii, NBS/NaHCO3/Et2O/H2O; iv, (MeO)3P/ethyl azodicarboxylate; v, R3SnH; vi, saponification with either NaOD/D2O/MeOD or Me3SiOK/THF
U46619 is shown in Figure 7.18, starting from suitably protected PGE2. Very mild methylenation of the PGE2 by sulphoximine anion addition, followed by reductive elimination, gave the exocyclic olefin 51, thereby avoiding the dehydration of the PGE ring to PGB. Highly selective hydroboration of the exocyclic alkene with the sterically constrained 9-borabicyclononane (9-BBN) led to the alcohol 53, from which was prepared the methane sulphonate (mesylate) ester 54. After removing the protecting silyl groups, the generation of the alkoxide anion at the 11-position caused elimination of the mesyloxy group and cyclization to the desired structure. This final treatment with excess base was extended in order to hydrolyse the carboxylic ester group to 55, the free acid (Bundy 1975). A more elaborate synthesis of U46619 from more basic raw materials was performed later by Trost et al (1978). In a similar cyclization, the other epoxymethano-cogener, U44069 (9b-carba-PGH2) 56, has also been synthesized by the Upjohn company (Bundy 1975). U-44069 is less potent and in some biological systems acts as a partial agonist, producing slightly less than the maximum response of the biological system. Other molecules employed as TXA2 mimetics include SQ 26655 57, EP 171 58 and thia-carba-TXA2 59 (named for convenience as STA2) (Figure 7.19). Further reviews and leading references are available in Tymkewycz et al (1991) and Wilson and Jones (1985). These TXA2 mimetics are available from Cayman Chemicals Inc., a company specializing in the production and marketing of all eicosanoids.
THE SYNTHESIS OF PGI2 (PROSTACYCLIN) AND ITS STABLE ANALOGUES There are a number of variations of the chemical syntheses of the antiplatelet and vasodilatory PGI2, all employing virtually the same approach from PGF2a. The typical reaction sequence is shown in Figure 7.20. Iodination of the methyl ester of PGF2a in the presence of mild base gave reaction of the 9-position hydroxyl with the 5,6-double bond, yielding the two isomers 60 and 61. These products underwent dehydroiodination with diazabicyclononane (DBN) to the desired Z-alkene 62. Hydrolysis of the ester with base gave the final product as the stable sodium salt of PGI2 63 (Johnson et al 1977, 1978). This synthesis showed that a transaddition, followed by a trans-elimination, both normally expected in this type of reaction, would give rise to the Z-geometry of the newly created olefinic bond. The structure of PGI2 was thereby confirmed. Instead of iodine, other similar cyclization methods employing bromine, phenylselenyl chloride or mercuric acetate have been shown to give the desired product (see also Corey et al 1977; 1978; To¨mo¨sko¨zi et al 1977, 1978). The Upjohn chemists have also shown that PGI2, along with other olefinic isomers, can be isolated by heating the biologically deactivated hydration metabolite, 6-keto-PGF1a, in benzene, with Dean and Stark water removal (Johnson et al 1978). PGI2 has also been prepared by cyclization of the 5,6-didehydroPGF2a analogue, obtainable by conjugate addition of the a-chain 33 in the three-component coupling reaction (Figures 7.14 and 7.15). The ring formation step was achieved, in high yield, using a palladium chloride complex, followed by depalladation with
SYNTHETIC EICOSANOIDS
81
Figure 7.18 Synthesis of U-46619. Reagents: i, N-methylphenylsuphonimidoylmethylmagnesium chloride/THF/7788C; ii, Al/Hg, AcOHH2O; iii, Ph3SiCl (TPSCl); iv, 9-BBN, H2O2/OH7; v, MsCl/Et3N; vi, H3PO4/H2O/THF; vii, KOH/MeOH/H2O
Figure 7.19 Structures of other selected TX mimetics
ammonium formate or, with less stereochemical control, by mercuric trifluoroacetate followed by reductive demercuration with borohydride (Suzuki et al 1988). The use of PGI2 as a therapeutic agent is inhibited by its instability to acid. The associated administration of the alkaline sodium salt and the high rate of metabolism in mammals is a further impediment. Accordingly, medicinal chemists have prepared a vast array of PGI2 congeners, mostly based on the 6acarba-PGI2, usually named carbacyclin for convenience, 69 (Figure 7.21). It was also hoped that it would be possible to
find analogues with selectivity for the IP receptors on platelets, as opposed to those mediating hypotensive vascular activity, since the use of PGI2 to restore circulation in peripheral vascular disease, etc., requires the patient to be maintained in a supine posture to prevent loss of conciousness, due to low blood pressure. The reader is referred to the textbooks cited earlier for the huge array of synthetic efforts in this field and to more recent treatises by Nickolson et al (1985) and Collins and Djuric (1993). An example is outlined, in Figure 7.21, of an asymmetric synthesis of
82
BIOSYNTHESIS AND METABOLISM
Figure 7.20 Synthesis of PGI2. Reagents: i, KI/I2/Na2CO3/H2O/08C; ii, DBN/benzene/408C; iii, NaOH/H2O/MeOH
carbacyclin (Aristoff 1981), starting from a chiral Corey pathway intermediate (compounds 10, 11; Figure 7.8). The a-chain addition to 68 produced both Z- and E-alkenes. Studies to define the conditions to enhance formation of the desired isomer were performed by the Schering group (Rehwinkel et al 1988; Westermann et al 1992). Care is required in the nomenclature of these analogues, since the Z-isomer designation in PGI2 becomes the E-isomer, for designation of the naturally occurring geometry in carbacyclin, due to the alteration in precedence of groups around the exocyclic alkene. The two carbacyclin analogues having altered o-chains, Iloprost 70 (Skuballa and Vorbru¨ggen 1983; note that in the early literature Iloprost was called Ciloprost) and Cicaprost 71 (Skuballa et al 1986) (Figure 7.22), were synthesized and showed interesting structure–activity relationships (SAR). Cicaprost, which has a highly defined quite rigid structure from C-4 to C15 (PG numbering) due to the acetylenic bond and the relatively inflexible cis ring junction, is specific to IP receptors. However, Iloprost has considerable PGE-like actions (Sheldrick et al 1988) as well as PGI activity, like carbacyclin itself and PGI2. For further useful SAR discussion, see Nickolson et al (1985) and Wise and Jones (1996). Further investigation of PGI cogeners has turned up good evidence that there may be sub-types of the IP receptor. Radioligand binding of the isocarbacyclin analogue 73 (Suzuki et al 1996) showed specificity for binding sites in most rat
brain areas. Cicaprost (the most selective IP agonist) displayed much lower binding, except in selected other CNS regions. Isocarbacyclin 72 bound efficiently to both types of brain site and is potent as an antiaggregatory analogue on platelets. Cicaprost is highly potent on platelet inhibition, where compound 73 is very weak, and so the platelet receptor has been called IP1 and the preponderant rat CNS receptor IP2 (Takechi et al 1996). OVERVIEW OF PG SYNTHESIS AND BIOLOGY The breakthrough in understanding the receptors for PGs arose about 1980 with the discovery of TP receptor antagonists using the stable TX-mimetics (see section on PGI2 synthesis, above) (Coleman and Humphrey 1993). It was then possible to quantify the often considerable amount of TX activity in other PGs. The activity of the other PG receptors could then be observed uncontaminated by TP effects, and new PG analogues could be investigated for potentially dangerous TP agonist actions. For example, the introduction of the 16,16-dimethyl o-chain, a standard strategy to attentuate the standard 15-dehydrogenase deactivating metabolism of most PGs, also increases the TP agonist activity. PGE2 has approximately 200-fold less activity as a TX mimetic than U46619, whereas 16,16-dimethyl-PGE2 has only about 15-fold less activity. Even more amazing TP agonism
SYNTHETIC EICOSANOIDS
83
Figure 7.21. Asymmetric synthesis of carbacyclin. Reagents: i, LiCH2PO(OMe)2; ii, CrO3/pyridine; iii, K2CO3/18-crown-6/PhMe; iv, HCOO7HNEt3+/ PdC; v, a-chain Wittig reaction; vi, remove protecting groups
is produced by the introduction of a 16-p-fluorophenoxy-o-chain, as in EP 171. The increase in TP effects is dramatically large. The ring structure to which the 16-p-fluorophenoxy group is attached is of minor importance because of the large increase in TP effects (Jones et al 1987 and references therein). It is important to appreciate that most PGs can interact at least to some degree with several PG receptors and that selective antagonism of a particular receptor is the most useful criterion for classification. Currently, for prostanoid receptors there are good antagonists for TP, some EP and DP receptors. The search for powerful selective antagonists to the other PG receptors continues. For an introduction to eicosanoid receptors, see Wise and Jones (2000) and Coleman and Humphrey (1993). ISOPROSTANES These molecular species are derived from PUFAs by adventitious autoxidation. They comprise a large range of structures having features similar to PGs formed enzymatically by COX. A subgroup, the neuroprostane class, has been identified, derived from the very highly unsaturated fatty acids, particularly docosahexenoic acid, that are found in abundance in the brain. The classes of isoprostanes (iPGs) formed from arachidonic acid are shown in Figure 7.23; the nomenclature is that of Rokach et al (1997). The initial peroxide structures formed are reduced or isomerized to more stable compounds having hydroxyl or
carbonyl groups like the PGs. Note that Types I and II (not shown in Figure 7.23) constitute isoprostanes derived from fatty acids with even more degrees of unsaturation, such as timnodonic acid. Type III are close isomers of the naturally occurring enzymatically produced PGs; however, the non-enzymatic reactions give rise to a large number of racemic diasteroisomers. Many of these species have been identified in the body and it is apparent that the non-enzymatic oxidation favours formation of cisdeployed side-chains, leading to 8-iso- or 12-iso-PGs of the F, D and E ring types. Most chemical synthesis of PGs is designed to be asymmetric to reflect the natural COX activity. Since the isoprostanes are not enzymatically produced and are in consequence racemic mixtures of diastereoisomers (Morrow et al 1990), any chemically synthesized isoprostane will not reflect the natural mixture of stereoisomers. In fact, the chemical syntheses of isoprostanes will normally generate a limited number of stereoisomers. This creates a problem in ELISA and radioimmunoassay methods, where the crucial, selective enzyme or antibody detector is sensitive to a single enantiomer. Such an assay method will not reflect the full isoprostane synthesis. However, these methods can be validated to some degree by GC–MS, particularly as more of the possible isomers become available by synthetic methods. Already predominant isomers have been unambiguously made and it has been suggested that measurements, using some of these as standards, can be used as markers of the oxidative stress levels in an organism (Proudfoot et al 1999).
84
BIOSYNTHESIS AND METABOLISM
Figure 7.22 Structure of important PGI analogues
Rokach’s synthesis of the two lactones 78 and 80, derived from glucose, is indicated in Figure 7.24 (Hwang et al 1994). Wittigtype reactions identical or similar to those employed in the elaboration of the side-chains in the natural PGs have been used to obtain a large range of iPGs. Very few of the iPG stereoisomers have been examined for overt biological actions. Examples include 8-iso-PGF2a, which is available as a pure enantiomer and is recognized to be a highly active TP agonist (Kromer and Tippins 1996). Another example, 8-iso-PGE2, has also been shown to be formed by non-enzymatic autoxidation and has also been synthesized chemically and investigated for biological activity. This cogener has been shown to have strong TP activity on smooth muscle of the vasculature but only causes platelet aggregation at higher dose levels. It also inhibited the aggregation due to the standard powerful TXA2 mimetics (Longmire et al 1994). These authors speculate that this may be due to a unique receptor for iPGE. One complicating factor may be that the iPGE can activate IP receptors, thereby making the platelet response capricious. Alternatively, this iPGE can epimerize readily at the 8-position, adjacent to the ring carbonyl group, via the enol form, to give PGE2 (Figure 7.25). The equilibrium position for these epimers is 90:10 in favour of the natural PGE2 over 8-iso-PGE2 (Daniels et al 1968). How fast this epimerization occurs is pH-dependent and is normally done with very mild base in the case of PGE to minimize PGB formation. In principle, any protic system can catalyse this and it is interesting to speculate whether this can occur while the iPG is still attached to the lipid pool. Similarly, 8-iso-PGD2 can give rise to some nat-PGD2 by equilibration of the suitable diastereoisomer. In an organism suffering a high level of oxidative stress, cell membrane function is likely to be affected by the presence of these polar lipids and a sudden large release of these potentially bioactive substances could have catastrophic repercussions. For an overview of this interesting subject, see Rokach et al 1997.
INTRODUCTION TO LIPOXYGENASE PRODUCTS Distinct from the COX systems are the lipoxygenase (LOX) enzymes, which oxidize PUFA to hydroperoxyeicosatetraenoic acid (HPETE) compounds as an initial product. These are ubiquitous in both plants and animals. The HPETE initial product is usually reduced to the hydroxyeicosatetraenoic acid (HETE) but can also go on to more exotic epoxy species, as discussed later. The function of these metabolic pathways is not fully understood, but in the mammalian system research has shown that these pathways are involved in a vast array of crucial functions, including inflammation, control of the immune system, programmed cell death (apoptosis), fat metabolism and transcriptional control within the cell nucleus. The energy balance within an organism at a deep level appears to be modulated by PUFA, such as docosahecaenoic acid and arachidonic acid as well as oxidation products derived from them, and these substances have been implicated as ligands for peroxisome proliferator-activator receptors (PPARs) in the cell nucleus (for reviews of this area, see Desvergne and Wahli 1999; Kliewer et al 1999; Willson et al 2000). Different cell types have their own unique range of lipoxygenases and, in the case of polyunsaturated eicosanoic acids, nearly every possible position for oxidation has been observed. The LOX enzyme system is therefore categorized by the position at which the oxygen is introduced. In man the main lipoxygenases are 5-, 12- and 15- and these occur in leukocytes and lymphocytes, platelets, lung and several other tissue types (Maclouf 1996). A general review of lipoxygenases is available (Brash 1999). Dihomo-g-linolenic acid, the 1-series PG precursor, has no 5,6double bond, and therefore cannot undergo 5-lipoxygenation, whereas arachidonic and timnodonic acids, with their higher degrees of unsaturation, can form products of 5-LOX. All these acids are susceptible to the actions of LOX enzymes at other
SYNTHETIC EICOSANOIDS
85
Figure 7.23 Isoprostanes derived from arachidonic acid
Figure 7.24 Rokach synthesis of lactones leading to iPGs. Reagents: i, acetone/H3PO4, ZnCl2/208C; ii, NaH/THF/08C then CS2/208C then MeI/208C; iii, Bu3SnH/toluene/1118C; iv, H2O/AcOH/208C; v, thiocarbonyldiimidazole/dichloroethane/838C; vi, H+/H2O/THF/668C; vii, methyl (triphenylphosphoranylidene)acetate/base/THF; viii, Bu3SnH/AIBN/C6H6/808C; ix, TBDMSCl/imidazole/DMF/608C
86
BIOSYNTHESIS AND METABOLISM
Figure 7.25 The equilibration of 8-iso-PGE2 and its enantiomer to the more stable trans side-chain isomers
suitable positions. Experiments to alter diets, and hence perturb the lipid pools of the body, and the blocking of pathways such as the COX systems by aspirin-like drugs, which may lead to more lipoxygenase products, are under investigation (DeWitt 1996). It appears, therefore, that even when no biological role for a given metabolite is initially apparent, further research usually discovers a role, either a binding site (enzyme modulation) or a receptor (cell signalling), with important consequences for the organism. The contribution of synthetic chemistry in providing these mediators as standards for analytical work and pharmacological study cannot be overestimated.
HETEs, HPETEs, LEUCOTRIENES AND LIPOXINS An important lipoxygenase pathway in humans is the one involving 5-LOX (Samuelsson 1983). This metabolic route leads to the physiologically important leukotrienes (LTs) and is depicted in Figure 7.26. LTA4 is very unstable and adds water to form LTB4 or glutathione to form LTC4. These reactions are mediated by hydrolase or glutathione-S-transferase enzymes, respectively. The glutathione in the latter case is added to the epoxide by nucleophilic attack of the cysteine thiol group. The enzymatic routes are still used to obtain these esoteric compounds but chemical synthesis is also available and was required to prove the stereochemistry. The lipoxygenases mostly produce the S-enantiomer of the hydroperoxide and subsequent enzymic reactions are all stereospecific. Low-temperature HPLC is the key to obtaining pure material for biological study, particularly of the labile hydroperoxides. All the highly unsaturated lipoxygenase products suffer from rapid aerial autoxidation/polymerization.
LTB4 is important as a major chemotactic agent (Samuelsson et al 1987; Ford-Hutchinson 1990). The ketone derived from oxidation of 5-HETE (5-oxo-ETE) has similar actions to LTB4, but it acts by a different mechanism and is less potent (Powell et al 1995). LTC4 and the other peptido-LTs (Figure 7.26) have been identified with the slow-reacting substance of anaphylaxis (SRSA). This is released along with histamine in allergic responses and promotes hyper-reactivity of airways. The 12- and 15-lipoxygenase pathways, shown in Figures 7.27 and 7.28, appear to be more important than was first realized. Initially it was considered that the formation of these very polar lipids constituted methods for deactivation and elimination of very bioactive substances from the body, and this may still be true in some cases. However, it is now realized that virtually all of these products have some role to play in the signalling systems of the organism, e.g. 12-HPETE and LTB4 have been proposed to be endogenous ligands for the vanilloid (VR1) receptor which mediates the burning pain sensation caused either by heat or acid (Hwang et al 2000). These lipoxygenase routes also lead to the hepoxilins and lipoxins. These substances are examples of subtle mediators of important functions (see Lumin et al 1992; Reynaud et al 1995; Yoshimoto and Yamamoto 1995; Bratt et al 1995). The following section on the synthesis of the oxylipins is designed as an introduction to their chemistry. CHEMICAL SYNTHESIS OF LIPOXYGENASE-DERIVED EICOSANOIDS Chemical synthesis of many lipoxygenase products has been performed using standard aliphatic olefinic and acetylenic chemistry. Excellent monographs on the subject are available
SYNTHETIC EICOSANOIDS
87
Figure 7.26 The 5-lipoxygenase biosynthetic pathway
(Corey and Cheng 1989; Scheinmann and Ackroyd 1984; Pike and Morton 1985). In many cases these syntheses were done to prove the stereochemistry of molecules of biological prominence. Practical routes, mostly starting from the precursor fatty acids, are discussed as methods of obtaining these exotic substances for research.
SYNTHESIS OF HYDROPEROXY- AND HYDROXY-DERIVATIVES OF EICOSAPOLYENOIC ACIDS (HPETEs AND HETEs) Enzymatic methods have the advantage that the starting material is the readily available PUFA and even very labile products are easy to obtain under mild conditions by HPLC, since the reaction often forms a UV chromophore. The enzymic products are also normally chiral. Apart from the asymmetric aspect, the enzymatic systems can be imitated chemically by use of an excited state of dioxygen, singlet oxygen, which is conveniently generated from ground state dioxygen by dye sensitization. Irradiation of a solution of methylene blue dye with light of red wavelength causes excitation of the dye to the singlet electronic state, which is then transferred to oxygen dissolved in the solvent. Air can often be
used as the source of oxygen and is much less hazardous than pure oxygen gas. The singlet oxygen thus formed reacts with suitable substrates in the solution, in this case the methylene-interrupted polyunsaturated fatty acid. The ‘‘ene’’ reaction, thus engendered, produced HPETE products where the shifted double bond became trans. This general rule, that double bonds which move during a reaction become trans, only seems to be violated in the enzymic addition of water to LTA4 to form LTB4. The oxygenation reaction can be run at low temperatures, such that thermally labile products can be isolated. The general reaction is shown in Figure 7.29 (Porter et al 1979b) and when applied to arachidonic acid gave all possible HPETE products. These were characterized by HPLC separation guided by UV absorbing fractions. Hydrogenation of all double bonds and reduction of the hydroperoxy group to hydroxy followed by derivatization of the hydroxy-saturated acid to methyl ester, trimethylsilyl ether (Me, TMS), followed by gas chromatography–mass spectrometry, determined the position of the oxygen function (Porter et al 1979b). Milligram quantities of the pure racemic HPETE compounds were available by this onestep process. Woollard (1986) used this method to make standard HETEs and demonstrated that 12-HETE formed in psoriatic skin was the R-enantiomer, in contrast to the usual 12S-HETE formed in platelets. He was also able to chromatographically separate the
88
BIOSYNTHESIS AND METABOLISM
Figure 7.27 The 12-lipoxygenase biosynthetic pathway
two enantiomers using the diastereoisomeric dehydroabietylurethane derivatives, showing that this procedure can produce chiral material identical to that formed in biological systems. Arguably the most important synthetic approach to the lipoxygenase products is by the selective epoxidation of arachidonic acid. This can be adjusted to provide, directly, either the 5,6-epoxide 83 shown in Figure 7.30, or the 14,15-epoxide 85 shown in Figure 7.31 (Corey et al 1980a). The related HPETEs and HETEs can thus be obtained. Furthermore these 5,6- and 14,15-epoxides can be used as starting points for the transfer of the oxygen function along the chain (Figure 7.32) to give, respectively, the 8,9-epoxide 89 (Falck and Manna 1982) and the 11,12-epoxide 91 (Corey et al 1980a). The highly efficient allylic epoxidation mediated by vanadium as catalyst provided the key step in this shift of the oxirane function to the adjacent double bond position. The HPETE/HETE systems related to these new epoxides were thus able to be obtained. HPETEs can be generated chemically at low temperature from HETEs by nucleophilic displacement, as in Figures 7.30 and 7.31, and the scheme can be used to obtain either the free acid or the methyl ester. The stereochemistry of these products was determined by comparison with enzymatically produced compounds from either plant or animal sources.
LTA4, LTB4 AND LTC4 SYNTHESIS FROM 5-HPETE 5-HPETE was converted to the labile leukotriene A4 (LTA4) by a hydroperoxide-oxiranylcarbinol rearrangement (Corey and Barton 1982; Corey et al 1984). LTA4 has been converted to LTB4 by enzymatic addition of water (Maycock et al 1982). LTC4 and the other peptido-LTs have been obtained by addition of glutathione, etc. to LTA4 (see Figure 7.26) (Corey et al 1980b). Lipoxins and Hepoxilins Synthesis of these esoteric and highly unstable species has been crucial to prove the stereochemistry and make them available for biological studies. Lipoxin A4 (LxA4) was most conveniently prepared from 15HETE, which can in turn be obtained in chiral form, and in large amount, from oxidation of arachidonic acid by soya bean lipoxygenase. The 15-HETE hydroxyl group was protected, and then the carboxyl group was utilized in an iodolactonization reaction to introduce the 5-hydroxy function in an analogous method to that shown in Figure 7.30. Vicinal oxidation directed by the 5-hydroxyl then allowed addition of the final hydroxyl.
SYNTHETIC EICOSANOIDS
89
Figure 7.28 The 15-lipoxygenase biosynthetic pathway
Figure 7.29 Formation of hydroperoxides by singlet oxygen
5S,6S,15S-Trihydroxy-7,9,11,13-eicosatetraenoic acid (LxA4) was obtained as a 1:1 mixture with the 5R,6R-diastereoisomer and was identical to biological material from human neutrophils (Corey and Su 1985a; see also Corey et al 1989). Lipoxin B4 (LxB4) was prepared analogously. The crucial step was the conversion of the hydroperoxy soya bean lipoxygenase product, 15-HPETE, to a mixture of threo- and erythroepoxyalcohols with titanium tetraisopropoxide. The threo form was then treated with base to produce the required 14,15dihydroxy compound which, after protecting as the trimethylsilyl ether, underwent the iodolactonosation process to introduce the hydroxy group at the 5-position. One of the 5-epimers was shown by HPLC retention times to be identical to natural LxB4 (Corey et al 1985b). Other elegant but longer syntheses of the lipoxins are known which further confirm the assigned structure. Most employ standard alphatic chemistry techniques, involving coupling of alkynes and/or chiral starting points to control the stereochemistry. An example is given by Gravier-Pelletier et al (1991), where the starting material is D-isoascorbic acid. A key step is a modified
Wittig-type reaction involving the use of arsonium ylide. The arsenic moiety acts as a leaving group but the lower affinity for oxygen allows the starting material carbonyl oxygen to be retained as an epoxide. Both LxA4 and LxB4 were synthesized. The synthesis of hepoxilin A3 (HxA3) (Chabert et al 1989) is depicted in Figure 7.33 as an example of the use of this approach using arsonium ylides. An interesting, almost biomimetic, synthesis of HxA3 and, by hydrolysis, the associated trioxilin A3 (TxA3) has also been described by Lumin et al (1992). THE MONOOXYGENASE PATHWAY (CYTOCHROME P450) A range of oxidized derivatives are known to be formed by the hepatic cytochrome P450 enzyme systems, which require molecular oxygen and NADPH. The enzymes are site-specific and each double bond of the polyunsaturated eicosanoic acids can be attacked, usually to form a monoepoxide (EET), with almost total
90
BIOSYNTHESIS AND METABOLISM
Figure 7.30 Alternative synthesis of 5-HETE, 5-HPETE and the 5,6-epoxide of arachidonic acid. Reagents: i, KI3/K2CO3/THF/H2O/O8C; ii, DBU; iii, Et3N/MeOH; iv, LiOH then CH2N2; v, MsCl/Et3N/CH2Cl2/7658C then H2O2/71108C
Figure 7.31 The selective 14,15-epoxidation of arachidonic acid and conversion to 15-HPETE. Reagents: i, (imidazolyl)2CO/CH2 Cl2 then H2O2/lithium imidazolide then KHSO4/CH2Cl2; ii, CH2N2/Et2O; iii, Mg(NR2)2; iv, MsCl/Et3NCH2Cl2/7428C then TBDMSOOH; v, H+/H2O
enantiofacial selectivity. In the case of arachidonic acid, all epoxides are known (see previous section) and are further modified by ubiquitous epoxide hydrolases to form vicialdihydroxyeicosatrienoic acids (DHETEs). These EETs are often formed with the unsaturated acid still incorporated in the phospholipids and thus can be released by phospholipases. The consequences to membrane structure and the pathophysiology of these metabolites is still under investigation (Capdevilla et al
1992). The DHETEs and some epoxyhydroxy derivatives have mostly been characterized by GC–MS, and the DHETEs synthesized by addition of water across the various epoxides discussed in the previous section. Addition of water or glutathione to various positional isomers of LTA4 produced by non-enzymatic oxidations could, in principle, lead to a large range of isoleukotrienes and related substances. This would be analogous to the
SYNTHETIC EICOSANOIDS
91
Figure 7.32 Synthesis of the 8,9- and 11,12-epoxides of arachidonic acid. Reagents: i, KBr/THF/H2O/AcOH; ii, ButOOH/VO(acac)2/PhH; iii, (CF3SO2)2O/pyridine/HMPT/CH2Cl2
Figure 7.33 Use of arsonium ylides in synthesis of HxA3
Figure 7.34 Structures of the endogenous cannabinoids and D9-tetrahydrocannabinol
92
BIOSYNTHESIS AND METABOLISM
formation of the isoprostanes. There are some reports of these substances occurring in vivo; however, the pathophysiological significance of these reactions remains to be determined (Rokach et al 1997).
ANANDAMIDE AND ENDOGENOUS CANNABINOIDS In 1992 it was shown (Devane et al 1992) that cannabinoid receptors existed in the mammalian system and that the endogenous ligand was an eicosanoid, namely arachidonylethanolamide. Since then another ligand 2-arachidonylglycerol has been shown to be endogenous (for a review of this field, see Pertwee 1998). These natural ligands have affinity for the cannabinoid receptors similar to tetrahydrocannabinol, the main pharmacologically active constituent of the traditional marijuana plant. The chemical constitution of the endogenous and the plant ligands is enormously different, and these investigations have led to a large SAR study resulting in new agonists, antagonists and inverse agonists for these receptors. The endogenous ligands can be prepared in large amounts by simple derivatization of arachidonic acid. Review papers (Martin et al 1999; Hillard 2000) are available which discuss the chemical synthesis and SAR of anandamide and its analogues. More problems attend the discovery of 2-arachidonylglycerol as a ligand for CB receptors, since it is known that 2 acylglycerols can easily rearrange to the 1substituted compound. 1-Arachidonylglycerol is much weaker as a CB agonist than 2-arachidonylglycerol. The chemical synthesis and activity is discussed in a series of papers (Sugiura et al 1999 and references therein). Both endogenous ligands, so far discovered, are achiral substances, unlike the plant cannabinoids (phytocannabinoids), where it is known that only the naturally occurring enantiomeric substance has high biological activity. However, chiral analogues of the endogenous CB ligands have been prepared and exhibit a eudesmic effect and some have advantages since they are not metabolized as readily.
REFERENCES Adlof RO (1999) Lipid Synthesis and Manufacture, Gunstone FD (ed.). Sheffield Academic Press, Sheffield, UK, pp. 46–93. Armstrong RA and Wilson NH (1995) Aspects of the thromboxane receptor system. Gen Pharmacol, 26, 463–472. Aristoff PA (1981) Practical synthesis of 6a-carbaprostaglandin I2. J Org Chem, 46, 1954–1957. Bagwat SS, Hamann PR and Still WC (1985a) Synthesis of thromboxane A2. J Am Chem Soc, 107, 6372–6376. Bagwat SS, Hamann PR, Still WC et al (1985b) Synthesis and structure of the platelet aggregation factor thromboxane A2. Nature, 315, 511–513. Barrow SE, Heavey DJ, Ennis M et al (1984) Measurement of PGD2 and identification of metabolites in human plasma during intravenous infusion. Prostaglandins, 28, 743–754. Bernady KF, Floyd, MB, Poletto JF and Weiss MJ (1979) Prostaglandins and cogeners. 20. Synthesis of prostaglandins via conjugate addition of lithium trans-1-alkenyltrialkylalanate reagents. A novel reagent for conjugate 1,4-additions. J Org Chem, 44, 1438–1446. Bindra JS and Bindra R (1977) Prostaglandin Synthesis. Academic Press, New York. Brash AR (1999) Lipoxygenases: occurrence, functions, catalysis and acquistion of substrate. J Biol Chem, 274, 23679–23682. Bratt D, Lerner R, Ringertz B and Palmblad J (1995) Mechanisms for lipoxin A4-induced neutrophil-dependent cytotoxicity for human endothelial cells. In Advances in Prostaglandin, Thromboxane and Leucotriene Research, vol 23. Raven, New York, 419–421.
Brown ED, Clarkson R, Leeney TJ and Robinson GE (1978) The acetoxyfulvene synthesis of prostaglandins. Part 1. Synthesis of the Corey aldehyde. J Chem Soc, 1507–1511. Bundy GL (1975) The synthesis of prostaglandin endoperoxide analogs. Tetrahed Lett, 1957–1960. Bundy GL, Lincoln FH, Nelson NA et al (1971) Novel prostaglandin syntheses. Ann NY Acad Sci, 180, 76–90. Capdevilla J, Falck JR and Estabrook RW (1992) Cytochrome P450 and the arachidonic cascade. FASEB J, 6, 731–736. Chabert P, Mioskowski C and Falck JR (1989) Synthesis of (+/7) hepoxilin A3 utilising arsonium ylides. Tetrahed Lett, 30, 2545–2548. Clayton RB, Henbest HB and Smith M (1957) Aspects of stereochemistry. Part V. Reactions of cyclohexane-1,3-diol monoarenesulphonates with alkali. J Chem Soc, 1982–1993. Coleman RA and Humphrey PPA (1993) Prostanoid receptors: their function and classification. In Therapeutic Applications of Prostaglandins, Vane JR and O’Grady J (eds). Edward Arnold, London, 15–36. Collins PW and Djuric SW (1993) Synthesis of therapeutically useful prostaglandin and prostacyclin analogs. Chem Rev, 93, 1533–1564. Corey EJ, Koelliker U and Neuffer J (1971a) Methoxy-methylation of thallous cyclopentadienide. a simplified preparation of a key intermediate for the synthesis of prostaglandins. J Amer Chem Soc, 93, 1489–1490. Corey EJ, Ravindranathan T and Terashima S (1971b) A new method for the 1,4-addition of the methylenecarbonyl unit to dienes. J Amer Chem Soc, 93, 4326–4327. Corey EJ, Albonico SM, Koelliker U et al (1971c) New reagents for stereoselective carbonyl reduction. An improved synthetic route to the primary prostaglandins. J Am Chem Soc, 93, 1491–1493. Corey EJ, Shirahama H, Yamamoto H et al (1971d) Stereospecific total synthesis of prostaglandin E3 and F3a. J Am Chem Soc, 93, 1490–1491. Corey EJ and Kim CU (1972) A new and highly effective method for the oxidation of primary and secondary alcohols to carbonyl compounds. J Am Chem Soc, 94, 7586–7587. Corey EJ, Becker KB and Varma RK (1972) Efficient generation of the 15S configuration in prostaglandin synthesis. Attractive interactions in stereochemical control of carbonyl reduction. J Am Chem Soc, 94, 8616–8618. Corey EJ and Kim CU (1973) Improved synthetic routes to prostaglandins utilising sulfide-mediated oxidation of primary and secondary alcohols. J Org Chem, 38, 1233. Corey EJ and Suggs WJ (1975) A method for catalytic dehalogenations via trialkyltin hydrides. J Org Chem, 40, 2554–2555. Corey EJ, Keck GE and Sze´kely I (1977) Synthesis of Vane’s prostaglandin X, 6,9a-oxido-9a,15a-dihydroxyprosta-(Z)5,(E)13dienoic acid. J Am Chem Soc, 99, 2006–2008. Corey EJ, Pearce HL, Sze´kely I and Ishiguro M (1978) Configuration at C-6 of 6,9a-oxido-bridged prostaglandins. Tetrahed Lett, 1023–1026. Corey EJ, Marfat A, Falck JR and Albright JO (1980a) Controlled chemical synthesis of the enzymatically produced eicosanoids 11-, 12and 15-HETE form arachidonic acid and conversion into the corresponding hydroperoxides (HPETE). J Am Chem Soc, 102, 1433–1435. Corey EJ, Clark DA, Goto G et al (1980b) Stereospecific total synthesis of a ‘‘slow reacting substance’’ of anaphyaxis, leucotriene C1. J Am Chem Soc, 102, 1436–1439. Corey EJ and Barton AE (1982) Chemical conversion of arachidonic acid to slow reacting substances. Tetrahed Lett, 23, 2351–2354. Corey EJ, Su W-G and Mehrotra MM (1984) An efficient and simple method for the conversion of 15-HPETE to 14,15-EPETE (lipotriene A) and 5-HPETE to leucotriene A as the methyl ester. Tetrahed Lett, 25, 5123–5126. Corey EJ and Su W-G (1985a) Simple synthesis and assignment of stereochemistry of lipoxin A. Tetrahed Lett, 26, 281–284. Corey EJ, Mehrotra MM and Su W-G (1985b) On the synthesis and structure of lipoxin B. Tetrahed Lett, 26, 1919–1922. Corey EJ, Bakshi RK, Shibata S et al (1987) A stable and easily prepared catalyst for the enantioselective reduction of ketones. Applications to multistep syntheses. J Am Chem Soc, 109, 7925–7926. Corey EJ and Cheng X-M (1989) The Logic of Chemical Synthesis, Chapters 11 and 12. Wiley Interscience, New York.
SYNTHETIC EICOSANOIDS Corey EJ, Su W-G and Cleaver MB (1989) Simple and efficient synthesis of (7E,9E,11Z,13E-5S,6R,15S)-trihydroxyeicosatetraenoic acid (6Rlipoxin A). Tetrahed Lett, 30, 4181–4184. Daniels EG, Krueger WC, Kupiecki FP et al (1968) J Am Chem Soc, 90, 5894. Desvergne B and Wahli W (1999) Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev, 20, 649–688. Devane WA, Hanus L, Breuer A et al (1992) Isolation and structure of a brain constitutent that binds to the cannabinoid receptor. Science, 258, 1946–1949. DeWitt DL (1996) The physiological and therapeutic significance of separate pathways for prostaglandin synthesis. In Eicosanoids, Folco GC et al (eds), Series A, Life Sciences. Plenum, New York, 283, 21–36. Falck JR and Manna S (1982) 8,9-Epoxyarachidonic acid: a cytochrome P450 metabolite. Tetrahed Lett, 23, 1755–1756. Fitzpatrick FA and Wynalda MA (1983) Albumin-catalysed metabolism of prostaglandin D2. Identification of products formed in vitro. J Biol Chem, 258, 11713–11718. Ford-Hutchinson AW (1990) Leucotriene B4 in inflammation. Crit Rev Immunol, 10, 1–12. Gravier-Pelletier C, Dumas J, Le Merrer Y and Depezay JC (1991) Total synthesis of lipoxins A4 and B4 from D-isoascorbic acid. Tetrahed Lett, 32, 1165–1168. Hamberg M and Samuelsson B (1973) Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Natl Acad Sci USA, 70, 899–903. Hamberg M, Svensson J and Samuelsson B (1975) Thromboxanes: a new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Natl Acad Sci USA, 72, 2994–2998. Harre M, Raddatz P, Walenta R and Winterfeldt E (1982) 4-oxo-2Cyclopentenyl acetate—a synthetic intermediate. Angew Chem Int Ed, 21, 480–492. Heather JB, Sood R, Price P et al (1973) Total synthesis of prostaglandins V. A synthesis of (7)prostaglandin E2 via a totally asymmetric process. Tetrahed Lett, 2313–2316. Hillard CJ (2000) Biochemistry and pharmacology of the endocannabinoids arachidonyl ethanolamide and 2-arachidonylglycerol. Prostagland Lipid Mediat, 61, 3–18. Hwang S-W, Adiyaman M, Khanapure SP et al (1994) Total synthesis of 8-epi-PGF2a. A novel strategy for the synthesis of isoprostanes. J Am Chem Soc, 116, 10829–10830. Hwang SW, Cho H, Kwak J et al (2000) Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci USA, 97, 6155–6160. Isakson P, Seibert K, Masferrer J et al (1995) Discovery of a better aspirin. In Advances in Prostaglandin, Thromboxane and Leucotriene Research, Samuelsson B, Ramwell PW, Paoletti R et al (eds). Raven, New York, 23, 49–54. Johnson RA, Lincoln FH, Thompson JL et al (1977) Synthesis and stereochemistry of prostacyclin and synthesis of 6-keto-prostaglandin F1a. J Am Chem Soc, 99, 4182–4184. Johnson RA, Lincoln FH, Nidy EG et al (1978) Synthesis and characterization of prostacyclin, 6-keto-PGF1a, PGI1 and PGI3. J Am Chem Soc, 100, 7690–7705. Jones RL, Cammock S and Horton EW (1972) Partial purification of cat plasma prostaglandin A isomerase. Biochim Biophys Acta, 280, 588– 601. Jones RL and Wilson NH (1978) Tautomerism of the enedione system of 15-oxo-PGD2. J Chem Soc, 211–214. Jones RL, Wilson NH, Armstrong RA and Tymkewycz PM (1987) Developments in the characterisation of prostaglandin endoperoxide/ thromboxane receptors. In Biology of Icosanoids, Lagarde M (ed.), 152, INSERM, France, 335–344. Kliewer SA, Lehmann JM and Willson TM (1999) Orphan nuclear receptors: shifting endocrinology into reverse. Science, 284, 757–760. Kluge AF, Untch KG and Fried JH (1972) Synthesis of prostaglandin models and prostaglandins by conjugate addition of a functionalized organocopper reagent. J Am Chem Soc, 94, 7827–7832. Kromer BM and Tippins JR (1996) Coronary artery constriction by the isoprostane, 8-epi-prostaglandin F2a. Br J Pharmacol, 119, 1276–1280. Longmire AW, Jackson Roberts L and Morrow JD (1994) Actions of the E2-isoprostane, 8-iso-PGE2 on the platelet thromboxane/endoperoxide receptor in humans and rats. Prostaglandins, 48, 247–256.
93
Lumin S, Falck JR, Capdevilla J and Karara A (1992) Palladiummediated allylic Mitsunobu displacement: stereocontrolled synthesis of hepoxilin A3 and trioxilin A3 methyl esters. Tetrahed Lett, 33, 2091– 2094. Maclouf J (1996) Transcellular biosynthesis of arachidonic acid metabolites: from in vitro investigations to in vivo reality. Eicosanoids, vol 283, Series A, Life Sciences. Plenum, New York, 77– 88. Maycock AL, Anderson MS, DeSousa DM and Kuehl FA (1982) Leucotriene A4: preparation and enzymatic conversion in a cell free system to leucotriene B4. J Biol Chem, 257, 13911–13914. Mahmud I, Smith DL, Whyte MA et al (1984) On the identification and biological properties of prostaglandin J2. Prostagland Leucotrienes Med, 16, 131–146. Mallion KB and Walker ERH (1975) The acetoxyfulvene synthesis of prostaglandins: a new versatile synthesis of D, L-PGF1a, D, L-PGF2a, D, L-PGE1 and D, L-PGE2 from a common intermediate. Synthet Commun, 5, 221–226. Martin BR, Mechoulam R and Razdan RK (1999) Discovery and characterisation of endogenous cannabinoids. Life Sci, 65, 573–595. Morrow JD, Hill KE, Burk RF et al (1990) A series of prostaglandin F2like compounds are produced in vivo in humans by a noncyclooxygenase, free radical-catalysed mechanism. Proc Natl Acad Sci USA, 87, 9383–9387. Moncada S, Gryglewski RJ, Bunting S and Vane JR (1976) Nature, 263, 663–665: idem. Prostaglandins, 12, 715–737. Narumiya S and Fukushima M (1985) D12-Prostaglandin J2, an ultimate metabolite of prostaglandin D2 exerting cell growth inhibition. Biochem Biophys Res Comm, 127, 739–745. Needleman P, Moncada S, Bunting S et al (1976) Identification of an enzyme in platelet microsomes which generates thromboxane A2 from prostaglandin endoperoxides. Nature, 261, 558–560. Newton RF and Roberts SM (1980) Steric control in prostaglandin synthesis involving bicyclic and tricyclic intermediates. Tetrahedron, 36, 2163–2196. Nickolson RC, Town MH and Vorbru¨ggen H (1985) Prostacyclinanalogs. Med Res Rev, 5, 1–53. Nugteren DH and Hazelhof E (1973) Isolation and properties of intermediates in prostaglandin biosynthesis. Biochim Biophys Acta, 326, 448–461. Patterson JW and Fried JH (1974) Synthesis of prostaglandins by conjugate addition and alkylation of a directed enolate ion. 11Deoxyprostaglandins. J Org Chem, 39, 2506–2509. Pertwee R (1998) Cannabinoid receptor pharmacology. In Cannabis, Brown DT (ed.), Harwood Academic, Amsterdam, 125–174. Pike JE and Morton DR (1985) Chemistry of the prostaglandins and leucotrienes. In Advances in Prostaglandin, Thromboxane and Leucotriene Research. Raven, New York, 14. Pike JE, Lincoln FH and Schneider WP (1969) Prostanoic acid chemistry. J Org Chem, 34, 3552–3557. Polet H and Levine L (1975) Metabolism of prostaglandins E, A and C in serum. J Biol Chem, 250, 351–357. Porter NA, Byers JD, Mebane RC et al (1978) Prostaglandin H2 methyl ester. J Org Chem, 43, 2088–2090. Porter NA, Byers JD, Holden KM and Menzel DB (1979a) Synthesis of prostaglandin H2. J Am Chem Soc, 101, 4319–4322. Porter NA, Byers JD, Ali AE and Eling TE (1980) Prostaglandin G2. J Am Chem Soc, 102, 1183–1184. Porter NA, Logan J and Kontoyiannidou V (1979b) Preparation and purification of arachidonic acid hydroperoxides of biological importance. J Org Chem, 44, 3177–3181. Powell WS, Gravel S, Stamatiou P and Macleod RT (1995) Effects of 5oxo-ETE on neutrophils, eosinophils and intestinal epithelial cells. In Advances in Prostaglandin Thromboxane and Leukotriene Research, Sammuelson B, Ramwell PW, Paoletti R et al (eds), vol 23. Raven, New York, 421–433. Poyser NL (1995) The control of prostaglandin production by the endometrium in relation to luteolysis and menstruation. Prostagland Leucotrienes Essent Fatty Acids, 53, 147–195. Price CC and Pappalardo JA (1961) b-Chlorovinyl-isoamyl ketone. Org Syn Coll, 4, 186–188. Proudfoot J, Barden A, Mori TA et al (1999) Measurerment of urinary F2isoprostanes as markers of in vivo lipid peroxidation—a comparison of
94
BIOSYNTHESIS AND METABOLISM
enzyme immunoassay with gas chromatography/mass spectrometry. Anal Biochem, 272, 209–215. Ranganathan S, Ranganathan D and Mehrotra AK (1977) Ketene equivalents. Synthesis, 289–296 Rehwinkel H, Skupsch J and Vorbru¨ggen H (1988) E or Z-selective Horner–Wittig reactions of substituted bicyclo[3.3.0]octane-3-ones with chiral phosphonoacetates. Tetrahed Lett, 29, 1775–1776. Reynaud D, Demin PM and Pace-Asciak CR (1995) Enzymic formation of hepoxilin A3. In Advances in Prostaglandin Thromboxane and Leucotriene Research, Samuelsson B, Ramwell PW, Paoletti R et al (eds), vol 23. Raven, New York, 183–185. Roberts SM and Scheinmann F (1982) New Synthetic Routes to Prostaglandins and Thromboxanes. Academic Press, London. Rokach J, Khanapure SP, Huang S-W et al (1997) The isoprostanes: a perspective. Prostaglandins, 54, 823–851. Samuelsson B (1983) Leucotrienes: mediators of immediate hypersensitivity reactions and inflammation. Science, 220, 568–575. Samuelsson B, Dahlen S-E and Lindgren JA (1987) Leucotrienes and lipoxins: structures, biosynthesis and biological effects. Science, 237, 1171–1176. Schaaf TK and Corey EJ (1972) A total synthesis of prostaglandins F1a and E1. J Org Chem, 37, 2921–2922. Scheinmann F and Ackroyd J (1984) Leucotriene Syntheses: a New Class of Biologically Active Compounds Including SRS-A. Raven, New York. Schwartz J, Loots MJ and Kosugi H (1980) Nickel-catalysed conjugate addition of alkenylzirconium species to ab-unsaturated ketones. J Am Chem Soc, 102, 4182–4184. Sheldrick RIG, Coleman RA and Lumley P (1988) Iloprost, a potent FPIreceptor and IP-receptor agonist. Br J Pharmacol, 94, 334. Shibasaki M, Nishida A and Ikegami S (1980) A novel synthetic route to 6-oxabicyclo[3.1.1]heptane skeleton. Tetrahed Lett, 2, 3061–3064. Sih CJ, Heather JB, Peruzzotti GP et al (1973) Total synthesis of prostaglandins IV. A completely stereospecific synthesis of prostaglandin E1. J Am Chem Soc, 95, 1676–1677. Sih CJ, Salmon RG, Price P et al (1975a) Stereospecific total synthesis of prostaglandins via reaction of a-alkylcyclopentenones with organocuprates. J Am Chem Soc, 97, 857–865. Sih CJ, Heather JB, Sood R et al (1975b) Asymmetric total synthesis of (7)-prostaglandin E1 and (7)-prostaglandin E2. J Am Chem Soc, 97, 865–874. Skuballa W, Schillinger E, Sturzebecher C-St and Vorbru¨ggen H (1986) Synthesis of a new chemically and metabolically stable prostacyclin analogue with high and long-lasting oral activity. J Med Chem, 29, 313–315. Skuballa W and Vorbru¨ggen H (1983) Synthesis of Ciloprost (ZK 36374): a chemically stable and biologically potent prostacyclin analog. In Advances in Prostaglandin, Thromboxane and Leucotriene Research, Samuelsson B, Ramwell PW, Paoletti R (eds). Raven, New York, 303– 305. Sugiura T, Kodaka T, Nakane S et al (1999) Evidence that the cannabinoid CB1 receptor is a 2-arachidonyl-glycerol receptor. J Biol Chem, 274, 2794–2801. Suzuki M, Kato K, Noyori R et al (1996) (15R)-16-m-Tolyl-17,18,19,20tetranorisocarbacyclin: a stable ligand with high binding affinity and
selectivity for a prostacyclin receptor in the central nervous system. Angew Chem Int Ed, 35, 334–336. Suzuki M, Yanagisawa A and Noyori R (1988) The three-component coupling synthesis of prostaglandins. J Am Chem Soc, 110, 4718–4726. Suzuki M, Morita Y, Koyano H et al (1990) Three-component coupling synthesis of prostaglandins. A simplified general procedure. Tetrahedron, 46, 4809–4822. Takechi H, Matsumura K, Watanabe Y et al (1996) A novel subtype of the prostacyclin receptor expressed in the central nervous system. J Biol Chem, 271, 5901–5906. To¨mo¨sko¨zi I, Gruber L, Kovacs, G et al (1976) RegiospecificPrins Reaction, a new way to prostanoids. Tetrahed Lett, 4639–4642. To¨mo¨sko¨zi I, Galambos G, Simonidesz V and Kovacs G (1977) A simple synthesis of PGI2. Tetrahed Lett, 2627–2628. To¨mo¨sko¨zi I, Galambos G, Simonidesz V and Kovacs G (1978) Stereochemistry of 5-bromo- and 5-iodo-5,6-dihyroprostacyclin and related bicyclic ethers. Tetrahed Lett, 581–584. Trost BM, Timko JM and Stanton JL (1978) An enantio-convergent approach to prostanoids. Chem Commun, 436–438. Tymkewycz PM, Jones RL and Wilson NH (1991) Heterogeneity of thromboxane A2 receptors: evidence from antagonist but not agonist potency measurements. Br J Pharmacol, 102, 607–614. Vane JR and Botting RM (1995) A better understanding of antiinflammatory drugs based on isoforms of cyclooxygenase (COX-1 and COX-2). In Advances in Prostaglandin, Thromboxane and Leucotriene Research, Samuelsson B et al (eds). Raven, New York, 41–48. Vane JR and O’Grady J (1993) Therapeutic Applications of Prostaglandins. Edward Arnold, London. Weinheimer AJ and Spraggins RL (1969) The occurrence of two new prostaglandin derivatives (15-epi-PGA2 and its acetate methyl ester) in the gorgonian Plexaura homomalla: chemistry of coelenterates, XV. Tetrahed Lett, 5185–5188. Westermann J, Harre M and Nickisch K (1992) E- or Z-selective Wittig reactions in the synthesis of the carbacyclin Iloprost. Tetrahed Lett, 33, 8055–8056. Willson TM, Brown PJ, Sternbach DD and Henke BR (2000) The PPARs: from orphan receptors to drug discovery. J Med Chem, 43, 527–550. Wilson NH and Jones RL (1985) Prostaglandin endoperoxide and thromboxane A2 analogues. Advances in Prostaglandin, Thromboxane and Leucotriene Research, Pike JE and Morton DR (eds). Raven, New York, 393–425. Wise H and Jones RL (1996) Focus on prostacyclin and its novel mimetics. Trends Pharmacol Sci, 17, 17–21. Wise H and Jones RL (2000) Prostacyclin and Its Receptors. Kluwer Academic/Plenum, Dordrecht. Woollard PM (1986) Stereochemical differences between 12-hydroxy5,8,10,14-eicosatetraenoic acid in platelets and psoriatic lesions. Biochem Biophys Res Commun, 136, 169–175. Yoshimoto T, Yamamoto S, Okuma M and Hayaishi O (1977) Solubilisation and resolution of thromboxane synthesizing system from microsomes of bovine blood platelets. J Biol Chem, 252, 5871– 5874. Yoshimoto T and Yamamoto S (1995) Arachidonate 12-lipoxygenase. J Lipid Mediat Cell Sign, 12, 195–212.
Section Two Analytical Methods
8 Perspectives of Analytical Methods for Eicosanoids Jay Y. Westcott1 and Angelo Sala2 1National
Jewish Medical and Research Center, Denver, CO; and 2Department of Pharmacological Sciences, Milan, Italy
The methods that have been utilized to measure eicosanoids are diverse and reflect the chemistry and physiology of these bioactive lipids (Andersen et al, 1985). Early investigators utilized bioassay techniques to discover arachidonic acid metabolites in biological fluids. Later, mass spectrometry was utilized to determine the chemical structure of these lipids we now call prostaglandins and leukotrienes. Immunoassays, first radioimmunoassays and then enzyme immunoassays, became the workhorse methods for the quantification of eicosanoids in a wide variety of biological fluids, cell culture supernatants and tissues. Immunoassay technology continues to advance with the development of time-resolved fluoroimmunoassays, which can be easier, faster and more sensitive than the more traditional immunoassays. This chapter briefly discusses some general considerations for the analysis of eicosanoids before focusing on each of the four analytical techniques mentioned above. GENERAL CONSIDERATIONS Eicosanoids are hydrophobic lipids, typically being more soluble in organic solvents than in aqueous media. This hydrophobicity in some cases can be extreme, with the eicosanoid sticking to glass or plastic rather than staying soluble in an aqueous buffer. This is often most pronounced in aqueous media with very low protein, and the inclusion of even 0.1% albumin is usually enough to keep eicosanoids in solution. Stock solutions of eicosanoids are usually kept in organic solvents, such as methanol or ethanol, to help keep them in solution. Keeping them in these solvents can at times increase the stability of eicosanoids, although this is not always the case. When adding even small concentrations (0.5%) of solvents with eicosanoids to biological preparations, it is important to do a vehicle control, as these solvents can have a variety of effects on their own (Westcott and Murphy 1985). The enzymatic catabolism of eicosanoids in vivo can be very rapid. Prostaglandin (PG)E2 is quickly cleared from the circulation during a single pass through the lung (Robinson et al 1985). Peptidoleukotriene metabolizing enzymes are prevalent both on the cell surface and in fluids such as blood and urine (Kuo et al 1984). Thus, it is important to measure the correct eicosanoid in the correct biological fluid, e.g. in urine leukotriene (LT)E4 is the correct cysteinyl-leukotriene to measure, since its precursors would be metabolized to this product in either blood or urine (Westcott et al 1991). In some fluids, such as epithelial lining fluid in the lung, there could be a mix of leukotrienes and an assay system that can measure LTC4, LTD4 and LTE4 is desirable (Wenzel et al 1990). In urine there can be both the prostacyclin The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
metabolite 6-keto-PGF1a (felt to be an index of kidney prostacyclin production) as well as another metabolite 2,2-dinor-6-ketoPGF1a (an index of extrarenal prostacyclin production). The urinary prostacyclin metabolite to be measured is dependent on what compartment the investigator desires to investigate. In addition to enzymatic degradation, eicosanoids are subject to non-enzymatic oxidation and hydrolysis. Arachidonic acid can be oxidized to a variety of hydroxy-acids, which can be shown to be different from the enzymatic products by being a fairly equal racemic mix rather than one specific stereoisomer (Fretland and Djuric 1989). Arachidonic acid can also oxidize to form isoprostanes, which are thought to be indicators of the oxidative stress in a system (Morrow and Roberts II 1996). Prostacyclin has a half-life of minutes in plasma before it is hydrolysed (Lucas et al 1986). The cysteinyl leukotrienes also become oxidized to a variety of products (Westcott et al 1984) that can catalyse further degradation, such that once degradation begins it can accelerate quickly. The warning here (particularly for leukotriene quantification) is to be very wary of eicosanoid standard concentrations, especially if they have been stored for 6 months or more. For any analytical technique it is important to perform some type of recovery experiment. In these experiments a spiked amount of the compound to be measured is added to various components of the system and the resultant amount recovered at the end of analysis is determined. These types of studies are often performed early on in assay development and neglected by later investigators who often take this work for granted. However, there are situations in eicosanoid analysis in which the individual investigator should perform additional assay validation. For example, when a complex system is being studied and no eicosanoid is found, it is imperative to add the eicosanoid to the system to show that it could be found if it were indeed present in the system. Metabolism or uptake of the eicosanoid could have occurred, which would have prevented finding the eicosanoid in the sample although it was indeed produced in the system. A second situation is when a sample has to undergo substantial purification or concentration prior to quantification. Since eicosanoids are both ‘‘sticky’’ and prone to degradation/metabolism, it is important to have some idea of what recovery is. In instances where multiple samples are routinely run, it is a reasonable practice to spike one sample with the eicosanoid to be measured, treat it as an unknown and run it through all the analytical steps. For example, this should be done in the quantification of urinary LTE4, where an immunoaffinity resin is used to initially bind the leukotriene followed by centrifugation to remove the resin, washing, methanol elution, evaporation and final quantification by enzyme immunoassay (Westcott et al 1998).
98
ANALYTICAL METHODS
BIOASSAY The bioassay represents a unique technique which is capable of detecting biological activities in samples (Kellaway and Trethewie 1940). Although today a number of sophisticated techniques are available for the immunological or physicochemical detection of eicosanoids, it is important to remember that properly mastered bioassays contributed crucial information to the development of the eicosanoid field. For example, the discoveries of prostacyclin (PGI2) and thromboxane (TXA2) were driven by their unique behaviours in bioassay systems (Piper and Vane 1969) and it is questionable whether the presence of such unstable compounds would have been recognized without bioassay techniques. This was clearly recognized by the assignment of the Nobel Prize for Physiology or Medicine to Sir John Vane in 1982 (Vane 1983). Bioassays were also utilized to elucidate the mechanism of action of non-steroidal antiinflammatory drugs (NSAIDs) and in the identification of early antagonists of slow-reacting substances of anaphylaxis (Augstein et al 1973; Vane 1971). Numerous examples exist for the use of bioassays in the quantitation of eicosanoids (Fleisch et al 1979). Before the development of specific immunoassays, bioassays were among the most accessible methods for sensitive semi-quantitative determination of biologically active compounds. A possible problem associated with the use of bioassays for quantitative analysis is that additional endogenous substances can also be present which affect the responsiveness of isolated organs to specific eicosanoids, and vice versa (Gryglewski and Korbut 1976). For example, platelet activating factor (PAF), when released from pulmonary tissues following hypersensitivity reactions, may cause a 10-fold increase in the contracting activity of cysteinyl leukotrienes on the isolated guinea-pig ileum, even at sub-nanomolar concentrations (Clancy and Hugli 1983). While this can be regarded as a significant pitfall in quantification, the occurrence of this phenomenon presents a clearer picture of the final biological relevance (in terms of the biological activity measured) linked to the production of a given metabolite. Considerable specificity can be achieved using bioassay methodology. An appropriate selection of assay organs can be set up in a cascade so that the substance to be tested sequentially superfuses a series of different isolated organ preparations. The typical pattern of biological activities (contraction, relaxation or no effect) observed in the different preparations can provide a unique fingerprint, allowing discrimination between different eicosanoids, e.g. while both PGI2 and PGE2 relax the rabbit mesenteric artery, the latter contracts the bovine coronary artery while the former relaxes it. By utilizing five or six different isolated organ preparations, it is indeed possible to achieve remarkable specificity. The specificity of very simple bioassays can also be substantially increased by using appropriate mixtures of receptor antagonists/synthesis inhibitors added to the superfusion buffer (Gilmore et al 1968). It is also common for NSAIDs to be supplemented to systems to inhibit the generation of endogenous prostaglandins that might occur in response to superfusion. There exist numerous variations on the typical isolated organ bioassay. An interesting evolution of a typical superfusion cascade was introduced by Vane (1964), in which heparinized blood taken from an artery by means of a peristaltic pump was used to superfuse an isolated organ cascade and then returned to the animal by gravity or using another pump. In another bioassay variation, increased sensitivity was achieved by a laminar flow superfusion technique (Ferreira and De Souza Costa 1976). This modification of the classical superfused organ cascade lowers the superfusion flow to 0.1–0.3 ml/min, substantially prolonging the contact time between tested substances and the isolated organs used for detection. This leads to a significant increase in
sensitivity, such that as little as 5–10 pg PGE2 or 50–100 pg cysteinyl-leukotrienes can be easily detected. The technique of bioassay represents not only ‘‘a way of biological thinking’’, as stated by Professor R. Gryglewski, but is a versatile tool in the hands of experienced pharmacologists. Bioassays can provide highly reliable quantitative and qualitative information of unique importance in the discovery of biologically active unstable products. Recent technological advances have provided additional valuable support for the use of bioassays in the analysis of eicosanoids. For example, the use of computerassisted video-microscopy provides enhanced sensitivity by detecting changes in smooth muscle size of the order of a few microns while providing recording of smooth muscle activity in conditions very similar to that observed in vivo (Wilkens et al 1992). Parallel use of analytical techniques providing structural information, such as liquid chromatography–mass spectrometry (LC–MS), with bioassay will continue to represent a powerful approach for the unequivocal identification of bioactive arachidonic acid metabolites (Subbanagounder et al 2002). Mass Spectrometry The structure elucidation of prostaglandins and their metabolites in the 1960s was one of the most successful early applications of gas chromatography–mass spectrometry (GC–MS) to the solution of biological structural problems (Bergstrom et al 1962). Less than 20 years later, MS again played a crucial role in the structural identification of the slow reacting substance of anaphylaxis (SRS-A) as a cysteine containing oxygenated arachidonic acid derivative named leukotriene C (Murphy et al 1979). The importance of MS in the field of eicosanoid research can easily be observed in the several thousand publications pertaining to MS of eicosanoids. In addition to the pivotal role of MS in the identification of prostaglandins and leukotrienes, this technique has also been widely applied to the quantitative analysis of arachidonic acid metabolites and in validating other methodologies. Over the years, MS analysis has represented the ‘‘golden standard’’ for the validation of different analytical techniques and confirming (or not) immunoassay results. For example, during the early 1970s several laboratories reported peripheral plasma levels of PGA compounds (PGA1+PGA2) measured by RIA at about 1 ng/ml and higher, suggesting that PGA might represent a circulating hormone (Attallah et al 1974; Pletka and Hickler 1974). However, no PGA metabolites could be detected (detection limit 5 pg/ml) when specific GC–MS techniques were utilized, strongly implicating the presence of interfering cross-reacting substances in the immunoassays (Green and Steffenrud 1976). The basis for MS methods in the quantitation of eicosanoids is the use of appropriate internal standards added immediately after collection of samples. These internal standards are usually stable isotope-labelled analogues of the eicosanoids to be measured, i.e. synthetic eicosanoids in which a given number of hydrogen, carbon-12 or oxygen-16 isotopes are replaced by the corresponding stable isotopes deuterium, carbon-13 or oxygen-18. After purification, derivatization and chromatographic separation, the ratio between the ions arising from stable isotope-labelled and -unlabelled molecules is measured by mass spectrometers, allowing calculation of the amount of unlabelled compound present in a given sample. These stable isotope-labelled analogues have physical/chemical properties almost identical to the unlabelled molecules and thus co-elute during purification and derivatize with similar efficiency. This appropriate use of stable isotope-labelled analogues of eicosanoids as internal standards corrects for any loss of analyte arising from less than quantitative extraction or derivatization, and also
PERSPECTIVES provides a significant carrier effect, helping to improve recovery of trace amounts of the compound of interest (Green et al 1973). Although capable of providing an excellent analytical approach for the specific and sensitive quantitation of eicosanoids, the use of MS has several problems or disadvantages. Probably the biggest drawback is that MS still requires very expensive instrumentation and highly trained personnel that are not available in all laboratories. Second, while extremely sensitive, many GC–MS applications require extensive extraction and purification of samples before analysis, in particular when complex biological matrices, such as plasma or urines, are involved. Fairly simple solid phase extraction on normal or reverse phase cartridges has been successfully adopted in a few cases, resulting in significantly improved signal:noise ratios. However, in many cases either thin layer chromatography (TLC) or high pressure liquid chromatography (HPLC) are required to attain satisfactory purification of the sample. In addition to the need to purify many samples, another major drawback for methodologies based on GC is the absolute requirement for volatility and thermal stability of the compounds to be analysed at the temperatures used for GC. This prerequisite initially precluded the use of GC–MS for the analysis of polar or thermally labile eicosanoids, such as the cysteinyl-leukotrienes. Nevertheless, catalytic desulphurization to yield the corresponding 5-hydroxyeicosatetraenoic acid (5-HETE) followed by derivatization and GC–MS was successfully used to quantitate the total amount of LT in biological samples (Murphy and Sala 1990). Different chemical reactions involving the functional groups present on eicosanoids have also been used to increase the volatility and thermal stability of eicosanoids, such as esterification of the carboxy group, methoximation of the keto group and trimethylsilylation of the hydroxy group. The rapid progress of analytical methods observed with other technologies has also occurred in the field of MS. The introduction of more sensitive instrumentation coupled with new ionization techniques has allowed significant advancement in the analysis of eicosanoids in complex matrices. Multiple-sector mass spectrometers allow for selected reaction monitoring which, together with chromatographic resolution, results in extreme selectivity while still maintaining excellent signal:noise ratios as a result of a very low background (Dawson et al 1988; Hughes et al 1988). New ionization techniques, such as electrospray ionization and negative ion/electron capture-chemical ionization (Barrow et al 1982), have created new opportunities for the analysis of compounds previously not suitable for MS as intact molecules, such as LTE4 (Wu et al 1996). A major advantage of electrospray ionization is that it allows the direct sampling of the effluent from HPLC analysis into tandem mass spectrometers (LC–MS–MS). In addition to expanding the number of eicosanoids that can be analysed by mass spectrometry, LC–MS–MS substantially decreases the required sample manipulation, resulting in a significant decrease in the total analysis time and allowing for a completely automated analytical protocol to be utilized (Kishi et al 2001). The introduction in recent years of new, lower-cost mass spectrometers with MS–MS capability is also an advantage that might allow for wider utilization of MS techniques. Enzyme Immunoassays The first reported immunoassay was a radioimmunoassay (RIA) for insulin in 1959 (Yalow and Berson 1959). This new methodology was quickly accepted and its use expanded to thousands of other large and small molecular weight compounds (including eicosanoids) in both clinical and basic research applications (Dray et al 1975). This technique was considered so important that Rosalyn Yalow was awarded the Nobel Prize for
99
Physiology or Medicine in 1977 (Yalow 1978). Numerous variations of this RIA have appeared, of which enzyme immunoassays (EIA) became widespread in use after their first description in 1971 (Engvall and Perlman 1971; Van Weemen and Schuurs 1971). The use of EIAs to quantify prostaglandins and leukotrienes rapidly replaced radioimmunoassays (RIA) for the measurement of these lipids. The success of immunoassays has been due largely to the specificity of antibodies and the sensitivity of detection systems utilized. Both RIAs and EIAs for eicosanoids utilize a specific antibody for the eicosanoid to be measured, a labelled tracer eicosanoid modified so that it can be easily detected, and a method to separate what is bound to the antibody from what is not bound. A competition takes place between the labelled tracer and the eicosanoid itself (either in an unknown sample or a standard) for the small amount of specific antibody available. For RIA and the eicosanoid could be labelled with radiolabelled iodine or tritium, the separation of bound from unbound tracer taking place by absorption of the small, unbound component to added charcoal. The charcoal is removed by centrifugation and the radioactivity in the supernatant measured using a scintillation counter. In EIAs, the tracer is the eicosanoid covalently linked to an enzyme (usually acetylcholinesterase with eicosanoids). The antibody is made to adhere to a polystyrene plate (often utilizing an additional species-specific antibody), separating what is bound to it from what remains in solution. After allowing a suitable time (2–24 h) for binding to occur, what is not bound is washed away. The amount of tracer bound is then determined after adding substrate by measuring enzyme activity (the enzymatic conversion of a colourless substrate to a coloured product). Figure 8.1 shows the main steps in a simple eicosanoid EIA. The specific antibody can be produced in a rabbit, mouse, goat or other species and the same antibody will often work in both RIAs and EIAs. However, just because an antibody works in an RIA does not mean it necessarily will work in an EIA. One reason for this is that in EIAs the small molecular weight eicosanoid is linked to a much larger enzyme, such that now an antibody might completely fail to recognize it, or an antibody might recognize it much better than the unlabelled compound. In either case there is not a good ‘‘competition’’ of tracer and unlabelled eicosanoid for the antibody and the assay will not work well. Pradelles, Grassi and Maclouf first reported the use of EIAs in the quantification of prostaglandins in 1985 (Pradelles et al 1985). One unique aspect of this assay was the utilization of acetylcholinesterase as the detection enzyme instead of horseradish peroxidase (HRP) or alkaline phosphatase. Acetylcholinesterase has a variety of advantages, including great stability, allowing synthesized tracer to be frozen (unlike HRP) or lyophilized and remain usable for years (usually until the eicosanoid degrades). Acetylcholinesterase is also very active in producing product and is not inactivated during activity (like HRP). Finally, the enzyme can be linked in subunits of two to four, which effectively amplifies the activity of the enzyme complex. These advantages are especially important for competitive assays in which enzyme activity and stability are crucial for a sensitive assay. It should be stressed that commercial EIAs utilizing other enzymes are available and can work well. How well an assay works depends not only on the type of tracer utilized but, most importantly, on how specific and how avid the antibody is. Although competitive EIAs are generally utilized to quantify eicosanoids, another type of immunoassay called a sandwich enzyme-linked immunosorbent assay (ELISA) can be utilized for the quantification of larger molecular weight antigens. Two antigen-specific antibodies are utilized in sandwich ELISAs, one to trap the antigen on a surface and the other to detect the antigen. Thus, although sandwich ELISAs can be utilized to measure enzymes involved in prostaglandin synthesis, such as
100
ANALYTICAL METHODS
Figure 8.1 Methodology of competitive enzyme immunoassay of eicosanoids utilizing an acetylcholinesterase-labelled tracer (figure provided by Dr Kirk Maxey, Cayman Chemical, Ann Arbor, MI)
cyclooxygenase-1 and -2 (Creminon et al 1995), competitive EIAs are generally required for the eicosanoids themselves. Compared to sandwich ELISAs, competitive EIAs have increased risk of cross-reactivity problems, since only one specific antibody for an eicosanoid is utilized. Competitive immunoassays are inherently more affected by matrix effects, especially the pH and protein concentration of the buffer, and are also typically less sensitive than sandwich ELISAs. Finally, unlike sandwich ELISAs in which the antibodies are typically present in excess, competitive assays have a trade-off in which the assay can be made more sensitive by decreasing the antibody or tracer concentration but at the cost of the assay taking longer to develop. It is often unclear whether samples need to be purified prior to performing immunoassays for eicosanoids. The answer is a definitive—it depends. Typically, cell culture supernatants do not have to be purified before analysis. That having been said, the presence of serum in these samples can cause increased background in some cases (TXB2 or PGE2 can be present in serum) or can add metabolic enzymes (e.g. in the quantification of LTC4). Serum samples are usually not utilized to quantify what is in blood because of low endogenous levels and the risk for production during collection. However, blood samples can be utilized ex vivo to provide an approximation of the total metabolic capacity of blood cells (often in the presence of in vivo treatment with potential inhibitors). Urine samples require purification for some eicosanoids but not for others, so it is advisable to check with the assay vendor for the validation of data available under different conditions (purified/unpurified). Urine LTE4 is one eicosanoid that requires fairly extensive purification prior to quantification (Westcott and Taylor 1998). Other fluids, e.g. bronchoalveolar lavage fluid, do not require purification as much as concentration, so that they can be processed efficiently using solid phase extraction cartridges. Tissue homogenates are occasionally utilized to measure eicosanoids but, as with serum, it is very difficult to get endogenous eicosanoid levels in tissue; rather, one can get a synthetic capability approximation. Tissue can be homogenized in an aqueous buffer (with or without an inhibitor) or in methanol to extract eicosanoids and precipitate proteins. EIAs are the most common method utilized for the quantification of eicosanoids and most likely will continue to be for at least the next 5–10 years. There are several reasons for this. The assay is
sensitive (often in the low pg/ml range), which makes it typically more sensitive than other methods. The assay is relatively easy to perform and can be performed quickly or at least does not require much labour/time. This type of analysis does not require unique expensive equipment, although a plate reader costing as little as $3000 is required. EIAs utilize reagents that are commercially available and should provide consistency over prolonged periods. The cost of these reagents when compared to other analytical alternatives makes it reasonably priced. Although this technique is not as specific as GC–MS (often considered the gold standard for analysis), many antibodies are very specific and are quite suitable for this desired task. TIME-RESOLVED FLUOROIMMUNOASSAYS Fluoroimmunoassays represent the most recently developed variation of the RIA theme. In this version, the labelled tracer used in competitive assays (or the enzyme/biotin-labelled detection antibody in sandwich ELISAs) has been replaced with a fluorescent-labelled molecule (Soini and Hemmila 1979). This type of assay has the potential advantages of increased sensitivity, faster assay time and use of stable reagents. Fluoroimmunoassays have been utilized to measure a wide variety of compounds, including bendazac and metabolites (Staley et al 1988), cytokines (Kimura et al 2001), steroids (Fiet et al 2001) and immunoglobulin E (Yuan et al 1997), to name a few. Hundreds of published papers describing fluoroimmunoassays have appeared over the past 10 years, but only a few have been concerned with the analysis of eicosanoids. Luke and Schlegel (1990) described an early assay for PGF2a and later other assays for metabolites of PGE and PGF (Luke and Schlegel 1992). These assays utilized biotin to label the prostanoid and streptavidin to add to the fluorescent europium (Luke and Schlegel 1992). Detection sensitivities of these assays were between 0.2 and 1.2 pg/ assay which, although better than RIAs, were similar to EIAs. Fluoroimmunoassays were also utilized to quantify cytochrome P450 epoxygenase products of arachidonic acid (Nithipatikom et al 1997). This assay utilized fluorescein-labelled 14,15-dihydroxyeicosatrienoic acid as the labelled tracer to compete with the unlabelled eicosanoid. The assay sensitivity was approximately 3 pg/ml. Finally, a two-site sandwich-type immunofluorimetric
PERSPECTIVES assay was developed for PGD synthase, which was sensitive (50 pg/ml), accurate and very specific for this enzyme (Melegos et al 1996). Although the first report of the use of fluoroimmunoassays for prostaglandins was 10 years ago, there has been only minimal subsequent work reported utilizing this type of assay. There have been several reasons for this, but probably the main reason is the lack of pre-made commercial reagents. Although the tracer reagent is supposedly easy and quick to synthesize, most investigators do not want to spend this time and effort, especially if it is required to be made often. Second, this assay utilizes a specialized fluorescent reader for microplates that many laboratories do not have. Finally, this assay is not that much more sensitive than the commercial EIAs that utilize acetylcholinesterase. This is in large part due to the advantages of the acetylcholinesterase-labelled tracer compared to radioisotopes used in RIAs. There are some advantages that could drive considerable investments in fluoroimmunoassays of eicosanoids. One of the unique aspects of fluoroimmunoassays is that multiple different fluorescent tags (for different antigens) with different spectral emissions can be utilized at the same time and quantified in the same sample (Merio et al 1996; Zerwes et al 2002). For example, Zerwes et al (2002) utilized europium, samarium and terbium to label antibodies recognizing VCAM-1, E-selectin and ICAM-1 in a multiparameter screening assay. This saves time and sample volume compared to assaying each compound separately. The availability of standard methods to prepare reagents and universally used reagents could also make the assay much more user-friendly. Recently, a universal streptavidin-labelled europium reagent has been described that could be utilized in a wide range of assays that currently utilize biotin-labelled tracers and intermediates (Scorilas and Diamandis 2000). In the competitive assays of eicosanoids, biotin-linked eicosanoids would be synthesized and used with the streptavidin–europium in a fashion very similar to that described by Luke and Schlegel (1992). The biotin-labelled eicosanoids should be fairly easy to synthesize and would be quite stable.
CONCLUSION Eicosanoids play an important role in normal physiology as well as in pathophysiology. In addition to the numerous known metabolic pathways and arachidonate metabolites, new metabolic enzymes and new products will undoubtedly continue to be discovered over the coming years. In fact, a new cyclooxygenase-3 that is specifically inhibited by acetaminophen was recently reported (Chandrasekharan et al 2002). Analytical techniques, including bioassay, GC–MS, EIA and fluoroimmunoassay, will be extremely important in discovering these new products and in determining their biological roles in health and disease. While this present chapter has presented a brief introduction to these four analytical techniques commonly used in eicosanoid analysis, the following four chapters will examine these techniques in more detail.
REFERENCES Andersen NH, Hartzell CJ and De B (1985) Chemistry and structure of cyclooxygenase-derived eicosanoids: a historical perspective. Adv Prostaglandin Thromboxane Leukotriene Res, 14, 1–43. Attallah A, Payakkapan W, Lee J, Carr A and Brazelton E (1974) Letter: PGA: fact, not artifact. Prostaglandins, 5, 69–72. Augstein J, Farmer JB, Lee TB et al (1973) Selective inhibitor of slow reacting substance of anaphylaxis. Nature New Biol, 245, 215–217.
101
Barrow SE, Waddell KA, Ennis M et al (1982) Analysis of picomolar concentrations of 6-oxo-prostaglandin F1a in biological fluids. J Chromatogr, 239, 71–80. Bergstrom S, Ryhage R, Samuelsson B and Sjovall J (1962) Prostaglandins and related factors. J Biol Chem, 238, 3555–3564. Chandrasekharan NV, Dai H, Roos KL et al (2002) COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure, and expression. Proc Natl Acad Sci USA, 99, 13926–13931. Clancy RM and Hugli TE (1983) The extraction of leukotrienes (LTC4, LTD4 and LTE4) from tissue fluids: the metabolism of these mediators during IgE-dependent hypersensitivity reactions in lung. Anal Biochem, 133, 30–39. Creminon C, Frobert Y, Habib A et al (1995) Immunological studies of human constitutive cyclooxygenase (COX-1) using enzyme immunometric assay. Biochim Biophys Acta, 1254, 333–340. Dawson M, McGee CM, Brooks PM et al (1988) Quantification of leukotriene B4 in synovial fluid by gas chromatography/tandem mass spectrometry. Biomed Environ Mass Spectrom, 17, 205–211. Dray F, Charbonnel B and Maclouf J (1975) Radioimmunoassay of prostaglandins Fa, E1 and E2 in human plasma. Eur J Clin Invest, 5, 311–318. Engvall E and Perlman P (1971) Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of immunoglobulin G. Immunochemistry, 8, 871–874. Ferreira SH and De Souza Costa F (1976) A laminar flow superfusion technique with much increased sensitivity for the detection of smooth muscle-stimulating substances. Eur J Pharmacol, 39, 379–381. Fiet J, Giton F, Boudi A et al (2001) Development of a sensitive and specific new plasma 4-androstene-3,17-dione time-resolved fluoroimmunoassay (TR-FIA). Steroids, 66, 609–614. Fleisch JH, Zaborowsky BR, Cerimele BJ and Spaethe SM (1979) Computerization of a bioassay: quantitation of slow reacting substance of anaphylaxis (SRS-A). J Pharmacol Exp Therapeut, 209, 238–243. Fretland DJ and Djuric SW (1989) 12(R)- and 12(S)hydroxyeicosatetraenoic acids: chemistry, biology and pharmacology. Prostagland Leukotrienes Essent Fatty Acids, 38, 215–228. Gilmore NJ, Vane JR and Wyllie JH (1968) Prostaglandin released by the spleen. Nature, 218, 1135–1141. Green K, Granstrom E, Samuelsson B and Axen U (1973) Methods for quantitative analysis of PGF2, PGE2, 9,11-dihydroxy-15-keto-prost-5enoic acid and 9,11,15-trihydroxy-prost-5-enoic acid from body fluids using deuterated carriers and gas chromatography–mass spectrometry. Anal Biochem, 54, 434–453. Green K and Steffenrud S (1976) Method for quantitative analysis of PGA2 from plasma using deuterated carrier and gas–liquid chromatography–mass spectrometry. Anal Biochem, 76, 606–614. Gryglewski RJ and Korbut R (1976) Bioassay of histamine in the presence of prostaglandins. Br J Pharmacol, 56, 39–41. Hughes H, Nowlin J, Gaskell SJ and Parr VC (1988) Selected reaction monitoring during gas chromatography/mass spectrometry of eicosanoids. Biomed Environ Mass Spectrom, 16, 409–413. Kellaway CH and Trethewie ER (1940) The liberation of a slowreacting smooth muscle substance in anaphylaxis. Qu J Exp Physiol, 30, 121–145. Kimura H, Suzui M, Nagao F and Matsumoto K (2001) Highly sensitive determination of plasma cytokines by time-resolved fluoroimmunoassay: effect of bicycle exercise on plasma level of interleukin-1a (IL-1a), tumor necrosis factor-a (TNF-a), and interferon g (IFN-g). Anal Sci, 17, 593–597. Kishi N, Mano N and Asakawa N (2001) Direct injection method for quantitation of endogenous leukotriene E4 in human urine by liquid chromatography/electrospray ionization tandem mass spectrometry with a column-switching technique. Anal Sci, 17, 709–713. Kuo CG, Lewis MT and Jakschik BA (1984) Leukotriene D4 and E4 formation by plasma membrane bound enzymes. Prostaglandins, 28, 929–938. Lucas FV, Skrinska VA, Chisolm GM and Hesse BL (1986) Stability of prostacyclin in human and rabbit whole blood and plasma. Thrombosis Res, 43, 379–387. Luke FJ and Schlegel W (1990) A time-resolved fluoroimmunoassay for the determination of prostaglandin F2a. Clin Chim Acta, 189, 257–265.
102
ANALYTICAL METHODS
Luke FJ and Schlegel W (1992) Determination of prostaglandin metabolites in biological samples by competitive time-resolved fluoroimmunoassay. J Immunol Methods, 148, 217–223. Melegos DN, Diamandis EP, Oda H et al (1996) Immunofluorometric assay of prostaglandin D synthase in human tissue extracts and fluids. Clin Chem, 42, 1984–1991. Merio L, Pettersson K and Lovgren T (1996) Monoclonal antibody-based dual-label time-resolved fluorometric assays in a simplified one-step format. Clin Chem, 42, 1513–1517. Morrow JD and Roberts LJ II (1996) The isoprostanes. Current knowledge and directions for future research. Biochem Pharmacol, 51, 1–9. Murphy RC, Hammarstrom S and Samuelsson B (1979) Leukotriene C: a slow reacting substance from murine mastocytoma cells. Proc Natl Acad Sci USA, 76, 4275–4279. Murphy RC and Sala A (1990) Quantitation of sulfidopeptide leukotrienes in biological fluids by gas chromatography–mass spectrometry. Methods Enzymol, 187, 90–98. Nithipatikom K, Falck JR, Bhatt RK et al (1997) Determination of 14,15epoxyeicosatrienoic acid and 14,15-dihydroxyeicosatrienoic acid by fluoroimmunoassay. Anal Biochem, 246, 253–259. Piper PJ and Vane JR (1969) Release of additional factors in anaphylaxis and its antagonism by antiinflammatory drugs. Nature, 223, 29–35. Pletka P and Hickler RB (1974) Blood prostaglandin A (PGA) levels in normal human subjects. Prostaglandins, 7, 107–115. Pradelles P, Grassi J and Maclouf J (1985) Enzyme immunoassays of eicosanoids using acetylcholine esterase as label: an alternative to radioimmunoassay. Anal Chem, 57, 1170–1173. Robinson C, Hardy CC and Holgate ST (1985) Pulmonary synthesis, release, and metabolism of prostaglandins. J Allergy Clin Immunol, 76, 265–271. Scorilas A and Diamandis EP (2000) Polyvinylamine–streptavidin complexes labelled with a europium chelator: a universal detection reagent for solid-phase time resolved fluorometric applications. Clin Biochem, 33, 345–350. Soini E and Hemmila I (1979) Fluoroimmunoassay: present status and key problems. Clin Chem, 25, 353–361. Staley K, Coxon R, Gallacher G and Landon J (1988) Development of separation fluoroimmunoassays for bendazac and 5-hydroxy bendazac. Therapeut Drug Monit, 10, 321–326. Subbanagounder G, Wong JW, Lee H et al (2002) Epoxyisoprostane and epoxycyclopentenone phospholipids regulate monocyte chemotactic protein-1 and interleukin-8 synthesis. Formation of these oxidized phospholipids in response to interleukin-1b. J Biol Chem, 277, 7271– 7281.
Van Weemen BK and Schuurs AH (1971) Immunoassay using antigenenzyme conjugates. FEBS Lett, 15, 232–236. Vane JR (1964) The use of isolated organs for the detection of active substances in circulating blood. Br J Pharmacol, 23, 360–367. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231, 232–235. Vane JR (1983) Nobel lecture, 8 December 1982. Adventures and excursions in bioassay: the stepping stones to prostacyclin. Br J Pharmacol, 79, 821–838. Wenzel SE, Larsen GL, Johnston K et al (1990) Elevated levels of leukotriene C4 in bronchoalveolar lavage fluid from atopic asthmatics after endobronchial allergen challenge. Am Rev Resp Dis, 142, 112– 119. Westcott JY, Clay KL and Murphy RC (1984) Decomposition of leukotriene C4. J Allergy Clin Immunol, 74, 363–368. Westcott JY, Maxey KM, MacDonald J and Wenzel SE (1998) Immunoaffinity resin for purification of urinary leukotriene E4. Prostagland Lipid Mediat, 55, 301–321. Westcott JY and Murphy RC (1985) Effect of alcohols on arachidonic acid metabolism in murine mastocytoma cells and human polymorphonuclear leukocytes. Biochim Biophys Acta, 833, 262–271. Westcott JY, Smith HR, Wenzel SE et al (1991) Urinary leukotriene E4 in patients with asthma. Effect of airways reactivity and sodium cromoglycate. Am Rev Resp Dis, 143, 1322–1328. Westcott JY and Taylor IK (1998) In Five Lipoxygenase Products in Asthma, vol 120 Drazen JM, Lee TH and Dahlen S-E (eds). Marcel Dekker, New York, 245–281. Wilkens JH, Becker A, Wilkens H et al (1992) Bioassay of a tracheal smooth muscle-constricting factor released by respiratory epithelial cells. Am J Physiol, 263, L137–L141. Wu Y, Li LY, Henion JD and Krol GJ (1996) Determination of LTE4 in human urine by liquid chromatography coupled with ionspray tandem mass spectrometry. J Mass Spectrom, 31, 987–993. Yalow RS (1978) Radioimmunoassay: a probe for the fine structure of biologic systems. Science, 200, 1236–1245. Yalow RS and Berson SA (1959) Assay of plasma insulin in human subjects by immunological methods. Nature, 184, 1648–1649. Yuan J, Wang G, Kimura H and Matsumoto K (1997) Highly sensitive time-resolved fluoroimmunoassay of human immunoglobulin E by using a new europium fluorescent chelate as a label. Anal Biochem, 254, 283–287. Zerwes HG, Peter JC, Link M et al (2002) A multiparameter screening assay to assess the cytokine-induced expression of endothelial cell adhesion molecules. Anal Biochem, 304, 166–173.
9 Enzyme Immunoassays of Metabolites and Enzymes Using Acetylcholinesterase as Label Christophe Cre´minon1 and The Late Jacques Maclouf 1CEA-Saclay,
2*
Gif s/Yvette Cedex; and 2U348 INSERM, Hoˆpital Lariboisie`re, Paris, France
Eicosanoids, including prostaglandins (PGs), thromboxanes (TXs), leukotrienes (LTs) and hydroxyeicosatetranoic acids (HETEs), are a class of molecules derived from 20-carbon essential fatty acids. The oxidative modifications of the arachidonic acid backbone (this compound being the main precursor in humans) can follow two different pathways, either enzymatic or non-enzymatic, leading to the production of eicosanoids or isoeicosanoids, respectively. Our knowledge of eicosanoids has grown considerably over the last 10 years. Different biosynthetic pathways have been demonstrated, viz. constitutive, inducible and transcellular. The list of non-enzymatically generated isoeicosanoids is constantly being enriched by newly identified compounds. While the physiological importance of eicosanoids has been known for many years, these recent findings have opened up new lines of research for numerous scientists of differing backgrounds. In addition to the scientific interest of these different themes, these new molecules appear to be promising targets for pharmacological purposes, due to their probable involvement in pathophysiological processes, particularly in vascular diseases. Identification and characterization of eicosanoids are generally achieved using molecular biology or mass spectrometry techniques for macromolecules (receptors or enzymes) and small active mediators (isoeicosanoids), respectively. Routine analysis of these different molecules clearly requires suitable analytical methods. Antibodies thus appear as reagents of choice due to their high specificity and affinity. These properties have been advantageously exploited for over 30 years in quantitative and specific immunoassays, in combination with labelled molecules of high specific activity. Initially developed using (125I or 3H) radiolabelled molecules (Maclouf et al 1976), these methods now tend to involve non-isotopic labelling, particularly enzymatic tracers (Hayashi et al 1983), which are at least as sensitive, more stable and subject to fewer safety-related constraints. We have been developing such systems for many years and have applied this strategy to measure PGs, LTs and TXs using acetylcholinesterase (AChE; EC 3.1.1.7) from Electrophorus electricus as a label (Pradelles et al 1985, 1990), in combination with a chromogenic non-toxic substrate for the staining step. The use of acetylcholinesterase for the labelling of molecules (haptens, antigens or antibodies) presents numerous advantageous features, such as a high specific activity, a very low non-specific binding (in spite of a 320 kDa molecular weight) and the possibility of using different chemical strategies to label molecules over a large pH range. Moreover this enzyme *Deceased, 14 July 1998 The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
provides a continuous signal for more than 24 h, offering the possibility of increasing the precision and the sensitivity of the immunoassay by allowing the enzymatic reaction to proceed for longer periods. In this chapter, we will focus on findings concerning recently identified molecules (specific immunoassays for 8-epi-PGF2a, cyclooxygenases 1 and 2) and the development of a new method to assay LTC4. DEVELOPMENT OF COMPETITIVE ENZYME IMMUNOASSAY FOR 8-epi-PGF2a Background Recent observations have shown that lipid peroxides accumulate in certain diseases due to attack by free radicals produced in atherosclerosis or myocardial infarction (Patrono and FitzGerald 1997). The recent discovery of a new family of biologically active derivatives of arachidonic acid (the F2-isoprostanes), formed by a non-enzymatic mechanism catalysed by free radicals, may yield information on illnesses linked to oxidative stress. One of these compounds, 8-epi-PGF2a, acts as a powerful vasoconstrictor and its action appears to involve participation of the TXA2/PGH2 vascular receptor. We have produced polyclonal antibodies against this compound to develop a competitive enzyme immunoassay (EIA). Strategy All eicosanoids are low molecular weight substances (haptens) which must be covalently attached to a macromolecule (antigenic carrier) to elicit antibody production. Like most eicosanoids, 8epi-PGF2a has but one useful chemical group, a carboxylic function, which was activated to allow covalent coupling to KLH (immunogen for antibody production) or to AChE (enzyme tracer). This strategy previously proved to be successful for developing a competitive enzyme immunoassay for a large set of eicosanoids (Table 9.1). Rabbit polyclonal antibodies were obtained and tested using the 8-epi-PGF2a-AChE tracer. A competitive enzyme immunoassay was developed (Wang et al 1995). Assay sensitivity and specificity were analysed. The assay was applied to the measurement of 8-epi-PGF2a in biological samples and validated by correlation with other analytical methods (RIA, GC–MS).
104
ANALYTICAL METHODS
Table 9.1 Sensitivity of the different competitive enzyme immunoassays of eicosanoids using acetylcholinesterase as label Compound PGB2 PGD2-MO PGE2 Bicyclo-PGE2 6-Keto-PGF1a 11b-PGF2a 13,14-Dehydro-PGF2a PGF2a TXB2 2,3-Dinor-TXB2 11-Dehydro-TXB2 LTB4 LTC4 LTE4 12-HETE 15-HETE
IC50 (pg/ml) 38 80 110 10 60 55 400 70 70 200 180 30 100 140 1000 740
Results As shown in Figure 9.1A, the competitive enzyme immunoassay was very sensitive (B/B0 50% &8 pg/ml). It was also highly specific, with less than 0.1% cross-reactivity with related compounds such as 8-epi-PGE2 or PGF2a. Analysis of the same urine samples by two independent methods, EIA and GC–MS, yielded very similar results, thus validating the assay. Although no circadian variation of the compound in urine was observed (Figure 9.1B), preliminary results revealed a positive correlation between dose and the subject’s age. Urinary elimination was not modified by the treatment of volunteers with two non-steroidal antiinflammatory drugs (NSAIDs). However, the analysis of plasma samples indicated low but real production of 8-epi-PGF2a by an enzymatic mechanism (cyclooxygenase-dependent), a synthesis that was completely suppressed by aspirin.
PRODUCTION OF SPECIFIC ANTIBODIES AND DEVELOPMENT OF SPECIFIC ENZYME IMMUNOMETRIC ASSAYS FOR CYCLOOXYGENASES 1 AND 2, KEY ENZYMES IN ARACHIDONIC ACID METABOLISM Background Cyclooxygenase (COX, also known as PGH synthase; EC 1.14.99.1) is the first enzyme that intervenes in the biosynthesis of prostaglandins and thromboxanes starting with arachidonic acid. This activity therefore controls the synthesis of molecules acting as mediators in numerous cell functions. Thus an increase in prostaglandins accompanies the majority of clinical inflammation symptoms. Recent data have shown that there are two isoforms of this enzyme (Smith et al 1996), which is the target of numerous NSAIDs, such as aspirin, indomethacin and ibuprofen. The first form (COX-1) appears to be constitutively expressed and its production is only slightly affected by stimuli and hormones. This 70 kDa protein was cloned, purified and sequenced from various tissue and animal sources. More recently, a new form (COX-2) of similar molecular weight was identified as the product of an early response gene. Its expression can be induced by numerous external ligands. This novel form was cloned and its primary sequence determined for different species. COX-1 presents more than 90% between-sequence homology, whereas the homology is about 80% for COX-2. Comparison of the COX1 and COX-2 sequences in a given species shows around 60% similarity, with extensive conservation of the major residues involved in enzymatic activity. However, a significant difference is observed between these two sequences. There is a significant deletion in the amino-terminal part of COX-2 (including a large part of the COX-1 peptide signal sequence) compensated for by the insertion of an 18-residue sequence in the C-terminus (absent in COX-1). Strategy
Prospects This method could facilitate a rationally-based evaluation of antioxidant drugs in humans. Moreover, these compounds have a longer life than hydroperoxides, for example, and could therefore be used as stable markers in clinical studies of oxidizing reactions.
Detailed comparative analysis of human COX-1 and COX-2 sequences has identified different limited zones specific to each of the two isoforms. We have thus raised several polyclonal and monoclonal antibodies against synthetic peptides (Figure 9.2). Purified COX-1 (from ram seminal vesicles) was used to raise polyclonal and monoclonal antibodies, since the extensive conservation of the inter-species primary structure could lead to
Figure 9.1 (Panel A) Dose–response curve of 8-epi-PGF2a enzyme immunoassay. (Panel B) Daily variation of urinary 8-epi-PGF2a excretion analysed by competitive enzyme immunoassay in 10 healthy subjects
IMMUNOASSAYS USING AChE AS LABEL
105
Figure 9.2 Scheme of the h-COX-1 and h-COX-2 peptides used for development of specific immunometric assays. SP: signal peptide
recognition of human COX-1. Polyclonal antibodies were also raised against a 12-residue peptide specific to the human COX-1 amino-terminus. Our strategy for COX-2 was to produce monoclonal antibodies directed against the specific 18-residue insert. Polyclonal antibodies were raised against a 12-residue peptide located immediately upstream of this insert, a sequence completely modified with respect to COX-1. These different antibodies were first used and characterized in Western blot experiments. Most were also used in competition assays. Subsequently, two-site immunometric assays specific to each of the two isoenzymes were developed and used to study the production of these two forms in a model of human umbilical vein endothelial cells (HUVEC). Results We obtained two immunosera against COX-1, one raised against the first 12 residues of the human enzyme (anti-N-COX-1) and the other against the total sequence of purified ovine COX-1 (antiCOX-1). We also developed six mAbs with the same protein (Cre´minon et al 1995a). We obtained polyclonal antibodies and 10 mAbs raised against the C1 peptide of COX-2 (580–596 sequence of COX-2, specific insert of this form). Polyclonal antibodies directed against the C2 peptide (570–581 sequence) were also prepared (Cre´minon et al 1995b). Western blot experiments, designed to characterize the usefulness and specificity of these antibodies (Habib et al 1993), showed that COX-2 can be induced in HUVEC, since this isoform had been previously observed only at the transcriptional and genomic level. In HUVEC stimulated by interleukin-1a (IL-1a), or by a phorbol ester (PMA), COX enzymatic activity increased in parallel with a protein doublet detected using anti-C1 antibodies, while COX-1 immunoreactivity remained practically unchanged. Induction kinetics experiments with IL-1a and PMA revealed
different patterns of COX-2 production: faster and shorter for PMA and longer for IL-1a. We carried out immunoprecipitations with cells stimulated by IL-1a, incubated with Met 35S. Using an anti-C1 antiserum, a 35S-labelled protein doublet appeared, with a molecular weight of around 70 kDa, while the anti-COX-1 polyclonal antibodies showed only a very slight difference between the reference and stimulated cells. COX-2 immunoreactivity detected in these experiments was completely suppressed by preincubation of antibodies with the C1 peptide used in their production. We also verified that COX-1 and COX-2 immunoprecipitates exhibited COX enzymatic activity inhibited by a NSAID, flurbiprofen. Treatment of COX-1 and -2 by endoglycosidase-H revealed the glycosyl character of these enzymes (molecular weight decrease) without suppressing the doublet observed for COX-2 immunoreactivity. In parallel with the Western blot experiments, we developed immunological assays for COX-1 and COX-2. Two of the six mAbs did not recognize human COX-1 in Western blotting. All the polyclonal and monoclonal antibodies used in a competitive type assay (using AChE-labelled ram COX1 as tracer) were insufficiently sensitive, thus ruling out any assays of biological samples. We then developed two-site immunometric assays using one antibody coated onto the solid phase to ensure capture of the molecule, with staining by another AChE-labelled antibody (in a Fab’ form). We first analysed all the possible binding complementarities between these different antibodies by using ovine COX-1 or human platelet extracts (which contain only COX-1). Various combinations with good sensitivities [minimum detectable concentration (MDC) 20–60 pg/ml with ovine COX-1] were eliminated due to limited recognition of human COX-1. Polyclonal anti-COX-1 and anti-N-COX-1 antibodies were shown to be good capture antibodies, used with AChE-labelled CX-110 mAb as tracer. As shown in Figure 9.3A, the assay combining coated polyclonal anti-COX-1 antibodies and CX-110-AChE as tracer gave an MDC of 113 pg/ml with ovine COX-1, with good accuracy and good recognition of the human enzyme (parallelism with serial dilutions of the sample).
106
ANALYTICAL METHODS
Figure 9.3 (Panel A) Standard curve obtained using polyclonal anti-COX-1 IgGs as capture antibody and AChE-labelled mAb CX-110 as enzyme tracer, ram COX-1 as standard (full line) and serial dilutions (1/10, 1/20, 1/40, arbitrarily positioned on the abscissa) of human platelets extract (dotted line). (Panel B) Standard curve obtained using affinity-purified polyclonal anti-C2 peptide antibodies as capture antibody and AChE-labelled (anti-C1 peptide) mAb CX-329 as enzyme tracer, C3 peptide as standard (full line) and serial dilutions (1/20, 1/40, 1/80, 1/160, arbitrarily positioned on the abscissa) of PMA-stimulated HUVEC extract (dotted line)
The assay with the anti-N-COX-1 polyclonal antibody as capture antibody was less sensitive (MDC&1 ng/ml) and was set aside. This sandwich assay indicated 100 ng COX-1/mg proteins (&6.666108 platelets) in platelet extracts, a value near that reported by other authors. All the anti-C1 antibodies detected COX-2 in Western blot experiments, but showed appreciable labelling intensity fluctuations. The best results were obtained with the CX-308, -329 and -367 antibodies. The anti-C1 polyclonal and monoclonal antibodies were tested in a competition assay (with the AChE-labelled C1 peptide as tracer). The results were very disappointing, with detection thresholds of a few pmol/ml (corresponding to a minimum concentration of 70 ng/ml for the complete protein). This assay did not detect COX-2 production after stimulation of HUVECs. On the other hand, the use of anti-C2 polyclonal antibodies (total IgG fraction) as capture antibodies provided good complementarity with three AChE-labelled anti-C1 mAbs (CX-320, -329 and -394 mAbs), both with the C3 peptide as standard (human COX-2 570–595 sequence) and PMA-stimulated HUVEC samples. Assay sensitivity was greater using affinity-purified anti-C2 antibodies (Figure 9.3B). This assay was used to measure (with respect to the C3 peptide) COX-2 production in HUVECs under different conditions. The assay does not recognize COX-1 in human platelet extracts. These two assays were validated by molecular sieve chromatography experiments. The two isoforms were eluted at very high molecular weights (&300 kDa), while electrophoretic analysis of these immunoreactive peaks gave only the expected value of 70 kDa. The high molecular weight peaks very probably resulted from non-covalent associations between COXs or between COXs and other (unidentified) entities, since after a simple SDS denaturing treatment the eluted immunoreactivity was found at the expected molecular weight. Comparison of the measured reactivities of stimulated and reference cells confirmed the very high production of COX-2, in contrast with the very small changes recorded for COX-1. A combination of these two assays yielded different data for the HUVECs (Figure 9.4). An excellent correlation was seen between the parallel increase in poststimulation enzyme activity (measured by 6-keto-PGF1a production; Figure 9.4C) and COX-2 immunoreactivity (detected by both sandwich assay and Western blot; Figure 9.4A), while COX1 immunoreactivity (measured by the same techniques; Figure
9.4B) varied only slightly. We verified a dose–response effect with PMA and the sensitivity of COX-2 production (PMA-induced) to a PKC and a transcription inhibitor. Recent application of this assay to recombinant human COX-2 gave an MDC close to 250 pg/ml (unpublished results). Prospects These two assays are sensitive and high-performance tools for routine measurements of numerous samples (in comparison with Western blot techniques). However, two major problems remain: (a) the detected immunoreactivity for these two proteins corresponds to non-covalent aggregates; (b) the difficulty of obtaining suitable standards, i.e. recombinant or purified proteins. The first problem appears to be difficult to solve, since the denaturing conditions used (SDS) are generally incompatible with the formation of the antigen–antibody complex (and also with use of AChE). However, the good correlation between the Western blot results, the COX immunoassays, and the assays of PG production suggests that this aggregation does not introduce marked bias. The second problem will probably soon be resolved since various groups have prepared purified or recombinant human proteins which will be useful for calibrating the assays. These COX assays are also of potential value in examining putative interactions (Habib et al 1997) between the COX system and the NO-synthases system (also essential at the cardiovascular level). In fact, the two enzyme systems appear to share certain features and be complementary in function, and there may be interdependency in cardiovascular regulation. This example demonstrates the wide-ranging potential application of these immunoanalytical methods. ENZYME IMMUNOMETRIC ASSAY OF LTC4 Background Since the theoretical bases of competition assays were laid down in the early 1960s, several different types of immunological assays have been developed. However, two major general principles have emerged: competitive and non-competitive assays. Competitive
IMMUNOASSAYS USING AChE AS LABEL
107
Figure 9.4 Time-course of immunoreactivity of lysates from PMA-stimulated HUVEC. Cells were incubated in the absence (open blocks) or presence (full blocks) of 20 nM PMA for the indicated times. (Panel A) Immunometric analysis of COX-2 in the lysates performed using affinity-purified polyclonal anti-C2 peptide antibodies as capture antibody and AChE-labelled (anti-C1 peptide) mAb CX-329 as enzyme tracer. The inset corresponds to Western blot analysis of COX-2 synthesized by the same lysates using anti-C1 peptide mAb CX-329. (Panel B) Immunometric analysis of COX-1 in the same lysates performed using polyclonal anti-COX-1 IgGs as capture antibody and AChE-labelled mAb CX-110 as enzyme tracer. The inset corresponds to Western blot analysis of the same membranes as in Panel A (after stripping) using anti-COX-1 mAb CX-111. (Panel C) COX activity was evaluated in the same cells before lysis
assays use a labelled molecule and a limited concentration of a specific corresponding antibody. Immunoassays of small molecules (haptens, including the eicosanoids) are founded on this principle (see 8-epi-PGF2a immunoassay, above). In noncompetitive assays, the most widespread method corresponds to two-site immunometric assays (also called sandwich assays), which use two antibodies that simultaneously bind the molecule assayed. One of these antibodies (immobilized on a solid phase) captures the molecule, while the other is labelled and serves as a tracer. Both antibodies are used in excess. It should be specified that these methods, which are extensively used to assay high molecular weight or antigen molecules (45–10 kDa), are much more sensitive (see the COX-1 and COX-2 immunoassays above). It is generally admitted, however, that they are not applicable to small haptens, which cannot simultaneously bind two antibody molecules. Nonetheless, by disregarding this assumption we first demonstrated that this assay format was functional with small peptides (Cre´minon et al 1995c). However, in spite of the successful development of two-site immunometric assays for small peptides, it is clear that this procedure cannot be applied to the lowest molecular weight haptens (51000 Da). In fact, the joint and simultaneous binding of two complementary antibodies becomes impossible, due to steric hindrance alone. A new approach must therefore be envisaged if this assay format (which is more sensitive due to use of reagents in excess) is to be applied to small haptens. This is why we developed another methodology (Pradelles et al 1994) based on SPIE-IA (solid-phase immobilized epitope immunoassay; SPI-CEA patent). Theoretically, this technique
allows an immunometric format assay of molecules of any size, using just a single antibody (preferably monoclonal). This new technique is particularly suitable for haptens bearing a free amino group, since this function is generally used for conjugation with the antigenic carrier during immunogen preparation. It is assumed that this group is not directly involved in recognition by the antibodies (as supported by the observation that chemical modification of this amino group does not alter antibody binding) and is thus available for application of our new methodology. General principle This assay includes four major consecutive steps which are shown schematically in Figure 9.5. The first step is the immunological capture of the hapten by antibodies immobilized on a solid phase. After washing, the second step corresponds to covalent binding of the hapten to the antibodies. This involves the reaction of a homobifunctional reagent reactive with amino groups, thereby crosslinking the hapten to the proteins on the solid phase. After further washing, the third step (epitope release) uses a denaturing reagent to disrupt the interactions between the antibody binding sites and the hapten (still covalently bound to the antibody), thus allowing recovery of epitope accessibility. Ideally, the capture antibody is inactivated in this step. Finally, the antibody tracer addition step is an incubation with an mAb–AChE conjugate, the tracer antibody used being identical to that employed in the immunological capture step. Addition of the enzyme substrate after washing
108
Figure 9.5
ANALYTICAL METHODS
General scheme of SPIE-IA
yields a typical immunometric calibration curve with a signal proportional to the amount of molecule introduced. Although the initial and final steps are common to standard immunometric assays, the originality of the procedure derives from the covalent binding and epitope release steps, which allow the use of a single antibody in a sandwich assay.
5 min reaction with 2.5% glutaraldehyde and a further reduction of remaining aldehyde active groups by treatment with 0.1 M NaBH4. Epitope release is achieved in 2 min with 0.1 M HCl before AChE-labelled anti-LTC4 mAb tracer is added for 1 h incubation. A final 30 min or 1 h revelation of AChE bound to the solid phase completes the assay.
Strategy
Results
This new procedure was tested with a monoclonal anti-LTC4 antibody (Volland et al 1994), which is not specific to LTC4 as it also cross-reacts with LTD4 and LTE4. The whole SPIE-IA takes less than 3.5 h. The capture step requires a 1 h reaction of standard or biological samples. The cross-linking step involves a
Different experimental conditions were studied for both the covalent binding and the epitope release steps, which constitute the originality of this method. After optimization, this assay gave a linear standard curve typical of immunometric assay (Figure 9.6A) and, as classically observed with immunometric assays
Figure 9.6 (Panel A) Standard curve obtained for SPIE-IA of LTC4. (Panel B) Time-course of LTC4 after addition of LTA4 to human platelets
IMMUNOASSAYS USING AChE AS LABEL involving AChE tracers, assay sensitivity and accuracy were improved when the enzymatic reaction was allowed to proceed for a longer period: MDC close to 5 pg/ml and 2 pg/ml for the 30 min and 1 h reaction, respectively. This assay is 60 times more sensitive than the competitive assay performed with the same mAb and LTC4–AChE tracer. SPIE-IA appears more specific, since the cross-reactivity with other LTs decreased (15% and 9% with LTD4 and LTE4, respectively) by comparison with the competitive assay. This probably results from steric hindrance due to the smaller size of LTE4 and LTD4, which are less accessible after covalent immobilization to the tracer antibody. The time-course of LTC4 production by platelets after addition of LTA4 was studied by SPIE-IA (Figure 9.6B) and correlated with competitive assay results. Validation was performed by HPLC fractionation experiments.
109
the Hoˆpital Lariboisie`re. He was an ardent advocate of immunoanalytical methods since using them in his doctoral work at the Institut Pasteur, and of the use of acetylcholinesterase as an enzymatic marker. Jacques Maclouf was both a respected scientist and an accomplished sportsman but, rather than participate in the usual scientific ‘‘competition’’, he preferred to establish a large network of contacts with both friends and colleagues, an approach which better suited his deep human qualities. The perennial smile of this marathon-running scientist and gentleman in everyday life will be missed by all his friends and above all by his family, and our thoughts go out to his wife Nicole and their children, Be´atrice, Antoine and Guillaume. Farewell, Jacques . . .
REFERENCES Prospects This procedure appears to be limited in the case of eicosanoids to leukotrienes, which alone have a reactive amino function. However, experiments using photoactivation, a technique of great potential, are being carried out with different study models so as to extend the procedure to other chemical groups. UV and laser irradiation experiments are now under way, using various molecules. CONCLUSIONS Eicosanoid research is constantly evolving, with the identification of new molecules (or even new biosynthetic pathways, as shown for COX-2) and the reappraisal of the role of previously identified molecules. These constant changes are mainly due to the increasingly complex image emerging from the investigations of these biological systems. The cardiovascular system presents all the characteristics of this complexity in which all the cells, both circulating and constituting the vascular wall, interact dynamically and cannot be dissociated from other fundamental phenomena, such as proliferation, programmed death or cell activation. In this context, antibody affinity and specificity are undeniably very useful, and immunolocalization (Western blot, immunocytochemistry and immunohistochemistry) is most informative. Moreover, these antibodies can be used to develop routine, analytically efficient immunoassays. Lastly, the use of AChE as labelling enzyme yields inexpensive and reliable tools which will improve our understanding of numerous eicosanoids and isoeicosanoids. IN MEMORIAM Jacques Maclouf passed away on 14 July 1998. As a pharmacist, and above all as a practised and judicious scientist, he quickly understood the full potential of eicosanoids in the cardiovascular field, and this represented the main research thrust of his team at
Cre´minon C, Frobert Y, Habib A et al (1995a) Immunological studies of human constitutive cyclooxygenase (COX-1) using enzyme immunometric assay. Biochim Biophys Acta, 1254, 333–340. Cre´minon C, Habib A, Maclouf J et al (1995b) Differential measurements of constitutive (COX-1) and inducible (COX-2) cyclooxygenase expression in human umbilical vein endothelial cells using specific immunometric enzyme immunoassays. Biochim Biophys Acta, 1254, 341–348. Cre´minon C, De´ry O, Frobert Y et al (1995c) Two-site immunometric assay for substance P with increased sensitivity and specificity. Anal Chem, 67, 1617–1622. Habib A, Cre´minon C, Frobert Y et al (1993) Demonstration of an inducible cyclooxygenase in human endothelial cells using antibodies raised against the C-terminal region of cyclooxygenase-2. J Biol Chem, 268, 23448–23454. Habib A, Bernard C, Lebret M et al (1997) Regulation of the expression of cyclooxygenase-2 by nitric oxide in rat peritoneal macrophages. J Immunol, 158, 3845–3851. Hayashi Y, Ueda N, Yokota K et al (1983) Enzyme immunoassay of thromboxane B2. Biochim Biophys Acta, 750, 322–329. Maclouf J, Pradel M, Pradelles P and Dray F (1976). 125I derivatives of prostaglandins. A novel approach in prostaglandin analysis by radioimmunoassay. Biochim Biophys Acta, 431, 139–146. Patrono C and FitzGerald GA (1997) Isoprostanes: potential markers of oxidant stress in atherothrombotic disease. Arterioscler Thromb Vasc Biol, 17, 2309–2315. Pradelles P, Grassi J and Maclouf J (1985) Enzyme immunoassays of eicosanoids using acetylcholine esterase as label: an alternative to radioimmunoassay. Anal Chem, 57, 1170–1173. Pradelles P, Grassi J and Maclouf J (1990) Enzyme immunoassays of eicosanoids using acetylcholinesterase. Methods Enzymol, 187, 24–34. Pradelles P, Grassi J, Cre´minon C et al (1994) Immunometric assay of low molecular weight haptens containing primary amino groups. Anal Chem, 66, 16–22. Smith WL, Garavito RM and DeWitt DL (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem, 271, 33157–33160. Volland H, Vulliez Le Normand B, Mamas S et al (1994) Enzyme immunometric assay for leukotriene C4. J Immunol Methods, 175, 97– 105. Wang Z, Ciabattoni G, Cre´minon C et al (1995) Immunological characterization of urinary 8-epi-prostaglandin F2a excretion in man. J Pharm Exp Therapeut, 275, 94–100.
10 Bioassay of Eicosanoids Robert L. Jones The Chinese University of Hong Kong, Shatin, NT, Hong Kong
Bioassay has played a fundamental role in the discovery of the eicosanoids and their development into a major research area. Goldblatt (1933) and von Euler (1934) first described the smooth muscle stimulant and vasodilator actions of acidic lipid extracts of various reproductive tissues. Many years later, Bergstro¨m’s group in Sweden isolated PGE1 and PGF1a from sheep seminal vesicles (Bergstro¨m and Sjo¨vall 1960a, 1960b; Bergstro¨m et al 1963), and the corresponding arachidonate-derived products, PGE2 and PGF2a, were soon recognized as local hormones of great biological relevance. Subsequently, the primary products of cyclooxygenase (COX) activity, PGG2 and PGH2, and their biologically active metabolites thromboxane A2 (TXA2) and prostacyclin (PGI2) were initially detected and quantified by bioassay methods. In a similar but slightly later scenario, cobra venom challenge to normal guinea-pig lung or antigen challenge to sensitized lung resulted in the release of lipid mediators with smooth muscle spasmogenic activity. The term ‘‘slow-reacting substance of anaphylaxis’’ (SRS-A) was coined for one of these principles, but its identity remained elusive for many years (Feldberg and Kellaway 1938; Kellaway and Trethewie 1940; Brocklehurst 1953; Orange and Austen 199). Eventually, the massive advances in chemical analysis techniques that occurred in the 1970s led to the identification of two peptidoleukotrienes, LTC4 and LTD4, as the main components of SRS-A, and to the identification of arachidonate 5-lipoxygenase as the initiating enzyme (Murphy et al 1979; Shimizu et al 1984, 1986; Ueda et al 1986). While LTC4 and LTD4 are powerful smooth muscle stimulants (Dahle´n et al 1981), particularly of the bronchial tree, an alternative 5-lipoxygenase pathway leads to LTB4, a nonpeptide mediator with potent chemotactic activity (see FordHutchinson, 1990). This chapter mainly deals with smooth muscle bioassays for eicosanoids, of which there are two main types. In a conventional assay, muscle tension is recorded from the assay tissue immersed in physiological salt solution in an organ bath. Standard and unknown agents are added to the bath in sequence, with wash and rest periods between each dose. This type of assay is best suited to the estimation of stable eicosanoids in either collected tissue fluid or solvent extracts/chromatography fractions reconstituted in water. In a superfusion assay, physiological salt solution drips continuously over the assay tissue and agents are added to the inflow stream over a few seconds (Gaddum 1953; Figure 10.1). In a cascade superfusion system, two or more assay tissues are arranged in a vertical bank and the effluent from the upper one drips over the next lower tissue. This arrangement may be attached to the outflow of a perfused organ (the generator in Figure 10.1), thereby affording real-time estimation of mediator release from the organ. Both stable and unstable mediators can be The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
detected by this method, which was pioneered by Sir John Vane over many years (Vane 2000). CONVENTIONAL BIOASSAYS FOR EICOSANOIDS Practical Considerations The purity of standards and unknowns is a critical factor in bioassay. The earliest PGE and PGF standards were derived from incubation of precursor fatty acids with crude prostaglandin synthase obtained from sheep seminal vesicles. However, the fatty acids were not always of the highest purity, and as a graduate student in Eric Horton’s laboratory in the late 1960s, I recall our concern over the 1–2% PGE1 present in our precious few milligrams of PGE2, in the light of work showing that PGE1 and PGE2 had opposite actions on blood platelet aggregation (Kloeze 1967). We now know, of course, that PGE1 is a much more potent agonist than PGE2 at the prostacyclin (IP) receptor (see Wise and Jones 2000). Fortunately, the Upjohn Company was soon supplying researchers with pure PGE2 and PGF2a prepared by chemical synthesis. In those early days, purification of tissue fluids or homogenates involved a solvent extraction process followed by a fairly crude chromatography process. In one commonly used extraction process, the retained phases (italics) sequentially contained low– medium polarity, neutral–acidic lipids (pH 3 water:ethyl acetate), then low–medium polarity, acidic lipids (ethyl acetate:pH 8 water; pH 3 water:ethyl acetate) and then medium polarity, acidic lipids (petroleum spirit:67% ethanol) (Jouvenaz et al 1970). The Karolinska group had developed reversed-phase partition chromatography systems with an adsorbed stationary phase (as opposed to the modern bonded stationary phases) for the purification of prostaglandins (see Gre´en et al 1978). Although these systems had great separating power, they were not suitable for purifying many samples containing only nanogram quantities of prostaglandins. Consequently, most groups used adsorption chromatography on small columns of silicic acid. The material supplied by commercial manufacturers was notoriously variable in its retention and solvent flow characteristics, and much time was spent removing fines by sieving and obtaining the desired absorptivity by careful heating. Nevertheless, good separation and recovery of PGE2 and PGF2a could usually be obtained (see Salmon and Karim 1976). The low quantities of prostaglandin present in most extracts did not allow the use of a 2+2 Latin square design (see Rang and Dale 1991). A bracketing assay was the preferred method, in which two or three doses of unknown were sandwiched between low and high doses of standard. This meant that confidence limits
112
ANALYTICAL METHODS
Figure 10.1 Bioassay using the cascade superfusion procedure. T1, T2 and T3 represent assay tissues. The mediator generator can be an immobilized enzyme system, an isolated organ, or an intact animal
for the estimate were not usually presented. I have memories of A. S. Milton standing behind a postgraduate struggling to obtain the optimum tissue load or dose cycle and saying ‘‘It takes 10 years to become a competent bioassayist’’. Indeed, the process of increasing the steepness of the lower portion of the dose–response curve was mastered by some, whilst remaining a complete mystery to others. A ‘‘good preparation’’ would often be taken over by several workers (including ‘‘The Prof’’), well into the early hours of the morning. The rat stomach fundus, which was developed to assay 5-HT (Vane 1957), was probably the most widely used preparation for prostaglandins. It responded well to both PGE2 and PGF2a and therefore one had to have confidence in the preceding chromatographic separation. Other useful tissues included the guinea-pig ileum and the colon of the gerbil or jird (Meriones trystiami and M. libycus), which were more sensitive to PGE2 than PGF2a. The rabbit jejunum, which shows rhythmic spontaneous activity, was sometimes used to assay PGF2a. Turning to the leukotrienes, the commonest assay preparation for SRS-A has been the guinea-pig ileum, prepared as either a tube or a longitudinal strip, in the presence of a histamine H1 receptor antagonist (Kellaway and Trethewie 1940; Brocklehurst 1953); its utility has extended into studies that followed the structural elucidation of the peptidoleukotrienes and the provision of authentic standards. LTD4 is about 10 times more potent than LTC4 on guinea-pig ileum (Piper et al 1981). SRS-A was distinguished from bradykinin, PGE2 and PGF2a by its greater activity on guinea-pig ileum relative to rat uterus and gerbil colon (Orange and Austen 1969). In later studies, SRS-A in tissue extracts was assayed against an in-house laboratory standard (Conroy et al 1976). The Contribution of Conventional Bioassays to Discoveries in the Eicosanoid Field For about 10 years from 1965, conventional bioassay played a major part in establishing several physiological and pathological
roles for PGE2 and PGF2a. Starting with the pro-inflammatory actions of PGE2, Ambache and colleagues (1965) demonstrated the presence of unsaturated hydroxy fatty acids in the aqueous humour of the rabbit eye following irritation of the iris. They found that the gerbil colon had little spontaneous activity and, because of this, was a better assay preparation than the rat colon. Later workers reported increased production of PGE1, PGE2 and PGF2a in anterior uveitis in the rabbit using the rat stomach fundus as a bioassay preparation (Eakins et al 1972). Discrimination between PGE and PGF-like activity was partly based on the instability of PGE under mild alkaline conditions, while PGE1 and PGE2 were separated by thin-layer chromatography on silverimpregnated silica or paper. A considerable number of studies were performed on eicosanoid function in skin inflammation, e.g. A¨ngga˚rd et al. (1970) and Arturson et al (1973) used rat stomach fundus, gerbil colon and guinea-pig ileum to show the presence of PGE2-like material in scald and burn fluids. The group led by Malcolm Greaves was in the forefront of studies on human skin, using thin-layer chromatography (TLC) to separate E and F prostaglandins, followed by assay on the rat uterus preparation. His group showed that flucinolone, but not hydrocortisone, dramatically reduced the biosynthesis of PGE2 and PGF2a from arachidonic acid in human breast skin (Greaves and McDonald-Gibson 1972). They also incorporated tritium-labelled arachidonic acid into the incubation medium so that appropriate zones on the TLC plate could be accurately defined, and the extent of biosynthesis could also be confirmed. It is of interest that during my collaboration with Amersham Radiochemicals to produce heptatritio-PGE2 and heptatritio-PGF2a for radioimmunoassay purposes, the mass component for calculation of specific activity was determined by assay on the guinea-pig ileum and rabbit jejunum, respectively. It had to be assumed that the introduction of seven tritium atoms into a prostaglandin molecule does not alter its biological potency. One of the major effects of NSAIDs, such as aspirin, indomethacin and meclofenamic acid, is to reduce fever. In 1970, Milton and Wendlandt reported that PGE1 injected into the
BIOASSAY third ventricle of the conscious cat produced shivering and a rapid rise in body temperature. They noted that the fever was similar in character to that produced by bacterial pyrogen, except for its shorter duration. In collaboration with Feldberg’s group, cerebrospinal fluid collected from the third ventricle of cats during Shigella dysenteriae fever was assayed on the rat fundus strip preparation and found to contain elevated levels of PGE1-like activity (Feldberg et al 1973). Significantly, indomethacin, aspirin and paracetamol reduced the fever and also reduced the prostaglandin concentration to below the level of detection; this led to general acceptance of the role of prostaglandins in the generation of bacterial fever and of the inhibition of prostaglandin biosynthesis as the antipyretic mechanism of NSAIDs. A fuller description of the early work on the role of prostaglandins in fever has been written by Milton (1976). Turning to PGF2a, the paper presented by Eric Horton’s group at the New York Academy of Sciences meeting in 1970 (Horton et al 1971) illustrated the value of bioassay for detecting small quantities of a mediator of unknown structure, followed by identification of the mediator in a bulked sample by a sophisticated physicochemical method. Internal distension of the isolated uterine horn of the guinea-pig resulted in release of PGF2a-like material into the bathing medium, but there was little change in the release of PGE2, as assayed on the rat stomach strip. Extracts from distended and non-distended horns from 35 guineapigs on day 3 of the oestrous cycle were then subjected to combined gas chromatography–mass spectrometry after methyl ester–trimethysilyl ether derivatization. Ten times more PGF2a was found in the distension sample compared to the nondistension sample, and this supported the emerging hypothesis that PGF2a is a natural luteolytic agent in several animal species (see Horton and Poyser 1976). CASCADE SUPERFUSION BIOASSAYS FOR EICOSANOIDS Practical Considerations Figure 10.1 shows the basic superfusion unit and several ways of using the cascade and accessory equipment to measure mediator release and metabolism. The essence of the cascade is that the assay tissues are chosen to detect specific mediators or combinations of mediators (as a consequence of their different complements of cell surface receptors). For example, the rabbit aorta contracts to nanogram quantities of PGH2 and TXA2, while the rat stomach strip contracts to similar amounts of PGE2. Receptor antagonists can be applied to one or more of the assay tissues to improve the specificity of the detection process. For example, in many prostanoid studies, an antagonist cocktail of atropine (M receptors), mepyramine (H1 receptors), phenoxybenzamine (a1-adrenoceptors and others), propranolol (badrenoceptors) and methysergide (some 5-HT receptors) has been employed. A delay coil can be inserted into the cascade to determine the stability of the mediator under the conditions of the assay (say, pH 7.4 and 378C). Also the effluent from the generator can be split and run over two banks of assay tissues, one of which may be treated with antagonist(s). In a small number of studies, a blood-bathed system has been used. The generator is an anaesthetized animal or a blood-perfused isolated organ, such as the spleen. Blood is taken from the carotid artery of the animal or the venous outflow of the isolated organ to the top of the cascade, and instead of running to waste at the bottom it is returned to a reservoir and then to the animal or organ. It is essential to have good anticoagulation with heparin, otherwise the assay tissues become coated with a layer of fibrin and their sensitivity to mediators becomes depressed. Release of ADP and
113
5-HT from aggregating platelets can also produce variation in the basal tone of the assay tissues. The whole cascade system must be thermostatically controlled to obtain reproducible responses to standards and released agents. It should be realized that the exposure profiles of assay tissues to standard agents and mediators may not be similar (as they are in conventional assays) and this may affect the accuracy of an assay. In addition, release of a mixture of mediators from the generator may lead to physiological antagonism in the assay tissues, with underestimation of the level of one or both of the mediators (see next section, final paragraph). Finally, while a highly unstable agonist may lose activity during passage down the cascade, the reverse may also occur. For example, contractile responses to LTC4 increased in size on passage over a bank of guinea-pig ileum preparations, whereas those to LTD4 remained virtually unchanged (Piper et al 1982); the explanation lay in the conversion of LTC4 to the more potent LTD4 by g-glutamyl transpeptidase present in the assay tissues. Despite these caveats, the cascade superfusion method has worked remarkably well for the bioassay of eicosanoids. The Contribution of Cascade Superfusion Bioassays to Discoveries in the Eicosanoid Field The cascade superfusion method was first used in eicosanoid research to show the release of prostaglandins from the Krebsperfused spleen of the dog during stimulation of the sympathetic nerve supply (Gilmore et al 1968). Assay bank 1 contained rat stomach strip, rat colon and chick rectum in the presence of phenoxybenzamine and propranolol; strong contractions of the first two assay tissues indicated the release of both PGE2 and PGF2a; assay bank 2 contained the same preparations continuously contracted with a combination of PGE2, angiotensin IIamide and 5-HT and was used to demonstrate catecholamine release through relaxant responses of the tissues. Contraction of the spleen induced by adrenaline is also accompanied by the release of prostaglandins (conventional bioassay) (Davies et al 1967) and in the cascade experiments the infusion of adrenaline into the chronically-denervated spleen still caused contraction of the assay tissues. Thus, it was concluded that the prostaglandins derive from activated smooth muscle cells in the spleen (or other cells with an adrenergic innervation) and not from nerves supplying the spleen. Probably the greatest utility of the cascade superfusion method in the eicosanoid field has been to detect and quantify unstable prostanoids such as PGH2, TXA2 and PGI2 (prostacyclin). Our story starts with attempts to determine the nature of ‘‘rabbit aorta-contracting substance’’ (RCS), which could be generated from lung tissue by several different stimuli and appeared to derive from enzyme action on membrane phospholipids. Gryglewski and Vane (1972) allowed arachidonic acid to interact with a crude COX preparation from dog spleen for several minutes in a silicone rubber coil (generator), and directed the effluent over rabbit aorta and rat stomach strip preparations. Arachidonate infusion alone caused weak contraction of the rabbit aorta, which was dramatically enhanced when COX was also added (Figure 10.2). PGE2 generation could also be detected using the rat stomach strip, but was complicated by a contractile response to COX alone. Addition of meclofenamic acid or indomethacin to the generator reduced the contractions of both preparations, indicating COX as the source of RCS. Further experiments showed that the contractile activity of RCS on rabbit aorta decayed considerably following passage through a delay coil, whereas activity on rat stomach strip increased. The conclusion that RCS is simply PGG2/PGH2, which decays spontaneously to give mainly PGE2, did not, however, fit all the facts; the half-lives
114
ANALYTICAL METHODS
Figure 10.2 Record from a cascade superfusion assay of rabbit aortacontacting substance (RCS) generated by the interaction of arachidonic acid with a crude cyclo-oxygenase (COX) enzyme preparation from dog spleen during passage through a silicone rubber coil. Modified from Gryglewski and Vane (1972) with permission of the author and publisher
of RCS (52 min) and the prostaglandin endoperoxides (ca. 5 min) were different. This led Samuelsson’s group to search for an even more unstable COX product, which they identified as thromboxane A2 (TXA2) (Hamberg et al 1975). TXA2 is several times more potent than PGH2 on the rabbit aorta (Needleman et al 1976). Rabbit coeliac or mesenteric arteries show biphasic responses, contractile progressing to relaxant, to the prostaglandin endoperoxides, while they only contract to TXA2. This profile allows the discrimination of PGH2 and TXA2, and was used to show that TXA2 is the major component of RCS released from guinea-pig lung incubated with arachidonic acid (Moncada and Vane 1977). Jones et al (1982) used the same guinea-pig lung–rabbit aorta system to show that EP 045, one of the first selective TP receptor antagonists to be synthesized, could block the contractile actions of TXA2 and U-46619 (a synthetic TP receptor agonist) to the same extent, while having no effect against angiotensin II. RCS activities generated in rabbit platelets during aggregation (Vargaftig 1977) and rabbit polymorphonuclear leukocytes during phagocytosis (Higgs et al 1976) were also shown to be due to TXA2. In following up the dual action of prostaglandin endoperoxides on rabbit coeliac and mesenteric arteries, Bunting et al (1976) detected a new natural prostanoid generated by a microsomal enzyme preparation (generator) from vascular tissue. This product, which was initially called PGX, did not contract rabbit aorta (which does not contain relaxant prostanoid receptors), had weak contractile activity on rat stomach strip and relaxed rabbit coeliac and mesenteric arteries. PGX was subsequently characterized and renamed prostacyclin, and is a potent vasodilator and inhibitor of platelet aggregation through agonist actions on IPreceptors (see Wise and Jones, 2000 for a fuller account of the discovery of prostacyclin). In the original studies, PGX did not affect the superfused rat colon, but in later experiments, purified material was reported to relax and reduce the spontaneous activity of the colon (Moncada et al 1978). The inhibitory action of prostacyclin on the colon is due to activation of IP receptors on non-adrenergic non-cholinergic (NANC) nerves, causing the release of NO and a second unidentified inhibitory transmitter (Qian et al 1995). In some preparations, spontaneous activity decays completely over several hours, and this may account for the variable action of PGX. Prostacyclin is one of three endothelium-derived relaxant factors (EDRFs) released by agents acting on the intimal surface of vascular endothelium cells, the others being nitric oxide (NO) and endothelium-derived hyperpolarizing factor (EDHF) (see
Mombouli and Vanhoutte 1997). Endothelium-denuded vessel rings have been used in a cascade superfusion system to differentiate and assay NO and prostacyclin released from endothelium cells cultured on microcarrier beads packed into a column (generator). The underlying principle involves choosing two assay tissues, both of which relax well to NO, while only one relaxes to prostacyclin. Cocks et al (1985) used the pig and dog coronary arteries for this purpose, while Gryglewski et al (1986) used the rabbit aorta and the cow coronary artery. An example from the latter study is shown in Figure 10.3, where bradykinininduced release of both NO and prostacyclin from pig aorta endothelial cells, and the much shorter half-life of NO relative to prostacyclin, can be seen by comparing the responses of the first and second tissues of each type in the cascade. Tone in the preparations was generated by inclusion of U-46619 in the superfusate. The name ‘‘slow-reacting substance of anaphylaxis’’ (SRS-A) was derived from the very slow relaxation of the guinea-pig ileum to baseline tension following washout of the mediator from the organ bath. Thus, it was natural to attempt the bioassay of SRS-A by cascade superfusion due to the efficient washing of the assay tissues in this technique. In addition to the guinea-pig ileum, various respiratory smooth muscle preparations, such as tracheal and bronchial rings and strips of lung parenchyma, have been used on the basis of their high contractile sensitivity to the peptidoleukotrienes. The simultaneous release of RCS and peptidoleukotrienes from guinea-pig lung was measured using an initial bank of rat stomach strip/guinea-pig trachea/rabbit aorta, followed by a delay circuit, and then by a second bank of the same three preparations (Bakhle et al 1985). RCS activity decayed during the delay, whereas LT-like activity did not. The Ca2+ ionophore A23187 induced much greater release of LT-like material than either bradykinin or arachidonic acid, prompting the authors to suggest that different cell types in the lung have different complements of biosynthetic enzymes. Surprisingly, the bradykinin- and arachidonic acid-released mediators that contracted the guinea-pig trachea appeared to be COX products, but
Figure 10.3 Record from a cascade superfusion assay of both prostacyclin and NO released from pig aorta endothelial cells after challenge with bradykinin (BK). Equivalent sensitivity of the assay tissues to NO is shown by the relaxant responses to sodium nitroprusside (NP). The ring preparations were denuded of endothelium and therefore did not relax to direct application of bradykinin. Modified from Gryglewski et al (1986) with permission of the author and publisher
BIOASSAY
115
REFERENCES
Figure 10.4 Assay of LTB4 extracted and purified from 500 mg of human psoriatic skin scale using chemokinesis of human peripheral blood polymorphonuclear leukocytes. Modified from Brain et al (1984) with permission of the author and publisher
their actions were blocked by the LT receptor antagonist FPL 55712. FPL 55712 blocked authentic LTC4, but not PGE2, on guinea-pig ileum, and blocked both agonists on guinea-pig trachea, leading the authors to question the specificity of FPL 55712. However, the guinea-pig trachea contains both EP1 receptors responsible for contraction and EP2 receptors that mediate relaxation (Dong et al 1986). Consequently, just a small degree of EP1 receptor block by FPL 55712 may allow the EP2 system to become dominant, leading to a false impression of antagonist specificity.
ASSAYS FOR CHEMOTACTIC ACTIVITY Stimulation of rat neutrophils with A23187 caused the release of a lipoxygenase product with chemoattractant and aggregating activity on neutrophils (Bray et al 1980). Isolation of the product by HPLC was followed by its identification as leukotriene B4 (Ford-Hutchinson et al 1980). The agarose micro-droplet chemokinesis system employed for assay of LTB4 involves mixing neutrophils with agarose solution, placing droplets of the suspension in a 96-well micro-titre plate and allowing solidification of the agarose. Aliquots of solutions containing potential chemoattractant are then added to the wells and, after waiting for several hours, the movement of the white cells is measured microscopically (Smith and Walker 1980). Figure 10.4 shows log concentration–response curves for the LTB4 standard and LTB4-like material isolated from a bulked sample of human psoriatic skin scale (Brain et al 1984); the high sensitivity and good linearity and parallelism of this assay are readily apparent.
CONCLUDING REMARKS Classical bioassay methods are rarely used in laboratories today. Nevertheless, we should not abandon them altogether, since they are capable of detecting new biologically active natural agents, which other methods cannot.
Ambache N, Kavanagh L and Whiting J (1965) Effect of mechanical stimulation on rabbits’ eyes: release of active substances in anterior chamber perfusates. J Physiol, 176, 378–408. Arturson G, Hamberg M and Jonsson C-E (1973) Prostaglandins in human burn blister fluid. Acta Physiol Scand, 87, 270–276. A¨ngga˚rd E, Arturson G and Jonsson C-E (1970) Efflux of prostaglandins in lymph from scalded tissues. Acta Physiol Scand, 80, 46A–47A. Bakhle YS, Moncada S, de Nucci G and Salmon JA (1985) Differential release of eicosanoids by bradykinin, arachidonic acid and calcium ionophore AH 23187 in guinea-pig isolated perfused lung. Br J Pharmacol, 86, 55–62. Bergstro¨m S and Sjo¨vall J (1960a) The isolation of prostaglandin E from sheep prostate glands (a). Acta Chem Scand, 14, 1701–1705. Bergstro¨m S and Sjo¨vall J (1960b) The isolation of prostaglandin F from sheep prostate glands (b). Acta Chem Scand, 14, 1693–1700. Bergstro¨m S, Ryhage R, Samuelsson B and Sjo¨vall J (1963) The structures of prostaglandins E1, F1a and F1b. J Biol Chem, 238, 3555–3564. Brain SD, Camp RDR, Cunningham FM et al (1984) Leukotriene B4-like material in scale of psoriatic skin lesions. Br J Pharmacol, 83, 313–317. Bray, NA, Ford-Hutchinson AW, Shipley ME and Smith MJH (1980) Calcium ionophore A23187 induces release of chemotactic and aggregating factors from polymorphonuclear leucocytes. Br J Pharmacol, 71, 507–512. Brocklehurst WE (1953) Occurrence of an unidentified substance during anaphylactic shock in cavy lung. J Physiol, 120, 16P–17P. Bunting S, Gryglewski R and Moncada S (1976) Arterial walls generate from prostaglandin endoperoxides a substance (prostaglandin X) which relaxes strips of mesenteric and coeliac arteries and inhibits platelet aggregation. Prostaglandins, 12, 897–913. Cocks TM, Angus JA, Campbell JH and Campbell GR (1985) Release and properties of endothelium-derived relaxing factor (EDRF) from endothelial cells in culture. J Cell Physiol, 123, 310–320. Conroy MC, Organe RP and Lichtenstein LM (1976) Release of slow reacting substance of anaphylaxis (SRS-A) from human leukocytes by the calcium ionophore A23187. J Immunol, 116, 1677–1681. Dahle´n SE, Hedqvist P and Samuelsson B (1981) Leukotrienes are potent constrictors of human bronchi. Nature, 288, 484–486. Davies BN, Horton EW and Withrington PG (1967) The occurrence of prostaglandin E2 in splenic venous blood of the dog following splenic nerve stimulation. J Physiol, 188, 38P–39P. Dong YJ, Jones RL and Wilson NH (1986) Prostaglandin E receptor subtypes in smooth muscle: agonist activities of stable prostacyclin analogues. Br J Pharmacol, 87, 97–107. Eakins KE, Whitelocke RA, Perkins ES et al (1972) Release of prostaglandins in ocular inflammation in the rabbit. Nature New Biol, 239, 248–249. Feldberg W and Kellaway CH (1938) Liberation of histamine and formation of lysolethicin-like substances by cobra venom. J Physiol, 94, 187–226. Feldberg W, Gupta KP, Milton AS and Wendlandt S (1973) Effect of pyrogen and antipyretics on prostaglandin activity in cisternal CSF of unanesthetised cats. J Physiol, 234, 279–303. Ford-Hutchinson AW, Bray NA, Doig MV et al (1980) Leukotriene B: a potent chemokinetic and aggregating substance released from polymorphonuclear leucocytes. Nature, 286, 264–265. Ford-Hutchinson AW (1990) Leukotriene B4 in inflammation. Crit Rev Immunol, 10, 1–12. Gaddum JH (1953) The technique of superfusion. Br J Pharmacol Chemother, 8, 321–326. Gilmore N, Vane JR and Wyllie JH (1968) Prostaglandins released by the spleen. Nature, 218, 1135–1140. Goldblatt MW (1933) A depressor substance in seminal fluid. J Soc Chem Indust, 52, 1056–1057. Greaves MW and McDonald-Gibson W (1972) Prostaglandin biosynthesis in human skin and its inhibition by corticosteroids. Br J Pharmacol, 46, 172. Gre´en K, Hamberg M and Samuelsson B (1978) Extraction and chromatographic procedures for purification of prostaglandins, thromboxanes, prostacyclin and their metabolites. In Advances in Prostaglandin and Thromboxane Research, 5th edn, Fro¨lich JC (ed). Raven, New York, 15–18.
116
ANALYTICAL METHODS
Gryglewski R and Vane JR (1972) The generation from arachidonic acid of rabbit aorta contracting substance (RSC) by a microsomal; enzyme preparation which also generates prostaglandins. Br J Pharmacol, 46, 449–457. Gryglewski R, Moncada S and Palmer RMJ (1986) Bioassay of prostacyclin and endothelium-derived relaxing factor (EDRF) from porcine aortic endothelial cells. Br J Pharmacol, 87, 685–694. Hamberg M, Svensson J and Samuelsson B (1975) Thromboxanes: a new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Nat Acad Sci USA, 72, 2994–2998. Higgs GA, Bunting S, Moncada S and Vane JR (1976) Polymorphonuclear leucocytes produce thromboxane A2-like activity during phagocytosis. Prostaglandins, 12, 749–757. Horton EW, Thompson C, Jones R and Poyser NL (1971) Release of prostaglandins. Ann NY Acad Sci, 180, 351–361. Horton EW and Poyser NL (1976) Uterine luteolytic hormone: a physiological role for PGF2a. Physiol Rev, 56, 595–651. Jones RL, Peesapati V and Wilson NH (1982) Antagonism of the thromboxane-sensitive contractile systems of the rabbit aorta, dog saphenous vein and guinea-pig trachea. Br J Pharmacol, 76, 423–438. Jouvenaz GH, Nugteren DH, Beerthuis RK and Van Dorp DA (1970) A sensitive method for the determination of prostaglandins by gas chromatography with electron capture detection. Biochimica Biophysica Acta, 202, 231–234. Kellaway CH and Trethewie ER (1940) The liberation of slow-reacting smooth muscle-stimulating substance in anaphylaxis. Qu J Exp Physiol, 30, 121–145. Kloeze J (1967) Influence of prostaglandins on platelet adhesiveness and platelet aggregation. In Nobel Symposium II. Prostaglandins, Bergstro¨m S and Samuelsson B (eds). Interscience, New York, 241–252. Milton AS and Wendlandt S (1970) A possible role for prostaglandin E1 as a modulator for temperature regulation in the central nervous system. J Physiol, 207, 76–77. Milton AS (1976) Modern views on the pathogenesis of fever and the mode of action of pyretic drugs. Journal of Pharmacy and Pharmacology, 28, 393–399. Mombouli JV and Vanhoutte PM (1997) Endothelium-derived hyperpolarizing factor(s): updating the unknown. Trends Pharmacol Sci, 18, 252–256. Moncada S and Vane JR (1977) The discovery of prostacyclin—a fresh insight into arachidonic acid metabolism. In Biochemical Aspects of Prostaglandins and Thromboxanes, Kharasch N and Fried J (eds). Academic Press, New York, USA, 155–177. Moncada S, Ferreira SH and Vane JR (1978) Bioassay of prostaglandins and biologically active substances from arachidonic acid. Adv Prostagland Thromboxane Res, 5, 211–236. Murphy RC, Hammarstro¨m S and Samuelsson B (1979) Leukotriene C: a slow reacting substance from murine mastocytoma cells. Proc Nat Acad Sci USA, 76, 4275–4279.
Needleman P, Moncada S, Bunting S et al (1976) Identification of an enzyme in platelet microsomes which generates thromboxane A2 from prostaglandin endoperoxides. Nature, 261, 558–560. Orange RP and Austen KF (1969) Slow-reacting substances of anaphylaxis. Adv Immunol, 10, 105–144. Piper PJ, Letts LG, Samhoun MN et al (1982) Actions of leukotrienes on vascular, airway and gastrointestinal; smooth muscle. Adv Prostagland Thromboxane Leukotriene Res, 9, 169–181. Piper PJ, Samhoun MN, Tippins JR et al (1981) Pharmacological studies on pure SRS-A, and synthetic leukotrienes C4 and D4. In SRS-A and Leukotrienes, Piper PJ, (ed). Wiley, New York, 81–99. Qian YM and Jones RL (1995) Inhibition of rat colon contractility by prostacyclin (IP-) receptor agonists: involvement of NANC neurotransmission. Br J Pharmacol, 115, 163–171. Rang HP and Dale MM (1991) Measurement in pharmacology. In Pharmacology. Churchill Livingstone, Edinburgh, 49–71. Salmon JA and Karim SMM (1976) Methods for analysis of prostaglandins. In Prostaglandins: Chemical and Biochemical Aspects, Karim SMM (ed). MTP Press, Lancaster, 25–85. Shimizu T, Izumi T, Seyama Y et al (1986) Characterisation of leukotriene A4 synthase from murine mast cells: evidence for its identity as arachidonate 5-lipoxygenase. Proc Nat Acad Sci USA, 83, 4175–4179. Shimizu T, Radma˚rk O and Samuelsson B (1984) Enzyme with dual lipoxygenase activities catalyses leukotriene A4 synthesis from arachidonic acid. Proc Nat Acad Sci USA, 81, 689–693. Smith MJH and Walker JR (1980) The effect of some antirheumatic drugs on an in vitro model of human polymorphonuclear leucocyte chemokinesis. Br J Pharmacol, 69, 473–478. Ueda N, Kaneko S, Yoshimoto T and Yamamoto S (1986) Purification of arachidonate 5-lipoxygenase from porcine leukocytes and its reactivity with hydroperoxyeicosatetraenoic acids. J Biol Chem, 261, 7982–7988. Vane JR (1957) A sensitive method for the assay of 5-hydroxytryptamine. Br J Pharmacol Chemother, 12, 344–349. Vane JR (2000) The Second Gaddum Memorial Lecture: the release and fate of vaso-active hormones in the circulation. Br J Pharmacol, 131, 27–60. Vargaftig BB (1977) Carrageenan and thrombin trigger prostaglandin synthetase-independent aggregation of rabbit platelets: inhibition by phospholipase A2 inhibitors. J Pharm Pharmacol, 29, 222–228. Von Euler US (1934) Zur kenntnis der pharmakalogischen wirkungen von nativsekreten und extrekten mannlicher accessorsicher geschlechsdrusen, Naunyn Schmiedeberg’s Arch Exp Path Pharmacol, 175, 78–84. Wise H and Jones RL (2000) An introduction to prostacyclin and its receptors. Prostacyclin and Its Receptors. Kluwer Academic, New York, 1–27.
11 Gas Chromatography and Mass Spectrometry in Eicosanoid Analysis Michinao Mizugaki1, Takanori Hishinuma2, Naoto Suzuki2 and Junichi Goto2 1Tohoku
Pharmaceutical University and 2Tohoku University Hospital, Sendai, Japan
Since the discovery of the structure of prostaglandins, thromboxanes, leukotrienes, other eicosanoids and their metabolites, one of the most challenging tasks in eicosanoids research has been to define the role of eicosanoids in human health and disease. Since the measurement of eicosanoids and their specific metabolites is a useful approach to assess in vivo formation of these compounds, numerous analytical techniques have been developed for their analysis in biological samples. These include radioimmunoassay (RIA) and enzyme immunoassay (EIA), gas chromatography– mass spectrometry (GC–MS) using different ionization techniques, liquid chromatography (LC) in combination with ultraviolet (UV), fluorescence and electrochemical detection and, more recently, LC combined with MS utilizing different ionization techniques. In addition to these methods, bioassays have also been developed. The most widely utilized methods now for the detection and quantitation of eicosanoids in biological samples are sensitive RIA, EIA and GC–MS. More recently, LC–MS has become utilized by many investigators. Moreover, as biological fluids such as human plasma and urine consist of extremely low concentrations of arachidonic acid metabolites and large amounts of contaminants, extensive sample extraction and purification procedures are required prior to the analysis. Therefore, significant advances in methodologies for extraction and purification, i.e. solid-phase extraction (SPE), high-performance liquid chromatography (HPLC), thin-layer chromatography (TLC), immunoaffinity purification and combinations of these have been used. In this section, the recent developments of analytical methods, extraction and purification procedures for measurement of eicosanoids and their metabolites in biological fluids are reviewed. SAMPLE EXTRACTION AND PURIFICATION The extraction and purification of eicosanoids from biological samples are necessary for their analysis by any analytical methods, including RIA, EIA, GC–MS and LC–MS, because of the low concentration of analytes and the interference by abundant contaminants. Extraction and purification of eicosanoids and their metabolites in biological samples prior to their analysis are usually complicated and have involved methods that utilize several chromatographic extraction/purification steps, including SPE, TLC and HPLC, often being time-consuming and resulting in a considerable loss of sample yield. From this point of view, various extraction and purification methodologies that have advantages in terms of the rapidity and high recovery have been The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
developed, including a combination of efficient and applicable SPE and automated technologies. Furthermore, immunoaffinity purification, which is based on the use of an immobilized specific antibody against the analyte and has the ability to concentrate samples, has also been reported for the purification of prostaglandins, thromboxanes, leukotrienes and their metabolites. Solid-phase Extraction (SPE) The first step of analytical methods in the analysis of eicosanoids is their extraction from a biological sample. Although two extraction techniques have been reported, i.e. solvent extraction and SPE, solvent extraction of eicosanoids has been largely substituted by SPE. Up to now, SPE has been developed to be the indispensable extraction procedure in immunoassays, GC–MS and LC–MS and other analytical techniques for isolation of metabolites of arachidonic acid from biological samples because of its rapidity and applicability. In general, eicosanoids in biological samples are extracted by reverse-phase cartridges consisting of octadecylsilica (ODS). Other reverse-phase extraction cartridges, consisting of phenylboronic acid (PBA) (Lawson et al 1985; Lorenz et al 1989) and normal-phase aminopropyl (NH2) cartridges (Kikawa et al 1990; Bessard et al 2001), have been shown to be highly selective for some eicosanoids. Extraction of 11-dehydro-TXB2 (Lorenz et al 1989) and 2,3-dinor-TXB2 (Lawson et al 1985) using PBA cartridges is specific because it is based on the principle that boronates have been known to form very stable complexes with b-diols in carbohydrates and eicosanoids. Extraction of 8-iso-PGF2a, however, has been unsuccessful, despite its 1,3-diol structure (Tsikas et al 1998). The application of the normal-phase NH2 cartridge for purification of LTE4 (Kikawa et al 1990) and 15-iso-PGF2a (Bessard et al 2001) in human urine after extraction by ODS cartridge has been reported. Furthermore, the Bond-Elut Certify IITM cartridge, which contains mixed-phase material with lipophilic anion-exchange parts, has also been used for a simple one-step SPE of 11-dehydro-TXB2 in urine prior to analysis by EIA (Perneby et al 1999). Extraction with these materials yields highly purified extracts and considerably simplifies and shortens analysis of some eicosanoids. More recently, the EmporeTM extraction disk (Wu et al 1996) and the membrane-filter type EmporeTM extraction cartridge were used for extraction of LTE4 (Mita et al 1997; Mizugaki 1999) and 8-iso-PGF2a (Ohashi and Yoshikawa 2000; Murai et al 2000) from human urine. This methodology has several advantages over the conventional extraction cartridge: improved recovery, a smaller amount of solvent and
118
ANALYTICAL METHODS
less contamination of particles from cartridges into the eluent. Using these SPE cartridges, moreover, an automated SPE system has been recently developed. A commercial automated SPE system has been evaluated for cyclooxygenase metabolites of arachidonic acid, PGE2, PGF2a, 6-keto-PG1a and TXB2 in urine samples (Hotter et al 1992). TLC and HPLC Despite the significant developments in SPE described above, further chromatographic purification methods are required for the analysis of eicosanoids and their metabolites in biological samples, because purification by SPE only is generally insufficient, e.g. an additional TLC purification has been required for 2,3dinor-TXB2 and TXB2 in human urine, despite selective extraction using PBA cartridges (Lawson et al 1985). For the analysis of urinary 8-iso-PGF2a, the introduction of HPLC has been necessary to ensure adequate purity of the sample to be injected into the GC–MS and acceptable accuracy (Ferretti and Flanagan 1997). Both reverse- and normal-phase HPLC and TLC have often been used, therefore, for the further purification of eicosanoids in eluent from the SPE cartridge. Although the TLC technique is often time-consuming, complicated (i.e. scraped off the plate) and leads to poor recovery, simple and inexpensive TLC is a widely employed technique in the analysis of eicosanoids and other lipids (Myher and Kuksis 1995). Numerous applications have appeared for purification and separation of native and derivatized eicosanoids (Ferretti and Flanagan 1997; Morrow and Roberts 1994; Tsikas et al 2000; Schwedhelm et al 2000). HPLC is also a widely utilized purification method for the analysis of eicosanoids (Tsikas et al 2000; Schwedhelm et al 2000; Qiu et al 1996; Kumlin 1996; O’Sullivan et al 1999; Walter et al 2000). HPLC methodologies provide an effective clean-up and good yield over the TLC technique. One further advantage of HPLC is that it is easily automated by the use of an autoinjector, autosampler and fraction collector. A fully automated HPLC method which employs a column-switching system was developed for extraction and isolation of various lipoxygenase metabolites of arachidonic acid from protein-containing biological fluids (Haas and Buchanan 1988). Immunoaffinity Purification The immunoaffinity technique, which is based on the principle that the analyte is extracted by an immobilized specific antibody, is a powerful tool for rapid and specific purification of small amounts of compounds in complex biological fluids. This methodology was utilized for purification of eicosanoids and their metabolites in plasma and urine, including 2,3-dinor-TXB2 (Tagari et al 1994), 2,3-dinor-6-keto-PGF1a (Hiramatsu et al 1994), PGE2 (Matsumoto et al 1993), 8-iso-PGF1a (Bachi et al 1996), 11-dehydro-TXB2 (Hayashi et al 1990; Djurup et al 1993), LTC4 (Matsumoto et al 1993) and LTE4 (Westcott et al 1998), e.g. Westcott et al. (1998) developed two immunopurification methods for purification of urinary LTE4. One is an easy and reasonable immunoaffinity chromatography utilizing commercially available immunoaffinity resin, and the other is a new immunofiltration purification (Westcott et al 1997). In immunoaffinity chromatography, a specific antibody against an analyte is attached to a solid support, which allows the separation of what is bound from what is not bound. These columns can be prepared at a reasonable cost but are often washed and reused, with the potential for sample contamination. In immunoaffinity purification, on the other hand, excess antibody is added to a sample to bind specific compounds. Bound compounds are separated from low molecular
weight unbound components by passage through a molecular weight cut-off filter. This unique technique has both the advantage that individual filters are utilized for each sample, so there is no possibility of sample contamination, and the disadvantage that a relatively large amount of antibody is required. Another unique immunoaffinity purification method has utilized a specific monoclonal antibody against cis-3-hexen-1-ol, which was used as a model compound of the o3-olefin unit of D17-6-keto-PGF1a (Hishinuma et al 1992). Very low concentrations of D17-6-ketoPGF1a (6 pg/ml) in human sera can be concentrated by simple one-step immunoaffinity purification prior to analysis by GC–MS. ANALYTICAL METHODS Recently, many groups dealing with the analysis of eicosanoids in biological fluids have developed more sophisticated methods. The recent trend has consisted of the development of more novel simple methods and improvements in the already existing methods in terms of simplicity, routine analysis, versatility and effectiveness. These include ‘‘high-throughput’’, ‘‘automated’’, ‘‘simultaneous’’, ‘‘rapid’’, ‘‘streamlined’’, ‘‘improved’’ analytical methods, which make it possible, for example, to determine many eicosanoids in a single biological sample within a single analytical procedure. HPLC Since eicosanoids have no specific absorption with minor exceptions (HETEs and LTs etc.), UV detection appears to be unsuitable. The HPLC technique is insufficient to detect and quantitate eicosanoids and their metabolites, especially in plasma and urine. Nevertheless, some studies have reported the analysis of eicosanoids using HPLC with various detectors, such as UV (Chavis et al 1999), fluorescence (Beil et al 1998; Maier et al 2000) and chemiluminescence detectors (Chiba et al 1999) after derivatization using labelling reagents, photodiode array (Eberhard et al 2000) and light-scattering detectors (Schepky et al 1997). Eberhard et al reported the detection and quantitative analysis of four arachidonic acid metabolites, PGE2, LTB4, 12HETE and 15-HETE, in biopsies of human periodontal tissues by HPLC combined with a photodiode array detector, with the detection limit of their method being adequate to detect the metabolites of arachidonic acid (Chiba et al 1999). In the combination of a light-scattering and photodiode array detector, Schepky et al (1997) simultaneously determined the levels of lipids, prostaglandins, HETEs and other metabolites of the inflammation cascade in vitro. The analytical methods of PGE1 in plasma were developed by the improvement of conventional analysis of eicosanoids and other fatty acids using 9-anthryldiazomethane (ADAM) as a fluorescence-labelling reagent (Tsutsumiuchi et al 1997). As a result of the improvement of the elimination procedure of excess reagents, including SPE and HPLC, the femtomolar level of derivatized PGE1 can be determined by laser-induced fluorometric detection. 20-HETE and other cytochrome P-450 metabolites of arachidonic acid in urine, tissue and interstitial fluid have been measured after labelling with 2-(2,3-naphthalimino)ethyl trifluoromethanesulphonate (Maier et al 2000). These sensitive methods are expected to be useful for simultaneous determination of other eicosanoids and their metabolites. Immunoassay In contrast to the mass spectrometric analysis described later, the immunoassay methodologies, especially EIA, are simple and
GAS CHROMATOGRAPHY AND MASS SPECTROMETRY inexpensive to determine the levels of eicosanoids and so lead to the analysis of large numbers of samples. RIA is sensitive but there are several limitations in laboratory use because of the use of radiolabelled compounds (Battistini et al 1998; John et al 1998). General EIA methodologies have been established and EIA kits for many eicosanoids are now commercially available (although not for all eicosanoids), which have been utilized for routine analysis in numerous studies (Qiu et al 1996; Kumlin 1996; Battistini et al 1998). These techniques, however, may suffer from a lack of selectivity and sensitivity according to some cases, particularly when complicated biological samples containing structurally related compounds and interfering substances such as plasma and urine are analysed. Therefore, the use of RIA and EIA methods for the analysis of eicosanoids in plasma and urine often requires their validation by GC–MS methods (Bessard et al 2001; Perneby et al 1999; O’Sullivan et al 1999; Djurup et al 1993; Proudfoot et al 1999). Recently, Bessard et al (2001) compared 15iso-PGF2a levels in human urine estimated using commercially available EIA kits with those estimated by GC–MS following the same extraction procedure, in order to determine the agreement between the two methods, resulting in higher values determined by EIA for all samples. Similarly, values of 9a,11b-PGF2 determined by EIA were in the same range but consistently higher than those obtained by GC–MS (O’Sullivan et al 1999). Although analysis of GC–MS or GC–MS–MS after extensive purification and derivatization procedures provides the most reliable results, detection by EIA may be reliable if sample extraction and purification is appropriate. Apart from a slight overestimation by EIA at the highest levels of 11-dehydro-TXB2, comparison of 11-dehydro-TXB2 levels in urine determined by EIA after selective SPE with those determined by GC–MS has shown good agreement between the two methods (Perneby et al 1999). These results appear to suggest that the appropriate extraction and purification procedure must necessarily be used to obtain accurate values of eicosanoids in biological samples by EIA. Of the several few new approaches for developing more sensitive immunometric methodologies compared with conventional EIA, solid-phase-immobilized epitope immunoassay (SPIEIATM) was developed by Grassi et al (1996). The SPIE-IATM procedure, based on the use of a single monoclonal antibody for capture and an enzyme-linked tracer antibody, involves covalent cross-linking of the hapten to the solid phase (i.e. the microtitre plate). This methodology has been applied to the analysis of LTC4, which contains a highly reactive primary amino group, allowing efficient covalent cross-linking (Volland et al 1994). This assay has been at least 60-fold more sensitive than conventional competitive EIA using similar reagents and conditions, including the monoclonal anti-LTC4 antibody, LTC4-acetylcholine esterase and incubation conditions. In addition, the procedure is very simple and rapid. This SPIE-IATM has been applied to the analysis of other small compounds containing primary amino groups, including endothelin (Lawson et al 1985), thyroxine (Etienne et al 1995), substance P (Lawson et al 1985) and angiotensin II (Grassi et al 1996). With each hapten, SPIE-IATM was also 70–300-fold more sensitive than conventional competitive EIA. Although it is difficult to achieve efficient immobilization with small haptens that have no primary amino groups, this unique method appears to have great potential for the development of more sensitive methods for eicosanoid analysis. GC–MS and GC–MS–MS At the beginning of eicosanoid research, GC–MS in the electron impact (EI) ionization mode (GC–EI–MS), which leads to high fragmentation, was the analytical method of choice for the
119
quantitative determination of index metabolites in the plasma and urine of humans. This ionization is less sensitive than negative-ion chemical ionization (NICI) because of the high fragmentation, but is often reliable for the identification of structurally unknown metabolites and the separation of isomers which shared the same formula. For example, Bessard et al (2001) recently described the possibility of the simultaneous analysis of four regioisomers of F2isoprostanes in human urine using GC–EI–MS. On the other hand, the very sensitive GC–MS in the NICI mode (GC–NICI– MS), in which an intense carboxylate anion but less fragmentation is generally observed in eicosanoid analysis, is currently the most powerful technique for analysis of numerous eicosanoids in biological fluids. More recently, tandem mass spectrometry (MS– MS) instruments have been developed for coupling with GC (Tsikas 1998). The MS–MS instrument in the selected reaction monitoring (SRM) or multiple reaction monitoring (MRM) mode, in which the selected product (or daughter) ions produced by collision-induced dissociation (CID) of precursor (or parent) ions are monitored, results in an increase of the specificity of eicosanoid analysis in complex biological fluids. In addition, SRM may require less laborious and less time-consuming extraction/ purification steps compared with GC–MS in the selected ion monitoring mode because of its high sensitivity and selectivity. Due to these developments of instruments and the extraction/ purification procedures described above, specific and accurate measurement of eicosanoid formation in vivo in humans is currently possible using only a few millilitres of urine, plasma or other biological fluids by GC–MS and in particular by GC–MS– MS. Therefore, numerous applications of GC–MS and GC–MS– MS have appeared for the analysis of eicosanoids in vitro and in vivo, including prostaglandin metabolites (Tsikas et al 1998, 2000; Schweer et al 1994; Callaghan et al 1994; Wu¨bert et al 1997; Hammes et al 1999; Obata et al 1999; Thevenon et al 2001), thromboxane metabolites (Schweer et al 1994; Callaghan et al 1994; Wu¨bert et al 1997; Obata et al 1999; Ishibashi et al 1994), leukotriene metabolites (Mita et al 1997; Tsikas et al 1993; Takamoto et al 1995) and others (Fulton et al 1998; Watzer et al 2000; Powell et al 2001). In particular, measurements of the recently discovered F2-isoprostanes by GC–MS and GC–MS–MS, which may provide a means of assessing oxidative stress status in vivo, have been reported in many studies (Tsikas et al 1998; Schwedhelm et al 2000; Wu¨bert et al 1997; Schweer et al 1997; Lawson et al 1998; Morrow et al 1999; Morales et al 2001). Morrow and Roberts (1994) quantified F2-isoprostanes in isolated plasma by GC–NICI–MS, with extraction and purification by ODS and silica SPE cartridges followed by TLC twice. Ferretti and Flanagan (1997) also developed a similar method using GC– NICI–MS, with the replacement of one TLC step with HPLC. An improved method for the measurement of urinary and plasma F2-isoprostanes, using a combination of ODS and silica SPE cartridges, HPLC and GC–NICI–MS, has been reported (Mori et al 1999). Furthermore, Walter et al (2000) reported a method for the analysis of F2-isoprostane in plasma and urine samples by solvent extraction, HPLC with the amino column, followed by GC–NICI–MS, which provides a sensitive procedure requiring less sample volume. Recently, a more sensitive and selective GC–MS–MS method has also been utilized, with extraction and purification by a combination of the ODS SPE cartridge, TLC or HPLC (Tsikas et al 1998; Schwedhelm et al 2000; Schweer et al 1997). At present, methods based on GC–MS and GC–MS–MS are indispensable analytical tools to reliably assess eicosanoid formation in vivo. GC–MS and GC–MS–MS are very sensitive and selective; however, this instrumentation may not be readily available to most investigators and an extensive sample purification and derivatization procedure is required prior to analysis, making it impractical for routine analysis when large numbers of
120
ANALYTICAL METHODS
samples are to be measured. In particular, eicosanoids have to be derivatized using several derivatizing agents in order to increase thermal stability and volatility in GC, and to improve sensitivity in MS. Since eicosanoids have several functionalities in their molecules, including carboxyl, oxo and hydroxyl groups, esterification, methoxylation and etherification, respectively, are required, according to their structures (Tsikas 1998). In addition, not only the appropriate selection of the derivatizing agents but also efficient removal of excess agents after derivatization are necessary for accurate and sensitive analysis by GC–MS. LC–MS and LC–MS–MS Due to the recent significant developments in MS instruments coupled with HPLC utilizing various ionization techniques, i.e. particle beam (PB), thermospray (TSP) and fast-atomic bombardment (FAB) ionization and the more recently developed atmospheric pressure chemical ionization (APCI) and electrospray or ionspray ionization (ESI or ISP) interfaces, LC–MS and especially LC–MS–MS have rapidly been becoming indispensable methods for analysis of drugs (Brewer and Henion 1998) and other compounds, including eicosanoids. In early studies, LC– TSP–MS was utilized for the direct analysis of major human seminal prostaglandins (Abia´n and Gelpı´ 1987). Yamane and Abe (1991) also succeeded in detecting 20 kinds of derivatized prostaglandins and thromboxanes by LC–TSP–MS. LC–PB–MS was similarly utilized for the determination of derivatized hydroxyeicosatetraenoic acids, 5-HETE, 12-HETE and 15HETE (Galimberti et al 1992). More advanced LC–MS and LC–MS–MS coupled with the APCI and ESI systems (which have several advantages over other ionization systems, such as higher sensitivity and fewer limitations in analytical conditions) have been applied for the analysis of small amounts of eicosanoids and their metabolites in complicated biological fluids in numerous studies. In contrast with GC–MS and GC–MS–MS in particular, these methodologies could be easily applicable to direct analysis of non-volatile or thermally labile compounds, such as leukotrienes (Wu et al 1996; Mizugaki et al 1999; Kishi et al 2001) and other eicosanoids (Griffiths et al 1996; Newby and Mallet 1997; Hall and Murphy 1998; Mallat et al 1999), with simple extraction/ purification steps and without any derivatizations. For example, some studies have reported similar simple and selective methods of direct analysis of native LTE4 in human urine by simple extraction followed by LC–ISP–MS–MS (Wu et al 1996) or LC– ESI–MS–MS (Mizugaki et al 1999; Kishi et al 2001), although in GC–MS analysis cysteinyl leukotrienes must be catalytically hydrogenized and desulphurized to obtain volatile derivatives, as described previously (Tsikas et al 1993). LTC4, LTD4, LTE4 and the recently discovered 5-oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG7) were analysed by LC–ESI–MS–MS (Hevko and Murphy 2001). 5-Lipoxygenase products of arachidonic acid produced by human neutrophils, LTB4, 5-HETE and o-oxidation metabolites of LTB4 have been quantitated by LC–ESI–MS–MS (Wheelan and Murphy 1997). Epoxyeicosatrienoic acids (EETs) and HETEs in human red blood cells have also been analysed by LC–ESI–MS–MS (Nakamura et al 1997). These mythologies have been utilized for the analysis of F2-isoprostanes in biological fluids (Ohashi and Yoshikawa 2000; Murai et al 2000; Li et al 1999) as well as the GC–MS and GC–MS–MS described above. Moreover, a unique application of ESI–MS–MS (not LC–ESI–MS–MS) for the analysis of eicosanoids was reported by Margalit et al (1996). This method required not only no derivatization procedures but also no chromatographic separations by HPLC, with only simple extraction using the SPE cartridge. Extracted samples were injected directly into a mass spectrometer and quantitative analysis of eicosanoids was based on the correlation between
the mass spectrometer response in the SRM mode and their concentration. This new methodology was capable of evaluating a wide variety of eicosanoid production in two inflammation models, lipopolysaccaride (LPS)-stimulated human blood and carrageenan-challenged rat air pouch, with high-throughput and high selectivity. Although LC–MS and especially LC–MS–MS have several advantages, including the ability to analyse many analytes simultaneously in the same sample, as well as GC–MS and GC– MS–MS and no requirement for special derivatization steps or tedious extraction/purification procedures, there are some restrictions with regard to the mobile phases, additives and flow rates typically used in conventional HPLC. In addition, compared with MS coupled with GC, LC–MS and LC–MS–MS are often inferior at present in terms of the sensitivity, resolution and the ability of chromatographic separation. Further developments are expected in the field of chromatography, such as nanoflow-HPLC and capillary electrophoresis (CE) coupled with MS (Petersson et al 1999), a novel ionization interface and mass spectrometer, such as an ion trap mass spectrometer (Liminga and Oliw 2000), as well as effective extraction and purification procedures.
CONCLUSIONS As described above, recent trends in the field of eicosanoid analysis have consisted of the development of sensitive, simple and high-throughput methods. The development of ‘‘userfriendly’’ methods may contribute to the further understanding of the role of eicosanoids in normal cellular responses, in immune reactions and especially in pathophysiological events such as inflammation and cardiovascular diseases.
REFERENCES Abia´n J and Gelpı´ E (1987) J Chromatogr, 394, 147–153. Bachi A, Zuccato E, Baraldi M et al (1996) Free Radic Biol Med, 20, 619– 624. Battistini B, Picard S, Borgeat P and Sirois P (1998) Methods Mol Biol, 105, 187–199. Battistini B, Picard S, Borgeat P and Sirois P (1998) Methods Mol Biol, 105, 201–207. Beil D, Kinder H, Paschke A et al (1998) J Chromatogr Sci, 36, 284–286. Bessard J, Cracowski JL, Stanke-Labesque F and Bessard G (2001) J Chromatogr B Biomed Sci Appl, 754, 333–343. Brewer E and Henion J (1998) J Pharm Sci, 87, 395–402. Callaghan DH, Yergey JA, Rousseau P and Masson P (1994) Pulm Pharmacol, 7, 35–41. Chavis C, Fraissinet L, Chanez P et al (1999) Anal Biochem, 271, 105–108. Chiba N, Imai H and Nakagawa Y (1999) J Chromatogr B Biomed Sci Appl, 728, 35–40. Djurup R, Chiabrando C, Jorres A et al (1993) Clin Chem, 39, 2470–2477. Eberhard J, Jepsen S, Albers HK and Ac¸il Y (2000) Anal Biochem, 280, 258–263. Etienne E, Cre´minon C, Lamourette P et al (1995) Anal Biochem, 225, 34– 38. Ferretti A and Flanagan VP (1997) J Chromatogr B Biomed Sci Appl, 694, 271–276. Fulton D, Falck JR, McGiff JC et al (1998) J Lipid Res, 39, 1713–1721. Galimberti R, Lecchi P, De Angelis L et al (1992) Anal Biochem, 201, 356– 361. Grassi J, Cre´minon C, Frobert Y et al (1996) Clin Chem, 42, 1532–1536. Griffiths WJ, Yang Y, Sjovall J and Lindgren JA (1996) Rapid Commun Mass Spectrom, 10, 183–196. Haas TA and Buchanan MR (1988) J Chromatogr, 430, 1–9. Hall LM and Murphy RC (1998) J Am Soc Mass Spectrom, 9, 527–532. Hammes W, Bu¨chsler U, Kinder P and Bo¨kens H (1999) J Chromatogr A, 847, 187–202.
GAS CHROMATOGRAPHY AND MASS SPECTROMETRY Hayashi Y, Shono F, Yamamoto S, et al (1990) Anal Biochem, 187, 151– 159. Hevko JM and Murphy RC (2001) J Am Soc Mass Spectrom, 12, 763–771. Hiramatsu M, Hayashi Y, Yamamoto S et al (1994) Prostagland Leukotrenes Essent Fatty Acids, 50, 69–79. Hishinuma T, Shimomura K, Nishikawa M et al (1992) Prostaglandins, 44, 329–338. Hotter G, Gomez G, Ramis I et al (1992) J Chromatogr, 607, 239–243. Ishibashi M, Nakagawa Y, Harima N et al (1994) Biol Mass Spectrom, 23, 612–620. John H, Cammann K and Schlegel W (1998) Prostagland Lipid Mediat, 56, 53–76. Kikawa Y, Nakai A, Shigematsu Y and Sudo M (1990) J Chromatogr, 532, 387–393. Kishi N, Mano N and Asakawa N (2001) Anal Sci, 17, 709–713. Kumlin M (1996) J Chromatogr A, 725, 29–40. Lawson JA, Brash AR, Doran J and FitzGerald GA (1985) Anal Biochem, 150, 463–470. Lawson JA, Li H, Rokach J et al (1998) J Biol Chem, 273, 29295–29301. Li H, Lawson JA, Reilly M et al (1999) Proc Natl Acad Sci USA, 96, 13381–13386. Liminga M and Oliw E (2000) Lipids, 35, 225–232. Lorenz R, Helmer P, Uedelhoven W et al (1989) Prostaglandins, 38, 157– 170. Maier KG, Henderson L, Narayanan J et al (2000) Am J Physiol Heart Circ Physiol, 279, H863–871. Mallat Z, Nakamura T, Ohan J et al (1999) J Clin Invest, 103, 421–427. Margalit A, Duffin KL and Isakson PC (1996) Anal Biochem, 235, 73–81. Matsumoto S, Hayashi Y, Kinoshita I et al (1993) Ann Clin Biochem, 30, 60–68. Mita H, Oosaki R, Mizushima Y et al (1997) J Chromatogr B Biomed Sci Appl, 692, 461–466. Mizugaki M, Hishinuma T and Suzuki N (1999) J Chromatogr B Biomed Sci Appl, 729, 279–285. Morales CR, Terry ES, Zackert WE et al (2001) Clin Chim Acta, 314, 93– 99. Mori TA, Croft KD, Puddey IB and Beilin LJ (2001) Anal Biochem, 268, 117–125. Morrow JD and Roberts LJ II (1994) Methods Enzymol, 233, 163–174. Morrow JD, Zackert WE, Yang JP et al (1999) Anal Biochem, 269, 326– 331. Murai Y, Hishinuma T, Suzuki N et al (2000) Prostagland Lipid Mediat, 62, 173–181. Myher JJ and Kuksis A (1995) J Chromatogr B Biomed Appl, 671, 3–33. Nakamura T, Bratton DL and Murphy RC (1997) J Mass Spectrom, 32, 888–896. Newby CS and Mallet AI (1997) Rapid Commun Mass Spectrom, 11, 1723– 1727. Obata T, Nagakura T, Maeda H et al (1999) J Chromatogr B Biomed Sci Appl, 731, 73–81.
121
Ohashi N and Yoshikawa M (2000) J Chromatogr B Biomed Sci Appl, 746, 17–24. O’Sullivan S, Mueller MJ, Dahle´n SE and Kumlin M (1999) Prostagland Lipid Mediat, 57, 149–165. Perneby C, Granstro¨m E, Beck O et al (1999) Thromb Res, 96, 427–436. Petersson MA, Hulthe G and Fogelqvist E (1999) J Chromatogr A, 854, 141–154. Powell WS, Boismenu D, Khanapure SP and Rokach J (2001) Anal Biochem, 295, 262–266. Proudfoot J, Barden A, Mori TA et al (1999) Anal Biochem, 272, 209–215. Qiu DW, Hui KP, Lee CW et al (1996) J Chromatogr B Biomed Appl, 677, 152–155. Schepky AG, Siegner R and Diembeck W (1997) Inflamm Res, 46, 417– 419. Schwedhelm E, Tsikas D, Durand T et al (2000) J Chromatogr B Biomed Sci Appl, 744, 99–112. Schweer H, Watzer B and Seyberth HW (1994) J Chromatogr, 652, 221– 227. Schweer H, Watzer B, Seyberth HW and Nusing RM (1997) J Mass Spectrom, 32, 1362–1370. Tagari P, Callaghan DH, Black C and Yerge JA (1994) Prostaglandins, 47, 293–306. Takamoto M, Yano T, Shintani T and Hiraku S (1995) J Pharm Biomed Anal, 13, 1465–1472. Thevenon C, Guichardant M and Lagarde M (2001) Clin Chem, 47, 768– 770. Tsikas D (1998) J Chromatogr B Biomed Sci Appl, 717, 201–245. Tsikas D, Fauler J, Gutzki FM et al (1993) J Chromatogr, 622, 1–7. Tsikas D, Schwedhelm E, Fauler J et al (1998) J Chromatogr B Biomed Sci Appl, 716, 7–17. Tsikas D, Schwedhelm E, Gutzki FM and Fro¨lich JC (1998) Anal Biochem, 261, 230–232. Tsikas D, Gutzki FM, Bo¨hme M et al (2000) J Chromatogr A, 885, 351– 359. Tsutsumiuchi R, Saito H, Imagawa T et al (1997) Anal Chem, 69, 5006– 5010. Volland H, Vulliez Le Normand B, Mamas S, et al (1994) J Immunol Methods, 175, 97–105. Walter MF, Blumberg JB, Dolnikowski GG and Handelman GJ (2000) Anal Biochem, 280, 73–79. Watzer B, Reinalter S, Seyberth HW and Schweer H (2000) Prostagland Leukotrienes Essent Fatty Acids, 62, 175–181. Westcott JY, Sloan S and Wenzel SE (1997) Anal Biochem, 248, 202–210. Westcott JY, Maxey KM, MacDonald J and Wenzel SE (1998) Prostagland Lipid Mediat, 55, 301–321. Wheelan P and Murphy RC (1997) Anal Biochem, 244, 110–115. Wu Y, Li LY, Henion JD and Krol GJ (1996) J Mass Spectrom, 31, 987– 993. Wu¨bert J, Reder E, Kaser A et al (1997) Anal Chem, 69, 2143–2146. Yamane M and Abe A (1991) J Chromatogr, 568, 11–24.
12 Time-resolved Fluoroimmunoassay in Eicosanoid Analysis Werner Schlegel1 and Harald John2 1Universita¨tsklinikum
Mu¨nster, Mu¨nster; and 2IPF PharmaCeuticals GmbH, Hannover, Germany
Prostaglandins are important chemical messengers capable of regulating cellular behaviour in mammalian tissues. Arachidonic acid is the precursor of the prostaglandin biosynthesis system that is localized in the endoplasmatic reticulum. From an endoperoxide pool the cells form all prostaglandins, via special enzymes or spontaneously. However, the capacity of the biosynthesis of the prostaglandins may differ from one cell type to the other. Prostaglandins are considered as paracrine/autocrine factors being metabolized next to the place where they exert their activity (Schlegel and Greep 1976). Therefore, it is wise to measure the primary prostaglandins together with their metabolites, when the total output of prostaglandins has to be calculated. Radioimmunoassays (RIA) have been used successfully for routine measurements of prostaglandins; they are sensitive, practicable and reasonably reliable (Jaffe and Behrman 1974; Schlegel et al 1977; Granstro¨m et al 1982; Maclouf et al 1976; John et al 1998). In recent years, non-radioactive labelling methods have been developed. The last step of a non-isotopic immunoassay normally includes an enzymatic reaction resulting in a coloured (Hayashi and Yamamoto 1982; Pradelles et al 1985; Meyer et al 1989) or chemiluminescent (Weerasekera et al 1983) end product. These procedures are preferable, since radioactive material can be avoided. The rare earth metal europium is a useful non-radioactive marker which is becoming a popular choice in immunoassay methods (Soini and Hemmila¨ 1979), since it can be measured with high sensitivity by time-resolved fluorimetry. This methodology detects the specific signal of the europium metal only after nonspecific background fluorescence has rapidly declined to a low level, and this effect markedly increases the sensitivity of the technique. The time-resolved fluorimetry was designed primarily for use with fluorescent chelates of europium. In our first attempts we introduced prostaglandin F2a–polylysine europium chelate conjugates as labelled antigen (Lu¨ke and Schlegel 1990). However, this conjugation method failed when using other prostaglandins for this labelling procedure. Therefore, we decided to change our labelling methodology to enable a highly sensitive detection for all measured prostaglandins (Lu¨ke and Schlegel 1992; Schlegel et al 1997). The present chapter describes several sensitive time-resolved fluoroimmunoassays (TR–FIAs) for the determination of prostaglandins, based on the use of biotinylated prostaglandins and europium-labelled streptavidin. The application of these assays for the measurement of prostaglandin concentrations in tissue is also reported. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
REAGENTS Anti-rabbit IgG (R2004), biocytin (Ne-biotinyl-L-lysine), streptavidin, diethylene-triamino-penta-acetic acid (DTPA) anhydride, 13,14-dihydro-15-keto prostaglandin F2a (PGFM) and 13,14dihydro-15-keto prostaglandin E2 (PGEM) were purchased from Sigma (Deisenhofen, Germany). All other prostaglandins were obtained from Paesel (Frankfurt am Main, Germany). EuCl3 was purchased from Fluka (Buchs, Switzerland). The enhancement solution (100 mM acetate buffer adjusted to pH 3.2 with potassium hydrogen phthalate, containing 15 mM 2-naphthoyltrifluoroacetone, 50 mM tri-n-octylphosphine oxide, and 0.1% Triton X-100) was kindly provided by Wallac Oy (Turku, Finland). All other chemicals were obtained from common suppliers (in general from Sigma). Microtitre strips were purchased from Eflab Helsinki, Finland. Because of the chemical instability of the metabolite PGEM under assay conditions, it was quantitatively converted to the more stable 11-deoxy-13,14-dihydro-15-keto-11b,16x-cycloprostaglandin E2 (PGEMcyc) as described by Granstro¨m et al (1982). This was achieved by adding 0.5 ml methanol and 0.2 ml 1 M NaOH to a 0.2 ml aliquot of sample solution. After 2 h at room temperature the solution was acidified to pH 3.5 and extracted with 5 ml ethyl acetate. ANTISERA The polyclonal antisera against PGE2, PGF2a, PGEMcyc, PGFM, thromboxane B2, 6-keto-PGF1a and 12-S-hydroxyheptadecatrienoic acid were raised in New Zealand White rabbits by immunization with thyroglobulin- or bovine serum albuminconjugated antigens, as previously described in detail (Schlegel et al 1982, John et al 1998). PREPARATION OF THE LABELLED LIGANDS Labelling of Streptavidin with Europium Streptavidin (1 ml) was dissolved in 1 ml 50 mM carbonate buffer (pH 8.0). DTPA anhydride was added in 200-fold molar excess, and the reaction allowed to proceed overnight at 48C; 150 mg EuCl3, dissolved in 500 ml H2O, were added to the reaction mixture, and the solution was stirred for 1 h. Eu3+-labelled streptavidin was purified by dialysis to equilibrium against 50 mM
124
ANALYTICAL METHODS
Tris–HCl buffer (pH 7.8). The average number of Eu3+ ions incorporated in each streptavidin molecule was about 13, determined by fluorescence measurement against a Eu3+ standard solution. The conjugate was stored in aliquots at 7208C. Synthesis and Purification of Biocytinyl Ligands Prostaglandin (3 mmol) plus 4 mmol N-hydroxysuccinimide were dissolved in 250 ml dioxane prior to the addition of 4 mmol N,N ’dicyclohexylcarbodiimide dissolved in 250 ml dioxane. The reaction mixture was stirred for 4 h at room temperature. The mixture was combined with 8 mmol biocytin, dissolved in 300 ml water. After stirring overnight, the crude product was purified by thinlayer chromatography (TLC), using precoated TLC plates with Silica gel 60 (Merck, Darmstadt, Germany). The solvent system was methanol:acetic acid (98:2 v/v). Fifteen zones were excised, eluted with methanol and evaporated. The presence of biocytinyllabelled and immunoreactive ligands in the fraction was checked using a TR–FIA binding test. The diluted fractions were incubated in the ligand-corresponding antibody-coated microtitre wells. After washing, the wells were incubated with the streptavidin–europium conjugate and then, after a second washing, the fluorescence was measured (see TR–FIA procedure). Further purification of the biocytin–ligand conjugate was achieved by HPLC on a reverse-phase silica column (Ultropac TSK ODS-120T, 5 mm, 25064.6 mm i.d.; LKB, Bromma, Sweden). Elution was performed with a linear gradient: methanol:water (acidified with phosphoric acid to pH 4.5) from 40:60 to 90:10 within 25 min; flow rate 1 ml/min; 1 ml fractions; 408C. The presence of biocytin-labelled ligands in the HPLC fractions was checked as described above. Synthesis of the Tracer The two compounds of the tracer (30 ng of the biocytin-labelled ligand and 1.2 mg of the europium-labelled streptavidin) were mixed in 10 ml assay buffer to produce the required ligand– biocytin–streptavidin–europium conjugate. The specific and nonspecific binding (BN) were tested in a TR–FIA test, as mentioned below, with various amounts of ligand–biocytin and streptavidin– europium conjugates. The optimal signal:noise ratio was 55:1 [maximum of specific binding (B0)=110.000 cps; BN=2000 cps]. It is recommended to prepare the tracer immediately before the assay is run. COATING OF THE MICROTITRE WELLS Anti-rabbit IgG was immobilized by adsorption to the well walls of polystyrene microtitre strips. The wells were coated overnight with 200 ml 5 mg/l anti-rabbit IgG solution in 0.05 M NaHCO3, pH 9.6. After coating, the wells were washed with distilled water. Saturation of the well walls was achieved by adding 300 ml saturation buffer [0.05 M Tris–HCl, 0.15 M NaCl, 5 g/l bovine serum albumin (BSA); pH 7.4]. The wells were stored in plastic bags at 7208C until use. TR–FIA PROCEDURE Figure 12.1 shows the principle of the solid phase competitive immunoassay. Standard solution or extracted biological samples (50 ml of each) were added to the coated microtitre wells, then 50 ml tracer solution were added. Finally, 50 ml antibody solution diluted in assay buffer (1/20 000–1/100 000, depending on the
ligand) were added to each well, and the mixture was incubated with continuous shaking for 2 h at room temperature. The immunoreaction was stopped by washing the wells four times with buffer containing 0.05 M Tris–HCl, 0.15 M NaCl and 0.01% Tween 20 at pH 7.8. The wells were treated with 200 ml enhancement solution. The europium bound to the solid phase was determined as its 2-naphthoyltrifluoroacetone chelate, using a single-photon counting time-resolved fluorimeter (Arcus 1230; Perkin-Elmer Life Sciences, Turku, Finland). Figure 12.2 shows the typical standard curves for the prostaglandins obtained from TR–FIA. The values of the response curves were calculated with the log-lin-transformation (Wood and Sokolowski, 1981). For all assays we obtained almost the same quality. The specificity was very high for all antibodies. Cross-reactions with other structurally related compounds were below 0.1% for most compounds, including the primary prostaglandins. The limit of detection [three standard deviations (SD) of B0 below the zero standard] calculated from 12 consecutive assays was 0.2–1.2 pg/assay. A displacement of over 90% was obtained with the highest standards. The dynamic ranges of the assays were 0.5–1000 pg/well as documented by the precision profiles. The interassay coefficients of variation ranged from 6.7% to 11.4%. The intra-assay coefficients of variation, determined by analysing three biological samples fortified with known amounts of prostaglandins (3–200 pg/assay) in eight replicates during the same run, ranged from 4.6% to 9.3%. Recently, we also succeeded in developing a TR–FIA for the detection of 6-keto-prostaglandin F1a (Schlegel et al 1997). The standard curve for this compound looks similar to the calibration curves shown in Figure 12.2. The standards were used in a range of 1.1–2430 pg/assay tube. The antibody was highly specific and was applied into the assay in a dilution of 1/10 000. The intra- and interassay coefficients of variation were both 57%. ANALYSIS OF PGF2a, PGE2, PGFM and PGEM IN BIOLOGICAL SAMPLES BY THE TR–FIAs Samples of tumour and adjacent tissue from breast cancer patients were obtained during surgery. The biopsies were examined histologically and residual material was used for prostaglandin determination. All tissues were frozen to 7208C until extraction. For the extraction of prostaglandins, the tissues were homogenized in 2 ml isotonic NaCl solution with a Potter homogenizer. The homogenates were centrifuged at 2000 6g for 5 min, and 1 ml aliquots of the supernatants were acidified with 0.1 M HCl to pH 3.5; 5 ml ethyl acetate were used for each extraction. The organic phase was separated by centrifugation, transferred with a Pasteur pipette into a polypropylene tube and dried under a gentle stream of nitrogen. Aliquots (2 ml) of assay buffer (0.05 M Tris– HCl, 0.15 M NaCl, 0.01% Tween 20, 2 g/l BSA, 0.03 M NaN3; pH 7.7) were added to each tube, the tubes were shaken intensively, and aliquots of the solution were stored at 7208C until use. We measured the concentrations of PGF2a, PGE2, PGFM and PGEMcyc in human tissue from seven breast cancer tumours (primary carcinoma) and adjacent non-cancerous tissue with our established TR–FIAs. The results are expressed as amount of prostaglandin per amount of protein in the tissue homogenates. The mean values +SD for the tumour samples were as follows: PGF2a (0.6+0.7 ng/mg); PGE2 (2.0+1.3 ng/mg); PGFM (0.7+0.7 ng/mg); and PGEMcyc (2.1+1.9 ng/mg). In the adjacent tissue we found for PGF2a (0.2+0.2 ng/mg), for PGE2 (0.8+0.5 ng/mg), for PGFM (0.2+0.2 ng/mg) and for PGEMcyc (1.1+0.8 ng/mg). Using Student’s t-test for comparison of values
TIME-RESOLVED FLUOROIMMUNOASSAY
125
Figure 12.1 A general outline of the TR–FIA method as used for routine work
from tumour samples with values from non-tumour samples, the differences were significant higher in the tumour samples (PGF2a, p50.0481; PGE2, p50.0058; PGFM, p50.0376; and PGEMcyc, p50.0263).
ml6106 cells (n=number of animals). The rabbit luteal cells produced mainly PGI2 during the early phase of pseudopregnancy. The addition of 15 mM arachidonic acid to the medium caused a significant stimulation of the PGI2 output.
ANALYSIS OF 6-keto-PGF1a IN BIOLOGICAL SAMPLES BY THE TR–FIAs
PRACTICAL CONSIDERATIONS
The 6-keto-PGF1a –TR–FIA was introduced into animal studies. We wanted to know whether isolated and cultivated luteal cells from pseudopregnant rabbits are able to synthesize PGI2 (for details, see Schlegel et al 1997). The secretion rates of PGI2 by rabbit luteal cells from days 0, 1, 2, 3 and 4 of pseudopregnancy under standard conditions or with exogenously applied 15 mM arachidonic acid (AA) are summarized in Table 12.1. Media were collected every 24 h and the results were plotted as cumulative values (mean+SD). The PGI2 levels are expressed as pmol/
The aim of our study was to develop practical and sensitive TR– FIAs for the determination of prostaglandin levels in biological samples as a real alternative to the widely used RIAs. Europium ions as labels and time-resolved fluorescence as the detection principle are among the most promising alternatives in the field of non-isotopic immunological methods. The high specific activity of the europium label reduces the volume of biological samples and opens new possibilities for investigations when a high sensitivity is required.
126
ANALYTICAL METHODS
Figure 12.2 Dose–response curves for the TR–FIA method in the presence of various amounts of authentic prostaglandins as indicated. Each point is the mean of triplicate determinations
Table 12.1 Production of PGI2 (pmol/ml6106 cells) by isolated rabbit luteal cells during pseudopregnancy on days 0, 1, 2, 3 and 4 under standard conditions or with exogenously applied arachidonic acid (AA) d
n
PGI21
PGI2 (AA)1
0 1 2 3 4
3 3 4 3 3
18.5+2.0 31.1+3.6* 35.7+4.4* 29.6+6.9* 22.6+7.0
29.7+3.1 50.6+5.8* 41.8+7.1* 47.6+11.2* 36.1+9.2
1 PGI2 was measured as 6-keto PGF1a; d, day of pseudopregnancy; n, number of animals; *significantly increased (p50.05) by comparison with day 0; cumulative values (24 h) are presented as mean+SD; the addition of 15 mM arachidonic acid resulted in significantly increased (p50.05) production rates of almost all treatment groups compared with the non-stimulated ones.
As demonstrated by Lu¨ke and Schlegel (1992), prostaglandins are easily linked to biocytin, which has a high affinity to the streptavidin–europium molecule, resulting in a fairly stable tracer. Both the biotinylation method and the conjugation of streptavidin with the europium chelate are standard procedures and can be performed using commercially available chemicals. The potential of our new assays was demonstrated by measuring prostaglandins in tissues from humans and rabbits where low concentrations had to be expected. The outcome of our results are promising for further investigations. REFERENCES Granstro¨m E, Fitzpatrick FA and Kindahl H (1982) Radioimmunologic determination of 15-keto-13,14-dihydro-PGE2: a method for its stable
TIME-RESOLVED FLUOROIMMUNOASSAY degradation product 11-deoxy-15-keto-13,14-dihydro-11b,16x-cycloPGE2. In Colowick SP and Kaplan NO (eds), Methods in Enzymology, vol 86, Prostaglandins and Arachidonate Metabolites. New York: Academic Press, 306–320. Hayashi Y and Yamamoto S (1982) Enzyme immunoassay of PGF2a. In Colowick SP and Kaplan NO (eds), Methods in Enzymology, vol 86, Prostaglandins and Arachidonate Metabolites. New York: Academic Press, 269–273. Jaffe BM and Behrman HR (1974) Prostaglandins E, A and F. In Methods of Hormone Radioimmunoassays, Jaffe BM and Behrman HR (eds). Academic Press, New York and London, 1934. John H, Cammann K and Schlegel W (1998) Development and review of radioimmunoassay of 12-S-hydroxyheptadecatrienoic acid. Prostagland Lipid Mediat, 56, 53–76. Lu¨ke FJ and Schlegel W (1990) A time-resolved fluoroimmunoassay for the determination of prostaglandin F2a. Clin Chim Acta, 189, 257–266. Lu¨ke FJ and Schlegel W (1992) Determination of prostaglandin metabolites in biological samples by competitive time-resolved fluoroimmunoassay. J Immunol Methods, 148, 217–223. Maclouf J, Pradel M, Pradelles P and Dray F (1976) Iodine-125 derivates of prostaglandins: a novel approach in prostaglandin analysis by radioimmunoassay. Biochim Biophys Acta, 431, 139–146. Meyer HHD, Eisele K and Osaso J (1989) A biotin–streptavidin amplified enzyme immunoassay for 13,14-dihydro-15-ketoprostaglandin F2a. Prostaglandins, 38, 375–383.
127
Pradelles P, Grassi J and Maclouf J (1985) Enzyme immunoassays of eicosanoids using acetylcholinesterase as label: an alternative to immunoassay. Anal Chem, 57, 1170–1173. Schlegel W and Greep RO (1976) Kinetic studies on 15hydroxyprostaglandin dehydrogenase from human placenta. In Advances in Prostaglandin and Thromboxane Research, Samuelsson B and Paoletti R (eds). Raven, New York, 159–162. Schlegel W, Ammermann D and John H (1997) Prostaglandin- and thromboxane-producing activity of isolated luteal cells from pseudopregnant rabbits. Horm Metab Res, 29, 237–241. Schlegel W, Urdinola J and Schneider HPG (1982) Radioimmunoassay for 13,14-dihydro-15-ketoprostaglandin F2a and its application in normoand anovulatory women. Acta Endocrinol, 100, 98–104. Schlegel W, Wenk K, Dollinger HC and Raptis S (1977) Concentrations of prostaglandin A, E, F-like substances in gastric mucosa of normal subjects and of patients with various gastric diseases. Clin Sci Mol Med, 52, 255–258. Soini E and Hemmila¨ I (1979) Fluoroimmunoassay: present status and key problems. Clin Chem, 25, 353–361. Weerasekera DA, Koullapis EN, Kim JB et al (1983) Chemiluminescence immunoassay of thromboxane B2. In Advances in Prostaglandin, Thromboxane, and Leukotriene Research, vol 11, Samuelsson B, Paoletti R and Ramwell P (eds). Raven, New York, 185–190. Wood WG and Sokolwski G (1981) Radioimmunassay in Theory and Practice. Schnetztor, Konstanz, 60–70.
Section Three Biochemical and Molecular Pharmacology
13 Perspectives and Clinical Significance of the Biochemical and Molecular Pharmacology of Eicosanoids Subhash P. Khanapure and L. Gordon Letts NitroMed Inc., Bedford, MA, USA
Metabolism of C-20 polyunsaturated fatty acids (PUFAs) by enzymatic pathways, including prostaglandin-H synthases (PGH), epoxygenases and 5-, 12- and 15-lipoxygenases (LOs), leads to the production of eicosanoids. Of these enzymatic oxidative pathways the most significant from a biological standpoint are PGH synthases, which initiate the formation of prostaglandins (PGs) (Willis 1987a, 1987b), prostacyclin (Whittle et al 1978) and thromboxanes (Moncada and Vane 1978). 5-LO is responsible for the formation of leukotrienes (LTs) (Ford-Hutchinson et al 1980; Rokach 1989) and other 5-oxygenated eicosanoids, e.g. 5(S)hydroxy-6,8,11,14-eicosatetraenoic acid (5-HETE) and 5-oxo6,8,11,14-eicosatetraenoic acid (5-oxo-ETE) (O’Flaherty et al 1993, 1996; Powell et al 1993). PGH synthases, which are more commonly termed cyclooxygenases (COXs), exist in two isoforms: COX-1 and COX-2. Exhaustive research in the cyclooxygenase field has helped to understand and separate the roles of the two isoforms. COX-1 is constitutively expressed, whereas inflammatory stimuli and growth factors regulate COX-2 expression (Vane and Botting 2001). In addition to the enzymatic pathways, a parallel, nonenzymatic free-radical-mediated biochemical pathway is operative, which leads to a new class of compounds, the isoprostanes, that are the products of free-radical-induced peroxidation of arachidonic acid (AA). In this overview, recent developments related to the clinical significance of the biochemical and molecular pharmacology of eicosanoids have been reviewed and are presented in three sections: (a) lipoxygenase pathway; (b) cyclooxygenase pathway; and (c) non-enzymatic pathway of AA metabolism. LIPOXYGENASE PATHWAY OF ARACHIDONIC ACID (AA) METABOLISM The introduction of the term ‘‘leukotrienes’’ (LTs) represented the culmination of an over 40-year search to define the structure of slow-reacting substances (SRS and SRS-A). Since the discovery of SRS by Feldberg and Kellaway (1938; Kellaway and Trethewie 1940), several research groups have observed that slow-contracting smooth muscle stimulating substance was released by a lung challenged with cobra venom. The history of how SRS became a slow reacting substance of anaphylaxis (SRS-A) and then became leukotrienes has been very well described in one of the recent reviews (Holgate and Dahlen 1996). The announcement of the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
discovery of leukotriene-C4 at the Washington Prostaglandin Meeting in May 1979 created excitement in the scientific field. The paper was published in the same year (Murphy et al 1979) and it was soon realized that SRS-A was a mixture of cysteinylleukotrienes: leukotriene C4 (LTC4), leukotriene D4 (LTD4) and leukotriene E4 (LTE4) (Corey et al 1980; Samuelsson 1980; Samuelsson et al 1980). Since then the relentless efforts of different research groups have helped to increase the understanding of the cellular and molecular basis of human inflammatory processes, including asthma. For a comprehensive monograph on LTs, see Rokach (1989). The biochemical origin of leukotriene A4 (LTA4), the biologically most important leukotriene intermediate, is from 5-SHPETE (Figure 13.1). The opening of the epoxide by glutathione S-transferase (LTC4 synthase) leads to the formation of LTC4, while the enzymatic conjugate hydrolysis with a stereospecific rearrangement of the triene, by LTA4-hydrolase, produces leukotriene B4 (LTB4). LTC4 is rapidly transformed to LTD4 and then to LTE4 by transpeptidase enzymes. There are three different mammalian lipoxygenases, which catalyse the insertion of molecular oxygen into arachidonic acid at positions 5, 12 or 15, respectively (Figure 13.1). The initial product of the reaction is hydroperoxyeicosatetraenoic acid (e.g. 5-, 12- or 15-HPETE), which is then reduced to the corresponding hydroxyeicosatetraenoic acid (e.g. 5-, 12- or 15-HETE) (Rokach 1989). The 15-LO is found in leukocytes and is involved in the production of lipoxins (Figure 13.2) (Fitzsimmons et al 1985), whereas 12-LO, whose precise function is not yet clear, is found in platelets and leukocytes. The 5-LO pathway has received the most attention, since this is the one involved in leukotriene synthesis and has been identified as a molecular pathway suitable for drug modification. Recently, new biochemical pathways for the production of oxoeicosanoids have been discovered. 5-HETE is converted to biologically active 5-oxo-ETE (Figure 13.1) by 5-hydroxyeicosanoid dehydrogenase (Khanapure et al 2000), whereas 12-HETE is converted to biologically active 12-R-HETE and 12-R-HETrE (Figure 13.3) via a common intermediate, 12-oxo-ETE (Wainwright and Powell 1991). Each of these proinflammatory mediators may have a pathological role in certain inflammatory diseases. New pathways for the metabolism of LTB4 in neutrophils by microsomal 12-hydroxyeicosanoid dehydrogenase to give 12-oxo-LTB4 has also been identified (Figure 13.4; Wainwright and Powell 1991). Generation of 12-oxo-LTB4 is followed by reduction of the 10,11-double bond by a cytosolic
132
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.1 Metabolism of arachidonic acid by lipoxygenase pathways
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
133
Figure 13.2 15-LO-initiated pathway for lipoxin biosynthesis
olefin reductase to give 10,11-dihydro-12-oxo-LTB4 (Khanapure et al 1997). In addition to 12-hydroxyeicosanoid dehydrogenasemediated production of 12-oxo-LTB4, keratinocytes metabolize LTB4 to 12-oxo-LTB4 (Wheelan et al 1993; Khanapure et al 1995), and subsequent glutathione conjugation by glutathione transferase leads to the formation of c-LTB3 and d-LTB3 compounds via the intermediate 12-oxo-c-LTB3 (Wheelan et al 1993). There is a high degree of structural analogy between c-LTB3 and LTC4 as well as d-LTB3 and LTD4. In some recent reports LTC4 and LTD4 have been shown to have proliferative effect on keratinocytes (Kragballe et al 1985), and they are potent mitogens for cultured human neonatal melanocytes (Morelli et al 1989). In vitro experiments have demonstrated that LTC4 also stimulates melanocyte migration and leads to the formation of structures resembling tumour spheroids (Morelli et al 1992; Medrano et al 1993). Source of 5-Lipoxygenase (5-LO) and Its Mechanism of Action 5-LO is found primarily in cells of myeloid origin, e.g. mast cells, neutrophils, basophils and macrophages. It is important to note that the cellular presence of 5-LO is not a requirement for the production of leukotrienes (LTs). Few cells actually possess the enzyme 5-LO, whereas virtually all cells possess LTA4-hydrolase (Samuelsson and Funk 1989; Lewis et al 1990). Consequently, cells devoid of 5-LO activity can synthesize LTB4 in the presence of activated monocytes that provide them LTA4 (Jakobsson et al
1991; Serhan 1991). In human whole blood, LTs are synthesized predominantly by polymorphonuclear neutrophils (PMNs) (Fradin et al 1989). In contrast to PMNs, human eosinophils from asthmatics produce significantly more LTC4 than eosinophils from healthy donors (Schauer et al 1989; Aizawa et al 1990). Further, and unlike other lipoxygenases, 5-LO requires calcium for the activation and initiation of leukotriene biosynthesis in intact cells. The native 5-LO enzyme is inactive and is normally found in the ferrous state (Fe2+). Upon activation by either calcium, hydroperoxides or adenosine triphosphate (ATP), the 5LO enzyme is converted to the active ferric form (Fe3+) and this active form translocates either within the nucleus to the nuclear envelope (Woods et al 1995) or from the cytosol to the cell membrane (Rouzer and Kargman 1988), where it docks with an inner trans-membrane protein known as five lipoxygenase activating protein (FLAP). The understanding of the mechanism of activation of 5-LO was confirmed by studies on the mechanism of action of MK-886 (Gillard et al 1989; Rouzer et al 1990). Once docked to the cell membrane, FLAP presents arachidonic acid (AA), released after the action of phospholipases (PLs), to 5-LO, where it acts as a substrate. The first committed step in the 5-LOmediated oxygenation of AA is the removal of pro-(S) C-7 hydrogen radical from AA, which is then followed by trapping of the radical at the C-5 position to initially give 5-(S)-hydroperoxyeicosatetraenoic acid (5S-HPETE). The enzyme 5-LO then catalyses the second step to convert 5(S)-HPETE to LTA4. The mechanism of this second step, i.e. the LTA4-synthase reaction, involves removal of the pro-(R) hydrogen radical from the C-10 position of 5(S)-HPETE, double bond migrations, and
134
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.3 Biosynthetic pathway for 12(R)-HETE and 12(R)-HETrE
elimination of the hydroxyl moiety from the hydroperoxy group to generate LTA4.
predominant synthesis of LTC4 (Verhgen et al 1984; Bruynzeel et al 1985a, 1985b), while the addition of exogenous AA also results in the synthesis of 5-HETE.
LT Synthesis in PMNs LT Synthesis in Platelets Arachidonic acid is metabolized in neutrophils primarily by the 5-LO pathway, resulting in the formation of 5-(S)-hydroxy6,8,11,14-eicosatetraenoic acid (5S-HETE) and LTB4 (Borgeat and Samuelsson, 1979a, 1979b, 1979c). 5(S)-HETE is further metabolized to the biologically active compound 5-oxo-ETE by these cells, whereas, LTB4 is converted to o-oxidation products with reduced activity (Ford-Hutchinson et al 1983; Powell 1984). 5-oxo-ETE is a potent chemotactic factor for eosinophils (Stamatiou et al 1998). Hence, it may be an important inflammatory mediator in asthma. The relative abilities of LTB4 metabolites to activate neutrophils may also be important in determining the in vivo biological effects (Powell et al 1996). LT Synthesis in Eosinophils Human eosinophils responding to inflammatory stimuli synthesize distinct LTs compared with PMNs. Purified human eosinophils, upon stimulation with ionophore A23187, result in the
In contrast to neutrophils and eosinophils, platelets do not possess 5-LO. Alternatively, and in addition to COX-1, platelets possess 12-LO and hence are capable of converting AA to products of 12-LO. The major product formed by these cells is 12(S)-hydroxy-5(Z),8(E),10(E),14(Z)-eicosatetraenoic acid (12SHETE). Although platelets cannot synthesize 5-LO products from AA, they can metabolize a number of neutrophil-derived products, e.g. platelets convert 5-HETE to 5,12-DiHETE by the action of 12-LO (Marcus et al 1982). Platelets can also convert neutrophil-derived LTA4 to LTC4 because, unlike neutrophils, they contain LTC4-synthase (Pace-Asciak et al 1986; Edenius et al 1988; Maclouf and Murphy 1988). Stimulated platelets convert 5-oxo-ETE to its biologically less active 12-hydroxy metabolite, 5-oxo-12(S)-hydroxy-6(E),(Z),10(E),14(Z)-eicosatetraenoic acid (5-oxo-12-HETE) (Khanapure et al 1998a). Thus, platelets could make an important contribution to the biological inactivation of neutrophil-derived chemotactic 5-oxo-ETE (Powell et al 1998).
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
135
Figure 13.4 Metabolism of LTB4 in keratinocytes
Leukotriene Receptors Classification of leukotriene receptors is mainly derived from functional responses in smooth muscle cell assays of the effects of specific agonists and antagonists.
Leukotriene-B4 Receptors Soon after discovery, it was shown that LTB4 is a potent stimulus for the activation of leukocytes, including chemotactic responses (Ford-Hutchinson et al 1980). In vivo, LTB4 increases leukocyte rolling and adhesion to the venular endothelium (Dahlen et al 1981). This initial response is followed by migration of leukocytes into the extravascular space (Dahlen et al 1981). During initial short-time exposure to LTB4, PMNs are mainly recruited (Bray et al 1981), while upon prolonged exposure, additional actions include secretion of superoxide anion and degranulation (Hafstrom et al 1981; Rae and Smith 1981). LTB4 has also been observed to have an effect on expression of low-affinity receptors for immunoglobin E (IgE) on B lymphocyte cell-lines (Odlander et al 1988). Biological activities of the LTB4 are mediated by the BLT receptor. Two subclasses of BLT receptors have been identified and designated as either BLT1 receptor or BLT2 receptor. Pharmacological evidence suggests that BLT1 and BLT2 receptors belong to the family of G-protein-coupled receptors (GPCRs) (Votta and Mong 1990). BLT receptors have been cloned and species differences have been observed in their sequences. Prior to the discovery of the BLT2 receptor, the BLT1 receptor was referred to as BLT receptor. The guinea-pig cloned BLT receptor contains 348 amino acids and shares 73% and 70% identity with human and mouse BLT receptors, respectively (Boie et al 1999). Besides LTB4, other LTs also have affinity to bind BLT receptors. The
rank order potency for LTs in competition for tritium-labelled LTB4-specific binding at the recombinant guinea-pig BLT receptor is LTB4412(S)-HETE420-COOH-LTB444LTC4= LTD4=LTE4. For the human BLT receptor, the rank order of 12(S)-HETE and 20-COOH-LTB4 is reversed. Subsequently, a new and novel GPCR for LTB4, designated as the BLT2 receptor, was identified and cloned (Wang et al 2000; Yokomizo et al 2000). The BLT2 receptor is a low-affinity receptor which is expressed ubiquitously, in contrast to BLT1, which is expressed predominantly in leukocytes (Toda et al 2002). Tissue expression of BLT2 is broad but the highest is in the liver, intestine, spleen and kidney (Yokomizo et al 2001). Radioligand binding studies have demonstrated that BLT2 has high affinities for LTB4 and LTB5, and intermediate affinities for 12(S)-HETE, 15(S)-HETE and 12-oxo-ETE. These findings suggest distinct biological roles for the BLT2 and BLT1 receptors.
Cysteinyl-Leukotriene (Cys-LT) Receptors Since the identification of SRS in 1979, the peptido-leukotrienes (p-LTs), also referred to as cysteinyl-leukotrienes (cys-LTs), have been shown to have a prominent role in inflammatory diseases, including asthma, cystic fibrosis, allergic rhinitis, rheumatoid arthritis, psoriasis and inflammatory bowel disease. Peptidoleukotrienes are proinflammatory mediators and their biological actions are mediated by Cys-LT receptors and cause bronchoconstriction. The development of Cys-LT receptor antagonists for the treatment of asthma reflects that they have a role in asthma. The p-LTs, LTC4, LTD4 and LTE4 activate contractile and inflammatory processes via interaction with specific G-proteincoupled LT receptors (Evans 2002).
136
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Even before the information about the molecular structure of the receptors for LTs was available, several potent antagonists of Cys-LT receptors were developed as a new therapy for asthma by conventional screening of newly synthesized compounds. Pharmacological characterizations identified two subtypes of cysteinyl leukotriene (Cys-LT) receptors, based on their agonist and antagonist potency for biological responses. On the basis of this evidence, the International Union of Pharmacological Sciences (IUPHAR) introduced the names Cys-LT1 and Cys-LT2 to describe these responses. The rank order potency of agonist activation for the Cys-LT1 receptor is LTD44LTC44LTE4 and that for the Cys-LT2 receptor (previously called the LTC4 receptor) is LTC4=LTD44LTE4. The Cys-LT1 receptor antagonists are efficacious in the treatment of asthma. p-LTs induce migration and enhance degranulation in eosinophils via the Cys-LT1 receptor (Ohshima et al 2002). Hence, interaction of p-LTs and Cys-LT1 receptor is predominantly expressed in human bronchi, peripheral blood leukocytes, lung smooth muscle cells, interstitial lung macrophages and spleen. Cys-LT2 receptor is mainly expressed in heart, adrenal medulla, lungs (Labat et al 1992) and peripheral blood leukocytes (Labat et al 1992; Evans 2002). Human mast cells (hMCs) also express the Cys-LT1 receptor (Mellor et al 2001). There are both tissue and species differences observed in the potency of the antagonists, e.g. a significant difference in the potency of ICI-198,615 between rat lung and guinea-pig trachea has been observed (Tudhope et al 1994). It appears that the presence of additional receptor subtypes may contribute to some differential observation. Recently, the possibility of an additional Cys-LT receptor has been proposed in the involvement of p-LTs in the enteric nervous system (Liu et al 2003). Both human Cys-LT1 and Cys-LT2 receptors have been cloned and characterized (Heise et al 2000). The Cys-LT2 receptor is a 346 amino acid protein with 38% amino acid identity to the Cys-LT1 receptor (Nothacker et al 2000). The cloning of the new Cys-LT2 receptor should prove to be beneficial in understanding its functional role. One of the recent findings suggests that the Cys-LT2 receptor may have biological significance in the cardiovascular system (Kamohara et al 2001). It is detected at high levels in the human atrium and ventricle and at intermediate levels in the coronary artery, whereas CysLT1 was barely detected. Additional evidence supports that there is a possibility of another new subtype Cys-LT receptor in human pulmonary arteries (Walch et al 2002) distinct from Cys-LT1 and Cys-LT2 receptors. The physiological role of Cys-LT2 receptor is not yet defined, although Hui et al (2001) have cloned and characterized mouse Cys-LT2 receptor from heart tissue. Recently, Ogasawara et al (2002) have also cloned and characterized mouse receptors for p-LTs mCys-LT1 and mCys-LT2, which were composed of 309 amino acids with 87.3% identity and 309 amino acids with 73.4% identity to human orthologues, respectively (Ogasawara et al 2002). Different pharmacological characteristics of Cys-LT2 between human, mouse (Hui et al 2001) and guinea-pig (Sakata and Back 2002) and different distributions of p-LTs suggest that a careful choice and interpretation are necessary for a study of p-LTs in animal models. Pathophysiology of 5-LO-derived Products in Disease States Lung Diseases Earlier studies both in vitro and in vivo demonstrated that administration of p-LTs (LTC4, LTD4 and LTE4) resulted in the constriction of airways in a number of different animal species. In
guinea-pig isolated preparations of large airways (tracheal smooth muscle), LTC4 and LTD4 were equipotent. They exhibited similar contractile profiles, which were several orders of magnitude more potent than histamine and which were unaffected by antihistamines but blocked by p-LT receptor antagonist FPL 55712 (Piper et al 1981; Folco et al 1982). LTE4 was found to be less potent than LTC4 but more potent than histamine in contracting guineapig isolated tracheal strips; LTC4 and LTD4 do not contract rat trachea (Krell et al 1982). In in vivo experiments, intravenous administration of LTC4 or LTD4 increases airways overflow and aerosol administration of LTC4 in monkeys produces bronchoconstriction. When histamine is also administered via the intravenous route, LTC4 has a longer duration of action and is approximately 80 times more potent (Smedegard et al 1982). Since synthetic leukotrienes first became available, human airways were one of the first target tissues tested for contractile activity. Dahlen et al reported that at a concentration of 1 nM, LTC4 or LTD4 produced contractile effects similar to that observed after exposure to 1000 nM histamine. Studies on isolated human airways reflected that the constrictor effect of LTC4 and LTD4 was Cys-LT1 receptor-mediated (Jones et al 1982; Buckner et al 1986). The current data indicates that the Cys-LT1 receptor mediates intrinsic airway tone in isolated human bronchi (Ellis and Undem 1994) and asthmatic airway narrowing is primarily mediated by the p-LTs in patients with asthma. Bronchoconstrictor actions of p-LTs in humans in vivo have been extensively studied in normal and asthmatic patients following inhalation of aerosolized p-LTs and measuring lung function by recording the forced expiratory volume in 1 s (FEV1). Asthmatic subjects were more responsive to LTC4 and LTD4 (Davidson et al 1987). With regard to bronchoconstrictor effects of LTE4 relative to the potency of histamine, there are conflicting reports. Davidson et al (1987) reported that LTE4 was 14 times more potent than histamine, while O’Hickey et al (1988) have reported that LTE4 was 112 times more potent than histamine. The reasons for these differences are not clear. The three Cys-LT antagonists, zafirlukast (Accolate), montelukast (Singulair) and Pranlukast (ONO-1078), and the 5-LO inhibitor zileuton (A-64077) have received regulatory approval for the treatment of asthma. The clinical data obtained from these drugs are consistent and complementary (Misson et al 1999). Clinical data, as well as the molecular cloning of human Cys-LT1 and Cys-LT2 receptors, confirms that p-LTs are undoubtedly synthesized in the lung following antigen provocation, are elevated in asthma and other lung diseases, and that their interaction with Cys-LT receptors mediates the inflammatory process and bronchoconstriction. Peptido-LTs are also involved in other diseases of the lung. Urinary levels of LTE4 in adult respiratory distress syndrome (ARDS) patients reflect that there is persistent generation of pLTs in ARDS (Bernard et al 1991). Similarly, LT levels in urine are elevated in children with cystic fibrosis (Sampson et al 1990). These results indicate that 5-LO-derived LTs play a role in the pathophysiology of asthma and other lung diseases. Thus, LT biosynthesis inhibitors (5-LO inhibitors), as well as Cys-LT antagonists, will provide a useful therapeutic approach for the treatment of lung diseases. 5-oxo-ETE is a strong activator of human eosinophils, with a chemotactic potency comparable to those of eotaxin and RANTES (regulated on activation, normal T cell expressed and secreted), both of which act synergistically to enhance 5-oxo-ETEinduced chemotaxis of eosinophils (Powell et al 2001). Thus, 5-oxo-ETE and CC chemokines may collectively contribute to inducing pulmonary eosinophilia in asthmatic patients and their antagonism may provide an alternative treatment of airway disorders.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY Inflammatory Diseases The involvement of leukotrienes plays a pathogenic role in human inflammatory diseases such as inflammatory bowel disease (IBD), Crohn’s disease and psoriasis. Extremely high levels of LTB4 are observed in the colonic mucosa in patients with IBD and Crohn’s disease (Sharon and Stenson 1984). The high levels of LTB4 suggest that 5-LO-derived eicosanoid mediators play an important role in these diseases. In psoriatic lesions, leukotrienes are potent mediators of inflammation. The presence of LTB4, LTC4 and mono-HETEs has been characterized (Brain et al 1982, 1985; Kragballe et al 1985). Among the eicosanoid mediators, formation of 12(R)hydroxy-5,8,10,14-eicosatetraenoic acid (12R-HETE) is increased in psoriasis, and it is present in the highest amounts in psoriatic lesions (Woollard 1986). 12(R)-HETE found in psoriatic lesions is stereochemically different from the 12-lipoxygenase-derived eicosanoid mediator, which produces 12(S)-HETE. The levels of 12(R)-HETE have been found to be 1000-fold higher than those of LTB4 (Woollard 1986). Both LTB4 and 12(R)-HETE are neutrophil chemoattractants when injected intradermally (Fretland and Djuric 1989). The biochemical pathway for the formation of 12(R)-HETE involves 12-oxo-ETE as an intermediate, formed from 12(S)-HETE and/or 12(S)-HPETE by enzyme 12-hydroxyeicosanoid dehydrogenase (Wainwright and Powell 1991). The reduction of 12-oxo-ETE by 12-keto-reductase leads to the production of 12(R)-HETE and 12(S)-HETE (Figure 13.2). Recently there has been increasing evidence to suggest that 12(R)HETE, 12-oxo-ETE, 12(R)-HETrE and 5-oxo-ETE are potent chemotactic factors for neutrophils. The biochemical pathways for the formation of these intermediates have been confirmed using synthetic materials (Wang et al 1994; Khanapure et al 1998b). The addition of platelet-derived 12(S)-HETE to human cultured epidermal keratinocytes results in the stimulation of DNA synthesis, 68% at 1077 molar concentration (Kragballe et al 1985). These data suggest that eicosanoid-induced platelet activation occurs in psoriasis, and that release of inflammatory and mitogenic compounds by activated platelets may play a role in the pathophysiology of psoriasis (Kragballe and Fallon 1986). A recent report on the inflammatory effect of LTC4 on melanocytes (Morelli et al 1992; Medrano et al 1993) and the discovery that keratinocytes metabolize LTB4 to LTC4-like products in keratinocytes, e.g. c-LTB3 (Figure 13.3), has complicated the picture of involvement of proinflammatory eicosanoids in psoriasis. Prominent features of psoriasis are excessive infiltration of PMNs, epidermal hyperplasia and angiogenesis. Indications are that the eicosanoids involved in the pathology of psoriasis exert their biological actions via receptors Cys-LT1, Cys-LT2, BLT1 and BLT2 and two new proposed receptors for 12(R)-HETE and 5-oxo-ETE. Cardiovascular Diseases Peptido-LTs are also implicated in the pathophysiology of cardiovascular diseases (Hand et al 1981; Lefer 1986). In vitro experiments have demonstrated that LTC4 and LTD4 contract guinea-pig, rabbit, rat and human isolated pulmonary arteries (Burke et al 1982; Woodman and Dusting 1982; Letts and Piper 1983). p-LTs have also been shown to be vasoconstrictors in isolated perfused hearts of guinea-pig, rat, cat and rabbit (Letts and Piper 1981, 1982; Sibelius et al 2003). In vivo effects of p-LTs on the cardiovascular system have received comparatively little attention. The biological half-life of LTC4 and LTD4 is very short, and they do not appear to survive one circulatory passage (Letts et al 1981). This is possibly due to
137
factors such as metabolism and binding. The recent finding that the Cys-LT2 receptor is expressed more in the cardiovascular system, and that the binding affinity of LTC4 is high for Cys-LT2 receptor, may have more implications in the contractile effect of LTC4 in pulmonary arteries. Further, a new Cys-LT receptor is also proposed in pulmonary arteries (Walch et al 2002). Studies with Cys-LT2 receptor antagonists are necessary to elucidate the role of LTC4 in cardiovascular disease states. In addition to the effects of p-LTs, non-peptido-leukotriene, LTB4, can increase leukocyte endothelium adhesion and may have significance in ischaemia reperfusion injury.
Renal Diseases There are reports that p-LTs have been implicated in the pathogenesis of different experimental models of glomerulonephritis. They decrease the glomerular filtration rate and renal blood flow (Ardaillou 1990). Menegatti et al (1999) have shown that in human glomerulonephritis with nephrotic syndrome, renal tissue co-expressed 5-LO and LTA4-hydrolase. Clinical and immunological data showed that these patients had impaired renal function that correlated with 5-LO expression. It has been reported that leukotriene production is also stimulated during acute renal allograft rejection (Spurney et al 1990), and 5-LO inhibitors may prevent renal allograft rejection (Anderson and Mangino 1991). LTB4 production is also elevated during nephritis and the increased levels of LTB4 may therefore participate in the inflammatory reactions. In a rat model of renal ischaemic reperfusion injury, it has been demonstrated that LTB4 plays a role (Noiri et al 2000). In rat kidney, LTC4 is metabolized to LTE4 (Bernstrom and Hammarstrom 1981), and LTB4 is metabolized to the biologically less active 20-OH-LTB4 metabolite (Lianos 1986).
Immune System There are examples where LT production is increased and there is a reported correlation between increasing malignancy and increased LT production (Rojers 1989; Simmet et al 1989). 5HETE has also been found to stimulate malignant and normal cell growth. This effect can be blocked by 5-LO inhibitors. Pharmacology of Leukotrienes in Skin, GI Tract, and Eye Skin Inflammation Skin inflammation is a complex reaction leading to increases in vascular permeability, blood flow, pain and proinflammatory cell infiltration. A number of studies have shown that leukotrienes may have a physiological and pathological role in skin; of particular interest is the possibility that leukotrienes may be involved in the pathogenesis of psoriasis.
Skin Blood Flow LTs have been shown to have potent vascular effects in human skin. When injected intradermally in nanomolar amounts in human skin, p-LTs induce significant flare reactions. The effect of p-LTs in human skin is equipotent to histamine. The dilator actions of p-LTs in human skin are unusual, considering the fact that p-LTs induce profound contractions on most vascular preparations.
138
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
There are considerable species variations with regard to the vascular actions of leukotrienes in skin. Leukotrienes are potent vasoconstrictors in guinea-pig skin, whereas they are potent vasodilators in pig skin. The porcine skin is the only animal model that reflects the potent vasodilator actions of LTs seen in human skin. The responses in porcine skin and human skin are different, however, in terms of their duration of action. Shorter duration was observed for porcine skin compared to human skin. The responses are also dose-dependent and in porcine skin the vasodilator effect is observed at low doses (2 pM), whereas in human skin a dose of more than 2 nM of LTC4 or LTD4 is required to observe a similar effect. It has also been proposed that the vasoconstrictor actions of leukotrienes are a result of the direct action of LTs on vascular smooth muscle cells, whereas any vasodilator actions of LTs are mediated by secondarily released mediators from platelets (Letts et al 1985). Vascular Permeability in Skin There are very few human skin studies that involve the direct measurement of changes in vascular permeability. Intradermal injections of p-LTs into human skin induce dose-related wheal reactions. In contrast to the p-LTs, LTB4 induces significantly less wheal formation (Juhlin and Hammarstrom 1983) but produces characteristic areas of induration (Camp et al 1983; Soter et al 1983). In porcine skin only LTB4 (but not p-LTs) induces increases in vascular permeability (Chan and Ford-Hutchinson 1985). The permeability effects of LTB4 are potentiated by the co-administration of PGE2. The enhancement of vascular permeability in humans is associated with the influx of neutrophils (Camp et al 1983; Soter et al 1983). The recruited neutrophils by LTB4 at the inflammatory site render the neutrophils more adherent to vascular endothelium. Epidermal Proliferation In both animal and human models in vitro and in vivo, LTs have been shown to have a profound enhancing effect on epidermal cell proliferation. LTB4 as well as p-LTs stimulate tritiated thymidine incorporation into DNA in cultured human epidermal keratinocytes (Kragballe et al 1985, 1987). The effect of LTB4 is stereospecific, as other stereoisomers were inactive, suggesting the involvement of specific LTB4 binding sites. P-LTs were 10-fold less potent than LTB4 and their effect was blocked by the Cys-LT antagonist FPL 55712. Epidermal hyperplasia is a characteristic feature of psoriasis (Weinstein et al 1983). LTs in Eye and Ocular Inflammation Non-peptido LTB4 and p-LTs (LTC4 or LTD4) have different effects in the eye. LTB4 induces infiltration of PMNs in the aqueous humour (Bhattacherjee et al 1982; Stjernschantz et al 1984). In contrast to what is observed in skin, LTB4 has no vascular effect in the eye. Intraocular injection of LTB4 into the rabbit eye produced no changes in intraocular pressure (Bhattacherjee et al 1982; Stjernschantz et al 1984), suggesting that LTB4 has no direct effect on the blood–aqueous barrier in the ocular chamber. Topically applied nanomolar amounts of LTD4 in guinea-pig eye resulted in marked extravasation of protein in the conjunctiva (Woodward and Ledgard 1985). This effect was blocked by specific LTD4 antagonists L-649,923 and L-648,051, which confirmed that the effect of LTD4 was receptor-mediated. In
rabbit eye, 12(R)-HETrE has been found to be a potent angiogenic factor (Davis et al 1990; Nishimura et al 1991). The observation that LTs are produced in human and animal ocular tissues suggests that LTs may have a role in ocular inflammatory reactions. These initial responses are followed by the infiltration of inflammatory cells (Allansmith et al 1984). Studies with a 5-LO inhibitor (REV-5901) have shown that 5-LO inhibition has a greater effect on the late phase (3.5 h) than on the immediate response. Specific 5-LO inhibitors or LT antagonists may therefore be useful in prophylactic or chronic therapy of ocular inflammation.
LTs in GI Tract In the gastrointestinal (GI) tract, LTB4 and p-LTs (LTC4, LTD4 and LTE4) have differential effects. P-LTs have potent contractile actions in GI smooth muscles and may modify gastric function. On the other hand, LTB4 has little effect in the GI smooth muscles and gastric functions but may play a pathological role in inflammatory bowel disease. Recently, a special issue of Prostaglandins and Other Lipid Mediators was dedicated to reviewing the role of PGs and LTs in the physiology and pathology of the digestive tract (Dubois 2000). Levels of LTB4 and LTC4 are increased in the mucosal biopsies of patients with oesophagitis (reflux disease). These results are consistent with the concept that an increase in LTs reflects the degree of inflammation in the oesophagus and that these levels decrease towards baseline levels after treatment with omeprazole 20 mg daily. The roles of leukotrienes in gastric mucosa have been studied extensively. LTs stimulate pepsin secretion, reduce mucosal blood flow and interfere with gastric emptying. Gyomber et al (1996) have shown that pretreatment with LO inhibitor and LT antagonist exerted protective effects on acute and chronic gastric damage in ulcer models. These findings suggest that LT-inhibitors may be useful in the treatment of gastric ulcers (Rainsford 1987; Rogers et al 1987; Wallace 1990). In Crohn’s disease the small intestine is the major site of inflammation (Lashner 1995). About 70–80% of patients with Crohn’s disease have colonic inflammation of the small intestine. Pathogenesis of IBD involves the chronic inflammation of the large intestine. Along with the overexpression of COX-2, the pathological role of LO-derived products in IBD has been studied extensively. In vivo models suggest that LT activity may be an important contributing factor. LTs will stimulate neutrophil migration and adhesion of leukocytes to endothelial cells, cause smooth muscle contractions and increase vascular permeability and mucus secretion (Eberhart and Dubois 1995). In animal models LT inhibitors decreased inflammation and accelerated GI healing (Wallace and Keenan 1990). The current effective treatment of IBD with sulphasalazine significantly reduces LTB4 production in colonic tissues (Allgayer and Stenson 1988), which indicates that 5-LO-derived products of AA metabolism may play an important role in the pathogenesis of IBD.
Inhibitors of LT Production Elucidation of the biochemical production of LTs and the pharmacology of their actions have led to the development of a number of therapeutic approaches for the treatment of inflammatory diseases, such as asthma. These approaches either prevent the synthesis of LTs, prevent the action of FLAP, prevent the catalytic action of 5-LO or inhibit the biological actions of LTs at Cys-LT receptors.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
139
Figure 13.5 FLAP inhibitors
Figure 13.6 5-LO inhibitors and dual-acting 5-LO inhibitors
FLAP Inhibitors FLAP and the mechanism of FLAP inhibition were discovered when a series of compounds (represented by MK-886) that were observed to be potent inhibitors of intact cell-stimulated LT biosynthesis were also found to be inactive in broken cell 5-LO inhibition assays. The MK-886 compound was discovered from indole libraries derived from indomethacin COX inhibitor research (Figure 13.5; Gillard et al 1989). This compound was a more potent inhibitor in human neutrophil assay (IC50=3–5 nM) but was less potent in a human blood assay (IC50=2.1 mM). MK886 was clinically evaluated in asthmatics but was discontinued due to unremarkable results. Another FLAP inhibitor selected for clinical evaluation was MK-0591. However, the degree of improvement observed in clinical studies was also not satisfactory, which led to discontinuation of the development of this class of compounds. 5-LO Inhibitors Hydroxamic Acid 5-LO Inhibitors The starting point for the design of 5-LO inhibitors was the evaluation of compounds interacting with the catalytically important iron moiety. Earlier compounds evaluated for 5-LO inhibition were hydroxamic acid derivatives, e.g. BW A4C (shown in Figure 13.6) that was evaluated in a Phase I clinical trial. Oral administration of 400 mg three times a day resulted in prolonged
inhibition of ex vivo stimulated LTB4 production in blood samples. A major problem associated with BW A4C was its extensive metabolism and the accumulation of metabolites. Due to limited pharmacokinetic properties of hydroxamic acids in humans, it was therefore discontinued. To prevent drug metabolism, vast arrays of different pharmacophores were evaluated for 5-LO inhibition. One array resulted in the identification of the N-hydroxyurea series of 5-LO inhibitors. Zileuton (Figure 13.6) was selected from hundreds of compounds studied, for clinical evaluation in asthmatic patients in a controlled clinical setting using a variety of stimuli, such as allergens, exercise, cold dry air or aspirin. The positive results from clinical studies resulted in FDA approval of zileuton for the treatment of asthma. ABT-761 is a follow-up compound for zileuton. It is a second-generation 5-LO inhibitor with a 150-fold increased potency over zileuton and longer duration of action (Brooks et al 1995). ABT-761 showed potent inhibition of LT formation, both in vitro and in vivo. More importantly, it demonstrated activity in reducing bronchoconstriction and pulmonary inflammation in a rodent model of airway disease (Bell et al 1997). Currently it has entered Phase III clinical trial for the treatment of asthma. Besides specific 5-LO inhibition, compounds possessing both 5LO inhibitory activity and COX inhibitory activity were designed, e.g. compound ML-3000 has been shown to inhibit COX as well as 5-LO (Laufer et al 1994) and is also currently completing Phase III trials. The observation that di-tert-butylphenol derivatives possess non-ulcerogenic antiinflammatory properties generated potent dual-acting COX-2/5-LO inhibitors, and an example of
140
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.7 LTA4-Hydrolase inhibitors
this series is PD 138387 (Song et al 1999a, 1999b). An interesting new and novel modification of the 5-LO template includes the incorporation of additional H-1 receptor antagonist activities into the molecule (Cai et al 2002). The compound UCB 35440 (shown in Figure 13.6) is a dual-acting 5-LO inhibitor possessing H-1 receptor antagonist properties and is completing Phase I clinical trials.
LTA4-Hydrolase Inhibitors LTA4-hydrolase inhibitors specifically block the generation of LTB4. There is some research data supporting this action in human PMNs and animal models with compounds such as captopril and Bayer x1005 (Muller-Peddinghaus et al 1993; Figure 13.7), but there is no clinical data available for LTA4-hydrolase inhibitors.
Cys-LT Receptor Antagonists (Cys-LTRAs) When early development began to develop Cys-LT antagonists, little was known about the receptors and the possibility of interspecies variability. At first the guinea-pig model served as a species, since it was known that guinea-pig trachea, lung parenchyma and ileum were all known to contract upon exposure
Figure 13.8 LTD4 antagonists (hydroxyacetophenone derivatives)
to SRS-A. The first LTD4 antagonist, FPL 55712 (Figure 13.8), was discovered in 1973, 6 years before the LT structures were defined, by screening crude SRS-A in a guinea-pig assay (Augstein et al 1973). For many years it was practically the only lead LT antagonist. It was hypothesized that of the three molecular segments of LTD4 (lipid, peptide and acid), hydroxyacetophenone and chromanone groups of FPL 55712 were mimicking the triene region and the peptide segment of LTD4, respectively. This approach led to the design of several compounds by different research groups in the hydroxyacetophenone series of LTD4 antagonists, as shown in Figure 13.8, several of which progressed into clinical development but were later discontinued due to adverse side-effects. From the antagonist drug design viewpoint, LY 171883 (Tomelukast; Fleisch et al 1985) was interesting as it showed that the carboxylic acid group can be replaced with bioisosteric tetrazole (Thornber 1979). LY 171883 was the first orally active LTD4 antagonist that showed efficacy in patients with mild asthma (Phillips et al 1987) but it also had to be suspended due to side-effects. Once the structures of SRS-A were elucidated, many pharmaceutical companies started drug discovery programs to design LTD4 antagonists (Figure 13.9). These first-generation LT antagonists, which can be classified as LT structure analogues, had problems of complicated syntheses, chemical stability and short duration of action. Nevertheless, these compounds were
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
141
Figure 13.9 First-generation LTD4 antagonist
viewed as valuable tools and proved the concept of designing LTD4 antagonists as a new and novel class of drug therapy for the treatment of asthma. As shown in the substance called Bay x7195 (see Figure 13.9), the early observations were that simpler rigid groups could be replaced and used to anchor the lipid or peptide region of LTD4 (Gorenne et al 1995). The second generation of LT antagonists were substituted indole and indazole derivatives (Figure 13.10). Indole derivatives are known for improved oral bioavailability. Among the indole and indazole derivatives, ICI-198,615 (Krell et al 1987) and ICI204,219 (Adkins and Brogden 1998) proceeded into clinical trials. ICI-204,219, zafirlukast (Accolate) was the first LTD4 antagonist to receive regulatory approval for the treatment of asthma (Adkins and Brogden 1998). The Merck group used quinoline to mimic the triene region of LTD4 and developed quinoline-based LTD4 antagonists. Two compounds, MK-0571 and its single enantiomer MK-0649 (Figure 13.10), were evaluated in the clinic, but were discontinued due to side-effects observed in the liver. Ultimately, Merck’s efforts to retain the potency and reduce the liver toxicity led to the discovery of MK-0476, montelukast (Singulair), which received the US FDA approval for the treatment of asthma (Jones et al 1995; Labelle et al 1995). Pranlukast was discovered by random screening and marketed in Japan as Cys-LTRA (Nabe et al 1994).
LTB4 Antagonists Current advancements for the treatment of inflammatory diseases, such as rheumatoid arthritis (RA), include the development of COX-2 inhibitors; however, they are not satisfactory treatments
for inflammatory diseases such as IBD and psoriasis (Cole and Hawkey 1992). A conceptually novel mechanistic approach to new treatments is to inhibit the infiltration of leukocytes into the inflammed tissue. This infiltration is mediated by chemotactic factors that control chemotaxis and upregulation of cell surface adhesion molecules. One such major factor is the production of the chemotactic agent LTB4. To inhibit the chemotaxis, LTB4 antagonists were evaluated (Brooks and Summers 1996), but in clinical trials they did not show beneficial effects like Cys-LT antagonists. The research efforts dedicated to the development of LTB4 antagonists have been minor in comparison to Cys-LT antagonists. The recent discovery of BLT2 receptor should help in developing the new BLT2-specific receptor antagonists and understanding the role of LTB4. Biomarkers for 5-LO in Disease States Generally, LTB4 is used as a biomarker of 5-LO activation in disease states and an EIA kit for the quantification of LTB4 is commercially available. 5-LO inhibitory activity of inhibitors is tested by determining LTB4 in cell cultures, pretreated with the compounds of interest (Bossu et al 1999). Stimulation of blood with LPS at 1 mg/ml results in a dramatic increase in the production of LTs, the major products being LTB4 and 5HETE. The priming activity of LPS has been detected with as little as 1–10 ng LPS/ml blood. This also supports the role of LTs in pathological states involving LPS (Surette et al 1993). In asthmatic patients, 5-LO activation of PMNs leads to 5,15diHETE and lipoxins in addition to LTB4, which indicates an increased transcellular mechanism. PMNs from asthmatic
142
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.10 LTD4 antagonists
patients generate more 5,15-diHETE and lipoxins, whereas under the same experimental conditions no detectable amounts of these compounds are observed in PMNs from healthy subjects. Hence, in asthmatic patients, measurements of 5,15-diHETE and lipoxins could be a more specific inflammatory biomarker than LTB4 (Chavis et al 1995). Aspirin-induced asthma is associated with overproduction of LTs, with a shift to the 5-LO products of AA. In aspirin-intolerant asthmatic (AIA) patients, urinary levels of LTE4 are increased. Hence, in the evaluation of LT receptor antagonists in the pathogenesis of AIA, measurement of urinary LTE4 provides a useful biomarker (Kumlin and Dahlen 2000).
b2-agonists and/or decrease in the use of inhaled steroid treatment as a rescue medication, and the improvement in the quality of life, e.g. decrease in nocturnal awakenings. The 5-LO inhibitor zileuton is administered orally in 400 or 600 mg tablets, 4 times daily for the treatment of asthma in patients aged 12 years and older. Results in the clinical trial revealed that the effect of 1.6 g/day and/or 2.4 g/day proved modestly effective compared to placebo group (Israel et al 1996). Improvements in the clinical end-points were assessed based upon improved mean FEV1 and decrease in the use of inhaled b2agonists.
Clinical Studies of 5-LO Inhibitors and Cys-LTRAs
ABT-761
Zileuton
Zileuton has a short duration of action (t1/2=2.5 h). The secondgeneration 5-LO inhibitor ABT-761 has a longer duration of action and has reached the Phase III stage of clinical evaluation for the treatment of asthma. In Phase I clinical studies it showed excellent oral bioavailability. A single 200 mg dose provided more than 95% inhibition of ex vivo-stimulated LTB4 formation in blood samples taken up to 18 h post-dosing. It has shown excellent oral bioavailability and, due to its longer duration of
At present, zileuton is the only marketed drug with a specific mechanism of 5-LO inhibition (Israel et al 1996). Three CysLTRAs, zafirlukast, montelukast and pranlukast (Figure 13.10), have received regulatory approval for the treatment of asthma in various markets. For the evaluation of all these drugs, the clinical end-points were the improvements in FEV1, decrease in the use of
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY action (t1/2=15 h), only once-daily dosing (50–200 mg/day) is required. This compound is showing significant beneficial effects against exercise- and adenosine-induced bronchoconstriction in asthmatics (Reid 2001). The drug is well tolerated and shows linear pharmacokinetics according to body weight (VanSchoor et al 1997; Wong et al 1998a, 1998b). The clinical investigation of this second-generation 5-LO inhibitor will help to clarify outstanding issues regarding the therapeutic benefit of 5-LO inhibitors. Zafirlukast The Cys-LTRA zafirlukast (AstraZeneca) was discovered by a mechanism-based drug discovery approach. On the basis of a beneficial clinical effect, it became the first LTRA to be approved for the treatment of asthma. Zafirlukast is administered orally and clinical efficacy has been demonstrated when 20–40 mg was given twice daily for the treatment of asthmatic patients 12 years and older. In a pivotal clinical trial, patients who received zafirlukast treatment showed an 11% improvement in FEV1 over placebo, and the number of days without use of b-agonists increased from 6/month to 13/month. Montelukast Montelukast is an orally bioavailable drug and has received US FDA approval for the treatment of asthma in children aged 2 years and older (Merck & Co. 1999). Doses of 10–200 mg have similar efficacy, while a dose of 2 mg/day produced a suboptimal response (Altman et al 1998; Noonan et al 1998). The mean improvement in lung function during montelukast treatment was 6–13% over placebo and the use of b2-agonist rescue medicines was reduced by approximately 1 puff/day and nocturnal awakenings by approximately 1 night/week. Paediatric studies with montelukast have been completed. In a clinical trial of 689 asthmatic children 2–5 years old, treatment with a 4 mg tablet once daily for 12 weeks (Bisgaars 2001) showed beneficial effects. A smaller percentage of patients in the montelukast treatment than in the placebo group reported adverse respiratory experiences and they needed fewer days of b2-agonist use along with a decreased use of corticosteroid rescue medicine. Pranlukast Pranlukast is more effective in patients with moderate asthma or in those patients with severe asthma who are not treated with oral steroids (Tomari et al 2001). It also has demonstrated efficacy in patients with aspirin-intolerant asthma (Yoshida et al 2000). Asthmatic patients treated with pranlukast for 4 weeks were found to have decreased levels of exhaled nitric oxide (NO) (Kobayashi et al 1999). Pranlukast has been reported to exhibit not only Cys-LTRA activity but also pharmacological activity, including antieosinophilic effects at 225 mg/twice daily. This additional activity raises further therapeutic possibilities and supports for further investigation of new approaches for the treatment of asthma. Pranlukast and montelukast, but not zafirlukast, are able to interact with the high-affinity binding site for tritiated LTC4 (Ravasi et al 2002). This again raises the possibility that, for a better understanding of the clinical efficacy of the drugs (besides their Cys-LT1 receptor antagonist potency), their pharmacological differences with respect to LTC4 and LTD4, specifically, may be important.
143
In summary, LTRAs have proved moderately effective and their effect appears to be complementary to current treatments. CYCLOOXYGENASE PATHWAY OF AA METABOLISM Prostaglandins (PGs) produced by COXs play physiological and pathophysiological roles in inflammation and nociception. COX exists in two isoforms: COX-1 and COX-2. Both COX isoforms metabolize AA to PGH2, the common substrate for thromboxane A2 (TXA2), prostacyclin (PGI2) and prostaglandin E2 (PGE2) synthesis. TXA2, PGI2 and PGE2 play important roles in the maintenance of cardiovascular homeostasis. TXA2 is primarily synthesized by COX-1 in platelets. COX-1 also derives constitutive PGs, which are cytoprotective in the gastrointestinal (GI) tract. Inducible COX-2 is expressed at sites of inflammation by monocytes and macrophages and plays a key role in mediating the inflammatory process (Smith et al 1998). PG Production (COX-Mediated AA Pathway) An identical catalytic mechanism of AA metabolism is operative for both COX isoforms, as shown in Figure 13.11. The formation of the labile endoperoxide PGH2 occurs in a two-step reaction. In the first step, two molecules of oxygen are inserted into the substrate in a bisoxygenase reaction leading to the formation of PGG2 (Hamberg and Samuelsson 1973; Nugteren and Hazelhof 1973). In the second step, a hydroperoxidase reaction reduces the 15-hydroperoxyl group in PGG2 to PGH2. The endoperoxide PGH2 is enzymatically transformed to PGD2, PGE2, PGF2, TXA2 and PGI2. Endoperoxide metabolism mechanisms for the formation of TXA2 and PGI2 have also been discovered (Moncada et al 1971; Moncada and Vane 1978; Hecker and Ullrich 1989). Pharmacological Activities of PGs PG Receptor and Receptor Subtype Classification PGs are synthesized upon cell stimulation and act on cells in the close vicinity of their synthesis to exert their actions. Biochemical and pharmacological studies have indicated the presence of specific receptors for the actions of various PGs in different tissues. Some PG actions are associated with changes in the cAMP levels, but their bioactivity correlates to the binding affinity to the receptors. Based on binding affinity studies of PGs, a comprehensive classification of PG receptors was proposed and adopted (Coleman et al 1990). The PG receptors based on specificity for TXA2, PGI2, PGE2, PGF2 and PGD2 are classified as TP, IP, EP, FP and EP receptors, respectively (Coleman et al 1990; Narumiya et al 1999). The TP receptor was the first to be isolated and cloned (Hirata et al 1994) from purified human blood platelets (Ushikubi et al 1989). These studies revealed that the TXA2 receptor is a Gprotein-coupled receptor with seven transmembrane domains. Two subtypes of TP receptors have been identified and designated as TPt (responsible for smooth cell contraction) and TPa (responsible for platelet aggregation) (Saussy et al 1985; Mais et al 1989; Masuda et al 1991). The EP receptor has been classified into four subtypes, EP1, EP2, EP3 and EP4, all of which respond to PGE2 but differ in their actions and their responses to various PGE analogues. On the basis of pharmacological action, PG receptors can be classified into three groups: the relaxant receptors, the contractile receptors, and the inhibitory receptors. The IP, DP, EP2 and EP4 receptors mediate increases in cAMP and induce smooth muscle
144
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.11 Cyclooxygenase pathway of AA metabolism
relaxation, and thus belong to the relaxant group of receptors. The TP, FP and EP1 receptors belong to the contractile group of receptors. They mediate Ca2+ mobilization and induce smooth muscle contraction. EP3 is an inhibitory receptor that mediates decreases in cAMP and inhibits smooth muscle relaxation.
venous smooth muscle. Hence, PGI2 is an important eicosanoid that is involved in the cardiovascular system.
Cardiovascular Actions of PGs Expression and Pharmacological Activities of PG Receptors The DP receptor is expressed in specific cells in the brain and is responsible for the sleep–wake cycle. PGD2 released from mast cells in the lung may have a pathological role in allergic asthma. EP receptor subtypes have been studied in detail. EP1 is expressed moderately in the muscularis mucosa layer of the stomach and is involved in local folding of the mucosa. Among the four subtypes, EP2 is least expressed and its expression is upregulated by LPS. EP3 and EP4 receptors are widely distributed in smooth muscle cells throughout the body and mediate contraction of smooth muscles and platelet shape change. PGE2 mediates fever by acting on EP3 receptors. In the gastrointestinal tract, EP3 is expressed in the longitudinal muscle layer but not in the circular muscle layer. It has been suggested that EP3 is involved in acid secretion. EP4 is highly expressed in the gland of the gastric antrum, indicating that EP4 is probably involved in PGE2-mediated mucus secretion. FP receptor is expressed in the corpora lutea and is involved in luteolysis. The FP receptor is also expressed in the cortical tubules of the kidney and in the stomach, where PGF2a exerts constrictor effects. IP receptor is highly expressed in megakaryocytes. It is also liberally expressed in the smooth muscles of arteries, but not in
PGE2, PGD2 and PGI2 are vasodilators, and also inhibit platelet aggregation by stimulating the production of cAMP (Moncada and Vane 1978). PGF2a, however, causes vasoconstriction. TXA2, produced by aggregating platelets, is a potent vasoconstrictor and mediator of platelet aggregation (Moncada and Vane 1978; Saussy et al 1985). The most important action of TXA2 is its proaggregatory effect on platelets, which is the pharmacological basis for the use of low-dose aspirin for the treatment and prevention of cardiovascular diseases.
Renal Actions of PGs In animals, activation of renal PG synthesis by infusion of the COX substrate AA leads to enhanced renal blood flow (particularly in the medulla) and the release of renin. This latter effect is thought to be mediated by PGI2 via interaction with the IP receptor (Larsson et al 1978). Renal failure is associated with enhanced production of vasoconstrictor TXA2 (Morrison et al 1978). Long-term intake of NSAIDs can have a deleterious effect on kidney function, particularly in diseases such as congestive heart failure.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
145
Figure 13.12 Non-selective NSAIDs
NSAIDs/Aspirin/PG Receptor Antagonists NSAIDs In addition to catalysing AA metabolism and transformation to physiologically important PGs, COX is an important pharmacological target. Aspirin (acetylsalicylic acid) was the first drug, introduced in 1889, for the treatment of arthritis and several decades later new NSAIDs such as indomethacin (1963) and ibuprofen (1969) were introduced. In 1971, Vane discovered that aspirin and virtually all other non-steroidal antiinflammatory drugs (NSAIDs) inhibit PG formation and demonstrated that their relative COX-inhibitory potency in vitro correlated to their antiinflammatory activity in vivo (Vane 1971). This important discovery provided a biochemical basis for the antipyretic, analgesic and antiinflammatory actions of NSAIDs and a pharmacological tool to evaluate the role of PGs in pathophysiology. Vane also proposed that both beneficial actions and sideeffects of NSAIDs could be explained by the inhibition of PG synthesis, because PGs, TXA2 and PGI2 are involved in a broad range of physiological and pathophysiological responses of NSAIDs. This proposal led to the rational design of some new antiinflammatory drugs, as shown in Figure 13.12. NSAIDs are widely used for reducing pain and inflammation. However, chronic NSAID use is known to cause serious GI side effects, such as ulceration. NSAIDs induce GI damage by multiple mechanisms and vary in ulcerogenic activity in different regions of the GI tract. Both PG-dependent and PG-independent factors are responsible for the NSAID-induced gastric toxicity. PG-dependent factors include the receptor-mediated actions of PGs on mucus-bicarbonate secretion, regulation of acid secretion and blood flow.
Aspirin Serine-530 is the site for acetylation by aspirin in COX-1, which is also the binding site of the substrate AA. Acetylation by aspirin therefore blocks arachidonate binding to COX-1 (DeWitt et al 1990). One major difference in the aspirin-mediated inhibition of
PG production by the two COX enzymes is the ability of aspirin acetylated-COX-2 (Ac-COX-2) to oxygenate AA to 15(R)-HETE (Lecomte et al 1994; Xiao et al 1997). Acetylated COX-1 is not able to carry out this transformation.
Ac-COX-2-Mediated 15-epi-Lipoxin Biosynthesis When COX-2 is induced in the presence of aspirin, it is acetylated. The Ac-COX-2 switches its catalytic activity and generates 15(R)HETE, which is released and transformed via a transcellular route to 15-epi-lipoxins, as shown in Figure 13.13, by adherent leukocytes (Serhan 1997). 15-epi-Lipoxins-A4 (15-epi-LXA4) have been demonstrated to be more potent than lipoxins-A4 in inhibiting neutrophil adhesion and 15-epi-LXB4 is an inhibitor of cell proliferation (Serhan 1997). Thus, when aspirin is taken when COX-2 is likely to be in place (or even aspirin intake can induce COX-2 in the GI tract), it will acetylate the serine residue of COX-2. In the resulting Ac-COX-2, arginine-120 appears to be the binding site for AA, and this changes the conformation of the AA in the enzyme’s active binding site to give 15(R)-HETE, which can lead to the biosynthesis of 15-epi-lipoxin (15-epi-LXs) (Figure 13.13). 15epi-Lipoxins can then serve as potential endogenous antiinflammatory mediators for beneficial actions, including prevention of GI ulcer formation, renal inflammation and myocardial infarction. The biological actions of lipoxins are in sharp contrast to those of most other proinflammatory LTs. Lipoxins inhibit both neutrophil and eosinophil chemotaxis at nanomolar concentrations (Lee et al 1989, 1991). The most important issue to address is the effect of a combination of aspirin and a COX-2 inhibitor in humans. A COX-2 inhibitor would still be able to block the AcCOX-2 to prevent the AA metabolism to 15-(R)-HETE and its subsequent transformation to 15-epi-lipoxins. Animal model studies have demonstrated that the combination of aspirin and COX-2 inhibition prevents the beneficial actions of 15-epi-lipoxins and can lead to ulcer formation in the rat (Fiorucci et al 2002). This is one of the important issues to be investigated in humans in vivo. The COX-2 is expressed in monocytes and, if 15-LO is also present, then in addition to a transcellular mechanism there is a
146
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.13 Ac-COX-2-mediated 15-epi-lipoxin biosynthesis
high possibility that 15-epi-lipoxins would be formed within the same cells, i.e. COX-2 expressed monocytes.
PG Receptor Ligands Cloning of PG receptor subtypes has helped towards the design of ligands to study the properties of the agonist and antagonist molecules. A few examples are illustrated in Figure 13.14. This is a complex picture and it should be noted that there is a crossreactivity observed with IP receptor ligands on the EP3 receptor. There have been extensive research efforts in generating TXA2synthase inhibitors, but they all have failed to achieve therapeutic potential in clinical trials. The disappointing results were probably because TXA2-synthase inhibition leads to accumulation of PGH2, which shares a common receptor with TXA2 and can exhibit the same pharmacological profile (Bhagwat et al 1985). For the development of TP receptor antagonists, many structural analogues of TXA2 or PGH2, such as SQ-28668, SQ-29549, ICI180,080 and u-46619, were also developed, as shown in Figure 13.14. Indole derivatives such as L-655,240 and L-670,596 were developed as TP receptor antagonists (Hall et al 1987; FordHutchinson et al 1989). Whereas the clinical efficacy of TP receptor antagonists is yet to be demonstrated, a few remain for experimental studies (Ogletree et al, 1992, 1993).
Involvement of PGs in Inflammation, Pain and Fever Inflammation Classic signs of acute inflammation are swelling, increased blood flow and vascular permeability. PGs are primarily involved in
vasodilation in the inflammatory process and synergize with other proinflammatory mediators, such as histamine and bradykinin, to cause an increase in vascular permeability and oedema. Among the PGs, PGE2 and PGI2 are the most powerful vasodilators and they are present at high concentrations at the site of inflammation (Davies et al 1984). In a recent study, classical carageenan-induced paw swelling was used to test the role of PGI2 and PGE2 using IP-deficient mice. In this model indomethacin was used to treat wild-type mice and decreased the swelling by *50%. IP-deficient mice developed swelling only to levels comparable to those observed in indomethacin-treated wild-type mice, and indomethacin treatment of IP-deficient mice did not induce any further decrease in swelling (Murata et al 1997). Intradermally injected PGE2 could synergize with bradykinin to induce increases in vascular permeability in both wild-type and IP-deficient mice. These results indicate that PGI2 and the IP receptor play important roles in inflammation, and the PGE2 and EP receptor system are utilized depending on the stimulus and site of inflammation.
Pain Inflammatory pain is partly due to the involvement of PGs in nociception. The main site of PG action lies in the peripheral nervous system, in which PGs are believed to sensitize the free ends of neurons. PGE2 and PGI2 exert stronger effects than other PGs, which indicates the involvement of EP and IP receptors in inducing inflammatory pain (Bley et al 1998). A recent study using a selective monoclonal antibody against the EP receptor showed that it inhibits phenylbenzoquinone-induced writhing in mice and carageenan-induced paw hyperalgesia in rats to the same extent as
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
147
Figure 13.14 NSAID ligands (agonists/antagonists)
indomethacin (Portanova et al 1996). This study suggests the involvement of EP receptors in inflammatory pain.
Recent Developments Discovery of COX-2 in Inflammation
Fever Fever is one of the important components of the acute phase response to inflammatory stimuli or infectious organisms, such as LPS. Exogenous pyrogens stimulate the production of cytokines such as IL-1, IL-6, TNFa, IFN-a, and IFN-g. These cytokines can then act on the preoptic area which stimulates the neural pathways that raise body temperature. Fever can be suppressed by NSAIDs, which clearly reflects that PGs are important in fever generation. PGE2 has been identified as a mediator of fever. Levels of PGE2 are increased in the brain during LPS-induced fever. Studies in rats using the EP1 and EP3 receptor agonist, 17phenyl-PGE2, indicated that PGE2-induced fever in rats is mediated via interaction with the EP1 receptor, whereas EP receptor-deficient studies demonstrated that PGE2 mediates fever generation by acting on the EP3 receptor. Both results clearly indicate that PGE2 has a central role in development of PGdependent fever. Non-PG-dependent fever is induced by IL-8 and macrophage inflammatory protein-1b (Zampronio et al 1994).
For many years, COX was thought to exist as a single enzyme in platelets and other cells in organs and tissues. The pioneering work of Needleman (Fu et al 1990; Masferrer et al 1990) and Herschman (1996, 1999) provided convincing evidence for the inducible isomeric COX enzyme, later named cyclooxygenase-2 (COX-2). The expression of COX-2 is upregulated by a variety of inflammatory stimuli, growth factors and bacterial endotoxin. The differential distribution of COX-2 raised the possibility of separating the roles of COX-1 and COX-2. The search for safe NSAIDs with fewer side-effects has greatly intensified and this has led to a better understanding of the mechanisms involved in inflammation. Since COX-2 is induced at the sites of inflammation, it was suggested that the antiinflammatory effects of NSAIDs were due to their COX-2 inhibition, whereas adverse effects were due to the inhibition of constitutive COX-1 (Laneuville et al 1994; Mitchell et al 1994). This hypothesis has been validated by the demonstration that selective COX-2 inhibitors are antiinflammatory and analgesic, with reduced GI toxicity when compared to non-selective NSAIDs. For a
148
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
comprehensive monograph on COX-2, see Vane and Botting (2001). COX-2 Structural Profile COX-1 and COX-2 active sites are almost identical and have only slight differences. Critical hydrophilic residues in the active site are arginine-120, serine-530 and tyrosine-385. Isoleucine-530 in COX1 is substituted for valine-523 in COX-2, and it is this smaller side chain in this residue that provides access to the hydrophobic side channel that is exploited by COX-2 selective drugs. Both NSAIDs and AA enter the active cyclooxygenase catalytic site through the same channel. The entrance of this channel originates in the lipid bilayer of the endoplasmic reticulum. NSAIDs form complex interactions within the COX active catalytic site. These include an important hydrogen bonding interaction of carboxylic acid with arginine-120. These drugs compete for binding to the active site better than AA and thus prevent PGG2 synthesis. Significance of COX-2 Side Pocket The first major difference in COX-1 and COX-2 is the substitution of isoleucine-523 in COX-1 by a smaller valine residue. This valine substitution opens up access to a ‘‘side pocket’’ adjacent to the main COX-2 channel. Additionally, exchange of histamine (513) in COX-1 for arginine (499) in COX-2 allows hydrogen bonding with the sulphone pharmacophore of a COX-2 inhibitor. The second difference at the top of the channel is the replacement of phenylalanine (503) in COX-1 by smaller amino acid leucine (489), which creates extra space in COX-2 to accommodate larger inhibitors to bind in COX-2 than in COX-1. Medicinal chemists have explored this side-pocket in COX-2 for the development of COX-2 selective inhibitors (Dannhardt and Laufer 2000; Ryn et al 2000). Relevance of COX-2 Selectivity The selectivity of the COX-2 inhibitors is determined by the ratio of COX-1:COX-2 inhibitory potencies. It should be noted that the ratios are highly variable, dependent upon the assay system, e.g. pure isolated enzyme, intact cell or whole blood assay (Ryn et al 2000). Considering these limitations, the whole blood assay is the most appropriate to determine COX-1/COX-2 selectivity, which more accurately reflects physiological conditions. The clinical relevance of the COX-2 selectivity is another issue which still needs to be evaluated thoroughly (Patrono 2001; Patrono et al 2001). Development of COX-2 Selective Inhibitors During the last decade, there have been remarkable efforts concerning the synthesis of a variety of structural types of COX-2 inhibitors, identification of selective COX-2 inhibitors with an attractive pharmacological profile, and development of lead compounds for clinical evaluations. The in-depth review of structure–activity relationship (SAR) efforts is beyond the scope of this overview. Interested readers are advised to refer to other recent review articles (deLeval et al 2000; Kalgutkar and Zhao 2001; Rodrigues et al 2002). It is worth mentioning that the Patent Cooperation Treaty (PCT) literature on COX-2 inhibitors is enormous.
extensive libraries for identification of novel COX-2 selective inhibitors (Dannhardt and Kiefer 2001; Kalgutkar and Zhao 2001). The hypothesis is that COX-2 inhibition should block the PG production in inflammatory cells, while not interfering with the production of gastroprotective PGs in the GI tract by COX-1. Prior to the proposal of the COX-2 hypothesis, compounds such as NS-398 and DuP-697, as shown in Figure 13.15, demonstrated antiinflammatory activity with GI sparing profiles in animal models (Gans et al 1990; Futaki et al 1993; Galbraith 1992). Later it was confirmed that DuP-697 and NS-398 preferentially inhibit COX-2 over COX-1 (Copeland et al 1994; Futaki et al 1994). This led to the exploitation of five-membered ring templates for the development of selective COX-2 inhibitors. Structural biology (Marnett et al 1999; Kiefer et al 2000) and X-ray crystallography (Picot et al 1994; Kurumbali et al 1996) provided additional useful information. For good COX-2 inhibitory activity and selectivity, compounds require a key pharmacophore (Dannhardt and Laufer 2000), the para-(methylsulphonyl)phenyl group, attached to a five-membered ring in which additional vicinal ring substitution is present. The para-methylsulphonyl group may be replaced by a sulphonamide moiety, albeit with a loss of COX-2 selectivity. A variety of five-membered ring templates were utilized to give ortho-diaryl substituted compounds with rigid conformations. Major SAR efforts were developed towards alteration of the fivemembered ring templates, which include: pyrazoles (Penning et al 1997; Ochi and Goto 2000), furanones (Prasit et al 1999), cyclopentenes (Reitz et al 1994; Li et al 1995b), cyclopentenones (Black et al 1999), imidazoles (Khanna et al 1997a, 2000), isoxazoles (Talley et al 2000), pyrroles (Khanna et al 1997b), oxazoles (Matsushita et al 1997) and oxazolones (Puig et al 2000). The five-membered ring has also been replaced by six-membered aryl (Li et al 1996) and heteroaryl rings (Davies et al 2001). Etoricoxib is an example of a diaryl-substituted pyridine COX-2 selective inhibitor (Davies et al 2000, 2001). In a new approach to vicinal-substituted COX-2 selective inhibitors, the effect of incorporating a spacer group between the aryl group and the benzo-1,3-dioxolane central ring (structure A, Figure 13.15) was found to play an important role in COX-2 inhibitory potency (Khanapure et al 2003; Khanapure et al 2002). Modification of NSAIDs into Selective COX-2 Inhibitors Classical NSAID templates were exploited to convert a nonselective COX inhibitor into a COX-2 selective inhibitor. This approach was entirely based on modification of a NSAID drug; there was no COX-2 pharmacophore in the template molecule which could fit into the COX-2 side-pocket to induce COX-2 selectivity. It has been demonstrated that neutralization of the carboxylic acid moiety in indomethacin by conversion to the corresponding amides imparts COX-2 selectivity into these derivatives (Li et al 1995a; Kalgutkar et al 2000). COX-2 selectivity was achieved in the diethoxy derivative of flurbiprofen (Bayly et al 1999) due to preferential accommodation of the diethoxy groups into the more accommodating cyclooxygenase-2 catalytic domain. The Novartis discovery group developed a COX-2 selective inhibitor, COX-189 (lumiracoxib; Figure 13.15), which is a modification of the diclofenac template. Role of COX-2 Inhibitors in Renal Function
Pharmacological and Clinical Profile of COX-2 Inhibitors Since its discovery, selective inhibition of COX-2 has proved to be a useful therapeutic target (Turini and DuBois 2002). As a result, a number of pharmaceutical companies have developed and tested
Vasodilator actions of PGs are important to prevent renal ischaemia in patients with renal insufficiency. Both NSAIDs and COX-2 inhibitors decrease the renal producion of PGI2, as reflected by decreased levels of urinary 2,3-dinor-6-keto-PGF1a, a metabolite of PGI2. In the kidney, COX-2 is also expressed
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
149
Figure 13.15 COX-2 inhibitors
constitutively in the macula densa (Harris 1996) and COX-2derived PGI2 may be responsible for activating the renin– angiotensin system (Schneider and Stahl 1998).
Role of COX-2 Inhibitors in Brain In the central nervous system, COX-2 is also expressed constitutively in the brain and spinal cord (Hoffmann 2000). The treatment of degenerative brain disorders with selective COX-2 inhibitors is thus an interesting experimental hypothesis, e.g. cerebral ischaemia leads to upregulation of COX-2. Expression of COX-2 is increased in the sensile plaques characteristic of Alzheimer’s disease (AD). At present, there is limited information on the effects of selective COX-2 inhibitors in the brain and further studies are needed to provide information on the potential beneficial effects.
Clinical Confirmation of COX-2 Hypothesis The clinical efficacy and impressive GI safety of COX-2 inhibitors resulted in the US FDA’s approval of two COXIBs, celecoxib and rofecoxib, as the first generation of selective COX-2 inhibitors marketed for the treatment of inflammatory diseases. Specifically for the treatment of acute pain, osteoarthritis and rheumatoid arthritis (Penning et al 1997; Chan et al 1999; Prasit et al 1999), they seem to be as effective as classical NSAIDs, with reduced deleterious GI side-effects (Jackson and Hawkey 1999; FitzGerald
and Patrono 2001). Recently, valdecoxib received FDA approval as a COX-2 selective antiinflammatory drug (Talley et al 2000; Yuan et al 2002) and another COX-2 selective inhibitor, JTE-522 (Hashimoto et al 2002), is in a Phase II clinical trial, while COX189 (lumiracoxib; Rordorf et al 2003) and etoricoxib are completing Phase III clinical studies. Recently, two large clinical studies have been completed, in which the GI safety in patients taking rofecoxib (VIGOR) (Bombardier et al 2000) or celecoxib (CLASS) (Silverstein et al 2000) were compared with naproxen, ibuprofen or diclofenac. In the VIGOR study of rofecoxib vs. naproxen, there were clinically significantly fewer occurrences of upper GI ulcers in patients taking rofecoxib compared to patients taking naproxen. In the class study of celecoxib vs. ibuprofen or diclofenac, in contrast to the VIGOR study, patients (*21%) were allowed to take aspirin at the cardiovascular dose. In all patients, including those taking aspirin, there was a reduction in the incidence of ulcers from 3.5% to 2.8%. In summary, the COX-2 hypothesis has been clinically confirmed. The launch of selective COX-2 inhibitors as a new drug class has been one of the most successful drug launches in the modern drug development era.
NON-ENZYMATIC PATHWAY OF AA METABOLISM Non-enzymatic free radical-catalysed metabolism of AA leads to the production of isoprostanes (iPs), which has been the subject of intense scientific research over the past decade. Earlier work was
150
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.16 Biochemical formation of iPs. ROS, reactive oxygen species
focused on the renal vasoconstrictor effect of iPF2a-III (previously called 8-iso-PGF2a). At that time considerable attention was paid towards the synthesis of iPs for their biological evaluation, as well as to use authentic synthetic materials and radiolabelled compounds for the development of sensitive methodologies to quantify their levels in biological fluids. iPs have also been exploited to understand the role of free radicals in human physiology and pathophysiology, as well as to measure the extent of lipid peroxidation in several disease states. Since iPs are the stable byproducts of lipid peroxidation, their quantification represents an accurate, sensitive, non-invasive and reliable index of in vivo lipid peroxidation and oxidant stress. Increased levels of plasma and urinary iPs have been reported in several syndromes and their increased levels have been correlated with the severity of the disease. The most recent developments indicate that production of iPs is a complex process and that the iPs produced may be involved in a number of pathophysiological conditions. iPs are thought to play important pathological roles in cardiovascular dysfunction, renal, lung and neurodegenerative diseases. This section will review the pharmacological and clinical significance of iPs. Biochemical Formation of iPs from Phospholipids The formation of prostaglandin-like compounds by peroxidation of polyunsaturated fatty acids (PUFAs) in vitro was first demonstrated more than 26 years ago (Pryor et al 1976). However, it was Roberts and co-workers who, for the first time in 1990, demonstrated that isomeric prostaglandin-like compounds are produced in humans, and named these PG-like compounds as ‘‘F2 isoprostanes’’ (Morrow et al 1990). Isoprostanes are formed by a non-enzymatic free radical-catalysed peroxidation of AA, either in its free form or while still esterified to membrane phospholipids, and subsequently released by the action of phospholipases (PLs) (Figure 13.16). Mechanism of iPs Formation The mechanism by which iPs are formed by peroxidation of AA is shown in Figure 13.17. Initially, three arachidonyl radicals at the 7, 10 and 13 positions are formed, which then undergo endocyclization to form four bicyclic endoperoxide intermediate regioisomers, which lead to the formation of four types of regioisomers as shown in Figure 13.17. Nomenclature When first discovered, iPs were named according to the prostaglandin nomenclature, with the first isoprostane being
named 8-iso-prostaglandin-F2a or 8-epi-prostaglandin-F2a. The production of iPs is a complex process, giving rise to four types of products as shown in Figure 13.17. Within each type, 16 stereoisomers can be formed, eight cis-isoprostane isomers (major) and eight trans-isoprostane isomers (minor); thus, in the F2 series there are 64 iPs. To differentiate the numerous stereoand regioisomeric structures, two nomenclatures were proposed. In 1997 Taber et al (1997) proposed a new nomenclature for isoprostanes, and in the same year Rokach et al (1997a) modified their original proposed nomenclature to bring more logic into it and to be applicable for all newly synthesized iPs (Rokach et al 1997b, 1998) and products derived from o3 and o6 PUFAs. A new classification system for iPs was introduced, based on the o system of counting the numbers of PUFA double bonds. During the last 5–6 years, more than 500 publications have appeared in this field and the use of all three different nomenclatures is confusing, not only for the non-specialist in the eicosanoid area but also for the specialists themselves. In this review, we have used the nomenclature proposed by Rokach et al (1997a) and only in a few places is the old or other nomenclatures specified in parentheses. Nomenclature of representative examples from each type (in which two side-chains are cis to each other and anti to hydroxyl groups) and their all syn isomers (in which two side-chains are cis to each other and syn to hydroxyl groups) are given in Figure 13.18.
Pharmacological Activities of the iPs Biological Properties The majority of iPs are initially formed esterified to phospholipids and are then released in free form. Once released from cell membranes by PLs, they circulate in plasma and are excreted in the urine. Studies exploring the biological activities of iPs have been limited by the availability of synthetic compounds to the broad scientific community. Most studies have focused on the biological activity of iPF2a-III (also referred to as 15-F2t-IsoP, 8iso-PGF2a or 8-epi-PGF2a) which is one of the few iPs commercially available. iPF2a-III causes potent contractions of isolated blood vessels (Kromer and Tippins 1996; Gardan et al 2000; Oliviera et al 2000), lymphatics (Sinzinger et al 1997), airway (Kawikova et al 1996) and myometrial smooth muscle (Crankshaw 1995). The constrictor effect of iPF2a-III is increased in a pathophysiological setting. Thus, its effect is more pronounced after ischaemia (Kromer and Tippins 1996, 1999), endothelial damage and nitric oxide synthase (NOS) inhibition (Jourdan et al 1997; Sametz et al 1999). Metabolism of iPF2a-III leads to two major metabolites in human tissues: 2,3-dinor iPF2a-III and 2,3-dinor-iPF1a-III (Chiabrando et al 1999; Burke et al 2000). There are very few
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY
151
Figure 13.17 Non-enzymatic AA metabolism
reports about the biological activities of these metabolites. In rat thoracic aorta (Roberts et al 1996; Chiabrando et al 1999; Cracowski et al 2002) neither of these metabolites have any vasoconstrictor or dilator effect, but in porcine brain microvessels (Hou et al 2001) 2,3-dinor,5,6-dihydro-iPF2a-III was found to exhibit a constrictor effect. An unresolved issue regarding biological activity is whether the effect observed in vitro is consistently observed in vivo at pathophysiological concentrations and whether these effects contribute to pathological states in vivo. Plasma concentrations of iPF2a-III have been found to be 100–500 pg/ml ranges. These concentrations are unlikely to induce vasoconstrictor effects, considering the in vitro EC50 values of iPF2a-III; however, if iPs are produced at the site of injury and subsequently enter the circulation to be excreted, the local concentration may be high enough to induce vasoconstriction. Nanomolar (nM) concentrations of iPs, e.g. iPF2a-III and iPF2a-VI, were observed in the coronary sinus following coronary angioplasty (Reilly et al 1997; Iuliano et al 2001).
Pharmacology and Role in the Lungs Effect on smooth muscle. Isoprostanes can exert diverse biological effects on smooth muscles, depending upon the species and the tissue. The majority of published reports describe the effect of iPF2a-III on rat tissues and very little is known about their effects in human tissues. iPF2a-III and iPE2 are potent constrictors in several tissues, including the aorta (Kromer and Tippins 1998, 1999), coronary (Kromer and Tippins, 1996, 1998; Mobert and Becker 1998), carotid (Mohler et al 1996), pulmonary (Janssen 2000, 2001), retinal (Lahaie et al 1998; Michoud et al 1998) and renal arteries (Takahashi et al 1992; Fukunaga et al 1993b). They also constrict airway (Banerjee et al 1992; Kang et al 1993; John and Valentin 1997) and intestinal smooth muscle (Elmhurst et al 1997). In rat pulmonary artery, iPF2a-III causes vasoconstriction via TP receptor activation, as well as vasodilation via a non-TP receptor (Jourdan et al 1997). In the airways the contractions induced by iPs are unaffected by LTantagonists or LT-inhibitors (Janssen 2000). In guinea-pigs the
152
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 13.18 Nomenclature of iPs
intratracheal instillation of iPF2a-III at 1–10 nM doses causes dose-dependent increases in airflow obstruction and plasma exudation. The TP receptor antagonist Bay u3405 abolished the airway effects. The levels of TXB2, the stable metabolite of TXA2, were increased in bronchoalveolar lavage fluid (BALF) after instillation of iPF2a-III. Collectively, these data suggest that airflow obstruction may be mediated by the TP receptor and in the guinea-pig a secondary generation of TXA2 may be involved in response to iPF2a-III (Okazawa et al 1997). Canine and porcine tissues respond differently; in canine and porcine airway smooth muscle or pulmonary arteries, F2-iPs were found to be ineffective, whereas iPE2 exerts EP receptor-directed c-AMP-dependent relaxation in canine and porcine airway smooth muscle (Janssen 2001; Janssen and Tazzeo 2002). Recently, Janssen et al (2000) reported that iPs have an excitatory effect in the human airway, pulmonary artery and pulmonary vein. An interesting initial observation was that iPE2 was found to be 10–100-fold more potent than iPF2a-III. A high degree of specificity was also observed, since iPF3a-III, which possesses one more extra double bond at the o3 position and can be derived from peroxidation of EPA in vivo, evoked dosedependent relaxation (Janssen 2000). Mediators of free radical damage in the lung. Recently, much attention has been broadly focused on the effects of free radicals on smooth muscle. In fact, all of these effects may be the combined result of free radical-induced iP production on membrane phospholipids and the pharmacological actions observed after their release. It is well known that oxidative stress is caused by excessive production of free radicals that will lead to increased metabolism of AA and produce acute lung injury (Janssen 2000).
suggesting the potential role for iPs as quantitative markers of oxidant stress (Delanty et al 1997). Atherosclerosis. Oxidative modification of low-density lipoprotein (LDL) is a major event in the development of atherogenesis. Production of oxidatively modified LDL in the sub-endothelial space exhibits biological actions relevant to development of atherosclerotic processes, compared with native LDL. Increased local levels of both minimally modified (MM)LDL and iPs are observed during an inflammatory response involving neutrophils, leading to tissue injury in the pathogenesis of ischaemia-reperfusion syndrome and restenosis after PTCA (Roque et al 2000; Simon et al 2000; Iuliano et al 2001). MMLDL and iPs are also present in human atherosclerotic plaques of coronary arteries from patients with acute coronary syndromes. During acute coronary reperfusion, injury is largely initiated by free radicals generated by neutrophils and is mediated by adhesive interactions with platelets and fissured arterial vascular surfaces. It has been demonstrated that b2 integrin-dependent neutrophil adhesion induced by MM-LDL is mediated by iPF2a-III (Fontana et al 2002). Animal studies have demonstrated that blockade of neutrophil adhesion to endothelium attenuates ischaemiareperfusion injury. It is known that, despite standard aspirin therapy, leukocyte activation and platelet adherence still occurs after coronary angioplasty; aspirin does not prevent PTCAinduced vasopasm or restenosis after PTCA (Thornton et al 1984). It also does not prevent iP formation. The production of iPs in atherosclerotic plaque may have important consequences in vivo and contribute to the pathophysiology of atherosclerosis. In human carotid atherosclerosis it has been shown that iPs contribute to plaque instability (Mallat et al 1999).
Alcohol-induced Liver Disease (AILD) Involvement of iPs in Cardiovascular Dysfunction Myocardial infarction. It is known that myocardial infarction is the result of damage due to excessive production of free radicals. Potential sources of free radicals during myocardial postischaemia reperfusion injury are related to reactive oxygen species (ROS) produced by activated neutrophils. Another potential source of ROS is the enzyme xanthine oxidase, which is localized within the vascular endothelial cells. Increased iP production has been observed during coronary reperfusion,
Earlier studies in a rodent model of AILD reported elevated levels of iPF2a-III (Nanji et al 1993, 1994a, 1994b). Recently, in a randomized, placebo-controlled trial FitzGerald and co-workers reported that alcohol ingestion induces oxidant stress in both healthy human volunteers and patients with AILD (Meagher et al 1996, 1999). In this study it was found that alcohol consumption increased urinary excretion of iPF2a-III and iPF2a-VI in a dosedependent manner, with a significant correlation between the peak plasma alcohol concentrations and the urinary F2-isoprostane
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY levels. In the excretion of iPs in individuals with chronic AILD, both iPF2a-III and iPF2a-VI were markedly increased in patients with acute alcoholic hepatitis. In general, urinary iPF2a-III and iPF2a-VI levels were significantly elevated in patients relative to controls, when cirrhosis was induced by alcohol rather than by hepatitis C. In addition to the increment in the urinary iPF2a-III, it has also been reported that excretion of 2,3-dinor-5,6-dihydroiPF2a-III, which may be formed as a metabolite of iPF2a-III (Burke et al 2000), is also increased in patients with chronic AILD. This suggests that the increments in urinary iPF2a-III represent an increase in its generation in vivo, rather than oxidative metabolism due to impairment of hepatic function. Renal Diseases (Renovascular Hypertension) Since its discovery, iPF2a-III (8-iso-PGF2a) has been studied extensively under pathophysiological conditions involving radical generation. Intrarenal infusion of iPF2a-III in rodents causes an abundant reduction in glomerular filtration rate and blood flow, which clearly indicates that it acts as a glomerular vasoconstrictor (Takahashi et al 1992). The cortex region of the kidney is the major source for the generation of TXA2 and PGI2, in addition to PGF2a, PGE2 and PGD2. The medullar region is more restricted to the generation of PGE2. Klein et al (2001b) addressed the issue of iP generation in isolated rat glomeruli cells, and demonstrated that the production of vasoconstrictor iPF2a-III is increased under oxidative stress condition. Recent clinical studies demonstrated that oxidative stress is enhanced in renovascular disease (RVD) and is reflected by increased levels of iPs. RVD in most cases is associated with activation of the renin–angiotensin system due to a fall in renal blood flow and renal artery stenosis. The end-stage of this disease is renal failure, due to increased angiotensin II activity, resulting in vasoconstriction, increased endothelin release, atherogenesis and glomerulosclerosis (Minuz et al 2002). Neurological Disorders Alzheimer’s disease (AD) is the most common form of neurodegenerative disease, with dementia, in elderly patients. Since the discovery of iPs, both studies performed in living patients and the development of transgenic animal models of AD– amyloidosis have contributed greatly to our understanding of the role of ROS in AD pathogenesis. The recent information points towards an earlier involvement than previously thought of oxidative stress in the pathogenesis of this disease, as indicated by increased levels of iPs (Pratico et al 1998c; Montine et al 2002; Pratico 2002). Diabetes Studies in a rat model of diabetes (Shinomiya et al 2002), as well as recent studies in human patients with type 2 diabetes (Gopaul et al 1995; Davi et al 1999; Sampson et al 2002), showed a significantly higher production of iPF2a-III. These data suggest that increased production of iPs, as a result of hyperglycaemia, may be attributed to complications of diabetes and cardiovascular diseases. Receptors and Interaction with PG Receptors Earlier studies on iPs were focused on vasoconstrictor and bronchoconstrictor activities of iPF2a-III on smooth muscle and endothelial cells. Generally, iPF2a-III acts on smooth muscle via
153
TXA2a-selective prostanoid receptors. Fukunaga et al (1993a) reported that iPF2a-III displaces the binding of TP-receptor-acting ligands with much less potency but stimulates IP3 production and 3 H thymidine incorporation with higher potency than TP agonists. Binding experiments indicated the presence of both low-affinity and high-affinity binding sites for iPF2a-III. Studies were also performed to determine the biological activity of iPF2aIII on platelets. In platelets, iPF2a-III is only a partial agonist for the TP receptors (Morrow et al 1992; Longmire et al 1994; Yin et al 1994). In human blood, iPF2a-III has an antiaggregatory effect (Cranshaw et al 2001). These activities suggest the possibility of another putative receptor for iPF2a-III that is distinct from the TP receptor. So far, the existence of a new receptor for iPF2a-III has not yet been confirmed. In isolated guinea-pig hearts, iPF2a-III and iPE2-III are potent vasoconstrictors and their actions appear to be mediated by TXA2 receptors. 8,12-iso-iPF2a-III (previously also called 12-iso-PGF2a) is a potent agonist for the FP receptor (Kunapuli et al 1997). Recent studies on the E-2 series iPs demonstrated that iPE2-III (15-F2tiso-P-E2) exerts its biological activity via interaction with both TP and EP3 receptor stimulation (Sametz et al 2000; Janssen and Tazzeo 2002). In contrast to iPF2a-III, the type VI iPs, e.g. iPF2aVI, possess no vasomotor effect (Marliere et al 2002) and as such are unlikely to be involved in the pathogenesis of vascular diseases. It could be used as a clinical marker of lipid peroxidation in vascular diseases, since elevated levels of iPF2a-VI have been described (Lawson et al 1998; Pratico et al 1998a). Biomarkers to Measure Oxidative Stress The concentration of iPs in plasma and urine of normal humans can be several orders of magnitude higher than COX-derived prostaglandins, and are further elevated during oxidative stress. Considerably increased generation of iPs has been reported in a variety of disease states and pathological conditions associated with oxidant stress. These include atherosclerosis, Alzheimer’s disease (Pratico 2002), lung disease (Janssen 2001), chronic obstructive pulmonary disease (Pratico et al 1998b) and diabetes (Davi et al 1999; Sampson et al 2002). Measured levels of iPs and their metabolites are increased in the plasma, urine, BALF and tissues of cigarette smokers (Morrow et al 1995; Reilly et al 1996) and patients with asthma (Dworski et al 2000) and cystic fibrosis (Wood et al 2003) during ventilated ischaemia (Becker et al 1998) during exposure to allergen (Montuschi et al 1999). Likewise, levels of iPs are elevated in cardiovascular conditions such as renal (Lerman et al 2001; Minuz et al 2002) and myocardial PTCA ischaemia-reperfusion injury (Iuliano et al 2001). Plasma levels of free and esterified F2-iPs were significantly higher in smokers than in non-smokers, and these levels of circulating iPF2a-III were decreased by 35% 2 weeks after cessation of smoking (Morrow et al 1995; Reilly et al 1996). Thus, smoking is associated with enhanced iP production, which in turn reflects that smoking causes oxidative damage. This may explain the causative link between smoking and a high risk for the development of chronic obstructive pulmonary disease, atherosclerosis and cancer. In general, sensitive, accurate and reliable methods have been developed for the measurement of iPs and their metabolites in plasma, urine and cerebrospinal fluid (CSF) and their increased levels have been correlated with the disease severity. They can be detected in exhaled breath condensates (Montuschi and Barnes 2002). There are several approaches that can be utilized to assess the endogenous production of iPs, each of which has certain advantages and disadvantages. Excellent reviews have appeared on the development of methods to measure iPs, and issues such as why iPs should be measured and which iPs should be measured
154
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
have been addressed (Rokach et al 1997b; Lawson et al 1999; Roberts and Morrow 2000). In this review, we will not attempt to discuss this in detail; as an example the measurement of iPs in atherosclerosis is discussed. Two distinct iPs, e.g. iPF2a-III and iPF2a-VI, were investigated for their potential as indices of oxidative stress in human atherosclerotic plaques obtained from patients undergoing carotid and arterectomy (Pratico et al 1997). Atherosclerotic plaques contained higher levels of both iPF2a-III and iPF2a-VI (previously called iPF2a-I) than the healthy vessels. The modification of iPF2aIII in atherosclerotic plaque was 2.25 pM/mM phospholipids. The corresponding median value in the normal vessel was 0.09 25 pM/ mM phospholipids. The levels of iPF2a-VI in atherosclerotic plaque were higher than those of iPF2a-III, with a median of 9.00 pM/mM phospholipids; the corresponding median in normal vessels was 0.28 pM/mM phospholipids. These observations support the hypothesis of measurement of iPs and may provide an accurate and sensitive index of oxidant stress in atherosclerotic diseases. Clinical Marker Isoprostanes as Clinical Markers for the Evaluation of Antioxidants Several research groups have shown that vitamin E (a-tocopherol) supplements reduce the risk of cardiovascular disease in humans (Rimm et al 1993; Stampfer et al 1993). a-Tocopherol is the principal lipid-soluble antioxidant in plasma and in LDL and is a free radical scavenger. Ascorbate (vitamin C) is a water-soluble free-radical scavenger that regenerates a-tocopherol from its chromanoyl radical form (Levine et al 1996; Berger et al 1997). Measurement of F2-iPs in biological fluids represents a valuable pharmacological tool for the evaluation of antioxidant therapy and determination of rational dose selection of antioxidants. In the limited clinical studies, vitamin E supplementation was found to reduce urinary levels of iPF2a-III in patients with type 2 diabetes (Davi et al 1999) and cystic fibrosis (Ciabattoni et al 2000). In addition, dose-dependent reduction in iP levels was observed in hypercholesterolaemia patients (Davi et al 1997). Clinical studies in patients with chronic alcohol-induced liver disease and hepatic cirrhosis demonstrated that vitamin E (or vitamin C) reduced urinary F2-isoprostane levels (Meagher et al 1999). These clinical studies clearly indicate that vitamin E has antioxidant effects in patients associated with a high rate of lipid peroxidation (Meagher et al 2001). A clinical study of patients 6 months after successful angioplasty of the stenotic renal arteries indicated that urinary iPF2a-III was significantly higher in patients with RVD than in patients with essential hypertention or in healthy subjects. These results reflect that lipid peroxidation is considerably enhanced in hypertensive patients with RVD and is related to the renin– angiotensin system. Persistent platelet activation triggered by iPF2a-III may contribute to cardiovascular and renal damage in this setting. These studies reflect a significant correlation between plasma renin activity, iPF2a-III and RVD, and support a link between the renin–angiotensin system and oxidative stress (Minuz et al 2002). CONCLUDING REMARKS In the past 10–15 years, research efforts in the field of eicosanoids have resulted in the discoveries of new biochemical pathways of eicosanoid mediators which have helped to further understand their biochemical origin and their receptor-mediated actions in
both physiological and pathophysiological conditions. These include: (a) the discovery of cyclooxygenase-2; (b) in the lipoxygenase field, new additional biochemical pathways for oxo-eicosanoids; (c) identification, characterization and cloning of LT receptors; and (d) the discovery of isoprostanes. The discovery of COX-2 was a scientific breakthrough and provided a new therapeutic target for the antiinflammatory action of NSAIDs, with reduced GI side-effects. Two antiinflammatory COX-2 inhibitors, celecoxib and rofecoxib, are marketed for the treatment of arthritis, dental pain and inflammatory diseases. A third COX-2 inhibitor, valdecoxib, has received the US FDA’s approval and is poised to enter the market soon. Pharmaceutical companies are aggressively pursuing the goal of introducing additional COX-2 inhibitors; etoricoxib and COX-189 are currently completing Phase III clinical trials. Several selective COX-2 inhibitors have been shown to possess anti-cancer properties. However, the precise role of COX-2 in cancer pathogenesis and the mechanism of anti-cancer properties of COX-2 inhibitors remain to be evaluated. Nonetheless, there is an overexpression of COX-2 in human cancers and the availability of the substrate arachidonic acid (AA) is significantly higher, due to the breakdown of the cellular membranes of the cancerous cells, which suggests that COX-2 may have a functional role in the early stages of the disease. Several approaches have been used in the discovery of LT modulators. These have resulted in new classes of compounds: (a) inhibitors of FLAP; (b) inhibitors of 5-LO; (c) peptido-LTRAs; (d) LTA4-hydrolase inhibitors; (e) LTB4 antagonists; (f) dual acting 5-LO/COX inhibitors; and (g) 5-LO/H1-RA. The 5-LO inhibitors have the potential to offer a broader therapeutic benefit in 5-LO-derived LT-mediated pathological conditions. It should be emphasized that the biological role of the newly discovered 5LO-derived AA metabolite, 5-oxo-ETE, to induce eosinophilia has only been recently elucidated and its relevance to the disease process in humans is not yet identified. The 5-LO inhibitor, zileuton, is available for asthmatic patients, but its weak potency and fast clearance are therapeutic drawbacks. A potent and long-acting 5-LO inhibitor is warranted in the future. The inhibition of LTA4-hydrolase blocks the formation of LTB4, but none of the LTA4-hydrolase inhibitors have yet progressed to a full clinical evaluation. FLAP inhibitors were evaluated but unfortunately, due to their side-effects, were discontinued. The problem with this approach is that FLAP is necessary for LT production when endogenous AA is the source of substrate. In the broken cellular assay, FLAP inhibitors did not prevent LT production. In the area of inflammation, one aspect that has not been studied in detail is that the neutrophils have less capacity to produce the substrate AA, but they have the capacity to uptake the AA liberated by other activated cells, such as platelets. Under inflammatory situations, such as LPS-stimulated whole blood, platelets release considerable amounts of AA, which has been confirmed by the overproduction of eicosanoids and isoeicosanoids by LPS-stimulated blood (McAdam et al 2000). Substrate specificity of PUFAs for 5-LO is another aspect that has received scant attention. At equimolar concentrations of AA and EPA, it has been reported that EPA is a better substrate than AA for 5-LO (Grimminger et al 2000). The early disappointments in the 5-LO inhibitor approach directed major efforts of pharmaceutical companies to the design of Cys-LTRAs to modulate LTD4-mediated airway function. The regulatory approval for three Cys-LTRAs, zafirlukast, montelukast and pranlukast, clearly demonstrated the value of these drugs for the treatment of asthma. Since the discovery of isoprostanes 10 years ago, there has been an exponential growth in the number of publications pertaining to isoprostanes. In the last 2 years the literature has doubled. Initially, emphasis was particularly focused on the synthesis of iPs,
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY structural elucidation, evaluation of biological activity and development of analytical methods for their measurement as biomarkers in disease states. Now there is a growing interest in iPs as important mediators in cardiovascular and lung pathophysiology. In the iP area, further work is necessary to clarify whether new putative iP receptors are involved or whether there are interactions of iPs with different PG receptor subtypes. With the current therapies available, it should be possible to inhibit their production by using radical scavengers. Studies with vitamin supplementation have shown a proportional decrease in iP levels in pathophysiological conditions. The iPs have been used as clinical markers for the evaluation of antioxidants and may also prove to be useful as clinical markers in the evaluation of new drugs. The recent discovery of the BLT2 receptor should be advantageous in designing BLT2-specific receptor antagonists. Differences in human-BLT2 and mouse-BLT2 are now confirmed, hence BLT2 receptor antagonists need to be reevaluated and some nonselective BLT receptor antagonists should be considered for evaluation. One such compound, Bill 284, a pro-drug of the nonselective BLT receptor, is currently under clinical evaluation. The expression of Cys-LT2 receptors in pulmonary arteries is another area to exploit for specific Cys-LT2 antagonists. It is noteworthy that this receptor seems to be different from the CysLT2 receptors expressed in the lungs. Co-expression of PGE2 synthase with COX-2 in human rheumatoid synovial cells has been identified recently (Kojima et al 2002). It is now well known that PGE2 is the major PG metabolite produced by COX-2. Microsomal PGE2 synthase (m-PGEs) catalyses the final step of PGH2 to PGE2 synthesis. Previously, inducible human PGE2 synthase has been identified as a microsomal glutathione-dependent enzyme (Jakobsson et al 1999). The idea of the possibility of a functional coupling of COX-2 and PGE2 synthase has also been proposed. These discoveries reflect that PGE2 synthase could be a future potential target for antiinflammatory drugs. The complexities of the human chronic inflammatory process are continuously being appreciated and new findings are reported regularly. In the past decade there have been exciting discoveries in the clinical significance of the biochemical and molecular pharmacology of eicosanoids. Hopefully there will be even more interesting findings in the next decade. REFERENCES Adkins JC and Brogden RN (1998) Zafirlukast, A review of its pharmacology and therapeutic potential in the management of asthma. Drugs, 1998, 55. Aizawa T, Tamura G, Ohtsu H and Takishima T (1990) Eosinophil and neutrophil production of leukotriene C4 and B4: comparision of cells from asthmatic subjects and healthy donors. Ann Allergy, 64, 287–292. Allansmith MR, Baird RS, Grenier JV and Bloch KJ (1984) Late-phase reactions in ocular anaphylaxis in the rat. J Allergy Clin Immunol, 73, 49–55. Allgayer H and Stenson WF (1988) A comparision of effects of sulfasalazine and its metabolites on the metabolism of endogenous vs. exogenous arachidonic acid. Immunopharmacology, 15, 39–46. Altman LC, Munk Z, Seltzer J et al (1998) A placebo-controlled, doseranging study of montelukast, a once-daily leukotriene receptor antagonist. J Allergy Clin Immunol, 102, 50–56. Anderson CB and Mangino MJ (1991) Arachidonate 5-lipoxygenase inhibition and acute renal allograft rejection. Transpl Proc, 23, 640– 642. Ardaillou R (1990) Synthesis and effects of leukotrienes and other lipoxygenase products of arachidonic acid in the kidney. Eur J Pharmacol, 183, 142. Augstein J, Farmer JB, Lee TB et al (1973) Selective inhibitor of slow reacting substance of anaphylaxis. Synthesis of a series of chromone-2carboxylic acids. Nature (Lond) New Biol, 245, 215–217.
155
Banerjee M, Kang KH, Morrow JD et al (1992) Effect of a novel prostaglandin, 8-epi-PGF2a, in rabbit lung in situ. Am J Physiol Heart Circ Physiol, 263, H660–H663. Bayly CI, Black WC, Leger S et al (1999) Structure-based design of COX-2 selectivity into flurbiprofen. Bioorg Med Chem Lett, 9, 307–312. Becker PM, Sanders SP, Price P and Christman BW (1998) F2-isoprostane generation in isolated ferret lungs after oxidant injury or ventilated ischemia. Free Rad Biol Med, 25, 703–711. Bell RL, Harris RR, Malo PE et al (1997) ABT-761 Attenuates bronchoconstriction and pulmonary inflammation in rodents. J Pharmacol Exp Therapeut, 280, 1366–1373. Berger TM, Polidori MC, Dabbagh A et al (1997) Antioxidant activity of vitamin C in iron-overloaded human plasma. J Biol Chem, 272, 15656– 15660. Bernard GR, Korley V, Che P et al (1991) Persistent generation of peptido leukotrienes in patients with the adult respiratory distress syndrome. Am Rev Resp Dis, 144, 263–267. Bernstrom K and Hammarstrom S (1981) Metabolism of leukotriene D by porcine kidney. J Biol Chem, 256, 9579–9582. Bhagwat SS, Hamann PR, Still WC et al (1985) Synthesis and structure of the platelet aggregation factor thromboxane A2. Nature (Lond), 315, 511–513. Bhattacherjee P, Hammond B, Salmon JA and Eakins KE (1982) Adv Prostagland, Thromboxane Leukotriene Res, 9, 325–330. Bisgaars H (2001) Leukotriene modifiers in pediatric asthma management. Pediatrics, 107, 381–390. Black WC, Brideau C, Chan C et al (1999) 2,3-Diarylcyclopentenones as orally active, highly selective cyclooxygenase-2 inhibitors. J Med Chem, 42, 1274–1281. Bley KR, Hunter JC, Eglen RM and Smith JAM (1998) The role of IP prostanoid receptors in inflammatory pain. Trends Pharmacol Sci, 19, 141–147. Boie Y, Stocco R, Sawyer N et al (1999) Characterization of the cloned pig leukotriene B4 receptor: comparision to its human orthologue. Eur J Pharmacol, 380, 203–213. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med, 343, 1520–1528. Borgeat P and Samuelsson B (1979a) Arachidonic acid metabolism in polymorphonuclear leukocytes: unstable intermediate in formation of dihydroxy acids. Proc Natl Acad Sci USA, 76, 3213–3217. Borgeat P and Samuelsson B (1979b) Metabolism of arachidonic acid in polymorphonuclear leukocytes. J Biol Chem, 254, 7865–7869. Borgeat P and Samuelsson B (1979c) Transformation of arachidonic acid by rabbit polymorphonuclear leukocytes. J Biol Chem, 254, 2643–2646. Bossu E, Agliano AM, Desideri N et al (1999) LTB4 as marker of 5-LO inhibitor activity of two new N-o-ethoxycarbonyl-4-quinolones. J Pharm Biomed Anal, 19, 539–548. Brain SD, Camp RDR, Black AK et al (1985) Leukotrienes C4 and D4 in psoriatic skin lesions. Prostaglandins, 29, 611–619. Brain SD, Camp RDR, Leigh IM and Ford-Hutchinson AW (1982) The synthesis of leukotriene B4-like material by cultured human keratinocytes. J Invest Dermatol, 78, 328. Bray MA, Ford-Hutchinson AW and Smith MJH (1981) Leukotriene B4. An inflammatory mediator in vivo. Prostaglandins, 22, 213–222. Brooks CDW, Stewart AO, Basha A et al (1995) (R)-(+)-N-[5-(4fluorophenyl)methyl]-2-thienyl]-1-methyl-2-propynyl]-N-hydroxyurea (ABT-761), a second generation 5-lipoxygenase inhibitor. J Med Chem, 38, 4768–4775. Brooks CDW and Summers JB (1996) Modulators of leukotriene biosynthesis and receptor activation. J Med Chem, 39, 2629–2654. Bruynzeel PL, Kok PT, Hamelink ML et al (1985a) Exclusive leukotriene C4 synthesis by purified human eosinophils induced by opsonized zymosan. Fed Eur Biochem Soc, 189, 350–354. Bruynzeel PLB, Kok PTM, Vietor RJ and Verhagen J (1985b) On the optimal conditions of LTC4 formation by human eosiniphils in vitro. Prostagland Leukotrienes Med, 20, 11–22. Buckner CK, Krell RD, Laravuso RB et al (1986) Pharmacological evidence that human intralobar airways do not contain different receptors that mediate contractions to leukotriene C4 and leukotriene D4. J Pharmacol Exp Therapeut, 237, 558–562. Burke A, Lawson JA, Meagher EA and Rokach J (2000) Specific analysis in plasma and urine of 2,3-dinor-5,6-dihydro-isoprostane F2a-III, a
156
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
metabolite of isoprostane F2a-III and an oxidation product of glinolenic acid. J Biol Chem, 275, 2499–2504. Burke JA, Levi R, Guo Z and Corey EJ (1982) Leukotrienes C4, D4 and E4: effects on human and guinea-pig cardiac preparations in vitro. J Pharmacol Exp Therapeut, 221, 235–241. Cai X, Arrington M, Bayless L et al (2002) Discovery of UCB 35440: a potent orally active dual acting 5-lipoxygenase inhibitor and H1 receptor antagonist. In 224th National Meeting of the American Chemical Society, Division of Medicinal Chemistry, Abstract No. 317, Boston, August 14–22. Camp RDR, Coutts AA, Greaves MW et al (1983) Responses of human skin to intradermal injection of leukotrienes C4, D4, B4. Br J Pharmacol, 80, 497–502. Chan C-C, Bouce S, Brideau C et al (1999) Rofecoxib [Vioxx, MK-0966; 4-(4’-methylsulfonylphenyl)-3-phenyl-2-(5H)-furanone]: a potent and orally active cyclooxygenase-2 inhibitor. Pharmacological and biochemical profiles. J Pharmacol Exp Therapeut, 290, 551–560. Chan C-C and Ford-Hutchinson AW (1985) Effects of synthetic leukotrienes on local blood flow and vascular permeability in porcine skin. J Invest Dermatol, 84, 154–157. Chavis C, Chanez P, Vachier I et al (1995) 5,15-di-HETE and lipoxins generated by neutrophils from endogenous arachidonic acid as asthma biomarkers. Biochem Biophys Res Commun, 207, 273–279. Chiabrando C, Valagussa A, Rivalta C et al (1999) Identification and measurement of endogenous b-oxidation metabolites of 8-epiprostaglandin F2a. J Biol Chem, 274, 1313–1319. Ciabattoni G, Davi G, Collura M et al (2000) In vivo lipid peroxidation and platelet activation in homozygous homocystinuria. Am J Resp Crit Care Med, 162, 1195–1201. Cole AT and Hawkey CJ (1992) New treatments in inflammatory bowel disease. Br J Hosp Med, 47, 581–590. Coleman RA, Kennedy I, Humphery PPA et al (1990) Prostanoids and Their Receptors. Pergamon, Oxford, UK. Copeland RA, Williams JM, Giannaras J et al (1994) Mechanism of selective inhibition of the inducible isoform of prostaglandin G/H synthase. Proc Natl Acad Sci USA, 91, 11202–11206. Corey EJ, Clark DA, Goto G et al (1980) Stereospecific synthesis of a ‘‘slow-reacting substance’’ of anaphylaxis, leukotriene C1. J Am Chem Soc, 102, 1436–1439. Cracowski JL, Camus L, Durand T et al (2002) Response of rat thoracic aorta to F2-isoprostane metabolites. J Cardiovasc Pharmacol, 39, 396– 403. Crankshaw D (1995) Effects of the isoprostane 8-epi-prostaglandin F2a on the contractility of the human myometrium. Eur J Pharmacol, 285, 151–158. Cranshaw JH, Evans TW and Mitchell JA (2001) Characterization of the effects of isoprostanes on platelet aggregation in human whole blood. Br J Pharmacol, 132, 1699–1706. Dahlen S-E, Bjork J, Hedqvist P et al (1981) Leukotrienes promote plasma leakage and leukocyte adhesion in postcapillary venules: in vivo effects with relevance to the acute inflammatory response. Proc Natl Acad Sci USA, 78, 3887–3891. Dannhardt G and Kiefer W (2001) Cyclooxygenase inhibitors—current status and future prospects. Eur J Med Chem, 36, 109–126. Dannhardt G and Laufer S (2000) Structural approaches to explain the selectivity of COX-2 inhibitors: is there a common pharmacophore? Curr Med Chem, 7, 1101–1112. Davi G, Ciabattoni G, Consoli A et al (1999) In vivo formation of 8-isoprostaglandin F2a and platelet activation in diabetes mellitus: effects of improved metabolic control and vitamin E supplementation. Circulation, 99, 224–229. Davi G, Alessandrini P, Mezzetti A et al (1997) In vivo formation of 8-epiprostaglandin F2a is increased in hypercholesterolemia. Arterioscler Thromb Vasc Biol, 17, 3230–3235. Davidson AB, Lee TH, Scanlon PD et al (1987) Bronchoconstrictor effect of leukotriene E4 in normal and asthmatic subjects. Am Rev Respir Dis, 135, 333–337. Davies IW, Marcoux J-F, Corley EG et al (2000) A practical synthesis of a COX-2-specific inhibitor. J Org Chem, 65, 8415–8420. Davies IW, Taylor M, Marcoux J-F et al (2001) Preparation and novel reduction reactions of vinamidinium salts. J Org Chem, 66, 251–255. Davies P, Bailey PJ, Goldenberg MM and Ford-Huchinson AW (1984) The role of arachidonic acid oxygenation products in pain and inflammation. Ann Rev Immunol, 2, 335–357.
Davis KL, Dunn MW and Schwartzman ML (1990) Hormonal stimulation of 12(R)-HETE, a cytochrome P450 arachidonic acid metabolite in the rabbit cornea. Curr Eye Res, 9, 661–667. Delanty N, Reilly M, Lawson JA et al (1997) 8-epi-PGF2a generation during coronary reperfusion: a potential quantitative marker of oxidant stress in vivo. Circulation, 95, 2492–2499. de Leval X, Delarge J, Somers F et al (2000) Recent advances in inducible cyclooxygenase (COX-2) inhibition. Curr Med Chem, 7, 1041–1062. DeWitt DL, El-Harith EA, Kraemer SA et al (1990) J Biol Chem, 265, 5192–5198. Dubois A (2000) The role of prostaglandins and leukotrienes in the physiology and pathology of the digestive tract. Prostagland Lipid Mediat, Special Issue, 1–174. Dworski R, Roberts LJ, Murray JJ et al (2000) Assessment of oxidant stress in allergic asthma by measurement of the major urinary metabolite of F2-isoprostane 15-F2t-IsoP (8-iso-PGF2a) (abstr). Am J Resp Crit Care Med, 161, A331. Eberhart CE and Dubois RN (1995) Eicosanoids and the gastrointestinal tract. Gastroenterology, 109, 285–301. Edenius C, Heidvall K and Lindgren JA (1988) Novel transcellular interaction: conversion of granulocyte-derived leukotriene A4 to cysteinyl-containing leukotrienes by human platelets. Eur J Biochem, 178, 81. Ellis JL and Undem BJ (1994) Role of cysteinyl-leukotrienes and histamine in mediating intrinsic tone in isolated human bronchi. Am J Resp Crit Care Med, 149, 118–122. Elmhurst JL, Betti PA and Rangachari PK (1997) Intestinal effects of isoprostanes: evidence for the involvement of prostanoid EP and TP receptors. J Pharmacol Exp Therapeut, 283, 1198–1205. Evans JF (2002) Cysteinyl leukotriene receptors. Prostagland Lipid Mediators, 68–69, 587–597. Feldberg W and Kellaway CH (1938) Liberation of histamine and formation of lysocithin-like substances by cobra venom. J Physiol, 94, 187–197. Fiorucci S, de Lima J, Mencarelli A et al (2002) Cyclooxygenase-2-derived lipoxin A4 increases gastric resistance to aspirin-induced damage. Gastroenterology, 123, 1598–1606. FitzGerald GA and Patrono C (2001) The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med, 345, 433–442. Fitzsimmons BJ, Adams J, Evans JF et al (1985) The lipoxins: stereochemical identification and determination of their biosynthesis. J Biol Chem, 260, 13008–13012. Fleisch JH, Cloud ML and Marshall WS (1985) A brief review of preclincal and clinical studies with LY 171883 and some comments on newer cysteinyl leukotriene receptor antagonists. Ann NY Acad Sci, 524, 356–368. Folco GC, Omino C, Vigano T et al (1982) Biological activity of leukotriene C4 in guinea pigs: in vitro and in vivo studies. In Leukotriene and Other Lipoxygenase Products, Samuelsson B and Paoletti R (eds). Raven, New York, 153–158. Fontana L, Giafulli C, Cominacini L et al (2002) b2 integrin-dependent neutrophil adhesion induced by minimally modified low-density lipoproteins is mainly mediated by F2-isoprostanes. Circulation, 106, 2434–2441. Ford-Hutchinson AW, Bray MA, Shipley ME et al (1980) Leukotriene B, a potent chemokinetic and aggregating substance released from polymorphonuclear leukocytes. Nature, 286, 264–265. Ford-Hutchinson AW, Girad Y, Lord A et al (1989) The pharmacology of L-670,596, a potent and selective thromboxane/prostaglandin endoperoxide receptor antagonist. Can J Physiol Pharmacol, 67, 989– 993. Ford-Hutchinson AW, Rackham A, Zamboni R et al (1983) Comparative biological activities of synthetic leukotriene B4 and its o-oxidation products. Prostaglandins, 25, 29–37. Fradin A, Zirroli JA, Maclouf J et al (1989) Platelet-activating factor and leukotriene biosynthesis in whole blood. J Immunol, 143, 3680–3685. Fretland DJ and Djuric SW (1989) 12(R)- and 12(S)-Hydroxyeicosatetraenoic acids: chemistry, biology and pharmacology. Prostagland Leukotriene Essent Fatty Acids, 38, 215–228. Fu JY, Masferrer JL, Seibert K et al (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY Fukunaga M, Makita N, Roberts LJ II et al (1993a) Evidence for the existence of F2-isoprostane receptors on rat vascular smooth muscle cells. Am J Physiol, 264, C1619–1624. Fukunaga M, Takahashi K and Badr KF (1993b) Vascular smooth muscle actions and receptor interactions of 8-iso-prostaglandin E2, an E2isoprostane. Biochem Biophys Res Commun, 195, 507–515. Futaki N, Takahashi S, Yokoyama M et al (1994) NS-398, a new antiinflammatory agent, selectively inhibits prostaglandin G/H synthase/cyclooxygenase (COX-2) activity in vitro. Prostaglandins, 47, 55–59. Futaki N, Yoshikawa K, Hamasaka Y et al (1993) NS-398, a novel nonsteroidal antiinflammatory drug with potent analgesic and antipyretic effects, which causes minimal stomach lesions. Gen Pharmacol, 24, 105–110. Galbraith R (1992) DuP 697 as an agent which may be a selective cyclooxygenase inhibitor. In Winter Prostaglandin Conference, Keystone, CO, USA, January. Gans KR, Galbraith W, Romans RJ et al (1990) Antiinflammatory and safety profile of DuP 697, a novel orally effective prostaglandin synthesis inhibitor. J Pharmacol Exp Therapeut, 254, 180–187. Gardan B, Cracowski J, Sessa C et al (2000) Vasoconstriction effects of isoprostaglandin F2a type-III on human saphenous veins. J Cardiovasc Pharmacol, 35, 729–734. Gillard J, Ford-Hutchinson AW, Chan C et al (1989) L-663, 536 (MK886) (3-[1-(4-chlorobenzyl)-3t-butyl-thio-5-isopropylindol-2-yl]-2,2dimethylpropanoic acid), a novel, orally active leukotriene biosynthesis inhibitor. Can J Physiol Pharmacol, 67, 456–464. Gopaul NK, Anggard EE, Mallet AI et al (1995) Plasma 8-epi-PGF2a levels are elevated in individuals with non-insulin dependent diabetes mellitus. Fed Am Soc Exp Biol, 368, 225–229. Gorenne I, Ortiz JL, Labat C et al (1995) Antagonism of leukotriene responses in human airways by BAY x7195. Eur J Pharmacol, 275, 207–212. Grimminger F, Mayer K, Kiss L et al (2000) PAF-induced synthesis of tetraenoic and pentaenoic leukotrienes in the isolated rabbit lung. Am J Physiol Lung Cell Mol Physiol, 278, L268–275. Gyomber E, Vattay P, Szabo S and Rainsford KD (1996) Effect of lipoxygenase inhibitors and leukotriene antagonists on acute and chronic gastric haemorrhagic mucosal lesions in ulcer models in the rat. J Gastroenterol Hepatol, 11, 922–927. Hafstrom I, Palmblad J, Malmsten C et al (1981) Leukotriene B4—a stereospecific stimulator for release of lysosomal enzymes from neutrophils. Fed Eur Biochem Soc, 130, 14–17. Hall RA, Gillard J, Guindon Y et al (1987) Pharmacology of L-655,240 (3[1-(4-chlorobenzyl)-5-fluoro-3-methylindol-2-yl]-2,2-dimethyl-propanoic acid); a potent, selective thromboxane/prostaglandin endoperoxide antagonist. Eur J Pharmacol, 135, 193–201. Hamberg M and Samuelsson B (1973) Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Natl Acad Sci USA, 70, 899–903. Hand JM, Will JA and Buckner CK (1981) Effects of leukotrienes on isolated guinea pig pulmonary arteries. Eur J Pharmacol, 76, 439–442. Harris RC (1996) The macula densa: recent developments. J Hypertens, 14, 815–822. Hashimoto H, Imamura K, Haruta J and Wakitani K (2002) 4-(4Cycloalkyl/aryl-oxazol-5-yl)benzenesulfonamides as selective cyclooxygenase-2 inhibitors: enhancement of the selectivity by introduction of a fluorine atom and identification of a potent, highly selective, and orally active COX-2 inhibitor JTE-522. J Med Chem, 45, 1511–1517. Hecker M and Ullrich V (1989) On mechanism of prostacyclin and thromboxane A2 biosynthesis. J Biol Chem, 264, 141–150. Heise CE, O’Dowd BF, Figueroa DJ et al (2000) Characterization of the human cysteinyl leukotriene2 receptor. J Biol Chem, 275, 30531–30536. Herschman HR (1996) Prostaglandin synthase 2. Biochim Biophys Acta, 1299(1), 125–140. Herschman HR (1999) Function and regulation of prostaglandin synthase 2. Adv Exp Med Biol, 469, 3–8. Hirata M, Hayashi Y, Ushikubi F et al (1994) Cloning and expression of cDNA for a human thromboxane A2 receptor. Nature, 349, 617–620. Hoffmann C (2000) COX-2 in brain and spinal cord—implications for therapeutic use. Curr Med Chem, 7, 1113–1120. Holgate ST and Dahlen S-E (1996) From slow reacting substance to leukotrienes: a testimony of scientific endeavour and achievement. In
157
SRS-A to Leukotrienes: the Dawning of a New Treatment, Holgate S and Dahlen S-E (eds). Blackwell Science, Oxford, 1–23. Hou X, Roberts LJ II, Taber DF et al (2001) 2,3-Dinor-5,6-dihydro-15F2t-isoprostane: a bioactive prostanoid metabolites. Am J Physiol Regul Integr Comp Physiol, 281, R391–400. Hui Y, Yang G, Galczenski H et al (2001) The murine cysteinyl leukotriene 2 (CysLT2) receptor cDNA and genomic cloning, alternative splicing, and in vitro characterization. J Biol Chem, 276, 47489–47495. Israel E, Cohn J, Dube L and Drazen JM (1996) Effect of treatment with zileuton, a 5-lipoxygenase inhibitor, in patients with asthma. J Am Med Assoc, 275, 931–936. Iuliano L, Pratico D, Greco C et al (2001) Angioplasty increases coronary sinus F2-isoprostane formation: evidence for in vivo oxidative stress during PTCA. J Am Coll Cardiol, 37, 76–80. Jackson LM and Hawkey CJ (1999) Gastrointestinal effects of COX-2 inhibitors. Exp Opin Invest Drugs, 8, 963–971. Jakobsson P-J, Odlander B and Claesson H-E (1991) Effects of monocyte– lymphocyte interaction on the synthesis of leukotriene B4. Eur J Biochem, 196, 395–400. Jakobsson P-J, Thoren S, Morgenstren R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96, 7220–7225. Janssen LJ (2000) Isoprostanes: generation, pharmacology, roles in freeradical-mediated effects in the lung. Pulmon Pharmacol Therapeut, 13, 149–155. Janssen LJ (2001) Isoprostanes: an overview and putative roles in pulmonary pathophysiology. Am J Physiol Lung Cell Mol Physiol, 280, L1067–L1082. Janssen LJ, Premji M, Netherton S et al (2000) Excitatory and inhibitory actions of isoprostanes in human and canine airway smooth muscle. J Pharmacol Exp Therapeut, 295, 506–511. Janssen LJ and Tazzeo T (2002) Involvement of TP and EP3 receptors in vasoconstrictor responses to isoprostanes in pulmonary vasculature. J Pharmacol Exp Therapeut, 301, 1060–1066. John GW and Valentin JP (1997) Analysis of the pulmonary hypertensive effects of the isoprostane derivative, 8-iso-PGF2a in rat pulmonary artery. Br J Pharmacol, 122, 899–905. Jones RT, Davis C and Daniel EE (1982) Pharmacological study of the contractile activity of leukotriene C4 and D4 on isolated human airway smooth muscle. Can J Physiol Pharmacol, 60, 638–643. Jones TR, Labelle M, Belley M et al (1995) Pharmacology of Montelukast sodium (SingulairTM), a potent and selective leukotriene D4 receptor and antagonist. Can J Physiol Pharmacol, 73, 191–201. Jourdan K, Evans T, Curzen N and Mitchell J (1997) Evidence for a dilator function of 8-prostaglandin F2a in rat pulmonary artery. Br J Pharmacol, 120, 1280–1285. Kalgutkar AS, Marnett AB, Crews BC et al (2000) Ester and amide derivatives of the non-steroidal antiinflammatory drug, indomethacin, as selective cyclooxygenase-2 inhibitors. J Med Chem, 43, 2860–2870. Kalgutkar AS and Zhao Z (2001) Discovery and design of selective cyclooxygenase-2 inhibitors as non-ulcerogenic, antiinflammatory drugs with potential utility as anti-cancer agents. Curr Drug Targets, 79–106. Kamohara M, Takasaki J, Matsumoto M et al (2001) Functional characterization of cysteinyl leukotriene CysLT2 receptor on human coronary artery smooth muscle cells. Biochem Biophys Res Commun, 287, 1088–1092. Kang HK, Morrow JD, Roberts LJ II et al (1993) Airway and vascular effect of 8-epi-prostaglandin F2a in isolated perfused rat lung. J Appl Physiol, 74, 460–465. Kawikova I, Barnes P, Takahashi T et al (1996) 8-epi-PGF2a, a novel noncyclooxygenase-derived prostaglandin, constricts airways in vitro. Am J Resp Crit Care Med, 153, 590–596. Kellaway CH and Trethewie ER (1940) The liberation of slow reacting smooth muscle-stimulating substance in anaphylaxis. Qu J Exp Physiol, 38, 121–145. Khanapure SP, Garvey DS, Young DV et al (2003) Synthesis and structure-activity relationship (SAR) of novel, highly potent, metharyl and methcycloalkyl cyclooxygenase-2 (COX-2) selective inhibitors. J Med Chem 46, 5484–5504. Khanapure SP, Garvey DS, Young D et al (2002) Discovery, synthesis, and structure–activity profile of novel, highly potent, metharyl COX-2
158
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
selective inhibitors. In 224th American Chemical Society National Meeting, Division of Medicinal Chemistry, abstr 315, Boston, MA. Khanapure SP, Manna S, Rokach J et al (1995) 12-oxo-LTB4, a key pivotal intermediate in LTB4 metabolism. J Org Chem, 60, 1806–1813. Khanapure SP, Powell WS and Rokach J (1998a) The total synthesis of 5oxo-12-(S)-hydroxy-6(E),8(Z),10(E),14(Z)-eicostetraenoic acid and its 8,9-trans-isomer and their identification in human platelets. J Org Chem, 63, 8976–8982. Khanapure SP, Saha G, Powell WS and Rokach J (2000) The design and synthesis of a 5-HETE affinity chromatography ligand for 5hydroxyeicosanoid dehydrogenase. Tetrahed Lett, 41, 5807–5811. Khanapure SP, Shi XX, Powell WS and Rokach J (1998b) Total synthesis of a potent proinflammatory 5-oxo-ETE and its 6,7-dihydro biotransformation product. J Org Chem, 63, 337–342. Khanapure SP, Wang SS, Powell WS and Rokach J (1997) Synthesis of 10,11-dihydro-12-oxo-LTB4, a key biochemical intermediate. J Org Chem, 62, 325–330. Khanna IK, Huff YYRM, Weier RM et al (2000) Selective cyclooxygenase-2 inhibitors: heteroaryl modified 1,2-diarylimidazoles are potent, orally active antiinflammatory agents. J Med Chem, 43, 3168–3185. Khanna IK, Weier RM, Xiang YY et al (1997a) 1,2-Diarylimidazoles as potent, cyclooxygenase-2 selective, and orally active antiinflammatory agents. J Med Chem, 40, 1634–1647. Khanna IK, Weier RM, Yu Y et al (1997b) 1,2-Diarylpyrroles as potent and selective inhibitors of cyclooxygenase-2. J Med Chem, 40, 1619– 1633. Kiefer JR, Pawlitz JL, Moreland KT et al (2000) Structural insights into the stereochemistry of the cyclooxygenase reaction. Lett Nature, 405, 97–101. Klein T, Neuhaus K, Reutter F and Nusing RM (2001a) Generation of 8epi-prostaglandin F2a in isolated rat kidney glomeruli by a radicalindependent mechanism. Br J Pharmacol, 133, 643–650. Klein T, Neuhaus K, Reutter F and Nusing RM (2001b) Generation of 8epi-prostaglandin F2a in isolated rat kidney glomeruli by a radicalindependent mechanism. Br J Pharmacol, 133, 643–650. Kobayashi H, Takahashi Y, Mitsufuji H et al (1999) Decreased exhaled nitric oxide in mild persistent asthma patients treated with a leukotriene receptor antagonist, pranlukast. Jap J Physiol, 49, 541– 544. Kojima F, Naraba H, Sasaki Y et al (2002) Coexpression of microsomal prostaglandin E synthase with cyclooxygenase-2 human rheumatoid synovial cells. J Rheumatol, 29, 1836–1842. Kragballe K, Desjarlais L and Voorhees JJ (1985) Leukotrienes B4, C4 and D4 stimulate DNA synthesis in cultured human epidermal keratinocytes. Br J Dermatol, 113, 43–52. Kragballe K and Fallon JD (1986) Increased aggregation and arachidonic acid transformation by psoriatic platelets: evidence that plateletderived 12-hydroxy-eicosatetraenoic acid increases keratinocyte DNA synthesis in vitro. Arch Dermatol Res, 278, 449–453. Kragballe K, Voorhees JJ and Goetzl EJ (1987) J Invest Dermatol, 88, 555–558. Krell RD, Giles RE, Yee YK and Snyder DW (1987) In vivo pharmacology of ICI 198,615: a novel, potent and selective peptide leukotriene antagonist. J Pharmacol Exp Therapeut, 243, 557–564. Krell RD, Osborn R, Vickery L et al (1982) Contraction of isolated airway smooth muscle by synthetic leukotrienes C4 and D4. Prostaglandins, 22, 387–407. Kromer B and Tippins J (1996) Coronary artery constriction by the isoprostane 8-epi-prostaglandin F2a. Br J Pharmacol, 119, 1276–1280. Kromer B and Tippins J (1999) The vasoconstrictor effect of 8-epiprostaglandin F2a in the hypoxic rat heart. Br J Pharmacol, 126, 1171– 1174. Kromer BM and Tippins JR (1998) Actions of 8-epi-prostaglandin F2a on isolated rat aorta. J Cardiovasc Pharmacol, 32, 471–478. Kumlin M and Dahlen B (2000) The challenge procedure influences the extent of allergen-induced urinary excretion of leukotriene E4. Clin Exp Allergy, 30, 585–589. Kunapuli P, Lawson JA, Rokach J and FitzGerald GA (1997) Functional characterization of the ocular prostaglandin F2a (PGF2a) receptor. Activation by the isoprostane, 12-iso-PGF2a. J Biol Chem, 272, 27147– 27154.
Kurumbali RG, Stevens AM, Gierse JK et al (1996) Structural basis for selective inhibition of cyclooxygenase-2 by inflammatory agents. Nature, 384, 644–648. Labat C, Ortiz JL, Norel X et al (1992) A second cysteinyl leukotriene receptor in human lung. J Pharmacol Exp Therapeut, 263, 800–805. Labelle M, Belley M, Gareau Y et al (1995) Discovery of MK-0476, a potent and orally active leukotriene D4 receptor antagonist devoid of peroximal enzyme induction. Bioorg Med Chem Lett, 5, 283–288. Lahaie I, Hardy P, Hou X et al (1998) A novel mechanism for vasoconstrictor action of 8-isoprostaglandin F2a on retinal vessels. Am J Physiol Regulat Integr Comp Physiol, 274, R1406–R1416. Laneuville O, Breuer DK, DeWitt DL et al (1994) Differential inhibition of human prostaglandin endoperoxide-H synthases-1 and -2 by nonsteroidal antiinflammatory drugs. J Pharmacol Exp Therapeut, 271, 927–934. Larsson C, Weber P and Anggard E (1978) Arachidonic acid increases and indomethacin decreases plasma renin activity in the rabbit. Eur J Pharmacol, 28, 391–394. Lashner BA (1995) Clinical features, laboratory findings, and course of Crohn’s disease. In Inflammatory Bowel Disease, Kirsner JB and Shorter RG (eds). Williams and Williams, Baltimore, 344–354. Laufer SA, Augustin J, Dannhardt G and Kiefer W (1994) (6,7Diaryldihydropyrrolizin-5-yl)acetic acids, a novel class of potent dual inhibitors of both cyclooxygenase and 5-lipoxygenase. J Med Chem, 37, 1894–1898. Lawson JA, Li H, Rokach J et al (1998) Identification of two major F2 isoprostanes, 8,12-iso- and 5-epi-8,12-iso-isoprostane F2a-VI, human urine. J Biol Chem, 273, 29295–29301. Lawson JA, Rokach J and FitzGerald GA (1999) Isoprostanes: formation, analysis and use as indices of lipid peroxidation in vivo. J Biol Chem, 274, 24441–24444. Lecomte M, Laneuville O, Ji C et al (1994) Acetylation of human prostaglandin endoperoxide synthase-2 (cyclooxygenase-2) by aspirin. J Biol Chem, 269, 13207–13215. Lee TH, Horton CE, Kyan-Aung U et al (1989) Lipoxin A4 and lipoxin B4 inhibit chemotactic responses of human neutrophils stimulated by leukotriene B4 inhibit B4 and N-formyl-L-leucyl-L-phenylalanine. Clin Sci, 77, 195. Lee TH, Lympany P, Crea AE and Spur BW (1991) Inhibition of leukotriene B4-induced neutrophil migration by lipoxin A4: structure– function relationships. Biochem Biophys Res Commun, 180, 1416. Lefer AM (1986) Leukotrienes as mediators of ischemia and shock. Biochem Pharmacol, 35, 123–127. Lerman LO, Nath KA, Rodriguez-Porcel M et al (2001) Increased oxidative stress in experimental renovascular hypertension. Hypertension, 37, 541–546. Letts LG, Cirino M, Yusko P et al (1985) Actions of synthetic leukotrienes on platelets and blood vessels in the anesthetized pig: the release of a platelet-derived vasodilator. Prostaglandins, 29, 1049–1062. Letts LG, Palmer MA and Piper PJ (1981) The fate of leukotrienes C4 and D4 in the pulmonary and coronary circulation in vitro. Br J Pharmacol, 75, 5P. Letts LG and Piper PJ (1981) The effects of leukotrienes (LT) D4 and C4 on guinea-pig isolated heart. J Physiol, 317, 94P. Letts LG and Piper PJ (1982) The actions of leukotriene C4 and D4 on guinea-pig isolated hearts. Br J Pharmacol, 76, 169–176. Letts LG and Piper PJ (1983) Cardiac actions of leukotrienes B4, C4, D4 and E in guinea-pig and rat in vitro. Adv Prostagland Thromboxane Leukotriene Res, 11, 391–396. Levine M, Conry-Cantilene C, Wang Y et al (1996) Vitamin C pharmacokinetics in healthy volunteers: evidence for a recommended dietary allowance. Proc Natl Acad Sci USA, 93, 3704–3709. Lewis RA, Austen KF and Soberman RJ (1990) Leukotrienes and other products of the 5-lipoxygenase pathway. N Engl J Med, 323, 645–655. Li C-S, Black WC, Chan C-C et al (1995a) Cyclooxygenase-2 inhibitors. Synthesis and pharmacological activities of 5-methanesulfonamido-1indanone derivatives. J Med Chem, 38, 4897–4905. Li JJ, Anderson GD, Burton EG et al (1995b) 1,2-Diarylcyclopentenes as selective cyclooxygenase-2 inhibitors and orally active antiinflammatory agents. J Med Chem, 38, 4570–4578. Li JJ, Norton MB, Reinhard EJ et al (1996) Novel terphenyls as selective cyclooxygenase-2 inhibitors and orally active anti-inflammatory agents. J Med Chem, 39, 1846–1856.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY Lianos EA (1986) Eicosanoids and the modulation of glomerular immune injury. Kidney Int, 29, 339. Liu S, Hu H-Z, Gao C et al (2003) Actions of cysteinyl leukotrienes in the enteric nervous system of guinea-pig stomach and small intestine. Eur J Pharmacol, 459, 27–39. Longmire AW et al (1994) Actions of the E2-isoprostane, 8-iso-PGE2, on the platelet thromboxane/endoperoxide receptor in humans and rats: additional evidence for the existence of a unique isoprostane receptor. Prostaglandins, 48, 247–256. Maclouf JA and Murphy RC (1988) Transcellular metabolism of neutrophil-derived leukotriene A4 by human platelets. A potential cellular source of leukotriene C4. J Biol Chem, 263, 174–181. Mais DE, Yoakim C, Guindon Y et al (1989) Photoaffinity labelling of the human platelet thromboxane A2/prostaglandin H2 receptor. Biochim Biophys Acta, 1012, 184–190. Mallat Z, Nakamura T, Ohan J et al (1999) The relationship of hydroxyeicosatetraenoic acids and F2-isoprostanes to plaque instability in human carotid atherosclerosis. J Clin Invest, 103, 421– 427. Marcus AJ, Broekman MJ, Safier LB et al (1982) Formation of leukotrienes and other hydroxy acids during platelet–neutrophil interactions in vitro. Biochem Biophys Res Commun, 109, 130. Marliere S, Cracowski J-C, Durand T et al (2002) The 5-series F2isoprostanes possess no vasomotor effects in the rat thoracic aorta, the internal mammary artery and the human saphenous vein. Br J Pharmacol, 135, 1276–1280. Marnett LJ, Rowlinson SW, Goodwin DC et al (1999) Arachidonic acid oxygenation by COX-1 and COX-2; mechanisms of catalysis and inhibition. J Biol Chem, 274, 22903–22906. Masferrer JL, Zweifel BS, Seibert K et al (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. Masuda A, Mais DE, Oatis JE Jr and Halushka PV (1991) Platelet and vascular thromboxane A2/prostaglandin H2 receptors: evidence for different subclasses in the rat. Biochem Pharmacol, 42, 537–544. Matsushita M, Masaki M, Yagi Y et al (1997) Pharmacological profile of JTE-522, a novel prostaglandin H synthase-2 inhibitor, in rats. Inflamm Res, 46, 461–466. McAdam BF, Mardini IA, Habib A et al (2000) Effects of regulated expression of human cyclooxygenase isoforms on eicosanoid and isoeicosanoid production in inflammation. J Clin Invest, 105, 1473– 1482. Meagher EA, Barry OP, Lawson JA et al (2001) Effects of vitamin E on lipid peroxidation in healthy persons. J Am Med Assoc, 285, 1178– 1182. Meagher EA, Lucey M, Bensinger S and FitzGerald GA (1996) Alcoholinduced formation of 8-epi-PGF2a, an index of oxidant stress in vivo. J Invest Med, 44, 223. Meagher EM, Barry OP, Burke A et al (1999) Alcohol-induced generation of lipid peroxidation products in humans. J Clin Invest, 104, 805–813. Medrano EE, Farooqui JZ, Boissy RE et al (1993) Chronic growth stimulation of human adult melanocytes by inflammatory mediators in vitro: implications for nevus formation and initial steps in melanocyte oncogenesis. Proc Natl Acad Sci USA, 90, 1790–1794. Mellor EA, Maekawa A, Austen KF and Boyce JA (2001) Cysteinyl leukotriene receptor 1 is also a pyrimidinergic receptor and is expressed by human mast cells. Proc Natl Acad Sci USA, 98, 9764–9769. Menegatti E, Roccatello D, Fadden K et al (1999) Gene expression of 5lipoxygenase and LTA4 hydrolase in renal tissue of nephrotic syndrome patients. Clin Exp Immunol, 116, 347–353. Merck & Co. (1999) Singulair, Manufacturer’s Prescribing Information. Whitehouse Station, NJ. Michoud E, Lecomte M, Lagarde M and Wiernsperger N (1998) In vivo effect of 8-epi-PGF2a on retinal circulation in diabetic and non-diabetic rats. Prostagland Leukotrienes Essent Fatty Acids, 59, 349–355. Minuz P, Patrignani P, Gaino S et al (2002) Increased oxidative stress and platelet activation in patients with hypertension and renovascular disease. Circulation, 106, 2800–2805. Misson J, Clark W and Kendall MJ (1999) Therapeutic advances: leukotriene antagonists for the treatment of asthma. J Clin Pharm Therapeut, 24, 17–24. Mitchell JA, Akarasereenont P, Thiemermann C et al (1994) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697.
159
Mobert J and Becker BF (1998) Cyclooxygenase inhibition aggravates ischemia-reperfusion injury in the perfused guinea-pig heart: involvement of isoprostanes. J Am Coll Cardiol, 31, 1687–1694. Mobert J, Becker BF, Zahler S and Gerlach E (1997) Hemodynamic effects of isoprostanes (8-iso-prostaglandin F2a and E2) in isolated guinea-pig hearts. J Cardiovasc Pharmacol, 29, 789–794. Mohler ER, Franklin MT and Adam LP (1996) Intracellular signaling by 8-epi-prostaglandin F2a is mediated by thromboxane A2/prostaglandin endoperoxide receptors in porcine carotid arteries. Biochem Biophys Res Commun, 225, 915–923. Moncada S, Gryglewski RJ, Bunting S and Vane JR (1971) An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature, 231, 663– 665. Moncada S and Vane JR (1978) Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2 and prostacyclin. Pharmacol Rev, 30, 293. Montine TJ, Neely MD, Quinn JF et al (2002) Lipid peroxidation in aging brain and Alzheimer’s disease. Free Rad Biol Med, 33, 620–606. Montuschi P and Barnes PJ (2002) Analysis of exhaled breath condensate for monitoring airway inflammation. Trends Pharmacol Sci, 23, 232– 237. Montuschi P, Curro D, Ragazzoni E et al (1999) Anaphylaxis increases 8iso-prostaglandin F2a from guinea-pig lung in vitro. Eur J Pharmacol, 365, 59–64. Morelli JG, Kincannon J, Yohn JJ et al (1992) Leukotriene C4 and TNF-a are stimulators of human melanocyte migration in vitro. J Invest Dermatol, 98, 290–295. Morelli JG, Yohn JJ, Lyons MB et al (1989) Leukotrienes C4 and D4 as potent mitogens for cultured human neonatal melanocytes. J Invest Dermatol, 93, 719–722. Morrison AR, Nishikawa K and Needleman P (1978) Thromboxane A2 biosynthesis in the ureter obstructed isolated perfused kidney of the rabbit. J Pharmacol Exp Therapeut, 205, 1–8. Morrow JD, Frei B, Longmire AW et al (1995) Increase in circulating products of lipid perioxidation (F2-isoprostanes) in smokers. Smoking as a cause of oxidative damage. N Engl J Med, 332, 1198–1203. Morrow JD, Hill KE, Burk RF et al (1990) A series of prostaglandin F2like compounds are produced in vivo in humans by a noncyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci USA, 87, 9383–9387. Morrow JD, Minton TA and Roberts LJ II (1992) The F2-isoprostane, 8epi-prostaglandin F2a, a potent agonist of the vascular thromboxane/ endoperoxide receptor, is a platelet thromboxane/endoperoxide receptor antagonist. Prostaglandins, 44, 155–163. Muller-Peddinghaus R, Kohlsdorfer C, Theisen-Popp P et al (1993) BAY X1005, a new inhibitor of leukotriene synthesis: in vivo inflammation pharmacology and pharmacokinetics. J Pharmacol Exp Therapeut, 267, 51–57. Murata T, Ushikubi F, Matsuoka T et al (1997) Altered pain perception and inflammatory response in mice lacking prostacyclin receptor. Nature, 388, 678–682. Murphy RC, Hammarstrom S and Samuelsson B (1979) Leukotriene C: a slow-reacting substance from murine mastocytoma cells. Proc Natl Acad Sci USA, 76, 4275–4279. Nabe T, Kohno S, Tanpo T et al (1994) Inhibitory effect of ONO-1078 on specific binding of peptide leukotrienes to human lung crude membrane. Prostagland Leukotrienes Essent Fatty Acids, 51, 163–171. Nanji AA, Khwaja S and Sadrzadeh SMH (1994a) Eicosanoid production in experimental alcoholic liver disease is related to vitamin E levels and lipid perioxidation. Mol Cell Biochem, 140, 85–89. Nanji AA, Khwaja S, Tahan SR and Sadrzadeh SMH (1994b) Plasma levels of a novel noncyclooxygenase-derived prostanoid (8-isoprostane) correlate with severity of liver injury in experimental alcoholic liver disease. J Pharmacol Exp Therapeut, 269, 1280–1285. Nanji AA, Sadrzadeh SMH, Khettry U et al (1993) Protective effects of a noel quinone derivative, (2E)-3-[5-(2, 3-dimethoxy-6-methyl-1, 4 benzoquinoyl)]-2-nonyl-2-propanoic acid on experimental alcoholic liver injury. J Pharmacol Exp Therapeut, 266, 1085–1090. Narumiya S, Sugimoto Y and Ushikubi F (1999) Prostanoid receptors: structures, properties, and functions. Physiol Rev, 79, 1193–1226. Nishimura M, Schwartzman ML, Falck JR et al (1991) Metabolism of 12(R)-hydroxy-5,8,10,14-eicosatetraenoic acid (12(R)-HETE) in
160
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
corneal tissues: formation of novel metabolites. Arch Biochem Biophys, 290, 326–335. Noiri E, Yokomizo T, Nakao A et al (2000) An in vivo approach showing the chemotactic activity of leukotriene B4 in acute renal ischemicreperfusion injury. Proc Natl Acad Sci USA, 97, 823–828. Noonan MJ, Chervinsky P, Brandont M et al (1998) Montelukast, a potent leukotriene receptor antagonist, causes dose-related improvements in chronic asthma. Eur Resp J, 11, 1232–1239. Nothacker H-P, Wang Z, Zhu Y et al (2000) Molecular cloning and characterization of a second human cysteinyl leukotriene receptor: discovery of a subtype selective agonist. Mol Pharmacol, 58, 1601– 1608. Nugteren DH and Hazelhof E (1973) Isolation and properties of intermediates in prostaglandin biosynthesis. Biochim Biophys Acta, 236, 448–461. O’Flaherty JT, Cordes J, Redman J and Thomas MJ (1993) 5-oxoEicosatetraenoate, a potent human neutrophil stimulus. Biochem Biophys Res Commun, 192, 129–134. O’Flaherty JT, Kuroki M, Nixon AB et al (1996) 5-oxo-Eicosanoids and hematopoietic cytokines cooperate in stimulating neutrophil function and the mitogen-activated protein kinase pathway. J Biol Chem, 271, 17821–17828. O’Hickey SP, Arm JP, Rees PJ et al (1988) The relative responsiveness to inhaled leukotriene E4, methacholine inhalation histamine in normal and asthmatic subjects. Eur Resp J, 1, 913–917. Ochi T and Goto T (2000) Anti-inflammatory effect of FR140423, a novel selective cyclooxygenase-2 inhibitor, in rat adjuvant arthritis without gastrointestinal side effects. J Pharmacokinet Pharmacol, 52, 553–560. Odlander B, Jakobsson PJ, Rosen A and Claesson HE (1988) Human B and T lymphocytes convert leukotriene A4 into leukotriene B4. Biochem Biophys Res Commun, 153, 203–208. Ogasawara H, Ishii S, Yokomizo T et al (2002) Characterization of mouse cysteinyl leukotriene receptors mCysLT1 and mCysLT2; differential pharmacological properties and tissue distribution. J Biol Chem, 277, 18763–18768. Ogletree ML, Harris DN, Schumacher WA et al (1993) Pharmacological profile of BMS 180,291: a potent, long-acting orally active thromboxane A2/prostaglandin endoperoxide receptor antagonist. J Pharmacol Exp Therapeut, 264, 570–578. Ogletree ML, O’Keefe EH, Durham SK et al (1992) Gastroprotective effects of thromboxane receptor antagonists. J Pharmacol Exp Therapeut, 263, 374–380. Ohshima N, Nagase H, Koshino T et al (2002) A functional study on CysLT(1) receptors in human eosinophils. Int Arch Allergy Appl Immunol, 129, 67–75. Okazawa A, Kawikova I, Cui Z-H et al (1997) 8-epi-PGF2a induces airflow obstruction and airway plasma exudation in vivo. Am J Resp Crit Care Med, 155, 436–441. Oliviera L, Stallwood N and Crankshaw D (2000) Effects of some isoprostanes on the human umbilical artery in vitro. Br J Pharmacol, 129, 509–514. Pace-Asciak CR, Klein J and Spielberg SP (1986) Metabolism of leukotriene A4 into C4 by human platelets. Biochem Biophys Acta, 68, 877. Patrono C (2001) Measurement of cyclooxygenase isozyme inhibition in humans: exploring the clinical relevance of biochemical selectivity. Rheumatology, 19, S45–S50. Patrono C, Patrignani P and Rodriguez LAG (2001) Cyclooxygenaseselective inhibition of prostanoid formation: transducing biochemical selectivity into clinical read-outs. J Clin Invest, 108, 7–13. Penning TD, Talley JJ, Bertenshaw SR et al (1997) Synthesis and biologicalevaluation of the 1,5-diarylpyrazole class of cyclooxygenase2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide (SC-58635, Celecoxib). J Med Chem, 40, 1347–1365. Phillips GD, Rafferty P and Holgate ST (1987) LY 171883 as an oral leukotriene D4 antagonist in non-asthmatic subjects. Thorax, 42, 723. Picot D, Loll PJ and Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature, 367, 243– 249. Piper PJ, Samhoun MN, Tippins JR et al (1981) Pharmacological studies on pure SRS-A, SRS and synthetic leukotriene C4 and D4. In SRS-A and Leukotrienes, Piper PJ (ed.). Chichester, Wiley, 81–101.
Portanova JP, Zhang Y, Anderson GD et al (1996) Selective neutralization of prostaglandin E2 blocks inflammation, hyperalgesia and interleukin 6 production in vivo. J Exp Med, 184, 883–891. Powell WS (1984) Properties of leukotriene B4 20-hydroxylase from polymorphonuclear leukocytes. J Biol Chem, 259, 3082–3089. Powell WS, Ahmed S, Gravel S and Rokach J (2001) Eotaxin and RANTES enhance 5-oxo-6,8,11,14-eicosatetraenoic acid-induced eosinophil chemotaxis. J Allerg Clin Immunol, 107, 272–278. Powell WS, Gravel S, Khanapure SP and Rokach J (1998) Biological inactivation of 5-oxo-6,8,11,14-eicosatetraenoic acid by human platelets. Blood (submitted for publication). Powell WS, Gravel S, MacLeod RJ et al (1993) Stimulation of human neutrophils by 5-oxo-6,8,11,14-eicosatetraenoic acid by a mechanism independent of the leukotriene B4 receptor. J Biol Chem, 268, 9280– 9286. Powell WS, Rokach J, Khanapure SP et al (1996) Effects of metabolites of leukotriene B4 on human neutrophil migration and cytosolic calcium levels. J Pharmacol Exp Therapeut, 276, 728–736. Prasit P, Wang Z, Brideau C et al (1999) The discovery of rofecoxib[MK 966, Vioxx, 4-(4’-methylsulfonylphenyl)3-phenyl-2(5H)-furanone], an orally active cyclooxygenase-2 inhibitor. Bioorg Med Chem Lett, 9, 1773–1778. Pratico D (2002) Alzheimer’s disease and oxygen radicals: new insights. Biochem Pharmacol, 63, 563–567. Pratico D, Barry OP, Lawson JA et al (1998a) IPF2a-I: an index of lipid peroxidation in humans. Proc Natl Acad Sci USA, 95, 3449–3454. Pratico D, Basili S, Vieri M et al (1998b) Chronic obstructive pulmonary disease is associated with an increase in urinary levels of isoprostane F2a-III, an index of oxidant stress. Am J Resp Crit Care Med, 158, 1709–1714. Pratico D, Iuliano L, Mauriello A et al (1997) Localization of distinct F2isoprostanes in human atherosclerotic lesions. J Clin Invest, 100, 2028– 2034, 2637. Pratico D, Lee VM-Y, Trojanowski JO et al (1998c) Increased F2isoprostanes in Alzheimer’s disease: evidence for enhanced lipid peroxidation in vivo. FASEB J, 12, 1777–1783. Pryor WA, Stanley JP and Blair E (1976) Autoxidation of polyunsaturated fatty acids: II. A suggested mechanism for the formation of TBAreactive materials from prostaglandin-like endoperoxides. Lipids, 11, 370–379. Puig C, Crespo MI, Godessart N et al (2000) Synthesis and biological evaluation of 3,4-diaryloxazolones: a new class of orally active cyclooxygenase-2 inhibitors. J Med Chem, 43, 214–223. Rae SA and Smith MJH (1981) The stimulation of lysosomal enzyme secretion from human polymorphonuclear leukocytes by leukotriene B. J Pharmacokinet Pharmacol, 33, 616–618. Rainsford KD (1987) The effects of 5-lipoxygenase inhibitors and leukotriene antagonists on the development of gastric lesions induced by non-steroidal anti-inflammatory drugs in mice. Agents Act, 21, 316– 319. Ravasi S, Capra V, Panigalli T et al (2002) Pharmacological differences among CysLT (1) receptor antagonists with respect to LTC(4) and LTD(4) in human lung parenchyma. Biochem Pharmacol, 63, 1537– 1546. Reid JJ (2001) ABT-761. Curr Opin Investig Drugs, 2, 68–71. Reilly M, Delanty N, Lawson JA and FitzGerald GA (1996) Modulation of oxidant stress in vivo in chronic cigarette smokers. Circulation, 94, 19–25. Reilly MP, Delanty N, Roy L et al (1997) Increased formation of the isoprostanes IPF2a-I and 8-epi-prostaglandin F2a in acute coronary angioplasty: evidence of oxidant stress during coronary reperfusion in humans. Circulation, 96, 3314–3320. Reitz DB, Li JL, Norton MB et al (1994) Selective cyclooxygenase inhibitors: novel 1,2-diarylcyclopentenes are potent and orally active COX-2 inhibitors. J Med Chem, 37, 3878–3881. Rimm EB, Stampfer MJ, Ascherio A et al (1993) Vitamin E consumption and the risk of coronary heart disease in men. N Engl J Med, 328, 1450–1456. Roberts II LJ, Moore KP, Zackert WE and Oates JA (1996) Identification of the major urinary metabolite of the F2-isoprostane 8-isoprostaglandin F2a in humans. J Biol Chem, 271, 20617–20620. Roberts LJ and Morrow JD (2000) Measurement of F2-isoprostanes as an index of oxidative stress in vivo. Free Rad Biol Med, 28, 505–513.
CLINICAL SIGNIFICANCE OF EICOSANOID PHARMACOLOGY Rodrigues CR, Veloso MP, Verli H et al (2002) Selective PGHS-2 inhibitors: a rational approach for the treatment of the inflammation. Curr Med Chem, 9, 849–867. Rogers C, Pihan G, Raza A and Szabo S (1987) Possible participation of leukotrienes in the pathogenesis of gastric mucosal lesions and vascular injury induced by ethanol, HCl, NaOH, or aspirin in the rat. Effects of the eicosapentaenoic acid or a lipoxygenase inhibitor, L-651,392. Gastroenterology, 92, 1808–1817. Rojers TJ (1989) Immunopharmacology of regulatory cells. In Immunopharmacology, Gilman SC and Rogers TJ (eds). Telford, Caldwell, NJ, 409–463. Rokach J (1989) Leukotrienes and Lipoxygenases, Chemical, Biological and Clinical Aspects. Elsevier Science, Amsterdam. Rokach J, Khanapure SP, Hwang SW et al (1997a) Nomenclature of isoprostanes: a proposal. Prostaglandins, 54, 853–873. Rokach J, Khanapure SP, Hwang SW et al (1997b) The isoprostanes: a perspective. Prostaglandins, 54, 823–851. Rokach J, Khanapure SP, Hwang SW et al (1998) Isoprostanes, a new class of natural products: synthesis and biosynthesis. Synthesis, 569– 580. Roque M, Fallon JT, Badimon JJ et al (2000) Mouse model of femoral artery denudation injury associated with the rapid accumulation of adhesion molecules on the luminal surface and recruitment of neutrophils. Arterioscler Thromb Vasc Biol, 20, 335–342. Rordorf C, Kellett N, Mair S et al (2003) Gastroduodenal tolerability of lumiracoxib vs. placebo and naproxen: a pilot endoscopic study in healthy male subjects. Aliment Pharmacol Ther, 18, 533–554. Rouzer CA, Ford-Hutchinson AW, Morton HE and Gillard JW (1990) MK886, a potent and specific leukotriene biosynthesis inhibitor blocks and reverses the membrane association of 5-lipoxygenase in ionophorechallenged leukocytes. J Biol Chem, 265, 1436–1442. Rouzer CA and Kargman S (1988) Translocation of 5-lipoxygenase to the membrane in human leukocytes challenged with ionophore A23187. J Biol Chem, 263, 10980–10988. Ryn JV, Trummlitz G and Pairet M (2000) COX-2 selectivity and inflammatory processes. Curr Med Chem, 7, 1145–1161. Sakata K and Back M (2002) Receptor preferences of cysteinylleukotrienes in the guinea-pig lung parenchyma. Eur J Pharmacol, 436, 119–126. Sametz W, Hennerbichler S, Glaser S et al (2000) Characterization of prostanoid receptor mediating actions of the isoprostanes, 8-isoPGE(2) and 8-iso-PGF(2a). in some isolated smooth muscle preparations. Br J Pharmacol, 130, 1903–1910. Sametz W, Prasthofer S, Wintersteiger R and Juan H (1999) Vascular effects of isoprostanes after endothelial damage. Prostagland Leukotrienes Essent Fatty Acids, 61, 369–372. Sampson AP, Spencer DA, Green CP et al (1990) Leukotrienes in the sputum and urine of cystic fibrosis children. Br J Clin Pharmacol, 30, 861–869. Sampson MJ, Gopaul N, Davies IR et al (2002) Plasma F2 isoprostanes: direct evidence of increased free radical damage during acute hyperglycemia in type 2 diabetes. Diabetes Care, 25, 537–541. Samuelsson B (1980) The leukotrienes: a new group of biologically active compounds including SRS-A. Trends Pharmacol Sci, 1, 227–230. Samuelsson B, Borgeat P, Hammarstrom S and Murphy RC (1980) A new group of biologically active compounds. Adv Prostagland Thromboxane Res, 6, 1–8. Samuelsson B and Funk CD (1989) Enzymes involved in the biosynthesis of leukotriene B4. J Biol Chem, 264, 19469–19472. Saussy DL, Mais DE, Knapp DR and Halushka PV (1985) Thromboxane A2 and prostaglandin endoperoxide receptors in platelets and vascular smooth muscle. Circulation, 72, 1202–1207. Schauer U, Eckhart A, Muller R et al (1989) Enhanced leukotriene C4 production by peripheral eosinophils. Int Arch Allerg Appl Immunol, 90, 201–206. Schneider A and Stahl RAK (1998) Cyclooxygenase-2 and the kidney: current status and potential perspectives. Nephrol Dial Transpl, 13, 10– 12. Serhan CN (1991) Compounds of the arachdonic acid signalling cascade: a brief update and hypothesis. Adv Rheumatol Inflamm, 1, 147–149. Serhan CN (1997) Lipoxins and novel aspirin-triggered 15-epi-lipoxins (ATL): a jungle of cell–cell interactions or a therapeutic opportunity? Prostaglandins, 53, 107–137.
161
Sharon P and Stenson WF (1984) Enhanced synthesis of leukotriene B4 by colonic mucosa in inflammatory bowel disease. Gastroenterology, 86, 453–460. Shinomiya K, Fukunaga M, Kiyomoto H et al (2002) A role of oxidative stress-generated eicosanoid in the progression of arteriosclerosis in type 2 diabetes mellitus model rats. Hypertens Res, 25, 91–98. Sibelius U, Grandel U, Buerke M et al (2003) Leukotriene-mediated coronary vasoconstriction and loss of myocardial contractility evoked by low doses of Escherichia coli hemolysin in perfused rat hearts. Crit Care Med, 31, 683–688. Silverstein FE, Faich G, Goldstein JL et al (2000) GI toxicity with celecoxib vs. NSAIDs for arthritis. J Am Med Assoc, 284, 1247–1255. Simmet T, Luck W, Winking M et al (1989) Formation of cysteinylleukotrienes by human intracranial tumors. In Leukotrienes and Prostanoids in Health and Disease, Zor U, Naor Z and Danon A (eds). Karger, Basel, 166–170. Simon DI, Chen Z, Seifert P et al (2000) Decreased neointimal formation in Mac-1 (7/7) mice reveals a role for inflammation in vascular repair after angiopasty. J Clin Invest, 105, 293–300. Sinzinger H, Oguogho A and Kaliman J (1997) Isoprostane 8-epiprostaglandin F2a is a potent contractor of human peripheral lymphatics. Lymphology, 30, 155–159. Smedegard G, Hedqvist P, Dahlen S et al (1982) Leukotriene C4 affects pulmonary and cardiovascular dynamics in monkeys. Nature, 295, 327–329. Smith CJ, Zhang Y, Koboldt CM et al (1998) Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc Natl Acad Sci USA, 95, 13313–13318. Song Y, Connor DT, Doubleday R et al (1999a) Synthesis, structure– activity relationships, and in vivo evaluations of substituted di-tertbutylphenols as a novel class of potent, selective, and orally active cyclooxygenase-2 inhibitors. 1. Thiazolone and oxazolone series. J Med Chem, 42, 1151–1160. Song Y, Connor DT, Sercel AD et al (1999b) Synthesis, structure–activity relationships, and in vivo evaluations of substituted di-tertbutylphenols as a novel class of potent, selective, and orally active cyclooxygenase-2 inhibitors. 2. 1,3,4- and 1,2,4-thiadiazole series. J Med Chem, 42, 1161–1169. Soter NA, Lewis RA, Corey EJ and Austen KF (1983) Local effects of synthetic leukotrienes (LTC4, LTD4, LTE4, and LTB4) in human skin. J Invest Dermatol, 80, 115–119. Spurney RF, Klotman PE and Coffman TM (1990) Leukotriene (LT) production by rejecting rat kidney allografts correlates with renal dysfunction. Clin Res, 38, 23A. Stamatiou P, Hamid Q, Taha R et al (1998) 5-oxo-6,8,11,14Eicosatetraenoic acid stimulates eosinophil infiltration into the lungs of brown Norway rats in an integrin-dependent manner. J Clin Invest, 102, 2165–2172. Stampfer MJ, Hennekens CH, Manson JE et al (1993) Vitamin E consumption and the risk of coronary disease in women. N Engl J Med, 328, 1444–1449. Stjernschantz J, Sherk T, Borgeat P and Sears M (1984) Intraocular effects of lipoxygenase pathway products in arachidonic acid metabolism. Acta Ophthalmol (Copenh), 62, 104–111. Surette ME, Palmantier R, Gosselin J and Borgeat P (1993) Lipopolysaccharides prime whole human blood and isolated neutrophils for the increased synthesis of 5-lipoxygenase products by enhancing arachidonic acid availability: involvement of the CD14 antigen. J Exp Med, 178, 1347–1355. Taber DF, Morrow JD and Roberts II LJ (1997) A nomenclature system for the isoprostanes. Prostaglandins, 53, 63–67. Takahashi K, Nammour TM, Fukunaga M et al (1992) Glomerular action of a free radical-generated novel prostaglandin, 8-epi-prostaglandin F2a in the rat. Evidence for interaction with thromboxane A2 receptors. J Clin Invest, 90, 136–141. Talley JA, Brown DL, Carter JS et al (2000) 4-[5-Methyl-3phenylisoxazol-4-yl]-benzenesulfonamide, valdecoxib: a potent and selective inhibitor of COX-2. J Med Chem, 43, 775–777. Thornber CW (1979) Isosterism and molecular modification in drug design. Chem Soc Rev, 563–580. Thornton MA, Gruentzig AR, Hollman J et al (1984) Coumadin and aspirin in prevention of recurrence after transluminal coronary angioplasty: a randomized study. Circulation, 69, 721–727.
162
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Toda A, Yokomizo T and Shimizu T (2002) Leukotriene B4 receptors. Prostagland Lipid Mediat, 68–69, 575–585. Tomari S, Shimoda T, Kawano T et al (2001) Effects of pranlukast, a systeinyl leukotriene receptor 1 antagonist, combined with inhaled beclomethasone in patients with moderate or severe asthma. Ann Allerg Asthma Immunol, 87, 156–161. Tudhope SR, Curthbert NJ, Abram TS et al (1994) Bay u9773, a novel antagonist of cysteinyl-leukotrienes with activity against two receptor subtypes. Eur J Pharmacol, 264, 317–323. Turini ME and DuBois RN (2002) Cyclooxygenase-2: a therapeutic target. Ann Rev Med, 53, 35–57. Ushikubi FM, Nakajima M, Hirata M et al (1989) Purification of the thromboxane A2/prostaglandin H2 receptor from human blood platelets. J Biol Chem, 264, 16496–16501. Vane JR (1971) Inhibition of prostaglandin biosynthesis as a mechanism of action for aspirin-like drugs. Nature, 231, 232–235. Vane JR and Botting RM (2001) Therapeutic Roles of Selective COX-2 Inhibitors. William Harvey Press, London. VanSchoor J, Joos GF, Kips JC et al (1997) The effect of ART-761, a novel 5-lipoxygenase inhibitor, on exercise- and adenosine-induced bronchoconstriction in asthmatic subjects. Am J Resp Crit Care Med, 155, 875–880. Verhgen J, Bruynzeel PLB, Koedam JA et al (1984) Fed Eur Biochem Soc, 168, 23–28. Votta B and Mong S (1990) Transition of affinity states for leukotriene B4 receptor in sheep lung membranes. J Pharmacol Exp Therapeut, 265, 841–847. Wainwright SL and Powell WS (1991) Mechanism for the formation of dihydro metabolites of 12-hydroxyeicosanoids. J Biol Chem, 266, 20899–20906. Walch L, Norel X, Back M et al (2002) Pharmacological evidence for a novel cysteinyl-leukotriene receptor subtype in human pulmonary artery smooth muscle. Br J Pharmacol, 137, 1339–1345. Wallace JL (1990) Lipid mediators of inflammation in gastric ulcer. Am J Physiol, 258, G1–11. Wallace JL and Keenan CM (1990) An orally active inhibitor of leukotriene synthesis accelerates healing in a rat model of colitis. Am J Physiol, 258, G527–G534. Wang S, Gustafson E, Pang L et al (2000) A novel hepatointestinal leukotriene B4 receptor. Cloning and functional charcterization. J Biol Chem, 275, 40686–40694. Wang SS, Rokach J, Powell WS et al (1994) Synthesis of 12-KETE and its 8,9-trans-isomer. Tetrahed Lett, 35, 4051–4054. Weinstein GD, Ross P, McCullough JL and Cotton A (1983) In Psoriasis: Cell Proliferation, Wright NA and Camplejohn RS (eds). Churchill Livingstone, London. Wheelan P, Zirrolli JA, Morelli JG and Murphy RC (1993) Metabolism of leukotriene B4 by cultured human keratinocytes. J Biol Chem, 268, 25439–25448. Whittle B, Moncada S and Vane J (1978) Formation of prostacyclin by the gastric mucosa and its action on gastric function. Prostaglandins, 15, 704–705.
Willis AL (1987a) CRC Handbook of Eicosanoids: Prostaglandins and Related Lipids, vol. 1, part A. CRC Press, Boca Raton, FL. Willis AL (1987b) CRC Handbook of Eicosanoids: Prostaglandins and Related Lipids, vol. 1, part B. CRC Press, Boca Raton, FL. Wong SL, Drajesk J, Chang MS et al (1998a) Dose-proportional pharmacokinetics of a new 5-lipoxygenase inhibitor, ABT-761, in healthy volunteers. Biopharm Drug Dispos, 19, 159–162. Wong SL, Kearns GL, Kemp JP et al (1998b) Pharmacokinetics of a novel 5-lipoxygenase inhibitor (ABT-761) in pediatric patients with asthma. Eur J Clin Pharmacol, 54, 715–719. Wood LG, Fitzgerald DA, Lee AK and Garg ML (2003) Improved antioxidant and fatty acid status of patients with cystic fibrosis after antioxidant supplementation is linked to improved lung function. Am J Clin Nutrit, 77, 150–159. Woodman OL and Dusting GJ (1982) Coronary vasoconstriction induced by leukotrienes in the anaesthetized dog. Eur J Pharmacol, 86, 125– 128. Woods JW, Coffey MJ, Brock TG and Singer P-G M I (1995) 5Lipoxygenase is located in the euchromatin of the nucleus in resting human alveolar macrophages and translocates to the nuclear envelope upon cell activation. J Clin Invest, 95, 2035–2046. Woodward DF and Ledgard SE (1985) Effect of LTD4 on conjunctival vasopermeability and blood–aqueous barrier integrity. Invest Opthalmol Vis Sci, 26, 481–485. Woollard PM (1986) Stereochemical difference between 12-hydroxy5,8,10,14-eicosatetraenoic acid in platelets and psoriatic lesions. Biochem Biophys Res Commun, 136, 169–176. Xiao G, Tsai A-L, Palmer G et al (1997) Analysis of hydroperoxideinduced tyrosyl radicals and lipoxygenase activity in aspirin-treated human prostaglandin H synthase-2. Biochemistry, 36, 1836–1845. Yin K, Halushka PV, Yan Y-T and Wong PY-K (1994) Antiaggregatory Activity of 8-epi-prostaglandin F2a and other F-series prostanoids and their binding to thromboxane A2/prostaglandin H2 receptors in human platelets. J Pharmacol Exp Therapeut, 270, 1192–1196. Yokomizo T, Kato K, Hagiya H et al (2001) Hydroxyeicosanoids bind to and activate the low-affinity leukotriene B4 receptor, BLT2. J Biol Chem, 276, 12454–12459. Yokomizo T, Kato K, Terawaki K et al (2000) A second leukotriene B4 receptor, BLT2. A new therapeutic target in inflammation and immunological disorders. J Exp Med, 192, 421–432. Yoshida S, Sakamoto H, Ishizaki Y et al (2000) Efficacy of leukotriene receptor antagonist in bronchial hyperresponsiveness and hypersensitivity to analgesic in aspirin-intolerant asthma. Clin Exp Allerg, 30, 64–70. Yuan JJ, Zhang D, Karim RB et al (2002) Disposition of a specific cyclooxygenase-2 inhibitor, valdecoxib, in human. Am Soc Pharmacol Exp Therapeut, 30, 1013–1021. Zampronio AR, Souza GEP, Silva AA et al (1994) Interleukin-8 induces fever by a prostaglandin-independent mechanism. Am J Physiol Regul Integr Comp Physiol, 266, R1670–R1674.
14 Eicosanoid Antagonists Kiyoshi Yasui and Akinori Arimura Discovery Research Laboratories, Shionogi & Co. Ltd, Osaka, Japan
Eicosanoids, i.e. prostaglandins and leukotrienes, display their various actions via selective receptors. In addition to naturally occurring eicosanoids, synthesized ligands (agonists and antagonists) against eicosanoid receptors have been developed and used for investigative and therapeutic purposes. Among the above two types of receptor ligands, this chapter will focus on receptor antagonists for eicosanoids, although there is another strategy for inhibition of eicosanoid actions—inhibition of eicosanoid biosynthesis. Eicosanoid receptors are seven-transmembrane Gprotein-coupled receptors and have been classified as DP, EP1, EP2, EP3, EP4, FP, IP, TP and CRTH2 for prostanoids, and BLT1, BLT2, Cys-LT1 and Cys-LT2 for leukotrienes. The compounds antagonizing eicosanoid functions at receptor sites can be valuable tools for not only classifying eicosanoid receptors but also characterizing their physiological and pathophysiological roles. Moreover, in cases where an eicosanoid plays a pathophysiological role in the induction of a disease, its antagonist can be useful for treatment. We will summarize recent advances in the development of eicosanoid antagonists.
Cys-LT1 and Cys-LT2 for leukotrienes. References concerning cloning, expression and characterization of eicosanoid receptors are summarized in Table 14.1. In humans, six isoforms of EP3 (Schmid et al 1995) and two isoforms of TP (Raychowdhury et al 1994) are also identified. Table 14.2 summarizes the distribution, function and antagonists of eicosanoid receptors. The chemical structures of the individual compounds are shown in Figures 14.1–14.4.
PROSTANOID RECEPTOR ANTAGONISTS DP Antagonists BW A868C
The first study concerning a prostaglandin inhibitor was reported in the 1950s (Ambache 1957, 1959). A substance, patulin, was used as an inhibitor of prostaglandins, but proved to be a nonselective compound (Eliasson 1958). Next, three different compounds, SC-19220, 7-oxo-13-prostynoic acid and polyphloretin, were reported to be selective and competitive antagonists of prostaglandins (Sanner 1969; Fried et al 1969; Eakins and Karim 1970). Augstein et al (1973) reported FPL 55712 to be the first selective inhibitor of the slow-reacting substance of anaphylaxis, SRS-A, a mixture of cysteinyl leukotrienes. In the 1980s, there was much development of potent, selective and competitive antagonists for thromboxane. For a review of the development of eicosanoid antagonists to the 1980s, see the review by Sanner (1988).
BW A868C has been reported to be a highly selective and competitive antagonist for DP. This compound was found to inhibit PGD2- and BW245C (a selective agonist for DP)-induced aggregation and cyclic AMP elevation in human platelets in vitro (Giles et al 1989; Trist et al 1989), and to suppress PGD2- and BW245C-induced decrease in systemic arterial blood pressure in rats by intravenous pretreatment (Hamid-Bloomfield and Whittle 1989). Although BW A868C exhibits partial agonistic action on recombinant human and mouse DP receptor (Boie et al 1995; Hirata et al 1994), it is widely accepted as a selective DP antagonist and is used in various experimental systems (Woodward et al 1996; Walch et al 1999; Norel et al 1999; Sharif et al 2000a, 2000b; Angeli et al 2001; Monneret et al 2001). The specific binding of [3H]PGD2 to recombinant human and mouse DP was suppressed by BW A868C with Ki values of 1.7 and 220 nM, respectively (Boie et al 1995; Kiriyama et al 1997). In addition, the affinity of BW245C for mouse DP was also relatively lower than that for human DP; the Ki values for human and mouse DP were 0.9 and 250 nM, respectively (Boie et al 1995; Kiriyama et al 1997). These findings suggest that there are species differences for DP.
CLASSIFICATION OF EICOSANOID RECEPTORS
S-5751
Eicosanoid receptors have been pharmacologically characterized by comparative studies regarding the potency of agonists and antagonists in various preparations from various species (Kennedy et al 1982; Coleman et al 1984, 1994, 1995). These receptors have been cloned in several species including human, mouse, rat and guinea-pig, and classified as DP, EP1, EP2, EP3, EP4, FP, TP and CRTH2 for prostanoids, and as BLT1, BLT2,
Recently, we have demonstrated S-5751 to be a novel, selective and orally active DP antagonist (Arimura et al 2001). S-5751 clearly suppressed both the specific binding of [3H]PGD2 to human platelet membrane fractions with a Ki value of 1.6 nM and PGD2-induced cyclic AMP accumulation in human washed platelets with an IC50 value of 0.9 nM. S-5751 also inhibited PGD2-induced conjunctival plasma exudation in vivo (ED50
HISTORICAL PERSPECTIVE
The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
164
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Table 14.1 References concerning cloning, expression and characterization of eicosanoid receptors Receptor Human
Mouse
Rat
DP EP1 EP2 EP3
Boie et al 1995 Funk et al 1993 Regan et al 1994 Adam et al 1994
Hirata et al 1994 Watabe et al 1993 Katsuyama et al 1995 Sugimoto et al 1992
EP4
Bastien et al 1994
FP IP TP CRTH2 BLT1
Abramovitz et al 1994 Boie et al 1994 Hirata et al 1991 Hirai et al 2001, Monneret et al 2001 Yokomizo et al 1997, Owman et al 1997
Honda et al 1993, Nishigaki et al 1995 Sugimoto et al 1994 Namba et al 1994 Namba et al 1992
Wright et al 1999 Okuda-Ashitaka et al 1996, Boie et al 1997 Boie et al 1997 Takeuchi et al 1993, Neuscha¨fer-Rube et al 1994, Boie et al 1997 Sando et al 1994, Boie et al 1997
BLT2 Cys-LT1
Yokomizo et al 2000, Kamohara et al 2000, Tryselius et al 2000 Lynch et al 1999, Sarau et al 1999
Cys-LT2
Heise et al 2000, Nothacker et al 2000
Huang et al 1998, Martin et al 1999
Toda et al 1999
ZK138357 ZK138357 was used as a putative DP antagonist in in vitro study (Chan et al 2000).
EP1 Antagonists SC-19220 The competitive antagonistic activity of SC-19220 against PGE2induced response was first studied on isolated guinea-pig ileum (Sanner 1969). This compound also inhibited PGE2-induced reactions in guinea-pig fundus, dog fundus and guinea-pig trachea (Farmer et al 1974; Lambley and Smith 1975; Kennedy et al 1982). SC-19220 competed for the specific binding of [3H]PGE2 to cloned human EP1 with the IC50 value of 6.7 mM (Funk et al 1993), but it was inactive to cloned human EP2 and EP4 (Bastien et al 1994; Davis and Sharif 2000), suggesting that this compound acts as a selective EP1 ligand. However, since SC-19220 had no affinity for mouse EP1 (Kiriyama et al 1997), these data may
Distribution, function and antagonists of eicosanoid receptors
Receptor Distribution (human)
Function
DP EP1
Retina, small intestine Kidney
cAMP: Ca2+:
EP2 EP3 EP4 FP IP TP
Uterus Kidney, uterus Small intestine, lung, kidney, thymus, uterus, brain Ovary, uterus, testis Kidney, lung, liver Platelet, placenta, lung
cAMP: cAMP; cAMP: Ca2+: cAMP: Ca2+:
CRTH2 BLT1
Th2 cell, Tc2 cell, eosinophil, basophil Peripheral blood leukocyte, spleen, thymus
Ca2+: Ca2+:, cAMP;
BLT2
Spleen, liver, ovary, peripheral blood leukocyte, heart, pancreas Spleen, lung, peripheral blood leukocyte
Ca2+:, cAMP;
Cys-LT1 Cys-LT2
Masuda et al 1999, Boie et al 1999
Martin et al 2001, Maekawa et al 2001 Hui et al 2001
value=0.099 mg/kg, p.o.), but it had no effect on histamine- and U-46619 (a selective TP agonist)-induced conjunctival plasma exudation and bronchoconstriction in guinea-pigs. Oral pretreatment of S-5751 dramatically inhibited various antigen-induced allergic responses, such as nasal blockage, eosinophil infiltration into nose and lung and conjunctival oedema, in guinea-pig allergic models, indicating that PGD2 plays a critical role in the development of allergic inflammation in guinea-pigs.
Table 14.2
Guinea-pig
Lung macrophage, airway smooth muscle, cardiac Purkinje cell, adrenal medulla cell, peripheral blood leukocyte, brain, placenta, lymph node
Ca2+: Ca2+:
Antagonists BW A868C, S-5751, ZK138357 SC-19220, SC-51089, SC-51322, ONO-8711, ONO-8713, ZM325802, AH6809 (DP/EP1/EP2) AH6809 (DP/EP1/EP2) — AH22921 (EP4/TP), AH23848 (EP4/TP), ONO-AE2-227 AL-8810, AL-3138 — SQ 29,548, GR32191, BM 13,505, BMS 180,291, S-145, AA-2414, BAY u3405, ICI 192,605, BM-531, BM-573, Z-335 BAY u3405 (TP/CRTH2) U-75302, ONO-4057 (BLT1/BLT2), LY293111, CP-105696, CP-195543 (BLT1/BLT2), SC-41930, SC-51146, LY223982, CGS 25019C, ZK158252 (BLT1/BLT2), BIIL 284 LY255283 ONO-1078, ICI 204,219, MK-476, MK-571, SK&F 104353, CGP 45715A, ICI 198,615, LY171883, CP-199,330, CP-199,331 YM158 (Cys-LT1/TP) BAY u9773 (Cys-LT1/Cys-LT2)
EICOSANOID ANTAGONISTS
Figure 14.1 Chemical structures of prostaglandin antagonists
165
166
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 14.2 Chemical structures of thromboxane A2 antagonists
reflect species differences in EP1 as well as in DP. The IL-1binduced febrile response in rats was clearly inhibited by intracerebroventricular injection of SC-19220 at a dose of 100 mg (Oka et al 1998), although the results obtained from experiments using four subtypes of EP-deficient mice suggested that the IL-1b-induced febrile response is mediated via EP3 (Ushikubi et al 1998). SC-51089 and SC-51322 SC-51089 and SC-51322 antagonize PGE2-induced contraction in guinea-pig ileum, a response thought to be mediated via EP1 (Hallinan et al 1996). Both compounds also exhibit EP1
antagonistic activity in a functional reporter gene assay for Gprotein-coupled receptors with pKB values of 8.8 for SC-51322 and 6.9 for SC-51089, respectively, reported by Durocher et al (2000). SC-51322 has higher affinity for cloned human EP1 than SC-51089; their Ki values are 13.8 and 1322 nM, respectively (Abramovitz et al 2000). SC-51089 and SC-51322, but not SC19220, were reported to inhibit the specific binding of [3H]PGE2 to cloned mouse EP1 with Ki values of 800 and 11 nM, respectively (Maruyama and Ohuchida 2000, published in Japanese). Intrathecal injection of SC-51089 (30–300 mg) dose-dependently suppressed the flinching behaviour evoked by formalin in rats (Malmberg et al 1994). Topical application of SC-51322 (4 mg) resulted in significant inhibition of allodynia induced in rats by the GABAA-receptor antagonist bicuxulline (Zhang et al 2001).
EICOSANOID ANTAGONISTS
167
Figure 14.3 Chemical structures of leukotriene B4 antagonists
ONO-8711 and ONO-8713 ONO-8711 and ONO-8713 are highly selective ligands for EP1; the Ki values for ONO-8717 are 0.6 and 1.7 nM for cloned human and mouse EP1, respectively, 67 nM for mouse EP3, 76 nM for human TP and more than 1000 nM for other receptors, such as
human IP and mouse DP, EP2, EP4 and FP (Watanabe et al 1999), the Ki values for ONO-8713 are 0.3 nM for both cloned human and mouse EP1 and more than 1000 nM for all other types of receptors (Watanabe et al 2000). Both compounds strongly inhibited PGE2-induced intracellular calcium rise in human and mouse EP1-expressing cells (Watanabe et al 1999, 2000),
168
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 14.4 Chemical structures of cysteinyl leukotriene antagonists
EICOSANOID ANTAGONISTS
169
indicating that they act as potent antagonists for EP1. Moreover, ONO-8711 and ONO-8713 are orally active, and administration of these compounds in diet dose-dependently reduced aberrant crypt foci induced in mice by azoxymethane, a colon carcinogen (Watanabe et al 1999, 2000). The chemopreventive effect of ONO8711 was also observed with the breast cancer model in rats (Kawamori et al 2001). Oral and subcutaneous administration of ONO-8711 resulted in significant suppression of rat models for neuropathic and postoperative pain, respectively (Kawahara et al 2001; Omote et al 2001).
Wise 1998; Noguchi et al 1999; Crider et al 2000; Crider and Sharif 2001; Jones and Chan 2001). Davis and Sharif (2000) demonstrated that AH22821 and AH23848 were weak inhibitors of [3H]PGE2 binding against cloned human EP4 with Ki values of 31800 and 2690 nM, respectively. In addition, according to a study that examined the selectivity of AH23848 against cloned human prostanoid receptors, this compound is most potent against TP (Ki value for TP=592 nM, Ki value for EP4=13727 nM) and has limited selectivity (Abramovitz et al 2000).
AH6809 (DP/EP1/EP2 Antagonist)
ONO-AE2-227
AH6809, as well as SC-19922, has been described as an EP1 antagonist, because it exhibited PGE2-antagonistic action in EP1sensitive preparations such as in guinea-pig ileum, guinea-pig fundus and dog fundus (Coleman et al 1987). AH6809 has been also shown to possess antagonistic action for the antiaggregatory activity of PGD2 (Keery and Lumley 1988). In addition, Woodward et al (1995) reported that AH6809 has affinity for recombinant human EP2 and behaves as an EP2 antagonist. Based on these results, AH6809 has been used as a DP/EP1/EP2 antagonist in recent studies (Reinheimer et al 1998; Spicuzza et al 1998; Pelletier et al 2001; Walch et al 2001). Indeed, against cloned human EP subtypes, AH6809 has low selectivity, showing similar affinity for DP, EP1 and EP2, as well as for EP3 (Abramovitz et al 2000). This compound showed no affinity for cloned mouse EP1, although it possesses weak affinity for mouse EP3 (Ki value=350 nM; Kiriyama et al 1997).
More recently, ONO-AE2-227 has been reported to be a potent, selective and orally active EP4 antagonist. This compound clearly inhibited the specific binding of [3H]PGE2 and PGE2-induced increase in cytosolic cyclic AMP levels in experiments using mouse EP4 expressing CHO cells, with a Ki value and median inhibitory concentration of 2.7 and 10 nM, respectively (Mutoh et al 2002). Although ONO-AE2-227 also had affinity for mouse EP3 with a Ki value of 21 nM, the Ki values for other mouse prostanoid receptors, DP, EP1, EP2, FP, IP and TP, were more than 1000 times higher than that for mouse EP4 (Mutoh et al 2002). ONOAE2-227 given in diet significantly reduced azoxymethaneinduced colon carcinogenesis (Mutoh et al 2002), as did ONO8711- and ONO-8713-selective EP1 antagonists. FP Antagonists AL-8810 and AL-3138
ZM325802 ZM325802 was used as an EP1 antagonist in in vitro study (Jenkins et al 2001). EP2 Antagonists As far as we know, there are no selective EP2 antagonists at present, although AH6809 (DP/EP1/EP2 antagonist) is available as a non-selective EP2 antagonist. EP3 Antagonists Although some compounds have been presented as EP3 antagonists at an international conference (Labelle et al 2000), there are no well-characterized EP3 antagonists available for biochemical and pharmacological studies. EP4 Antagonists
Recently, two PGF2a analogues, AL-8810 and AL-3138, were reported to be partial agonists of relatively low efficacy and functional antagonists at FP in A7r5 rat thoracic aorta smooth muscle cells and Swiss mouse 3T3 fibroblasts (Griffin et al 1999; Sharif et al 2000c). These cell lines express FP receptor coupled with phosphoinositide turnover and intracellular calcium mobilization (Griffin et al 1997, 1998). The AL-8810 and AL-3138 Ki values for phosphoinositide turnover induced by fluprostenol, a selective FP agonist, in A7r5 rat thoracic aorta smooth muscle cells were 426 and 296 nM, respectively (Griffin et al 1999; Sharif et al 2000c). Both compounds did not exhibit significant antagonistic action at DP, EP2, EP4 and TP in various cell lines. The affinities of AL-8810 and AL-3138 for cloned human, mouse or rat prostanoid receptors have not yet been elucidated. IP Antagonists The development of IP antagonists is delayed in comparison with that of other eicosanoid antagonists.
AH22921 and AH23848 (EP4/TP Antagonists)
TP Antagonists
The EP4 receptor was first discovered in piglet saphenous vein, in which PGE2 clearly induced smooth muscle relaxation, but selective agonists for EP1, EP2 and EP3 were all weak or inactive in inducing relaxation. This PGE2-induced response was concentration-dependently inhibited by AH22921 and AH23848, originally reported as TP antagonists (Boura et al 1986; Brittain et al 1985), with pA2 values of 5.3 and 5.4, respectively (Coleman et al 1994). From the above findings, both compounds have been used as weak but relatively selective EP4 antagonists to clarify the physiological and pathophysiological roles of EP4 (Ono et al 1998;
SQ 29,548 SQ 29,548 is one of the most useful ligands for studying TP, and its [3H]-labelled form is widely used as a standard radioligand for TP. This compound inhibited the specific binding of [3H]U-46619, a selective TP agonist, to unactivated, intact human platelets and U-46619-induced aggregation in human platelets with IC50 values of 7.9 and 50 nM, respectively (Ogletree et al 1985; Kattelman et al 1986). Intravenous injection of SQ 29,548 almost completely suppressed arachidonic acid-induced sudden death in rabbits and
170
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
U-46619-induced increase in lung resistance in cats at doses of 1 mg/kg and 0.5 mg/kg, respectively (Darius et al 1985; Underwood et al 1987). Although there are two isoforms of human TP, the platelet/placental receptor, termed TPa, and the endothelial receptor, TPb (Kinsella et al 1997), produced by different splicing events, SQ 29,548 has affinity and acts as an antagonist for both TPa and TPb (Raychowdhury et al 1994). SQ 29,548 has high affinity for cloned human TPa and mouse TP, with Ki values of 4.1 and 13 nM, and no cross-reactivity with other human prostanoid receptors (Kiriyama et al 1997; Abramovitz et al 2000). GR32191 (Vapiprost) GR32191, a close analogue of AH23848, potently suppressed U46619-induced platelet aggregation in human whole blood and human washed platelets, with pA2 values of 8.23 and 8.79, respectively (Lumley et al 1989). This compound also inhibited U46619-induced contraction of vascular and airway smooth muscle preparations from human, rat, guinea-pig, dog and rabbit with different potencies, yielding comparable pA2 values of 7.1–8.8 (Lumley et al 1989). The specific binding of [3H]SQ 29,548 to human platelets and guinea-pig platelet membranes was under concentration-dependent competition with GR32191 with IC50 values of 21.7 and 3.69 nM, respectively (Tanaka et al 1998). GR32191 antagonized U-46619-, PGD2- and PGF2a-induced contraction in guinea-pig airways but not the methacholineinduced one (Beasley et al 1989), indicating that these prostanoids cause airway smooth muscle contraction via TP, as previously suggested (Jones et al 1982; Coleman et al 1984). This finding was also confirmed by the result obtained from in vivo study in which oral pretreatment with GR32191 (80 mg) to asthmatic subjects significantly reduced PGD2-induced bronchoconstriction but not the methacholine-induced one (Beasley et al 1989). GR32191 could discriminate two subtypes of TP in human platelets; one linked to phospholipase C activation, mediating platelet aggregation, and the other linked to intracellular calcium elevation, resulting in platelet shape change (Takahara et al 1990). The affinity (Ki value) of GR32191 for cloned mouse TP was reported to be 12 nM (Kiriyama et al 1997). BM 13,505 (Daltroban) BM 13,505 antagonized the specific binding of [3H]SQ 29,548 to human platelets and guinea-pig platelet membranes with IC50 values of 86.0 and 293 nM, respectively (Tanaka et al 1998). This compound also suppressed [125I]IBOP, a TXA2 analogue, binding to guinea-pig lung membranes with an IC50 value of 184 nM (Saussy et al 1991). Platelet aggregation of humans and guineapigs caused by U-46619 was clearly inhibited by BM 13,505, with pIC50 values of 6.33 and 6.05, respectively (Tanaka et al 1998). BMS 180,291 (Ifetroban) U-46619-induced human platelet aggregation was inhibited by BMS 180,291, and the inhibitory effect of BMS 180,291 was timedependent. Briefly, 2.5, 30 and 120 min pre-incubation of BMS 180,291 with platelets before U-46619 stimulation led to progressively potent efficacy, with IC50 values of 21, 10 and 9 nM, respectively (Ogletree et al 1993). BMS 180,291 also suppressed U-46619- and PGD2-induced contraction in guinea-pig trachea and U-46619-induced contraction in rat aorta, with KB values of 0.1, 0.6 and 0.6 nM, respectively (Ogletree et al 1993). The Kd value of BMS 180,291 against the specific binding of [3H]SQ 29,548 to human platelet membranes was reported to be 4.03 nM
(Ogletree et al 1993). Gomoll et al (1994) demonstrated that bolus injection and subsequent infusion of BMS 180,291 significantly reduced myocardial infarction in dog and ferret models of myocardial ischaemia. S-145 (Domitroban) S-145 strongly inhibited the specific binding of [3H]U-46619 to human, rabbit and rat platelet membranes, with Ki values of 3.0, 1.0 and 1.8 nM, respectively (Hanasaki and Arita 1988). Its radiolabelled forms, [3H]S-145 and [125I]I-S-145-OH, were reported to be highly selective radioligands for platelet TP, with low non-specific binding and high binding affinity for various receptor preparations, such as washed human platelets, platelet membrane and solubilized TP (Ushikubi et al 1989). Using these radioligands, we have previously demonstrated the existence of TP in human and guinea-pig nasal mucosa (Arimura et al 1998). The specific binding of [3H]S-145 to cloned human and mouse TP was shown to be under concentration-dependent competition with S-145, with a Ki value of 0.68 for mouse TP (Hirata et al 1991; Kiriyama et al 1997). Contraction of isolated monkey and cat arteries, and guinea-pig lung parenchyma induced by U-46619, PGD2 and PGF2a, were significantly suppressed by S-145 and S1452 (a calcium salt of (+)-S-145) (Nakajima and Ueda 1989; Arimura et al 1992). Oral administration of S-1452 clearly inhibited U-46619-, PGD2-, PGF2a-, LTD4- and PAF-induced bronchoconstriction in guinea-pigs, with ED50 values of 0.006, 0.031, 0.112, 0.033 and 0.115 mg/kg, respectively (Arimura et al 1992). The superior inhibitory effect of S-1452 has been reported in animal models of various diseases, such as asthma (Arimura et al 1992, 1993, 1994a, 1994b), rhinitis (Yasui et al 1997, 2001), brain ischaemia (Matsuo et al 1993) and nephritis (Matsuo et al 1995). AA-2414 (Seratrodast, BRONICA) AA-2414 competitively antagonized the contractile response induced by U-46619 in guinea-pig trachea and lung parenchyma, and dog saphenous vein, with pA2 values of 7.69, 8.29 and 6.79, respectively (Ashida et al 1989). This compound also inhibited PGD2-, 9a,11b-PGF2- and PGF2a-induced contraction in guineapig trachea, with pA2 values of 7.20, 7.79 and 5.71, respectively (Ashida et al 1989). Oral pretreatment with AA-2414 significantly reduced U-46619-, LTD4- and PAF-induced bronchoconstriction in guinea-pigs (Ashida et al 1989). The specific binding of [3H]U46619 to guinea-pig washed platelets and cloned human TP was antagonized by AA-2414; the IC50 values were 8.2 and 60 nM, respectively (Imura et al 1990; Kurokawa et al 1994). The antiasthmatic and anti-rhinitic activities of AA-2414 have been reported in animals (Ashida et al 1991; Matsumoto et al 1994; Yamasaki et al 1997, 2001a, 2001b) and asthmatic subjects (Aizawa et al 1998; Hoshino et al 1999), and AA-2414 (BRONICA) has been available as an anti-asthmatic drug in Japan since 1995. BAY u3405 (Ramatroban, BAYNAS) BAY u3405 inhibited U-46619-induced vasoconstriction of rabbit aorta and platelet aggregation in human platelet rich plasma, with EC50 values of 0.38 and 0.46 mM, respectively (Rosentreter et al 1989). Airway smooth muscle contraction in human, guinea-pig, rat and ferret caused by U-46619 was concentration-dependently reduced by BAY u3405, with pA2 values between 8.0 and 8.9 (McKenniff et al 1991). This compound also suppressed
EICOSANOID ANTAGONISTS PGD2- and 9a,11b-PGF2-induced contraction in isolated guineapig trachea, but not carbachol, histamine, serotonin, LTC4 and LTD4 (McKenniff et al 1991). The antagonistic activity of BAY u3405 against U-46619-induced bronchoconstriction was observed when the drug was given to guinea-pigs by oral administration, intravenous injection and aerosol inhalation with the approximate ID50 values of 1.7 mg/kg, 600 mg/kg and 0.1% (w/v) 20 breaths, respectively (Francis et al 1991). The specific binding of [3H]SQ 29,548 to human platelet membranes was concentration-dependently displaced by BAY u3405, yielding the IC50 value of 68 nM (Darius et al 1992). The involvement of TXA2 in the pathogenesis of rhinitis (Narita et al 1996; Terada et al 1998), asthma (Nagai et al 1995; Aizawa et al 1996) and occlusive diseases (Perzborn et al 1990; Canale et al 1994) has been suggested, based on results obtained from experiments using BAY u3405 for animals and humans. BAY u3405 (BAYNAS) is now available for treatment of allergic rhinitis in Japan.
Other Compounds ICI 192,605 (Brewster et al 1988), BM-531 with TX synthase inhibitory activity (Dogne et al 2001), BM-573 with TX synthase inhibitory activity (Rolin et al 2001) and Z-335 (Tanaka et al 1998) etc. have also been reported to be TP antagonists.
CRTH2 Antagonist CRTH2, a seven-transmembrane G-protein-coupled receptor, is selectively expressed in T helper type 2 cells, T cytotoxic type 2 cells, eosinophils and basophils (Nagata et al 1999a, 1999b). Although its selective ligands were initially unknown, PGD2 was recently identified as a selective CRTH2 agonist (Hirai et al 2001; Monneret et al 2001). CRTH2 may be involved in PGD2-induced eosinophil infiltration and activation in vitro (Hirai et al 2001; Monneret et al 2001). Recently, Sugimoto et al (2003) reported that Bay u3405, a TP antagonist, also has an antagonizing activity against CRTH2. This was the first report concerning an orally active small molecule CRTH2 antagonist.
LEUKOTRIENE RECEPTOR ANTAGONISTS
171
U-75302 U-75302, a structural analogue of LTB4, competitively inhibited the specific binding of [3H]LTB4 to human neutrophils (Lin et al 1988). In addition, U-75302 antagonized LTB4-induced contractile response in isolated guinea-pig lung parenchyma, although this compound has a partial agonistic activity at a high concentration (Lawson et al 1989). Oral administration of U75302 (1, 10 and 30 mg/kg) dose-dependently reduced antigeninduced increase in the number of eosinophils in bronchoalveolar lavage fluid in actively sensitized guinea-pigs, whereas the compound did not affect neutrophil counts (Richards et al 1989). The specific binding of [3H]LTB4 to guinea-pig eosinophil membrane and LTB4-induced chemotactic response of guinea-pig eosinophils were significantly inhibited by U-75302, with a Ki value of 27.13 nM (binding) and an EC50 value of 11.5 mM (chemotaxis), respectively (Taylor et al 1991). Boie et al (1999) demonstrated that U-75302 had affinity to guinea-pig and human BLT, with Ki values of 886 and 444 nM, respectively. They also reported that U-75302 behaved as a full agonist of guinea-pig and human BLT, but not as an antagonist in intracellular calcium mobilization assay. ONO-4057 ONO-4057 competed for specific binding of [3H]LTB4 to human, guinea-pig and rat polymorphonuclear leukocytes, with Ki values of 6, 18 and 86 nM, respectively (Kishikawa et al 1990). LTB4induced calcium mobilization, aggregation, chemotaxis and degranulation of human neutrophils (Kishikawa et al 1992), and contraction of guinea-pig lung parenchyma (Kishikawa et al 1990) were also suppressed by ONO-4057, with IC50 values of 0.7, 3.0, 0.9, 1.6 and 0.7 mM, respectively. Oral administration of ONO4057 dose-dependently prevented LTB4-induced neutrophil accumulation in blood and skin with ED50 values of 25.6 and 5.3 mg/ kg, respectively (Kishikawa et al 1992). ONO-4057, administered orally (10, 30 and 100 mg/kg/day) or subcutaneously (30 mg/kg/ day), exerted an immunosuppressive effect on liver allograft in rats (Ii et al 1996; Tanaka et al 2000). ONO-4057 inhibited T cell proliferation caused by concanavalin A, anti-CD3 monoclonal antibody and IL-2, and cytokine production from activated T cells (Morita et al 1999). These findings suggest that LTB4 is involved in the immune response as well as the inflammatory response, and BLT antagonists have potential as immunosuppressive drugs.
BLT Antagonists
LY255283
At present, there are two LTB4 receptors, BLT1 and BLT2. The first, BLT1, is a high-affinity receptor preferentially expressed in leukocytes, and mediates LTB4-induced chemotaxis of granulocytes, adhesion of granulocytes to endothelial cells and degranulation of lysosomal enzymes (Yokomizo et al 1997). The other, BLT2, a low-affinity receptor, displays high homology to BLT1 and its open reading frame is located in the promoter region of BLT1 (Yokomizo et al 2000). BLT2 is more ubiquitously expressed than BLT1 (Yokomizo et al 2000). BLT2-expressing cells cause chemotaxis and calcium mobilization in response to LTB4. BLT2 may contribute to cellular functions in tissues other than leukocytes; however, its physiological and pathophysiological roles are still unclear. Using BLT1- and BLT2-expressing cells, Yokomizo et al (2000, 2001) divided some BLT antagonists into three groups: BLT1-selective, U-75302 and CP-105696; BLT1/ BLT2 dual antagonists, ONO-4057, CP-195543 and ZK158252; and BLT2-selective, LY255283.
LY255283 is an inhibitor of [3H]LTB4 binding to human neutrophils (IC50 value=10.5 mM) and of LTB4-induced aggregation of guinea-pig neutrophils (Jackson et al 1988). In guinea-pigs, this compound also suppressed both specific binding of [3H]LTB4 to lung membrane and LTB4-induced contraction of lung parenchyma, with a pKi of 7.0 and a pA2 value of 7.2, respectively (Silbaugh et al 1992). Intravenous and oral administration of LY255283 dose-dependently reduced airway obstruction in guinea-pigs induced by LTB4, but not by histamine and U46619, with ED50 values of 2.8 and 11.0 mg/kg, respectively (Silbaugh et al 1992). Antigen-induced lung eosinophilia in rats was inhibited by orally administered LY255283 (Richards et al 1991). LY255283 exhibited the ability to suppress the shock induced by endotoxin (Li et al 1991) and splanchnic artery occlusion in rats (Karasawa et al 1991). LY255283 (bolus i.v. and subsequent infusion) attenuated endotoxin-induced adult respiratory distress syndrome in the pig model (Wollert et al 1993).
172
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
LY293111
Cys-LT Antagonists
LY293111, a compound more potent in selectivity for inhibiting LTB4-induced responses and more active when administered orally than LY255283, inhibited the specific binding of [3H]LTB4 to human neutrophils, with the IC50 value of 17.6 nM (Jackson et al 1999). LTB4-induced human neutrophil aggregation, luminoldependent chemiluminescence, chemotaxis and superoxide production by adherent cells were also suppressed by LY293111 with IC50 values of 32, 20, 6.3 and 0.5 nM, respectively (Jackson et al 1999). LY293111, in guinea-pig lung, antagonized [3H]LTB4 binding (Ki value=7.1 nM) and LTB4-induced contraction in vitro (Jackson et al 1999). Intravenous injection or oral administration of LY293111 sodium dose-dependently inhibited LTB4induced airway obstruction in guinea-pigs with ED50 values of 14 mg/kg and 0.4 mg/kg, respectively (Silbaugh et al 2000). However, LY293111 sodium did not affect airway obstruction induced by histamine, LTD4 or U-46619. When administered orally, LY293111 sodium significantly inhibited LTB4- and calcium ionophore-induced granulocyte infiltration into guinea-pig lung (Silbaugh et al 2000). The inhibitory effect of orally administered LY293111 sodium on animal models for asthma and arthritis was also reported (Kuwabara et al 2000; Asanuma et al 2001).
Previous pharmacological studies have suggested that Cys-LTs, LTC4, LTD4 and LTE4 exhibit their actions through at least two distinct receptors (Snyder and Krell 1984; Gardiner et al 1990, 1993; Labat et al 1992). Recently, two receptors for Cys-LTs, designated Cys-LT1 and Cys-LT2, were cloned and characterized (Lynch et al 1999; Sarau et al 1999; Heise et al 2000; Nothacker et al 2000). The rank order potencies of Cys-LTs for cloned human receptors assessed by calcium mobilization assay are as follows: LTD44LTC44LTE4 for Cys-LT1, LTC4=LTD444LTE4 for Cys-LT2. As described below, there are now three types of antagonists known: Cys-LT1 selective (ONO-1078, ICI 204,219, MK-476 etc.), Cys-LT1/Cys-LT2 dual antagonist (BAY u9773) and Cys-LT1/TP dual antagonist (YM158). Cys-LT1 Antagonists ONO-1078 (Pranlukast, ONON)
CP-195543, a compound having relatively low plasma protein binding compared with CP-105696, inhibited the specific binding of [3H]LTB4 to human neutrophils and mouse spleen membranes with IC50 values of 6.8 and 37.0 nM, respectively (Showell et al 1998). This compound also suppressed LTB4-induced human and mouse neutrophil chemotaxis, and LTB4-induced CD11b upregulation in human neutrophils, with IC50 values of 2.4 and 7.5 nM, and a pA2 value of 7.66, respectively (Showell et al 1998). Oral administration of CP-195543 inhibited LTB4-induced neutrophil migration in guinea-pig and mouse skin, with ED50 values of 0.1 and 2.8 mg/ kg, respectively (Showell et al 1998). In addition, IL-1-accelerated collagen-induced arthritis in mice was prevented by CP-195543, administered by osmotic pump (Showell et al 1998).
ONO-1078 antagonized the specific binding of [3H]LTC4, [3H]LTD4 and [3H]LTE4 to guinea-pig lung membranes with Ki values of 5640, 0.99 and 0.63 nM, respectively (Obata et al 1992). LTC4- and LTD4-induced contractile responses in guinea-pig tracheal and lung parenchymal strips were inhibited by ONO1078 with the pA2 range of 7.70–10.71 (Obata et al 1992). In addition, in the presence of serine borate complex, an inhibitor of the conversion of LTC4 to LTD4, ONO-1078 also suppressed LTC4-induced contraction of guinea-pig trachea, with a pA2 value of 7.78 (Obata et al 1992). Intravenous and oral administration of ONO-1078 dose-dependently inhibited LTC4-, LTD4- and LTE4induced bronchoconstriction in guinea-pigs, with ED50 values of 3.8, 15.8 and 4.8 mg/kg (i.v.) and 0.59, 0.97 and 0.27 mg/kg (p.o.), respectively (Nakagawa et al 1992). ONO-1078 also suppressed LTD4-induced airway (trachea, main bronchi and intrapulmonary airways) and cutaneous microvascular permeability, with ED50 values (p.o.) of 0.74 (trachea), 0.40 (main bronchi), 0.69 (intrapulmonary airways) and 0.63 mg/kg (skin), respectively (Nakagawa et al 1992). ONO-1078 at an i.v. dose of 10 mg/kg did not affect bronchoconstriction caused by arachidonic acid, LTB4, PGD2, PGF2a, 9a,11b-PGF2, a stable TXA2 mimetic, STA2, PAF, histamine, serotonin and acetylcholine (Nakagawa et al 1992). The specific binding of [3H]LTD4 to cloned human and mouse Cys-LT1 concentration-dependently competed with ONO-1078, with an IC50 value of 4.4 nM and a Ki value of 0.25 nM, respectively (Sarau et al 1999; Martin et al 2001). LTD4-induced calcium mobilization in human Cys-LT1-expressing cells was also inhibited by ONO-1078, with an IC50 value of 0.1 nM (Sarau et al 1999). ONO-1078 exhibited only weak affinity for cloned human Cys-LT2 (Heise et al 2000). In the guinea-pig allergic model for asthma and rhinitis, ONO-1078 clearly exerted anti-allergic effects on various symptoms and inflammatory reactions (Nakagawa et al 1993; Fujita et al 1999). Antiasthmatic efficacy of ONO-1078 was also observed in patients with asthma (Fujimura et al 1993; Taniguchi et al 1993; Taki et al 1994; Yoshida et al 2000) and this compound was first launched in Japan for the treatment of adult asthma in 1995.
Other Compounds
ICI 204,219 (Zafirlukast, ACCOLATE)
There are several other BLT antagonists, as follows: SC-41930 (Tsai et al 1989), SC-51146 (Tsai et al 1994), LY223982 (Jackson et al 1992), CGS 25019C (Raychaudhuri et al 1995), ZK158252 (Matousek et al 2001) with BLT1/BLT2 dual antagonist activity, and BIIL 284 (Birke et al 2001).
ICI 204,219 competitively antagonized LTD4- and LTE4-induced contraction of guinea-pig trachea and lung parenchyma, with negative log molar KB values of 9.52 and 9.67 at a concentration of 3 nM (trachea), and 9.34 and 9.55 at a concentration of 3.3 nM (lung parenchyma), respectively (Krell et al 1990). This compound
CP-105696 CP-105696 antagonized [3H]LTB4 binding to human neutrophils with an IC50 value of 8.42 nM (Showell et al 1995). LTB4-induced human neutrophil chemotaxis and CD11b upregulation were also suppressed by CP-105696 with an IC50 value of 5.0 nM and a pA2 value of 8.03, respectively (Showell et al 1995). After oral administration, this compound showed an inhibitory effect on LTB4-induced dermal neutrophilia and eosinophilia in guineapigs with ED50 values of 0.3 and 0.3 mg/kg, respectively (Showell et al 1995). The contribution of LTB4 to the induction of experimental allergic asthma and encephalomyelitis was demonstrated using CP-105696 (Turner et al 1996, Gladue et al 1996). CP-105696 also ameliorated cardiac allograft rejection in mice and ischaemia/reperfusion injury in rats (Weringer et al 1999; Souza et al 2000). CP-195543
EICOSANOID ANTAGONISTS did not inhibit LTC4-induced contraction of guinea-pig trachea in the presence of a bioconversion inhibitor of LTC4 to LTD4, serine borate complex (Krell et al 1990). Binding of [3H]LTD4 and [3H]LTE4 to guinea-pig lung membranes was blocked by ICI 204,219, with Ki values of 0.34 and 0.23 nM, respectively (Krell et al 1990). Aerosolized LTD4-induced dyspnoea was inhibited by ICI 204,219 when administered orally and intravenously or by aerosol inhalation, with ED50 values of 0.52 mmol/kg (p.o.), 0.046 mmol/kg (i.v.) and 5.1 mM (aerosol), respectively (Krell et al 1990). Single oral pretreatment of ICI 204,219 at 2, 12 or 24 h before provocation significantly reduced LTD4-induced bronchoconstriction in normal subjects (Smith et al 1990). The affinity of ICI 204,219, for cloned human and mouse Cys-LT1 was reported to be 1.8 nM (IC50 value) and 0.72 nM (Ki value), respectively (Sarau et al 1999; Martin et al 2001). LTD4-induced calcium mobilization in cloned human Cys-LT1-expressing cells was also inhibited by ICI 204,219 with the IC50 value of 0.26 nM (Sarau et al 1999). This compound was inactive in the binding and calcium mobilization assay in cloned human Cys-LT2-expressing cells (Heise et al 2000; Nothacker et al 2000). ICI 204,219 has been used as an antiasthmatic drug worldwide, and Dunn and Goa (2001), in their review, have presented an update of zafirlukast’s efficacy against asthma.
MK-476 (Montelukast, SINGULAIR) MK-476 inhibited [3H]LTD4 binding to guinea-pig and sheep lung membranes, and dimethylsulphoxide-differentiated U937 cell plasma membranes, with Ki values of 0.18, 4 and 0.52 nM, respectively (Jones et al 1995). MK-476 antagonized the LTD4induced contractile response in isolated guinea-pig trachea, with a pA2 value of 9.3, but this compound had no effect on contraction of guinea-pig trachea induced by LTC4 (in the presence of serine borate), PGD2, histamine, serotonin or acetylcholine (Jones et al 1995). MK-476 clearly suppressed both LTD4-induced bronchoconstriction in anaesthetized guinea-pigs and antigen-induced respiratory distress in conscious rats, with ED50 values of 0.001 mg/kg (i.v.) and 0.032 mg/kg (p.o.), respectively (Jones et al 1995). Bronchoconstriction induced by antigen and LTD4 in conscious squirrel monkeys was also significantly inhibited by orally administered MK-476 (Jones et al 1995). In patients with asthma, a single oral administration of MK-476 4 h before provocation at doses of 5, 20, 100 and 250 mg reduced LTD4induced bronchoconstriction, and the inhibitory effect of MK-476 was observed even when administered 20 h before LTD4 challenge at a dose of 200 mg (De Lepeleire et al 1997). The affinity and potency of MK-476 against cloned human Cys-LT1 assessed by [3H]LTD4 binding assay and LTD4-induced calcium mobilization assay were as follows: the IC50 values for both assays were 4.9 and 2.3 nM, respectively (Sarau et al 1999). MK-476 was inactive in competing for the specific binding of [3H]LTD4 to cloned human Cys-LT2 (Heise et al 2000; Nothacker et al 2000). The usefulness of MK-476 in the treatment of asthma has been confirmed by clinical studies in adults and children. Jarvis and Markham (2000) recently summarized montelukast’s potential as an antiasthmatic drug.
Other Compounds MK-571 (Zamboni et al 1992; Lynch et al 1999; Sarau et al 1999; Maekawa et al 2001; Martin et al 2001), SK&F 104353 (pobilukast, Hay et al 1987; Sarau et al 1999), CGP 45715A (iralukast, Bray et al 1991), ICI 198,615 (Snyder et al 1987; Krell et al 1987, Bernstein 1998), LY171883 (Fleisch et al
173
1985), CP-199,330 and CP-199,331 (Chambers et al 1999) also possess the characteristics of being selective Cys-LT1 antagonists. Cys-LT1/Cys-LT2 Antagonist—BAY u9773 BAY u9773 clearly competed with the specific binding of [3H]LTD4 to guinea-pig lung membranes with a Ki value of 7.0 (Tudhope et al 1994). This compound antagonized LTC4- and LTD4-induced contraction of various tissue preparations, such as guinea-pig trachea, rat lung, ferret spleen and sheep bronchus, with pA2 values in the range 6.8–7.7. In contrast, the reference LT antagonists, ICI 198,615, MK-571 and SK&F 104353, potent inhibitors of LTD4-induced contraction of guinea-pig trachea and LTC4- and LTD4-induced contraction of rat lung, were weak or inactive for all other responses (Tudhope et al 1994). In addition, both LTC4- and LTD4-induced contractile responses were sensitive to BAY u9773 but insensitive to ICI 198,615, MK-571 and SK&F 104353 (Labat et al 1992). These results provide evidence that there are at least two isoforms of Cys-LT, and BAY u9773 is a dual Cys-LT1/Cys-LT2 antagonist. The dual antagonistic activity of BAY u9773 was also observed against cloned human Cys-LT1 and Cys-LT2; it inhibited LTD4-induced calcium mobilization in both cloned receptors, with IC50 values of 440 and 300 nM, respectively (Nothacker et al 2000). BAY u9773 had affinity for cloned Cys-LT2 with an IC50 value of 597 nM (Heise et al 2000). However, Nothacker et al (2000) suggest that BAY u9773 acts as a partial agonist for cloned human Cys-LT2. Cys-LT1/TP Antagonist—YM158 YM158 competitively antagonized [3H]LTD4 and [3H]U-46619 binding to guinea-pig lung membranes, with Ki values of 0.64 and 5.0 nM, respectively (Arakida et al 1998a). This compound showed no affinity for several other receptors, such as histamine, serotonin, bradykinin, neurokinin, PAF and so on (Arakida et al 1998a). Both LTD4- and U-46619-induced contraction of isolated guinea-pig trachea were suppressed by YM158 with pA2 values of 8.87 and 8.81, respectively (Arakida et al 1998b). YM158 also inhibited LTD4-induced contraction of guinea-pig iluem and U46619-induced aggregation of human and guinea-pig platelets, with IC50 values of 0.58, 290 and 530 nM, respectively (Arakida et al 1998b). Orally administered YM158 dose-dependently inhibited LTD4- and U-46619-induced increase in lung resistance and LTD4-induced skin reaction in guinea-pigs, with ED50 values of 8.6, 14 and 6.6 mg/kg, respectively (Arakida et al 2000b). These data suggest that YM158 is a potent Cys-LT1/TP dual antagonist. In guinea-pig models for asthma, YM158 administered orally exhibited a potent inhibitory effect (Arakida et al 1999, 2000a, 2000b, 2000c). CONCLUSION Since eicosanoid receptors have been cloned in succession, the development of selective eicosanoid antagonists has progressed. Particularly notable is the clinical efficacy of Cys-LT1 antagonists ONO-1078 (pranlukast, ONON), ICI 204,219 (zafirlukast, ACCOLATE) and MK-476 (motelukast, SINGULAIR) for the treatment of asthma. The antiallergic effect of DP antagonists, the chemopreventive effect of EP1 and EP4 antagonists, and the immunosuppressive effect of BLT1 antagonists are anticipated, as described above. Completion of the full line-up of eicosanoid antagonists should enable further classification of eicosanoid receptors and definition of the eicosanoid roles in physiological and pathophysiological conditions, probably resulting in discovery
174
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
of new candidates for drugs. Thus, continuous efforts are needed to develop highly potent and selective eicosanoid antagonists. REFERENCES Abramovitz M, Boie Y, Nguyen T et al (1994) Cloning and expression of a cDNA for the human prostanoid FP receptor. J Biol Chem, 269, 2632– 2636. Abramovitz M, Adam M, Boie Y et al (2000) The utilization of recombinant prostanoid receptors to determine the affinities and selectivities of prostaglandins and related analogs. Biochim Biophys Acta, 1483, 285–293. Adam M, Boie Y, Rushmore TH et al (1994) Cloning and expression of three isoforms of the human EP3 prostanoid receptor. FEBS Lett, 338, 170–174. Aizawa H, Shigyo M, Nogami H et al (1996) BAY u3405, a thromboxane A2 antagonist, reduces bronchial hyperresponsiveness in asthmatics. Chest, 109, 338–342. Aizawa H, Inoue H, Nakano H et al (1998) Effects of thromboxane A2 antagonist on airway hyperresponsiveness, exhaled nitric oxide, and induced sputum eosinophils in asthmatics. Prostagland Leukotriene Essent Fatty Acids, 59, 185–190. Ambache N (1957) The unsaturated nature of irin and its interaction with lactones. J Physiol, 140, 24P–25P. Ambache N (1959) Further studies on the preparation, purification and nature of irin. J Physiol, 146, 255–294. Angeli V, Faveeuw C, Roye O et al (2001) Role of the parasite-derived prostaglandin D2 in the inhibition of epidermal Langerhans cell migration during schistosomiasis infection. J Exp Med, 193, 1135– 1147. Arakida Y, Suwa K, Ohga K et al (1998a) In vitro pharmacologic profile of YM158, a new dual antagonist for LTD4 and TXA2 receptors. J Pharmacol Exp Ther, 287, 633–639. Arakida Y, Ohga K, Kobayashi S et al (1998b) Binding of YM158, a new dual antagonist for leukotriene D4 and thromboxane A2 receptors, to guinea-pig lung membranes. Eur J Pharmacol, 362, 229–233. Arakida Y, Ohga K, Suwa K et al (1999) Effects of lipid mediator antagonists on predominant mediator-controlled asthmatic reactions in passively sensitized guinea-pigs. J Pharmacol Exp Ther, 290, 1285– 1291. Arakida Y, Ohga K, Suwa K et al (2000a) Effect of YM158, a dual lipid mediator antagonist, on immediate and late asthmatic responses, and airway hyper-responsiveness in guinea-pigs. Jpn J Pharmacol, 82, 287– 294. Arakida Y, Ohga K, Suwa K et al (2000b) In vivo pharmacologic profile of YM158, a new dual antagonist for leukotriene D4 and thromboxane A2 receptors. Jpn J Pharmacol, 83, 63–72. Arakida Y, Ohga K, Okada Y et al (2000c) Effect of combined leukotriene D4 and thromboxane A2 receptor antagonist on mediator-controlled resistance in guinea-pigs. Eur J Pharmacol, 403, 169–179. Arimura A, Asanuma F, Kurosawa A and Harada M (1992) Antiasthmatic activity of a novel thromboxane A2 antagonist, S1452, in guinea-pigs. Int Arch Allergy Immunol, 98, 239–246. Arimura A, Asanuma F, Yagi H et al (1993) Involvement of thromboxane A2 in bronchial hyperresponsiveness but not lung inflammation induced by bacterial lipopolysaccharide in guinea-pigs. Eur J Pharmacol, 231, 13–21. Arimura A, Asanuma F, Kurosawa A and Harada M (1994a) Contribution of thromboxane A2 to the antigen-induced immediate asthmatic response mediated by IgG1 antibody by augmentation of bronchial responsiveness in guinea-pigs. Br J Pharmacol, 111, 339–345. Arimura A, Asanuma F, Matsumoto Y et al (1994b) Effect of the selective thromboxane A2 receptor antagonist, S-1452, on antigen-induced sustained bronchial hyperresponsiveness. Eur J Pharmacol, 260, 201– 209. Arimura A, Miwa M, Hasegawa H et al (1998) Characterization and localization of thromboxane A2 receptor in human and guinea-pig nasal mucosa using radiolabelled (+)-S-145. Br J Pharmacol, 124 795– 803. Arimura A, Yasui K, Kishino J et al (2001) Prevention of allergic inflammation by a novel prostaglandin receptor antagonist, S-5751. J Pharmacol Exp Ther, 298, 411–419.
Asanuma F, Kuwabara K, Arimura A et al (2001) Effects of leukotriene B4 receptor antagonist, LY293111Na, on antigen-induced bronchial hyperresponsiveness and leukocyte infiltration in sensitized guinea pigs. Inflamm Res, 50, 136–141. Ashida Y, Matsumoto T, Kuruki H et al (1989) A novel antiasthmatic quinone derivative, AA-2414, with a potent antagonistic activity against a variety of spasmogenic prostanoids. Prostaglandins, 38, 91– 112. Ashida Y, Matsumoto T, Kuruki H and Saijo T (1991) Effect of the prostaglandin endoperoxide antagonist AA-2414 on experimental allergic asthma. Ann NY Acad Sci, 629, 394–395. Augstein J, Farmaer JB, Lee TB et al (1973) Selective inhibitor for slowreacting substance of anaphylaxis. Nature New Biol, 245, 215–217. Bastien L, Sawyer N, Grygorczyk R et al (1994) Cloning, functional expression, and characterization of the human prostaglandin E2 receptor EP2 subtype. J Biol Chem, 269, 11873–11877. Beasley RCW, Featherstone RL, Church MK et al (1989) Effect of a thromboxane receptor antagonist on PGD2- and allergen-induced bronchoconstriction. J Appl Physiol, 66, 1685–1693. Bernstein PR (1998) Chemistry and structure–activity relationships of leukotriene receptor antagonists. Am J Respir Crit Care Med, 157, S220–S226. Birke FW, Meade CJ, Anderskewitz R et al (2001) In vitro and in vivo pharmacological characterization of BIIL 284, a novel and potent leukotriene B4 receptor antagonist. J Pharmacol Exp Ther, 297, 458– 466. Boie Y, Rushmore TH, Darmon-Goodwin A et al (1994) Cloning and expression of a cDNA for the human prostanoid IP receptor. J Biol Chem, 269, 12173–12178. Boie Y, Sawyer N, Slipetz DM et al (1995) Molecular cloning and characterization of the human prostanoid DP receptor. J Biol Chem, 270, 18910–18916. Boie Y, Stocco R, Sawyer N et al (1997) Molecular cloning and characterization of the four rat prostaglandin E2 prostanoid receptor subtypes. Eur J Pharmacol, 340, 227–241. Boie Y, Stocco R, Sawyer N et al (1999) Characterization of the cloned guinea pig leukotriene B4 receptor: comparison to its human orthologue. Eur J Pharmacol, 380, 203–213. Boura ALA, Gude NM, King RG et al (1986) Characterization of thromboxane A2 receptors in the human fetal placental vessels and umbilical vein. Clin Exp Pharmacol Physiol, 13, 83–86. Bray MA, Anderson WH, Subramanian N et al (1991) CGP 45715 A: a leukotriene D4 analogue with potent peptido-LT antagonist activity. Adv Prostagland Thromboxane Leukotriene Res, 21, 503–507. Brewster AG, Brown GR, Foubister AJ et al (1988) The synthesis of a novel thromboxane receptor antagonist 4(Z)-6-(2-o-chlorophenyl-4-ohydroxyphenyl-1,3-dioxan-cis-5-yl) hexenoic acid ICI 192605. Prostaglandins, 36, 173–178. Brittain RT, Boutal L, Carter MC et al (1985) AH23848: a thromboxane receptor-blocking drug that can clarify the pathophysiologic role of thromboxane A2. Circulation, 6, 1208–1218. Canale P, Squadrito F, Altavilla D et al (1994) Beneficial effects of BAY u3405, a novel thromboxane A2 receptor antagonist, in splanchnic artery occlusion shock. Pharmacology, 49, 376–385. Chambers RJ, Marfat A, Antognoli GW et al (1999) Discovery of CP199,330 and CP-199,331: two potent and orally efficacious cysteinyl LT1 receptor antagonists devoid of liver toxicity. Bioorg Med Chem Lett, 9, 2773–2778. Chan CL, Jones RL and Lau HYA (2000) Characterization of prostanoid receptors mediating inhibition of histamine release from anti-IgEactivated rat peritoneal mast cells. Br J Pharmacol, 129, 589–597. Coleman RA, Humphrey PPA, Kennedy I and Lumley P (1984) Prostanoid receptors—the development of a working classification. Trends Pharmacol Sci, 5, 303–306. Coleman RA, Kennedy I and Sheldrick RLG (1987) New evidence with selective agonists and antagonists for the subclassification of PGE2sensitive (EP) receptors. Adv Prostagland Thromboxane Leukotriene Res, 17, 467–470. Coleman RA, Smith WL and Narumiya S (1994) International Union of Pharmacology classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. Coleman RA, Eglen RM, Jones RL et al (1995) Prostanoid and leukotriene receptors: a progress report from the IUPHAR working
EICOSANOID ANTAGONISTS parties on classification and nomenclature. Adv Prostagland Thromboxane Leukotriene Res, 23, 283–285. Crider JY, Griffin BW and Sharif NA (2000) Endogenous EP4 prostaglandin receptor coupled positively to adenylyl cyclase in Chinese hamster ovary cells: pharmacological characterization. Prostagland Leukotriene Essent Fatty Acids, 62, 21–26. Crider JY and Sharif NA (2001) Functional pharmacological evidence for EP2 and EP4 prostanoid receptors in immortalized human trabecular meshwork and non-pigmented ciliary epithelial cells. J Ocular Pharmacol, 17, 35–46. Darius H, Smith JB and Lefer AM (1985) Beneficial effects of a new potent and specific thromboxane receptor antagonist (SQ-29,548) in vitro and in vivo. J Pharmacol Exp Ther, 235, 274–281. Darius H, Michael-Hepp J and Meyer J (1992) Receptor binding properties of the new and specific thromboxane receptor antagonist Bay u 3405. Agent Action, 37 (suppl), 157–161. Davis TL and Sharif NA (2000) Pharmacological characterization of [3H]prostaglandin E2 binding to the cloned human EP4 prostanoid receptor. Br J Pharmacol, 130, 1919–1926. De Lepeleire I, Reiss TF, Rochette F et al (1997) Montelukast causes prolonged, potent leukotriene D4-receptor antagonism in the airways of patients with asthma. Clin Pharmacol Ther, 61, 83–92. Dogne´ JM, Rolin S, de Leval X et al (2001) Pharmacology of the thromboxane receptor antagonist and thromboxane synthase inhibitor BM-531. Cardiovasc Drug Rev, 19, 87–96. Dunn CJ and Goa KL (2001) Zafirlukast. An update of its pharmacology and therapeutic efficacy in asthma. Drugs, 61, 285–315. Durocher Y, Perret S, Thibaudeau E et al (2000) A reporter gene assay for high-throughput screening of G-protein-coupled receptors stably or transiently expressed in HEK293 EBNA cells grown in suspension culture. Anal Biochem, 226, 349–354. Eakins KE and Karim SMM (1970) Polyphloretin phosphate—a selective antagonist for prostaglandins F1a and F2a. Life Sci, 9, 1–5. Eliasson R (1958) The spasmolytic effect of patulin. Experientia, 14, 460– 461. Farmer JB, Farrar DG and Wilson J (1974) Antagonism of tone and prostaglandin-mediated responses in a tracheal preparation by indomethacin and SC-19220. Br J Pharmacol, 52, 559–565. Fleisch JH, Rinkema LE, Haisch KD et al (1985) LY171883, 1-52hydroxy-3-propyl-4-54-(1H-tetrazol-5-yl)butoxy4phenyl4ethanone, an orally active leukotriene D4 antagonist. J Pharmacol Exp Ther, 233, 148–157. Francis HP, Greenham SJ, Patel UP et al (1991) BAY u3405 an antagonist of thromboxane A2- and prostaglandin D2-induced bronchoconstriction in the guinea-pig. Br J Pharmacol, 104, 596–602. Fried J et al (1969) Prostaglandin antagonists: synthesis and smooth muscle activity. Nature, 223, 208–210. Fujimura M, Sakamoto S, Kamio Y and Matsuda T (1993) Effect of a leukotriene antagonist, ONO-1078, on bronchial hyperresponsiveness in patients with asthma. Resp Med, 87, 133–138. Fujita M, Yonetomi Y, Shimouchi K et al (1999) Involvement of cysteinyl leukotrienes in biphasic increase of nasal airway resistance of antigeninduced rhinitis in guinea-pigs. Eur J Pharmacol, 369, 349–356. Funk CD, Furci L, FitzGerald GA et al (1993) Cloning and expression of a cDNA for the human prostaglandin E receptor EP1 subtype. J Biol Chem, 268, 26767–26772. Gardiner PJ, Abram TS and Cuthbert NJ (1990) Evidence for two leukotriene receptor types in the guinea-pig isolated ileum. Eur J Pharmacol, 182, 219–299. Gardiner PJ, Norman P, Cuthbert NJ et al (1993) Characterization of the peptide-leukotriene receptor PL2 on the ferret spleen strip. Eur J Pharmacol, 238, 19–26. Giles H, Leff P, Bolofo ML et al (1989) The classification of prostaglandin DP-receptors in platelets and vasculature using BW A868C, a novel, selective and potent competitive antagonist. Br J Pharmacol, 96, 291– 300. Gladue RP, Carroll LA, Milici AJ et al (1996) Inhibition of leukotriene B4–receptor interaction suppresses eosinophil infiltration and disease pathology in a murine model of experimental allergic encephalomyelitis. J Exp Med, 183, 1893–1898. Gomoll AW, Grover GJ and Ogletree ML (1994) Myocardial salvage efficacy of the thromboxane receptor antagonist ifetroban in ferrets and dogs. J Cardiovasc Pharmacol, 24, 960–968.
175
Griffin BW, Williams GW, Crider JY and Sharif NA (1997) FP prostaglandin receptor mediating inositol phosphates generation and calcium mobilization in Swiss 3T3 cells: a pharmacological study. J Pharmacol Exp Ther, 281, 845–854. Griffin BW, Magnino PE, Pang IH and Sharif NA (1998) Pharmacological characterization of an FP prostaglandin receptor on rat vascular smooth muscle cells (A7r5) coupled to phosphoinositide turnover and intracellular calcium mobilization. J Pharmacol Exp Ther, 286, 411– 418. Griffin BW, Klimko P, Crider JY and Sharif NA (1999) AL-8810: a novel prostaglandin F2a analog with selective antagonist effects at the prostaglandin F2a (FP) receptor. J Pharmacol Exp Ther, 290, 1278– 1284. Hallinan EA, Hagen TJ, Tsymbalov S et al (1996) Aminoacetyl moiety as a potential surrogate for diacylhydrazine group of SC-51089, a potent PGE2 antagonist, and its analogs. J Med Chem, 39, 609–613. Hamid-Bloomfield S and Whittle BJR (1989) Antagonism of PGD2 vasodepressor responses in the rat in vivo by the novel, selective antagonist, BW A868C. Br J Pharmacol, 96, 307–312. Hanasaki K and Arita H (1988) Characterization of a new compound, S145, as a specific TXA2 receptor antagonist in platelets. Thromb Res, 50, 365–376. Hay DWP, Muccitelli RM, Tucker SS et al (1987) Pharmacologic profile of SK&F 104353: a novel, potent and selective peptideleukotriene receptor antagonist in guinea-pig and human airways. J Pharmacol Exp Ther, 243, 474–481. Heise CE, O’Dowd BF, Figueroa DJ et al (2000) Characterization of the human cysteinyl leukotriene 2 receptor. J Biol Chem, 275, 30531– 30536. Hirai H, Tanaka K, Yoshie O et al (2001) Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J Exp Med, (193:) 255– 261. Hirata M, Hayashi Y, Ushikubi F et al (1991) Cloning and expression of cDNA for a human thromboxane A2 receptor. Nature, 349, 617–620. Hirata M, Kakizuka A, Aizawa M et al (1994) Molecular characterization of a mouse prostaglandin D receptor and functional expression of the cloned gene. Proc Natl Acad Sci USA, 91, 11192–11196. Honda A, Sugimoto Y, Namba T et al (1993) Cloning and expression of a cDNA for mouse prostaglandin E receptor EP2 subtype. J Biol Chem, 268, 7759–7762. Hoshino M, Sim J, Shimizu K et al (1999) Effect of AA-2414, a thromboxane A2 receptor antagonist, on airway inflammation in subjects with asthma. J Allerg Clin Immunol, 103, 1054–1061. Huang WW, Garcia-Zepeda EA, Sauty A et al (1998) Molecular and biological characterization of the murine leukotriene B4 receptor expressed on eosinophils. J Exp Med, 188, 1063–1074. Hui Y, Yang G, Galczenski H et al (2001) The murine cysteinyl leukotriene 2 (CysLT2) receptor. cDNA and genomic cloning, alternative splicing, and in vitro characterization. J Biol Chem, 276, 47489–47495. Ii T, Izumi R and Shimizu K (1996) The immunosuppressive effects of a leukotriene B4 receptor antagonist on liver allotransplantation in rats. Surg Today, 26, 419–426. Imura Y, Terashita Z, Shibouta Y et al (1990) Antagonistic action of AA2414 on thromboxane A2/prostaglandin endoperoxide receptor in platelets and blood vessels. Jpn J Pharmacol, 52, 35–43. Jackson WT, Boyd RJ, Froelich LL et al (1988) Inhibition of LTB4 binding and aggregation of neutrophils by LY255283 and LY223982. FASEB J, 2, A1110. Jackson WT, Boyd RJ, Froelich LL et al (1992) Specific inhibition of leukotriene B4-induced neutrophil activation by LY223982. J Pharmacol Exp Ther, 263, 1009–1014. Jackson WT, Froelich LL, Boyd RJ et al (1999) Pharmacological actions of the second-generation leukotriene B4 receptor antagonist LY293111: in vitro studies. J Pharmacol Exp Ther, 288, 286–294. Jarvis B and Markham A (2000) Montelukast. A review of its therapeutic potential in persistent asthma. Drugs, 59, 891–928. Jenkins DW, Feniuk W and Humphrey PPA (2001) Characterization of the prostanoid receptor types involved in mediating calcitonin generelated peptide release from cultured rat trigeminal neurons. Br J Pharmacol, 134, 1296–1302.
176
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Jones RL, Peesapati V and Wilson NH (1982) Antagonism of the thromboxane-sensitive contractile systems of the rabbit aorta, dog saphenous vein and guinea-pig trachea. Br J Pharmacol, 76, 423–438. Jones RL and Chan K (2001) Distinction between relaxations induced via prostanoid EP4 and IP1 receptors in pig and rabbit blood vessels. Br J Pharmacol, 134, 313–324. Jones TR, Labelle M, Belley M et al (1995) Pharmacology of montelukast sodium (Singulair2), a potent and selective leukotriene D4 receptor antagonist. Can J Physiol Pharmacol, 73, 191–201. Kamohara M, Takasaki J, Matsumoto M et al (2000) Molecular cloning and characterization of another leukotriene B4 receptor. J Biol Chem, 275, 27000–27004. Karasawa A, Guo JP, Ma XL et al (1991) Protective actions of a leukotriene B4 antagonist in splanchnic ischemia and reperfusion in rats. Am J Physiol, 261, G191–G198. Katsuyama M, Nishigaki N, Sugimoto Y et al (1995) The mouse prostaglandin E receptor EP2 subtype: cloning, expression, and Northern blot analysis. FEBS Lett, 372, 151–156. Kattelman EJ, Venton DL and Le Breton GC (1986) Characterization of U46619 binding in unactivated, intact human platelets and determination of binding site affinities of four TXA2/PGH2 receptor antagonists (13-APA, BM 13.177, ONO 3708, SQ 29,548). Thromb Res, 41, 471–481. Kawahara H, Sakamoto A, Takeda S et al (2001) A prostaglandin E2 receptor subtype EP1 receptor antagonist (ONO-8711) reduces hyperalgesia, allodynia, and c-fos gene expression in rats with chronic nerve constriction. Anesth Analg, 93, 1012–1017. Kawamori T, Uchiya N, Nakatsugi S et al (2001) Chemopreventive effects of ONO-8711, a selective prostaglandin E receptor EP1 antagonist, on breast cancer development. Carcinogenesis, 22, 2001–2004. Keery RJ and Lumley P (1988) AH6809, a prostaglandin DP-receptor blocking drug on human platelets. Br J Pharmacol, 94, 745–754. Kennedy I, Coleman RA, Humphrey PPA et al (1982) Studies on the characterization of prostanoid receptors: a proposed classification. Prostaglandins, 24, 667–689. Kinsella BT, O’Mahony DJ and FitzGerald GA (1997) The human thromboxane A2 receptor a isoform (TPa) functionally coupled to the G proteins Gq and G11 in vivo and is activated by the isoprostane 8-epi prostaglandin F2a. J Pharmacol Exp Ther, 281, 957–964. Kiriyama M, Ushikubi F, Kobayashi T et al (1997) Ligand binding specificities of the eight types and subtypes of the mouse prostanoid receptors expressed in Chinese hamster ovary cells. Br J Pharmacol, 122, 217–224. Kishikawa K, Matsunaga N, Maruyama T et al (1990) ONO-LB-457: a novel and orally active leukotriene B4 receptor antagonist. Adv Prostagland Thromboxane Leukotriene Res, 21, 407–410. Kishikawa K, Tateishi N, Maruyama T et al (1992) ONO-4057, a novel, orally active leukotriene B4 antagonist: effects on LTB4-induced neutrophil functions. Prostaglandins, 44, 261–275. Krell RD, Giles RE, Yee YK and Snyder DW (1987) In vivo pharmacology of ICI 198,615: a novel, potent and selective peptide leukotriene antagonist. J Pharmacol Exp Ther, 243, 557–564. Krell RD, Aharony D, Buckner CK et al (1990) The preclinical pharmacology of ICI 204,219. A peptide leukotriene antagonist. Am Rev Respir Dis, 141, 978–987. Kurokawa T, Matsumoto T, Ashida Y et al (1994) Antagonism of the human thromboxane A2 receptor by an anti-asthmatic agent AA-2414. Biol Pharm Bull, 17, 383–385. Kuwabara K, Yasui K, Jyoyama H et al (2000) Effects of the secondgeneration leukotriene B4 receptor antagonist, LY293111Na, on leukocyte infiltration and collagen-induced arthritis in mice. Eur J Pharmacol, 402, 275–285. Labat C, Ortiz JL, Norel X et al (1992) A second cysteinyl leukotriene receptor in human lung. J Pharmacol Exp Ther, 263, 800–805. Labelle M, Metters KM, Abramovitz M et al (2000) Potent and selective EP3 receptor antagonists. 11th International Conference on Advances in Prostaglandin and Leukotriene Research: Basic Sciences and New Clinical Applications. Abstract Book, Florence, 2000, 62. Lambley JE and Smith AP (1975) The effects of arachidonic acid, indomethacin, and SC-19220 on guinea-pig tracheal muscle tone. Eur J Pharmacol, 30, 148–153. Lawson CF, Wishka DG, Morris J and Fitzpatrick FA (1989) Receptor antagonism of leukotriene B4 myotropic activity by the 2,6
disubstituted pyridine analog U-75302: characterization on lung parenchyma strips. J Lipid Mediators, 1, 3–12. Li EJ, Cook JA, Wise WC et al (1991) Effect of LTB4 receptor antagonists in endotoxic shock in the rat. Circ Shock, 34, 385–392. Lin AH, Morris J, Wishka DG and Gorman RR (1988) Novel molecules that antagonize leukotriene B4 binding to neutrophils. Ann NY Acad Sci, 524, 196–200. Lumley P, White BP and Humphrey PPA (1989) GR32191, a highly potent and specific thromboxane A2 receptor blocking drug on platelets and vascular and airways smooth muscle in vitro. Br J Pharmacol, 97, 783–794. Lynch KR, O’Neill GP, Liu Q et al (1999) Characterization of the human cysteinyl leukotriene CysLT1 receptor. Nature, 399, 789–793. Maekawa A, Kanaoka Y, Lam BK and Austen KF (2001) Identification in mice of two isoforms of the cysteinyl leukotriene 1 receptor that result from alternative splicing. Proc Natl Acad Sci USA, 98, 2256– 2261. Malmberg AB, Rafferty MF and Yaksh TL (1994) Antinociceptive effect of spinally delivered prostaglandin E receptor antagonists in the formalin test on the rat. Neurosci Lett, 173, 193–196. Martin V, Ronde P, Unett D et al (1999) Leukotriene binding, signaling, and analysis of HIV co-receptor function in mouse and human leukotriene B4 receptor-transfected cells. J Biol Chem, 274, 8597–8603. Martin V, Sawyer N, Stocco R et al (2001) Molecular cloning and functional characterization of murine cysteinyl-leukotriene 1 (CysLT1) receptors. Biochem Pharmacol, 62, 1193–1200. Maruyama T and Ohuchida S (2000) Selective agonists and antagonists for prostaglandin E2 receptor subtypes. Protein Nucleus Enzyme, 45, 1001–1007 (published in Japanese). Masuda K, Yokomizo T, Izumi T and Shimizu T (1999) cDNA cloning and characterization of guinea-pig leukotriene B4 receptor. Biochem J, 342, 79–85. Matousek M, Mitsube K, Mikuni M and Bra¨nnstro¨m M (2001) Inhibition of ovulation in the rat by a leukotriene B4 receptor antagonist. Mol Hum Reprod, 7, 35–42. Matsumoto T, Ashida Y and Tsukuda R (1994) Pharmacological modulation of immediate and late airway response and leukocyte infiltration in the guinea-pig. J Pharmacol Exp Ther, 269, 1236–1244. Matsuo Y, Izumiyama M, Onodera H et al (1993) Effect of a novel thromboxane A2 receptor antagonist, S-1452, on postischemic brain injury in rats. Stroke, 24, 2059–2065. Matsuo Y, Takagawa I, Koshida H et al (1995) Antiproteinuric effect of a thromboxane receptor antagonist, S-1452, on rat diabetic nephropathy and murine lupus nephritis. Pharmacology, 50, 1–8. McKenniff MG, Norman P, Cuthbert NJ and Gardiner PJ (1991) BAY u3405, a potent and selective thromboxane A2 receptor antagonist on airway smooth muscle in vitro. Br J Pharmacol, 104, 585–590. Monneret G, Gravel S, Diamond M et al (2001) Prostaglandin D2 is a potent chemoattractant for human eosinophils that acts via a novel DP receptor. Blood, 98, 1942–1948. Morita H, Takeda K, Yagita H and Okumura K (1999) Immunosuppressive effect of leukotriene B4 receptor antagonist in vitro. Biochem Biophys Res Commun, 264, 321–326. Mutoh M, Watanabe K, Kitamura T et al (2002) Involvement of prostaglandin E receptor subtype EP4 in colon caricinogenesis. Cancer Res, 62, 28–32. Nagai H, Takeda H, Yamaguchi S et al (1995) The effect of a thromboxane A2 receptor antagonist BAY-u-3405 on experimental allergic reactions. Prostaglandins, 50, 75–87. Nagata K, Tanaka K, Ogawa K et al (1999a) Selective expression of a novel surface molecule by human Th2 cells in vivo. J Immunol, 162, 1278–1286. Nagata K, Hirai H, Tanaka K et al (1999b) CRTH2, an orphan receptor of T-helper-2-cells, is expressed on basophils and eosinophils and responds to mast cell-derived factor(s). FEBS Lett, 459, 195–199. Nakagawa N, Obata T, Kobayashi T et al (1992) In vivo pharmacologic profile of ONO-1078: a potent, selective and orally active peptide leukotriene (LT) antagonist. Jpn J Pharmacol, 60, 217–225. Nakagawa N, Obata T, Kobayashi T et al (1993) Effect of a peptide leukotriene receptor antagonist, ONO-1078, on guinea-pig models of asthma. Eur J Pharmacol, 235, 211–219. Nakajima M and Ueda M (1989) Actions of a novel thromboxane A2receptor antagonist, S-145, on isolated monkey and cat arteries. J Cardiovasc Pharmacol, 14, 502–509.
EICOSANOID ANTAGONISTS Namba T, Sugimoto Y, Hirata M et al (1992) Mouse thromboxane A2 receptor: cDNA cloning, expression and Northern blot analysis. Biochem Biophys Res Commun, 184, 1197–1203. Namba T, Oida H, Sugimoto Y et al (1994) cDNA cloning of a mouse prostacyclin receptor. Multiple signaling pathways and expression in thymic medulla. J Biol Chem, 269, 9986–9992. Narita S, Asakura K and Katakura A (1996) Effects of thromboxane A2 receptor antagonist (Bay u 3405) on nasal symptoms after antigen challenge in sensitized guinea-pigs. Int Arch Allergy Immunol, 109, 161–166. Neuscha¨fer-Rube F, DeVries C, Ha¨necke K et al (1994) Molecular cloning and expression of a prostaglandin E2 receptor of the EP3b subtype from rat hepatocytes. FEBS Lett, 351, 119–122. Nishigaki N, Negishi M, Honda A et al (1995) Identification of prostaglandin E receptor ‘EP2’ cloned from mastocytoma cells as EP4 subtype. FEBS Lett, 364, 339–341. Noguchi K, Iwasaki K, Shitashige M and Ishikawa I (1999) Prostaglandin E2 receptors of the EP2 and EP4 subtypes downregulate tumor necrosis factor a-induced intercellular adhesion molecule-1 expression in human gingival fibroblasts. J Periodont Res, 34, 478–485. Norel X, Walch L, Labat C et al (1999) Prostanoid receptors involved in the relaxation of human bronchial preparations. Br J Pharmacol, 126, 867–872. Nothacker HP, Wang Z, Zhu Y et al (2000) Molecular cloning and characterization of a second human cysteinyl leukotriene receptor: discovery of a subtype selective agonist. Mol Pharmacol, 58, 1601–1608. Obata T, Okada Y, Motoishi M et al (1992) In vitro antagonism of ONO1078, a newly developed anti-asthma agent, against peptide leukotrienes in isolated guiena pig tissues. Jpn J Pharmacol, 60, 227–237. Oka K, Oka T and Hori T (1998) PGE2 receptor subtype EP1 antagonist may inhibit central interleukin-1b-induced fever in rats. Am J Physiol, 275, R1762–R1765. Okuda-Ashitaka E, Sakamoto K, Ezashi T et al (1996) Suppression of prostaglandin E receptor signaling by the variant form of EP1 subtype. J Biol Chem, 271, 31255–31261. Ogletree ML, Harris DN, Greenberg R et al (1985) Pharmacological actions of SQ 29,548, a novel selective thromboxane antagonist. J Pharmacol Exp Ther, 234, 435–441. Ogletree ML, Harris DN, Schumacher WA et al (1993) Pharmacological profile of BMS 180,291: a potent, long-lasting, orally active thromboxane A2/prostaglandin endoperoxide receptor antagonist. J Pharmacol Exp Ther, 264, 570–578. Omote K, Kawamata T, Nakayama Y et al (2001) The effects of peripheral administration of a novel selective antagonist for prostaglandin E receptor subtype EP1, ONO-8711, in a rat model of postoperative pain. Anesth Analg, 92, 233–238. Ono K, Akatsu T, Murakami T et al (1998) Important role of EP4, a subtype of prostaglandin (PG) E receptor, in osteoclast-like cell formation from mouse bone marrow cells induced by PGE2. J Endocrinol, 158, R1–R5. Owman C, Sabirsh A, Boketoft A˚ and Olde B (1997) Leukotriene B4 is the functional ligand binding to and activating the cloned chemoattractant receptor, CMKRL1. Biochem Biophys Res Commun, 240, 162–166. Pelletier S, Dube´ J, Villeneuve A et al (2001) Prostaglandin E2 increases cyclic AMP and inhibits endothelin-1 production/secretion by guineapig tracheal epithelial cells through EP4 receptors. Br J Pharmacol, 132, 999–1008. Perzborn E, Fiedler VB, Seuter F et al (1990) Characterization of Bay U 3405, a novel thromboxane A2/endoperoxide receptor antagonist. Stroke, 21, IV-143–IV-145. Raychaudhuri A, Kotyuk B, Pellas TC et al (1995) Effect of CGS 25019C and other LTB4 antagonists in the mouse ear edema and rat neutropenia models. Inflamm Res, 44, (suppl 2), S141–S142. Raychowdhury MK, Yukawa M, Collins LJ et al (1994) Alternative splicing produces a divergent cytoplasmic tail in the human endothelial thromboxane A2 receptor. J Biol Chem, 269, 19256–19261 [published erratum appears in J Biol Chem, 270, 7011 (1995)]. Regan JW, Bailey TJ, Pepperl DJ et al (1994) Cloning of a novel human prostaglandin receptor with characteristics of the pharmacologically defined EP2 subtype. Mol Pharmacol, 46, 213–220. Reinheimer T, Harnack E, Racke K and Wessler I (1998) Prostanoid receptors of the EP3 subtypes mediate inhibition of evoked [3H]acetylcholine release from isolated human bronchi. Br J Pharmacol, 125, 271–276.
177
Richards IM, Griffin RL, Oostveen JA et al (1989) Effect of the selective leukotriene B4 antagonist U-75302 on antigen-induced bronchopulmonary eosinophilia in sensitized guinea-pigs. Am Rev Respir Dis, 140, 1712–1716. Rochards IM, Shields SK, Johnson RA and Smith HW (1991) Effects of the selective leukotriene B4 antagonists U-75302 and Ly255283 on bronchoalveolar eosinophilia induced by inhalation of leukotriene B4 or allergen in brown Norway rats. Ann NY Acad Sci, 629, 428–429. Rolin S, Dogne´ JM, Michaux C et al (2001) Activity of a novel dual thromboxane A2 receptor antagonist and thromboxane synthase inhibitor (BM-573) on platelet function and isolated smooth muscles. Prostagland Leukotriene Essent Fatty Acids, 65, 67–72. Rosentreter U, Bo¨shagen H, Seuter F et al (1989) Synthesis and absolute configuration of the new thromboxane antagonist (3R)-3-(4fluorophenylsulfonamido)-1,2,3,4-tetrahydro-9-carbazolepropanoic acid and comparison with its enantiomer. Arzneim Forsch Drug Res, 39, 1519–1521. Sando T, Usui T, Tanaka I et al (1994) Molecular cloning and expression of rat prostaglandin E receptor EP2 subtype. Biochem Biophys Res Commun, 200, 1329–1333. Sanner JH (1969) Antagonism of prostaglandin E2 by 1-acetyl-2-(8-chloro10,11-dihydrodibenz[b,f][1,4]oxazepine-10-carbonyl)hydrazine (SC19220). Arch Int Pharmacodyn, 180, 46–56. Sanner JH (1988) Eicosanoid antagonists. In Curtis-Prior PB (ed.), Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids. Churchill Livingstone, Edinburgh, 179–197. Sarau HM, Ames RS, Chanbers J et al (1999) Identification, molecular cloning, expression, and characterization of a cysteinyl leukotriene receptor. Mol Pharmacol, 56, 657–663. Saussy DL Jr, Mais DE, Dube´ GP et al (1991) Characterization of a thromboxane A2/prostaglandin H2 receptor in guinea-pig lung membranes using a radioiodinated thromboxane mimetic. Mol Pharmacol, 39, 72–78. Schmid A, Thierauch KH, Schleuning WD and Dinter H (1995) Splice variants of the human EP3 receptor for prostaglandin E2. Eur J Biochem, 228, 23–30. Sharif NA, Williams GW and , Davis TL (2000a) Pharmacology and autoradiography of human DP prostanoid receptors using [3H]-BW A868C, a DP receptor-selective antagonist radioligand. Br J Pharmacol, 131, 1025–1038. Sharif NA, Crider JY, Xu SX and Williams GW (2000b) Affinities, selectivities, potencies, and intrinsic activities of natural and synthetic prostanoids using endogenous receptors: focus on DP class prostanoids. J Pharmcol Exp Ther, 293, 321–328. Sharif NA, Crider JY and Davis TL (2000c) AL-3138 antagonizes FP prostanoid receptor-mediated inositol phosphates generation: comparison with some purported FP antagonists. J Pharm Pharmacol, 52, 1529–1539. Showell HJ, Pettipher ER, Cheng JB et al (1995) The in vitro and in vivo pharmacologic activity of the potent and selective leukotriene B4 receptor antagonist CP-105696. J Pharmacol Exp Ther, 273, 176–184. Showell HJ, Conklyn MJ, Alpert R et al (1998) The preclinical pharmacological profile of the potent and selective leukotriene B4 antagonist CP-195543. J Pharmacol Exp Ther, 285, 946–954. Silbaugh SA, Stengel WS, Cockerham SL et al (1992) Pulmonary actions of LY255283, a leukotriene B4 receptor antagonist. Eur J Pharmacol, 223, 57–64. Silbaugh SA, Stengel PW, Cockerham SL et al (2000) Pharmacologic actions of the second generation leukotriene B4 receptor antagonist LY293111: in vivo pulmonary studies. Naunyn Schmiedeberg’s Arch Pharmacol, 361, 397–404. Smith LJ, Geller S, Ebright L et al (1990) Inhibition of leukotriene D4induced bronchoconstriction in normal subjects by the oral LTD4 receptor antagonist ICI 204,219. Am Rev Resp Dis, 141, 988–992. Snyder DW and Krell RD (1984) Pharmacological evidence for a distinct leukotriene C4 receptor in guinea-pig trachea. J Pharmacol Exp Ther, 231, 616–622. Snyder DW, Giles RE, Keith RA et al (1987) In vitro pharmacology of ICI 198,615: a novel, potent and selective peptide leukotriene antagonist. J Pharmacol Exp Ther, 243, 548–556. Souza DG, Coutinho SF, Silveira MR et al (2000) Effects of a BLT receptor antagonist on local and remote reperfusion injures after transient ischemia of the superior mesenteric artery in rats. Eur J Pharmacol, 403, 121–128.
178
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Spicuzza L, Giembycz MA, Barnes PJ and Belvisi M (1998) Prostaglandin E2 suppression of acetylcholine release from parasympathetic nerves innervating guinea-pig trachea by interacting with prostanoid receptors of the EP3-subtype. Br J Pharmacol, 123, 1246–1252. Sugimoto H, Shichijo M, Iino T et al (2003) An orally bioavailable small molecule antagonist of CRTH2, ramatroban (BAY u3405), inhibits prostaglandin D2-induced eosinophil migration in vitro. J Pharmacol Exp Ther, 305, 347–352. Sugimoto Y, Namba T, Honda A et al (1992) Cloning and expression of a cDNA for mouse prostaglandin E receptor EP3 subtype. J Biol Chem, 267, 6463–6466. Sugimoto Y, Hasumoto K, Namba T et al (1994) Cloning and expression of a cDNA for mouse prostaglandin F receptor. J Biol Chem, 269, 1356–1360. Takahara K, Murray R, FitzGerald GA and FitzGerald DJ (1990) The response to thromboxane A2 analogues in human platelets. Discrimination of two binding sites linked to distinct effector systems. J Biol Chem, 265, 6836–6844. Takeuchi K, Abe T, Takahashi N and Abe K (1993) Molecular cloning and intrarenal localization of rat prostaglandin E2 receptor EP3 subtype. Biochem Biophys Res Commun, 194, 885–891. Taki F, Suzuki R, Torii K et al (1994) Reduction of the severity of bronchial hyperresponsiveness by the novel leukotriene antagonist 4-oxo8-[4-(4-phenylbutoxy)benzoylamino]-2-(tetrazol-5-yl)-4H-1-benzopyran hemihydrate. Arzneim Forsch Drug Res, 44, 330–333. Tanaka T, Fukuta Y, Higashino R et al (1998) Antiplatelet effect of Z-335, a new orally active and long-lasting thromboxane receptor antagonist. Eur J Pharmacol, 357, 53–60. Tanaka M, Tamaki T, Konoeda Y et al (2000) Effect of leukotriene B4 receptor antagonist (ONO4057) on hepatic allografting in rats. Transplant Proc, 32, 2340. Taniguchi Y, Tamura G, Honma M et al (1993) The effect of an oral leukotriene antagonist, ONO-1078, on allergen-induced immediate bronchoconstriction in asthmatic subjects. J Allerg Clin Immunol, 92, 507–512. Taylor BM, Crittenden NJ, Bruden MN et al (1991) Biological activity of leukotriene B4 analogs: inhibition of guinea-pig eosinophil migration in vitro by the 2,6-disubstituted pyridine analogs U-75,302 and U-75,485. Prostaglandins, 42, 211–224. Terada N, Yamakoshi T, Hasegawa M et al (1998) The effect of ramatroban (BAY u 3405), a thromboxane A2 receptor antagonist, on nasal cavity volume and minimum cross-sectional area and nasal mucosal hemodynamics after nasal mucosal allergen challenge in patients with perennial allergic rhinitis. Acta Otolaryngol 537, (suppl), 32–37. Toda A, Yokomizo T, Masuda K et al (1999) Cloning and characterization of rat leukotriene B4 receptor. Biochem Biophys Res Commun, 262, 806–812. Trist DG, Collins BA, Wood J et al (1989) The antagonism by BW A868C of PGD2 and BW245C activation of human platelet adenylate cyclase. Br J Pharmacol, 96, 301–306. Tryselius Y, Nilsson NE, Kotarsky K et al (2000) Cloning and characterization of cDNA encoding a novel human leukotriene B4 receptor. Biochem Biophys Res Commun, 274, 377–382. Tsai BS, Villani-Price D, Keith RH et al (1989) SC-41930: an inhibitor of leukotriene B4-stimulated human neutrophil functions. Prostaglandins, 38, 655–674. Tsai BS, Keith RH, Villani-Price D et al (1994) The in vitro pharmacology of SC-51146: a potent antagonist of leukotriene B4 receptors. J Pharmacol Exp Ther, 268, 1499–1505. Tudhope SR, Cuthbert NJ, Abram TS et al (1994) BAY u9773, a novel antagonist of cysteinyl-leukotrienes with activity against two receptor subtypes. Eur J Pharmacol, 264, 317–323. Turner CR, Breslow R, Conklyn MJ et al (1996) In vitro and in vivo effects of leukotriene B4 antagonism in a primate model of asthma. J Clin Invest, 97, 381–387. Underwood DC, Kriseman T, McNamara DB et al (1987) Blockade of thromboxane responses in the airway of the cat by SQ 29,548. J Appl Physiol, 62, 2193–2200. Ushikubi F, Nakajima M, Yamamoto M et al (1989) [3H]S-145 and [125I]IS-145-OH: new radioligands for platelet thromboxane A2 receptor with low non-specific binding and high binding affinity for various receptor preparations. Eicosanoids, 2, 21–27. Ushikubi F, Segi E, Sugimoto Y et al (1998) Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature, 395, 281–284.
Walch L, Labat C, Gascard JP et al (1999) Prostanoid receptor involved in the relaxation of human pulmonary vessels. Br J Pharmacol, 126, 859–866. Walch L, de Montpreville V, Brink C and Norel X (2001) Prostanoid EP1and TP-receptors involved in the contraction of human pulmonary veins. Br J Pharmacol, 134, 1671–1678. Watabe A, Sugimoto Y, Honda A et al (1993) Cloning and expression of cDNA for a mouse EP1 subtype of prostaglandin E receptor. J Biol Chem, 268, 20175–20178. Watanabe K, Kawamori T, Nakatsugi S et al (1999) Role of the prostaglandin E receptor subtype EP1 in colon carcinogenesis. Cancer Res, 59, 5093–5096. Watanabe K, Kawamori T, Nakatsugi S et al (2000) Inhibitory effect of a prostaglandin E receptor subtype EP1 selective antagonist, ONO-8713, on development of azoxymethane-induced aberrant crypt foci in mice. Cancer Lett, 156, 57–61. Weringer EJ, Perry BD, Sawyer PS et al (1999) Antagonizing leukotriene B4 receptors delays cardiac allograft rejection in mice. Transplantation, 67, 808–815. Wise H (1998) Activation of the prostaglandin EP4-receptor subtype is highly coupled to inhibition of N-formyl-methionyl-leucylphenylalanine-stimulated rat neutrophil aggregation. Prostagland Leukotriene Essent Fatty Acids, 58, 77–84. Wollert PS, Menconi MJ, O’Sullivan BP et al (1993) LY255283, a novel leukotriene B4 receptor antagonist, limits activation of neutrophils and prevents acute lung injury induced by endotoxin in pigs. Surgery, 114, 191–198. Woodward DF, Pepperl DJ, Burkey TH and Regan JW (1995) 6Isopropoxy-9-oxoxanthene-2-carboxylic acid (AH6809), a human EP2 receptor antagonist. Biochem Pharmacol, 50, 1731–1733. Woodward DF, Nieves AL and Friedlaender MH (1996) Characterization of receptor subtypes involved in prostanoid-induced conjunctival pruritus and their role in mediating allergic conjunctival itching. J Pharmacol Exp Ther, 279, 137–142. Wright DH, Nantel F, Metters KM and Ford-Hutchinson AW (1999) A novel biological role for prostaglandin D2 is suggested by distribution studies of the rat DP prostanoid receptor. Eur J Pharmacol, 377, 101– 115. Yamasaki M, Matsumoto T, Fukuda S et al (1997) Involvement of thromboxane A2 and histamine in experimental allergic rhinitis of guinea-pigs. J Pharmacol Exp Ther, 280, 1471–1479. Yamasaki M, Mizutani N, Sasaki K et al (2001a) Involvement of thromboxane A2 and peptide leukotrienes in early and late phase nasal blockage in a guinea-pig model of allergic rhinitis. Inflamm Res, 50, 466–473. Yamasaki M, Sasaki K, Mizutani N et al (2001b) Pharmacological characterization of the leukocyte kinetics after intranasal antigen challenge in a guinea-pig model of allergic rhinitis. Inflamm Res, 50, 474–482. Yasui K, Asanuma F, Furue Y and Arimura A (1997) Involvement of thromboxane A2 in antigen-induced nasal blockage in guinea-pigs. Int Arch Allerg Immunol, 112, 400–405. Yasui K, Asanuma F and Arimura A (2001) Inhibitory effect of a TPreceptor antagonist, S-1452, on antigen-induced nasal plasma exudation in guinea-pig. Pharmacology, 63, 65–70. Yokomizo T, Izumi T, Chang K et al (1997) A G-protein-coupled receptor for leukotriene B4 that mediates chemotaxis. Nature, 387, 620–624. Yokomizo T, Kato K, Terawaki K et al (2000) A second leukotriene B4 receptor, BLT2: a new therapeutic target in inflammation and immunological disorders. J Exp Med, 192, 421–431. Yokomizo T, Kato K, Hagiya H et al (2001) Hydroxyeicosanoids bind to and activate the low affinity leukotriene B4 receptor, BLT2. J Biol Chem, 276, 12454–12459. Yoshida S, Ishizaki Y, Shoji T et al (2000) Effect of pranlukast on bronchial inflammation in patients with asthma. Clin Exp Allerg, 30, 1008–1014. Zamboni R, Belley M, Champion E et al (1992) Development of a novel series of styrylquinoline compounds as high-affinity leukotriene D4 receptor antagonists: synthetic and structure-activity studies leading to the discovery of (+)-3-[[[3-[2-(7-chloro-2-quinolinyl)-(E)-ethenyl]phenyl] [[3-(dimethylamino)-3-oxopropyl]thio]methyl]thio]propionic acid. J Med Chem, 35, 3832–3844. Zhang Z, Hefferan MP and Loomis CW (2001) Topical bicuculline to the rat spinal cord induces highly localized allodynia that is mediated by spinal prostaglandins. Pain, 92, 351–361.
15 Biosynthesis and Degradation of Anandamide, an Endogenous Ligand of Cannabinoid Receptors Natsuo Ueda1 and Dale G. Deutsch2 1Kagawa
University School of Medicine, Kagawa, Japan; and 2State University of New York, Stony Brook, NY, USA
D9-Tetrahydrocannabinol is the psychoactive cannabinoid constituent in marijuana (Figure 15.1) (Pertwee 1997). Biological activities of the cannabinoids are generally attributable to signal transduction via two types of G protein-coupled cannabinoid receptors (CB1 and CB2). cDNAs of these receptors have been cloned in 1990 and 1993, respectively (Matsuda et al 1990; Munro et al 1993). CB1 is highly expressed in the brain, and low in peripheral tissues, while CB2 is absent from brain but highly expressed in immune cells, such as B cells and macrophages (Pertwee 1997). Shortly after the cDNA cloning of CB1, anandamide (N-arachidonoylethanolamine; Figure 15.1) was discovered as an endogenous ligand of the receptor (Devane et al 1992). Later, ethanolamides of other polyunsaturated fatty acids (N-dihomo-g-linolenoylethanolamine and N-docosatetraenoylethanolamine) (Hanus et al 1993), 2-arachidonoylglycerol (2-AG; Mechoulam et al 1995; Sugiura et al 1995), and 2arachidonyl glyceryl ether (referred to as Noladin ether; Hanus et al 2001) were also found to be the receptor ligands. The compounds so far reported as endogenous ligands of cannabinoid receptors are all polyunsaturated fatty acid derivatives, and are collectively referred to as endocannabinoids (Di Marzo 1998; Mechoulam et al 1998). Endocannabinoids function as cannabinoid receptor agonists and exhibit a variety of cannabimimetic activities, including inhibition of adenylyl cyclase activity, inhibition of voltage-gated calcium channels, sedation, catalepsy, analgesia and hypothermia (Di Marzo 1998; Howlett and Mukhopadhyay 2000). Since endocannabinoids as well as eicosanoids are generated from membrane phospholipids on demand (Di Marzo 1998), the regulation of the enzymes involved in the biosynthesis and degradation of endocannabinoids should play crucial roles for the control of the endocannabinoid levels. In this chapter we will focus on the enzymes related to the metabolism of anandamide.
enzymes involved in the anandamide metabolism we briefly mention the occurrence and biological activities of N-acylethanolamines other than anandamide. Prior to the discovery of anandamide, saturated and monounsaturated N-acylethanolamines have been isolated from animal tissues (Bachur et al 1965; Schmid et al 1990). N-Acylethanolamines and their precursors, N-acylphosphatidylethanolamines (N-acyl PEs; see below), have received attention for several reasons. First, these compounds remarkably increase in degenerating tissues and cells in conditions such as myocardial infarction (Epps et al 1979, 1980), glutamate-induced neuronal cytotoxicity (Hansen et al 1995), post-decapitative brain ischaemia (Schmid et al 1995; Moesgaard et al 2000), cadmium-induced testicular inflammation (Kondo et al 1998), and irradiation of epidermal cells by UV light (Berdyshev et al 2000). Second, N-palmitoylethanolamine, one of the most abundant N-acylethanolamines in animal tissues, has been an attractive target in pharmacological studies due to its various biological activities such as antiinflammatory, immunosuppressive, analgesic, neuroprotective, membrane-stabilizing, and antioxidant effects (for reviews, see Schmid et al 1990; Lambert and Di Marzo 1999; Lambert et al 2002). These findings suggest that N-acylethanolamines play a protective role in degenerating tissues. N-Oleoylethanolamine was also reported to suppress food intake (Rodriguez de Fonseca et al 2001). Although the mechanisms for these biological activities remain unidentified (Lambert and Di Marzo 1999; Lambert et al 2002), saturated and monounsaturated N-acylethanolamines were recently reported to stimulate phosphorylation of extracellular signal-regulated protein kinase (ERK) independent of cannabinoid receptor and enhanced activator protein-1 (AP-1)-dependent transcriptional activity (Berdyshev et al 2001).
OCCURRENCE AND BIOLOGICAL ACTIVITIES OF LONG-CHAIN N-ACYLETHANOLAMINES
As one of the biosynthetic pathways of long-chain N-acylethanolamines, including anandamide, several groups reported their enzymatic formation by the condensation of fatty acid with ethanolamine (Bachur and Udenfriend 1966; Schmid et al 1985; Deutsch and Chin 1993; Devane and Axelrod 1994; Kruszka and Gross 1994; Sugiura et al 1996a). It was suggested that the condensation was a reverse reaction of the hydrolysis catalysed by fatty acid amide hydrolase (FAAH; Schmid et al 1985; Ueda et al 1995a) (Figure 15.2). Later, it was demonstrated by the use of recombinant FAAH preparations (Arreaza et al 1997; Kurahashi
Anandamide is usually detected in animal tissues and cells, along with much larger amounts of saturated and monounsaturated long-chain N-acylethanolamines, which are inactive with cannabinoid receptors (Hansen et al 2000; Schmid 2000; Schmid and Berdyshev 2002). Since the biosynthetic and degradative pathways are generally thought to be common to various long-chain Nacylethanolamines including anandamide, before describing the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
BIOSYNTHETIC ROUTES OF ANANDAMIDE AND OTHER N-ACYLETHANOLAMINES
180
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 15.1 Cannabinoid and endocannabinoids
et al 1997; Katayama et al 1999). However, due to a very high Km value for ethanolamine (27–50 mM), this reverse reaction did not appear to be physiologically significant. Several lines of evidence reveal that long-chain N-acylethanolamines are generally biosynthesized in vivo in the ‘N-acylationphosphodiesterase pathway’ (Figure 15.2), which was recently reviewed in detail (Hansen et al 2000; Schmid 2000; Sugiura et al 2002). Anandamide was also shown to be formed through the same pathway (Di Marzo et al 1994; Sugiura et al 1996a, 1996b). The pathway starts from ethanolamine phospholipid and is composed of two successive reactions catalysed by N-acyltransferase and phosphodiesterase.
activity than adult rats (Natarajan et al 1986; Moesgaard et al 2000). When heart tissues of various animal species were compared, the N-acyltransferase activity was high in dog and cat while it was very low in other animals, such as human, rat and guinea-pig (Moesgaard et al 2002). Interestingly, the agedependency in the brain and the special difference in the heart, were different from those of the phosphodiesterase described below. Although the enzyme was solubilized from the particulate fraction of rat brain with 0.5% NP-40 and partially purified by a Mono Q column (Cadas et al 1997), molecular characterization of the enzyme has not yet been performed.
PHOSPHODIESTERASE N-ACYLTRANSFERASE The first reaction, catalysed by N-acyltransferase, is the transfer of a fatty acyl chain to the amino group of an ethanolamine phospholipid, generating N-acyl PE. The acyl chain at the sn-1 position of phosphatidylcholine, 1-acyl-lysophosphatidylcholine, PE and cardiolipin is utilized as acyl donor rather than free fatty acid or acyl-CoA (Natarajan et al 1982, 1983; Reddy et al 1983a, 1983b). It seems that the enzyme has no substrate specificity in terms of the fatty acid species of the sn-1 position of the acyl donor (Sugiura et al 1996a; Cadas et al 1997). Various ethanolamine phospholipids (diacyl-PE, alkylacyl-PE, plasmalogen and lyso-PE) can be substrates as acyl acceptors. The Nacyltransferase is a membrane-bound protein, and in canine brain the activity was the highest in the microsomal subcellular fraction, followed by synaptosomes and mitochondria (Natarajan et al 1983). The optimal pH was alkaline in most cases. ATP was not necessary for the transacylation, but Ca2+ could stimulate the activity and it could be replaced by Sr2+, Mn2+ and Ba2+ (Natarajan et al 1986). Although stimulation of cultured cells by calcium ionophores often leads to accumulation in N-acyl PE and N-acylethanolamines, it is unclear whether or not the enzyme is activated directly by the increased intracellular Ca2+ (Schmid 2000). We should also note that concentrations as high as millimolar levels of Ca2+ were required for the full activity. It was reported that cAMP potentiated the ionomycin-stimulated generation of N-acyl PE (Cadas et al 1996). Sulphhydryl reagents, alkylating reagents and several cations, including Be2+, Zn2+ and Ag+, inhibited the enzyme (Reddy et al 1984; Cadas et al 1997). Regarding the organ distribution in rats, the enzyme activity was the highest in brain, followed by testis, muscle and many other organs (Cadas et al 1997). The enzyme was widely distributed in rat brain regions, with the highest activity in the brain stem (Cadas et al 1997). In rat brain the enzyme activity changed during development, i.e. infant rats showed a several-fold higher
The second step of the anandamide biosynthesis is catalysed by a phosphodiesterase of the phospholipase D type (Figure 15.2). This enzyme was also membrane-bound, and the potency of the enzyme activity in subcellular fractions was in the order of microsomes 4 mitochondria in rat heart (Schmid et al 1983) or microsomes 4 synaptosomes 4 mitochondria in canine brain (Natarajan et al 1984). Although the enzyme hydrolysed N-acyl lyso-PE and glycerophospho(N-acyl) ethanolamine in addition to the diacyl type and plasmalogen type of N-acyl PE, the enzyme appeared to be almost inactive with phosphatidylcholine and PE (Schmid et al 1983; Natarajan et al 1984). The enzyme did not catalyse transphosphatidylation, a well-known reaction of the phospholipase Ds, in which phosphatidylalcohol is generated rather than phosphatidic acid in the presence of alcohol such as butanol and ethanol. Furthermore, none of the known activators reported for the previously characterized phospholipase Ds were effective with this enzyme. Based on these results, the enzyme appeared to be distinct from the known phospholipase D (Petersen and Hansen 1999). The enzyme did not discriminate between the fatty acid species of the N-acyl group (Sugiura et al 1996a, 1996b). Triton X-100 was found to be a potent activator, while Ca2+ was inhibitory or only slightly stimulatory with the microsomal enzyme (Schmid et al 1983; Natarajan et al 1984; Sugiura et al 1996a; Petersen and Hansen 1999). Bile salts, some other detergents (Tween 20 and cetyltrimethylammonium bromide), Zn2+ (Schmid et al 1983) and sulphhydryl reagent (p-chloromerucuriphenylsulphonyl fluoride) (Sasaki and Chang 1997) were inhibitory. Recently we solubilized this enzyme from the particulate fraction of rat heart with 1% octylglucoside and partially purified the enzyme by ion-exchange chromatography (Ueda et al 2001a; Liu et al 2002). After this partial purification we found that the specific enzyme activity increased to 50 nmol/min/mg protein at 378C with N-palmitoyl PE as substrate. The optimal pH was
BIOSYNTHESIS AND DEGRADATION OF ANANDAMIDE
181
Figure 15.2 Biosynthetic and degradative pathways of anandamide. PE, phosphatidylethanolamine; PC, phosphatidylcholine; PA, phosphatidic acid
around 7.5 and Triton X-100 was stimulatory. In contrast to the membrane-bound enzyme, the solubilized enzyme was highly stimulated in a dose-dependent manner by millimolar concentrations of Ca2+. Several divalent cations (Co2+, Mg2+, Mn2+, Ba2+, Sr2+ and Ni2+) could replace Ca2+, while other divalent cations (Hg2+, Cu2+, Fe2+ and Zn2+) were inhibitory (Ueda et al 2001a). Monovalent cations (Na+, K+ and Li+) hardly affected the activity. We also found that polyamines such as spermine, spermidine, and putrescine functioned as activators (Liu et al 2002). Spermine was the most potent, with an EC50 of about 0.1 mM. However, a synergistic effect of spermine and the activators Ca2+ or Triton X-100 was not observed. Ca2+ and spermine also activated the conversion of N-arachidonoyl PE to anandamide, catalysed by this enzyme. However, the mechanism of the activation by these cations and its physiological significance remain unclear. Interestingly, Km values for N-palmitoyl PE were considerably different between the Ca2+-stimulated enzyme and the Triton X-100-stimulated enzyme (8 mM vs. 45 mM). Organ
distribution of the phosphodiesterase was examined with rats. The highest specific activity was observed in heart, followed by testis and brain. The other organs tested also showed a low but detectable activity (Schmid et al 1983; Liu et al 2002). DEGRADATION OF ANANDAMIDE AND OTHER LONG-CHAIN N-ACYLETHANOLAMINES Anandamide administered to mice is rapidly metabolized in the body (Willoughby et al 1997). The degradation of anandamide and other long-chain N-acylethanolamines is generally attributable to the hydrolysis to fatty acid and ethanolamine, catalysed by an amidase referred to as FAAH (Figure 15.2). Some anandamide analogues, including (R)-methanandamide (Figure 15.3), that are resistant to the hydrolysis by FAAH were reported to reveal potent cannabimimetic activities in vivo (Abadji et al 1994). The central role of FAAH in the anandamide
182
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 15.3 Representative substrates and inhibitors of FAAH. 1Substrate; 2anandamide derivative resistant to FAAH; 3inhibitor
degradation in vivo was strongly confirmed by recent studies with FAAH gene-disrupted mice (Cravatt et al 2001; Lichtman et al 2002). Exogenous anandamide caused intense cannabimimetic behavioural activities, including hypomotility, analgesia, catalepsy and hypothermia in FAAH7/7 mice. Endogenous level of anandamide in the brain of the FAAH7/7 mice was 15-fold higher than that of the wild-type mice (775+113 pmol/g vs. 50+10 pmol/g of tissue). The mice also displayed reduced pain sensitivity in the tail immersion, hot plate, and formalin tests, which was reversed by SR141716A, an antagonist of the CB1 receptor. FAAH is one of the best-characterized of the endocannabinoidrelated enzymes, and a number of reviews focusing on this enzyme were recently published (Ueda and Yamamoto 2000; Ueda et al 2000; Fowler et al 2001; Giuffrida et al 2001; Patricelli and Cravatt 2001a; Bisogno et al 2002; Deutsch et al 2002). We recently found another N-acylethanolamine-hydrolysing enzyme that is active at acidic pH, as described below (Ueda et al 1999, 2001b). CATALYTIC PROPERTIES OF FATTY ACID AMIDE HYDROLASE Before the discovery of anandamide an ‘‘N-acylethanolamine amidohydrolase’’ was extensively characterized, with saturated and monounsaturated N-acylethanolamines as substrates, by Schmid’s group (Schmid et al 1985, 1990). After the characterization of anandamide as an endocannabinoid, several groups examined the catalytic properties of the enzyme hydrolysing anandamide to arachidonic acid and ethanolamine, and it was called an ‘‘amidase’’ (Deutsch and Chin 1993) or ‘‘anandamide amidohydrolase’’ (Desarnaud et al 1995; Ueda et al 1995a) or oleamide hydrolase (Cravatt et al 1995). In 1996 this enzyme was cloned (Cravatt et al 1996). This enzyme, as expressed in COS-7 cells, catalysed the hydrolysis of anandamide as well as the endogenous sleep-inducing factor oleamide, and was called fatty acid amide hydrolase (FAAH). These enzymes described above are now considered to be the same enzyme protein, and FAAH is the most general name, although other names are still used. FAAH is a membrane protein composed of 579 amino acids, and the molecular mass is 63 kDa (Cravatt et al 1996). In earlier studies on the characterization of the enzyme, microsomes or
solubilized proteins were mostly used (Schmid et al 1985; Deutsch and Chin 1993; Desarnaud et al 1995; Hillard et al 1995). We partially purified the enzyme from the particulate fraction of porcine brain by hydrophobic chromatography (Ueda et al 1995a). Later, Cravatt et al highly purified the enzyme from rat liver by affinity chromatography, using an inhibitor as ligand (Cravatt et al 1996). After cDNA cloning, recombinant FAAH became available, and was highly purified from Escherichia coli cells (Patricelli et al 1998) or insect Sf9 cells (Katayama et al 1999). One of the notable catalytic properties of FAAH is its wide substrate specificity toward fatty acid derivatives (Figure 15.3). The enzyme hydrolyses the amide bond of not only various longchain N-acylethanolamines (Schmid et al 1985; Ueda et al 1995a) but also primary fatty acid amides, including oleamide (Maurelli et al 1995; Cravatt et al 1996). Among long-chain N-acylethanolamines, anandamide was the best substrate. As mentioned above, the hydrolysis of anandamide by FAAH is reversible, although the physiological significance of the reverse reaction is unclear. Oleamide did not bind to cannabinoid CB1 receptor, but revealed cannabimimetic behavioural activity in vivo. It was suggested that this effect, referred to as an ‘‘entourage effect’’, was derived from the potentiation of the biological activity of endogenous anandamide by competing with anandamide for FAAH (Mechoulam et al 1997). However, in FAAH-deficient mice, oleamide still displayed cannabimimetic activity (Lichtman et al 2002). Furthermore, the enzyme acted as an esterase for monoacylglycerol, including 2-AG, another endocannabinoid (Di Marzo et al 1998; Goparaju et al 1998; Lang et al 1999), and methyl esters of fatty acids (Kurahashi et al 1997; Patricelli and Cravatt 1999). The physiological significance of the 2-AG hydrolysis by FAAH may be minor, if any, since a monoacylglycerol lipase-like hydrolase appeared to be a major enzyme degrading 2-AG in the porcine brain (Goparaju et al 1999b) and FAAH-deficient mice were shown to have a normal brain level of 2-AG (Lichtman et al 2002). Very recently, the wide distribution of monoacylglycerol lipase in rat brain regions was reported (Dinh et al 2002). The enzyme is active in a range of neutral and alkaline pH, with an optimal pH of 8.5–10. An activating factor has not been reported, although a low concentration of detergent was required for the purified FAAH (Patricelli et al 1998). Development of inhibitors for FAAH has been of great interest due to their potential usefulness as pharmacological tools.
BIOSYNTHESIS AND DEGRADATION OF ANANDAMIDE Serendipitously, phenylmethylsulphonyl fluoride (PMSF), a non-specific inhibitor of serine hydrolases, was found to inhibit FAAH (Deutsch and Chin 1993) and to potentiate the biological activities of anandamide (Childers et al 1994; Adams et al 1995; Pertwee et al 1995; Compton and Martin 1997). A large number of FAAH inhibitors have been reported and they are summarized in the recent reviews (Pertwee 1998; Fowler et al 2001; Patricelli and Cravatt 2001a; Deutsch et al 2002). Among them, methyl arachidonyl fluorophosphonate (MAFP) (De Petrocellis et al 1997; Deutsch et al 1997b), palmitylsulphonyl fluoride (Deutsch et al 1997a) and a-keto heterocycle derivatives of fatty acids (Boger et al 2000) should be noted due to their extreme potency (Figure 15.3). FAAH inhibitors are not necessarily selective, e.g. MAFP was first reported as an inhibitor of cytosolic PLA2 (Huang et al 1993) and later also as an irreversible antagonist of cannabinoid receptor (Deutsch et al 1997b; Fernando and Pertwee 1997). Several reports have revealed a wide distribution of FAAH in mammalian tissues. The organ distribution of the anandamide hydrolysing activity and FAAH mRNA was examined in the rat (Deutsch and Chin 1993; Cravatt et al 1996; Katayama et al 1997). The specific enzyme activity was highest in the liver, followed by small intestine, testis, cerebrum, stomach and others. However, the distribution of FAAH mRNA in human organs was markedly different; it was most abundant in pancreas, brain, kidney and skeletal muscle (Giang and Cravatt 1997). Uterus (Paria et al 1996) and various regions of the eye (Matsuda et al 1997; Bisogno et al 1999; Yazulla et al 1999) were also rich in FAAH. A variety of cell lines, including neuronal and leukaemia cells, which express endogenous FAAH, were utilized for functional analyses of FAAH, as summarized previously (Ueda et al 2000). With respect to the high expression level and functional importance of cannabinoid CB1 receptor in the brain, regional distribution of FAAH in the brain was of great interest. The FAAH activity was relatively high in the hippocampus, cerebral cortex and cerebellum of the rat (Desarnaud et al 1995; Hillard et al 1995; Thomas et al 1997), which were also enriched with the CB1 receptor. However, FAAH-expressing cell types (pyramidal cells in the cerebral cortex and hippocampus, Purkinje cells in the cerebellar cortex, and mitral cells in the olfactory bulb) were different from CB1-expressing cell types. This complementary pattern in the expression of FAAH and CB1 suggested a role for FAAH in regulating the endocannabinoid tone (Egertova et al 1998; Tsou et al 1998). The distribution of FAAH in human brain resembled that in rat brain (Romero et al 2002). FAAH was also expressed in epithelial cells of the choroid plexus, consistent with FAAH regulating the concentration of the sleep inducer oleamide in cerebrospinal fluid (Egertova et al 2000). MOLECULAR PROPERTIES OF FAAH FAAH was cloned first from rat liver (Cravatt et al 1996) and later from mouse liver (Giang and Cravatt 1997), human liver (Giang and Cravatt 1997), and porcine brain (Goparaju et al 1999a). The deduced primary structures are all composed of 579 amino acids, and amino acid identity was greater than 80% among these four species. The sequences are characterized by the presence of an ‘‘amidase signature sequence’’ between amino acids 215–257, which is well conserved in the amidase signature family (Cravatt et al 1996). The sequence between residues 9 and 29 appeared to be involved in self-association of the FAAH protein (Patricelli et al 1998). A deletion mutant lacking amino acids 307–315, containing an Src homology 3 (SH3)-binding domain, was immunocytochemically distributed diffusely in the cell, in contrast to the perinuclear localization of the wild-type (Arreaza and Deutsch 1999). Catalytically important amino acid
183
residues of FAAH were identified by site-directed mutagenesis experiments. Point mutation of Ser-217 or Ser-241 (both of which were highly conserved in the amidase signature sequence) to alanine resulted in complete losses of the catalytic activity and the binding activity to 14C-diisopropyl fluorophosphate (Goparaju et al 1999a; Omeir et al 1999; Patricelli et al 1999). Affinity-labelling of FAAH with ethoxy oleoyl fluorophosphonate identified Ser-241 as a catalytic nucleophile (Patricelli et al 1999), confirming FAAH as one of the serine hydrolases. On the other hand, the mutation of Ser-218 (another highly conserved serine residue) to alanine led to a marked decrease (Goparaju et al 1999a; Patricelli et al 1999) or slight reduction (Omeir et al 1999) in the catalytic activity. Moreover, the K142A and R243A mutants also exhibited a more than 100-fold reduced catalytic activity without impacting the structural integrity of the protein (Patricelli and Cravatt 2000). A catalytic triad composed of serine, histidine, and aspartic acid residues is generally found as the active site in serine hydrolases. However, FAAH did not seem to fit this model, since mutation of the highly conserved histidine residues did not abolish the activity (Patricelli et al 1999). Instead, Lys-142 was revealed to be a catalytic base (Patricelli and Cravatt 1999). Interestingly, the mutation of this lysine residue to alanine brought a more than 500-fold higher esterase activity with methyl oleate than its amidase activity with oleamide (Patricelli and Cravatt 1999). Ile-491 was shown to participate in hydrophobic interaction with medium-chain substrates (Patricelli and Cravatt 2001b). Very recently, a single nucleotide polymorphism of human FAAH was discovered (Sipe et al 2002). This abnormality resulted in a missense mutation, causing the conversion of Pro-129 to threonine. The variant enzyme displayed a normal catalytic activity but an enhanced sensitivity to proteolysis. Interestingly, it was proposed that this genetic mutation was significantly associated with street drug use and drug/alcohol abuse. The genes of the human and mouse FAAH were assigned to chromosome 1p34-p35 and chromosome 4, respectively, and clarified to contain 15 exons (Wan et al 1998). The 5’-flanking region of the mouse FAAH gene was isolated, and multiple or single transcription site(s) were determined (Puffenbarger et al 2001; Waleh et al 2002). The molecular size of the FAAH transcript was estimated as 2.6 kb. The region lacked an obvious TATA-box, and contained putative transcription factor-binding sites, including oestrogen response elements and glucocorticoid response elements (Puffenbarger et al 2001; Waleh et al 2002). Related to this finding, the expression of FAAH was downregulated in mouse uterus by sex hormones (Maccarrone et al 2000) and upregulated in human lymphocytes by progesterone (Maccarrone et al 2001b). Interestingly, transfection of human neuroblastoma cells (SY5Y) with oestrogen receptor or glucocorticoid receptor resulted in ligand-independent downregulation of the transcriptional activity of the FAAH gene (Waleh et al 2002). Recently, ovarian steroid hormones were suggested to upregulate the expression of FAAH mRNA in epithelial cells and circular myometrium of the non-pregnant rat uterus (Xiao et al 2002). Furthermore, FAAH was downregulated by lipopolysaccharide in human peripheral lymphocytes (Maccarrone et al 2001a) and by N-palmitoylethanolamine in human breast cancer cells (Di Marzo et al 2001). The promoter activity of the FAAH gene also showed tissue specificity; the 5’-flanking region was active in N18TG2 neuroblastoma cells and C6 glioma cells, which express FAAH natively, while it was inactive in C2C12 and L6 myogenic cells that lack the FAAH activity (Puffenbarger et al 2001). As monitored during development in rat whole brain, the expression level of FAAH mRNA increased progressively from embryonic day 14 until postnatal day 10, when maximal levels were observed (Thomas et al 1997).
184
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Prior to the intracellular hydrolysis by FAAH, anandamide must enter the cells. This uptake is believed to be mediated by facilitated diffusion, although the transporter protein has not been identified (for review, see Hillard and Jarrahian 2000; Giuffrida et al 2001). It was recently shown that the cellular uptake of anandamide was functionally coupled to its intracellular hydrolysis by FAAH (Deutsch et al 2001; Day et al 2001), i.e. uptake of anandamide into the FAAH-expressing cells was decreased by FAAH inhibitors such as MAFP. On the other hand, uptake into the cells lacking endogenous FAAH was increased by transfection with the cDNA encoding FAAH. These results revealed that the intracellular degradation of anandamide by FAAH drives its cellular uptake. Recently, the existence of an anandamide transporter in certain cell types has been challenged (Glaser et al 2003; Fasia et al 2003). AN ACID AMIDASE THAT HYDROLYSES ANANDAMIDE AND OTHER N-ACYLETHANOLAMINES Although FAAH-deficient mice demonstrated a central role of FAAH in the degradation of anandamide, there was still a possibility that an enzyme(s) other than FAAH was involved in the hydrolysis of anandamide and other N-acylethanolalmines. We recently reported that an enzyme catalytically distinguishable from FAAH hydrolysed anandamide in a human megakaryoblastic cell line (CMK; Ueda et al 1999). We observed that the intact CMK cells could convert exogenous anandamide to polar and non-polar lipids. However, the cell homogenate could not hydrolyse anandamide at pH 9, which is optimal for FAAH; the hydrolysis occurred only at acidic pH. When subcellular distribution of the hydrolase was examined with CMK cells, the highest activity was seen in the 12 0006g pellet. The enzyme was readily solubilized from this fraction by freezing and thawing without detergent. The solubilized enzyme was most active around pH 5 and was almost completely inactive at alkaline pH. Dithiothreitol dose-dependently enhanced the anandamide hydrolase activity up to six-fold. In contrast, FAAH of rat basophilic leukaemia cells (RBL-1) was not stimulated with dithiothreitol. When two FAAH inhibitors (PMSF and MAFP) were tested, their IC50 values for the inhibition of the CMK cell enzyme were 3 mM and 410 mM, respectively, while FAAH of RBL-1 cells was inhibited by PMSF and MAFP, with IC50 values of 20 mM and 3 nM under the same conditions. Its substrate specificity was also different from that of FAAH; the CMK cell enzyme hydrolysed N-palmitoylethanolamine 1.5-fold faster than anandamide. Based on these results, the presence of an anandamide-hydrolysing enzyme distinct from FAAH was strongly suggested. The subcellular distribution and pH dependence also suggested that the enzyme was one of the lysosomal hydrolases. We next investigated whether or not this enzyme is physiologically expressed in rats (Ueda et al 2001b). We solubilized proteins from the 12 0006g pellet of various rat organs by freezing and thawing, and incubated the solubilized proteins with N-palmitoylethanolamine at pH 5 in the presence of 1 mM MAFP. The results revealed wide distribution of the enzyme, and the specific enzyme activity was highest in the lung, followed by spleen, small intestine, thymus and caecum. Isolated rat alveolar and peritoneal macrophages also displayed a high activity. This distribution in rat tissues was clearly different from that of FAAH. The enzyme was purified from the solubilized proteins of rat lung to a specific activity of about 1.8 mmol/min/mg protein by acid treatment and a series of chromatographic steps using Phenyl-Sepharose, HiTrap Heparin, hydroxyapatite and HiTrap
Butyl. As analysed by sodium dodecylsulphate-polyacrylamide gel electrophoresis, the purified enzyme gave a major protein band at a molecular mass of 31 kDa. The enzyme activity was enhanced by Triton X-100, with an optimal concentration of 0.1–0.3%. When the purified enzyme was allowed to react with various N-acylethanolamines in the presence of 0.1% Triton X-100, N-palmitoylethanolamine was the most active, followed by N-myristoylethanolamine. Anandamide was hydrolysed at about 10% of the rate of the N-palmitoylethanolamine hydrolysis. The reactivity of the enzyme with saturated Nacylethanolamines was markedly reduced in the absence of Triton X-100. However, the hydrolytic rates with unsaturated Nacylethanolamines, including anandamide, decreased only slightly by the removal of Triton X-100. Based on the different patterns of tissue distribution and substrate specificity, this enzyme may play a unique role different from that of FAAH. Especially, its high expression level in macrophages suggests the involvement in inflammation and immunity. OXYGENATION OF ANANDAMIDE Apart from hydrolysis, the arachidonate moiety of anandamide can be directly oxygenated by the oxygenases involved in the arachidonate cascade (Burstein et al 2000; Kozak and Marnett 2002). 12-Lipoxygenase of mammalian tissues and 15-lipoxygenase of mammalian tissues and soybean oxygenated the C-12 or C-15 position of the arachidonate moiety to generate 12- or 15-hydroperoxy-anandamide, respectively (Ueda et al 1995b; Hampson et al 1995; Edgemond et al 1998). Anandamide was inactive with leukocyte 5-lipoxygenase (Ueda et al 1995b), but was converted predominantly to the 11S-hydroperoxy derivative by plant 5-lipoxygenases (van Zadejhoff et al 1998). Cytochrome P-450 also converted anandamide to a variety of oxygenated compounds (Bornheim et al 1993, 1995). Furthermore, cyclooxygenase-2 oxygenated anandamide, producing prostaglandin ethanolamides (Yu et al 1997). Prostaglandin E2 ethanolamide was stable in plasma and was slowly converted to prostaglandin B2 ethanolamide (Kozak et al 2001). However, the physiological significance of the metabolism of anandamide in the arachidonate cascade is not well understood. PERSPECTIVES Since the discovery of anandamide as an endocannabinoid, its biosynthetic and degradative pathways have been extensively studied. The greatest progress was achieved in the catalytic, molecular biological and pharmacological studies on FAAH. Thus, the central role of FAAH in the inactivation of anandamide was confirmed. Further analysis of FAAH-deficient mice and development of specific FAAH inhibitors will contribute to the elucidation of physiological significance of anandamide and other fatty acid amides. The second enzyme hydrolysing anandamide should also be investigated more extensively. The enzymes involved in the biosynthesis of anandamide await further catalytic and molecular biological analyses. REFERENCES Abadji V, Lin S, Taha G et al (1994) (R)-Methanandamide: a chiral novel anandamide possessing higher potency and metabolic stability. J Med Chem, 37, 1889–1893. Adams IB, Ryan W, Singer M et al (1995) Evaluation of cannabinoid receptor binding and in vivo activities for anandamide analogs. J Pharmacol Exp Ther, 273, 1172–1181.
BIOSYNTHESIS AND DEGRADATION OF ANANDAMIDE Arreaza G and Deutsch DG (1999) Deletion of a proline-rich region and a transmembrane domain in fatty acid amide hydrolase. FEBS Lett, 454, 57–60. Arreaza G, Devane WA, Omeir RL et al (1997) The cloned rat hydrolytic enzyme responsible for the breakdown of anandamide also catalyses its formation via the condensation of arachidonic acid and ethanolamine. Neurosci Lett, 234, 59–62. Bachur NR and Udenfriend S (1966) Microsomal synthesis of fatty acid amides. J Biol Chem, 241, 1308–1313. Bachur NR, Masek K, Melmon KL and Udenfriend S (1965) Fatty acid amides of ethanolamine in mammalian tissues. J Biol Chem, 240, 1019– 1024. Berdyshev EV, Schmid PC, Dong Z and Schmid HHO (2000) Stressinduced generation of N-acylethanolamines in mouse epidermal JB6 P+ cells. Biochem J, 346, 369–374. Berdyshev EV, Schmid PC, Krebsbach RJ, et al (2001) Cannabinoidreceptor-independent cell signalling by N-acylethanolamines. Biochem J, 360, 67–75. Bisogno T, Delton-Vandenbroucke I, Milone A et al (1999) Biosynthesis and inactivation of N-arachidonoylethanolamine (anandamide) and Ndocosahexaenoylethanolamine in bovine retina. Arch Biochem Biophys, 370, 300–307. Bisogno T, De Petrocellis L and Di Marzo V (2002) Fatty acid amide hydrolase, an enzyme with many bioactive substrates. Possible therapeutic implications. Curr Pharm Design, 8, 533–547. Boger DL, Sato H, Lerner AE et al (2000) Exceptionally potent inhibitors of fatty acid amide hydrolase: the enzyme responsible for degradation of endogenous oleamide and anandamide. Proc Natl Acad Sci USA, 97, 5044–5049. Bornheim LM, Kim KY, Chen B and Correia MA (1993) The effect of cannabidiol on mouse hepatic microsomal cytochrome P450dependent anandamide metabolism. Biochem Biophys Res Commun, 197, 740–746. Bornheim LM, Kim KY, Chen B and Correia MA (1995) Microsomal cytochrome P450-mediated liver and brain anandamide metabolism. Biochem Pharmacol, 50, 677–686. Burstein SH, Rossetti RG, Yagen B and Zurier RB (2000) Oxidative metabolism of anandamide. Prostagland Lipid Mediat, 61, 29–41. Cadas H, Gaillet S, Beltramo M et al (1996) Biosynthesis of an endogenous cannabinoid precursor in neurons and its control by calcium and cAMP. J Neurosci, 16, 3934–3942. Cadas H, di Tomaso E and Piomelli D (1997) Occurrence and biosynthesis of endogenous cannabinoid precursor, N-arachidonoyl phosphatidylethanolamine, in rat brain. J Neurosci, 17, 1226–1242. Childers SR, Sexton T and Roy MB (1994) Effects of anandamide on cannabinoid receptors in rat brain membranes. Biochem Pharmacol, 47, 711–715. Compton DR and Martin BR (1997) The effect of the enzyme inhibitor phenylmethylsulfonyl fluoride on the pharmacological effect of anandamide in the mouse model of cannabimimetic activity. J Pharmacol Exp Ther, 283, 1138–1143. Cravatt BF, Prospero-Garcia O, Siuzdak G et al (1995) Chemical characterization of a family of brain lipids that induce sleep. Science, 268, 1506–1509. Cravatt BF, Giang DK, Mayfield SP et al (1996) Molecular characterization of an enzyme that degrades neuromodulatory fattyacid amides. Nature, 384, 83–87. Cravatt BF, Demarest K, Patricelli MP et al (2001) Supersensitivity to anandamide and enhanced endogenous cannabinoid signaling in mice lacking fatty acid amide hydrolase. Proc Natl Acad Sci USA, 98, 9371– 9376. Day TA, Rakhshan F, Deutsch DG and Barker EL (2001) Role of fatty acid amide hydrolase in the transport of the endogenous cannabinoid anandamide. Mol Pharmacol, 59, 1369–1375. De Petrocellis L, Melck D, Ueda N et al (1997) Novel inhibitors of brain, neuronal, and basophilic anandamide amidohydrolase. Biochem Biophys Res Commun, 231, 82–88. Desarnaud F, Cadas H and Piomelli D (1995) Anandamide amidohydrolase activity in rat brain microsomes. Identification and partial characterization. J Biol Chem, 270, 6030–6035. Deutsch DG and Chin SA (1993) Enzymatic synthesis and degradation of anandamide, a cannabinoid receptor agonist. Biochem Pharmacol, 46, 791–796.
185
Deutsch DG, Lin S, Hill WAG et al (1997a) Fatty acid sulfonyl fluorides inhibit anandamide metabolism and bind to the cannabinoid receptor. Biochem Biophys Res Commun, 231, 217–221. Deutsch DG, Omeir R, Arreaza G et al (1997b) Methyl arachidonyl fluorophosphonate: a potent irreversible inhibitor of anandamide amidase. Biochem Pharmacol, 53, 255–260. Deutsch DG, Glaser ST, Howell JM et al (2001) The cellular uptake of anandamide is coupled to its breakdown by fatty-acid amide hydrolase. J Biol Chem, 276, 6967–6973. Deutsch DG, Ueda N and Yamamoto S (2002) The fatty acid amide hydrolase (FAAH). Prostagland Leukotriene Essent Fatty Acids, 66, 201–210. Devane WA and Axelrod J (1994) Enzymatic synthesis of anandamide, an endogenous ligand for the cannabinoid receptor, by brain membranes. Proc Natl Acad Sci USA, 91, 6698–6701. Devane WA, Hanus L, Breuer A et al (1992) Isolation and structure of a brain constituent that binds to the cannabinoid receptor. Science, 258, 1946–1949. Di Marzo V (1998) ‘‘Endocannabinoids’’ and other fatty acid derivatives with cannabimimetic properties: biochemistry and possible physiopathological relevance. Biochim Biophys Acta, 1392, 153–175. Di Marzo V, Fontana A, Cadas H et al (1994) Formation and inactivation of endogenous cannabinoid anandamide in central neurons. Nature, 372, 686–691. Di Marzo V, Bisogno T, Sugiura T et al (1998) The novel endogenous cannabinoid 2-arachidonoylglycerol is inactivated by neuronal- and basophil-like cells: connections with anandamide. Biochem J, 331, 15– 19. Di Marzo V, Melck D, Orlando P et al (2001) Palmitoylethanolamide inhibits the expression of fatty acid amide hydrolase and enhances the anti-proliferative effect of anandamide in human breast cancer cells. Biochem J, 358, 249–255. Dinh TP, Carpenter D, Leslie FM et al (2002) Brain monoglyceride lipase participating in endocannabinoid inactivation. Proc Natl Acad Sci USA, 99, 10819–10824. Edgemond WS, Hillard CJ, Falck JR et al (1998) Human platelets and polymorphonuclear leukocytes synthesize oxygenated derivatives of arachidonylethanolamide (anandamide): their affinities for cannabinoid receptors and pathways of inactivation. Mol Pharmacol, 54, 180–188. Egertova M, Giang DK, Cravatt BF and Elphick MR (1998) A new perspective on cannabinoid signalling: complementary localization of fatty acid amide hydrolase and the CB1 receptor in rat brain. Proc R Soc Lond B, 265, 2081–2085. Egertova M, Cravatt BF and Elphick MR (2000) Fatty acid amide hydrolase expression in rat choroid plexus: possible role in regulation of the sleep-inducing action of oleamide. Neurosci Lett, 282, 13–16. Epps DE, Schmid PC, Natarajan V and Schmid HHO (1979) NAcylethanolamine accumulation in infarcted myocardium. Biochem Biophys Res Commun, 90, 628–633. Epps DE, Natarajan V, Schmid PC and Schmid HHO (1980) Accumulation of N-acylethanolamine glycerophospholipids in infarcted myocardium. Biochim Biophys Acta, 618, 420–430. Fasia L, Karava V and Siafaka-Kapadai A (2003) Uptake and metabolism of [3H]anandamide by rabbit platelets. Eur J Biochem, 270, 3498–3506. Fernando SR and Pertwee RG (1997) Evidence that methyl arachidonyl fluorophosphonate is an irreversible cannabinoid receptor antagonist. Br J Pharmacol, 121, 1716–1720. Fowler CJ, Jonsson K-O and Tiger G (2001) Fatty acid amide hydrolase: biochemistry, pharmacology, and therapeutic possibilities for an enzyme hydrolyzing anandamide, 2-arachidonoylglycerol, palmitoylthanolamide, and oleamide. Biochem Pharmacol, 62, 517–526. Giang DK and Cravatt BF (1997) Molecular characterization of human and mouse fatty acid amide hydrolases. Proc Natl Acad Sci USA, 94, 2238–2242. Giuffrida A, Beltramo M and Piomelli D (2001) Mechanisms of endocannabinoid inactivation: biochemistry and pharmacology. J Pharmacol Exp Ther, 298, 7–14. Glaser ST, Abumrad NA, Fatade F, Kaczocha M, Studholme KM and Deutsch DG (2003) Evidence against the presence of an anandamide transporter. Proc Natl Acad Sci USA, 100, 4269–4274.
186
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Goparaju SK, Ueda N, Yamaguchi H and Yamamoto S (1998) Anandamide amidohydrolase reacting with 2-arachidonoylglycerol, another cannabinoid receptor ligand. FEBS Lett, 422, 69–73. Goparaju SK, Kurahashi Y, Suzuki H, Ueda N and Yamamoto S (1999a) Anandamide amidohydrolase of porcine brain: cDNA cloning, functional expression and site-directed mutagenesis. Biochim Biophys Acta, 1441, 77–84. Goparaju SK, Ueda N, Taniguchi K and Yamamoto S (1999b) Enzymes of porcine brain hydrolyzing 2-arachidonoylglycerol, an endogenous ligand of cannabinoid receptors. Biochem Pharmacol, 57, 417–423. Hampson AJ, Hill WAG, Zan-Phillips M et al (1995) Anandamide hydroxylation by brain lipoxygenase: metabolite structures and potencies at the cannabinoid receptor. Biochim Biophys Acta, 1259, 173–179. Hansen HS, Lauritzen L, Strand AM et al (1995) Glutamate stimulates the formation of N-acylphosphatidylethanolamine and Nacylethanolamine in cortical neurons in culture. Biochim Biophys Acta, 1258, 303–308. Hansen HS, Moesgaard B, Hansen HH and Petersen G (2000) NAcylethanolamines and precursor phospholipids—relation to cell injury. Chem Phys Lipids, 108, 135–150. Hanus L, Gopher A, Almog S and Mechoulam R (1993) Two new unsaturated fatty acid ethanolamides in brain that bind to the cannabinoid receptor. J Med Chem, 36, 3032–3034. Hanus L, Abu-Lafi S, Fride E et al (2001) 2-Arachidonyl glyceryl ether, an endogenous agonist of the cannabinoid CB1 receptor. Proc Natl Acad Sci USA, 98, 3662–3665. Hillard CJ and Jarrahian A (2000) The movement of Narachidonoylethanolamine (anandamide) across cellular membranes. Chem Phys Lipids, 108, 123–134. Hillard CJ, Wilkison DM, Edgemond WS and Campbell WB (1995) Characterization of the kinetics and distribution of Narachidonylethanolamine (anandamide) hydrolysis by rat brain. Biochim Biophys Acta, 1257, 249–256. Howlett AC and Mukhopadhyay S (2000) Cellular signal transduction by anandamide and 2-arachidonoylglycerol. Chem Phys Lipids, 108, 53– 70. Huang Z, Liu S, Street I et al (1993) Methyl arachidonyl fluorophosphonate, a potent irreversible cPLA2 inhibitor, blocks the mobilization of arachidonic acid in human platelets and neutrophils. Mediat Inflamm, 3, 307–308. Katayama K, Ueda N, Kurahashi Y et al (1997) Distribution of anandamide amidohydrolase in rat tissues with special reference to small intestine. Biochim Biophys Acta, 1347, 212–218. Katayama K, Ueda N, Katoh I and Yamamoto S (1999) Equilibrium in the hydrolysis and synthesis of cannabimimetic anandamide demonstrated by a purified enzyme. Biochim Biophys Acta, 1440, 205–214. Kondo S, Sugiura T, Kodaka T et al (1998) Accumulation of various N-acylethanolamines including N-arachidonoylethanolamine (anandamide) in cadmium chloride-administered rat testis. Arch Biochem Biophys, 354, 303–310. Kozak KR and Marnett LJ (2002) Oxidative metabolism of endocannabinoids. Prostagland Leukotriene Essent Fatty Acids, 66, 211–220. Kozak KR, Crews BC, Ray JL et al (2001) Metabolism of prostaglandin glycerol esters and prostaglandin ethanolamides in vitro and in vivo. J Biol Chem, 276, 36993–36998. Kruszka KK and Gross RW (1994) The ATP- and CoA-independent synthesis of arachidonoylethanolamide. A novel mechanism underlying the synthesis of the endogenous ligand of the cannabinoid receptor. J Biol Chem, 269, 14345–14348. Kurahashi Y, Ueda N, Suzuki H et al (1997) Reversible hydrolysis and synthesis of anandamide demonstrated by recombinant rat fatty-acid amide hydrolase. Biochem Biophys Res Commun, 237, 512–515. Lambert DM and Di Marzo V (1999) The palmitoylethanolamide and oleamide enigmas: are these two fatty acid amides cannabimimetic? Curr. Med. Chem., 6, 757–773. Lambert DM, Vandevoorde S, Jonsson K-O and Fowler CJ (2002) The palmitoylethanolamide family: a new class of anti-inflammatory agents? Curr Med Chem, 9, 663–674. Lang W, Qin C, Lin S et al (1999) Substrate specificity and stereoselectivity of rat brain microsomal anandamide amidohydrolase. J Med Chem, 42, 896–902.
Lichtman AH, Hawkins EG, Griffin G and Cravatt BF (2002) Pharmacological activity of fatty acid amides is regulated, but not mediated, by fatty acid amide hydrolase in vivo. J Pharmacol Exp Ther, 302, 73–79. Liu Q, Tonai T and Ueda N (2002) Activation of N-acylethanolaminereleasing phospholipase D by polyamines. Chem Phys Lipids, 115, 77– 84. Maccarrone M, De Felici M, Bari M et al (2000) Downregulation of anandamide hydrolase in mouse uterus by sex hormones. Eur J Biochem, 267, 2991–2997. Maccarrone M, De Petrocellis L, Bari M et al (2001a) Lipopolysaccharide downregulates fatty acid amide hydrolase expression and increases anandamide levels in human peripheral lymphocytes. Arch Biochem Biophys, 393, 321–328. Maccarrone M, Valensise H, Bari M et al (2001b) Progesterone upregulates anandamide hydrolase in human lymphocytes: role of cytokines and implications for fertility. J Immunol, 166, 7183–7189. Matsuda LA, Lolait SJ, Brownstein MJ et al (1990) Structure of a cannabinoid receptor and functional expression of the cloned cDNA. Nature, 346, 561–564. Matsuda S, Kanemitsu N, Nakamura A et al (1997) Metabolism of anandamide, an endogenous cannabinoid receptor ligand, in porcine ocular tissues. Exp Eye Res, 64, 707–711. Maurelli S, Bisogno T, De Petrocellis L et al (1995) Two novel classes of neuroactive fatty acid amides are substrates for mouse neuroblastoma ‘anandamide amidohydrolase’. FEBS Lett, 377, 82–86. Mechoulam R, Ben-Shabat S, Hanus L et al (1995) Identification of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. Biochem Pharmacol, 50, 83–90. Mechoulam R, Fride E, Hanus L et al (1997) Anandamide may mediate sleep induction. Nature, 389, 25–26. Mechoulam R, Fride E and Di Marzo V (1998) Endocannabinoids. Eur J Pharmacol, 359, 1–18. Moesgaard B, Petersen G, Jaroszewski JW and Hansen HS (2000) Agedependent accumulation of N-acyl-ethanolamine phospholipids in ischemic rat brain: a 31P-NMR and enzyme activity study. J Lipid Res, 41, 985–990. Moesgaard B, Petersen G, Mortensen SA and Hansen HS (2002) Substantial species differences in relation to formation and degradation of N-acyl-ethanolamine phospholipids in heart tissue: an enzyme activity study. Comp Biochem Physiol B, 131, 475–482. Munro S, Thomas KL and Abu-Shaar M (1993) Molecular characterization of a peripheral receptor for cannabinoids. Nature, 365, 61–65. Natarajan V, Reddy PV, Schmid PC and Schmid HHO (1982) NAcylation of ethanolamine phospholipids in canine myocardium. Biochim Biophys Acta, 712, 342–355. Natarajan V, Schmid PC, Reddy PV et al (1983) Biosynthesis of Nacylethanolamine phospholipids by dog brain preparations. J Neurochem, 41, 1303–1312. Natarajan V, Schmid PC, Reddy PV and Schmid HHO (1984) Catabolism of N-acylethanolamine phospholipids by dog brain preparations. J Neurochem, 42, 1613–1619. Natarajan V, Schmid PC and Schmid HHO (1986) N-Acylethanolamine phospholipid metabolism in normal and ischemic rat brain. Biochim Biophys Acta, 878, 32–41. Omeir RL, Arreaza G and Deutsch DG (1999) Identification of two serine residues involved in catalysis by fatty acid amide hydrolase. Biochem Biophys Res Commun, 264, 316–320 (published erratum, 2000 267, 676). Paria BC, Deutsch DG and Dey SK (1996) The uterus is a potential site for anandamide synthesis and hydrolysis: differential profiles of anandamide synthase and hydrolase activities in the mouse uterus during the periimplantation period. Mol Reprod Dev, 45, 183–192. Patricelli MP and Cravatt BF (1999) Fatty acid amide hydrolase competitively degrades bioactive amides and esters through a nonconventional catalytic mechanism. Biochemistry, 38, 14125–14130. Patricelli MP and Cravatt BF (2000) Clarifying the catalytic roles of conserved residues in the amidase signature family. J Biol Chem, 275, 19177–19184. Patricelli MP and Cravatt BF (2001a) Proteins regulating the biosynthesis and inactivation of neuromodulatory fatty acid amides. Vit Horm, 62, 95–131.
BIOSYNTHESIS AND DEGRADATION OF ANANDAMIDE Patricelli MP and Cravatt BF (2001b) Characterization and manipulation of the acyl chain selectivity of fatty acid amide hydrolase. Biochemistry, 40, 6107–6115. Patricelli MP, Lashuel HA, Giang DK et al (1998) Comparative characterization of a wild-type and transmembrane domain-deleted fatty acid amide hydrolase: identification of the transmembrane domain as a site for oligomerization. Biochemistry, 37, 15177–15187. Patricelli MP, Lovato MA and Cravatt BF (1999) Chemical and mutagenic investigations of fatty acid amide hydrolase: evidence for a family of serine hydrolases with distinct catalytic properties. Biochemistry, 38, 9804–9812. Pertwee RG (1997) Pharmacology of cannabinoid CB1 and CB2 receptors. Pharmacol Ther, 74, 129–180. Pertwee RG (1998) Pharmacological, physiological and clinical implications of the discovery of cannabinoid receptors. Biochem Soc Trans, 26, 267–272. Pertwee RG, Fernando SR, Griffin G et al (1995) Effect of phenylmethylsulphonyl fluoride on the potency of anandamide as an inhibitor of electrically evoked contractions in two isolated tissue preparations. Eur J Pharmacol, 272, 73–78. Petersen G and Hansen HS (1999) N-Acylphosphatidylethanolaminehydrolysing phospholipase D lacks the ability to transphosphatidylate. FEBS Lett, 455, 41–44. Puffenbarger RA, Kapulina O, Howell JM and Deutsch DG (2001) Characterization of the 5’-sequence of the mouse fatty acid amide hydrolase. Neurosci Lett, 314, 21–24. Reddy PV, Natarajan V, Schmid PC and Schmid HHO (1983a) NAcylation of dog heart ethanolamine phospholipids by transacylase activity. Biochim Biophys Acta, 750, 472–480. Reddy PV, Schmid PC, Natarajan V and Schmid HHO (1983b) The role of cardiolipin as an acyl donor in dog heart N-acylethanolamine phospholipid biosynthesis. Biochim Biophys Acta, 751, 241–246. Reddy PV, Schmid PC, Natarajan V et al (1984) Properties of canine myocardial phosphatidylethanolamine N-acyltransferase. Biochim Biophys Acta, 795, 130–136. Rodriguez de Fonseca F, Navarro M, Gomez R et al (2001) An anorexic lipid mediator regulated by feeding. Nature, 414, 209–212. Romero J, Hillard CJ, Calero M and Rabano A (2002) Fatty acid amide hydrolase localization in the human central nervous system: an immunohistochemical study. Mol Brain Res, 100, 85–93. Sasaki T and Chang MCJ (1997) N-Arachidonylethanolamine (anandamide) formation from N-arachidonylphosphatidylethanolamine in rat brain membranes. Life Sci, 61, 1803–1810. Schmid HHO (2000) Pathways and mechanisms of N-acylethanolamine biosynthesis: can anandamide be generated selectively? Chem Phys Lipids, 108, 71–87. Schmid HHO and Berdyshev EV (2002) Cannabinoid receptor-inactive Nacylethanolamines and other fatty acid amides: metabolism and function. Prostagland Leukotriene Essent Fatty Acids, 66, 363–376. Schmid HHO, Schmid PC and Natarajan V (1990) N-Acylated glycerophospholipids and their derivatives. Prog Lipid Res, 29, 1–43. Schmid PC, Reddy PV, Natarajan V and Schmid HHO (1983) Metabolism of N-acylethanolamine phospholipids by a mammalian phosphodiesterase of the phospholipase D type. J Biol Chem, 258, 9302–9306. Schmid PC, Zuzarte-Augustin ML and Schmid HHO (1985) Properties of rat liver N-acylethanolamine amidohydrolase. J Biol Chem, 260, 14145–14149. Schmid PC, Krebsbach RJ, Perry SR et al (1995) Occurrence and postmortem generation of anandamide and other long-chain Nacylethanolamines in mammalian brain. FEBS Lett, 375, 117–120 (published correction, 1996 385, 125–126). Sipe JC, Chiang K, Gerber AL et al (2002) A missense mutation in human fatty acid amide hydrolase associated with problem drug use. Proc Natl Acad Sci USA, 99, 8394–8399. Sugiura T, Kondo S, Sukagawa A et al (1995) 2-Arachidonoylglycerol: a possible endogenous cannabinoid receptor ligand in brain. Biochem Biophys Res Commun, 215, 89–97.
187
Sugiura T, Kondo S, Sukagawa A et al (1996a) Transacylase-mediated and phosphodiesterase-mediated synthesis of N-arachidonoylethanolamine, an endogenous cannabinoid-receptor ligand, in rat brain microsomes. Comparison with synthesis from free arachidonic acid and ethanolamine. Eur J Biochem, 240, 53–62. Sugiura T, Kondo S, Sukagawa A et al (1996b) Enzymatic synthesis of anandamide, an endogenous cannabinoid receptor ligand, through Nacylphosphatidylethanolamine pathway in testis: involvement of Ca2+-dependent transacylase and phosphodiesterase activities. Biochem Biophys Res Commun, 218, 113–117. Sugiura T, Kobayashi Y, Oka S and Waku K (2002) Biosynthesis and degradation of anandamide and 2-arachidonoylglycerol and their possible physiological significance. Prostagland Leukotriene Essent Fatty Acids, 66, 173–192. Thomas EA, Cravatt BF, Danielson PE et al (1997) Fatty acid amide hydrolase, the degradative enzyme for anandamide and oleamide, has selective distribution in neurons within the rat central nervous system. J Neurosci Res, 50, 1047–1052. Tsou K, Nogueron MI, Muthian S et al (1998) Fatty acid amide hydrolase is located preferentially in large neurons in the rat central nervous system as revealed by immunohistochemistry. Neurosci Lett, 254, 137– 140. Ueda N and Yamamoto S (2000) Anandamide amidohydrolase (fatty acid amide hydrolase). Prostagland Lipid Mediat, 61, 19–28. Ueda N, Kurahashi Y, Yamamoto S and Tokunaga T (1995a) Partial purification and characterization of the porcine brain enzyme hydrolyzing and synthesizing anandamide. J Biol Chem, 270, 23823– 23827. Ueda N, Yamamoto K, Yamamoto S et al (1995b) Lipoxygenasecatalysed oxygenation of arachidonylethanolamide, a cannabinoid receptor agonist. Biochim Biophys Acta, 1254, 127–134. Ueda N, Yamanaka K, Terasawa Y and Yamamoto S (1999) An acid amidase hydrolyzing anandamide as an endogenous ligand for cannabinoid receptors. FEBS Lett, 454, 267–270. Ueda N, Puffenbarger RA, Yamamoto S and Deutsch DG (2000) The fatty acid amide hydrolase (FAAH). Chem Phys Lipids, 108, 107–121. Ueda N, Liu Q and Yamanaka K (2001a) Marked activation of the Nacylphosphatidylethanolamine-hydrolyzing phosphodiesterase by divalent cations. Biochim Biophys Acta, 1532, 121–127. Ueda N, Yamanaka K and Yamamoto S (2001b) Purification and characterization of an acid amidase selective for Npalmitoylethanolamine, a putative endogenous antiinflammatory substance. J Biol Chem, 276, 35552–35557. van Zadelhoff G, Veldink GA and Vliegenthart JFG (1998) With anandamide as substrate plant 5-lipoxygenases behave like 11lipoxygenases. Biochem Biophys Res Commun, 248, 33–38. Waleh NS, Cravatt BF, Apte-Deshpande A et al (2002) Transcriptional regulation of the mouse fatty acid amide hydrolase gene. Gene, 291, 203–210. Wan M, Cravatt BF, Ring HZ et al (1998) Conserved chromosomal location and genomic structure of human and mouse fatty-acid amide hydrolase genes and evaluation of clasper as a candidate neurological mutation. Genomics, 54, 408–414. Willoughby KA, Moore SF, Martin BR and Ellis EF (1997) The biodisposition and metabolism of anandamide in mice. J Pharmacol Exp Ther, 282, 243–247. Xiao AZ, Zhao YG and Duan EK (2002) Expression and regulation of the fatty acid amide hydrolase gene in the rat uterus during the estrous cycle and peri-implantation period. Mol Hum Reprod, 8, 651–658. Yazulla S, Studholme KM, Mcintosh HH and Deutsch DG (1999) Immunocytochemical localization of cannabinoid CB1 receptor and fatty acid amide hydrolase in rat retina. J Comp Neurol, 415, 80–90. Yu M, Ives D and Ramesha CS (1997) Synthesis of prostaglandin E2 ethanolamide from anandamide by cyclooxygenase-2. J Biol Chem, 272, 21181–21186.
16 Inhibitors of Eicosanoids K. D. Rainsford Sheffield Hallam University, Sheffield, UK
It may seem odd, but there are inhibitors as well as stimulators of eicosanoids in our diet, in the cosmetics and some of the preparations employed for skin and intestinal conditions, as well as in a wide range of drugs. Furthermore, some agents in the environment, including carcinogens, can affect eicosanoid metabolism. Literally, inhibitors and stimulators of eicosanoid metabolism are all about us. So when it comes to considering the therapeutic actions of the antiinflammatory, analgesic, antipyretic and antithrombotic group of drugs, it is a challenge to see how these agents can act so specifically on pathways of eicosanoid metabolism when there are so many agents in the environment and our diet that can influence this metabolism or could be capable of having similar inhibitory effects, and so could influence the therapeutic responses to these drugs. In all probability the environmental and dietary agents do influence the therapeutic responses of these drugs. What then is the difference in ingesting foods (e.g. high phenolic-containing plants, fish or marine oils or meats, or plants with high concentrations of certain unsaturated fatty acids) and the actions of drugs, such as non-steroidal antiinflammatory drugs (NSAIDs), analgesics (e.g. paracetamol) and corticosteroids, in the therapy of inflammatory and painful conditions? The challenge by nutraceuticals, which are formulations or extracts of some of these naturally-containing plant or animal products for use in place of conventional drugs, has now assumed considerable importance because their inhibitory effects on various pathways of eicosanoid metabolism have been shown by these nutraceuticals. Also, many components have been isolated and shown to have potency as inhibitors of eicosanoid metabolism comparable with, or even greater than, that of established drugs such as the NSAIDs (Shah et al 1999). In this chapter, the actions of these naturally-derived inhibitors of eicosanoid metabolism present in some of the commonly-used foods, nutraceuticals or therapeutic dietary products, cosmetics and skin protectants will be considered, along with those in the wider variety of drugs, including the classical inhibitors of selective pathways of the metabolism of prostaglandins, leukotrienes and thromboxanes that characterize their actions (e.g. corticosteroids, NSAIDs, analgesics and antithrombotics). Much of our present knowledge on the actions of these drugs and the recent development of highly specific or selective inhibitors of prostaglandins that are elicited in response to inflammatory and painful stimuli, via induction of inducible cyclooxygenase (COX-2) and development of the so-called ‘‘COX-2 preferential or selective’’ inhibitors, has come from spectacular advances in the molecular biology of the genes controlling the production and regulation of the cyclooxygenases (Vane et al 1998; Crofford 1997; Dubois et al 1998), details of which will be found elsewhere in this book. Concepts of the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
differential effects of NSAIDs on the activities of the constitutive COX-1 isoform that regulates physiological functions, compared with that on the inducible COX-2 isoform expressed during inflammation and some pain pathways, are outlined in Figure 16.1. Advances in the understanding of the regulation and molecular biological properties of the phospholipases that control the release of arachidonic acid from membrane phospholipids and phospholipid metabolism have provided important insight into how eicosanoids are produced under various physiopathological conditions. As with all advances in the science of regulation, this knowledge of eicosanoid metabolism has gone through stages of complexity before some essential features emerge. Furthermore, knowledge of the mechanisms of inhibition by drugs, natural products or endogenous inhibitors on eicosanoid pathways has yielded some puzzling features and properties. For example, before the advent of the COX isoenzymes, it was thought that aspirin acted as an irreversible inhibitor of cyclooxygenases by acetylating key amino acid residue Ser530 at the entrance to the active site of the enzyme, so blocking the entry of the substrate arachidonic acid to its site of oxidation. While this has now been proven for the COX-1 isoform (e.g. present in platelets), when it came to examining the effects of aspirin on the newly identified inducible isoform, COX-2, it emerged that the acetylation of the homologous serine residue to that in COX-1 led to the conversion of the enzyme to a 15lipoxygenase (Figure 16.2), an unexpected outcome! Moreover, this proved to be a key enzyme involved in the formation of another range of lipids, the aspirin-activated lipoxins or epilipoxins (see Figure 16.3 for their formation) which are related to the lipoxins, formed as shown in Figures 16.4a,b. These have proved to have antiinflammatory and immunoregulatory effects, along with the other naturally occurring lipoxins produced by 50 -, 12- and 15-lipoxygenases (Figure 16.4a,b). Clearly, these advances in knowledge of the actions of inhibitors of eicosanoid metabolism have led to important challenges to the views, such as those that were held in the 1970s and 1980s, that inhibition of prostaglandins explained the actions of NSAIDs or analgesics, as well as other actions of inhibitory agents. These earlier views have been covered in the chapter on eicosanoid inhibitors in the first issue of this book, published in 1988, and the reader is referred to this for historical details. INHIBITORS OF CYCLOOXYGENASES As indicated above, studies of the molecular biology of the cyclooxygenase (COX) enzymes in the late 1980s–early 1990s uncovered the existence of two separate genes that control the
Figure 16.1 Concepts of the pathways of eicosanoid metabolism and the sites of actions of analgesic and antiinflammatory drugs, from Rainsford (2004)
190 BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
EICOSANOID INHIBITORS
191
Figure 16.2 Formation of 15-lipoxygenase activity by inhibition of COX-2 by aspirin. From Xiao et al (1997). Published by the American Chemical Society
production of two COX-isoforms (Vane et al 1998; Crofford 1997). The classical view developed from these studies was that one of these, COX-1, governs the production of physiologically important prostaglandins (PGs), while the other, the inducible form COX-2, is amplified during inflammatory reactions and governs the production of inflammation-related prostaglandins. Since the discovery of these two isoforms of the COXs there has been an immense effort to discover new drugs that could selectively inhibit the inflammation-related COX-2 activity while sparing COX-1, which is important in physiological functions in the gastrointestinal tract, renal, reproductive and haemostatic systems, where inhibition of this enzyme is linked to development of adverse reactions attributed to NSAIDs (Dubois et al 1998; Dannhardt and Kiefer 2001). This has led to substantial investment by the pharmaceutical industry in the development of selective COX-2 inhibitors, among them the so-called ‘‘coxibs’’, which are now being marketed extensively worldwide. They present an interesting challenge to the traditional NSAIDs, some of which have varying degrees of COX-2 activity relative to effects on the COX-1 system. Some of the long-term data on the serious gastrointestinal (GI) events from the coxibs compared with established NSAIDs (diclofenac, ibuprofen, naproxen) has shown that, although the coxibs have lower endoscopicallyobserved GI lesions in short-term trials in volunteers and arthritic subjects, the incidence of various GI complications is not appreciably different with coxibs than with established drugs (Schoenfeld 1999). Moreover, coxibs appear to have a higher risk
of cardiovascular events, while renal complications appear about as frequent as with conventional NSAIDs. A recent evaluation of the cost–benefits and overall safety of coxibs by the UK National Institute for Clinical Excellence (NICE) has concluded that coxibs are of only marginal benefit and that their use should be restricted to certain elderly ‘‘at risk’’ patients (http://phrmacotherapy. medscape.com/reuters/prof/2001/07/07.27/20010726rglt009.html). Contrasted with the coxibs, the range of COX-2-inhibitory NSAIDs (etodolac, nimesulide) that were identified as having some degree of COX-2 selectivity (Glaser et al 1995) after they had been discovered and introduced clinically have had relatively good GI safety compared with other NSAIDs. The second approach for the development of NSAIDs with low GI or renal toxicity has been the development of dual cyclooxygenase/5lipoxygenase inhibitors (COX/LOX), nitric oxide (NO) donating cyclooxygenase inhibitory NSAIDs (or CIONODs) and the prodrugs, e.g. nabumetone (Rainsford 1999a, 2001). While nabumetone has shown favourable GI benefits in long-term studies (Rainsford 1999a), long-term safety data from the COX/ LOX and CIONODs has not accumulated sufficiently to enable a clear determination of their long-term cost-efficacy. Phospholipase A2 inhibitors and glucocorticoids represent other approaches to controlling eicosanoid production in rheumatic and other chronic inflammatory diseases. While glucocorticoids are well established for use in rheumatic, gastrointestinal, respiratory and dermatological conditions, concerns about their long-term safety limit their widespread use.
192
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 16.3 Formation of aspirin-activated lipoxins. From Serhan (1997), with permission. Published by Elsevier Science
While a considerable number of phospholipase A2 inhibitors have been developed over the years, few of these have been introduced into clinical trials and none have to any extent found their way into long-term use. In this chapter the major emphasis will be on the development and properties of COX-2 selective drugs and COX/LOX inhibitors, in view of their current preeminence as inhibitors of eicosanoid metabolism. Much interest has been shown in the recent discovery of another isoform of cyclooxygenase, COX-3, a splice variant of COX-1 that appears to be abundant in cerebral cortex and other neural tissues and which is particularly sensitive to inhibitory
effects by paracetamol and a range of NSAIDs (Chandrasekharan et al 2002). The inhibition of COX-3 may account for the central mechanisms of the analgesic activity of these drugs Chandrasekharan et al 2002). MECHANISMS OF COX ENZYME INHIBITION BY NSAIDs The inhibition by various NSAIDs of COX-1 differs quantitatively and mechanistically from that of COX-2 (Gierse et al 1995; Vane et al 1998). This is related to extensive studies that have been
EICOSANOID INHIBITORS undertaken on the basic properties of the three-dimensional (3-D) structures of the two isoforms of cyclooxygenase and modelling of NSAIDs to these active sites, using different isoforms and muteins, kinetic studies of various inhibitors (Gierse et al 1995) and studies on the substrate, arachidonate, conformation and interactions with different isoforms. Of particular interest is the time-dependent inhibition of COX-2 activity, which sets COX-2 selective drugs aside from their effects on COX-1 (Smith et al 1994). Structural differences between the active sites of COX-1 and COX-2 affect the interactions between arachidonate and various NSAIDs, as summarized in Table 16.1 (Trummlitz 2003). A major structural difference between the activity sites of COX-1 and COX-2 is the slightly larger binding site for NSAIDs in the latter compared with that in COX-1. The structural modifications are mainly due to the difference in the amino acids at positions 513 and 523 in COX-1 and the homologous position in COX-2 (i.e. His/Arg 513; Ile/Val at 523; Table 16.1). The prominent role of the monocationic guanidinium group of Arg120, with subtle changes in Arg277 and Gln358, along with conditions such as solvation in aqueous solution and molecular volumes of the inhibitors probably contribute to differences in the interactions of NSAIDs with COX-1 and COX-2 enzymes. Additionally, there are important quantitative differences in enzymic activity and responses to inhibitors which arise from different assay methodologies (see next section, Differential Effects of NSAIDs on COX-1 and COX-2). Thus, substrate availability at or near the intracellular sites of enzymes (in their location in the endoplasmic reticulum with both COX-1 and COX-2 and in the nuclear envelope with COX-2) markedly influences the effects of inflammatory stimuli. The kinetics of intracellular uptake and accumulation of the drugs will influence their actions at the cellular level. Moreover, drug pharmacokinetics can markedly affect drug actions in vivo (Fenner 1997; Pairet and van Ryn 1998; Harada et al 1998) and also the interpretation of concentration relationships at which COX-1 and COX-2 inhibition is achieved in vivo by different NSAIDs. The overall expression of COX-isoform activities in vivo also depends on the various regulatory systems influencing their activities. Thus, the negative control of COX-2 activity by nitric oxide, produced in response to co-induction of both iNOS and COX-2 by inflammagens and differential effects of various NSAIDs on iNOS activity, can markedly influence overall COX-2 activity (Swierkosz et al 1995; Narayanan et al 2003; Ribeiro et al 2003). There is also evidence for ‘‘COX cross-talk’’ during different phases of the inflammatory response and regulation by cyclopentanone prostaglandins and end products of COX-2 activity in vivo (Willoughby et al 2000). Overall, it is clear that there are many factors accounting for the essential differences in the properties of the two enzymes, the mechanisms and the pharmacokinetics of the inhibitory actions of NSAIDs as well as coxibs, and the competing roles of the substrate, arachidonate, in the presence of the inhibitors and intracellular regulation of the enzymic activities. The mechanisms of inhibition of COX-1 and COX-2 by aspirin are, intriguingly, quite different than with other NSAIDs. With COX-1, aspirin acetylates Ser530 (sheep seminal enzyme) or Ser529 (human enzyme), which are located at the entrance to the active site of this enzyme, effectively blocking entry of the substrate, arachidonate, to the active site. The acetylation by aspirin of the serine residue is covalent and irreversible. In cells such as platelets, where there is no capacity to renew the COX-1 enzyme, the consequences of this are that the blockade of the production of prostanoids is irreversible for the entire lifetime (5 days) of the platelet. Thus, production of prothombotic thromboxane A2 in aspirin-treated
193
subjects is long-lasting and this forms the basis for the antithrombotic, cardioprotective and ischaemia–stroke-preventive actions of aspirin. The effects of aspirin on COX-2 are, in contrast to those of COX-1, such that the enzyme is converted into a form that catalyses the conversion of arachidonate to 15-hydroperoxyeicosatetraenoic acid (15-HPETE) and then 15hydroxyeicosatetraenoic acid (15-HETE). This conversion gives the COX-2 enzyme characteristics of a 15-lipoxygenase (15-LOX). The 15-HETE thus formed is then metabolized in some cells via 5- or 12-lipoxygenases to yield ‘‘aspirin-triggered’’ epilipoxins A and B (Figure 16.3). These products have important antiinflammatory properties. So aspirin, far from having the capacity to reduce prostanoids as part of its antiinflammatory actions, also exerts non-prostaglandin-related antiinflammatory effects via the production of epi-lipoxins. The precise quantitative significance of PG inhibition to epi-lipoxin formation is not known with aspirin, but clearly the production of the latter affords considerable novelty to the actions of aspirin. The mechanisms of inhibition of COX-1/2 activities were investigated during the 1970s before the identity of COX isoforms was established. Some retrospective identity of the isoforms may be presumptively established, based on what is known about either the currently known properties of those isolated enzyme preparations (e.g. ovine or bovine seminal vesicle microsomal lysozymes as COX-1 preparations) or cell-based assays in which COX-2 as well as COX-1 activities are known to be present, depending on the stimuli employed. A review by Flower (1974) discussed the mechanisms of inhibition of enzymic activity (principally in COX1-containing seminar vesicle microsomal enzyme preparations) by ‘‘established’’ or ‘‘conventional’’ NSAIDs, e.g. ibuprofen, indomethacin, naproxen, phenylbutazone. Some drugs, such as indomethacin, that are tight binding inhibitors of COX activity have been regarded as having pseudo-irreversible inhibition, which in fact is a reflection of partial enzyme inactivation, perhaps due to peroxidative activity during the conversion of PGG2 to PGH2 whereby an oxyradical species is released following cleavage of the 15-hydroperoxy group to form a 15hydroxy derivative. Phenolic compounds (salicylate, MK-447), as well as phenol itself, are thought to scavenge the oxyradicals (Ow. ) released in the PGG2 to PGH2 conversion, and so may act to protect the COX enzyme from self-destruction as a consequence of Ow. . Salicylate is known to be converted to 3- or 5-hydroxy derivatives [2,3-dihydroxybenzoic acid or 2,5-dihydroxybenzoic (=gentisic) acid, respectively] following oxyradical attack, but it is not known whether this reaction occurs during the interaction of this drug with the peroxidase component enzymic activity that constitutes PGH synthase. The direct inhibition of COX activity by salicylate is relatively weak and it is presumed that the interactions with the peroxidative enzyme contribute to the overall effects of the drug on total COX activity by accelerating PGG2 to PGH2. An interesting ‘‘model’’ antiinflammatory drug, MK-886, was also found to accelerate PGG2 to PGH2 conversion, possibly as well as inhibiting COX activity (Kuehl et al 1980). These effects were linked to the acute antiinflammatory actions of MK-886. Phenylbutazone and azapropazone were also found to affect peroxidative (PEROX) activity of PGH synthases, as well as inhibiting COX and LOX activities (Hughes et al 1988; Cucurou et al 1991). Phenolic compounds, including the natural products eugenol and rezveratrol, as well as guaiacol and paracetamol, also inhibit PEROX activity (Thompson and Eling 1989; Johnson and Maddipati 1998). Oxidation of the inhibitor, with consequent formation of a reactive intermediate, may occur during or as part of the mechanisms of inhibition of PEROX by these agents (Hughes et al 1988; Hsuanyu et al 1992; Forghani et al 1998). This may be a common mechanism of
194
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
action of many mono- and polyphenolic natural products that have been shown to have antiinflammatory activity. It should be noted that many NSAIDs and phenolic compounds also affect oxyradical production or actions in non-eicosanoid pathways, including the inhibition of leucocyte myeloperoxidase activity (Kimura 1997; Neve et al 2001).
Another aspect of the mechanisms of inhibition of COX activity was highlighted by the studies of Rome, Lands, Smith and DeWitt in a series of papers which showed that there were slow as well as fast inhibitors of COX activity among the conventional NSAIDs (DeWitt et al 1990; Laneuville et al., 1994), a feature confirmed by others (Callan and Swinney 1996; Selinsky et al, 2001). These
EICOSANOID INHIBITORS
195
Figure 16.4 Routes of formation of lipoxins by various pathways involving 5-, 12- and 15-lipoxygenases. (A) (opposite) Cell–cell lipoxin formation; 5-LO:12-LO interactions. (B) 15-LO-initiated LX biosynthesis. From Serhan (1997). Published by Elsevier Science
Table 16.1 Critical amino acids of COX-1 and COX-2 for the interaction with arachidonic acid and NSAIDs Amino acid
Enzyme site
Function
Tyr 385 Ser 530/529 Phe 518 (Ile/Val 434) Leu 384 (Phe/Leu 503) Ile/Val 523 His/Arg 513 Arg 120, Tyr 355
Catalytic site Aspirin site Flexible space COX-2 flexible Extra space COX-2 Side pocket COX-2 Side pocket COX-2 Constriction site
H-abstraction at C13 of arachidonic acid Acetylation by aspirin Secondary shell variation: channel volume in COX-2 Secondary shell variation: larger channel volume in COX-2 Primary shell variation allows access to the side pocket in COX-2 Secondary shell variation: binding in the side pocket in COX-2 Ionic/H-bonding interaction with arachidonic acid and NSAIDs
From Trummlitz (2003), with permission.
196
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
differences further underscored the general point, that there are fundamental differences in the inhibitory effects of various conventional NSAIDs on COX and PEROX activities of PGH synthase. As many of the early studies were performed with seminal vesicle microsomal preparations, in all probability these were the PGHS-1 enzyme containing COX-1 activity. Using mouse peritoneal macrophages, several authors examined the inhibitory effects of conventional and experimental NSAIDs that were available in the 1970s and 1980s (e.g. Bray and Gordon 1978; Brune et al 1981; Hoffman and Autor 1982; Humes et al 1981). It is now known that many of these primary macrophage cells probably had appreciable COX-2 activity as a consequence of stimulation by activators of PG production, such as phorbol esters, calcium ionophores, zymosan and/or the prior activation of resident macrophage populations in the peritoneal compartment by intraperitoneal injection of thioglycollate or zymosan. Activation of phospholipase A2 activity with release of arachidonate would also be expected with these treatments. Additionally, COX-1 activity would also be present, although the total contribution of this to overall PG production by activated macrophages would be relatively small. Among these earlier studies, those reported by Brune and colleagues (1981), who used non-serum-containing culture media, showed that (a) inhibition of PGE2 production by conventional and some experimental NSAIDs correlated with their acute antiinflammatory activities in the rat carrageenan assay and chronic antiarthritic activity in adjuvant arthritic rats; (b) that the concentration required for 50% inhibition of macrophage PGE2 production correlated well with the daily dose of these NSAIDs for relief of pain and swelling arthritic disease; and (c) that among a range of salicylates and phenols, there were clear physicochemical and quantum chemical features that could be identified to explain the mode of their inhibitory activity on PGE2 production (Brune et al 1981). The quantum chemical features underlying the inhibition of PG production by NSAIDs in macrophages have been examined by Mehler et al. The potency of NSAIDs as inhibitors of PGE2 production was found to be correlated with the highest occupied molecular orbital (HOMO) (Mehler et al 1982). As potency increases, so there is decrease in the binding strength of p-HOMO electrons, suggesting that charge transfer is important for binding the drugs to binding sites. In a later study with 19 benzoic acids, phenols and salicylic acids, the authors were able to refine these conclusions, invoking a two-step binding process to COXs, originating with: (a) alignment of the inhibitor molecule to the binding site, with stabilization originating in the electrostatic potential of the inhibitor and the receptor binding site; and (b) overlapping of the drug molecule’s frontier orbitals with those of the receptor binding site. Molecular orbital calculations by Zoete et al (1999), using the data of Habicht and Brune (1983) in essentially the same experimental system as employed by Mehler et al (1982, and Mehler and Gerhards 1987), confirmed the conclusions concerning HOMO interactions made by the latter authors, but additionally showed that there was improved correlation when drug lipophilicity Log P corrected to pH 7.2 was included in the analysis. There are some important underlying assumptions concerning the empirical application of molecular orbital calculations employed in these studies, inasmuch as the presumed interaction of the drug with the COX enzyme active site would not be expected to be limited by the uptake of the drug into the macrophage or distribution to the site of the enzyme in microsomes. Substrate availability and optimal concentration (i.e. at approximately the Km value) would also be assumed. While some of these assumptions may be unknown and so are limitations, the advantage of these studies is that they represent drug interactions in an intact cellular system. However, there is no
discrimination between drug effects on COX-1 from those on COX-2, since it is likely that the predominant cyclooxygenase activity in macrophages is COX-2 with a small amount of COX-1. Thus, it is probably a fair assumption that the drug molecular interactions are probably with COX-2. The semi-empirical analysis by Zoete et al (1999), using data on NSAID inhibition of PGE2 production by sheep seminal vesicleisolated enzyme preparations (which is presumably COX-1 activity) of Dewhirst (1980), showed that the inhibitory potency, like that of the mouse macrophage data, was correlated with HOMO energies and Log P values of the drugs, and that these correlations were superior to those where HOMO energies were compared alone. Using comparative molecular field analysis to establish a 3-D quantitative structure activity analysis of COX-2 inhibitors, Marot et al (2000) showed the role of electronic and steric factors of the drugs influencing their effects and were able to confirm expected and predicted values to molecular models they developed. Thus, it appears that both HOMO energies of the molecules and their lipophilicity, as well as steric interaction, are the major determinants of the inhibitory actions of NSAIDs in COX systems. Differential Effects of NSAIDs on COX-1 and COX-2 The identification, isolation and characterization of PGH synthases having COX-1 or COX-2 activities has permitted studies on the effects of NSAIDs on the different isoforms. A range of systems have been developed, among them human or other recombinant PGH Synthase-1 and -2 (COX-1 and COX-2 activities) cells, transfected with the recombinant enzymes, e.g. Chinese hamster ovary (CHO) and insect cells, and a range of primary cells (Vane and Botting 1995; Vane et al 1998). A variant of the latter is the in vitro whole blood assay techniques in which blood is incubated for 1 h with the drugs and thromboxane B2 determined by ELISA or RIA as a measure of platelet COX-1 activity. Incubation of whole blood with LPS or other stimuli together with the drugs for 24 h and subsequent assay of PGE2 by ELISA or RIA enables measurement of COX-2 activity (Tacconelli et al 2002). Proponents of this technique claim that it is the best representation of COX activities in circulation and can account for the free drug effects, allowing for their binding to plasma proteins and distribution in red blood cells. By taking blood samples from subjects who have ingested NSAIDs and incubating them for 1 h or 24 h (with LPS or other stimuli) and assaying the prostanoids, it is possible to determine the in vivo effects of the drugs in essentially an ex vivo system. With the in vitro whole blood assay, it is possible to relate the concentrations for effects of the drugs to their known concentrations in the circulation. Thus, linking the pharmacokinetics of the drugs to their pharmacodynamic activity is possible by the in vitro whole blood assay, while the ex vivo technique gives a related estimation of in vivo drug effects. These whole blood assays have been extensively employed to determine the COX-2 selectivity (i.e. from COX-2:COX-1 ratio of activities) of NSAIDs and the identification in particular of the selective effects of coxibs and other COX2 active drugs in human studies. The whole blood assays are not without limitations, although they are probably about the best available methods for determining clinically-relevant COX-2 selectivity (Brooks et al 1999). The limits are that in the COX-2 assay there will be: (a) some COX-1 activity from monocytes-macrophages and platelets in the circulation; (b) uncontrolled release of arachidonate, so that drugs with a degree of reversible activity may prove less potent than they would be if assayed in an isolated enzyme system; (c)
EICOSANOID INHIBITORS
197
Table 16.2 Comparison of the inhibitory activities of aspirin with other NSAIDs on COX-1 and COX-2 isoenzymes using human recombinant enzymes (transfected into CHO Chinese ovary, COX monkey kidney or S19 insect cells) and in whole human blood in vitro Human recombinant enzymes
Drug Aspirin Celecoxib Diclofenac sodium Etodolac Fenoprofen calcium Flurbiprofen Ibuprofen Ketoprofen Ketorolac Mefenamic acid Meloxicam Naproxen sodium Nimesulide Oxaprozin Piroxicam Rofecoxib Sulindac Tenoxicam Tolmetin
COX-1, IC50 (mM) 4.45 0.32 0.0034 50 1.96 0.0015 1.53 0.006 0.083 0.15 1.95 0.23 2.1 14.9 2.52 26.3 NR 17.8 0.58
COX-2, IC50 (mM)
Whole human blood
COX-2 selectivity COX-1/COX-2 (IC50)
COX-1, IC50 (mM)
COX-2, IC50 (mM)
COX-2 selectivity COX-1/COX-2 (IC50)
0.28 160 1.47 1219 0.07 0.58 0.13 0.05 6.9 0.54 33 0.07 6.7 50.02 6.5 77 NR 0.48 0.40
4.00 6.3 0.14 11.7 2.73 0.59 4.88 0.11 0.32 1.94 1.87 11.3 4.1 14.6 2.62 19 41.3 2.3 1.23
81 0.99 0.06 2.60 4.03 3.46 22.4 0.88 0.88 0.16 0.54 52.3 0.31 36.7 5.35 0.51 24.9 14.2 7.09
0.05 6.36 2.33 4.5 0.68 0.17 0.22 0.13 0.36 12.1 3.46 0.22 13.2 0.40 0.49 37.2 1.66 0.16 0.17
16 0.002 0.0023 0.041 26.3 0.0026 11.8 0.119 0.012 0.28 0.06 3.36 0.31 41000 0.39 0.34 NR 36.9 1.44
From Kulkarni et al (2000), with data derived from references therein and Prous Science MFLine1
Table 16.3 Potencies of compounds as inhibitors of prostanoid formation determined in the COX-1 assay, WBA-COX-2 IC50 ratios COX-1 IC50 (mM) NSAIDs 5-aminosalicylic acid Aspirin Diclofenac Diflunisal Flubiprofen Flufenamate Ibuprofen Indomethacin Ketoprofen Meclofenamate Mefenamic acid Naproxen Piroxicam Salicin Salicylaldehyde Sodium salicylate Sulfasalazine Sulindac Sulindac sulphide Valeryl salicylate COX-2 Inhibitors Celecoxib Etodolac Meloxicam Nimesulide L745,337 NS398 Rofecoxib SC58125 Analgesics (non-narcotic) Aminopyrone Paracetamol n.d., not determined. From Warner et al (1999)
410 1.7 0.075 113 0.075 3.0 7.6 0.013 0.047 0.22 25 9.3 2.4 4100 4100 4956 3242 4100 1.9 42 1.2 12 5.7 10 4100 6.9 63 4100 55 4100
WBA-COX-2 WHMA-COX-2 IC50 (mM) IC50 (mM) 61 4100 0.038 8.2 5.5 9.3 7.2 1.0 2.9 0.7 2.9 28 7.9 4100 4100 34440 2507 4100 55 2.3 0.83 2.2 2.1 1.9 8.6 0.35 0.84 2.0 203 49
Ranking at IC50 ratios
WBA COX-1
WHMA COX-1
WBA COX-1
WHMA COX-1
n.d. 7.5 0.020 134 0.77 n.d. 20 0.13 0.24 0.2 1.3 35 0.17 n.d. n.d. 482 n.d. 58 1.21 n.d.
0.15 4100 0.5 0.1 73 3.1 0.9 80 61 3.2 0.11 3.0 3.3 – – 6.9 0.8 – 29 0.053
n.d. 4.4 0.3 1.2 10 n.d. 2.6 10 5.1 0.91 0.049 3.8 0.1 n.d. n.d. 0.10 n.d. – 0.64 n.d.
n.d. 34 10 9 31 13 14 29 31 22 – 18 17 – – 16 15 – 20 –
n.d. 23 9 14 27 n.d. 20 24 25 11 – 22 13 n.d. n.d. 15 n.d. – 10 n.d.
0.34 0.94 0.23 0.39 1.3 0.042 0.31 n.d.
0.7 0.2 0.37 0.19 50.01 0.051 0.013 40.01
0.3 0.1 0.040 0.038 50.01 0.0061 0.0049 n.d.
3.7 –
1.5 –
85 64
8 6 11 7 1= 5 4 1= 24 –
7 5 6 8 1= 4 3 n.d. 19 –
198
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
antioxidant drugs, and those which may be prodrugs or undergo metabolism, may yield data for COX-2 activity which could be an underestimate as a result of drug instability. The reason for emphasizing these methodological aspects is also because the data that have been obtained in the different enzyme preparations with various NSAIDs show wide variability. Thus, inspection of the data in Tables 16.2 and 16.3 shows that the COX-2:COX-1 ratios, as well as the relative inhibitory potencies of various NSAIDs and coxibs, vary markedly with assay methodologies. It can be argued that each assay system has its own merits and utility. Thus, isolated human recombinant COX-enzyme preparations will give information on the enzyme kinetic properties, and the mechanisms of inhibition by the drugs are useful for defining their mode of action at the biochemical level. However, the data obtained may not be entirely predictive of what may happen in animals and humans, even if there are estimates of the free or total circulating levels of the drugs taken into account in estimating the likely drug effects. Moreover, even accounting for optimizing substrate concentrations for the enzyme assays may not account for the more dynamic situation regarding substrate availability and oxidant/antioxidant status in those compartments where COX activities occur in intact cells. Attempts to overcome these limitations by the use of transfected cell lines or primary cells yield information which may be more pharmacologically relevant; the transfected cell lines enable a direct measure of COX-1 or COX-2 activity without the unknown possibility of the alternative enzyme which may be present in primary cells. Yet primary cells do, of course, enable understanding of drug activities in cells, which are not as artificial as transfected systems and can give information where applied to, say, synovial or mucosal cells which is relevant to pharmacological activities. Finally, as indicated above, the in vitro/ex vivo whole blood assays are attempts to obtain clinicallyrelatable or related data. An important consideration is, however, that these systems will in all probability be poor predictors of drug effects in the gastrointestinal tract, kidney or inflamed tissues in which side- and therapeutic effects of NSAIDs are manifested. Here there is no escape from the need to obtain direct measurements of drug effects in these sites, but unfortunately this inevitably requires considerable expense and resources. SAFER NSAIDs THROUGH COX-2 SELECTIVITY Various strategies have been employed to overcome the major problem of the severe gastrointestinal (GI) adverse effects (bleeding and ulceration) that occurs with the traditional NSAIDs (Whitehouse and Rainsford 1982; Rainsford 1997, 1999a,b; Wallace and Cirino 1994; Wallace 1999). Among the most popular of these in recent years has been the development of COX-2-selective drugs (e.g. nimesulide, meloxicam, celecoxib and rofecoxib; Hawkey 1999; Rainsford 1999b,c; Feldman and McMahon 2000). The rationale behind the development of these drugs is that they selectively inhibit the production of the COX-2 enzyme that is specifically upregulated during inflammation, so producing prostaglandins involved in expression of inflammatory events, while sparing the constitutive COX-1 that produces prostaglandins that are involved in regulation of physiologically important processes (blood flow and haemodynamics, renal function) and gastric mucosal protection (e.g. gastric blood flow, regulation of acid, nitric oxide and mucus secretion) (Smith et al 1994; Gierse et al 1995; Dubois et al 1998; Cullen et al 1998; Vane et al 1998). While undoubtedly demonstrating proof of the principle that human gastric mucosal prostaglandin production is spared by drugs such as nimesulide (Cullen et al 1998; Shah et al 1999; Bjarnason and Thjodleifsson 1999), celecoxib (Feldman and McMahon 2000) and rofecoxib coincident with
reduction in gastric ulcers in patients with rheumatic diseases receiving these drugs (Bjarnason and Thjodleifsson 1999; Laine et al 1999; Langman et al 1999; Schoenfeld 1999; Feldman and McMahon 2000; Hawkey et al 2000), with the newer COX-2 preferential or selective drugs (meloxicam, celecoxib, rofecoxib) there has not been an appreciable reduction in upper GI symptoms (nausea, vomiting, heartburn, epigastric pain) compared with traditional NSAIDs (e.g. ibuprofen, naproxen, diclofenac; Langman et al 1999; Hawkey et al 2000). This alone is compelling evidence for the need to search for drugs that reduce GI symptoms, since these are a major cause of withdrawal from use of COX-2 selective drugs (Langman et al 1999; Schoenfeld 1999) like that of traditional NSAIDs (Rainsford and Quadir 1995). Moreover, there is increasing concern that the long-term appearance of severe adverse reactions with the newer COX-2 selective drugs may only just be starting to appear. Among these are hypertension, oedema, renal failure, abnormal renal function and an increased risk of coronary-vascular events (heart attack) and stroke with rofecoxib (Pierson 2000) and gastropathy from celecoxib (Mohammed and Croom 1999; Reuben and Steinberg 1999), as well as media reports of deaths linked to celecoxib. There have also been challenges to evidence that some traditional COX-1/COX-2 acting NSAIDs (e.g. nabumetone, ibuprofen) may have a lower incidence of severe gastroduodenopathy (peptic ulcer bleeds) than observed with other established NSAIDs (Freston 1999; Huang et al 1999), thus questioning the rationale that COX2 selective NSAIDs will a priori be the way forward for safer GI therapy in the future (Bjarnason et al 2003). Even the pharmacological rationale for COX-2 selectivity has been challenged with evidence that: (a) COX-2 is present in the stomach (Rainsford et al 1995; Sawaoka et al 1997; Schmassmann et al 1998) and the inhibition of this enzyme results in delayed ulcer healing (Halter et al 1997; Wallace 1999); (b) the production of the endothelial antiplatelet-aggregating prostanoid, PGI2 or prostacyclin, is derived from COX-2 (McAdam et al 1999); and (c) COX-2 regulates renin production (Stichtenoth et al 1998) and during salt depletion in volume-depleted subjects (Rossat et al 1999), so that COX-2 inhibition may have pronounced consequences for renal function. Indeed COX-2 knockout mice have been shown to have pronounced nephropathy (Morham et al 1995). Recent studies have also shown that COX-2 selective NSAIDs exhibit pronounced enteropathy in COX-1 knockout mice (Bjarnason and Thjodleifsson 1999; Bjarnason et al 2003). Finally, examination of the physicochemical properties of the newer COX-2 selective drugs reveals that their high pKa (45.5), presence of sulphonyl instead of carboxyl moieties (the latter being a common feature of conventional NSAIDs such as aspirin, diclofenac, indomethacin, ibuprofen, naproxen) and high relative lipophilicity sets them aside from conventional NSAIDs, and may be predicted to lead to less focal accumulation in gastric mucosal cells and disruption of the protective phospholipid mucus barrier than seen with conventional NSAIDs (Rainsford 1999b). Thus, the rationale for COX-2 selectivity sparing the stomach from the prostaglandin-related effects of NSAIDs may also have more to do with the non-COX-2 related physicochemical properties of these drugs (Rainsford et al 1999b). DEVELOPMENT OF SPECIFIC COX-2 SELECTIVE INHIBITORS The development of the new class of COX-2 selective inhibitors has been comprehensively reviewed by a number of authors (Prasit and Riendeau 1997; Dannhardt and Laufer 2000; Dannhardt et al 2000; de Leval et al 2000; Dannhardt and Kiefer 1994), to which the reader is referred for detailed consideration of the medicinal chemistry and overall views of the discovery at
EICOSANOID INHIBITORS
Figure 16.5. General structure of methane sulphonanilide COX-2 inhibitors (R, alleyl, X, O, S, Ar, aromatic [substituted, unsubstituted] group, Cyc, cyclohexyl ring, Het, heterocycle group, Ewg, electronwithdrawing substituent) (de Leval et al 2000)
different stages in the development of these agents. The design of these compounds obviously has benefited considerably from knowledge of the three-dimensional structure of the COX isoenzymes, as well as the application of high-throughput screening and other fast screening procedures. Early development of these compounds benefited from knowledge about the actions of methane sulphonanilide inhibitors, whose general structures are shown in Figure 16.5. Of these, nimesulide, NS398, the Schering compound flosulide (or formerly the Ciba-Geigy compound, CGP28238) T614 and diflumidone are representative drugs (Figure 16.6) (de Leval et al 2000). The general chemical features of the sulphonanilide compounds (Figure 16.5) include: (a) an alkyl substituent on the sulphonyl group, which is usually a methyl substituent, although trifluoromethyl substituents have also been reported; (b) at the 2-position there may be aryloxy, arylsulphonyl, cycloalkyloxy or cyclooxysulphonyl groups; (c) at the 2-position of these aryl substituents there may be present either an aromatic group, substituted or unsubstituted, cycloalkyl or a heterocyclic ring structure; and (d) at the 4-position of the alkyl sulphonyl moiety there are usually present electron withdrawing groups, and in the case of nimesulide and NS398 (Flosulide) (Figure 16.6). Of these sulphonanilides, only nimesulide has achieved clinical success.
199
However, identification of the characteristics of the methylsulphonylanilides led to the development of a number of drugs, amongst them L745,337 (Figure 16.6). Trials with flosulide were disappointing. Even though this drug appeared effective, it did unfortunately result in nephrotoxicity in an appreciable number of patients, so leading to its withdrawal. Another early progenitor of COX-2 selective inhibitors was the indoxol oxaprozin (not itself a COX-2 selective drug) and some related heterocyclic compounds. These led to the development of a range of diaryl substituted heterocyclic compounds, although the relative influence of these earlier drugs is not entirely clear. Some structures of these diaryl substituted heterocyclics that attracted particular interest in the early development of COX-2 selective drugs are shown in Figure 16.7. Among this class were compounds for the central four-membered ring (Figure 16.7), diaryl substituted heterocyclics with the five-membered central cyclic ring (Figure 16.7), those with the central 5-membered heterocyclic group comprising thiophenes, pyrols, furans, oxazoles and isoxazoles, and various 5- and 6-membered ring derivatives (Figure 16.7). From these finally emerged the clinically successful 1,5-diarylpyrozole drug celecoxib (Figure 16.8), and the 3,4-diaryl substituted furan, rofecoxib (Figure 16.8), both of which have been introduced worldwide in most markets (Figure 16.8). Further information about these drugs from the clinical and experimental viewpoint can be found in a number of reviews and on the respective FDA websites: for celecoxib (2001): www.fda.gov/ohrms/dockets/ac/01/slides/3677s1-01-sponsor.pdf; and for rofecoxib (2001): www.fda.gov/ohrms/dockets/ac/01/ slides/3677s2-01-sponsor.pdf. The relative potencies of rofecoxib and celecoxib compared with other NSAIDs, including COX-2 inhibitory agents with a lower COX-2 selectivity, are shown in Tables 16.2 and 16.3. The relative inhibitory activities and degree of COX-2 selectivity (as measured by the IC50 values for COX-1 compared with COX-2 inhibition) vary considerably according to the assay methodology. Claims have been made that data based on whole blood assays is a closer representation of the degree of inhibition of COX-1 compared with COX-2 in vivo but this does not give an indication
Figure 16.6 Some representative COX-2 sulphonanilide inhibitors. Of these, only nimesulide is used clinically
200
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 16.7 Diaryl-substituted heterocycles with varying central ring substituents that have been developed with COX-2 inhibitory activity (based on de Leval et al 2000)
Figure 16.8 Currently employed COX-2-selective inhibitors
of the inhibition at inflamed sites or in the central nervous system, where COX effects are considered significant. Of the other experimental COX-2 inhibitors, a range of di-tertiary butyl phenols were developed at Parke-Davis (Li et al 1999; Song et al 1999a,b), the structures of which are shown in
Figure 16.9. Many of these di-tert-butyl phenols that were based on tebufelone have antioxidant activity, and this in some respects complicates the mechanism of inhibition of prostaglandin production by the COX isoenzymes. This is because of the relative effects of antioxidants in affecting the oxygen radical-
EICOSANOID INHIBITORS
201
Figure 16.9 Di-tert-butyl phenols based on tebufelone
mediated reactions that characterize the dioxygenation of arachidonic acid. Since the development and successful introduction of celecoxib and rofecoxib there have been a number of new, so-called secondgeneration COX-2 selective inhibitors that have been introduced clinically (Stichtenoth and Fro¨hlich 2003). Among these new drugs are etoricoxib (MK-0663); valdecoxib, parecoxib and lumaricoxib (Stichtenoth and Fro¨lich 2003). Parecoxib (Figure 16.8) has been developed as an injectable COX-2 specific inhibitor and the pharmacokinetics of this drug given i.v. to rats have been compared with that of the analgesic activity (Talley et al 1999). The latter results show that parecoxib is cleared relatively rapidly in rats, with a half time of approximately 1 h, so this presumably favours short-term analgesia in comparison with other injectable NSAIDs, such as ketorolac. Etoricoxib, among the most selective of COX-2 inhibitors (Patrignani et al 2003) was studied in patients with osteoarthritis, rheumatoid arthritis and gout. These studies have shown that etoricoxib is at least comparable with standard NSAIDs (Leung et al 2002; Capone et al 2003). In all of these cases the perceived safety benefits were favourable. In the context of GI safety and long-term effects on the renal system, obviously longer-term trials are awaited with this drug, along with the new coxibs, in order to establish whether they really are more favourable than either established NSAIDs or the first generation of coxibs (celecoxib, rofecoxib). It is of particular interest that etoricoxib, given at a high dose of 120 mg o.d., was as effective as indomethacin 50 mg t.i.d. in the treatment of gout (Schumacher et al 2002). This implies that gout may be considered to be a condition driven largely by COX-2, although classical studies have indicated that lipoxygenase activity is increased in this condition. Valdecoxib (Figure 16.8) was studied in the treatment of osteoarthritis of the hip, rheumatoid arthritis, a range of acute pain conditions encountered in oral surgery and in primary dysmenorrhoea (Benson et al 2002; Kivitz et al 2001; Makarowski et al 2002)). Valdecoxib has been claimed not to impair platelet function and produced less endoscopically-confirmed ulcers in
patients with osteoarthritis than naproxen or in comparison with ibuprofen and diclofenac in patients with osteoarthritis (Sikes et al 2002). Because of concerns about upper GI ulceration from parenteral NSAIDs, such as ketorolac, the GI effects of parecoxib sodium were evaluated in comparison with ketorolac, naproxen and placebo in elderly subjects. Upon injection, the pro-drug, parecoxib, is metabolized to valdecoxib (Figure 16.8; Talley et al 1999), so it would be expected that there would be protection by the drug being a pro-drug. The results showed that after 7 days of treatment with parecoxib (10 mg i.v., b.i.d.) there were no upper GI lesions or ulcers, in comparison with ketorolac 15 mg i.v./ q.i.d., where all four subjects who received this drug had ulcers or lesions (Harris et al 2001). Parecoxib given intravenously has been shown to be effective in postoperative pain following gynaecological laparotomy surgery where this drug as well as ketorolac spared the use of morphine in analgesia when given as a single injection (Barton et al 2002). The most recent addition to the coxib drugs is lumaricoxib, which is an acidic drug (in contrast to the more basic character of the other coxibs) which may lead to its greater accumulation in inflamed tissues (Capone et al 2003; Stichtenoth and Fro¨lich 2003). It is probably the most selective of the COX-2 inhibitors (Capone et al 2003; Stichtenoth and Frolich 2003). Lumaricoxib has been shown to be effective in treatment of a number of rheumatic conditions. Overall, studies on the development and clinical applications of COX-2 selective drugs have been instructive in highlighting the significance of COX-2 inhibition for the control of pain and chronic inflammation. The concept that there may be some common structural features that underlie the selectivity of COX inhibition (Dannhardt and Kiefer 1994) is of particular interest. Whether the newer class of highly selective COX-2 inhibitors known as the coxibs deserve classification as a separate group of NSAIDs is a moot point, and until further investigations have been undertaken concerning the interrelationships between COX-2 inhibition and the consequences in reducing prostaglandin production have been fully evaluated it will not be possible to support the contention for separate classification.
202
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 16.10 Regulation of vascular-leucocyte reactions by leukotrienes and relation to activated leucocyte products
NITRIC-OXIDE-DONATING NSAIDs Alternatives to highly selective COX-2 inhibitors, which have as their mode of action selective modulation of the production of inflammatory mediators (arachidonic acid metabolites, nitric oxide, leukocyte accumulation and activation) have been, and continue to be, developed as GI-safe or safer antiinflammatory, analgesic and antipyretic agents. Among these agents, a group that have recently attracted much interest has been the nitricoxide-donating NSAIDs, originally developed by Del Soldato, Wallace and their co-workers, whose properties have been reviewed by Wallace (Wallace 1999; Wallace and Del Soldata 2003; Tam et al 2000). The nitrobutoxyl esters of established NSAIDs (NO–NSAIDs), as described by these authors, have as their main principle the property to produce NO from the presumed hydrolysis of the nitrobutoxyl-moiety released from the NSAID, the NO of which protects the gastric mucosa from the vascular injury and effects of accumulated leukocytes that are induced by the NSAIDs. In the case of NO–aspirin, there may be additional benefits of the NO released from NO–aspirin when employed as an antithrombotic agent, in that this may additionally prevent the adhesion of the platelets of arteriosclerotic subjects to endothelial cells of blood vessels, so conferring added benefits to aspirin as an antithrombotic agent. Other NOdonating NSAIDs are being developed with more selective redox properties that enable more controlled release of NO. The clinical development of consequent progress with these NO– NSAIDs is awaited with much interest. It should be emphasized that these are derivatives of established ulcerogenic NSAIDs, many of which have variable effects on arachidonic acid metabolism and for which the addition of the NO-donating group is designed to overcome their ulcerogenic effects—a case of addition of an antidote whose properties are to promote localized vasodilation with some inflammatory properties. The nitrobutoxyl moieties may also confer properties of blocking the acidic
functionality of the carboxylate group of NSAIDs, the gastroprotective effects of which were well described in the case of methyl- and higher alkyl esters (Rainsford and Whitehouse 1976, 1980a,b; Whitehouse and Rainsford 1980).
DUAL COX/LOX INHIBITORS The potential importance has been examined of another group of NSAIDs that have a different mode of action for modulating arachidonic acid metabolism and leucocyte accumulation and activation, as a consequence of their properties as dual inhibitors of cyclooxygenase(s)/5-lipoxygenase (COX/LOX) inhibitors (Celotti and Laufer 2001; Bertolini et al 2001). The basis for the development of COX/LOX inhibitors occurred about 20–25 years ago from a number of lines of research. Probably the major impetus at that time was the discovery of the leukotrienes derived from 5-lipoxygenase activity and their involvement in the inflammatory process (Figures 16.10 and 16.11). The potent effects of leukotriene B4 (LTB4) in leukocyte (principally neutrophil) accumulation and activation to release a range of inflammatory mediators and enzymes was, and is, still considered a key element in the development of the inflammatory process (Figures 16.10 and 16.11). From the other arm of the leukotriene metabolic pathway, the peptido- (or cysteinyl-) leukotrienes (LTC4, LTD4 and LTE4) produced vascular effects, especially dilatation of post-capillary venules, which also became recognized as another key factor in the development of inflammation (Figure 16.10: Issekutz et al 1986). The logical drug development which evolved was to combine the cyclooxygenase inhibitory activity with that of selective inhibition of the 5-lipoxygenase enzyme (Figure 16.11). As it happened, a number of NSAIDs developed at that time were discovered serendipitously to have these dual COX/LOX properties.
EICOSANOID INHIBITORS
203
Figure 16.11 Interrelationships between cyclooxygenase and lipoxygenase pathways and their feedback regulation by IL-1
Among the first of these COX/LOX inhibitors was BW775c (3-amino-[m-(trifluoromethyl)-phenyl]-2-pyrazoline), which was developed by Wellcome Research Laboratories (Beckenham, Kent, UK) and studied extensively for its antiinflammatory properties, and had low gastric irritancy. While this compound was never progressed because of inherent liver toxicity, it proved to be a useful agent for investigating the potential of LOX inhibitory activity to be combined with COX inhibition. This compound also showed the importance of blockade of 5-LOX activity leading to reduction in the accumulation and subsequent activation of polymorphonuclear leucocytes (PMNs) in the control of the development of the acute phase of inflammation. At about the same time, the Lilly Research Centre (Windlesham, Surrey, UK), who had a very active leukotriene programme centring on the development of antiasthmatic as well as antiinflammatory drugs, discovered and developed the propionic acid NSAID, benoxaprofen (Dawson 1980). This discovery had a major impact, for it was the first COX/LOX inhibitor showing leukocyte-inhibitory properties that was employed for use in the treatment of arthritic conditions. Extensive studies in laboratory animal models (Rainsford 1981; Rainsford and Willis 1982; Rainsford et al 1982, 1984) and studies in humans showed that benoxaprofen had low GI mucosal ulcerogenic effects and bleeding compared with established NSAIDs. Unfortunately, the long plasma half-life of elimination exhibited by this drug led to substantial accumulation in some elderly subjects (possibly those with impaired renal function), which resulted in a number of deaths from hepato-renal failure and GI bleeding, such that in the early 1980s this drug had to be withdrawn. A number of important lessons had been drawn from the failure of benoxaprofen, among them that the low GI mucosal potential exhibited by this drug may have been lost as a consequence of pronounced drug accumulation and hepatorenal adverse reactions in the elderly. Nonetheless, several other companies developed COX/LOX inhibitors, having recognized the important potential for control-
ling both the 5-LOX as well as the COX arms of arachidonic and mediated inflammatory processes. The additional benefits, of not only achieving more balanced control of inflammatory processes but also markedly reducing the potential for leukotriene-mediated inflammatory events occurring in the GI mucosa, also progressively became evident at and after the development of benoxaprofen and subsequent NSAIDs. The basis for recognizing the importance of leukotrienes in gastric pathophysiology and the development of mucosal lesions and ulceration was initially from observations on the microvascular effects of peptido-leukotrienes, but then it became recognized that they also have other effects. Thus, vascular injury as well as leucocyte adhesion to endothelia, infiltration and activation became recognized as important factors in the pathogenesis of GI ulceration induced by ethanol, NSAIDs and various other necrotizing agents (Rainsford et al 1984). Production of lipid mediators of inflammation, principally the leukotrienes, was observed during gastric ulceration in rats induced by ethanol. Inhibitors of leukotriene production, while having some variable protective effects on ethanol-induced gastric ulceration (Wallace and Whittle 1985), was significant. In some cases the variability was possibly related to the mechanism of action of the drugs and the concentration, dose and timing of ethanol administration. It was also found that parenterally administered peptidoleukotrienes could induce gastric vasoconstriction, increase vascular permeability, promote mucosal breakdown and stimulate the secretion of gastric acid and pepsin. Thus, peptido-leukotrienes have become prime candidates for mediating gastric mucosal damage. The earlier studies showed that some inhibitors of 5-lipoxygenase (5-LO) and leukotriene antagonists partly protected gastric mucosal lesions induced by some NSAIDs in normal rats and cholinomimetic treated mice, implying that vasoconstrictor leukotrienes could have a role in the development of gastric injury by these drugs. An unspecific inhibitor of the 5-LO
204
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
pathway, nordihydroguaiaretic acid, has not been found to protect the gastric mucosa of normal rats given NSAIDs. An explanation of some of the variations in these results is that some of the 5-LO inhibitors available at this time (mid-1980s) were relatively unspecific, their mechanism of action was not well understood, and they lacked the specificity of action of the more recently developed drugs (e.g. MK-886, zileuton). Moreover, the low sensitivity of the GI tract of normal fasted rats employed in some studies is also a major limitation. Further studies to demonstrate the importance of leukotrienes in the development of NSAID-induced vascular injury were undertaken by Gyo¨mber et al (1996). Indomethacin s.c. was found to cause vascular injury, evidenced by Monastral blue dye labelling of blood vessels and the electron microscopic appearance of the opening of endothelial cell tight junctions and extravasation of red blood cells into the surrounding interstitium before the development of gastric mucosal lesions (Gyo¨mber et al 1996). Similar evidence of vascular injury was obtained after oral dosing with indomethacin, although the appearance of mucosal lesions was faster in onset than from s.c. dosing of the drug, presumably due to some topical effects of the drug on the mucosa following oral administration. Under similar experimental conditions, indomethacin- and aspirin-induced gastric mucosal injury was inhibited by prior-administered 5-lipoxygenase inhibitors, e.g. the five-lipoxygenase activating protein (FLAP) inhibitor, MK886, or the direct enzyme inhibitor L-6555,224. However, the LTD4 antagonists MK-571 and L-660,711 were not effective in protecting the mucosa against aspirin- or indomethacin-induced gastric lesions. Thus, blockade of the production of all 5-LO products, rather than antagonism of the actions of peptido-leukotrienes alone, effectively prevented mucosal damage. These results suggest that NSAID-induced vascular injury is a primary early event in gastric mucosal injury, and that it is mediated by a range of 5lipoxygenase products, not peptido-leukotrienes alone. The role of leukocyte accumulation in these events involving vascular injury could also be of significance. Detailed pharmacological analysis of the role of leukotrienes, calcium ions and platelet activating factors (the latter two of which could contribute to the vascular and leukocyte-mediated gastric and intestinal injury) in disease-relevant states was examined by employing specific inhibitors or antagonists of these mediators in NSAID-induced injury in cholinomimeticstimulated mice and arthritic rats (Rainsford 1999b). The results showed that 5-LO products and intracellular calcium play key roles in the development of mucosal injury. The role of intracellular calcium may be extensive in mediating cellular signalling and could include the activation of phospholipase A2. The question of the mechanism by which 5-lipoxygenase products are involved in the development of NSAIDassociated mucosal injury has, at its basis, a postulate that cyclo-oxygenase inhibition by NSAIDs may cause diversion of arachidonic acid through the 5-lipoxygenase pathway, leading to increased production of peptido-LTB4. This would result in accumulation and activation of PMNs and monocytes, while the latter would be expected to cause dilatation of postcapillary venules in the mucosa, thus accounting for the observed extravasation of red blood cells and accumulated leucocytes into the interstitial regions adjacent to damaged blood vessels. That increased production of 5-LO products occurs with treatment by indomethacin has been shown using GC–MS and RIA assays of leukotrienes in both the gastric mucosa of rodents (Rainsford 1999b) and the gastric efferent circulation of pigs. More recent extensive studies since this early postulate have highlighted the role of both vascular and leucocyte adhesion molecules, whose increased expression in
the mucosa occurs during early stages following administration of NSAIDs. Activation of signal transduction pathways that mediate proinflammatory cytokine production, expression of adhesion molecules and proteosome activation have added further important understanding to the cellular and molecular events underlying the actions of NSAIDs in the interaction of vascular injury and leukocyte activation, which may in part be a consequence of activation of the 5-LO pathway by these drugs. It is clear from these extensive studies and significant observations that there is a sound rationale for the development of dual inhibitors of both 5-LO and COX as potential GI-safer NSAIDs. The added benefits of controlling LOX-mediated inflammatory events may have considerable significance for controlling the underlying disease process in those states where there is appreciable involvement of leukocyte activation and invasion. Some NSAIDs that are COX-inhibitors also have inhibitory activity on LOX pathways (e.g. meclofenamic acid; Stadler et al 1994) but their relative potency as LOX inhibitors is appreciably lower than that on COX pathways. The essential point about determining the optimal dual COX/LOX inhibitors is that their potency on both pathways should be about equivalent, especially if there is low GI irritancy and potent antiinflammatory activity that is achieved optimally in the one molecule. The question of whether the inhibition of COX-1 or COX-2 plays a major role in causing the diversion of arachidonate to produce excess 5-LO products is unresolved. Some insights may come from the actions of newly-developed COX/LOX inhibitors with varying degrees of COX-1 or COX-2 inhibitory activity. Development of Dual COX/5-LOX Inhibitors Since the earlier studies with BW-755c, benoxaprofen and some experimental agents (e.g. ETYA) with dual COX/5-LOX activity, several companies and groups have embarked upon programmes to discover and develop specific COX/5-LOX inhibitory NSAIDs. Among these developments have been: a-(3,5-di-tert-butyl-4hydroxybenzylidene)-g-butyrolactone (KME-4; Hidaka et al 1985); 2-acetylthiophene-2-thiazolylhydrazone (CBS-1108; Sincholle et al 1985); a range of styrylpyrazoles, styryloxazoles and styrylisothiazoles (Flynn et al 1991) modelled on curcumin and yakuchinone, which are dual COX/5-LOX inhibitors (Flynn et al 1991; Huang et al 1991); some NSAID quinoline hybrid drugs, e.g. quinoline-etodolac (WAY-120,739; 1,8-diethyl-1,3,4,9-tetrahydro-6-(2-quinolinylmethoxy)-pyrano-[3,4]-indole acetic acid; Kreft et al 1993); some C5-derivatives of 6,7-diphenyl-2,3dihydro-1H-pyrrolizines (Dannhardt and Keifer 1994), N-[5-(4flurophenoxy)-thien-2-yl]-methane sulphonamide (RWJ 63556; Kirchner et al 1997b); tepoxalin (5-(4-chlorophenyl)-N-hydroxy1(4-methoxyphenyl)-N-methyl-1H-pyrazole-3-propanamide (Figure 16.9); Knight et al 1996; Kirchner et al 1997a); a series of 5-ketosubstituted 7-tert-butyl-2,3-dihydro-3,3-dimethylfurans (Janusz et al 1998), a series of N-hydroxyurea and hydroxamic acid derivatives of tepoxalin (Connolly et al 1999); and an acetylated derivative of isaindigotone, 3-(40 -acetoxy-30 ,50 -dimethyloxy)-benzylidene-1,2-didropyrrolo-[2,1b]-quinazoline-9-ione (PQ; Rioja et al 2002). It is of interest that RWJ 63556 and PQ are selective COX-2/5-LOX inhibitors, whereas some others, e.g. licofelone (ML-3000) are COX-1/COX-2/5-LOX inhibitors (Tries et al 1997; Laufer 2001). Dannhardt et al (2000) examined the role of the pyrrole moiety as a template for development of COX/5-LOX inhibitors among a range of aroyl- and thiophene-substituted pyrrole derivatives. The pyrrole template confers both COX-1 and COX-2 inhibitory effects on these derivatives and Dannhardt et al
EICOSANOID INHIBITORS
Figure 16.12 Licofelone (ML-300)—a COX–LOX inhibitor that has been successfully employed in treating arthritic conditions and has demonstrated low GI toxicity
(2000) consider that the balanced inhibition of both these isoenzymes may confer advantages over that on COX-2 inhibitors alone. Among the most successful of the balanced or dual COX/5LOX inhibitors has been licofelone (ML-3000; 2,2-dimethyl-6-(4chlorophenyl)-7-phenyl-2,3-ihydro-1H-pyrrolizine-5-yl]-acetic acid; Figure 16.12) (Tries and Laufer 2001; Laufer et al 2003). Laufer (2001) has pointed out that the failure of some COX/5-LOX inhibitors (e.g. tepoxalin, KME-4, tebufelone) because of liver toxicity is probably due to their redox potential. Licofelone does not have this potential and does not appear to have shown marked liver toxicity in clinical trials. GLUCOCORTICOIDS AND PHOSPHOLIPASE INHIBITORS Glucocorticoids are multi-acting potent antiinflammatory agents that are devoid of the analgesic and antipyretic activities seen with the NSAIDs (Robinson 1989; Schleimer 1993; Rainsford 1994; Garcia-Leme 1996). Earlier views suggested that glucocorticoids inhibited the production of all arachidonic acid-derived lipids and platelet-activating factor (PAF) release by a receptor-mediated induced increase in a specific protein inhibitor (lipocortin) of phospholipase A2 (Flower and Blackwell 1979; Garcia-Leme 1996). More recent studies suggested that mRNA coding for the production of liportin I (renamed calpactin II and annexin I) was not induced by glucocorticoids (Bronnegard et al 1988), and other evidence questioned the role for lipocortins (Bienkowski et al 1989; Isacke et al 1989; Solito et al 1990). Some data suggested that in some in vitro systems the inhibitory activity of lipocortin I/ annexin I may be to affect substrate binding rather than by inhibition of the enzymatic activity of phospholipase A2 (Aarsman et al 1987; Davidson et al 1987). Other data have clearly established, however, that glucorticoids do regulate the secretion of annexin I (Solito et al 1991, 1994; Vishwanath et al 1992, 1993; Goulding et al 1990; Christmas et al 1991) and that annexin I has acute antiinflammatory activity (Cirono et al 1987, 1989). Annexin I-binding proteins have been shown on human monocytes and mast cells to produce annexin I and this is modulated by dexamethasone. Annexin I-like glucocorticoids reduce myocardial ischaemia reperfusion injury (D’Amico et al 2000; La et al 2001a,b). These data attest to the role of annexin I as part of the mode of action of glucocorticords. The discovery of a glucorticoid-regulated cyclooxygenase (later shown to be COX-2) that negatively regulated its induction by growth factors, IL-1 or other inflammagens and mitogens (Pash and Bailey 1988; DeWitt et al 1990; O’Banion et al 1991, 1992; O’Sullivan et al 1993; Winn et al 1993; Yamagata et al 1993: Masferrer et al 1994) has been shown to be a gene-based regulation by a so-called glucocorticoid response element (gre), located in the upstream of the main transcription site for COX-2 (Crofford 1997).
205
Thus, it can be seen that glucocorticoids have two potential inhibitory sites in eicosanoid metabolism, that involving the release of arachidonic acid from phospholipids via phospholipase A2, and that involved in the induction of the synthesis of COX-2 enzyme protein. A large number of steroid antiinflammatory analogues have been developed and explored for in vivo effects in inflammation assays and some in vitro activities, including steroid-receptor binding activity (Lee et al 1989), but many of these analogues have not been evaluated for their effects on phospholipase A2 or COX-2 inhibitory effects. Molecular features of the recognition sites of the glucocorticoid receptor have been identified (Anzali et al 1998) and no doubt this has been, and will be in future, of use in identifying novel inhibitors. An interesting development recently is the nitric oxide-donating analogue of prednisolone (NCX-1015), which has more potent antiinflammatory effects than prednisolone in in vivo rat collageninduced arthritis (Paul-Clark et al 2002) and a rat model of colitis induced by intrarectal 2,4,6-trinitrobenzene sulphonic acid (Fiorucci et al 2002). It is suggested by these authors that the additive effects of nitric oxide are due to its inhibitory effects on leukocyte activation and chemokine expression (Fiorucci et al 2002). These additive effects with the broad-acting effect of the steroid may account for the enhanced acute antiinflammatory activities of the NO derivative over that of the steroid in the two models of inflammation. Among the effects that are not related to direct effects on eicosanoid metabolism that may indirectly influence the release of these lipids are the inhibition of cytokine production (GarciaLeme 1996) and nuclear factor-kB (NF-kB) transcription factor, which may integrate with the glucocorticoid receptor and which regulates production of phospholipase A2, inducible nitric oxide synthase and a whole range of proinflammatory mediators and tissue-destructive metalloproteinases (Baek et al 2000; Triggiani et al 2000; Croxtall et al 2002). The search for selective phospholipase A2 inhibitors has proceeded now for two or three decades and has been based on the premise of developing corticosteroid inhibitors of both prostaglandin and leukotriene pathways of arachidonic acid metabolism and PAF, while at the same time avoiding the systemic complications from immunosuppressive effects of the corticosteroids. The identification of large amounts of phospholipase A2 in synovial fluid and from unstimulated chondrocytes in vitro (Pruzanski et al 1990), and the fact that intracellular production from chondrocytes is a major source of this enzyme in synovial fluid (Gilman and Chang 1990), together with observations that IL-1 could enhance its production (Lyons-Giordano et al 1989), gave impetus for the search for specific inhibitors of phospholipase A2. While most conventional antirheumatic drugs were found inactive against the synovial fluid enzyme (with the exception of moderate to weak effects by indomethacin, sulindac sulphide and MK-886; Marshall et al 1991; Lobo and Hoult 1994), a large number of inhibitors were screened for inhibitory effects derived from marine and other natural product sources and some moderately potent inhibitors were identified (Ferrandiz et al 1994; de las Heras et al 1994; Lorenzen et al 1995; Schrier et al 1996; Gil et al 1997; Lindahl and Tagesson 1997; Saini et al 1997; Witter et al 1998; Randazzo et al 1998: De Rosa et al 1998a,b; Garcia-Pastor et al 1999; Benrezzouk et al 1999; De Marino et al 2000; Posadas et al 2001; Shimoyama et al 2001). Many of these have been found to have antiinflammatory activity in vivo in animal models. Among the compounds identified, as examples of the range of types of natural inhibitors, were: sesquiterpenes from marine sponges (Ferrandiz et al 1994; Giannini et al 2001); a furanone from fermentations from Calyptella species (Lorezen et al 1995); plant-derived flavanoids and biflavanoids (Gil et al 1997; Lindahl and Tagesson 1997); and
206
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
a tripeptide, penidiamide, containing dehydrotryptamine, glycine and anthranilic acid isolated from Penicillium cultures (Witter et al 1998); and an inhibitor derived from python serum (Thwin et al 2002). Among synthetic derivatives that have been developed is the tri-aryl indomethacin-type analogue, WAY-121,520 (Kreft et al 1993; Glaser et al 1993), which was shown to have antiinflammatory effects in acute ear inflammation in mice and prevented leukotriene-induced bronchoconstriction in guinea-pigs (Glaser et al 1993). A prostaglandin analogue, PX-52, was found to be an inhibitor of phospholipase A2 in a number of in vitro systems (Franson and Rosenthal 1997). Other compounds include a novel triterpenoid, dysidotronic acid (Posadas et al 2001). It is of interest that among the actions of chondroitin sulphate in osteoarthritis it is claimed that the inhibition of phospholipase A2 may be part of its action (Ronca et al 1998). The 3-D protein structural determination of the crystalline state has been reported of various types of soluble phospholipase A2 from synovial fluid (Christensen et al 1993; Hariprasad and Kulkarni 1996; Church et al 2000) and molecular field analyses (Ortiz et al 1997) have provided a basis for the development of selective inhibitors of synovial phospholipase A2. It does not appear that inhibitors of this enzyme have been successfully developed for therapeutic use in arthritic diseases, but further developments are awaited with much interest.
CONCLUSIONS An immense number and type of inhibitors of eicosanoid metabolism have been developed, especially in the past two decades. Many of these have been developed on the basis of elegant structure determinations of the enzymes involved in the metabolism of arachidonic acid through the various oxygenase pathways, as well as from extensive structure–activity analyses with model compounds and derivatives. The specificity of drugs developed to selectively target certain enzymes has been quite striking, but debate continues about whether highly specific targets (e.g. COX-2 selective drugs) are necessarily the best approach for the control of certain diseases (e.g. chronic inflammatory conditions and pain), where it is clearly evident that multiple attacks on various pathways of arachidonic acid metabolism (e.g. COX/LOX inhibitors) may have benefit in the broader therapeutic sense. We await the development of novel modulators/regulators of the production of the newer class of arachidonate metabolites, the lipoxins, with much interest, since these may have important modulatory roles in a variety of inflammatory conditions.
REFERENCES Aarsman A et al (1987) Lipocortin inhibition of extracellular and intracellular phospholipases A2 is substrate concentration dependent. FEBS Lett, 219, 176–180. Anzali S, Gasteiger J, Hozgrabe U et al (1998) The use of self-organising neural networks in drug design. In Kuvinyi H, Folkers G, Martin YC (eds), 3D QSAR in Drug Design, vol 2, Ligand–Protein Interactions and Molecular Similarity. Kluwer/ESCOM, Dordrecht. Baek SH, Kwon TK, Lim JH et al (2000) Secretory phosholipase A2potentiated inducible nitric oxide synthase expression by macrophages requires NF-k B activation. J Immunol, 164(12), 6359–6365. Barton SF, Langeland FF, Snabes MC, LeComte D, Kuss ME, Dhadda SS, Hubbard RC (2002). Efficacy and safety of intravenous paracoxib sodium in relieving acute postoperative pain following gynecologic laparotomy surgery. Anesthesiology, 97(2): 306–314.
Benrezzouk R, Terencio MC, Ferrandiz ML et al (1999) Inhibition of human sPLA2 and 5-lipoxygenase activities by two neo-clerodane diterpenoids. Life Sci, 64(19), PL205–PL211. Benson W, Weaver A, Espinoza L, Zhao WW, Riley W, Paperiello B, Recker DP (2002) Efficacy and safety of valdecoxib in treating the signs and symptoms of rheumatoid arthritis: a randomized, controlled comparison with placebo and naproxen. Rheumatology, 41(9): 1008– 1016. Bertolini A, Ottani A, Sandrini M (2001) Dual acting anti-inflammatory drugs: a reappraisal. Pharmacological Research, 44(6): 437–450. Bienkowski MJ, Petro MA and Robinson LJ (1989). Inhibition of thromboxane A synthesis in U937 cells by glucocorticoids. Lack of evidence for lipocortin 1 as the second messenger. J Biol Chem, 264, 6536–6544. Bjarnason I and Thjodleifsson B (1999) Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs: the effect of nimesulide compared with naproxen on the human gastrointestinal tract. Rheumatology, 38(suppl 1), 24–32. Bjarnason I, Walley M, Hotz-Behofsits C (2003). COX-selective agents: the bad and not so bad news. In: International Conference on Inflammopharmacology, 22–24 April 2003. Bray MA, Gordon D (1978) Prostaglandin production by macrophages and the effect of anti-inflammatory drugs. Br J Pharmacol, 63(4): 635– 642. Bronnegard M et al (1988) Human calpactin II (lipocortin I) messenger ribonucleic acid is not induced by glucocorticoids. Mol Endocrinol, 2, 732–739. Brooks P, Emery P, Evans JF, Fenner H, Hawkey CJ, Patrono C, Smolen J, Breedveld F, Day R, Dougados M, Ehrich EW, Gijon-Banos J, Kvien TK, Van Rijswijk MH, Warner T, Zeidler H (1999) Interpreting the clinical significance of the differential inhibition of cyclooxygenase1 and cyclooxygenase-2. Rheumatology, 38(8): 779–788. Brune K, Rainsford KD, Wagner K and Peskar BM (1981) Inhibition by antiinflammatory drugs on prostaglandin production in cultured macrophages: factors influencing the apparent drug actions. Naunyn Schmiedeberg’s Arch Pharmacol, 315, 269–276. Callan OH, So OY, Swinney DC (1996) The kinetic factors that determine the affinity and selectivity for slow binding inhibition of human prostaglandin H synthase 1 and 2 by indomethacin and flurbiprofen. J Biol Chem, 271(7): 3548–3554. Capone ML, Tacconelli S, Sciulli MG et al. (2003) Clinical pharmacology of selective COX-2 inhibitors. Int J Immunopath Ph, 16(2): 49–58. Celotti F, Laufer S (2001) Anti-inflammatory drugs: new multitarget compounds to face an old problem. The dual inhibition concept. Pharmacological Research, 43(5): 429–436. Chandrasekharan NV, Dai H, Roos KL et al (2002) COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure and expression. Proc Natl Acad Sci USA, 99, 13926–13931. Christensen IT, Jorgensen FS, Svensson LA and Hogberg T (1993) Threedimensional structures of human phospholipase A2 from pancreas and synovial fluid by model building. Drug Dev Discov, 10(2), 101–113. Christmas P, Callaway J, Fallon J et al (1991) Selective secretion of annexin 1, a protein without a signal sequence, by the human prostate gland. J Biol Chem, 266, 2499–2507. Church WB, Lei PW, Ogg DJ and Scott KF (2000). Crystallization and preliminary X-ray diffraction studies of a new crystal form of human secretory type IIA phospholipase A2. Acta Crystallogr D Biol Crystallogr, 56(P11), 1482–1484. Cirino G, Flower RJ, Browning JL et al (1987) Recombinant human lipocortin 1 inhibits thromboxane release from guinea-pig isolated perfused lung. Nature (Lond), 328, 270–272. Cirino G, Peers SH, Flower RJ et al (1989) Human recombinant lipocortin 1 has acute local anti-inflammatory properties in the rat paw edema test. Proc Natl Acad Sci USA, 86, 3428–3432. Connolly PJ, Wetter SK, Beers KN et al (1999) N-Hydroxyurea and hydroxamic acid inhibitors of cyclooxygenase and 5-lipoxygenase. Bioorg Med Chem Lett, 9, 979–984. Crofford LJ (1997) COX-1 and COX-2 tissue expression: implications and predictions. J Rheumatol, 24(suppl 49), 15–19. Croxtall JD, van Hal PT, Choudhury Q et al (2002) Different glucocorticoids vary in their genomic and non-genomic mechanism of action in A549 cells. Br J Pharmacol, 135(2), 511–519.
EICOSANOID INHIBITORS Cucurou C, Battioni JP, Thang DC et al (1991) Mechanisms of inactivation of lipoxygenases by phenidone and BW755C. Biochemistry, 30(37), 8964–8970. Cullen L, Kelly L, Connor SO and Fitzgerald DJ (1998) Selective cyclooxygenase-2 inhibition by nimesulide in man. J Pharmacol Exp Ther, 287, 578–582. D’Amico M, Di Filippo C, La M et al (2000) Lipocortin 1 reduces myocardial ischemia-reperfusion injury by affecting local leukocyte recruitment. FASEB J, 14(13), 1867–1869. Dannhardt G and Kiefer W (1994) Non-steroidal antiinflammatory agents, XXXVIII: C-5 functionalized 6,7-diphenyl-2,3-dihydro-1Hpyrrolizines as inhibitors of bovine cyclooxygenase and 5-lipoxygenase. Arch Pharm (Weinheim), 327, 9–14. Dannhardt G, Laufer S (2000) Structural approaches to explain the selectivity of COX-2 inhibitors: is there a common pharmacophore? Current Medicinal Chemistry, 7, 1101–1112. Dannhardt G, Kiefer W, Kra¨mer G, Maehrlein S, Nowe U, Fiebich B (2000) The pyrrole moiety as a template for COX-1/COX-2 inhibitors. Eur J Med Chem, 35: 499–510. Dannhardt G and Keifer W (2001) Cyclooxygenase inhibitors—current status and future prospects. Eur J Med Chem, 36, 109–126. Davidson FF et al (1987) Inhibition of phospholipase A2 by ‘‘lipocortins’’ and calpactins; an effect of binding to substrate phospholipids. J Biol Chem, 262, 1698–1705. Dawson W (1980). Mechanisms of action of antiinflammatory drugs. In Samuelsson B, Ramwell PW, Paolett R (eds), Advances in Prostaglandin and Thromboxane Research, vol 8. Raven, New York. De las Heras B, Villar A, Vivas JM and Hoult JR (1994) Novel antiinflammatory plant labdanes: comparison of in vitro properties with aspirin and indomethacin. Agents Actions, 41(1–2), 114–117. De Leval X, Delarge J, Somers F, de Tullio P, Henrotin Y, Pirotte B, Dogne´ J-M (2000). Recent advances in inducible cyclooxygenase (COX-2) inhibition. Current Medicinal Chemistry, 7, 1041–1062. De Marino S, Iorizzi M, Zollo F et al (2000) New pyridinium alkaloids from a marine sponge of the genus Spongia with a human phospholipase A(2) inhibitor profile. J Nat Prod, 63(3), 322–326. De Rosa M, Crispino A, De Giulio A et al (1998a). A new cacospongionolide inhibitor of human secretary phospholipase A2 from the Tyrrhenian sponge Fasciospongia cavernosa and absolute configuration of cacospongionolides. J Nat Prod, 61(7), 931–935. De Rosa M, Giordano S, Scettri A et al (1998b). Synthesis and comparison of the anti-inflammatory activity of manoalide and cacospongionolide B analogues. J Med Chem, 41(17), 3232–3238. Dewhirst FE (198) Structure-activity relationships for inhibition of prostaglandin cyclooxygenase by phenolic compounds. Prostaglandins, 20(2), 209–222. DeWitt DL et al (1990) Serum induction and superinduction of PGG/H synthase mRNA levels in 3T3 fibroblasts. In Samuelsson B, Paoletti R and Ramwell PW (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Research, vol 21. Raven, New York, 65–68. Dubois RN, Abramson SB, Crofford L et al (1998) Cyclooxygenase in biology and disease. FASEB J, 12, 1063–1071. Feldman M and McMahon AT (2000) Do cyclooxygenase-2 inhibitors provide benefit similar to those of traditional non-steroidal antiinflammatory drugs, with less gastrointestinal toxicity? Ann Intern Med, 132, 134–143. Fenner H (1997) Differentiating among nonsteroidal antiinflammatory drugs by pharmacokinetic and pharmacodynamic profiles. Semin Arthritis Rheum, 26 (Suppl 1), 28–33. Ferrandiz ML, Sanz MJ, Bustos G et al (1994) Avarol and avarone, two new anti-inflammatory agents of marine origin. Eur J Pharmacol, 253(1–2), 75–82. Fiorucci S, Antonelli E, Distrutti E et al (2002) NCX-1015, a nitric-oxide derivative of prednisolone, enhances regulatory T cells in the lamina propria and protects against 2,4,6-trinitrobenzene sulfonic acid-induced colitis in mice. Proc Natl Acad Sci USA, 99(24), 15770– 15775. Flower RJ (1974) Drugs which inhibit prostaglandin biosynthesis. Pharmacol Rev, 26, 33–67. Flower RJ and Blackwell GJ (1979) Anti-inflammatory steroids induce biosynthesis of a phospholipase A2 inhibitor which prevents prostaglandin generation. Nature, 278, 456–459.
207
Flynn DL, Belliotti TR, Boctor AM et al (1991) Strylpyrazoles, styrylisoxazoles, and styrylisothiazoles. Novel 5-lipoxygenase and cyclooxygenase inhibitors. J Med Chem, 34, 518–525. Forghani F, Ouellet M, Keen S et al (1998) Analysis of prostaglandin G/H synthase-2 inhibition using peroxidase-induced luminol luminescence. Anal Biochem, 264(2), 216–221. Franson RC and Rosenthal MD (1997) PX-52, a novel inhibitor of 14 kDa secretory and 85 kDa cytosolic phospholipases A2. Adv Exp Med Biol, 400A, 365–373. Freston JW (1999) Rationalizing cyclooxygenaase (COX) inhibition for maximal efficacy and minimal adverse events. Am J Med, 107, 78S– 89S. Garcia-Leme J (1996) Recent advances in the understanding of the mode of action of glucocorticoids. In Rainsford KD (ed.), Advances in Antirheumatic Therapy. CRC Press, Boca Raton, FL. Garcia-Pastor P, Randazzo A, Gomez-Paloma L et al (1999) Effects of petrosaspongiolide M, a novel phospholipase A2 inhibitor, on acute and chronic inflammation. J Pharmacol Exp Ther, 289(1), 166–172. Giannini C, Debitus C, Lucas R et al (2001) New sesquiterpene derivatives from the sponge Dysidea species with a selective inhibitor profile against human phospholipase A2 and other leukocyte functions. J Nat Prod, 64(5), 612–615. Gierse JK, Hauser SD, Creely DP et al (1995) Expression and selective inhibition of the constitutive and inducible forms of human cyclooxygenase. Biochem J, 305, 479–484. Gil B, Sanz MJ, Terencio MC et al (1997) Morelloflavone, a novel biflavonoid inhibitor of human secretary phospholipase A2 with antiinflammatory activity. Biochem Pharmacol, 53, 733–740. Gilman SC and Chang J (1990) Characterization of interleukin 1 induced rabbit chondrocyte phospholipase A2. J Rheumatol, 17(10), 1392– 1396. Glaser KB, Carlson RP, Sung A et al (1993) Pharmacological characterization of WAY-121.520: a potent anti-inflammatory indomethacin-based inhibitor of 5-lipoxygenase (5-LO)/ phospholipase A2 (PLA2). Agents Act, 39, C30–C32. Glaser KB (1995) Cyclooxygenase selectivity and NSAIDs: cyclooxygenase-2 selectivity of etodolac (Lodine). Inflammopharmacology, 3, 335–345. Goulding NJ, Godolphil JL, Sharland PR et al (1990) Anti-inflammatory lipocortin 1 production by peripheral blood leucocytes in response to hydrocortisone. Lancet, 335, 1416–1418. Gyo¨mber, E, Vattay P, Szabo S and Raiinsford KD (1996) Peptic ulcer: pathogenesis, healing and anti-ulcer drugs. Effect of lipoxygenase inhibitors and leukotriene antagonists on acute and chronic gastric haemorrhagic mucosal lesions in ulcer models in the rat. J Gastroenterol Hepatol, 11, 922–927. Habicht J, Brune K (1983) Inhibition of prostaglandin E2 release by salicylates, benzoates and phenols: a quantitative structure-activity study. J Pharm Pharmacol, 35(11): 718–723. Halter F, Rainsford KD, Sirko SP, Schmassmann A (1997) NSAIDinduced mucosal injury: Analysis of gastric toxicity of new generation NSAIDs: Ulcerogenicity compared with ulcer healing. Yale J Biol Med, 70: 33–43. Harada Y, Kawamura M, Hatanaka K, Saito M, Ogino M, Ohno T, Ogino K, Yang Q (1998) Differing profiles of prostaglandin formation inhibition between selective prostaglandin H synthase-2 inhibitors and conventional NSAIDs in inflammatory and non-inflammatory sites of the rat. Prostaglandins and other Lipid Mediators, 55, 345–358. Hariprasad V and Kulkarni VM (1996) A molecular dynamics study of the three-dimensional model of human synovial fluid phospholipase A2transition state mimic complexes. J Mol Recognit, 9(2), 95–102. Harris SI, Kuss M, Hubbard RC, Goldstein JL (2001) Upper gastrointestinal safety evaluation of parecoxib sodium, a new parenteral cyclooxygenase-2-specific inhibitor, compared with ketorolac, naproxen, and placebo. Clin Ther, 23(9): 1422–1428. Hawkey C, Laine L, Simon T et al (2000) Comparison of the effect of rofecoxib (a cyclooxygenase 2 inhibitor), ibuprofen, and placebo on the gastroduodenal mucosa of patients with osteoarthritis. A randomized, double-blind, placebo-controlled trial. Arthr Rheum, 43, 370–377. Hawkey CJ (1999) COX-2 inhibitors. Lancet, 353, 307–314. Hidaka T, Takeo K, Hosoe K et al (1985) Inhibition of polymorphonuclear leukocyte 5-lipoxygenase and platelet cyclooxygenase by a-(3,5-di-tert-butyl-4-hydroxybenzylidene)-g-
208
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
butyrolactone (KME-4), a new anti-inflammatory drug. Japan J Pharmacol, 38, 267–272. Hoffman M, Autor AP (1982) Effect of cyclooxygenase inhibitors and protease inhibitors on phorbol-induced stimulation of oxygen consumption and superoxide production by rat pulmonary macrophages. Biochem Pharmacol, 31(5): 775–780. Hsuanyu YC and Dunford HB (1992) Reduction of prostaglandin H synthase compound II by phenol and hydroquinone and the effect of indomethacin. Arch Biochem Biophys, 292(1), 213–220. Huang J-Q, Sridhar S and Hunt RH (1999) Am J Med, 107, 55S–64S. Huang MT, Lysz T, Ferraro T et al (1991) Inhibitory effects of curcumin on in vitro lipoxygenase and cyclooxygenase activities in mouse epidermis. Cancer Res, 51(3), 813–819. Hughes MF, Mason RP and Eling TE (1988) Prostaglandin hydroperoxidase-dependent oxidation of phenylbutazone: relationship to inhibition of prostaglandin cyclooxygenase. Mol Pharmacol, 34(2), 186–193. Humes J, Winter CA, Sadowski SJ, Kuehl FA (1981) Multiple sites on prostaglandin cyclooxygenase are determinants in the action of nonsteroidal antiinflammatory agents. Proc Natl Acad Sci, 78(4): 2053–2056. Isacke CM, Lindberg RA and Hunter T (1989) Synthesis of p36 and p35 is increased when U937 cells differentiate in culture but expression is not inducible by glucocorticoids. Mol Cell Biol, 9, 232–240. Issekutz AC, Szpejda M and Ripley M (1986) Effects of lipoxygenase and cyclooxygenase inhibitors on neutrophil margination, emigration, and related inflammatory responses. Adv Inflamm Res, 11, 31–45. Johnson JL and Maddipati KR (1998) Paradoxical effects of resveratrol on the two prostaglandin H synthases. Prostagland Lipid Mediat, 56(2–3), 131–143. Kimura K (1997) Mechanisms of active oxygen species reduction by nonsteroidal anti-inflammatory drugs. Int J Biochem Cell Biol, 29(3), 437– 446. Kirchner T, Aparicio B, Argentieri DC et al (1997a) Effects of tepoxalin, a dual inhibitor of cyclooxygenase/5-lipoxygenase, on events associated with NSAID-induced gastrointestinal inflammation. Prostagland Leukotrienes Essent Fatty Acids, 56(6), 417–423. Kirchner T, Argentieri DC, Barbone AG et al (1997b) Evaluation of the anti-inflammatory activity of a dual cycloxygenase-2 selective/5lipoxygenase inhibitor, RWJ 63556 in a canine model of inflammation. J Pharmacol Exp Therapeut, 282, 1094–1101. Kivitz A, Moskovitz RW, Woods E, Hubbard RC, Verburg KM, Lefkowith JB, Geis GS (2001) Comparative efficacy and safety of celecoxib and naproxen in the treatment of osteoarthritis of the hip. J Int Med Res, 29(6): 467–479. Knight EV, Kimball JP, Keenan CM et al (1996) Preclinical toxicity evaluation of tepoxalin, a dual inhibitor of cyclooxygenase and 5lipoxygenase, in Sprague–Dawley rats and beagle dogs. Fund Appl Toxicol, 33(1), 38–48. Kreft A, Nelson J, Musser J et al (1993) Structure–activity relationships leading to WAY-121,520, a tris aryl-type, indomethacin-based, phospholipase A2 (PLA2)/leukotriene biosynthesis inhibitor. Agents Act, 39, C33–C35. Kuehl FA Jnr, Humes JL, Ham EA et al (1980) Inflammation: the role of peroxidase-dervied products. Adv Prostaglandin Thromboxane Res, 6, 77–86. Kulkarni SK, Jain NK, Singh A (2000) Cyclooxygenase isoenzymes and newer therapeutic potential for selective COX-2 inhibitors. Methods Find Exp Clin Pharmacol, 22(5): 291–298. La M, D’Amico M, Bandiera S et al (2001a) Annexin 1 peptides protect against experimental myocardial ischemia-reperfusion: analysis of their mechanism of action. FASEB J, 15(12), 2247–2256. La M, Tailor A, D’Amico M et al (2001b) Analysis of the protection afforded by annexin 1 in ischaemia-reperfusion injury: focus on neutrophil recruitment. Eur J Pharmacol, 429(1–3), 263–278. Laine L, Harper S, Simon T et al (1999) A randomized trial comparing the effect of robecoxib, a cyclooxygenase 2-specific inhibitor, with that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology, 117, 776–783. Laneuville O, Breuer DK, De Witt DL et al. (1994) Differential inhibition of human prostaglandin endoperoxide-H synthase-1 and synthase-2 by nonsteroidal antiinflammatory drugs. J Pharmacol Exp Ther, 271(2): 927–934.
Langman MJ, Jensen DM, Watson DJ et al (1999) Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. J Am Med Assoc, 282, 1929–1933. Laufer S (2001) Discovery and development of ML3000. Inflammopharmacology, 9, 101–112. Lee HJ, Heiman AS and Taraporewala IB (1989) New steroidal antiinflammatory drugs. In Rainsford KD and Velo GP (eds), New Developments in Antirheumatic Therapy. Inflammation and Drug Therapy Series, vol III. Kluwer Academic, Lancaster, 153– 186. Leung AT, Malmstrom K, Gallacher AE, Sarembock B, Poor G, Beaulieu A, Castro R, Sanchez M, Detora LM, Ng J (2002) Efficacy and tolerability profile of Etoricoxib in patients with osteoarthritis: a randomized, double-blind, placebo and active-comparator controlled 12-week efficacy trial. Curr Med Res Opin, 18(2): 49–58. Li JJ, Black WC, Brideau C, Chan CC, Charleson S, Cromlish WA, Claveau D, Gauthier JY, Gordon R, Greig G, Grimm E, Guay J, Lau CK, Riendeau D, The´rien M, Visco DM, Wong E, Xu L, Prasit P (1999). A new structural variation on the methanesulfonylphenyl class of selective cyclooxygenase-2 inhibitors. Bioorganic & Medicinal Chemistry Letters, 9: 3181–3186. Lindahl M and Tagesson C (1997) Flavonoids as phospholipase A2 inhibitors: importance of their structure for selective inhibition of group II phospholipase A2. Inflammation, 21(3), 347–356. Lobo IB and Hoult JR (1994) Groups I, II and III extracellular phospholipases A2: selective inhibition of group II enzymes by indomethacin but not other NSAIDs. Agents Act, 41(1–2), 111– 113. Lorenzen K, Anke T, Konetschyny-Rapp S and Scheuer W (1995) 5hydroxy-3-vinyl-2(5H)-furanone—a new inhibitor of human synovial phospholipase A2 and platelet aggregation from fermentations of a Calyptella species (Basidiomycetes). Z Naturforsch [C], 50(5–6), 403– 409. Lyons-Giordano B, Davis GL, Galbraith W et al (1989) Interleukin-1 b stimulates phospholipase A2 mRNA synthesis in rabbit articular chondrocytes. Biochem Biophys Res Commun, 164(1), 488–495. Makarowski W, Zhao WW, Bevirt T, Recker DP (2002) Efficacy and safety of the COX-2 specific inhibitor valdecoxib in the management of osteoarthritis of the hip: a randomized, double-blind, placebocontrolled comparison with naproxen. Osteoarthritis Cartilage, 10(4): 290–296. Marot C, Chavatte P, Lesieur D (2000) Comparative molecular field analysis of selective cyclooxygenase-2 (COX-2) inhibitors. Quant Struct-Act Relat, 19: 127–134. Marshall LA, Bauer J, Sung ML and Chang JY (1991) Evaluation of antirheumatic drugs for their effect in vitro on purified human synovial fluid phospholipase A2. J Rheumatol, 18(1), 59–65. Masferrer JL, Reddy ST, Zweifel BS et al (1994) In vivo glucocorticoids regulate cyclooxygenase-2 but not cyclooxygenase-1 in peritoneal macrophages. J Pharmacol Exp Therapeut, 270(3), 1340–1344. McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA, Fitzgerald GA (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: The human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96: 272–277. Mehler EL, Gerhards J (1987) Electronic determinants of the antiinflammatory action of benzoic and salicylic acids. Mol Pharmacol, 31(3): 284–293. Mehler EL, Habicht J, Brune K (1982) Quantum chemical analysis of structure-activity relationships in nonsteroidal anti-inflammatory drugs. Mol Pharmacol, 22(3): 525–528. Mohammed S and Croom DW (1999) Gastropathy due to celecoxib. N Engl J Med, 340, 2005. Morham SG, Langenbach R, Loftin CD et al (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 83, 473–482. Narayanan BA, Narayanan NK, Simi B et al (2003) Modulation of inducible nitric oxide synthase and related proinflammatory genes by the omega-3 fatty acid docosahexaenoic acid in human colon cancer cells. Cancer Res, 63, 972–979. Neve J, Parij N and Moguilevsky N (2001) Inhibition of the myeloperoxidase chlorinating activity by non-steroidal antiinflammatory drugs investigated with a human recombinant enzyme. Eur J Pharmacol, 417(1–2), 37–43.
EICOSANOID INHIBITORS O’Banion MK, Sadowski HB, Winn V and Young DA (1991) A serumand glucocorticoid-regulated 4-kilobase mRNA encodes a cycloxygenase-related protein. J Biol Chem, 266(34), 23261–23267. O’Banion MK, Winn VD and Young DA (1992) CDNA cloning and functional activity of glucocorticoid-regulated inflammatory cyclooxygenase. Proc Natl Acad Sci USA, 89, 4888–4892. O’Sullivan MG, Huggins EM Jr and McCall CE (1993) Lipopolysaccharide-induced expression of prostaglandin H synthase2 in alveolar macrophages is inhibited by dexamethasone but not by aspirin. Biochem Biophys Res Commun, 191(3), 1294–1300. Ortiz AR, Pastor M, Palomer A et al (1997) Reliability of comparative molecular field analysis models: effects of data scaling and variable selection using a set of human synovial fluid phospholipase A2 inhibitors. J Med Chem, 40, 1136–1148 (see also p 4168). Pairet M and van Ryn J (1998) Experimental models used to investigate the differential inhibition of cyclooxygenase-1 and cyclooxygenase-2 by non-steroidal antiinflammatory drugs. Inflam Res, 47 (suppl 2), S93– S101. Pash JM and Bailey JM (1988) Inhibition by corticosteroids of epidermal growth factor-induced recovery of cyclooxygenase after aspirin inactivation. FASEB J, 2(9), 2613–2618. Patrignani P, Capone ML, Tacconelli S (2003) Clinical pharmacology of Etoricoxib: a novel selective COX-2 inhibitor. Expert Opin Pharmaco, 4(2): 265–284. Paul-Clark MJ, Mancini L, Del Soldato P et al (2002) Potent antiarthritis properties of a glucocorticoid derivative, NCX-1015, in an experimental model of arthritis. Proc Natl Acad Sci USA, 99(3), 1677–1682. Pierson R (2000) US Regulators likely to scrutinize Vioxx adverse events. Reuters Medical News, 5 March 2000. Posadas I, Terencio MC, De Rosa S and Paya M (2000) Cavernolide: a new inhibitor of human sPLA2 sharing unusual chemical features. Life Sci, 67(24), 3007–3014. Posadas I, Terencio MC, Giannini C et al (2001) Dysidotronic acid, a new sesquiterpenoid, inhibits cytokine production and the expression of nitric oxide synthase. Eur J Pharmacol, 415(2–3), 285–292. Prasit P and Riendeau D (1997) Selective cyclooxygenase-2 inhibitors. In Annual Reports in Medicinal Chemistry-32, Academic Press. Pruzanski W, Bogoch E, Stefanski E et al (1990) Synthesis and release of phospholipase A2 by unstimulated human articular chondrocytes. J Rheumatol, 17(10), 1386–1391. Rainsford KD (1994) Non-steroidal anti-inflammatory drugs. In Dale MM, Foreman JC and Fan T-P (eds), Textbook of Immunopharmacology. Blackwell, London, 309–319. Rainsford KD (1981) Comparison of the gastric ulcerogenic activity of new non-steroidal anti-inflammatory drugs in stressed rats. Br J Pharmacol, 73, 79c–80c. Rainsford KD (1997) Gastrointestinal adaptation, regulation of eicosanoids, and mucosal protection from NSAIDs. In Rainsford KD (ed.), Side Effects of Anti-Inflammatory Drugs, IV, Proceedings of the IVth International Meeting on Side Effects of Anti-inflammatory Drugs, Sheffield, 7–9 August 1995; 197–205. Rainsford KD (1999a) Profile and mechanism of gastrointestinal and other side effects of non-steroidal anti-inflammatory drugs. Am J Med, 107, 27S–36S. Rainsford KD (1999b) Inhibition by leukotriene inhibitors, and calcium and platelet-activating factor antagonists, of acute gastric and intestinal damage in arthritic rats and in cholinomimetic mice. J Pharm Pharmacol, 51, 331–339. Rainsford KD (1999c) Relationship of nimesulide safety to its pharmacokinetics: assessment of adverse reactions. Rheumatol, 38(suppl 1), 4–10. Rainsford KD (2001) The ever-emerging anti-inflammatories. Have there been any real advances? J Physiol (Paris), 95, 11–19. Rainsford KD (2004) Pharmacology and Biochemistry of Salicylates and Related Drugs. In: Rainsford KD (ed). Aspirin and Related Drugs. Taylor & Francis, London, 215–366. Rainsford KD, Fox SA and Osborne DJ (1984) Comparative effects of some non-steroidal anti-inflammatory drugs on the ultrastructural integrity and prostaglandin levels in the rat gastric mucosa: relationship to drug uptake. Scand J Gastroenterol, 19, (suppl 101), 55–68.
209
Rainsford KD and Quadir M (1995) Gastrointestinal damage and bleeding from non-steroidal anti-inflammatory drugs. I. Clinical and epidemiological aspects. Inflammopharmacology, 3, 169–190. Rainsford KD, Tsang S, Hunt RH, Al-Jehani N (1995) Effects of nonsteroidal anti-inflammatory drugs on prostaglandin H synthase isoenzyme 2 (cyclooxygenase 2) production by porcine gastric mucosa in organ culture. Inflammopharmacology, 3, 299–310. Rainsford KD and Whitehouse MW (1976) Gastric irritancy of aspirin and its congeners: anti-inflammatory activity without this side-effect. J Pharm Pharmacol, 28, 599–601. Rainsford KD and Whitehouse MW (1980a) Anti-inflammatory/antipyretic salicylic acid esters with low gastric ulcerogenic activity. Agents Act, 10, 451–456. Rainsford KD and Whitehouse MW (1980b) Esterification of acidic antiinflammatory drugs suppresses their gastro-toxicity without adversely affecting their anti-inflammatory activity in rats. J Pharm Pharmacol, 32, 795–796. Rainsford KD and Willis C (1982) Relationship of gastric mucosal damage induced in pigs by anti-inflammatory drugs to their effects on prostaglandin production. Dig Dis Sci, 27, 624–635. Rainsford KD, Willis CM, Walker SA and Robins PG (1982) Electronmicroscopic observations comparing the gastric mucosal damage induced in rats and pigs by benoxaprofen with aspirin, reflecting their differing actions as prostaglandin synthesis inhibitors. Br J Exp Pathol, 63, 225–234. Randazzo A, Debitus C, Minale L et al (1998). Petrosaspongiolides M-R: new potent and selective phospholipase A2 inhibitors from the New Caledonian marine sponge Petrosaspongia nigra. J Nat Prod, 61, 571– 575. Reuben SS and Steinberg R (1999) Gastric perforation associated with the use of celecoxib. Anesthesiology, 91, 1548–1549. Ribeiro ML, Cella M, Farina M et al (2003) Cross talk between nitric oxide synthase and cyclooxygenase metabolites in the oestrogenized rat uterus. Prostaglandins leukot Essent Fatty Acids, 68, 285–290. Rioja I, Terencio MC, Ubeda A et al (2002) A pyrroloquinazoline derivative with anti-inflammatory and analgesic activity by dual inhibition of cyclo-oxygenase-2 and 5-lipoxygenase. Eur J Pharmacol, 434, 177–185. Robinson DR (1989) Eicosanaoids, inflammation, and anti-inflammatory drugs. Clin Exp Rheumatol, 7(suppl 3), S155–S161. Ronca F, Palmieri L, Panicucci P and Ronca G (1998) Anti-inflammatory activity of chondroitin sulfate. Osteoarthrit Cart, 6(suppl A), 14– 21. Rossat J, Maillard M, Nussberger J et al (1999) Renal effects of selective cyclooxygenase-2 inhibition in normotensive salt-depleted subjects. Clin Pharmacol Ther, 66, 76–84. Saini SS, Peterson JW and Chopra AK (1997) Melittin binds to secretary phospholipase A2 and inhibits its enzymatic activity. Biochem Biophys Res Commun, 238(2), 436–442. Sawaoka H, Tsujii M, Nakama A et al (1997) Expression of the cyclooxygenase-2 gene in gastric epithelium. J Clin Gastroenterol, 25(suppl1), S105–S110. Schleimer RP (1993) An overview of glucocorticoid anti-inflammatory actions. Eur J Clin Pharmacol, 45(suppl 1), S3–S7. Schmassmann A, Peskar BM, Stettler C, Netzer P, Stroff T, Flogerzi B, Halter F (1998) Effects of inhibition of prostaglandin endoperoxide synthetase-2 in chronic gastro-intestinal ulcer models in rats. Br J Pharmacol, 123(5): 795–804. Schoenfeld P (1999) Gastrointestinal safety profile of meloxicam: a metaanalysis and systematic review of randomized controlled trials. Am J Med, 107, 48S–54S. Schrier DJ, Flory CM, Finkel M et al (1996) The effects of the phospholipase A2 inhibitor, manoalide, on cartilage degradation, stromelysin expression, and synovial fluid cell count induced by intraarticular injection of human recombinant interleukin-1 a in the rabbit. Arthritis Rheum, 39(8), 1292–1299. Schumacher HR Jr, Boice JA, Daikh DI, Mukhopadhyay S, Malmstrom K, Ng J, Tate GA, Molina J (2002) Randomised double-blid trial of Etoricoxib and indomethacin in treatment of acute gouty arthritis. BMJ, 324(7352), 1488–1492. Selinsky BS, Gupta K, Sharkey CT et al. (2001) Structural analysis of NSAID binding by prostaglandin H-2 synthase: time-dependent and time-independent inhibitors elicit identical enzyme conformations. Biochemistry, 40(17), 5172–5180.
210
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Serhan CN (1997). Lipoxins and novel aspirin-triggered 15-epi-lipoxins (ATL): a jungle of cell–cell interactions or a therapeutic opportunity? Prostaglandins, 53, 107–137. Shah AA, Murray FE and Fitzgerald DJ (1999) The in vivo assessment of nimesulide cyclooxygenase-2 selectivity. Rheumatology, 38(suppl1), 19– 23. Shimoyama Y, Sakamoto R, Akaboshi T et al (2001) Characterization of secretary type IIA phospholipase A2 (sPLA2-IIA) as a glycyrrhizin (GL)-binding protein and the GL-induced inhibition of the CK-IImediated stimulation of sPLA2-IIA phospholipase A2 (sPLA2-IIA) as a glycyrrhizin (GL)-binding protein and the GL-induced inhibition of the CK-II-mediated stimulation of sPLA2-IIA activity in vitro. Biol Pharm Bull, 24(9), 1004–1008. Sikes DH, Agrawal NM, Zhao WW, Kent JD, Recker DP and Verburg KM (2002) Incidence of gastroduodenal ulcers associated with valdecoxib compared with that of ibuprofen and diclofenac in patients with osteoarthritis. Eur J Gadstroenterol Hepatol, 14(10), 1101–1111. Sincholle D, Bertez C, Legrand A, Conduzorgues JP, Bonne C (1985) Anti-inflammatory activity of a dual inhibitor of cyclooxygenase and lipoxygenase pathways, CBS-1108 (2-acetylthiophene-2thiazolylhydrazone). Arzneimittelforschung, 35, 1260–1263. Smith WL, Meade EA and DeWitt DL (1994) Interaction of PGH synthase isozymes-1 and –2 with NSAIDs. Ann NY Acad Sci, 744, 50– 57. Solito E et al. (1990) Effect of dexamethasone and phorbol myristate acetate on lipocortin 1,2 and 5 mRNA and protein synthesis. In Samuelsson B, Paoletti R and Ramwell PW (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Research, vol. 21. Raven, New York, 291–294. Solito E, Raugei G, Melli M and Parente L (1991) Dexamethasone induces the expression of the mRNA of lipocortin 1 and 2 and the release of lipocortin 1 and 5 in differentiated but not undifferentiated U-937 cells. FEBS Lett, 291, 238–244. Solito E, Nuti S and Parente L (1994) Dexamethasone-induced translocation of lipocortin (annexin) 1 to the cell membrane of U937 cells. Br J Pharmacol, 112, 347–348. Song Y, Connor DT, Sercel AD, Sorenson RJ, Doubleday R, Unangst PC, Roth BD, Beylin VG, Gilbertsen RB, Chan K, Schrier DJ, Guglietta A, Bornemeier DA, Dyer RD (1999a) Synthesis, structureactivity relationships and in vivo evaluations of substituted Di-tertbutylphenols as a novel class of potent, selective and orally active cyclooxygenase-2 inhibitors. 2. 1,3,4- and 1,2,4-thiadiazole series. J Med Chem, 42, 1161–1169. Song Y, Connor DT, Doubleday R, Sorenson RJ, Sercel AD, Unangst PC, Roth BD, Gilbertsen RB, Chan K, Schrier DJ, Guglietta A, Bornemeier DA, Dyer RD (1999b) Synthesis, structure-activity relationships and in vivo evaluations of substituted Di-tertbutylphenols as a novel class of potent, selective and orally active cyclooxygenase-2 inhibitors. 1. Thiazolone nd Oxazolone Series. J Med Chem, 42, 1151–1160. Stadler I, Kapui Z and Ambrus JL (1994) Study on the mechanisms of action of sodium meclofenamic acid (Meclomen) a ‘‘double inhibitor’’ of the arachidonic acid cascade. J Med, 25, 371–382. Stichtenoth DO, Wagner B and Froehlich JC (1998) Effect of selective inhibition of the inducible cyclooxygenase on renin release in healthy volunteers. J Investig Med, 46, 290–296. Stichtenoth DO, Fro¨lich JC (2003) The second generation of COX-2 inhibitors: what advantage do the newest offer? Drugs, 63(1), 33–45. Surierkosz TA, Mitchell JA, Warner TD et al (1995) Co-induction of nitric oxide synthase and cyclo-oxygenase: interactions between nitric oxide and prostanoids. Br J Pharmacol, 114, 1335–1342. Tacconelli S, Capone ML, Sciulli MG, Ricciotti E and Patrignani P (2002) The biochemical selectivity of novel COX-2 inhibitors in whole blood assays of COX-isozyme activity. Curr Med Res Opin, 18(8), 503–511. Talley JJ, Bertenshaw SR, Brown DL, Carter JS, Graneto MJ, Koboldt CM, Masferrer JL, Norman BH, Rogier DJ, Zweifel BS and Seibert K (1999). 4,5-diaryloxazole inhibitors of cyclooxygenase-2 (COX-2). Med Res Rev, 19(3), 199–208. Tam SW, Saha JK, Garvey DS, Schroeder JD, Shelekhin TE, Janero DR, Chen L, Glavin A and Letts LG (2000). Nitrosothiol-based NO-donors inhibit the gastrointestinal mucosal damaging actions of NSAIDs. Inflammopharmacology, 8, 81–88.
Thompson D and Eling T (1989) Mechanism of inhibition of prostaglandin H synthase by eugenol and other phenolic peroxidase substrates. Mol Pharmacol, 36(5), 809–817. Thwin MM, Satish RL, Chan ST and Gopalakrishnakone P (2002) Functional site of endogenous phospholipase A2 inhibitor from python serum. Eur J Biochem, 269(2), 719–727. Tries S and Laufer S (2001) The pharmacological profile of ML3000: a new pyrrolizine derivative inhibiting the enzymes cyclo-oxygenase and 5-lipoxygenase. Inflammopharmacology, 9, 113–124. Tries S, Laufer S and Wallace JL (1997) ML 3000 A non-redox dual Inhibitor of COX and 5-LOX: stomach sparing properties and proposed mechanism of action. In International Conference on Inflammopharmacology, San Francisco, 15–17 March 1997. Triggiani M, Granata F, Oriente A et al (2000) Secretory phospholipases A2 induce beta-glucuronidase release and IL-6 production from human lung macrophages. J Immunol, 164(9), 4908–4915. Trummlitz G (2003) Insight into the molecular basis of selective COX-2 inhibition. In Inflammopharmacology 2003 International Conference, Royal College of Physicians of Edinburgh, Edinburgh, 22–24 April 2003. Abstr 38, 58. Vane JR and Botting RM (1995) New insights into the mode of action of anti-inflammatory drugs. Inflam Res, 44(1), 1–10. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenase 1 and 2. Ann Rev Pharmacol Toxicol, 38, 97–120. Vishwanath BS, Frey FJ, Bradbury M et al (1992) Adrenalectomy decreases lipocortin-I messenger ribonucleic acid and tissue protein content in rats. Endocrinology, 130, 585–591. Vishwanath BS, Frey FJ, Bradbury MJ et al (1993) Glucocorticoid deficiency increases phospholipase A2 activity in rats. J Clin Invest, 92(4), 1974–1980. Wallace JL (1999) Selective COX-2 inhibitors: is the water becoming muddy? TiPS, 20, 4–6. Wallace JL and Cirino G (1994) The development of gastrointestinalsparing non-steroidal anti-inflammatory drugs. TiPS, 15, 405–406. Wallace JL, Del Soldato P (2003) The therapeutic potential of NONSAIDs. Fundam Clin Pharmacol, 17(1), 11–20. Wallace JL and Whittle BJR (1985) Role of prostanoids in the protective actions of BW 755C on the gastric mucosa. Eur J Pharmacol, 115, 45– 51. Warner TD, Giuliano F, Vojnovic I, Bukasa A, Mitchell JA and Vane JR (1999) Nonsteroid drug selectiveities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2 are associated with human gastrointestinal toxicity: a full in vitro analysis. Proc Natl Acad Scie USA, 96(13), 7563–7568. Whitehouse MW and Rainsford KD (1980). Esterification of acidic antiinflammatory drugs suppresses their gastrotoxicity without adversely affecting their anti-inflammatory activity in rats. J Pharm Pharmacol, 32(11), 795–796. Whitehouse MW and Rainsford KD (1982) A comparison of gastric ulcerogenic activities of different salicylates. In Pfeiffer CJ (ed.), Drugs and Peptic Ulcer, vol II. CRC Press, Boca Raton, FL, 127–141. Willoughby DA, Moore AR and Colville-Nash PR (2000) COX-1, COX-2 and COX-3 and the future treatment of chronic inflammatory disease. Lancet, 355(9204), 646–648. Winn VD, O’Banion K and Young DA (1993) Anti-inflammatory glucocorticoid action: inhibition of griPGHS, a new cyclooxygenase. J Lipid Mediat, 6, 101–111. Witter L, Anke T and Sterner O (1998) A new inhibitor of synovial phospholipase A2 from fermentations of Penicillium sp. 62–92. Z Naturforsch [C], 53(1–2), 60–74. Xiao G, Tsai AL, Palmer G et al (1997) Analysis of hydroperoxideinduced tyrosyl radicals and lipoxygenase activity in aspirin-treated human prostaglandin H synthase-2. Biochemistry, 36, 1836–1845. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Zoete V, Bailly F, Maglia F, Rougee M and Bensasson RV (1999) Molecular orbital theory applied to the study of nonsteroidal antiinflammatory drug efficiency. Free Radic Biol Med, 26, 1261–1266. Zoete V, Maglia F, Rouge´e M and Bensasson RV (2000) Mechanism of action in a 4,5-diarylpyrrole series of selective cyclo-oxygenase-2 inhibitors. Free Radical Biology & Medicine, 28(11), 1638–1641.
17 Biology and Chemistry of Products of the Isoprostane Pathway L. Jackson Roberts II and Jason D. Morrow Vanderbilt University, Nashville, TN, USA
One of the major targets of free radicals are polyunsaturated lipids, which undergo peroxidation reactions. A plethora of products are produced by free radical-induced lipid peroxidation (Gardner 1989; Porter et al 1995). In 1990, we reported the discovery that a series of prostaglandin (PG) F2-like compounds are produced by free radical-induced peroxidation of arachidonic acid in vivo (Morrow et al 1990). Although the initial discovery of F2-IsoPs was a novel finding in that it elucidated new products of lipid peroxidation, the importance of this discovery has evolved considerably over the last several years, encompassing several aspects of the biology and chemistry of IsoPs which will be reviewed in this chapter. Another very important aspect of the discovery of F2-IsoPs relates to measuring F2-IsoPs to assess oxidative stress. Free radicals have been hypothesized to play an important role in the pathogenesis of a wide variety of disease processes (Halliwell and Gutteridge 1990; Southorn and Powis 1988). However, a major impediment in translating these hypotheses to fact has been the lack of a reliable non-invasive approach to assess oxidative stress status in vivo in human beings (Halliwell and Grootveld 1987). In this regard, quantification of F2-IsoPs has proved to be a major advance in this area. This aspect of the discovery of F2-IsoPs has been discussed in detail elsewhere and therefore will not be reviewed in this chapter (Roberts and Morrow 1997, 2000; Roberts and Morrow 1996, 1997; Moore and Roberts 1998).
BIOCHEMISTRY OF THE FORMATION OF ISOPROSTANES The mechanism by which isoprostanes are formed is shown in Figure 17.1. Because these compounds are isomeric to PGF2a derived from the cyclooxygenase, they were termed F2-isoprostanes (F2-IsoPs). A unique aspect of the formation of IsoPs is that they are initially formed in situ on phospholipids and then released by phospholipase action, which can be a very dynamic process (Morrow et al 1992a). As noted in Figure 17.1, three arachidonyl radicals give rise to four F2-IsoP regioisomers, each of which are comprised of eight racemic diastereomers. Each regioisomer is designated by the carbon number on which the side chain hydroxyl group is located with the carboxyl carbon designated as C-1. This is in accordance with the nomenclature system established for IsoPs that has been approved by the Eicosanoid Nomenclature Committee, sanctioned by JCBN of IUPAC (Taber et al 1997). We carried out a study to establish the relative abundance The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
of the different F2-IsoP regioisomers. The results of this study suggested that there is no preferential formation of regioisomers or isomeric compounds within a regioisomer group other than what would be predicted from a non-enzymatic mechanism (Waugh et al 1997). Central in the pathway of formation of IsoPs are PGH2-like endoperoxides (H2-IsoPs). H2-IsoPs are reduced to form F-ring IsoPs (Figure 17.1). Recently we reported the discovery that glutathione is an important effector of the reduction of H2-IsoPs to F2-IsoPs and that other thiols could substitute for glutathione in this reaction (Morrow et al 1994b). In aqueous solutions, PGH2 is unstable and undergoes rearrangement to form PGE2 and PGD2, with a t1=2 of approximately 5 min (Nugteren and Hazelhof 1973). Therefore, we explored the possibility that the reduction of the H2-IsoP endoperoxides in vivo to F2-IsoPs may not be completely efficient. In this regard we found that abundant quantities of E-ring and D-ring IsoPs (E2/D2-IsoPs) are also produced in vivo (Morrow et al 1994b). Moreover, PGH2 can also undergo rearrangement to form a thromboxane ring and we have shown that isothromboxanes are also produced in vivo (Morrow et al 1996). ISOPROSTANES AS BIOEFFECTORS OF OXIDANT INJURY IsoPs have been found to exert potent biological activity and potentially mediate some of the adverse effects of oxidant injury. First, as mentioned, IsoPs are initially formed esterified on phospholipids. Molecular modelling of IsoP-containing phospholipids reveals them to be remarkably distorted molecules (Morrow et al 1992a). Thus, the formation of these abnormal phospholipids would be expected to exert profound effects on membrane fluidity and integrity, which is a well-known sequela of oxidant injury. Receptor-mediated Biological Actions One of the unique characteristics of IsoPs that contrasts with prostaglandins formed by the cyclooxygenase enzyme is that the side chains are predominantly orientated cis in relation to the cyclopentane ring (O’Connor et al 1984). Two IsoPs that were first available for biological testing were 15-F2t-IsoP (8-iso-PGF2a) and 15-E2t-IsoP (8-iso-PGE2). These differ from their respective counterparts derived from the cyclooxygenase by inversion of the upper side chain stereochemistry. We have previously shown
212
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 17.1 Pathways of formation of F2-isoprostanes
that both 15-F2t-IsoP and 15-E2t-IsoP are produced in vivo (Morrow et al 1994a, 1998). 15-F2t-IsoP and/or 15-E2t-IsoP have been shown to be potent vasoconstrictors in a variety of vascular beds, including the kidney (Morrow et al 1990, 1994b; Takahashi et al 1992; Fukunaga et al 1993b), lung (Kang et al 1993; Banerjee et al 1992; Janssen et al 2001; Janssen and Tazzeo 2002; Jourdan et al 1997; Kawikova et al 1996; John and Valentin 1997; Okazawa et al 1997), heart (Mobert et al 1997; Kromer and Tippins 1999), portal vein (Marley et al 1997), umbilical vein (Oliveira et al 2000), brain (Hoffman et al 1997; Hou et al 2000), retina (Lahaie et al 1998; Michoud et al 1998; Beauchamp et al 2001), carotid (Mohler et al 1996), aorta (Kromer and Tippins 1998; Wagner et al 1997; Zhang et al 1996) and lymphatics (Sinzinger et al 1997). We had previously explored the metabolic fate of 15-F2t-IsoP in humans and identified the major urinary metabolite as 2,3-dinor-5,6dihydro-15-F2t-IsoP (Roberts et al 1996). Metabolites of prostaglandins have been found to be biologically inactive, with the exception of the initial metabolite of PGD2, 9a,11b-PGF2 (Liston and Roberts 1985). Interestingly, we found that 2,3-dinor-5,6dihydro-15-F2t-IsoP is a potent constrictor of porcine retinal and brain microvessels, comparable to that observed with 15-F2t-IsoP (Hou et al 2001). In contrast, 2,3-dinor-5,6-dihydro-15-F2t-IsoP and 2,3-dinor-15-F2t-IsoP were found to be devoid of vascular actions in rat thoracic aorta (Cracowski et al 2002). These authors also found that 15-keto-15-F2t-IsoP did cause constriction of rat thoracic aorta, as did 15-F2t-IsoP. However, results from the
study of the metabolism of 15-F2t-IsoP in humans indicated that C15 dehydrogenation is not a quantitatively important pathway of the metabolic disposition of 15-F2t-IsoP in humans (Roberts et al 1996). 5-F2t-IsoP and its 5-epimer have also been found to possess no vasomotor effects in the rat thoracic aorta, human internal mammary artery and saphenous vein (Marliere et al 2002). 15-F2t-IsoP and 15-E2t-IsoP have both been found to have potent contractile activity in airway smooth muscle of a number of animal species, including human (Kang et al 1993; Kawikova et al 1996; Janssen et al 2000). In addition, 15-F2t-IsoP induces endothelin release and proliferation of vascular smooth muscle cells (Fukunaga et al 1993a, 1995; Yura et al 1999; Lahaie et al 1998). 15-F2t-IsoP has also been found to stimulate endothelial cells to bind monocytes but not neutrophils. Interestingly, this involves activation of PKA and MAP kinases in the absence of enhanced expression of VCAM-1 (Leitinger et al 2001). Moreover, epoxy-E2-IsoPs and epoxy-A2-IsoPs esterified in phosphatidylcholine activate endothelial cells to produce MCP-1 and IL-8 (Subbanagounder et al 2002). 15-E2t-IsoP has also been shown to activate intestinal epithelial cells (Elmhurst et al 1997) and both 15-E2t-IsoP and 15-F2t-IsoP exert contractile activity on gastrointestinal smooth muscle (Sametz et al 2000; Unmack et al 2001). These IsoPs also produce nociception and sensitize neurons (Evans et al 2000). 15-E2t-IsoP has also been shown to induce alkaline phosphatase activity and differentiation of calcifying vascular cells (Parhami et al 1997).
PRODUCTS OF THE ISOPROSTANE PATHWAY In porcine retinal and periventricular brain microvessels, 15F2t-IsoP has no effect on vascular smooth muscle cells. Rather, it has been shown to exert its contractile effects by inducing the endothelial cell formation of thromboxane (Lahaie et al 1998; Hou et al 2000) and to a lesser extent by inducing endothelin release (Lahaie et al 1998). This effect is unique to retinal and brain microvessels, whereas in other vascular beds its contractile action does not involve the generation of thromboxane (Takahashi et al 1992). However, results from initial experiments suggested that the vascular effects of IsoPs in vasculature other than the retina and brain may result from an interaction with thromboxane receptors. This was based on the finding that the vasoconstriction could be abrogated by thromboxane receptor antagonists (Takahashi et al 1992). Abrogation of the vascular and other effects of 15-F2t-IsoP and 15-E2t-IsoP by a variety of thomboxane receptor antagonists has since been observed by many other investigators (Janssen et al 2000; Kawikova et al 1996; Jourdan et al 1997; Banerjee et al 1992; Elmhurst et al 1997; Fukunaga et al 1993b; John and Valentin 1997; Kang et al 1993; Kromer and Tippins 1996, 1998, 1999; Mohler et al 1996; Sametz et al 2000; Lahaie et al 1998; Oliveira et al 2000; Okazawa et al 1997; Michoud et al 1998; Zhang et al 1996). However, a number of lines of evidence obtained subsequently suggests that these IsoPs do not interact with thromboxane receptors. This notion initially emerged from our finding that high concentrations of 15F2t-IsoP and 15-E2t-IsoP only induced reversible platelet aggregation, whereas at low concentrations they acted primarily as an antagonist of thromboxane agonist-induced platelet aggregation (Morrow et al 1992c; Longmire et al 1994). It was then shown that 15-F2t-IsoP and 15-E2t-IsoP do not compete for binding of thromboxane receptor ligands to thromboxane receptors (Pratico et al 1996; Fukunaga et al 1993a, 1993b). IsoP radioligand binding studies also suggested the presence of a distinct IsoP receptor (Fukunaga et al 1993a, 1995, 1997; Yura et al 1999). Moreover, functional differences between thromboxane receptor agonists and 15-F2t-IsoP have been observed (Hou et al 2000; Lahaie et al 1998). Although many of the biological responses to 15-E2t-IsoP are antagonized by thromboxane receptor antagonists, it appears to also exert effects via activation of PGE2 (EP) receptors in some tissues (Elmhurst et al 1997; Unmack et al 2001). Moreover, in a number of tissues in which 15-E2t-IsoP exerts biological activity, 15-F2t-IsoP is inactive (Elmhurst et al 1997; Evans et al 2000; Parhami et al 1997; Janssen et al 2000). So it is currently unclear whether the effects of these IsoPs result from interaction with thromboxane receptors, other known receptor(s), or a novel ‘‘IsoP receptor(s)’’. More recently, 15-F2c-IsoP (12-isoPGF2a) has also become available for biological testing and has been found to activate the PGF2a (FP) receptor, but relatively high concentrations were required and it has also been shown to induce hypertrophy of cardiac smooth muscle cells (Kunapuli et al 1997, 1998). The forthcoming availability of additional compounds for biological testing will likely contribute in a valuable way to our understanding of receptor-mediated actions of IsoPs as effectors of oxidant injury. Receptor-independent Biological Effects of Reactive Compounds Generated by the IsoP Pathway We have reported the discovery of two new groups of novel compounds that are capable of exerting biological effects due to their inherent chemical reactivity. One of these is the group cyclopentenone IsoPs (Chen et al 1999). These compounds are formed by dehydration of E2-isoPs and D2-IsoPs, analogous to the formation of PGA2 and PGJ2 by dehydration of cyclooxygenase-derived PGE2 and PGD2, respectively. Accordingly, these cyclopentenone IsoPs are termed A2/J2-IsoPs. As with the
213
other IsoP series of compounds, four regioisomers of both A-ring and J-ring IsoPs are formed. The unique feature of these compounds is that they are a,b-unsaturated carbonyls, which confer reactivity, in particular rendering them highly susceptible to Michael addition reactions (Boyland and Chasseaud 1968; Atsmon et al 1990; Honn and Marnett 1985). Cyclopentenone prostaglandins derived from the cyclooxygenase have been the subject of considerable interest because of the unique biological actions they exert, which has been attributed to the reactive a,bunsaturated carbonyl moiety (Honn and Marnett 1985). Specifically, they have been shown to inhibit cellular proliferation and induce differentiation, an effect that may be related to their ability to modulate a variety of growth-related and stress-induced genes (Fukushima 1990, 1992). These cytostatic effects can be reversible, but higher concentrations are cytotoxic and induce apoptosis (Fukushima 1990; Kim et al 1993; Fukushima et al 1989). Although the biological effects exerted by PGA2 and PGJ2 have been extensively studied for many years, the extent to which these compounds are formed in vivo has been the subject of continuing controversy for over two decades (Middledtich 1975; Jonsson et al 1976; Attallah et al 1974). Recently D12-PGJ2 was identified in human urine, but whether this arose from dehydration of PGD2 in the genitourinary tract prior to voiding or from systemic sources is not clear (Hirata et al 1988). Recently, we demonstrated that in normal rat liver A2/J2-IsoPs could be detected esterified in phospholipids at a level of 5.1+2.3 ng/g liver (Chen et al 1999). In the same livers, levels of E2/D2-IsoPs were 28.0+4.3 ng/g liver. Following administration of CCl4 to induce an oxidant injury in the liver, levels of A2/J2-IsoPs and E2/ J2-IsoPs increased strikingly and to a similar extent, 23.9-fold and 21.2-fold, respectively. One of the A2-IsoPs, 15-A2t-IsoP, was found to readily undergo Michael addition with glutathione in the presence of glutathione-S-tranferase; approximately 70% had conjugated within 2 min and the conjugation was complete by 8 min. In addition, using albumin as a model, it was demonstrated that 15-A2t-IsoP forms covalent adducts with proteins. F2- and E2/D2-IsoPs reach very high levels in the circulation following administration of CCl4 (Morrow et al 1990, 1992b, 1994b). However, A2/J2-IsoPs could not be detected in free form in the circulation, even after administration of CCl4. This can likely be explained by the finding that almost all the radioactivity excreted into urine following administration of radiolabelled 15-A2-IsoP to a human volunteer had undergone conjugation to form polar conjugates, presumably derived from glutathione (Chen et al 1999). These data are consistent with our previous findings, indicating that formation of polar conjugates is a major pathway of metabolic disposition of D12-PGJ2 in the rat (Atsmon et al 1990). This may explain the difficulty of others to demonstrate conclusively the formation of PGA2 or PGJ2 in vivo. However, in the case of A2/J2-IsoPs, we were able to detect them in their esterified form in the liver because when they are esterified in membrane phospholipids they are unable to adduct to glutathione and other thiols in the cytosol. In summary, these studies have elucidated a new class of reactive compounds formed as products of the IsoP pathway that exert biological effects relevant to the pathogenesis of oxidant injury. We recently reported the discovery of another class of compounds formed as products of the IsoP pathway, the acyclic g-ketoaldehydes, which are even more reactive than cyclopentenone IsoPs (Brame et al 1999). In 1985, Salomon and colleagues described the formation of g-ketoaldehydes from rearrangement of PGH2 derived from the cylooxygenase in vitro (Salomon et al 1984). Because of the structural similarities to levulinaldehyde, these compounds were termed levuglandins (LG) E2 and D2. Thus, we explored the hypothesis that levuglandin-like compounds may also be formed by rearrangement of H2-IsoP endoperoxides (Figure 17.2). To distinguish g-ketoaldehydes
214
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 17.2 Formation of IsoKs as rearrangement products of IsoP endoperoxides
formed via the IsoP pathway from levuglandins, we termed these compounds ‘‘isoketals’’ (IsoKs). Because, as shown in Figure 17.1, there are four H2-IsoP endoperoxide regioisomers, four regioisomers of both E2- and D2-IsoKs are also formed. Oxidation of arachidonic acid in vitro yielded a series of compounds that were confirmed to be IsoKs, using a number of mass spectrometric approaches including electron impact mass spectral analysis. Interestingly, the amounts of IsoLGs formed in vitro were found to be intermediate between the amounts of F2IsoPs and E2/D2-IsoPs formed, indicating that the amount of these compounds formed was not trivial. Nonetheless, we could not detect their formation in biological systems in vitro, i.e. oxidation of liver microsomes and LDL, neither could we detect them in plasma, urine or in the liver following administration of CCl4 to rats. We hypothesized that this could be due to very rapid adduction to proteins. It should be mentioned that other reactive aldehydes that are generated as products of lipid peroxidation, e.g. 4-hydroxynonenal and malondialdehyde, can be detected in their free non-adducted form in biological fluids and tissues (Esterbauer et al 1991). To obtain support for this hypothesis, we assessed the rate of adduction of a synthetic E2-IsoK to proteins using albumin and compared this with the rate of adduction of 4hydroxynonenal. The results obtained were most informative. Rate of adduction was determined by assessing the percentage decline in free levels of compounds measured in aliquots removed during incubations with albumin over time. As shown in Figure 17.3, E2-IsoK underwent adduction with extreme rapidity; 60% had adducted within the first 20 s of incubation. In striking contrast, approximately 50% of 4-hydroxynonenal still remained unadducted after 1 h. Of note is that the free level of E2IsoK does not decline completely to zero but plateaus near zero. This is likely due to the presence of some E2-IsoK, in which the double bond on the lower side chain has migrated from the D10 to the D9 position, rendering the molecule less reactive (Iyer et al 1990). These data indicate that IsoKs adduct to proteins at a rate that is at least an order of magnitude faster than 4-hydroxynonenal, which is considered one of the most reactive aldehydes formed as a product of lipid peroxidation. Moreover, IsoKs exhibit a remarkable proclivity to cross-link proteins (Iyer et al 1989). Therefore, we turned to developing methods to detect the formation of IsoKs as protein adducts using liquid chromatography (LC) electrospray ionization (ESI) tandem mass spectrometry (MS–MS). Salomon and colleagues had obtained evidence that LGE2 forms a pyrrole adduct with lysine residues on proteins (Iyer et al 1990). However, LC–ESI–MS analysis
Figure 17.3 Comparative rates of adduction of E2-IsoK and 4hydroxynonenal during incubation with albumin. Formation of covalent adducts was assessed by the decline in levels of free compounds measured in aliquots removed at various times indicated
Figure 17.4 Time course of formation of lactam and Schiff base adducts during incubation of E2-IsoK with lysine
following incubation of E2-IsoK with lysine did not yield evidence for the presence of compounds with the predicted MH+ ion for a lysyl E2-IsoK pyrrole adduct. However, full-spectrum scanning analysis revealed major MH+ ions present 16 and 32 Da higher than the MH+ ion for the E2-IsoK lysyl pyrrole. These compounds were consistent with lactam and hydroxylactam adducts formed by facile autoxidation of highly substituted pyrroles (Smith and Jensen 1967). This was confirmed by tandem mass spectrometric analyses of adducts formed with various lysine analogues and [13C6] lysine. The analyses of these adducts were performed following prolonged incubation of E2-IsoK with lysine. We then analysed earlier time points and a new adduct species appeared. This had the appropriate MH+ ion for a Schiff base adduct. Evidence confirming that this was a lysyl LGE2 Schiff base adduct was obtained by tandem mass spectrometric analysis and finding that expected products were formed following treatment with NaCN, methoxyamine.HCL, and NaBH4 (Smith and Jensen 1967). The time course of formation of these various adducts was then assessed (Figure 17.4). As noted, the Schiff base adduct is formed very rapidly but is unstable and disappears over time. In contrast, the lactam adducts accumulate much more slowly. The time course of formation of these adducts is consistent with the proposed mechanism of adduct formation depicted in Figure 17.5.
PRODUCTS OF THE ISOPROSTANE PATHWAY
215
Figure 17.6 Formation of eight regioisomers of F4-neuroprostanes from oxidation of DHA. Abstraction of specific allylic hydrogens results in the formation of the individual regioisomers, designated by carbon atom numbers (C)
Isoprostane-like Compounds from Other Fatty Acids Figure 17.5 Mechanism of the formation of E2-IsoK lysyl Schiff base and lactam adducts
IsoK and lysine are initially converted via an intermediate to a reversible Schiff base adduct. This intermediate also proceeds through an irreversible pathway, leading to the formation of a pyrrole adduct, which then undergoes autoxidation to form lactam and hydroxylactam adducts. With this information in hand, we turned to explore whether we could detect the formation of IsoK adducts on Apo-B protein during oxidation of LDL in vitro. In these experiments, the Apo-B protein was enzymatically digested to individual amino acids and then analysed for lysyl E2-IsoK lactam adducts. This was considered a key experiment because previously we could not detect the formation of IsoKs in free form during oxidation of LDL. IsoK lactam adducts could not be detected in native LDL but intense signals were detected for lysyl lactam and hydroxylactam IsoLG adducts on ApoB protein following oxidation of LDL (Brame et al 1999). The key question is whether IsoK adducts are formed in vivo. Data obtained from recent experiments indicate that IsoK adducts: (a) can be detected in normal rat liver; (b) levels in rat liver increase significantly following induction of an oxidant injury by administration of CCl4; and (c) can be detected in normal human plasma (manuscript submitted). Insights into the potential toxicity of IsoKs emerged from recent experiments in which we explored their effects on proteasome function (Davies et al 2002). We found that IsoK-adducted proteins are degraded very poorly by the proteasome and that IsoK-adducted proteins also potently inhibit chymotrypsin-like activity of the 20S proteasome. Moreover, incubation of IsoK with neuroglial cells dose-dependently inhibited proteasome activity (IC50=330 nM) and induced cell death (LD50=670 nM).
The basic requirement for cyclization to occur to form a cyclopentane ring that characterizes IsoPs during oxidation of unsaturated fatty acids is the presence of at least three double bonds. Thus, oxidation of linoleic acid (C18:2) does not generate IsoP-like compounds, whereas oxidation of linolenic acid (C18:3) would generate F1-IsoPs. However, the relevance of the formation of F1-IsoPs is dubious because linolenic acid is normally present in only minor quantities in humans and animals. In plants, however, arachidonic acid is generally absent and the most abundant fatty acid is a-linolenic acid. Interestingly, F-ring and E/D-ring IsoPlike compounds derived from oxidation of a-linolenic acid are present in abundant quantities in plants, which have been termed phytoprostanes (Parchmann and Mueller 1998; Imbusch and Mueller 2000). F3-IsoPs have recently been described as products of free radical-induced peroxidation of eicosapentaenoic acid (C20:5) (Nourooz-Zadeh et al 1997). Again, however, this is not an abundant fatty acid under normal circumstances. The formation of F3-IsoPs may be of interest, however, in situations where the ingestion of eicosapentaenoic acid is high, such as high dietary intake of fish and dietary supplementation with fish oil. Docosahexaenoic acid (DHA) (C22:6, o3) has been a focus of interest because it is present in abundant quantities in the brain, particularly in grey matter, where it comprises up to 25–35% of total fatty acids in aminophospholipids (Skinner et al 1993; Salem et al 1986). We recently described the formation of F-ring IsoPlike compounds with four double bonds during free radicalinduced peroxidation of DHA both in vitro and in vivo (Roberts et al 1998). Because DHA is uniquely highly enriched in neurons, we have termed these compounds F4-neuroprostanes (F4-NPs). Oxidation of DHA leads to the formation of eight F4-NP regioisomers, each of which is theoretically comprised of eight racemic diastereomers for a total of 128 compounds (Figure 17.6). Subsequently, we also showed that E- and D-ring NPs are also produced in vivo (Reich et al 2000) and more recently IsoK-like compounds (neuroketals) (Bernoud-Hubac et al 2001). We considered the possibility that measurement of F4-NPs may
216
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Figure 17.7 The spectrum of compounds formed by the isoprostane pathway
represent a novel and sensitive marker of oxidative neuronal injury. Of considerable interest is that we have found that levels of F4-NPs are elevated in cerebrospinal fluid from patients with Alzheimer’s disease, providing considerable support for a role of free radicals in neuronal injury in this disease (Roberts et al 1998). Interestingly, we found an excellent correlation between levels of F2-IsoPs and F4-NPs in cerebrospinal fluid from patients with Alzheimer’s disease and age-matched normal controls over the range of concentrations present (r=0.88, p=0.0003; Montine et al 1998). We, and others, have found that NPs esterified in brain lipids are increased in Alzheimer’s disease but, interestingly, F2IsoPs are not increased (Reich et al 2001; Nourooz-Zadeh et al 1999). The reason for these differences between the relative increases in tissue and the cerebrospinal concentrations of F2IsoPs and F4-NPs remains unclear. In summary, the initial discovery of F2-IsoPs was of biochemical interest but the importance of this discovery has become increasingly apparent over the last 12 years. First, quantification of F2-IsoPs has proved to be a major advance in our ability to assess oxidative stress status reliably in vivo. Our understanding of the biochemistry of the IsoP pathway has also expanded greatly. This is of interest in that almost the entire spectrum of compounds produced by the cyclooxygenase pathway have now been shown to be produced non-enzymatically via the IsoP pathway (Figure 17.7). In addition, IsoP-like compounds can also be generated from other unsaturated fatty acids, and of particular interest in this regard are NPs formed from oxidation of DHA. IsoPs and related compounds produced by the IsoP pathway are also of potential relevance in that they can exert potent biological actions. This involves apparent receptor-mediated actions as well as, in the
case of cyclopentenone IsoPs, IsoKs and NKs, actions attributed to inherent chemical reactivity, which can lead to covalent modification of critical biomolecules. Thus, the current understanding of the IsoP pathway has diverged into a variety of areas of potential biochemical and biological importance that are likely to continue to expand as new avenues for scientific inquiry regarding these unique molecules emerge.
ACKNOWLEDGEMENTS This work was supported by Grants GM42056, GM15431, DK48831, CA77839, DK26657 and CA68485 from the National Institutes of Health, USA.
REFERENCES Atsmon J, Sweetman BJ, Baertschi SW et al (1990) Biochemistry, 29, 3760– 3765. Attallah A, Payakkapan W, Lee J et al (1974) Prostaglandins, 5, 69–72. Banerjee M, Kang KH, Morrow JD et al (1992) Am J Physiol, 263, H660– 663. Beauchamp, MH, Martinez-Bermudez, AK, Gobeil, F Jr et al (2001) J Appl Physiol, 90, 2279–2288. Bernoud-Hubac N, Davies SS, Boutaud O et al (2001) J Biol Chem, 276, 30964–30970. Boyland E and Chasseaud LF (1968) Biochem J, 109, 651–661. Brame CJ, Salomon RG, Morrow JD and Roberts LJ II (1999) J Biol Chem, 274, 13139–13146.
PRODUCTS OF THE ISOPROSTANE PATHWAY Chen Y, Morrow JD and Roberts LJ II (1999) J Biol Chem, 274, 10863– 10868. Cracowski JL, Camus L, Durand T et al (2002) J Cardiovasc Pharmacol, 39, 396–403. Davies SS, Amarnath V, Montine KS et al (2002) FASEB J, 16, 715–717. Elmhurst JL, Betti PA and Rangachari PK (1997) J Pharmacol Exp Ther, 282, 1198–1205. Esterbauer H, Schaur RJ and Zollner H (1991) Free Radic Biol Med, 11, 81–128. Evans AR, Junger H, Southall MD et al (2000) J Pharmacol Exp Ther, 293, 912–920. Fukunaga M, Makita N, Roberts LJ II et al (1993a) Am J Physiol, 264, C1619–1624. Fukunaga M, Takahashi K and Badr KF (1993b) Biochem Biophys Res Commun, 195, 507–515. Fukunaga M, Yura T and Badr KF (1995) J Cardiovasc Pharmacol, 26, S51–52. Fukunaga M, Yura T, Grygorczyk R and Badr KF (1997) Am J Physiol, 272, F477–483. Fukushima M (1990) Eicosanoids, 3, 189–199. Fukushima M (1992) Prostagland Leukotrienes Essent Fatty Acids, 47, 1–12. Fukushima M, Kato T, Narumiya S et al (1989) Adv Prostagland Thrombox Leukotriene Res, 19, 415–418. Gardner HW (1989) Free Radic Biol Med, 7, 65–86. Halliwell B and Grootveld M (1987) FEBS Lett, 213, 9–14. Halliwell B and Gutteridge JM (1990) Methods Enzymol, 186, 1–85. Hirata Y, Hayashi H, Ito S et al (1988) J Biol Chem, 263, 16619–16625. Hoffman SW, Moore S and Ellis EF (1997) Stroke, 28, 844–849. Honn KV and Marnett LJ (1985) Biochem Biophys Res Commun, 129, 34– 40. Hou X, Gobeil F Jr, Peri K et al (2000) Stroke, 31, 516–524; discussion 525. Hou X, Roberts LJ II, Taber DF et al (2001) Am J Physiol Regul Integr Comp Physiol, 281, R391–400. Imbusch R and Mueller MJ (2000) Free Radic Biol Med, 28, 720–726. Iyer RS, Ghosh S and Salomon RG (1989) Prostaglandins, 37, 471–480. Iyer RS, Miller DB and Salomon RG (1990) J Org Chem, 55, 3175–3180. Janssen LJ, Premji M, Netherton S et al (2000) J Pharmacol Exp Ther, 295, 506–511. Janssen LJ, Premji M, Netherton S et al (2001) Br J Pharmacol, 132, 127– 134. Janssen LJ and Tazzeo T (2002) J Pharmacol Exp Ther, 301, 1060–1066. John GW and Valentin JP (1997) Br J Pharmacol, 122, 899–905. Jonsson HT Jr, Middleditch BS, Schexnayder MA and Desiderio DM (1976) J Lipid Res, 17, 1–6. Jourdan KB, Evans TW, Curzen NP and Mitchell JA (1997) Br J Pharmacol, 120, 1280–1285. Kang KH, Morrow JD, Roberts LJ II et al (1993) J Appl Physiol, 74, 460– 465. Kawikova I, Barnes PJ, Takahashi T et al (1996) Am J Respir Crit Care Med, 153, 590–596. Kim IK, Lee JH, Sohn HW et al (1993) FEBS Lett, 321, 209–214. Kromer BM and Tippins JR (1996) Br J Pharmacol, 119, 1276–1280. Kromer BM and Tippins JR (1998) J Cardiovasc Pharmacol, 32, 471–478. Kromer BM and Tippins JR (1999) Br J Pharmacol, 126, 1171–1174. Kunapuli P, Lawson JA, Rokach J and FitzGerald GA (1997) J Biol Chem, 272, 27147–27154. Kunapuli P, Lawson JA, Rokach JA et al (1998) J Biol Chem, 273, 22442– 22452. Lahaie I, Hardy P, Hou X et al (1998) Am J Physiol, 274, R1406–R1416. Leitinger N, Huber J, Rizza C et al (2001) FASEB J, 15, 1254–1256. Liston TE and Roberts LJ II (1985) Proc Natl Acad Sci USA, 82, 6030– 6034. Longmire AW, Roberts LJ and Morrow JD (1994) Prostaglandins, 48, 247–256. Marley R, Harry D, Anand R et al (1997) Gastroenterology, 112, 208–213. Marliere S, Cracowski JL, Durand T et al (2002) Br J Pharmacol, 135, 1276–1280. Michoud E, Lecomte M, Lagarde M and Wiernsperger N (1998) Prostagland Leukotrienes Essent Fatty Acids, 59, 349–355. Middledtich B (1975) Prostaglandins, 9, 409–411. Mobert J, Becker BF, Zahler S and Gerlach E (1997) J Cardiovasc Pharmacol, 29, 789–794.
217
Mohler ER, Franklin MT and Adam LP (1996) Biochem Biophys Res Commun, 225, 915–923. Montine TJ, Markesbery WR, Morrow JD and Roberts LJ II (1998) Ann Neurol, 44, 410–413. Moore K and Roberts LJ II (1998) Free Radic Res, 28, 659–671. Morrow JD, Awad JA, Boss HJ et al (1992a) Proc Natl Acad Sci USA, 89, 10721–10725. Morrow JD, Awad JA, Kato T et al (1992b) J Clin Invest, 90, 2502–2507. Morrow JD, Awad JA, Wu A et al (1996) J Biol Chem, 271, 23185–23190. Morrow JD, Hill KE, Burk RF et al (1990) Proc Natl Acad Sci USA, 87, 9383–9387. Morrow JD, Minton TA, Badr KF and Roberts LJ II (1994a) Biochim Biophys Acta, 1210, 244–248. Morrow JD, Minton TA, Mukundan CR et al (1994b) J Biol Chem, 269, 4317–4326. Morrow JD, Minton TA and Roberts LJ II (1992c) Prostaglandins, 44, 155–163. Morrow JD and Roberts LJ (1997) Prog Lipid Res, 36, 1–21. Morrow JD and Roberts LJ II (1996) Biochem Pharmacol, 51, 1–9. Morrow JD, Scruggs J, Chen Y et al (1998) J Lipid Res, 39, 1589–1593. Nourooz-Zadeh J, Halliwell B and Anggard EE (1997) Biochem Biophys Res Commun, 236, 467–472. Nourooz-Zadeh J, Liu EH, Yhlen B et al (1999) J Neurochem, 72, 734– 740. Nugteren DH and Hazelhof E (1973) Biochim Biophys Acta, 326, 448–461. O’Connor D, Mihelich ED and Coleman MC (1984) J Am Chem Soc, 106, 3577–3584. Okazawa A, Kawikova I, Cui ZH et al (1997) Am J Respir Crit Care Med, 155, 436–441. Oliveira L, Stallwood NA and Crankshaw DJ (2000) Br J Pharmacol, 129, 509–514. Parchmann S and Mueller MJ (1998) J Biol Chem, 273, 32650–32655. Parhami F, Morrow AD, Balucan J et al (1997) Arterioscler Thromb Vasc Biol, 17, 680–687. Porter NA, Caldwell SE and Mills KA (1995) Lipids, 30, 277–290. Pratico D, Smyth EM, Violi F and FitzGerald GA (1996) J Biol Chem, 271, 14916–14924. Reich EE, Markesbery WR, Roberts LJ II et al (2001) Am J Pathol, 158, 293–297. Reich EE, Zackert WE, Brame CJ et al (2000) Biochemistry, 39, 2376– 2383. Roberts LJ II, Montine TJ, Markesbery WR et al (1998) J Biol Chem, 273, 13605–13612. Roberts LJ II, Moore KP, Zackert WE et al (1996) J Biol Chem, 271, 20617–20620. Roberts LJ II and Morrow JD (1997) Biochim Biophys Acta, 1345, 121– 135. Roberts LJ and Morrow JD (2000) Free Radic Biol Med, 28, 505–513. Salem N Jr, Kim HY and Yergery JA (1986) In Simopoulos AP, Kifer RR and Martin RE (eds), Health Effects of Polyunsaturated Fatty Acids in Sea Foods. Orlando, FL: Academic Press, pp. 263–317. Salomon RG, Miller DB, Zagorski MG and Coughlin DJ (1984) J Am Chem Soc, 106, 6049–6060. Sametz W, Hennerbichler S, Glaser S et al (2000) Br J Pharmacol, 130, 1903–1910. Sinzinger H, Oguogho A and Kaliman J (1997) Lymphology, 30, 155–159. Skinner ER, Watt C, Besson JA and Best PV (1993) Brain, 116, 717–725. Smith EB and Jensen HB (1967) J Org Chem, 32, 3330–3333. Southorn PA and Powis G (1988) Mayo Clin Proc, 63, 390–408. Subbanagounder G, Wong JW, Lee H et al (2002) J Biol Chem, 277, 7271– 7281. Taber DF, Morrow JD and Roberts LJ II (1997) Prostaglandins, 53, 63– 67. Takahashi K, Nammour TM, Fukunaga M et al (1992) J Clin Invest, 90, 136–141. Unmack MA, Rangachari PK and Skadhauge E (2001) J Pharmacol Exp Ther, 296, 434–441. Wagner RS, Weare C, Jin N et al (1997) Prostaglandins, 54, 581–599. Waugh RJ, Morrow JD, Roberts LJ II and Murphy RC (1997) Free Radic Biol Med, 23, 943–954. Yura T, Fukunaga M, Khan R et al (1999) Kidney Int, 56, 471–478. Zhang R, Ogletree ML and Moreland S (1996) Eur J Pharmacol, 317, 91– 96.
18 Insight into Prostanoid Functions: Lessons from Receptor-knockout Mice Yukihiko Sugimoto, Shuh Narumiya and Atsushi Ichikawa Kyoto University, Kyoto, Japan
PROSTANOID RECEPTORS Prostaglandins (PGs) and thromboxanes (TXs) are the eicosanoids synthesized via the cyclooxygenase (COX) pathway. The collective term for this family of eicosanoids is ‘‘the prostanoids’’. Prostanoids are synthesized in a variety of cells in response to various physiological and pathological stimuli, and are then quickly released from the cells and act as local hormones in the vicinity of their production site to maintain local homeostasis (Halushka et al 1989). Prostanoids exert a wide variety of actions in the body, which are mediated by specific receptors on the plasma membranes of target cells. Prostanoid receptors were initially characterized pharmacologically in several bioassay systems, including contraction–relaxation assays on various smooth muscles and the aggregation of platelets. These receptors are classified into five basic types, termed DP, EP, FP, IP and TP, on the basis of their sensitivity to the five primary prostanoids, PGD2, PGE2, PGF2a, PGI2 and TXA2, respectively. Furthermore, EP is subdivided into four subtypes, EP1, EP2, EP3 and EP4, on the basis of their responses to various agonists and antagonists (Kennedy et al 1982; Coleman et al 1990). The prostanoid receptors have also been characterized biochemically using radioactive specific ligands (Coleman et al 1994b). Biochemical studies showed that the actions of prostanoids are mediated by G proteins, and the ligand-binding properties indicated that a variety of prostanoids cross-react with each receptor, suggesting the structural similarity of the receptors. It has been reported repeatedly that the actions of prostanoids are associated with changes in second messenger levels. Some prostanoid actions had been noticed to be associated with changes in cyclic AMP (cAMP) levels, phosphatidylinositol turnover or free calcium ion concentrations in the cell. However, none of the receptors had been isolated and cloned until the TXA2 receptor, TP, was purified from human blood platelets in 1989 (Ushikubi et al 1989) and its cDNA cloned in 1991 (Hirata et al 1991). These studies revealed that the TP was a G-proteincoupled, rhodopsin-type receptor with seven transmembrane domains. Homology screening of mouse cDNA libraries subsequently identified the structures of all of the eight types and subtypes of the prostanoid receptors. These receptors have been expressed and their ligand-binding properties and signal transduction mechanisms have been examined in homogenous receptor populations in heterologous expression systems. In addition, the tissue and cell distributions of the receptors have been studied by Northern blot and in situ hybridization analyses of their mRNA The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
expression. The correlation of this knowledge with the findings that have accumulated from pharmacological studies, using COX inhibitors and various prostanoid analogues with agonistic and antagonistic activities, have helped to define the actions of each type of receptor (Coleman et al 1990) as well as helped to reveal novel actions of these receptors. The accumulated knowledge from these analyses on the structure, pharmacological and biochemical properties and cellular distribution of the prostanoid receptor molecules has been described elsewhere (Narumiya et al 1999; Sugimoto et al 2000) and some is summarized in Table 18.1. In recent years, the method of inactivating the function of a gene specifically and completely in mice has become a routine procedure. Gene disruption by the creation of a targeting vector and its introduction into embryonic stem cells was established in the 1980s (Doetschman et al 1987; Capecchi 1989). To date, targeted gene disruptions have been reported, not only in the enzymes engaged in prostanoid synthesis but also in the prostanoid receptors. Mice deficient in each prostanoid receptor have been generated, and initial analyses of the EP1- (Ushikubi et al 1998), EP2- (Kennedy et al 1999; Tilley et al 1999; Hizaki et al 1999), EP3- (Ushikubi et al 1998), EP4- (Nguyen et al 1997; Segi et al 1998), DP- (Matsuoka et al 2000), FP- (Sugimoto et al 1997), IP- (Murata et al 1997) and TP- (Thomas et al 1998) deficient mice have been reported. Such progress in strategy has enabled us to confirm the existing knowledge from pharmacological and biochemical analyses, to uncover novel prostanoid functions, and to answer questions that otherwise could not have been addressed. This section summarizes the phenotypes observed in prostanoid receptor-deficient mice compared with those observed in mice with altered prostanoid synthesis (Table 18.2), and presents various important insights into the mechanisms of the physiological actions of the prostanoids via their receptors. MICE DEFICIENT IN EACH EP SUBTYPE (EP1, EP2, EP3 AND EP4) EP4 was the most recent of the subtypes to be pharmacologically identified, having been identified in 1994 (Coleman et al 1994a), but this receptor subtype is thought to be responsible for many of the actions of the PGE2, such as the dilatation of smooth muscles, inhibition of immune responses and regulation of mucus secretion (Coleman et al 1990). Since both EP2 and EP4 are coupled to the stimulation of adenylate cyclase, the two receptors have been
220
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Table 18.1 Properties of the mouse prostanoid receptors1 Receptor type
Kd, nM (radioligand)
Rank order of binding affinity2
Signalling
Gene locus (mouse/human)
Alternatively spliced isoforms
EP1
21 ([3H]PGE2) 27 ([3H]PGE2) 3 ([3H]PGE2)
PGE24iloprost4PGE1
[Ca2+]:
8/19p13.1
2 (Rat)3
PGE2=PGE14butaprost
cAMP:
14/14q22
None
PGE2=PGE14iloprost
cAMP; [Ca2+]:
3/1p31.2
11 ([3H]PGE2) 40 ([3H]PGD2) 1.3 ([3H]PGF2a) 4.5 ([3H]iloprost) 3.3 ([3H]S-145)
PGE2=PGE1
cAMP:
15/5p13.1
3 (Mouse)3 7 (Human) 4 (Bovine)3 None
PGD24BW245C
cAMP:
14/14q21.3
None
EP2 EP3 EP4 DP FP IP TP
PGF2a4PGD2
[Ca ]:
3/1p31.1
2 (Ovine)
cicaprost4iloprost4PGE1
cAMP: [Ca2+]: [Ca2+]: cAMP;
7/19q13.3
None
10/19p13.3
2 (Human)3
S-1454STA24U46619
2+
1
References for cDNA cloning, chromosomal mapping and multiple receptor isoforms are summarized in previous reviews (Narumiya et al 1999; Sugimoto et al 2000). Basic prostanoids and their derivatives with low K1 values (51076 M) are indicated. Cicaprost and iloprost are stable IP agonists, STA2 and U46619 are stable TP agonists, and S-145 is a stable TP antagonist. Butaprost and BW245C are selective EP2 and DP agonists, respectively. For details of the binding characters, see Kiriyama et al (1997). 3 Some alternatively spliced receptor isoforms have been found to differ in their signal transduction pathways. 2
suspected to function in a similar manner. However, a drastic induction of EP2 gene expression in response to hormonal and proinflammatory stimuli, but not that of EP4, has been identified in various kinds of cells, suggesting that the two receptors have rather different roles in various physiological processes (Sugimoto et al 2000). It is interesting in this respect to study the phenotypes that appear in the EP2- and EP4-knockout mice. In contrast, EP1 and EP3 have been shown to be coupled to an increase in intracellular CA2+ mobilization. It should be noted that EP3 is the only prostanoid receptor that inhibits adenylate cyclase. To date, EP3 has been shown to be involved in pyrogen-induced fever generation and in mucosal defence of the gastrointestinal tract. However, the roles of EP1 in the body remain to be clarified. Ductus Arteriosus In EP4-deficient mice from an inbred 129 strain, the ductus arteriosus (DA) fails to close after birth, and this is followed by death in the early neonatal period. The DA is an arterial connection in the foetus that directs the blood to be oxygenated away from the pulmonary circulation and toward the placenta.
Thus, in wild-type animals, the drop in PGE2 that acts as a trigger for DA closure in the neonate is sensed through EP4. It is worth noting that when the gene disruption occurs on a mixed genetic background a small percentage of mice survive, suggesting that alleles at other loci can provide an alternative mechanism for DA closure (Nguyen et al 1997; Segi et al 1998). Loftin et al (2001) examined this issue by using mice deficient in either or both COX isoforms. The absence of only COX-1 did not affect closure of the DA. However, 35% of COX-2-deficient mice die with a patent DA. The mortality and patent DA incidence due to the absence of COX-2 is significantly increased (79%) when one copy of the COX-1 gene is also inactivated. Furthermore, 100% of the mice deficient in both isoforms die with a patent DA. These results indicate the dominant contribution of COX-2 to DA closure, but this effect can be partly compensated by the COX-1 isoform. Ovulation and Fertilization Recent studies on mice deficient in COX-2 showed multiple failures in female reproduction, including ovulation and fertilization,
Table 18.2 Major phenotypes of mice deficient in the prostanoid receptors Disrupted gene
Phenotypes
DP EP1 EP2
Reduced allergic responses in ovalbumin-induced bronchial asthma Decreased aberrant crypt foci formation in response to azoxymethane Impaired ovulation and fertilization Decreased intestinal polyp formation in ApcD716 mice Salt-sensitive hypertension Impaired osteoclastogenesis in vitro Impaired febrile response to pyrogens Impaired duodenal bicarbonate secretion Increased bleeding tendency and decreased susceptibility to thromboembolism Patent ductus arteriosus Impaired mucosal integrity and enhanced immune response in colitis Decreased inflammatory bone resorption Loss of parturition Thrombotic tendency Decreased inflammatory swelling Decreased acetic acid writhing Bleeding tendency and resistance to thromboembolism
EP3 EP4 FP IP TP
Gene disruption showing similar phenotypes COX-2(7/7) COX-2(7/7) COX-2(7/7), cPLA2(7/7) COX-2(7/7) COX-2(7/7)/COX-1(7/7), COX-2(7/7) COX-2(7/7), COX-1(7/7) COX-1(7/7), cPLA2(7/7)
INSIGHT INTO PROSTANOID FUNCTIONS
221
suggesting that PGs play essential roles in multiple processes occurring during early pregnancy (Davis et al 1999; Dinchuck et al 1995; Lim et al 1997). Three groups independently generated mice lacking EP2, which showed a failure during early pregnancy. Kennedy et al (1999) and Tilley et al (1999) found that EP2deficient female mice consistently deliver fewer pups than their wild-type counterparts irrespective of the genotypes of the mating males. They detected slightly impaired ovulation and a dramatic reduction in fertilization in EP2-deficient mice and concluded that reproduction failures in Cox-2-deficient mice is at least partly due to the dysfunction of EP2. Hizaki et al (1999) further found that this phenotype is due to impaired expansion of the cumulus oophorus. Since EP2 and COX-2 are induced in the cumulus in response to gonadotropins, and since PGE2 can induce cumulus expansion by elevating cAMP (Eppig 1981), the authors suggest that the PGE2 and EP2 systems work as a positive-feedback loop to induce oophorus maturation required for fertilization during and after ovulation. Indeed, unovulated eggs remaining in the corpora lutea were observed at a higher frequency in EP2-deficient mice. It is interesting in this respect that indomethacin treatment has been reported to result in the formation of luteinized unruptured follicles in humans (Priddy et al 1990).
Sonoshita et al (2001) reported that the homozygous disruption of EP2 in Apc-knockout mice caused significant decreases in the number and size of the intestinal polyps, showing similar effects to those induced by the COX-2 gene disruption. Regarding the mechanism, the authors indicate that an increased cellular cAMP level through EP2 signalling amplifies COX-2 expression and stimulates the expression of vascular endothelial growth factor in the polyp stroma. In a separate paper, carcinogeninduced formation of aberrant crypt foci, putative preneoplastic lesions of the colon, was examined; foci formation was decreased in EP1-deficient mice to *60% of the level in wildtype mice (Watanabe et al 1999). Furthermore, partial reduction of foci formation was observed by the administration of an EP1-antagonist in the diet of azoxymethane-treated wildtype mice. Similar treatment also reduced the number of polyps in Min mice, suggesting that PGE2 contributes to carcinogeninduced colon foci formation through its action on EP1. Thus, there is an apparent discrepancy regarding the identity of the EP subtypes acting in carcinogenesis, which awaits further study for clarification.
Fever Generation
The current hypothesis regarding the medicinal usage of aspirinlike drugs is that the inhibition of COX-2 is responsible for their beneficial effects, whereas the inhibition of COX-1 is responsible for their adverse effects, the most common being gastric ulceration (Langenbach et al 1999). However, neither COX-1-deficient nor COX-2-deficient mice showed spontaneous ulcer formation, although gastric PG levels in COX-1 null mice were greatly reduced to levels observed following an ulcerative dose of indomethacin (Langenbach et al 1995; Moham et al 1995). Thus, elimination of COX-1-derived PGs alone was not sufficient to cause gastric ulcers. In contrast to the understanding of the contribution of the COX enzymes in the ulcerative process, which prostanoid receptor is involved in the protective actions against ulcerative stimuli is poorly understood. EP and other prostanoid receptor-knockout mice will be used to clarify this issue. Indeed, Takeuchi et al (1999) recently found that EP3 but not EP1 is involved in acid-induced duodenal bicarbonate secretion, which is physiologically important in the mucosal defence against acid injury. Prostanoids, especially the E-type PG, have been suggested to contribute to mucosal defence in gastrointestinal inflammation, such as in inflammatory bowel disease. Indeed, genetic absence of COX-1 or COX-2 exacerbated the extent of dextran sodium sulphate (DSS)-induced colitis (Morteau et al 2000). This treatment increased intestinal PGE2 production in a COX-2dependent manner. Among the EP-deficient mice, only EP4deficient mice showed a greatly increased susceptibility to a low dose (3%) of DSS that caused only mild colonic injury in wildtype mice (Kabashima et al 2002). The phenotype was mimicked in wild-type mice by administration of an EP4selective antagonist. The EP4 deficiency caused impaired mucosal defence and aggregation of neutrophils and lymphocytes in the colon. A high dose (7%) of DSS elicited severe colitis in wild-type mice, but an EP4-agonist reversed these effects of DSS. An EP4-antagonist suppressed recovery from colitis and induced significant proliferation of CD4+ T cells. In the colon isolated from EP4-deficient mice with DSS-induced colitis, gene expression of epidermal growth factor was reduced and the expression of chemoattractants increased, compared with wild-type mice treated with DSS. Thus, PGE2 contributes to maintain intestinal homeostasis via EP4 by promoting epithelial regeneration and also by inhibiting intestinal immune responses.
The E-type PG is a powerful inducer of fever when injected into the brain, and the level of PGE2 increases in the preoptic area (POA) during lipopolysaccharide (LPS)-induced fever. In addition, indomethacin completely abolishes both the LPSinduced fever and the increased levels of PGE2 in the POA (Kluger 1991; Saper and Breder 1994). The febrile responses to PGE2, interleukin (IL)-1b and LPS in mice lacking EP1, EP2 and EP4 were comparable to those in wild-types. The EP3deficient mice failed to show a febrile response to all of these stimuli (Ushikubi et al 1998). Thus, PGE2 mediates fever generation in response to both exogenous and endogenous pyrogens by acting on EP3. It has also been reported that COX-2-deficient mice also show impaired febrile responses, suggesting that COX-2 is involved in fever generation (Li et al 1999). Indeed, intravenous administration of IL-1b induces expression of both COX-2 and microsomal PGE synthase in endothelial cells of the brain microvessels (Ek et al 2001). The resultant PGE2 appears to act on EP3 in the POA, especially in the region surrounding the organum vasculosa lamina terminalis (OVLT), which is the most sensitive area of the brain for microinjected PGE2 to produce fever (Elmquist et al 1997). In fact, the mRNA for EP3 is particularly abundant in the regions surrounding the OVLT (Sugimoto et al 1994) and EP3 immunoreactivity is also present in the cell bodies of these neurons, with a distribution pattern similar to that of EP3 mRNA (Nakamura et al 1999). Thus, EP3 expressed in the neurons surrounding the OVLT appears to work as an initial input of ‘pyrogenic’ PGE2 to alter the set-point of thermal regulation. Colorectal Cancer COX-2 has been implicated in the progression of colorectal cancer. Supporting evidence comes from a study in which COX-2-deficient mice were crossed with mice with a truncated Apc gene (ApcD716), used as a model of human familial adenomatous polyposis (Oshima et al 1996). The ApcD716 heterozygous/COX-2-deficient mice have a dramatically reduced number and size of intestinal polyps. This provides direct genetic evidence for the role of COX-2 in tumorigenesis.
Gastrointestinal Functions
222
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Vascular Homeostasis
DP-KNOCKOUT MICE
PGE2 also elicits contractile and/or relaxant responses of vascular smooth muscles in vitro. Kennedy et al (1999) administered PGE2 and its analogues intravenously into wild-type and EP2-deficient mice and examined their responses in vivo. Infusion of PGE2 or an EP2 agonist, butaprost, induces transient hypotension in wild-type mice. In EP2-deficient mice, butaprost failed to elicit hypotension but, unexpectedly, PGE2 evoked considerable hypertension. The authors discussed that the absence of EP2 abolishes the ability of the mouse vasculature to vasodilate in response to PGE2 and unmasks the contractile response via the vasoconstrictor PGE receptor(s). Moreover when fed on a high-salt diet the EP2deficient mice develop significant hypertension, with a concomitant increase in urinary excretion of PGE2. Thus, PGE2 is produced in the body in response to a high-salt diet and works to negatively regulate blood pressure via the relaxant EP2. Interestingly, the relative contribution of each EP subtype appears to be different between males and females (Audoly et al 1999). In females, EP2 and EP4 mediate the major portion of the vasodepressor response to PGE2. In males, EP2 plays only a modest role, and most of the vasodepressor effect is mediated by the phospholipase C-coupled EP1. In addition, in male mice, EP3 actively opposes the vasodepressor actions of PGE2. Thus the haemodynamic actions of PGE2 are mediated through complex interactions of several PGE receptors.
Allergic Asthma
Bone Remodelling The E-type PG can also affect bone remodelling, in both bone formation and resorption. The bone resorptive activity of PGE2 is associated with the occurrence of an increased number of osteoclasts. Sakuma et al (2000) and Miyaura et al (2000) reported impaired osteoclast formation in culture cells from EP4-deficient mice. They found that osteoclast formation is most potently induced by analogues with EP4-agonistic activity. Indeed, PGE2-induced osteoclast formation was impaired in osteoblast cultures from the EP4-deficient mice and osteoclast precursors from the spleen of wild-type mice. Suzawa et al (2000) further found that bone resorption by PGE2 was greatly decreased in bone from EP4-deficient mice, which showed an equal level of response to dibutyryl cAMP added to the culture as the bones from control mice. These studies clearly established the role of the EP4 subtype of PGE receptors in PGE2-mediated bone resorption. On the other hand, Li et al (2000) reported that the osteoclastogenic response to PGE2 and other stimulants is reduced significantly in culture cells from EP2-deficient mice. This apparent discrepancy may reflect redundant roles of the two relaxant PGE receptors. Sakuma and Miyaura et al found a small but significant PGE2-dependent response in EP4-deficient mice, and Li et al reported a further decrease in osteoclastogenesis when an EP4-selective antagonist was added to EP2-deficient cells. Exogenous PGE2 has been shown to induce not only bone resorption, but also bone formation. Yoshida et al (2002) examined the effects of PGE2 infusion into the periosteal region of the femur for 6 weeks in wild-type or mice deficient in each EP. PGE2 induced callus formation on the femur at the site of infusion in wild-type mice, but not in EP4-deficient mice. Consistently, bone formation was induced in wild-type mice by infusion of an EP4-selective agonist, but not by agonists specific for other EP subtypes. The EP4-agonist completely blocked the bone loss induced by ovariectomy or immobilization, and restored the bone mass with an increased density of osteoblasts lining the bone surface. These results suggest that EP4 is responsible for both bone resorption and bone formation induced by PGE2 and that activation of EP4 induces bone remodelling in vivo.
Allergic responses are often associated with an increase in prostanoid formation. PGD2 is the major prostanoid generated by mast cells upon allergen challenge and is produced abundantly in allergic diseases such as asthma, allergic dermatitis and conjunctivitis. However, the specific role played by PGD2 in allergy has been unclear. Matsuoka et al (2000) focused on the specific role of PGD2 in allergy by subjecting DP-deficient mice to ovalbumin-induced allergic asthma. They found a marked reduction in airway inflammation, obstruction and hypersensitivity in DP-deficient animals, suggesting that PGD2, acting via DP, works as a mediator of allergy. Interestingly, DP expression was seen in bronchiolar and alveolar epithelial cells only in antigen-challenged mice and not in mice immunized before the antigen challenge. On the contrary, Gavett et al (1999) reported that both COX-1- and COX-2-deficient mice showed enhanced allergic lung responses in a similar asthmatic model. This study by Gavett et al highlighted the beneficial aspects of prostanoids in an asthmatic model. The idea that prostanoids play protective roles in asthma was originally raised upon the finding that aspirin is not beneficial for allergy and can even cause asthmatic attacks in certain individuals. It is suspected that other prostanoids normally antagonize the action of PGD2, resulting in aspirin treatment having complex effects on the disease pathway. Another factor of consideration in this issue is the existence of another PGD2 receptor, CRTH2, which is also a plasma membrane-type receptor but more closely related to the Nformyl peptide receptor superfamily than to the other prostanoid receptors (Hirai et al 2001). This receptor is expressed preferentially in Th2 cells and has been shown to mediate chemotactic movement in response to PGD2. The exact roles of PGD2 via this receptor in allergy remain to be clarified. Sleep PGD2 is a potent endogenous sleep promoting substance in rats and other mammals including humans (Hayaishi et al 2000). PGD2 infused into the subarachnoid space underlying the rostral basal forebrain was effective in inducing sleep. Mizoguchi et al (2001) infused PGD2 into the lateral ventricle of wild-type and DP-deficient mice and determined the amounts of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. In wild-type mice, PGD2 infusion significantly increased NREM sleep. In DP-deficient mice, however, the amount of neither NREM nor REM sleep was altered at all by PGD2 infusion. Thus, PGD2 predominantly increased NREM sleep in wild-type mice and DP is crucially involved in PGD2-induced NREM sleep. The authors further demonstrated that the activation of DP elicited an increase in the extracellular adenosine content in the subarachnoid space of the rostral basal forebrain after PGD2 infusion. The amount of PGD2-induced sleep was reduced by pretreatment with an adenosine A2A receptor-specific antagonist (Satoh et al 1996), while administration of an adenosine A2A receptor-selective agonist into the subarachnoid space induced sleep (Satoh et al 1999). These results suggested that PGD2-induced sleep is mediated by the adenosine A2A receptor system. FP-KNOCKOUT MICE Luteolysis and Parturition FP-deficient mice do not show any abnormalities during early pregnancy or any changes in the oestrous cycle (Sugimoto et al
INSIGHT INTO PROSTANOID FUNCTIONS 1997). However, FP-deficient pregnant mice do not perform parturition, apparently due to the lack of labour. FP-deficient mice do not undergo parturition, even when given exogenous oxytocin, and show no prepartum decline in progesterone. A reduction in progesterone levels due to ovariectomy 24 h before term caused an upregulation of uterine receptors for oxytocin and normal parturition in the FP-deficient mice. These experiments indicate that the luteolytic action of PGF2a is required in mice to diminish progesterone levels and thus permit the initiation of labour. Indeed, ovarian expression of 20a-hydroxysteroid dehydrogenase, a catabolic enzyme for progesterone, is absent in FPdeficient mice, while this enzyme is induced at the mRNA and protein levels on day 19 of pregnancy in wild-type mice (Stocco et al 2000). The luteolytic role for PGF2a in the induction of labour in mice is also supported by the finding that mice lacking the gene encoding COX-1 also exhibited a similar parturition failure (Gross et al 1998). In these mice, production of PGF2a in intrauterine tissues during late pregnancy is significantly reduced and the administration of PGF2a on day 19 is able to restore normal parturition. In wild-type mice, the uterine expression of COX-1 mRNA gradually increases from day 15 of pregnancy, reaches maximal levels on day 17, and rapidly decreases after day 20, the day when parturition normally occurs. Tsuboi et al (2000) found that the uterine expression of COX-1 mRNA was still at high levels on day 20 in FP-deficient mice. This observation suggests that progesterone withdrawal serves as a negative feedback system for uterine COX-1 expression. Considering the phenotypes of the EP2- and FP-knockouts as well as the COX-2- and COX-1-knockouts, it can be concluded that prostanoids play essential roles in multiple processes in female reproduction. This is also supported by research on cytosolic phospholipase A2 (cPLA2), another key enzyme in prostanoid synthesis which catalyses cleavage of the eicosanoid precursor, arachidonic acid, from phospholipids. cPLA2-deficient females have smaller litter sizes and delayed parturition, which are interpreted as phenotypes equivalent to those seen in EP2-deficient and FP-deficient mice, respectively (Bonventre et al 1997; Uozumi et al 1997). Moreover, the administration of a progesterone receptor antagonist (RU-486) to mice at term to substitute for the luteolytic decline in progesterone corrected the defect in labour seen in the cPLA2-deficient mice.
223
blood–brain barrier (Samad et al 2001). PGE2 then enters the brain and cerebrospinal fluid and induces prostanoid receptor activation on neurons and microglia. This increases neuronal excitability and leads to non-painful stimuli becoming painful, basically converting a peripheral injury to a central pain response without nerve impulse transmission. At present, the possible involvement of EP1, EP3 and IP in pain has also been suggested (Minami et al 2001; Ueno et al 2001). Because the dorsal root ganglion expresses several types of prostanoid receptor mRNAs, including IP, EP1, EP3 and EP2 (Oida et al 1995), the exact contribution of receptors other than IP to pain generation should be carefully determined. Haemostasis PGI2 and TXA2, produced abundantly by vascular endothelial cells and platelets, respectively, are a potent vasodilator and vasoconstrictor, respectively. Mice lacking IP are viable, normotensive and reproductive; however, their susceptibility to thrombosis is increased (Murata et al 1997). Their platelets no longer respond to the PGI2 agonist cicaprost, neither does vascular smooth muscle relax upon this treatment, effectively demonstrating that a single IP subtype mediates both platelet and smooth muscle cell effects. The thrombotic tendencies of the IPdeficient mice were tested in a model of arterial thrombosis. IPknockouts demonstrated more extensive thrombus formation than wild-type animals after injury induced by ferric chloride. These findings confirmed the long-proposed role of PGI2 as an endogenous antithrombotic agent, and suggested that this antithrombotic system is activated in response to vascular injury to minimize its effects. TP-deficient mice showed an increased bleeding tendency and were resistant to cardiovascular shock induced by intravenous infusion of a TP agonist, U-46619, and arachidonic acid (Thomas et al 1998). Interestingly, endogenous PGE2 is likely to contribute to platelet aggregation via EP3; gene disruption of this PGE receptor resulted in an increased bleeding tendency and decreased susceptibility to thromboembolism (Ma et al 2001). CONCLUDING REMARKS
IP- AND TP-KNOCKOUT MICE Inflammation and Pain Vasodilation and pain sensation are two classic features of acute inflammation to which prostanoids appear to contribute. Aspirinlike drugs suppress these responses, and PGE2 and PGI2 can mimic these actions. Carrageenan-induced paw oedema and acetic acid-induced writhing are representative models for acute inflammation and pain, respectively. In IP-deficient mice, both responses are completely absent (Murata et al 1997). Thus, PGI2, acting on IP, works as a physiological mediator of these responses. However, it remains to be seen whether PGI2 and IP play important roles in other types of inflammation and pain. Regarding pain, PGs are involved not only in hyperalgesia, an increased sensitivity to a painful stimulus, but also in allodynia, a pain response to a usually non-painful stimulus (Malmberg and Yaksh 1992). The former is caused by sensitizing the free end of pain neurons at the site of peripheral inflammation, while the latter condition is frequently seen in neuropathic pain and is thought to occur in the spinal cord (Bley et al 1998). The circulating IL-1b cytokine, which originates at the site of peripheral injury and cannot pass the blood–brain barrier, induces both COX-2 and mPGES activities in cells lining the
Prostanoids have been suggested to have roles in many physiological processes, based on the various actions of aspirinlike drugs and exogenously added agonists. Until recently, however, the specific receptor involved in each process was unclear due to the failure to elucidate the molecular characteristics of each prostanoid receptor. The study of mice null for each prostanoid receptor has seen a remarkable development in the past several years. Some unexpected findings have raised many new questions about the traditional views based on the actions of aspirin-like drugs. However, post-genomic approaches such as SNP and expression profile analyses should reveal complete answers to those questions in the near future. For example, a polymorphic variation of the human TP gene has been identified and its connection to allergic diseases has been discussed (Unoki et al 2000). Thus, this exciting field of study will no doubt bring about many novel findings related not only to the prostanoids but also to the eicosanoids in the new millennium. REFERENCES Audoly LP, Tilley SL, Goulet J et al (1999) Identification of specific EP receptors responsible for the hemodynamic effects of PGE2. Am J Physiol, 277, H924–H930.
224
BIOCHEMICAL AND MOLECULAR PHARMACOLOGY
Bley KR, Hunter JC, Eglen RM and Smith JAM (1998) The role of IP prostanoid receptors in inflammatory pain. Trends Pharmacol Sci, 19, 141–147. Bonventre JV, Huang Z, Taheri MR et al (1997) Reduced fertility and postischaemic brain injury in mice deficient in cytosolic phospholipase A2. Nature, 390, 622–625. Capecchi MR (1989) Altering the genome by homologous recombination. Science, 244, 1288–1292. Coleman RA, Kennedy I, Humphrey PPA et al (1990) Prostanoids and their receptors. In Comprehensive Medicinal Chemistry, vol 3: Membranes and Receptors, Emmett JC (ed.). Pergamon, Oxford, 643–714. Coleman RA, Grix SP, Head SA et al (1994a) A novel inhibitory receptor in piglet saphenous vein. Prostaglandins, 47, 151–168. Coleman RA, Smith WL and Narumiya S (1994b) Classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. Davis BJ, Lennard DE, Lee CA et al (1999) Anovulation in cyclooxygenase-2-deficient mice is restored by prostaglandin E2 and interleukin-1b. Endocrinology, 140, 2685–2695. Dinchuk JE, Car BD, Focht RJ et al (1995) Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature, 378, 406–409. Doetschman T, Gregg RG, Maeda N et al (1987) Targeted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature, 330, 576–578. Ek M, Engblom D, Saha S et al (2001) Inflammatory response: pathway across the blood–brain barrier. Nature, 410, 430–431. Elmquist JK, Scammell TE and Saper CB (1997) Mechanisms of CNS response to systemic immune challenge: the febrile response. Trends Neurosci, 20, 565–570. Eppig JJ (1981) Prostaglandin E2 stimulates cumulus expansion and hyaluronic acid synthesis by cumuli oophori isolated from mice. Biol Reprod, 25, 191–195. Gavett SH, Madison SL, Chulada PC et al (1999) Allergic lung responses are increased in prostaglandin H synthase-deficient mice. J Clin Invest, 104, 721–732. Gross GA, Imamura T, Luedke C et al (1998) Opposing actions of prostaglandins and oxytocin determine the onset of murine labor. Proc Natl Acad Sci USA, 95, 11875–11879. Halushka PV, Mais DE, Mayeux PR and Morinelli TA (1989) Thromboxane, prostaglandin and leukotriene receptors. Annu Rev Pharmacol Toxicol, 10, 213–239. Hayaishi O (2000) Molecular mechanisms of sleep–wake regulation: a role of prostaglandin D2. Phil Trans R Soc Lond B Biol Sci, 355, 275–280. Hirai H, Tanaka K, Yoshie O et al (2001) Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J Exp Med, 193, 255–261. Hirata M, Hayashi Y, Ushikubi F et al (1991) Cloning and expression of cDNA for a human thromboxane A2 receptor. Nature, 349, 617–620. Hizaki H, Segi E, Sugimoto Y et al (1999) Abortive expansion of the cumulus and impaired fertility in mice lacking the prostaglandin E receptor subtype EP2. Proc Natl Acad Sci USA, 96, 10501–10506. Kabashima K, Saji T, Murata T et al (2002) Impaired mucosal integrity and enhanced immune response in colitis in mice lacking prostaglandin E receptor EP4. J Clin Invest, 109, 883–893. Kennedy CRJ, Zhang Y, Brandon S et al (1999) Salt-sensitive hypertension and reduced fertility in mice lacking the prostaglandin EP2 receptor. Nature Med, 5, 217–220. Kennedy I, Coleman RA, Humphrey PPA et al (1982) Studies on the characterization of prostanoid receptors: a proposed classification. Prostaglandins, 24, 667–689. Kiriyama M, Ushikubi F, Kobayashi T et al (1997) Ligand binding specificities of the eight types and subtypes of the mouse prostanoid receptors expressed in Chinese hamster ovary cells. Br J Pharmacol, 122, 217–224. Kluger MJ (1991) Fever: role of pyrogens and cryogens. Physiol Rev, 71, 93–127. Langenbach R, Loftin C, Lee C and Tiano H (1999) Cyclooxygenase knockout mice: models for elucidating isoform-specific functions. Biochem Pharmacol, 58, 1237–1246. Langenbach R, Morham SG, Tiano HF et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced
inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Li S, Wang Y, Matsumura K et al (1999) The febrile response to lipopolysaccharide is blocked in cyclooxygenase-2(7/7), but not in cyclooxygenase-1(7/7) mice. Brain, 825, 86–94. Li X, Okada Y, Pilbeam CC et al (2000) Knockout of the murine prostaglandin EP2 receptor impairs osteoclastogenesis in vitro. Endocrinology, 141, 2054–2061. Lim H, Paria BC, Das SK et al (1997) Multiple female reproductive failures in cyclooxygenase 2-deficient mice. Cell, 91, 197–208. Loftin CD, Trivedi DB, Tiano HF et al (2001) Failure of ductus arteriosus closure and remodeling in neonatal mice deficient in cyclooxygenase-1 and cyclooxygenase-2. Proc Natl Acad Sci USA, 98, 1059–1064. Ma H, Hara A, Xiao CY et al (2001) Increased bleeding tendency and decreased susceptibility to thromboembolism in mice lacking the prostaglandin E receptor subtype EP(3). Circulation, 104, 1176–1180. Malmberg AB and Yaksh TL (1992) Hyperalgesia mediated by spinal glutamate or substance P receptor blocked by spinal cyclooxygenase inhibition. Science, 257, 1276–1279. Matsuoka T, Hirata M, Tanaka H et al (2000) Prostaglandin D2 as a mediator of allergic asthma. Science, 287, 2013–2017. Minami T, Nakano H, Kobayashi T et al (2001) Characterization of EP receptor subtypes responsible for prostaglandin E2-induced pain responses by use of EP1 and EP3 receptor knockout mice. Br J Pharmacol, 133, 438–444. Miyaura C, Inada M, Suzawa T et al (2000) Impaired bone resorption to prostaglandin E2 in prostaglandin E receptor EP4-knockout mice. J Biol Chem, 275, 19819–19823. Mizoguchi A, Eguchi N, Kimura K et al (2001) Dominant localization of prostaglandin D receptors on arachnoid trabecular cells in mouse basal forebrain and their involvement in the regulation of non-rapid eye movement sleep. Proc Natl Acad Sci USA, 98, 11674–11679. Moham SG, Langenbach R, Loftin CD et al (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 83, 473–482. Morteau O, Morham SG, Sellon R et al (2000) Impaired mucosal defense to acute colonic injury in mice lacking cyclooxygenase-1 or cyclooxygenase-2. J Clin Invest, 105, 469–478. Murata T, Ushikubi F, Matsuoka T et al (1997) Altered pain perception and inflammatory response in mice lacking prostacyclin receptor. Nature, 388, 678–682. Nakamura K, Kaneko T, Yamashita Y et al (1999) Immunohistochemical localization of prostaglandin EP3 receptor in the rat hypothalamus. Neuroscience Lett, 260, 117–120. Narumiya S, Sugimoto Y and Ushikubi F (1999) Prostanoid receptors: structures, properties, and functions. Physiol Rev, 79, 1193–1226. Nguyen MT, Camenisch T, Snouwaert JN et al (1997) The prostaglandin receptor EP4 triggers remodelling of the cardiovascular system at birth. Nature, 390, 78–81. Oida H, Namba T, Sugimoto Y et al (1995) In situ hybridization studies of prostacyclin receptor mRNA expression in various mouse organs. Br J Pharmacol, 116, 2828–2837. Oshima M, Dinchuk JE, Kargman SL et al (1996) Suppression of intestinal polyposis in ApcD716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell, 87, 803–809. Priddy AR, Killick SR, Elstein M et al (1990) The effect of prostaglandin synthetase inhibitors on human preovulatory follicular fluid prostaglandin, thromboxane, and leukotriene concentrations. J Clin Endocrinol Metab, 71, 235–242. Sakuma Y, Tanaka K, Suda M, et al (2000) Crucial involvement of the EP4 subtype of prostaglandin E receptor in osteoclast formation by proinflammatory cytokines and lipopolysaccharide. J Bone Min Res, 15, 218–227. Samad TA, Moore KA, Sapirstein A et al (2001) Interleukin-1b-mediated induction of COX-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature, 410, 471–475. Saper CB and Breder CD (1994) The neurologic basis of fever. N Engl J Med, 330, 1880–1886. Satoh S, Matsumura H, Suzuki F and Hayaishi O (1996) Promotion of sleep mediated by the A2-adenosine receptor and possible involvement of this receptor in the sleep induced by prostaglandin D2 in rats. Proc Natl Acad Sci USA, 93, 5980–5984.
INSIGHT INTO PROSTANOID FUNCTIONS Satoh S, Matsumura H, Koike N et al (1999) Region-dependent difference in the sleep-promoting potency of an adenosine A2A receptor agonist. Eur J Neurosci, 11, 1587–1597. Segi E, Sugimoto Y, Yamasaki A et al (1998) Patent ductus arteriosus and neonatal death in prostaglandin receptor EP4-deficient mice. Biochem Biophys Res Commun, 246, 7–12. Sonoshita M, Takaku K, Sasaki N, et al (2001) Acceleration of intestinal polyposis through prostaglandin receptor EP2 in Apc(D716) knockout mice. Nature Med, 7, 1048–1051. Stocco CO, Zhong L, Sugimoto Y et al (2000) Prostaglandin F2a-induced expression of 20a-hydroxysteroid dehydrogenase involves the transcription factor NUR77. J Biol Chem, 275, 37202–37211. Sugimoto Y, Shigemoto R, Namba T et al (1994) Distribution of the messenger RNA for the prostaglandin E receptor subtype EP3 in the mouse nervous system. Neuroscience, 62, 919–928. Sugimoto Y, Yamasaki A, Segi E et al (1997) Failure of parturition in mice lacking the prostaglandin F receptor. Science, 277, 681–684. Sugimoto Y, Narumiya S and Ichikawa A (2000) Distribution and function of prostanoid receptors: studies from knockout mice. Prog Lipid Res, 39, 289–314. Suzawa T, Miyaura C, Inada M et al (2000) The role of prostaglandin E receptor subtypes (EP1, EP2, EP3, and EP4) in bone resorption: an analysis using specific agonists for the respective EPs. Endocrinology, 141, 1554–1559. Takeuchi K, Ukawa H, Kato S et al (1999) Impaired duodenal bicarbonate secretion and mucosal integrity in mice lacking prostaglandin E-receptor subtype EP3. Gastroenterology, 117, 1128–1135. Thomas DW, Mannon RB, Mannon PJ et al (1998) Coagulation defects and altered hemodynamic responsiveness in mice lacking receptors for thromboxane A2. J Clin Invest, 102, 1994–2001.
225
Tilley SL, Audoly LP, Hicks EH et al (1999) Reproductive failure and reduced blood pressure in mice lacking the EP2 prostaglandin E2 receptor. J Clin Invest, 103, 1539–1545. Tsuboi T, Sugimoto Y, Iwane A et al (2000) Uterine expression of prostaglandin H2 synthase in late pregnancy and during parturition in prostaglandin F receptor-deficient mice. Endocrinology, 141, 315–324. Ueno A, Matsumoto H, Naraba H (2001) Major roles of prostanoid receptors IP and EP(3) in endotoxin-induced enhancement of pain perception. Biochem Pharmacol, 62, 157–160. Unoki M, Furuta S, Onouchi Y et al (2000) Association studies of 33 single nucleotide polymorphisms (SNPs) in 29 candidate genes for bronchial asthma; positive association with a T924C polymorphism in the thromboxane A2 receptor gene. Hum Genet, 106, 440–446. Uozumi N, Kume K, Nagase T et al (1997) Role of cytosolic phospholipase A2 in allergic response and parturition. Nature, 390, 618–622. Ushikubi F, Nakajima M, Hirata M et al (1989) Purification of the thromboxane A2/prostaglandin H2 receptor from human blood platelets. J Biol Chem, 264, 16496–16501. Ushikubi F, Segi E, Sugimoto Y et al (1998) Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature, 395, 281–284. Watanabe K, Kawamori T, Nakatsuji S et al (1999) Role of the prostaglandin E receptor subtype EP1 in colon carcinogenesis. Cancer Res, 59, 5093–5096. Yoshida K, Oida H, Kobayashi T et al (2002) Stimulation of bone formation and prevention of bone loss by prostaglandin E EP4 receptor activation. Proc Natl Acad Sci USA.
Section Four Immunology, Endocrinology and Metabolic Regulation
19 Perspectives and Clinical Significance of Eicosanoids in Immunology, Endocrinology and Metabolic Regulation Milan R. Henzl Department of Obstetrics and Gynaecology, Stanford University, Palo Alto, CA, USA
ESSENTIAL FATTY ACIDS AND EICOSANOIDS Compounds known by the collective name of eicosanoids are present in virtually every tissue and cell of the human body. Because of this diverse universality, eicosanoids are critical for the activities of virtually all organ systems. The purpose of Section Four is to outline the metabolic pathways by which eicosanoids are generated and discuss in detail the molecular mechanisms by which they regulate and participate in immunity and endocrine and metabolic regulations, as well as in atherothrombosis and diseases of ageing. The genesis of eicosanoids is intimately related to essential fatty acids, therefore a discussion on their metabolism has been included in this Section (Min and Crawford, Chapter 22, this volume). In humans, two fatty acids, linoleic acid and linolenic acid, have been designated ‘‘essential’’ since they are indispensable to certain key life-sustaining processes. Nutrition is the only source that provides essential fatty acids to the human body, since mammals lack the ability to synthesize them de novo. Moreover, human tissues cannot convert linoleic acid into linolenic acid and vice versa. Both linoleic and linolenic acids are polyunsaturated fatty acids (PUFAs), i.e. they contain two or more double bonds in their molecules. To understand the biological implications of linoleic and linolenic acids fully, we shall stress a few points in the relationship between the chemical structures and biological functions of these molecules. Linoleic acid is a chain of 18 carbons with two double bonds, the second on the sixth carbon from the terminal methyl group of the molecule. In chemical shorthand, the complex formula of linoleic acid is abbreviated as 18:2 o-6 (Figure 19.1). Linolenic acid also comprises an 18-carbon chain but has three double bonds, the third one three carbons from the terminal methyl group (18:3 o-3). In the structure of essential fatty acids, the position of the last double bond, closest to the terminal methyl group, has biological consequences. Humans cannot convert an o-6 fatty acid into an o-3 fatty acid or vice versa. This is important for designing artificial nutrition, primarily bottlefeeding formulas for newborns, especially premature babies, and intravenous nutrition for adults that require it, particularly elderly patients. Artificial nutrition must contain well-balanced amounts of o-6 and o-3 fatty acids. Linoleic acid is a precursor of arachidonic acid (AA)—the starting point of formation of eicosanoids. Equally important is linolenic acid—the precursor of docosahexaenoic acid (DHA), The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
which is essential for several key functions of the organism, including normal brain and retinal development. However, AA has 20 carbons and four double bonds (20:4 o-6), while DHA consists of a 22-carbon chain with six double bonds (22:6 o-3). In humans, elongation of the carbon chains and increase in the number of double bonds is accomplished by enzyme systems known as elongases and desaturases. Defects in desaturases of essential fatty acids have been implicated in various diseases, including cardiovascular disease, obesity, non-insulin-dependent diabetes mellitus, hypertension, neurological diseases, immune disorders and cancer (Ntambi 1999). Once formed, AA is rapidly incorporated into phospholipids of the cell membrane. A group of phospholipases attacks the membrane and frees AA to initiate the AA cascade, which comprises two fundamental metabolic pathways. Enzymes of the cyclooxygenase metabolic pathway, cyclooxygenases 1 and 2 (COX-1, COX-2), convert AA into prostaglandin H2 (PGH2). PGH2 is the starting point for the formation of other prostanoids: tissue-specific synthases convert PGH2 into the remaining prostaglandins namely PGD2, PGE2, PGF2a, PGI2 (prostacyclin) and thromboxane (TX) A2. The second metabolic pathway converts AA into leukotrienes and lipoxins by the action of lipoxygenases. Products of both metabolic pathways are continuously synthesized in cell membranes of virtually all tissues but, unlike most other signalling molecules, they are not stored within the cell, but reach the cell exterior as soon as they are formed. Generally, it is accepted that under physiological conditions the prevalent enzymes catalysing the formation of prostanoids are phospholipase A2 and COX-1. COX-1 is present constitutively in most tissues and generates adequate prostanoids to maintain homeostatic processes of the organism. Tissue production of eicosanoids is accelerated when biosynthetic processes of the AA cascade are stimulated by various tissue injuries. Damaged cells release substances such as epinephrine, thrombin and bradykinin, which activate phospholipase A2. This enzyme produces additional AA from cell membrane phospholipids. The increased release of AA is a challenge to the AA metabolic pathway, which responds by induction of an isoenzyme, cyclooxygenase-2 (COX2), produced near the site of injury. There is a continuous demand for AA in both physiological and pathological conditions, and a continuous nutritional supply of linoleic acid is needed. Various enzymes generated in an organelle, the peroxisome, participate in the metabolism of the essential fatty acids, AA, and prostanoids. Absence of intact peroxisomes, disturbances in their
230
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 19.1 Linoleic acid (18:2o6). The 18 carbons in the molecule are counted from the COOH group (C-1) to the terminal CH3 group (C-18). There are two double-bonds in the molecule, the last one removed six carbons (pointed out by circles) from the terminal (C-18) methyl group. In chemical shorthand for 18:2o-6, 18 indicates the number of carbons in the molecule, 2 indicates the number of double-bonds, and o-6 indicates the distance of the first double bond from the terminal C-18 methyl group.
structure and assembly result in peroxisomal diseases, which will be discussed later (Diczfalusy et al 1991; Gordon et al 1993; Beier 2003). Clinical pharmacology studies have demonstrated that not only human adults but also infants born as early as in the 32nd week of gestation have the ability to synthesize AA and docosahaexanoic acids from their respective precursors, linoleic and linolenic acids (Salem et al 1996, 1999). However, these studies also revealed that adult human tissues are more efficient in converting linoleic acid into AA than converting linolenic acid into DHA. The difference in the rate of metabolism of the two essential fatty acids is even more pronounced in newborn infants, both full-term and premature. This finding has clinical implications for babies fed infant formulas containing linolenic acid but no DHA. Autopsy studies have shown that such infants have lower amounts of DHA in the brain than breast-fed babies (Farquardson et al 1992; Makrides et al 1994). Low circulating levels of DHA have been associated with poorer retinal development and decreased visual acuity in infants fed with a DHA-deficient formula. At 12 months, breast-fed babies performed better in intelligence tests than infants who were bottle-fed (Birch et al 1992; Carlson et al 1994). Moreover, it was found that a diet rich in linolenic acid alone is not able to support optimal visual function. Therefore, addition of preformed DHA is important in formulas for bottlefed premature babies (Birch et al 1992; Uauy 2001). During human pregnancy, there is a high demand for linoleic and linolenic acids, AA and DHA, and adequate amounts of these substances are necessary for proper prenatal and neonatal development of brain and other neural structures. Low concentrations of AA and DHA in cord blood have been associated with low birth weight. For infants, maternal breast milk is an excellent source of eicosanoids and essential fatty acids. The main nutritional source of o-6 fatty acids, principally linoleic acid, are plant oils, such as sunflower and soy bean oil, while fish oils are the main dietary source of o-3 fatty acids, principally linolenic acid. In experimental animals, dietary lack of linoleic acid has been associated with growth retardation, kidney malfunction, impaired reproduction and scaly skin. Clinical consequences of DHA deprivation during intrauterine development can lead to neurological abnormalities that have been linked to disturbances of peroxismal enzymes. The final step in the DHA biosynthesis is accomplished by enzymes released by peroxisomes, single-membrane organelles similar to lysosomes
which are present in nearly all eukaryotic cells. Peroxismal enzymes also catalyse the oxidation of long- and very long-chain fatty acids, and play a decisive role in the catabolism of prostaglandins (Diczfalusy 1994). Human developmental disorders can result from multiple enzyme deficiencies associated with defects of peroxisomes, caused by either faulty assembly of peroxisomes from proteins of the cell plasma or the absence of functional peroxisome proteins (‘‘peroxisome biogenesis disorders’’). There is also a set of disorders that result from deficiencies in only one enzyme produced by peroxisomes. In general, such ‘‘isolated’’ enzyme deficiencies affect only one of the numerous metabolic pathways that depend on peroxisomes. Of the peroxisome biogenesis disorders developing in utero, the Zellweger spectrum of syndromes is outstanding. It consists of an array of abnormalities that differ in their severity. The extreme form, cerebrohepatorenal syndrome, shows hypomyelination in cerebral white matter and includes global developmental delay, mental retardation, eye abnormalities, craniofacial deformities, hepatomegaly, renal cysts and impaired adrenocortical function. The results of basic research on disturbances of the metabolic pathways of a-linolenic acid and DHA were translated into a therapeutic approach to the Zellweger spectrum. Daily dosing of children with DHA proved beneficial to these individuals and became an established mode of therapy for this disease. Infants with defects in a single peroxisomal enzyme may suffer from disorders that are similar to those encountered in the Zellweger spectrum. Clinical syndromes may include craniofacial dysmorphism, severe retardation, hepatomegaly and other expressions of the Zellweger spectrum; however, other peroxisomal functions are normal. The nutritional importance of essential fatty acids, AA and DHA has been known since the 1970s. It is also recognized that dietary habits with respect to fat intake occurring in the last century have been the basis of the rise in mortality from a cluster of chronic non-communicable diseases that are the major cause of mortality worldwide. Cumulative mortality from the abovementioned diseases is higher than from infectious diseases, such as tuberculosis, HIV/AIDS and malaria. Key experimental and clinical data on the nutritional importance of unsaturated fatty acids (summarized in Chapter 23, this volume) have opened broad possibilities for health improvement of the population by correcting dietary habits. The American Heart Association recommends at least two servings/week of o-3rich fatty acid fish and claims a lower risk of heart attack for senior citizens who follow such a diet. To augment the intake of unsaturated fatty acids, we must resolve the question of whether to produce vegetable/fruit/animal species with a high content of these desirable macronutrients, either by natural breeding or genetic engineering. In an ongoing large-scale experiment, feeding farm-raised fish by diets high in o3 unsaturated fatty acids increased their tissues content of these macronutrients to amounts higher than are found in any other animal (Brown 2003). A remaining special objective is to continue designing an optimal diet for pregnant women and to compose effective formulas for preterm and hypotrophic babies. It is plausible to achieve these goals within the next decade. However, further clinical research is still needed to expand our knowledge on the biological functions of essential and polyunsaturated fatty acids, with the goal of defining nutritional requirements with optimal effects on human health. HYPERPROSTAGLANDINAEMIA Spontaneously increased levels of prostaglandins have not been encountered among adult populations. However, the antenatal form of Bartter syndrome, a rare congenital disease, is associated
PERSPECTIVES AND CLINICAL SIGNIFICANCE with marked overproduction of renal and systemic PGE2. Bartter syndrome is a set of closely related inherited renal tubular disorders resulting in hypokalaemic alkalosis, with similar clinical and biochemical features. Intrauterine development in Bartter syndrome is characterized by polyhydramnios, prematurity, dehydration at birth, hypercalciuria with subsequent nephrocalcinosis, failure to grow and mental retardation. The clinical symptomatology is driven by the excess of PGE2, which aggravates renal salt and water losses and accounts for systemic manifestations, such as fever, vomiting, secretory diarrhoea and failure to thrive (Gill et al 1976; Seyberth et al 1985). The role of PGE2 in the pathogenesis of the disease has been demonstrated by suppressing prostaglandin E synthesis with the cyclooxygenase inhibitor indomethacin. This intervention resulted in resolution of hypokalaemia, amelioration of symptoms and normal growth (Seyberth et al 1987). Genetic studies have identified mutations in four different genes and allowed prenatal diagnosis of hyperprostaglandin E syndrome. Prenatal indomethacin therapy has shown beneficial effects on the natural course of hyperprostaglandin E syndrome. Significantly, the treatment decelerated the progression of polyhydramnios, so that extreme prematurity could be prevented. Early postnatal diagnosis of Bartter syndrome allows effective water and electrolyte substitution before severe volume depletion occurs (Konrad et al 1999). IMMUNITY Eicosanoids have major effects in the regulation of inflammation and immunity. Originally, prostaglandins were regarded as simply inhibiting immune responses. As scientific knowledge about the biology of eicosanoids has expanded, it has been recognized that they intervene in the processes of inflammation and immunity in a complex and divergent fashion. Clinical and basic research during the last three decades has revealed the intricacy of the immune system, in which individual processes are interwoven and none is static. Immune responses change with time, age, various external influences and multiple other factors. Moreover, individual mediators exercise multiple actions, and none acts in isolation. Reviews of the molecular mechanisms of eicosanoids’ engagement in the immune system have recently appeared in the literature (Harris et al 2002; Fleming and Kelly, Chapter 20, this volume). The following paragraphs point out some basic functions of two prostaglandins involved in immune processes, namely PGE2 and a prostaglandin J2 derivative, PGD-2–15-deoxy-D12,14-PGJ2 (15-d-PGJ2). PGE2 interacts with T cells, B cells, and the antigen-presenting cells (APCs). At least four types of receptors have been identified to bind with PGE2 (EP1–EP4); in humans, at least seven splice variants of EP3 have been recognized. PGE2 receptors belong to the superfamily of G protein-coupled receptors. The extracellular domain of the receptor recognizes PGE2 molecules and binds them with high affinity. Binding of PGE2 to the receptor activates the G-protein which, in turn, triggers a chain of reactions resulting in either an influx of calcium ions or modulation of cAMP levels; calcium and cAMP being second messengers by which the prostaglandin signal is tranduced. PGE2 affects T cells throughout their life in a variety of ways, from the regulation of positive and negative selection in the thymus to the regulation of their proliferation. Well established is the fact that PGE2 inhibits proliferation of T cells. One of the mechanisms of this action is suppression of calcium release. Furthermore, PGE2 regulates apoptosis and cytokine production of mature T cells, and primes them to recognize invading organisms. With respect to B cells, PGE2 negatively affects proliferation of immature B cells and promotes their apoptosis, while it has no such effect on mature B cells. On the other hand, certain types of
231
B cells carry cyclooxygenases and produce PGE2. Thus, the B lineage cells can modulate immune responses both by reacting to PGE2 and by producing this prostaglandin. Antigen-presenting cells play a key role in B and T cellmediated immune response, and PGE2 modulates their function. PGE2 affects the activities of dendritic cells and macrophages. In all cell types involved in the immune responses, PGE2 modulates the production of cytokines. The role of PGE2 in inflammation is complex. In addition to its anti-inflammatory properties noted above, it is a mediator of pain and fever (see Ch 42, Ch 43 of this book) and is pro-inflammatory in certain models of arthritis. The functions of 15-d-PGJ2 have been explored only recently, and the role of this prostaglandin in the regulation of inflammation and neoplastic diseases remains under intensive investigation. 15-d-PGJ2 is abundantly produced by mast cells, platelets and alveolar macrophages and has been proposed as a key immunoregulatory lipid mediator. 15-d-PGJ2 is a potent antiinflammatory agent that downregulates inflammatory and immune responses by multiple mechanisms. One of these is promotion of the apoptosis of T and B cells. On the other hand, under certain conditions, 15-dPGJ2 can stimulate the production of proinflammatory mediators. Moreover, 15-d-PGJ2 is emerging as a potent agent that inhibits tumour cell growth both in vitro and in vivo. This is in contrast to PGE2, which promotes neoplasia. Furthermore, 15-d-PGJ2 inhibits vascular endothelial growth, with consequent inhibition of angiogenesis in vivo. The cloning of the receptors for the leukotrienes and the generation of mice lacking specific leukotriene receptors has broadened our appreciation of their actions. The cysteinyl leukotriene receptors (CysLTR) are expressed not only on smooth muscle but also on various leukocytes, and LTC4 and LTD4 elicit cytokine release from human eosinophils and mast cells. Disruption of the genes encoding LTC4 synthase and of CysLT1R revealed the contribution of the cysteinyl leukotrienes to plasma exudation in zymosan-induced peritonitis. Novel studies of mice lacking the high affinity LTB4 receptor revealed a role for LTB4 in recruitment of effector T cells to sites of inflammation. ASPIRIN-INTOLERANT ASTHMA (AIA) An example of the complexity of immune responses of the human body is anaphylactic reactions to drugs. Altered metabolism of eicosanoids has been postulated in aspirin-induced anaphylaxis, with its most serious clinical manifestation—aspirin-intolerant asthma (AIA) (Sampson and Holgate, Chapter 21, this volume). AIA is a crucial problem, since aspirin (see Figure 19.2) is one of the most frequently used drugs to treat common disorders such as pain, fever and inflammation, and the drug is extensively
Figure 19.2 Structure of aspirin. Aspirin consists of salicylic acid and the acetyl group. In humans, the acetyl group is split off and binds firmly (covalent bond) with serine of the enzyme cyclooxygenase-1 (see Figure 19.3)
232
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
employed for long-term treatment and prevention of diseases such as rheumatism and coronary heart disease (Berges-Gimeno et al 2002). Adverse respiratory reactions are not uncommon in aspirin users, and approximately 10% of middle-aged asthmatics develop aspirin intolerance. Many of these patients present with the triad of aspirin sensitivity, chronic rhinosinusitis with associated nasal polyposis, and asthma. The asthma attacks are often severe and may be life-threatening. The pathophysiological mechanism of aspirin intolerance has not been fully defined. A plausible explanation is that aspirin suppresses the cyclooxygenase pathway of arachidonic acid metabolism in favour of the lipoxygenase pathway, with overproduction of leukotrienes, principally the potent bronchoconstrictor cysteinyl leukotriene (cys-LT). Suppression of cyclooxygenases results in inhibited production of PGE2, a potent inhibitor of leukotrienes, and is sometimes called the ‘‘PGE2 brake’’. This PGE2 brake is operative in eosinophils in vitro. High numbers of eosinophils in nasal polyps and biopsies of bronchial mucosa, as well as in peripheral blood, are marked in aspirin-sensitive patients vs. non-aspirin-sensitive asthmatics and normal subjects. AIA patients tolerate selective COX-2 inhibitors, such as nimesulide, meloxicam and rofecoxib, but do not tolerate NSAIDs with greater selectivity for COX-1, such as aspirin and idomethacin. This raises the possibility that the PGE2 brake is mediated by constitutive COX-1. Patients with intolerance can be desensitized, and crossdesensitization takes place. This phenomenon is important, since low-dose NSAIDs other than the one that caused the reaction can be used for desensitization. Since it has been realized that leukotrienes are noxious agents precipitating asthma attacks, small molecule leukotriene antagonists have been synthesized and tested clinically. Such compounds include montelukast and zafirlukast. Their action is based on selective and competitive binding to leukotriene receptors. Both compounds have been approved in the USA for clinical use and they represent a new approach to the treatment of asthma. Although currently they are not considered as the first-line therapy for asthma, they are efficacious in many patients with moderate disease and have an acceptable side-effect profile. REYE’S SYNDROME Epidemiological data suggest that Reye’s syndrome is associated with aspirin use. The extreme clinical expression of this syndrome is characterized by acute encephalopathy, with fatty infiltration of the liver, kidney and sometimes heart and pancreas. Typically, Reye’s syndrome occurs in children younger than 16 years recovering from a viral disease. The precise pathophysiology is unknown. First described in 1963, the prevalence of this syndrome decreased dramatically when the use of aspirin in children was replaced by that of acetaminophen. ATHEROTHROMBOSIS Current opinion stresses the role of endothelial dysfunction in the pathogenesis of atherosclerosis, with significant diagnostic, therapeutic and prognostic implications. It also has been recognized that leukocytes initiate plaque formation, while platelets are decisive to the late occlusive events triggered by plaque rupture. Several powerful systems are involved in cardiovascular haemostasis and in the formation of atheromatous plaques and vascular occlusion. In accordance with the mission of this monograph, we will focus on certain prostanoids and their relation to blockers of thrombocyte action (Chlopicky and Gryglewski, Chapter 23, this volume).
Two prostanoids participate significantly in the regulation of cardiovascular homeostasis: the platelet-derived thromboxane A2 (TXA2) and prostacycline (PGI2), which originates in the endothelium. TXA2 is a potent amplifier of platelet aggregation and a vasoconstrictor that stimulates vascular smooth muscle proliferation. In contrast, PGI2 is a potent inhibitor of platelet aggregation and a vasodilator that inhibits smooth muscle cell proliferation. In healthy endothelium, PGI2 synthesis is catalysed by COX-1, although COX-2 has also been implicated in endothelial cells activated by inflammatory mediators. Biosynthesis of TXA2 in platelets depends exclusively on COX-1. Both PGI2 and TXA2 bind to seven transmembrane domain G protein-coupled receptors on the surface of the membrane of their respective target cells. Activated receptors transduce the message into the intracellular domain of the receptor, which activates a Gprotein. This triggers a chain of intracellular reactions that ultimately induce the target cells to exercise their respective functions. The regulatory system of TXA2$PGI2 involves several feedback mechanisms and counter-regulation which maintain the dynamic equilibrium between these two mediators. For example, TXA2 released by activated platelets stimulates endothelial cells to release PGI2, which in turn inhibits platelet activity (Marcus and Hajjar 1993). Alteration of the homeostatic balance between PGI2 and TXA2 can lead to atherosclerotic changes. The extent to which COX-2 participates on the synthesis of PGI2 could have therapeutic implications: the use of specific COX-2 inhibitors could suppress PGI2 and the consequent prevalence of TXA2 can lead to an imbalance promoting thrombus formation. Some studies have claimed that in healthy volunteers COX-2 inhibitors profoundly suppress endothelial PGI2 (McAdam et al 1999). Analysis of certain clinical trials has shown that the relative risk of developing a thrombotic cardiovascular event (myocardial infarction, cardiac thrombus, and ischaemic stroke) among rofecoxib users was 2.38 when compared to naproxen users (p=0.002). There was no significant difference in rates of similar cardiovascular events between patients treated with celecoxib and other NSAIDs (Mukherjee et al 2001). In another large-scale clinical trial, the cardiovascular thrombotic events were not more frequent in patients treated with rofecoxib (a selective COX-2 inhibitor) than in subgroups receiving placebo or nonselective COX-1/COX-2 inhibitors, e.g. nabumetone and diclofenac, although the relative risk vs. naproxen was 1.69 (Konstam et al 2001). The controversy has been partially resolved by a populationbased retrospective cohort study, comparing the occurrence of serious coronary heart disease among users of rofecoxib, other NSAIDs, and non-users of NSAIDs. Users of high-dose rofecoxib (425 mg/day) were 1.7 times more likely than non-users to have coronary heart disease. Users of lower doses of rofecoxib (25 mg/ day) and users of other NSAIDs had no increased risk of coronary heart disease (Ray et al 2001). These studies indicate that long-term dosing with high doses of rofecoxib should not be recommended. Platelet activation is another mechanism that contributes to clot formation. NSAIDs that have anti-platelet properties are associated with fewer cardiovascular complications. Such NSAIDs are aspirin, widely used for its cardioprotective action, and naproxen. Patients receiving naproxen showed fewer cardiovascular adverse events than those receiving rofecoxib (Kostam et al 2001), although this finding has been disputed (Ray et al 2002). The antiplatelet action is mediated through inhibition of glycoprotein IIb–IIIa receptors on the surface of activated platelets. Fibrinogen binds to these receptors, allowing platelets to aggregate and initiate the blood coagulation cascade. Medications designed specifically to block the final step in platelet aggregation by interfering with the binding of fibrinogen to the IIb–IIIa receptors are more potent than
PERSPECTIVES AND CLINICAL SIGNIFICANCE aspirin and other NSAIDs. Examples of IIb–IIIa receptor inhibitors include tirofiban (Aggrestat1), eptifibatide (Integrilin1), abciximab (ReoPro1), clopidogrel and ticlopidine. These agents have significantly reduced the incidence of death or nonfatal myocardial infarction. Specific glycoproteins IIb–IIIa directly stimulate vascular endothelium to produce PGI2. However, these agents are available only in an intravenous form; therefore, orally administered aspirin has retained its therapeutic value as a cardioprotective agent and in stroke prevention. The catalytic site of the platelet COX-1, together with thromboxane synthase, metabolizes AA into TXA2, the highly potent thrombocyte aggregator and vasoconstrictor. Aspirin (acetylsalicylic acid) irreversibly blocks access of AA to the catalytic site of the COX-1 molecule and by this action substantially suppresses the generation of TXA2. This process is schematically depicted in Figures 19.2 and 19.3. In humans, the acetyl group of the aspirin molecule is split off and enters the channels of the enzyme PGH synthase. There, the acetyl group binds chemically (covalent bonds) with the amino-acid serine 529 of the binding site of PGHS. In this way, the acetyl group blocks access of AA to the PGH synthase subunit, where AA is converted into PGH and other prostaglandins, prostacyclins and thromboxanes. Platelets have only limited numbers of PGHS molecules and cannot synthesize proteins. An aspirin dose as low as 30 mg (the aspirin content of one tablet is 325 mg) can suppress thromboxane production in all platelets
233
and prevent their haemostatic action for the duration of the platelets’ life. Other cells can synthesize cyclooxygenases, including PGH synthase, and within a few hours after aspirin ingestion the cyclooxygenase metabolic pathway is reinstituted (Garavito 1999; Catella-Lawson 2001). The question has arisen whether other NSAIDs would compete with aspirin for the access to the catalytic site of COX-1 and possibly reduce the effects of aspirin. The answer to this question is of clinical importance, since patients with cardiovascular disease and arthritis frequently receive both low-dose aspirin and other NSAIDs. A clinical pharmacology study has demonstrated that COX-2 inhibitors rofecoxib and diclofenac do not interfere with the platelet inhibitory action of aspirin, while ibuprofen, a prevalently COX 1 inhibitor, antagonizes the irreversible platelet inhibition induced by aspirin (Catella-Lawson 2001). A recent study (MacDonald and Wei 2003) has shown that patients with cardiovascular disease have increased risk of cardiovascular mortality [risk ratio (RR) 1.73] as compared to patients using aspirin alone (RR 1.0), aspirin plus diclofenac (RR 0.80) or aspirin plus other NSAIDs (RR 1.03). Prescribing physicians should be aware that concomitant use of ibuprofen with aspirin may limit aspirin’s cardioprotective effect in patients with cardiovascular risk (Catella-Lawson et al 2001). Aspirin per se does not perturb biosynthesis of PGI2. However, new information indicates that COX-2 is constitutively expressed in several tissues, and selective COX-2 inhibition has been
Figure 19.3 ‘‘Anti-prostaglandin’’ action of aspirin and ibuprofen in platelets. (A) The platelet cyclooxygenase-1 acts as prostaglandin H synthase. The enzyme is depicted as a dimer. Arachidonic acid (AA) is substrate for cyclooxygenase-1 and gains access to the core of the enzyme through the access channel. At the catalytic site AA is converted into prostaglandin H—the starting point for biosynthesis of other prostaglandins, including prostacyclins, and thromboxanes. (B) During dosing with aspirin, the acetyl group enters the access channel and binds chemically to the serine residue at position 529. The acetyl–serine bond is a firm covalent bond, which is irreversible for the life-time of the thrombocyte. (C) Ibuprofen competes with aspirin for the entry through the access channel to the catalytic site of the enzyme. However, ibuprofen does not bind chemically to the serine residue, therefore its blocking action lasts only for the life time of ibuprofen. T1/2 of ibuprofen is 2 h and elimination is completed 24 h after the last dose. Adapted by permission from Catella-Lawson et al (2001)
234
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
associated in animal studies with inhibition of PGI2, an effect that inhibits vasodilatation without inhibiting platelet aggregation. In humans, theoretically, NSAIDs selectively inhibiting COX-2 could create an imbalance between PGI2 and TXA2 in favour of TXA2. The lack of PGI2 would fail to oppose platelet aggregation and vasoconstriction, thus reducing the therapeutic efficacy of aspirin. Whether this theoretical supposition has any clinical validity remains to be explored in long-term human studies (Freston 1999). The efficacy of aspirin in prevention of myocardial infarction, stroke or venous thrombosis has been studied by regular followup of apparently healthy men by measurements of C-reactive protein, a marker of inflammation. The study subjects were randomly assigned to receive aspirin or placebo. The importance of inflammation in the genesis of cardiovascular diseases has been underscored by the fact that C-reactive protein was higher in men who developed myocardial infarction or ischaemic stroke during the approximately 8-year observation period. The study has also shown that aspirin decreases significantly (by 56.7%) the risk of cardiovascular events but only among men in the highest quartile of C-reactive protein values (Ridker et al 1997). Diabetes Mellitus The participation of eicosanoids in the pathogenesis of diabetes mellitus has been intensively studied by both traditional types of experimentation and methods of molecular biology (Robertson 1998). Metabolites of the COX pathway that have been consistently identified in pancreatic islet cells include PGE2 and PGF2a. Moreover, it has been revealed that in diabetes mellitus the islet tissues overexpress the COX-2, but not the COX-1, pathways of AA metabolism. Increased expression of COX-2 leads to overproduction of PGE2, which binds to a specific receptor in pancreatic islet cells, decreases cyclic AMP concentrations within these cells and consequently reduces insulin secretion. Upregulation of COX-2 may thus lead to insufficient insulin release and glucose intolerance, thereby playing a role in the pathogenesis of type 2 diabetes. Type 1 diabetes, or insulin-dependent diabetes mellitus (IDDM), is considered to be a T cell-mediated autoimmune disease associated with b-cell destruction. Cytokines produced in the autoimmune response induce overexpression of COX-2 and stimulate the formation of prostaglandins. Cytokines and overproduction of prostaglandins probably initiate and maintain inflammation, which is one factor in the destruction of the b-cells in the islets. Beneficial effects of most NSAIDs in diabetes mellitus, with the exception of indomethacin, can be explained by the dominance of COX-2 in b cells of pancreatic islets. Interestingly, sodium salicylate was the first NSAID found to have glucose-lowering effects (Ebstein 1876) and most modern NSAIDs boost insulin secretion and improve glucose tolerance. The pathophysiology of diabetes mellitus has become increasingly complex by the discovery of prostaglandin receptor subtypes for each prostanoid of the cyclooxygenase pathway. At least four subtypes of receptors have been identified for PGE2. G-proteins, another important mediator of PGE2induced inhibition of insulin secretion, have been shown to exhibit about 11 types of a-subunits. The lipoxygenase pathway of AA metabolism may also have effects in diabetes mellitus; it has been shown that 12-HPETE potentiates glucose-induced insulin secretion, while other products of the lipoxygenase pathway either do not affect insulin secretion or inhibit it. Nevertheless, drugs inhibiting lipoxygenase activity also suppress insulin secretion (Metz et al 1984). The involvement of COX-2 and the beneficial effect of NSAIDs on glucose tolerance indicate that the latter may have application in the treatment of diabetic patients. Although we cannot hope that NSAIDs will cure diabetes mellitus, it is reasonable to
propose that NSAIDs with a more specific action on the b cells of the pancreatic islets may reduce the insulin requirement in type 1 diabetes and improve tissue sensitivity to insulin in type 2 diabetes. Since such compounds would have to be used chronically, a wide dissociation between the beneficial effect on insulin secretion and undesirable adverse reactions on the cardiovascular and other systems has to be achieved. In designing therapeutic strategies for diabetes mellitus that would employ NSAIDs, further detailed studies of the effects of NSAIDs on the production and effects of prostaglandins in pancreatic islet cells would be helpful. BONE METABOLISM Studies of eicosanoids have improved our understanding of bone remodelling—a process of contiguous bone resorption and formation executed by osteoclasts and the osteoblastic activity of osteocytes. Remodelling of the human adult skeleton is directed by calcium-regulating hormones, parathormone, 1,25-dihydroxy vitamin D and calcitonin, as well as sex hormones. The effects of these hormones on bone are mediated through the action of local cytokines, growth factors and eicosanoids that are products of the cyclooxygenase as well as the lipoxygenase pathway of AA metabolism. Recently, the mechanisms of bone remodelling and the complexity of eicosanoid involvement in this process have been exhaustively reviewed (Kenny and Raisz 2002; Pilbeam and Raisz, Chapter 25, this volume). When evaluating the effects of eicosanoids on bone, we must distinguish between in vitro experimentation, animal studies and human clinical data. In extrapolating the results of experimental animal studies to humans, we have to bear in mind that processes of bone growth and formation differ substantially among various species, e.g. in humans, bone maturation is completed by the closure of ossification centres of the skeleton, while in rats, growth of bones continues as long as the animals live. These and other differences can explain discrepancies between experimental data and clinical findings. Prostaglandins play a dual role in bone remodelling, i.e. under certain circumstances they enhance bone resorption, while under other circumstances they stimulate bone formation. Exogenously administered prostaglandin E2 acts as a potent stimulator of bone growth in experimental animals. Administration of prostaglandin E2 to aged male rats increased the rate of bone formation on the periosteal and endocortical surfaces of the tibiae. The bone gain was accompanied by the reappearance of osteoblasts which could not be detected in pretreatment controls (Yao et al 1999). The effects of misoprostol on bone have been of great interest, since this derivative of PGE1 is clinically used to protect the gastrointestinal mucosa from the deleterious effects of NSAIDs. The compound is also used for ‘‘menses induction’’. Misoprostol prevented bone loss in ovariectomized rats in a dose-related manner (Sonmez et al 1999). Human data are not available. Adverse effect of hormonal replacement therapy with oestrogens and progestogens in menopausal women and the search for non-hormonal treatment of osteoporosis in men have stimulated an interest in prostaglandins as bone anabolic agents. Synthetic efforts resulted in analogues of prostaglandin F, which binds with human prostaglandin receptors in vitro and in vivo restores bone loss caused by ovariectomy in adult rats (Soper et al 2001). Of interest was also the relationship between selective oestrogen receptor mediators (SERMs), which decrease bone turnover, and prostaglandins, which stimulate bone turnover. Experimentation on aged ovariectomized rats has shown that droloxiphen—a SERM oestrogen—did not antagonize the bone anabolic response to prostaglandins (Ke et al 1999). The interaction of prostaglandin E2 with risedronate, a bisphosphonate approved for the treatment of Paget’s disease and osteoporosis, has also been
PERSPECTIVES AND CLINICAL SIGNIFICANCE studied. When given simultaneously, the effects of PGE2 were potentiated by risedronate during a 60 day treatment period. During the following 60 treatment-free days, the newly formed cortical bone remained largely preserved (Ma et al 1997). Since NSAIDs are widely used for long-term treatment, and since they inhibit synthesis of prostaglandins, the question was posed whether NSAIDs would have a deleterious effect on bone density and/or on bone healing of fractures. A large observational study of women aged 44–98 years has shown that regular daily use of propionic acid NSAIDs with or without simultaneous use of oestrogen may be helpful in preventing bone loss in older women. Acetic acid NSAIDs did not show this effect (Morton et al 1998). These study results require validation in prospective, randomized and double-blind clinical trials. The effects of NSAIDs on bone formation after injury were studied in numerous animal models. Recently, an experimental study in (NZ white) rabbits has compared the effects of oral administration of naproxen, a preferential COX-1 inhibitor, and rofecoxib, a selective COX-2 inhibitor, in a system that simulates bone healing after fractures. Control animals received plain water. Both NSAIDs were found to suppress bone formation after a bone lesion (Goodman et al 2002). Whether these and similar animal experiments are applicable to humans is not clear, since there are few properly conducted clinical studies. One of the human studies addressed the effects of piroxicam on the healing of dislocated Colles’ fractures in 42 postmenopausal women. This clinical trial was conducted according to a randomized double-blind study design. Piroxicam, an NSAID with mostly COX-1 inhibitory activity, did not decrease the rate of fracture healing. Patients treated with piroxicam showed a mean bone mineral decrease of 7% and 5% in the radius and the ulna, respectively, vs. 10% in the radius and 7% in the ulna in the placebo group—a non-significant difference. The patients who received piroxicam had significantly less pain. In their previous experiments in rabbits, the authors found a bone-sparing effect of piroxicam (Adolphson et al 1993). Another clinical group compared 32 patients with non-union of a fracture of the diaphysis of the femur to 67 patients whose fracture had united. There was a marked association between non-union and the use of NSAIDs after injury. Moreover, among patients receiving NSAIDs whose fractures had united, a high proportion showed delayed healing (Giannoudis et al 2000). Since treatment with NSAIDs is widespread, particularly in elderly patients, these conflicting results should stimulate larger collaborative clinical trials. ALZHEIMER’S DEMENTIA Involvement of eicosanoids in diseases of ageing have been reviewed in detail in Section IV (Hornych, Chapter 26, this volume). The following paragraphs focus on Alzheimer’s dementia (AD), perhaps the most frightening disorder of ageing. The collective burden of AD is underscored by its high prevalence and skyrocketing costs. Currently, approximately 4 million Americans suffer from this disease; it has been calculated that this number will increase to 14 million within the next 50 years unless a prevention and/or cure is found. One in 10 persons over 65 and nearly half of those over 85 have AD, but the disease rarely affects individuals in their 30s. Data from Europe show the prevalence of AD to be 5–6 million cases. The costs of AD in the USA are at least $100 billion a year. Only 5% of AD cases can be linked to the detectable familial forms associated with mutations. However, there is no significant difference in the concordancy rate between monozygotic and dizygotic twins. These findings strongly suggest that epigenetic and/or environmental factors are operative in familial AD, as well as in its common sporadic form.
235
Current research data point to neuroinflammation as the central feature of AD. The histopathological picture of AD is dominated by loss of synapses, intracellular neurofibrillary tangles, extracellular amyloid deposits and neuronal cell death. However, the pathophysiological relationships between these four lesions are not well understood. Post mortem immunohistochemical studies have revealed a state of chronic inflammation, a number of inflammatory mediators, proinflammatory cytokines and activated complement products. The described changes are limited to the regions of the brain that are affected in AD. The inflammatory mediators appear to be produced locally and selectively elevated in the affected areas of AD brains. Systemic inflammatory mechanisms are not involved in these processes. A key feature is an enhanced activity of phospholipase A2, which releases AA from phospholipids of the neuronal cell membrane. The released AA is metabolized by enzymes with the formation of highly active prostaglandins. AA is also metabolized along the lipoxygenase pathway into leukotrienes, lipoxins and HPETE. Further AA products include isoprostanes and other compounds that promote and sustain inflammatory responses and the prostaglandin-generating cyclooxygenases COX-1 and COX2. Overactivation of PLA2 and PG production are among the earliest initiating events in triggering brain-damage pathways. A number of clinical observations and basic research data have suggested that the use of NSAIDs may help to prevent and/or ameliorate AD (Hull et al 2000, 2002a, 2002b). Only recently, in a prospective population-based cohort study, 6989 subjects aged 55 or older who were free of dementia at base line have been followed up for 10 years. The risk of AD was estimated in relation to the use of NSAIDs, as documented in pharmacy records. The longterm use (24 months or more) of NSAIDs showed protection against AD but not against vascular dementia. Short and intermediate use (41 month, and 51 and 424 months) did not provide this protection (In t0 Veld et al 2001). A prodromal stage of Alzheimer’s disease is the mild cognitive impairment syndrome. Patients would benefit from defining the early symptoms and the development of reliable tests that would capture the prodromal stage of the disease. Moreover, there is a need to find biological markers of AD as well as of its prodromal stage. Early suspicion and diagnosis of AD might allow timely treatment and possibly prevent progression of the disease into its extreme clinical manifestations. Immunohistochemical and molecular biological studies on immune system components in the AD brain are revealing the complexities of the innate immune reaction of which the complement system, microglia and cytokines are the key components. However, this very complexity may offer points of therapeutic intervention (McGeer and McGeer 2002). By better understanding inflammatory and immunoregulatory processes in AD, it may be possible to develop new types of antiinflammatory agents. Although a definitive cure of AD may not be achieved, the new antiinflammatory approaches might help to slow the progression and/or delay the onset of this devastating disorder. ACKNOWLEDGEMENTS The author wishes to express his gratitude to Anthony Allison, DSc, FRCP (London), Rhona R. Simmonds, PhD and Jonathan P. Arm, MD, for their invaluable input into this chapter. REFERENCES Adolphson P et al (1993) No effects of piroxicam on osteopenia and recovery after Colles’ fracture. A randomized, double-blind, placebocontrolled, prospective trial. Arch Orthop Trauma Surg, 112(3), 127–130.
236
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Beier K (2003) Peroxisomes. http://www.unibas.ch/anatomie/beier/ bericht3.htm (accessed 18 February). Berges-Gimeno MP, Simon RA and Stevenson DD (2002) The natural history and clinical characteristics of aspirin-exacerbated respiratory disease. Ann Allerg Asthma Immunol, 89(5), 474–478. Birch EE, Birch DG, Hoffman DR and Uauy R (1992) Dietary essential fatty acid supply and visual acuity development. Invest Ophthalmol Vis Sci, 33, 3242–3253. Brown P (2003) Researcher Design Fishes’ Diets to Make You Healthier. Purdue University, USA. Carlson SE, Werkman SH, Peeples JM and Wilson WM III (1994) Growth and development of premature infants in relation to o 3 and o 6 fatty acid status. World Rev Nutr Diet, 75, 63–69. Catella-Lawson F, Reilly MP, Kapoor SC et al (2001) Cyclooxygenase inhibitors and the antiplatelet effects of aspirin. N Engl J Med, 345(25), 1809–1817. Diczfalusy U, Kase BF, Alexson SE and Bjorkhem I (1991) Metabolism of prostaglandin F2 a in Zellweger syndrome. Peroxisomal b-oxidation is a major importance for in vivo degradation of prostaglandins in humans. J Clin Invest, 88(3), 978–984. Diczfalusy U (1994) b-Oxidation of eicosanoids. Prog Lipid Res, 33(4), 403–428. Ebstein W (1876) U¨ber Diabetes. Leipzig: Deutsches Archiv fu¨r klinische Medicin, XXVIII, XXX pp. Farquharson J, Cockburn F, Patrick WA et al (1992) Infant cerebral cortex phospholipid fatty-acid composition and diet. Lancet, 340, 810–813. Freston JW (1999) Rationalizing cyclooxygenase (COX) inhibition for maximal efficacy and minimal adverse events. Am J Med, 107, 78S–88S. Garavito RM (1999) Aspirin. Sci Am, May, 108. Giannoudis PV, MacDonald DA, Matthews SJ et al (2000) Non-union of the femoral diaphysis. The influence of reaming and non-steroidal antiinflammatory drugs. J Bone Joint Surg Br, 82(5), 655. Gill JR Jr, Frolich JC, Bowden RE et al (1976) Bartter’s syndrome: a disorder characterized by high urinary prostaglandins and a dependence of hyperreninemia on prostaglandin synthesis. Am J Med, 61, 43–51. Goodman S et al (2002) COX-2 selective NSAID decreases bone ingrowth in vivo. J Orthop Res, 20(6), 1164–1169. Gordon JA, Warnock LJ and Spector AA (1993) Attenuated prostaglandin formation in peroxisomal-deficient human skin fibroblasts. J Clin Invest, 92(1), 169–178. Harris SG et al (2002) Prostaglandins as modulators of immunity. Trends Immunol, 23, 144–150. Harris SG and Phipps RP (2002) Prostaglandin D(2), its metabolite 15-dPGJ(2), and peroxisome proliferator activated receptor-g agonists induce apoptosis in transformed, but not normal, human T lineage cells. Immunology, 105(1), 23–34. Hull M, Lieb K and Fiebich BL (2000) Antiinflammatory drugs: a hope for Alzheimer’s disease? Exp Opin Investig Drugs, 9, 671–683. Hull M, Lieb K and Fiebich BL (2002a) Pathways of inflammatory activation in Alzheimer’s disease: potential targets for diseasemodifying drugs. Curr Med Chem, 9, 83–88. Hull M, Lieb K and Fiebich BL (2002b) Non-steroidal drugs and Alzheimer’s disease. N Engl J Med, 346, 1171–1173. In t’ Veld BA, Ruitenberg A, Hofman A et al (2001) Non-steroidal antiinflammatory drugs and the risk of Alzheimer’s disease. N Engl J Med, 345, 1515–1521. Kaplan MM (1994) Reye’s syndrome. In Stein JH (ed.), Internal Medicine, 4th edn. St Louis, MO: Mosby 629. Ke HZ, Crawford DT, Qi H et al (1999) Droloxifene does not blunt bone anabolic effects of prostaglandin E2, but maintains prostaglandin E2restored bone in aged, ovariectomized rats. Bone, 24, 41–47. Kenny AM and Raisz LG (2002) Mechanisms of bone remodeling: implications for clinical practice. J Reprod Med, 47(1, suppl), 63–70. Konrad M, Leonhardt A, Hensen P et al (1999) Prenatal and postnatal management of hyperprostaglandin E syndrome after genetic diagnosis from amniocytes. Pediatrics, 103, 678–683. Konstam MA, Weir MR, Reicin A et al (2002) Cardiovascular thrombotic events in controlled, clinical trials of rofecoxib. Circulation, 106(5), e18. Ma YF, Pan Z, Jee WS et al (1997) Intermittent on/off prostaglandin E2 and risedronate are equally anabolic as daily PGE2-alone treatment in cortical bone of ovariectomized rats. J Bone Miner Res, 12, 2108–2112. MacDonald TM and Wei L (2003) Effect of ibuprofen on cardioprotective effect of aspirin. Lancet, 361, 573–574.
Makrides M, Neumann MA, Byard RW et al (1994) Fatty acid composition of brain, retina, and erythrocytes in breast- and formula-fed infants. Am J Clin Nutr, 60, 189–194. Marcus AJ and Hajjar DP (1993) Vascular transcellular signaling. J Lipid Res, 34(12), 2017. Martı´ n-Garcı´ a C et al (2002) Safety of a cyclooxygenase-2 inhibitor in patients with aspirin-sensitive asthma. Chest, 121, 1812–1817. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96(1), 272–277. McGeehan MK and Bush RK (2002) The mechanisms of aspirinintolerant asthma and its management. Curr Allerg Asthma Rep, 2, 117–125. McGeer PL and McGeer EG (2002) Local neuroinflammation and the progression of Alzheimer’s disease. J Neurovirol, 8, 529–538. Metz SA, Murphy RC and Fujimoto W (1984) Effects on glucose-induced insulin secretion of lipoxygenase-derived metabolites of arachidonic acid. Diabetes, 33, 119–124. Morton DJ, Barrett-Connor EL and Schneider DL (1998) Non-steroidal antiinflammatory drugs and bone mineral density in older women: the Rancho Bernardo study. J Bone Min Res, 13, 1924–1931. Mukherjee D, Nissen SE and Topol EJ (2001) Risk of cardiovascular events associated with selective COX-2 inhibitors. J Am Med Assoc, 286, 954–959. Ntambi JM (1999) Regulation of stearoyl-CoA desaturase by polyunsaturated fatty acids and cholesterol. J Lipid Res, 40, 1549–1558. Picado C (2002) Aspirin-intolerant asthma: role of cyclooxygenase enzymes. Allergy, 57(suppl 72), 58–60. Purdue Center for Enhancing Foods to Protect Health (2003) http:// www.agriculture.purdue.edu (6 Jan 2003; accessed 5 February 2003). Ray WA, Stein CM, Daugherty JR et al (2002a) COX-2 selective nonsteroidal antiinflammatory drugs and risk of serious coronary heart disease. Lancet, 360, 1071–1073. Ray WA, Stein CM, Hall K et al (2002b) Non-steroidal antiinflammatory drugs and risk of serious coronary heart disease: an observational study. Lancet, 359, 118–123. Ridker PM, Cushman M, Stampfer MJ et al (1997) Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med, 336, 973–979. Robertson RP (1998) Dominance of cyclooxygenase-2 in the regulation of pancreatic islet prostaglandin synthesis. Diabetes, 47, 1379–1383. Salem N Jr, Wegher B, Mena P and Uauy R (1996) Arachidonic and docosahexaenoic acids are biosynthesized from their 18-carbon precursors in human infants. Proc Natl Acad Sci USA, 93(1), 49–54. Salem N Jr, Pawlosky R, Wegher B and Hibbeln J (1999) In vivo conversion of linoleic acid to arachidonic acid in human adults. Prostagland Leukotrienes Essent Fatty Acids, 60, 407–410. Seyberth HW, Koniger SJ, Rascher W et al (1987) Role of prostaglandins in hyperprostaglandin E syndrome and in selected renal tubular disorders. Pediatr Nephrol, 1(3), 491–497. Seyberth HW, Rascher W, Schweer H et al (1985) Congenital hypokalemia with hypercalciuria in preterm infants: a hyperprostaglandinuric tubular syndrome different from Bartter syndrome. J Pediatr, 107, 694–701. Sonmez AS, Birincioglu M, Ozer MK et al (1999) Effects of misoprostol on bone loss in ovariectomized rats. Prostagland Lipid Mediat, 57(2–3), 113–118. Soper DL, Milbank JB, Mieling GE (2001) Synthesis and biological evaluation of prostaglandin-F alkylphosphinic acid derivatives as bone anabolic agents for the treatment of osteoporosis. J Med Chem, 44, 4157–4169. Uauy R, Hoffman DR, Peirano P et al (2001) Essential fatty acids in visual and brain development. Lipids, 36, 885–995. Uauy R and Mena P (1999) Requirements for long-chain polyunsaturated fatty acids in the preterm infant. Curr Opin Pediatr, 11, 115–120. Uauy-Dagach R and Mena P (1995) Nutritional role of o-3 fatty acids during the perinatal period. Clin Perinatol, 22, 157–175. Wilhelm Ebstein, German internist. http://www.whonamedit.com (accessed 18 February 2003). Yao W, Jee WS, Zhou Het al (1999) Anabolic effect of prostaglandin E2 on cortical bone of aged male rats comes mainly from modelingdependent bone gain. Bone, 25(6), 697–702.
20 Prostaglandins and the Immune Response Diana C. Fleming and Rodney W. Kelly University of Edinburgh Centre for Reproductive Biology, Edinburgh, UK
The one prostaglandin with major and far-reaching effects on the immune system is prostaglandin E2 (PGE2) and this chapter will mainly deal with this compound and its relevance in different loci. PGE is a proinflammatory agent by virtue of its action on blood vessels and their surrounding cells, where it enhances oedema (Williams and Morley 1973) and neutrophil ingress into tissue (Rampart and Williams 1987). However, by affecting the differentiation and function of many cells of the immune system, PGE acts as a counter-measure to an activated immune system prior to and during an inflammatory process, such as that initiated by infection, eventually promoting the healing phase and restoring the normal state of homeostasis. The most effective response to infection involves the priming of T cells to recognize specific epitopes on the particular invading organism—the adaptive immune system. Antigen is processed and presented by antigen-presenting cells, which include dendritic cells, macrophages and epithelial cells. Processed antigen is presented in conjunction with major histocompatibility complex (MHC) to T cells, which are then primed to recognize that specific antigen on subsequent exposure. Thus, a second encounter with bacteria or viruses results in the clonal expansion of highly specific T cells. In the expansion of T cells, the monocyte may act as an ‘‘accessory’’ cell that can secrete stimulatory cytokines, such as IL-12, or inhibitory ones, such as PGE and IL-10. However, such responses are relatively slow and the innate immune system allows a more rapid activation, which can also sensitize the adaptive response. The innate immune system can respond to infection by recognizing specific, conserved, bacterial or viral products.
ANTIGEN-PRESENTING CELLS
signal (either humoral or based on cell–cell contact) which allows the dendritic cell to recognize that the antigen is likely to originate from a pathogen. In addition, a third signal has been proposed (Kalinski et al 1999) which, by determining the cytokine secretion pattern of the primed T cell, will generate either a Th1 or Th2 cell. This third signal is thought to be related to the IL-12 secretion of the dendritic cell, IL-12 being the predominant cytokine produced by dendritic cells. Those mature dendritic cells that secrete IL-12 will generate Th1 cells, whereas in the absence of IL-12, Th2 cells will be formed. This effect is determined by exposure to humoral factors at the site of antigen capture; interferon-g (IFN-g) leads to a dendritic cell that secretes IL-12 (Kalinski et al 1999) and PGE2 gives rise to a dendritic cell that secretes low levels of IL-12 (Kalinski et al 1998; Kraan et al 1995). Kalinski et al have elucidated that the IL-12 p40 subunit is produced without the p35 subunit, under the influence of PGE2, and this acts as an antagonist to IL-12 (Kalinski et al 2001). They go on to suggest that the susceptibility of the mature dendritic cell to functional modulation by these cytokines appears to be lost or strongly reduced. Others suggest that there is still modulation; addition of PGE2 to matured dendritic cells has been shown to increase the expression of Th1 cytokines, particularly IFN-g (Jonuleit et al 1997; Steinbrink et al 2000). It has also been suggested that PGE affects the expression of some cell surface markers, such as B7 and ICAM-1 (Harizi et al 2001; MenetrierCaux et al 1999; Steinbrink et al 2000) but the effect is not clear and depends on the time of stimulation and the cytokine environment at that time. This change in cell surface markers could further influence T cell stimulation by altering ‘‘signal 2’’, the co-stimulatory signal (Bartik et al 1994; Saha et al 1994), e.g. B7 is a modulator of anergy and, by inhibiting its expression, PGE could downregulate the adaptive immune response.
Dendritic Cells
Monocytes/Macrophages
The dendritic cell is the most effective antigen-presenting cell. Prostaglandins act on the dendritic cell, probably during maturation (Vieira et al 2000), altering its cytokine profile and hence affecting the cytokine profile of the T lymphocytes. Since these cells also possess the COX-1 and COX-2 enzymes (Whittaker et al 2000), they have the capacity to produce prostaglandins. In skin and mucosal surfaces the immature cell captures antigen and matures while migrating to the lymph node, where the antigen is presented to T cells. PGE2 has been shown to enhance the yield, maturation, migration and immunostimulatory capacity of dendritic cells in vitro (Jonuleit et al 1997; Kurth et al 2001). At least two signals are required for successful priming of naı¨ ve T cells, the processed antigen itself and a co-stimulatory
Prostaglandin Production
The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
The monocyte and its further differentiated phenotype, the macrophage, are major participants in both the innate and adaptive immune systems. In both cases, PGE generated by the monocyte downregulates the system to limit the inflammatory response and to induce repair-phase events. The monocyte can arguably be described as the most important source of stimulatory cytokines, such as IL-1 and IL-12, and suppressive agents, such as TGFb, IL-10 and PGE. The monocyte is a major source of PGE (Humes et al 1977) but also can secrete thromboxane A2 (TXA2) (Reale et al 1996; Remick et al 1987). Cytokines such as IFN-g stimulate PGE
238
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
preferentially (Nichols and Garrison 1987) and PGE suppresses TXA production (Widomski et al 1991). The main reason for TXA production in the monocyte is unclear, although it is reported to induce adhesion molecule expression on the cell surface (Wagner et al 1996). It has been suggested that the eicosanoid produced from a macrophage is influenced by the predominating COX isoform, COX-1 giving rise to PGI2, TXA2, PGD2 and 12-HPETE with minimal PGE2, while if COX-2 is upregulated, PGE2 and PGI2 predominate (Brock et al 1999). Monocytes play a major role in recognizing the presence of Gram-negative bacteria, using receptors for the bacterial cell wall component, lipopolysaccharide (LPS). This recognition of molecules originating from pathogens allows a rapid upregulation of the monocyte and synthesis of lipids and cytokines that initiate a paracrine defence against the infection. An important reason for understanding the mechanisms downstream of LPS activation is that excessive response to bacterial infection and LPS can result in Gram-negative septic shock. One major component of the LPS response is the stimulation of PGE2 in monocytes and macrophages by a cyclooxygenase-2-dependent mechanism. In monocytes the importance of PGE is underlined by the observation that endoperoxides are preferentially metabolized to PGE by a simultaneous induction by LPS of COX-2 and PGE synthase (Matsumoto et al 1997). LPS also stimulates phospholipase A2, the enzyme necessary for the release of arachidonic acid from phospholipid pools (Shankavaram et al 1998). Thus, the release of substrate, the stimulation of active cyclooxygenase and the specific PGE-forming enzyme are all induced with LPS. LPS interacts with a receptor complex on the surface of monocyte and macrophages. Although CD14 binds LPS (enhanced by LPS binding protein), this glycoprotein has no intracellular segment that could mediate signal transduction. CD14 therefore interacts with Toll-like receptors (TLR), such as TLR4 for the Gram-negativederived LPS (Chow 1999; Rhee and Hwang 2000). The mammalian TLRs are homologues of the Drosophila Toll and are transmembrane proteins with an intracellular domain with homology to the IL-1 receptor (Rock et al 1998). These proteins, when stimulated with ligand, are major activators of the innate immune system (Kaisho and Akira 2000). The involvement of NF-kB in signal transduction downstream of LPS leading to COX-2 expression is controversial. Unravelling the mechanisms involved are made more complicated because a second phase of COX-2 induction by LPS has been proposed (Caivano et al 2001). The experimental approaches have largely used the promoter region of COX-2 coupled to a luciferase reporter. One such study suggests that COX-2 expression in murine monocytes responding to LPS is mediated by cAMP response element (CRE) and NF-IL-6 sequences in the promoter region, but is NF-kB independent (Wadleigh et al 2000). However, other reports, including a similar study using a more complete upstream promoter region for COX-2, suggest that activation of NF-kB is required for efficient COX-2 expression (Hwang et al 1997; Hwang 2000). Such reports are not as contradictory as they might seem; a more recent characterization of the complex mechanisms that are necessary for control of LPS activation of COX-2 shows a redundancy in the system (Mack Strong et al 2001). Thus, NF-IL-6, NF-kB, and CRE are all involved in activating the promoter of COX-2 in monocytes but optimum transcription is only induced when any two of the three promoter sequences are activated. The second delayed phase of COX-2 transcription is then controlled by different mechanisms and the C/EBP transcription factors c/EBPb and C/EBPd are involved (Caivano et al 2001). The COX-2 message is subject not only to transcriptional control but also to post-transcriptional events. The posttranscriptional control, as studied in mouse mesangial cells, is
mainly governed by the first 60 nucleotides of the 3’ UTR, which contain multiple AUUUA sites, known to destabilize the message. However, other sites distal to this region also contributed to message instability and translational efficiency and not all of these contain the AUUUA motif (Cok and Morrison 2001). PGE levels in monocytes are controlled by catabolism as well as by synthesis and the metabolic enzyme 15-hydroxy prostaglandin dehydrogenase (PGDH) is induced in macrophages during their development from monocytes (Pichaud et al 1997). When indomethacin was added to phorbol ester matured HL60 cells, COX-2 remained unchanged but PGDH was increased (Frenkian et al 2001). This suggests an additional mechanism for nonsteroidal antiinflammatory drugs (NSAIDs) to reduce active prostaglandin levels, although a previous report examined the effects of indomethacin on the ductus arteriosus and found that PGDH was reduced by this agent (Takizawa et al 1996). The recent report showing the promoter sequence for PGDH should accelerate further studies on the control of this critical enzyme (Greenland et al 2000). Inhibitory cytokines that may be involved in the resolution of the immune response, such as IL-1 receptor antagonist (Porreca et al 1996) IL-4 (Mertz et al 1996) and IL-10 (Mertz et al 1994), all reduce COX-2 stimulated by LPS. These effects can be reversed by agents that raise intracellular cAMP, including PGE (Hinz et al 2000; Mertz et al 1994), but from what we know about the promoter sequence of COX-2 this reversal is likely to be partial. Actions of PGE on the Monocyte The important actions of PGE on the monocyte are mediated by an increase in cAMP and mainly use the EP4 receptor (Mori et al 1996). A rise in intracellular cAMP levels can be achieved with PGE or PGI, experimentally with cAMP analogues such as dibutyryl cAMP or with agents that block the catabolic enzyme phosphodiesterase. Raised intracellular cAMP in the monocyte is generally suppressive but this should be qualified, since early in the immune response the effects of cAMP on T cell priming by neighbouring dendritic cells will be to generate Th2 or Tr cells whilst in the restorative phase of the response the action may be to stimulate extracellular matrix remodelling. In the presence of LPS, PGE has a powerful effect on monocytes in stimulating IL-10 release (Strassmann et al 1994) and inhibiting IL-12 (Kraan et al 1995). These changes are cAMPmediated and a similar response is seen with phosphodiesterase inhibitors stimulating IL-10 (Kambayashi et al 1995) and inhibiting IL-12 (Moller et al 1997). Increased cAMP also inhibits TNFa synthesis (Souness et al 1996; Verghese et al 1995) and IL-1 release is attenuated by phosphodiesterase inhibitors, although the message is increased (Verghese et al 1995). A major role for PGE in both the monocyte and the mucosal epithelial surfaces of the body is the control of MHC expression (Figure 20.1), the molecular mechanism necessary for antigen presentation. The effect of IFN-g released by intraepithelial lymphocytes is to increase the expression of MHC class II by stimulating CIITA (MHA class II transactivator) through its inducible AIR-1 promoter, which in turn is activated through the JAK/STAT–IRF-1 pathway (Accolla et al 2001). CIITA is disabled through phosphorylation and in monocytes this step has been shown to be a PKA action stimulated by PGE (Li et al 2001). This mechanism explains the previously reported results showing downregulation of HLA by prostaglandins (Helbig et al 1990; Piacibello et al 1986). In summary, it has been hypothesized that the ratio IL-12:PGE secreted from the antigen presenting cells, and particularly dendritic cells, is a determinant of the nature of the subsequently activated T cell (Figure 20.2). It has also been suggested that PGE
PROSTAGLANDINS AND THE IMMUNE RESPONSE
239
Figure 20.1 Influence of PGE on CIITA inactivation through a cAMP-dependent pathway (Li et al 2001)
can affect HLA expression, resulting in alterations to the adaptive immune response.
cAMP levels. Indomethacin can reverse the changes induced in NK cells by PGE.
NATURAL KILLER CELLS
T LYMPHOCYTES
Natural killer (NK) cells are major effector cells of the innate immune system. Their function is regulated by cytokines, particularly IL-2, IL-15 and IL-12 (natural killer cell stimulatory factor). IL-12 produced by the monocyte is stimulatory for NK cells. PGE has been shown to play a part in downregulating the cell-mediated cytotoxicity of NK cells in both humans and rodents (Baxevanis et al 1993; Joshi et al 2001; Linnemeyer and Pollack 1993; Liu et al 2000), possibly via the inhibition of IL-12 by PGE. The main source of the PGE is the monocytes, and the PGE effects are thought to be mediated by changes in cytoplasmic
There is a large amount of evidence that has accumulated on the interaction of prostaglandin, particularly PGE2, and T and B lymphocytes. Many issues are still unresolved and evidence often unclear. Resting and activated peripheral blood T cells and Jurkat T cells possess the cyclooxygenase enzyme systems (Iniguez et al 1999; Pablos et al 1999). The evidence in B cells is less conclusive, where B lymphocyte precursors have not shown evidence of COX 1 or 2 mRNA expression. Whether presence of COX enzyme translates into PG production is still debated. Webb and
240
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 20.2 The dendritic cell is the most prominent antigen-presenting cell but cytokines released by monocytic cells will influence the outcome of T cell/ dendritic cell interaction. Two of the critical cytokines are IL-10 and IL-12 and these are controlled differentially by PGE which raises IL-10 and inhibits IL-12
Nowowiejski (1978) reported release of PGE into culture medium after PHA stimulation, while Pablos et al (1999) could detect no PGE from T cells in vitro. Others have found little or no PG production from normal lymphocytes and suggest the PG that others have detected may be from contaminating cells (Bankhurst et al 1981; Kennedy et al 1980). For the PG to exert a biological effect, it has to act via a receptor, and lymphocytes have been shown to have high affinity binding sites for PG (Goodwin et al 1979). There have been six PGE receptors identified; EP1, EP2, EP3a, EP3b, EPg and EP4. The evidence to date suggests that effects on lymphocytes are predominantly mediated via the EP2 and EP4 receptors, which are both cyclic adenosine monophosphate (cAMP)-dependent (Fedyk and Phipps 1996). Receptor coupling results in stimulation of adenylate cyclase and thus conversion of ATP into cAMP. The actions of PGE are not entirely reliant on this path, as evidence has also shown the calcium-dependent EP1 receptor to play a role in rats, where inhibition of calcium stopped the inhibitory effect of PGE on T cell proliferation. Downstream from these second messengers, PKA and/or tyrosine phosphorylations of substrate proteins are then responsible for the activation of DNA binding sites, e.g. CRE and AP-2 (Micali et al, 1996). PGE2 is a complex immunomodulator. Initially PGE was seen to inhibit T cell proliferation through inhibition of IL-2 (Rappaport and Dodge 1982; Smith et al 1971). However, PGE may play a more pervasive role by directing the cytokine profile of the T cell during development and activation. Such an effect acting predominantly on the cellular immune response (CD4+ T cells) shifts the balance away from a Th1 and towards a Th2 response (Betz and Fox 1991). Naı¨ ve T cells are capable of producing a wide range of cytokines and are known as Th0 cells. During development and subsequent activation, the cytokine response from Th cells and their progeny can be modified to a point where two subgroups can be identified at either end of a spectrum: Th1 and Th2 cells. Th1 lymphocytes produce cytokines such as IL-2 and IFN-g and promote cellular immunity by stimulating cytotoxic and phagocytic functions in effector cells, such as cytotoxic T cells, natural
killer (NK) cells and macrophages. Th2 lymphocytes, on the other hand, secrete cytokines such as IL-4 and IL-10, which support B cell development and production of antibodies (FernandezBotran et al 1988). IFN-g and IL-2 production are inhibited from human, mouse and bovine T lymphocytes (Chouaib et al 1985; Elliott et al 1996; Walker et al 1983) after exposure to PGE. IL-2 receptor expression is also inhibited (Krause and Deutsch 1991), stopping the autocrine action of IL-2. IL-2-dependent proliferation of lymphocytes is consequently impaired. Conversely, IL-10 expression is upregulated in both mice and humans (Ayala et al 1994; Benbernou et al 1997), leading to T cell inhibition. The effect on IL-4, IL-6 and GM-CSF has been less consistent and, depending on the conditions, can be either up- or downregulated by PGE (Benbernou et al 1997; Sottile et al 1996). Thus, PG modulates the balance of cytokines, with a trend towards downregulation of the Th1 cytokines and maintenance or upregulation of the Th2 cytokines (Figure 20.2). As discussed previously, PGE has been shown to inhibit the production of IL-12 (Kelly 1997; Kraan et al 1995), the major cytokine produced from APC. By affecting its expression, PGE may be a determinant of the cytokine profile of the subsequently activated T cell. The effect of PGE in stimulating monocyte IL-10 production (Strassmann et al 1994) will also be a major determinant of the cytokine environment affecting T cell differentiation. Prostaglandins affect expression of co-stimulatory molecules on APC. Anergic T cells arise when presented with antigen in the absence of co-stimulatory signals. These cells then recognize antigen, but do not mount a response. When treating anti-CD3 mAb-stimulated human peripheral blood T lymphocytes with both dexamethasone and PGE2, anergy was induced. This was thought to be via a cAMP-independent mechanism, and could be reversed by the addition of IL-2 (Elliott and Levay 1997). In some circumstances these cells may act as suppressor or regulatory cells, another mechanism whereby PGE may be immunomodulatory (Chouaib et al 1984; Fischer et al 1985; Fulton and Levy 1981). Fischer et al demonstrated a subset of PGE2-responsive human
PROSTAGLANDINS AND THE IMMUNE RESPONSE
241
Figure 20.3 Release of enzymes and natural anti-microbial compounds by the neutrophils represent a major contribution to disease control. The synergistic interaction between chemotactic agents and PGE has a major effect on extravasation of these cells
peripheral blood lymphocytes, which exerted a strong suppressor activity on mitogen, or allogeneic cell-induced lymphocyte proliferation in vitro. Whether these are true suppressor cells or rather regulatory cells is debatable. In the presence of IL-10, antigen-specific anergic cells can be stimulated to proliferate clonally (Groux et al 1996) and are then termed ‘‘regulatory cells’’. PGE inhibits IL-12 formation and the absence of IL-12 is a prerequisite for anergic cell formation (Van Parijs et al 1997). As PGE stimulates IL-10 production and inhibits IL-12, a role for it in affecting the regulatory functions of T cells is likely (Figure 20.2). Data on the effect of prostaglandins on cytotoxic CD8+ T cells is limited. Initially they were thought to be unaffected by PG (Jordan et al 1987), but Ouelette et al (1999) has shown a decrease in CD8+ expression in human peripheral blood lymphocytes after treatment with 10 ng/ml PGE. Apoptosis of both T and B lymphocytes has been shown to be mediated by PGE in some circumstances. Thymic stromal cells express COX-1 and COX-2 and produce prostaglandins. Rocca et al (1999) suggest that the prostaglandin produced may regulate the development of the thymocytes by modulating their interactions with the stromal cells. While PGE2 protects a human thymocyte cell line from apoptosis (Goetzl et al 1995), it has also been reported to initiate apoptosis in neonatal primary thymocytes (Pica et al 1996). Saiagh et al (1994) demonstrate PGE2 having two roles, promoting differentiation of some thymocytes, while increasing or accelerating the apoptosis of differentially marked thymocytes. In B cells evidence shows that an immature B cell lymphoma undergoes apoptosis in response to PGE2, while its mature counterparts survive (Brown and Phipps 1996). B LYMPHOCYTES Early work in this field demonstrated an increase in plaqueforming responses in the presence of PGE (Webb and Osheroff 1976; Zimecki and Webb 1976). Fedyk et al (1996) then published evidence that immature B cell lines were sensitive to the inhibitory
effect of PGE2, while in contrast mature B cell lines lost this sensitivity. The inhibitory effect has been shown to be cAMPdependent (Roper et al 1994), involving EP2 and EP4 receptors (Fedyk and Phipps 1996). This inhibitory effect may be as a result of the effect of PGE2 on co-stimulatory molecules such as CD40 ligand (Splawski et al 1996). Activation without the appropriate co-stimulatory signal could lead to no further proliferation/ activation. PGE2 suppressing the clonal expansion of activated B lymphocytes (Simkin et al 1987) could affect antibody production. Recent work has shown that PGE can exert an effect on the production of IgE, IgG1 and IgM. In mice, LPS and PGE2 have been shown to synergize with IL-4 to stimulate a 26-fold increase in IgE and IgG1, while at the same time inhibiting IgM (Roper et al 1990). This effect occurs in vitro with antigen-specific stimulation as well (Ohmori et al 1990). Evidence has suggested that this class switching is due to an increase in the number of B cells, as opposed to an increase in the number of antibodies from a smaller number of cells. The regulation of the isotype class switch to IgE could be via two possible routes. First, PGE has been shown to directly affect recombination of Ig heavy chain at the level of transcription. Second, it may be an indirect effect as a result of the effects of PGE on cytokine production. The evidence with regard to T lymphocytes demonstrates that PGE2 can enhance a Th2 response with no effect on a Th1 response. This leads to a cytokine milieu in which cytokines such as IL-4 and IL-5 predominate and it is these cytokines which have been shown to have an effect on the immunoglobulin production from B lymphocytes. THE VASCULATURE AND NEUTROPHILS MMPs The monocyte secretes matrix metalloproteinases (MMPs) in order to transverse basement membrane, in addition, MMPs are involved in tissue remodelling, which is necessary in the resolution
242
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
phase of any inflammatory event. Activated human monocytes release MMPs through a cAMP- and PGE-dependent pathway. LPS stimulation induces release of MMP-1, MMP-9 and the inhibitor of matrix metalloproteinase, TIMP-1, by a largely PGEdependent mechanism. In contrast, cytokine stimulation of MMP1 by TNFa, IL-1 and GM-CSF is PGE-dependent, but not the stimulation of MMP-9 and TIMP-1 (Zhang et al 1998). Thus, we have a mechanism whereby PG influences MMPs and allows tissue remodelling at sites of inflammation.
to food antigen. Such a hypothesis is strengthened by reports that NSAIDs break tolerance (Louis et al 1996; Scheuer et al 1997). It is suggested that PGE is generated from the (non-bone marrowderived) stromal cells in the lamina propria, which continuously express COX-2 (Newberry et al 2001). The actions of PGE in stimulating IL-10 and inhibiting IL-12 are very relevant to establishing immune tolerance.
Infection Transendothelial Migration, Monocytes, Neutrophils and T Cells Prostaglandins are released by endothelial cells and pericytes, cells with a smooth muscle character which surround the blood vessels. In this location, PGE and PGI modulate both vessel dilatation and permeability. In the small blood vessels, where PGE is the major prostaglandin, there is a distinct synergy of prostaglandin and chemotactic agents. Recent work has demonstrated that monocytes pretreated with PGE2 show enhanced mobilization to breast and kidney-expressed chemokine (Kurth et al 2001). This synergy has also been clearly shown for neutrophil entry into tissue (Colditz 1990; Foster et al 1989; Rampart and Williams 1987). The effect is paralleled by an increase in oedema in response to agents such as bradykinin (Williams and Morley 1973; Williams 1979). The major effect of neutrophils entering into tissue is likely to be phagocytosis of pathogens and dead cells. But the neutrophil has a variety of granules, which can to some extent be selectively released. This allows the neutrophil to be a source of lytic enzymes such as collagenase and elastase. In addition, natural antimicrobial compounds such as the a-defensins are contained within neutrophil granules and thus the neutrophil can play a major innate immune defence role by releasing compounds that will kill bacteria and viruses (Figure 20.3). At present little is known about the mechanisms active in controlling selective release from the neutrophil. Early work by Weissmann showed that the activation of the neutrophil was suppressed by PGE through a cAMP-dependent process (Weissmann et al 1971; Smolen et al 1982; Weissmann et al 1980). Although initial chemotaxis occurs, PG then inhibits neutrophil adherence to endothelium and neutrophil activation; without PG, an inflammatory response is maintained (Wallace 1993; Wallace and Tigley 1995). Modulation of transendothelial migration of T cells has been investigated as an area where PGE2 may play an antiinflammatory role. Oppenheimer-Marks et al (1994) showed that transendothelial migration of T cells through an endothelial cell barrier was inhibited by PGE2, and this effect could be mimicked by cAMPelevating agents. This effect has been observed in rats where it seemed to be independent of an effect on adhesion molecules (Mesri et al 1996). Conflicting evidence has been published by Hailer et al (2000), who have shown that PGE2 induces P-selectin surface expression on HUVEC and increases T cell adhesion, and therefore possibly enhances the recruitment of inflammatory cells.
Agents that infect monocytes may partially thwart an immune response that could be detrimental to their survival. Several infecting organisms raise the production of PGE by the host monocyte; examples include Mycobacterium avium (Venkataprasad et al 1990), leprosy (Misra et al 1995), cytomegalovirus (CMV; Nokta et al 1996) and human immunodeficiency virus (HIV; Longo et al 1993). In HIV infection, PGE is raised even in nonsymptomatic patients and the viral component GP120 is thought to be responsible for this (Denis 1994). An elevated PGE will ensure a reduction in Th1 type responses due to the decrease in IL-12 and the increase in IL-10, and this may contribute to the increased survival of these intracellular organisms. In the case of Epstein–Barr virus (EBV), where the viral genome encodes for a variant of the suppressive cytokine IL-10, PGE synthesis is inhibited by the organism (Savard et al 2000). The resulting imbalance may play a part in allowing unrestricted expansion of B cell populations. The effects of PGE may be beneficial or detrimental, depending on the circumstances. In septic shock, high levels of PGE are thought to be an attempt to inhibit the widespread, detrimental inflammatory effect but may lead to immunosuppression (Junger et al 1996). Conversely, in atopic dermatitis high levels of PGE2 result in inhibition of Th1 cells, while causing predominance of Th2 and cytokines, inducing IgE production, which is at the core of the pathology of the disease (Chan et al 1996). Similarly, patients with hyper-E syndrome have extremely high serum levels of IgE and monocytes that produce abnormally high levels of PGE2 (Leung et al 1988). A widely used treatment for inflammatory conditions is the use of NSAIDs. These target the COX enzymes, thereby decreasing prostaglandins and leukotrienes. Given the myriad of damping effects of prostaglandins on the main cells of the immune system, it should be no surprise that these non-specific NSAIDs have unwanted additional effects, particularly chronic gastric inflammation and possible subsequent ulceration. More selective COX inhibitors are now in use to attempt to decrease these side-effects.
SUMMARY In general, the effects of PGE on immune cells are to downregulate the cell-mediated immune response. This applies to antigen presentation, T cell priming, class switching, and NK cell cytotoxicity. By understanding the mechanisms underlying these responses, we can understand the disease states observed and target therapies more successfully.
CLINICAL CONSIDERATIONS AND APPLICATIONS Tolerance The prime example of tolerance to antigens is oral tolerance, a phenomenon that is better established in animals than in man (Jones et al 2001). There has been good reason to suppose that the presence of PGE in the gut may play a part in inducing tolerance
REFERENCES Accolla RS, De Lerma Barbaro A, Mazza S et al (2001) The MHC class II transactivator: prey and hunter in infectious diseases. Trends Immunol, 22, 560–563. Ayala A, Lehman DL, Herdon CD and Chaudry IH (1994) Mechanism of enhanced susceptibility to sepsis following hemorrhage. Interleukin-10
PROSTAGLANDINS AND THE IMMUNE RESPONSE suppression of T-cell response is mediated by eicosanoid-induced interleukin-4 release. Arch Surg, 129, 1172–1178. Bankhurst AD, Hastain E, Goodwin JS and Peake GT (1981) The nature of the prostaglandin-producing mononuclear cell in human peripheral blood. J Lab Clin Med, 97, 179–186. Bartik MM, Bauman GP, Brooks WH and Roszman TL (1994) Costimulatory signals modulate the antiproliferative effects of agents that elevate cAMP in T cells. Cell Immunol, 158, 116–130. Baxevanis CN, Reclos GJ, Gritzapis AD et al (1993) Elevated prostaglandin E2 production by monocytes is responsible for the depressed levels of natural killer and lymphokine-activated killer cell function in patients with breast cancer. Cancer, 72, 491–501. Benbernou N, Esnault S, Shin HC et al (1997) Differential regulation of IFN-g, IL-10 and inducible nitric oxide synthase in human T cells by cyclic AMP-dependent signal transduction pathway. Immunology, 91, 361–368. Betz M and Fox BS (1991) Prostaglandin E2 inhibits production of Th1 lymphokines but not of Th2 lymphokines. J Immunol, 146, 108–113. Brock TG, McNish RW and Peters-Golden M (1999) Arachidonic acid is preferentially metabolized by cyclooxygenase-2 to prostacyclin and prostaglandin E2. J Biol Chem, 274, 11660–11666. Brown DM and Phipps RP (1996) Bcl-2 expression inhibits prostaglandin E2-mediated apoptosis in B cell lymphomas. J Immunol, 157, 1359–1370. Caivano M, Gorgoni B, Cohen P and Poli V (2001) The induction of cyclooxygenase-2 mRNA in macrophages is biphasic and requires both C/EBP(b) and C/EBP (D) transcription factors. J Biol Chem, 276, 48693–48701. Chan S, Henderson WR, Li SH and Hanifin JM (1996) Prostaglandin E(2) control of T-cell cytokine production is functionally related to reduced lymphocyte proliferation in atopic dermatitis. J Allerg Clin Immunol, 97, 85–94. Chouaib S, Chatenoud L, Klatzmann D and Fradelizi D (1984) The mechanisms of inhibition of human IL-2 production. II. PGE2 induction of suppressor T lymphocytes. J Immunol, 132, 1851–1857. Chouaib S, Welte K, Mertelsmann R and Dupont B (1985) Prostaglandin E2 acts at two distinct pathways of T lymphocyte activation: inhibition of interleukin 2 production and downregulation of transferrin receptor expression. J Immunol, 135, 1172–1179. Chow JC, Young DW, Golenbock DT et al (1999) Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J Biol Chem, 276, 23179–23185. Cok SJ and Morrison AR (2001) The 30 -untranslated region of murine cyclooxygenase-2 contains multiple regulatory elements that alter message stability and translational efficiency. J Biol Chem, 276, 23179– 23185. Colditz IG (1990) Effect of exogenous prostaglandin E2 and actinomycin D on plasma leakage induced by neutrophil activating peptide-1/ interleukin-8. Immunol Cell Biol, 68, 397–403. Denis M (1994) Envelope glycoprotein (Gp120) from HIV-1 enhances Mycobacterium avium growth in human bronchoalveolar macrophages—an effect mediated by enhanced prostaglandin synthesis. Clin Exp Immunol, 98, 123–127. Elliott LH, Levay AK, Sparks B et al (1996) Dexamethasone and prostaglandin E2 modulate T-cell receptor signaling through a cAMPindependent mechanism. Cell Immunol, 169, 117–124. Elliott LH and Levay AK (1997) Co-stimulation with dexamethasone and prostaglandin E2: a novel paradigm for the induction of T-cell anergy. Cell Immunol, 180, 124–131. Fedyk ER and Phipps RP (1996) Prostaglandin E(2) receptors of the Ep(2) and Ep(4) subtypes regulate activation and differentiation of mouse B lymphocytes to IGE-secreting cells. Proc Nat Acad Sci USA, 93, 10978–10983. Fedyk ER, Ripper JM, Brown DM and Phipps RP (1996) A molecular analysis of PGE receptor (EP) expression on normal and transformed B lymphocytes: co-expression of EP1, EP2, EP3b and EP4. Mol Immunol, 33, 33–45. Fernandez-Botran R, Sanders VM, Mosmann TR and Vitetta ES (1988) Lymphokine-mediated regulation of the proliferative response of clones of T helper 1 and T helper 2 cells. J Exp Med, 168, 543–558. Fischer A, Le Deist F, Durandy A and Griscelli C (1985) Separation of a population of human T lymphocytes that bind prostaglandin E2 and exert a suppressor activity. J Immunol, 134, 815–819.
243
Foster SJ, Aked DM, Schroder JM and Christophers E (1989) Acute inflammatory effects of a monocyte-derived neutrophil activating peptide in rabbit skin. Immunology, 67, 181–183. Frenkian M, Pidoux E, Baudoin C et al (2001) Indomethacin increases 15PGDH mRNA expression in HL60 cells differentiated by PMA. Prostaglandin Leukotriene Essent Fatty Acids, 64, 87–93. Fulton AM and Levy JG (1981) The induction of non-specific T suppressor lymphocytes by prostaglandin E1. Cell Immunol, 59, 54–60. Goetzl EJ, An SZ and Zeng L (1995) Specific suppression by prostaglandin-E(2) of activation-induced apoptosis of human Cd4(+)Cd8(+) T-lymphoblasts. J Immunol, 154, 1041–1047. Goodwin JS, Wiik A, Lewis M et al (1979) High-affinity binding sites for prostaglandin E on human lymphocytes. Cell Immunol, 43, 150–159. Greenland KJ, Jantke I, Jenatschke S et al (2000) The human NAD+dependent 15-hydroxyprostaglandin dehydrogenase gene promoter is controlled by ETS and activating protein-1 transcription factors and progesterone. Endocrinology, 141, 581–597. Groux H, Bigler M, de Vries JE and Roncarolo MG (1996) Interleukin-10 induces a long-term antigen-specific anergic state in human CD4+ T cells. J Exp Med, 184, 19–29. Hailer NP, Oppermann E, Leckel K et al (2000) Prostaglandin E2 induces expression of P-selectin (CD62P) on cultured human umbilical vein endothelial cells and enhances endothelial binding of CD47 T-cells. Transplantation, 70, 236–240. Harizi H, Juzan M, Grosset C et al (2001) Dendritic cells issued in vitro from bone marrow produce PGE(2) that contributes to the immunomodulation induced by antigen-presenting cells. Cell Immunol, 209, 19–28. Helbig H, Gurley RC, Reichl RJ et al (1990) Induction of MHC class II antigen in cultured bovine ciliary epithelial cells. Graefes Arch Clin Exp Ophthalmol, 228, 556–561. Hinz B, Brune K and Pahl A (2000) Cyclooxygenase-2 expression in lipopolysaccharide-stimulated human monocytes is modulated by cyclic AMP, prostaglandin E(2), and non-steroidal antiinflammatory drugs. Biochem Biophys Res Commun, 278, 790–796. Humes JL, Bonney RW, Pelus L et al (1977) Macrophages synthesize and release prostaglandins in response to inflammatory stimuli. Nature, 269, 149–151. Hwang D, Jang BC, Yu G and Boudreau M (1997) Expression of mitogeninducible cyclooxygenase induced by lipopolysaccharide: mediation through both mitogen-activated protein kinase and NF-kB signaling pathways in macrophages. Biochem Pharmacol, 54, 87–96. Iniguez MA, Punzon C and Fresno M (1999) Induction by cyclooxygenase-2 on activated T lymphocytes: regulation of T cell activation by cyclooxygenase-2 inhibitors. J Immunol, 163, 111–119. Jones CA, Vance GH, Power LL et al (2001) Co-stimulatory molecules in the developing human gastrointestinal tract: a pathway for fetal allergen priming. J Allergy Clin Immunol, 108, 235–241. Jonuleit H, Kuhn U, Muller G et al (1997) Pro-inflammatory cytokines and prostaglandins induce maturation of potent immunostimulatory dendritic cells under fetal calf serum-free conditions. Eur J Immunol, 27, 3135–3142. Jordan ML, Hoffman RA, Debe EF et al (1987) Prostaglandin E2 mediates subset-specific effects on the functional responses of allosensitized T lymphocyte clones. Transplantation, 43, 117–123. Joshi PC, Zhou X, Cuchens M and Jones Q (2001) Prostaglandin E2 suppressed IL-15-mediated human NK cell function through downregulation of common g-chain. J Immunol, 166, 885–891. Junger WG, Hoyt DB, Liu FC et al (1996) Immunosuppression after endotoxin-shock—the result of multiple antiinflammatory factors. J Trauma Inj Infect Crit Care, 40, 702–709. Kaisho T and Akira S (2000) Critical roles of Toll-like receptors in host defense. Crit Rev Immunol, 20, 393–405. Kalinski P, Schuitemaker J and Kapsenberg ML (1998) Reciprocal regulation of IL-12 p40 vs. IL-12 p70 production in dendritic cells. Regulation of IL-12 subunits by PGE(2) and IL-4. J Leuk Biol, H2. Kalinski P, Hilkens CM, Wierenga EA and Kapsenberg ML (1999) T-cell priming by type-1 and type-2 polarized dendritic cells: the concept of a third signal. Immunol Today, 20, 561–567. Kalinski P, Vieira PL, Schuitemaker JH et al (2001) Prostaglandin E(2) is a selective inducer of interleukin-12 p40 (IL-12p40) production and an inhibitor of bioactive IL-12p70 heterodimer. Blood, 97, 3466–3469. Kambayashi T, Jacob CO, Zhou D et al (1995) Cyclic-nucleotide phosphodiesterase type-IV participates in the regulation of IL-10 and
244
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
in the subsequent inhibition of TNF-a and IL-6 release by endotoxinstimulated macrophages. J Immunol, 155, 4909–4916. Kelly RW (1997) Prostaglandins in primate semen: biasing the immune system to benefit spermatozoa and virus. Prostagland Leukotrienes Essent Fatty Acids, 57, 113–118. Kennedy MS, Stobo JD and Goldyne ME (1980) In vitro synthesis of prostaglandins and related lipids by populations of human peripheral blood mononuclear cells. Prostaglandins, 20, 135–145. Kraan TCTMV, Boeije LCM, Smeenk RJT et al (1995) Prostaglandin-E2 is a potent inhibitor of human interleukin-12 production. J Exp Med, 181, 775–779. Krause DS and Deutsch C (1991) Cyclic AMP directly inhibits IL-2 receptor expression in human T cells: expression of both p55 and p75 subunits is affected. J Immunol, 146, 2285–2296. Kurth I, Willimann K, Schaerli P et al (2001) Monocyte selectivity and tissue localization suggests a role for breast and kidney-expressed chemokine (brak) in macrophage development. J Exp Med, 194, 855–862. Leung DY, Key L, Steinberg JJ et al (1988) Increased in vitro bone resorption by monocytes in the hyper-immunoglobulin E syndrome. J Immunol, 140, 84–88. Li G, Harton JA, Zhu X and Ting JP (2001) Downregulation of CIITA function by protein kinase a (PKA)-mediated phosphorylation: mechanism of prostaglandin E, cyclic AMP, and PKA inhibition of class II major histocompatibility complex expression in monocytic lines. Mol Cell Biol, 21, 4626–4635. Linnemeyer PA and Pollack SB (1993) Prostaglandin E2-induced changes in the phenotype, morphology, and lytic activity of IL-2-activated natural killer cells. J Immunol, 150, 3747–3754. Liu H, Cunnick JE and Hendrich S (2000) Opposing effects of prostaglandin E(2) and F(2a) on rat liver—associated natural killer cell activity in vitro. Prostagland Leukotriene Essent Fatty Acids, 63, 153–158. Longo N, Zabay JM, Sempere JM et al (1993) Altered production of PGe(2), IL-1b and TNFa by peripheral-blood monocytes from HIVpositive individuals at early stages of HIV-infection. J Acqu Immune Defic Syndr, 6, 1017–1023. Louis E, Franchimont D, Deprez M et al (1996) Decrease in systemic tolerance to fed ovalbumin in indomethacin-treated mice. Int Arch Allerg Immunol, 109, 21–26. Mack Strong VE, Mackrell PJ, Concannon EM et al (2001) NS-398 treatment after trauma modifies NF-kB activation and improves survival. J Surg Res, 98, 40–46. Matsumoto H, Naraba H, Murakami M et al (1997) Concordant induction of prostaglandin E2 synthase with cyclooxygenase-2 leads to preferred production of prostaglandin E2 over thromboxane and prostaglandin D2 in lipopolysaccharide-stimulated rat peritoneal macrophages. Biochem Biophys Res Commun, 230, 110–114. MenetrierCaux C, Bain C, Favrot MC et al (1999) Renal cell carcinoma induces interleukin 10 and prostaglandin E-2 production by monocytes. Br J Cancer, 79, 119–130. Mertz PM, DeWitt DL, Stetler Stevenson WG and Wahl LM (1994) Interleukin 10 suppression of monocyte prostaglandin H synthase-2. Mechanism of inhibition of prostaglandin-dependent matrix metalloproteinase production. J Biol Chem, 269, 21322–21329. Mertz PM, Corcoran ML, McCluskey KM et al (1996) Suppression of prostaglandin-H synthase-2 induction in human monocytes by in vitro or in vivo administration of interleukin-4. Cell Immunol, 173, 252–260. Mesri M, Liversidge J and Forrester JV (1996) Prostaglandin E2 and monoclonal antibody to lymphocyte function-associated antigen-1 differentially inhibit migration of T lymphocytes across microvascular retinal endothelial cells in rat. Immunology, 88, 471–477. Micali A, Medici N, Sottile A et al (1996) Prostaglandin E2 induction of binding activity to CRE and AP-2 elements in human T lymphocytes. Cell Immunol, 174, 99–105. Misra N, Selvakumar M, Singh S et al (1995) Monocyte-derived IL-10 and PGE(2) are associated with the absence of Th1 cells and in vitro T-cell suppression in lepromatous leprosy. Immunol Lett, 48, 123–128. Moller DR, Wysocka M, Greenlee BM et al (1997) Inhibition of human interleukin-12 production by pentoxifylline. Immunology, 91, 197–203. Mori K, Tanaka I, Kotani M et al (1996) Gene expression of the human prostaglandin E receptor EP4 subtype: differential regulation in monocytoid and lymphoid lineage cells by phorbol ester. J Mol Med, 74, 333–336.
Newberry RD, McDonough JS, Stenson WF and Lorenz RG (2001) Spontaneous and continuous cyclooxygenase-2-dependent prostaglandin e(2) production by stromal cells in the murine small intestine lamina propria: directing the tone of the intestinal immune response. J Immunol, 166, 4465–4472. Nichols FC and Garrison SW (1987) Interferon-g-potentiation of lipopolysaccharide-induced eicosanoid release from human monocytes. J Interferon Res, 7, 121–129. Nokta MA, Hassan MI, Loesch K and Pollard RB (1996) Human cytomegalovirus-induced immunosuppression—relationship to tumor necrosis factor-dependent release of arachidonic acid and prostaglandin E(2) in human monocytes. J Clin Invest, 97, 2635–2641. Ohmori H, Hikida M and Takai T (1990) Prostaglandin E2 as a selective stimulator of antigen-specific IgE response in murine lymphocytes. Eur J Immunol, 20, 2499–2503. Oppenheimer-Marks N, Kavanaugh AF and Lipsky PE (1994) Inhibition of the transendothelial migration of human T lymphocytes by prostaglandin E2. J Immunol, 152, 5703–5713. Ouellette MJ, St-Jacques S and Lambert RD (1999) Cd8 membrane expression is downregulated by transforming growth factor (TGF)-b1, TGF-b2, and prostaglandin E2. Am J Reprod Immunol, 41, 183–191. Pablos JL, Santiago B, Carreira PE et al (1999) Cyclooxygenase-1 and -2 are expressed by human T cells. Clin Exp Immunol, 115, 86–90. Piacibello W, Rubin BY and Broxmeyer HE (1986) Prostaglandin E counteracts the g-interferon induction of major histocompatibility complex class II antigens on U937 cells and induction of responsiveness of U937 colony-forming cells to suppression by lactoferrin, transferrin, acidic isoferritins, and prostaglandin E. Exp Hematol, 14, 44–50. Pica F, Franzese O, D’Onofrio C et al (1996) Prostaglandin E2 induces apoptosis in resting immature and mature human lymphocytes: a cMyc-dependent and Bel-2-independent associated pathway. J Pharmacol Exp Ther, 277, 1793–1800. Pichaud F, Roux S, Frendo JL et al (1997) 1,25-Dihydroxyvitamin D-3 induces NAD(+)-dependent 15-hydroxyprostaglandin dehydrogenase in human neonatal monocytes. Blood, 89, 2105–2112. Porreca E, Reale M, Difebbo C et al (1996) Downregulation of cyclooxygenase-2 (cox-2) by interleukin-1 receptor antagonist in human monocytes. Immunology, 89, 424–429. Rampart M and Williams TJ (1987) The effect of PGE2 on polymorphonuclear leukocyte accumulation in rabbit skin. Br J Pharmacol, 90, 56P. Rappaport RS and Dodge GR (1982) Prostaglandin E inhibits the production of human interleukin 2. J Exp Med, 155, 943–948. Reale M, Barbacane RC, Frydas S et al (1996) Human recombinant interleukin-1b induces thromboxane A2 release in polymorphonuclear leukocytes, macrophages and platelets: effect of IL-1 receptor antagonist. Mol Cell Biochem, 159, 163–168. Remick DG, Larrick JW, Nguyen DT and Kunkel SL (1987) Stimulation of prostaglandin E2 and thromboxane B2 production by human monocytes in response to interleukin-2. Biochem Biophys Res Commun, 147, 86–93. Rhee SH and Hwang D (2000) Murine TOLL-like receptor 4 confers lipopolysaccharide responsiveness as determined by activation of NFkB and expression of the inducible cyclooxygenase. J Biol Chem, 275, 34035–34040. Rocca B, Spain LM, Ciabattoni G et al (1999) Differential expression and regulation of cyclooxygenase isozymes in thymic stromal cells. J Immunol, 162, 4589–4597. Rock FL, Hardiman G, Timans JC et al (1998) A family of human receptors structurally related to Drosophila Toll. Proc Natl Acad Sci USA, 95, 588–593. Roper RL, Conrad DH, Brown DM et al (1990) Prostaglandin E2 promotes IL-4-induced IgE and IgG1 synthesis. J Immunol, 145, 2644–2651. Roper RL, Ludlow JW and Phipps RP (1994) Prostaglandin E2 inhibits B lymphocyte activation by a cAMP-dependent mechanism: PGEinducible regulatory proteins. Cell Immunol, 154, 296–308. Saha B, Das G, Vohra H et al (1994) Macrophage–T cell interaction in experimental mycobacterial infection. Selective regulation of costimulatory molecules on Mycobacterium-infected macrophages and its implication in the suppression of cell-mediated immune response. Eur J Immunol, 24, 2618–2624.
PROSTAGLANDINS AND THE IMMUNE RESPONSE Saiagh S, Rigal D and Monier JC (1994) Effects of PGE2 upon differentiation and programmed cell death of suspension cultured CD4–Cd8-thymocytes. Int J Immunopharmacol, 16, 775–786. Savard M, Belanger C, Tremblay MJ et al (2000) EBV suppresses prostaglandin E2 biosynthesis in human monocytes. J Immunol, 164, 6467–6473. Scheuer WV, Hobbs MV and Weigle WO (1987) Interference with tolerance induction in vivo by inhibitors of prostaglandin synthesis. Cell Immunol, 104, 409–418. Shankavaram UT, DeWitt DL and Wahl LM (1998) Lipopolysaccharide induction of monocyte matrix metalloproteinases is regulated by the tyrosine phosphorylation of cytosolic phospholipase A2. J Leukoc Biol, 64, 221–227. Simkin NJ, Jelinek DF and Lipsky PE (1987) Inhibition of human B cell responsiveness by prostaglandin E2. J Immunol, 138, 1074–1081. Smith JW, Steiner AL and Parker CW (1971) Human lymphocytic metabolism. Effects of cyclic and non-cyclic nucleotides on stimulation by phytohemagglutinin. J Clin Invest, 50, 442–448. Smolen JE, Korchak HM and Weissmann G (1982) Stimulus-response coupling in neutrophils. Trends Pharmacol Sci, 3, 483–485. Sottile A, Venza I, Venza N et al (1996) PGE(2)-induced immunoregulation mediated by cytokine production from cultures of human peripheral T-lymphocytes. Immunopharmacol Immunotoxicol, 18, 27–36. Souness JE, Griffin M, Maslen C et al (1996) Evidence that cyclic-AMP phosphodiesterase inhibitors suppress TNF-a generation from human monocytes by interacting with a low-affinity phosphodiesterase-4 conformer. Br J Pharmacol, 118, 649–658. Splawski JB, Nishioka J, Nishioka Y and Lipsky PE (1996) CD40 ligand is expressed and functional on activated neonatal T-cells. J Immunol, 156, 119–127. Steinbrink K, Paragnik L, Jonuleit H et al (2000) Induction of dendritic cell maturation and modulation of dendritic cell-induced immune responses by prostaglandins. Arch Dermatol Res, 292, 437–445. Strassmann G, Paul Koota V, Finkelman F et al (1994) Evidence for the involvement of interleukin 10 in the differential deactivation of murine peritoneal macrophages by prostaglandin E2. J Exp Med, 180, 2365– 2370. Takizawa T, Ikeda Y, Togashi H et al (1996) Inhibitory effect of indomethacin on neonatal lung catabolism of prostaglandin E2: possible mechanism of the re-opening of the ductus arteriosus after indomethacin therapy. J Toxicol Sci, 21, 243–248. Van Parijs L, Perez VL, Biuckians A et al (1997) Role of interleukin 12 and co-stimulators in T cell anergy in vivo. J Exp Med, 186, 1119–1128. Venkataprasad N, Shiratsuchi H, Johnson JL and Ellner JJ (1996) Induction of prostaglandin E(2) by human monocytes infected with Mycobacterium avium complex—modulation of cytokine expression. J Infect Dis, 174, 806–811. Verghese MW, McConnell RT, Strickland AB et al (1995) Differential regulation of human monocyte-derived TNF-a and IL-b-IVphosphodiesterase (cAMP-PDE) inhibitors. J Pharmacol Exp Therapeut, 272, 1313–1320.
245
Vieira PL, de Jong EC, Wierenga EM et al (2000) Development of Th1inducing capacity in myeloid dendritic cells requires environmental instruction. J Immunol, 164, 4507–4512. Wadleigh DJ, Reddy ST, Kopp E et al (2000) Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. J Biol Chem, 275, 6259–6266. Wagner RS, Halushka PV and Cook JA (1996) Activation of thromboxane A2 receptors alters lipopolysaccharide-induced adherence of THP-1 cells. Shock, 5, 41–46. Walker C, Kristensen F, Bettens F and de Weck AL (1983) Lymphokine regulation of activated (GI) lymphocytes. I. Prostaglandin E2-induced inhibition of interleukin 2 production. J Immunol, 130, 1770–1773. Wallace JL (1993) Gastric ulceration: critical events at the neutrophil– endothelium interface. Can J Physiol Pharmacol, 71, 98–102. Wallace JL and Tigley AW (1995) Review article: new insights into prostaglandins and mucosal defence. Aliment Pharmacol Ther, 9, 227– 235. Webb DR and Osheroff PL (1976) Antigen stimulation of prostaglandin synthesis and control of immune responses. Proc Natl Acad Sci USA, 73, 1300–1304. Webb DR and Nowowiejski I (1978) Mitogen-induced changes in lymphocyte prostaglandin levels: a signal for the induction of suppressor cell activity. Cell Immunol, 41, 72–85. Weissmann G, Zurier RB, Spieler PJ and Goldstein IM (1971) Mechanisms of lysosomal enzyme release from leukocytes exposed to immune complexes and other particles. J Exp Med, 134(suppl), 149s. Weissmann G, Smolen JE and Kaorchak H (1980) Prostaglandins and inflammation: receptor/cyclase coupling as an explanation of why PGEs and PGI2 inhibit functions of inflammatory cells. In Samuelsson B, Ramwell PW, Paoletti R (eds), Prostaglandins and Thromboxane Research, vol 8. Raven, New York, 1637–1653. Whittaker DS, Bahjat KS, Moldawer LL and Clare-Salzler MJ (2000) Autoregulation of human monocyte-derived dendritic cell maturation and IL-2 production by cyclooxygenase-2-mediated prostanoid production. J Immunol, 165, 4298–4304. Whittaker DS, Bahjat KS, Moldawer LL and Clare-Salzler MJ (2000) Autoregulation of human monocyte-derived dendritic cell maturation and IL-12 production by cyclooxygenase-2-mediated prostanoid production. J Immunol, 165, 4298–4304. Widomski DL, Walsh RE, Baron DA et al (1991) Effects of the prostaglandin analogue misoprostol on inflammatory mediator release by human monocytes. Agents Act, 34, 30–31. Williams TJ and Morley J (1973) Prostaglandins as potentiators of increased vascular permeability in inflammation. Nature, 246, 215–217. Williams TJ (1979) Prostaglandin E2, prostaglandin I2 and the vascular changes of inflammation. Br J Pharmacol, 65, 517–524. Zhang Y, McCluskey K, Fujii K and Wahl LM (1998) Differential regulation of monocyte matrix metalloproteinase and TIMP-1 production by TNF-a, granulocyte–macrophage CSF, and IL-1b through prostaglandin-dependent and -independent mechanisms. J Immunol, 161, 3071–3076. Zimecki M and Webb DR (1976) The regulation of the immune response to T-independent antigens by prostaglandins and B cells. J Immunol, 117, 2158–2164.
21 Leukotrienes in Aspirin-intolerant Asthma Anthony P. Sampson and Stephen T. Holgate Southampton University School of Medicine, UK
ASPIRIN-INTOLERANT ASTHMA Aspirin and other non-steroidal antiinflammatory drugs (NSAIDs) are extensively used as mild analgesics and antiinflammatory agents, and aspirin has an additional clinical role in preventing thrombosis. NSAIDs are associated with a number of adverse reactions, including prolonged bleeding, nephritis and gastric ulceration, and it has long been recognized that in some patients with asthma NSAIDs may also precipitate acute exacerbations (Cooke 1991; Smith 1977) that can be lifethreatening (Picado et al 1989). This chapter will focus on evidence from the last 10–15 years that indicates a central role for eicosanoids in these exacerbations, particularly the cysteinylleukotrienes (LTC4, LTD4, LTE4), and on the cellular, biochemical, and genetic mechanisms that may underlie this distinct asthma phenotype.
Clinical Features The classic aspirin-intolerance syndrome has been described as a triad of rhinosinusitis (often with nasal polyps), asthma and aspirin sensitivity (Samter and Beers 1968). Clinically, the aspirintolerant asthma (AIA) syndrome most often emerges in the third or four decades of life as rhinosinusitis resembling a persistent viral infection (Stevenson and Simon 1993; Szczeklik et al 2000). Patients report nasal congestion, anosmia and rhinorrhoea, and many develop nasal polyps, often with secondary infection of the paranasal sinuses. Bronchoconstriction and airway inflammation usually emerge later, and this becomes severe, chronic, perennial asthma, with about 75% of AIA patients needing oral or inhaled corticosteroids to maintain control of their symptoms. Acute respiratory reactions to NSAIDs may be accompanied by rhinoconjunctival symptoms and by dermal symptoms such as facial flushing and exacerbation of pre-existing urticaria. In some NSAID-sensitive patients, acute bronchoconstriction may occur as part of an anaphylactoid reaction to a specific NSAID, while other NSAIDs are tolerated. These patients appear normal, without underlying asthma, when not exposed to the NSAID to which they are sensitive. Such idiosyncratic reactions may be related to IgE-mediated allergy, although NSAID-specific IgE has not been clearly demonstrated. In contrast, patients with aspirin-intolerant asthma (AIA) show cross-reactivity to a wide range of NSAID compounds with widely varying molecular structures. AIA patients do not have positive skin test reactions to NSAIDs nor NSAID-specific IgE (Weltman et al 1978; Slepian et al 1985). They typically have persistent severe asthma even in the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
avoidance of NSAIDs for many years, suggesting that AIA is a specific inflammatory syndrome which is exacerbated, but not initiated, by the common pharmacological mechanism of NSAIDs. Aspirin Challenge Tests Diagnosis of aspirin sensitivity is based upon NSAID challenge under controlled conditions. Lung function is monitored while patients ingest incremental oral doses of aspirin, or inhaled doses of lysine–aspirin conjugate or sulpyrine. Use of b2-agonists, cromones and inhaled corticosteroids may mask responses to NSAID challenge, leading to a high rate of false-negative results (Szczeklik and Serwonska 1979; Stevenson 1988; Nizankowska and Szczeklik 1989; Phillips et al 1989). In a typical 3-day oral challenge protocol (Stevenson and Simon 1993), incremental doses of aspirin are given at 3-hourly intervals up to a maximum of 650 mg. The challenge is terminated when forced expiratory volume in 1 s (FEV1) falls by at least 20%. Reactions occur around 50 min after oral aspirin ingestion, ranging from 20 to 120 min (McDonald et al 1972). A shorter protocol involves the inhalation of incremental aerosolized doses of lysine–aspirin, which is soluble and non-irritant to the bronchial mucosa (Bianco et al 1977). Respiratory reactions to inhaled lysine–aspirin often occur within 1 min, so with dosing intervals of 30–60 min the entire challenge procedure can be completed within 1 day (Bianco et al 1981; Phillips et al 1989). Inhaled lysine–aspirin challenges are safer than oral challenges, as the reactions are not systemic but localized to the airways, and are more easily reversible with inhaled b2-agonists. The sensitivity of inhaled lysine–aspirin challenge is similar to oral aspirin challenge (Dahle´n and Zetterstrom 1990). Prevalence of AIA Aspirin intolerance is generally underdiagnosed. Reported prevalence is 3–5% in adult asthmatics based on patient history alone (Giraldo et al 1969), but this rises to 19% when consecutive adult asthmatics are challenged with oral aspirin (Spector et al 1979). Similarly, aspirin sensitivity in asthmatic children has a prevalence of less than 2% based on history alone, but this rises to 13–16% when aspirin-challenged (Vedanthan et al 1977; Stevenson and Simon 1993). In the European Network on Aspirin-induced Asthma (AIANE) study, patients with full-blown AIA were most likely to be female and non-atopic, with onset occurring typically
248
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
in early middle age (Szczeklik et al 2000). Aspirin sensitivity is over-represented in the severe asthmatic population. Among patients who have experienced a near-fatal acute asthma exacerbation requiring treatment in the intensive care unit, 24% are aspirin-sensitive based on history alone (Marquette et al 1992). The true proportion based on aspirin challenge is likely to be higher. Many life-threatening acute exacerbations in AIA patients are due to inadvertent use of NSAIDs, but over 40% cannot be attributed to NSAID ingestion (Picado et al 1989), illustrating the severity of the underlying chronic asthma in these patients. PATHOGENESIS OF AIA It is likely that genetic and environmental factors contribute to aspirin sensitivity. A significantly enhanced frequency of the human leukocyte antigen HLA-DQw2 has been reported in AIA patients compared to aspirin-tolerant asthmatics (ATA) (Mullarkey et al 1986). Other studies suggest associations between HLAA74 and nasal polyposis but not aspirin intolerance (Luxenberger et al 2000). An altered distribution of IgG subclasses in AIA patients (Szczeklik et al 1992) has suggested that AIA may be caused by chronic viral infection (Szczeklik 1988). Anomalies in platelet function have also been reported in AIA (Ameisen et al 1985). Platelets from AIA patients, but not those from ATA patients, produce cytotoxic activity and a burst of chemiluminescence when challenged in vitro with NSAIDs. However, other workers report no evidence of abnormal chemiluminescence in AIA platelets, neither does aspirin activate AIA platelets as judged by release of b-thromboglobulin (Williams et al 1990). The Cyclooxygenase Theory of AIA The most successful model to explain the pharmacological and clinical features of AIA is the cyclooxygenase theory (Szczeklik 1990), which suggests that acute respiratory reactions to NSAIDs are related directly to their pharmacological activity in inhibiting cyclooxygenase (COX). Intolerance to an individual NSAID can be predicted by its potency in inhibiting COX in vitro, with strong inhibitors (including aspirin, indomethacin, mefenamic acid, ibuprofen, and piroxicam) being common precipitants of adverse reactions, while weak inhibitors (such as sodium salicylate) precipitate reactions rarely or only at high doses (Szczeklik et al 1975). Analgesic and antiinflammatory drugs that are not COX inhibitors, such as opiates and glucocorticosteroids, do not precipitate adverse reactions in AIA patients more often than in aspirin-tolerant patients, and desensitization to aspirin does not cross-desensitize AIA patients to corticosteroids (Szczeklik et al 1985; Feigenbaum et al 1995). In contrast, desensitization of an AIA patient to one NSAID confers protection against reactions to other NSAIDs, even if the drugs belong to diverse structural groups. The COX theory thus predicts that abnormalities in prostanoid metabolism underlie adverse responses to NSAIDs. Cysteinyl-leukotrienes The structural elucidation of cysteinyl-leukotrienes (cys-LTs) in the late 1970s (Murphy et al 1979) and characterization of their bronchoconstrictor and pro-inflammatory role in asthma added an extra component to the COX theory. Like the prostanoids generated by COX, leukotrienes are products of arachidonic acid, but are derived by the interaction of 5-lipoxygenase (5-LO) and its activating protein (FLAP). The unstable intermediate LTA4 can be converted by LTA4 hydrolase to the neutrophil chemotaxin
LTB4, or by LTC4 synthase to the first of the cys-LTs, LTC4 (Yoshimoto et al 1988; Lam et al 1994). The cys-LTs LTC4, LTD4, and LTE4 cause long-lasting bronchoconstriction, mucus hypersecretion and airway oedema (Arm and Lee 1993), and more recently they have been recognized as potent and specific chemotaxins for human eosinophils both in vitro (Spada et al 1994) and in vivo (Laitinen et al 1993; Diamant et al 1997). CysLT levels are elevated in the bronchoalveolar lavage fluid (BALF), plasma, and urine following allergen challenge and after acute asthma exacerbations in adults and children (Taylor et al 1989; Wenzel et al 1990; Sampson et al 1995). Studies with leukotriene synthesis inhibitors (e.g. zileuton) and cys-LT receptor antagonists (e.g. montelukast, zafirlukast) have implicated cys-LTs as the predominant mediators of early and late bronchoconstrictor responses to allergen challenge, and also of responses to exercise, cold air and sulphur dioxide (Holgate et al 1996; Lazarus et al 1997). Anti-leukotriene Drugs and AIA The prominent role of cys-LTs in adverse respiratory and rhinitic reactions to NSAIDs in most AIA patients has been confirmed with placebo-controlled double-blind clinical trials of specific leukotriene modifier drugs. The 5-LO inhibitors zileuton and ZD2138 markedly blocked the rise in urinary LTD4 and the fall in FEV1 following oral aspirin challenge of AIA patients (Israel et al 1993; Nasser et al 1994). Rhinoconjunctival and dermal reactions to oral aspirin are also blocked by zileuton (Israel et al 1993; Fischer et al 1994). The cys-LT receptor antagonists pobilukast (SKF 104,353), verlukast (MK-0679) and pranlukast (ONO-1078) block oral NSAID-induced respiratory reactions in AIA patients (Christie et al 1991a; Dahle´n et al 1993a; Yamamoto et al 1994). In contrast, the potent histamine H1 receptor antagonist, terfenadine, has no effect on respiratory reactions to inhaled lysine–aspirin (Phillips et al 1989). In multiple-dose studies, both montelukast and pranlukast produced consistent clinical improvements in AIA patients that were additive to those achieved with corticosteroids (Yoshida et al 2000; Dahle´n et al 2002). The ‘‘Shunting’’ and ‘‘PGE2 Brake’’ Models of AIA Since the prostanoids and leukotrienes are both derived from arachidonic acid, blockade of the prostanoid pathway by NSAIDs was initially proposed to shunt arachidonate towards the formation of cys-LTs, leading to acute bronchoconstriction. However, arachidonate utilized by the COX and 5-LO pathways may be generated by distinct cytosolic (85 kDa) and secretory (14 kDa) phospholipase A2 enzymes, as demonstrated in monocytes (Marshall et al 1997), making the shunting mechanism less probable. Moreover, in isolated eosinophils, the total increase in 5-LO pathway products following incubation with an NSAID appears greater in molar amounts than the decrement in prostanoid synthesis caused by COX inhibition (Tenor et al 1996). The discrepancy argues against a simple shunting of arachidonate from the COX to the 5-LO pathway by NSAIDs. An alternative mechanism is that NSAIDs block the formation of PGE2, which is known to inhibit leukotriene synthesis by leukocytes in vitro (Ham et al 1983; Elliott et al 1991; Tenor et al 1996). NSAIDs may thus liberate the 5-LO pathway from suppression by endogenous PGE2 in vivo (Szczeklik 1995). Indeed, in the absence of NSAIDs, eosinophils generate sufficient PGE2 to inhibit platelet-activated factor-induced LTC4 synthesis by about 90% (Tenor et al 1996). In the presence of indomethacin, LTC4 synthesis is greatly enhanced and is returned to baseline by exogenous PGE2. The PGE2 brake may act in an autocrine or
LEUKOTRIENES IN ASPIRIN-INTOLERANT ASTHMA paracrine manner at EP2 receptors on the eosinophil surface, followed by an increase in intracellular cAMP, but the mechanism by which this inhibits 5-LO pathway enzyme activity is unknown. That this braking mechanism may occur only in restricted populations of leukocyte subtypes is suggested by the failure of NSAIDs to stimulate cys-LT synthesis in mixed leukocyte populations from AIA or ATA subjects (Pierzchalska et al 2000). Prostanoid and Leukotriene Synthesis In Vivo in AIA It is difficult to confirm either the ‘shunting’ hypothesis or the ‘PGE2 brake’ hypothesis in the absence of specific PGE2 receptor (EP) antagonists, but in either case, inhibition of COX by NSAIDs is thought to lead to acute overproduction of the potent bronchoconstrictor cys-LTs. The phenomenon has been convincingly demonstrated following NSAID challenge in vivo (Figure 21.1). Levels of LTC4 in nasal lavage fluid rise in some AIA subjects who experience bronchoconstriction and rhinitic symptoms in response to low-dose (60 mg) oral aspirin challenge, but not in those subjects who experienced bronchoconstriction without rhinitic symptoms, or in aspirin-tolerant subjects (Ferreri et al 1988). Although no changes in PGE2 levels in nasal lavage were detectable with 60 mg aspirin, larger doses (650 mg) reduced PGE2 in the ATA and normal groups without any increase in LTC4 release. Instillation of aspirin into the nasal passages of 10 AIA, 10 ATA, and 7 normal subjects reduced PGE2 levels in all groups 60 min after challenge, but rises in cys-LTs and in nasal symptoms were only observed in the AIA group (Picado et al 1992). Urinary LTE4 levels are a marker of whole-body cys-LT production. Urinary LTE4 levels rise three- to seven-fold after oral aspirin challenge of AIA patients, but not after methacholineinduced bronchoconstriction or placebo challenge, and are accompanied by a fall in urinary 11-dehydro-thromboxane A2, a marker of prostanoid synthesis (Sladek and Szczeklik 1993). Urinary LTE4 levels also rise after inhaled lysine–aspirin challenge (Christie et al 1992; Kumlin et al 1992). Urinary LTE4 levels do not rise after oral aspirin challenge of ATA patients (Christie et al 1991b; Knapp et al 1992; Sladek and Szczeklik 1993). Bronchoscopic challenge with lysine–aspirin reduces PGE2 and thromboxane A2 levels in the BALF of AIA and ATA patients, and a dramatic rise in BALF cys-LTs is seen in the AIA group but not the ATA group (Szczeklik et al 1996b). The PGE2 brake hypothesis has received further support from studies with inhaled PGE2. In AIA patients, preinhalation of PGE2 before inhaled lysine–aspirin challenge completely ablated the bronchoconstriction and the rise in urinary LTE4 that occurred after lysine–aspirin challenge (Sestini et al 1996). The protective effect of inhaled PGE2 against inhaled lysine–aspirin does not correlate with its relatively weak bronchodilator activity (Szczeklik et al 1996a), suggesting that inhaled PGE2 protects by restoring the suppression of cys-LT synthesis, not by dilating the airways directly. Eicosanoid Enzymes in the AIA Airway The cyclooxygenase theory and the PGE2 brake hypothesis do not explain why NSAIDs increase cys-LT synthesis in AIA patients but not in aspirin-tolerant subjects. Aspirin effectively inhibits PGE2 synthesis in both patient groups (Szczeklik et al 1996b). It is also unclear why cys-LT synthesis is chronically elevated in AIA patients even in the absence of exposure to NSAIDs, as demonstrated in BALF and urinary cys-LT levels (Christie et al 1991b, 1992; Kumlin et al 1992; Smith et al 1992; Cowburn et al 1998). This persistent overproduction has marked effects on lung
249
function in AIA patients, as the cys-LT receptor antagonist verlukast improves baseline lung function in AIA patients by an average of 18%, ranging up to 34%, with the improvement correlating with the severity of asthma and aspirin sensitivity (Dahle´n et al 1993b). These data imply chronic overactivity of the cys-LT pathway in AIA or a defect in its downregulation by the prostanoid pathway, and indicate that NSAIDs exacerbate, but do not create, the anomaly. These questions have been investigated by immunohistochemical analysis of eicosanoid pathway enzyme expression in bronchial biopsies. In biopsies from unchallenged AIA and ATA patients and normal subjects, there are no differences in the numbers of cells immunostaining for 5-LO or FLAP (Nasser et al 1996b; Sampson et al 1997; Cowburn et al 1998). However, counts of cells immunostaining for LTC4 synthase, the terminal enzyme for cys-LT synthesis, are five-fold higher in AIA biopsies than in ATA biopsies and 18-fold higher than in normal biopsies (Sampson et al 1997; Cowburn et al 1998). Baseline levels of cys-LTs are significantly higher in the BALF of AIA patients compared to ATA patients, and correlate exclusively with the numbers of LTC4 synthase-positive cells in the bronchial mucosa (Cowburn et al 1998). Persistent cys-LT overproduction in steadystate AIA may therefore be due to overexpression of LTC4 synthase in the bronchial wall. The same study showed that bronchial responsiveness to inhaled lysine–aspirin correlates exclusively with the numbers of LTC4 synthase-positive cells in the bronchial mucosa, and not with immunostaining for the eicosanoid pathway enzymes, or with counts of leukocyte subtypes (Cowburn et al 1998). By increasing cellular capacity for LTC4 synthesis when PGE2 suppression is removed by NSAIDs, overexpression of LTC4 synthase in the bronchial wall may be an important determinant of acute respiratory reactions to NSAIDs (Figure 21.2). Abnormal Regulation of LTC4 Synthase in AIA LTC4 synthase is a homodimeric, integral membrane protein of 16.6 kDa subunits with sequence homology to FLAP (Lam et al 1994), which conjugates the unstable epoxide LTA4 with reduced glutathione to form LTC4 (Yoshimoto et al 1988). Its gene lies on human chromosome 5q35, telomeric to other genes including those of cytokines and receptors implicated in allergic inflammation (Penrose et al 1996). The majority of LTC4 synthase-positive cells in AIA bronchial biopsies are eosinophils (71%), with the remainder being mast cells and macrophages, and the proportion of eosinophils expressing LTC4 synthase is significantly higher in AIA biopsies (51%) than in ATA biopsies (21%) (Cowburn et al 1998). LTC4 synthase expression within eosinophils and other cells may therefore be enhanced by the cytokine microenvironment in the AIA lung and/or by genetic variation in its regulation. During the maturation of eosinophils from cord blood mononuclear cells, 5-LO, FLAP and LTC4 synthase are expressed sequentially in response to interleukin (IL)-5 and IL-3 (Boyce et al 1996). Counts of cells expressing IL-5 are significantly higher in AIA biopsies than in ATA biopsies, but there are no differences in IL-3 or GM-CSF (Lams et al 1997; Cowburn et al 1998). The local increase in IL-5 within the bronchial mucosa is not reflected in the serum of AIA patients (Mastalerz et al 2001). Increased LTC4 synthase expression may alternatively reflect a recently-described biallelic polymorphism in the LTC4 synthase gene promoter (Sanak et al 1997). A single base-pair (bp) transversion (A–C) at a site 444 bp upstream of the transcription start site creates an additional recognition sequence for the AP-2 nuclear transcription factor, suggesting an increased susceptibility of the variant allele to induction by cytokines or other factors. Furthermore, genotyping in a Polish population showed that the
250
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 21.1 The two principal eicosanoid anomalies in the nasal and bronchial airways of patients with aspirin-intolerant asthma (AIA), compared to aspirin-tolerant asthmatics (ATA) and normal subjects. Schematic diagram based on experimental data in Christie et al (1991b), Picado et al (1992), Christie et al (1992), Kumlin et al (1992), Smith et al (1992), Sladek and Szczeklik (1993), Szczeklik et al (1996b) and Cowburn et al (1998). First, in all subjects, NSAID exposure reduces levels of PGE2 and other prostanoids and their metabolites in BALF, nasal lavage fluid and urine, but only the AIA patients show a consequent rise in the synthesis of cysteinyl-leukotrienes. Studies with leukotriene modifier drugs, including the specific cys-LT receptor antagonists montelukast and zafirlukast, indicate that the rise in cys-LTs is the principal mediator of acute bronchoconstriction after NSAID exposure. Second, at baseline, in the absence of any NSAID exposure, AIA patients have significant elevations in cys-LT synthesis compared to the ATA and normal groups
frequency of the variant allele was doubled in AIA patients compared to aspirin-tolerant asthmatic or normal subjects (Sanak et al 1997). Among AIA patients, those with at least one variant allele produce significantly more urinary LTE4 after aspirin challenge than those who are homozygous for the normal allele, and their blood eosinophils show enhanced transcription of LTC4 synthase mRNA (Sanak et al 2000a). Thus, genetic variation in LTC4 synthase regulation may underlie the enhanced LTC4 synthase expression in AIA bronchial biopsies, resulting in elevated cys-LT synthesis at baseline, and an increased capacity for cys-LT synthesis when NSAID exposure liberates the pathway from suppression by the endogenous PGE2 brake. However, other workers find no difference in the prevalence of the A444C LTC4 synthase polymorphism in AIA, ATA, and normal subjects from Japanese and US Caucasian populations (Van Sambeek et al 2000). The LTC4 synthase mutation may instead be a marker of cys-LT overproduction and involvement in severe asthma, irrespective of aspirin sensitivity (Sampson et al 2000).
The Cellular Source of cys-LT Overproduction in AIA In the airways, cys-LTs may be generated by eosinophils, mast cells, basophils, or monocyte-macrophages. Cell-specific activation markers are therefore required to identify the site of action of NSAIDs and to localize the source of excess cys-LT synthesis in AIA patients.
Figure 21.2 NSAIDs are proposed to trigger the acute release of cysteinyl-leukotrienes by reducing prostanoid synthesis via inhibition of cyclooxygenase, particularly COX-1. This may shunt the common substrate, arachidonic acid, into the 5-lipoxygenase pathway, or more likely remove the suppressive effects of PGE2 on cysteinyl-LT synthesis. In this schematic diagram, the PGE2 brake is postulated to elevate cyclic cAMP via EP2 receptor activation, leading to protein kinase A inihibition of one or more enzymes of the 5-LO/LTC4 synthase pathway. The excess cysteinyl-LT synthesis in aspirin-intolerant asthma patients after NSAIDs, and also the elevated basal synthesis of cysteinyl-LT in these patients, may be related to excess expression in airway leukocytes of LTC4 synthase in AIA patients. Anomalies in COX-1 or COX-2 that make the prostanoid pathway less active at baseline and more vulnerable to inhibition by NSAIDs have also been postulated
LEUKOTRIENES IN ASPIRIN-INTOLERANT ASTHMA In the nasal airway, local or oral aspirin challenge of sensitive patients decreases prostanoid levels and increases cys-LT levels, and this is associated with increments in nasal tryptase and histamine, indicating mast cell activation (Ferreri et al 1988; Picado et al 1992; Fischer et al 1994; Kowalski et al 1996). Tryptase and histamine levels also rise in the serum of patients experiencing systemic reactions to oral aspirin, but not in those with localized respiratory reactions (Bosso et al 1991). Release of tryptase is also a recognized marker of mast cell activation in the pulmonary airway, and occurs within 5 min of allergen challenge in allergic asthmatics (Dahle´n et al 1993a). However, in AIA patients challenged with inhaled or endobronchial lysine–aspirin, the expected fall in PGE2 and rise in cys-LT levels in BALF are not accompanied by increased tryptase (Sladek et al 1994; Szczeklik et al 1996b). A correlation between cys-LT and histamine levels perhaps suggests that basophils may be involved. A rise in urinary levels of the PGD2 metabolite 9a,11b-PGF2 following endobronchial lysine–aspirin challenge has been interpreted as evidence for mast cell activation in AIA (O’Sullivan et al 1996). The paradox that prostanoid synthesis seems to be stimulated by a COX inhibitor has not been satisfactorily explained, and it is not known whether urinary 9a,11b-PGF2 also rises after NSAID exposure in ATA patients or normals. The possibility that increases in prostanoid levels in BALF or in urine may simply reflect displacement by aspirin from non-specific binding sites on plasma albumin or on tissue proteins has not been evaluated. Blood eosinophilia is a consistent finding in patients with AIA (Stevenson and Simon 1993). Eosinophils are also found in high numbers in nasal polyps in aspirin-sensitive subjects (Yamashita et al 1989; Hamilos et al 1995; Ogata et al 1999). Alongside increases in tryptase, increases in eosinophils and ECP have been described in the nasal airways of aspirin-sensitive patients after nasal lysine–aspirin challenge (Kowalski et al 1996). Oral aspirin reduces blood eosinophil counts in AIA patients and tends to increase plasma ECP levels, suggestive of eosinophil activation and migration (Sladek and Szczeklik 1993). In BALF, eosinophil counts and ECP levels are higher in AIA patients than in ATA patients (Sladek et al 1994). However, reported responses to lysine–aspirin challenge are contradictory, with ECP in BALF either falling (Sladek et al 1994) or remaining unchanged (Szczeklik et al 1996b). Lysine–aspirin challenge increases eosinophil counts in BALF in AIA but not ATA patients (Szczeklik et al 1996b). Immunohistochemical studies in bronchial biopsies have confirmed a marked airway eosinophilia in unchallenged AIA patients, with eosinophil counts 2.4–4-fold higher than in aspirin-tolerant asthmatics and 10–15-fold higher than in normals (Nasser et al 1996b; Cowburn et al 1998). In contrast, there are no meaningful differences between AIA patients and aspirin-tolerant asthmatics in counts of mast cells, neutrophils, monocytemacrophages or subsets of T lymphocytes. Twenty minutes after endobronchial lysine–aspirin challenge, a decrease in tryptasepositive mast cell counts and a modest increase in ECP-secreting eosinophils have been reported in bronchial biopsies of AIA patients (Nasser et al 1996a), suggesting that mast cells and eosinophils are both activated. Other studies contradict these changes in cell counts, but show that inhaled lysine–aspirin causes the rapid release of histamine and IL-5 in BALF, probably from mast cell granules, and an efflux of eosinophils into the airway lumen, suggesting that activation of both mast cells and eosinophils occurs, leading to cys-LT generation (Szczeklik et al 1996b; Cowburn et al 1998). IL-5 expression is elevated in AIA bronchial biopsies, and most IL-5-expressing cells are mast cells (Cowburn et al 1998). Since both cys-LTs and IL-5 are specific eosinophil chemoattractants, an influx of eosinophils into the bronchial mucosa might become apparent at later time-points after NSAID challenge.
251
PGE2 suppresses, and NSAIDs enhance, leukotriene synthesis in a number of inflammatory cell-types in vitro, including eosinophils, neutrophils, basophils and macrophages (Ham et al 1983; Peters et al 1986; Elliott et al 1991; Tenor et al 1996). This does not appear to occur in human lung mast cells (Peters et al 1985, 1986). Moreover, lysine–aspirin challenge fails to reduce BALF and urine levels of mast cell-distinctive prostanoids in AIA patients (O’Sullivan et al 1996; Szczeklik et al 1996b). A similar phenomenon is observed in the nasal airways (Picado et al 1992). This refractoriness of the prostanoid pathway in mast cells to inhibition by NSAIDs is intriguing, but may suggest that the mast cell is not the site of action of NSAIDs in AIA. In contrast to the mast cell, suppression of cys-LT synthesis by autologous and exogenous PGE2, and enhancement of cys-LT synthesis by indomethacin, has been described in human eosinophils (Tenor et al 1996). The ‘mast cell stabilizing’ drug sodium cromoglycate, the antiviral agent acyclovir, the antibiotic roxithromycin, and the long-acting b2-agonist salmeterol have all been reported to block not only the bronchial responses to inhaled NSAIDs but also the associated rise in urinary LTE4 (Szczeklik et al 1998; Yoshida et al 1998a, 1998b; Shoji et al 1999). The mechanism of these drugs in this context is difficult to understand, but may involve prevention of mast cell degranulation, consistent with the mast cell being the source of acute cys-LT release following NSAID exposure. In contrast, their lack of effect on urinary LTE4 levels in unchallenged AIA patients may suggest that chronic overproduction of cys-LT in AIA is related to eosinophils. The Prostanoid Pathway in AIA The PGE2 brake has been proposed as an important factor in responses to allergic and other stimuli, as well as to NSAIDs (Pavord and Tattersfield 1994). Since AIA patients tolerate the selective COX-2 inhibitors, including nimesulide, meloxicam and rofecoxib (Bianco et al 1993; Quaratino et al 2000; Szczeklik et al 2000), but most often respond adversely to NSAIDs with a greater selectivity for COX-1, such as aspirin and indomethacin, the PGE2 brake in the lung may be produced by constitutive COX-1. COX-1 is expressed in a large number of cell types, including mast cells, eosinophils, macrophages, vascular endothelial cells, bronchial epithelium and bronchial smooth muscle (Belvisi et al 1997). The cellular sources of the putative PGE2 brake, its precise cellular targets and the receptors involved remain unclear. Indeed, a 6-week clinical trial of the stable PGE1 analogue misoprostol at two doses demonstrated no clinical benefit on asthma control in 17 AIA patients (Wasiak and Szmidt 1992). AIA patients may overproduce cys-LTs chronically and after NSAID exposure, because of a defect in the PGE2 brake. However, baseline BALF levels of PGE2 and of other prostanoids, including PGD2, PGF2a and thromboxane A2, are not consistently different between AIA and ATA patients (Sladek et al 1994; Szczeklik et al 1996b). Baseline levels of the PGD2 metabolite 9a,11b-PGF2 in the urine are not different between AIA, ATA and normal subjects (O’Sullivan et al 1996). Inhaled lysine–aspirin can reduce BALF PGE2 levels in ATA patients to the same extent as in AIA patients, but only the AIA patients generate significant cys-LTs as a result (Szczeklik et al 1996b). Overall, these data do not support a generalized anomaly with the COX pathway in AIA at baseline. Similarly, an abnormal expression or cellular distribution of COX-2 and COX-1 might cause altered responses due to the different sensitivities of the two isoenzymes to inhibition by various NSAIDs. However, immunohistochemical studies of bronchial biopsies have found little evidence for a meaningful anomaly in COX isoenzyme expression in AIA. There are no
252
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
differences in the counts of cells expressing COX-1 in biopsies from AIA, ATA and normal subjects, and no study has found an overall difference in COX-2 expression between ATA and AIA biopsies (Demoly et al 1997; Sousa et al 1997; Cowburn et al 1998). Although bronchial COX-2 expression is reported to be generally higher in asthmatics in one study (Sousa et al 1997), other studies find no differences in COX-2 between AIA, ATA and normal biopsies (Demoly et al 1997; Cowburn et al 1998). Despite the lack of difference in total COX-2-positive cells, a significantly greater proportion of total COX-2 has been reported to be expressed in mast cells and eosinophils in AIA biopsies compared to ATA biopsies (Sousa et al 1997), but it is not clear which other cell type has less COX-2 expression to compensate. This finding requires confirmation in lung mast cells and eosinophils isolated from the two patient groups. When aspirin acetylates COX-2 but not COX-1, 15R-hydroxyeicosatetraenoic acid (15R-HETE) is produced (Lecomte et al 1994). This has been proposed to underlie AIA via enhanced production of 15-epi-lipoxins if COX-2 expression is enhanced in AIA lung (Mitchell and Belvisi 1997). However, as described above, the total number of cells expressing COX-2 is not different in AIA and ATA biopsies (Sousa et al 1997; Cowburn et al 1998). Moreover, 15R-HETE itself is biologically inactive and 15-epilipoxins are largely bronchodilator and antiinflammatory mediators, while NSAID-induced reactions are known to be blocked by specific antileukotriene drugs. Only aspirin and not other NSAIDs generate 15R-HETE from COX-2 (Lecomte et al 1994), yet AIA patients show cross-reactivity to a wide range of structurally unrelated NSAIDs. It is therefore unlikely that 15HETE or its metabolites can account for acute adverse respiratory reactions to aspirin or other NSAIDs (Sampson et al 1998). Other workers report that whole blood from AIA patients in fact has a reduced capacity for the synthesis of lipoxins and 15-epi-lipoxins compared to ATA blood (Sanak et al 200b), suggesting that absence of the antiinflammatory activities of these mediators may be involved in chronic AIA. The most convincing evidence of a fundamental anomaly in COX isozyme expression in aspirin intolerance comes from the nasal epithelia of patients with asthma and rhinitis; nasal epithelial cells from those patients with aspirin intolerance had markedly lower COX-2 mRNA levels than those with aspirintolerant disease, while COX-1 mRNA was not different (Picado et al 1999). Although the pharmacological evidence suggests that inhibition of PGE2 synthesis by COX-1 is the critical factor in triggering acute NSAID reactions, a failure of COX-2 to contribute to the production of protective PGE2 may also be involved. Perhaps PGE2 from both isozymes is required, so that adverse reactions occur when COX-1 is inhibited in the putative absence of COX-2 expression in AIA patients. Further work is required to determine whether an anomaly at the heart of the COX pathway exists in AIA.
Aspirin Desensitization Acute respiratory reactions to NSAIDs in AIA patients are always followed by a refractory period lasting 2–5 days during which further reactions to NSAIDs cannot be induced (Stevenson and Simon 1993). Sensitivity to NSAIDs re-emerges within a week, but regular low-dose NSAIDs can maintain the refractory state indefinitely. Daily or alternate-day dosing of NSAIDs is therefore used clinically to desensitize AIA patients to inadvertent ingestion of NSAIDs and to allow NSAID therapy of concomitant diseases such as arthritis. Cross-desensitization occurs such that repeated dosing with one NSAID provides protection against adverse reactions to other NSAIDs.
The effect of aspirin desensitization on responses to other stimuli is unclear. Bronchial responsiveness to methacholine is not altered immediately after desensitization (Stevenson et al 1980), but responsiveness to histamine is reduced in some AIA patients after 4 weeks (Kowalski et al 1986). Desensitization improves symptoms of rhinosinusitis in 68% of AIA patients, but only 31% experience an improvement in asthma symptoms (Kowalski 1992). Chronic desensitization reduces the numbers of acute exacerbations and hospital admissions in AIA patients compared to a control group of AIA patients who avoided NSAIDs, and this was associated with reductions in corticosteroid use (Sweet et al 1999). Aspirin desensitization may thus improve chronic airway inflammation in aspirin-sensitive patients. The airways of AIA patients have been reported to be more responsive to inhaled LTE4 than those of ATA patients, and this bronchial hyperresponsiveness (relative to inhaled histamine) is significantly reduced by aspirin desensitization (Arm et al 1989). Following desensitization with low doses of NSAIDs, the amounts of cys-LTs generated by a large dose of NSAID are somewhat lower than before desensitization, and the bronchial response is greatly reduced (Nasser et al 1998). It is possible that low doses of NSAIDs may enhance cys-LT synthesis sufficiently to cause suicide inactivation of 5-LO and/or to downregulate cysLT receptors on bronchial smooth muscle, but without significant effect on bronchial tone. However, the capacity of leukocytes to generate cys-LTs and the sensitivity of the airways to cys-LTs may be gradually reduced. Further studies are required to monitor changes in inflammatory cell populations, 5-LO pathway enzyme expression, cys-LT release and cys-LT receptor density during aspirin desensitization. In summary, aspirin-intolerant asthma can be seen as a chronic, severe non-allergic form of asthma associated with exaggerated airway eosinophilia and usually a dependence on corticosteroids. Overexpression of LTC4 synthase in eosinophils within AIA biopsies may drive chronic cys-LT overproduction, associated with polymorphism in the LTC4 synthase gene and with excessive IL-5 production. This is exacerbated when NSAIDs inhibit COX1-derived PGE2 synthesis, leading to an acute surge in cys-LT synthesis, possibly in mast cells. A failure of COX-2 transcription or expression has also been implicated in making the nasal epithelia vulnerable to NSAID-induced reactions. In patients with aspirin-intolerance, the development of specific oral cys-LT antagonists, such as zafirlukast and montelukast (Sampson and Holgate 1999), may provide effective control of the acute exacerbations and the underlying airway inflammation, particularly in combination with corticosteroids, while the development of COX-2-selective NSAIDs, such as rofecoxib and celecoxib, may allow safe treatment of concurrent inflammatory and autoimmune diseases such as rheumatoid arthritis in these patients.
REFERENCES Ameisen JC, Capron A, Maclouf J et al (1985) Aspirin-sensitive asthma: abnormal platelet response to drugs inducing asthmatic attacks. Diagnostic and physiopathological implications. Int Arch Allerg Appl Immunol, 78, 438–448. Arm JP and Lee TH (1993) Sulphidopeptide leukotrienes in asthma. Clin Sci, 84, 501–510. Arm JP, O’Hickey SP, Spur BW and Lee TH (1989) Airway responsiveness to histamine and leukotriene E4 in subjects with aspirin-induced asthma. Am Rev Resp Dis, 140, 148–153. Belvisi MG, Saunders MA, Haddad EB et al (1997) Induction of cyclooxygenase-2 by cytokines in human airway smooth muscle cells: novel inflammatory role of this cell-type. Br J Pharmacol, 120, 910– 916.
LEUKOTRIENES IN ASPIRIN-INTOLERANT ASTHMA Bianco S, Robuschi M and Petrigni G (1977) Aspirin-induced tolerance in aspirin-asthma detected by a new challenge test. J Med Sci, 5, 129–134. Bianco S, Robuschi M and Petrigni G (1981) Aspirin insensitivity in asthmatics. Br Med J, 282, 146. Bianco S, Robuschi M, Petrigni G et al (1993) Efficacy and tolerability of nimesulide in asthmatic patients intolerant to aspirin. Drugs, 46(suppl 1), 115–120. Bosso JV, Schwartz LB and Stevenson DD (1991) Tryptase and histamine release during aspirin-induced respiratory reactions. J Allerg Clin Immunol, 88, 830–837. Boyce JA, Lam BK, Penrose JF et al (1996) Expression of LTC4 synthase during the development of eosinophils in vitro from cord blood progenitors. Blood, 88, 4338–4347. Christie PE, Smith CM and Lee TH (1991a) The potent and selective sulfidopeptide leukotriene antagonist, SK&F 104353, inhibits aspirininduced asthma. Am Rev Resp Dis, 144, 957–958. Christie PE, Tagari P, Ford-Hutchinson AW et al (1991b) Urinary leukotriene E4 concentrations increase after aspirin challenge in aspirin-sensitive asthmatic subjects. Am Rev Resp Dis, 143, 1025–1029. Christie PE, Tagari P, Ford-Hutchinson AW et al (1992) Urinary leukotriene E4 after lysine–aspirin inhalation in asthmatic subjects. Am Rev Resp Dis, 146, 1531–1534. Cooke RA (1919) Allergy in drug idiosyncrasy. J Am Med Assoc, 73, 759–760. Cowburn AS, Sladek K, Soja J et al (1998) Overexpression of leukotriene C4 synthase in bronchial biopsies from patients with aspirin-intolerant asthma. J Clin Invest, 101, 834–846. Dahle´n B and Zetterstrom O (1990) Comparison of bronchial and per oral provocation with aspirin in aspirin-sensitive asthmatics. Eur Resp J, 3, 527–534. Dahle´n B, Kumlin M, Margolskee DJ et al (1993a) The leukotrienereceptor antagonist MK-0679 blocks airway obstruction induced by inhaled lysine–aspirin in aspirin-sensitive asthmatics. Int Resp J, 6, 1018–1026. Dahle´n B, Margolskee DJ, Herstrom O and Dahle´n SE (1993b) Effect of the leukotriene receptor antagonist HK-0679 on baseline pulmonary function in aspirin-sensitive asthmatic subjects. Thorax, 48, 1205–1210. Dahle´n SE, Malmstrom K, Nizankowska E et al (2002) Improvement of aspirin-intolerant asthma by montelukast, a leukotriene antagonist. A randomized, double-blind, placebo-controlled trial. Am J Resp Crit Care Med, 165, 9–14. Demoly P, Jaffuel D, Lequeux N et al (1997) Prostaglandin H synthase 1 and 2 immunoreactivities in the bronchial mucosa of asthmatics. Am J Resp Crit Care Med, 156, 670–675. Diamant Z, Hiltermann JT, Van Rensen EL et al (1997) The effect of inhaled leukotriene D4 and methacholine on sputum cell differentials in asthma. Am J Resp Crit Care Med, 155, 1247–1253. Elliott GR, Lauwen AP and Bonta IL (1991) Leukotriene B4 stimulation of macrophage cyclooxygenase metabolite synthesis. Agents Act, 32, 73–74. Feigenbaum BA, Stevenson DD and Simon RA (1995) Hydrocortisone sodium succinate does not cross-react with aspirin in aspirin-sensitive patients with asthma. J Allerg Clin Immunol, 96, 545–548. Ferreri NR, Howland WC, Stevenson DD and Spiegelberg HL (1988) Release of leukotrienes, prostaglandins, and histamine into nasal secretions of aspirin-sensitive asthmatics during reaction to aspirin. Am Rev Resp Dis, 137, 847–854. Fischer AR, Rosenberg MA, Lilly CM et al (1994) Direct evidence for a role of the mast cell in the nasal response to aspirin in aspirin-sensitive asthma. J Allerg Clin Immunol, 94, 1046–1056. Giraldo B, Blumenthal MN and Spink WW (1969) Aspirin intolerance and asthma: clinical and immunological study. Ann Int Med, 71, 479–496. Ham EA, Soderman DD, Zanetti ME et al (1983) Inhibition by prostaglandins of leukotriene B4 release from activated neutrophils. Proc Natl Acad Sci USA, 80, 4349–4353. Hamilos DL, Leung DYM, Wood R et al (1995) Evidence for distinct cytokine expression in allergic vs. non-allergic chronic sinusitis. J Allerg Clin Immunol, 96, 537–544. Holgate ST, Bradding P and Sampson AP (1996) Leukotriene antagonists and synthesis inhibitors: new directions in asthma therapy. J Allerg Clin Immunol, 98, 1–13. Israel E, Fischer AR, Rosenberg MA et al (1993) The pivotal role of 5lipoxygenase products in the reaction of aspirin-sensitive asthmatics to aspirin. Am Rev Resp Dis, 148, 1447–1451.
253
Knapp HR, Sladek K and Fitzgerald GA (1992) Increased excretion of leukotriene E during aspirin-induced asthma. J Lab Clin Med, 119, 48–51. Kowalski ML, Grzelewska Rzymowska I, et al (1986) Clinical efficacy of aspirin in ‘‘desensitised’’ aspirin-sensitive asthmatics. Eur J Resp Dis, 69, 219–225. Kowalski ML (1992) Management of aspirin-sensitive rhinosinusitis– asthma syndrome: what role for aspirin desensitization? Allerg Proc, 13, 175–184. Kowalski ML, Grzegorczyk J, Wojciechowska B and Poniatowska M (1996) Intranasal challenge with aspirin induces cell influx and activation of eosinophils and mast cells in nasal secretions of ASAsensitive patients. Clin Exp Allerg, 26, 807–814. Kumlin M, Dahle´n B, Bjorck T et al (1992) Urinary excretion of leukotriene E4 and 11-dehydro-thromboxane B2 in response to bronchial provocations with allergen, aspirin, leukotriene E4, and histamine in asthmatics. Am Rev Resp Dis, 146, 96–103. Laitinen LA, Laitinen A, Haahtela T et al (1993) Leukotriene E4 and granulocytic infiltration into asthmatic airways. Lancet, 341, 989–990. Lam BK, Penrose JF, Freeman GJ and Austen KF (1994) Expression cloning of a cDNA for human leukotriene C4 synthase, an integral membrane protein conjugating reduced glutathione to leukotriene A4. Proc Natl Acad Sci USA, 91, 7663–7667. Lams BEA, Sousa AR, Pfister R et al (1997) Expression of interleukin-5 (IL-5) and granulocyte-macrophage colony stimulating factor (GMCSF) in aspirin-sensitive (ASA) and non-aspirin-sensitive (NASA) asthmatic airways. Am J Resp Crit Care Med, 155 (abstr), A974. Lazarus SC, Wong HH, Watts MJ et al (1997) The leukotriene receptor antagonist zafirlukast inhibits sulfur dioxide-induced bronchoconstriction in patients with asthma. Am J Resp Crit Care Med, 156, 1725–1730. Lecomte M, Laneuville O, Ji C et al (1994) Acetylation of human prostaglandin endoperoxide synthase-2 (cyclooxygenase-2) by aspirin. J Biol Chem, 269, 13207–13215. Luxenberger W, Posch U, Berghold A et al (2000). HLA patterns in patients with nasal polyposis. Eur Arch Otorhinolaryngol, 257, 37–39. Manning ME and Stevenson DD (1991) Aspirin sensitivity: a distressing reaction that is now often treatable. Postgrad Med, 90, 227–233. Marquette CH, Saulnier F, Leroy O et al (1992) Long-term prognosis of near-fatal asthma: a 6-year follow-up of 145 asthmatic patients who underwent mechanical ventilation for a near-fatal attack of asthma. Am Rev Resp Dis, 146, 76–81. Marshall LA, Bolognese B, Winkler JD and Roshak A (1997) Depletion of human monocyte 85 kDa phospholipase A2 does not alter leukotriene formation. J Biol Chem, 272, 759–765. Mastalerz L, Sanak M and Szczeklik A (2001) Serum IL-5 in aspirininduced asthma. Clin Exp Allerg, 31, 1036–1040. McDonald JR, Mathison DA and Stevenson DD (1972) Aspirin intolerance in asthma: detection by oral challenge. J Allerg Clin Immunol, 50, 198–207. Mitchell JA and Belvisi MG (1997) Too many COX (cyclo-oxygenase) spoil the broth: aspirin-sensitive asthma and 5-lipoxygenase. Thorax, 52, 933–935. Mullarkey MF, Thomas PS, Hansen JA et al (1986) Association of aspirin-sensitive asthma with HLA-DQw2. Am Rev Resp Dis, 133, 261–263. Murphy RC, Hammarstrom S and Samuelsson B (1979) Leukotriene C: a slow-reacting substance from murine mastocytoma cells. Proc Natl Acad Sci USA, 76, 4275–4279. Nasser SM, Bell GS, Hawksworth RJ et al (1994) Effect of the 5lipoxygenase inhibitor ZD2138 on allergen-induced early and late asthmatic responses. Thorax, 49, 743–748. Nasser S, Christie PE, Pfister R et al (1996a) Effect of endobronchial aspirin challenge on inflammatory cells in bronchial biopsy samples from aspirin-sensitive asthmatic subjects. Thorax, 51, 64–70. Nasser SM, Patel M, Bell GS and Lee TH (1995) The effect of aspirin desensitization on urinary leukotriene E4 concentrations in aspirinsensitive asthma. Am J Resp Crit Care Med, 151, 1326–1330. Nasser SM, Pfister R, Christie PE et al (1996b) Inflammatory cell populations in bronchial biopsies from aspirin-sensitive asthmatic subjects. Am J Resp Crit Care Med, 153, 90–96. Nizankowska E and Szczeklik A (1989) Glucocorticosteroids attenuate aspirin-precipitated adverse reactions in aspirin-intolerant patients with asthma. Ann Allerg, 63, 159–162.
254
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Ogata Y, Okinaka Y and Takahashi M (1999) Detection of activated eosinophils in nasal polyps of an aspirin-induced asthma patient. Rhinology, 37, 16–20. O’Sullivan S, Dahle´n B, Dahle´n SE and Kumlin M (1996) Increased urinary excretion of the prostaglandin D2 metabolite 9a,11bprostaglandin F2 after aspirin challenge supports mast cell activation in aspirin-induced airway obstruction. J Allerg Clin Immunol, 98, 421– 432. Pavord ID and Tattersfield AE (1994) Bronchoprotective role for endogenous prostaglandin E2. Lancet, 345, 436–438. Penrose JF, Spector J, Baldasaro M et al (1996) Molecular cloning of the gene for human leukotriene C4 synthase: organization, nucleotide sequence, and chromosomal localization to 5q35. J Biol Chem, 271, 11356–11361. Peters SP, MacGlashan DWJ, Schleimer RP et al (1985) The pharmacologic modulation of the release of arachidonic acid metabolites from purified human lung mast cells. Am Rev Resp Dis, 132, 367–373. Peters SP, Naclerio RM, Schleimer RP et al (1986) The pharmacologic control of mediator release from human basophils and mast cells. Respiration, 50(suppl 2), 116–122. Phillips GD, Foord R and Holgate ST (1989) Inhaled lysine–aspirin as a bronchoprovocation procedure in aspirin-sensitive asthma: its repeatability, absence of a late-phase reaction, and the role of histamine. J Allerg Clin Immunol, 84, 232–241. Picado C, Castillo JA, Montserrat JM and Agusti Vidal A (1989) Aspirinintolerance as a precipitating factor of life-threatening attacks of asthma requiring mechanical ventilation. Eur Resp J, 2, 127–129. Picado C, Fernandez-Morata JC, Juan M et al (1999) Cyclooxygenase-2 mRNA is downexpressed in nasal polyps from aspirin-sensitive asthmatics. Am J Resp Crit Care Med, 160, 291–296. Picado C, Ramis I, Rosello J et al (1992) Release of peptide leukotriene into nasal secretions after local instillation of aspirin in aspirinsensitive asthmatic patients. Am Rev Resp Dis, 145, 65–69. Pierzchalska M, Mastalerz L, Sanak M et al (2000). A moderate and unspecific release of cysteinyl leukotrienes by aspirin from peripheral blood leucocytes precludes its value for aspirin sensitivity testing in asthma. Clin Exp Allerg, 30, 1785–1791. Quaratino D, Romano A, Di Fonso M et al (2000). Tolerability of meloxicam in patients with histories of adverse reactions to nonsteroidal antiinflammatory drugs. Ann Allerg Asthma Immunol, 84, 613–617. Reiss TF, Chervinsky P, Dockhorn RJ et al (1998) Montelukast, a oncedaily leukotriene receptor antagonist, in the treatment of chronic asthma: a multicenter, randomized, double-blind trial. Arch Int Med, 158, 1213–1220. Sampson AP, Castling DP, Green CP and Price JF (1995) Persistent elevation in plasma and urinary leukotrienes after acute asthma. Arch Dis Child, 73, 221–225. Sampson AP, Cowburn AS, Sladek K et al (1997) Profound overexpression of leukotriene C4 synthase in bronchial biopsies from aspirin-intolerant asthmatic patients. Int Arch Allerg Immunol, 113, 355–357. Sampson AP and Holgate ST (1999) Leukotriene Modifiers in Asthma Treatment. Martin Dunitz, London. Sampson AP, Holgate ST, Austen KF and Szczeklik A (1998) Cyclooxygenase (letter). Thorax, 53, 719–720. Sampson AP, Siddiqui S, Buchanan D et al (2000) Variant LTC4 synthase allele modifies cysteinyl-leukotriene synthesis in eosinophils and predicts clinical response to zafirlukast. Thorax, 55(suppl 2), S28–S31. Samter M and Beers RF (1968) Intolerance to aspirin: clinical studies and consideration of its pathogenesis. Ann Intern Med, 68, 975–983. Sanak M, Simon HU and Szczeklik A (1997) Leukotriene C4 synthase promoter polymorphism and risk of aspirin-induced asthma. Lancet, 350, 1599–1600. Sanak M, Pierzchalska M, Bazan-Socha S and Szczeklik A (2000a) Enhanced expression of the leukotriene C4 synthase due to overactive transcription of an allelic variant associated with aspirin-intolerant asthma. Am J Resp Cell Mol Biol, 23, 290–296. Sanak M, Levy BD, Clish CB et al (2000b) Aspirin-tolerant asthmatics generate more lipoxins than aspirin-intolerant asthmatics. Eur Resp J, 16, 44–49.
Sestini P, Armetti L, Gambaro et al (1996) Inhaled PGE2 prevents aspirininduced bronchoconstriction and urinary LTE4 excretion in aspirinsensitive asthma. Am J Resp Crit Care Med, 153, 572–575. Shoji T, Yoshida S, Sakamoto H et al (1999). Antiinflammatory effect of roxithromycin in patients with aspirin-intolerant asthma. Clin Exp Allerg, 29, 950–956. Sladek K, Dworski R, Soja J et al (1994) Eicosanoids in bronchoalveolar lavage fluid of aspirin-intolerant patients with asthma after aspirin challenge. Am J Resp Crit Care Med, 149, 940–946. Sladek K and Szczeklik A (1993) Cysteinyl leukotrienes overproduction and mast cell activation in aspirin-provoked bronchospasm in asthma. Eur Resp J, 6, 391–399. Slepian IK, Mathews KP and McLean JA (1985) Aspirin-sensitive asthma. Chest, 87, 386–391. Smith AP (1977) Response of aspirin-allergic patients to challenge by some analgesics in common use. Br Med J, 2, 494–496. Smith CM, Hawksworth RJ, Thien FC et al (1992) Urinary leukotriene E4 in bronchial asthma. Eur Resp J, 5, 693–699. Sousa AR, Pfister R, Christie PE et al (1997) Enhanced expression of cyclooxygenase isoenzyme 2 (COX-2) in asthmatic airways and its cellular distribution in aspirin-sensitive asthma. Thorax, 52, 940–945. Spada CS, Nieves AL, Krauss AH and Woodward DF (1994) Comparison of leukotriene B4 and D4 effects on human eosinophil and neutrophil motility in vitro. J Leuko Biol, 55, 183–191. Spector SL, Wangaard CH and Farr RS (1979) Aspirin and concomitant idiosyncrasies in adult asthmatic patients. J Allerg Clin Immunol, 64, 500–506. Stevenson DD (1988) Oral challenges to detect aspirin and sulfite sensitivity in asthma. N Engl Allerg Proc, 9, 135–142. Stevenson DD and Simon RA (1993) Sensitivity to aspirin and nonsteroidal antiinflammatory drugs. In Middleton F, Reed CF, Ellis EF et al (eds), Allergy: Principles and Practice, 4th edn. Mosby YearBook, St Louis, MO, 1747–1765. Stevenson DD, Simon RA and Mathison DA (1980) Aspirin-sensitive asthma: tolerance to aspirin after oral aspirin challenges. J Allerg Clin Immunol, 66, 82–88. Sweet JM, Stevenson DD, Simon RA and Mathison DA (1990) Long-term effects of aspirin desensitization treatment for aspirin-sensitive rhinosinusitis–asthma. J Allerg Clin Immunol, 85, 59–65. Szczeklik A (1988) Aspirin-induced asthma as a viral disease. Clin Allerg, 18, 15–20. Szczeklik a (1990) The cyclooxygenase theory of aspirin-induced asthma. Eur Resp J, 3, 588–593. Szczeklik A (1995) Prostaglandin E2 and aspirin-induced asthma. Lancet, 345, 1056. Szczeklik A, Dworski R, Mastalerz L et al (1998) Salmeterol prevents aspirin-induced attacks of asthma and interferes with eicosanoid metabolism. Am J Resp Crit Care Med, 158, 1168–1172. Szczeklik A, Gryglewski RJ and Czerniawska-Mysik G (1975) Relationship of inhibition of prostaglandin biosynthesis of analogues to asthma attacks in aspirin sensitive patients. Br Med J, 1, 66–69. Szczeklik A, Mastalerz L, Nizankowska E and Cmiel A (1996a) Protective and bronchodilator effects of prostaglandin E and salbutamol in aspirin-induced asthma. Am J Resp Crit Care Med, 153, 567–571. Szczeklik A, Nizankowska E, Czerniawska-Mysik G and Sek S (1985) Hydrocortisone and airflow impairment in aspirin-induced asthma. J Allerg Clin Immunol, 76, 530–536. Szczeklik A, Nizankowska E and Duplaga M (2000) Natural history of aspirin-induced asthma. AIANE Investigators. European Network on Aspirin-induced Asthma. Eur Resp J, 16, 432–436. Szczeklik A, Schmitz Schumann M, Nizankowska E et al (1992) Altered distribution of IgG subclasses in aspirin-induced asthma: high IgG4, low IgG1. Clin Exp Allerg, 22, 283–287. Szczeklik A and Serwonska M (1979) Inhibition of idiosyncratic reactions to aspirin in asthmatic patients by clemastine. Thorax, 34, 654–657. Szczeklik A, Sladek K, Dworski R, et al (1996b) Bronchial aspirin challenge causes specific eicosanoid response in aspirin-sensitive asthmatics. Am J Resp Crit Care Med, 154, 1608–1614. Taylor GW, Taylor IK, Black PN et al (1989) Urinary leukotriene E4 after allergen challenge and in acute asthma and allergic rhinitis. Lancet, i, 584–588. Tenor H, Hatzelmann A, Church MK, et al (1996) Effects of theophylline and rolipram on leukotriene C4 (LTC4) synthesis and chemotaxis of
LEUKOTRIENES IN ASPIRIN-INTOLERANT ASTHMA human eosinophils from normal and atopic subjects. Br J Pharmacol, 118, 1727–1735. Van Sambeek R, Stevenson DD, Baldasaro M et al (2000). 5’Flanking region polymorphism of the gene encoding leukotriene C4 synthase does not correlate with the aspirin-intolerant asthma phenotype in the United States. J Allerg Clin Immunol, 106, 72–76. Vedanthan PR, Menon MM, Bell JD et al (1977) Aspirin and tartrazine oral challenge: incidence of adverse responses in chronic childhood asthma. J Allerg Clin Immunol, 60, 8–13. Wasiak W and Szmidt M (1999). A six week double-blind, placebocontrolled, crossover study of the effect of misoprostol in the treatment of aspirin sensitive asthma. Thorax, 54, 900–904. Weltman JK, Szaro RP and Settipane GA (1978) An analysis of the role of IgE in intolerance to aspirin and tartrazine. Allergy, 34, 273–281. Wenzel SE, Fowler AA and Schwartz LB (1988) Activation of pulmonary mast cells by bronchoalveolar allergen challenge: in vitro release of histamine and tryptase in atopic subjects with and without asthma. Am Rev Resp Dis, 137, 1002–1008. Wenzel SE, Larsen GL, Johnston K et al (1990) Elevated levels of leukotriene C4 bronchoalveolar lavage fluid from atopic asthmatics after endobronchial allergen challenge. Am Rev Resp Dis, 142, 112–119.
255
Williams WR, Pawlowicz A and Davies BH (1990) In vitro tests for the diagnosis of aspirin-sensitive asthma. J Allerg Clin Immunol, 86, 445–451. Yamamoto H, Nagata M, Kuramitsu K et al (1994) Inhibition of analgesic-induced asthma by leukotriene receptor antagonist ONO1078. Am J Resp Crit Care Med, 150, 254–257. Yamashita T, Tsuji H, Maeda N et al (1989) Etiology of nasal polyps associated with aspirin-sensitive asthma. Rhinol Suppl, 8, 15–24. Yoshida S, Sakamoto H, Ishizaki Y et al (2000) Efficacy of leukotriene receptor antagonist in bronchial hyperresponsiveness and hypersensitivity to analgesic in aspirin-intolerant asthma. Clin Invest Allergy, 30, 64–70. Yoshida S, Amayasu H, Sakamoto H et al (1998a) Cromolyn sodium prevents bronchoconstriction and urinary LTE4 excretion in aspirininduced asthma. Ann Allerg Asthma Immunol, 80, 171–176. Yoshida S, Sakamoto H, Yamawaki Y et al (1998b) Effect of acyclovir on bronchoconstriction and urinary leukotriene E4 excretion in aspirininduced asthma. J Allerg Clin Immunol, 102, 909–914. Yoshimoto T, Soberman RJ, Spur B and Austen KF (1988) Properties of highly purified leukotriene C4 synthase of guinea-pig lung. J Clin Invest, 81, 866–871.
22 Essential Fatty Acids Yoeju Min and Michael A. Crawford London Metropolitan University, London, UK
Essential fatty acids are straight-chain monocarboxylic acids with an even number of carbon atoms. There are two families of essential fatty acids, o6 and o3. Linoleic (18:2o6, LA) and alinolenic (18:3o3, ALA) acids are the parent compounds of the o6 and o3 families, respectively. By convention, the formula for fatty acids is abbreviated as X:YoZ. X refers to the number of carbon atoms, Y to the number of double bonds, and Z to the position of the first double bond counting from the terminal methyl (CH3) group. The term ‘‘essential fatty acids’’ originated from the discovery that absence of certain fatty acids in the diet leads to clinical abnormalities. The essentiality of LA was first described many years ago by Burr and Burr (1929). They observed that a highly purified fat-free diet caused scaliness in the skin, feet and tail, growth retardation, kidney malfunction and impaired reproduction in rats. This observation was initially attributed to a deficiency of a vitamin or other dietary component. Later, they confirmed that it was the common fatty acid, LA, that caused the deficiency (Burr and Burr 1930). Similar clinical symptoms of LA deficiency have been reported in humans (Hansen et al 1958, 1963; Collins et al 1971). In contrast to LA deficiency, animals, with the exception of fish, fed on a diet devoid of o3 fatty acids grow normally (Tinoco 1982). However, o3 fatty acid deficiency is associated with fatty liver, alopecia (Fiennes et al 1973) and severe behavioural disturbance in capuchins (Fiennes et al 1973), cerebellar degeneration in chickens (Budowski et al 1987), impaired visual acuity in rhesus monkeys (Connor et al 1984; Neuringer et al 1986) and poorer learning in rats (Lamptey and Walker 1976; Yamamoto et al 1987). These functional impairments in animals are primarily due to deficiency in docosahexaenoic acid (22:6o3, DHA), the major metabolite of ALA. Neurological abnormalities associated with o3 fatty acid deficiency have also been reported in humans (Holman et al 1982; Bjerve et al 1987).
DIETARY SOURCE FOR o6 AND o3 FATTY ACIDS Humans and mammals do not have enzymes which desaturate fatty acids at the D12 and D15 positions (Figure 22.1). Consequently, humans and mammals need to obtain LA and ALA from the diet. Linoleic acid is widely found in nuts (e.g. walnuts, peanuts, pistachios, almonds, pumpkin seeds) and seed oils (e.g. cotton, maize, sesame, peanut, sunflower, safflower, soybean) (Tinoco 1982). Evening primrose, borage and blackcurrant oils, which are rich in LA, also contain significant amounts of g-linolenic acid (18:3o6, GLA). Animal products (e.g. liver, kidney, etc.), eggs and tropical fish provide appreciable The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
amounts of arachidonic acid (20:4o6, AA), the major metabolite of LA (Pauletto et al 1996). Linseed, canola, soybean, perilla and walnut oils are typical sources of ALA (Tinoco 1982; Drevon 1992). The metabolites of ALA, namely eicosapentaenoic acid (20:5o3, EPA) and DHA, are abundant in fish, shellfish and marine plants such as diatoms, giant kelp and seaweeds (Hepburn et al 1986; Ackman 1989; Drevon 1992). The seaweeds are used in traditional foods in the Far East, Wales, Ireland and some other parts of coastal Europe. Human milk has been studied from various populations with different habitual diets and is shown to be a good source for longchain o6 and o3 polyunsaturated fatty acids (Crawford et al 1976; van Beusekom et al 1995). However, the levels of fatty acids are influenced by maternal diet (Finley et al 1985; Francois et al 1998).
DESATURATION AND ELONGATION OF LINOLEIC AND a-LINOLENIC ACIDS Desaturation and elongation of fatty acids occurs predominantly in the endoplasmic recticulum. This metabolic synthesis also takes place in the mitochondria and microsomes. Elongation involves condensation of acyl-CoA groups with malonyl-CoA. The o6 and o3 fatty acids use the same enzymes for elongation and desaturation (Figure 22.2) (Sprecher et al 1982). The first desaturation step of AA and DHA from LA and ALA requires D6 desaturase. This enzyme is rate limiting (Marcel et al 1968; Hassam et al 1977) and has a greater affinity for ALA than LA (Mohrhauer et al 1967). The final desaturation step for the synthesis of DHA requires the enzyme D6 desaturase. The elongated metabolite of 22:5o3, 24:5o3 is desaturated at position 6 (24:6o3), using the D6 desaturase, and then b-oxidized to DHA in the peroxisome (Hiltunen et al 1986; Voss et al 1991). Oleic acid (18:1o9), a non-essential fatty acid, uses the same desaturases as o6 and o3. If both LA and ALA intakes are inadequate, oleic acid is elongated and desaturated to Mead acid (20:3o9). Moreover, the ratio of Mead acid to AA increases, partly due to a fall in AA but mainly due to a rise in Mead acid (Innis 1991). Hence, Mead acid and the Mead acid/AA ratio (triene/tetraene) are used as biochemical markers for essential fatty acid deficiency (Holman 1977). Diet either deficient in o3 fatty acids or high in LA results in a reduction in DHA and increase in o6-docosatetraenoic (22:4o6) and docosapentaenoic (22:5o6) acids (Bourre et al 1992; Schiefermeier and Yavin 2002). The ratios of DHA to 22:5o6 (22:5o6/DHA) and 22:4o6 to 22:5o6 (22:5o6/22:4o6) are often used as indicators for DHA deficiency (Al et al 2000). The increased synthesis of o6-DPA is
258
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 22.1 Diagram of linoleic and a-linolenic acids
Figure 22.2 Synthesis of long-chain polyunsaturated fatty acids and formation of eicosanoids. D6, D6 desaturase; D5, D5 desaturase; E, elongase; b, boxidation; b*, peroxisomal b-oxidation; n, cyclooxygenase pathway; k, lipoxygenase pathway
thought to be a compensatory reaction to maintain membrane fluidity by keeping the overall lipid unsaturation level. Desaturation is perturbed in some pathological conditions. Diabetes is a well-known metabolic disorder characterized by impaired D6 and D5 desaturase activities, which causes a diminution of conversion to their long-chain metabolites (Holman
et al 1983; El Boustani et al 1989; Gordon et al 1995). A reduction in the plasma phospholipid AA and/or DHA levels due to absent or insufficient desaturase is also a common feature seen in Crohn’s disease (Clemmesen et al 2000), cirrhosis (Geerling et al 1999), cystic fibrosis (Strandvik et al 2001), Sjo¨gren–Larsson syndrome (Hernell et al 1982) and Zellweger’s syndrome (Martinez 1992).
ESSENTIAL FATTY ACIDS
259
Figure 22.3 Levels of linoleic (LA), arachidonic (AA) and docosahexaenoic (DHA) acids in the plasma choline phosphoglycerides (pCPG) and red cell choline (rCPG) and ethanolamine (rEPG) phosphoglycerides of Korean mothers (n=40) and their neonates (n=40) at birth
AA AND DHA IN THE PRE- AND POST-NATAL PERIODS Arachidonic acid is a major constituent of the endothelial cell membrane (Crawford et al 1997) and DHA is a key component of the photoreceptors (Anderson and Maude 1972; Anderson et al 1992) and synaptosomes (Suzuki et al 1997). During pregnancy, there is a high demand for AA and DHA for the development of the foetal brain and other vital tissues. Consequently, the accretion of long-chain o6 and o3 metabolites by the foetal brain increases progressively in the latter part of gestation (Clandinin et al 1980; Martinez 1992). Brain has a strong preference for preformed AA and DHA (Sinclair 1975; Cunnane et al 2001). The maternal–foetal gradient of o6 and o3 fatty acids has been consistently reported (Friedman et al 1978; Al et al 1995; Otto et al 1997; Zeijdner et al 1997; Min et al 2001). In cord plasma and foetal red cell phospholipids, the proportion of LA is only half that of the mother’s and ALA is almost non-detectable. In contrast, the levels of AA and DHA are higher in the foetus (Figure 22.3). This finding suggests that AA and DHA are selectively transferred from the maternal circulation, which may lead to a drain on the maternal membrane fatty acid store. Indeed, this was reflected in a significant decrease in the red cell phospholipid AA and DHA in healthy pregnant women (Figure 22.4; Ghebremeskel et al 2000). There is evidence that the foetus and neonates are capable of synthesizing AA and DHA from LA and ALA, respectively (Chambaz et al 1985; Poisson et al 1993). However, the key issue is that the rate of synthesis in relation to the growth demand is not fast enough to keep pace with the rapid foetal and neonatal growth and development (Clandinin et al 1980; Koletzko et al 1996). Indeed, in preterm infants Leaf et al (1992a) have observed
that the plasma AA is reduced to one-third, despite a three-fold increase in the precursor LA between birth and 3 weeks of age. Koletzko et al (1996), by the use of an isotope balance equation, have established that endogenous synthesis contributed only about 23% of total plasma AA in full-term babies. Subsequently, the foetus and neonates rely on maternal placental supply and milk, respectively, for these fatty acids. There is a growing consensus that non-communicable diseases such as coronary heart disease, hypertension, insulin resistance and non-insulin-dependent diabetes may have their origins in the foetal period (Langley-Evans 1996; Barker 1997). Deficit of AA and DHA during the prenatal period has profound implications for foetal vascular and neural development. Rats which were exposed to a high saturated fat diet during the intrauterine period have been associated with low levels of AA in the arteries and vascular dysfunction (Koukkou et al 1998; Ghosh et al 2001; Ozaki et al 2001). Inadequate DHA in the retina and brain was related to impaired visual function and reduced learning ability, respectively, in animals (Benolken et al 1973; Lamptey and Walker 1976; Neuringer et al 1984; Bourre et al 1986). There is also significant evidence in humans. Preterm babies are born with little fat store. As a result, they are more susceptible to deficiencies in AA and DHA. Martinez (1987) and Farquharson et al (1995) have reported that preterm babies fed on formula with no DHA had reduced retinal and brain DHA compared with full-term babies. Studies of maternal and cord blood phosphoglycerides showed that low concentrations of AA and DHA were associated with low birth weight and gestational age, respectively (Leaf et al 1992a, 1992b). In addition, low birth weight babies whose tissue level of DHA was low displayed poorer photoreceptor sensitivity and visual acuity (Uauy et al 1990; Birch et al 1992). Improved performance in the intelligence test in breast-fed preterm infants
260
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 22.4 Proportional difference of linoleic (LA), arachidonic (AA) and docosahexaenoic (DHA) acids in plasma choline phosphoglycerides (pCPG) and red cell choline (rCPG) and ethanolamine (rEPG) phosphoglycerides between non-pregnant (n=40) and pregnant Korean women (n=40)
has also been reported (Lucas et al 1992). Preterm infants fed breast milk scored higher (103.0 vs. 92.8) in the intelligence test at 7.5–8 years compared to those who received no maternal milk. The high score remained even after adjustment for the confounding factors, such as maternal education and social class. Better visual acuity and stereoacuity was also found in healthy term infants at 52 weeks who were weaned from breast-feeding at 6 weeks to formula that contained long-chain polyunsaturated fatty acids than did infants who were switched to formula without those fatty acids (Birch et al 2002). This result further emphasizes that dietary supply of AA and DHA can affect the maturation of cortical function beyond 6 weeks of age. On the contrary, Auestad et al (2001) claimed that adding AA and DHA had no effect on growth and visual development in term babies. However, their conclusion is questionable, as the amount of DHA (0.14%) they opted to add to the formula was the lowest level reported (range 0.4–0.9% AA, 0.1–0.9% DHA; Jensen 1999). By using 0.1% DHA, this dose is most likely to give a null effect. Birch et al (2002) used 0.36% DHA and 0.72% AA, values which are more consistent with reality. The discrepancy in the results between the two studies could simply be a dose–response factor. Nevertheless, the importance of AA and DHA in infant nutrition was recognized as early as the 1970s, as the WHO/FAO Expert Committee recommended adding preformed AA and DHA to formulas destined for preterm infants (FAO 1978). The same message was again put forward in 1994. A Success Story of DHA Treatment for Children with Zellweger’s Syndrome Zellweger’s syndrome is a congenital disease that is characterized by the absence of peroxisomes or a defect in peroxisomal b-
oxidation (Gordon et al 1995), which is the last step in the synthesis of DHA (Voss et al 1991). As a result, patients with Zellweger’s syndrome have a low concentration of blood DHA (Martinez 1992). Developmental delay and mental retardation is common to all these patients and there is subsequent loss of vision and hearing. The DHA treatment (DHA ethyl ester) initiated by Dr Marnuela Martinez has proved to be beneficial for patients with Zellweger’s syndrome; in patients treated with 100–600 mg/ day of DHA, levels of DHA in the plasma and erythrocyte phospholipids normalized and subsequently improved clinical symptoms such as liver function, floppiness and vision. Furthermore, magnetic resonance imaging revealed that remyelination had occurred in these patients (Martinez and Vazquez 1998). BIOLOGICAL FUNCTIONS OF ESSENTIAL FATTY ACIDS o6 and o3 Fatty Acids and Prostaglandins ‘‘Prostaglandin’’ is the collective name for unsaturated lipids synthesized from 20-carbon polyunsaturated fatty acids via cyclooxygenase metabolic pathways (Slater and McDonald-Gibson 1987). Each prostaglandin is designated by a letter, A–J, indicating the nature and position of substituents on the cyclopentane ring. A numerical subscript (1, 2 or 3) indicates the number of double bonds in the alkyl side chains. Subscript a or b is added for the PGF series to indicate the spatial position of the hydroxyl group in the cyclopentane ring. Prostaglandin activity was discovered at the same time as the essential fatty acids. In 1930, two American gynaecologists, R. Kurzrok and C. C. Lieb, reported that fresh semen provoked either strong contraction or relaxation of uterus in humans. Five
ESSENTIAL FATTY ACIDS years later, Ulf S. von Euler in Sweden and M. W. Goldblatt in England discovered a substance that influenced blood vessels and muscle tissues. This substance was named prostaglandin, as it was initially thought to be synthesized from the prostate gland. The purified form and structure of the first two prostaglandins, E1 and F1a, were determined by Sune Bergstro¨m and Jan Sjio¨vall in 1958. The initial identification of the series of different prostaglandins and unravelling of the biosynthetic pathway was done entirely by two research groups: those led by David van Dorp in Holland and by Sune Bergstro¨m and Bengt Samuelsson in Sweden. It was in the mid-1960s that the metabolic link between prostaglandins and essential fatty acids was finally brought to light. The prostaglandin 1, 2, and 3-series are synthesized from dihomo-g-linolenic acid (20:3o6, DGLA), AA and EPA, respectively. They stimulate various biological activities in platelets, vascular smooth muscle, bronchopulmonary function, gastrointestinal integrity and reproduction (Schettler 1986; Dutta-Roy 1994; Caserta et al 1998). Prostaglandin E1 was the most powerful anti-thrombotic agent known until PGI2 was discovered. Under stimulation, platelets produce PGE2 but a negligible amount of PGE1 (Lagarde et al 1981). This can be explained by the paucity of DGLA in the platelet phospholipids as compared to AA and specific phospholipase A2 for AA. However, without stimulation platelets produce more PGE1 than PGE2 (Lagarde et al 1981). This difference makes sense as under normal conditions platelets need to be protected against the buffeting between the other cells and vessel walls in circulation. However, in the face of trauma the immediate response is to aggregate and adhere to the damaged vessel wall so as to stop blood loss. During the metabolic conversion a very small amount of EPA is available for prostaglandin synthesis. This leaves AA as the major precursor for prostaglandins (Weber et al 1984). AA rapidly incorporates into membrane phospholipids and the concentration of free AA is very low in tissues (<1076 M; Horrobin 1983). The membrane AA is used for eicosanoid synthesis when membrane is attacked by phospholipase A2 to free AA (Slater and McDonaldGibson 1987). Alternatively, phospholipase C catalyses phospholipids, which forms diacylglycerol, and part of this is degraded by diacylglycerol lipase to yield AA. Subsequently, membrane phospholipids contribute AA from their reserves for prostaglandin production. However, overproduction of AA-derived prostaglandins may be partly responsible for atherothrombotic, inflammatory and autoimmune conditions.
Inositol Phosphoglycerides, Arachidonic Acid and Signalling Ionic fluxes across cellular membrane are important for muscle contraction, exocrine cell secretion and the generation of neuronal excitability in vertebrates. These fluxes are achieved through activation and deactivation of gating, in other words opening and closure of ion channels. Before the discovery of the phosphoinositide cycle, the inositol phosphoglycerides (IPGs) were considered to serve mainly as a component of the cell membrane. AA comprises half of the total fatty acids of IPGs and almost all of them are at the b-position. Therefore, it is tempting to suggest that IPG is the major source for AA. However, the release of AA from IPG is not straightforward. The first step in the breakdown of IPG is not the release of AA but the stripping of inositol, which is followed by the breakdown of phosphatidic acid, leaving diglyceride (Bell et al 1979; Lapetina et al 1981). Diacylglycerol itself acts as a second messenger by activating protein kinase C (PKC). Consequently, AA release is accomplished by diglyceride hydrolase, which has been reported in the platelets (RittenhouseSimmons 1979; Lapetina et al 1981). The chain of events leading to AA release is a simultaneous and rapid reaction. There is now
261
evidence of the specific function of AA, as with diglycerides, in the activation of PKC (Hindenes et al 2000). Lipid Rafts and Caveolae Lipid rafts are microdomains within the exoplasmic leaflet of the plasma membrane that are enriched with cholesterol and sphingolipids. This concept was proposed by Simons and Ikonen in 1997 from a series of studies on epithelial cell polarity. Sphingolipids are relatively non-polar and tend to aggregate with cholesterol, forming ‘‘rafts’’. These rafts are insoluble in non-ionic detergents, and therefore are often referred to as detergent-resistant membrane domains. The distribution of lipid rafts is dependent on cell types. In polarized cells, lipid rafts are mainly found in the apical plasma membrane, whilst they are abundant in the axonal membrane of neurons (Simons and Toomre 2000). Once lipid rafts are assembled in the Golgi, they move to the plasma membrane accompanying protein. The transport of protein is one of the important functions of lipid rafts. Proteins such as glycosylphophatidyl inositol-anchored proteins, doubly acylated peripheral membrane proteins, cholesterol-linked proteins and transmembrane proteins are delivered in this manner (Brown and London 1998). Once the delivery of protein to the surface of the membrane is completed, the rafts are either endocytosed or recycled (Puri et al 1999). Lipid rafts are also involved in signalling and the activation of immune responses (Langlet et al 2000). The immunosuppressive properties of o3 long-chain polyunsaturated fatty acids, particularly EPA, have long been recognized and used to treat inflammatory diseases (Belluzzi et al 1996). Evidence suggests that the inhibitory effect of EPA in T cell activation is through altering the lipid composition of the lipid rafts, rather than displacing the acylated proteins from the rafts (Stulnig et al 2001). Moreover, recent studies have demonstrated that influenza virus, bacteria and parasites use the rafts to populate the host cells (Scheiffele et al 1999; Wang et al 2000). The presence of caveolae was first described in the early 1950s. They are described as small flask-shaped invaginations in the plasma membrane, enriched with glycolipids, sphingolipids, cholesterol and caveolins. They were originally detected in endothelial cells but are now known to be present in most cells and are particularly rich in adipocytes, smooth muscle cells and epithelial cells (Couet et al 2001). The role of caveolae remained elusive until the discovery of caveolin, now called caveolin-1a, in the early 1990s (Ikonen and Parton 2000). Caveolins are highly expressed in the caveolae. At least three distinctive caveolins have been identified so far, including caveolin-1(a, b), -2, and -3. Caveolin-1 and -2 are abundant in adipocytes, endothelia, fibroblasts, epithelia, hepatocytes and some neuronal cells, while caveolin-3 is highly expressed in skeletal and cardiac muscle (Paron 1996). The proposed functions of caveolae and caveolin are transcytosis, potocytosis, signal transduction regulation and cholesterol transport (Murata et al 1995; Smart et al 1996; Massimino et al 2002). A number of reports have suggested that the active involvement of caveolae and caveolin in signalling might contribute to the aetiology of several diseases, such as muscular dystrophy, cancer, diabetes, atherosclerosis and Alzheimer’s disease (Campbell et al 2001). o3 Fatty Acids Regulation of Gene Expression Increasing evidence suggests that essential fatty acids, AA and DHA in particular, play an active role in gene expression. The results of our work consistently imply that AA and DHA are not simply membrane constituents. The specific connection of DHA,
262
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
for example, in cell signalling cannot be explained by liquidity (Bloom et al 1999). This specificity has occurred despite the fact that its immediate precursor, o3-docosapentaenoic acid (22:5o3), is more abundant in primary products used in the non-neural organs of many species (Crawford et al 1969). We have suggested that DHA has unique structural and signalling properties (Bloom et al 1999; Crawford et al 1999). The link between DHA and human cerebral expansion which we have proposed (Crawford et al 2001; Broadhurst et al 2002) almost certainly means that DHA is involved in the regulation of gene expression. Evidence in favour of this concept has emerged from the work of de Urquiza et al (2000), who have reported a role of DHA as a ligand for the retinoid X receptor (RXR) which is obligatory in the activation of the nuclear receptors. Dietary o3 long-chain metabolites suppress hepatic lipogenesis and triglyceride synthesis/secretion, while inducing peroxisomal and microsomal fatty acid oxidation (Ikeda et al 1998; Yoshida et al 1999). These effects on lipid metabolism are due to changes in gene expression by regulating the activity of transcription factors (Jump 2002). They include peroxisome proliferator activated receptor (PPARa,b,g), liver X receptors (LXRa,b), hepatic nuclear factor-4 (HNF-4a), sterol regulatory element binding proteins (SREBP1,2) and retinoid X receptors (RXR). These transcription factors play a major role in hepatic carbohydrate, fatty acid, triglyceride, cholesterol and bile acid metabolism. Therefore, ligands that specifically target RXR are emerging as potential pharmaceutical therapies for cancer, diabetes, and the lowering of circulatory cholesterol. The PPAR acts on regulation of lipid and lipoprotein metabolism and glucose homeostasis, and influences cellular proliferation, differentiation and apoptosis (Wolf 1998; Chinetti et al 2000). PPARa is highly expressed in the liver, muscle, kidney and heart, where it stimulates the b-oxidation of fatty acids. PPARg, which triggers adipocyte differentiation and promotes lipid storage, is predominantly expressed in intestine and adipose tissue. Activation of PPAR-mediated transcription is achieved through PPAR–RXR heterodimers, which bind DNA motifs called peroxisome proliferator response elements (PPRE), located in the promoters of target genes (Berger and Moller 2002). Several studies have demonstrated that EPA and DHA are ligands to activate PPARs (Devchand et al 1996; Ren et al 1997) and RXRs (de Urquiza et al 2000), respectively. Walt Schmidt at the USDA, who described the molecular dynamics of DHA (Crawford et al 1999), has commented on the three-dimensional structure of DHA as being like half of vitamin A, i.e. it could fit in the RXR activation site. In normal endothelial cells, retinoid concentration is barely detectable. However, the amount of DHA is large. It would swamp any trace amounts of retinoid present. Moreover, membrane DHA would be related to diet. Even in the brain, despite its strong protection against change, variations could be predicted, especially during embryonic, foetal and early neonatal development. It is therefore plausible that as far as normal systems are concerned the retinoid receptors are misnamed. They should perhaps be called DHA-receptors or DXR. CONCLUSION The concept that the essential fatty acids play a pivotal role in regulation of gene expression and even prenatal programming indicates a new paradigm in biomedical science research, food and agricultural policy. The outcome of the human genome project has made it evident that there are many genes which, like DHA in signalling sites, have been highly conserved. J Craig Ventor (founder President and CSO of Celera Genomics Group, Rockville, Maryland, USA) commented that there are also too few
genes to fully explain human behaviour, therefore it has to be that the environment has a major impact on regulation. Hence it is certainly not the genome that is changing, as is evident from the small 1.5% difference between humans and chimpanzees despite at least 5 million years of separation. It is absolutely clear from the change in shape, size and disease pattern in the last century, and more so from current change, with the sharp rise in mental ill-health amongst young people. We are witnessing a changing paradigm which is being loosely referred to as ‘‘post-genome’’. However, the environmental or nutritional impact, which has so far been neglected in biomedical research, must now come to the forefront. It is evident that the first point of contact between nutrition and the cell is the plasma membrane. The gene compositionally fixes the proteins in the membrane. However, the lipids are the aspect that is subject to change by diet. The concept that the lipid bilayer is a uniform barrier has now to be discarded. The bilayer is made up of different molecular species of lipids, each with their own specific functions, and domains whereby lipids are associated with the proteins, rafts or caveolae. These domains play a pivotal role in membrane protein function and signalling, with their properties influenced by diet and the fatty acids of the individual molecular species. It is now not difficult to conceive, from the evidence presented in this review, that the fatty acids are the key components that regulate gene expression, with the changes in dietary fats during the last century underlying the rise in mortality from the Western cluster of chronic diseases. The fatty acids are the major dietary variables in the membrane. Such variation can explain change and regulation of function, nuclear receptor activity and gene expression. The evidence of the causative role of adverse dietary fats in the aetiology of heart disease is just one pointer to the significance of the membrane lipids and essential fatty acid nutrition. The non-communicable diseases are now the major cause of mortality worldwide. They are responsible for more deaths than tuberculosis, gastrointestinal disease, malaria, AIDS and all other infectious diseases combined. The rise in mental ill-health amongst young people is a sinister new development in this changing panorama of disease. It is our view that the challenge to biomedical science, world food production and health lies in solving the true role of lipids and applying the message from the cell to policy.
REFERENCES Ackman RG (1989) Nutritional composition of fats in seafoods. Progr Food Nutr Sci, 13, 161–289. Al MDM, van Houwelingen AC, Kester ADM et al (1995) Maternal essential fatty acid patterns during normal pregnancy and their relationship to the neonatal essential fatty acid status. Br J Nutr, 74, 55–68. Al MD, van Houwelingen AC and Hornstra G (2000) Long-chain polyunsaturated fatty acids, pregnancy, and pregnancy outcome. Am J Clin Nutr, 71(suppl 1), 285S–291S. Anderson RE and Maude MB (1972) Lipids of ocular tissues: VII. The effects of essential fatty acid deficiency on the phospholipids of the photoreceptor membranes of rat retina. Arch Biochem Biophys, 151, 270–276. Anderson RE, O’Brien PJ, Weigand RD et al (1992) Conservation of docosahexaenoic acid in the retina. In Bazan NG, Murphy MG and Toffano G (eds), Advances in Experimental Medical Biology. vol 318, Neurobiology of Essential Fatty Acids. Plenum, New York, 285–294. Auestad N, Halter R, Hall RT et al (2001) Growth and development in term infants fed long-chain polyunsaturated fatty acids: a doublemasked, randomized, parallel, prospective, multivariate study. Pediatrics, 108, 372–381.
ESSENTIAL FATTY ACIDS Barker DJ (1997) Fetal nutrition and cardiovascular disease in later life. Br Med Bull, 53, 96–108. Bell RL, Kennerly DA, Stanford N and Majerus PW (1979) Diglyceride lipase: a pathway for arachidonate release from human platelets. Proc Natl Acad Sci USA, 76, 3238–3241. Belluzzi A, Brignola C, Campieri M et al (1996) Effect of an enteric-coated fish oil preparation on relapses in Crohn’s disease. N Engl J Med, 334, 1557–1560. Benolken RM, Anderson RE and Wheeler TG (1973) Membrane fatty acids associated with the electrical response in visual excitation. Science, 182, 1253–1254. Berger J and Moller DE (2002) The mechanisms of action of ppars. Ann Rev Med, 53, 409–435. Birch DG, Birch EE, Hoffman DR and Uauy RD (1992) Retinal development in very-low-birth-weight infants fed diets differing in o-3 fatty acids. Invest Ophthalmol Vis Sci, 33, 2365–2376. Birch EE, Hoffman DR, Castaneda YS et al (2002) A randomized controlled trial of long-chain polyunsaturated fatty acid supplementation of formula in term infants after weaning at 6 wk of age. Am J Clin Nutr, 75, 570–580. Bjerve KS, Fischer S and Alme K (1987) a-Linolenic acid deficiency in patients on long-term gastric tube feeding: estimation of linolenic acid and long-chain unsaturated n-3 fatty acid requirement in man. Am J Clin Nutr, 45, 66–77. Bloom M, Linseisen F, Lloyd-Smith J and Crawford MA (1999) Insight from NMR on the functional role of polyunsaturated lipids in the brain. In Magnetic Resonance and Brain Function—Approaches from Physics. Proceedings of the 1998 Enrico Fermi International School of Physics, Enrico Fermi Lecture, Course #139, Maraviglia B (ed), Varenna, Italy. Bourre JM, Dumont O, Piciotti M et al (1986) Polyunsaturated fatty acids of the n-3 series and nervous system development. In Galli C and Fedeli E (eds), Fat Production and Consumption (Technologies and Nutritional Implications), NATO ASI Series, Series A: Life Sciences, vol 131. New York: Plenum, 159–175. Bourre JM, Bonneil M, Chaudiere J et al (1992) Structural and functional importance of dietary polyunsaturated fatty acids in the nervous system. In Bazan NG (ed.), Neurobiology of Essential Fatty Acids. Plenum, New York, 211–229. Brenner RR (1971) The desaturation step in the animal biosynthesis of polyunsaturated fatty acids. Lipids, 6, 567–575. Broadhurst L, Wang Y, Crawford MA et al (2002) Brain-specific lipids from marine, lacustrine or terrestrial food resources: potential impact on early African Homo sapiens. Comp Biochem Physiol B Mol Biol, 131, 653–673. Brown DA and London E (1998) Functions of lipid rafts in biological membranes. Ann Rev Cell Dev Biol, 14, 111–136. Budowski P, Leighfield MJ and Crawford MA (1987) Nutritional encephalomalacia in the chick: an exposure of the vulunerable period for cerebellar development and the possible need for both o6- and o3fatty acids. Br J Nutr, 58, 511–520. Burr GO and Burr MM (1929) A new deficiency disease produced by the rigid exclusion of fat from the diet. J Biol Chem, 82, 345–367. Burr GO and Burr MM (1930) The nature and role of the fatty acids essential in nutrition. J Biol Chem, 86, 587–621. Campbell L, Gumbleton M and Ritchie K (2001) Caveolae and the caveolins in human disease. Adv Drug Deliv Rev, 49, 325–335. Caserta G, De Graaf OF and Di Renzo GC (1998) Mechanisms of initiation of human parturition. In Kurjak A (ed.), Textbook of Perinatal Medicine: A Comprehensive Guide to Modern Clinical Perinatology. New York: Parthenon, 1307–1330. Chambaz J, Ravel D, Manier MC et al (1985) Essential fatty acids interconversion in the human foetal liver. Biol Neonate, 47, 136–140. Chinetti G, Fruchart JC and Staels B (2000) Peroxisome proliferatoractivated receptors (PPARs): nuclear receptors at the crossroads between lipid metabolism and inflammation. Inflamm Res, 49, 497–505. Clandinin MT, Chappell JE, Leong S et al (1980) Intrauterine fatty acid accretion rates in humna brain: implications for fatty acid requirements. Early Hum Dev, 4, 121–129. Clemmesen JO, Hoy CE, Jeppesen PB and Ott P (2000) Plasma phospholipid fatty acid pattern in severe liver disease. J Hepatol, 32, 481–487. Collins FD, Sinclair AJ, Royle JP et al (1971) Plasma lipids in human linoleic acid deficiency. Nutr Metab, 13, 150–167.
263
Connor WE, Neuringer M, Barstad L and Lin DS (1984) Dietary deprivation of linolenic acid in rhesus monkeys: effects on plasma and tissue fatty acid composition and on visual function. Trans Assoc Am Physic, 97, 1–9. Couet J, Belanger MM, Roussel E and Drolet MC (2001) Cell biology of caveolae and caveolin. Adv Drug Deliv Rev, 49, 223–235. Crawford MA, Gale MM and Woodford MH (1969) Linoleic acid and linolenic acid elongation products in muscle tissue of Syncerus caffer and other ruminant species. Biochem J, 115, 25–27. Crawford MA, Hassam AG, Williams G and Whitehouse WL (1976) Essential fatty acids and foetal brain growth. Lancet, i, 452–453. Crawford MA, Costeloe K, Ghebremeskel K and Phylactos AC (1997) Are deficits of arachidonic and docosahexaenoic acids responsible fot the neural and vascular complications of preterm babies? Am J Clin Nutr, 66, 1032S–1041S. Crawford MA, Bloom M, Broadhurst CL et al (1999) Evidence for the unique function of DHA during the evolution of the modern hominid brain. Lipids, 34, S39–S47. Crawford MA, Bloom M, Cunnane S et al (2001) Docosahexaenoic acid and cerebral evolution. In Hamazaki T and Okuyama H (eds), Fatty Acids and Lipids—New Findings. World Review of Nutrition and Diet. Basel, Karger, 88, 6–17. Cunnane SC, Nadeau CR and Likhodii SS (2001) NMR and isotope ratio mass spectrometry studies of in vivo uptake and metabolism of polyunsaturates by the developing rat brain. J Mol Neurosci, 16, 173– 180. de Urquiza AM, Liu S, Sjoberg M et al (2000) Docosahexaenoic acid, a ligand for the retinoid X receptor in mouse brain. Science, 290, 2140– 2144. Devchand PR, Keller H, Peters JM et al (1996) The PPARa-leukotriene B4 pathway to inflammation control. Nature, 384, 39–43. Drevon CA (1992) Marine oils and their effects. Nutr Rev, 50, 38–45. Dutta-Roy AK (1994) Insulin mediated processes in platelets, erythrocytes and monocytes/macrophage: effects of essential fatty acid metabolism. Prostagland, Leukotrienes Essent Fatty Acids, 51, 385–399. El Boustani S, Caausse JE, Descomps B et al (1989) Direct in vivo characterization of D5 desaturase activity in humans by deuterium labeling: effect of insulin. Metabolism, 38, 315–321. FAO (1978) The role of dietary fats and oils in human nutrition. Rapporteur for a Joint FAO/WHO Expert Consultation, Nutrition Report No. 3. Farquharson J, Jamieson EC, Abbasi KA et al (1995) Effect of diet on the fatty acid composition of the major phospholipids of infant cerebral cortex. Arch Dis Child, 72, 198–203. Fiennes RNTW, Sinclair A and Crawford MA (1973) Essential fatty acid studies in primates. Linoleic acid requirements of capuchins. J Med Primates, 2, 155–169. Finley DA, Lonnerdal B, Dewey KG and Grivetti LE (1985) Breast milk composition: fat content and fatty acid composition in vegetarians and non-vegetarians. Am J Clin Nutr, 41, 787–800. Francois CA, Connor SL, Wander RC and Conner WE (1998) Acute effects of dietary fatty acids on the fatty acids of human milk. Am J Clin Nutr, 67, 301–308. Friedman Z, Danon A, Lamberth EL and Mann WJ (1978) Cord blood fatty acid composition in infants and in their mothers during the third trimester. J Pediatr, 92, 461–466. Geerling BJ, van Houwelingen AC, Badart-Smook A (1999) Fat intake and fatty acid profile in plasma phospholipids and adipose tissue in patients with Crohn’s disease, compared with controls. Am J Gastroenterol, 94, 410–417. Ghebremeskel K, Min Y, Crawford MA et al (2000) Blood fatty acid composition of pregnant and non-pregnant Korean women: red cells may act as a reservoir of arachidonic acid and docosahexaenoic acid for utilization by the developing foetus. Lipids, 35, 567–574. Ghosh P, Bitsanis D, Ghebremeskel K et al (2001) Abnormal aortic fatty acid composition and small artery function in offspring of rats fed a high fat diet in pregnancy. J Physiol, 533(3), 815–822. Gordon JA, Heller SK, Rhead WJ et al (1995) Formation of a novel arachidonic acid metabolite in peroxisomes. Prostagland Leukotrienes Essent Fatty Acids, 52, 77–81. Hansen AE, Haggard ME, Boelsche AN et al (1958) Essential fatty acids in infant nutrition. III. Clinical manifestations of linoleic acid deficiency. J Nutr, 66, 565–576.
264
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Hansen AE, Wiese HF, Boelsche AN et al (1963) Role of linoleic acid in infant nutrition; clinical and chemical study of 428 infants fed on milk mixtures varying in kind and amount of fat. Pediatrics, suppl(2), 171– 192. Hassam AG, Rivers JP and Crawford MA (1977) The failure of the cat to desaturate linoleic acid; its nutritional implications. Nutr Metab, 21, 321–328. Hepburn FN, Exler J and Weihrauch JL (1986) Provisional tables on the content of o-3 fatty acids and other fat components of selected foods. J Am Dietet Assoc, 86, 788–793. Hernell O, Holmgren G, Jagell SF et al (1982) Suspected faulty essential fatty acid metabolism in Sjogren–Larsson syndrome. Pediatr Res, 16, 45–49. Hiltunen JK, Karki T, Hassinen IE and Osmundsen H (1986) b-oxidation of polyunsaturated fatty acids by rat liver peroxisomes. A role for 2,4dienoyl-coenzyme A reductase in peroxisomal b-oxidation. J Biol Chem, 261, 16484–16493. Hindenes J-O, Nerdal W, Guo W et al (2000) Physical properties of the transmembrane signal molecule, sn-1-stearoyl 2-arachidonoylglycerol. J Biol Chem, 275, 6857–6867. Holman RT (1977) Essential fatty acids in human nutrition. Adv Exp Med Biol, 83, 515–534. Holman RT, Johnson SB and Hatch TF (1982) A case of human linolenic acid deficiency involving neurological abnormalities. Am J Clin Nutr, 35, 617–623. Holman RT, Johnson SB, Gerrard JM et al (1983) Arachidonic acid deficiency in streptozotocin-induced diabetes. Proc Natl Acad Sci USA, 80, 2375–2379. Horrobin DF (1983) The regulation of prostaglandin biosynthesis by the manipulation of essential fatty acid metabolism. Rev Pure Appl Sci, 4, 339–383. Ikeda I, Cha JY, Yanagita T et al (1998) Effects of dietary a-linolenic, eicosapentaenoic and docosahexaenoic acids on hepatic lipogenesis and b-oxidation in rats. Biosci, Biotechnol Biochem, 62, 675–680. Ikonen E and Parton RG (2000) Caveolins and cellular cholesterol balance. Traffic, 1, 212–217. Innis SM (1991) Essential fatty acids in growth and development. Progr Lipid Res, 30, 39–103. Jensen RG (1999) Lipids in human milk. Lipids, 34, 1243–1271. Jump DB (2002) Dietary polyunsaturated fatty acids and regulation of gene transcription. Curr Opin Lipidol, 13, 155–164. Koletzko B, Decsi T and Demmelmair H (1996) Arachidonic acid supply and metabolism in human infants born at full term. Lipids, 31, 79–83. Koukkou E, Ghosh P, Lowy C and Poston L (1998) Offspring of normal and diabetic rats fed saturated fat in pregnancy demonstrate vascular dysfunction. Circulation, 98, 2899–2904. Lagarde M, Guichardand M and Dechavanne M (1981) Human platelet PGE and dihomogammalinolenic acid. Comparison to PGE2 and arachidonic acid. Progr Lipid Res, 20, 439–443. Lamptey M and Walker BL (1976) A possible essential role for dietary linolenic acid in the development of the young rat. J Nutr, 106, 86–93. Langlet C, Bernard AM, Drevot P and He HT (2000) Membrane rafts and signaling by the multichain immune recognition receptors. Curr Opin Immunol, 12, 250–255. Langley-Evans SC (1996) Intrauterine programming of hypertension in the rat: nutrient interactions. Comp Biochem Physiol, 114, 327–333. Lapetina EG, Billah MM and Cuatrecasas P (1981) The initial action of thrombin on platelets. Conversion of phosphatidylinositol to phosphatidic acid preceding the production of arachidonic acid. J Biol Chem, 256, 5037–5040. Leaf AA, Leighfield MJ, Costeloe KL and Crawford MA (1992a) Factors affecting long-chain polyunsaturated fatty acid composition of plasma choline posphoglycerides in preterm infants. J Pediatr Gastroenterol Nutr, 14, 300–308. Leaf AA, Leighfield MJ, Costeloe KL and Crawford MA (1992b) Long chain polyunsaturated fatty acids and foetal growth. Early Hum Dev, 30, 183–191. Lucas A, Morley R, Cole TJ et al (1992) Breast milk and subsequent intelligence quotient in children born preterm. Lancet, 339, 261–264. Marcel YL, Christiansen K and Holman RT (1968) The preferred metabolic pathway from linoleic acid to arachidonic acid in vitro. Biochim Biophys Acta, 164, 25–34. Martinez M (1987) Dietary polyunsaturated fatty acids in relation to neural development in humans. In Galli C and Simopoulos AP (eds),
Dietary o-3 and o-6 Fatty Acids: Biological Effects and Nutritional Essentiality. Series A: Life Sciences, vol 171, 123–133. Martinez M (1992) Abnormal profiles of polyunsaturated fatty acids in the brain, liver, kidney and retina of patients with peroxisomal disorders. Brain Res, 583, 171–182. Martinez M and Vazquez E (1998) MRI evidence that docosahexaenoic acid ethyl ester improves myelination in generalized peroxisomal disorders. Neurology, 51, 26–32. Massimino ML, Griffoni C, Spisni E et al (2002) Involvement of caveolae and caveolae-like domains in signaling, cell survival and angiogenesis. Cell Signal, 14, 93–38. Min Y, Ghebremeskel K, Crawford MA et al (2001) Maternal–foetal n-6 and n-3 polyunsaturated fatty acids gradient in plasma and red cell phospholipids. Int J Vit Nutr Res, 71, 286–292. Mohrhauer H, Christensen K, Gan MV et al (1967) Chain elongation of linoleic acid and its inhibition by other fatty acids in vitro. J Biol Chem, 242, 4507–4514. Murata M, Peranen J, Schreiner R et al (1995) VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci USA, 92, 10339– 10343. Neuringer M, Connor WE, van Petten C and Barstad L (1984) Dietary o-3 fatty acid deficiency and visual loss in infant rhesus monkeys. J Clin Invest, 73, 272–276. Neuringer M, Connor WE, Lin DS et al (1986) Biochemical and functional effects of prenatal and postnatal o-3 fatty acid deficiency on retina and brain in rhesus monkeys. Proc Nat Acad Sci USA, 83, 4021–4025. Otto SJ, van Houwelingen AC, Antal M et al (1997) Maternal and neonatal essential fatty acid status in phospholipids: an international comparative study. Eur J Clin Nutr, 51, 232–242. Ozaki T, Nishina H, Hanson MA and Poston L (2001) Dietary restriction in pregnant rats causes gender-related hypertension and vascular dysfunction in offspring. J Physiol, 530(1), 141–152. Paron RG (1996) Caveolae and caveolins. Curr Opin Cell Biol, 8, 542–548. Pauletto P, Puato M, Caroli MG et al (1996) Blood pressure and atherogenic lipoprotein profiles of fish-diet and vegetarian villagers in Tanzania: the Lugalawa study. Lancet, 348, 784–788. Poisson JP, Dupuy RP, Sarda P et al (1993) Evidence that liver microsomes of human neonates desaturate essential fatty acids. Biochim Biophys Acta, 1167, 109–113. Puri V, Watanabe R, Dominguez M et al (1999) Cholesterol modulates membrane traffic along the endocytic pathway in sphingolipid-storage diseases. Nature Cell Biol, 1, 386–388. Ren B, Thelen AP, Peters JM et al (1997) Polyunsaturated fatty acid suppression of hepatic fatty acid synthase and S14 gene expression does not require peroxisome proliferator-activated receptor a. J Biol Chem, 272, 26827–26832. Rittenhouse-Simmons S (1979) Production of diglyceride from phosphatidylinositol in activated human platelets. J Clin Invest, 63, 580–587. Schiefermeier M and Yavin E (2002) n-3-Deficient and docosahexaenoic acid-enriched diets in critical periods of the developing prenatal rat brain. J Lipid Res, 43, 124–131. Scheiffele P, Rietveld A, Wilk T and Simon K (1999) Influenza viruses select ordered lipid domains during budding from the plasma membrane. J Biol Chem, 274, 2038–2044. Schettler G (1986) Relevance of fatty acids and eicosanoids to clinical and preventive medicine. Progr Lipid Res, 25, 1–4. Simons K and Ikonen E (1997) Functional rafts in cell membranes. Nature, 387, 569–572. Simons K and Toomre D (2000) Lipid rafts and signal transduction. Nature Rev Mol Cell Biol, 1, 31–39. Sinclair AJ (1975) Incorporation of radioactive polyunsaturated fatty acids into liver and brain of developing rat. Lipids, 10, 175–184. Slater TF and McDonald-Gibson RG (1987) Introduction to the eicosanoids. In Benedetto C, McDonald-Gibson RG, Nigam S and Slater TF (eds), Prostaglandins and Related Substances (A Practical Approach). Oxford: IRL Press, 1–4. Smart EJ, Ying YS, Donzell WC and Anderson RG (1996) A role for caveolin in transport of cholesterol from endoplasmic reticulum to plasma membrane. J Biol Chem, 271, 29427–29435. Sprecher H, van Rollins M, Sun F et al (1982) Dihomo-prostaglandin and thromboxanes: a novel prostaglandin family from adrenic acid that may specifically be synthesized in the kidney. J Biol Chem, 257, 3912.
ESSENTIAL FATTY ACIDS Strandvik B, Gronowitz E, Enlund F et al (2001) Essential fatty acid deficiency in relation to genotype in patients with cystic fibrosis. J Pediatr, 139, 650–655. Stubbs CD and Smith AD (1984) The modification of mammalian polyunsaturated fatty acid composition in relation to membrane fluidity and function. Biochim Biophys Acta, 779, 89–137. Stulnig TM, Hubert J, Leitinger N et al (2001) Polyunsaturated eicosapentaenoic acid displaces proteins from membrane rafts by altering raft lipid composition. J Biol Chem, 276, 37335–37340. Suzuki H, Manabe S, Wada O and Crawford MA (1997) Rapid incorporation of docosahexaenoic acid from dietary sources into brain microsomal, synaptosomal and mitochondrial membranes in adult mice. Int J Vit Nutr Res, 67, 272–278. Tinoco J (1982) Dietary requirements and functions of a-linolenic acid in animals. Progr Lipid Res, 21, 1–45. Uauy R (1990) Are o-3 fatty acids required for normal eye and brain development in the human? J Pediatr Gastroenterol Nutr, 11, 296–300. van Beusekom CM (1995) Lipids, Carbohydrates and Micronutrients in the Perinatal Period. PhD Thesis, University of Groningen, Amsterdam.
265
Voss A, Reinhart M, Sankarappa S and Sprecher S (1991) The metabolism of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,13,16,19docosahexaenoic acid in rat liver is independent of a 4-desaturase. J Biol Chem, 266, 19995–20000. Wang Y, Storlien LH, Jenkins AB et al (2000) Dietary variables and glucose tolerance in pregnancy. Diabet Care, 23, 460–464. Weber PC, Fischer S, Lorenz R et al (1984) Formation of eicosanoids after dietary long chain n-3 fatty acids in man. Proceedings of University of Reading Conference, 107–111. Wolf G (1998) Fatty acids bind directly to and activate peroxisome proliferator-activated receptors a and g. Nutr Rev, 56, 61–63. Yamamoto N, Saitoh M, Moriuchi A et al (1987) Effect of dietary alinolenate/linoleate balance on brain lipid compositions and learning ability of rats. J Lipid Res, 28, 144–151. Yoshida H, Mawatari M, Ikeda I et al (1999) Effect of dietary seal and fish oils on triacylglycerol metabolism in rats. J Nutr Sci Vitaminol, 45, 411–421. Zeijdner EE, van Houwelingen AC, Kester ADM and Hornstra G (1997) Essential fatty acid status in plasma phospholipids of mothers and neonates after multiple pregnancy. Prostagland Leukotrienes Essent Fatty Acids, 56, 395–401.
23 Endothelial Secretory Function and Atherothrombosis Stefan Chlopicki and Richard J. Gryglewski Jagiellonian University Medical College, Krako´w, Poland
Atherothrombosis is a major cause of ischaemic heart disease, stroke and peripheral vascular disease and, therefore, the leading cause of death. Despite decades of research, the mechanisms leading to the formation of atherosclerotic plaque and its progression are far from being understood. Various concepts have been put forward, some of them as early as in the nineteenth century. In 1848 Carl Rokitansky suggested intra-arterial clot formation to be a primary factor involved in development of atherosclerotic plaque (encrustation theory). A few years later Rudolf Virchow presented an opposing view, which says that the major culprit in development of atherosclerosis is the process of lipid infiltration into the vascular wall (infiltration theory), while intra-arterial clot formation is a secondary event. Virchow’s concept was supported experimentally by Anichkov’s elegant experiments, in which atherosclerosis was evoked by feeding rabbits with a high-cholesterol diet (Keaney 2000). Since then the view that atherosclerotic plaque results from accumulation of plasma lipoproteins, which progressively reduces the lumen of the vessels, has dominated cardiovascular medicine. Low-density lipoproteins (LDL) and later oxidized LDL were suspected to play a major role in foam cell formation and development of atherosclerotic plaque. Accordingly, it has been assumed for decades that lowering the level of cholesterol represents the mainstay of antiatherosclerotic treatment (Levine et al 1995; Libby, Aikawa and Schonbeck 2000). Recently, the classical view on atherosclerosis as cholesteroldriven pathology has given way to the understanding of atherosclerosis as a chronic inflammatory disease of the vascular wall. In this, vascular pathology lipid, thrombotic and inflammatory pathways intertwine in a complex network (Ross 1999a; Lusis 2000; Hansson 2001; Blake and Ridker 2001; Libby 2001; Libby and Simon 2001; Libby, Ridker and Maseri 2002). These mechanisms are far from being entirely understood. It is becoming clear, however, that LDL alone cannot be the major culprit of atherosclerosis. Paradoxically, the clinical success of inhibitors of 3-hydroxy 3methylglutaryl coenzyme A (HMG-CoA) reductase (statins) in the prevention of atherosclerosis (Maron, Fazio and Linton 2000; Fox 2001; Plutzky and Ridker 2001; Muntwyler and Gutzwiller 2002) offered strong support for that notion. In fact, statins introduced to the clinic as hypocholesterolaemic agents later emerged as potent antiatherosclerotic drugs, with properties related only partially to their LDL-cholesterol lowering effect. The vascular protection offered by statins seems to result in major part from their pleiotropic actions, including the reversal of endothelial dysfunction followed by antiinflammatory, antithrombotic and thrombolytic effects (Faggiotto and Paoletti 1999; Maron, Fazio and Linton 2000; Lefer, Scalia and Lefer 2001; Gryglewski et al 2001; Takemoto and Liao 2001).
The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
Indeed, there is increasing evidence that endothelial dysfunction plays a key role in the formation and progression of atherosclerotic plaque (Quyyumi 1998; Drexler 1999). Endothelial dysfunction has gained not only therapeutic but also diagnostic (Raitakari and Celermajer 2000b) and prognostic significance in atherosclerosis (Suwaidi et al 2000; Schachinger, Britten and Zeiher 2000; Heitzer et al 2001; Perticone et al 2001). Historically, Ross and Glomset, in their ‘‘response to injury’’ hypothesis of atherosclerosis, envisaged overt endothelial injury as responsible for local activation of platelets and release of growth factors, ultimately leading to formation of atherosclerotic lesion (Ross, Glomset and Harker 1977). When Bunting et al (1976) demonstrated prostacyclin synthesis in endothelium to be selectively impaired by lipid peroxides, the first possible mechanism was proposed linking functional alteration of endothelial integrity to development of atherosclerosis. It was suggested that a decreased capacity of endothelial cells to produce PGI2, a powerful antiplatelet mediator, would subsequently lead to enhanced production of TXA2 by platelets, and to platelet–vessel wall interactions that importantly contributed to the initiation of atherosclerotic plaque formation (Gryglewski 1980a, 1980b). The discovery of EDRF, which was later identified as nitric oxide (NO; Furchgott 1996), started a new era in the understanding of the antiatherosclerotic properties of endothelium. Since endothelial NO displayed potent antiadhesive activity towards leukocytes, along with antiproliferative, anti-thrombotic and vasculoprotective properties, this molecule emerged as the most important endothelium-derived body defender against the development of atherosclerosis (Drexler 1999; Kelm and Rath 2001; Napoli and Ignarro 2001). This notion is in line with the currently accepted view that leukocytes, but not platelets, play a key role in the initiation of plaque development, while platelets contribute to the late occlusive events triggered by plaque rupture (Lusis 2000; Libby 2001). Consequently PGI2, with its potent antiplatelet but weak antileukocyte properties (Gryglewski, Szczeklik and Wandzilak 1987; Hickey and Kubes 1997), was shifted to a remote position in the early antiatherosclerotic defence afforded by endothelium. In this chapter we review the role of endothelial PGI2 in regulating the healthy cardiovascular system, and we reevaluate the significance of PGI2 deficiency in the pathogenesis of atherosclerosis. In particular, we summarize recent evidence for the role of platelets in the formation of atherosclerotic plaque, as well as evidence for the contribution of PGI2-deficiency to platelet-dependent enhancement of the inflammatory process in the vascular wall. Finally, we present our perspective on the pharmacological correction of endothelial dysfunction in prevention and treatment of atherothrombosis.
268
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
BIOSYNTHESIS OF PGI2 AND TXA2 Prostanoids such as PGI2 and TXA2, as well as prostaglandins (PGE2, PGD2, PGF2a PGJ2), are generated upon cell activation, which may be induced by a variety of physiological and pathological stimuli. Various cell types may produce PGI2 and TXA2 (Ullrich, Zou and Bachschmid 2001); however, major cellular sources of these mediators in the circulation are vascular endothelium for PGI2 and blood platelets for TXA2. Three enzymes are involved in biosynthesis of PGI2 and TXA2: phospholipase A2 (PLA2), cyclooxygenase (COX) and then PGI2 or TXA2 synthases (PGI-S and TXA-S, respectively). After cleavage of arachidonic acid (AA) from membrane phospholipids by PLA2 (which may require previous activation by PLC), AA is converted by the cyclooxygenase function of PGH synthase (PGHS) to an unstable prostaglandin endoperoxide (PGG2). PGG2, by the peroxidase function of PGHS, is converted to PGH2, which finally is isomerized to PGI2 or TXA2 by PGI-S or TXA-S, respectively, or to prostaglandins by cell-specific isomerases. Details of how these enzymes are orchestrated to yield the final products remain largely unknown; however, structure and cellular localization, as well as enzymatic reactions for each of these enzymes, are well characterized (Funk 2001). Although several other phospolipases, such as iPLA2 and sPLA2, may also play a role in cleavage of arachidonic acid, Ca2+dependent cPLA2 appears to be the dominant phospholipase involved in arachidonic acid liberation for PGI2 and TXA2 biosynthesis (Fitzpatrick and Soberman 2001). There are two isoforms of COX: COX-1 and COX-2 (Vane, Bakhle and Botting 1998; Patrono, Patrignani and Garcia Rodriguez 2001; FitzGerald and Patrono 2001; Hinz and Brune 2002). Each isoform of this enzyme catalyses exactly the same two reactions: insertion of two oxygen molecules into arachidonic acid, leading to formation of prostaglandin endoperoxide PGG2 (cyclooxygenase activity), and the subsequent reduction of PGG2 to PGH2 (peroxidase activity). The important biological difference between these isoforms, encoded by separate genes, is that COX-1 is constitutively expressed in most tissues, whereas COX-2 is usually an inducible enzyme. Some time ago it was suggested that COX-2 is undetectable in physiological situations, unless induced by pathological stimuli, such as lipopolysaccharide or cytokines (TNFa, IL-1b or IFNg), as well as by growth factors. At the moment, however, the classic hypothesis stating that COX-1 and COX-2 provide, respectively, physiological and pathological sources of prostanoids has to be revised. Indeed, it is apparent that COX-1 may play a certain role in inflammation, while COX-2 may be expressed constitutively and possess an important regulatory role, e.g. in brain, kidney and, possibly, in vascular endothelium (Smith and Langenbach 2001; FitzGerald and Patrono 2001; Hinz and Brune 2002). Although it was well known that under certain conditions COX-2 might be induced in the endothelium, it was a dogma that in healthy endothelium COX-1, but not COX-2, determined PGI2 production (Vane, Bakhle and Botting 1998). However, not long ago this view was challenged, when it was reported that there was a pronounced expression of COX-2 in cultured endothelial cells exposed to laminar shear stress, which would suggest a physiological role of endothelial COX-2 products (Topper et al 1996). Moreover, it was claimed that COX-2 inhibitors profoundly suppressed endothelial PGI2 production in healthy volunteers, as assessed by a fall in the 2,3-dinor-6-ketoPGF1a level in urine (McAdam et al 1999). These observations have fuelled recent discussions on the importance of COX-2 in the biosynthesis of PGI2 by healthy endothelium, which has an important clinical context, as COX-2 selective inhibitors (coxibs) are increasingly in use (Patrono, Patrignani and Garcia Rodriguez 2001; FitzGerald and Patrono 2001).
It may well be that COX-2 provides an additional source for PGH2 and subsequent PGI2 synthesis in endothelium in pathological circumstances. However, the role of COX-2 in the generation of PGI2 by healthy endothelium does not appear to be convincingly documented. Our objections are as follows. Healthy endothelium of conduit vessels does not express COX-2, according to an elegant immunocytochemical analysis by Shonbeck et al (1999) and that of Baker et al (1999). This means that the presence of COX-2 described in vitro in human umbilical vein endothelial cells (HUVEC) may not be representative for arterial endothelium, and the presence of COX-2 in cultured endothelium from conduit-type vessels may be due to the induction of this enzyme by culturing conditions. Moreover, systemic vascular endothelium may not necessarily be the cellular source of 2,3-dinor-6-ketoPGF1a measured in urine as a marker of endothelial PGI2 synthesis as indicated by the case of acetaminophen, a weak COX inhibitor. To our knowledge it has not been demonstrated that acetaminophen affects COX-1 or COX-2 in systemic endothelium (Botting 2000). Nonetheless, ingestion of 500 mg of acetominophen was associated with a reduction of 2,3dinor-6-ketoPGF1a levels in urine (Green, Drvota and Vesterqvist 1989). Accordingly, one may assume that the major source of PGI2 in healthy endothelium is not COX-2 but rather COX-1 or another form of cyclooxygenase (Simmons et al 1999; Botting 2000). In fact, there exists clinical evidence against a major role of endothelial COX-2 in PGI2 production in humans. In controlled clinical trials with almost 28 000 patients, cardiovascular thrombotic events were not more frequent in a subgroup treated with rofecoxib (a selective COX-2 inhibitor) than in subgroups administered placebo, nabumeton or diclofenac (non-selective COX-1/COX-2 inhibitors). Fewer cardiovascular complications reported in patients treated with naproxen as compared with those given rofecoxib might have resulted from the known antiplatelet properties of naproxen (Konstam et al 2001). Notably, while PGI2 is a major COX-product of endothelium from conduit vessels, the endothelium of microvessels synthesizes predominantly PGE2 (Gerritsen and Printz 1981). The latter, brewed up by a recently identified inducible PGE synthase (PGES), is rather coupled to COX-2, and not to COX-1 (Jakobsson et al 1999). COX-2 also shows up in endothelium of the adventitia vessels (vasa vasorum) (Stemme et al 2000); however, neither its products nor their roles have been defined. In contrast with this ambiguity concerning the relative contribution of COX isoforms to endothelial PGI2 synthesis, it is clear that TXA2-synthase activity in platelets is functionally coupled to COX-1 only. The final step in the synthesis of TXA2 and PGI2 involves two isomerases of distinct intracellular localization, encoded by different genes (Ullrich, Zou and Bachschmid 2001). The PGI2 synthase in endothelium is associated with caveolae (Spisni et al 2001), while TXA2 synthase in platelets is a microsomal enzyme. Interestingly, during TXA2 synthesis almost equimolar amounts of two by-products of undetermined significance are formed (heptadecatrienoic acid, or HHT; and malondialdehyde, or MDA), while synthesis of PGI2 leads to formation of a single end-product. Despite this difference, PGI2 and TXA2 synthases show some biochemical similarities. In each case, the enzymatic reaction involves a haem-thiolate-dependent mechanism (Ullrich, Zou and Bachschmid 2001). Notably, the rate of formation of TXA2 exceeds that of PGI2 synthesis. It is suggested that this difference is compensated by a relatively high level of enzyme in the PGI2-synthesizing cells (Ullrich, Zou and Bachschmid 2001). Like other eicosanoids, both PGI2 and TXA2 are not stored within the cell but are immediately released to the extracellular space. It is unknown whether any transporting system is involved in their release (Pucci et al 1999), as it is in the case of some other eicosanoids, e.g. LTC4 (Keppler, Jedlitschky and Leier 1994).
ENDOTHELIAL SECRETORY FUNCTION AND ATHEROTHROMBOSIS Interestingly, in contrast with other prostanoids, which are enzymatically inactivated during a single passage through the lung (Bakhle 1990), both PGI2 and TXA2 break down spontaneously. They disintegrate into biologically inactive, stable products—6-keto-PGF1a, and TXB2, respectively (Bunting, Moncada and Vane 1983; Ullrich, Zou and Bachschmid 2001). Each of them may be further enzymatically transformed, e.g. to 2,3-dinor-6-ketoPGF1a or 11-dehydroTXB2, respectively (FitzGerald, Pedersen and Patrono 1983). Measurement of stable first products of spontaneous decomposition of PGI2 and TXA2 in serum (6-keto-PGF1a, TXB2), is widely used and offers a good possibility of tracing the fast appearance of biological activity of PGI2 and TXA2 in the circulation, next to application of drugs or experimental procedures. In biological fluids both PGI2 and TXA2 are unstable, with half-lives close to 4 min and 30 s, respectively, and this is why they are thought to act locally in an autocrine and paracrine fashion, near the site of their production. However, PGI2 released from pulmonary endothelium was proposed to act in an endocrine manner and to influence not too distant vascular beds (e.g. cerebral or coronary circulation; Gryglewski et al 1978b). PGI2 and TXA2 exert their biological actions via membrane IP and TP receptors, respectively (Maclouf, Folco and Patrono 1998; Narumiya and FitzGerald 2001; Funk 2001). Both types of receptors belong to the family of G-protein coupled receptors and typically possess seven transmembrane domains. Only one type of IP receptor is known today, but two splice variants of TP receptors, coupled to different G proteins, have been described (TPa, TPb). IP receptor stimulation triggers a rise in cAMP levels, while TP receptor stimulation is coupled to phospholipase C activation and calcium mobilization. Interestingly, PGI2 is the only endogenous agonist for IP receptors, whereas TP receptors also respond to PGH2 and isoprostanes (Ullrich, Zou and Bachschmid 2001). The latter are formed from arachidonic acid by non-enzymatic oxidative mechanisms. Indeed, isoprostane levels in serum are elevated in conditions of oxidative stress, such as ischaemia/reperfusion, hypercholesterolaemia and atherosclerosis (Patrono and FitzGerald 1997). Accordingly, it may well be that TP receptors, when responding to PGH2 and isoprostane stimulation, represent a gateway for transmitting signals of oxidative stress into cells. REGULATION OF CARDIOVASCULAR HOMEOSTASIS BY PGI2 AND TXA2 Although PGI2 and TXA2 released by other types of cells have been implicated in a variety of biological processes, regulation of cardiovascular homeostasis by platelet-derived TXA2 and endothelium-derived-PGI2 seems to represent the most important biological role for these mediators. There is an apparent physiological antagonism between these two mediators. PGI2 is a potent inhibitor of platelet aggregation and a vasodilator that inhibits also smooth muscle cell proliferation and migration. In contrast, TXA2 is a potent amplifier of platelet aggregation and a vasoconstrictor that stimulates vascular smooth muscle cell migration and proliferation (Moncada and Vane 1978; Gryglewski 1980a; Bunting, Moncada and Vane 1983; Jones, Benjamin and Linseman 1995; Ullrich, Zou and Bachschmid 2001). It is of biological importance that the regulatory system of TXA2–PGI2 involves several counter-regulation and feedback mechanisms which maintain the dynamic equilibrium between these two mediators. There are several possible mechanisms of reciprocal regulation between the activities of TXA2 and PGI2. For example, TXA2 released by activated platelets stimulates endothelium to release PGI2, which in turn inhibits platelet activity (Marcus
269
and Hajjar 1993). Moreover, transcellular transfer of PGH2 released from activated platelets provides a substrate for endothelial PGI2 biosynthesis (Gryglewski 1999). It seems that in both in vitro and in vivo experimental conditions, when activated platelets interact with endothelium, PGH2, but not arachidonic acid, is the major source of PGI2 synthesis (Gryglewski 1999). Also, in healthy volunteers, redirection of PGH2 metabolism from platelet to endothelium and subsequent increased synthesis of PGI2 are observed (Nowak and FitzGerald 1989). In the light of the above, one may assume that in some circumstances, when biosynthesis of TXA2 by platelets increases, healthy endothelium would accommodate to this situation by a concomitant increase in PGI2 production to maintain homeostatic balance and undisturbed function of cardiovascular system. Recent studies with genetically modified mice have clearly shown some consequences of lost equilibrium in homeostatic regulation between PGI2 and TXA2 in response to vascular injury. In IP receptor knockout mice (IP-KO), smooth muscle proliferative response to wire-induced vascular injury was augmented, along with a concomitant rise in TXA2 biosynthesis (FitzGerald et al 2000). However, when TP receptors were additionally deleted in these mice, proliferation of smooth muscle cells was abolished, suggesting that this vascular proliferative response was mediated by the unopposed action of TXA2 (FitzGerald et al 2000; FitzGerald, Cheng and Austin 2001). There are many clinical conditions, most importantly atherosclerosis (Gryglewski 1980a, 1980b; Gryglewski and Szczeklik 1983), which are associated with alterations in homeostatic balance between PGI2 and TXA2. DEFICIENCY OF ENDOTHELIAL PGI2 IN ATHEROSCLEROSIS In one of the original papers on the discovery of PGI2, it was demonstrated that 15-HPETE suppressed formation of PGI2 (Bunting et al 1976). Later it was shown that PGI2 synthesis by endothelium was also impaired by other lipid peroxides, including oxLDL (Szczeklik and Gryglewski 1980; Gryglewski and Szczeklik 1983). These observations were pivotal for putting forward a hypothesis that increased lipid peroxidation might promote development of atherosclerosis owing to the selective impairment of prostacyclin synthesis in endothelial cells and subsequent activation of platelets (Gryglewski 1980a, 1980b; Gryglewski and Szczeklik 1983). This concept was supported experimentally. It was demonstrated that in aorta and various arteries from rabbits fed with a high-cholesterol diet, production of PGI2 was substantially suppressed as early as 1–4 weeks after institution of atherogenic diet, and that this was associated with increased activation of platelets and enhanced TXA2 biosynthesis (Dembinska-Kiec et al 1977; Gryglewski et al 1978a; Gryglewski and Szczeklik 1983). Interestingly, suppression of PGI2 generation had been detected before any anatomical changes in arteries were visible. Impairment of PGI2 production in atherosclerotic vessels (D’Angelo et al 1978; Sinzinger et al 1979a, 1979b, 1980; Udvardy, Torok and Rak 1987; Kyrle et al 1989), as well as a concomitant increase in TXA2 biosynthesis by platelets (Patrono, Davi and Ciabattoni 1992), were also detected in humans. Other authors, however, postulated atherosclerosis to be associated with increased levels of PGI2 metabolites in urine (FitzGerald et al 1984). It was found that in various clinical conditions associated with platelet activation and increased TXA2 activity, e.g. in unstable angina, there would be a compensatory response of endothelium leading to an increase in PGI2 generation (Fitzgerald et al 1986). Interestingly, similar findings of a parallel increase in TXA2 and PGI2 biosynthesis by platelets and endothelium, respectively, were observed in two experimental mouse models of atherosclerosis (Pratico et al 2000). It has to be
270
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
emphasized, however, that in all of these studies PGI2 production by endothelium was assessed by measurements of 2,3-dinor-6ketoPGF1a in urine, but not of 6-ketoPGF1a in serum. The significance of 2,3-dinor-6-ketoPGF1a metabolites in urine still needs to be defined for clear-cut interpretation of these findings. Interestingly, in apoE mice, which develop atherosclerotic lesions in the aorta and at the branch point of the carotid and intercostal arteries, PGI2 synthase activity was not compromised in the coronary circulation, as assessed by 6-ketoPGF1a levels in coronary effluent (Godecke et al 2002). So, it may well be that PGI2 deficiency in atherosclerosis is confined to particular vessels or even to particular regions of atherosclerotic blood vessels. This remains to be determined. Impairment of PGI2 generation in atherosclerotic vessels is associated with a number of alterations in responsiveness and in the profile of prostanoids produced by the vascular wall, e.g. an increase in susceptibility of platelet adenylate cyclase and endothelial adenylate cyclase to exogenous prostacyclin was reported (Gryglewski 1980b). Importantly, since impairment of PGI2 synthesis coincides with increased expression of COX-2 in atherosclerotic vessels (Schonbeck et al 1999), excessive levels of PGH2, TXA2 and PGE2 may be generated instead of PGI2. Indeed, when COX-2 is induced in endothelium in vitro, transfer of PGH2 from activated endothelium to platelets and increased TXA2 formation by platelets may occur (Karim et al 1996). On the other hand, it is also suggested that endothelial COX-2 may be coupled to PGI2 synthase expressed in smooth muscle cells, which remains silent in physiological conditions but may transform excessive levels of COX-2-derived PGH2 to PGI2 (Ullrich, Zou and Bachschmid 2001). This may offer additional vascular protection. Finally, coupling of COX-2 to PGE2 within atherosclerotic plaque may lead to the progression of atherosclerosis. Indeed, overexpression of COX-2/PGE-S occurs in macrophages of human plaque and is associated with activation of metalloproteinases (MMP-2 and MMP-9), which are responsible for destabilization of atheromatic plaque (Cipollone et al 2001). The selective COX-2 inhibitor NS-398 prevents this unfavourable sequence of events (Cipollone et al 2001). It may seem that overexpression of COX-2/PGE-S in atherothrombotic arteries may be considered to be a general sign of arterial inflammatory response, characterized also by an increase in blood of endothelin1 (ET-1; Ihling et al 2001), C-reactive protein (hs-CRP; Chew et al 2001) and soluble cytoadhesins such as sVCAM-1, sICAM-1 and sE-selectin (Blankenberg et al 2001). Each of these signs of arterial inflammation including overexpression of COX-2/PGE-S constitutes signum omnis malis for patients with atherothrombosis and should be pharmacologically treated. As in atherosclerotic vessels, in vessels subjected to ballooninduced vessel injury, deficiency of PGI2 synthesis and induction of COX-2 coincide, leading to increased formation of PGH2, TXA2 and PGE2 (Yamada et al 2002). In this model, transfer of PGI2 synthase gene reversed the imbalance between COX-derived prostanoids. Decreased formation of PGE2, TXA2 and increased formation of PGI2 brought about by PGI2 synthase gene transfer led to the inhibition of neointimal formation induced by vascular injury (Yamada et al 2002). Whether this approach could be of therapeutic significance in atherosclerosis remains to be established. The original hypothesis formulated by Gryglewski in the late 1970s depicted lipid peroxides, in particular oxLDL, as killers of endothelial PGI2 synthesis (Gryglewski 1980a, 1980b; Gryglewski and Szczeklik 1983). Indeed, for some time oxidized LDL were considered to play a key role in development of atherosclerosis (Steinberg 1997). Potential mechanisms of atherogenic activity of oxLDL include stimulation of chemotaxis of monocytes and their adhesion to endothelium, stimulation of macrophage transformation to foam cells, and stimulation of smooth muscle proliferation and migration (Steinberg et al 1989).
Nowadays it is claimed that the oxidation of LDL involved in the formation of atherosclerotic plaque formation takes place within the artery wall (Diaz et al 1997). Accordingly, attention is turned to oxidative events within the vascular wall. A variety of markers of oxidative stress have been detected in human atherosclerotic lesions, including F2-isoprostanes and nitrotyrosine (Gaut and Heinecke 2001). Increasing evidence indicates that vascular NAD(P)H oxidase is a major source of oxidant stress within the vascular wall (Griendling, Sorescu and Ushio-Fukai 2000). It is postulated that NAD(P)H oxidase is involved in the oxidation of LDL within the vascular wall and in the development of atherosclerosis (Meyer and Schmitt 2000; Sorescu, Szocs and Griendling 2001). Furthermore, it was not long ago demonstrated that oxidation of LDL by endothelial cells in vitro could be attenuated by superoxide dismutase (SOD), suggesting the involvement of superoxide anions generated within the vascular wall in that process (Fang et al 1998). It is noteworthy that, in addition to NAD(P)H-oxidase, 12/15 lipoxygenase was also suggested to contribute to LDL peroxidation in vivo (Funk and Cyrus 2001). Moreover, other types of reactive species, such as HOCl7 formed from hydrogen peroxide by myeloperoxidase (MPO) of neutrophils, or peroxynitrite (ONOO7) formed in the reaction between NO and O7 2 , may also play a role (Gaut and Heinecke 2001). Thus, exact identities of oxidative species responsible for LDL oxidation and their sources are still not fully defined. In the light of these recent findings, it may be suggested that oxLDL is only a marker of the local oxidative stress within the vascular wall and sometimes its transmitter. Interestingly, increased oxidant stress within the vascular wall may lead to the impairment of prostacyclin formation directly by peroxynitrite, without participation of oxLDL. Indeed, it has been shown that exogenous peroxynitrite at suprisingly low concentrations (<0.1 mM) inhibited purified PGI2 synthase activity due to the nitration of a tyrosine residue in the active site of the enzyme (Zou, Martin and Ullrich 1997). PGI2 synthase was claimed to be the only protein known, at present to be nitrated at such low levels of exogenous peroxynitrite (Ullrich, Zou and Bachschmid 2001). Nitration of PGI2 synthase was detected in atherosclerotic vessels, and at the sites of vascular injury induced by hypoxia/reoxygenation. Thus, evoked PGI2 deficiency was associated either with impaired endothelium-dependent vasorelaxation or with augmented vessel vasoconstriction reversible with TP receptor antagonist (Zou, Leist and Ullrich 1999; Zou and Bachschmid 1999). Not long ago, the impairment of PGI2 synthesis in endothelial cells exposed to high glucose concentration was also described. In this setting, increase in the expression of adhesion molecules and a high rate of endothelial cells apoptosis were also abolished by TP receptors antagonist (Zou, Shi and Cohen 2002). Altogether, these data suggest that deficiency of PGI2 in endothelium may lead to the excessive stimulation of TP receptors in endothelium and vascular smooth muscle cells by TXA2, PGH2 or other eicosanoids, and to subsequent vasoconstriction, platelet aggregation and inflammatory response of the endothelium, as well as endothelial apoptosis. Thus, PGI2 deficiency and stimulation of the inflammatory response of endothelial cells by TP receptordependent mechanisms are closely linked. Furthermore, deficiency of PGI2 may trigger or enhance the vascular inflammatory process, which is now considered the key element of atherosclerosis. The above mechanisms seem to operate in two murine models of atherosclerosis. Inhibition of TXA2 biosynthesis retarded atherogenesis in LDL receptor knockout mice (Pratico et al 2001). Interestingly, a fall in TXA2 generation was associated with decreased levels of soluble intercellular adhesion molecule-1 (sICAM-1), as well as monocyte chemoattractant protein-1 (MCP-1), suggesting the antiatherosclerotic effect of platelet inhibition to be mediated by a decrease in the inflammatory response of the vascular wall (Cyrus et al 2001). Furthermore, in
ENDOTHELIAL SECRETORY FUNCTION AND ATHEROTHROMBOSIS ApoE-deficient mice atherosclerosis was delayed by TP receptor antagonist. Importantly, this effect was also associated with the reduction of inflammatory response, as evidenced by a decrease in levels of sICAM (Cayatte et al 2000). These results suggest TP receptor involvement in the progression of atherosclerosis, most likely via a stimulatory effect on the inflammatory process within the vascular wall. POSSIBLE ROLE OF PLATELETS IN EARLY INFLAMMATORY EVENTS OF ATHEROSCLEROSIS When Ross and Glomset formulated their ‘‘response to injury’’ hypothesis of atherosclerosis, they were pointing out that activated platelets in response to vascular injury would release mitogenic factors such as PDGF, triggering the proliferation and migration of smooth muscle cells into the atherosclerosis plaque (Ross, Glomset and Harker 1977). Later the role of platelets in early atherosclerosis was dismissed, mainly on the basis of their absence in the descriptive morphology of atherosclerotic plaques in humans (Davies 1996). Today, evidence indicates that activated platelets may influence plaque progression by a variety of mechanisms which may enhance inflammatory process within the vascular wall. Interaction between leukocytes and endothelial cells represents an initial event in the inflammatory response, which eventually leads to the development of atherosclerotic plaque (Ross 1999b; Lusis 2000; Libby 2001; Libby, Ridker and Maseri 2002). Indeed, there is good evidence suggesting that recruitment of mononuclear leukocytes through vascular leukocyte adhesion molecules and chemokines is involved in atherosclerosis. In murine models of atherosclerosis, genetic deletions of P-E-selectins, intercellular adhesion molecule-1 (ICAM-1; Collins et al 2000), vascular cell adhesion molecule-1 (VCAM-1; Ley and Huo 2001), as well as monocyte chemoattractant protein-1 (MCP-1; Gosling et al 1999) and its receptor CCR-2 (Boring et al 1998), each retard the development of atherosclerosis. Interestingly, platelets may also play a role in these early inflammatory events of atherosclerosis as they are involved in the recruitment of leukocytes and their trafficking across the vascular wall. Various potential mechanisms emerge, involving direct adhesion molecule-dependent bridging of platelets with endothelium or with leukocytes, as well as secretory products released by platelets. For example, P-selectin-dependent adhesion of platelets to monocytes induces the expression and secretion of key inflammatory products such as MCP-1 and IL-8 by monocytes (Weyrich et al 1996), while CD40 ligand on platelet surface triggers an inflammatory response in endothelial cells (Henn et al 1998). Also, activated platelets release proinflammatory cytokines, e.g. IL-1b or the chemokine RANTES (regulated upon activation, normal T cell expressed and secreted), resulting in endothelial cell activation and monocyte recruitment, respectively (Hawrylowicz, Howells and Feldmann 1991; von Hundelshausen et al 2001). Recent experimental data suggest that platelet-derived RANTES adhere to the surface of inflamed endothelium to trigger monocyte arrest and adhesion (von Hundelshausen et al 2001). Moreover, bioactive lipids shed in the form of microparticles from activated platelets can also importantly modulate the local environment (Barry et al 1997). These data confirm a hypothesis that blocking of the activity of platelets may offer an efficacious approach to the prevention of atherosclerosis progression. What could support a hypothesis that platelet activation contributes to atherosclerotic plaque formation in patients? It is well known that platelet-dependent vascular occlusion is considered to be the pathophysiological basis of final clinical events of atherothrombosis, such as acute coronary syndrome and ischae-
271
mic stroke. Support for this hypothesis comes from the efficacy of antiplatelet drugs such as aspirin (a COX-1 inhibitor), clopidogrel (a metabolite of which is supposed to be an ADP P2Y12 receptor antagonist) and GpIIb/IIIa antagonists (fibans) in the treatment of coronary and cerebrovascular disease (Patrono et al 2001). Much less is known, however, about the relevance of platelets to the development and progression of atherosclerotic plaques in humans. Apart from one observational study, which reported a beneficial effect of aspirin on atherosclerotic plaque progression (Ranke et al 1993), no other direct clinical data would confirm the involvement of platelets in atherosclerotic plaque progression. Interestingly, it was found in retrospective analysis that aspirin treatment was associated with a reduction of C reactive protein (CRP) blood levels in atherosclerotic, but not in healthy, individuals (Kennon et al 2001; Feldman et al 2001). CRP, which represents a systemic marker of inflammation, contributes to the vascular inflammatory process in atherosclerosis and may predict the risk of cardiovascular events (Blake and Ridker 2001; Chew et al 2001). It seems unlikely that this CRP-lowering effect of aspirin results from its direct antiinflammatory action. It may well be, however, that this is due to the inhibition of plateletdependent stimulation of the vascular inflammatory response. Also, treatment with other antiplatelet drugs goes along with suppression of blood markers of vascular inflammatory response (Ikonomidis et al 1999; Lincoff et al 2001). It must be mentioned, however, that interpretation of these data is blurred by the pleiotropic actions of antiplatelet drugs. For example, aspirin and salicylate show an antiinflammatory action that is associated with promotion of NF-kB decomposition (Yin, Yamamoto and Gaynor 1998), whereas clopidogrel and ticlopidine directly stimulate the vascular endothelium to produce PGI2 (Gryglewski et al 1996, 1999). As a matter of fact, thienopyridines may serve as basic chemical structures for development of even more potent stimulators of endothelial function (Gryglewski et al 2000). Despite these concerns, evidence, mostly from experimental studies, suggests that platelets are causally involved not only in the late occlusive events of atherothrombosis but also in the progression of inflammatory atherosclerotic lesions. Accordingly, the capacity of endothelium to inhibit platelet adhesion and activation seems to play an important part in the defence system afforded by endothelium against platelet-dependent activation of inflammatory processes within the vascular wall. Endogenous antiplatelet mediators derived from endothelium, particularly PGI2, the most powerful antiplatelet agent, as well as NO (Gryglewski 1994) and endothelial membrane-bound ADP-ase (CD39) (Marcus et al 1997), may be of great importance in that respect. ENDOTHELIAL DYSFUNCTION IN ATHEROSCLEROSIS AND ITS PHARMACOLOGICAL CORRECTION Abundant clinical evidence supports the notion that the development of atherothrombosis may be initiated or facilitated by the loss of endothelial vasculoprotective properties (Quyyumi 1998; Drexler 1999). However, it is difficult to associate the relative contribution of a single endothelial defect with the broad scope of endothelial dysfunction. Clinically, endothelial dysfunction is frequently diagnosed as an impairment of endothelial NO-dependent vasodilation (Raitakari and Celermajer 2000a, 2000b). Impairment of NO-dependent vasodilation seems to predict future cardiovascular events in patients with coronary artery disease or hypertension (Suwaidi et al 2000; Schachinger, Britten and Zeiher 2000; Heitzer et al 2001; Perticone et al 2001). Changes in the release of other endothelial secretogogues, such as PGI2 or t-PA, may be difficult
272
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
to measure in clinical practice, and this is why they are not frequently monitored in parallel to NO-dependent vasodilation. However, in patients with atherosclerosis there were reported: impairment of PGI2 synthesis (Sinzinger et al 1980; Udvardy, Torok and Rak 1987; Kyrle et al 1989); decreased t-PA activity (Newby et al 2001); and increased plasma levels of PAI-1 (JuhanVague et al 1989) and of von Willebrand factor (Schumacher et al 2002). One may presume that these patients also had impaired capacity to generate endothelial NO. Certainly, the impairment of NO-dependent function goes along with increased vascular production of superoxide anion (Heitzer et al 2001), with elevated plasma levels of endothelin-1 (Lerman et al 1995) as well as several markers of inflammation, such as cytokines (e.g. IL-6), soluble adhesion molecules (e.g. P-selectin or sICAM-1) and, finally, hs-CRP (Fichtlscherer et al 2000; Blake and Ridker 2001; Chew et al 2001; Blankenberg et al 2001). Some of the above alterations may be the consequence of NO-deficiency, some not. It is out of the scope of the present review to characterize in detail the alterations of endothelial function in atherosclerosis. However, it must be stressed that, at present, one cannot definitely distinguish between the relative contribution of NO and other endothelial mediators to the clinical consequences of endothelial dysfunction, such as development of atherothrombosis. Most likely functional alterations of numerous endothelial mediators are causally involved. Moreover, it seems that pharmacological substitution or correction of one endothelial mediator is insufficient to exert substantial clinical benefit in prevention and treatment of atherothrombosis. Indeed, L-arginine, a substrate for NO synthesis, and stable analogues of PGI2 (e.g. iloprost) might only alleviate some symptoms but do not bring unequivocal clinical benefit in treatment of atherothrombosis. On the contrary, the clinical effectiveness of HMG-CoA reductase inhibitors (statins) and angiotensin-converting enzyme inhibitors (ACE-I) is impressive in both the prevention and treatment of atherothrombosis (Maron, Fazio and Linton 2000; Fox 2001; Plutzky and Ridker 2001; Dagenais et al 2001; Muntwyler and Gutzwiller 2002). Paradoxically, in contrast with L-arginine or PGI2 analogues, statins (e.g. simvastatin, lovastatin, atorvastatin) and tissue-type ACE-I (e.g. quinapril, perindopril, ramipril)—the drugs introduced to medical practice to lower cholesterol level or blood pressure, respectively—possess a pronounced pleiotropic action on endothelium, which is not limited to correction of a single endothelial mediator (Faggiotto and Paoletti 1999; Maron, Fazio and Linton 2000; Lefer, Scalia and Lefer 2001; Takemoto and Liao 2001; Munzel and Keaney 2001). The pleiotropic effects of statins and ACE-I include antiinflammatory, antithrombotic, antidiabetic, antistroke, antiischaemic, vasoprotective or cardioprotective actions, reported in numerous clinical and basic research papers (Faggiotto and Paoletti 1999; Maron, Fazio and Linton 2000; Lefer, Scalia and Lefer 2001; Takemoto and Liao 2001; Dagenais et al 2001; Munzel and Keaney 2001; Wilson et al 2002; Chan et al 2002). Not long ago we (Gryglewski et al 2001) proposed a common pleiotropic endothelial link between the two classes of drugs. This link may explain unexpected and similar potential for both classes of drugs. In the case of statins, the molecular mechanism of their endothelial action is not clear. Statin-induced increase in NOS-3 (Laufs et al 1998), t-PA expression (Essig et al 1998), a decrease in expression of PAI-1 (Bourcier and Libby 2000) of endothelin-1 (Hernandez-Perera et al 2000) and AT1 receptors (Wassmann et al 2001b), as well as a decrease in the activity of NAD(P)H oxidase (Wagner et al 2000; Wassmann et al 2001a), were all proposed to be mediated by HMG-CoA reductase-dependent inhibition of prenylation of small signalling proteins. As we reported, an
immediate endothelial action of statins seems to be independent from their inhibitory effects on prenylation of signalling proteins. This is rather supposed to be initiated by a rise in [Ca2+]i in vascular endothelium and to involve synthesis and release of PGI2, NO or t-PA (Gryglewski et al 2001). In the case of ACE-I, the most likely mechanism for their pleiotropic endothelial action would be accumulation of bradykinin in the vicinity of endothelial B2 kinin receptors (Remme 1997; Heusch, Rose and Ehring 1997; Faggiotto and Paoletti 1999; Munzel and Keaney 2001). Following chronic administration of ACE-I, induction of endothelial B1 receptors may also take place (Marin-Castano et al 2002). On the other hand, inhibition of angiotensin II formation by ACE-I may also contribute to vasculoprotective and antiinflammatory action of these drugs, e.g. via decreased NAD(P)H oxidase expression and diminished O7 2 formation by vascular endothelium (Munzel and Keaney 2001; Nickenig and Harrison 2002). We claim that in Wistar rats in vivo thrombolytic action of statins and ACE-I is associated with the endothelial release of PGI2 but not of NO or t-PA (Gryglewski et al 2001). In that respect, we hardly observed any ‘‘class effect’’ for these drugs. For instance, thrombolytic and PGI2-releasing potencies in vivo for statins decreased in the order atorvastatin>simvastatin>>> cerivastatin, in a manner unrelated to their inhibitory potencies of HMG-CoA reductase. For ACE-I this order was quinapril= perindopril>>captopril. The drug-induced release of endogenous NO had rather a weak permissive effect on thrombolysis by statins or ACE-I, while the delayed appearance of t-PA antigen in blood feebly correlated with their thrombolytic potencies. The potency to induce thrombolysis in vivo by the most active ACE-I greatly surpassed that observed for statins. Effective doses were 30 mg/kg vs. 3 mg/kg, respectively (Gryglewski et al 2001). In clinical research, the pleiotropic actions of statins or ACE-I are usually monitored by assays of endothelial NO, such as acetylcholine- or flow-induced vasodilation (Mancini et al 1996; O’Driscoll, Green and Taylor 1997), or sometimes by markers of inflammation, e.g. CRP (Blake and Ridker 2001; Ridker et al 2001), by markers of thrombolysis, e.g. PAI-1 (Pahor et al 2002), or by markers of the coagulation cascade (Undas et al 2001). We would like to add to this list blood levels of 6-keto-PGF1a, a stable product of the spontaneous breakdown of endothelial PGI2. It proved to be an important functional marker of pleiotropic endothelial effects of statins and ACE-I, associated with duration and intensity of their thrombolytic action. In this fashion, the prostacyclin ring has closed again around the pathogenesis of atherothrombosis. In summary, the wide spectrum of endothelial actions of statins and ACE-I seems to offer a good justification for their clinical effectiveness. Although evidence is mostly indirect, it leaves little doubt that correction of endothelial function by statins and ACE-I contributes significantly to their clinical benefits, such as inhibition of progression of atherothrombosis and decrease in mortality from myocardial infarct and stroke. Importantly, it is achieved by correction of various functions of the endothelium, not only of the NOS-3–NO or COX–PGI2 pathways. The endothelial potential of these two classes of cardiovascular drugs undoubtedly has a strong impact on the current approach to the treatment of atherothrombosis. CONCLUSION In summary, we have witnessed a considerable evolution of concepts on the pathogenesis of atherothrombosis and on the role of endothelial dysfunction in it. From the era of PGI2-orientated research through an NO-orientated decade, we begin to appreciate the complexity of endothelial function and dysfunction. It has
ENDOTHELIAL SECRETORY FUNCTION AND ATHEROTHROMBOSIS become obvious that treatment of endothelial dysfunction targeted toward replacement of a single endothelial mediator is not likely to be successful, as shown by the limited clinical benefit of nitrites, L-arginine or PGI2 and its analogues. It may well be that successful pharmacological correction of endothelial dysfunction may be achieved by the removal of the detrimental influence of activated platelets and leukocytes and by the suppression of systemic inflammatory and prothrombotic reactions. However, major strategy should rely on stimulation of natural endothelial defensive mechanisms, which are mediated through NO, PGI2, t-PA, antioxidants (e.g. SOD) and antiapoptotic pathways. This should be associated with or complemented by suppression of excessive action of prostaglandin endoperoxide 7 (PGH2), superoxide anion (O7 2 ), peroxynitrite (ONOO ), plasminogen activator inhibitor (PAI-1), endothelin-1 (ET-1), various chemokines, cytokines and cytoadhesins. Only then one may expect that the endothelium will effectively display its defensive properties against atherothrombosis. The lesson learned from the clinical success of statins, ACE-I and thienopyridines should direct future endothelial research towards the development of new efficient endotheliotropic drugs for the prevention and treatment of atherothrombosis. REFERENCES Baker CS, Hall RJ, Evans TJ et al (1999) Cyclooxygenase-2 is widely expressed in atherosclerotic lesions affecting native and transplanted human coronary arteries and colocalizes with inducible nitric oxide synthase and nitrotyrosine particularly in macrophages. Arterioscler Thromb Vasc Biol, 19, 646–655. Bakhle YS (1990) Pharmacokinetic and metabolic properties of lung. Br J Anaesth, 65, 79–93. Barry OP, Pratico D, Lawson JA and FitzGerald GA (1997) Transcellular activation of platelets and endothelial cells by bioactive lipids in platelet microparticles. J Clin Invest, 99, 2118–2127. Blake GJ and Ridker PM (2001) Novel clinical markers of vascular wall inflammation. Circ Res, 89, 763–771. Blankenberg S, Rupprecht HJ, Bickel C et al (2001) Circulating cell adhesion molecules and death in patients with coronary artery disease. Circulation, 104, 1336–1342. Boring L, Gosling J, Cleary M and Charo IF (1998) Decreased lesion formation in CCR27/7 mice reveals a role for chemokines in the initiation of atherosclerosis. Nature, 394, 894–897. Botting RM (2000) Mechanism of action of acetaminophen: is there a cyclooxygenase 3? Clin Infect Dis, 31 (Suppl 5), S202–S210. Bourcier T and Libby P (2000) HMG CoA reductase inhibitors reduce plasminogen activator inhibitor-1 expression by human vascular smooth muscle and endothelial cells. Arterioscler Thromb Vasc Biol, 20, 556–562. Bunting S, Gryglewski R, Moncada S and Vane JR (1976) Arterial walls generate from prostaglandin endoperoxides a substance (prostaglandin X) which relaxes strips of mesenteric and coeliac ateries and inhibits platelet aggregation. Prostaglandins, 12, 897–913. Bunting S, Moncada S and Vane JR (1983) The prostacyclin– thromboxane A2 balance: pathophysiological and therapeutic implications. Br Med Bull, 39, 271–276. Cayatte AJ, Du Y, Oliver-Krasinski J et al (2000) The thromboxane receptor antagonist S18886 but not aspirin inhibits atherogenesis in apoE-deficient mice: evidence that eicosanoids other than thromboxane contribute to atherosclerosis. Arterioscler Thromb Vasc Biol, 20, 1724–1728. Chan AW, Bhatt DL, Chew DP et al (2002) Early and sustained survival benefit associated with statin therapy at the time of percutaneous coronary intervention. Circulation, 105, 691–696. Chew DP, Bhatt DL, Robbins MA et al (2001) Incremental prognostic value of elevated baseline C-reactive protein among established markers of risk in percutaneous coronary intervention. Circulation, 104, 992–997. Cipollone F, Prontera C, Pini B et al (2001) Overexpression of functionally coupled cyclooxygenase-2 and prostaglandin E synthase in
273
symptomatic atherosclerotic plaques as a basis of prostaglandin E(2)dependent plaque instability. Circulation, 104, 921–927. Collins RG, Velji R, Guevara NV et al (2000) P-Selectin or intercellular adhesion molecule (ICAM)-1 deficiency substantially protects against atherosclerosis in apolipoprotein E-deficient mice. J Exp Med, 191, 189–194. Cyrus T, Tang LX, Rokach J et al (2001) Lipid peroxidation and platelet activation in murine atherosclerosis. Circulation, 104, 1940–1945. D’Angelo V, Villa S, Mysliwiec M et al (1978) Defective fibrinolytic and prostacyclin-like activity in human atheromatous plaques. Thromb Haemost, 39, 535–536. Dagenais GR, Yusuf S, Bourassa MG et al (2001) Effects of ramipril on coronary events in high-risk persons: results of the Heart Outcomes Prevention Evaluation Study. Circulation, 104, 522–526. Davies MJ (1996) The contribution of thrombosis to the clinical expression of coronary atherosclerosis. Thromb Res, 82, 1–32. Dembinska-Kiec A, Gryglewska T, Zmuda, A and Gryglewski RJ (1977) The generation of prostacyclin by arteries and by the coronary vascular bed is reduced in experimental atherosclerosis in rabbits. Prostaglandins, 14, 1025–1034. Diaz MN, Frei B, Vita JA and Keaney JF (1997) Antioxidants and atherosclerotic heart disease. N Engl J Med, 337, 408–416. Drexler H (1999) Nitric oxide and coronary endothelial dysfunction in humans. Cardiovasc Res, 43, 572–579. Essig M, Nguyen G, Prie D et al (1998) 3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors increase fibrinolytic activity in rat aortic endothelial cells. Role of geranylgeranylation and Rho proteins. Circ Res, 83, 683–690. Faggiotto A and Paoletti R (1999) State-of-the-Art lecture. Statins and blockers of the renin–angiotensin system: vascular protection beyond their primary mode of action. Hypertension, 34, 987–996. Fang X, Weintraub NL, Rios CD et al (1998) Overexpression of human superoxide dismutase inhibits oxidation of low-density lipoprotein by endothelial cells. Circ Res, 82, 1289–1297. Feldman M, Jialal I, Devaraj S and Cryer B (2001) Effects of low-dose aspirin on serum C-reactive protein and thromboxane B2 concentrations: a placebo-controlled study using a highly sensitive Creactive protein assay. J Am Coll Cardiol, 37, 2036–2041. Fichtlscherer S, Rosenberger G, Walter DH et al (2000) Elevated Creactive protein levels and impaired endothelial vasoreactivity in patients with coronary artery disease. Circulation, 102, 1000–1006. Fitzgerald DJ, Roy L, Catella F and FitzGerald GA (1986) Platelet activation in unstable coronary disease. N Engl J Med, 315, 983–989. FitzGerald GA, Austin S, Egan K et al (2000) Cyclooxygenase products and atherothrombosis. Ann Med, 32, 21–26. FitzGerald GA, Cheng Y and Austin S (2001) COX-2 inhibitors and the cardiovascular system. Clin Exp Rheumatol, 19(6, suppl 25), S31–S36. FitzGerald GA and Patrono C (2001) The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med, 345, 433–442. FitzGerald GA, Pedersen AK and Patrono C (1983) Analysis of prostacyclin and thromboxane biosynthesis in cardiovascular disease. Circulation, 67, 1174–1177. FitzGerald GA, Smith B, Pedersen AK and Brash AR (1984) Increased prostacyclin biosynthesis in patients with severe atherosclerosis and platelet activation. N Engl J Med, 310, 1065–1068. Fitzpatrick FA and Soberman R (2001) Regulated formation of eicosanoids. J Clin Invest, 107, 1347–1351. Fox R (2001) American Heart Association 2001 scientific sessions: latebreaking science—statins: the new aspirin? Circulation, 104, E9051– E9052. Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science, 294, 1871–1875. Funk CD and Cyrus T (2001) 12/15-lipoxygenase, oxidative modification of LDL and atherogenesis. Trends Cardiovasc Med, 11, 116–124. Furchgott RF (1996) The 1996 Albert Lasker Medical Research Awards. The discovery of endothelium-derived relaxing factor and its importance in the identification of nitric oxide. J Am Med Assoc, 276, 1186–1188. Gaut JP and Heinecke JW (2001) Mechanisms for oxidizing low-density lipoprotein. Insights from patterns of oxidation products in the artery wall and from mouse models of atherosclerosis. Trends Cardiovasc Med, 11, 103–112.
274
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Gerritsen ME and Printz MP (1981) Sites of prostaglandin synthesis in the bovine heart and isolated bovine coronary microvessels. Circ Res, 49, 1152–1163. Godecke A, Ziegler M, Ding Z and Schrader J (2002) Endothelial dysfunction of coronary resistance vessels in apoE7/7 mice involves NO but not prostacyclin-dependent mechanisms. Cardiovasc Res, 53, 253–262. Gosling J, Slaymaker S, Gu L et al (1999) MCP-1 deficiency reduces susceptibility to atherosclerosis in mice that overexpress human apolipoprotein B. J Clin Invest, 103, 773–778. Green K, Drvota V and Vesterqvist O (1989) Pronounced reduction of in vivo prostacyclin synthesis in humans by acetaminophen (paracetamol). Prostaglandins, 37, 311–315. Griendling KK, Sorescu D and Ushio-Fukai M (2000) NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res, 86, 494–501. Gryglewski RJ (1980a) Prostaglandins, platelets, and atherosclerosis. CRC Crit Rev Biochem, 7, 291–338. Gryglewski RJ (1994) Prostacyclin and nitric oxide. Acta Haematol Pol, 25, 75–81. Gryglewski RJ (1999) Transcellular biosynthesis of eicosanoids. Pol J Pharmacol, 51, 113–117. Gryglewski RJ, Dembinska-Kiec A, Zmuda A and Gryglewska T (1978a) Prostacyclin and thromboxane A2 biosynthesis capacities of heart, arteries and platelets at various stages of experimental atherosclerosis in rabbits. Atherosclerosis, 31, 385–394. Gryglewski RJ, Dupin JP, Uracz W et al (2000) Thrombolysis by thienopyridines and their congeners. J Physiol Pharmacol, 51, 683–693. Gryglewski RJ, Korbut R, Ocetkiewicz A et al (1978b) Lungs as a generator of prostacyclin—hypothesis on physiological significance. Naunyn Schmiedebergs Arch Pharmacol, 304, 45–50. Gryglewski RJ, Korbut R, Swies J et al (1996) Thrombolytic action of ticlopidine: possible mechanisms. Eur J Pharmacol, 308, 61–67. Gryglewski RJ, Korbut R, Swies J and Uracz W (1999) Effect of ticlopidine on streptokinase-induced thrombolysis in rats. Wien Klin Wochenschr, 111, 98–102. Gryglewski RJ and Szczeklik A (1983) Prostacyclin and atherosclerosis— experimental and clinical approach. In Rheinisch Westfalische Akademie der Wissenschaften. Abhandlung Band 70, Second Munster Arteriorsclerosis Symposium: Clinical Implications of Recent Research Results in Arteriosclerosis, Westdeutscher Verlag, 213–226. Gryglewski RJ, Szczeklik A and Wandzilak M (1987) The effect of six prostaglandins, prostacyclin and iloprost on generation of superoxide anions by human polymorphonuclear leukocytes stimulated by zymosan or formyl-methionyl-leucyl-phenylalanine. Biochem Pharmacol, 36, 4209–4213. Gryglewski RJ, Uracz W, Swies J et al (2001) Comparison of endothelial pleiotropic actions of angiotensin converting enzyme inhibitors and statins. Ann NY Acad Sci, 947, 229–245. Gryglewski RJ (1980b) Prostacyclin and atherosclerosis. TIPS, 1, 164–168. Hansson GK (2001) Immune mechanisms in atherosclerosis. Arterioscler Thromb Vasc Biol, 21, 1876–1890. Hawrylowicz CM, Howells GL and Feldmann M (1991) Platelet-derived interleukin 1 induces human endothelial adhesion molecule expression and cytokine production. J Exp Med, 174, 785–790. Heitzer T, Schlinzig T, Krohn K et al (2001) Endothelial dysfunction, oxidative stress, and risk of cardiovascular events in patients with coronary artery disease. Circulation, 104, 2673–2678. Henn V, Slupsky JR, Grafe M et al (1998) CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells. Nature, 391, 591–594. Hernandez-Perera O, Perez-Sala D, Soria E and Lamas S (2000) Involvement of Rho GTPases in the transcriptional inhibition of preproendothelin-1 gene expression by simvastatin in vascular endothelial cells. Circ Res, 87, 616–622. Heusch G, Rose J and Ehring T (1997) Cardioprotection by ACE inhibitors in myocardial ischaemia/reperfusion. The importance of bradykinin. Drugs, 54, 31–41. Hickey MJ and Kubes P (1997) Role of nitric oxide in regulation of leucocyte-endothelial cell interactions. Exp Physiol, 82, 339–348. Hinz B and Brune K (2002) Cyclooxygenase-2—10 years later. J Pharmacol Exp Ther, 300, 367–375. Ihling C, Szombathy T, Bohrmann B et al (2001) Coexpression of endothelin-converting enzyme-1 and endothelin-1 in different stages of human atherosclerosis. Circulation, 104, 864–869.
Ikonomidis I, Andreotti F, Economou E et al (1999) Increased proinflammatory cytokines in patients with chronic stable angina and their reduction by aspirin. Circulation, 100, 793–798. Jakobsson PJ, Thoren S, Morgenstern R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96, 7220–7225. Jones DA, Benjamin CW and Linseman DA (1995) Activation of thromboxane and prostacyclin receptors elicits opposing effects on vascular smooth muscle cell growth and mitogen-activated protein kinase signaling cascades. Mol Pharmacol, 48, 890–896. Juhan-Vague I, Alessi MC, Joly P et al (1989) Plasma plasminogen activator inhibitor-1 in angina pectoris. Influence of plasma insulin and acute-phase response. Arteriosclerosis, 9, 362–367. Karim S, Habib A, Levy-Toledano S and Maclouf J (1996) Cyclooxygenase-1 and -2 of endothelial cells utilize exogenous or endogenous arachidonic acid for transcellular production of thromboxane. J Biol Chem, 271, 12042–12048. Keaney JF Jr (2000) Atherosclerosis: from lesion formation to plaque activation and endothelial dysfunction. Mol Aspects Med, 21, 99–166. Kelm M and Rath J (2001) Endothelial dysfunction in human coronary circulation: relevance of the L-arginine–NO pathway. Basic Res Cardiol, 96, 107–127. Kennon S, Price CP, Mills PG et al (2001) The effect of aspirin on Creactive protein as a marker of risk in unstable angina. J Am Coll Cardiol, 37, 1266–1270. Keppler D, Jedlitschky G and Leier I (1994) Transport and metabolism of leukotrienes. Adv Prostagland Thromboxane Leukotriene Res, 22, 83– 89. Konstam MA, Weir MR, Reicin A et al (2001) Cardiovascular thrombotic events in controlled, clinical trials of rofecoxib. Circulation, 104, 2280– 2288. Kyrle PA, Minar E, Brenner B et al (1989) Thromboxane A2 and prostacyclin generation in the microvasculature of patients with atherosclerosis—effect of low-dose aspirin. Thromb Haemost, 61, 374–377. Laufs U, La FV, Plutzky J and Liao JK (1998) Upregulation of endothelial nitric oxide synthase by HMG CoA reductase inhibitors. Circulation, 97, 1129–1135. Lefer AM, Scalia R and Lefer DJ (2001) Vascular effects of HMG CoAreductase inhibitors (statins) unrelated to cholesterol lowering: new concepts for cardiovascular disease. Cardiovasc Res, 49, 281–287. Lerman A, Holmes DR Jr, Bell MR et al (1995) Endothelin in coronary endothelial dysfunction and early atherosclerosis in humans. Circulation, 92, 2426–2431. Levine GN, Keaney JF Jr and Vita JA (1995) Cholesterol reduction in cardiovascular disease. Clinical benefits and possible mechanisms. N Engl J Med, 332, 512–521. Ley K and Huo Y (2001) VCAM-1 is critical in atherosclerosis. J Clin Invest, 107, 1209–1210. Libby P (2001) Current concepts of the pathogenesis of the acute coronary syndromes. Circulation, 104, 365–372. Libby P, Aikawa M and Schonbeck U (2000) Cholesterol and atherosclerosis. Biochim Biophys Acta, 1529, 299–309. Libby P, Ridker PM and Maseri A (2002) Inflammation and atherosclerosis. Circulation, 105, 1135–1143. Libby P and Simon DI (2001) Inflammation and thrombosis: the clot thickens. Circulation, 103, 1718–1720. Lincoff AM, Kereiakes DJ, Mascelli MA et al (2001) Abciximab suppresses the rise in levels of circulating inflammatory markers after percutaneous coronary revascularization. Circulation, 104, 163–167. Lusis AJ (2000) Atherosclerosis. Nature, 407, 233–241. Maclouf J, Folco G and Patrono C (1998) Eicosanoids and isoeicosanoids: constitutive, inducible and transcellular biosynthesis in vascular disease. Thromb Haemost, 79, 691–705. Mancini GB, Henry GC, Macaya C et al (1996) Angiotensin-converting enzyme inhibition with quinapril improves endothelial vasomotor dysfunction in patients with coronary artery disease. The TREND (Trial on Reversing ENdothelial Dysfunction) Study. Circulation, 94, 258–265. Marcus AJ, Broekman MJ, Drosopoulos JH et al (1997) The endothelial cell ecto-ADPase responsible for inhibition of platelet function is CD39. J Clin Invest, 99, 1351–1360.
ENDOTHELIAL SECRETORY FUNCTION AND ATHEROTHROMBOSIS Marcus AJ and Hajjar DP (1993) Vascular transcellular signaling. J Lipid Res, 34, 2017–2031. Marin-Castano ME, Schanstra JP, Neau E et al (2002) Induction of functional bradykinin b-(1)-receptors in normotensive rats and mice under chronic angiotensin-converting enzyme inhibitor treatment. Circulation, 105, 627–632. Maron DJ, Fazio S and Linton MF (2000) Current perspectives on statins. Circulation, 101, 207–213. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. Meyer JW and Schmitt ME (2000) A central role for the endothelial NADPH oxidase in atherosclerosis. FEBS Lett, 472, 1–4. Moncada S and Vane JR (1978) Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2, and prostacyclin. Pharmacol Rev, 30, 293–331. Muntwyler J and Gutzwiller F (2002) Statins and mortality. Eur Heart J, 23, 183–184. Munzel T and Keaney JF Jr (2001) Are ACE inhibitors a ‘‘magic bullet’’ against oxidative stress? Circulation, 104, 1571–1574. Napoli C and Ignarro LJ (2001) Nitric oxide and atherosclerosis. Nitric Oxide, 5, 88–97. Narumiya S and FitzGerald GA (2001) Genetic and pharmacological analysis of prostanoid receptor function. J Clin Invest, 108, 25–30. Newby DE, McLeod AL, Uren NG et al (2001) Impaired coronary tissue plasminogen activator release is associated with coronary atherosclerosis and cigarette smoking: direct link between endothelial dysfunction and atherothrombosis. Circulation, 103, 1936–1941. Nickenig G and Harrison DG (2002) The AT(1)-type angiotensin receptor in oxidative stress and atherogenesis: part I: oxidative stress and atherogenesis. Circulation, 105, 393–396. Nowak J and FitzGerald GA (1989) Redirection of prostaglandin endoperoxide metabolism at the platelet–vascular interface in man. J Clin Invest, 83, 380–385. O’Driscoll G, Green D and Taylor RR (1997) Simvastatin, an HMGcoenzyme A reductase inhibitor, improves endothelial function within 1 month. Circulation, 95, 1126–1131. Pahor M, Franse LV, Deitcher SR et al (2002) Fosinopril vs. amylodipine comparative treatments study: a randomized trial to assess effects on plasminogen activator inhibitor-1. Circulation, 105, 457–461. Patrono C, Coller B, Dalen JE et al (2001) Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest, 119, 39S–63S. Patrono C, Davi G and Ciabattoni G (1992) Thromboxane biosynthesis and metabolism in relation to cardiovascular risk factors. Agents Act Suppl, 37, 10–17. Patrono C and FitzGerald GA (1997) Isoprostanes: potential markers of oxidant stress in atherothrombotic disease. Arterioscler Thromb Vasc Biol, 17, 2309–2315. Patrono C, Patrignani P and Garcia Rodriguez LA (2001) Cyclooxygenase-selective inhibition of prostanoid formation: transducing biochemical selectivity into clinical read-outs. J Clin Invest, 108, 7–13. Perticone F, Ceravolo R, Pujia A et al (2001) Prognostic significance of endothelial dysfunction in hypertensive patients. Circulation, 104, 191– 196. Plutzky J and Ridker PM (2001) Statins for stroke: the second story? Circulation, 103, 348–350. Pratico D, Cyrus T, Li H and FitzGerald GA (2000) Endogenous biosynthesis of thromboxane and prostacyclin in 2 distinct murine models of atherosclerosis. Blood, 96, 3823–3826. Pratico D, Tillmann C, Zhang ZB et al (2001) Acceleration of atherogenesis by COX-1-dependent prostanoid formation in low density lipoprotein receptor knockout mice. Proc Natl Acad Sci USA, 98, 3358–3363. Pucci ML, Bao Y, Chan B et al (1999) Cloning of mouse prostaglandin transporter PGT cDNA: species-specific substrate affinities. Am J Physiol, 277, R734–R741. Quyyumi AA (1998) Endothelial function in health and disease: new insights into the genesis of cardiovascular disease. Am J Med, 105, 32S–39S. Raitakari OT and Celermajer DS (2000a) Flow-mediated dilatation. Br J Clin Pharmacol, 50, 397–404.
275
Raitakari OT and Celermajer DS (2000b) Testing for endothelial dysfunction. Ann Med, 32, 293–304. Ranke C, Hecker H, Creutzig A and Alexander K (1993) Dose-dependent effect of aspirin on carotid atherosclerosis. Circulation, 87, 1873–1879. Remme WJ (1997) Bradykinin-mediated cardiovascular protective actions of ACE inhibitors. A new dimension in anti-ischaemic therapy? Drugs, 54, 59–70. Ridker PM, Rifai N, Clearfield M et al (2001) Measurement of C-reactive protein for the targeting of statin therapy in the primary prevention of acute coronary events. N Engl J Med, 344, 1959–1965. Ross R (1999a) Atherosclerosis—an inflammatory disease. N Engl J Med, 340, 115–126. Ross R (1999b) Atherosclerosis is an inflammatory disease. Am Heart J, 138, S419–S420. Ross R, Glomset J and Harker L (1977) Response to injury and atherogenesis. Am J Pathol, 86, 675–684. Schachinger V, Britten MB and Zeiher AM (2000) Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation, 101, 1899–1906. Schonbeck U, Sukhova GK, Graber P et al (1999) Augmented expression of cyclooxygenase-2 in human atherosclerotic lesions. Am J Pathol, 155, 1281–1291. Schumacher A, Seljeflot I, Sommervoll L et al (2002) Increased levels of endothelial haemostatic markers in patients with coronary heart disease. Thromb Res, 105, 25–31. Simmons DL, Botting RM, Robertson PM et al (1999) Induction of an acetaminophen-sensitive cyclooxygenase with reduced sensitivity to nonsteroid antiinflammatory drugs. Proc Natl Acad Sci USA, 96, 3275–3280. Sinzinger H, Clopath P, Silberbauer K and Auerswald W (1979a) Prostacyclin synthesis in human and experimental atherosclerosis. Atherosclerosis, 34, 345–347. Sinzinger H, Kaliman J, Silberbauer K and Oppolzer R (1980) Diminished prostacyclin formation in forearm venous tissue of patients with peripheral occlusive arteriopathy. Exp Pathol (Jena), 18, 254–255. Sinzinger H, Silberbauer K, Feigl W et al (1979b) Prostacyclin activity is diminished in different types of morphologically controlled human atherosclerotic lesions. Thromb Haemost, 42, 803–804. Smith WL and Langenbach R (2001) Why there are two cyclooxygenase isozymes. J Clin Invest, 107, 1491–1495. Sorescu D, Szocs K and Griendling KK (2001) NAD(P)H oxidases and their relevance to atherosclerosis. Trends Cardiovasc Med, 11, 124–131. Spisni E, Griffoni C, Santi S et al (2001) Colocalization prostacyclin (PGI2) synthase–caveolin-1 in endothelial cells and new roles for PGI2 in angiogenesis. Exp Cell Res, 266, 31–43. Steinberg D (1997) Lewis A. Conner Memorial Lecture. Oxidative modification of LDL and atherogenesis. Circulation, 95, 1062–1071. Steinberg D, Parthasarathy S, Carew TE et al (1989) Beyond cholesterol. Modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med, 320, 915–924. Stemme V, Swedenborg J, Claesson H and Hansson GK (2000) Expression of cyclooxygenase-2 in human atherosclerotic carotid arteries. Eur J Vasc Endovasc Surg, 20, 146–152. Suwaidi JA, Hamasaki S, Higano ST et al (2000) Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation, 101, 948–954. Szczeklik A and Gryglewski RJ (1980) Low density lipoproteins (LDL) are carriers for lipid peroxides and inhibit prostacyclin (PGI2) biosynthesis in arteries. Artery, 7, 488–495. Takemoto M and Liao JK (2001) Pleiotropic effects of 3-hydroxy-3methylglutaryl coenzyme a reductase inhibitors. Arterioscler Thromb Vasc Biol, 21, 1712–1719. Topper JN, Cai J, Falb D and Gimbrone MA Jr (1996) Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese superoxide dismutase, and endothelial cell nitric oxide synthase are selectively upregulated by steady laminar shear stress. Proc Natl Acad Sci USA, 93, 10417–10422. Udvardy M, Torok I and Rak K (1987) Plasma thromboxane and prostacyclin metabolite ratio in atherosclerosis and diabetes mellitus. Thromb Res, 47, 479–484. Ullrich V, Zou MH and Bachschmid M (2001) New physiological and pathophysiological aspects on the thromboxane A(2)-prostacyclin regulatory system. Biochim Biophys Acta, 1532, 1–14.
276
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Undas A, Brummel KE, Musial J et al (2001) Simvastatin depresses blood clotting by inhibiting activation of prothrombin, factor V, and factor XIII and by enhancing factor Va inactivation. Circulation, 103, 2248– 2253. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenases 1 and 2. Ann Rev Pharmacol Toxicol, 38, 97–120. von Hundelshausen P, Weber KS, Huo Y et al (2001) RANTES deposition by platelets triggers monocyte arrest on inflamed and atherosclerotic endothelium. Circulation, 103, 1772–1777. Wagner AH, Kohler T, Ruckschloss U et al (2000) Improvement of nitric oxide-dependent vasodilatation by HMG-CoA reductase inhibitors through attenuation of endothelial superoxide anion formation. Arterioscler Thromb Vasc Biol, 20, 61–69. Wassmann S, Laufs U, Baumer AT et al (2001a) HMG–CoA reductase inhibitors improve endothelial dysfunction in normocholesterolemic hypertension via reduced production of reactive oxygen species. Hypertension, 37, 1450–1457. Wassmann S, Laufs U, Baumer AT et al (2001b) Inhibition of geranylgeranylation reduces angiotensin II-mediated free radical production in vascular smooth muscle cells: involvement of angiotensin AT1 receptor expression and Rac1 GTPase. Mol Pharmacol, 59, 646–654. Weyrich AS, Elstad MR, McEver RP et al (1996) Activated platelets signal chemokine synthesis by human monocytes. J Clin Invest, 97, 1525– 1534.
Wilson SH, Herrmann J, Lerman LO et al (2002) Simvastatin preserves the structure of coronary adventitial vasa vasorum in experimental hypercholesterolemia independent of lipid lowering. Circulation, 105, 415–418. Yamada M, Numaguchi Y, Okumura K et al (2002) Prostacyclin synthase gene transfer modulates cyclooxygenase-2-derived prostanoid synthesis and inhibits neointimal formation in rat balloon-injured arteries. Arterioscler Thromb Vasc Biol, 22, 256–262. Yin MJ, Yamamoto Y and Gaynor RB (1998) The anti-inflammatory agents aspirin and salicylate inhibit the activity of I(k)B kinase-b. Nature, 396, 77–80. Zou M, Martin C and Ullrich V (1997) Tyrosine nitration as a mechanism of selective inactivation of prostacyclin synthase by peroxynitrite. Biol Chem, 378, 707–713. Zou MH and Bachschmid M (1999) Hypoxia-reoxygenation triggers coronary vasospasm in isolated bovine coronary arteries via tyrosine nitration of prostacyclin synthase. J Exp Med, 190, 135–139. Zou MH, Leist M and Ullrich V (1999) Selective nitration of prostacyclin synthase and defective vasorelaxation in atherosclerotic bovine coronary arteries. Am J Pathol, 154, 1359–1365. Zou MH, Shi C and Cohen RA (2002) High glucose via peroxynitrite causes tyrosine nitration and inactivation of prostacyclin synthase that is associated with thromboxane/prostaglandin H(2) receptor-mediated apoptosis and adhesion molecule expression in cultured human aortic endothelial cells. Diabetes, 51, 198–203.
24 Molecular Regulation of Pancreatic Islet Prostaglandin Synthesis and its Relevance to Diabetes Mellitus R. Paul Robertson Pacific Northwest Research Institute, Seattle, WA, USA
Eicosanoids and diabetes mellitus share the characteristic of universality. Eicosanoids are synthesized by every tissue in the mammalian body and play hugely diverse roles in maintaining physiological functions. Diabetes mellitus is found in all modern societies and in the USA estimations of prevalence reach 7–9%, with another 5% of affected individuals going undetected. On the other hand, diabetes mellitus is an ancient disease, whereas the eicosanoids were discovered in the twentieth century. Only in the latter part of the twentieth century was it established that these arachidonic acid metabolites play modulatory roles in carbohydrate physiology. Evidence has also accumulated that certain eicosanoids might play a role in the pathophysiology of diabetes. The last decade has witnessed a rapid evolution of molecular biological information about prostaglandin synthesis which, in turn, has provided important new insights into the roles of eicosanoids in mammalian physiology and pathophysiology. This chapter will focus on the interrelationships between prostaglandins and diabetes, with emphasis on recent molecular observations that pertain to the pancreatic islet.
DIABETES: CHARACTERIZATION AND PATHOPHYSIOLOGY The human pancreas is comprised of exocrine and endocrine tissue. The endocrine pancreas contains over 1 million discrete islets of Langerhans, which are distributed throughout the entire pancreas. Each islet can be likened to a tennis ball that contains b cells within its core and a, d and PP cells on its surface. Insulin is produced by the b cells, whereas glucagon, somatostatin and pancreatic polypeptide are made by the a, d and PP cells, respectively. Glucose is the primary stimulator of insulin secretion but other substances are also secretagogues, such as arginine, leucine, secretin and glucagon. Inhibitors of insulin secretion include epinephrine, somatostatin and prostaglandin E2. The most important physiological actions of insulin are to increase glycogenesis and glucose uptake in tissues and to decrease lipolysis, ketogenesis, gluconeogenesis, glycogenolysis and glycaemia. Most patients with diabetes mellitus fall into the category of type 1 or type 2. Type 1 diabetes is synonymous with juvenileonset diabetes or insulin-dependent diabetes mellitus whereas type 2 diabetes mellitus is synonymous with adult-onset diabetes or non-insulin-dependent diabetes mellitus. The prevalence of type 1 and type 2 diabetes mellitus varies greatly among social groups but, generally speaking, type 2 is 10–20 times more frequent than The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
type 1. Patients with type 1 diabetes tend to be younger, more easily ketotic and lean, whereas type 2 patients tend to be older, less prone to ketosis and obese. Type 1 diabetes is an autoimmune disease accompanied by circulating islet cell antibodies and specific HLA-association, with a 40–50% concordance rate in identical twins. Type 2 patients have twice this concordance rate in identical twins but do not have an HLA association or increased circulating islet cell antibodies. Insulin treatment is absolutely required for patients with type 1 diabetes but not for type 2 diabetes. The secondary complication rates of diabetes for retinopathy, nephropathy, neuropathy, micro- and macrovascular disease are similar in both types of diabetes when corrected for the variables of glycaemic control and duration of disease. The pathogenesis of type 1 diabetes involves interactions among genetic susceptibility, environmental factors (such as viruses) and an autoimmune response. The autoimmune nature of type 1 diabetes is of particular relevance to eicosanoid pharmacology because cytokines are involved in autoimmunity. As will be developed below, cytokines can regulate prostaglandin synthesis and cytokine actions can be blocked by prostaglandin synthesis inhibitors. The result of these interactions is b cell injury and, eventually, death of b cells. Type 2 diabetes also involves genetic susceptibility but is not an autoimmune disease and does not result in b cell death. A major component of type 2 diabetes is insulin resistance, i.e. decreased sensitivity to insulin action in tissues such as muscle, fat and liver. Neither defective insulin secretion nor insulin resistance alone is sufficient to explain the pathophysiology of type 2 diabetes (Figure 24.1). Rather, a combination of both defects is necessary (Robertson 1992).
PROSTAGLANDINS AND DIABETES: A HISTORICAL PERSPECTIVE In 1876, decades before prostaglandins were discovered, an investigator named Ebstein reported in the German literature that sodium salicylate given orally decreased the amount of glucose appearing in the urine of diabetic patients. In 1971 a classic work by Vane and his colleagues reported that nonsteroidal, antiinflammatory drugs had, as a major mechanism of action, the inhibition of the cyclooxygenase pathway. For those interested in assessing the effects of prostaglandins on insulin secretion, this observation set into motion studies of the available non-steroidal antiinflammatory drugs (NSAIDs) as well as evaluation of the effects of the ever-increasing number of arachidonic acid metabolites on insulin secretion. The earliest
278
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 24.1 Hypothetical construct in which the pathogenesis of type 2 diabetes mellitus is the arithmetic product of two factors, defective glucoseinduced insulin secretion and defective tissue response to insulin. The combined effect of these products leads to a sequence of events in various tissues that eventually establishes a state of hyperglycaemia. Continued hyperglycaemia, in turn, has adverse effects on the pancreatic islet b cell referred to as ‘‘glucose toxicity’’, as well as adverse effects in body tissues expressed at multiple levels which result in insulin resistance. Eventually, glucose toxicity leads to further impairment of b cell function, which further diminishes glucose-induced insulin secretion. Implicit in this hypothesis is that both of the two initial factors, decreased glucose-induced insulin secretion and decreased tissue response to insulin, are required for the pathogenesis of type 2 diabetes, and that if either factor is missing, diabetes will not result. From Robertson (1992), with permission
studies of arachidonic acid metabolism were quite varied. Spellacy (1971) studied normal, pregnant women; Bressler et al (1968) studied live mice; and Johnson et al (1973) experimented with isolated rat islets. No uniform result was obtained in these early studies. Spellacy failed to find any effect of intravenous infusions of PGE2 and PGF2a in pregnant normal humans. Bressler et al observed that intraperitoneal PGE2 injections increased insulin and glucose levels in mice. Johnson et al reported that PGE1, PGE2 and PFG2a increased insulin secretion from isolated rat islets. One year after Johnson et al published their work, work from our laboratory provided evidence that prostaglandin E can inhibit basal and glucose-induced first-phase insulin secretion in vivo (Robertson et al 1974). These studies were performed in intact, anaesthetized dogs in which PGE1 or PGE2 was infused intravenously at a rate that developed four-fold increases in pancreatic arterial PGE concentrations. However, this rate of PGE infusion also caused hypotension. Consequently, these studies were repeated in animals who had been treated with aadrenergic blockade to test whether the inhibitory effects of PGE on insulin secretion were direct or mediated by hypotensioninduced release of epinephrine, a known inhibitor of insulin secretion (Robertson et al 1974). Despite the presence of aadrenergic blockade, PGE still inhibited glucose-induced insulin secretion (Figure 24.2). At about the same time, Burr and Sharp (1974) observed similar inhibition of first-phase insulin secretion when isolated islets were perifused with high glucose concentrations. These observations created renewed interest in the earlier work by Ebstein and other early investigators (Ebstein 1876; Williamson 1901; Field et al 1967; Hyams et al 1971), who demonstrated that salicylate can ameliorate clinical diabetes. Many reports
appeared subsequently demonstrating that NSAIDs augment insulin secretion and improve glucose tolerance. We examined first-phase glucose-induced insulin secretion because its absence is a specific metabolic defect (Figure 24.3) (Robertson and Porte 1973) in hyperglycaemic type 2 diabetic patients. The reason that this defect in glucose responsiveness is considered specific is because other secretagogues, such as arginine, isoproterenol, secretin and glucagon, stimulate first-phase insulin secretion in type 2 diabetic subjects who do not have this response to glucose. We observed that absent glucose-induced first-phase insulin secretion in type 2 diabetic patients can be partially restored by intravenous infusion of sodium salicylate (Figure 24.4) and that this was accompanied by improved glucose tolerance (Robertson and Chen 1977). These findings led us to hypothesize that part of the pathogenesis of impaired insulin secretion in type 2 diabetes may be related to islet overproduction of, or islet hypersensitivity to, PGE2 (Robertson 1986). At that time, however, no drugs were available for human use that specifically antagonized PGE2 action or inhibited the synthesis of PGE2 exclusively, so this hypothesis could not be fully tested. However, important in vitro information that the beneficial effect of sodium salicylate on glucose-induced insulin secretion is mediated by a decrease in PGE2 availability was provided by Metz et al (1981). In these experiments, sodium salicylate was shown to have a concentration-related effect to decrease PGE2 synthesis by neonatal rat islets that was accompanied by an increase in insulin secretion, which was reversed by adding exogenous PGE2 (Figure 24.5). Subsequently, reports from other investigators (Giugliano et al 1981; Micossi et al 1978) confirmed that salicylate and aspirin partially restore first-phase glucose-induced insulin secretion in type 2 diabetic patients. By 1986, 16 studies had been reported using aspirin, sodium salicylate or ibuprofen; 15 of these reported that these drugs
REGULATION OF ISLET PROSTAGLANDIN SYNTHESIS
279
glucose-induced first-phase insulin secretion (for review, see Robertson 1986). That indomethacin is the only exception to the rule that NSAIDs augment insulin secretion may be related to its ability to interfere with intracellular calcium flux and cAMPdependent protein kinase activity, properties that would be expected to diminish insulin secretion. Metabolites of the cyclooxygenase pathway that have been consistently identified in pancreatic islets include PGE2 and PGF2a. This was first demonstrated in 1977 by Hamandzic and Malik (1977), who demonstrated that adrenergic stimulation of the perfused rat pancreas releases these two prostanoids. Ten subsequent reports (Metz et al 1981, 1984; Banschbach and Nokin-Neaverson 1980; Dunlop et al 1984; Evans et al 1983; Horie et al 1984; Kelly and Laychock 1981; Marshall et al 1981; Morgan and Pek 1984; Turk et al 1984) also identified at least one of these two prostaglandins. The dominant metabolite from the lipoxygenase pathway was first shown by Metz et al to be 12-HETE (Metz et al 1983). This finding was later confirmed by other work (Turk et al 1984, 1985; Metz 1984). Virtually all drugs that decrease lipoxygenase activity also inhibit glucose-induced insulin secretion (Dunlop et al 1984; Morgan and Pek 1984; Metz et al 1983, 1982a, 1982b; Falck et al 1983; MacAdams et al 1984; Walsh and Pek 1984; Yamamoto et al 1982, 1983, 1985). By the end of 1983, efforts to characterize arachidonic acid metabolites made by the pancreatic islets and their effects on insulin secretion subsided greatly. At that time it could be concluded that:
Figure 24.2 Inhibition by PGE1 of glucose responses to intravenous glucose in anaesthetized dogs pretreated with phentolamine, an aadrenergic blocker. From Robertson et al (1974), with permission
augmented glucose-induced first-phase insulin secretion in diabetic and non-diabetic humans (for review, see Robertson 1986). On the other hand, five studies using indomethacin had been published; four of these reported inhibitory effects of this drug on
1. Pancreatic islets synthesize arachidonic acid metabolites via the cyclooxygenase and lipoxygenase pathways that exert regulatory influences over insulin secretion. 2. PGE2 consistently inhibits glucose-induced insulin secretion and worsens glucose homeostasis. 3. All inhibitors of the cyclooxygenase pathway, with the sole exception of indomethacin, augment glucose-induced insulin secretion and improve glucose homeostasis. 4. Defective glucose-induced first-phase insulin secretion in type 2 diabetic patients is partially restored by certain inhibitors of the cyclooxygenase pathway. 5. All drugs that inhibit lipoxygenase activity also inhibit insulin secretion.
Figure 24.3 Comparison of acute insulin responses to intravenous glucose and intravenous isoproterenol in type 2 diabetic patients. Left, glucose, 5 g i.v., failed to stimulate insulin secretion. Right, isoproterenol, 2 mg intravenously, successfully stimulated insulin release in the same subjects who failed to respond to intravenous glucose. These data indicate that the failure of glucose-induced insulin secretion cannot be explained by a general failure of the pancreatic islet to synthesize insulin or release it through exocytosis. From Robertson and Porte (1973), with permission
280
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 24.4 Restoration of absent first-phase, and defective second-phase, glucose-induced insulin secretion by an intravenous infusion of sodium salicylate, 40 mg/min, in type 2 diabetic patients. From Robertson and Chen (1977), with permission
Figure 24.5 Concentration–response curves for sodium salicylate effects on PGE2 synthesis and insulin secretion by neonatal rat islets. The insulin release potentiated by sodium salicylate was reversed when exogenous PGE2 (1076 M) was added to the 20 mg/dl concentration of sodium salicylate. From Metz et al (1981), with permission
MECHANISM OF ACTION: PGE2 RECEPTOR AND G-PROTEINS Prostanoid receptors were first demonstrated in fat cells by Kuehl and Humes (1972). Thereafter, receptors were demonstrated in many tissue types, as will be remarked upon later in this chapter. In 1987 we identified a mechanism of action of PGE2 in pancreatic islets and in HIT-T15 cells that involved a ligand– receptor interaction (Figure 24.6) (Robertson et al 1987) with pertussis toxin-sensitive substrates. Our data indicated that PGE2 binds to a specific receptor, decreases cyclic AMP concentrations (Figure 24.7), and decreases insulin secretion (Figure 24.8). Importantly, the binding affinity constant and the IC50s for decreasing cyclic AMP and glucose-induced insulin secretion were approximately the same (1079 M). In the same line of experiments, pertussis toxin was found to ADP-ribosylate a
40 kDa substrate in HIT-T15 cell membranes. Seaquist et al (1989) performed a study using perifused HIT-T15 cells to determine whether pertussis toxin would interfere with the ability of PGE2 to specifically inhibit glucose-induced first-phase insulin secretion. It was observed that pretreatment of the cells with the toxin almost completely prevented PGE2 inhibition of phasic insulin release (Figure 24.9). During these experiments, it was again observed that cyclic AMP levels were decreased by PGE2 and that pretreatment with pertussis toxin could partially prevent this effect. Soon thereafter, Kowluru and Metz (1994) demonstrated that PGE2 stimulates GTPase activity in a secretory granule-rich fraction of both rat and human islets and that this PGE2 effect is blocked by pertussis toxin treatment. They concluded that there may be a PGE2-stimulatable inhibitory G-protein located on the secretory granules that regulates exocytosis.
REGULATION OF ISLET PROSTAGLANDIN SYNTHESIS
Figure 24.6 Scatchard analysis of PGE2 binding data in HIT-T15 cells, a pancreatic islet b cell line. The data indicate saturable binding with a single class of binding sites and a Kd of 161079 M. From Robertson et al (1987), with permission
Evidence for many heterotrimeric G-protein a subunits has been found within the pancreatic islet or in islet cell lines. Identification of G-proteins has generally been performed by immunodetection of the protein by a specific antiserum. Identified a subunits of G-proteins include Gsa45, Gsa52, Gia1, Gia2, Gia3, Goa, Gaq, G11a, G14a, Gza and Gt2a (Walseth et al 1989; Zigman et al 1994; Hsu et al 1990; Schmidt et al 1991; Cormont et al 1991; Seaquist et al 1992; Berrow et al 1992; Baffy et al 1993; Poitont et al 1995). Thus, participation of G-proteins in mechanisms of islet exocytosis in general, and in PGE2-induced inhibition of insulin secretion specifically, is well established (Figure 24.10).
281
Figure 24.7 Inhibitory effect of increasing concentrations of PGE2 on cAMP accumulation in buffer during HIT-T15 cell incubations in the presence of 11.1 mM glucose. IC50%161079 M
THE MOLECULAR ERA AND PROSTAGLANDINS The conventional pharmacological approach extant in the early 1980s had several limitations. For example, all of the drugs used to block prostanoid synthesis had other non-prostanoid-related effects. Moreover, these drugs block the synthesis of more than one prostanoid because their site of action usually involved cyclooxygenase, an enzyme located early in the prostaglandin synthetic pathway. Very few antagonists directed specifically against prostanoid receptors were available. Fortunately, in the past decade, molecular biology has ushered in new investigative tools that have brought radical changes to our thinking about prostaglandin pharmacology and physiology (Robertson 1995; Smith and Dewitt 1966; Coleman et al 1994). One of the pivotal discoveries was the finding of two separate forms of cyclooxygenase, constitutive and inducible (Figure 24.11). One of the earliest clues that two forms of cyclooxygenase exist was provided by Rosen et al (1989) through experiments in epithelial cells isolated from sheep tracheal mucosa, demonstrating that a 70 kDa cyclooxygenase protein and a 2.8 kb cyclooxygenase mRNA were coordinately expressed. It was noted that the mRNA levels did not increase as expected when cyclooxygenase activity was stimulated. However, rehybridization
Figure 24.8 Inhibition of glucose-induced insulin secretion during static incubations of HIT-T15 cells with 11.1 mM glucose. PGE2 (IC50 161079 M) was more potent than epinephrine (EPI) but less so than somatostatin (SRIF). From Robertson et al (1987), with permission
282
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 24.9 Inhibition of glucose-induced insulin secretion from HIT-T15 cells by PGE2 and its prevention by pretreatment of HIT-T15 cells with pertussis toxin. From Seaquist et al (1989), with permission
Figure 24.10 Mechanism of action for PGE2 inhibition of insulin secretion. After PGE2 binds to its receptor (step 1), it associates with GDP-binding Gprotein (comprised of a-, b- and d-subunits). This union promotes exchange of GTP for GDP (step 2). This unstable complex quickly disassociates into free PGE2-receptor complex (step 3), frees the b-d-subunits (step 4), and activates the a-subunit to which GTP is bound (step 5). In the case of PGE2, this activated species operates through Gi/o subunits to inhibit insulin secretion. The activated species is inactivated by intrinsic GTPase activity of the asubunit, which hydolyses bound GTP to GDP (step 7). This allows for reassociation of a-GDP with the b-d-subunit (step 8). This completes the cycle and returns the G-protein to its quiescent state. Cholera toxin (CTx) irreversibly ADP ribosylates Gsa and Gta, blocks intrinsic GTPase activity and consequently prevents step 7. Pertussis toxin irreversibly ADP ribosylates G-proteins such as Gia and Goa and prevents interaction with PGE2-receptor complex with GDP-binding G-proteins
REGULATION OF ISLET PROSTAGLANDIN SYNTHESIS
283
Figure 24.11 Molecular regulation of prostaglandin synthesis. The COX-1 and COX-2 genes control synthesis of the COX-1 and COX-2 enzymes, respectively. The two genes are located on separate chromosomes and give rise to mRNAs of different sizes. The two enzymes are approximately the same size and have approximately 60% homology. Both enzymes use arachidonic acid as substrate to generate biologically active prostanoids, here represented by PGE2. Once PGE2 is formed, it can interact with any one subtype or all four subtypes of EP receptors. IL-1 increases gene expression of COX-2 mRNA but does not affect COX-1 mRNA levels. Dexamethasone (DEX) diminishes the induction of COX-2 gene expression. Acetylsalicylic acid (ASA) and sodium salicylate (SS) decrease the mass of COX-2 enzyme translated and processed after COX-2 mRNA levels are increased. These two NSAIDs, as well as others, represented by indomethacin (INDO), also inhibit activation of the COX-1 and COX-2 enzymes. None of the NSAIDs affect expression of the COX-1 gene or translating and processing of the COX-1 enzyme. From Robertson (1992), with permission
of Northern blots at lower stringency identified an additional 4.0 kb mRNA cyclooxygenase species whose expression did increase coordinately with cyclooxygenase activity. This led the authors to speculate that two genes might exist that are responsible for synthesis of cyclooxygenase. At about the same time, work by Fu et al (1990) suggested the existence of two pools of cyclooxygenase in human monocytes. Several years later, separate COX-1 and COX-2 genes were cloned (Herschman 1996), thereby fully establishing that two forms of the enzyme exist. Other observations added new dimensions to concepts about sites of action for NSAIDs (Figure 24.11). Aspirin and sodium salicylate were shown in human umbilical vein endothelial cells to diminish the mass of cyclooxygenase induced by IL-1, whereas indomethacin had no apparent effects on enzyme levels (Wu et al 1991). The authors also reported that aspirin suppressed expression of 2.7 kb cyclooxygenase mRNA. Crofford et al (1994) observed that dexamethasone markedly suppressed induction of cyclooxygenase-2 mRNA responsiveness to IL-1. This observation suggested that a major site for the inhibitory action of
corticosteroids on prostaglandin synthesis involves gene expression of cyclooxygenase-2. This was a timely observation because of increasing scepticism about the conventional concept that the sole site of action for glucocorticoids on prostanoid synthesis is inhibition of phospholipase A2 activity. Yet another new twist about actions of NSAIDs was provided by reports that aspirin, indomethacin, ibuprofen, acetaminophen and sodium salicylate were more potent in inhibiting cyclooxygenase-1 than cyclooxygenase-2, whereas diclofenac and naproxen were more potent in inhibiting cyclooxygenase-2 than cyclooxygenase-1 (Mitchell et al 1994). Since aspirin and indomethacin, often used to treat inflammation, are more effective in inhibiting cyclooxygenase-1, which usually modulates physiological processes, this increased efficacy against COX-1 may account for some of the undesirable clinical side effects of these two drugs when they are used to decrease inflammation. Sodium salicylate and aspirin were reported to have another potential site of action. The transcription factor NF-kB was observed to be inhibited by sodium salicylate and aspirin (Kopp
284
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
and Ghosh 1994). This is of importance to molecular regulation of prostaglandin synthesis because the cyclooxygenase-2 gene has an NF-kB binding site in its promoter region. Thus, the promoter region of cyclooxygenase-2 is regulated by a transcription factor whose availability appears to be limited by at least two NSAIDs. The discovery that there are prostaglandin receptor subtypes provided major insights into existing concepts about prostaglandin physiology. Previously, studies utilizing radiolabelled prostaglandins and various tissues demonstrated receptors in cellular derivatives of fat, liver, adrenal cortex and medulla, ovary (corpus luteum), uterus, kidney, stomach, ileum, thymus, skin, brain, lung, pancreatic islets, and various blood products, including platelets, red blood cells, neutrophils, macrophages and monocytes (Robertson 1986). A new classification of prostanoid receptors has been proposed involving five types of receptors, based upon sensitivity to the five naturally occurring prostanoids, PGD2, PGE2, PGF2a, PGI2 and TXA2 (Coleman et al 1994). This nomenclature refers to the receptors as ‘‘P-receptors’’ and the preceding letter indicates the prostanoid to which each receptor is most sensitive. EP receptors have been subdivided into four subtypes and termed EP1, EP2, EP3 and EP4. The mechanism of action of DP receptors involves stimulation of adenylyl cyclase. The EP1, EP2, EP3 and EP4 receptors are coupled to transduction systems that alter phosphatidylinositide hydrolysis and adenylyl cyclase activity. FP receptors, like EP1 receptors, are coupled to stimulation of phosphatidylinositide hydrolysis. IP receptors, like EP2 and EP4 receptors, involve stimulation of adenylyl cyclase activity. TP receptors, like EP1 and FP receptors, involve increases in phosphatidylinositide turnover. EP3 receptors are mediated by decreases in adenylyl cyclase activity. Thus, although there are specific receptors for prostaglandins and, in the case of PGE2, at least four different receptor subtypes, some of these different receptors are coupled to shared biochemical pathways (Coleman et al 1994).
Figure 24.12 RT–PCR analysis of COX-1 and COX-2 expression in Syrian hamster tissues. COX-1 mRNA in HIT-T15 cells and Syrian hamster islets was barely detectable, whereas it was readily detectable in 3T3 cells and in Syrian hamster spleen, kidney, and liver cultured in 0.2% foetal bovine serum. In contrast, RT–PCR readily demonstrated COX-2 mRNA in HIT-T15 cells and hamster islets, whereas much smaller amounts were observed in the other tissues cultured in 0.2% foetal bovine serum. From Sorli et al (1998), with permission
CYCLOOXYGENASE GENES AND THE PANCREATIC ISLET: DOMINANCE OF COX-2 Given the new information about molecular regulation of prostaglandin synthesis, we began in 1996 to assess how well the pancreatic islet conformed to the new dogma. Surprisingly, we found only cyclooxygenase-1 mRNA in various non-stimulated and stimulated preparations of pancreatic islet b cells, including the HIT-T15 cell line, Syrian hamster islets and human isolated islets (Figure 24.12) (Sorli et al 1998). No evidence was found for cyclooxygenase-1. However, cyclooxygenase-2 gene expression was easily detected under non-stimulated conditions and yet also was modestly increased with IL-1 treatment and decreased by dexamethasone treatment (Figure 24.13). Evidence for the dominance of cyclooxygenase-2 gene expression in islet tissue was consistently found using Northern blot analysis, RT–PCR, Western blot analysis and a cyclooxygenase-2 specific inhibitor (Figure 24.14). Our experiments investigating the effects of IL-1 on expression of the cyclooxygenase-2 gene revealed a biphasic effect, with an initial increase followed by a later decrease below basal levels of COX-2 mRNA. We also observed surprisingly high levels of NF-IL-6 binding, and only modest amounts of NF-kB binding, in the non-stimulated state (Sorli et al 1998). Studies with reporter gene techniques demonstrated that cyclooxygenase-2 promoter activity was greatly reduced under basal conditions when mutating its promoter region to delete the binding site for NF-IL-6 (Figure 24.15). No differences were observed with the NF-kB mutant. After stimulations with IL-1, levels of NF-IL-6 binding gradually decreased, which corresponded temporally with biphasic changes in cyclooxygenase-2 mRNA. We concluded that pancreatic islet cyclooxygenase-2 is dominantly expressed in the
Figure 24.13 RT–PCR for COX-2 expression in human pancreatic islets. COX-2 mRNA was readily detectable in islets in the presence of 0.2% foetal bovine serum. Interleukin-1 (IL-1B) treatment increased COX-2 mRNA levels; these levels were suppressed by dexamethasone (DEX). From Sorli et al (1998), with permission
non-stimulated and in early stimulated conditions, a situation reported previously only for rat brain (Yamagata et al 1993), renal tissue (Harris et al 1994), bronchial tissue (Asano et al 1996) and granulosa cells (Narko et al 1997). Shortly thereafter, islet COX-2 dominance was confirmed by other investigators (Kwon et al 1998). We hypothesized that this unusual situation in the islet might be related to high basal levels of NF-IL-6 binding to the COX-2 promoter.
REGULATION OF ISLET PROSTAGLANDIN SYNTHESIS
285
Figure 24.14 Stimulation of PGE2 release from HIT-T15 cells by interleukin-1 (IL-1) and inhibition of basal and stimulated PEG2 levels by NS-398, a COX-2-specific synthesis inhibitor. From Sorli et al (1998), with permission
Figure 24.15 Comparison of the effects of mutating the COX-2 promoter at the NF-kB and the NF-IL-6 binding sites. Mutation of the NF-IL-6 binding site, but not the NF-kB binding site, caused a marked decrease in reporter gene activity. From Sorli et al (1998), with permission
IMPLICATIONS OF COX-2 DOMINANCE FOR PATHOPHYSIOLOGY AND TREATMENT OF DIABETIC PATIENTS Three new considerations, i.e. (a) that there are two forms of cyclooxygenase; (b) that NSAIDs and corticosteroids have more sites of action on the prostaglandin synthetic pathway than previously appreciated; and (c) that there are at least four PGE2 receptors with different post-receptor consequences, carry important implications for the interpretation of previously reported
data about prostaglandins and pancreatic islet function. Many of the physiological and pathophysiological experiments performed in the 1970s and the 1980s need to be repeated now that more specific inhibitors of COX-1 and COX-2 are available. The discoveries that prostaglandin E2 has four different receptors and three different mechanisms of action make more understandable the variable results that are sometimes observed when the effects of prostaglandin E on pancreatic islet function are studied (Robertson 1983). In terms of normal physiology, that PGE2 synthesis is stimulated by glucose (Turk et al 1984) suggests that endogenous PGE2 production serves to negatively modulate physiological release of insulin under normal stimulatory situations, because PGE2 is an inhibitor of glucose-induced insulin secretion (Robertson and Chen, 1977). On the other hand, should cyclooxygenase-2 activity be upregulated excessively, consequent overproduction of PGE2 might lead to insufficient insulin release and glucose intolerance, thereby playing a role in the pathogenesis of type 2 diabetes, as has been suggested (Robertson 1986). In the case of type 1 diabetes, in which there is an autoimmune response that appears to involve IL-1, the high basal cyclooxygenase-2 activity normally present might be increased to even higher levels by IL-1. To the extent that b-cell rest might stave off full development of type 1 diabetes, IL-1-stimulated PGE2 production could represent a host defence mechanism. On the other hand, inhibition of insulin release by PGE2 could also have negative effects, by virtue of worsening the developing hyperglycaemia through inhibition of insulin secretion during the onset of type 1 diabetes. These speculations need to be subjected to extensive evaluation through the use of the newer COX-2-specific antagonists at the time of onset of type 1 diabetes and as treatment for type 2 diabetes.
ACKNOWLEDGEMENTS This work was supported by NIH Grant RO1-39994. Appreciation is expressed to Shannon Greer for manuscript preparation.
286
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
REFERENCES Asano K, Lilly CM and Drazen JM (1996) Prostaglandin G/H synthase-2 is the constitutive and dominant isoform in cultured human lung epithelial cells. Am J Physiol, 271, L126–L131. Baffy G et al (1993) G-protein specificity in signaling pathways that mobilize calcium in insulin-secreting cell line derived from transgenic mice. Diabetes, 42, 1878–1882. Banschbach MW and Hokin-Neaverson M (1980) Acetylcholine promotes the synthesis of prostaglandin E in mouse pancreas. FEBS Lett, 117, 131–133. Berrow NS, Milligan G and Morgan NG (1992) Immunological characterization of the guanine-nucleotide binding proteins Gi and Go in rat islets of Langerhans. J Mol Endocrinol, 8, 103–108. Bressler R, Vargas-Cordon M and Lebovitz HE (1968) Tranylcypromine: a potent insulin secretagogue and hypoglycemic agent. Diabetes, 17, 617–624. Burr IM and Sharp R (1974) Effects of prostaglandin E1 and of epinephrine on the dynamics of insulin release in vitro. Endocrinology, 94, 835–839. Coleman RA, Smith WL and Narumiya S (1994). VIII International union of pharmacology classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. Cormont M et al (1991) Identification of G protein a-subunits in RINm5F cells and their selective interaction with galanin receptor. Diabetes, 40, 1170–1176. Crofford LJ et al (1994) cyclooxygenase-1 and 2 expression in rheumatoid synovial tissues: effects of interleukin-1 b, phorbol ester, and corticosteroids. J Clin Invest, 93, 1095–1101. Dunlop M et al (1984) Effects of inhibitors of eicosanoid synthesis on insulin release by neonatal pancreatic islets. Biochim Biophys Acta, 801, 10–15. Ebstein W (1876) Zur therapei des diabetes mellitus, insbesondere uber die anwendung des salicylsauren natron bei demselben. Klin Wochenschr, 13, 337–340. Evans MH, Pace CS and Clements RSJ (1983) Endogenous prostaglandin synthesis and glucose-induced insulin secretion from the adult rat pancreatic islet. Diabetes, 32, 509–515. Falck JR et al (1983) Epoxyeicosatrienoic acids stimulate glucagon and insulin release from isolated rat pancreatic islets. Biochem Biophys Res Commun, 114, 743–749. Field JB, Boyle C and Remer A (1967) Effect of salicylate infusion on plasma-insulin and glucose tolerance in healthy persons and mild diabetics. Lancet, 1, 1191–1194. Fu J et al (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740. Giugliano D et al (1981) Impaired insulin secretion in human diabetes mellitus. II. A possible role for prostaglandins. Prostagland Med, 6, 41–50. Hamamdzic M and Malik KU (1977) Prostaglandins in adrenergic transmission of isolated perfused rat pancreas. Am J Physiol, 232, E201–E209. Harris RC et al (1994) Cyclooxygenase-2 is associated with the macula dense of rat kidney and increases with salt restriction. J Clin Invest, 94, 2504–2510. Herschman HR (1996) Prostaglandin synthase 2. Biochim Biophys Acta, 1299, 125–140. Horie H et al (1984) Presence of prostaglandin D2, E2 and F2a in rat pancreatic islets. Prostagland Leukotrienes Med, 16, 39–44. Hsu WH et al (1990) Molecular cloning of a novel splice variant to the asubunit of the mammalian Go protein. J Biol Chem, 265, 11220–11226. Hyams DE et al (1971) The effect of 3-methyl salicylic (O-cresotinic) acid on plasma insulin and glucose tolerance in diabetic and non-diabetic subjects. Diabetologia, 7, 94–101. Johnson DG, Fujimoto WY and Williams RH (1973) Enhanced release of insulin by prostaglandins in isolated pancreatic islets. Diabetes, 22, 658–663. Kelly KL and Laychock SG (1981) Prostaglandin synthesis and metabolism in isolated pancreatic islets of the rat. Prostaglandins, 21, 759–769. Kopp E and Ghosh S (1994) Inhibition of NF-kB by sodium salicylate and aspirin. Science, 265, 956–959.
Kowluru A and SA Metz (1994) Stimulation of prostaglandin E2 of a highaffinity GTPase in the secretory granules of normal rat and human pancreatic islets. Biochem J, 297, 399–406. Kuehl FA and Humes JL (1972) Direct evidence for a prostaglandin receptor and its application to prostaglandin measurements. Proc Natl Acad Sci USA, 69, 480–484. Kwon G et al (1998) Evidence for involvement of the proteasome complex (26S) and NFkB in IL-1b-induced nitric oxide and prostaglandin productin by rat islets and RINm5F cells. Diabetes, 47, 583–591. MacAdams MR, Pek SB and Lands WEM (1984) The effect of flurbiprofen, a potent inhibitor of prostaglandin synthesis, on insulin and glucagon release from isolated rat pancreas. Endocrinology, 114, 1364–1370. Marshall PJ, Boatman DE and Hokin LE (1981) Direct demonstration of the formation of prostaglandin E2 due to phosphatidylinositol breakdown associated with stimulation of enzyme secretion in the pancreas. J Biol Chem, 256, 844–847. Metz SA (1985) Glucose increases the synthesis of lipoxygenase-mediated metabolites of arachidonic acid in intact rat islets. Proc Natl Acad Sci USA, 82, 198–202. Metz SA et al (1983) Lipoxygenase pathway in islet endocrine cells. J Clin Invest, 71, 1191–1205. Metz SA, Fujimoto W and Robertson RP (1982a) Modulation of insulin secretion by cyclic AMP and prostaglandin E: the effects of theophylline, sodium salicylate and tolbutamide. Metabolism, 31, 1014–1022. Metz SA, Fujimoto WF and Robertson RP (1982b) Lipoxygenation of arachidonic acid: a pivotal step in stimulus-secretion coupling in the pancreatic b cell. Endocrinology, 111, 2141–2143. Metz SA, Murphy RC and Fujimoto W (1984) Effects on glucose-induced insulin secretion of lipoxygenase-derived metabolites of arachidonic acid. Diabetes, 33, 119–124. Metz SA, Robertson RP and Fujimoto WY (1981) Inhibition of prostaglandin E synthesis augments glucose-induced insulin secretion in cultured pancreas. Diabetes, 30, 551–557. Micossi P et al (1978) Aspirin stimulates insulin and glucagon secretion and increases glucose tolerance in normal and diabetic subjects. Diabetes, 27, 1196–1204. Mitchell JA et al (1994) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697. Morgan RO and Pek SB (1984) Role of arachidonate lipoxygenase and cyclooxygenase products in insulin and glucagon secretion from rat pancreatic islets. Metabolism, 33, 928–935. Narko K, Ritvos O and Ristimaki A (1997) Induction of cyclooxygenase-2 and prostaglandin F2a receptor expression by interleukin-1b in cultured human granulosa-luteal cells. Endocrinology, 138, 3638–3644. Poitout V et al (1995) Morphologic and functional characterization of bTC6 cells, an insulin-secreting cell line derived from transgenic mice. Diabetes, 44, 306–313. Robertson RP (1983) PGE, carbohydrate homeostasis, and insulin secretion: a suggested resolution of the controversy. Diabetes, 32, 231–234. Robertson RP (1986) Arachidonic acid metabolite regulation of insulin secretion. Diabetes/Metab Rev, 2, 261–296. Robertson RP (1986) Characterization and regulation of prostaglandin and leukotriene receptors: an overview. Prostaglandins, 31, 395–411. Robertson RP (1992) Defective insulin-secretion in NIDDM: integral part of a multiplier hypothesis. J Cell Biochem, 48, 227–233. Robertson RP (1995) Molecular regulation of prostaglandin synthesis: implications for endocrine systems. Trends Endo Metab, 6, 293. Robertson RP and Chen M (1977) A role for prostaglandin E in defective insulin secretion and carbohydrate intolerance in diabetes mellitus. J Clin Invest, 60, 747–753. Robertson RP and Porte DJ (1973) The glucose receptor: a defective mechanism in diabetes mellitus distinct from the b-adrenergic receptor. J Clin Invest, 52, 870–876. Robertson RP et al (1974) Inhibition of in vivo insulin secretion by prostaglandin E1. J Clin Invest, 54, 310–315. Robertson RP et al (1987) Receptor-mediated adenylate cyclase-coupled mechanism for PGE2 inhibition of insulin secretion in HIT cells. Diabetes, 36, 1047–1053.
REGULATION OF ISLET PROSTAGLANDIN SYNTHESIS Rosen GD et al (1989) Identification of a cyclooxygenase-related gene and its potential role in prostaglandin formation. Biochem Biophys Res Commun, 164, 1358–1365. Schmidt A et al (1991) Involvement of pertussis toxin-sensitive G-proteins in the hormonal inhibition of dihydropyridine-sensitive Ca2+ currents in an insulin secreting cell line (RIN 5mF). J Biol Chem, 266, 18025– 18033. Seaquist ER et al (1989) Pertussis toxin-sensitive G protein mediation of PGE2 inhibition of cAMP metabolism and phasic glucose-induced insulin secretion in HIT cells. Diabetes, 38, 1439–1445. Seaquist ER et al (1992) G-proteins and hormonal inhibition of insulin secretion from HIT-T15 cells and isolated rat islets. Diabetes, 41, 1390– 1399. Smith WL and Dewitt DL (1966) Prostaglandin endoperoxide H synthases-1 and 2. Adv Immunol, 62, 167. Sorli CH et al (1998) Basal expression of cyclooxygenase-2 and NIF-IL-6 are dominant and coordinately regulated by IL-1 in the pancreatic islet. Proc Natl Acad Sci USA, 95, 1788–1793. Spellacy WN, Buhi WC and Holsinger KK (1971) The effect of prostaglandin F2a and E2 on blood glucose and plasma insulin levels during pregnancy. Am J Obstet Gynecol, 111, 239–243. Turk J et al (1984) Arachidonic acid metabolism in isolated pancreatic islets. I. Identification and quantitation of lipoxygenase and cyclooxygenase products. Biochim Biophys Acta, 794, 110–124. Turk J et al (1984) Arachidonic acid metabolism in isolated pancreatic islets. II. The effects of glucose and of inhibitors of arachidonic metabolism on insulin secretion and metabolite synthesis. Biochim Biophys Acta, 794, 125–136. Turk J, Colca JR and McDaniel ML (1985) Arachidonic acid metabolism in isolated pancreatic islets. III. Effects of exogenous lipoxygenase
287
products and inhibitors on insulin secretion. Biochim Biophys Acta, 834, 23–36. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature (New Biol), 231, 232–235. Walseth TF et al (1989) Increase in Gs and cyclic AMP generation in HIT cells. J Biol Chem, 264, 21106–21111. Walsh MF and SB Pek (1984) Possible role of endogenous arachidonic acid metabolites in stimulated release of insulin and glucagon from the isolated, perfused rat pancreas. Diabetes, 33, 929–936. Williamson RT (1901) On the treatment of glycosuria and diabetes mellitus with sodium salicylate. Br Med J, 760–762. Wu KK et al (1991) Aspirin inhibits interleukin 1-induced prostaglandin H synthase expression in cultured endothelial cells. Proc Natl Acad Sci USA, 88, 2384–2387. Yamagata K et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Yamamoto S et al (1982) Prevention of glucose-induced insulin secretion by lipoxygenase inhibitor. Eur J Pharmacol 78, 225–227. Yamamoto S et al (1983) Insulinotropic effects of exogenous phospholipase A2 and C in isolated pancreatic islets. Eur J Pharmacol. 86, 121–124. Yamamoto S et al (1983) Modulation of insulin secretion by lipoxygenase products of arachidonic acid. J Biol Chem, 258, 12149–12152. Yamamoto S et al (1985) Lipoxygenase inhibitors and cyclic AMPmediated insulin secretion caused by forskolin, theophylline and dibutyrl cyclic AMP. J Pharmacol Exp Ther, 233, 176–180. Zigman JM et al (1994) Expression of cone transducin, G2a, and other Gprotein a-subunit messenger ribonucleic acids in pancreatic islets. Endocrinology, 135, 31–37.
25 Prostaglandins, Leukotrienes and Bone Carol C. Pilbeam and Lawrence G. Raisz University of Connecticut Health Center, Farmington, CT, USA
In the 30 years since PGE2 was first shown to stimulate cyclic AMP production and resorption in bone organ cultures, evidence that prostaglandins (PGs) and other eicosanoids have important physiological and pathological roles in skeletal remodelling has continued to accumulate. Bone is a dynamic tissue that is remodelled throughout life in humans. Remodelling permits the skeleton to respond to mechanical stresses with changes in mass and structure, in order to achieve a better balance between stress and load-bearing capacity and to respond to changing metabolic conditions in order to maintain stable extracellular calcium concentration. This remodelling, or bone turnover, consists of cycles of resorption followed by formation that remove and replace discrete packets of bone throughout the skeleton. The net balance of these cycles determines whether or not the skeletal bone mass is maintained. Loss of bone mass increases skeletal fragility and risk for fracture and can result in the disease of osteoporosis. Clinical paradigms for preventing and treating osteoporosis depend on our identifying and understanding the action of factors that determine the balance of bone turnover. Among these factors are prostaglandins, which are abundant in bone and can regulate both bone resorption and bone formation. There are two major lineages of cells involved in bone turnover, osteoclasts and osteoblasts (Figure 25.1). Osteoclasts have as their
major function the resorption of bone. In response to a variety of activating signals, they differentiate from a haematopoietic stem cell, eventually fusing into large multinuclear cells with the ability to resorb mineralized matrix. When their resorptive activity is completed, they undergo apoptosis and are eventually replaced by osteoblasts, the bone-forming cells (Takahashi et al 2002). Osteoblasts differentiate from a mesenchymal stromal stem cell in the bone marrow (Aubin and Triffitt 2002). Mature osteoblasts produce the bone matrix, consisting largely of type I collagen, which will subsequently become mineralized. After their boneforming activity ceases, some osteoblasts will undergo apoptosis, some will become quiescent cells lining the newly formed matrix, and some will become incorporated into the newly formed matrix as an interconnecting network of terminally differentiated cells called osteocytes. Osteocytes are thought to be the cells that sense strains in the mineralized matrix and send out signals that result in adaptive bone remodelling in response to mechanical loading. The observation that resorption is generally followed by formation has been characterized as the ‘‘coupling’’ of resorption to formation. Potential coupling agents are factors such as prostaglandins, which can stimulate the differentiation of both the osteoclast and osteoblast lineage, or factors released from the collagen matrix during resorption, which can stimulate osteoblastic
Figure 25.1 Schematic diagram of bone turnover. The resorption/formation cycle begins with the arrival of osteoclasts, derived from haematopoietic stem cells, that will resorb mineralized bone matrix. After several weeks of activity, osteoclasts undergo apoptosis. Osteoclasts are followed by osteoblasts, derived from a mesenchymal or stromal stem cell in the bone marrow. Osteoblasts lay down osteoid or unmineralized bone matrix over a period of several months. Some osteoblasts will undergo apoptosis, while others will become lining cells or be incorporated into the bone matrix as osteocytes The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
290
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
differentiation and chemotaxis. Although osteoclasts appear to initiate the cycle, the differentiation of mature osteoclasts from haematopoietic stem cells is dependent on contact of osteoclastic precursors with cells of the osteoblastic lineage and is regulated by soluble factors produced by the osteoblasts. Multiple model systems for studying bone metabolism in vitro have been established (Majeska and Gronowicz 2002). Measures of both bone resorption and formation can be studied in organ cultures of foetal or neonatal rodent long bones and calvariae. Osteoblastic cells can be obtained from freshly isolated calvarial bone, expanded in culture and used to study collagen production and gene expression. Osteocytes, on the other hand, are terminal cells that do not replicate in culture. Osteoclasts are also terminal cells and do not replicate in culture, but isolated osteoclasts can be cultured on bone slices and their resorptive activity studied in short-term culture. Transformed cell models or tumour cell lines exist for both osteoblastic and osteoclastic lineages. One of the most commonly used clonal osteoblastic cell lines, MC3T3-E1 cells, was immortalized from neonatal murine calvaria. Differentiation of mature osteoblasts or osteoclasts from early progenitor cells in the bone marrow can be studied in cultured marrow from rodent long bones. PG PRODUCTION IN BONE PGs have been the eicosanoids most studied in bone because they are highly expressed in osteoblastic cells and can have marked effects on both bone resorption and formation. Although PGE2 has been the prostanoid most extensively studied in bone, other prostanoids produced by bone cells include PGF2a and 6-ketoPGF1a the metabolite of PGI2, as well as some PGD2 and thromboxane (Feyen et al 1984; Raisz et al 1979). Stimulators of PG Production in Bone Early studies showed that complement-sufficient antisera could increase resorption of foetal rat long bones by stimulating PG production (Raisz et al 1974). Subsequently, many agonists were found to increase bone PG production in osteoblastic cells. Most commonly used osteoblastic cell models constitutively express cyclooxygenase-1 (COX-1) but can be induced to express cyclooxygenase-2 (COX-2), except for the rat osteosarcoma cell line, ROS 17/2.8, which expresses only COX-1 (Pilbeam et al 1997b). However, the induction of COX-2 in osteoblastic cells appears necessary for most stimulated PG responses (Pilbeam et al 2002). Agonists shown to stimulate PGE2 production associated with COX-2 induction in osteoblastic cells include: (a) cytokines—IL-1 (Harrison et al 1994; Kawaguchi et al 1994, 1996; Min et al 1998; Pilbeam et al 1997a), TNF-a (Kawaguchi et al 1996) and IL-6 (Tai et al 1997); (b) growth factors—TGFa (Harrison et al 1994; Pilbeam et al 1997a), TGFb (Pilbeam et al 1993, 1997a), bone morphogenetic protein-2 (BMP-2) (Chikazu et al 2002) and basic fibroblast growth factor (FGF-2) (Kawaguchi et al 1995a); (c) systemic hormones—PTH (Kawaguchi et al 1994; Tetradis et al 1997) and 1,25(OH)2 vitamin D3 (Okada et al 2000a); and (d) fluid shear stress or mechanical loading (Forwood et al 1998; Klein-Nulend et al 1997; Pavalko et al 1998). Serum is also a potent stimulator of COX-2 expression and PG production in cultured osteoblasts (Pilbeam et al 1993). In contrast to bone organ cultures, many osteoblastic cell cultures (Pilbeam et al 1994; Kawaguchi et al 1994, 1996) produce little PGE2 despite induction of COX-2 expression unless serum is present or arachidonic acid is added, suggesting that substrate availability is limiting. Some agonists, such as TGFb (Pilbeam et al 1997a) and PDGF (Chen et al 1997), stimulate more PGE2 production in serum-free cultures
of osteoblasts than other agonists, probably secondary to their effects on phospholipases. Many PGs can themselves induce COX-2 expression and can, therefore, amplify PG responses to other agonists, e.g. PGF2a was found to increase PGE2 production in organ culture, thereby enhancing the resorptive effects of PGE2 (Raisz et al 1990). Cyclic AMP (cAMP) increases COX-2 expression and PG production and PGs increase cAMP production, which is the basis for a positive feedback system (Raisz et al 1990; Oshima et al 1991; Tetradis et al 1997). PGs also induce COX-2 expression through protein kinase C (PKC) pathways (Pilbeam et al 1994; Takahashi et al 1994; Suda et al 1998). PGE2 can also enhance its own generation in osteoblasts via effects on phospholipases (Murakami et al 1997). The autoamplification of PGs could be important in mediating the prolonged effects of short periods of impact loading or mechanical strain in skeletal tissue or the effects of cytokines in inflammatory diseases. Inhibitors of PG Production in Bone Glucocorticoids are potent inhibitors of stimulated PG production. They have a major inhibitory effect on inducible COX-2 mRNA and protein expression, which accounts for the majority of their effects on PG production in bone and other tissues (Kawaguchi et al 1994; Pilbeam et al 1993). Retinoic acid is a potent transcriptional inhibitor of COX-2 gene expression and PG production in osteoblasts (Pilbeam et al 1995). The cytokines IL-4 and IL-13 also inhibit COX-2 expression and PG production in bone organ and cell cultures (Kawaguchi et al 1996; Onoe et al 1996). Non-steroidal antiinflammatory drugs (NSAIDs) inhibit PG production by competing directly with arachidonic acid for binding to the cyclooxygenase catalytic site. NS-398 was the first selective COX-2 NSAID to become widely available for in vitro and animal studies. In rodent osteoblastic cells, NS-398 at a concentration of 0.01 mM was selective for inhibition of COX-2 activity; however, at concentrations of 0.1 and 1 mM, NS-398 also inhibited COX-1 activity by 60% and 85%, respectively (Pilbeam et al 1997b). Hence, the selectivity of NS-398 was lost at higher doses. Because PGs themselves can induce COX-2 expression, NSAIDs can also decrease COX-2 expression by reducing PGmediated auto-amplification. Oestrogen may decrease cytokine-stimulated PG production in bone. The bone loss of oestrogen withdrawal may be mediated in part by increased cytokine activity (Kimble et al 1995) and cytokines can stimulate PG production in bone. PG production by calvaria cultured briefly from ovariectomized rats was increased, and this effect was reversed by oestradiol administration in vivo (Feyen and Raisz 1987). Marrow supernatants from ovariectomized mice produced a greater stimulation of bone resorption, secondary to a greater increase in COX-2 expression and PG production, in calvarial organ cultures compared to marrow supernatants from sham-operated mice (Kawaguchi et al 1995b; Miyaura et al 1995). The induction of COX-2 expression was due to increased IL-1 activity in the supernatants and was reversed by in vivo treatment of the ovariectomized mice with oestradiol (Kawaguchi et al 1995b). In vivo, partial reversal of bone loss by an NSAID in ovariectomized rats has been reported, but the effect was not sustained (Kimmel et al 1992). Transcriptional Regulation of COX-2 Expression in Osteoblasts COX-2 mRNA is generally expressed at low or undetectable levels in osteoblasts, and induction of expression by most agonists is
PROSTAGLANDINS, LEUKOTRIENES AND BONE rapid, transient and independent of new protein production (Min et al 1998; Kawaguchi et al 1995a; Pilbeam et al 1993, 1997a). In murine osteoblastic cells stably or transiently transfected with murine COX-2 promoter–luciferase reporter constructs, the 371 bp of the 5’-flanking COX-2 gene proximal to the transcription start site are adequate to mediate transcriptional induction by many of the agonists discussed above (Harrison et al 2000; Kawaguchi et al 1995a; Okada et al 2000b; Pilbeam et al 1997a; Wadleigh and Herschman 1999). cis-Acting sequences in this region which mediate induction of COX-2 expression in osteoblastic cells, as in a number of other cells, include a cyclic AMP response element (CRE) (Okada et al 2000b; Wadleigh and Herschman 1999), an activator protein-1 (AP-1 or Fos/Jun) binding site (Okada et al 2000b; Ogasawara et al 2001) and CCAAT enhancer binding protein (C/EBP) sites (Harrison et al 2000; Wadleigh and Herschman, 1999; Yamamoto et al 1995). Of specific relevance to osteoblasts is the ability of BMP-2 to induce COX-2 expression (Chikazu et al 2002). BMP-2 belongs to the TGFb superfamily and is a potent anabolic agent for bone, stimulating osteoblast differentiation, accelerating fracture healing and forming ectopic bone (Rosen and Wozney 2002; Yamaguchi et al 2000). In murine osteoblastic cells, the BMP-2 induction of COX-2 expression is mediated largely via a core binding factor activity 1 (Cbfa1) consensus sequence (5’AACCACA-3’) at 7267 to 7261 bp in the 5’-flanking region of the murine COX-2 gene (Chikazu et al 2002). Cbfa1, also known as Runx2, is an essential transcriptional factor for osteoblast differentiation (Ducy et al 1997). Cbfa1-deficient mice lacked both intramembranous and endochondral ossification because of the failure of osteoblasts to differentiate (Komori et al 1997). The nuclear factor of the activated T cells (NFAT) family of transcription factors may also play a role in the regulation of COX-2 gene transcription in osteoblasts by PTH, one of the most important systemic regulators of bone turnover. There are three NFAT core consensus sequences (5’-GGAAA-3’) at 7111 to 7107 bp, 790 to 786 bp and 777 to 773 bp in the murine COX-2 promoter (Chikazu et al 2001). The 777 to 773 bp sequence (NFAT77) is adjacent to the AP-1 site (5’-AGAGTCA3’) at 769 to 763 bp. NFAT and AP-1 (Fos/Jun) proteins interact cooperatively in other systems to mediate transcription, and the 777 to 763 bp sequence, GGAaagacagagTCA, is consistent with the minimum consensus sequence, GGA(N)9TCA, reported for NFAT/AP-1 composite sites. In murine osteoblastic MC3T3-E1 cells stably transfected with 7371 to +70 bp of the COX-2 5’-flanking region fused to a luciferase reporter, PTH luciferase activity was decreased 90% by mutation of this composite site (Chikazu et al 2001). Extracellular calcium ion concentration ([Ca2+]E) can be markedly elevated (up to 40 mM) in the region immediately surrounding actively resorbing osteoclasts (Silver et al 1988). Many tissues express a specific G-protein receptor for [Ca2+]E (Brown and MacLeod 2001). It has been proposed that [Ca2+]E is an important signalling factor in bone metabolism and may be one of the agents that couples bone resorption to formation. We have found that [Ca2+]E is a transcriptional inducer of COX-2 expression in primary osteoblastic cells and that the induction is dependent in part on an ERK1/2 pathway (Choudhary et al 2001). To study the transcriptional regulation of COX-2 in vivo, we generated mice transgenic for 7371 to +70 bp of the COX-2 5’flanking DNA fused to a luciferase reporter (Freeman et al 1999). Injection of these mice with lipopolysaccharide (LPS) induced COX-2 mRNA and luciferase activity in multiple tissues; however, the LPS induction of both luciferase activity and of endogenous COX-2 mRNA in calvarial bone was among the highest of all the tissues sampled, including brain, colon, heart,
291
lung and uterus. Hence, it is possible that osteoblasts are an important source of PGs that may influence neighbouring cells in the bone marrow and in the vascular network. ROLE OF PGs IN BONE RESORPTION Resorption in organ culture can be measured by the release of stable calcium or previously incorporated 45Ca from bone. PGs of the E series are the most potent of the prostaglandins in stimulating bone resorption in organ culture, with an effective concentration range of 1 nM to 10 mM (Klein and Raisz 1970; Raisz and Martin 1983). PGF2a is less effective than PGE2, stimulating resorption at concentrations of 0.1 mM and above; PGI2 can also stimulate resorption, while PGD2 is ineffective. Many factors that stimulate PG production also stimulate resorption in organ culture, and the resorption stimulated by such factors may be mediated in part by PG production (Akatsu et al 1991; Stern et al. 1985; Tashjian et al 1982, 1987), but the dependence of resorption in organ culture on stimulated PG production is quite variable. Role of COX-2-associated PGs in Osteoclastogenesis Formation of bone-resorbing osteoclasts requires a contactdependent interaction of osteoclastic precursors with cells of the osteoblastic lineage (Suda et al 1999). The molecule mediating this interaction, originally cloned as receptor activator of NF-kB (RANK) ligand or RANKL (Anderson et al 1997), is identical to TNF-related activation-induced cytokine (TRANCE) (Wong et al 1997) and osteoclast differentiating factor (ODF) (Yasuda et al 1998). RANKL is also a ligand for osteoprotegerin (OPG), a decoy receptor for RANKL that acts as a natural inhibitor of RANKL action, and is, therefore, sometimes called OPGL (Yasuda et al 1998). Osteoblastic cells produce both RANKL and OPG, while osteoclastic cells express the receptor RANK (Figure 25.2). Expression of RANKL is highly inducible by multiple stimulators of resorption. The ability of an agonist to stimulate resorption is roughly proportional to its ability to stimulate expression of RANKL and inhibit expression of OPG. Induction of RANKL has been shown to be essential for resorption by PGE2 (Tsukii et al 1998). Because bone marrow contains progenitors for both the osteoclastic and osteoblastic lineage, resorption agonists can stimulate the differentiation of new osteoclasts in cultured bone marrow. Osteoclastic cells are identified as multinucleated cells that stain for tartrate-resistant acid phosphatase, express calcitonin receptors, and resorb pits in mineralized bone. These types of studies consistently demonstrate a dependence of osteoclast formation on PG production. PGE1 and PGE2, but not PGF2a, stimulate osteoclast formation in marrow cultures (Collins and Chambers 1991; Kaji et al 1996). Agonists that stimulate PGdependent osteoclast formation in marrow cultures include IL-1 (Akatsu et al 1991; Lader and Flanagan 1998; Sato et al 1996), TNFa (Lader and Flanagan 1998), PTH (Inoue et al 1995; Sato et al 1997), 1,25-(OH)2D3 (Collins and Chambers 1992), IL-11 (Girasole et al 1994; Morinaga et al 1998), IL-6 (Tai et al 1997), IL-17 (Kotake et al 1999), phorbol ester (Amano et al 1994), and FGF-2 (Hurley et al 1998). Many stimulators of osteoclastogenesis also induce COX-2 expression, and PGs produced via the induction of COX-2 enhance the stimulated osteoclastogenesis. We used mice with the COX-2 gene disrupted (Morham et al 1995) and their wild-type littermates to study the role of COX-2 in osteoclastogenesis. In marrow cultures from COX-2 knockout (COX-27/7) mice, osteoclast formation stimulated by 1,25-(OH)2D3 or PTH was
292
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
reduced 60–70% compared to wild type (COX-2+/+) cultures (Okada et al 2000a). PGE2 production was markedly reduced in the cultures from COX-27/7 mice compared to cultures from wild type mice, and addition of exogenous PGE2 to the COX-2 7/7 cultures reversed the deficit in osteoclast formation. Treatment of COX-2+/+ cultures with NSAIDs (indomethacin and NS-398) mimicked the results observed in COX-27/7 cultures. The reduced response to 1,25-(OH)2D3 and PTH in COX-27/7 cultures was associated with reduced expression of RANKL. We have found similar results for osteoclastogenesis stimulated by FGF-2 and IL1 (Pilbeam, unpublished data). Osteoclasts can be formed in spleen cultures, which contain osteoclast precursors, if the spleen cells are co-cultured with osteoblastic cells which produce both the macrophage colony stimulating factor (M-CSF) and RANKL that are necessary for osteoclastogenesis. Alternatively, spleen cultures can be treated with M-CSF and soluble RANKL to replace the need for supporting osteoblastic cells. In spleen cell-only cultures, PGE2 has been shown to enhance the combined effects of RANKL and M-CSF to stimulate osteoclast formation (Wani et al 1999). In a similar system, we found that osteoclast formation was reduced 50% when the spleen cells came from COX-27/7 mice compared to cells from COX-2+/+ mice. This was associated with increased expression of granulocyte-macrophage colony stimulating factor (GM-CSF), an inhibitor of osteoclast formation in these cultures (Okada et al 2000a). In the presence of antibody to GM-CSF, osteoclastogenesis was stimulated equally in COX-2+/+ and COX27/7 spleen cultures. GM-CSF may act by diverting progenitor cells into the macrophage pathway that might otherwise differentiate into osteoclasts. Hence, PGs can enhance osteoclastogenesis by acting on osteoclast supporting cells to increase RANKL and on haematopoietic cells to decrease GM-CSF (Figure 25.2).
Inhibition of Osteoclast Activity by PGE2 PGE2 transiently inhibits the resorptive activity of osteoclasts when added to isolated osteoclasts in vitro (Fuller and Chambers 1989). Similarly, a transient inhibition of bone resorption and lysosomal enzyme release in mouse calvarial bone cultures by PGs has also been observed (Lerner et al. 1987). The physiological relevance of this transient inhibition is unknown.
PG Receptors and Bone Resorption EP receptor knockout mice and selective EP receptor agonists are being used to define the pathways by which PGE2 acts on bone resorption. Osteoclast formation in bone marrow cultures and mixed cultures of osteoblasts and spleen cells was impaired when cells from either EP2 or EP4 knockout animals were used (Li et al 2000; Sakuma et al 2000a, 2000b), confirming earlier results using selective EP agonists and antagonists (Ono et al 1998). A selective EP4 receptor antagonist can inhibit bone resorption in organ culture and osteoclast formation in marrow cultures (Tomita et al. 2002) Cultured calvaria from EP4 knockout animals showed an impaired resorptive response to PGE2 (Miyaura et al 2000). While both EP2 and EP4 are involved in the PGE2 effects on the osteoblastic support cells to stimulate RANKL and decrease OPG, EP2 is also involved in the effects of PGE2 on haematopoietic precursors to decrease GM-CSF production and stimulate osteoclastic dirfferentiation. Spleen cells cultured from EP2 knockout animals and treated with M-CSF, RANKL and PGE2 had an impaired osteoclastogenic response (Li et al 2000). The transient inhibitory effect of PGE2 on osteoclast activity may also involve both the EP4 and EP2 receptors (Mano et al 2000).
Figure 25.2 Putative roles for PGE2 in stimulating osteoclast formation and differentiation. Resorption agonists, such as PTH and 1,25(OH)2D3, as well as PGE2 itself, stimulate the expression of both COX-2 and RANKL in stromal cells. Many also inhibit the expression of the decoy receptor for RANKL, OPG. Interaction of RANKL on stromal cells with RANK on osteoclastic precursor cells is required for the osteoclastic precursor cells to differentiate into mature osteoclasts. PGE2 may also act on haematopoietic cells to inhibit GM-CSF expression, resulting in increased entrance of progenitor cells into the osteoclastic differentiation pathway
PROSTAGLANDINS, LEUKOTRIENES AND BONE Studies using cells from EP2 and EP4 knockout animals confirm the dual pathways for PGE2 effects. The RANKL response to PGE2 is reduced when either receptor is absent (Li et al 2002a). In vivo studies show that the hypercalcaemic response to PGE2 is blunted in EP2 knockout mice (Li et al 2002b), while the response to LPS is blunted in EP4 knockout mice (Sakuma et al 2000a, 2000b).
Role of PGs in Bone Loss In Vivo PG production is frequently increased in inflammatory processes, such as rheumatoid arthritis (Crofford et al 1994). Because PGs are potent stimulators of bone resorption, it is often assumed that PGs contribute to the bone loss and cartilage destruction associated with inflammatory diseases. Some support for this conclusion comes from studies showing that NSAIDs decrease alveolar bone loss in periodontitis (Howell et al 1991; Jeffcoat et al 1993). However, cytokines are elevated in inflammatory diseases and are themselves potent inducers of bone resorption independent of their effects on PG production. In addition, cytokines can inhibit bone matrix proteins independently of cytokine-induced PG production (Rosenquist et al 1996). PGs may enhance and prolong the resorptive responses to proinflammatory agents. Although the absence of COX-2 expression causes no marked abnormality in skeletal bone mass in young COX-27/7 mice, the response to resorption agonists applied acutely is decreased in COX-27/7 mice (Okada et al 2000a). We injected PTH at high doses above the calvaria of COX-27/7 and COX-2+/+ mice, a protocol that has been shown to stimulate localized resorption at the site of injection resulting in systemic hypercalcaemia (Zhao et al 1999). After 3 days of injection, PTH caused hypercalcaemia in COX-2+/+ but not COX-27/7 mice (Okada et al 2000a). In another study, the prolonged in vivo resorptive response to IL-1 was also found to depend on PGs (Boyce et al 1989). In vivo injection of IL-1 above the calvaria in mice for 3 days stimulated PG-independent resorption when the mice were sacrificed 24 h after the last injection. However, after the IL-1 injections were stopped, resorption continued for 3–4 weeks and this resorption was PGdependent. The role of PGs in the resorption stimulated by proinflammatory agents may be most evident when the generation of new osteoclasts has become the factor limiting the rate of resorption. On the other hand, since PGs can stimulate new bone formation (see below), it is possible that PGs might have a compensatory effect on the bone loss induced by proinflammatory agents.
ROLE OF PGs IN BONE FORMATION Biphasic Effects of PGs on Bone Formation Studies in cell and organ culture have shown that PGE2 can have both stimulatory and inhibitory effects on bone formation, while in vivo studies in rats have constantly demonstrated a potent anabolic effect of PGE2. Similar to the effects of PTH, PGE2 can increase both periosteal and endosteal bone formation in the rat and increase total bone mass (Jee and Ma 1997; Lin et al 1994; Suponitzky and Weinreb 1998). Anabolic effects are also seen in foetal rat calvarial organ cultures, in which PGE2 can stimulate both cell replication and differentiation (Woodiel et al 1996). At high concentrations, PGs can inhibit collagen synthesis in cell and organ culture (Fall et al 1994). This inhibitory effect appears to occur largely via transcriptional inhibition of collagen and to be mediated by the FP receptor rather than an EP receptor.
293
Role of COX-2-associated PGs in Osteoblastogenesis The most consistent anabolic effect of PGE2 is to increase osteoblastic differentiation. PGE2 stimulates the formation and differentiation of osteoblastic colonies in marrow stromal cell and primary calvarial cell cultures (Flanagan and Chambers 1992; Scutt and Bertram 1995). PGE2 given in vivo enhances osteoblastic differentiation in explanted marrow stromal cells (Weinreb et al 1997). Marrow stromal cells or primary calvarial cells cultured from COX-27/7 mice have markedly inhibited osteoblastic differentiation compared to cells from COX-2+/+ mice (Okada et al 2000c). This inhibition can be mimicked in COX-2+/+ cells by the addition of NSAIDs, and differentiation in COX-27/7 cultures can be increased to that seen in COX-2+/+ cultures by addition of exogenous PGE2. In primary calvarial cell cultures the increased osteoblastic differentiation seen in cells from COX-2+/+ mice relative to cells from COX-27/7 mice is not due to increased cell proliferation. Examination of cell numbers and 3H-thymidine incorporation into newly replicating DNA shows that cultured COX-27/7 calvarial osteoblasts proliferate more rapidly than COX-2+/+ osteoblasts (unpublished data, Pilbeam). Our preliminary studies in COX-27/7 mice in vivo have suggested that absence of COX-2 may result in a small deficit in bone formation in vivo (Okada et al 2000c). However, studies of in vivo bone turnover in adult COX-27/7 mice are complicated by renal abnormalities in these mice (Morham et al 1995; Dinchuk et al 1995). These abnormalities are reported to kill 20% or more of adult mice in a C57Bl/66129 background (Norwood et al 2000). Renal abnormalities may have important consequences for parathyroid hormone and vitamin D metabolism, major regulators of bone turnover. Hence, we are trying to breed the COX-2 knockout into other genetic backgrounds that may be free of this problem for further studies. The receptor pathway for the anabolic effect is still being elucidated. Based on early studies with selective agonists in rat calvarial organ culture and marrow cell culture, a role for the EP2 receptor was suggested, but subsequent studies in marrow cultures and in vivo support mediation by the EP4 receptor (Machwate et al 2001; Weinreb et al 1999; Yoshida et al 2002). Role of PGs in the Anabolic Effects of Mechanical Loading on Bone Many studies have suggested that PGs mediate some of the anabolic effects of mechanical loading on bone. Mechanical loading of bone explants can increase PG production and result in new bone formation (Cheng et al 1997; Lanyon 1992; Rawlinson et al 1991). In humans, in situ microdialysis showed a 2.5–3.5-fold increase in the release of PGE2 in the tibial metaphysis after loading (Thorsen et al 1997). Loading induced new bone formation in an isolated avian ulna preparation (Pead and Lanyon 1989) and in externally loaded tail vertebrae in vivo of rats (Chow and Chambers 1994) was blocked by NSAIDs. In vivo, an increase in endosteal bone formation following a single short period of bending applied to rat tibiae was prevented by a selective COX-2 inhibitor (NS-398) (Forwood 1996). Interstitial fluid flow in the lacunar–canalicular network in bone is thought to play a role in transducing the response to physiological loading (Hillsley and Frangos, 1993; Piekarski and Munro 1977; Turner et al 1994). Osteocytes, in contact with each other and the osteoblasts lining the mineralized matrix via extended cell processes connected by gap junctions, are thought to be the main strain sensing network (Kufahl and Saha 1990). Osteoblasts and osteocytes subjected to fluid shear stress produce PGs (Klein-Nulend et al 1997; Pavalko et al 1998; Reich and Frangos 1993). When osteoblastic cells are subjected to fluid flow,
294
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
there is an early release of PGE2, probably due to release of arachidonic acid converted to PGs by constitutively expressed COX-1, followed by sustained PG production due to induction of COX-2 (Klein-Nulend et al 1997; Pavalko et al 1998). We have shown that fluid shear stresses as low as 0.1 dyne/cm2 rapidly induce new COX-2 gene transcription in osteoblasts and that the induction is dependent on the mitogen-activated protein signalling pathway that activates ERK 1 and 2 (Wadhwa et al 2002). This transcriptional activation may be mediated by a combination of C/EBP-b, CRE and AP-1 trans-acting sites (Ogasawara et al 2001). EFFECTS OF NSAIDs ON BONE IN HUMANS A few studies have examined the effects of non-selective NSAIDs on bone mineral density (BMD) in humans, and there are no published studies examining the effects of selective COX-2 inhibitors on BMD. In a study of post-menopausal women 65 years of age or older, there was no significant difference in Ntelopeptide cross-link excretion, a marker of bone resorption, between self-reported users of NSAIDs or aspirin compared to non-users (Lane et al 1997). In a similar cohort, the use of aspirin or NSAIDs was found to be associated with small but significant increase in hip and spine BMD, but there was no clinically significant protective effect on risk for fracture (Bauer et al 1996). In a study of older women from Rancho Bernardo, CA, the regular use of propionic acid NSAIDs (ibuprofen, naproxen, ketoprofen), but not acetic acid NSAIDs (indomethacin, diclofenac, sulindac, tolmetin), was associated with higher BMD at multiple skeletal sites (Morton et al 1998). When women with selfreported osteoarthritis were excluded, significantly higher BMD was observed in the hip in propionic acid NSAID users. It is possible that continuous inhibition of COX-2 activity might not have been attained in these studies with non-selective NSAIDs. In addition, non-selective NSAIDs in vitro can have biphasic effects—low concentrations can actually increase PG production (Lindsley and Smith 1990; Raisz et al 1989). These studies may also have been biased by the inclusion of subjects with conditions that might be associated with higher levels of proinflammatory cytokines. Studies with selective inhibitors of COX-2 at doses that effectively inhibit COX-2 in healthy subjects may help to clarify the role of PGs in bone metabolism. Non-selective NSAIDs have also been used to examine the effects of PGs after orthopaedic surgical procedures. Several studies have shown that perioperative treatment with aspirin or other NSAIDs can prevent heterotropic ossification, a complication of hip arthroplasty that can adversely affect the outcome (Kienapfel et al 1999; Neal et al 2000; Nilsson and Persson 1999). Because PGs may enhance the bone remodelling needed for fracture healing, there is concern that NSAIDs might adversely affect fracture healing. Studies suggest that NSAIDs can inhibit repair of fractured femurs (Altman et al 1995; Reikeraas and Engebretsen, 1998) and spinal fusions (Dimar et al 1996) in rats and the fixation of implants in femora of rabbits (Jacobsson et al 1994), while exogenously applied PGs can stimulate callus formation in rabbits (Keller et al 1993). It is not yet clear whether selective COX-2 inhibitors will have different effects from the nonselective NSAIDs. LEUKOTRIENES AND BONE Products of the 5-lipoxygenase (5-LO) pathway have been far less studied than the products of the COX pathway in bone. Most studies have examined effects on bone resorption. Leukotrienes, peptido-leukotrienes and hydroxyeicosatetraenoic acids (HETEs)
were shown to have biphasic effects on resorption in neonatal mouse calvarial cultures, with low doses stimulating and high doses inhibiting resorption (Meghji et al 1988). LTB4, LTC4, LTE4 and 5-HETE were more potent stimulators of resorption than 12HETE and LTD4. The effects of LTB4, LTC4 and 12-HETE were partially inhibited by indomethacin, suggesting mediation by PGs. Multinucleated cells in giant cell tumours of bone were found to produce an activity, identified as leukotrienes and 5-HETE, which stimulated isolated osteoclasts to resorb bone in vitro (Gallwitz et al 1993). Peptido-leukotrienes (LTC4, LTD4 and LTE4) stimulated isolated avian osteoclasts to form resorption lacunae (Garcia et al 1996a). LTB4 increased bone resorption and osteoclast numbers in mice when injected locally above the calvariae and stimulated resorption in mouse calvarial organ cultures (Garcia et al 1996b). Avian osteoclasts were shown to express both high- and lowaffinity LTB4 receptors, and LTB4 increased pit formation by these osteoclasts (Flynn et al 1999). Hence, 5-LO metabolites generally stimulate resorption in organ culture and isolated osteoclast activity, while PGs stimulate resorption in organ culture but inhibit isolated osteoclast activity. Less is known about the effects of 5-LO metabolites on bone formation but limited data suggest that, in contrast to PGs, they may inhibit bone formation. 5-Hydroxyeicosatetraenoic acid (HETE) and leukotriene B4 (LTB4) have been shown to inhibit BMP-2-induced bone formation in mouse calvarial organ culture and to block BMP-2-induced formation of mineralized nodules in foetal rat calvarial cell culture (Traianedes et al 1998). Bonewald et al (1997) found increased cortical bone thickness in 5’lipoxygenase null mice. SUMMARY Osteoblasts produce abundant PGs in bone, and this production is highly regulated by local and systemic regulators of bone metabolism. PGs are complex regulators of bone cell function in vitro, being able to stimulate or inhibit both bone resorption and formation, and the net balance of these effects under physiological or pathological conditions in vivo is not yet clear. Some of the complexity of PG actions on bone may be explained by the multiplicity of receptors for PGs. Exogenously applied PGE2 is a potent stimulator of new bone formation and could be a useful therapeutic agent for osteoporosis if the unwanted side-effects of systemically administered PGE2 could be eliminated. Further studies to clarify the specific pathways of PG action in bone may make it possible to manipulate PG pathways to achieve these beneficial effects. ACKNOWLEDGEMENTS The authors’ participation in this chapter was supported by National Institutes of Health Grants DK-38933 and AR-47673 to C.P. and AR-18063 to L.G.R, and a Donaghue Foundation Investigator Award to C.P. REFERENCES Akatsu T, Takahashi N, Udagawa N et al (1991) Role of prostaglandins in interleukin-1-induced bone resorption in mice in vitro. J Bone Min Res, 6, 183–189. Altman RD, Latta LL, Keer R et al (1995) Effect of non-steroidal antiinflammatory drugs on fracture healing: a laboratory study in rats. J Orthop Trauma, 9, 392–400. Amano S, Hanazawa S, Kawata Y et al (1994) Phorbol myristate acetate stimulates osteoclast formation in 1a,25-dihydroxyvitamin D3-primed
PROSTAGLANDINS, LEUKOTRIENES AND BONE mouse embryonic calvarial cells by a prostaglandin-dependent mechanism. J Bone Min Res, 9, 465–472. Anderson DM, Maraskovsky E, Billingsley WL et al (1997) A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature, 390, 175–179. Aubin JE and Triffitt JT (2002) Mesenchymal stem cells and osteoblast differentiation. In Bilezikian JP, Raisz LG and Rodan GA (eds), Principles of Bone Biology. New York, Academic Press, 59–81. Bauer DC, Orwoll ES, Fox KM et al (1996) Aspirin and NSAID use in older women: effect on bone mineral density and fracture risk. Study of Osteoporotic Fractures Research Group. J Bone Min Res, 11, 29–35. Bonewald LF, Flynn M, Qiao M et al (1997) Mice lacking 5-lipoxygenase have increased cortical bone thickness. Adv Exp Med Biol, 433, 299– 302. Boyce BF, Aufdemorte TB, Garrett IR et al (1989) Effects of interleukin-1 on bone turnover in normal mice. Endocrinology, 125, 1142–1150. Brown EM and MacLeod RJ (2001) Extracellular calcium sensing and extracellular calcium signaling. Physiol Rev, 81, 239–297. Chen QR, Miyaura C, Higashi S et al (1997) Activation of cytosolic phospholipase A2 by platelet-derived growth factor is essential for cyclooxygenase-2-dependent prostaglandin E2 synthesis in mouse osteoblasts cultured with interleukin-1. J Biol Chem, 272, 5952–5958. Cheng MZ, Zaman G, Rawlinson SC et al (1997) Enhancement by sex hormones of the osteoregulatory effects of mechanical loading and prostaglandins in explants of rat ulnae. J Bone Min Res, 12, 1424–1430. Chikazu D, Voznesensky O, Kream B et al (2001) Parathyroid hormone induces cyclooxygenase-2 in osteoblasts via a composite NFAT/AP-1 binding site. J Bone Min Res, 16(S1), S307 (abstr). Chikazu D, Li X, Kawaguchi H et al (2002) BMP-2 induces COX-2 in osteoblasts via a Cbfa1 binding site: role in effects of BMP-2 in vitro and in vivo. J Bone Min Res. Choudhary S, Godwin S, Mellas S et al (2001) Extracellular calcium transriptionally induces cyclooxygenase-2 in osteoblasts via an ERK signaling pathway. J Bone Min Res, 16(S1), S427 (abstr). Chow JW and Chambers TJ (1994) Indomethacin has distinct early and late actions on bone formation induced by mechanical stimulation. Am J Physiol, 267, E287–E292. Collins DA and Chambers TJ (1991) Effect of prostaglandins E1, E2, and F2a on osteoclast formation in mouse bone marrow cultures. J Bone Min Res, 6, 157–164. Collins DA and Chambers TJ (1992) Prostaglandin E2 promotes osteoclast formation in murine hematopoietic cultures through an action on hematopoietic cells. J Bone Min Res, 7, 555–561. Crofford LJ, Wilder RL, Ristimaki AP et al (1994) Cyclooxygenase-1 and -2 expression in rheumatoid synovial tissues. J Clin Invest, 93, 1095– 1101. Dimar JR, Ante WA, Zhang YP and Glassman SD (1996 ) The effects of non-steroidal antiinflammatory drugs on posterior spinal fusions in the rat. Spine, 21, 1870–1876. Dinchuk JE, Car BD, Focht RJ et al (1995) Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature, 378, 406–409. Ducy P, Zhang R, Geoffroy V et al (1997) Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell, 89, 747–754. Fall PM, Breault DT and Raisz LG (1994) Inhibition of collagen synthesis by prostaglandins in the immortalized rat osteoblastic cell line, Py1a: structure activity relations and signal transduction mechanisms. J Bone Min Res, 9, 1935–1943. Feyen JHM and Raisz LG (1987) Prostaglandin production by calvariae from sham operated and oophorectomized rats: effects of 17b-estradiol in vivo. Endocrinology, 121, 819–821. Feyen JH, van der WG, Moonen P et al (1984) Stimulation of arachidonic acid metabolism in primary cultures of osteoblast-like cells by hormones and drugs. Prostaglandins, 28, 769–781. Flanagan AM and Chambers TJ (1992) Stimulation of bone nodule formation in vitro by prostaglandins E1 and E2. Endocrinology, 130, 443–448. Flynn MA, Qiao M, Garcia C et al (1999) Avian osteoclast cells are stimulated to resorb calcified matrices by and possess receptors for leukotriene B4. Calcif Tissue Int, 64, 154–159. Forwood MR (1996) Inducible cyclo-oxygenase (COX-2) mediates the induction of bone formation by mechanical loading in vivo. J Bone Min Res, 11, 1688–1693.
295
Forwood MR, Kelly WL and Worth NF (1998) Localisation of prostaglandin endoperoxide H synthase (PGHS)-1 and PGHS-2 in bone following mechanical loading in vivo. Anat Rec, 252, 580–586. Freeman A, Mallico E, Voznesensky O et al (1999) Transcriptional regulation of prostaglandin G/H synthase-2 in transgenic mice. J Bone Min Res, 14S(1), S472 (abstract). Fuller K and Chambers TJ (1989) Effect of arachidonic acid metabolites on bone resorption by isolated rat osteoclasts. J Bone Min Res, 4, 209– 215. Gallwitz WE, Mundy GR, Lee CH et al (1993) 5-Lipoxygenase metabolites of arachidonic acid stimulate isolated osteoclasts to resorb calcified matrices. J Biol Chem, 268, 10087–10094. Garcia C, Qiao M, Chen D et al (1996a) Effects of synthetic peptidoleukotrienes on bone resorption in vitro. J Bone Min Res, 11, 521–529. Garcia C, Boyce BF, Gilles J et al (1996b) Leukotriene B4 stimulates osteoclastic bone resorption both in vitro and in vivo. J Bone Min Res, 11, 1619–1627. Girasole G, Passeri G, Jilka RL and Manolagas SC (1994) Interleukin-11: a new cytokine critical for osteoclast development. J Clin Invest, 93, 1516–1524. Harrison JR, Lorenzo JA, Kawaguchi H et al (1994) Stimulation of prostaglandin E2 production by interleukin-1 and transforming growth factor-a in osteoblastic MC3T3-E1 cells. J Bone Min Res, 9, 817–823. Harrison JR, Kelly PL and Pilbeam CC (2000) Involvement of C/EBP transcription factors in prostaglandin G/H synthase-2 gene induction by interleukin-1 in osteoblastic MC3T3-E1 cells. J Bone Min Res, 15, 1138–1146. Hillsley MV and Frangos JA (1993) Review: bone tissue engineering: the role of interstitial fluid flow. Biotechnol Bioeng, 43, 573–581. Howell TH, Jeffcoat MK, Goldhaber P et al (1991) Inhibition of alveolar bone loss in beagles with the NSAID naproxen. J Periodont Res, 26, 498–501. Hurley MM, Lee SK, Raisz LG et al (1998) Basic fibroblast growth factor induces osteoclast formation in murine bone marrow. Bone, 22, 309– 316. Inoue H, Tanaka N and Uchiyama C (1995) Parathyroid hormone increases the number of tartrate-resistant acid phosphatase-positive cells through prostaglandin E2 synthesis in adherent cell culture of neonatal rat bones. Endocrinology, 136, 3648–3656. Jacobsson SA, Djerf K, Ivarsson I and Wahlstrom O (1994) Effect of diclofenac on fixation of hydroxyapatite-coated implants. An experimental study. J Bone Joint Surg Br, 76, 831–833. Jee WS and Ma YF (1997) The in vivo anabolic actions of prostaglandins in bone. Bone, 21, 297–304. Jeffcoat MK, Reddy MS, Moreland LW and Koopman WJ (1993) Effects of non-steroidal antiinflammatory drugs on bone loss in chronic inflammatory disease. Ann NY Acad Sci, 696, 292–302. Kaji H, Sugimoto T, Kanatani M et al (1996) Prostaglandin E2 stimulates osteoclast-like cell formation and bone-resorbing activity via osteoblasts: role of cAMP-dependent protein kinase. J Bone Min Res, 11, 62–71. Kawaguchi H, Raisz LG, Voznesensky OS et al (1994) Regulation of the two prostaglandin G/H synthases by parathyroid hormone, interleukin-1, cortisol and prostaglandin E2 in cultured neonatal mouse calvariae. Endocrinology, 135, 1157–1164. Kawaguchi H, Pilbeam CC, Gronowicz G et al (1995a) Regulation of inducible prostaglandin G/H synthase mRNA levels and promoter activity by basic fibroblast growth factor in osteoblastic cells. J Clin Invest, 96, 923–930. Kawaguchi H, Pilbeam C, Vargas S et al (1995b) Ovariectomy enhances and estrogen replacement inhibits the activity of marrow factors which stimulate prostaglandin production in cultured mouse calvariae. J Clin Invest, 96, 539–548. Kawaguchi H, Nemoto K, Raisz LG et al (1996) Interleukin-4 inhibits prostaglandin G/H synthase-2 and cytosolic phospholipase A2 induction in neonatal mouse parietal bone cultures. J Bone Min Res, 11, 358–366 Keller J, Klamer A, Bak B and Suder P (1993) Effect of local prostaglandin E2 on fracture callus in rabbits. Acta Orthop Scand, 64, 59–63. Kienapfel H, Koller M, Wust A et al (1999) Prevention of heterotopic bone formation after total hip arthroplasty: a prospective randomised study comparing postoperative radiation therapy with indomethacin medication. Arch Orthop Trauma Surg, 119, 296–302.
296
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Kimble RB, Matayoshi AB, Vannice JL et al (1995) Simultaneous block of interleukin-1 and tumor necrosis factor is required to completely prevent bone loss in the early postovariectomy period. Endocrinology, 136, 3054–3061. Kimmel DB, Coble T and Lane N (1992) Long-term effect of naproxen on cancellous bone in ovariectomized rats. Bone, 13, 167–172. Klein-Nulend J, Burger EH, Semeins CM et al (1997) Pulsating fluid flow stimulates prostaglandin release and inducible prostaglandin G/H synthase mRNA expression in primary mouse bone cells. J Bone Min Res, 12, 45–51. Klein DC and Raisz LG (1970) Prostaglandins: stimulation of bone resorption in tissue culture. Endocrinology, 86, 1436–1440. Komori T, Yagi H, Nomura S et al (1997) Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell, 89,755–764. Kotake S, Udagawa N, Takahashi N et al (1999) IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J Clin Invest, 103, 1345–1352. Kufahl RH and Saha S (1990) A theoretical model for stress-generated fluid flow in the canaliculi-lacunae network in bone tissue. J Biomechanics, 23, 171–180. Lader CS and Flanagan AM (1998) Prostaglandin E2, interleukin 1a, and tumor necrosis factor-a increase human osteoclast formation and bone resorption in vitro. Endocrinology, 139, 3157–3164. Lane NE, Bauer DC, Nevitt MC et al (1997) Aspirin and non-steroidal antiinflammatory drug use in elderly women: effects on a marker of bone resorption. The Study of Osteoporotic Fractures Research Group. J Rheumatol, 24, 1132–1136. Lanyon LE (1992) Control of bone architecture by functional load bearing. J Bone Min Res, 7, S369–S375. Lerner UH, Ransjo M and Ljunggren O (1987) Prostaglandin E2 causes a transient inhibition of mineral mobilization, matrix degradation, and lysosomal enzyme release from mouse calvarial bones in vitro. Calcif Tissue Int, 40, 323–331. Li X, Okada Y, Pilbeam CC et al (2000) Knockout of the murine prostaglandin EP2 receptor impairs osteoclastogenesis in vitro. Endocrinology, 141, 2054–2061. Li X, Pilbeam CC, Pan L et al (2002a). Effects of prostaglandin E2 on gene expression in primary osteoblastic cells from prostaglandin receptor knockout mice. Bone, 30, 567–573. Li X, Tomita M, Pilbeam CC et al (2002b) Prostaglandin receptor EP2 stimulated hypercalcemia in mice in vivo. Prostagland Lipid Mediat, 67, 173–180. Lin BY, Jee WSS, Ma YF et al (1994) Effects of prostaglandin E2 and risedronate administration on cancellous bone in older female rats. Bone, 15, 489–496. Lindsley HB and Smith DD (1990) Enhanced prostaglandin E2 secretion by cytokine-stimulated human synoviocytes in the presence of subtherapeutic concentrations of non-steroidal antiinflammatory drugs. Arthritis Rheum, 33, 1162–1169. Machwate M, Harada S, Leu CT et al (2001) Prostaglandin receptor EP4 mediates the bone anabolic effects of PGE2. Mol Pharmacol, 60, 36–41. Majeska RJ and Gronowicz GA (2002) Current methodologic issues in cell and tissue culture. In Bilezikian JP, Raisz LG and Rodan GA (eds), Principles of Bone Biology. New York, Academic Press, 1529–1539. Mano M, Arakawa T, Mano H et al (2000) Prostaglandin E2 directly inhibits bone-resorbing activity of isolated mature osteoclasts mainly through the EP4 receptor. Calcif Tissue Int, 67, 85–92. Meghji S, Sandy JR, Scutt AM et al (1988) Stimulation of bone resorption by lipoxygenase metabolites of arachidonic acid. Prostaglandins, 36, 139–149. Min YK, Rao Y, Okada Y et al (1998) Regulation of prostaglandin G/H synthase-2 expression by interleukin-1 in human osteoblast-like cells. J Bone Min Res, 13, 1066–1075. Miyaura C, Kenichiro K, Masuzawa T et al (1995) Endogenous boneresorbing factors in estrogen deficiency: cooperative effects of IL-1 and IL-6. J Bone Min Res, 10, 1365–1373. Miyaura C, Inada M, Suzawa T et al (2000) Impaired bone resorption to prostaglandin E2 in prostaglandin E receptor EP4-knockout mice. J Biol Chem, 275, 19819–19823. Morham SG, Langenbach R, Loftin CD et al (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 83, 473–482.
Morinaga Y, Fujita N, Ohishi K et al (1998) Suppression of interleukin11-mediated bone resorption by cyclooxygenases inhibitors. J Cell Physiol, 175, 247–254. Morton DJ, Barrett-Connor EL and Schneider DL (1998) Non-steroidal antiinflammatory drugs and bone mineral density in older women: the Rancho Bernardo study. J Bone Min Res, 13, 1924–1931. Murakami M, Kuwata H, Amakasu Y et al (1997) Prostaglandin E2 amplifies cytosolic phospholipase A2- and cyclooxygenase-2-dependent delayed prostaglandin E2 generation in mouse osteoblastic cells. Enhancement by secretory phospholipase A2. J Biol Chem, 272, 19891–19897. Neal BC, Rodgers A, Clark T et al (2000) A systematic survey of 13 randomized trials of non-steroidal antiinflammatory drugs for the prevention of heterotopic bone formation after major hip surgery. Acta Orthop Scand, 71, 122–128. Nilsson OS and Persson PE (1999) Heterotopic bone formation after joint replacement. Curr Opin Rheumatol, 11, 127–131. Norwood VF, Morham SG and Smithies O (2000) Postnatal development and progression of renal dysplasia in cyclooxygenase-2 null mice. Kidney Int, 58, 2291–2300. Ogasawara A, Arakawa T, Kaneda T et al (2001) Fluid shear stressinduced cyclooxygenase-2 expression Is mediated by C/EBPb, CREB, and AP-1 in osteoblastic MC3T3-E1 cells. J Biol Chem 276, 7048–7054. Okada Y, Lorenzo JA, Freeman AM et al (2000a). Prostaglandin G/H synthase-2 is required for maximal formation of osteoclast-like cells in culture. J Clin Invest, 105, 823–832. Okada Y, Voznesensky O, Herschman H et al (2000b) Identification of multiple cis-acting elements mediating the induction of prostaglandin G/H synthase-2 by phorbol ester in murine osteoblastic cells. J Cell Biochem, 78, 197–209. Okada Y, Tomita M, Gronowicz G et al (2000c) Effects of cyclooxygenase-2 gene disruption on osteoblastic function. J Bone Min Res, 15(S1), S217 (abstr). Ono K, Akatsu T, Murakami T et al (1998) Important role of EP4, a subtype of prostaglandin (PG) E receptor, in osteoclast-like cell formation from mouse bone marrow cells induced by PGE2. J Endocrinol, 158, R1–R5. Onoe Y, Miyaura C, Kaminakayashiki T et al (1996) IL-13 and IL-4 inhibit bone resorption by suppressing cyclooxygenase-2-dependent prostaglandin synthesis in osteoblasts. J Immunol, 156, 758–764. Oshima T, Yoshimoto T, Yamamoto S et al (1991) cAMP-dependent induction of fatty acid cyclooxygenase mRNA in mouse osteoblastic cells (MC3T3-E1). J Biol Chem, 266, 13621–13626. Pavalko FM, Chen NX, Turner CH et al (1998) Fluid shear-induced mechanical signaling in MC3T3-E1 osteoblasts requires cytoskeleton– integrin interactions. Am J Physiol, 275, C1591–C1601. Pead MJ and Lanyon LE (1989) Indomethacin modulation of load-related stimulation of new bone formation in vivo. Calcif Tissue Int, 45, 34–40. Piekarski K and Munro M (1977) Transport mechanism operating between blood supply and osteocytes in long bones. Nature, 269, 80– 82. Pilbeam CC, Kawaguchi H, Hakeda Y et al (1993) Differential regulation of inducible and constitutive prostaglandin endoperoxide synthase in osteoblastic MC3T3-E1 cells. J Biol Chem, 268, 25643–25649. Pilbeam CC, Raisz LG, Voznesensky O et al (1994) Autoregulation of inducible prostaglandin G/H synthase in osteoblastic cells by prostaglandins. J Bone Min Res, 10, 406–414. Pilbeam C, Bernecker P, Harrison J et al (1995) Retinoic acid inhibits induction of prostaglandin G/H synthase-2 mRNA and promoter activity in MC3T3-E1 osteoblastic cells. J Bone Min Res, 10(S1), S496 (abstr). Pilbeam C, Rao Y, Voznesensky O et al (1997a) Transforming growth factor-b1 regulation of prostaglandin G/H synthase-2 expression in osteoblastic MC3T3-E1 cells. Endocrinology, 138, 4672–4682. Pilbeam CC, Fall PM, Alander CB and Raisz LG (1997b) Differential effects of nonsteroidal antiinflammatory drugs on constitutive and inducible prostaglandin G/H synthase in cultured bone cells. J Bone Min Res, 12, 1198–1203. Pilbeam CC, Harrison JR and Raisz LG (2002) Prostaglandins and bone metabolism. In Bilezikian JP, Raisz LG and Rodan GA (eds), Principles of Bone Biology. New York, Academic Press, 979–994. Raisz LG, Sandberg AL, Goodson JM et al (1974) Complementdependent stimulation of prostaglandin synthesis and bone resorption. Science, 185, 789–791.
PROSTAGLANDINS, LEUKOTRIENES AND BONE Raisz LG, Vanderhoek JY, Simmons HA et al (1979) Prostaglandin synthesis by fetal rat bone in vitro: evidence for a role of prostacyclin. Prostaglandins, 17, 905–914. Raisz LG and Martin TJ (1983) Prostaglandins in bone and mineral metabolism. In Peck, WA (ed.), Bone and Mineral Research, Annual 2. Amsterdam, Elsevier Science, 286–310. Raisz LG, Simmons HA and Fall PM (1989) Biphasic effects of nonsteroidal antiinflammatory drugs on prostaglandin production by cultured rat calvariae. Prostaglandins, 37, 559–565. Raisz LG, Alander CB, Fall PM and Simmons HA (1990) Effects of prostaglandin F2a on bone formation and resorption in cultured neonatal mouse calvariae: role of prostaglandin E2 production. Endocrinology, 126, 1076–1079. Rawlinson SCF, El-Haj AJ, Minter SL et al (1991) Loading-related increases in prostaglandin production in cores of adult canine cancellous bone in vitro: a role for prostacyclin in adaptive bone remodeling? J Bone Min Res, 6, 1345–1352. Reich KM and Frangos JA (1993) Protein kinase C mediates flow-induced prostaglandin E2 production in osteoblasts. Calcif Tissue Int, 52, 62– 66. Reikeraas O and Engebretsen L (1998) Effects of ketoralac tromethamine and indomethacin on primary and secondary bone healing. An experimental study in rats. Arch Orthop Trauma Surg, 118, 50–52. Rosen V and Wozney JM (2002) Bone morphogenetic proteins. In Bilezikian JP, Raisz LG, and Rodan GA (eds), Principles of Bone Biology. New York, Academic Press, 919–928. Rosenquist JB, Ohlin A and Lerner UH (1996) Cytokine-induced inhibition of bone matrix proteins is not mediated by prostaglandins. Inflamm Res, 45, 457–463. Sakuma Y, Tanaka K, Suda M et al (2000a) Impaired bone resorption by lipopolysaccharide in vivo in mice deficient in the prostaglandin E receptor EP4 subtype. Infect Immun, 68, 6819–6825. Sakuma Y, Tanaka K, Suda M et al (2000b) Crucial involvement of the EP4 subtype of prostaglandin E receptor in osteoclast formation by proinflammatory cytokines and lipopolysaccharide. J Bone Min Res, 15, 218–227. Sato T, Morita I, Sakaguchi K et al (1996) Involvement of prostaglandin endoperoxide H synthase-2 in osteoclast-like cell formation induced by interleukin-1b. J Bone Min Res, 11, 392–400. Sato T, Morita I and Murota S (1997) Prostaglandin E2 mediates parathyroid hormone induced osteoclast formation by cyclic AMP independent mechanism. Adv Exp Med Biol, 407, 383–386. Scutt A and Bertram P (1995) Bone marrow cells are targets for the anabolic actions of prostaglandin E2 on bone: induction of a transition from non-adherent to adherent osteoblast precursors. J Bone Min Res, 10, 474–487. Silver IA, Murrills RJ and Etherington DJ (1988) Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp Cell Res, 175, 266–276. Stern PH, Krieger NS, Nissenson RA et al (1985) Human transforming growth factor-a stimulates bone resorption in vitro. J Clin Invest, 76, 2016–2019. Suda M, Tanaka K, Yasoda A et al (1998) Prostaglandin E2 (PGE2) autoamplifies its production through EP1 subtype of PGE receptor in mouse osteoblastic MC3T3-E1 cells. Calcif Tissue Int, 62, 327–331. Suda T, Takahashi N, Udagawa N et al (1999) Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocrine Rev, 20, 345–357. Suponitzky I and Weinreb M (1998) Differential effects of systemic prostaglandin E2 on bone mass in rat long bones and calvariae. J Endocrinol, 156, 51–57. Tai H, Miyaura C, Pilbeam CC et al (1997) Transcriptional induction of cyclooxygenase-2 in osteoblasts is involved in interleukin-6-induced osteoclast formation. Endocrinology, 138, 2372–2379. Takahashi Y, Taketani Y, Endo T et al (1994). Studies on the induction of cyclooxygenase isozymes by various prostaglandins in mouse osteoblastic cell line with reference to signal transduction pathways. Biochim Biophys Acta, 1212, 217–224.
297
Takahashi N, Udagawa N, Takami M and Suda T (2002) Cells of bone: osteoclast generation. In Bilezikian JP, Raisz LG, and Rodan GA (eds), Principles of Bone Biology. New York, Academic Press, 109–126. Tashjian AHJ, Hohmann EL, Antoniades HN and Levine L (1982) Platelet-derived growth factor stimulates bone resorption via a prostaglandin-mediated mechanism. Endocrinology, 111, 118–124. Tashjian AH, Voelkel EF, Lazzaro M et al (1987) Tumor necrosis factor-a (cachectin) stimulates bone resorption in mouse calvaria via a prostaglandin-mediated mechanism. Endocrinology, 120, 2029–2036. Tetradis S, Pilbeam CC, Liu Y et al (1997) Parathyroid hormone increases prostaglandin G/H synthase-2 transcription by a cyclic adenosine 3’,5’monophosphate-mediated pathway in murine osteoblastic MC3T3-E1 cells. Endocrinology, 138, 3594–3600. Thorsen K, Kristoffersson AO, Lerner UH and Lorentzon RP (1997) In situ microdialysis in bone tissue: stimulation of prostaglandin E2 release by weight-bearing mechanical loading. J Clin Invest, 98, 2446–2449. Tomita M, Li X, Okada Y et al (2002) Effects of selective prostaglandin EP4 receptor antagonist on osteoclast formation and bone resorption in vitro. Bone, 30, 159–163. Traianedes K, Dallas MR, Garrett IR et al (1998) 5-Lipoxygenase metabolites inhibit bone formation in vitro. Endocrinology, 139, 3178– 3184. Tsukii K, Shima N, Mochizuki S et al (1998) Osteoclast differentiation factor mediates an essential signal for bone resorption induced by 1a,25-dihydroxyvitamin D3, prostaglandin E2, or parathyroid hormone in the microenvironment of bone. Biochem Biophys Res Commun, 246, 337–341. Turner CH, Forwood MR and Otter MW (1994) Mechanotransduction in bone: do bone cells act as sensors of fluid flow? FASEB J, 8, 875–878. Wadhwa S, Godwin SL, Peterson DR et al (2002) Fluid flow induction of cyclooxygenase-2 gene expression in osteoblasts depends on the ERK signaling pathway. J Bone Min Res, 17, 266–274. Wadleigh DJ and Herschman HR (1999) Transcriptional regulation of the cyclooxygenase-2 gene by diverse ligands in murine osteoblasts. Biochem Biophys Res Commun, 264, 865–870. Wani MR, Fuller K, Kim NS et al (1999) Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: synergistic activation of differentiation, cell spreading, and fusion. Endocrinology, 140, 1927–1935. Weinreb M, Suponitzky I and Keila S. (1997) Systemic administration of an anabolic dose of PGE2 in young rats increases the osteogenic capacity of bone marrow. Bone, 20, 521–526. Weinreb M, Grosskopf A and Shir N (1999) The anabolic effect of PGE2 in rat bone marrow cultures is mediated via the EP4 receptor subtype. Am J Physiol, 276, E376–383. Wong BR, Rho J, Arron J et al (1997) TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-Jun N-terminal kinase in T cells. J Biol Chem, 272, 25190–25194. Woodiel FN, Fall PM and Raisz LG (1996) Anabolic effects of prostaglandins in cultured fetal rat calvariae: structure–activity relations and signal transduction pathway. J Bone Min Res, 11, 1249–1255. Yamaguchi A, Komori T and Suda T (2000) Regulation of osteoblast differentiation mediated by bone morphogenetic proteins, hedgehogs, and Cbfa1. Endocrine Rev, 21, 393–411. Yamamoto K, Arakawa T, Ueda N and Yamamoto S (1995) Transcriptional roles of nuclear factor kB and nuclear factorinterleukin-6 in the tumor necrosis factor a-dependent induction of cyclooxygenase-2 in MC3T3-E1 cells. J Biol Chem, 270, 31315–31320. Yasuda H, Shima N, Nakagawa N et al (1998) Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA, 95, 3597–3602. Yoshida K, Oida H, Kobayashi T et al (2002) Stimulation of bone fomation and prevention of bone loss by prostaglandin E EP4 receptor activation. Proc Natl Acad Sci USA, 99, 4580–4585. Zhao W, Byrne MH, Boyce BF and Krane SM (1999) Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J Clin Invest, 103, 517–524.
26 Ageing and Prostaglandins A. Hornych European Hospital George Pompidou, Paris, France
Ageing is associated with the appearance of several cardiovascular complications, such as atherosclerosis, ischaemic heart disease, an increase in arterial blood pressure, stroke, a decrease in renal function and others. Senescence is also characterized by the atrophy of multiple tissues, decreased proliferative capacity of cells and reduced immune responses. The homeostatic adaptation of the ageing organism and its integrity depends on the capacity of cells and organs to compensate (or not) for deficient metabolic and endocrine regulatory systems. Recent decades have shown that prostaglandins (PGs), as real local hormones (autacoids) with autocrine and paracrine functions, represent an important regulatory system implicated in cell replication, migration, adhesion and aggregation, smooth muscle cell motility, endothelial cell function, lipid metabolism, signal transmission and many other functions (Samuelsson 1976; Mamas 1997). All these mechanisms are involved in the pathogenesis of age-related diseases. Therefore it is possible that the agedependent alterations of prostaglandin synthesis may play a role in senescence. This chapter is concentrated on the influence of age on prostaglandin synthesis and metabolism in experimental and clinical studies, with implications in the appearance of clinical syndromes.
BIOSYNTHESIS OF PROSTAGLANDINS Prostaglandins of the 1,2,3 series are derivatives of essential polyunsaturated fatty acids (PUFA), such as dihomo-g-linolenic acid (DGLA, 20:3n-6), arachidonic acid (AA, 20:4n-6) and eicosapentaenoic acid (EPA, 20:5n-3) (Figure 26.1). Their synthesis depends on the availability of the corresponding substrate. These PUFA are not normally present in the food, but are generated by metabolic pathways from two essential fatty acids of o-6 and o-3 series, which are linoleic acid (18:2n-6) and alinolenic acid (18:3n-3) (Horrobin 1993; Needleman et al 1979) (Figure 26.1). The availability of PG substrates is dependent, in the first phase, on the enzymatic activity of D6-desaturase, elongases and D5-desaturase (Figure 26.1). The generated fatty acids are, in general, rapidly incorporated into cell membrane phospholipids as phosphatidylcholine and phosphatidylinositol. The synthesis of prostaglandins necessitates free acids. Therefore, the availabilities of PG substrates depend, in the second phase, upon the activity of phospholipase A2 (PLA2) and phospholipase C (PLC) being able to generate directly (PLA2), or via diacylglycerol and monoacylglycerol lipase (PLC), free arachidonate (Smith 1992) (Figure 26.2). Finally, the generation of prostaglandins depends, in the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
third phase, on the activity of cyclooxygenases and specific PG synthetases (Figure 26.2). The biological activity of the prostanoids generated depends, in the last phase, on the activity of metabolic enzymes allowing the persistence or rapid cessation of the action (Figure 26.3). This short biochemical outline shows four different active enzymatic steps which may be influenced by ageing and may play a role in age-dependent diseases. The first two steps are treated briefly below, with greater attention to the third phase.
EFFECT OF AGEING ON ENZYMATIC SYSTEM Ageing and Metabolism of Polyunsaturated Fatty Acids (PUFA) Brenner’s group (Ayala et al 1973) reported that aged animals exhibited reduced capacity to convert linoleic acid (LA) to glinolenic acid (GLA) (Figure 26.1). Horrobin (1981) proposed that the resulting deficiency in the normal flow of GLA and its metabolites might be one of the key factors involved in ageing. Indeed an impaired desaturation with ageing was described in elderly humans (Darcet et al 1980) and a decline in 6-desaturase activity for both LA and a-linolenic acid (ALA) was also described in ageing rats (Hrelia et al 1989). Therefore, deficient transformation of LA to GLA and then to DGLA may reduce the generation of PGE1 (Figure 26.1), an antiaggregant and vasodilator prostaglandin, which may favour the increase of platelet aggregation and vasoconstriction in aged subjects. Conversely, the exogenous supply of GLA and LA to aged subjects considerably increases PGE1 synthesis by platelets (Darcet et al 1980). Similarly, the exogenous supply of ethyl-dihomo-g-linolenate increases PGE1 synthesis in rabbits (Oelz et al 1976). On the other hand, deficient D6- and D5-desaturase (Darcet et al 1980) may reduce the availability of EPA necessary for the generation of PGI3—also antiaggregant and vasodilator prostaglandin (Needleman et al 1979). These enzymatic deficits are less important for the generation of PG series 2, because arachidonic acid (AA) is in general supplied by nutrients, especially meat and eggs, if consumed by aged people. The deficiency in arachidonate may also be compensated by an exogenous supply of AA by enriching the prostaglandin precursor pool, allowing the generation of PG series 2 (Seyberth et al 1975). The tissue content of DGLA, which is generally very low and even lower in aged subjects, can be highly increased by dietary supplementation (Darcet et al 1980; Hansen 1983; Hornych et al 2001). This could be achieved by oral intake of vegetable oils rich in GLA, such as evening primrose, blackcurrant or borage oil.
300
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 26.1 Enzymatic transformation of essential fatty acids and the generation of prostaglandins 1, 2, 3 series by the cyclooxygenase pathway from corresponding polyunsaturated fatty acids (PUFA)
However, evening primrose oil seems to be a particularly desirable source because it contains 7–10% of GLA and raises the production of PGE1 more effectively than the other oils (Horrobin 1990). It also increases the production of prostacyclin (PGI2) (Horrobin 1990; Guivernan et al 1994). However it should be kept in mind that: (a) excessive GLA intake may inhibit elongation of GLA to DGLA (Navarette et al 1992); (b) very high intake of o-3 essential fatty acids, e.g. as fish oil, may inhibit the metabolism of o-6 essential fatty acids and the desaturation step (Horrobin 1990). Therefore, an equilibrated dietetic intake of essential fatty acids in aged subjects is necessary. In conclusion, dietary essential fatty acids may help PG production in vivo in ageing people and overcome enzymatic defects, especially in desaturases. Phospholipases The biosynthesis of prostaglandins depends on the availability of the substrate, predominantly arachidonic acid (AA). This acid is contained in membrane phospholipids, mainly esterified at the
sn-2 position of the glycerol backbone of phosphatidylinositol (Pi), phosphatidylcholine (PCh) and phosphatidylethanolamine (Pe). The release of AA in response to different stimuli occurs through the activation of phospholipase A2 (PLA2) or phospholipase C (PLC) and eventually phospholipase D (PLD) (Smith 1992). With regard to ageing, there are several difficulties in resolving which lipases are involved in AA release, because there are multiple forms of PLA2, at least six (Dennis 1994), that are not highly specific inhibitors for any of these enzymes (Smith 1992). Similarly, there are at least nine distinct isozymes of PLC (Wahl and Carpenter 1991). Recently isolated high molecular weight cytosolic (c) PLA2 seems important, activated by low physiological concentration of calcium (Wahl and Carpenter 1991), which may provide predominantly AA for the synthesis of PGs and other eicosanoids. However, cytosolic PLA2 activity is regulated only partially by gene expression, but rather by rapid enzyme modification via phosphorylation and changes in intracellular calcium concentration (Michel 2000). The cell-specific pattern of prostanoids is further dependent on the activity of distal enzymes (see Figure 26.2). Actually there are growing data concerning the effects of ageing on different phospholipases.
AGEING AND PROSTAGLANDINS
301
Figure 26.2 Hormone-activated prostanoid biosynthesis from membrane phospholipids (PCh, phosphatidylcholine; Pi, phosphatidylinositol; Pe, phosphatidylethanolamine) via the cyclooxygenase pathway
Phospholipase A2 Kimura et al (1995) have noted that renal phospholipase A2 activity increases with age in the stroke-prone spontaneously hypertensive rat, while the membraneous phospholipids decreased with increasing age. Similarly, in rat cerebral cortex synaptic plasma membrane-bound PLA2 and PLC or diacylglycerol (DAG) lipase of the aged brain (27 months) exert significantly
higher activity in degradation of Pi as compared to their activities in adult brain (4 months). Conversely, the activity of the cytosolic enzymes involved in the degradation of Pi are not significantly changed in the senescent cerebral cortex as compared to the adult. Ageing significantly modifies the activity of membrane-bound, Ca2+-dependent phospholipase(s) degrading Pi, which may influence the formation and accumulation of potent lipid messengers such as AA and DAG (Strosznajder et al 1994). The
302
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Figure 26.3 Metabolism of prostaglandins by three enzymatic steps, prostaglandin E2 (PGE2) taken as an example
data of Yu et al (1992) are also consistent with increased phospholipase A2 activity with ageing in rats. Similarly, Petkova et al (1986) have found membrane-bound PLA2 activity enhanced with ageing, correlating well with reduced membrane fluidity. Kim et al (1997) described increased cerebral cortical phospholipase A2 activity in an age-dependent manner in the mouse. In contrast, Morita and Murota (1980) have found in rat liver that the activity incorporating AA into phospholipids increases gradually with ageing as a mirror image of decreasing PG synthesis. Williams et al (1994) observed that PLA2 activity in cerebral microvessels of aged mice (21–24 months) was significantly less than in younger animals (6 months). PLA2 activity was found to be selectively decreased in the hippocampus of aged apoE-deficient mice, but not in aged-matched control mice (Patrick et al 2000). In man, the data of Taylor et al (1989) would be consistent with increasing PLA2 activity and decreasing stability of cell membranes with age. Ito et al (1996) have found no effect of ageing on the PLA2 pathway in neutrophils from elderly human donors.
Phospholipase C PLC was studied by Narang et al (1996) in cerebral cortical membranes of young, middle-aged and old rats. They also separately analysed PLC isozymes, PLC-a, PLC-b and PLC-g, which are localized differentially in brain regions. They found increased PLC-b mRNA in the frontal cortex and superficial cortical layers of aged rats but decreased PLC-a mRNA in the
hippocampal regions of older rats. In contrast, Undie et al (1995) observed decreased PLC-b immunoreactivity and phosphoinositide metabolism in senescent F-344 rat brain but no changes in PLC-g levels in aged animals. Grossmann et al (1995) described decreased calcium mobilization and tyrosine phosphorylation of PLC-g1 in T lymphocytes from aged mice. The decrease of antioxidants with age may be responsible for this inhibitory effect. Mizutani et al (1998) analysed phospholipase C (PLC) and phospholipase D (PLD) isoforms in spleen, brain and kidney of aged rats. PLC isozymes were downregulated in different tissues during ageing, but not PLD. In man, Di Pietro et al (2000) observed altered response of the b(2) isoform of PLC in unstimulated T cells and of the b(2) and g(2) isoforms in stimulated T cells of healthy individuals over 65 years, in comparison with healthy donors below 35 years.
Phospholipase D Phospholipase D-induced generation of phosphatidic acid was studied in human neutrophils from young and old donors. The activity of this PLD was significantly higher in old than in young (Ito et al 1996). However, the recently discovered circulating enzyme, phosphoinositol-specific phospholipase D, measured in the sera of healthy individuals was strongly negatively correlated with age (Raymond et al 1994). In summary, the activities of phospholipases are differently modified by ageing but the final response, i.e. the availability of
AGEING AND PROSTAGLANDINS
303
substrate for the AA cascade, depends on different tissues, organs, species and stimulating or inhibitory factors.
analogous results (Larrue et al 1991; Hasegawa and Yamamoto 1993).
Effect of Age on Cyclooxygenases and PG Synthetases
Mechanisms of Prostaglandin Synthesis Inhibition with Ageing
Prostaglandin synthesis occurs in three phases: (a) mobilization of AA; (b) conversion of AA in the prostaglandin endoperoxides PGG2 and PGH2 by prostaglandin endoperoxide synthetase or PGH synthase, an enzyme with cyclooxygenase and hydroperoxidase activity; and (c) rearrangement or reduction of PGH2 in biologically active prostanoids by different PG synthases (Hornych 2000) (Figure 26.2). The synthesis of different PGs depends on the specific enzymatic equipment of cells and organs. PG endoperoxide synthase is an enzyme which autodestroys after several oxygenation cycles (‘‘suicide enzyme’’) by covalent binding to hydroperoxy lipids produced by cyclooxygenase activity of the enzyme (Hemler 1979). Renewal necessitates new proteosynthesis. All these steps of the AA cascade could be influenced by ageing. In general, the synthesis of eicosanoids decreases with ageing, particularly of PGI2, therefore the steps of its biosynthesis were analysed very early (Austin and Garett 2000). Kent et al (1981) were the first to show that PGI2 synthesis in the aorta of aged pigs (2–4 years) was decreased in comparison with young animals (6 months) and this result was the same also with an excess of exogenous AA in culture medium. These data imply an altered function of active enzymes metabolizing the free substrate (AA) at the level of either cyclooxygenase (COX) or PGI2 synthase. Therefore, this decrease of PGI2 synthesis is independent of the composition of fatty acids in membrane phospholipids. Chang and Tai (1983) have found similar results in the aortic tissue of adult (1 year) and aged (2 years) rats (Dekmyn et al 1983). Larrue et al (1991) further analysed the mechanism of decreased PG synthesis in vascular tissue with ageing. They measured the synthesis of PGI2 and PGE2 by rabbit aorta fragments from animals aged 1, 2 and 4 years in the presence of 14 C-AA. They found decreased PGI2 synthesis in rabbits 2 years old, but increased PGE2 synthesis. This implies a normal function of cyclooxygenase but decreased activity of PGI2 synthase. In animals aged 4 years, the activity of PGI2 synthase was the same (similarly low) as in animals aged 2 years, but the activity of cyclooxygenase was strongly depressed, as shown by an important decrease of PGE2 synthesis. According to these data, ageing in vascular tissue is associated with an early decrease of PGI2 synthase activity, with all the pathophysiological consequences, and later with the inhibition of the total cyclooxygenase pathway of AA metabolism. Chang et al (1980) observed a similar reduction of PGI2 synthesis in smooth muscle cell culture from aged rats. The primary inhibition of PGI2 synthase was also confirmed by the use of PGH2 endoperoxide in the culture as a substrate instead of AA (Chang et al 1980). Polgar’s group (Menconi et al 1987) have found similar results in cultivated cells from newborn, adult (12 month) and old rats (30 month), i.e. a relative increase of PGE2 synthesis at the expense of PGI2 in adult rats, but a decrease of PG synthesis in old rats. Moreover, senescent cells release more AA after stimulation by different agonists, but convert less of the freed fatty acids to prostaglandins (Polgar and Taylor 1984). The results in man, obtained by the culture of arterial cells from perioperative samples from the ascending aorta of patients 2–78 years old, confirm the experimental data showing a decrease of PGI2 synthase activity in aged patients without a modification of cyclooxygenase activity (Larrue et al 1991). Several other studies reproducing ageing by successive passages of different cells cultivated in vitro (human fibroblasts, pig endothelial and smooth muscle cells, human vein umbilical endothelial cells) described
Several mechanisms seem to be implicated: 1. An excess of lipid peroxidation and the generation of free radicals (Hemler et al 1979; Davidge et al 1996; Roberts and Reckelhoff 2001). 2. The reduction of the systems of protection against peroxidation: (a) enzymatic systems such as superoxide dismutase and glutathione peroxidase; (b) reduction of co-factors such as selenium; (c) deficit in natural antioxidants such as glutathione, vitamins C, E and A. 3. Pathophysiological factors such as atherosclerosis, with all metabolic alterations. Larrue et al (1982) have shown that experimental hypercholesterolaemia in rabbits reduces PG synthesis in smooth muscle cells early, compared with this occurring later in the normal processes of ageing. It looks as though experimental atherosclerosis induces the acceleration of ageing. 4. The alteration of proteosynthesis of the new enzyme is probably not implicated, because the tissue concentration of the cyclooxygenase was similar in adult and old rats (Chang and Tai 1983). However, subtle alterations at the level of the transcription of the PGH synthase gene may interfere with the replacement of PGH synthase (Smith 1992). 5. The decrease of substrate is probably not involved in reduced PG synthesis because in several conditions the biosynthesis of prostanoids may be increased with ageing, e.g. as for thromboxane A2 (TXA2) and PGF2a in human lung fibroblasts (Menconi et al 1987) or an increase of endothelium-derived contracting factor (EDCF) identified as PGH2 (Lu¨scher et al 1993). Recently, a new isoenzyme was identified, PGH synthase-2 (cyclooxygenase-2, COX-2), which is inducible (Fu et al 1990) because its synthesis is induced by inflammation, cytokines, LPS and mitogenic stimuli and is inhibited by glucocorticoids. This COX-2 contrasts with the previously described cyclooxygenase, which is constitutive, present in numerous tissues and denominated as cyclooxygenase-1 (COX-1, PGH synthase-1). The latter is not inhibited by glucocorticoids. COX-2 is present in several human tissues, as described by O’Neil and Ford-Hutchinson (1993), but can coexist within a single cell (Austin and Garett 2000). The two cyclooxygenases are differently regulated during their development in different organs (Peri et al 1995; Brannon et al 1994, 1998) or during maturation (Yamagata et al 1993), as well as during ageing: Heymes et al (2000) observed that aortic endothelial cells from aged, but not young, rats express the COX2 isoform, while COX-1 labelling was observed in endothelial cells from both young and aged rats. Matz et al (2000) studied the relative contribution of endothelial nitric oxide (NO) and cyclooxygenase metabolites in relaxation to acetylcholine with ageing in the aorta and the small mesenteric artery (SMA) of the rat. The authors observed endothelium-dependent relaxation to acetylcholine reduced in 70–100 week old rats. In these rats, exposure to the COX inhibitor indomethacin, but not to the selective COX-2 inhibitor NS-398, potentiated response to acetylcholine. The TXA2/PGH2 receptor anatgonist, GR 32191B, enhanced relaxation to acetylcholine in the aorta but not in the small mesenteric artery (SMA). Morover, acetylcholine increased thromboxane B2 (TXB2) production in aorta but not in the SMA. Finally, they found enhanced expression of COX-1 and COX-2 in the two arteries with ageing. They concluded that the
304
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
decrease in acetylcholine-induced relaxation with ageing involves reduced NO-mediated dilatation and increased generation of vasoconstrictor prostanoids, most likely from COX-1. They also point out the vascular bed heterogeneity related to the nature of prostanoids involved. The results of Chung et al (1999) found that another tissue, i.e. rat kidney, showed upregulation of COX-2 during ageing, which may play a role in age-related diseases. In summary, ageing may affect the enzymatic system of prostaglandin biosynthesis in three phases: (a) maintained activity of cyclooxygenase but the inhibition of PGI2 synthase activity without affecting the synthesis of other prostanoids; on the contrary, on some occasions the synthesis of TXA2 or PGE2 and PGF2a may be enhanced, especially when COX-2 is upregulated; (b) decreased activity of cycloxygenase with reduced activity of PGI2 synthase and with or without concomitant increase or decrease of other PGs; (c) decreased synthesis of all prostanoids. METABOLISM OF PROSTAGLANDINS Metabolic degradation of prostaglandins is ensured by different enzymatic steps, shown in Figure 26.3, but several transformations may be non-enzymatic, e.g. the metabolism of PGI2 in 6keto-PGF1a or TXA2 in TXB2. In general, the effect of ageing on the enzymes of PG metabolism is less pronounced. However, Chang and Tai (1984) described in rat kidneys (from animals aged 2, 12 and 24 months) a progessive considerable decrease of NAD+dependent 15-hydroxyprostaglandin dehydrogenase as a function of age, which may be responsible for decreased metabolism of TXA2 in aged kidneys, because its synthesis is not affected by ageing. On the other hand, Ueno et al. (1985) have found the activities of the enzymes responsible for PG metabolism (NADP-dependent 15-hydroxy-PGD2 dehydrogenase) unchanged postmaturationally in rat brain. Markov et al (1983) observed more intensive catabolism of PGE2 in the lungs and kidneys of young rats than in older rats, whereas PGF1a degradation did not differ in various age groups but decreased somewhat with age in the lungs. PaceAsciak and Edwards (1980) and Pace-Asciak (1976) have shown that there are age-dependent types of PG catabolism and different extents of PG catabolism, since the profile of PG metabolites changed as a function of age. In general, metabolic enzymes seem to be active in senescence to metabolize efficiently released prostaglandins, but their activity may be insufficient in the presence of elevated biosynthesis of PGH2, TXA2 or PGF2a, prothrombotic and vasoconstrictive PGs whose synthesis is often increased by ageing (Hornych et al 2000; Menconi et al 1987; Lu¨scher et al 1993; Todd et al 1994). AGEING AND PROSTAGLANDINS, SYSTEMS AND DISEASES Ageing does not have an identical effect on prostaglandin biosynthesis and metabolism in different tissues and organs. For this reason known observations are treated according to different systems.The experimental data in vitro and in vivo are treated first and the human or clinical data thereafter. Cardiovascular system Ageing is associated, in general, with an important deterioration of the cardiovascular system and the appearance of major pathological complications, such as atherosclerosis, thrombosis, hypertension, stroke, myocardial infarction, peripheral occlusive disease and others. The discovery of vasoconstrictive and
aggregatory TXA2 (Hamberg et al 1975) and vasodilator and antiaggregatory prostacyclin (Bunting et al 1976) revealed their important role in cardiovascular homeostasis (Moncada and Vane 1979). Since their discoveries, the key question has been whether the cardiovascular complications of ageing are due to the deficit of PGI2 or to an excess of TXA2 or both. Several research studies were orientated to this fundamental question. Since the greatest quantitites of PGI2 are synthesized by vascular endothelial and smooth muscle cells (SMC), the first publications dealt with these tissues. Chang et al (1980) measured PG biosynthetic activity in cultured rat SMC in young and old rats. Cells from young rats produced more PGI2 than PGE2, whereas cells from old rats produced more PGE2 than PGI2. Therefore, there is an agerelated decrease of PGI2 synthesis in aortic SMC which may, according to these authors, contribute to the pathogenesis of atherosclerosis. Similarly, Kent et al (1981) found an age-related decrease of PGI2 synthesis in intimal strips from swine aortas, independent of fatty acid composition in the aortic tissue (which was virtually identical in young and mature animals). The intimal tissue produced essentially PGI2, and only insignificant amounts of PGE2 and PGF2a. Chang and Tai (1983) investigated the PGI2 formation in rat aorta strips and TXB2 biosynthesis in platelets. Their results indicate an age-related decrease of PGI2 production by aortas but no significant change in TXA2 biosynthesis in platelets from mature (12 month) to senescent rats (24 months). On the other hand, there are some experimental data showing the dissociated biosynthesis of PGs in different tissues. Lennon and Poyser (1986) found significantly lower amounts of 6-ketoPGF1a synthesized by vascular endothelial cell suspension in old compared to young male rats; this difference was not found in female rats. However, the amounts of 6-keto-PGF1a synthesized by homogenates of aortic smooth muscle cells were considerably higher in old compared to young male and female rats. In the preparation of in vitro perfused aortas or mesenteric arteries, PGI2 output did not decrease with age. Also the increases in output of 6-keto-PGF1a from the mesenteric arterial bed in response to noradrenaline and angiotensin II were not diminished with age in either male or female rats. Conversely, the vascular output of PGF2a was higher in old rats. Since the arterial blood pressure (BP) of old rats was significantly higher compared to the young animals, the increase of vasoconstrictive PGF2a may be related to the increase of BP of old rats. In accordance with the previous work, the coronary vascular response to endothelin-1 (ET-1) in the rat perfused heart, investigated by Katano et al (1993), was associated with a marked increase of 6-keto-PGF1a, which was identical in all three age groups. The basal and ET-1-stimulated PGI2 release (measured as 6-keto-PGF1a) was measured also in another circulation: Katoh et al (1991) described a spontaneous release of PGI2 from isolated hind legs of WKY rats (aged 5, 10 and 40 weeks) falling with advancing age, but it remained unchanged in spontaneously hypertensive rats (SHR). ET-1 increased PGI2 release in a dosedependent fashion in both strains, regardless of age. These three experimental data together suggest that the vascular PGI2 response to vasopressor agonists is not significantly modified in naturally ageing animals. The effect of ageing on PG synthesis was also investigated on cells cultivated in vitro, in which ageing was induced by numerous passages. Takeuchi et al (1987) measured mesenteric artery SMC production of PGI2, PGE2, PGF2a and TXB2. PGI2 was the major product among these PGs. PG synthesis in these cells decreased with in vitro ageing, but the distribution pattern of PG synthesis did not change up to passage level 56. The authors suggest that the imbalance among PGs may not be directly implicated in vascular diseases in ageing.
AGEING AND PROSTAGLANDINS The experimental data of Takahashi et al (1990) are very interesting because they evoke clinical pathology. These authors observed an age-associated reduction of vascular PGI2 production and thrombin-stimulated TXA2 production in the blood of rats aged 64 weeks in comparison with rats aged 11 weeks. The 64 week-old rats had significantly higher plasma cholesterol and plasma triglycerides in comparison with young animals. Moreover there was a negative significant correlation between vascular 6-keto-PGF1a and plasma cholesterol levels, and thrombin-stimulated TXB2 and plasma cholesterol. These results suggest that age-associated changes of blood cholesterol are closely linked with vasoactive eicosanoid synthesis. They suggest also that age-associated hypercholesterolaemia may promote the atherosclerotic process, not only by the accumulation of cholesterol in vascular tissue but also due to reduction of vasoprotective prostacyclin production. The age-dependent reduction of eicosanoid synthesis is not due to changes of their precursor essential fatty acids, because no difference in aorta and platelet phospholipids was found with age (Takahashi and Horrobin 1988). The vascular function may be altered by an age-related increase of lipid peroxidation, as shown by Davidge et al (1996) in female rats. This alteration was due in part to a cyclooxygenasedependent vasoconstrictor (probably PGH2). The same endothelium-derived contracting factor, most likely PGH2, is released in SHR by acetylcholine and stretch (Lu¨scher et al 1992). According to Lu¨scher et al (1992), the functional alteration of endothelium does occur with ageing, hypercholesterolaemia and hypertension. Lipid peroxidation seems to play an important role in agedependent alterations of the cardiovascular system. The cyclooxygenase-catalysed AA cascade is an important source of reactive oxygen species. Kim et al (2001) studied the implication of COX-1 (constitutive) and COX-2 (inducible) in rat heart during ageing. COX-2 mRNA and protein levels increased with age, whereas those of COX-1 showed no change. The COX activity measured by PGE2 production increased with age. The authors also measured changes in the generation of reactive oxygen species and lipid peroxidation. They found an age-dependent increase of both. They concluded that COX-2 activity increases with age and can play a role in oxidative alteration in the aged heart. Similarly, Hayek et al (1997) observed that macrophages from old mice produce more PGE2 than those from young mice. Analysis of proteins showed no age-related difference in COX-1 protein levels, but macrophages from old mice had higher LPS-stimulated levels of COX-2 mRNA compared with those from young mice. They concluded that an age-associated increase in LPS-stimulated PGE2 production is due to increased COX activity resulting from higher COX-2 protein and mRNA expression. The influence of ageing on the human cardiovascular system was investigated in different ways: (a) in cells cultivated in vitro or in homogenates of tissues; (b) measuring plasma or urinary prostaglandin levels or their metabolites in different age-groups of subjects. Cell and Tissue Prostaglandins Murota et al (1979) observed a reduction in the synthesis of PGI2 in the culture of ageing human lung fibroblasts. Polgar and Taylor (1984) have found the same result in human lung and skin fibroblasts. Larrue et al (1991) have shown decreased capacity of prostacyclin biosynthesis in aortic arterial cells from elderly subjects. Tokunaga et al (1991) measured the biosynthesis of prostacyclin from in vitro-cultivated human aortic endothelial cells obtained from infants through to aged subjects. They observed progressively decreasing PGI2 synthesis in the older groups. According to these authors, the decreased synthesis of
305
prostacyclin with age may play an important role in the development and advancement of thrombosis and atherosclerosis. The only different results were reported by Lennon et al (1988), who measured PG production in homogenates of human aorta and saphenous vein from patients undergoing coronary bypass surgery (age range 31–80 years). The 6-keto-PGF1a, PGF2a and PGE2 production by homogenates of aorta was unaffected by age, sex or smoking habits. 6-keto-PGF1a production by homogenates of male saphenous vein was 20% lower in smokers and exsmokers than in non-smokers. In spite of these changes, the basal outputs of prostaglandins, particularly of 6-keto-PGF1a, from the saphenous vein were little affected by age, sex or smoking habits. However, these results were influenced by the concomitant treatments because most of the patients (80 out of 127) were treated by drugs either stimulating (b-adrenoceptor blocking drugs, nitrate vasodilator drugs, thiazide and loop diuretics) or inhibiting (calcium anatgonists) 6-keto-PGF1a production in homogenates used by the authors and tested by themselves. HUVEC The effect of ageing was also investigated in human umbilical vein endothelial cells (HUVEC), cultivated in vitro with multiple passages from the primoculture. Sato et al (1993) compared the PGI2, TXA2 and ET-1 secretion by HUVEC after 7 and 67 phases of culture. ET-1 secretion increased three-fold, PGI2 secretion sixfold and TXA2 secretion 18-fold during the same interval. The ratio PGI2:TXA2 decreased three-fold. These data indicate that the antithrombotic and antivasoconstrictive role of endothelial cells may decrease during in vitro ageing. Similarly, Hasegawa and Yamamoto (1993) have shown an age-related decrease of PGI2 production by in vitro ageing HUVEC, which was due to decreased conversion of AA to PGI2 in the cytoplasm. Plasma Plasma and urinary prostaglandins were measured very early in man, but most attention was attached to the levels of prostacyclin and TXA2 and their relationship to the cardiovascular diseases. The initial techniques for their measurement by bioassay evolved to more sensitive and specific methods, such as radioimmunoassay with previous extraction, HPLC and gas chromatography–mass spectrometry. Plasma PGI2. Neri Serneri et al (1981) were the first to measure plasma PGI2 levels in man using Vane’s bioassay technique. They described a progressive age-dependent decrease of circulating plasma PGI2 levels, both in men and women, from 20–30 year to 60–70 year age groups. Ylikorkala et al (1982), using the radioimmunoassay technique, measured plasma 6-keto-PGF1a in 140 subjects aged 10–90 years. They found dissimilar results because the highest plasma 6-keto-PGF1a levels were found in the 10–20 year age group in both sexes, lower in other age groups, but the 6-keto-PGF1a levels in subjects aged 71–90 years did not differ significantly from those in subjects under 20 years of age. The levels of prostacyclin metabolites in women over 70 years were higher than those in men of the same age. The authors concluded that PGI2 generation in vivo, as measured by 6-ketoPGF1a levels in peripheral plasma, is higher in adolescence and in elderly females than in healthy adults. Conversely, Gotoh et al (1983) described decreased plasma 6-keto-PGF1a in elderly normotensive men. Plasma TXB2. Hornych et al (1982a) measured plasma TXB2 concentration in young healthy volunteers (28+2 years) and in
306
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
older subjects (45+5 years). The peripheral venous plasma TXB2 was 35+8 pg/ml in young subjects and 101+14 pg/ml in an older group. The significant increase of plasma TXB2 with age may have a pathophysiological significance. Similarly, Todd et al (1994) measured plasma TXB2 levels in young and older men to determine the combined effect of age and acute exercise. Resting TXB2 was 53+9 pg/ml in young men and 79+18 pg/ml in older men. Also post-exercise plasma TXB2 levels were significantly higher in older than in younger men. These data suggest that the older men have a greater increase in TXB2 after exercise and may be more predisposed to platelet activation. Urinary Prostaglandins These were measured in different age groups. Urinary PGE2 and PGF2a reflect renal synthesis of these eicosanoids (Fro¨lich et al 1975), while the urinary 6-keto-PGF1a and TXB2 reflect the predominant renal synthesis of PGI2 and TXA2, respectively, but also a part of extrarenal synthesis if pathologically increased (Catella et al 1986). The measurement of urinary 2,3-dinormetabolites of 6-keto-PGF1a and TXB2 reflects more the wholebody (extrarenal) synthesis of PGI2 and TXA2 (FitzGerald et al 1983). The absolute amounts of PGI2 and TXA2 synthesis are important for cardiovascular pathology of the elderly but also the ratio of PGI2:TXA2 (or their respective metabolites), because they have quite opposite biological activities. This ratio is also analysed in some reports. PGE2. Mackenzie et al (1984) observed decreased urinary PGE2 excretion in older essential hypertensive males but not in older normotensives. Hou et al (1992) also found no age-related changes in urinary PGE2 excretion in normal subjects. Similarly, in our experience, urinary PGE2 excretion was not affected by ageing in either men or women until late age (71–95 years) and may even increase (Hornych et al 1991). The preserved renal PGE2 synthesis in the elderly may be explained by the disparity between the involution of the cortex (main site of PGI2 and TXA2 synthesis in man) and relative sparing of the medulla with increasing age, which is the main site of PGE2 synthesis (Sraer et al 1982). PGF2a. Lijnen et al (1983) measured urinary PGF2a excretion in a population of children and adults in the age range 3–41 years. They observed a biphasic urinary PGF2a excretion, increasing until the age of 20–30 years and then moderately decreasing. In our experience urinary PGF2a excretion moderately declined with increasing age from 20 until 95 years, and this decrease was significant in the 41–70 year age group (Hornych et al 1991). are prostaglandin F2-like 8-epi-PGF2a. F2-isoprostanes compounds that are formed in vivo directly by free radicalcatalysed lipid peroxidation (Morrow et al 1990). One of the compounds, 8-epi-PGF2a , is a potent vasoconstrictor and its excretion is increased as a function of age (Wang et al 1995). These data support the free radical theory of ageing (Roberts and Reckeloff 2001). PGI2 and TXA2. Two metabolites were measured in different age groups of normal subjects, 6-keto-PGF1a and 2,3-dinor-6-ketoPGF1a for PGI2, and TXB2 and 2,3-dinor-TXB2 for TXA2. In our study of 45 normotensive subjects of both sexes, divided into three age groups, 20–40, 41–70 and 71–95 years, we observed a significant (p50.01–0.001) decrease of urinary 6-keto-PGF1a excretion from 220+21 pg/min in the first group to 106+26 pg/ min in the 2nd group and to 61+8 pg/min in the oldest group. The decrease was also significant when expressed in pg/mg
creatinine. The decrease of urinary 6-keto-PGF1a was significantly negatively correlated with age (p50.001) (Hornych et al 1991). In the same study, urinary TXB2 increased significantly (p50.05– 0.001) with age, from 152+22 pg/mg creatinine to 220+31 pg/mg creatinine in the 2nd group and to 379+63 pg/mg creatinine in the oldest group, therefore the ratio 6-keto-PGF1a:TXB2 decreased significantly (p50.001). Arterial blood pressure (BP) increased progressively with age from 113+2/75+3 mmHg to 116+3/ 76+3 mmHg in the 2nd group and to 136+3/79+1 mmHg in the oldest group; the increase of systolic BP was significant (p50.001). Systolic BP was positively correlated with age (p50.001) and negatively with the ratio of urinary 6-ketoPGF1a:TXB2 (p50.01). These data may suggest that progressively decreasing renal PGI2 and increasing TXA2 synthesis with ageing may contribute to the increase of arterial blood pressure in the elderly. Similar results for urinary TXB2 excretion were found by Chiba et al (1984): when urinary excretion of TXB2 was expressed as a function of urinary creatinine excretion, it was significantly higher in the elderly (60–93 years) than in the younger subjects (19–59 years). Wennmalm et al (1990) measured urinary 2,3-dinor-6-ketoPGF1a (PGI2-metabolite, PGI-M) and 2,3-dinor-TXB2 (TXA2metabolite, TX-M) in 385 non-smoking men born in 1968–1969, i.e. 22–23 years old, and in 31 men born in 1913 or 1923, i.e. 77 or 67 years old, in relation to platelet activity and inheritance and environmental factors. They found significantly (p50.001) higher excretion of TX-M in old men than in younger (563 vs. 128 pg/mg creatinine), but also higher excretion of PGI-M (163 vs. 130 pg/mg creatinine, p50.01)#; both metabolites were correlated to the urinary output of noradrenaline and adrenaline. As most of the TX-metabolite is provided by the platelets, they concluded that advancing age and sympathoadrenal tone are positively correlated to platelet activity in randomly sampled men. Increased levels of PGI-M in older men are probably the consequence of platelet activation, because enhanced prostacyclin biosynthesis was described in patients with severe atherosclerosis and platelet activation in vivo (FitzGerald et al 1984).
Platelets Platelets play a central role in atherosclerosis (Cuevas 2000). Several studies have provided evidence of the in vivo activation of platelets in the elderly: enhanced platelet release (Zahavi et al 1980; Sie et al 1981), reduced platelet survival time (Abrahamsen 1968), shortened bleeding time (Jorgensen et al 1980) and increased platelet sensitivity to various agonists (Yokoyama et al 1985). Since platelets are the most important source of TXA2 in the body (e.g. 200–300 ng/ml sera), they appear particularly implicated in cardiovascular ischaemic–thrombotic events, as in unstable angina (FitzGerald et al 1986). For all these reasons the influence of ageing on platelet function and eicosanoid biosynthesis was also studied in man. Reilly and FitzGerald (1986) investigated a population of 16 healthy volunteers (21–39 years) and 20 old healthy volunteers (50–88 years) with the question of whether platelet function was altered in vivo with increasing age. They measured: (a) different indices of platelet and vascular function; (b) 2,3-dinor-TXB2, the major urinary metabolite of TXA2, reflecting endogenous formation of TXA2 as a non-invasive index of platelet activation in vivo; (c) serum TXB2, reflecting TXB2 formation ex vivo; (d) 2,3-dinor6-keto-PGF1a as an index of prostacyclin synthesis in vivo. They #
The ratio of PGI2-M:TXA2-M is 1.01 in younger men and decreases to 0.28 in older men.
AGEING AND PROSTAGLANDINS found significantly reduced bleeding time in older subjects, but no other significant differences in platelet function; also, serum TXB2 levels were similar. Conversely, they observed significantly higher (p50.005) urinary 2,3-dinor-TXB2 excretion in older subjects in comparison with younger ones and this TX-M excretion was significantly (p50.01) correlated with increasing age. At the same time, they observed a significantly higher (p50.005) urinary 2,3dinor-6-keto-PGF1a excretion in older than in younger subjects, and this excretion was also correlated (p50.05) with increasing age. The authors concluded that increased levels of TX-M reflect the presence of platelet activation in vivo increasing with age in apparently healthy individuals. The simultaneous increase of PGIM is probably due to increased frequency and/or intensity of platelet–vessel wall interactions occurring in the presence of in vivo platelet activation and primary loss of vascular integrity (FitzGerald et al 1984). As the platelet function in older subjects was not significantly modified, they also concluded that the urinary PG metabolites are more sensitive indices of platelet and vascular function in vivo. Another study was carried out by Todd et al (1994), who found significantly higher (p50.05) plasma TXB2 levels after 30 min exercise in older men (50–65 years) than in younger males (25–35 years). They concluded that older subjects may be more predisposed to platelet activation. In summary, most of the experimental and clinical data concur that ageing is associated with decreasing synthesis of vasodilator, antiaggregating and cytoprotective prostacyclin and increased synthesis of vasoconstrictor and proaggregating TXA2. Therefore, the biologically important PGI2/TXA2 balance seems to be progressively shifted to the predominance of vasoconstrictor, proaggregating and potentially prothrombotic factors, which may contribute to the development of atherosclerosis, hypertension and major ischaemic events during senescence. However, the activation of platelets with TXA2 release may also stimulate PGI2 synthesis and release in old subjects. On the other hand, the synthesis of PGE2 is less affected by ageing. Pulmonary System The influence of ageing on the pulmonary system was analysed almost exclusively in experimental studies. One special model was used in two studies: lung fibroblasts, isolated from the human foetus, which are known to have a limited life-span in vitro. Taylor et al (1981) showed that the production of prostacyclin by human embryo lung fibroblasts upon senescence in culture drops dramatically in response to arachidonic acid or ascorbic acid. Murota et al (1983), using a human diploid fibroblast strain, showed that prostacyclin synthesis was considerably higher in young cells (25 passages) than in old cells (54 passages). In contrast, TXA2 synthesis, after stimulation of these cells, was much higher in old cells than in young ones. Quantitatively, PGI2 predominated in young cells and TXA2 synthesis in old cells. The authors concluded that there is an age-related shift from the biosynthesis of PGI2 to TXA2. These changes should operate at the level of substrate transformation, because there are no significant differences between young and old cells in fatty acid composition, in cellular phospholipids or in arachidonic acid incorporation. Ibe and Rai (1992) studied the metabolism of endogenous AA using isolated perfused lung of ferrets when stimulated with Ca ionophore A 23187 at two ages, neonatal (2–3 weeks old) and adult (over 6 months). Compared with adult lungs, neonatal lungs produced more 6-keto-PGF1a but less PGE2. There was no difference in TXB2 production by neonatal and adult lungs. Ibe et al (1996) studied prostacyclin and TXA2 production using isolated segments of intrapulmonary arteries and veins of eight near-term foetal lambs and eight ewes. During
307
normotension, foetal arteries synthesized more PGI2 than adult arteries of the same size (43 mm diameter). Small foetal arteries (51 mm) synthesized less PGI2 than adult arteries (51 mm). PGI2 production by veins (43 mm) was similar in the foetus and adult but in veins (51 mm) was greater in adult than in foetal vessels. Production of TXA2 by segments of foetal and adult vessels was less than 50% of that of prostacyclin. These data show that there is heterogeneity in the production of prostacyclin and thromboxane along the ovine pulmonary vascular tree. Clinical data are lacking. Only Edlund et al (1981) studied the pulmonary formation of PGI2, as reflected by the difference in concentration of pulmonary and systemic arterial radioimmunoassayed 6-keto-PGF1a in healthy subjects, aged 25–44 years. In steady-state conditions there was no evidence of pulmonary formation and release of 6-keto-PGF1a. During artificial ventilation in two older men, 43 and 54 years, and two older women, 57 and 71 years, the arterial levels of 6-keto-PGF1a increased suggesting that pulmonary formation of PGI2 was stimulated. This was also the case after pulmonary reperfusion. The available data concerning the pulmonary system are concordant with the data from the cardiovascular system, that ageing may reduce the pulmonary synthesis of PGI2 and enhance TXA2 synthesis. However, the clinical data of Edlund et al (1981) show that pulmonary synthesis of PGI2 in older subjects may still be efficient in stimulated conditions. Renal System The renal system is closely related to the cardiovascular system because it regulates blood volume and arterial blood pressure through Na, Cl and water homeostasis, and also through the renal vasopressor and vasodepressor systems (Hornych 1991). Since renal function declines with age due to the predominant atrophy of the cortex with the reduction of filtering glomeruli (with less effect on medulla), ageing differently affects renal prostaglandin synthesis. In man the cortex generates predominantly PGI2 and TXA2, while the medulla synthesizes 5–10 times more PGs, especially PGE2 and PGF2a, less of PGI2 (Sraer et al 1982). However, the cortex is the predominant site of PG inactivation. Chang and Tai (1984) investigated the effect of ageing on the prostacyclin and thromboxane biosynthesis and prostaglandin catabolic enzyme activity in rat kidney. They observed an agedependent progressive decrease of prostacyclin biosynthesis in kidneys from rats 2, 12 and 24 months old, while thromboxane biosynthetic activity showed no significant change. Since the activity of cyclooxygenase was not changed by ageing, the progressive decrease of renal PGI2 biosynthesis was due to the defect in prostacyclin synthase in the aged kidney. The renal catabolic enzyme activity, NAD+-dependent 15-hydroxy-PG-dehydrogenase, progressively decreased as a function of age, which leads to a decrease in the metabolism of TXA2 in aged kidneys. Therefore, the decrease of both activities, in renal PGI2 synthesis and in the catabolic activity by 15-hydroxy-PG dehydrogenase, might contribute to a progressive decrease in renal function in the elderly. Other mechanisms may also reduce renal function, e.g. for oxidative stress: Reckelhoff et al (1998) were interested whether age-related alterations of renal haemodynamics and morphology were associated with oxidative stress and whether this could be attenuated by chronic administration of vitamin E. Rats aged 13 months were given either a control diet (50 IU/kg vitamin E) or a high-vitamin E diet (5000 IU/kg) for 9 months. Rats aged 3–4 months served as young controls. Ageing was accompanied by a 60% reduction in glomerular filtration rate (GFR) and a threefold increase in renal F2 isoprostanes—vasoconstrictive PGs generated by free radical-mediated lipid peroxidation. Renal ageing was also associated with an increase in oxidant-sensitive
308
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
haem oxygenase and advanced glycosylation end products and their receptors, whose interaction has been shown to induce oxidative stress. With the high vitamin E diet GFR increased by 50%, F2 isoprostanes were suppressed and glomerular sclerosis was attenuated. The authors concluded that age-related decline in renal function is accompanied by oxidative stress and that administration of antioxidants, such as vitamin E, could attenuate the decline in renal function. Sato et al (1989) investigated papillary PG synthesis in cultured renal papillary collecting tubule cells from young (4 weeks) and aged (16 weeks) Wistar–Kyoto rats. In cells from the aged rats, basal levels of PGE2 and cyclic AMP, as corrected for cellular protein, were significantly lower than those from young rats. Another group, Rathaus et al (1993), studied the effect of sodium loading on renal prostaglandins in old rats. Sodium loading did not significantly change PG synthesis in the kidneys from young rats. In the aged rats, glomerular and cortical TXB2 decreased, whereas medullary and papillary 6-keto-PGF1a increased. Sodium loading transiently increased urinary PGE2 excretion in young rats, but not in old rats. Another approach was chosen by Millatt and Siragy (2000), who measured in renal interstitial fluid levels of cGMP, cAMP, PGE2 and PGF2a from young (4 weeks) and adult (6 months) rats submitted for 5 days to low (0.04% NaCl), normal (0.28% NaCl) or high (4.0%) diet. A low-salt diet significantly increased and a high-salt diet significantly decreased plasma renin activity, with no significant differences in the two age groups.The low-salt diet significantly increased all four mediators measured in interstitial fluid with no differences between the responses of the two age groups. The high-salt diet increased also all four mediators but cAMP, cGMP and PGE2 production was significantly greater and PGF2a production was significantly lower in young rats compared with adult rats. These data show that there are age-related changes in renal response to sodium loading. In man, there are several studies of PG biosynthesis from newborn to adults. Ignatowska-Switalska and Januszewicz (1980) showed progressive increase of urinary PGE2 and PGF2a excretion. Leonhardt et al (1992) measured urinary excretion of PGE2, F2a, 6-keto-PGF1a and TXB2 (renal origin) and also PG metabolites reflecting total body synthesis of PGE2, PGI2 and TXA2 in 83 healthy subjects aged 1 day to 37 years. The excretion rates of all PGs increased with advancing age. After correction for 1.73 m2 body surface area, only urinary excretion of PGE-M metabolite and 6-keto-PGF1a depends on age. Since the reduction of renal function in man takes place within advanced age, some research groups studied the functional relations between the kidney function and renal prostaglandins. Chiba et al (1984) show increased urinary TXB2 excretion in the elderly (60–93 years). As TXA2 is a strong renal vasoconstrictor, it may be implicated in functional impairment of the aged kidney. Wilson et al (1989) demonstrated greater natriuresis after furosemide in older persons (50 years and older) associated with significantly (p50.05) higher excretion of TXB2 in comparison with younger subjects (18–30 years), while 6-keto-PGF1a excretion was not different. The renal functional response to furosemide in old subjects is more active in TXA2 synthesis. In our work (Hornych et al 1991), we have found a significant decrease of renal PGI2 synthesis (p50.001) and a significant increase of renal TXA2 in the old population (71–95 years) with a significant decrease of urinary 6-keto-PGF1a:TXB2 ratio. At the same time, we observed in this age group a decrease of GFR to a mean value of 37+6 ml/min/1.73 m2. There was a highly significant negative correlation between age and GFR (p50.001); urinary 6-keto-PGF1a was positively (p50.001) and urinary TXB2 negatively (p50.01) correlated with GFR. These data suggest that an age-dependent decrease of renal PGI2 synthesis with a concomitant increase in TXA2 synthesis may contribute to the
decrease of GFR in the elderly. These results also emphasize the importance of renal prostaglandins for the maintenance of renal function in senescence. This conclusion is reinforced by several reports showing that the treatment of old subjects with potent non-steroidal antiinflammatory drugs (NSAIDs), inhibitors of cyclooxygenase, is associated with increased risk of renal failure (Hornych 1984; Carmichael and Shankel 1985). The discovery of two different cyclooxygenases, one constitutive (COX-1) and the other inducible (COX-2), introduced the possibility for the development of selective inhibitors and safer NSAIDs (DeWitt and Smith 1993; Mitchell et al 1994). Indeed, COX-1 plays a greater role in the maintenance of essential physiological functions, such as gastric protection, vascular homeostasis and renal function, while COX-2 is implicated more in inflammation, tumoural angiogenesis, atherosclerosis, some cerebral alterations and other diseases. In the foetal kidney, COX-2 is essential to the differentiation and maturation of nephrons. In the adult kidney, COX-2 serves vascular resistance in the cortex and salt and water homeostasis in the medulla. Thus, more selective inhibitors of COX-2 may spare renal function in treated subjects. However, it was shown that targeted disruption of the COX-2 allele in mice has resulted in severe renal problems, suggesting that COX-2 inhibitors may also produce adverse effects (Swan et al 2000). Several new drugs have been developed for clinical use, such as nimesulide, meloxicam, celecoxib and rofecoxib, 100–400 times more selective for COX2 than for COX-1, and their renal actions were studied. Nimesulide induced an acute but transient decrease in indices of renal haemodynamics and reduced urinary excretion of PGE2 (Steinhauslin et al 1993). In another clinical study Brunel et al (1995) observed in two of eight healthy volunteers a strong and immediate reduction in the excretion of prostaglandins, but overall the two doses (5 and 25 mg b.i.d.) did not produce a statistically significant effect on glomerular filtration rate. The renal effect of meloxicam was studied in 25 patients (mean age 70 years) with rheumatic disease and mild renal impairment (creatinine clearance 25–60 ml/min). These patients were treated for 28 days with 15 mg meloxicam/day (Bevis et al 1996).The creatinine clearance, the excretion of N-acetyl-b-glucosaminidase (a marker of renal tubular damage) and urea and potassium were not significantly modified by the treatment. Rofecoxib in a daily dose of 12.5 or 25 mg was compared with indomethacin 50 mg three times daily and with placebo-treated subjects in a randomized, parallel-group, multiple-dose study of 60 patients, aged 60–80 years (Swan et al 2000); also in 15 patients a single dose of 250 mg rofecoxib and 75 mg indomethacin was studied. Compared with placebo, a single dose of rofecoxib decreased GFR significantly by 13.8 ml/min, and indomethacin by 10.8 ml/min. The multiple doses of rofecoxib 12.5 mg/day decreased GFR by 8.4 ml/min (p=0.019), 25 mg/day decreased GRF by 7.8 ml/min (p=0.029) and indomethacin 3650 mg daily decreased GFR by 6.0 ml/min (p=0.086). The authors concluded that the effects of COX-2 inhibition on renal function are similar to those observed with non-selective NSAIDs. The observations of Woywodt et al (2001), using rofecoxib in two patients with impaired renal function, confirm that COX-2 inhibitors, as a class, can be as nephrotoxic as their non-selective predecessors. The effect of celecoxib and naproxen was investigated in 29 healthy elderly subjects in a single-blind randomized, cross-over study by Whelton et al (2000). Subjects received either celecoxib 200 mg twice daily for 5 days and then 400 mg twice daily for the next 5 days, or they received naproxen 500 mg twice daily for 10 days. The GFR decreased significantly more with naproxen (77.5 ml/min/1.73 m2) than with celecoxib (71.1 ml/min/1.73 m2) on the sixth day of the treatment (p=0.004). Urinary PGE2 and 6-keto-PGF1a were significantly reduced with both celecoxib and naproxen (p40.004) without a significant difference between these
AGEING AND PROSTAGLANDINS two drugs. The decrease of sodium excretion was only transient (p50.05). This study suggests that celecoxib may spare renal haemodynamic action in healthy elderly subjects, although the effects on sodium excretion, as well as urinary PGE2 and 6-ketoPGF1a excretion, appear to be similar to those of non-specific cyclooxygenase inhibitors such as naproxen. However, the comparison of renal adverse reactions between rofecoxib and celecoxib shows significantly less adverse renal effects with celecoxib treatment in man according to the WHO/ Uppsala Monitoring Centre Safety database (Zhao et al 2001). In an additional analysis celecoxib was shown to have a similar renal safety profile to that of diclofenac and ibuprofen. According to this report rofecoxib has significantly greater renal toxicity than celecoxib or traditional NSAIDs. In summary, renal prostaglandins are important for the regulation of water and electrolyte homeostasis, arterial blood pressure and renal function, especially the glomerular filtration rate. According to experimental and clinical results, ageing is associated, in general, with reduced cortical PGI2 and unchanged or increased TXA2 synthesis, with concomitant decrease of GFR and an increase of arterial blood pressure. The medullary functions, as well as medullary synthesis of prostaglandins, seem less influenced by ageing and also (possibly for this reason) water and electrolyte homeostasis is relatively well preserved in senescence. However, the use of NSAIDs in old subjects may considerably decrease renal prostaglandin synthesis, with the danger of acute renal failure (Jacquot et al 1982), also including COX-2 inhibitors (Woywodt et al 2001). New NSAIDs, sparing COX-1 and preferentially inhibiting COX-2 may represent progress in the treatment of elderly subjects, but to date a clear renal sparing effect has not been achieved and the cited drugs in clinical use share risks for adverse renal effects similar to those of NSAIDs (Brater 1999). Therefore, further, more selective, COX-2 inhibitors have been developed, such as etoricoxib, valdecoxib and others (Rieudeau et al 2001), but only future clinical studies will show which could be the possible ‘‘ideal’’ NSAID for elderly subjects.
Hepato-gastrointestinal System The liver has an essential metabolic function in the organism, in metabolic degradation of prostaglandins coming from splanchnic or arterial circulation, but also in the local synthesis of prostaglandins. Morita and Murota (1978) have shown that liver synthesizes PGF2a and PGE2 but they have also shown that the PG synthesis in liver from rats older than 72 weeks decreases gradually. At 92 weeks the enzymatic activity was 60% compared to juvenile level, but the PGE2:PGF2a ratio remained constant. On the other hand, the activity of acylation enzymes enabling arachidonic acid incorporation into phospholipids increased gradually with ageing as a mirror image of PG synthesis (Murota and Morita 1980). With the progress of ageing, the prostaglandin synthase activity declines and, in addition, the acylation of arachidonic acid increases, therefore the supply of substrate for PG synthesis decreases and PG production decreases with ageing. Morover, the activation of PG synthase by different stimuli is also reduced in old animals. Park et al (1986) studied the developmental profile of PGsynthesizing enzymes in rat liver from the foetus to 2 year-old animals. In the neonatal period the activities of PGD2 and PGE2 synthases were predominant but later during adult to old ages only PGE2-synthesizing enzyme further increased. Regardless of age, the site of PGE2 production was the hepatocytes, while PGD2 was produced in non-hepatocytes.
309
The liver is often exposed to various hepatotoxic agents which may be more harmful in senescence. Kmiec (1994) investigated the effect of ageing on the cytoprotective properties of prostaglandins. Hepatocytes from old rats (24–28 months) were more susceptible to suppression of protein synthesis by galactosamine than the hepatocytes of young ones (4–6 months). The preincubation of hepatocytes with PGE1 or 9b-methylcarbacyclin (an analogue of PGI2) led to a partial recovery of protein synthesis in both age groups. PGE1 and PGI2 analogues elevated cAMP content in the hepatocytes of young but not old rats. Glucagon and forskolin similarly increased cAMP content in the cells of both young and old animals. These in vitro results suggest that PGE1 and some prostacyclin analogues might protect the hepatocytes of both young and old rats from chemical damage. The cytoprotective effect of prostaglandins has been studied much more in the gastrointestinal system in relation to ageing. The gastrointestinal system is particularly rich in eicosanoids because there are multiple functions in this system, including motility, gastric and intestinal secretion, blood flow and maintenance of mucosal integrity by cytoprotective and gastric acid antisecretory properties. It was also shown that cyclooxygenase inhibitors, such as aspirin, indomethacin and other NSAIDs, may induce gastric ulcer by reducing PG synthesis and, consequently, the cytoprotective action of prostaglandins (Euler 1989). The effect of ageing on gastrointestinal PG biosynthesis and its biological consequences have been studied by several experimental and clinical research workers. Lee and Feldman (1994) investigated the effect of ageing on gastric mucosal eicosanoid formation and aspirin-induced injury in Fischer 344 rats aged 3, 12 and 24 months. Gastric mucosal production of prostaglandins decreased with ageing. Aspirin caused significant mucosal injury in all age groups but significantly more in older rats. These observations indicate that gastric mucosal synthesis decreases with ageing in rats and that aged animals are more susceptible to aspirin-induced acute gastric mucosal injury. More detailed analysis was carried out by Uchida et al (1990), who studied the formation of gastric lesions, induced by orally administered aspirin, in 4–86 week-old male rats. They measured gastric mucosal PGI2 levels and gastric secretion. Gastric lesions reached maximum value in 7 week-old rats and were lowest in 60 week-old animals. Acid output also reached the maximum in 7 week-old rats. PGI2 level showed the maximum in 20 week-old rats and decreased thereafter to 35% levels in 86 week-old rats. The authors observed a linear positive correlation between the formation of aspirin-induced gastric lesions and acid secretion. They concluded that the formation of these lesions is closely related to gastric acidity and may be partly associated with reduced PGI2 level. Human studies complete the experimental results. Goto et al (1992) measured PG content in human gastric mucosa (biopsy samples) of 40 subjects divided into five age groups of eight persons each. The contents of 6-keto-PGF1a, PGF2a, PGE2 and PGD2 in the under-40 age group were 638+39, 97+16, 468+68 and 497+86 pg/mg tissue, respectively. No significant differences in PG contents among groups aged under 70 years were observed. Conversely, significantly low PG contents in the group older than 70 years were observed; the contents of 6-keto-PGF1a, PGF2a, PGE2 and PGD2 were 311+58, 36+8, 196+48 and 171+40, respectively. Thus gastric mucosal PG contents decrease significantly in over-70 year-old subjects and this might be a contributing factor in the pathogenesis of gastric ulcers in elderly people. Another study of Cryer et al (1992) extended the investigation also to duodenal mucosal PG secretion. They measured gastric and duodenal PGE2 and PGF2a concentration in 46 healthy subjects, 35 relatively young (21–40 years) and 11 older (52–72 years). The measurements were performed in biopsy
310
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
specimens obtained endoscopically from the fundus, antrum, duodenal bulb and postbulbar duodenum. A second study in 20 additional subjects (9 younger and 11 older) associated identical PG measurements with the measurement of gastric acid secretion. The results show that older age is associated with significantly lower fundic, antral and postbulbar duodenal PG concentrations, but with significantly higher mean basal acid output. It may be concluded that gastric and duodenal prostaglandin concentrations decline with ageing in humans and that the decline in fundic mucosal prostaglandins is associated with an increase of gastric acid secretion—all factors that may predispose to the development of mucosal lesions. It should be noted that stomach contains both cyclooxygenases, constitutive (COX-1) and inducible (COX-2) (O’Neil and FordHutchinson 1993; Vogiagis et al 2000). It was suggested that it is the constitutive COX-1 activity that protects the gastrointestinal tract. The inhibition of this activity by current NSAIDs may lead to gastrointestinal toxicity. This hypothesis was examined pharmacologically by comparing the ability of non-selective inhibitors of cyclooxygenase, such as indomethacin, and the selective inhibitors of COX-2, sparing COX-1, such as NS-398 and SC-58125, to inhibit gastric production and to cause ulcers (Masferrer et al 1994; Isakson et al 1995). A close correlation between reduction of stomach PG levels by non-selective COX inhibitors and the appearance of gastrointestinal lesions was observed. In contrast, selective COX-2 inhibitors had no effect on gastric PG production and did not cause gastric lesions. These data provide strong evidence that COX-1 is responsible for the production of cytoprotective prostaglandins in the gastrointestinal tract. On the other hand, these experimental results suggest that the inhibitors of COX-2 are potent antiinflammatory agents which do not produce the typical side-effects, i.e. gastric ulcer. The perspectives for better and safer treatment of elderly persons with rheumatic and other painful inflammatory diseases are very optimistic, especially with new drugs such as meloxicam, celecoxib or rofecoxib. In summary, prostaglandins are produced in the liver, as in the whole gastrointestinal tract, and they have multiple functions. One is the cytoprotective effect against several hepatotoxic agents, in the stomach and gut against mucosal injury. In general, ageing reduces hepatic as well as gastrointestinal production of protective prostaglandins, mainly PGI2 and PGE2, which increases the vulnerability of these organs against potentially toxic agents, especially non-specific NSAIDs. The development of more and more selective cyclooxygenase-2 inhibitors may reduce the untoward side-effects of these drugs. Haemato-immunological System Ageing may reduce haematopoiesis, the immunological response in the elderly and alter the defence system against different nocif agents, exogenous or endogenous. What is the role of ageing on prostaglandin function in this system? PGE plays an important role in this immunoregulation, either in the natural progression of immune cell precursors or in many stages of immune reactions, e.g. PGE inhibits T cell but not B cell mitogenesis, it inhibits and indomethacin enhances lymphokine production as well as the proliferation of lymphocytes. The macrophage is the predominant eicosanoid-producing cell in human peripheral blood, but experimental studies have shown that B cells, T cells and macrophages can all synthesize eicosanoids (Fletcher 1989). Licastro and Walford (1986) analysed the proliferative capacity of phytohaemagglutinin- and concanavalin A-stimulated spleen lymphocytes from young and old long-lived mice under the effect of indomethacin and prostaglandins E1 and E2. Mitogen-activated
splenocytes from old mice showed a different sensitivity to these drugs than splenocytes from young animals. They suggested that PGE may play a role in the pathogenesis of age-associated immune impairment. Hayek et al (1994) used also mouse splenocytes, isolated from young (4 months) and old mice (24 months) in order to determine the contribution of suppressive factors, secreted from macrophages, to the age-associated decline in T cell-mediated mitogenic responses, in experiments to characterize eicosanoid and H2O2 production, total cellular fatty acids and vitamin E composition of splenocytes. They observed an age-related increase in Ca2+ ionophore (A 23187)-stimulated ex vivo production of PGE2, leukotrienes B4, C4 and in concanavalin A-stimulated PGE2 production (p50.01). The age-related increase of PGE2 production was also observed in lipopolysaccharide (LPS)-stimulated peritoneal macrophages of mice. Inhibition of cyclooxygenase with indomethacin resulted in increased concanavalin A-stimulated proliferation of splenocytes from old mice (p50.01), while 5-lipoxygenase inhibition did not have any effect on mitogen-induced proliferation. Furthermore, PGE2 addition to purified splenic T-cells decreased their proliferation. No age-related differences were observed in total cellular fatty acid composition, vitamin E level or ex-vivo H2O2 production from stimulated splenocytes. These data indicate that ageing is associated with increased production of PGE2 and leukotrienes from activated splenocytes. Inhibition of PGE2 but not of leukotriene production enhances mitogenic responses of old mice, suggesting a contributory role for PGE2 in the ageassociated decline of T cell responsiveness to polyclonal mitogens. The same conclusion was published by Beharka et al (1997), who demonstrated, moreover, that vitamin E improved T cell responsiveness in old mice, mostly by reducing macrophage PGE2 production. Frainfeld et al (1995) measured PGE2 production in splenocyte cultures of mice aged 1–34 months. Similarly, they have found a significant increase in PGE2 production in response to polyclonal stimulation by concanavalin A in all age-groups. Activated splenocytes from oldest mice (24–48 months) produced two- to three-fold greater amounts of PGE2 as compared with younger mice, and a trend towards an age-associated increase was apparent from the age of 18 months onwards. The authors have not found significant age-related differences in PGE2 levels in unstimulated cultures. The results corroborate the hypothesis that PGE2 is implicated in age-related changes in lymphocyte functions. Increased PGE2 production in response to mitogenic stimuli may affect the profile of cytokines and may limit cellmediated immune response in ageing, such as lymphocyte proliferation and natural killer cell activities. There are clinical studies which show some different results. Riancho et al (1994) were interested by cytokine secretion in healthy young and old subjects and they also measured PGE2 produced by peripheral blood mononuclear cells isolated from 55 healthy volunteers aged 23–73 years. These cells were cultured in the presence or absence of LPS and 1,25-dihydroxy vitamin D3. They have not found age-related differences in the secretion of PGE2 in any culture condition. Peripheral mononuclear cells were used in the study of Rall et al (1996) to determine whether 12 weeks of progressive resistance strength training modifies in vivo and in vitro immune parameters in a controlled study of eight subjects with rheumatoid arthritis, eight healthy young (22–30 years) and eight healthy elderly (65–80 years) individuals. Six healthy elderly (65–80 years) non-training control subjects were also evaluated. The results show that training did not induce changes in peripheral blood mononuclear cell subsets IL-1, IL-1b, TNF, IL-6, IL-2 or PGE2 production in any of the training groups, compared with control subjects. Delfraissy et al (1982) observed that the abolished in vitro antibody response of peripheral lymphomononuclear cells from individuals aged over
AGEING AND PROSTAGLANDINS 70 years was restored by removal of nylon-adherent cell or phagocytic cells. A plastic-adherent, radioresistant non-T cell could suppress this response. The addition of indomethacin completely prevented its effect. Aged peripheral lymphomononuclear cells did not produce more PGE2 but they displayed a higher sensitivity to the suppressive effect of exogenous PGE2. The elimination of radiosensitive T suppressor cells could restore normal helper function in aged T lymphocytes. These observations show that T helper and B cell functions are normal in aged humans but suppressed by an exaggerated prostaglandinmediated T suppressor circuit. Prostaglandins are known to be involved in the maturation and differentiation of the progenitor cells of the bone marrow and erythropoietin-mediated erythropoiesis (Das 1990). However, knowledge of the role of ageing in this system is scarce. Ageing may influence platelet function, the haemostatic system and endothelium-dependent vascular regulation. Reilly and FitzGerald (1986) have shown that ageing is associated with platelet activation in vivo in healthy volunteers, characterized by increased excretion of 2,3-dinor-TXB2. The haemostatic system in the elderly is marked by increased plasma levels of fibrinogen, factor VII, VIII, b-thromboglobulin and increased production of TXA2 but a decreased number of platelet prostacyclin and TXA2 receptors—all elements that are risk factors for thrombotic disease. Finally, clinical and experimental evidence suggests that endothelium could play a central role in haemostatic alterations in the elderly (Abbate et al 1993). The data of Lu¨scher et al (1992) indicate the importance of endothelium as a target organ of cardiovascular risk factors in the elderly, because ageing reduces the production of prostacyclin and nitric oxide and increases the release of endothelin-1 and endothelinderived contracting factor, most likely PGH2 (Michel 2000; Austin and Garett, 2000; Cuevas, 2000). In summary, the immunological response in the elderly is reduced. In contrast to other systems, the production of PGE2 by stimulated immunocompetent cells seems to be increased in elderly animals, which has an inhibitory effect on lymphocyte proliferation and natural killer cell activity and may therefore contribute to the reduced immunological defence in senescence. Ageing is also associated with the alterations in the haemostatic system and vascular endothelium, which enhance the risk of thrombo-embolic complications. Central Nervous System Cerebral ageing is associated with the reduction of the number of neurons and progressive deterioration of cerebral functions. Primary prostaglandins are produced by the vasculature, choroid plexus, neuronal and glial components, and have not only a local haemodynamic regulatory role but especially a neuromodulatory one (Wolfe 1982; Murphy and Pearce 1988; Hedqvist 1977). Ageing is complicated by frequent cerebral ischaemia, which causes a massive release of arachidonic acid with subsequent generation of TXA2 and PGF2a, PGE2 and PGD2. Damaged endothelium may synthesize and release PGI2, which may counteract the effect of vasoconstrictor and proaggregating PGH2 and TXA2 (Rosenbaum et al 1989). Again, the PGI2:TXA2 balance seems to be of importance for the outcome of cerebral hypoperfusion. However, an important role may also be played by other vasoactive substances, as shown by the data of Mayhan et al (1990), who analysed the haemodynamic responses of cerebral arterioles in aged rats. The vasoconstriction produced by the TXA2 analogue U 46619 was similar in adult (6–8 months) and aged (22–24 months) rats. Indomethacin did not affect responses to acetylcholine, serotonin and ADP in aged rats. The dilator responses of cerebral arterioles to agonists that may release EDRF were altered in
311
aged compared with adult rats. Thus, impaired vasodilatation in aged rats does not appear to be related to the production of a cyclooxygenase constrictor substance. Ageing modifies prostaglandin metabolism in the brain. Ueno et al (1985) investigated the enzymatic formation of PGD2, PGE2 and PGF2a in the developing rat brain and also the activity of PG metabolizing enzymes with ageing. Although the activities of NADPdependent 15-hydroxy-PG dehydrogenases were unchanged post-maturationally, the maximal concentrations of the binding sites on the synaptic membrane for both PGD2 and PGE2 decreased with constant affinity to less than one-sixth with age from 1 week to 24 months after birth. These results indicate that prostaglandins may play an important role during maturation and ageing in the rat brain. From the clinical point of view, epidemiological evidence suggests that inhibition of COX may slow down the progress of Alzheimer’s disease (Pasinetti 2000). Raised levels of COX-2 have recently been shown in the frontal cortex of brains of Alzheimer patients at post mortem (Yasojima et al 1999) and also in hippocampal pyramidal neurons, as indicators of progression of dementia (Ho et al 2001). Therefore the most appropriate prophylactic action in ageing could be achieved by the use of specific inhibitors of COX-2, namely celecoxib and rofecoxib (Ferencik et al 2001). These data raised great interest in the role of COX-2 in the ageing process of the central nervous system (Kaufmann et al 1997) and stimulated several experimental studies. Indeed, COX-2, the inducible isoform of cyclooxygenase, is selectively expressed in neurons of the cerebral cortex, hippocampus and amygdala (Andreasson et al 2001). In models of acute excitotoxic neuronal injury, elevated and sustained levels of COX2 have been shown to promote neuronal apoptosis, indicating that upregulated COX-2 activity is injurious to neurons. The same authors developed a model of transgenic mice with overexpressed COX-2 in neurons and produced elevated levels of prostaglandins in brain. In cross-sectional behavioural studies, COX-2 transgenic mice developed an age-dependent deficit in spatial memory at 12 and 20 months but not at 7 months, and a deficit in aversive behaviour at 20 months of age. These behavioural changes were associated with a parallel age-dependent increase in neuronal apoptosis occurring at 14 and 22 months but not at 8 months of age and astrocytic activation at 24 months of age. These findings suggest that neuronal COX-2 may contribute to the pathophysiology of age-related diseases such as Alzheimer’s disease. In the rat, Baek et al (2001) observed significantly increased COX activity in 24 month-old rats compared with 6 month-old rats, but mRNA and protein levels of COX-2 showed little corresponding age-related change. According to McGeer (2000), animal experiments suggest that COX-2 may be performing adaptative functions associated with normal nervous and protective functions associated with stressed neurons. He means that the appropriate target for NSAID trials in Alzheimer’s disease is COX-1, but there is no contraindication to the use of those traditional NSAIDs that have mixed COX-1/COX-2. Some supplementary data show that other mechanisms may play a role in brain ageing, such as lipid peroxidation with the formation of F2-isoprostanes (Pratico et al 1999) or the increased activity of the 5-lipoxygenase system (Qu et al 2000). The above-mentioned data show the important influence of ageing on eicosanoid metabolism in the central nervous system but they also show possible target mechanisms for the preventive treatment of degenerative diseases of the central nervous system. Epidermis Normal skin maturation and skin tissue function rely on the maintenance of the permeability and integrity of all tissue
312
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
membranes. Prostaglandins are essential in the normal maturation processes of the skin and for the maintenance of normal cell function, particularly the relationship between PGE2 and PGF2a (Heggers and Robson 1989). Different tissue components also produce other PGs, such as PGI2, PGD2 or TXA2. Ikai et al (1989) studied the developmental change of enzymes involved in PG synthesis in rat skin from birth to 1.5 years old. The predominant prostaglandin formed was PGD2, whose synthesis was highest at 3 weeks after birth and then gradually decreased up to 1.5 years old. The activities of PGE2 and PGF2a synthases in rat skin were almost unchanged during development and ageing. Polgar and Taylor (1984) have shown that prostacyclin production by skin fibroblasts from human donors aged 30–95 years declined as the age of the cell donors increased. Ageing affects epidermal prostaglandin synthesis, as in other tissues. CLINICAL APPLICATIONS Multiple Role of Prostaglandins in Age-dependent Diseases This review analysed the experimental and clinical data about ageing and prostaglandins, mostly in non-pathological situations or in normally ageing humans. However, several age-dependent deficits or excesses in PG synthesis may be among the basic pathophysiological mechanisms contributing to the development of diseases in senescence. This is especially true for cardiovascular diseases because a deficit of prostacyclin synthesis or an excess of TXA2, or both, have been described in atherosclerosis (Larrue et al 1982; Grylewski et al 1978; Pomerantz and Hajjar 1989), coronary heart disease (FitzGerald et al 1983, 1986; Montalescot et al 1991), cerebrovascular diseases (Pickard and Walker 1984), peripheral obliterative diseases (Sizinger et al 1979; Szceklik et al 1979) and arterial hypertension (Hornych 1991; Grose et al 1983; Hornych et al 1989). A similar deficit of PGI2 with increased synthesis of TXA2 has been described in renal diseases with deteriorated renal function, such as lupus nephritis, chronic glomerulonephritis or other nephropathies (Patrono et al 1985). The decreased renal synthesis of PGI2 in aged subjects (Hornych et al 1991) increases their vulnerability to non-specific NSAIDs and the risk of renal failure (Hornych 1984; Carmichael and Shankel 1985). The same is valid for the gastrointestinal tract, where the use of non-specific NSAIDs in elderly subjects is often associated with the development of ulcers or gastrointestinal bleeding (Euler 1989). On the other hand, increased production of PGE2 by immunocompetent cells in elderly subjects may reduce their natural defence mechanisms. The objective of this review is not to enumerate all pathologies in which prostaglandins are implicated, especially in the elderly population. This field enlarges every year because new metabolites of arachidonic acid are discovered and their physiological and pathophysiological properties become progressively understood. However, it is clear that prostaglandins play an important role in the principal diseases of senescence. Diagnostic Possibilities Actual knowledge of the role of prostaglandins in the development of pathological events in the elderly suggests that the measurement of prostaglandins in biological fluids, such as plasma, serum, urine and cerebrospinal fluid, may offer promising diagnostic possibilites (Mamas 1997; Hornych et al 1982a; Fro¨lich et al 1975; Catella et al 1986; FitzGerald et al 1983; Wennmalm et al 1990; Leonhardt et al 1992) with consequent specific therapeutic and preventive interventions.
The measurement of urinary prostaglandins and their metabolites is especially useful in acute renal failure, in haemostatic and thrombo-embolic complications, in coronary heart diseases, in the post-transplantation period, in the control of the efficacy of aspirin treatment, in gastrointestinal and diabetic complications, but also in many other pathologies (see review by Mamas 1997). It is beyond the scope of this chapter to enumerate all indications. Therapeutic Possibilities Ageing in humans is associated with different defects in the cascade of arachidonic acid. These defects may be treated or ameliorated by dietetic or pharmacological treatment. Dietetic treatment. The target is to furnish the potentially deficient substrates for eicosanoid synthesis, which are dihomog-linolenic acid, arachidonic acid and eicosapentaenoic acid (Figure 26.1). These essential fatty acids cannot be synthesized de novo in humans. They should be ingested or synthesized through desaturation and elongation from linoleic acid (o-6 series) or a-linolenic acid (o-3 series). Humans cannot interconvert the o-3 and o-6 series. Impaired desaturation with ageing (Darcet et al 1980; Hrelia et al 1989) may be compensated by the supply of g-linolenic acid (Darcet et al 1980; Hansen 1983; Hornych et al 2001) as was described previously. Another target of dietetic treatment is to promote the preferential synthesis of triene prostaglandins (Needleman et al 1979) because PGI3 is fully biologically active as a vasodilator and powerful antiaggregant, while TXA3 is biologically inactive. This is achieved by the supply of oils rich in o-3 fatty acids, such as cod liver oil (Lorenz et al 1983) or temperate water fish oils (Exler and Weinrauch 1976) or different preparations of fish oil concentrates (maxEPA). The beneficial effect of o-3 fatty acids has been demonstrated several times, especially in the cardiovascular field (Weber and Leaf 1991; Singer 1991). Pharmacological treatment. Senescence is characterized, in general, by a deficit in PGI2 synthesis, in some cases in PGE2 synthesis and by an excess of TXA2 production. Pharmacological treatment in the case of deficits is based on the use of drugs, (a) able to stimulate deficient PGI2 or PGE2 production or (b) able to act as natural prostaglandins by the use of synthetic prostaglandins or their analogues. Tables 26.1, 26.2 and 26.3 show some of these drugs with their therapeutic indications. An excess of TXA2 or its predominant action may be treated by the use of: (a) inhibitors of TXA2 synthase; (b) competetive antagonists of TXA2 receptors; or (c) drugs with double action (Hornych 1997). Several inhibitors of TXA2 synthase have been developed, such as dazoxiben, dazmegrel, pirmagrel, furegrelate and others, but their clinical efficacy is limited (Fiddler and Lumley 1990). For this reason new drugs have been introduced in clinical trials able to block the action of TXA2 as competitive antagonists of TXA2/PGH2 receptors. Some of them have clinical usefulness, such as vapiprost, solutroban, daltroban or AH 23848. Since high concentrations of TXA2, e.g. after a massive disruption of platelets, may displace the antagonists from their binding to the TXA2 receptor, mixed compounds with dual action have been developed, e.g. TXA2 synthase inhibitors and TXA2 receptor antagonists such as Ridogrel (also being explored as treatment for ulcerative colitis) or Picotamide (also reduces microalbuminuria in type-2 diabetes) (Gresele et al 1991). Ridogrel was successfully used as an adjunct to thrombolysis with streptokinase in 907 patients with acute myocardial infarction (The RAPT Investigators 1994). Progress in the treatment of painful rheumatoid diseases in aged patients was achieved by the introduction of
AGEING AND PROSTAGLANDINS
313
Table 26.1 Drugs used in the treatment of PGI2 and PGE2 deficits Drug
Stimulated PG
Indication
Reference
Cicletanine Furosemide
I2 I2, E2
Hypertension Hypertension, renal failure, heart failure
Captopril Enalapril Nitroglycerine Propranolol Ca antagonists—bepridil, diltiazem, verapamil
E2, I2 I2 I2, E2 I2 I2
Hypertension Hypertension Coronary heart disease Hypertension, coronary heart disease Hypertension, coronary heart disease
(Hornych et al 1989; Guinot and Fro¨lich 1985) (Brater et al 1980; Ciabattoni et al 1979; Patrono et al 1982) (Hornych et al 1982b) (Oparil et al 1987) (Levin et al 1981; Schro¨r et al 1981) (Beckman et al 1988) (Boeynaems 1988)
Table 26.2 PGI2 analogues used in clinical practice Drug
Indication
Reference
Epoprostenol (Prostacyclin) Iloprost
Pulmonary hypertension, haemolytic–uraemic syndrome Peripheral occlusive disease, coronary heart disease Ischaemic heart disease, peripheral vascular disease
Cicaprost Beraprost
Cardiovascular diseases, Raynaud’s phenomenon Peripheral vascular disease
(Rubin et al 1990; Hautekete et al 1986) (Szceklik et al 1979; Uchida et al 1983) (Collins and Djuric 1993; Norgren et al 1990; Fiessenger and Scha¨ffer 1990) (Collins and Djuric 1993; Lau et al 1991) (Collins and Djuric 1993; Saragumi et al 1989)
Table 26.3 PGE analogues used in clinical practice Drug
PG
Indication
Marketed as
Reference
Alprostadil Alprostadil Misoprostol Enprostil Ornoprostol Limaprost
E1 E1 E1 E2 E0 E2
Peripheral vascular disease Induction of erection Antiulcer Antiulcer Antiulcer Antihypertensive
Prostin VR Caverject Cytotec Gardrin Ronak Opalman
(Collins and Djuric (Stackl et al 1988) (Collins and Djuric (Collins and Djuric (Collins and Djuric (Collins and Djuric
specific COX-2 inhibitors such as nimesulide, meloxicam, celecoxib or rofecoxib, sparing the gastrointestinal system but not devoid of possible renal adverse effects. Preventive Possibilities Epidemiological studies have shown that different individual life habits may contribute to the appearance of different diseases during senescence. This is especially valid with regard to diet, salt consumption, nicotine abuse and the lack of physical activity. The potential deleterious effect of these factors seem also to be prostaglandin-dependent. There are several published data to these topics but they go beyond this issue. We mention only the essential points. Diet. We have shown in previous chapters that a balanced diet enriched with g-linolenic acid and o-3 polyunsaturated fatty acids, mainly eicosapentaenoic and docosapentaenoic acids, is beneficial in senescence. These acids are contained in vegetable and fish oils but not in animal fats, which contain mostly saturated fatty acids. NaCl. A high salt diet increases arterial blood pressure and decreases the synthesis of vasodilator and natriuretic PGE2 in normotensive subjects; conversely, a low salt diet increases renal PGE2 synthesis (Campbell et al 1982) and also PGI2 synthesis, measured as urinary 6-keto-PGF1a (Klemm et al 1991). The preventive benefit of a lower salt diet may be of importance in elderly subjects with increasing arterial blood pressure and/or altered renal function with sodium retention.
1993; Gruss et al. 1982) 1993) 1993) 1993) 1993)
Nicotine. Smoking is a high risk factor for the development of coronary heart disease. Cigarette smoking elicits an increased formation of TXA2, indicating platelet activation, in both men and women (Wennmalm et al 1993; Rangemark et al 1992) and even more in older women (Rangemark et al 1992). The synthesis of PGI2 may be unchanged (Wennmalm et al 1993), in women moderately increased (Rangemark et al 1992), but also significantly decreased (Nadler et al 1983). The cessation of smoking normalizes TXA2 synthesis (Wennmalm et al 1993). Therefore, the beneficial effect of quitting smoking is evident. Physical activity. Physical activity plays an important role in the prevention of ischaemic heart disease. Acute exercise increases plasma levels of 6-keto-PGF1a, the stable metabolite of prostacyclin, by 100% (Demers et al 1981) or up to 224% (Mehta et al 1983), according to different authors. However, a low to moderate intensity walking–jogging programme for two months produced a favourable 12.5% increase in plasma prostacyclin and a 36% decrease in serum TXB2 concentrations (Rauramaa et al 1984). It is clear that moderate regular physical activity has a favourable effect on the PGI2:TXA2 balance and on cardiovascular homeostasis in senescence.
CONCLUSION Ageing is associated with several metabolic alterations, including the metabolism of arachidonic acid and the formation of different eicosanoids. These alterations have pathophysiological
314
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
consequences in most organs in senescence. Increasing knowledge in this field has considerably enriched our diagnostic, therapeutic and preventive possibilities, all of which may improve the health of elderly people and prolong their lives. The introduction of new drugs, such as prostacyclin or PGE analogues, TXA2 inhibitors or selective inhibitors of cyclooxygenase-2 represent very promising methods in the treatment of elderly subjects. Ageing is an inevitable biological process but discoveries in the field of eicosanoids may help ‘‘good’’ ageing and the ‘‘well-being’’ of ageing populations.
REFERENCES Abbate R, Prisco D, Rostagno C et al (1993) Age-related changes in hemostatic system. Int J Clin Lab Res, 23, 1–3. Abrahamsen AF (1968) Platelet survival studies in man with special reference to thrombosis and atherosclerosis. Scand J Haematol, 3(suppl), 1–53. Andreasson KI, Savonenko A, Vidensky S et al (2001) Age-dependent cognitive deficits and neuronal apoptosis in cyclooxygenase-2 transgenic mice. J Neurosci, 21, 8198–8209. Austin SC and Garett AF (2000) Prostacyclin. In Bikfalvi A (ed.), Encyclopedic Reference of Vascular Biology and Pathology. Heidelberg, Springer Verlag, 228–236. Ayala S, Gaspar G, Brenner RR et al (1973) Fate of linolenic, arachidonic and docosa-7,10,13,16-tetraenoic acids in rat testicles. J Lipid Res, 14, 296–305. Baek BS, Kim JW, Lee JH et al (2001) Age-related increase of brain cyclooxygenase activity and dietary modulation of oxidative status. J Gerontol A Biol Sci Med Sci, 56, B426–431. Beckman ML, Gerber JG, Byyny RL et al (1988) Propranolol increases prostacyclin synthesis in patients with essential hypertension. Hypertension, 12, 582–588. Beharka AA, Wu D, Han SN et al (1997) Macrophage prostaglandin production contributes to the age-associated decrease in T cell function which is reversed by the dietary antioxidant vitamin E. Mech Ageing Dev, 93, 59–77. Bevis PJ, Bird HA and Lapham G (1996) An open study to assess the safety and tolerability of meloxicam 15 mg in subjects with rheumatic disease and mild renal impairment. Br J Rheumatol, 35 (1), 56–60. Boeynaems J-M (1988) Drugs influencing the vascular production of prostacyclin. Prostagland Leukotrienes Essent Fatty Acids, 34, 197–204. Brannon TS, North AJ, Wells LB and Shaul PW (1994) Prostacyclin synthesis in ovine pulmonary artery is developmentally regulated by changes in cyclooxygenase-1 gene expression. J Clin Invest, 93, 2230– 2235 Brannon TS, Mac Ritchie AN, Jaramillo MA et al (1998) Ontogeny of cyclooxygenase-1 and -2 gene expression in ovine lung. Am J Physiol, 274, L 66–71. Brater DC (1999) Effects of nonsteroidal antiinflammatory drugs on renal function: focus on cyclooxygenase-2-selective inhibition. Am J Med, 107(6A), 65s–70s; discussion 70s–71s. Brater DC, Beck JM, Adams BV and Campbell WB (1980) Effects of indomethacin on furosemide stimulated urinary PGE2 excretion in man. Eur J Pharmacol, 65, 213–219. Brunel P, Hornych A, Guyene TT et al (1995) Renal and endocrine effects of flosulide, after single and repeated administration to healthy volunteers. Eur J Clin Pharmacol, 49, 193–201. Bunting S, Grylewski R, Moncada S and Vane JR (1976) Arterial walls generate from prostaglandin endoperoxides a substance (prostaglandin X) which relaxes strips of mesenteric and coeliac arteries and inhibits platelet aggregation. Prostaglandins, 12, 897–913. Campbell WB, Holland OB, Adams BV and Gomez-Sanchez CE (1982) Urinary excretion of prostaglandin E2, prostaglandin F2a and thromboxane B2 in normotensive and hypertensive subjects on varying sodium intakes. Hypertension, 4, 735–741. Carmichael J and Shankel SW (1985) Effect of non-steroidal antiinflammatory drugs on prostaglandins and renal function. Am J Med, 78, 992–1000. Catella F, Nowak J and Fitzgerald GA (1986) Measurement of renal and non-renal eicosanoid synthesis. Am J Med, 81(suppl 2B), 23–29.
Chang WC and Tai HH (1983) Changes in arachidonate metabolism in aortas and platelets of aging rats. Prostagland Leukotrienes Med, 12, 149–157. Chang WC and Tai HH (1984) Changes in prostacyclin and thromboxane biosynthesis and their catabolic enzyme activity in the kidneys of aging rats. Life Sci, 34, 1269–1280. Chang WC, Murota SI, Nakao J and Orimo H (1980) Age related decrease in prostacyclin biosynthetic activity in rat aortic smooth muscle cells. Biochim Biophys Acta, 620, 159–166. Chiba S, Abe K, Kudo K et al (1984) Sex and age-related differences in the urinary excretion of TxB2 in normal human subjects: a possible pathophysiological role of TxA2 in the aged kidney. Prostagland Leukotrienes Med, 16, 347–358. Chung HY, Kim HJ, Shim KH and Kim KW (1999) Dietary modulation of prostanoid synthesis in the aging process: role of cyclooxygenase-2. Med Ageing Dev, 111, 97–106. Ciabattoni G, Pugliese F, Cinotti GA et al (1979) Characterisation of furosemide-induced activation of renal prostaglandin system. Eur J Pharmacol, 60, 181–187. Collins PE and Djuric JW (1993) Synthesis of therapeutically useful prostaglandin and prostacyclin analogs. Chem Rev, 93, 1533–1564. Cryer B, Redfern JS, Goldschmiedt M et al (1992) Effect of aging on gastric and duodenal mucosal prostaglandin concentrations in humans. Gastroenterology, 102, 1118–1123. Cuevas P (2000) Atherosclerosis. In Bikfalvi (ed), Encyclopedic Reference of Vascular Biology and Pathology. Heidelberg, Springer Verlag, 30–36 Darcet P, Driss F, Mendy F and Delhaye N (1980) Exploration du me´tabolisme des acides gras des lipides totaux plasmatiques et de l’agre´gation plaquettaire d’une population d’hommes age´s a` l’aide d’une alimentation enrichie en acide g-linole´nique. Ann Nutr Alim, 34, 277–290. Das VN (1990) Agranulocytosis, aplastic anemia and leukemia. Relevance to arachidonic acid metabolism. Prostagland Leukotrienes Essent Fatty Acids, 39, 217–220. Davidge ST, Hubel CA and McLaughlin MK (1996) Impairment of vascular function is associated with an age-related increase of lipid peroxidation in rats. Am J Physiol, 271, R1625–R1631. Dekmyn H, Font L, Van Hemelen C et al (1983) Low prostacyclin synthetase activity of fetal rat aorta. Progressive increase after birth. Life Sci, 33, 1491–1497. Delfraissy JF, Galanaud P, Wallon C et al (1982) Abolished in vitro antibody response in elderly: exclusive involvement of prostaglandininduced T suppressor cells. Clin Immunol Immunopathol, 24, 3779– 3785. Demers LM, Harrison TS, Halbert DR and Santen RJ (1981) Effect of prolonged exercise on plasma prostaglandin levels. Prostagland Med, 6, 413–418. Dennis EA (1994) Diversity of group types regulation and function of phospholipase A2. J Biol Chem, 269, 13057–13060. DeWitt DL and Smith WL (1993) PGH synthase isoenzyme selectivity: the potential for safer non-steroidal antiinflammatory drugs. Am J Med, 95(suppl 2A), 40S–44S. Di Pietro R, Miscia S, Cataldi A and Rana R (2000) Age-dependent variations in the expression of PLC isoforms upon mitogenic stimulation of peripheral blood T cells from healthy donors. Br J Haematol, 114, 1209–1214. Edlund A, Bomfin W, Kaiser L et al (1981) Pulmonary formation of prostacyclin in man. Prostaglandins, 22, 323–332. Euler AR (1989) Gastroenterology: eicosanoids and the gastrointestinal tract. In David Watkins W et al (eds), Prostaglandins in Clinical Practice. New York, Raven, 97–129. Exler J and Weinrauch JM (1976) Comprehensive evaluation of fatty acids in foods. VIII. Finfish. J Am Diet Assoc, 69, 243–248. Ferencik M, Novak M, Rovensky J and Rybar I (2001) Alzheimer’s disease, inflammation and non-steroidal antiinflammatory drugs. Bratisl Lek Listy, 102, 123–132. Fiddler GI and Lumley P (1990) Preliminary clinical studies with thromboxane synthase inhibitors and thromboxane receptor blockers. A Review. Circulation, 81, 69–78. Fiessenger JN and Scha¨fer M (1990) Trial of iloprost vs. aspirin treatment for critical limb ischemia of thromboangiitis obliterans. Lancet, 335, 555–557. FitzGerald DJ, Roy L, Catella F and FitzGerald GA (1986) Platelet activation in unstable coronary disease. N Engl J Med, 315, 983–989.
AGEING AND PROSTAGLANDINS FitzGerald GA, Pedersen AK and Patrono C (1983) Analysis of prostacyclin and thromboxane biosynthesis in cardiovascular disease. Circulation, 67, 1174–1177. FitzGerald GA, Smith B, Pedersen AK and Brash AR (1984) Increased prostacyclin biosynthesis in patients with severe atherosclerosis and platelet activation. N Engl J Med, 310, 1065–1068. Fletcher JR (1989) Hematology, Oncology. In David Watkins W et al (eds), Prostaglandins in Clinical Practice, New York, Raven, 59–78. Frainfeld V, Kaplanski J, Kukulansky T and Globerson A (1995) Increased prostaglandin E production by concanavalin A-stimulated splenocytes of old mice. Gerontology, 41, 129–133. Fro¨lich JC, Wilson TW, Sweetman BJ et al (1975) Urinary prostaglandins: identification and origin. J Clin Invest, 55, 763–770. Fu JY, Masferrer JLK, Raz A and Needleman P (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740. Goto H, Sugiyama S, Ohara A et al (1992) Age-associated decrease in prostaglandin contents in human gastric mucosa. Biochem Biophys Res Commun, 186, 1443–1448. Gotoh S, Ogihara T, Nakamaru M et al (1983) Levels of plasma 6-ketoPGF1a in normotensive and essential hypertensive males with and without family history of hypertension. Prostagland Leukotrienes Med, 10, 27–32. Gresele P, Deckmyn H, Nenci GG and Vermylen J (1991) Thromboxane synthase inhibitors, thromboxane receptor antagonists and dual blockers in thrombotic disorders. Trends Pharmacol Sci, 12, 158–163. Grose JH, Lebel M and Gbeassor FM (1983) Imbalanced prostacyclin and thromboxane A2 production in essential hypertension. In Samuelsson B, Paoletti R, Ramwell P (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Research, vol 11. New York, Raven, 413–415. Grossmann A, Rabinovitch PS, Kavanagh TJ et al (1995) Activation of murine T-cells via phospholipase-C g1-associated protein tyrosine phosphorylation is reduced with aging. J Gerontol A Biol Sci Med Sci, 50, B205–B212. Gruss JD, Bartels D, Ohta T et al (1982) Conservative treatment inoperable arterial occlusions of the lower extremities with intraarterial prostaglandin E1. Br J Surg, 69, S11–S13. Grylewski R, Dembinska-Kiec A, Zmuda A and Grylewska T (1978) Prostacyclin and thromboxane A2 biosynthesis capacities of heart, arteries, and platelets at various stages of experimental atherosclerosis in rabbits. Atherosclerosis, 31, 385–394. Guinot P and Fro¨lich JC (1985) Study of the effect of cicletanine on prostanoids. Arzneimittelforschung Drug Res, 35, 1714–1716. Guivernan M, Meza N, Barja P and Roman O (1994) Clinical and experimental study on long-term effect of dietary g-linolenic acid on plasma lipids, platelet aggregation, thromboxane formation and prostacyclin production. Prostagland Leukotrienes Essent Fatty Acids, 51, 311–316. Hamberg M, Svensson J and Samuelsson B (1975) Thromboxanes: a new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Natl Acad Sci USA, 72, 2994–2998. Hansen HS (1983) Dietary essential fatty acids and in vivo prostaglandin production in mammals. World Rev Nutr Diet, 42, 102–134. Hasegawa N and Yamamoto K (1993) A step in the process of prostacyclin production whose decline leads to the age-related decrease in production by human vein umbilical endothelial cells in culture. Mech Ageing Dev, 69, 167–178. Hautekete ML, Nagler JM, Cuykens JJ et al (1986) 6-keto-PGF1a levels and prostacyclin therapy in two adult patients with hemolytic–uremic syndrome. Clin Nephrol, 26, 157–159. Hayek MG, Meydani SN, Meydani M and Blumberg JB (1994) Age differences in eicosanoid production of mouse splenocytes: effects of mitogen-induced T-cell proliferation. J Gerontol, 49, B197–B207. Hayek MG, Mura C, Wu D et al (1997) Enhanced expression of inducible cyclooxygenase with age in murine macrophages. J Immunol, 159, 2445–2451. Hedqvist P (1977) Basic mechanisms of prostaglandin action on autonomic neurotransmission. Ann Rev Pharmacol Toxicol, 17, 259– 279. Heggers JP and Robson MC (1989) Eicosanoids in wound healing. In David Watkins W et al (eds), Prostaglandins in Clinical Practice. New York, Raven, 183–194.
315
Hemler ME, Cook HW and Lands WEM (1979) Prostaglandin biosynthesis can be triggered by lipid peroxides. Arch Biochem Biophys, 193, 340–345. Heymes C, Habib A, Yang D et al (2000) Cyclooxygenase-1 and -2 contribution to endothelial dysfunction in ageing. Br J Pharmacol, 131, 804–810. Ho L, Purohit D, Haroutunian V et al (2001) Neuronal cyclooxygenase-2 expression in hippopocampus formation as a function of the clinical progression of Alzheimer disease. Arch Neurol, 58, 487–492. Hornych A (1984) Role of prostaglandins in drug nephrotoxicity. Contr Nephrol, 42, 220–232. Hornych A (1991) The deficit of prostacyclin in the pathogenesis of hypertension. Cor Vasa, 33, 492–505. Hornych A (1997) Les inhibiteurs des diffe´rentes voies de biosynthe`se des eicosanoı¨ des. In S. Mamas (ed.), Prostaglandines et Thromboxanes. Applications Cliniques et Pharmacologiques. Paris, Masson, 393–452. Hornych A (2000) Prostaglandins. In Bikfalvi A (ed.), Encyclopedic Reference of Vascular Biology and Pathology. Heidelberg, Springer Verlag, 236–244. Hornych A, Krief C and Barie´ty J (1982a) Radioimmunoassay measurement of thromboxane B2 in human plasma and urine. Prostagland Leukotrienes Med, 8, 467–480. Hornych A, Safar M, Simon A et al (1982b) Effects of captopril on prostaglandins and natriuresis in patients with essential hypertension. Am J Cardiol, 49, 1524–1526. Hornych A, Boisson J and Gaschard JC (1989) Diminution de la synthe`se re´nale des prostaglandines chez l’hypertendu essentiel. Arch Mal Coeur, 82(IV), 85–90. Hornych A, Forette F, Barie´ty J et al (1991) The influence of age on renal prostaglandin synthesis of man. Prostagland Leukotrienes Essent Fatty Acids, 43, 191–195. Hornych A, Girault F, Oravec S et al (2001) Effect of g-linolenic acid on plasma and membrane lipids and renal prostaglandins in old subjects. Klin Biochem Metab, 9, 28–34. Horrobin DF (1981) Loss of D-6-desaturase activity as a key factor in aging. Med Hypotheses, 7,1211–1220. Horrobin DF (1990) g-linolenic acid: an intermediate in essential fatty acid metabolism with potential as an ethical pharmaceutical and as a food. Rev Contemp Pharmacother, 1, 1–45. Horrobin DF (1993) Fatty acids metabolism in health and disease: the role of D-6-desaturase. Am J Clin Nutr, 57(suppl), 732S–737S. Hou J, Abe K and Yoshinaga K (1992) The investigation about agerelated changes in vasoactive substances of normal subjects and of patients with essential hypertension. Nippon Jinzo Gakkai Shi, 34, 287– 293. Hrelia S, Bordoni A, Celadon M et al (1989) Age-related changes in linoleate and a-linolenate desaturation by rat liver microsomes. Biochem Biophys Res Commun, 163, 348–355. Ibe BO and Rai JU (1992) Endogenous arachidonic acid metabolism by calcium ionophore stimulated ferret lungs. Effect of age, hypoxia. Lab Invest, 66, 370–377. Ibe BO, Hillyard RM and Rai JU (1996) Heterogeneity in prostacyclin and thromboxane synthesis in ovine pulmonary vascular tree: effect of age and oxygen tension. Exp Lung Res, 22, 351–374. Ignatowska-Switalska H and Januszewicz P (1980) Urinary prostaglandins E2 and F2a in healthy newborns, infants, children and adults. Prostagland Med, 5, 289–296. Ikai K, Ujihara M and Urade Y (1989) Changes of the activities of enzymes ionvolved in prostaglandin synthesis in rat skin during development and aging. Arch Dermatol Res, 281, 433–436. Isakson P, Seibert K, Masferrer JL et al (1995) Discovery of a better aspirin. In Samuelsson B et al (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Res, vol. 23, New York, Raven, 49–54. Ito Y, Ponnappan U and Lipschitz DA (1996) Excess formation of lysophosphatidic acid with age inhibits myristic acid-induced superoxide anion generation in intact human neutrophils. FEBS Lett, 394, 149–152. Jacquot C, Hornych A, d’Auzac de la Martinie C and Barie´ty J (1982) Insuffisance re´nale aigue¨ par la clometacine. Rechute sous diclofe´nac. Inhibition de la synthe`se des prostaglandines. Rev Med Int, 3, 349–351. Jorgensen KA, Dyeberg J, Olesen AS and Stoffersen E (1980) Acetylsalicylic acid, bleeding time and age. Thrombosis Res, 19, 799– 805.
316
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Katano Y, Ishihata A, Morinobu S and Endoh M (1993) Modulation by aging of the coronary vascular response to endothelin-1 in the rat isolated pefused heart. Naunyn Schmiedebergs Arch Pharmacol, 348, 82–87. Katoh K, Mizuno K, Sanada H et al (1991) Effect of aging on enothelininduced prostaglandin I2 release in isolated hind legs of the spontaneously hypertensive rat. Res Commun Chem Pathol Pharmacol, 74, 185–199. Kaufmann WE, Andreasson KI, Isakson PC and Worley PF (1997) Cyclooxygenases and the central nervous system. Prostaglandins, 54, 601–624. Kent RS, Kitchell BB, Shand DG and Whorton AR (1981) The ability of vascular tissue to produce prostacyclin decreases with age. Prostaglandins, 21, 483–490. Kim JW, Baek BS, Kim YK et al (2001) Gene expression of cyclooxygenase in the aging heart. J Gerontol A Biol Med Sci, 56, B350–355. Kim SS, Kim DK and Suh YH ( 1997) Cerebral cortical phospholipase A2 activity of senescence-accelerated mouse is increased in an agedependent manner. Neurosci Res, 29, 269–272. Kimura S, Minami M, Saito H et al (1995) Dietary docosahexaenoic acid (22:6n-3) prevents the development of hypertension in SHRSP. Clin Exp Pharmacol Physiol, 22(suppl 1), S308– S 309. Klemm SA, Hornych AF, Tunny TJ and Gordon RD (1991) The syndrome of hypertension and hypercalcaemia with normal glomerular filtration rate: is there a deficiency in vasodilator prostaglandins? Clin Exp Pharmacol Physiol, 18, 309–313. Kmiec Z (1994) Prostaglandin cytoprotection of galactosamine-incubated hepatocytes isolated from young and old rats. Ann NY Acad Sci, 717, 216–225. Larrue J, Leroux C and Bricaud H (1982) Decreased prostaglandin production in cultured smooth muscle cells from atherosclerotic rabbit aorta. Biochem Biophys Acta, 710, 257–263. Larrue J, Daret D, Henri J and Razaka G (1991) Eicosanoı¨ des et vieillissement arte´riel. Ann Cardiol Ange´iol, 40, 259–264. Lau CS, McLaren M, Saniabadi A et al (1991) The pharmacological effects of cicaprost, an oral analogue, in patients with Raynaud’s syndrome secondary to systemic sclerosis—a preliminary study. Clin Exper Rheumatol, 9, 271–273. Lee M and Feldman M (1994) Age-related reductions in gastric mucosal prostaglandin levels increase susceptibility to aspirin-induced injury in rats. Gastroenterology, 107, 1746–1750. Lennon EA and Poyser NL (1986) Effect of age on vascular prostaglandin production in male and female rats. Prostagland Leukotrienes Med, 25, 1–5. Lennon EA, Ansell I and Poyser NL (1988) The effect of age and smoking on vascular prostaglandin production in men and women. Prostagland Leukotrienes Essent Fatty Acids, 32, 15–21. Leonhardt A, Busch C, Schweer H and Seyberth HW (1992) Reference intervals and developmental changes in urinary prostanoid excretion in healthy newborns, infants and children. Acta Pediatr, 81, 191–196. Levin RI, Jaffe RA, Weksler BB and Tack-Goldman K (1981) Nitroglycerin stimulates synthesis of prostacyclin by cultured human endothelial cells. J Clin Invest, 67, 762–769. Licastro F and Walford RL (1986) Effects exerted by prostaglandins and indomethacin on the immune response during aging. Gerontology, 32, 1–9. Lijnen P, Van Damme-Lombaerts R, Proesmans W et al (1983) Urinary excretion of immunoreactive prostaglandins F2a in healthy children and adults. Prostaglandins, 26, 761–772. Lorenz R, Spengler V, Fischer S et al (1983) Platelet function, thromboxane formation and blood pressure control during supplementation of the Western diet with cod liver oil. Circulation, 67, 504–511. Lu¨scher TF, Tanner FC and Dohi Y (1992) Age, hypertension and hypercholesterolemia alter endothelin-dependent vascular regulation. Pharmacol Toxicol, 70, S32– S39. Lu¨scher TF, Boulanger CM, Yang Z et al (1993) Interactions between endothelium-derived relaxing and contracting factors in health and cardiovascular disease. Circulation, 87(suppl V), V36–V44. Mackenzie T, Zawada ET, Johnson MD and Green S (1984) The importance of age on prostaglandin E2 excretion in normal and hypertensive men. Nephron, 38, 178–182.
Mamas S (1997) Prostaglandins and Thromboxanes. Applications Cliniques et Pharmacologiques. Paris, Masson, 1–552. Markov KM, Ivanova IA and Zadkova GF (1983) Age-related characteristics of prostaglandin system in certain organs. Fiziol Zh SSSR, 69, 832–837. Masferrer JL, Zweifel BS, Manning PT et al (1994) Selective inhibition of inducible cyclooxygenase-2 in vivo is anti-inflammatory and nonulcerogenic. Proc Natl Acad Sci USA, 91, 3228–3232. Matz RL, de Sotomayor MA, Schott C et al (2000) Vascular bed heterogeneity in age-related endothelial dysfunction with respect to NO and eicosanoids. Br J Pharmacol, 131, 303–311. Mayhan WG, Faraci FM, Baumbach GL and Heistad DD (1990) Effects of aging on responses of cerebral arterioles. Am J Physiol, 258, H1138– H1143. McGeer PL (2000) Cyclooxygenase-2 inhibitors: rationale and therapeutic potential for Alzheimer disease. Drug Aging, 17, 1 –11. Mehta J, Mehta P and Hovalek C (1983) The significance of platelet-vessel wall prostaglandin equilibrium during exercise-induced stress. Am Heart J, 105, 895–900. Menconi M, Taylor M, Martin B and Polgar P (1987) A review: prostaglandins, aging and blood vessels. J Am Geriat Soc, 35, 239–247. Michel JB (2000) Vasomotor tone regulation, molecular mechanism of. In Bikfalvi A (ed.), Encyclopedic Reference of Vascular Biology and Pathology. Heidelberg, Springer Verlag, 345–351. Millatt LJ and Siragy HM (2000) Age-related changes in renal cyclic nucleotides and eicosanoids in response to sodium intake. Hypertension, 35, 643–647. Mitchell JA, Akarasereenont P, Thiemermann C et al (1994) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697. Mizutani T, Nakashima S and Nozawa Y (1998) Changes in the expression of protein kinase (PKC), phosphokinases C (PLC) and D (PLD) isoforms in spleen, brain, and kidney of the aged rat: RT–PRC and Western blot analysis. Mech Ageing Dev, 105, 151–172. Moncada S and Vane JR (1979) Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2, and prostacyclin. Pharmacol Rev, 30, 293–331. Montalescot G, Drobinski G, Maclouf J et al (1991) Evaluation of thromboxane production and complement activation during myocardial ischemia in patients with angina pectoris. Circulation, 84, 2054–2062. Morita I and Murota S (1978) Prostaglandin-synthesizing system in rat liver. Eur J Biochem, 90, 441–449. Morita I and Murota S (1980) Effect of aging on the prostaglandinsynthesizing system in rat liver. Prostagland Med, 4, 45–52. Morrow JD, Hill KE, Bork RF et al (1990) A series of prostaglandin F2like compounds are produced in vivo in humans by a noncyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci USA, 87, 9383–9387. Murota S and Morita I (1980) Prostaglandin-synthesizing system in rat liver: changes with aging and various stimuli. In Samuelsson B, Ramwell P, Paoletti VR (eds), Advances in Prostaglandin and Thromboxane Research, vol 8. New York, Raven Press, 1495–1506. Murota S, Mitsui Y and Kawamura M (1979) Effect of in vitro aging on 6ketoprostaglandin F1a-producing activity in cultured human diploid lung fibroblasts. Biochim Biophys Acta, 574, 351–355. Murota S, Matsuoka K, Mitsui Y et al (1983) Age-related shift from the biosynthesis of prostacyclin to thromboxane A2 and change in cyclic AMP product in response to prostacyclin in human diploid fibroblasts. In Samuelsson B, Paoletti R, Ramwell, P (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Res, vol 12. New York, Raven, 319–325. Murphy S and Pearce B (1988) Eicosanoids in CNS: sources and effects. Prostagland Leukotrienes Essent Fatty Acids Rev, 31, 165–170. Nadler JL, Velasco JS and Horton R (1983) Cigarette smoking inhibits prostacyclin formation. Lancet, 1, 1248–1250. Narang N, Joseph JA, Ayyagari PV et al (1996) Age-related loss of cholinergic–muscarinic coupling to PLC: comparison with changes in brain regional PLC subtypes mRNA distribution. Brain Res, 708, 143– 152. Navarette R, Tang W and Ziboh VA (1992) Dietary intake of concentrated g-linolenic acid (GLA)-enriched oil suppresses cutaneous level of dihomo-g-linolenic acid (DGLA): Possible in vivo
AGEING AND PROSTAGLANDINS inhibition of microsomal elongation of GLA to DGLA. Prostagland Leukotrienes Essent Fatty Acids, 46, 139–144. Needleman P, Raz A, Minkes MS et al (1979) Triene prostaglandins: Prostacyclin and thromboxane biosynthesis and unique biological properties. Proc Natl Acad Sci USA, 76, 944–948. Neri Serneri GG, Masotti G, Gensini GF et al (1981) Prostacyclin, thromboxane and ischemic heart disease. In Johnson Hegyeli R (ed), Prostaglandins and Cardiovascular Disease, Raven Press, New York, pp. 139–157. Norgren L, Alumark A, A¨ngqvist KA et al [ 22 authors, multicenter study] (1990) A stable prostacyclin analogue (Iloprost) in the treatment of ischemic ulcers of the lower limb. Eur J Vasc Surg, 4, 463–467. O’Neil GP and Ford-Hutchinson AW (1993) Expresion of mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human tissues. FEBS, 330, 156–160. Oelz O, Seybert HW, Knapp HR et al (1976) Effects of feeding ethyldihomo-g-linolenate on prostaglandin biosynthesis and platelet aggregation in the rabbit. Biochem Biophys Acta, 431, 268–277. Oparil S, Horton R, Wilkins LH et al (1987) Antihypertensive effect of enalapril in essential hypertension: Role of prostacyclin. Am J Med Sci, 30, 395–402. Pace-Asciak CR (1976) Biosynthesis and catabolism of prostaglandins during animal development. In Samulesson B and Paoletti R(eds), Advances in Prostaglandin and Thromboxane Res, vol. 1, Raven Press, New York, pp. 35–46. Pace-Asciak CR and Edwards MS (1980) Distinct age-related changes in the urinary catabolic profile in prostaglandin F2a in the rat. J Biol Chem, 255, 6106–6110. Park SW, Ueno R and Hayaishi O (1986) Prostaglandin synthesizing enzyme activities of hepatocytes and non-hepatocytes in rat liver as a function of age. Biochem Int, 13, 827–834. Pasinetti GM (2001) Cyclooxygenase and Alzheimer’s disease: implications for preventive initiatives to slow the progression of clinical dementia. Arch Gerontol Geriatr 33, 13–28. Patrick CB, Krzywkowski P, Ramassany C et al (2000) Phospholipase A2 activity is decreased selectively in the hippocampus of aged apolipoprotein E deficient mice. Neurosci Lett A, 288, 211–214. Patrono C, Pugliese F, Ciabattoni G et al (1982) Evidence for a direct stimulatory effect of prostacyclin on renin release in man. J Clin Invest, 69, 231–239. Patrono C, Ciabattoni G, Remuzzi G et al (1985) Functional significance of renal prostacyclin and thromboxane A2 production in patients with systemic lupus erythematosus. J Clin Invest, 76, 1011–1018. Peri KG, Hardy P, Li DY et al (1995) Prostaglandin G/H synthase-2 is a major contributor of brain prostaglandins in newborn. J Biol Chem, 270, 24615–24620. Petkova DH, Momchilova AB and Koumanov KS (1986) Age-related changes in rat liver plasma membrane phospholipase A2 activity. Exp Gerontol, 21,187–193. Pickard JD and Walker V (1984) Current concepts of the role of prostaglandins and other eicosanoids in acute cerebrovascular disease. In MacKenzie ET, Seyaz J, Be`s A (eds) L.E.R.S. Monograph Series vol 2: Neurotransmitters and the Cerebral Circulation. New York, Raven, 191–218. Polgar P and Taylor L (1984) Prostaglandins: participants in cellular aging. In Ordy JM, Harman D, Alfin-Slater V (eds), Nutrition in Gerontology. New York, Raven, 119–139. Pomerantz KB and Hajjar DP (1989) Eicosanoid metabolism in cholestrol-enriched arterial smooth muscle cells: reduced arachidonate release with concomitant decrease in cyclooxygenase products. J Lipid Res, 30, 1219–1231. Pratico D, Rokach J and Tangirala RK (1999) Brains of aged apolipoprotein E-deficient mice have increased levels of F2isoprostanes, in vivo markers of lipid peroxidation. J Neurochem, 73, 736–741. Qu T, Uz T and Manev H (2000) Inflammatory 5-LOX mRNA and protein are increased in brain of aging rats. Neurobiol Aging, 21, 647– 652. Rall LC, Roubenoff R, Cannon JG et al (1996) Effects of progressive resistance training on immune response in aging and chronic inflammation. Med Sci Sports Exerc, 28, 1356–1365. Rangemark C, Benthin G, Granstro¨m EF et al (1992) Tobacco use and urinary excretion of thromboxane A2 and prostacyclin metabolites in women stratified by age. Circulation, 86, 1495–1500.
317
Rathaus M, Greenfeld Z, Podjarny E et al (1993) Sodium loading and renal prostaglandins in old rats. Prostagland Leukotrienes Essent Fatty Acids, 49, 815–819. Rauramaa R, Salonen JT, Kukkonen-Harjula K et al (1984) Effect of mild physical exercice on serum lipoproteins and metabolites of arachidonic acid: a controlled randomized trial in middle-aged men. Br Med J, 288, 603–606. Raymond FD, Fortunato G, Moss DW et al (1994) Inositol-specific phospholipase D activity in health and disease. Clin Sci, 86, 447–451. Reckelhoff JF, Kanji V, Racusen LC et al (1998) Vitamin E ameliorates enhanced renal lipid peroxidation and accumulation of F2isoprostanes in aging kidneys. Am J Physiol, 274, R767–774. Reilly IAG and FitzGerald GA (1986) Eicosanoid biosynthesis and platelet function with advancing age. Thrombosis Res, 41, 545–554. Riancho JA, Zarrabeitia MT, Amado JA et al (1994) Age-related differences in cytokine secretion. Gerontology, 40, 8–12. Riendeau D, Percival MD, Brideau C et al (2001) Etoricoxib (MK-0663): preclinical profile and comparison with other agents that selectively inhibit cyclooxygenase-2. J Pharmacol Exp Ther, 296, 558–566. Roberts LJ and Reckeloff JF (2001) Measurement of F(2)-isoprostanes unveils profound oxidative stress in aged rats. Biochem Biophys Res Comm, 287, 254–256. Rosenbaum DM, McKenzie JD, Pettigrew LC and Yatsu FM (1989) Neurology. In David Watkins W et al (eds), Prostaglandins in Clinical Practice. New York, Raven, 211–225. Rubin LJ, Mendoza J, Hood M et al (1990) Treatment of primary pulmonary hypertension with continous intravenous prostacyclin (Epoprostenol). Ann Int Med, 11, 485–491. Samuelsson B (1976) New trends in prostaglandin research. In Samuelsson B, Paoletti R (eds), Advances in Prostaglandin and Thromboxane Research, vol. 1. New York, Raven, 1–6. Saragumi S, Tanaba T, Mishima Y et al (1989) Clinical evaluation of the efficacy of beraprost sodium, a prostacyclin derivative, on chronic arterial occlusion. Med Consult New Rem, 26(9), 32. Sato I, Kaji K, Morita I et al (1993) Augmentation of endothelin-1, prostacyclin and thromboxane A2 secretion associated with in vitro ageing in cultured human umbilical vein endothelial cells. Mech Ageing Dev, 71, 73–84. Sato M, Abe K, Takeuchi K et al (1989) PGE2 synthesis in cultured renal papillary collecting tubule cells from young and aged spontaneously hypertensive rats. Adv Exp Med Biol, 247B, 501–507. Schro¨r K, Grodzinska L and Darius H (1981) Stimulation of coronary vascular prostacyclin and inhibition of human platelet thromboxane A2 after low-dose nitroglycerin. Thromb Res, 23, 59–67. Seybert HW, Oelz O, Kennedy T et al (1975) Increased arachidonate in lipids after administration to man. Effects on prostaglandin biosynthesis. Clin Pharmacol Ther, 18, 521–529. Sie P, Montagut J, Blanc M et al (1981) Evaluation of some platelet parameters in a group of elderly people. Thromb Haemost, 45, 197–199. Singer P (1991) Blood pressure-lowering effect of o-3 polyunsaturated fatty acids in clinical studies. In Simopoulos AP, Kifer RR, Martin RE and Barlow SM (eds), Health Effects of o-3 Polyunsaturated Fatty Acids in Seafoods. World Reviews of Nutrition and Diet, vol 66. Basel, Karger, 329–348. Sizinger H, Feigh W and Silbauer K (1979) Prostacyclin geneneration in atherosclerotic arteries (letter). Lancet, 2, 469. Smith WL (1992) Prostanoid biosynthesis and mechanism of action. Am J Physiol, 263, F181–F191. Sraer J, Ardaillou R, Sraer JD and Ardaillou R (1982) In vitro prostaglandin synthesis by human glomeruli and papillae. Prostaglandins, 23, 855–864. Stackl V, Hasun R and Marberger M (1988) Intracavernous injection of prostaglandin E1 in impotent men. J Urol, 140, 66–68. Steinhauslin F, Munafo A, Buclin T et al (1993) Renal effects of nimesulide in furosemide-treated subjects. Drugs, 46(suppl 1), 257–262. Strosznajder J, Samochocki M, Wikiel H and Malecki A (1994) Aging modulates calcium-dependent phosphatidylinositol degradation by cerebral cortex synaptic plasma membrane phospholipases. Mol Chem Neuropathol, 21, 95–107. Swan SK, Rudy DW, Lasseter KC et al ( 2000) Effect of cyclooxygenase-2 inhibition on renal function in elderly persons trial. Arch Int Med, 133, 1–9.
318
IMMUNOLOGY, ENDOCRINOLOGY AND METABOLIC REGULATION
Szceklik A, Skawinski S, Gluszko P et al (1979) Sucessful therapy of advanced arteriosclerosis obliterans with prostacyclin. Lancet, 1, 1111– 1114. Takahashi R and Horrobin DF (1988) Linoleic acid-induced fatty acid changes in platelet and aorta of the rat: effect of age and cholesterol. Lipids, 23, 885–890. Takahashi R, Manku MS, Jenkins K and Horrobin DF (1990) Ageassociated reduction of prostacyclin and thromboxane synthesis is inversely related to plasma cholesterol levels: Modulation by dietary cholesterol supplementation. Prostagland Leukotrienes Essent Fatty Acids, 39, 65–67. Takeuchi K, Abe K, Sato M et al (1987) Effect of in vitro aging on prostaglandin synthesis in cultured rat vascular smooth muscle cells. Agents Act Suppl, 22, 49–54. Taylor L, Schneider E, Smith J and Polgar P (1981) Prostaglandin production and cellular aging. Mech Ageing Dev, 16, 311–317. Taylor RR, Sturm M, Kendrew PJ et al (1989) Plasma levels of the lysoderivative of platelet-activating factor are related to age. Clin Sci, 76, 195–198. The RAPT Investigators (1994) Randomized trial of Ridogrel, a combined thromboxane A2 synthase inhibitor and thromboxane A2/ prostaglandin endoperoxide receptor antagonist, vs. aspirin as adjunct to thrombolysis in patients with acute myocardial infarction. Circulation, 89, 588–595. Todd MK, Goldfarb AH, Kaufman RD and Burleson C (1994) Combined effects of age and exercise on thromboxane B2 and platelet activation. J Appl Physiol, 76, 1548–1552. Tokunaga O, Yamada T, Fan J and Watanabe T (1991) Age-related decline in prostacyclin synthesis by human aortic endothelial cells. Am J Pathol, 138, 941–949. Uchida M, Misaki N, Kawano O et al (1990) Change in formation of gastric lesions by aspirin during aging in rats. J Pharmacobiodyn, 13, 493–496. Uchida Y, Hanai T, Hasagawa K et al (1983) Recanalisation of obstructed coronary artery by intracoronary administration of prostacyclin in patients with acute myocardial infarction. In Samuelsson B, Paoletti R, Ramwell P (eds), Advances in Prostaglandin, Thromboxane and Leukotriene Research, vol 11. New York, Raven, 377–383. Ueno R, Osama H, Urade Y and Hayaishi O (1985) Changes of enzymes involved in prostaglandin metabolism and prostaglandin binding proteins in rat brain during development and aging. J Neurochem, 45, 483–489. Undie AS, Wang HY and Friedman E (1995) Decreased phospholipase Cb immunoreactivity, phosphoinositide metabolism, and protein kinase C activation in senescent F-344 rat brain. Neurobiol Aging, 16, 19–28. Vogiagis D, Glare EM, Misaion A et al (2000) Cyclooxygenase-1 and alternatively spliced mRNA in the rat stomach: effects of aging and ulcers. Am J Physiol, 278, G820–827. Wahl M and Carpenter G (1991) Selective phospholipase C activation. Bioassays, 13, 107–113.
Wang Z, Ciabattoni G, Creminon C et al (1995) Immunological characterisation of urinary 8-epi-prostaglandin F2a excretion in man. J Pharmacol Exp Ther, 275, 94–100. Whelton A, Schulman G, Wallemark C et al (2000) Effects of celecoxib and naproxen on renal function in the elderly. Arch Intern Med, 160, 1465–1470. Weber PC and Leaf A (1991) Cardiovascular effects of o-3 fatty acids. In Simopoulos AP, Kifer RR, Martin RE and Barlow SM (eds), Health Effects of o-3 Polyunsaturated Fatty Acids in Seafoods. World Review of Nutritional Diet, vol 66, Basel, Karger, 218–232. Wennmalm A˚, Benthin G, Caidahl K et al (1990) Non-invasive assessement of the cardiovascular eicosanoids, thromboxane A2 and prostacyclin, in randomly sampled males, with special reference to the influence of inheritance, and environmental factors. Clin Sci, 79, 639– 645. Wennmalm A˚, Benthin G, Granstro¨m EF et al (1993) Maintained hyperexcretion of thromboxane A2 metabolite in healthy young cigarette smokers: results from a prospective study of randomly sampled males with stable smoking habits. Clin Physiol, 13, 257–264. Williams WM, Chang MC and Rapoport SI (1994) Cerebral microvessel phospholipase A2 activity in senescent mouse. Neurochem Res, 19, 317– 320. Wilson TW, McCaulay FA and Waslen TA (1989) Effects of aging on response to furosemide. Prostaglandins, 38, 675–687. Wolfe LS (1982) Eicosanoids: prostaglandins, thromboxanes, leukotrienes, and other derivatives of carbon-20 unsaturated fatty acids. J Neurochem, 38, 1–14. Woywodt A, Schwartz A, Mengel M et al ( 2001) Nephrotoxicity of selective COX-2 inhibitors. J Rheumatol, 28, 2133–2135. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Yasojima K, Schwab C, McGeer EG and McGeer PL ( 1999) Distribution of cyclooxygenase-1 and cyclooxygenase-2 mRNAs and proteins in human brain and peripheral organs. Brain Res, 830, 226–236. Ylikorkala O, O¨sterman T, Linde´n I-B and Viinikka L (1982) The effect of age on circulating 6-keto-prostaglandin F1a in humans. Prostagland Leukotrienes Med, 9, 569–575. Yokoyama M, Kusui A, Sakamoto S and Fukuzaki H (1985) Ageassociated increments in human platelet a-adrenoreceptor capacity. Possible mechanisms for platelet hyperactivity to epinephrine in aging man. Thrombosis Res, 34, 387–395. Yu BP, Suescun EA and Yang SY (1992) Effect of age-related lipid peroxidation on membrane fluidity and phospholipase A2: modulation by dietary restriction. Mech Ageing Dev, 65, 17–33. Zahavi J, Jones NAG, Leyjon J et al (1980) Enhanced in vivo platelet ‘‘release reaction’’ in old healthy individuals. Thrombosis Res, 17, 329– 336. Zhao SZ, Reynold MW, Lejkowith J et al ( 2001) A comparison of renalrelated adverse drug reactions between rofecoxib and colecoxib, based on the World Health Organization/Uppsala Monitoring Centre Safety database. Clin Ther, 23, 1478–1491.
Section Five Inflammation
27 Perspectives and Clinical Significance of Eicosanoids in Pain and Inflammation Burkhard Hinz and Kay Brune Friedrich Alexander University Erlangen-Nu¨rnberg, Erlangen, Germany
INTRODUCTION In 1971, Vane showed that the antiinflammatory action of nonsteroidal antiinflammatory drugs (NSAIDs) rests in their ability to inhibit the activity of the cyclooxygenase (COX) enzyme, which in turn results in a diminished synthesis of proinflammatory prostaglandins (Vane 1971). This action is considered to be not the sole but a major factor of the mode of action of NSAIDs. The pathway leading to the generation of prostaglandins has been elucidated in detail. Within this process, the COX enzyme (also referred to as prostaglandin H synthase) catalyses the first step of the synthesis of prostanoids by converting arachidonic acid into prostaglandin H2, which is the common substrate for specific prostaglandin synthases. The enzyme is bifunctional, with fatty acid COX activity (catalysing the conversion of arachidonic acid to prostaglandin G2) and prostaglandin hydroperoxidase activity (catalysing the conversion of prostaglandin G2 to prostaglandin H2) (Figure 27.1). In the early 1990s, COX was demonstrated to exist as two distinct isoforms (Fu et al 1990; Xie et al 1991). COX-1 is constitutively expressed as a ‘‘housekeeping’’ enzyme in nearly all tissues, and mediates physiological responses (e.g. cytoprotection of the stomach, platelet aggregation). On the other hand, COX-2 expressed by cells that are involved in inflammation (e.g. macrophages, monocytes, synoviocytes) has emerged as the isoform that is primarily responsible for the synthesis of the prostanoids involved in pathological processes, such as acute and chronic inflammatory states. Accordingly, many of the side effects of NSAIDs (e.g. gastrointestinal ulceration and bleeding, platelet dysfunctions) can be ascribed to a suppression of COX-1-derived prostanoids, whereas inhibition of COX-2-dependent prostaglandin synthesis accounts for the antiinflammatory, analgesic and antipyretic effects of NSAIDs (Figure 27.1). Consequently, the hypothesis that specific inhibition of COX-2 might have therapeutic actions similar to those of NSAIDs, but without causing the unwanted side-effects, was the rationale for the development of specific inhibitors of the COX-2 enzyme as a new class of antiinflammatory and analgesic agents with improved gastrointestinal tolerability.
REGULATION OF COX-2 EXPRESSION DURING INFLAMMATION The genes for COX-1 and COX-2 are located on human chromosomes 9 and 1, respectively (Kraemer et al 1992). Whereas The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
COX-1 represents a housekeeping gene which lacks a TATA box (Kraemer et al 1992), the promotor of the immediate-early gene COX-2 contains a TATA box and binding sites for several transcription factors including nuclear factor-kB (NF-kB), the nuclear factor for interleukin-6 expression (NF-IL-6) and the cyclic AMP response element binding protein (CREB) (Appleby et al 1994). Thus, the expression of COX-2 is regulated by a broad spectrum of mediators involved in inflammation. Whereas lipopolysaccharide, proinflammatory cytokines, e.g. interleukin(IL)-1b, tumour necrosis factor (TNF), and growth factors may induce COX-2, glucocorticoids, IL-4, IL-13 and the antiinflammatory cytokine IL-10 have been reported to inhibit the expression of this enzyme (Lee et al 1992; Onoe et al 1996; Niiro et al 1997). Moreover, evidence is emerging to suggest that products of the COX-2 pathway may cell-dependently exert regulatory feedback actions on the expression of their biosynthesizing enzymes. Accordingly, a recent study using the rat model of carrageenan-induced inflammation (Nantel et al 1999a) has shown that indomethacin may block COX-2 expression in the inflamed paw, implying that prostaglandins produced at sites of inflammation may potentiate COX-2 expression via a positive feedback loop. In agreement with this finding, the major COX-2 product prostaglandin E2 has been shown to upregulate COX-2 expression by virtue of its cAMP-elevating capacity in a variety of cell types, including human blood monocytes (Hinz et al 2000a), rat microglia cells (Minghetti et al 1997), murine macrophages (Hinz et al 2000b) and murine keratinocytes (Maldve et al 2000). COX-2 is also regulated at the post-transcriptional level. Recently, a 3’-untranslated region of its mRNA has been shown to contain multiple copies of adenlylate- and uridylate-rich elements (AREs) that may confer post-transcriptional control of COX-2 expression by acting as a mRNA instability determinant or as a translation inhibitory element (Dixon et al 2000). Loss of this post-transcriptional regulation of COX-2 through mutation of proteins that specifically interact with the COX-2 ARE may lead to COX-2 overexpression and has been proposed as a crucial factor involved in colon carcinogenesis (see also section on Role of COX-2 in Cancer, below).
ROLE OF COX-2 IN PAIN PERCEPTION Inflammation causes an increased synthesis of COX-2-dependent prostaglandins, which sensitize peripheral nociceptor terminals and produce localized pain hypersensitivity. Recently, it has been shown that prostaglandins regulate the sensitivity of so-called
322
INFLAMMATION
Figure 27.1 Physiological and pathophysiological effects of the COX isoenzymes – COX hypotheses and concepts of the years 1990 and 2003
polymodal nociceptors that are present in nearly all tissues. A significant portion of these nociceptors cannot be easily activated by physiological stimuli such as (mild) pressure or (some) increase of temperature (Schaible and Schmidt 1988). However, following tissue trauma and subsequent release of prostaglandins, ‘‘silent’’ polymodal nociceptors become excitable to pressure, temperature changes and tissue acidosis (Neugebauer et al 1995). This process results in a phenomenon called hyperalgesia—in some instances allodynia. Prostaglandin E2 and other inflammatory mediators facilitate the activation of tetrodotoxin (TTX)-resistant Na+ channels in dorsal root ganglion neurons. A certain type of TTX-resistant Na+ channel has recently been cloned (Akopian et al 1996) that appears to be selectively expressed in small and medium-sized dorsal root ganglion neurons. Compelling evidence indicates that these small dorsal root ganglion neurons are the somata which give rise to thinly and unmyelinated C and Ad nerve fibres, both conducting nociceptive stimuli. Increased opening of these Na+ channels involves activation of the adenylyl cyclase enzyme and increases in cyclic AMP, possibly leading to protein kinase A-dependent phosphorylation of the channels. On the basis of this mechanism, prostaglandins produced during inflammatory states may significantly increase the excitability of nociceptive nerve fibres, thereby contributing to the activation of ‘‘sleeping’’ nociceptors. As such, it appears reasonable that at least a part of the peripheral antinociceptive action of acidic antipyretic analgesics arises from prevention of this sensitization. Within the past few years it has become increasingly clear that, apart from sensitizing peripheral nociceptors, prostaglandins may also act in the central nervous system to produce hyperalgesia. Experimental data suggest that COX inhibitors act primarily in the dorsal horn to cause analgesia (for review, see Brune et al 1999). Here, nociceptor signals are transferred to secondary neurons, which propagate the signals to the higher centres of the central nervous system. The sensation of pain is then assembled in the cortex. COX-2 is expressed constitutively in the dorsal horn of the spinal cord, and becomes upregulated briefly after a trauma, such as damage to a limb, in the corresponding sensory segments of the spinal cord (Beiche et al 1996). Compelling evidence suggests that the induction of spinal cord COX-2 expression may
facilitate transmission of the nociceptive input. Within its broad spectrum of actions, a direct depolarization of spinal neurons by spinal cord-generated prostanoids has been recently suggested to contribute the nociceptive action of prostaglandin E2 in this process (Baba et al 2001). In line with a role of COX-2 in central pain perception, Smith et al (1988) reported that the specific COX-2 inhibitor, celecoxib, suppressed inflammation-induced prostaglandin levels in cerebrospinal fluid, whereas the selective COX-1 inhibitor SC-560 was inactive in this regard. These observations were substantiated by recent findings that show a widespread induction of COX-2 expression in spinal cord neurons and in other regions of the central nervous system following peripheral inflammation (Samad et al 2001). In the same study, IL-1b was demonstrated to be the major inducer of COX-2 upregulation in the central nervous system. Accordingly, intraspinal administration of an interleukin-converting enzyme or COX-2 inhibitor was accompanied by decreases in both inflammationinduced central prostaglandin E2 levels and mechanical hyperalgesia (Samad et al 2001). In contrast to the acidic antipyretic acids and the selective COX-2 inhibitors, the analgesic and antipyretic mode of action of acetaminophen has been a matter of debate for more than three decades now. Due to its non-acidic chemistry, the drug reaches higher concentrations within the central nervous system as compared to the acidic antipyretic analgesics (e.g. aspirin, indomethacin) that accumulate in peripheral compartments with acidic extracellular pH (e.g. inflamed tissue) (Brune and Lanz 1985). Evidence supporting a central analgesic mode of action has already previously been demonstrated by the finding showing that acetaminophen may inhibit nociception-induced spinal prostaglandin synthesis (Muth-Selbach et al 1999). More recently, the concentration of hydroperoxide has been proposed to confer the cellular selectivity of acetaminophen in inhibiting COX-2 activity (Boutaud et al 2002). Recently, acetaminophen has been shown to inhibit a newly discovered COX isoform, derived from the same gene as COX-1 and referred to as COX-3 (Chandrasekharan et al 2002). However, until now COX-3 has been detected at relatively small amounts in the cerebral cortex of dogs and humans. Importantly, the enzyme possesses glycosylation-dependent COX
CLINICAL SIGNIFICANCE IN PAIN AND INFLAMMATION activity. However, studies indicating a pharmacologically different regulation of COX-3 as compared to the other two isoenzymes have been performed using canine COX-3 and murine COX-1 and -2 enzymes. To clarify the impact of acetaminophen on COX-3dependent prostaglandin production, further experiments have to be performed using the human COX isoenzymes. In line with the dilemma of interpreting IC50 values derived from different assays and performed on enzymes of different origin, acetaminophen, for instance, has been shown to inhibit COX-2 in the human whole blood assay (Cryer and Feldman 1998) at concentrations below those inhibiting COX-3 in the study by Chandrasekharan et al (2002). Moreover, prostaglandins derived from COX-2 but not from COX-1 have been proposed to be involved in the maintenance of central analgesic states (Smith et al 1998). Postmortem analysis of human tissue will probably help to characterize the specific pattern of COX-3 expression under both physiological and pathophysiological conditions. Considering the development and launch of selective COX-3 inhibitors for the treatment of pain and fever, more intensive research and clarification is needed. ROLE OF COX-2 IN INFLAMMATORY NEURODEGENERATIVE DISORDERS A connection between the COX pathway and Alzheimer’s disease has been based mostly on epidemiological studies. In the recently published Baltimore Longitudinal Study of Aging (Stewart et al 1997) with 1686 participants, the risk of developing Alzheimer’s disease was significantly reduced among users of NSAIDs, particularly when NSAIDs were taken for 2 years or more. The apparent protective effect of NSAIDs suggests that COX might be involved in neurodegenerative mechanisms. A role for COX-2 in this process has been established by several lines of evidence (for review, see Pasinetti 2001). In Alzheimer’s disease, COX-2 is upregulated in brain areas related to memory (hippocampus, cortex), with the amount of COX-2 correlating with the deposition of ß-amyloid protein in the neuritic plaques. ßAmyloid is thought to be elaborated as part of an inflammatory process in which activated microglia, the predominant source of COX-2-dependent prostanoids, participate. Elevation of COX-2 expression in hippocampal neurons during the early phase (mild dementia) of Alzheimer’s disease dementia is considered to favour the later inflammatory neurodegenerative process. Moreover, emerging evidence suggests that COX-2-derived prostanoids potentiate glutamate excitotoxicity, thereby accelerating neurodegeneration. Accordingly, primary neuron cultures derived from transgenic mice with neuronal overexpression of human COX-2 are more susceptible to excitotoxic and synthetic aggregated ßamyloid-mediated neuronal death (Pasinetti 2001). However, the precise role of COX-2 in Alzheimer’s disease remains to be clarified in further mechanistic studies. Ongoing studies, e.g. large National Institutes of Health (NIH)-supported trials, under way are evaluating whether selective COX-2 inhibitors may control the destructive progression of Alzheimer’s disease. ROLE OF COX-2 IN CANCER Compelling evidence suggests that COX-2 plays a crucial role in carcinogenesis. The capacity of NSAIDs, such as aspirin and sulindac, to reduce colorectal cancer mortality was already reported in epidemiological studies during the late 1980s. Recent investigations indicate that specific COX-2 inhibitors possess a strong chemopreventive action against colon carcinogenesis in rats, inhibiting tumours to a greater degree than conventional NSAIDs (Kawamori et al 1998). With regard to the action of
323
COX-2, Tsujii et al (1998) found that COX-2-derived prostaglandins may modulate the production of angiogenic factors by colon cancer cells, thereby inducing newly formed blood vessels that sustain tumour cell viability and growth. Moreover, overexpression of COX-2 in epithelial cells has been shown to result in resistance to apoptosis, which in turn leads to dysregulation of growth and normal cell death (Tsujii et al 1998). However, a possible role of COX-1 in colorectal cancer was also emphasized in the same study that showed that COX-1 activity in endothelial cells plays an important role in the modulation of angiogenesis (Tsujii et al 1998). Prostaglandins produced by COX-1 in endothelial cells could be important in regulating genes required for endothelial tube formation and may be a relevant target for cancer prevention or treatment in tumours lacking COX-2 expression. As such, NSAIDs may inhibit angiogenesis by inhibition of COX-2 activity in colon carcinoma cells and downregulating production of angiogenic factors, by induction of apoptosis and by inhibiting COX-1 activity in endothelial cells. Recent studies have further indicated that COX-2 overexpression is not necessarily unique to cancer of the colon, but may be a common feature of other epithelial cells. Increased COX-2 levels have been identified in lung, breast, gastric and prostate cancer, as well as in pancreatic adenocarcinomas (for review, see Prescott and Fitzpatrick 2000). On the basis of these data, it is conceivable that specific COX-2 inhibitors might be used as adjuvants in the treatment of tumours, as well as in cancer prevention. SELECTIVE COX-2 INHIBITORS IN THERAPEUTIC USE Celecoxib and rofecoxib have been shown to be effective antiinflammatory and analgesic substances in patients with rheumatoid arthritis and osteoarthritis. Celecoxib was approved in December 1998 by the US Food and Drug Administration for relief of the signs and symptoms of osteoarthritis (recommended oral dose is 200 mg/day, administered as a single dose or as 100 mg twice daily) and rheumatoid arthritis in adults (recommended oral dose is 100–200 mg twice daily). Rofecoxib became available in 1999 and is indicated for relief of the signs and symptoms of osteoarthritis (recommended starting dose is 12.5 mg/day; maximum recommended daily dose is 25 mg) and rheumatoid arthritis (recommended daily dose is 25 mg), for the management of acute pain in adults, and for the treatment of primary dysmenorrhoea (recommended initial doses are 50 mg once daily; use of rofecoxib at this dose for more than 5 days in the management of pain has not been studied). Both of these specific COX-2 inhibitors have been demonstrated to possess analgesic potency comparable to that of traditional NSAIDs (Ehrich et al 1999a,b; Lefkowith 1999). Celecoxib and rofecoxib were shown to cause a significantly lower incidence of upper gastrointestinal adverse effects (perforations, ulcers and bleeds) than conventional NSAIDs. In the Vioxx Gastrointestinal Outcomes Research (VIGOR) study (Bombardier et al 2000), treatment with rofecoxib at twice the approved maximal dose for long-term use resulted in significantly lower rates of clinically important upper gastrointestinal events and complicated upper gastrointestinal events than did treatment with a standard dose of naproxen. Moreover, the incidence of complicated upper gastrointestinal bleeding and bleeding from beyond the duodenum was significantly lower among patients who received rofecoxib. In the Celecoxib Longterm Arthritis Safety Study (CLASS) (Silverstein et al 2000), incidences of symptomatic ulcers and/or ulcer complications were not significantly different in patients taking celecoxib vs. NSAIDs who were also taking a concomitant low-dose of aspirin, indicating that the use of low-dose aspirin may abrogate the gastrointestinal-sparing effects of celecoxib. By contrast, analysis of
324
INFLAMMATION
non-aspirin users alone demonstrated that celecoxib, at a dosage two- to four-fold greater than the maximum therapeutic dosages, was associated with a significantly lower incidence of symptomatic ulcers and/or ulcer complications compared with NSAIDs. However, in recent years evidence has increased that suggests that a constitutively expressed COX-2 may play a role in physiological renal functions. In the human kidney, COX-2 immunoreactivity was observed in the renal vasculature, medullary interstitial cells and the macula densa, whereas COX-1 was detected in the collecting ducts, thin loops of Henle and portions of the renal vasculature (Nantel et al 1999b). The involvement of COX-2 in human renal functions was also emphasized by clinical studies (McAdam et al 1999; Ehrich et al 1999b; Catella-Lawson et al 1999; Rossat et al 1999; Whelton et al 2000; Brater et al 2001) that showed that specific COX-2 inhibitors, similar to other NSAIDs, may cause peripheral oedema, hypertension and exacerbation of pre-existing hypertension by inhibiting water and salt excretion by the kidneys. Moreover, in healthy elderly volunteers, specific COX-2 inhibitors decreased renal prostacyclin production and led to a significant transient decline in urinary sodium excretion (McAdam et al 1999; Catella-Lawson et al 1999). However, although decreases in sodium excretion were comparable between NSAIDs and specific COX-2 inhibitors, only NSAIDs were shown to reduce the glomerular filtration rate in volunteers with normal renal function (Catella-Lawson et al 1999). Collectively, the data with both celecoxib and rofecoxib are consistent with the expectation that specific COX-2 inhibitors do not spare the kidney. In conclusion, it seems plausible to use specific COX-2 inhibitors with caution in patients with fluid retention, hypertension and heart failure. In the context that selective COX-2 inhibitors diminish systemic prostacyclin production, it is interesting to note that specific COX-2 inhibitors that do not inhibit platelet COX-1 might (at least in theory) unfavourably alter the thromboxane–prostacyclin balance by inhibiting COX-2-dependent synthesis of vasoprotective prostacyclin in endothelial cells However, hitherto published clinical studies have yielded discrepant results in this regard. In the CLASS trial, no difference was noted in the incidence of cardiovascular events (cerebrovascular accident, myocardial infarction, angina) between celecoxib and NSAIDs (ibuprofen, diclofenac) (Silverstein et al 2000). On the other hand, in the VIGOR study, patients receiving rofecoxib had a significant four-fold increase in the incidence of myocardial infarctions, as compared with patients randomized to naproxen (Bombardier et al 2000). However, as both compounds are known to cause a similar inhibition of systemic prostacyclin production without altering platelet-derived thromboxane synthesis, the apparent discrepancy of these studies in terms of cardiovascular outcome is most likely due to differences in the study protocols (e.g. eligibility criteria, study population, study duration) and the use of different NSAID comparators (Fitzgerald et al 2000). Accordingly, 22% of the patients included in the CLASS trial took aspirin as a cardioprotective agent, whereas the entry criteria for the VIGOR study precluded aspirin consumption. In addition, the VIGOR study was performed on patients with rheumatoid arthritis, a condition that has been associated with an enhanced rate of cardiovascular events. By contrast, in the CLASS trial patients with osteoarthritis were included that have not been associated with an increased risk of cardiovascular complications. As a consequence, a possible thrombogenicity of specific COX-2 inhibitors deserves further well-controlled studies. Accordingly, the use of specific COX-2 inhibitors rather than traditional NSAIDs should be preferred in patients at increased risk of serious upper gastrointestinal complications. These patients include individuals older than 60 years, those with a
history of peptic ulcer disease and those taking glucocorticoids (with a high-dose NSAID) and anticoagulants. Collectively, new insights into the biological functions of COX2 caution against the uncritical use of COX-2 inhibitors. In view of the numerous physiological functions of COX-2, further close monitoring of the effects of specific COX-2 inhibitors is necessary to ensure their safety. In particular, well-controlled studies are needed to define the clinical utility of specific COX-2 inhibitions in patients at risk of renal disease, hypertension, cardiovascular diseases or chronic heart failure. Similarly, possible effects of specific COX-2 inhibitiors on reproductive functions, endothelial function and wound healing need to be evaluated in forthcoming clinical trials. On the other hand, the involvement of COX-2 in various pathological conditions implies additional clinical indications for specific COX-2 inhibitors. The utility of specific COX-2 inhibitors in colorectal cancer, colonic polyposis and Alzheimer’s disease is being investigated in ongoing studies. REFERENCES Akopian AN, Sivilotti L and Wood JN (1996) A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature, 379 257–262. Appleby SB, Ristimaki A, Neilson K et al (1994) Structure of the human cyclooxygenase-2 gene. Biochem J, 302: 23–727. Baba H, Kohno T, Moore KA and Woolf CJ (2001) Direct activation of rat spinal dorsal horn neurons by prostaglandin E2. J Neurosci, 21 1750–1706 Beiche F, Scheuerer S, Brune K et al (1996) Upregulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Lett, 390 165–169. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med, 343 1520– 1528. Boutaud O, Aronoff DM, Richardson JH et al (2002) Determinants of the cellular specificity of acetaminophen as an inhibitor of prostaglandin H2 synthases. Proc Natl Acad Sci USA, 99 7130–7135. Brater DC, Harris C, Redfern JS and Gertz BJ (2001) Renal effects of COX-2-selective inhibitors. Am J Nephrol, 21 1–15. Brune K and Lanz R (1985) Pharmacokinetics of non-steroidal antiinflammatory drugs. In Bonta IL, Bray MA, Parnham MJ (eds), Handbook of Inflammation, vol 5, The Pharmacology of Inflammation. Elsevier, Amsterdam, 413–449. Brune K, Zeilhofer HU and Hinz B (1999) Cyclooxygenase inhibitors: new insights. In Emery P (ed.), Fast Facts—Rheumatology Highlights 1998– 99. Oxford, Health Press, 18–24. Catella-Lawson F, McAdam B, Morrison BW et al (1999) Effects of specific inhibition of cyclooxygenase-2 on sodium balance, hemodynamics, and vasoactive eicosanoids. J Pharmacol Exp Ther, 289, 735–741. Chandrasekharan NV, Dai H, Lamar Turepu Roos K et al (2002) COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure, and expression. Proc Natl Acad Sci USA, 99, 13926–13931. Cryer B and Feldman M (1998) Cyclooxygenase-1 and cyclooxygenase-2 selectivity of widely used nonsteroidal anti-inflammatory drugs. Am J Med, 104, 413–421. Dixon DA, Kaplan CD, McIntyre TM et al (2000) Post-transcriptional control of cyclooxygenase-2 gene expression. The role of the 3’untranslated region. J Biol Chem, 275, 11750–11757. Ehrich EW, Dallob A, De Lepeleire I et al (1999a) Characterization of rofecoxib as a cyclooxygenase-2 isoform inhibitor and demonstration of analgesia in the dental pain model. Clin Pharmacol Ther, 65, 336– 347. Ehrich EW, Schnitzer TJ, McIlwain H et al (1999b) Effect of specific COX-2 inhibition in osteoarthritis of the knee: a 6-week double-blind, placebo-controlled pilot study of rofecoxib. J Rheumatol, 26, 2438– 2447. Fitzgerald GA, Austin S, Egan K et al (2000) Cyclooxygenase products and atherothrombosis. Ann Med, 32(suppl 1), 21–26.
CLINICAL SIGNIFICANCE IN PAIN AND INFLAMMATION Fu JY, Masferrer JL, Seibert K et al (1990) The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J Biol Chem, 265, 16737–16740. Hinz B, Brune K and Pahl A (2000a) Cyclooxygenase-2 expression in lipopolysaccharide-stimulated human monocytes is modulated by cyclic AMP, prostaglandin E2 and non-steroidal antiinflammatory drugs. Biochem Biophys Res Commun, 278, 790–796. Hinz B, Brune K and Pahl A (2000b) Prostaglandin E2 upregulates cyclooxygenase-2 expression in lipopolysaccharide-stimulated RAW 264.7 macrophages. Biochem Biophys Res Commun, 272, 744–748. Kawamori T, Rao CV, Seibert K and Reddy BS (1998) Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res, 58, 409–412. Kraemer SA, Meade EA and DeWitt DL (1992) Prostaglandin endoperoxide synthase gene structure: identification of the transcriptional start site and 5’-flanking regulatory sequences. Arch Biochem Biophys, 293, 391–400. Lee SH, Soyoola E, Chanmugam P et al (1992) Selective expression of mitogen-inducible cyclooxygenase in macrophages stimulated with lipopolysaccharide. J Biol Chem, 267, 25934–25938. Lefkowith JB (1999) Cyclooxygenase-2 specificity and its clinical implications. Am J Med, 106, 43S–50S. Maldve RE, Kim Y, Muga SJ and Fischer SM (2000) Prostaglandin E2 regulation of cyclooxygenase expression in keratinocytes is mediated via cyclic nucleotide-linked prostaglandin receptors. J Lipid Res, 41, 873–881. Minghetti L, Polazzi E, Nicolini A et al (1977) Up-regulation of cyclooxygenase-2 expression in cultured microglia by prostaglandin E2, cyclic AMP and non-steroidal anti-inflammatory drugs. Eur J Neurosci, 9, 934–940. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. Muth-Selbach US, Tegeder I, Brune K and Geisslinger G (1999) Acetaminophen inhibits spinal prostaglandin E2 release after peripheral noxious stimulation. Anesthesiology, 91, 231–239. Nantel F, Denis D, Gordon R et al (1999a) Distribution and regulation of cyclooxygenase-2 in carrageenan-induced inflammation. Br J Pharmacol, 128, 853–859. Nantel F, Meadows E, Denis D et al (1999b) Immunolocalization of cyclooxygenase-2 in the macula densa of human elderly. FEBS Lett, 457, 475–477. Neugebauer V, Geisslinger G, Ru¨menapp P et al (1995) Antinociceptive effects of R(7)- and S(+)-flurbiprofen on rat spinal dorsal horn
325
neurons rendered hyperexcitable by an acute knee joint inflammation. J Pharmacol Exp Ther, 275, 618–628. Niiro H, Otsuka T, Izuhara K et al (1997) Regulation by interleukin-10 and interleukin-4 of cyclooxygenase-2 expression in human neutrophils. Blood, 89, 1621–1628. Onoe Y, Miyaura C, Kaminakayashiki T et al (1996) IL-13 and IL-4 inhibit bone resorption by suppressing cyclooxygenase-2-dependent prostaglandin synthesis in osteoblasts. J Immunol, 156, 758–764. Pasinetti GM (2001) Cyclooxygenase and Alzheimer’s disease: implications for preventive initiatives to slow the progression of clinical dementia. Arch Gerontol Geriatr, 33, 13–28. Prescott SM and Fitzpatrick FA (2000) Cyclooxygenase-2 and carcinogenesis. Biochim Biophys Acta, 1470, M69–M78. Rossat J, Maillard M, Nussberger J et al (1999) Renal effects of selective cyclooxygenase-2 inhibition in normotensive salt-depleted subjects. Clin Pharmacol Ther 66, 76–84. Samad TA, Moore KA, Sapirstein A et al (2001) Interleukin-1b-mediated induction of Cox-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature, 410, 471–475. Schaible HG and Schmidt RF (1988) Time course of mechanosensitivity changes in articular afferents during a developing experimental arthritis. J Neurophysiol, 60, 2180–2195. Silverstein FE, Faich G, Goldstein JL et al (2000) Gastrointestinal toxicity with celecoxib vs. non-steroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. J Am Med Assoc, 284, 1247–1255. Smith CJ, Zhang Y, Koboldt CM et al (1998) Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc Natl Acad Sci USA, 95, 13313–13318. Stewart WF, Kawas C, Corrada M and Metter EJ (1997) Risk of Alzheimer’s disease and duration of NSAID use. Neurology, 48, 626– 632. Tsujii M, Kwano S, Tsuji S et al (1998) Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell, 93, 705–716. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231 232–235. Whelton A, Maurath CJ, Verburg KM and Geis GS (2000) Renal safety and tolerability of celecoxib, a novel cyclooxygenase-2 inhibitor. Am J Ther, 7, 159–175. Xie W, Chipman JG, Robertson DL et al (1991) Expression of a mitogenresponsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA, 88, 2692–2696.
28 Antiinflammatory Steroids Franc¸oise Russo-Marie BIONEXIS, Gif sur Yvette, France
Corticosteroids are potent antiinflammatory agents used in almost all types of inflammatory diseases. Although their first use as antiinflammatory agents was reported in 1949 (Hench et al 1949), the way they exert their potent effect is still a matter of debate. Corticosteroids are hormones formed in the adrenal gland in response to a great number of physiological stimuli. Their endogenous secretion is tightly controlled and regulated. The hypothalamo-pituitary system, CRF (corticotrophin releasing factor) and ACTH (adrenocorticotrophic hormone) control the level of the glucocorticoid hormone. CRF, formed in the hypothalamus, is connected to outer and inner stimuli such as stress, cold, pain, light; it gives a positive signal to the pituitary cells, which subsequently form ACTH. ACTH interacts with adrenocortical cells to induce the synthesis of cortisol from cholesterol. Cortisol is not stored in the cells and is immediately released in the circulation, where it exerts its physiological effects on its target tissues. One normal adult person secretes about 20 mg cortisol per day and the circulating concentration of cortisol varies (4–6 mg/100 ml or 110–440 nM) during the nycthemeral cycle. The latter is regulated by the CRF–ACTH– cortisol loop. The increase in cortisol concentration exerts a feedback regulation on the level of both CRF and ACTH, but the CRF level is also regulated by some external stimuli that are responsible of the circadian cycle for ACTH and cortisol (elevated in the morning and low in the evening). Besides this hormonal control, the level of cortisol is also regulated by the plasma concentration of two proteins, transcortin and albumin. Transcortin has a high affinity for cortisol but a low capacity, whereas albumin has a low affinity but a high capacity for cortisol. Transcortin saturation regulates the concentration of free cortisol, the only active form of the hormone. In normal conditions, cortisol is rapidly metabolized and eliminated from the circulation: its plasma half-life is 90 min. More than 95% of circulating cortisol is bound to plasmatic proteins (Haynes 1990; Goldstein et al 1992). Besides their physiological function as hormones, glucocorticoids are also among the most potent antiinflammatory drugs available at the present time. Their first use as antiinflammatory drugs was reported by Hench et al (1949) and used for the treatment of rheumatoid arthritis. Their first use was so spectacular that within a few years they became widely prescribed antiinflammatory drug. However, their molecular mechanism of action was unknown and, very rapidly, deleterious side-effects appeared, e.g. Cushing’s syndrome, osteoporosis, growth retardation. Although much is now known about the molecular mechanism of action of these hormones, the extreme potency of their antiinflammatory action raises still more questions than answers. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
The aim of this article is to shed some modern light on the antiinflammatory role of corticosteroids, among which their effect on eicosanoid synthesis represents only a part of their antiinflammatory action. GLUCOCORTICOIDS (GCs) ARE HORMONES AND DRUGS ACTING THROUGH THE GLUCOCORTICOID RECEPTOR (GR) One of most fascinating aspects of glucocorticoid action to apprehend is the fact that all the reported effects, i.e. physiological and pharmacological (antiinflammatory), involve the same hormone and the same receptor. The differences observed reside in the time and the localization for the effect to occur. This differential and somewhat versatile behaviour is the consequence of the physiological status of the host and of the presence or absence of diseased tissues. Physiological Effects Glucocorticoids, as mentioned above, are circulating hormones whose levels vary following a circadian cycle, elevated in the morning, the level decreasing in the evening. Glucocorticoids also rise in response to stress, cold, trauma and infection. Although one may survive in their absence, although only in protected conditions, they are necessary for host defence. In physiological conditions, they influence the metabolism of carbohydrate, proteins and lipids, control hydroelectrolytic status, some cardiovascular, renal, cerebral and muscular functions, and also control the number of red cells and some sets of leukocytes (Haynes 1990; Goldstein et al 1992). Antiinflammatory Effects While the physiological effects of glucocorticoids are always present, the antiinflammatory effects of glucocorticoids occur only when the host is in a diseased condition. Glucocorticoids participate first in defence and then in the repair system that will help the host to maintain its homeostasis and integrity. Since their first use as anti-inflammatory drugs, all observations have shown that they have the ability to control all the steps of an acute inflammatory reaction, both the very early events and some of the delayed effects. Indeed, they prevent all the symptoms found in the inflammatory reaction, such as oedema, vasodilation, leukocyte migration, oxidative burst and phagocytosis, and by
328
INFLAMMATION
their positive effect on the synthesis of acute-phase proteins they participate in the repair and resolution of inflammation (Solito and Russo-Marie 1998). If the acute inflammatory reaction can be schematically described as being composed of a cellular (i.e. activation of the inflammatory cells) and a vascular reaction (i.e. activation of enzymatic cascades, involving circulating proteins), it can be seen that glucocorticoids control the cellular component almost completely, whereas they have an indirect effect on the vascular component. Indeed, glucocorticoids control the activation of proinflammatory cells, such as monocytes-macrophages, epithelial cells, fibroblasts: they inhibit all the proinflammatory functions of these cells by controlling the synthesis of numerous proteins. As an example, they inhibit the formation of IL-1, IL-2, IL-5, IL-6 and IL-8, TNF, interferon (IFN)-g, GM–CSF, adhesion molecules (ICAM-1, VCAM-1, E-selectins), chemokines (MIP-1, MCP-1), eotaxin, RANTES, enzymes such as COX-2, cPLA2, iNOS and matrix metallo-proteases. They induce the formation of antiinflammatory proteins, acute phase proteins, annexin A1, IL1-RA and LBP (Russo-Marie 1999). Very recently, it has been shown that glucocorticoids exert acute cardiovascular protective effects by inducing the activation (and not the synthesis) of endothelial NOS, provoking nitric oxide vasorelaxation (HafeziMoghadam et al 2002). Glucocorticoids deactivate any amplification circuitry that occurs during the inflammatory reaction. In parallel, they participate in the resolution of the inflammatory reaction by controlling, together with other cytokines, the production of antiinflammatory molecules such as acute phase proteins. The latter have an important role in the deactivation of the circulating proteins responsible for the vascular part of the inflammatory reaction. Glucocorticoids, by the control they exert on all the mediators of the inflammatory reaction, either by repressing or inducing their formation, possess the most powerful anti-inflammatory properties one may expect from a drug.
It is known that in order to complete all its actions, the hormone glucocorticoid, delivered either as the natural circulating hormone (cortisol) or more bioactive molecules (modified forms of cortisol), needs to interact with the type II glucorticoid receptor: GRII. GRII, once bound to its ligand, changes its conformation, and it is the activated GR (GR*) which is responsible for all the described actions of the hormone (Pepin et al 1992). The question to be addressed now is, how does the GR*, one unique moiety, initiate these finely tuned and targeted responses necessary for the host to maintain its integrity and recover the best homeostasis? GLUCOCORTICOID RECEPTORS ARE MEMBERS OF THE STEROID RECEPTOR FAMILY The GR belongs to the superfamily of nuclear receptors to which also belong receptors for other hormones such as mineralocorticoids, progesterone, oestrogens, androgens, thyroid hormones, vitamin D and retinoic acid. All these receptors share a similar organization: a specific N-terminal domain, a conserved DNA binding domain (DBD) and a C-terminal hormone binding domain. In the absence of ligand, the GR is located in the cytoplasm, bound to proteins, among which are found heat shock proteins, namely HSP 90 and HSP 70 (Mangelsdorf et al 1995). GR* FORMS HOMODIMERS AND LEADS TO TRANSACTIVATION (Figure 28.1) Upon binding to its specific hormone, the GR changes its conformation and translocates to the nucleus. Once in the nucleus, the GR may homodimerize and bind to a specific DNA domain, glucocorticoid responsive elements (GREs), formed by palindromic sequences in the cis-regulatory region of target genes. This homodimerization is necessary for the GR* to participate in
Figure 28.1 Classical mechanism of action of the GR* on protein synthesis: GR* forms homodimers, interacts with the glucocorticoid responsive element and initiates transcription
ANTIINFLAMMATORY STEROIDS the activation complex that will initiate the transcription machinery. Only when present as a homodimer will the GR* initiate the synthesis of mRNA and proteins. The latter will perform most of the physiological reported effects of glucocorticoids. The time-lag necessary for the hormone to bind to its receptor and to complete protein synthesis is generally over 2 h. This mechanism takes into account the transactivation of the genes under the control of glucocorticoids. However, most of the reported antiinflammatory effects involve not transactivation but transrepression of target genes and they require a shorter delay to be observed (Beato et al 1995; Reichardt et al 1998). GR* CAN ALSO LEAD TO TRANSREPRESSION Whereas the mechanism of transactivation is well characterized, transrepression of target genes is much less understood. It had been noticed that most genes that are regulated by the GR* do not contain a classical GRE. To explain this negative regulation, distinct modes of action using different classes of response elements, viz. negative, composite and tethering GREs, have been proposed to account for transrepression. Negative GREs require DNA binding of the GR* as exemplified by the POMC (promelanocortin) gene. At composite elements (as in the proliferin gene), the GR* has to contact DNA together with another transcription factor, whereas at tethering elements, repression is mediated by protein–protein interaction without direct DNA binding of the GR* (Beato et al 1995; Lefstin and Yamamoto 1998, Go¨ttlicher et al 1998). Negative effects: AP-1; NF-kB AND INFLAMMATORY CYTOKINES (Figure 28.2) Of particular relevance are tethering interactions for genes that are regulated by AP-1 and NF-kB. Interstitial collagenase (collagenase types I and III), whose expression is modulated by AP-1, was one of the first genes for which repression by
329
interaction with another transcription factor at a tethering element was demonstrated. The repressive effect is probably mediated by GR* monomers and mutational analysis showed that the ligand-binding domain and the DNA-binding domain (DBD) participate in this mode of action. By introducing mutation points into the GR, it was possible to dissociate transactivation and transrepression. Using a mutational and cellular transfection approach, it could be proposed that all genes regulated by AP-1 and/or NF-kB would use the tethering mechanism. This mechanism could account for a wide number of genes that would be repressed. Under the control of AP-1, TNF, IL-1, IL-2, IL-5, IFN-g, ICAM1, E-selectin, GM-CSF and MMP are downregulated by glucocorticoids (Table 28.1). Involving NF-kB in their promoter sequence, almost all genes upregulated during inflammation were demonstrated to be downregulated by glucocorticoids: TNF-a, IL-1, IL-2, IL-6, IL-8, ICAM-1, V-CAM-1, E-selectin, MIP-1, MCP-1, eotaxin, RANTES, COX-2, cPLA2 and i-NOS (Table 28.2). It was demonstrated that the DBD of the GR* would interact directly with the proteins c-Jun and c-Fos (forming the heterodimer of the AP-1 complex); however, in the case of AP-1 formed by c-Jun homodimers, the direct interaction of the GR would lead to a synergistic activation of the gene (as reported for proliferin) in the composite element model. In the case of transrepression for NF-kB, a direct interaction between the GR* and NF-kB has been observed in vitro, again involving the DBD of the GR* and an interaction with the p65 subunit of the NF-kB complex (Lefstin and Yamamoto 1998; Go¨ttlicher et al 1998). This effect of glucocorticoids on the proinflammatory NF-kB activity was confirmed in mice carrying a DNA-binding-defective GR, confirming the importance of the tethering mechanism of action for the antiinflammatory action of glucocorticoids. These studies bring additional evidence that the proposed induction of IkBa expression is dispensable for this repressive effect (Reichardt et al 1998, 2001; Karin 1998). However, whereas many mechanisms were reported to account for the transrepression of glucocorticoids on the various inflammatory
Figure 28.2 Interaction of GR* (as a monomer) with NF-kB and AP-1, preventing the formation of proinflammatory proteins under the control of NFkB and AP-1
330
INFLAMMATION
Table 28.1 Activators of AP-1 and genes under the control of AP-1 during the inflammatory reaction Activators
Regulated genes
TNF, IL-1 UV, H2O2, stress LPS, growth factors TCR
TNF, IL-1, IL-2, IL-5 IFN-g, ICAM-1 E-Selectin GM-CSF Matrix metalloproteases
Table 28.2 Activators of NF-kB and genes under the control of NF-kB during the inflammatory reaction Activators
Regulated genes
TNF, IL-1 UV, H2O2 LPS Virus
TNFa, IL-1, IL-2 IL-6, IL-8 ICAM-1, V-CAM-1 E-Selectin MIP-1, MCP-1 Eotaxin, RANTES COX-2, cPLA2, i-NOS
genes involving a modulation by AP-1 and NF-kB, it was still difficult to propose a relevant mechanism of action. The versatility of the GR* activities could not rely only on various transrepression models. POSITIVE EFFECTS: IL-6, Stat-3: ACUTE PHASE PROTEINS (Figure 28.3) During the inflammatory reaction, glucocorticoids are not the only molecules involved in restoring host integrity and homeostasis; acute phase proteins (APPs) are also important actors for
repairing the host after injury. APP synthesis is increased during inflammation. APPs are synthesized in the liver but they can also be formed in macrophages, epithelia, brain and bone marrow. APPs are circulating plasma proteins belonging to various families, such as complement, coagulation and fibrinolytic proteins, antiproteases, copper and iron transport proteins. They are able to modulate mostly the vascular part of the inflammatory response. APP plasma levels are augmented 1.5–1000 times. Their nomenclature was initially rather complex, but a recent agreement has been obtained: class I APPs are induced by IL-1 and TNF, and class II APPs are induced by IL-6 and related cytokines (IL-11, LIF, OSM, CNTF). However, it should be noticed that IL-6 and its related cytokines have a synergistic effect with IL-1 to induce class I APPs (Baumann and Gaudie 1994; Steel and Whitehead 1994; Russo-Marie 1999) (see Tables 28.3 and 28.4). Glucocorticoids induce the synthesis of all APPs, by themselves, but they have a major synergistic effect with all cytokines to induce the synthesis of APP. Besides the well-known APPs, such as C reactive protein, serum amyloid A, a-glycoprotein, C3, haptoglobin, haemopexin, fibrinogens a and b, a-macroglobulin, a-antitrypsin, antichymotrypsin, ceruloplasmin, new APPs have been described that are regulated as APP. Included in this new group are IL-1RA, annexin A1, LBP, sPLA2, procalcitonin, angiogenin, metallothionein, contrapsins and PAP-I. All these proteins participate in host defence and repair. The common denominator of these APPs is to be under the positive control of glucocorticoids and IL-6. A direct interaction of the GR* with the IL-6-induced transcription factors, Stat-3 and C/EBP, has been reported (Russo-Marie 1999). Although the inhibitor of NF-kBa (IkBa) does not belong to the APPs, some in vivo experiments indicate that the inhibition of NF-kB in lymphoid cells by glucocorticoids involves IkBa gene transcription. This effect is due to the GR* but there is no GRE present in the promoter of the IkBa gene and GR*–DNA interaction is not required (Karin 1998).
Figure 28.3 Synergistic effects of GR* with IL-6 induced transcription factors: Sat-3 and C/EBP will increase the synthesis of APPs
ANTIINFLAMMATORY STEROIDS Table 28.3 Acute phase proteins Type 1 APP
Type 2 APP
C-reactive protein Serum amyloid A Glycoprotein acid-a C3 Haptoglobin (rat) Haemopexin (rat)
Fibrinogen-a, -b and -g Macroglobulin-a Antitrypsin-a Antichymotrypsin Haptoglobin (man) Haemopexin Ceruloplasmin
Table 28.4 Regulation of acute phase protein production Effectors
Response
IL-1, TNF IL-6, IL-11, LIF, OSM, CNTF
APP type 1 All PPA but major activator of type 2 APP synergize with IL-1 of type 1 APP Minor stimulation per se major synergistic effect with cytokines on all APP
Glucocorticoids
OTHER TRANSCRIPTION FACTORS: UNKNOWN MECHANISM OF ACTION The interaction of GR* with quite a large number of nuclear factors has been reported. The importance of these interactions on the inflammatory pathway has not always been addressed but they may play a yet unknown role in the control of the inflammatory reaction. Apart from AP-1 and NF-kB, synergism with the GR* has been demonstrated for the following DNA-binding proteins: NF-1, CCAAT box-binding factor, CAACC box-binding factor, SP-1, Oct-1/OTF-1, CREB, HNF-3, C-EBP and Stat-5. It has also been shown that the GR* cooperates positively with AP-1 if it is formed of c-Jun homodimers or the heterodimer c-Jun/Fra-1. How these interactions interfere with the formation of inflammatory mediators remains to be elucidated (McEwan et al 1997).
PROPOSED MECHANISM OF ACTION: HISTONE ACETYLATION; CO-REPRESSION AND/OR CO-STIMULATION To increase the complexity of the role of the GR*, one must take into account the fact that all nuclear receptors have been postulated to regulate gene expression via their association with histone acetylase (HAT) or deacetylase complexes. These interactions will modify the accessibility of DNA to the transcriptional machinery. Therefore, one may predict that GR* dimerization or heterodimerization with other transcription factors may modify the formation of large complexes that will induce or repress histone acetylation, leading consequently to gene activation and/or repression. However, this model would require the interaction of GR* dimers or heterodimers of GR* with other transcription factors with DNA (not necessarily at the GRE) and this has not yet been demonstrated (Maldonado and Hampsey 1999; Strahl and Allis 2000; Chen et al 1999; Freedman 1999).
INTERACTIONS OF GR* WITH THE CYTOPLASMIC SIGNALLING MACHINERY Even more intriguing has been the observation that some protective effects of glucocorticoids are mediated by nontranscriptional activation of endothelial nitric oxide synthase.
331
This effect links GR* with PI3K (PI3 kinase) (Hafezi-Moghadam et al 2002). Other recent observations have linked GR* with the MAP-kinase pathway for the secretion of annexin A1 (Solito et al 2002). These effects may explain some very early responses of glucocorticoids during the inflammatory reaction. CONCLUSION Glucocorticoids exert all their effects after they have interacted with their specific receptor, the GR (glucocorticoid receptor), forming an activated GR (GR*). The GR* is responsible for all the effects of the glucocorticoids, i.e. to permit host defence. How the GR* promotes all these effects is still not clear. Indeed GR*, by interacting with quite a large number of different proteins in either the cytoplasm or the nucleus, induces all the different effects reported. However, it is still not clear how one hormone and one protein can permit such a finely tuned and versatile response. The proposed explanations of combinatorial mechanisms involving large protein complexes are not satisfactory. One must think in a simpler way that would reconcile host defence and glucocorticoid effects as a simple response induced by a common denominator. REFERENCES Baumann H and Gaudie J (1994) The acute phase response. Immunol Today, 15, 74–80. Beato M, Herrlich P and Schu¨tz G (1995) Steroid hormone receptors: many actors in search of a plot. Cell, 49, 729–739. Chen H, Lin RJ, Xie W et al. (1999) Regulation of hormone-induced histone hyperacetylation and gene activation via acetylation of an acetylase. Cell, 98, 675–686. Freedman LP (1999) Increasing the complexity of coactivation in nuclear receptor signalling. Cell, 97, 5–8. Goldstein RA, Bowen DL and Fauci AS (1992) Adrenal corticosteroids. In Gallin JI, Goldstein IM and Snyderman R (eds), Inflammation: Basic Principles and Clinical Correlates. New York, Raven Press, 1061– 1081. Go¨ttlicher M, Keck S and Herrlich P (1998) Transcriptional cross-talk, the second mode of steroid hormone receptor action. J Mol Med, 76, 480– 489. Hafezi-Moghadam A, Simoncini T, Yang Z et al (2002) Acute cardiovascular protective effects of corticosteroids are mediated by a non-transcriptional activation of endothelial nitric oxide synthase. Nature Med, 8, 473–479. Haynes RC Jr (1990). Adrenocorticotropic hormone; adrenocortical steroids and their synthetic analogues; inhibitors of the synthesis and action of adrenocortical hormones. In Gilman AG, Rall TW, Nies AS, Taylor P (eds), Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 8th edn. New York, Pergamon Press, 1431–1462. Hench PS, Kendall EC, Slocumb CH and Polley HF (1949) The effect of a hormone of the adrenal cortex (17-hydroxy-11-dehydrocorticosterone; compound E) and of pituitary adrenocorticotropic hormone on rheumatoid arthritis. Proc Staff Meet Mayo Clin, 24, 181–197. Karin M (1998) New twists in gene regulation by glucocorticoid receptor: is DNA binding dispensable? Cell, 93, 487–490. Lefstin JA and Yamamoto KR (1998) Allosteric effects of DNA on transcriptional regulators. Nature, 392, 885–888. Maldonado E and Hampsey M (1999) Repression, targeting the heart of the matter. Cell, 99, 455–458. McEwan IJ, Wright APH and Gustafsson JA (1997) Mechanism of gene expression by the glucocorticoid receptor: role of protein-protein interactions. BioEssays, 19, 153–160. Mangelsdorf DJ, Thummel C, Beato M et al (1995) The nuclear receptor superfamily : the second decade. Cell, 83, 835–839. Pepin MC, Pothier F and Barden N (1992) Impaired type II glucocorticoid-receptor function in mice bearing antisense RNA transgene. Nature, 355, 725–728. Reichardt HM, Kaestner KH, Tukermann J et al (1998) DNA binding of the glucocorticoid receptor is not essential for survival. Cell 93, 531–541.
332
INFLAMMATION
Reichardt HM, Tuckermann JP, Gottlicher M et al. (2001) Repression of inflammatory response in the absence of DANN binding to the glucocorticoid receptor. EMBO J, 20, 7168–7173. Russo-Marie F (1999) Glucocorticoı¨ des et proteines de la phase aigue¨. Soc Biol, 193, 375–380. Solito E and Russo-Marie F (1998) Anti-inflammatoires ste´roı¨ diens. In Russo-Marie F, Polla B and Peltier A (eds), Inflammation. Paris, John Libbey Eurotext, 540–550.
Solito E, Mulla A, Morris JF et al (2003) Dexamethasone induces rapid serine phosphorylation and membrane translocation of annexin 1 in a human folliculer-stellate cell line via a novel non genomic mechanism involving the glucocorticoid receptor, PKC, PI3 kinase and MAP kinase. Endocrinology, 144, 1164–1174. Steel DA and Whitehead AS (1994) The major acute phase reactants. Immunol Today, 15, 81–88. Strahl BD and Allis CD (2000) The language of covalent histone modifications. Nature, 403, 41–45.
29 Eicosanoids and Algesia in Inflammation Lynne Murray, Henry Sarau and Kristen E. Belmonte 1Imperial
College School of Medicine, London, UK; and 2GlaxoSmithKline, King of Prussia, PA, USA
As early as the second century AD there is evidence that the Roman encyclopaedist Aulus Celus studied and described the qualities of inflammation as rubor et calor cum tumor et dolor (redness and heat combined with swelling and pain). These sensations result from a series of complex interactions integrated at a variety of levels, including the level of the periphery, spinal cord and cerebrum. Pain, or nociception, is often due to an overproduction of the mediators involved in sensory transmission. This chapter will describe the basic mechanisms and structures involved in nociception. In addition, recent evidence elucidating the role that prostaglandins play in enhancing and intensifying nociceptive pathways will be described. Finally, pharmacological interventions that modify the action of eicosanoids and thereby cause analgesia will be discussed. Pain can be classified into three general types, based on the source of the injury, as well as the fibres, mediators and mechanisms involved. Acute or physiological pain can be initiated by thermal, chemical or mechanical stimuli. This type of pain serves mainly as a warning system, alerting the recipient of potentially dangerous contact or stimuli. As such, it is characterized by having a high threshold, and by being highly localized to the area of injury. This type of pain is generally transient, and subsides as the injury heals. Neuropathic pain is a pathological condition that results from damage to the spinal cord or peripheral nerves. In this type of pain, normally innocuous stimuli can generate a pain sensation known as allodynia. Inflammatory pain results from acute or chronic tissue insult. Tissue injury results in the local release of a mixture of inflammatory mediators, including cytokines, chemokines, neuropeptides and eicosanoids. These inflammatory mediators are chemoattractants for various leukocyte populations that infiltrate the area and produce additional mediators, thus potentiating nociception. The overall effect of the interaction of cells and mediators is increased vascular permeability, resulting in oedema and redness. Inflammatory pain also involves pain hypersensitivity. These inflammatory mediators modify the function of nociceptive fibres and receptors, sensitizing them and leaving them susceptible to activation at much lower thresholds. Hypersensitivity to pain is a consequence of early post-translational changes in the peripheral terminals of nociceptors and in the dorsal horn neurons, and later transcription-dependent changes in effector genes in primary sensory and dorsal horn neurons (Woolf and Costigan 1999). This phenomenon is known as hyperalgesia and is evident from as early as a few hours after the inflammatory insult. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
NOCICEPTION: PHYSIOLOGICAL PROCESS Nociception is the detection of noxious or tissue-damaging stimuli. Whether detected topically at the level of the skin or deep within the viscera, the pain sensation results from the stimulation of a circuit of neurons that originate within the tissues and relay through the central nervous system. This circuit begins with activation of nociceptive receptors at the source of insult, which then transmit the signal through the primary afferent neurons and synapses within specific laminae in the dorsal horn of the spinal cord. There, sensory information is relayed to the thalamus and brainstem. There are two main types of ascending pathways—monosynaptic, those that project directly into the higher structures within the cerebrum; and polysynaptic, those that ascend through a relay of second-order neurons before reaching the higher structures. The relay of information occurs via complex networks of sensory fibres, specialized to conduct the nociceptive message. TRANSMISSION OF THE NOCICEPTIVE STIMULUS Primary afferent neurons can be loosely classified into three types, based on their diameter, structure and conduction velocity (see Table 29.1). Ab fibres are large (410 mm) fibres that are myelinated, and thus have a fast conduction velocity (430– 100 m/s); they have a low threshold for stimulation and act as mechanoreceptors, relaying information to the spinal cord. Ad fibres are also myelinated, although they are not as large (2–6 mm), and thus have an intermediate conduction velocity (12–30 m/s). The C fibres are the smallest of the three types (0.4–1.2 mm); they are unmyelinated, and thus have the slowest conduction velocity (0.5–2 m/s). While each of these fibre types is able to transmit sensory information, they display differential sensitivity to innocuous as well as noxious stimuli. All three types of fibres can transmit innocuous stimuli, but under normal physiological conditions only the C and Ad fibres transmit noxious messages. Stimulation of Ad fibres leads to a pricking pain sensation, whereas stimulation of C fibres results in a burning or dull pain Table 29.1 Sensory nerve fibres involved with nociception Fibre
Myelinated?
Size
Conduction velocity
Ab Ad C
Yes Yes No
410 mm 2–6 mm 0.4–1.2 mm
430–100 m/s 12–30 m/s 0.5–2 m/s
334
INFLAMMATION
sensation. At the cellular level, various receptors and ion channels, collectively known as nociceptors, carry out the transmission of pain. NOCICEPTORS Until very recently, there was no specific structure that could be identified histologically within the tissue as a nociceptive receptor. Instead, the idea that free unmyelinated sensory nerve endings, embedded in the skin, joints, muscles, and within the viscera, served as nociceptive receptors was, and to a degree still remains, the convention. Despite the fact that no distinct structure exists to receive or initiate the sensation, there are differential specializations that allow for the detection of a normal, low-intensity stimulus or an intense noxious one. Non-painful stimuli, such as simple touch or movement, are sensed with a high degree of specificity by nociceptive receptors that display the unique ability to amplify weak signals, followed by a rapid adaptation to intensification. This allows for more acute sensory perception, and has a built-in mechanism for preventing non-painful signal duplication. It is this adaptive mechanism which prevents, for example, the simple act of putting on a shirt or standing under a running shower from becoming a painful or traumatic experience. The case is entirely different when painful stimuli are involved; here, the primary objective is not so much sensory as it is defensive. As such, these high threshold receptors are considered to be polymodal, as they respond equally to mechanical, thermal or chemical stimuli. Furthermore, in contrast to the rapidly adapting receptors that function to detect non-painful sensations, when these receptors are repeatedly activated they are susceptible to a phenomenon known as sensitization, whereby their threshold is further decreased, resulting in pain perception following stimuli that would have previously been thought to be inconsequential. An example of sensitization would be the increased sensitivity that is perceived to the touch following a burn—even a somewhat slight burn—to the skin. The concept of, and mechanisms involved in, pain sensitization, with particular focus on the role that eicosanoids play in the process, will be explored in much greater detail later in this chapter. Recently, the discovery of a host of novel ion channel-coupled receptors expressed on sensory neurons has helped to elucidate many of the more advanced mechanisms involved in nociception. These receptors include the vanilloid or capsaicin receptors, as well as purinergic and proton-activated acid-sensing receptors. In addition, a deeper understanding of sodium and calcium channels—those ion channels once thought to be involved solely in sensor nerve fibre conductance—have now brought them to the fore as key players in nociception. An example of each of these types of novel receptors is discussed below. The vanilloid receptors have recently been identified as specific thermosensitive receptors. Known as VR-1, this receptor type is activated by capsaicin, the active ingredient in hot peppers. VR-1 is a ligand-gated, non-selective cation channel that is expressed primarily on small diameter primary afferent neurons, and recently it has been suggested that significant levels of VR-1 may be present in the brain. Apart from capsaicin, VR-1 has been shown to have activity in response to anandamide and the lipoxygenase product, 12-(S)-HPETE (Hwang et al 2000). While clearly VR-1 is responsive to a host of ligands, it also responds to thermal stimuli (threshold at approximately 43 8C) and to protons, suggesting that its activity may be enhanced in the acidic environment that is present in inflamed tissues. The sodium channels found on neurons can be classified based on their sensitivity to tetrodotoxin (TTX), and these sodium channels are differentially expressed. Large diameter fibres only express sodium channels that are TTX-sensitive. Small diameter
fibres, however, express the TTX-resistant sodium channels along with those that are TTX-sensitive. Another major difference between TTX-resistant and TTX-sensitive channels is the current the channels support. TTX-resistant channels propagate slow currents, whereas TTX-sensitive channels are involved in fast current transmission. There are two types of TTX-resistant sodium channels that have been cloned, SNS/PN3 and SNS2/ NaN, and these types of channels have been shown to be sensory neuron-specific. The SNS/PN3 channel population, in particular, has been demonstrated in significant numbers within the dorsal root ganglia and in association with other nociceptors, and expression of this channel is upregulated in states of chronic pain (Cummins et al 2000). A recent report has shown that these receptors have the potential of being upregulated especially during inflammatory situations (Okuse et al 2002). One unique property that has been ascribed to certain C fibres is the concept of sleeping or silent nociceptors. This term describes a particular type of C fibre that is essentially inactive under normal physiological conditions. However, under conditions of inflammation and tissue injury, these silent nociceptors become acutely chemosensitive (Rivera et al 2000). It is the activation of these silent nociceptors that creates the prolonged pain that accompanies extended inflammatory states. CHEMICAL MEDIATORS INVOLVED IN NOCICEPTION An additional level of specialization with respect to nociception occurs at the level of synaptic transmission within the spinal cord. Ad and C fibres release substance P and glutamate as their primary neurotransmitters. Innocuous stimuli are generally mediated by the excitatory amino acid, glutamate, acting on the Na+ channel-coupled a-amino-3-hydroxy-5-methylisoxazole (AMPA) receptors. AMPA receptors are localized to the dorsal horn of the spinal cord. They have low voltage dependence, rapid kinetics, and are rapidly desensitized following stimulation. As such, their affinity for glutamate is low. Repetitive stimulus, or stimuli resulting from tissue injury, results in the activation of a second glutamate receptor found on nociceptive neurons in the dorsal horn, the N-methyl-D-aspartate (NMDA) receptor, which is a Ca2+ channel that, under resting conditions, is quiescent. This is possible because the interior of the calcium channel features a negatively charged site that creates an interaction with a Mg2+ ion. The placement of this Mg2+ serves to block the channel’s permeability to Ca2+ and this interaction with the channel is voltage-dependent. Depolarization of the neuron causes the release of the Mg2+ blockade, resulting in a marked permeability to Ca2+. A massive influx of Ca2+ causes a depolarization of the post-synaptic terminal, thus resulting in neuronal transmission. Severe pain stimuli which activate C fibres result in the release of neuropeptides, such as tachykinins (substance P, NKA and NKB) and CGRP. The tachykinins are pronociceptive, and each acts at its own receptor (NK1, NK2 and NK3, respectively), while CGRP binds to its own receptors, CGRP1 and CGRP2. During intense stimuli, NK1 and NMDA receptors play a key role in the initiation and maintenance of the sensitized state. Activation of NK1 receptors in the dorsal horn results in the activation of protein kinase C (PKC), which phosphorylates NMDA receptors. Once phosphorylated, the Mg2+ blockade is lost, rendering NMDA receptors susceptible to activation at a much lower activation potential. This type of sensitization that takes place within the spinal cord is known as central sensitization. Local inducible nitric oxide synthase (iNOS) enzyme activity results in the production of nitric oxide (NO). NO can act on the presynaptic neuron, causing further release of substance P and glutamate, leading to increased neuronal transmission. Substance
EICOSANOIDS AND ALGESIA IN INFLAMMATION P is constitutively expressed in approximately 45% of all Ad and C fibres, however, after inflammation, Ab fibres have been shown to express substance P, switching the fibre to a typical Ad phenotype, thus increasing sensitization and potentiating nociception (Neumann et al 1996). Leukotriene B4, another lipid mediator produced from PLA2, interacts with C fibres and results in hyperalgesia. Although the precise mechanism is unknown, it is thought that LTB4 is able to decrease the thermal and mechanical thresholds for C fibres (Martin et al 1988; Martin 1990). One area that has been thoroughly investigated is the significant role that prostaglandins play in nociception, as prostaglandins are considered to be the key eicosanoids involved in pain mechanisms. EXPRESSION AND LOCALIZATION OF PROSTAGLANDINS, PROSTAGLANDIN RECEPTORS AND CYCLOOXYGENASES IMPLICATE PROSTANOIDS IN NOCICEPTION One of the first factors that must be taken into consideration when determining the role for prostaglandins in nociception is localization of prostaglandin receptors with respect to the local production of the prostaglandins, as well as the expression of the prostanoid synthesizing enzymes, COX-1 and COX-2. Prostaglandins have been shown to be produced in physiologically relevant levels in both the periphery and the central nervous system (CNS), with PGD2 being the most abundant prostaglandin in the brain (Hiroshima et al 1986). In the spinal cord, production of PGD2, PGE2, PGF2 and PGI2 has been demonstrated, but PGE2 and PGI2 have been the subject of the most study; thus, these two prostanoids appear as the key players for nociceptive mechanisms in the spinal cord. While there appears to be a minimal degree of basal prostaglandin release, tissue insult or injury is the primary mechanism by which prostanoid production is induced within the spinal cord and in the periphery. DP, EP, FP and IP receptors can all be found in the CNS and in the periphery. EP and IP receptors are identified as the major players in nociceptive mechanisms. The EP receptors are localized to the brain, spinal cord and periphery, and each of the subtypes is selectively expressed in a different area of the periphery and the CNS. In the spinal cord all four EP subtypes can be found, and since PGE2 is capable of signalling through four EP receptor subtypes, the consequences of receptor binding are thus dependent upon the differential expression of each receptor subtype. For example, in the spinal cord, EP1 receptors and mRNA can be found on about one-third of all neurons in the DRG, and can be found co-expressed with IP receptors (Oida et al 1995). EP2 mRNA is localized to neurons in laminae IV–VI within the dorsal horn of the spinal cord (Kawamura et al 1997). EP3 receptors and mRNA are found on over half of the neurons and in the dorsal root ganglia, but this expression is limited to neurons and is not found on glia (Sugimoto et al 1994; Oida et al 1995). EP3 expression is primarily seen in laminae I and II and somewhat in lamina V (Beiche et al 1998). Signalling through EP1 is associated with the hyperalgesic response; EP2 and EP4 mediate relaxation, whereas EP3 modulates neurostransmitter release from nerve fibres. Overall, the precise mechanisms that regulate the differential expression of EP receptors have yet to be elucidated. IP receptors and mRNA are also found on nearly half of the neurons in the DRG and are often found co-expressed with EP1 and, to a lesser extent, EP3 receptors (Oida et al 1995). DP and FP receptors have only been demonstrated in the spinal cord and dorsal root ganglia, and these receptors are not thought to play a significant role in nociception. Within the brain and spinal cord, constitutive expression of both COX-1 and COX-2 can be detected. Expression of both the
335
COX proteins and mRNAs has been described, and they have been localized to the neurons themselves, as well as to the associated glial cells. In the dorsal root ganglion, COX-1 protein and mRNA have been found; however, while there are reports of localization of COX-2 mRNA, no group has reported the presence of measurable COX-2 protein (Willingale et al 1997). In contrast, significant COX-1 and COX-2 mRNA and protein expression have been identified in the neurons of all laminae of the spinal cord (Beiche et al 1996; Resnick et al 1998). Prostaglandins Play a Significant Role in Nociception The non-steroidal antiinflammatory drug (NSAID) aspirin has been used to relieve pain ever since its discovery in the late 1890s, when its anti-nociceptive effects were first described (Wright 1993). It was not until 1971 that Vane demonstrated that the primary mechanism of action of aspirin and related drugs is through the inhibition of synthesis of prostaglandins from arachidonic acid (Vane 1971). It was these data that provided the first evidence for a significant role for prostaglandins in pain. Today there is a large body of evidence supporting the idea that prostaglandins are important mediators of pain. In addition to the observation that inhibitors of the synthetic enzyme cyclooxygenase provide effective pain relief, inhibition of nociception was also demonstrated by the administration of a neutralizing anti-PGE2 monoclonal antibody to mice exposed to phenylbenzoquinone (Mnich et al 1995). The level of inhibition is reported as similar to that seen with the non-selective cyclooxygenase inhibitor, indomethacin. Since phenylbenzoquinone is known to induce elevated prostaglandin levels at the site of application, these data indicate that PGE2 is largely involved with the nociceptive response. The reverse situation is also true—i.e. local application or injection of prostaglandins induces hyperalgesic responses (Ferreira 1972). PGF2, PGE2 and PGI2 exert the strongest effects, indicating the involvement of EP and/or IP receptors in inflammatory pain. More specifically, PGE2 was identified as the most potent hyperalgesic (Sciberras et al 1987) and thus became recognized as the primary mediator involved in pain (Bley et al 1998). Another indication that prostaglandins are involved in nociception in the CNS is the hyperalgesic response observed following intrathecal administration of PGE2 (Minami et al 1992, 1994). Intrathecal PGE2 could also induce allodynia, suggesting a dual role of this prostanoid. Overall, hyperalgesia is associated with an increase in PGD2 and PGE2, whereas allodynia is linked to increases in PGF2a and PGE2. However, a significant role for the IP receptors in nociception can also be demonstrated, since decreased peripheral inflammation and nociception was noted in PGI2 knockout mice (Murata et al 1997). Furthering the potential of PGI2 in nociception is the fact that IP receptors have been found on sensory neurons (Narumiya et al 1999). These data are summarized in Figure 29.1. Recently isoprostanes, a novel class of eicosanoids that are stereoisomers of prostaglandins, have been demonstrated to produce nociception in rat sensory neurons (Evans et al 2000). Isoprostanes have been shown to be produced at micromolar levels at sites of injury, especially in response to oxidative stress (Morrow and Roberts 2002). Interestingly, the synthesis of isoprostanes is largely cyclooxygenase-independent. Stimulation of C fibres by the direct application of capsaicin results in a rise in prostaglandin synthesis by increased COX-2 gene expression in the dorsal horn (Hay and de Belleroche 1997; Hua et al 1997; Willingale et al 1997). Since administration of COX-inhibiting drugs to the spinal region dampens sensory nerve reflexes and thus attenuates the pain sensation, it has been concluded that elevated levels of prostaglandins play a key role in pronociception (Beiche et al 1996; Bianchi and Panerai 1996;
336
INFLAMMATION
Figure 29.1 Summary of expression of key prostanoid receptors in the CNS and in the periphery. Differential expression of receptor subtypes allows for a multiplicity of roles for the same prostaglandin
Inflammation is a very complex response involving the coordinated action of many different cell types and mediators. Various animal models of inflammatory pain have begun to allow for a dissection of the mechanisms underlying pain to be evaluated at both the receptor and the mediator level. During inflammation, PGE2 is released from multiple cell types in the surrounding tissue, as opposed to being released from nerve endings. Other instrumental mediators produced in response to an inflammatory insult are the cytokines, with IL-1b and TNFa being produced locally and rapidly. These cytokines may not have a direct effect on nociception; however, IL-1b has been shown to enhance directly the release of substance P from neurons within the spinal cord, thus potentiating pain transmission. In addition, TNFa sensitizes neurons to the potent pain-producing agent, bradykinin, by increasing receptor expression. Prostaglandins have been suggested to participate in the inflammatory response itself by synergizing with these and other local inflammatory mediators, such as bradykinin and histamine. The result is an enhancement in the vasodilatory response, vascular permeability and oedema, along with the enhanced nociceptive response at the site of injury.
or injury, release of mediators such as IL-1b, NO and TNFa from injured cells causes increased local expression of COX-2, leading to increased local production of prostaglandins (Samad et al 2001). Several groups have shown that peripheral administration of complete Freund’s adjuvant induces an upregulation of COX-2 mRNA in the CNS. This phenomenon was associated with a concomitant increase in PGE2 in the cerebrospinal fluid, implicating prostaglandins in pain transmission (Hay and de Belleroche 1997; Samad et al 20001). NO is released from a variety of cell types during the inflammatory response, including leukocytes, endothelial cells and nerve fibres, and its presence is significant because it also increases the formation of proinflammatory prostaglandins (Rettori et al 1992; Sautebin and Di Rosa 1994). These prostaglandins are then available to participate in and enhance the nociceptive response. In addition to COX-2 gene induction by inflammatory mediators, it has been shown that synaptic activity induces expression of COX-2, suggesting that the increased neural activity that accompanies pain and inflammation may result in COX-2 induction at sites near, as well as beyond, the immediate site of tissue insult (Yamagata et al 1993). Furthermore, the sensory neurotransmitter, substance P, released by sensory nerves at sites of tissue inflammation and injury can cause the degranulation of mast cells directly. These cells then release additional inflammatory mediators (Lieb et al 1996), resulting in the recruitment of specific leukocyte subsets (Hood et al 2000), which in turn potentiate the inflammatory response and cause an increase in production and local accumulation of PGE2. It is of note that none of the mechanisms described has been shown to result in induction of the COX-1 gene (Vane et al 1998).
INDUCTION OF COX-2 AND INCREASED PROSTAGLANDIN SYNTHESIS DURING INFLAMMATION
ROLE OF PROSTAGLANDINS IN SENSITIZING NERVE ENDINGS AND PERIPHERAL NERVE FIBRES
COX-1 and COX-2 are membrane-associated enzymes that are the products of distinct genes located on distinct chromosomes (Kraemer et al 1992). COX-1 is expressed in most tissues apart from the CNS in a constitutive manner, while COX-2 is only expressed at physiologically relevant levels, when its gene expression is induced by a variety of inflammatory cytokines and/or growth factors (Vane et al 1998). Following inflammation
Early evidence indicating that prostaglandins may be involved in sensitization of sensory nerves and nociception was demonstrated using a human skin blister base model. Low concentrations of PGE2 or PGI2 could not elicit an acutely painful sensation when added alone; however, when added in conjunction with other mediators, such as bradykinin or histamine, both agents could greatly potentiate the pain sensation (Hornton 1963; Ferreira
Willingale et al 1997). While the mechanisms whereby prostaglandins facilitate nociception in the dorsal horn are unclear, it is thought that prostaglandins may have a sensitizing effect on sensory neurons. This will be discussed in greater detail later in this chapter.
THE ROLE OF PROSTAGLANDINS IN INFLAMMATORY PAIN
EICOSANOIDS AND ALGESIA IN INFLAMMATION 1972). It was postulated that the presence of prostaglandins might have a sensitizing effect on the local neurons. Careful investigation has allowed for the elucidation of the cellular messengers involved in the sensitization process mediated by prostaglandins. Sensory nerve sensitization by prostaglandins is first initiated by the activation of PKA. This is thought to occur through activation of EP and IP receptors found on these nerves, each of which are coupled to adenylate cyclase (Ferreira and Nakamura 1979). This results in an increase in cAMP (and thus calcium release) and PKA levels in the neurons conducting the signal. Elevated basal levels of intracellular messengers during resting conditions result in a reduction in the amount of stimulation required to depolarize the neuron, therefore lowering the threshold for activation. The result is a potentiation of neuropeptide release from these sensitized neurons. An additional mechanism whereby sensory nerves can become sensitized by prostaglandins is through activation of PKC and PLC. It is thought that activation of IP receptors may result in activation of PLC. While this effect may be subtle, albeit significant, it is thought that this pathway may synergize with the PKA-activating pathway (Sugita et al 1997) (see Figure 29.2). Prostaglandins, particularly PGI2 and PGE2, sensitize neurons to thermal, mechanical and chemical stimuli, such as bradykinin, which are found at inflammatory sites (Handwerker and Neher 1976; Chahl and Iggo 1977; Martin et al 1988; Taiwo and Levine
337
1988; Schepelmann et al 1992; Khasar et al 1995; Gold et al 1996; Wang et al 1996; Damas et al 1997). By lowering the activation threshold for so many different types of receptors, the chances for propagation of action potentials are greatly increased, making this mediator extremely efficient at sensitization. At the neuronal level, PGE2 also increased the number of substance P-containing peripheral afferent fibres, which again will respond to bradykinin (Stucky et al 1996). There is comprehensive evidence to suggest a reciprocal sensitization with bradykinin whereby polymodal C and Ad fibres become sensitized to prostaglandins. This sensitization may result in the further release of substance P and glutamate, which can then act back on the neuron, thus increasing pain transmission. Bradykinin itself may also increase prostaglandin production through activation of phospholipase A2 from phospholipids (Prado et al 1997). PROSTAGLANDINS INVOLVED IN THE MAINTENANCE OF PAIN As well as playing an instrumental role in the initial sensation of pain and also in pain transmission, prostaglandins, particularly PGE2, have been shown to be involved in the maintenance of pain. By administering anti-PGE2 monoclonal antibodies after inflammation has been established, the hyperalgesic effect could
Figure 29.2 Sensitization of sensory nerve endings by prostaglandins produced at sites of tissue insult. Activation of PKA and PKC results in a synergistic effect on ion channels, lowering the threshold for action potential, resulting in increased neurotransmitter release
338
INFLAMMATION
be abolished, suggesting that this prostaglandin is produced in pathophysiologically significant levels for some time following an inflammatory insult. These data suggest a therapeutic role for neutralizing this prostaglandin during inflammation (Zhang et al 1997).
Inhibition of Cyclooxygenase Activity Results in Analgesia NSAIDs such as aspirin, indomethacin, salicylic acid and ibuprofen have become valuable therapeutic agents for the treatment of a variety of painful conditions, such as rheumatoid arthritis and osteoarthritis. It is widely accepted that the nonselective inhibition of cyclooxygenase, resulting in decreased prostaglandin production, is the primary mechanism—although not the only mechanism—by which NSAIDs are able to produce analgesia. NSAIDs are able to block cyclooxygenase activity by blocking the ability of arachidonic acid to bind to the enzyme’s active site through a variety of mechanisms, e.g. aspirin irreversibly acetylates a key conserved serine residue in the cyclooxygenase enzyme active site, whereas ibuprofen competes with arachidonic acid pharmacologically for the cyclooxygenase active site (Vane et al 1998). While cyclooxygenase blockade is considered to be the primary mechanism of NSAID action, it is thought that NSAIDs may also have a stabilizing effect on excitable membranes or cellular signalling mechanisms in inflammatory states, e.g. salicylic acid has the ability to hyperpolarize excitable cells by interacting physiochemically with the cell membrane, thus rendering it less permeable to chloride ions and more permeable to potassium ions (Levitan and Barker 1972). Salicylates also have an inhibitory effect on activation of the transcription factor nuclear factor-kB (NF-kB; Grilli et al 1996). This is significant because NF-kB is activated during inflammation, and it is responsible for regulating the transcription of a number of inflammatory cytokines, chemokines and mediators. Thus, NSAIDs can have an inhibitory effect on prostaglandin production, and thus the pain sensation, in a more indirect fashion by reducing neuronal excitability and generally reducing the inflammatory response. Since the elucidation of two independent isoforms of the cyclooxygenase enzymes, and the growing understanding of the role of COX-2 in nociception, a variety of pharmaceutical companies generated a host of COX-2-selective antiinflammatory drugs, such as celecoxib and rofecoxib. These COX-2 selective NSAIDs are able to achieve COX-2 selectivity based on a key amino acid difference between COX-1 and COX-2 (isoleucine 523 D valine 523); this allows for access to a ‘‘side pocket’’ near the COX-2-active site that is not available near the COX-1-active site. Thus, COX-2-selective inhibitors are quite bulky, and so do not fit in the COX-1-active region (Kurumbail et al 1996; Luong et al 1996), resulting in selectivity over COX-1. The COX-2-selective NSAIDs have been shown to be just as effective, and in some cases, more effective, as non-selective NSAIDs at inhibiting pain in a variety of models, but COX-2-selective inhibitors do not have the gastrointestinal side-effects that are seen with acidic classical non-selective cyclooxygenase inhibitors. COX-2 inhibitors have now become first line therapy for rheumatoid arthritis and osteoarthritis, as well as for back and tooth pain.
REFERENCES Beiche F, Scheuerer S et al (1996) Upregulation of cyclooxygenase-2 mRNA in the rat spinal cord following peripheral inflammation. FEBS Lett, 390(2), 165–169.
Beiche F, Klein T, Nusing R, Neuhuber W, Goppelt-Struebe M (1998) Localization of cyclooxygenase-2 and prostaglandin E2 receptor EP3 in the rat lumbar spinal cord. J Neuroimmunol, 89, 26–34. Bianchi M and Panerai AE (1996) The dose-related effects of paracetamol on hyperalgesia and nociception in the rat. Br J Pharmacol, 117(1), 130–132. Bley KR, Hunter JC (1998) The role of IP prostanoid receptors in inflammatory pain. Trends Pharmacol Sci, 19(4), 141–147. Chahl LA and Iggo A (1977) The effects of bradykinin and prostaglandin E1 on rat cutaneous afferent nerve activity. Br J Pharmacol, 59(2), 343–347. Cummins TR, Dib-Haji SD et al (2000) Sodium channels and the molecular pathophysiology of pain. Prog Brain Res, 129, 3–19. Damas J, Liegeois JF et al (1997) Involvement of 5-hydroxytryptamine and bradykinin in the hyperalgesia induced in rats by collagenase from Clostridium histolyticum. Naunyn Schmiedebergs Arch Pharmacol, 355(5), 566–570. Evans AR, Junger H et al (2000) Isoprostanes, novel eicosanoids that produce nociception and sensitize rat sensory neurons. J Pharmacol Exp Ther, 293(3), 919–920. Ferreira SH (1972) Prostaglandins, aspirin-like drugs and analgesia. Nature New Biol, 240(102), 200–203. Ferreira SH and Nakamura M (1979) I-Prostaglandin hyperalgesia, a cAMP/Ca2+-dependent process. Prostaglandins, 18, 1779–190. Gold MS, Reichling DB et al (1996) Hyperalgesic agents increase a tetrodotoxin-resistant Na+ current in nociceptors. Proc Natl Acad Sci USA, 93(3), 1108–1112. Grilli M, Pizzi M, Memo M and Spano P (1996) Neuroprotection by aspirin and sodium salicylate through blockade of NF-kB activation. Science, 247, 1383–1385. Handwerker HO and Neher KD (1976) Characteristics of C-fibre receptors in the cat’s foot responding to stepwise increase of skin temperature to noxious levels. Pflugers Arch, 365(2–3), 221–229. Hay C and de Belleroche J (1997) Carrageenan-induced hyperalgesia is associated with increased cyclooxygenase-2 expression in spinal cord. Neuroreport, 8(5), 1249–1251. Hiroshima O, Hayashi H et al (1986) Basal level of prostaglandin D2 in rat brain by a solid-phase enzyme immunoassay. Prostaglandins, 32(1), 63–80. Hood VC, Cruwys SC et al (2000) Differential role of neurokinin receptors in human lymphocyte and monocyte chemotaxis. Regul Pept, 96(1–2), 17–21. Horton EW (1963) Action of prostaglandin E1 on tissues which respond to bradykinin. Nature, 200, 892–895. Hua XY, Calcutt NA et al (1997) Neonatal capsaicin treatment abolishes formalin-induced spinal PGE2 release. Neuroreport, 8(9– 10), 2325–2329. Hwang SW, Cho H et al (2000) Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci USA, 97(11), 6155–6160. Kawamura T, Yamauchi T, Koyama M et al (1997) Expression of prostaglandin EP2 receptor mRNA in the rat spinal cord. Life Sci, 61, 601–624. Khasar SG, Miao FJ et al (1995) Inflammation modulates the contribution of receptor-subtypes to bradykinin-induced hyperalgesia in the rat. Neuroscience, 69(2), 685–690. Kraemer SA, Meade EA and DeWit DL (1992) Prostaglandin endoperoxide synthase gene structure: identification of the transcriptional start site and 5’ flanking regulatory sequences. Arch Biochem Biophys, 293, 391–400. Kurumbail RG, Stevens AM, Gierse JK et al (1996) Structural basis for selective inhibition of cyclooxygenase-2 by antiinflammatory agents. Nature (Lond), 384, 644–648. Levitan H and Barker JL (1972) Effect of non-narcotic analgesics on membrane permeability of molluscan neurons. Nature New Biol, 239, 55–57. Lieb K, Fiebich BL et al (1996) Effects of substance P and selected other neuropeptides on the synthesis of interleukin-1b and interleukin-6 in human monocytes: a re-examination. J Neuroimmunol, 67(2), 77–81. Luong C, Miller A, Barnett J et al (1996) Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nature Struct Biol, 3, 927–933.
EICOSANOIDS AND ALGESIA IN INFLAMMATION Martin HA, Basbaum AI et al (1988) Leukotriene B4 decreases the mechanical and thermal thresholds of C-fibre nociceptors in the hairy skin of the rat. J Neurophysiol, 60(2), 438–445. Martin HA (1990) Leukotriene B4 induced decrease in mechanical and thermal thresholds of C-fibre mechanonociceptors in rat hairy skin. Brain Res, 509(2)273–279. Minomi T, Uda R, Horiguchi S et al (1992) Allodynia evoked by intrathecal administration of PGF2a to mice. Pain, 50, 223–229. Mnich SJ, Veenhuizen AW (1995) Characterization of a monoclonal antibody that neutralizes the activity of prostaglandin E2. J Immunol, 155(9), 4437–4444. Morrow JD and Roberts LJ II (2002) Mass spectrometric quantification of F2-isoprostanes as indicators of oxidant stress. Methods Mol Biol, 186, 57–66. Murata T, Ushikubi F (1997) Altered pain perception and inflammatory response in mice lacking prostacyclin receptor. Nature, 388(6643), 678–682. Narumiya S, Sugimoto Y (1999) Prostanoid receptors: structures, properties, and functions. Physiol Rev, 79(4), 1193–1226. Neumann S, Doubell TP et al (1996) Inflammatory pain hypersensitivity mediated by phenotypic switch in myelinated primary sensory neurons. Nature, 384(6607), 360–364. Oida H, Namba T, Sugimoto Y et al (1995) In situ hybridization studies of prostacyclin receptor mRNA expression in various mouse organs. Br J Pharmacol, 116, 2828–2837. Okuse K, Malik-Hall M et al (2002) Annexin II light chain regulates sensory neuron-specific sodium channel expression. Nature, 417(6889), 653–656. Prado GN, Taylor L et al (1997) Effects of intracellular tyrosine residue mutation and carboxyl terminus truncation on signal transduction and internalization of the rat bradykinin B2 receptor. J Biol Chem, 272(23), 14638–14642. Resnick DK, Graham SH, Dixon CE et al (1998) Role of cyclooxygenase2 in acute spinal cord injury. J Neurotrauma, 15(12), 1005–1013. Rettori V, Gimeno M et al (1992) Nitric oxide mediates norepinephrineinduced prostaglandin E2 release from the hypothalamus. Proc Natl Acad Sci USA, 89(23), 11543–11546. Rivera L, Gallar J, Pozo MA and Belmonte C (2000) Responses of nerve fibres of the rat saphenous nerve neuroma to mechanical and chemical stimulation: an in vitro study. J Physiol, 527(pt 2), 305–313. Samad TA, Moore KA et al (2001) Interleukin-1b-mediated induction of Cox-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature, 410(6827), 471–476. Sautebin L and Di Rosa M (1994) Nitric oxide modulates prostacyclin biosynthesis in the lung of endotoxin-treated rats. Eur J Pharmacol, 262(1–2), 193–196.
339
Schepelmann K, Messlinger K et al (1992) Inflammatory mediators and nociception in the joint: excitation and sensitization of slowly conducting afferent fibers of cat’s knee by prostaglandin I2. Neuroscience, 50(1), 237–247. Sciberras DG, Goldenberg MM et al (1987) Inflammatory responses to intradermal injection of platelet activating factor, histamine and prostaglandin E2 in healthy volunteers: a double-blind investigation. Br J Clin Pharmacol, 24(6), 753–761. Stucky CL, Thayer SA et al (1996) Prostaglandin E2 increases the proportion of neonatal rat dorsal root ganglion neurons that respond to bradykinin. Neuroscience, 74(4), 1111–1123. Sugimoto Y, Shigemoto R, Namba T et al (1994) Distribution of the messenger RNA for the prostaglandin E receptor subtype EP3 in the mouse nervous system. Neuroscience, 62, 919–928. Sugita S, Baxter DA and Byrne JH (1997) Modulation of a cAMP/protein kinase A cascade by protein kinase C in sensory neurons of Aplysia. J Neurosci, 17, 7237–7244. Taiwo YO and Levine JD (1988) Characterization of the arachidonic acid metabolites mediating bradykinin and noradrenaline hyperalgesia. Brain Res, 458(2), 402–406. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231(25), 232–235. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenases 1 and 2. Annu Rev Pharmacol Toxicol, 38, 97–120. Wang JF, Khasar SG et al (1996) Sensitization of C-fibres by prostaglandin E2 in the rat is inhibited by guanosine 5’-O-(2thiodiphosphate), 2’,3’-dideoxyadenosine and Walsh inhibitor peptide. Neuroscience, 71(1), 259–263. Willingale HL, Gardiner NJ et al (1997) Prostanoids synthesized by cyclooxygenase isoforms in rat spinal cord and their contribution to the development of neuronal hyperexcitability. Br J Pharmacol, 122(8), 1593–604. Woolf CJ and Costigan M (1999) Transcriptional and posttranslational plasticity and the generation of inflammatory pain. Proc Natl Acad Sci USA, 96(14), 7723–7730. Wright V (1993) Historical overview of NSAIDs. Eur J Rheumatol Inflamm, 13(1), 4–6. Yamagata K, Andreasson KI, Kaufmann WE, et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Zhang Y, Shatter A et al (1997) Inhibition of cyclooxygenase-2 rapidly reverses inflammatory hyperalgesia and prostaglandin E2 production. J Pharmacol Exp Ther, 283(3), 1069–1075.
30 Cyclooxygenase-2 in Cancer Ovidiu C. Trifan and Jaime L. Masferrer Pharmacia Corp., Chesterfield, MO, USA
Cancer occurs as a multi-step, multi-factorial process that frequently takes many years to unfold. Environmental conditions, mutations in key genes and genetic predisposition have a critical impact on whether individuals exposed to particular carcinogens develop cancer, as well as the duration of exposure required for malignant transformation to occur. Arachidonic acid is metabolized by three major enzymatic pathways: the cyclooxygenase, lipoxygenase and cytochrome P450-dependent pathways. Several eicosanoids derived from these pathways have been suggested to play a role in the development of human neoplasia (Jaffe 1974; Steele 2000). The focus of this review will be on the role of cyclooxygenase-2 and the derived prostaglandins (PGs) in cancer. There are two cyclooxygenase (COX) isoenzymes, COX-1 and COX-2. In general, COX-1 is constitutively expressed in the majority of tissues, playing a housekeeping role. Most normal tissues barely constitutively express COX-2, but this enzyme is rapidly induced by inflammatory cytokines, tumour promoters, growth factors and oncogenes (Prescott and Fitzpatrick 2000). Non-steroidal antiinflammatory drugs (NSAIDs), such as aspirin, have been used for more than a century for the treatment of inflammation-related symptoms. Their therapeutic properties derive mainly from inhibiting COX enzymes’ ability to produce PGs (Vane 1971; Smith et al 1994). Some of the first evidence suggesting that PGs may play a role in human cancers was derived from epidemiological data. Several studies observed a 40–50% decrease in relative risk for colorectal cancer in persons who regularly use aspirin and other NSAIDs (Smalley and DuBois 1997). Clinical studies with NSAIDs in patients with familial adenomatous polyposis (FAP) demonstrated that NSAID treatment caused regression of pre-existing adenomas (Giardiello et al 1995), suggesting that the COX enzymes are important in polyp formation. Similarly, studies using animal models of colon cancer have also indicated a significant tumour reduction after NSAID treatment (Levy 1997; Sheng et al 1997). Moreover, azoxymethane (AOM)-induced colon tumorigenesis was also inhibited by treatment with NSAIDs (Reddy et al 1987). Studies aimed at finding the molecular explanation for the effects of NSAIDs in cancer revealed that COX-2 is overexpressed in colorectal and other forms of epithelial cancers (Koki et al 1999). Furthermore, early studies using the COX-2-specific inhibitor celecoxib showed a 93% reduction in the development of adenomas and adenocarcinomas in the AOM rat model (Reddy et al 1996). The COX-2 expression data, together with epidemiological and pharmacology data, supported the hypothesis that COX-2 plays a role in cancer growth and progression. Many other studies appear to confirm this view. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
For example, in vitro experiments have shown that rat intestinal epithelial cells that overexpress COX-2 develop increased adhesion to extracellular matrix and resist undergoing butyrateinduced apoptosis (Tsujii and DuBois 1995). This can be reversed by treatment with NSAIDs, suggesting that overexpression of COX-2 is probably involved in the development and progression of colonic neoplasms. In human prostate cancer LNCaP cells that overexpress COX-2, an induction of apoptosis and downregulation of bcl-2 gene expression was observed when cells were treated with the COX-2-specific inhibitor NS-398 (Liu et al 1998). Moreover, ras-induced transformation of C57/MG cells resulted in increased levels of COX-2 mRNA and protein and increased production of PGE2 (Subbaramaiah et al 1996). Genetic evidence supporting a role for COX-2 in the development of intestinal cancer has also been reported. Breeding ApcD716 mice (which develop hundreds of intestinal polyps) with COX-2 null mice results in ApcD716/COX-2(7/7) mice with a significantly reduced number of polyps when compared to the offspring of the ApcD716/COX-2-wild-type mice (Oshima et al 1996), suggesting that COX-2 is required for polyp formation. Using a transgenic system, Liu et al (2001) have shown that selective overexpression of the human COX-2 gene in the mouse mammary glands driven by the murine mammary tumour virus promoter results in precocious glandular differentiation and delayed mammary tissue involution. Treatment with the non-selective COX inhibitor indomethacin inhibited this phenotype and caused a significant reduction in PGE2 synthesis. Chronic expression of COX-2 in the breast tissue led to a greatly exaggerated incidence of focal mammary gland hyperplasia, dysplasia and transformation into metastatic tumours. One important conclusion of this study is that COX-2 overexpression is sufficient to trigger tumorigenic transformations in mouse mammary gland alone or in combination with naturally occurring risk factors. The results from these two independent experimental approaches using COX2 deletion or overexpression clearly indicate that COX-2 is very important during the oncogenic process. Furthermore, pharmacological evidence for the role of COX-2 in tumorigenesis was obtained using COX-2 specific inhibitors, e.g. a COX-2-selective inhibitor, SC-58125, decreased cell growth in both in vitro and in vivo assays only in cells expressing COX-2 (Sheng et al 1997, 1998). Moreover, numerous groups have shown a marked reduction in tumour growth after treatment with COX-2-specific inhibitors (Sheng et al 1997; Masferrer et al 2000). Summarizing, the most notable observations are: (a) epidemiological data with NSAIDs show a reduction in risk of developing cancer; (b) COX-2 and PGs are increased in numerous human
342
INFLAMMATION
epithelial cancers; (c) genetic evidence strongly supports a role for COX-2 in tumorigenesis; (d) COX-2-specific inhibitors have potent anti-tumour effects in several animal models. Various theories have been proposed to explain the exact mechanism(s) by which high levels of COX-2 foster cancer growth. The most obvious consequence of COX-2 overexpression is increased PG production. In this chapter we will discuss the contribution of COX-2 and COX-2-derived prostanoids on proliferation, apoptosis, modulation of immune response, angiogenesis, invasion and metastasis. ROLE OF COX-2 AND PGS IN PROLIFERATION AND APOPTOSIS Increased PG synthesis can lead to direct stimulation of cell growth, e.g. PGE2 can induce proliferation of rat hepatocytes (Hashimoto et al 1997; Kimura et al 2001) and it can also stimulate the proliferation of mammary epithelial cells in serumfree primary cultures in the presence of epidermal growth factor (EGF) (Bandyopadhyay et al 1987). In breast tissue, PGE2 may also indirectly increase cell proliferation by increasing expression and activity of aromatase (Zhao et al 1996; Harris et al 1999), leading to increased oestrogen synthesis. Thus, it is possible that PG-mediated oestrogen production may be an important organ site-specific consequence of COX-2 upregulation in breast cancer. PGE2 is also required for the growth and proliferation of human keratinocytes (Pentland and Needleman 1986). This proliferation is dependent on PG production, since it can be inhibited by NSAIDs and restored by the addition of PGE2. In rat seminal vesicles, intracellular PGE2 has been shown to be involved in the mitogenic effects of oestradiol and testosterone (Lyson 1984; McKanna et al 1998). Similarly, PGE2 and PGF2a can induce a mitogenic response in Balb/c 3T3 fibroblasts in synergy with EGF (Nolan et al 1988). PGF2a is also mitogenic for MC3T3-E1 osteoblasts (Quarles et al 1993) and Swiss 3T3 cells (Goin et al 1993). Topical application of 12-O-tetradecanylphorbol-13-acetate (a cancer-promoting agent) on mouse skin induces considerable PG synthesis at the site of administration. The epidermal hyperproliferation is inhibited by indomethacin and this inhibition can be reversed by the topical application of PGE2 (Fu¨rstenberger and Marks 1979; Verma et al 1980). It is interesting to note that some of these experimental results suggest that PGs are not sufficient to induce cell proliferation, but rather act as permissive factors allowing the mitogenic action of various growth factors. However, uncontrolled stimulation of cellular proliferation by PGs may contribute to tumorigenesis. In contrast to stimulating growth on a variety of cells, PGE2 can inhibit blastogenesis of T cells (Marnett 1992). This effect, together with inhibition of the cytotoxic activity of natural killer cells (Marnett 1992), may contribute to the immune suppression associated with increased PG synthesis (see below). A prostaglandin-independent effect of COX-2 ectopic expression on cell growth has also been reported (Trifan et al 1999). Cox-2 overexpression induced G0/G1 arrest in a variety of cell types, including ECV-304, BMEC, COS-7, NIH 3T3 and HEK 293. Pretreatment with indomethacin and NS-398 (a COX-2 inhibitor) had no effect on the cell cycle arrest caused by the COX2 transfection. However, such treatment resulted in virtually complete inhibition of prostanoid secretion. The dissociation of cell cycle arrest from prostanoid synthesis was also demonstrated utilizing two active site mutants of COX-2 (mimicking aspirin acetylation), S516Q and S516M (Lecomte et al 1994), that induced cell-cycle arrest. Such an observation might be explained through the bifunctional character of the COX-2 enzyme. The oxygenase activity of COX-2 oxidizes AA to PGG2, while the peroxidase activity transforms PGG2 in PGH2. The peroxidase
component of COX-2 can oxidize a wide range of chemical carcinogens (e.g. heterocyclic and aromatic amines and polycyclic aromatic hydrocarbons) (Smith et al 1991; Marnett 1992). Additionally, the peroxidase activity of COX-2 may regulate the redox-sensitive transcription factor NF-kB, as reported for COX1 by Munroe et al (1995). Another possible explanation is a novel uncharacterized protein–protein interaction, e.g. Ballif et al (1996) have identified the autoimmunity- and apoptosis-associated protein, nucleobindin, which specifically interacts with COX isoenzymes in a yeast two-hybrid system. Another mechanism by which COX-2 might contribute to tumorigenesis is by decreasing apoptosis. Defective control of apoptosis is generally viewed as one of the central mechanisms of tumorigenesis because it is thought to lead to carcinogenesis by allowing survival of cells that have acquired mutations. The ability of COX-2 to confer cell resistance to apoptosis was observed in rat intestinal epithelial cells stably transfected with COX-2 (Tsujii and DuBois 1995). Several studies have also established a direct role for PGs in rendering cells resistant to apoptosis, e.g. in human colon cancer cells, PGE2 inhibited programmed cell death caused by the selective COX-2 inhibitor SC-58125 as well as inducing the anti-apoptotic protein Bcl-2 (Sheng et al 1998). In human neutrophilic polymorphonuclear leukocytes, PGE2 inhibited apoptosis, possibly due to its role as a cAMP-promoting agent (Ottonello et al 1998). PGE1 effectively inhibited apoptosis in rat phaeochromocytoma PC12 cells deprived of nerve growth factor (Kawamura et al 1999). Moreover, addition of PGI2, PGD2 and PGE1 to rat hepatocytes cultured in collagen gel and treated with transforming growth factor-b1 or exposed to UV light decreased the frequency of apoptotic nuclei in a dose-dependent manner by up to 70–80% and suppressed DNA fragmentation (Kroll et al 1998). Decreased apoptotic cell death caused by PGs could clearly lead to enhanced tumour growth and may explain at least in part how NSAIDs prevent cancer development. Lately there is a growing interest in using NSAIDs in cancer prevention and treatment, particularly since the development of COX-2 specific inhibitors. A number of researchers found that NSAIDs (including selective COX-2 inhibitors) could induce apoptosis (Hara et al 1997; Sheng et al 1998; Ding et al 2000; Hida et al 2000). Several hypotheses have been advanced to account for these observations. The most obvious explanation would be that COX-2 expression and increased PG synthesis suppress apoptosis; thus, inhibition of COX-2 activity should be sufficient to induce apoptosis. However, in some studies, NSAID-induced apoptosis was independent of COX-2 expression (Hanif et al 1996; Elder et al 1996; Zhang et al 1999; Grosch et al 2001; Richter et al 2001), suggesting that NSAIDs stimulate apoptosis via both COX/PGsdependent and -independent mechanisms (Rigas and Shiff 2000; Richter et al 2001). Possible PG independent mechanisms include inhibition of the protein kinase Akt (Hsu et al 2000) and suppression of NF-kB activation (Grilli et al 1996; Yin et al 1998; Yamamoto et al 1999; Stark et al 2001). Another proposed explanation is that COX inhibition with NSAIDs leads to accumulation of arachidonic acid and stimulation of the conversion of sphingomyelin to ceramide, which then leads to apoptosis (Chan et al 1998). Clearly, the effects on proliferation and apoptosis associated with COX-2 overexpression and increased PGs synthesis play an important role in tumorigenesis. CHRONIC INFLAMMATION AND IMMUNE SUPPRESSION IN CANCER Chronic immune activation and inflammation are recognized as risk factors for carcinogenesis (Weitzman and Gordon 1990).
CYCLOOXYGENASE-2 IN CANCER Chronic activation of the immune system could lead to the development of lymphomas in chronic graft vs. host disease (Habeshaw et al 1992), Kaposi’s sarcoma and lymphomas in AIDS (Dalgleish 1992; Whitby and Boshoff 1998), liver cancer in hepatitis B and C virus infections (Imperial 1999) or stomach cancer in Helicobacter pylori infection (Williams and Pounder 1999). Interestingly, Kim et al (2000) showed that infecting cells with H. pylori alone might play a role in cellular transformation. In Hs746T gastric epithelial cells infected with H. pylori, COX-2 was upregulated at the mRNA and protein level. Inhibition of COX-2 with NS-398 resulted in a significant increase of caspase-3 activation and apoptosis. Moreover, the effect of NS-398 on H. pylori-induced apoptosis was reversed by PGE2 addition, suggesting that the antiapoptotic effect of COX-2 expression is mediated by an increase in production of PGE2. Cancer is also associated with chronic inflammation resulting from exposure to non-infectious agents, e.g. chronic oesophagitis, including Barrett’s oesophagus characterized by high levels of COX-2 (Morris et al 2001), could lead to carcinoma of the oesophago–gastric junction (Jankowski et al 1999; McCann 1999; Morris et al 2001). Similarly, chronic bronchitis due to cigarette smoking is recognized as a major risk factor for the development of lung cancer (Mayne et al 1999). Once the cancer is established, the associated immunosuppression is a common complication that frequently leads to poor prognosis in late-stage cancer patients. Immune suppression may also contribute to tumorigenesis, since this may allow tumours to avoid immune surveillance that might otherwise limit their growth. It has been shown that inhibition of COX activity may be associated with an enhanced immune response (Goodwin 1984) and reduced tumorigenesis. The plasma of certain cancer patients (Williams et al 1968) and some tumour-bearing animals (Siddiqui and Williams 1990) exhibits high levels of prostaglandins. The dominant prostaglandins are PGE2 and PGD2 (Scott et al 1982; Urade et al 1989) and they are mainly produced by tumour cells (Siddiqui and Williams 1990; Kalmar et al 1984) and activated specialized antigen-presenting cells (Leung and Mihich 1980). It has been suggested that PGE2 regulates immune function by acting as a negative feedback inhibitor for various processes, including proliferation of lymphocytes, lymphokine production and macrophage and natural-killer cell cytotoxicity (Cantarow et al 1978; Balch et al 1984; Goodwin 1984). PGE2 also inhibits the production of TNF-a while inducing production of the immunosuppressive interleukin-10 (Kambayashi et al 1995; Huang et al 1996). Similarly, PGD2 has been shown to have antiproliferative and immunosuppressive properties (Rocklin et al 1985; Ito et al 1989). Although these PGs are immunosuppressive per se, another possible explanation for their properties is that they could be transformed in cyclopentenone PGs (containing an a,b-unsaturated carbonyl group in the cyclopentane ring) by exposure to plasma, serum, or solutions containing albumin (Fitzpatrick and Wynalda 1983; Kikawa et al 1984). PGA2 results from PGE2 and PGJ2 results from dehydration of PGD2 and can isomerize to form D12-PGJ2 (Fitzpatrick and Wynalda 1983). Further dehydration of the latter compound leads to 15-deoxy-D12,14-PGJ2. Cyclopentenone prostaglandins could also be the effectors of PGmediated immunosuppression in cancer (Ohno et al 1988; Homem de Bittencourt and Curi 1992), raising the possibility that the immunosuppression of late-stage cancer patients could be due to a variety of PGs. ROLE OF COX-2 AND PGS IN ANGIOGENESIS Tumour growth beyond 2–3 mm in size is dependent on angiogenesis. Early experiments showing a significant reduction
343
in xenograft neovascularization after treatment with NSAIDs (Peterson 1983) suggested the involvement of cyclooxygenases and PGs. Recently it has been hypothesized that COX-2 might play a role in the regulation of angiogenesis associated with solid tumours. Consequently, COX-2-specific inhibitors may block the growth of blood vessels into developing tumours. Indeed, selective COX-2 inhibitors reduced angiogenesis in vivo in several models (Majima et al 1997; Daniel et al 1999; Masferrer et al 1999, 2000; Sawaoka et al 1999; Yamada et al 1999), while COX-1 inhibition with a specific inhibitor SC-560 had no effect on bFGFinduced angiogenesis (Masferrer et al 2000). Tsujii et al (1998) showed in vitro that COX-2 expression in colon cancer cells stimulates angiogenesis of co-cultured endothelial cells by upregulating expression of vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), transforming growth factor-b1, platelet-derived growth factor and endothelin-1. Treatment with NS-398 decreased colorectal cancer cells’ ability to secrete these factors. Moreover, COX-27/7 fibroblasts had a significantly reduced ability to produce VEGF relative to wild-type fibroblasts (Williams et al 2000). The molecular mechanisms underlying COX-2-mediated production of proangiogenic factors remain to be defined. Moreover, the authors of this study showed that Lewis lung carcinoma xenografts grew much slower in COX-2 null mice than similar tumours in wild-type mice. Interestingly, the tumours from COX2 knockout mice exhibited a reduced vascular density, indicating that COX-2 in host tissue is required for tumour angiogenesis. Furthermore, the expression patterns of COX-2 and VEGF in the stromal compartment of the isografts were very similar and VEGF expression was significantly reduced in tumours grown in COX-2-deficient mice. In human colorectal cancer samples, immunohistochemistry and Western blot analysis revealed that COX-2 expression correlates with VEGF expression and microvessel density (Cianchi et al 2001). Similarly, in human samples of head and neck cancers, COX-2 mRNA and protein expression was higher than in normal mucosa and it correlated with microvessel density and VEGF (Gallo et al 2001). These data suggest VEGF as a downstream mediator of COX-2 in angiogenesis, and this is not surprising considering that PGE2 has been shown to induce VEGF in the human monocytic THP-1 cell line (Hoper et al 1997) and in cultured rat Mu¨ller cells (Cheng et al 1998). Recently the published work of Leahy and coworkers (2002) provided further insight into the role of COX-2 in angiogenesis. Rat corneas implanted with basic-fibroblast growth factor (bFGF) pellets contain significantly larger amounts of PGE2 and TXB2. COX-2 was expressed in the newly formed blood vessels (endothelial cells, macrophages, polynuclear cells, keratocytes and other unidentified cells in the angiogenic areas) but not in the vascular cells of mature limbic blood vessels. Treatment with the highly specific COX-2 inhibitor celecoxib decreased the PGE2 and TXB2 levels in the bFGF-containing corneas and inhibited angiogenesis. The reduced angiogenesis could be explained by the inhibition of the proliferation in the COX-2 expressing cells and increased apoptosis in the newly formed blood vessels observed after treatment with celecoxib. A similar reduction in proliferation was observed in tumour cells and the vascular stroma of celecoxib-treated xenograft tumours. One difficult question is whether angiogenesis is affected by COX-2 expression in the tumour cells, endothelial cells or other stromal cells. The data existing so far in the literature suggest that COX-2 in all these cells may contribute to the angiogenic process. An appealing corollary of COX-2 involvement in angiogenesis is that COX-2 inhibition may be useful in the treatment of virtually all types of solid cancers, rather than only malignancies characterized by high levels of COX-2 expression.
344
INFLAMMATION
COX-2 AND PGS IN INVASIVENESS AND METASTASIS Contributing to the progression of solid tumours is the ability of cancerous cells to invade the adjacent tissue and to spread to distant sites. Although the molecular mechanism is not fully understood, COX-2 had been shown to contribute to increased invasiveness. Rat intestinal epithelial cells stably transfected with COX-2 showed increased adhesion to the extracellular matrix proteins (Tsujii and DuBois 1995). Additionally, Caco-2 cells or breast cancer cell line Hs578T stably expressing COX-2 had an increased ability to invade through a layer of Matrigel (Tsujii et al 1997; Takahashi et al 1999). The cell line C3L5 (derived from a C3H/ HeJ spontaneous mammary tumour) was shown to express high levels of COX-2 mRNA and protein, as detected by Northern blotting, Western blotting and immunostaining. In these cells PGE2 production was primarily due to COX-2. The proliferation of C3L5 cells in vitro was not influenced by PGs, yet their migratory and invasive abilities were inhibited with indomethacin, NS-398 or PG-receptor (EP1/EP2) antagonist AH6809 in a dosedependent manner. The indomethacin- and NS-398-mediated inhibition was partially reversed upon addition of exogenous PGE2 (Rozic et al 2001). A PLA2 inhibitor, ibuprofen (a COX non-specific inhibitor) and NS-398 inhibited invasion through Matrigel of DU-145 and PC-3 human prostate cancer cell lines in response to fibroblast conditioned medium. Cell invasive potential was restored by the addition of PGE2. However, PGE2 alone did not stimulate migration through Matrigel, which suggests that PGE2 is a permissive factor, rather than sufficient for inducing invasiveness. Cells treated with ibuprofen or NS398 also showed a significant reduction in the levels of secreted matrix metalloproteinases (MMP) proMMP-2, MMP-9, and proMMP-9 in the culture medium (Attiga et al 2000). A correlation between COX-2, PGE2 and increased metastatic potential was also reported in human cancers. For example, Chen et al (2001) found in humans that COX-2 expression in hepatic metastases correlates with the expression in the primary colorectal tumour. They also found that in vitro invasiveness of colon cancer cell lines correlates with COX-2 expression and PGE2 production. Invasive properties and PGE2 production were inhibited by etodolac at COX-2 selective concentrations. In human colorectal cancer samples, PGE2 levels were significantly higher in metastatic tumours than in non-metastatic ones (Cianchi et al 2001). Similarly, in human head and neck cancers COX-2 expression and PGE2 levels seem to correlate directly with lymph node metastasis (Gallo et al 2001). Human cancers such as breast and prostate cancers are characterized by a high incidence of bone metastasis (Yoneda 1998). One other possible mechanism by which COX-2 expression contributes to bone metastasis could be that PGE2 released by tumour cells induces activation of osteoclasts (Kawaguchi et al 1995), causing bone resorption that will allow development of the metastatic tumour. These data support that COX-2 and the derived PG may contribute in multiple ways to the metastatic process. CONCLUSION Current data indicate multiple roles for COX-2 in cancer, from the earliest stages of tumorigenesis through metastatic disease. COX2-derived PGs could be secreted and then signal in an autocrine or paracrine manner, providing a ‘‘landscaping’’ effect (Kinzler and Vogelstein 1998). The findings summarized here suggest that COX-2-derived prostanoids are involved in multiple steps and mechanisms that could lead to the prevention of apoptosis of transformed cells, increased invasiveness, immunosuppression and the stimulation of tumour-associated angiogenesis.
One very attractive consequence of COX-2 involvement in cancer is the possibility of specific pharmacological modulation. Recently developed COX-2 inhibitors brought the advantage of selectively inhibiting the COX-2 associated with pathological conditions while sparing the COX-1 required for normal function of the gastrointestinal tract (Prescott and Fitzpatrick 2000). There are already numerous reports of success for COX-2-selective inhibitors, such as celecoxib, in the prevention and therapy of cancer in animal models (Kawamori et al 1998; Masferrer et al 2000). Thus, COX-2 inhibitors are emerging as a novel class of anticancer agents. Still debated is whether the in vivo anticancer properties of COX-2 inhibitors are PG-mediated, PG-independent or both. However, the plasma concentrations achieved under regular treatment in animal models are highly consistent with inhibition of PG synthesis and inhibition of angiogenesis. The excellent safety profile of COX-2-specific inhibitors such as celecoxib opens the possibility of successfully combining them with other standard of care anticancer agents, such as ionizing radiation (Davis et al 2003) or CPT-11, a topoisomerase 1 inhibitor effective in treatment of a variety of cancers (Trifan et al 2002). Another promising drug combination is celecoxib and a monoclonal antibody against the ErbB receptor subtype HER-2/ neu. Mann et al (2001) have recently shown that this combination reduces colorectal carcinoma growth more effectively than either agent alone. Future clinical trials will be able to address the usage of such combinations in human patients.
REFERENCES Attiga FA, Fernandez PM, Weeraratna AT et al (2000) Inhibitors of prostaglandin synthesis inhibit human prostate tumor cell invasiveness and reduce the release of matrix metalloproteinases. Cancer Res, 60(16), 4629–4637. Balch CM, Dougherty PA, Cloud GA and Tilden AB (1984) Prostaglandin E2-mediated suppression of cellular immunity in colon cancer patients. Surgery, 95(1), 71–77. Ballif BA, Mincek NV, Barratt JT et al (1996) Interaction of cyclooxygenases with an apoptosis- and autoimmunity-associated protein. Proc Natl Acad Sci USA, 93(11), 5544–5549. Bandyopadhyay GK, Imagawa W, Wallace D and Nandi S (1987) Linoleate metabolites enhance the in vitro proliferative response of mouse mammary epithelial cells to epidermal growth factor. J Biol Chem, 262(26), 2750–2756. Cantarow WD, Cheung HT and Sundharadas G (1978) Effects of prostaglandins on the spreading, adhesion and migration of mouse peritoneal macrophages. Prostaglandins, 16(1), 39–46. Chan TA, Morin PJ, Vogelstein B and Kinzler KW (1998) Mechanisms underlying non-steroidal antiinflammatory drug-mediated apoptosis. Proc Natl Acad Sci USA, 95(2), 681–686. Chen WS, Wei SJ, Liu JM et al (2001) Tumor invasiveness and liver metastasis of colon cancer cells correlated with cyclooxygenase-2 (COX-2) expression and inhibited by a COX-2-selective inhibitor, etodolac. Int J Cancer, 91(6), 894–899. Cheng T, Cao W, Wen R et al (1998) Prostaglandin E2 induces vascular endothelial growth factor and basic fibroblast growth factor mRNA expression in cultured rat Mu¨ller cells. Invest Ophthalmol Vis Sci, 39(3), 581–591. Cianchi F, Cortesini C, Bechi P et al (2001) Upregulation of cyclooxygenase-2 gene expression correlates with tumor angiogenesis in human colorectal cancer. Gastroenterology, 121(6), 1339–1347. Dalgleish AG (1992) Immunotherapy of AIDS. Curr Opin Biotechnol, 3(6), 650–5. Daniel TO, Liu H, Morrow JD et al (1999) Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res, 59(18), 4574–4577. Davis TW, Hunter N, Trifan OC et al (2003) COX-2 inhibitors as radiosensitizing agents for cancer therapy. Am J Clin Oncol 26(4), S58–S61.
CYCLOOXYGENASE-2 IN CANCER Ding XZ, Tong WG and Adrian TE (2000) Blockade of cyclooxygenase-2 inhibits proliferation and induces apoptosis in human pancreatic cancer cells. Anticancer Res, 20(4), 2625–2631. Elder DJ, Hague A, Hicks DJ and Paraskeva C (1996) Different growth inhibition by the aspirin metabolite salicylate in human colorectal tumor cell lines: enhanced apoptosis in carcinoma and in vitrotransformed adenoma relative to adenoma cell lines. Cancer Res, 56(10), 2273–2276. Fitzpatrick FA and Wynalda MA (1983) Albumin-catalyzed metabolism of prostaglandin D2. Identification of products formed in vitro. J Biol Chem, 258(19), 11713–11718. Fu¨rstenberger G and Marks F (1978) Indomethacin inhibition of cell proliferation induced by the phorbolester TPA is reversed by prostaglandin E2 in mouse epidermis in vivo. Biochem Biophys Res Commun, 84(4), 1103–1111. Gallo O, Franchi A, Magnelli L et al (2001) Cyclooxygenase-2 pathway correlates with VEGF expression in head and neck cancer. Implications for tumor angiogenesis and metastasis. Neoplasia, 3(1), 53–61. Giardiello FM, Offerhaus GJ and DuBois RN (1995) The role of nonsteroidal antiinflammatory drugs in colorectal cancer prevention. Eur J Cancer, 31A(7–8), 1071–1076. Goin M, Pignataro O and Jimenez de Asua L (1993) Early cell cycle diacylglycerol (DAG) content and protein kinase C (PKC) activity enhancement potentiates prostaglandin F2a (PGF2a) induced mitogenesis in Swiss 3T3 cells. FEBS Lett, 316(1), 68–72. Goodwin JS (1984) Immunological effects of non-steroidal antiinflammatory drugs. Am J Med, 77, 7–15. Goodwin JS, Bankhurst AD and Messner RP (1977) Suppression of human T-cell mitogenesis by prostaglandin. Existence of a prostaglandin-producing suppressor cell. J Exp Med, 146(6), 719–734. Grilli M, Pizzi M, Memo M and Spano P (1996) Neuroprotection by aspirin and sodium salicylate through blockade of NF-kB activation. Science, 274(5291), 1383–1385. Grosch S, Tegeder I, Niederberger E et al (2001) COX-2-independent induction of cell cycle arrest and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib. FASEB J, 15(14), 2742–2744. Habeshaw J, Hounsell E and Dalgleish A (1992) Does the HIV envelope induce a chronic graft-vs.-host-like disease? Immunol Today, 13(6), 207–210. Hanif R, Pittas A, Feng Y et al (1996) Effects of non-steroidal antiinflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem Pharmacol, 52(2), 237–245. Hara A, Yoshimi N, Niwa M et al (1997) Apoptosis induced by NS-398, a selective cyclooxygenase-2 inhibitor, in human colorectal cancer cell lines. Jpn J Cancer Res, 88(6), 600–604. Harris RE, Robertson FM, Abou-Issa HM et al (1999) Genetic induction and upregulation of cyclooxygenase (COX) and aromatase (CYP19): an extension of the dietary fact hypothesis of breast cancer. Med Hypoth, 52(4), 291–292. Hashimoto N, Watanabe T, Ikeda Y et al (1997) Prostaglandins induce proliferation of rat hepatocytes through a prostaglandin E2 receptor EP3 subtype. Am J Physiol, 272(3/1), G597–G604. Hida T, Kozaki K, Muramatsu H et al (2000) Cyclooxygenase-2 inhibitor induces apoptosis and enhances cytotoxicity of various anticancer agents in non-small cell lung cancer cell lines. Clin Cancer Res, 6(5), 2006–2011. Homem de Bittencourt PI Jr and Curi R (1992) Interconversion of prostaglandins: a possible signal to control immune system function in cancer. Cancer Ther Control, 2, 217–222. Hoper MM, Voelkel NF, Bates TO et al (1997) Prostaglandins induce vascular endothelial growth factor in a human monocytic cell line and rat lungs via cAMP. Am J Resp Cell Mol Biol, 17(6), 748–756. Hsu AL, Ching TT, Wang DS et al (2000) The cyclooxygenase-2 inhibitor celecoxib induces apoptosis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem, 275(15), 11397–1403. Huang M, Sharma S, Mao JT and Dubinett SM (1996) Non-small cell lung cancer-derived soluble mediators and prostaglandin E2 enhance peripheral blood lymphocyte IL-10 transcription and protein production. J Immunol, 157(12), 5512–5520. Imperial JC (1999) Natural history of chronic hepatitis B and C. J Gastroenterol Hepatol, 14(suppl), S1–5.
345
Ito S, Narumiya S and Hayaishi O (1989) Prostaglandin D2: a biochemical perspective. Prostagland Leukotrienes Essent Fatty Acids, 37(4), 219–234. Jaffe BM (1974) Prostaglandins and cancer: an update. Prostaglandins, 6(6), 453–461. Jankowski JA, Wright NA, Meltzer SJ et al (1999) Molecular evolution of the metaplasia–dysplasia–adenocarcinoma sequence in the esophagus. Am J Pathol, 154(4), 965–973. Kalmar L, Gyires K and Gergely P (1984) Inhibitory effect of tumour cells on leukocyte motility. Immunol Lett, 8(4), 175–177. Kambayashi T, Alexander HR, Fong M and Strassmann G (1995) Potential involvement of IL-10 in suppressing tumor-associated macrophages. Colon-26-derived prostaglandin E2 inhibits TNF-a release via a mechanism involving IL-10. J Immunol, 154(7), 3383–3390. Kawaguchi H, Pilbeam CC, Harrison JR and Raisz LG (1995) The role of prostaglandins in the regulation of bone metabolism. Clin Orthop, 313, 36–46. Kawamori T, Rao CV, Seibert K and Reddy BS (1998) Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res, 58(3), 409–412. Kawamura T, Horie S, Maruyama T et al (1999) Prostaglandin E1 transported into cells blocks the apoptotic signals induced by nerve growth factor deprivation. J Neurochem, 72(5), 1907–1914. Kikawa Y, Narumiya S, Fukushima M et al (1984) 9-Deoxy-D9,D12-13,14dihydroprostaglandin D2, a metabolite of prostaglandin D2 formed in human plasma. Proc Natl Acad Sci USA, 81(5), 1317–1321. Kim JM, Kim JS, Jung HC et al (2000) Upregulated cyclooxygenase-2 inhibits apoptosis of human gastric epithelial cells infected with Helicobacter pylori. Dig Dis Sci, 45(12), 2436–2443. Kimura M, Osumi S and Ogihara M (2001) Prostaglandin E(2) [EP(1)] receptor agonist-induced DNA synthesis and proliferation in primary cultures of adult rat hepatocytes: the involvement of TGF-a. Endocrinology, 142(10), 4428–4440. Kinzler KW and Vogelstein B (1998) Landscaping the cancer terrain. Science, 280(5366), 1036–1037. Koki AT, Leahy KM and Masferrer JL (1999) Potential utility of COX-2 inhibitors in chemoprevention and chemotherapy. Expert Opin Investig Drugs, 8(10), 1623–1638. Kroll B, Kunz S, Tu N and Schwarz LR (1998) Inhibition of transforming growth factor-b1 and UV light-induced apoptosis by prostanoids in primary cultures of rat hepatocytes. Toxicol Appl Pharmacol, 152(1), 240–250. Leahy KM, Ornberg RL, Wang Y et al (2002) Cyclooxygenase-2 inhibition by celecoxib reduces proliferation and induces apoptosis in angiogenic endothelial cells in vivo. Cancer Res, 62(3), 625–631. Lecomte M, Laneuville O, Ji C et al (1994) Acetylation of human prostaglandin endoperoxide synthase-2 (cyclooxygenase-2) by aspirin. J Biol Chem, 269(18), 13207–13215. Leung KH and Mihich E (1980) Prostaglandin modulation of development of cell-mediated immunity in culture. Nature, 288(5791), 597–600. Levy GN (1997) Prostaglandin H synthases, non-steroidal antiinflammatory drugs, and colon cancer. FASEB J, 11(4), 234–247. Liu CH, Chang SH, Narko K et al (2001) Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem, 276(21), 18563–18569. Liu XH, Yao S, Kirschenbaum A and Levine AC (1998) NS398, a selective cyclooxygenase-2 inhibitor, induces apoptosis and downregulates bcl-2 expression in LNCaP cells. Cancer Res, 58(19), 4245–4249. Lyson K (1984) Indomethacin suppression of the estradiol-induced proliferative response of the seminal vesicles. Exp Clin Endocrinol, 84(2), 223–227. Majima M, Isono M, Ikeda Y et al (1997) Significant roles of inducible cyclooxygenase (COX)-2 in angiogenesis in rat sponge implants. Jpn J Pharmacol, 75(2), 105–114. Mann M, Sheng H, Shao J et al (2001) Targeting cyclooxygenase 2 and HER-2/neu pathways inhibits colorectal carcinoma growth. Gastroenterology, 120(7), 1713–1719. Mamett LJ (1992) Aspirin and the potential role of prostaglandins in colon cancer. Cancer Res, 52(20), 5575–5589. Masferrer JL, Koki A and Seibert K (1999) COX-2 inhibitors. A new class of antiangiogenic agents. Ann NY Acad Sci, 89, 84–86. Masferrer JL, Leahy KM, Koki AT et al (2000) Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res, 60(5), 1306–1311.
346
INFLAMMATION
Mayne ST, Buenconsejo J and Janerich DT (1999) Previous lung disease and risk of lung cancer among men and women non-smokers. Am J Epidemiol, 149(1), 13–20. McCann J (1999) Esophageal cancers: changing character, increasing incidence. J Natl Cancer Inst, 91(6), 497–498. McKanna JA, Zhang MZ, Wang JL et al (1998) Constitutive expression of cyclooxygenase-2 in rat vas deferens. Am J Physiol, 275(1/2), R227–233. Morris CD, Armstrong GR, Bigley G (2001) Cyclooxygenase-2 expression in the Barrett’s metaplasia–dysplasia–adenocarcinoma sequence. Am J Gastroenterol, 96(4), 990–996. Munroe DG, Wang EY, MacIntyre JP et al (1995) Novel intracellular signaling function of prostaglandin H synthase-1 in NF-kB activation. J Inflamm, 45(4), 260–268. Nolan RD, Danilowicz RM and Eling TE (1988) Role of arachidonic acid metabolism in the mitogenic response of BALB/c 3T3 fibroblasts to epidermal growth factor. Mol Pharmacol, 33(6), 650–656. Ohno K, Sakai T, Fukushima M et al (1988) Site and mechanism of growth inhibition by prostaglandins. IV. Effect of cyclopentenone prostaglandins on cell cycle progression of G1-enriched HeLa S3 cells. J Pharmacol Exp Ther, 245(1), 294–298. Oshima M, Dinchuk JE, Kargman SL et al (1996) Suppression of intestinal polyposis in APCD716 knockout mice by inhibition of cyclooxygenase-2 (COX-2). Cell, 87(5), 803–809. Ottonello L, Gonella R, Dapino P et al (1998) Prostaglandin E2 inhibits apoptosis in human neutrophilic polymorphonuclear leukocytes: role of intracellular cyclic AMP levels. Exp Hematol, 26(9), 895–902. Pentland AP and Needleman P (1986) Modulation of keratinocyte proliferation in vitro by endogenous prostaglandin synthesis. J Clin Invest, 77(1), 246–251. Peterson HI (1983) Effects of prostaglandin synthesis inhibitors on tumor growth and vascularization. Experimental studies in the rat. Invasion Metastasis, 3(3), 151–159. Prescott SM and Fitzpatrick FA (2000) Cyclooxygenase-2 and carcinogenesis. Biochim Biophys Acta, 1470(2), M69–78. Quarles LD, Haupt DM, Davidai G and Middleton JP (1993) Prostaglandin F2a-induced mitogenesis in MC3T3-E1 osteoblasts: role of protein kinase-C-mediated tyrosine phosphorylation. Endocrinology, 132(4), 1505–1513. Reddy BS, Maruyama H and Kelloff G (1987) Dose-related inhibition of colon carcinogenesis by dietary piroxicam, a non-steroidal antiinflammatory drug, during different stages of rat colon tumor development. Cancer Res, 47(20), 5340–5346. Reddy BS, Rao CV and Seibert K (1996) Evaluation of cyclooxygenase-2 inhibitor for potential chemopreventive properties in colon carcinogenesis. Cancer Res, 56(20), 4566–4569. Richter M, Weiss M, Weinberger I et al (2001) Growth inhibition and induction of apoptosis in colorectal tumor cells by cyclooxygenase inhibitors. Carcinogenesis, 22(1), 17–25. Rigas B and Shiff SJ (2000) Is inhibition of cyclooxygenase required for the chemopreventive effect of NSAIDs in colon cancer? A model reconciling the current contradiction. Med Hypoth, 54(2), 210–216. Rocklin RE, Thistle L and Audera C (1985) Decreased sensitivity of atopic mononuclear cells to prostaglandin E2 (PGE2) and prostaglandin D2 (PGD2). J Immunol, 135(3), 2033–2039. Rozic JG, Chakraborty C and Lala PK (2001) Cyclooxygenase inhibitors retard murine mammary tumor progression by reducing tumor cell migration, invasiveness and angiogenesis. Int J Cancer, 93(4), 497–506. Sawaoka H, Tsuji S, Tsujii M et al (1999) Cyclooxygenase inhibitors suppress angiogenesis and reduce tumor growth in vivo. Lab Invest, 79(12), 1469–1477. Scott WA, Pawlowski NA, Murray HW et al (1982) Regulation of arachidonic acid metabolism by macrophage activation. J Exp Med, 155(4), 1148–1160. Sheng H, Shao J, Kirkland SC et al (1997) Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest, 99(9), 2254–2259. Sheng H, Shao J, Morrow JD et al (1998) Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 treatment in human colon cancer cells. Cancer Res, 58(2), 362–366. Siddiqui RA and Williams JF (1990) Tentative identification of the toxohormones of cancer cachexia: roles of vasopressin, prostaglandin E2 and cachectin-TNF. Biochem Int, 20(4), 787–797. Smalley WE and DuBois RN (1997) Colorectal cancer and nonsteroidal antiinflammatory drugs. Adv Pharmacol, 39, 1–20.
Smith BJ, Curtis JF and Eling TE (1991) Bioactivation of xenobiotics by prostaglandin H synthase. Chem Biol Interact, 79(3), 245–264. Smith WL, Meade EA and DeWitt DL (1994) Interactions of PGH synthase isozymes-1 and -2 with NSAIDs. Ann NY Acad Sci, 744, 50–57. Stark LA, Din FV, Zwacka RM and Dunlop MG (2001) Aspirin-induced activation of the NF-kB signaling pathway: a novel mechanism for aspirin-mediated apoptosis in colon cancer cells. FASEB J, 15(7), 1273–1275. Steele VE, Holmes CA, Hawk ET et al (2000) Potential use of lipoxygenase inhibitors for cancer chemoprevention. Expert Opin Investig Drugs, 9(9), 2121–2138. Subbaramaiah K, Telang N, Ramonetti JT et al (1996) Transcription of cyclooxygenase-2 is enhanced in transformed mammary epithelial cells. Cancer Res, 56(19), 4424–4429. Takahashi Y, Kawahara F, Noguchi M et al (1999) Activation of matrix metalloproteinase-2 in human breast cancer cells overexpressing cyclooxygenase-1 or -2. FEBS Lett, 460(1), 145–148. Trifan OC, Durham WF, Salazar V et al. COX-2 inhibition with celecoxib enhances anti-tumor efficacy and reduces diarrhea side effect of CPT11. Cancer Res 62(20), 5778–5784. Trifan OC, Smith RM, Thompson BD and Hla T (1999) Overexpression of cyclooxygenase-2 induces cell-cycle arrest: evidence for a prostaglandinindependent mechanism. J Biol Chem, 274(48), 34141–34147. Tsujii M and DuBois RN (1995) Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell, 83(3), 493–501. Tsujii M, Kawano S and DuBois RN (1997) Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA, 94(7), 3336–3340. Tsujii M, Kawano S, Tsuji S et al (1998) Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell, 93(5), 705–716. Urade Y, Ujihara M, Horiguchi Y et al (1989) The major source of endogenous prostaglandin D2 production is likely antigen-presenting cells. Localization of glutathione-requiring prostaglandin D synthetase in histiocytes, dendritic, and Kupffer cells in various rat tissues. J Immunol, 143(9), 2982–2989. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231(25), 232–236. Verma AK, Ashendel CL and Boutwell RK (1980) Inhibition by prostaglandin synthesis inhibitors of the induction of epidermal ornithine decarboxylase activity, the accumulation of prostaglandins and tumor promotion caused by 12-O-tetradecanoylphorbol-13acetate. Cancer Res, 40(2), 308–315. Weitzman SA and Gordon LI (1990) Inflammation and cancer: role of phagocyte-generated oxidants in carcinogenesis. Blood, 76(4), 655–663. Whitby D and Boshoff C (1998) Kaposi’s sarcoma herpesvirus as a new paradigm for virus-induced oncogenesis. Curr Opin Oncol, 10(5), 405– 412. Williams CS, Tsujii M, Reese J et al (2000) Host cyclooxygenase-2 modulates carcinoma growth. J Clin Invest, 105(11), 1589–1594. Williams ED, Karim SM and Sandler M (1968) Prostaglandin secretion by medullary carcinoma of the thyroid. A possible cause of the associated diarrhoea. Lancet, 1(7532), 22–23. Williams MP and Pounder RE (1999) Helicobacter pylori: from the benign to the malignant. Am J Gastroenterol, 94(11, suppl), S11–16. Yamada M, Kawai M, Kawai Y and Mashima Y (1999) The effect of selective cyclooxygenase-2 inhibitor on corneal angiogenesis in the rat. Curr Eye Res, 19(4), 300–304. Yamamoto Y, Yin MJ, Lin KM and Gaynor RB (1999) Sulindac inhibits activation of the NF-kB pathway. J Biol Chem, 274(38), 27307–27314. Yin MJ, Yamamoto Y and Gaynor RB (1998) The antiinflammatory agents aspirin and salicylate inhibit the activity of I(k)B kinase-b. Nature, 396(6706), 77–80. Yoneda T (1998) Cellular and molecular mechanisms of breast and prostate cancer metastasis to bone. Eur J Cancer, 34(2), 240–245. Zhang X, Morham SG, Langenbach R and Young DA (1999) Malignant transformation and antineoplastic actions of non-steroidal antiinflammatory drugs (NSAIDs) on cyclooxygenase-null embryo fibroblasts. J Exp Med, 190(4), 451–459. Zhao Y, Agarwal VR, Mendelson CR and Simpson ER (1996) Estrogen biosynthesis proximal to a breast tumor is stimulated by PGE2 via cyclic AMP, leading to activation of promoter II of the CYP19 (aromatase) gene. Endocrinology, 137(12), 5739–5742.
31 Cytokines and Eicosanoids in Arthritis K. D. Rainsford Sheffield Hallam University, Sheffield, UK
Cytokines are key protein mediators that are expressed in the cells of the immunoinflammatory lineage, especially during chronic inflammatory phases of diseases, including arthritic conditions. They comprise pro- and antiinflammatory cytokines whose properties are described in Table 31.1. A major part of the responses elicited by proinflammatory cytokines, especially interleukins-1a and -1b (IL-1a and IL-1b), tumour necrosis factor-a (TNFa) and IL-6 involve the induction of: (a) phospholipases A2, C and D; (b) the inducible or inflammatory cyclooxygenase, COX-2 (or prostaglandin G/H-synthase-2, PGHS-2), and (c) lipoxygenases, culminating in the production of eicosanoids, prostaglandins, leukotrienes and lipoxins, plateletactivating factor (PAF), diacylglycerols, ceremide and other autocoids (Serhan et al 1996). They also contribute to cell signalling via intracellular events, e.g. induction of the NFkB–IkB pathway (Stylianou et al 1992; Malinin et al 1997) that controls the induction of COX-2 and other inducible enzymes, e.g. isoenzyme involved in production of nitric oxide (iNOS) and metalloproteinases involved in the control and expression of cellular inflammatory processes (Belt et al 1999; Scott et al 1999; Nakao et al 2000). In turn, the production of proinflammatory cytokines is negatively or positively regulated by products of the cyclooxygenase and lipoxygenase pathways (Lewis and Gordon 1985; Kunkel et al 1986; Greaves and Camp 1988; RolaPleszczynski et al 1990). These feedback loops for control of eicosanoid metabolism by proinflammatory cytokines represent important and in some cases subtle control elements in the regulation of cytokine–eicosanoid interactions, the implications of which are not yet completely understood. CYTOKINE PRODUCTION IN ARTHRITIC AND OTHER CHRONIC INFLAMMATORY DISEASES Cytokines are defined rather loosely as a group of protein/ polypeptide or glycoprotein regulators produced by leukocytes and some other cells involved in inflammatory processes. They contribute to the development of the classical signs of inflammation—heat (calor), redness (rubor), swelling (tumor) and pain (dolor), as well as the sign defined by Virchov as loss of function (functio laesa). The term ‘‘cytokine’’ was proposed to embrace the older term ‘‘lymphokine’’ (or products of sensitized lymphocytes) (Dumonde et al 1969), ‘‘monokines’’ (products of activated monocytes/macrophages) ‘‘interferons’’, haematopoietic colonystimulating and growth factors. Interleukins constitute proteins communicating signals between different populations of leukocytes (2nd International Lymphokine Workshop 1979). The cytokines have essentially paracrine or autocrine actions, many having a limited field of action, although a few key cytokines accumulate in the circulation. A major and puzzling feature of The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
cytokines is that many have overlapping or pleotropic and antagonistic actions (Tables 31.1 and 31.2). They are grouped into those which have proinflammatory activity and those which negatively regulate immuno-inflammatory reactions and are thus termed antiinflammatory cytokines. The general molecular–biochemical and cellular properties of cytokines are summarized in Tables 31.1–31.5. Different cytokines are variously produced in various arthritic diseases, although there are probably primary cytokines involved at key points in development of joint manifestations and lymphocyte regulation that ‘drive’ the major manifestations of different arthritic diseases. The property of those cytokines of interest in the joint destructive events and eicosanoid production are shown in Tables 31.2–31.5. To understand the relationship between cytokine actions and their role in regulating eicosanoids in arthritis, it is useful to consider first the pattern of eicosanoid production in the inflamed joints in different cells that contribute to the inflammatory reactions therein, and those proinflammatory cytokines that are largely responsible for initiating eicosanoid production. Figure 31.1 summarizes the pattern of different eicosanoids and principal proinflammatory cytokines produced by different cells in inflamed joints and the key roles of PGE2 and LTB4 in regulating production of IL-1, TNFa and IL-6 by these cells. Another concept that is important for understanding the interrelationship between eicosanoids, cytokines and cellular immune functions in arthritic diseases are links between helper T lymphocytes and macrophages. The messengers are key proinflammatory cytokines such as IL-1, TNFa, IL-2 and interferon-b (IFN-b) (Figure 31.2). Cytokine imbalance comprising an excess of proinflammatory over antiinflammatory cytokines is considered the major driving force in the development of rheumatoid arthritis and probably, in some modified form, many other arthritic conditions. Figure 31.3 outlines a scheme showing the individual influence of proinflammatory compared with antiinflammatory cytokines in mediating this imbalance. The actions of antiinflammatory cytokines and endogenous inhibitors or ‘‘soluble’’ receptors that ‘‘mop up’’ proinflammatory cytokines are principally negative regulators of those immunoinflammatory cells that are responsible for production of proinflammatory cytokines (Oppenheim and Feldmann 2000; Oppenheim 2001). They represent essentially indirect regulators of proinflammatory cytokines and, through their actions, production of eicosanoids by these cells. A central element in the perturbation on the regulation of cytokines in chronic autoimmune diseases, e.g. rheumatoid arthritis (RA), is what is known as the Th1–Th2 balance (Figure 31.4A,B; Miossec and van den Berg 1997; Aarden et al 1997). The essence of this concept is that the cellular immune functions that are promoted by IL-2 and interferon-a (IFN-a) in the Th1producing subset of CD4+ helper T cells are counterbalanced by
348
INFLAMMATION
Table 31.1 Summary of proinflammatory and antiinflammatory cytokines Cytokines
Role
TNFa
Proinflammatory
IL-1
Proinflammatory
IL-2 IL-4 IL-5 IL-6
T cell growth factor B cell growth factor/antiinflammatory
IL-7 IL-8 IL-10
Lymphoprotein/proinflammatory Proinflammatory Antiinflammatory
IL-11 IL-12 IL-13 IL-15 IL-16 IL-17 IL-18 IFN-g
Antiinflammatory Proinflammatory Antiinflammatory Proinflammatory Proinflammatory Proinflammatory Proinflammatory Proinflammatory/antiinflammatory
TGFb
Antiinflammatory/proinflammatory
"COX-2; "PGE2; "NO; "adhesion molecules; "IL-1; "IL-6; "chemokines; "collagenases; "cell death; "procoagulant activity "COX-2; "PGE2; "adhesion molecules; "TNFa; "IL-6; "chemokines; "collagenases; "procoagulant activity; "osteoclast activation; "angiogenic factors "T cell proliferation; "NK cells; "TNFa; "IFN-g "Antibody production; "IgE; #IL-1; #TNFa; "IL-Ra "Eosinophils and differentiation "Acute phase proteins; "antibody production; "anorexia; "fibroblast proliferation; #IL-1; #TNFa; #chemokines; "IL-1Ra; "soluble TNF-RI (p55) "IgG synthesis; "adhesion molecules; #lymphocyte death Chemokine for neutrophils; "neutrophil infiltration; "angiogenesis "Antibody production; #IL-1; #TNFa; #chemokines; #MHC Class II expression; #adhesion molecules; #PGE2; #NO; #PLA2; #NO; #PLA2; #gelatinase; #collagenases "Acute phase proteins; #IL-1; #TNFa; #IFN-g; "IL-12 "IFN-g; "NK cells; "IFN-a; "IL-1; "IL-18 #PGE2; #NO; #IL-1; #TNFa; "IL-1Ra "T cell proliferation; "chemokines; "TNFa; #T cell death Eosinophil chemoattractant; "monocyte; "T cell infiltration "PGE2 "NO; "chemokines; "osteoclast activation "IFN-g; "TNFa; "IL-1; "ICAM-1; "VCAM-1; "NK cells "MHC Class II expression; "adhesion molecules; "NO; "TNFa; "IL-1 receptors; #TNFa; "IL-1 receptors; #PGE2; #collagenases; #IL-1 "Collagen synthesis, immunosuppressant, antagonizes IL-1 and TNFa activities; #IL-1; #IL-2; #TNFa synthesis
Antiinflammatory/proinflammatory
Based on information in Thomson (1998), Oppenheim and Feldman (2000) and Oppenheim (2001)
Table 31.2 Pleotropic actions of cytokines IFN-a IFN-b IFN-g TNFa LT Cytotoxic for tumour cells Antitumour activity Cytostatic for various cells Mitogenic for various cells Active B cells Stimulate B cell proliferation Stimulate B cell differentiation Stimulate B cell differentiation Stimulate isotype selection Induce IgE receptor on B cells Active T cells Stimulate T cell proliferation Stimulate T cell differentiation Induce Class I MHC Induce Class II MHC
+ + +
+ + +
7 7
+
7 + +3
+
Activate macrophages Stimulate granulocyte activity Stimulate eosinophil activity Stimulate NK cell activity Stimulate LAK activity Stimulate osteoclastic bone resorption Induce cellular antiviral state Stimulate production of ECM proteins Induce chemotactic migration of cells Stimulate angiogenesis in vivo Induce adhesion molecules Effect on trace elements in vivo Induce acute phase proteins Adjuvanicity Induce fever and weight loss 1
2
+ + +1 +
+ 7 +
+ + + +But not B cells +
+ 7 7
+
+5
+
+ + + +
+ + + +
+ + + + IgG2
+ + +
+ + + 7
+ +
+
+
IL-1 IL-2 IL-3 IL-4 IL-5 IL-6 IL-7 G-CSF M-CSF CSF-GM + + + + + +
+ +
+ + 7 + + + + +2 7 + +
+
+ +
+
+ + + +IgA
+
+ + +
+ + + + +
+ +
+
+ +
+ 3
+
+
+
+
+
+4
+
+ +
+ + +
+
+
+
+ + 7
+ +
+ +
+
7
7
7
+
+ + 7 7 7
7 7 7 +Depends on cell system +Depends on cell system
+ 7 7 + +
6
+
+
+B cells
Very + weak PMNs + + + + + + + + + + +
TGFb
+ PMNs No weak PMNs+Mf + + + + + +
+ + + + +
+
+ +
4
Synovial cells. Induces cytotoxic activity in B cells. On some subspecies of monocytes. Activates murine macrophages for tumour cell killing. 5But not T cells. 6Including PMNs/not including PMNs—some disagreement here!
CYTOKINES AND EICOSANOIDS IN ARTHRITIS the negative humoral immune functions by IL-4, IL-5, IL-10 and IL-13 producing Th2, CD4+ T cells (Figure 31.4A,B). For example, in rheumatoid arthritis the primary stimuli (infectious agents) and heat shock proteins stimulate Th1 cells to produce IL2 and IFN-a and the latter further amplifies the Th1 response (Figure 31.4B). Both these cytokines inhibit Th2 production of cytokines, so contributing to suppression in the cytokine production by these cells (Figure 31.4B). Since there are cytokine effects on the regulation of macrophage functions and these cells in turn are principal sources of proinflammatory cytokines (outlined in Figure 31.2), the Th1 amplification of both IL-2 and IFN-b can lead to enhancement of TNFa, IL-1 and IP-10 production, and these cytokines can in turn promote their own production by macrophages, as well as causing upregulation of the major histocompatability Class II receptor expression (MHCII), so promoting further T cell activation (Figure 31.2). The situation concerning the immune responses in joint destruction in osteoarthritis (OA) and related conditions is less clear than in RA, although there are localized immunological factors (antibody, macrophage activation) that contribute to destructive events. Recently it was shown that IL-23 and IL-12 play a central role in autoimmune disease in the induction of experimental autoimmune encephalomyelitis (EAE) reactions in mice that control production of proinflammatory cytokines IL-1, IFN-b and TNFa (Cua et al 2003; Watford and O’Shea 2003), so providing a link between T cells and the initiating events involved in production of those cytokines that initiate macrophages and local tissue destruction by these cells. Table 31.1 and Figures 31.3 and 31.4 summarize the current status of the range and properties of pro- and antiinflammatory cytokines and their major actions in regulating chronic inflammatory diseases. For the purposes of understanding the significant roles of proinflammatory cytokines, principally IL-1a and IL-1b, in regulating eicosanoid metabolism, it can be seen in Figure 31.1 that IL-1a/b can stimulate phospholipase A2 (PLA2) and COX-2 but, at the same time, the products of both COX and 5lipoxygenase (5-LOX) pathways can regulate these stimulatory activities. The pattern of proinflammatory cytokine production differs principally in a quantitative sense in different arthritic conditions (Figure 31.4A,B). For instance, in rheumatoid arthritis (RA) there is emerging a view that TNFa and IL-1a and IL-1b, IL-6 and chemokines along with the shift in balance of Th1/Th2 cytokines in the direction of Th1-driven events. Th1 cytokines are the principal cytokine responses that are expressed in this disease. Polymorphisms in TNFa, IL-1 and IL-6 may be important in
349
determining the eventual outcome of the joint inflammatory reactions and systemic responses in RA and other inflammatory arthropathies (Cox and Duff 1996; Duff 1998; Perrier et al 1998), although this has yet to be fully determined. The production of soluble receptors or endogenous inhibitors of TNFa and IL-1 also varies in RA and may influence the outcome of this disease in individual patients. The relative importance of these soluble receptor/endogenous inhibitors in influencing the response of proinflammatory cytokines in RA can be seen from the pronounced clinical efficacy of the commercially-produced biological agents, anti-TNFa and IL-1 receptor antagonist (IL-1ra) (van den Berg 2002). INDUCTION OF PHOSPHOLIPASE A2 AND CYCLOOXYGENASE-2 Kunkel et al (1986) and Bonta and Elliott (1992) showed that IL-1 and TNFa could stimulate the production of prostaglandin E2. Following the identification of PLA2 and COX-2, it was shown by a number of authors (Maier et al 1990; O’Sullivan et al 1992; Kennard et al 1995; Wilborn et al 1995; Hulkower et al 1997; Kuwata et al 1998; Tada et al 1998; Murakami et al 1999; Rauk and Ciao 2000; reviewed by Smith and De Witt 1995 and Scott et al 1999) that stimulus-induced formation of prostanoids by cells involved the sequential and coordinated expression of phospholipase A2 (PLA2) and COX-2. Diaz et al (1992) showed that the expression of COX-2 protein could be differentially regulated by transforming growth factor b, (TGFb1), IL-1b, TNFa and the phorbol ester (phorbol 12myristate 13-acetate, or PMA) by eliminating one of four Nglycosylation sites by splicing exon 9 of the COX-2 gene products. With knowledge of the COX-2 promotor site (Crofford 1997), it can be seen that there are a number of intracellular transcription factor binding sites where there could be gene activation as a consequence of stimulating production of these factors. Chief among these, which has attracted much interest in the context of COX-2 induction in rheumatic diseases, has been nuclear factorkb (NF-kB; Vilcek, 1998), whose production is negatively regulated by a number of NSAIDs, including those that inhibit production of COX-2 protein (see Rainsford, Chapter 16, this volume). Expression of COX-2 protein has been shown immunochemically in human synovial explants or synoviocytes from patients with RA and in the hind paw homogenates of rats with mycobacterial adjuvant-induced polyarthritis (Crofford et al 1994; Kang et al 1996) in various joint tissues of patients with
Table 31.3 Properties and functions of interleukin 1 . With TNFa, a ‘‘primary’’ cytokine involved in expression of inflammatory responses. Overlapping properties (pleotropism) with TNFa, IL-6, IL-8. Interactions with other cytokines in eliciting inflammatory responses . Formerly described according to biological properties as: –Endogenous pyrogen/leukocyte endogenous mediator=EP/LEM –Lymphocyte/B cell/thymocyte activating factor –Mononuclear cell factor/catabolin osteoclast activating factor!increases PG and collagenase production; produces cartilage/bone resorption=MCF . IL-1a and IL-1b produced as 31 kDa precursors from two separate genes on the long arm of chromosome 2; both polypeptides cleaved by proteolysis to mature forms; MW 15 kDa; IL-1a is acid (pH 5.3), produced mostly by Langerhan’s cells (skin) and synoviocytes (joints); IL-1b is neutral (pH 7.2), produced mostly by monocytes/macrophages. 75–78% homology in sequences across spp.; only 25% homology within isotypes, yet both have similar effects . IL-1b produced as inactive Pro-IL-1b (34 kDa) and converted by cysteine protease (interleukin-1 converting enzyme=ICE; caspase 1) that has importance in apoptosis. ICE-knockout mice develop normally and are fertile but do not respond to endotoxin (LPS) induction of shock . IL-1 stimulates NF-kB gene activation to yield coordinated synthesis of other inflammatory mediators and metalloproteinases (MMPs; proteoglycanase, collagenases) and neutrophil oxyradicals that degrade cartilage in osteo- and rheumatoid-arthritic diseases. They also inhibit proteoglycan synthesis. MMPs also degrade matrix proteins in other chronic diseases . IL-1 induces enzymes involved in induction of the isoenzyme that is involved in production of inflammatory prostaglandins (PGHS-2=COX-2) and nitric oxide (iNOS)
350
INFLAMMATION
Table 31.4 Properties and functions of tumour necrosis factor-a . Formerly known as cachectin; accounts for increased wasting (=cachexia) in cancer, parasitic and chronic autoimmune diseases. Produced mostly by macrophages. Considered a major cytokine— ‘‘apex of cytokines’’ in a cascade that includes IL-1b, IL-6, GM-CSF and chemokines in rheumatoid arthritis. This view has, however, been challenged. . TNFa important in parasitic diseases (especially cerebral malaria), autoimmune diseases (e.g. rheumatoid arthritis, multiple sclerosis) and cancer . Mr=17 kDa (trimer Mr=45–55 kDa); glycoprotein; pH-5.8. Highly conserved sequence in mammals of IL-1 and TNFa . Similarities of actions . Immunological properties –T cell activation –Increased IL-2R expression –B cell activation –Natural killer activity: synergism with IL-2 and IFN –Lymphokine gene expression . Proinflammatory properties –Fever, sleep, anorexia, neuropeptide release –Gene expression for complement; suppression of P450 synthesis –Endothelial cell activation –Increased adhesion molecule expression –Neutrophil priming, eosinophil degranulation –Hypotension, myocardial infarction, shock, death –Neutrophil tissue infiltration (via IL-8) –b Islet cell cytotoxicity –Amino acid turnover; hyperlipidaemia –Cyclooxygenase and lipooxygenase gene expression –Synthesis of collagenase and collagens; osteoblast activation . Protective effects in: –Malaria –Bacterial infections –Lethal radiation –Hyperoxia . Similarities of actions
Table 31.5 Properties and functions of interleukin 6 (=hepatocyte stimulating factor, B cell stimulating factor) . Glycoprotein (Mr=21–28 kDa) produced by most antigen presenting cells (APCs), chondrocytes, astroglial cells . Stimulated by IL-1s Summary of functions of IL-6: . With IL-1, controls production of hepatocyte acute phase proteins or acute phase reactants. These are pathognomic for many chronic inflammatory diseases and are routine clinical biomarkers. IL-6 also involved with IL-1 in fever and thermogenesis . Activates T and B cells—" expression of IL-2 receptors . Increases proliferation of NK cells . Activates osteoblasts . Increases megakaryocyte formation of platelets and haematopoietic stem cell proliferation . Regulates ACTH
RA and other inflammatory arthropathies (Siegle et al 1998), as well as in cartilage specimens from osteoarthritis (OA) patients (Amin et al 1997). COX-2 in MRNA and COX-2 protein were markedly increased by IL-1b in RA synoviocytes (Crofford 1997). In cultured OA cartilage stimulated with lipopolysaccharide (LPS), IL-1b or TNFa, upregulation of COX-2 was coincident with ‘‘superinduction’’ of PGE2, whose production was attenuated by induced nitric oxide (NO) production, thus showing the importance of interrelationships between the COX-2 and iNOS pathways in the control of cytokine-mediated COX-2. The principal agent mediating gout, monosodium urate (MSU), which is a well-known stimulator of IL-1, IL-6, IL-8, TNFa and PGE2 (Duff et al 1983; Dayer et al 1987; Guerne et al 1989;
Di Giovine et al 1991; Terkeltaub et al 1991), has also been shown to cause induction of COX-2 (Pouliot et al 1998). This induction of COX-2 by MSU was found to be inhibited by the anti-gout agent colchicine, as well as by transcriptional and translational inhibitors and an inhibitor of p38 mitogen-activated protein kinase, showing that cell cycle activation and gene expression are affected by MSU activation of COX-2. COX-2 expression has also been found to be induced in carrageenan-induced inflammation in rats (Harada et al 1996; Nantel et al 1999) at times coincident with cytokine production. The well-known production of fever induced by IL-1 (endogenous pyrogen) and some other cytokines (Milton 1982) has been shown to be related to the induction of COX-2 mRNA in the brain vasculature of rats following systemic administration of IL-1b (Cao et al 1996). The transcriptional regulation of COX-2 was hitherto thought to be enhanced by IL-1b and TNFa through the NF-kB pathway (O’Neill 1995; Karin 1998; Belt et al 1999). It emerges that when inflammatory stimuli such as endotoxin are employed which lead to increase in production of IL-1b, TNFa, IL-6 and induction of COX-2, there may be more complex intracellular signalling events. Induction of COX-2 by IL-1b in endothelial cells is inhibited by PGE2 via cAMP (Akarasereenont et al 1999). Wadleigh et al (2000) showed that endotoxin (LPS) treatment of the murine macrophage cell line RAW 264.7 leading to induction of COX-2 required the cAMP-response element (CRE) and two nuclear factor interleukin-6 (NF-IL-6) sites (ccAAT/enhancer proteins b and d). There was no response to NF-kB inhibition, although previous studies in the same cell line showed that synthetic inhibitors of NF-kB blocked COX-2 induction. PLA2 activation and COX-2 induction can also involve the p38 MAP kinase pathway, and IL-1b can induce production of the bradykinin B2 receptor by this pathway (Schmidlin et al 2000). This may represent an important pathway linking proinflammatory cytokines COX-2 and iNOS and pain-eliciting responses mediated by peptides such as bradykinin in the development of pain responses. In human myometrial cells, IL-1b induced COX-2 and increased PGE2 production was inhibited by pertussis toxin, as was the degradation of IkB and the expression of the mitogenactivated protein (MAP) kinase family, ERK and JNK. Other studies with synthetic inhibitors confirm the activation of the MAP kinases ERK and p38 during COX-2 induction and PKC, although the co-stimulation by IL-6 of COX-2 does not appear to be affected by PKC inhibitors. Figure 31.5a,b shows a schematic representation of the major intracellular signalling pathways, with those that appear to be involved in proinflammatory reactions involving cytokine–eicosanoid production highlighted. The stimulation of COX-2 production by IL-1b also appears to involve an increase in mRNA stability, which is mediated by ERK1/2 signalling. Cytokines other than IL-1 and TNFa can influence or be involved in the regulation of COX-2, among them IL-10 (Pomini et al 1999), which negatively regulates COX-2 and PGE2 production induced by other cytokines. Moreover, IL-6 is induced by increased PGE2 from COX-2 (Hinson et al 1996), possibly via the activation of the EP3 subtype of the PGE2 receptor. This and the other PGE2 receptors, EP2 and EP4, are upregulated by LPS (Sugimoto et al 2000) and this may be mediated by effects on IL-1 and TNFa by this inflammogen. Activation of pain pathways is known to be accompanied by induction of COX-2 in those regions in both peripheral as well as central nerves (e.g. dorsal horn, cells of the spinothalamic pathway) that are involved in pain transmission signals (Figure 31.6). IL-1b leads to upregulation of COX-2 as well as PLA2 activities in the central nervous system (CNS) and is considered the principal regulator of inflammation-mediated hyperalgesia, pain and febrile responses in the CNS. These studies on the induction of COX-2 by cytokines in joint tissues and CNS pathways show
CYTOKINES AND EICOSANOIDS IN ARTHRITIS
351
Figure 31.1 Production of eicosanoids in the cells and synovial compartment in patients with arthritic conditions and the regulation of their production by the principal proinflammatory cytokines. Key features of these reactions are the activating reactions elicited by IL-1b, TNFa and IL-6 in regulating (via induction and activation of phospholipase A2, cyclooxygenase-2 and 5, 12 and 15-lipoxygenases) the production of PGE2, LTB4 and those other eicosanoids that are produced as shown on the left side. Negative feedback shown as (7) on the production of IL-1b (as well as of lymphocyte proliferation) is produced by PGE2, while enhanced production of this cytokine shown as (+) is obtained by LTB4. LTB4 is produced in exceptionally high quantities in the joints of patients with gout, monosodium urate (MSU) activation of polymorphs probably accounting for the high amounts of LTB4 that are formed.
the importance of this cytokine–COX-2 axis in the mediation of inflammatory reactions in rheumatic pain and fever conditions. Coincident with induction of COX-2, there is also increase in receptors for PGI2 (IP), PGE2 (EPI, EP3, EP4), along with expression of their mRNAs (Sugimoto et al 2000).
DIFFERENTIAL REGULATION OF PROINFLAMMATORY CYTOKINES BY EICOSANOIDS AND SIGNALLING PATHWAYS Kunkel et al (1986) and Brandwein (1986, 1990) showed that cytokine production (IL-1) in mouse macrophages could be negatively regulated by PGE2. Brandwein (1986, 1990) showed that agents that stimulate cyclic AMP (cAMP) production or inhibit its production could block the intracellular and extracellular production of IL-1 and, since PGE2 activates PGE2 receptormediated cAMP production, that this could represent the means by which IL-1 is regulated (Zitnik et al 1993). More recent studies have shown that the PGE2-induced suppression of IL-1 production is mediated in macrophages by the EP2 and EP4 PGE2 receptor subtypes, which upon activation lead to increased adenylate cyclase and increased cAMP production (Sugimoto et al 2000). While the 5-lipoxygenase (5-LOX) leukoattractant/leukoactivator, leukotriene B4, did not alone affect IL-1 expression, it enhances LPS-stimulated production of this cytokine (Brandwein 1990). Other authors have shown leukotriene B4, which is produced by macrophages in response to IL-1b (Pruimboom et al 1994), can enhance production of IL-1b as well as IL-6 (RolaPleszczynski and Stankova 1992). 5-Lipoxygenase inhibitors inhibit the production of IL-1 and other cytokines that are involved in cartilage degradation (Schade et al 1992; Rainsford et al 1993). Overall, these observations suggest that there may be
opposing effects of products of the 5-LOX compared with COX-2 pathways in regulating the responses to IL-1. Bonta and Elliott (1992) further developed this theme and presented a so-called ‘‘heuristic’’ model (Figure 31.7) perhaps better described as a ‘‘working’’ model, integrating the negative regulatory effects of PGE2 on IL-1 and TNFa and stimulating effects on LTB4 on the production of these cytokines. With the identification of the individual isoenzymes of the cyclooxygenases, it has been possible to establish the effects of PGE2 derived from COX-1 or COX-2 on the production by macrophages of cytokines. Thus, Williams and Shacter (1997) showed that IL-6 production was stimulated by PGE2 derived from COX-2 but not from COX-1. It would therefore be expected
Figure 31.2 Cytokines and eicosanoids in arthritis
352
INFLAMMATION
Figure 31.3 Cytokine imbalance from an excess production and actions of proinflammatory cytokines over those of antiinflammatory cytokines in rheumatoid arthritis. Some modifications of this scheme probably occur in other rheumatic conditions. Modified from Feldman and Maini (2002)
that COX-2-selective drugs would reduce IL-6 and thus diminish the acute phase response from this cytokine as well as IL-1. Tumour necrosis factor a (TNFa) has been shown to increase PLA2 (Clark et al 1988) and to promote sustained expression of COX-2, which has been shown to be blocked by the COX-2 inhibitor NS-398 (Perkins and Kniss 1997). Paradoxically, TNFa downregulates LTB4 receptors in polymorphonuclear leucocytes (O’Flaherty et al 1991). Basic fibroblast growth factor (bFGF) also induced COX-2 (Kawaguchi et al 1995), as does plateletderived growth factor (PDGF; Goppelt-Struebe et al 2000) and TGFb (Tahara et al 1995). In addition to the central role of the NF-kB–1kB signalling pathway mediating these cytokine responses, NF-IL-6 (Sorli et al 1998) and a range of other pathways that interconnect the actions of proinflammatory cytokines (e.g. IL-5 control of 5-LOX, Courburn et al 1999) as well as by antiinflammatory cytokines (Alaaeddine et al 1999; Seitz et al 1994; Spanbrock et al 2001) control PGE2 and cGMP by 5-lipoxygenase metabolites (Harbon et al 1984; Stankova et al 1992; Oh-ishi et al 1996) and serve as negative and positive regulators, respectively, of IL-1, TNFa and other cytokines (Brandwein 1986, 1990; Zitnik et al 1993). Antiinflammatory cytokines (e.g. IL-10) may also affect NF-kB (Brennan 1999) and so indirectly stimulate COX-2-derived PGE2, which then causes negative regulation of proinflammatory cytokine production. Of particular interest in relation to the COX-2-mediated production of PGE2 in the neural pathways regulating pain production is that bradykinin, which is a wellknown stimulant of PG production, stimulates NF-kB activation and IL-1b gene expression (Pan et al 1996). IP receptors for PGI2 also express the precursor to the pain-eliciting substance P (preprotachykinin A) (Sugimoto et al 2000) and, since the PGE2 receptors EP2 and EP4 as well as IP receptors are upregulated upon initiation of pain, there appears to be a close link between
tachykinin and PGE2/PgI2 production from COX-2 and upregulation of their receptors. PHARMACOLOGICAL MODULATION OF CYTOKINE PRODUCTION BY COX AND LOX INHIBITORS A number of studies with conventional and experimental inhibitors of COX and/or LOX pathways have shown that these can regulate production of proinflammatory cytokines. Flurbiprofen has been found to inhibit TNFa and IL-1 but not IL-6 production in LPS-stimulated monocytes, whereas all three cytokines were inhibited in THP-1 and U-937 cell lines (Lozanski et al 1992). High concentrations of naproxen have been shown to reduce TNFa production in unstimulated human synoviocytes and IL-6 in LPS-stimulated cultures, but in the latter IL-1 production was increased (Ounissi-Benkalha et al 1996). Aspirin has been found to inhibit NFkB production and downstream regulated cytokines in chlamydia-treated endothelial cells (Tiran et al 2002). Lee et al (1988) showed that the experimental COX/ LOX inhibitor SKF 86002 [5-(4-pyridyl)-6-(4 fluorophenyl)2,3(dihydroimidazo(2,1b)-thiazoline] could inhibit production by activated human monocytes of IL-1 with an IC50 of 1.3 mM coincident with inhibition of COX and LOX products. Other LOX inhibitors, except for the antioxidant nordihydroguaiaretic acid (NDGA), were inactive. These studies were confirmed by Hartman et al (1995). A later study by the same group with an imidazole-pyrrole prodrug, SKF105809 [2-(4-methylsulphinyl phenyl)-3-(4-pyridyl)-6,7-dihydro-(5H)-pyrrolo-(1,2-a)imidazole], and its sulphite metabolite, SKF105561, inhibited COX and LOX activities in vitro and in vivo, exhibited antiinflammatory activity and reduced LPS-stimulated IL-1 production by human
CYTOKINES AND EICOSANOIDS IN ARTHRITIS
353
A
B
Figure 31.4 (A) The Th1/Th2 cytokine imbalance in various allergic states and rheumatic and other autoimmune diseases in pregnancy and HIV infection. In rheumatoid arthritis and other autoimmune conditions there is a predominant Th1 profile, while in systemic lupus erythematosus (SLE), allergic conditions and scleroderma there is a predominant Th2 profile. (B) The biological functions of Th1/Th2 balance and consequent effects in cellular and humoral immunity as they are mediated by the Th1 cytokines IL-2 and IFNg and the Th2 cytokines IL-4, IL-5, IL-10 and IL-13. From Miossec and van den Berg (1997), with permission
354
INFLAMMATION
CYTOKINES AND EICOSANOIDS IN ARTHRITIS
355
Table 31.6 Properties and functions of chemokines . . . .
Family of heparin-binding cytokines originally identified for involvement in chemotaxis—neutrophils (IL-8) and mononuclear cells Four families: identity based on juxtaposition of cysteine residues in the protein N-terminal region in form of C-, C–C, C–X–C, and C–X3–C C–X–C include IL-8 and growth-related oncogene-a (GRO) C–C monocyte chemoattractant proteins (MCP-1, -2, -3), macrophage inflammatory proteins (MIP-1a and MCP-2b) and another known as regulated upon activation, normal T cell and expressed as RANTES . Exert effects by binding to G-protein coupled receptors, e.g. CXCR, CCR . Some chemokines expressed by IL-1 and TNFa . Chemokine receptors Receptor type
Subtype
Chemokine specificity
CXC
CXCR1 CRCR2 CXCR3 CXCR4 CXCR5 HVS ECRF3 CCR1 CCR2 CCR3 CCR4 CCR5 CCR6 CCR7 CCR8 CMV US28 CX3CR1 D6 Duffy KSHV GPCR
IL-8, GCP-2 IL-8, GCP-2, GROa, GROb, GROg, ENA-78, NAP-2 IP-10, Mig SDF-1a, SDF-1b BCA-1 IL-8, GROa, NAP-2 MIP1a, RANTES, MCP-3, leukotactin-1, HCC-1, MPIF-1, MCP-4 MCP-1, MCP-3, MCP-4, MCP-2, MCP-5 Eotaxin, eotaxin-2, RANTES, MCP-3 TARC, MDC MIP-1a, MIP-1b, RANTES, MCP-2 LARC ELC, SLC I-309 MIP-1a, RANTES, MIP-1b, MCP-1 CX3C chemokine (fractalkaline) RANTES, MCP-1, MCP-3, MIP-1b RANTES, MCP-1; IL-8, NAP-2, GROa RANTES; MCP-1, I-309; IL-8, NAP-2
CC
CX3C CC chemokine binding proteins CC and CXC binding proteins
IL-8, interleukin-8; GRO, growth-related oncogene; NAP, neutrophil-activating protein; MCP, monocyte chemoattractant protein; MIP, macrophage inflammatory protein; ENA, epithelial cell-derived neutrophil activator; SDF, stromal cell-derived factor; GCP, granulocyte chemotactic protein; IP, interferon-inducible protein; Mig, monokine inducible by g-interferon (IFN-g).
Figure 31.6 Control of afferent and efferent pathways of pain mediation by aspirin, paracetamol and other analgesics. Anatomical structures from Fields (1987) with permission Figure 31.5 (opposite) (A) The major signal transduction pathways related to inflammation. TNFa, tumour necrosis factor-a; IkB, inhibitor of nuclear factor kB; Ub, ubiquitin; JAK, janus-activated kinase; STAT, signal transducer and activator of transcription; CaN, calcineurin; NF-ATc, nuclear factor of activated T cells; AP-1, activator protein 1; IL-1, interleukin-1; JNK, Jun N-terminal kinase; SR, steroid receptor. (B) Schema for signal transduction through the MAP kinase pathways. PKC, protein kinase C; EGF, epidermal growth factor; Grb2, growth factor receptor binding protein 2; MKK, MAP kinase kinase; ERK, extracellular signal-regulated kinase; TRE, TPA-responsible element; LPS, lipopolysaccharide; UV, ultraviolet; PAK, p42-associated kinase; ATF-2, activating transcription factor 2; SAP-1, stress-activated protein 1; TCF, ternary complex factor; SRE, serum response element. From Firestein and Manning (1999) with permission
356
INFLAMMATION CONCLUSIONS It is apparent from this brief review that there are multiple sites for the regulation by cytokines of eicosanoid metabolism, involving a wide range of actions by proinflammatory and antiinflammatory cytokines. Moreover, products of eicosanoid oxidative metabolism can influence or regulate the production or actions of cytokines. Many of these pathways have implications in arthritic diseases, although some of the components of the cytokine–eicosanoid systems require further elucidation to establish their significance in individual arthritic conditions. While there are further details of the various pathways involved in cytokine–eicosanoid interactions that need to be determined, it is clear that the interplay between cytokines and eicosanoids has high potential significance in arthritic conditions. REFERENCES
Figure 31.7 Heuristic model of some aspects of balanced regulation of mediator producing functions of the macrophage. Dotted lines represent inhibitory influences. Endogenous PGE2 controls the synthesis of LTB4, IL-1 and TNFa. The latter two are important stimulators of PGE2, which thus is their self-induced inhibitor. This balanced regulation represents a homeostatic process-loop system. Removal of PGE2 by NSAIDs causes a breakdown of the homeostasis and accordingly results in uncontrolled production of the triad of harmful mediators. This oversimplified model, in common with other models, is a poor replica of the reality, in which many other factors contribute to homeostasis and the pathological or iatrogenic breakdown thereof. From Bonta and Elliott (1992), with permission
monocytes (IC50 2 mM) in the range in which COX and LOX products were reduced (Marshall et al 1991; Griswold et al 1991). Other authors have shown that tenidap, tepoxalin and licofelone, which are inhibitors of COX and 5-LOX activities, also reduce production of IL-1 and other pro-inflammatory cytokines (Mattey et al 1994; Kazmi et al 1995; Ritchie et al 1995; Bertolini et al 2001; Jovanovic et al 2001) and by flavonoid and other COX/ LOX-inhibiting antioxidants (Eugui et al 1994; Zhao et al 1999). Some selective effects of experimental 5-LOX inhibitors on IL-1 and/or TNFa production have been reported. Among these are some conventional antioxidants, NDGA, butylated hydroxyanisole (BHA), tetrahydropapaveroline (THP) and 10,11-dihydroxyapomorphine (DHA), while generalized antioxidants devoid of LOX activity did not affect cytokine production (Eugui et al 1994). A range of 5-LOX inhibitors, including the five lipoxygenase activating protein (FLAP) inhibitor, MK-886, and several direct-acting inhibitors, were found to be potent inhibitors of IL-1 production by human synovial tissues from patients with inflammatory arthropathies in organ culture (Rainsford et al 1993). 5-LOX inhibitors also decrease both the lymphocyte mRNAs and production of IL-2 and IL-6 (Dornand et al 1987; Dornand and Gerber 1991), probably by affecting activation of protein kinase C (Dornand et al 1987) and phosphoinositide pathways, and increases intracellular calcium (Dornand and Gerber 1991). These effects may account for reduced proliferation of T cells (Hata et al 1987; Schultz and Altom 1988; Liu et al 1989; Olsen et al 1992). LOX inhibitors have also been shown to inhibit mitogenic activity of spleen cell cultures, which, interestingly, was reversed by the addition of the putative tumour cell inhibitory intermediate of LOX activity, 13-hydroxyoctadecadienoic acid (13-HODE) but not by LTB4, LTC4 or 15-HETE (Elekes et al 1993). In DA-1 cells, however, COX/LOX inhibitors do affect IL-3-induced proliferation (Barton et al 1988). Other cytokines are variably affected by LOX or COX inhibitors. Thus, the antiinflammatory cytokines IL-4 and IL-10 regulate IL-8 production in rheumatoid synovial fluid mononuclear cells that produce LTB4 but not 15-HETE (Deleuran et al 1994).
Aarden LA, Brunner TK, Cerottini J-C et al (1997) Revised nomenclature for antigen-nonspecific T-cell proliferation and helper factors. J Immunol, 123, 2928–2929. Akarasereenont P, Techatrisak K, Chottewuttakorn S et al (1999) The induction of cyclooxygenase-2 in IL-1b-treated endothelial cells is inhibited by prostaglandin E2 through cAMP. Mediators Inflammation, 8, 287–294. Alaaeddine N, Di Battista JA, Pelletier J-P et al (1999) Inhibition of tumor necrosis factor a-induced prostaglandin E2 production by the antiinflammatory cytokines interleukin-4, interleukin-10, and interleukin13 in osteoarthritic synovial fibroblasts. Arthrit Rheumat, 42, 710–718. Amin AH, Attur M, Patel RN et al (1997) Superinduction of cyclooxygenase-2 activity in human osteoarthritis-affected cartilage. J Clin Invest, 99, 1231–1237. Barton BE, WoldeMussie E and Wheeler L (1988) The role of arachidonic acid metabolism in IL-3-induced proliferation. Immunopharmacol Immunotoxicol, 10, 35–52. Belt AR, Baldassare JJ, Molnar M et al (1999) The nuclear transcription factor NF-kB mediates interleukin-1b-induced expression of cyclooxygenase-2 in human myometrial cells. Am J Obstet Gynecol, 181, 359–366. Bertolini A, Ottani A and Sandrini M (2001) Dual acting antiinflammatory drugs: a reappraisal. Pharmacol Res, 44, 437–450. Bonta IL and Elliott GR (1992) Non-steroidal anti-inflammatory drugs and the augmented lipoxygenase pathway: conceivable impact on joint condition. In Rainsford KD and Velo G-P (eds), Side Effects of Antiinflammatory Drugs, vol 3. Dordrecht, Kluwer Academic, 269–274. Brandwein SR (1986) Regulation of interleukin-1 production by mouse peritoneal macrophages. Effects of arachidonic acid metabolites, cyclic nucleotides, and interferons. J Biol Chem, 261, 8624–8632. Brandwein SR (1990) Differential regulation of soluble interleukin-1 release and membrane expression by pharmacologic agents. Agents Act, 30, 381–392. Brennan FM (1999) Interleukin 10 and arthritis. Editorial. Rheumatology, 38, 293–297. Cao C, Matsumura K, Yamagata K et al (1996) Endothelial cells of the rat brain vasculature express cyclooxygenase-2 mRNA in response to systemic interleukin-1b: a possible site of prostaglandin synthesis responsible for fever. Brain Res, 733, 263–272. Clark MA, Chen M-J, Crooke ST and Bomolaski JS (1988) Tumor necrosis factor (cachectin) induces phospholipase A2 activity and synthesis of a phospholipase A2-activating protein in endothelial cells. Biochem J, 250, 125–132. Cowburn AS, Holgate ST, Sampson AP (1999) IL-5 increases expression of 5-lipoxygenase-activating protein and translocated 5-lipoxygenase to the nucleus in human blood eosinophils. J Immunol, 163, 456–465. Cox A and Duff GW (1996) Cytokines as genetic modifying factors in immune and inflammatory diseases. J Pediat Endocrinol Metabol, 9, 129–132. Crofford LJ (1997) COX-1 and COX-2 tissue expression: implications and predictions. J Rheumatol, 24, 15–19. Crofford LJ, Wilder RL, Ristima¨ki AP et al (1994) Cyclooxygenase- and -2 expression in rheumatoid synovial tissues. Effects of interleukin-1b, phorbol esters and corticosteroids. J Clin Invest, 93, 1095–1101.
CYTOKINES AND EICOSANOIDS IN ARTHRITIS Cua DJ, Sherlock J, Chen Y et al (2003) Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation in the brain. Nature, 421, 744–748. Dayer J-M, Evequoz V, Zavadil-Grob C et al (1987) Effect of synthetic calcium pyrophosphate and hydroxyapatite crystals on the interaction of human mononuclear cells with chondrocytes, synovial cells, and fibroblasts. Arthrit Rheumat, 30, 1372. Deleuran B, Iversen L, Kristensen M et al (1994) Interleukin-8 secretion and 15-lipoxygenase activity in rheumatoid arthritis: in vitro antiinflammatory effects by interleukin-4 and interleukin-10, but not by interleukin-1 receptor antagonist protein. Br J Rheumatol, 33, 520–525. Diaz A, Reginato AM and Jimenez SA (1992) Alternative splicing of human prostaglandin G/H synthase mRNA and evidence of differential regulation of the resulting transcripts by transforming growth factorb1, interleukin 1b, and tumor necrosis factor a. J Biol Chem, 267, 10816–10822. Di Giovine FS, Malawista SE, Thorton E and Duff GW (1991) Urate crystals stimulate production of tumor necrosis factor a from human blood monocytes and synovial cells. J Clin Invest, 78, 1375. Dornand J, Sekkat C, Mani JC and Gerber M (1987) Lipoxygenase inhibitors suppress IL-2 synthesis: relationship with rise of [Ca++]i and the events dependent on protein kinase C activation. Immunol Lett, 16, 101–106. Dornand J and Gerber M (1991) Mechanisms of IL-2 production impairment by lipoxygenase inhibitors in activated Jurkat cells. J Lipid Mediat, 4, 23–38. Duff GW, Atkins E and Malawista SE (1983) The fever of gout: urate crystals activate endogenous pyrogen reduction from human and rabbit mononuclear phagocytes. Trans Assoc Am Physicians, 96, 234. Duff GW (1998) Molecular genetics of cytokines. In Thomson A (ed.), The Cytokine Handbook, 3rd edn. San Diego, Academic Press, 21–33. Dumonde DC, Wolstencroft RA, Panayi GS et al (1969) ‘‘Lymphokines’’: non-antibody mediators of cellular immunity generated by lymphocyte activation. Nature, 224, 38–42. Elekes E, Jakobs D and Schade FU (1993) Suppression of endotoxin mitogenicity of spleen cells by lipoxygenase inhibitors and its reversal by 13-hydroxyoctadecadienoic acid. FEMS Immunol Med Microbiol, 6, 13–20. Eugui EM, DeLustro B, Rouhafza S et al (1994) Some antioxidants inhibit, in a coordinate fashion, the production of tumor necrosis factor-a, IL-b, and IL-6 by human peripheral mononuclear cells. Int Immunol, 6, 409–422. Feldmann M and Maini RN (2002) Discovery of TNF-a as a therapeutic target in rheumatoid arthritis: preclinical and clinical studies. Joint Bone Spine, 69, 12–18. Fields HL (1987) Pain. New York, McGraw Hill. Firestein GS and Manning AM (1999) Signal transduction and transcription factors in rheumatic disease. Arthrit Rheum, 42, 609–621. Goppelt-Struebe M, Rehm M and Schaefers H-J (2000) Induction of cyclooxygenase-2 by platelet-derived growth factor (PDGF) and its inhibition by dexamethasone are independent of NF-kB/IkB transcription factors. Naunyn-Schmiedeberg’s Arch Pharmacol, 361, 636–645. Griswold DE, Marshall PJ, Lee JC et al (1991) Pharmacology of the pyrroloimidazole, SK&F 105809-II. Antiinflammatory activity and inhibition of mediator production in vivo. Biochem Pharmacol, 42, 825– 831. Guerne PA, Terkeltaub RA, Zuraw B and Lotz M (1989) Inflammatory microcrystals stimulate interleukin-6 production and secretion by human monocytes and synoviocytes. Arthrit Rheumat, 32, 1443. Harada Y, Hatanaka K, Kawamura M et al (1996) Role of prostaglandin H synthase-2 in prostaglandin E2 formation in rat carrageenin-induced pleurisy. Prostaglandins, 51, 19–33. Harbon S, Tanfin Z, Leiber D et al (1984) Selective interactions of cyclooxygenase and lipoxygenase metabolites with the cyclic AMP and cyclic GMP systems in the myometrium. Adv Cycl Nucleotides Protein Phosphoryl Res, 17, 639–649. Hartman DA, Ochalski SJ and Carlson RP (1995) The effects of antiinflammatory and antiallergic drugs on cytokine release after stimulation of human whole blood by lipopolysaccharide and zymosan A. Inflamm Res, 44, 269–274. Hata S, Sugama K, You-Li Z et al (1987) Synthesis of sterols and 5lipoxygenase products are required for the G1–S phase transition of interleukin-2-dependent lymphocyte proliferation. Microbiol Immunol, 31, 1231–1244.
357
Hinson RM, Williams JA and Shacter E (1996) Elevated interleukin 6 is induced by prostaglandin E2 in a murine model of inflammation: possible role of cyclooxygenase-2. Proc Natl Acad Sci USA, 93, 4885– 4890. Hulkower KI, Otis ER, Li J et al (1997) Induction of prostaglandin H synthase-2 and tumor necrosis factor-a in human amniotic WISH cells by various stimuli occurs through distinct intracellular mechanisms. J Pharmacol Exp Therap, 280, 1065–1074. Jovanovic DV, Fernandes JC, Martel-Pelletier J et al (2001) In vivo dual inhibition of cyclooxygenase and lipoxygenase by ML-3000 reduces the progression of experimental osteoarthritis: suppression of collagenase 1 and interleukin-1b synthesis. Arthrit Rheumat, 44, 2320–2330. Kang RY, Freire-Moar J, Sigal E and Chu C-Q (1996) Expression of cyclooxygenase-2 in human and an animal model of rheumatoid arthritis. Br J Rheumatol, 35, 711–718. Karin M (1998) The NFkB activation pathway—its regulation and role in inflammation and cell survival. Cancer J Sci Am, 4, S92–S99. Kazmi SM, Plante RK, Visconti V et al (1995) Suppression of NF-kB activation and NF kB-dependent gene expression by tepoxalin, a dual inhibitor of cyclooxygenase and 5-lipoxygenase. J Cell Biochem, 57, 299–310. Kawaguchi H, Pilbeam CC, Gronowicz G et al (1995) Transcriptional induction of prostaglandin G/H synthase-2 by basic fibroblast growth factor. JClin Invest, 96, 923–930. Kennard EA, Zimmerman PD, Friedman CI and Kniss DA (1995) Interleukin-1b induces cyclooxygenase-2 in cultured human decidual cells. Am J Reprod Immunol, 34, 65–71. Kunkel SL, Chensue SW and Phan SH (1986) Prostaglandins as endogenous mediators of interleukin-1 production. J Immunol, 136, 186–192. Kuwata H, Nakatani Y, Murakami M and Kudo I (1998) Cytosolic phospholipase A2 is required for cytokine-induced expression of type IIA secretory phospholipase A2 that mediates optimal cyclooxygenase2-dependent delayed prostaglandin E2 generation in rat 3Y1 fibroblasts. J Biol Chem, 273, 1733–1740. Lee JC, Griswold DE, Votta B and Hanna N (1988) Inhibition of monocyte IL-1 production by the anti-inflammatory compound, SK&F 86002. Int J Immunopharmacol, 10, 835–843. Lewis GP and Gordon D (1985) Cytokines as potential targets for antirheumatic therapy: a pharmacologist’s view. Br J Rheumatol, 24(suppl1), 29–33. Lozanski G, Ballou SP, Kushner I (1992) Effect of flurbiprofen on cytokine productions by human monocytes and U-937 and THP-1 cell lines. J Rheumatol, 19, 921–926. Liu DS, Liew FY and Rhodes J (1989) Immunoregulatory properties of novel specific inhibitors of 5-lipoxygenase. Immunopharmacology, 17, 1–9. Maier JAM, Hla T and Maciag T (1990) Cyclooxygenase is an immediate– early gene induced by interleukin-1 in human endothelial cells. J Biol Chem, 265, 10805–10808. Malinin NL, Boldin MP, Kovalenko AV and Wallach D (1997) MAP3Krelated kinase involved in NF-kB induction by TNF, CD95 and IL-1. Nature, 385, 540. Marshall PJ, Griswold DE, Breton J et al (1991) Pharmacology of the pyrroloimidazole, SK&F 105809—I. Inhibition of inflammatory cytokine production and of 5-lipoxygenase- and cyclooxygenasemediated metabolism of arachidonic acid. Biochem Pharmacol, 42, 813–824. Mattey DL, Evans E and Dawes PT (1994) The effects of Tenidap on cytokine-induced proliferation of human synovial fibroblasts in vitro. Ann Rhem Dis, 53, 250–255. Milton AS (1982) Prostaglandins in fever and the mode of action of antipyretic drugs. In Milton AS (ed.), Pyretics and Antipyretics. Berlin, Springer-Verlag, 257–303. Miossec P and van den Berg W (1997) Th1/Th2 cytokine balance in arthritis. Arthrit Rheum, 40, 2105–2115. Murakami M, Kambe T, Shimbara S et al (1999) Different functional aspects of the group II subfamily (types IIA and V) and type X secretory phospholipase A2s in regulating arachidonic acid release and prostaglandin generation. J Biol Chem, 274, 31435–31444. Nakao S, Ogata Y, Shimizu-Sasaki E et al (2000) Activation of NFkB is necessary for IL-1b-induced cyclooxygenase-2 (COX-2) expression in human gingival fibroblasts. Mol Cell Biochem, 209, 113–118. Nantel F, Denis D, Gordon R et al (1999) Distribution and regulation of cyclooxygenase-2 in carrageenan-induced inflammation. Br J Pharmacol, 128, 853–859.
358
INFLAMMATION
O’Flaherty JT, Rossi AG, Redman JF et al (1991) Tumour necrosis factora regulates expression of receptors for formyl-methionyl-leucylphenylalanine, leukotriene B4 and platlet-activating factor. J Immunol, 147, 3842–3847. Oh-ishi S, Utsunomiya I, Yamamoto T et al (1996) Effects of prostaglandins and cyclic AMP on cytokine production in rat leukocytes. Eur J Pharmacol, 300, 255–259. Olsen SC, Atluru D, Atluru S et al (1992) Inhibition of equine mononuclear cell proliferation and leukotriene B4 synthesis by a specific 5-lipoxygenase inhibitor, A-63162. Am J Vet Res, 53, 1015– 1018. O’Neill LAJ (1995) Toward an understanding of the signal transduction pathways for interleukin-1. Biochim Biophys Acta, 1266, 31–44. Oppenheim JJ and Feldman M (eds) (2000) A Compendium of Cytokines and Other Mediators of Host Defense. New York, Academic Press. Oppenheim JJ (2001) Cytokines, past, present and future. Int J Hematol, 74, 3–8. O’Sullivan MG, Huggins EM, Meade EA et al (1992) Lipopolysaccharide induces prostaglandin H synthase-2 in alveolar macrophages. Biochem Biophys Res Commun, 187, 1123–1127. Ounissi-Benkalha H, Pelletier JP, Tardif G et al (1996) In vitro effects of 2 antirheumatic drugs on the synthesis and expression of proinflammatory cytokines in synovial membranes from patients with rheumatoid arthritis. J Rheumatol, 23, 16–23. Pan ZK, Zuraw BL, Lung C-C et al (1996) Bradykinin stimulates NF-kB activation and interleukin 1b gene expression in cultured human fibroblasts. J Clin Invest, 98, 2042–2049. Perkins DJ and Kniss D (1997) Tumor necrosis factor-a promotes sustained cyclooxygenase-2 expression: attenuation by dexamethasone and NSAIDs. Prostaglandins, 54, 727–743. Perrier S, Coussediere C, Dubost JJ et al (1998) IL-1 receptor antagonist (IL-1RA) gene polymorphism in Sjogren’s syndrome and rheumatoid arthritis. Clin Immunol Immunopathol, 87, 309–313. Pomini F, Caruso A and Challis JRG (1999) Interleukin-10 modifies the effects of interleukin-1b and tumor necrosis factor-a on the activity and expression of prostaglandin H synthase-2 and the NAD+-dependent 15-hydroxyprostaglandin dehydrogenase in cultured term human villous trophoblast and chorion trophoblast cells. J Clin Endocrinol Metab, 84, 4645–4651. Pouliot M, James MJ, McColl SR et al (1998) Monosodium urate microcrystals induce cylooxygenase-2 in human monocytes. Blood, 91, 1769–1776. Pruimboom WM, van Dijk JA, Tak CJ et al (1994) Interactions between cytokines and eicosanoids: a study using human peritoneal macrophages. Immunol Lett, 41, 255–260. Rainsford KD, Ying C and Smith F (1993) Selective effects of some 5lipoxygenase inhibitors on synovial interleukin-1 (IL-1) production compared with IL-1 synthesis inhibitors. Agents Act, (Special Conference Issue) 39, C186–C188. Rauk PN and Chiao J-P (2000) Interleukin-1 stimulates human uterine prostaglandin production through induction of cyclooxygenase-2 expression. Am J Reprod Immunol, 43, 152–159. Ritchie DM, Argentieri DC, Aparicio BL et al (1995) Cytokinemodulating activity of tepoxalin, a new potential antirheumatic. Int J Immunopharmacol, 17, 805–812. Rola-Pleszczynski M, Turcotte S and Gingras D (1990) Enhanced TNF production and cytotoxic activity of human monocyte primed with PAF: post-receptor events. In Openheim JJ, Powanda MC, Kluger M and, Dinarello CA (eds), Molecular and Cellular Biology of Cytokines. New York, Wiley–Liss, 105–110. Rola-Pleszczynski M and Stankova J (1992) Leukotriene B4 enhances interleukin-6 (IL-6) production and IL-6 messenger RNA accumulation in human monocytes in vitro transcriptional and posttranscriptional mechanisms. Blood, 80, 1004–1011. Schade FU, Engel R and Jakobs D (1992) Lipoxygenase inhibitors but not site-specific 5-lipoxygenase blockers protect against endotoxic shock and inhibit production of tumor necrosis factor. Eicosanoids, 5(suppl), S45–S47. Schmidlin F, Loeffler S, Bertrand C et al (2000) PLA2 phosphorylation and cyclooxygenase-2 induction, through p38 MAP kinase pathway, is involved in the IL-1b-induced bradykinin B2 receptor gene transcription. Naunyn Schmiedeberg’s Arch Pharmacol, 361, 247–254. Schultz RM and Altom MG (1988) Modulation of interleukin-1 activity on murine thymocytes by various inhibitors of arachidonic acid oxygenation. Immunopharmacol Immunotoxicol, 10, 21–34.
Scott KF, Bryant KJ and Bidgood MJ (1999) Functional coupling and differential regulation of the phospholipase A2-cyclooxygenase pathways in inflammation. J Leukocyte Biol, 66, 535–541. Seitz M, Loetscher P, Dewald B et al (1994) Production of interleukin-1 receptor antagonist, inflammatory chemotactic proteins and prostaglandin E by rheumatoid and osteoarthritic synoviocytes – regulation by IFN-gamma and IL-4. J Immunol, 152, 2060–2065. Serhan CN, Haeggstrom JZ and Leslie CC (1996) Lipid mediator networks in cell signalling: update and impact of cytokines. FASEB J, 10, 1147–1158. Siegle I, Klein T, Backman JT et al (1998) Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue. Arthrit Rheumat, 41, 122–129. Smith WL and DeWitt DL (1995) Biochemistry of prostaglandin endoperoxide H synthase-1 and synthase-2 and their differential susceptibility to nonsteroidal anti-inflammatory drugs. Semin Nephrol, 15, 179–194. Spanbroek R, Hildner M, Kohler A et al (2001 IL-4 determines eicosanoid formation in dendritic cells by down-regulation of 5-lipoxygenase and up-regulation of 15-lipoxygenase expression. Proc Natl Acad Sci USA, 98, 5152–5157. Sorli CH, Zhang H-J, Armstrong MB et al (1998) Basal expression of cyclooxygenase-2 and nuclear factor-interleukin 6 are dominant and coordinately regulated by interleukin 1 in the pancreatic islet. Proc Natl Acad Sci USA, 95, 1788–1793. Stankova J, Gagnon N, Rola-Pleszczynski M (1992) Leukotriene B4 augments interleukin-2 receptor-beta (IL-2R beta) expression and IL-2R beta-mediated cytotoxic response in human peripheral blood lymphocytes. Immunology, 76, 258–263. Stylianou E, O’Neill LAJ, Rawlinson L et al (1992) Interleukin 1 induces NF-kB through its type I but not its type II receptor by lymphocytes. J Biol Chem, 267, 15836–15841. Sugimoto Y, Narumiya S and Ichikawa A (2000) Distribution and function of prostanoid receptors: studies from knockout mice. Progr Lipid Res, 39, 289–314. Tada K, Murakami M, Kambe T and Kudo I (1998) Induction of cyclooxygenase-2 by secretory phospholipase A2 in nerve growth factor-stimulated rat serosal mast cells is facilitated by interaction with fibroblasts and mediated by a mechanism independent of their enzymatic functions. J Immunol, 161, 5008–5015. Tahara M, Tasaka K, Masumoto N et al (1980) Expression of messenger ribonucleic acid for epidermal growth factor (EGF), transforming growth factor-a (TGFa), and EGF receptor in human amnion cells: possible role of TGFa in prostaglandin E2 synthesis and cell proliferation. J Clin Endocrinol Metab, 80, 138–146. Terkeltaub RA, Zachariae C, Santoro D et al (1991) Monocyte-derived neutrophil chemotactic factor/interleukin-8 is a potential mediator of crystal-induced inflammation. Arthrit Rheumat, 34, 894. Thomson A (1998) The Cytokine Handbook, 3rd edition. A. Thomson (Ed). New York, London, Academic Press. Tiran A, Gruber HJ, Graier WF et al (2002) Aspirin inhibits chlamydia pneumoniae-induced nuclear factor-kB activation, cytokine expression and bacterial development in human endothelial cells. Arterioscler Thromb Vasc Biol, 22, 1075–1080. Van den Berg WB (2002) Is there a rationale for combined TNF and IL-1 blocking in arthritis? Clin Exp Rheumatol, 20, S21–S25. Vilce˘k J (1998) The cytokines: an overview. In Thomson A (ed.), The Cytokine Handbook, 3rd edn. San Diego, Academic Press, 1–20. Wadleigh DJ, Reddy ST, Kopp E et al (2000) Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. J Biol Chem, 275, 6259–6266. Watford WT and O’Shea JJ (2003) A case of mistaken identity. Nature, 421, 706–707. Wilborn J, DeWitt DL and Peters-Golden M (1995) Expression and role of cyclooxygenase isoforms in alveolar and peritoneal macrophages. Am J Physiol, 268, L294–L301. Williams JA and Shacter E (1997) Regulation of macrophage cytokine production by prostaglandin E2. J Biol Chem, 272, 25693–25699. Zhao J, Sharma Y and Agarwal R (1999) Significant inhibition by the flavonoid antioxidant silymarin against 12-O-tetradecanoylphorbol 13acetate-caused modulation of antioxidant and inflammatory enxymes, and cyclooxygenase 2 and interleukin-1a expression in SENCAR mouse epidermis; implications in the prevention of stage 1 tumor promotion. Mol Carcinogen, 26, 321–333. Zitnik RJ, Zheng T and Elias JA (1993) camp inhibition of interleukin-1induced interleukin-6 production by human lung fibroblasts. Am J Physiol, 264, L253–L260.
Section Six Circulatory System
32 Perspectives and Clinical Significance of Eicosanoids in the Circulatory System The Late James B. Lee University at Buffalo, School of Medicine & Biomedical Sciences, Buffalo, NY, USA
BACKGROUND The story of the eicosanoids, from their beginnings in 1930 to the present, represents an outstanding example of a progression of biomedical scientific knowledge taking place over the last 73 years. This was initially marked in 1930 by the discovery that human semen stimulated isolated human uterine strips in vitro (Kurzrok and Lieb 1930). In a subsequent classical series of early experiments, von Euler demonstrated that human semen also contained a vasodepressor component and, using intricate combinations of available chemical and pharmacological techniques, determined that, in man and certain animals, the substances were unique fatty acids and named them prostaglandins (von Euler 1934, 1935, 1936). Skilfully utilizing a combination of gas chromatography and mass spectrometry, Bergstrom and his coworkers subsequently determined the structures of PGE and PGF (Bergstrom and Sjovall 1960a, 1960b; Bergstrom et al 1962). Some 33 years later, after the findings of Kurzrok and Lieb and unaware of these early findings, we independently discovered marked vasodepressor activity in the rabbit kidney medulla (Lee et al 1962, 1963). Unlike the previous autopharmacological screening experiments searching for biological activity in various tissues, we were interested in the kidney as a source of vasodepressor activity responsible for maintaining normotension, a lack of which might be a causal factor in hypertension. Since the eicosanoids involve so many diverse biological systems, it is only natural that various investigators proceeded to pursue their studies according to their own particular interests and perspectives. One of these was the author’s involvement in pursuing the hypothesis that human hypertension may not be solely the result of overactive pressor pathways, such as the renal renin–angiotensin axis and the sympathetic nervous system, but may be the result of a deficiency of renal vasodepressor agents, allowing vasopressor systems to raise blood pressure even at normal levels of activity: the so-called ‘‘antihypertensive function of the kidney’’ or the ‘‘deprivation hypothesis’’. Since much of the material regarding this function and PGs is now well documented in the literature, I shall ask the reader’s indulgence to present, in narrative form, a personalized account of one individual’s interest and experiences in searching for biochemical and physiological mechanisms, which might underlie the antihypertensive function of the kidney and, thus, the circulatory system. This account emphasizes the early literature in order to give a perspective of how past research relates to the sophisticated, updated material presented in subsequent chapters on the circulatory system in this volume. Because of this, references The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
have been arbitrarily selective and have been kept to a minimum. Unfortunately, it was not possible to include more exhaustive earlier citations, revealing the many important contributions of others to the circulatory system. These citations can be obtained from references provided in this chapter as well as reviews by von Euler and Eliasson (1967), Attallah and Lee (1982) and Sparks (1971). THE DISCOVERY In the fall of 1958, I was a resident making medical rounds at the Pennsylvania Hospital in Philadelphia, with the Chief of Service, Dr Garfield Duncan, Professor of Medicine at Jefferson Medical College and an internationally recognized authority in metabolic diseases, particularly diabetes mellitus. Dr Duncan’s lesser-known interest was hypertension, since so many exophthalmic balding young men were succumbing to malignant hypertension and uraemia as a complication of their diabetes. On one such occasion, Dr Duncan mentioned that he was attracted to a little-known concept that hypertension might in the final analysis be the result, at least in part, of a renal antihypertensive agent or mechanism. This deficiency hypothesis was quite attractive to me, since in other systems where a causal agent for an expressed excess was found it was inevitably from deficiency. Thus, pernicious anaemia presents as a malignancy with excessive megaloblastic cells in marrow and blood, peripheral neuropathy, inanition and demise. Rather than a primary excess of malignant red cells, a deficiency of vitamin B12 and intrinsic factor preventing maturation of primitive cells and an accumulation of precursor cells was found to be the underlying causal factor. In 1960, as a fellow of Dr George F. Cahill in Dr George W. Thorn’s laboratory at the Peter Bent Brigham Hospital in Boston, I was studying the intermediary metabolism of 14C-labelled substrates in slices of rabbit kidney cortex and medulla in vitro. At that time, a colleague of mine, Dr Roger B. Hickler, was studying angiotensinase activity on the blood pressure of the tracheotomized, pentobarbital- and pentolinium-treated rat. Mindful of Dr Duncan’s comments, it seemed like a good opportunity to test the antihypertensive theory in this preparation. I suggested to Dr Hickler that we see if there was any fall in blood pressure following injection of media from our rabbit kidney cortical and medullary slices. We were taken aback and pleasantly surprised to observe that a sharp drop in blood pressure actually did occur with the medullary media, which
362
CIRCULATORY SYSTEM
gradually returned to normal over a period of 10 min (Figure 32.1). Renal cortical media resulted in the previously well-known rise in blood pressure attributed to its renin content. Extracts of rabbit spleen, liver, intestine and lung were without effect. Initial enthusiasm gave way to the realization that many factors could lower rat blood pressure (acetylcholine, bradykinin, ammonium ion, acid pH, nucleotides, etc.). In fact, one of our colleagues passing through the laboratory believed we could obtain the same effect with ‘‘sterile milk’’. To exclude such factors, it was found that they all produced a spiking fall in blood pressure with a rapid return to normal. In addition, the active material was eluted in the internal volume by Sephadex chromatography and separated from short-acting nucleotides by DEAE Sephadex chromatography. The blood pressure effect was dose-dependent and the active component(s) was dialysable, ethanol-soluble and resistant to peptide hydrolysis. It was concluded that rabbit kidney medulla possessed a unique non-protein ethanol-soluble substance(s) of low molecular weight (54500) with potent vasodepressor activity. Prior to publication of these findings (Lee et al 1962, 1963), a review of the literature revealed extremely interesting background information. First, it was observed that the original research forming the basis of the antihypertensive renal function was published in 1947 by Braun-Mendendez and the same U.S. von Euler who discovered and characterized the PGs (Braun-Mendendez and von Euler 1947). These individuals observed that bilateral nephrectomy in the rat was associated with hypertension (so-called renoprival hypertension), which obviously could not be ascribed to any renal prohypertensive renin mechanisms. Although Sokabe and Grollman (1962) believed that the antihypertensive activity resided in the renal cortex, Muirhead et al, in a long series of experiments, concluded that this activity was renomedullary in origin and that it was a neutral lipid. These experiments have been summarized and reviewed in detail by Muirhead et al (1967, 1977). ISOLATION, CHARACTERIZATION AND IDENTIFICATION With the unflagging support of Dr John F. Stapleton of Georgetown University School of Medicine, I accepted a position in 1962 at the Saint Vincent Hospital Research Foundation in Worcester, MA, USA. By a remarkable coincidence, a former college and graduate school classmate, Dr Benjamin Covino, had become Medical Director of Astra Pharmaceuticals in Worcester. He immediately became interested in our early studies and together we decided to isolate and determine the structure of the responsible compound(s) and to characterize their mechanisms of hypotensive action. It soon became obvious that, although vasodepressor activity could be observed with one rabbit medulla, the material was so potent that it would take kilograms of purified compound to chemically identify and physiologically characterize the responsible agents. Accordingly, 4 kg quick-frozen rabbit kidney medullas were initially obtained from Pel-Freeze Biologicals, Rogers, Arkansas, USA. At that time, over 5000 rabbits were sacrificed daily by Pel-Freeze for human frozen food consumption. Many rabbit tissues, including whole kidneys, were immediately frozen for scientific research. By ingeniously using an olive pitter, the Pel-Freeze Corporation was able to excise, immediately freeze and ship the renal medullas (4–10 kg amounts) in frozen food containers. When it is realized that one renal papilla weighs 300–500 mg, it is evident that this contribution by the company was quite a feat and, as you can imagine, greatly appreciated. The medullas were thawed and homogenized in 5 mM Na2HPO4. Following centrifugation, the filtrate was concentrated
Figure 32.1 Sustained vasodepressor effect in the pentolinium-treated rat following incubation with slices of rabbit renal medulla. From Lee et al (1963). Reproduced with permission from Lippincott Williams & Wilkins
by vacuum distillation. It was observed that vasodepressor activity was in the organic phase (chloroform, benzene and ethyl acetate) following acidification (pH 3) of the concentrated filtrate but not following alkalization (pH 9), the vasodepressor activity remaining in the aqueous phase. However, activity remained in the aqueous phase at pH 3 when partitioned against petroleum ether. From these observations, it was concluded that the active material was most likely a relatively polar acidic lipid. The experimental details of the subsequent isolation and identification of the active acidic polar lipids have been previously published (Lee et al 1965). Suffice it to say that it was initially accomplished by gradient elution DEAE-cellulose chromatography with Na2HPO4 and NaH2PO4. The overall purification flow sheet for relatively moderate amounts of medulla (4 kg) is given in Figure 32.2. An infrared pattern of medullin gave evidence of the existence of carboxyl and methylene groups, hydroxyl groups and trans-ethylene bonds. At this time, Dr Emil R. Smith (Smith et al 1964), a colleague of Dr Covino, observed that our material could be prostaglandin in nature, since he had read the early work of Bergstrom and Sjovall (1960a, 1960b) on the isolation and identification of PGE and PGF in the Swedish literature. This was the first time I had learned of the existence of these compounds. The relationship to PGs was heightened by common features of chromatography, infrared and ultraviolet spectra of medullin and PGs. Thus, our final isolations of medullin compound 1 and compound 2 were believed to be prostaglandins (Lee et al 1965): compound 1 was thought to be PGF, since it stimulated non-vascular smooth muscle; compound 2 possessed both vasodepressor and non-vascular smooth muscle stimulatory activity similar to PGE (Figure 32.3). However, medullin was much less polar, with deeply iodine-staining attributes, and therefore was an unknown PG and called medullin from its origin in the medulla (Figure 32.3). Confirmation of this was obtained by thin-layer chromatography of the mobilities of medullin and PGE1 (kindly donated by Dr Sune Bergstrom), which showed medullin to be much less polar than PGE1. From additional isolates of large amounts of medulla (10 kg), a final isolation of these three compounds was made in collaboration with Dr Bertil Takman, Director of Medicinal Chemistry at the Astra Pharmaceutical Company. To our dismay, these three final PG isolates were inadvertently misplaced and lost at a collaborating mass spectrometer facility, necessitating a retreat to square one and a re-extraction and purification from an additional 10 kg of frozen rabbit kidney medulla. The extraction of 10 kg renal medulla was carried out following deproteinization of the homogenate and partition with petroleum ether to remove non-polar neutral lipids. Following extraction from the homogenate at pH 3 with benzene, the organic phase was evaporated to a small volume and silicic acid chromatography was carried out with various concentrations of ethyl acetate and benzene. These chromatographies required large (95 cm65 cm) water-jacketed glass columns, yielding successive 100 ml elutes
CLINICAL SIGNIFICANCE OF EICOSANOIDS
363
Figure 32.2 Schematic representation of sequential steps utilized in the purification and isolation of medullin. From Lee et al (1965). Reproduced with permission from Lippincott Williams & Wilkins
Figure 32.3 Isolation of medullin, compound 1 and compound 2 from rabbit renal medulla by thin-layer chromatography. OR, origin; SF, solvent front. From Lee et al (1965). Reproduced with permission from Lippincott Williams & Wilkins
364
CIRCULATORY SYSTEM
Figure 32.4 Identification of compound 1 as PGF2a, compound 2 as PGE2 and medullin as PGA2, and structures of the isolated prostaglandins. Suggested fragmentations of all three compounds are indicated in the structure of PGA2. From Lee (1967) by permission of the Nobel Foundation, and Lee et al (1967), by permission
with a total elution volume of 10 l. Each 100 ml elute was evaporated to dryness, reconstituted with pH 8 phosphate buffer and tested for blood pressure-lowering activity in the pentoliniumtreated rat. The active fractions were pooled and followed by reverse-phase partition chromatography with ultimate isolation by thin-layer chromatography. The purified medullin was pooled and stored frozen for use in subsequent identification and physiological studies. An ultraviolet spectrum of medullin revealed maximum absorption at 215–217 mm, suggesting a double bond in conjugation with a keto group in a five-membered ring (Lee 1967). Since infrared spectra revealed the presence of trans-ethylene bonds, it seemed likely that medullin was a PGE with elimination of water from the cyclopentane ring. At that time, it seemed prudent to pursue additional identification procedures at a different mass spectrometer facility. In subsequent studies in collaboration with Dr Jack Gougoutas of the Department of Chemistry, Harvard University, mass spectroscopy revealed the structure of medullin to be D-10 PGE2, renamed PGA2, compound 1 to be PGF2a and compound 2 to be PGE2 (Figure 32.4; Lee 1967; Lee et al 1967).
FORMATION OF PGA2 FROM PGE2 At this point, it is important to note that storage of pure PGE2 from the renal medulla under nitrogen at 7208C gave rise to a second component compatible with small amounts of PGA2. This instability was consistent with a personal communication from Dr David A. van Dorp, Unilever Laboratories, Vlaardingen, The Netherlands, that medullin could be an artifact arising from ring dehydration of PGE2 during the isolation procedure. To test this, two fresh rabbit renal medullas were homogenized. One homogenate (extract A) was extracted without added PGE2, while another (extract B) was homogenized with addition of 1.65 mg PGE2. After extraction and isolation by thin-layer chromatography, PGE2 but not PGA2 was detected in extract A. However, although very small amounts of PGA2 (100–150 mg) were detected in extract B, much larger amounts of PGE2 (0.9 mg) were present
(Lee et al 1967). It was concluded that only very small amounts of PGA2 could spontaneously arise from PGE2 (510%) during isolation and extraction. The failure to detect PGA2 in extract A was likely due to the very small amounts of endogenous PGE2 (15 mg/g wet wt). Since our original observations (Lee et al 1967), a number of investigators have repeatedly concluded that PGA2 is an artifact of the extraction procedure, most notably Crowshaw (1973), Terragno et al (1973) and Frolich et al (1975). This has led to a widespread, but unfounded, belief that all PGA1 and PGA2 are dehydration artifacts. It has always been, and still is, our belief that although small amounts of PGA2 result from the extraction procedures, significant amounts of PGA2 are generated under various conditions that cannot be accounted for by spontaneous dehydration of PGE2, for the following reasons: 1. Braselton and Carr (1974) extracted human plasma according to our usual procedures and identified PGA1 and PGA2 by two different gas chromatographic–mass spectrometric systems. PGA1 and PGA2 were identified in similar amounts in the two systems (0.66 and 0.86 ng/ml). These values also agreed quite well with the same human extracts sent to us, where a highly specific radioimmunoassay (relative cross-reactivity: PGA2, 100%, PGA1, 80%, PGE2) revealed a concentration of 1.27 ng/ ml. 2. Hamburg and Samuelsson (1967) found very large amounts of PGEs and 19-OH PGEs in sheep seminal fluid but were unable to detect PGAs or 19-OH PGAs. Under similar conditions in the human, large amounts of PGEs, 19-OH PGEs and PGAs and 19-OH PGAs were detected. Furthermore, addition of tritium-labelled PGE1 prior to extraction of human semen resulted in approximately 90% of labelled PGE1 in the PGE1 fraction following extraction and silicic chromatography. Thus, the major part of PGA1 must have been present in the original sample. In their conclusions, we quote: ‘‘These evidences indicate that the isolation procedure applied in these studies and even the storage of the prostaglandins in the body do not per se result in
CLINICAL SIGNIFICANCE OF EICOSANOIDS dehydration of PGE to give the 9-keto D10,13 dienoic acid derivatives (PGA). The available evidence on the origin of PGA1 and PGA2 in human seminal plasma thus indicates that they are formed enzymatically from corresponding PGE compounds’’. 3. We have observed in unpublished observations that when labelled PGE2 is added to blood or tissue extract less than 4% of PGA2 is derived from PGE2, confirming the observations in our original report (Lee et al 1967) and those of Hamburg and Samuelsson (1967). 4. Dray and Charbonnel (1973) have determined PGE concentrations in circulatory human plasma to be in the low picogram range. Thus, it would seem implausible for a significant amount of PGA2 in the nanogram range to be generated by PGE2 in the picogram range, unless rapid conversion of PGE2 to PGA2 occurs in vivo. It is concluded that PGA2 is not an artifact arising spontaneously and quantitatively from endogenous PGE2. Furthermore, PGA2 synthesis may be species-specific (i.e. in human and rabbit but not in sheep) arising from PGE2 or by such synthetic mechanisms as the PG cyclooxygenase-2 (COX-2) pathway.
CIRCULATORY CHARACTERISTICS IN ANIMALS During the isolation procedure of medullin, studies were carried out to determine its mechanism of vasodepression with medullin (purified and partially purified) and authentic PGE1 generously donated by Dr Sune Bergstrom. The hypotensive effect of medullin and PGE1 was attributed to a direct effect on peripheral arterioles, since intravenous injections in the dog lowered systolic and diastolic blood pressure, with a concomitant increase in cardiac output secondary to a baroreceptor-mediated elevation in
365
heart rate (Lee et al 1965). No direct effect on cardiac output or heart rate with medullin or PGE1 was observed in the isolated perfused heart preparation. Although intraarterial injection increased blood flow to the carotid, mesenteric limb and renal vasculatures, it was only on infusion into the renal arteries that an observed change in function was evident. Infusion of PGE2 (or medullin) into the kidney resulted in a marked increase in renal blood flow, with an acceleration of urine formation from the infused but not the opposite control kidney (Figure 32.5). This was ultimately attributed to a rather profound loss of NaCl and an osmotically adjusted loss of water (Lee 1967). In elegant studies by Barger and Herd (1966), utilizing 85Kr washout techniques confirmed by autoradiography and silastic injection, it was shown that medullin (PGA2) resulted in an increase in cortical blood flow in the unanaesthetized dog. Figure 32.6 shows the silastic kidney from a normal dog with filling of the cortical vasculature and visualization of the peritubular capillaries surrounding the vasa rectae. Following intrarenal artery infusion of medullin, there was a marked decrease in medullary and cortical juxtamedullary peritubular capillaries and long loops of vasa rectae in the papilla. The net effect of PGA2 is to redistribute blood flow from medulla to cortex with an increase in cortical blood flow. It is believed that this major redistribution effect of PGA2 results in the observed natriuresis and diuresis in normotensive animals and in studies in patients with essential hypertension, discussed below.
CIRCULATORY EFFECTS IN HUMAN HYPERTENSION The first PGA to be infused intravenously into a hypertensive patient was PGA2 isolated from the kidney as medullin. This led to a fall in blood pressure to normotensive levels, which was the result of peripheral vasodilation (Lee 1967). Subsequently, PGA1,
Figure 32.5 Effect of PGE2 infusion in the renal artery on renal haemodynamics and urine flow. Measurements were recorded on a multichannelled Grass polygraph. Renal blood flow was measured using an electromagnetic flow meter (Biotronex). Urine formation from both experimental and control kidneys was measured using a Grass photoelectric transducer (PTTI) adapted for drop counting. Note the initial decrease in renal blood flow, followed by a secondary increase accompanied by enhanced urine formation from the experimental but not the control kidney. From Lee (1968), by permission
366
CIRCULATORY SYSTEM blood pressure. Of importance is that renal blood flow, sodium excretion and urine flow returned to control levels when blood pressure fell to or toward normotensive values. Thus, normotension was achieved with normalization of renal blood flow and sodium and water excretion. Although these pharmacological studies show that PGAs behave as ‘‘ideal’’ antihypertensive agents, they give little insight as to any normal physiological role of these compounds in response to various hypertensive and volume-depleting stimuli. INTERACTION WITH THE RENIN-ANGIOTENSIN AXIS
Figure 32.6 At the top is a silastic-injected kidney from a control unanaesthetized normotensive dog. c, cortex; om, outer medulla; im, inner medulla. Note the bundles of vasa rectae in outer medulla (arrows) surrounded by a dense peritubular capillary network extending into the inner medulla. At the bottom is a silastic-injected kidney from an unanaesthetized normotensive dog after infusion of PGA2 into the renal artery (0.2 mg/min). Note prominent bundles of vasa rectae (arrow) in contrast to the dark background, which is the result of a marked decrease in outer medullary and cortical juxtamedullary peritubular capillaries. This is also accompanied by a reduction in the vascular filling of the inner medulla. Reproduced by kind permission of Dr Barger from Lee (1968), with permission of the publisher
kindly supplied by the Upjohn Company, was infused into 20 hypertensive patients by Carr (1970) and Lee et al (1971). These results are documented in the literature and will only be discussed relevant to any possible role in human hypertension. Figure 32.7 shows that PGA1 infused intravenously results in an immediate elevation in renal blood flow accompanied by a marked increase in urine flow, sodium and potassium excretion (not shown) but not with a decline in blood pressure. This is followed by a decrease in systemic blood pressure secondary to peripheral arteriolar dilation, with a resultant baroreceptor-mediated increase in cardiac output and rate, a phenomenon similar to the normotensive dogs. Thus, the renal circulation is most sensitive to the vasodilating actions of PGAs, but does not appear to be a major PGA resistance bed, since there was no associated initial fall in
Since PGAs produce a profound natriuresis in normotensive animals and hypertensive humans, the hypothesis arose that they might represent a natriuretic ‘‘hormone’’ (Lee 1972). With the advent of a specific radioimmunoassay for PGA (Zusman et al 1972; Attallah and Lee 1973), Zusman et al (1973) showed that plasma PGA and plasma renin activity (PRA) were low in essential hypertensive patients compared to normotensive subjects at any given sodium intake. In both groups of patients, plasma PGA rose significantly on a low sodium diet (1973). Almost identical results for PGA were also obtained by Lee (1973), suggesting a possible antihypertensive role for circulating PGA in essential hypertension. In the studies of Payakkapan et al (1975), there was a concomitant rise in human plasma PGA, PRA and urinary aldosterone during low sodium intakes in normotensive human subjects. This phenomenon was all the more striking when plasma PGA was measured in the same individual placed upon a very low followed by a very high sodium diet (Figure 32.8). This unequivocal low salt effect was extremely puzzling to us, since the opposite was expected from the many previous animal and human studies showing PGEs and PGAs to be natriuretic. To investigate this low-salt effect further, animal studies were undertaken by Stahl et al (1979) in rabbits on low, normal and high sodium intakes. In rabbits on a low-sodium diet, there was again an increase in urinary PGE2 accompanied by a rise in de novo in vitro PGE2 biosynthesis by rabbit renomedullary slices (Figure 32.9), similar to previous studies showing a rise in papillary PGA2 in rabbits on a low-salt diet (Attallah and Lee 1973). It was soon observed that other conditions of volume depletion (diuretics, post-dialysis, etc.) produced a rise in urinary PGE2. In fact, the rise in urinary PGE2 with furosemide (Katayama et al 1984) was in large part the result of a direct stimulation of renomedullary PGE2 biosynthesis by this diuretic; this elevation of PGE2 biosynthesis of furosemide effects was inhibited by indomethacin. Similarly, in human hypertension, the antihypertensive and natriuretic effects of furosemide are blunted by concomitant administration of the PG synthesis inhibitor, indomethacin (Figure 32.10) (Patak et al 1975). Since renal blood flow and sodium excretion only become PGE2-dependent under conditions of a compromised renal blood flow (Lonigro et al 1973), we and others concluded that a primary stimulus for renal PGE stimulation in vitro is a reduction in renal cortical blood flow from loss of effective intravascular blood volume, such as occurs in renal disease, oedematous states with loss of fluid to extravascular sites and in non-oedematous states such as haemorrhage, loss of fluid during haemodialysis, low sodium diet, diuretics, etc. From a clinical point of view, administration of a PG synthesis inhibitor is contraindicated in these conditions with the production of oliguria and possible acute renal failure. To further elucidate the relationships between antihypertensive PGE2 or PGA2 and prohypertensive angiotensin (AII), the AII blocker analogue, Sar1-Ileu8-AII (S-AII) and AII converting enzyme inhibitors (captopril or teprotide) were administered subcutaneously to rabbits. In every instance, there was a marked
CLINICAL SIGNIFICANCE OF EICOSANOIDS
367
Figure 32.7 Effect of PGA1 on blood pressure and renal blood flow in essential hypertension. Note the rise in urine flow before maximal blood pressure decrease (period I) and return of elevated renal blood flow toward control values when blood pressure declines (period II). Adapted from Lee et al (1971)
decrease in de novo in vitro PGE2 synthesis (Attallah et al l982). Furthermore, when the S-AII analogue was administered subcutaneously on a low, normal and high sodium intake every 4 h for 8 days, the S-AII analogue acted as an antagonist to AII by decreasing PRA, urinary PGE2 and PGE2 synthesis in vitro (Figure 32.11; Katayama et al 1987). However, S-AII analogue acted as an agonist during high sodium intake resulting in a marked increase in PGE2 synthesis but with a fall in already low PRA to values indistinguishable from 0 levels. It was concluded that renal PGE2 production was dependent on AII levels, either endogenously produced or exogenously administered. This appeared to implicate a close relationship between the prohypertensive renin AII axis and antihypertensive renal PGE2, suggesting that AII is a physiological monitor of renal PGE2 biosynthesis. Under conditions of high sodium intake, the already low PRA levels became negligible with S-AII blockade, suggesting very low AII endogenous generation. Under these conditions, it is possible that S-AII may itself substitute for endogenous AII and stimulate PGE2 production. These results, together with many other observations of the same nature, led to the oversimplified schema shown in Figure 32.12. Important other modulators of intrarenal blood flow, such as the cholinergic nervous system, the kinin–kallikrein axis, the cytokines, etc., have been omitted for clarity. The major renal prohypertensive factors are the rapid-acting renal adrenergic system and the slow-acting renin–angiotensin axis. In the former situation, renin is directly released by the sympathetic stimulus with a rise in blood pressure. In the latter case, renin is released secondary to intravascular volume depletion with an unchanged or decrease in blood pressure. Studies showing the involvement of renin release by heightened renal adrenergic stimuli and subsequent interactions with prostaglandins and angiotensin II are beyond the scope of this discussion. In this instance, the precise role of the adrenergic system in the schema shown in Figure 32.12 must await clarification. In any case, angiotensin II is now believed to have three major actions: (a) a rise in systemic blood pressure; (b) an increase in
[Image not available in this electronic edition.]
Figure 32.8 Concentration of PGA1 in peripheral venous plasma of the same normotensive subjects on a low sodium diet followed by a high sodium diet. Individual changes in plasma PGA in response to low and high sodium intakes. Low-salt diet: 50 meq/24 h, high salt diet: 4175 meq/ 24 h. There is a highly significant rise in PGA (p50.001) in subjects ingesting low amounts of sodium. From Payakkapan et al (1975), by permission of Blackwell Publishing Co
368
CIRCULATORY SYSTEM
Figure 32.9 De novo biosynthesis of PGE2 at low (L), normal (N) and high (H) sodium chloride intakes. Each value represents mean+SE. Results are expressed as mg/g, n=9. From Stahl et al (1979), by permission of the American Physiological Society
Figure 32.11 Effects of Sar1-Ileu8-AII (200 mg/4 h, s.c. for 2 days) on de novo renal PGE2 biosynthesis. C= control animals injected with vehicle alone; ˇ=animals injected with Sar1-Ileu8=AII. Each value indicates mean+SEM (n=5). *P50.01 vs. controls during each sodium intake. Reprinted from Katayama et al (1987). Effect of sar1-Ileu8-angiotensin on renal biosynthesis and excretion. J Lab Clin Med, 105, 504–508, with permission from Elsevier
Figure 32.10 Antagonism of antihypertensive and natriuretic activity of furosemide by indomethacin. Administration of furosemide resulted in expected significant reduction of blood pressure and natriuresis in 10 hypertensive subjects on a normal sodium intake. The prostaglandin synthetase inhibitor indomethacin, administered to the same 10 subjects, resulted in a slight but significant increase in blood pressure, with little change in sodium excretion. The combination of furosemide and indomethacin either abolished or markedly diminished the antihypertensive and natriuretic effects of this diuretic, suggesting that prostaglandins may mediate such actions. From Patak et al (1975), published by Elsevier
aldosterone production; and (c) an enhanced PGE2/PGA2 synthesis and release to cortex, where the PGs are extensively metabolized. The mechanism whereby PGE2/PGA2 reach the cortex is unknown, although the long veins of Barger and Herd extending from the medullary vasa rectae to mid-cortex could provide an anatomical pathway (Figure 32.13). Alternatively, countercurrent exchange of the PGs from the arcuate vein to the arcuate artery might well occur, as exists from the ovarian vein to the ovarian artery for PGF2a (McCracken 1971). In summary, Figure 32.12 shows that final blood pressure may be a finely tuned balance of prohypertensive angiotensin II and PGE2/PGA2 on the renal vasculature. Any generation of angiotensin II by antihypertensive stimuli does not result in hypertension, suggesting dominance of PGE2/PGA2 effects. If the balance were tipped in this way, a redistribution of blood flow to cortex from medulla would be expected, leading to increased cortical blood flow, salt and water loss and normotension. Such a phenomenon might explain the blood pressure-lowering effects of
CLINICAL SIGNIFICANCE OF EICOSANOIDS
369
Figure 32.12 Hypothetical schema showing suggested interactions of prohypertensive and antihypertensive activities consonant with known information on the antihypertensive effects of renal PGA2 and PGE2
low-sodium intake and diuretics, when blood pressure would be expected to rise with the elevation of the renin–angiotensin II– aldosterone axis during these conditions. Any generation of angiotensin II by prohypertensive activity results in transient or sustained blood pressure, suggesting a preponderance of prohypertensive factors acting individually or in concert with angiotensin II—in other words, a relative, if not absolute, deficiency of PGE2/PGA2. Interestingly, it is well known that renal artery stenosis results in hypertension. This may reflect the fact that the kidney ‘‘senses’’ the post-stenotic low renal blood flow as being the result of systemic decrease in intravascular volume when, in fact, a normal or even elevated intravascular volume exists. Thus, the normal antihypertensive response to a decrease in intravascular volume does not occur and a preponderance of prohypertensive factors prevails. THE TRANSITION—COX-1 AND COX-2 Although the foregoing represents a very early personal perspective on the PGs and the circulation, recent studies and other avenues of research have yielded even more important possible PG circulatory roles from entirely different perspectives.
Since the independent and simultaneous discovery that aspirin and other non-steroidal agents inhibit cyclooxygenase (COX) synthesis (Smith and Willis 1971; Vane 1971), much attention has been directed to aspirin and its interactions with arterial endothelium and vascular thrombosis. The subsequent discovery that endoperoxides generate prostacyclin (PGI2) in arterial walls, which inhibits platelet aggregation and locally vasodilates (Bunting et al 1976), and that platelet thromboxane synthesis generates thromboxane A2 (TXA2), which is platelet aggregatory and locally vasocontricts (Needleman et al 1976), has led to an interaction of the two which is similar to that of angiotensin II and PGE2 in the kidney. The net effect of aspirin inhibition is to prolong the bleeding time, since a preponderance of vascular PGI2 over TXA2 by a complex series of mechanisms, including suppression of TXA2 for the lifetime of the platelets, allowing greater expression of PGI2. These actions and interactions among PGI2, TXA2 and non-steroidal antiinflammatory drugs (NSAIDs) are thoroughly reviewed by A. A. Weber in Chapter 33 (this volume). The clinical effects of small doses of aspirin have received wide acceptance for use in the prevention of coronary artery and deep venous thrombosis. A recent voluminous literature has appeared, beginning in the 1900s with Masferrer et al (1990), Xie et al (1991) and Kubuju et
370
CIRCULATORY SYSTEM
Figure 32.13 Silastic-injected renal venous vasculature of a normotensive unanaesthetized dog. Long veins from peritubular outer medullary capillaries penetrate to mid- and superficial cortex to join deep and superficial cortocal veins. From Barger and Herd (1973), published by OUP
al (1991) involving COX, which has given a dramatic impetus to eicosanoid research. It is now recognized there are two COX isoenzymes: COX-1 and COX-2. COX-1 is constitutively or constantly expressed as a basal COX pathway and is found in almost all cells. COX-2 is an isoenzyme, which can be induced by many factors including certain cytokines and is inhibited by glucocorticoids. The importance of COX-2 stems from the fact that this isoenzyme appears capable of directing PG synthesis toward a particular eicosanoid, depending on the nature and site of the stimulus. This ‘‘directed’’ synthesis, such as increased PGE2 and PGI2 but not TXA2 in human synovial cells, is produced by expression of PG synthesis enzymes distal to the site of COX enzymes. Finally, it is worthwhile to note that selective stimulation and/ or inhibition of COX-2 enzyme expression affords a unique opportunity to uncover mechanistic and causal relationships between the specific eicosanoid produced and the target of its production. Since COX-1 and COX-2 genes have been cloned to different chromosomes, it would appear that unlimited possibilities are available for fundamental research from a transition to a molecular biological and genetic approach.
ACKNOWLEDGEMENTS In retrospect, it has been my good fortune to have lived a lifetime with the entire prostaglandin story, from the year of my birth in 1930, when Kurzok and Lieb first published the existence of biological activity in human semen, and in 1934, when von Euler first published these activities in sheep seminal vesicles and named them prostaglandins. From a single to several papers a year in the beginning, the literature has become voluminous, with thousands of articles being published at the height of interest and
meticulously and accurately referenced by officials at the Upjohn Company, who generously distributed them to researchers in the field. It has also been my good luck to have been able to participate in this research by pursuing my interest in the kidney as a major endocrine and metabolic as well as an excretory organ. In that this undoubtedly will be my final contribution to this field, I wish to gratefully acknowledge the exceedingly generous professional support and collaborative help of my fellows, friends and colleagues over the years, including those not mentioned in the text or references: Evan Calkins, A. Crastes de Poulet, MarieJosef Duschesne, Rainer Dusing, Thomas F. Frawley, Harold Jeghers, Anna Kantoch, William J. Reddy, George E. Schreiner and Vernon K. Vance. With a very few notable exceptions, I have encountered a dedicated group of research scientists freely sharing ideas and information with common goals of making incisive contributions to a rapidly evolving field of research, particularly in the latter half of the past 73 years. I am indebted to Drs Paul J. Davis and Edward A. Stehlik for critically reviewing the manuscript and to Cheryl Carnrike for preparing the manuscript. I am especially indebted to my wife Audrey and my sons James Jr, John, David and Steven for making the other part of my life even more rewarding than any fruits of my labours. Last, I would like to dedicate this chapter to the late Dr Garfield G. Duncan, who said it all in the first place.
REFERENCES Attallah AA and Lee JB (1973) Radioimmunoassay of prostaglandin A. Intrarenal PGA2 as a factor mediating saline-induced natriuresis. Circ Res, 33, 696–703. Attallah AA and Lee JB (1982) Prostaglandins, renal function and blood pressure regulation. In Lee JB (ed.), Prostaglandins. New York: Elsevier North Holland, 251–301.
CLINICAL SIGNIFICANCE OF EICOSANOIDS Attallah AA, Payakkapan W and Lee JB (1980) The kidney cortex as a major site of PGA2 degradation. Life Sci, 14, 1521–1534. Attallah AA, Stahl RAK, Bloch DL et al (1982) Renal PGE2 synthesis: dependence on angiotensin II. Am J Cardiol, 49, 1521–1523. Barger AC and Herd JA (1966) Study of renal circulation in the unanesthetized dog with inert gases: external counting. In Third International Congress of Nephrology 1. New York: Karger, 174–187. Barger AC and Herd JA (1973) In Orloff J and Berliner R (eds), Handbook of Physiology, Washington, DC: American Physiological Society, 249– 257. Bergstrom S and Sjovall J (1960a) The isolation of prostaglandin F from sheep prostate glands. Acta Chim Scand, 14, 1693–1700. Bergstrom S and Sjovall J (1960b) The isolation of prostaglandin E from sheep prostate glands. Acta Chim Scand, 14, 1701–1705. Bergstrom S, Ryhage R, Samuelsson B and Sjovall J (1962) The structure of prostaglandin E, F1 and F2. Acta Chim Scand, 16, 501–502. Braselton WE and Carr AA (1974) In discussion of Lee, J.B. Prostaglandins and the renal antihypertensive and natriuretic endocrine function. In Rec Progr Horm Res, 30, 527–528. Braun-Menendez E and von Euler US (1947) Hypertension after bilateral nephrectomy in the rat. Nature, 160, 905. Bunting S, Gryglewski RJ, Moncada S and Vane JR (1976) Arterial walls generate from prostaglandin endoperoxides a substance (prostaglandin X) which relaxes strips of mesenteric and coeliac arteries and inhibits platelet aggregation. Prostaglandins, 12, 897–913. Carr AA (1970) Hemodynamic and renal effects of a prostaglandin PGA1 in subjects with essential hypertension. Am J Med Sci, 259, 21–26. Crowshaw K (1973) The incorporation of (1-14C) arachidonic acid into the lipids of rabbit renal slices and conversion to prostaglandins E2 and F2a. Prostaglandins, 3, 607–620. Dray F and Charbonnel B (1973) Dosage radioimmunologique des prostaglandines F2a et E1 dans le plasma peripherique de l’homme normal. In Les Prostaglandines. INSERM, Paris, 133–158. von Euler US (1934) Zur Kenntnis der pharmakologischen Wirkungen von natiosekreten extrakten mannlicher accessorischer Geschlectsdrusen. Arch Exp Pathol Pharmacol, 175, 78–84. von Euler US (1935) Uber die spezifische Prostata und Samenglasensekretes. Klin Wochschr, 14, 1182–1183. von Euler US (1936) On the specific vasodilating and plain muscle stimulating substances from accessory genital glands in man and certain animals (prostaglandin and vesiglandin). J Physiol, 88, 213– 234. von Euler US and Eliasson R (1967) Prostaglandins. New York: Academic Press, 1–164. Frolich JC, Sweetman BJ, Carr K and Gates JA (1975) Prostaglandin synthesis in rabbit renal medulla. Life Sci, 17, 1105–1112. Hamburg M and Samuelsson B (1967) New groups of naturally occurring prostaglandins. In Bergstrom S and Samuelsson B (eds), Prostaglandins Nobel Symposium 2. Stockholm: Almqvist and Wiksell, 63–69. Katayama S, Attallah AA, Stahl RAK, Bloch DL and Lee JB (1984) Mechanism of furosemide-induced natriuresis by direct stimulation of renal prostaglandin E2. Am J Physiol, 247, F555–F561. Katayama S, Attallah AA, Stahl RAK and Lee JB (1987) Effect of sar1– Ileu8–angiotensin on renal biosynthesis and excretion. J Lab Clin Med, 105, 504–508. Kubuju DA, Fletcher BS, Barnum BC et al (1991) TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J Biol Chem, 266, 12866–12872. Kurzrok R and Lieb CC (1930) The action of semen on the human uterus. Proc Soc Exp Biol Med, 268–272. Lee JB (1967) Chemical and physiological properties of renal prostaglandins: the antihypertensive effects of medullin in essential human hypertension. In Bergstrom S and Samuelsson B (eds), Prostaglandins, Nobel Symposium 2. Stockholm: Almquist and Wiksell, 197–210. Lee JB (1968) Cardiovascular implications of the renal prostaglandins. In Ramwell PW and Shaw JE (eds), Prostaglandin Symposium of the Worcester Foundation for Experimental Biology. New York: Interscience, 131–146.
371
Lee JB (1972) Natriuretic ‘‘hormone’’ and the renal prostaglandins. Prostaglandins, 1, 55–70. Lee JB (1973) Hypertension, natriuresis and the renomedullary prostaglandins. An overview. Prostaglandins, 3, 551–579. Lee JB, Hickler RB, Saravis CA and Thorn GW (1962) Sustained depressor effect of renal medullary extract in the normotensive rat. Circulation II, 26, 747. Lee JB, Hickler RB, Saravis CA and Thorn GW (1963) Sustained depressor effect of renal medullary extract in the normotensive rat. Circ Res, 13, 359–366. Lee JB, Covino BG, Takman BH and Smith ER (1965) Renomedullary vasodepressor substance, medullin: isolation, chemical characteristics and physiological properties. Circ Res, 4, 229–330. Lee JB, Crowshaw K, Takman BH et al (1967) The identification of prostaglandin E2, F2a and A2 from rabbit renal medulla. Biochem J, 105, 1251–1260. Lee JB, McGiff JC, Kannegiesser H et al (1971) Prostaglandin A1: antihypertensive and renal effect. Stud Patients Essent Hypertens, 74, 703–710. Lonigro AJ, Itskovitz HD, Crowshaw K and McGiff JC (1973) Dependency of renal blood flow on prostaglandin synthesis in the dog. Circ Res, 32, 712–717. Masferrer JL, Zweifel BS, Seibert K and Needleman P (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. McCracken J (1971) Prostaglandin F2a and corpeus luteum regression. Ann NY Acad Sci, 180, 456–472. McGiff JC, Terragno, NA, Strand JC et al (1969) Selective passage of prostaglandins across the lung. Nature, 223, 742–745. Muirhead EE, Daniels G, Pike JE and Hinman J (1967) Renomedullary antihypertensive lipids and the prostaglandins. In Bergstrom S and Samuelsson B (eds), Prostaglandins, Nobel Symposium 2. Stockholm: Almqvist and Wiksell, 183–196. Muirhead EE, Rightsel WA, Leach BE et al (1977) Reversal of hypertension by transplants and lipid extracts of cultured renomedullary interstitial cells. Lab Invest, 35, 162–172. Needleman P, Moncada S, Bunting S et al (1976) Identification of an enzyme in platelet microsomes which generate thromboxane A2 from prostaglandin endoperoxides. Nature, 261, 558–560. Patak RV, Mookerjee BK, Bentzel CJ et al (1975) Antagonism of the effects of furosemide by indomethacin in normal and hypertensive man. Prostaglandins, 10, 649–659. Payakkapan W, Attallah AA, Lee JB and Carr AA (1975) Effect of sodium on prostaglandin A, renal and aldosterone in normotensive patients. Kidney Int, 85, 283–290. Smith ER, Lee JB and Covino BG (1964) Effect of renomedullary extract on vascular and non-vascular smooth muscle. J New Drugs, 4, 229– 230. Smith JB and Willis AL (1971) Aspirin selectively inhibits prostaglandin production in human platelets. Nature New Biol, 231, 235–237. Sokabe H and Grollman A (1962) Localization of blood pressure regulating and erythropoietic functions in rat kidney. Am J Physiol, 203, 991. Sparks RM (1971) Prostaglandin Abstracts. New York: Plenum, 1–497. Stahl RAK, Aattallah AA, Bloch DL and Lee JB (1979) Stimulation of rabbit renal PGE2 biosynthesis by dietary sodium restriction. Am J Physiol, 273, F344–F349. Terragno DA, Strand JC, Pacholczyk DA and McGiff J (1973) Prostaglandin E2, an intrarenal hormone. In Les Prostaglandines. Paris: INSERM, 207–233. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature, 231, 232–236. Xie W, Chipman JG, Robertson DL et al (1991) Expression of a mitogenresponsive gene encoding prostaglandin synthetase is regulated by mRNA splicing. Proc Natl Acad Sci USA, 88, 2692–2696. Zusman RM, Caldwell BV, Speroff L and Behrman HR (1972) Radioimmunoassay for prostaglandin A. Prostaglandins, 2, 41–43. Zusman RM, Caldwell BV, Mulrow PJ and Speroff L (1973) The role of prostaglandin A in sodium and blood pressure homeostasis. Prostaglandins, 3, 679–690.
33 Aspirin and Activated Platelets Artur-Aron Weber Universita¨tsklinikum Du¨sseldorf, Germany
Thromboxane (TX) A2 is an important regulator of platelet function. Platelet TX formation is increased in patients with unstable angina (Fitzgerald et al 1986), peripheral arterial obstructive disease (Catella et al 1986; Catella and FitzGerald 1987) and cerebral ischaemia (van Kooten et al 1997, 1999). Thus, inhibition of platelet TX formation by aspirin is a logical approach to prevent atherothrombotic complications in patients with atherosclerosis (Schro¨r 1995). This review summarizes the basic platelet pharmacology of aspirin and other TX-modifying drugs. EICOSANOIDS AND PLATELET FUNCTION Mechanisms of Platelet Aggregation Platelet aggregation is an important feature of haemostasis (George 2000). The initial step in platelet-dependent haemostasis is the adhesion of platelets to matrix-bound von Willebrand factor (vWF) via the glycoprotein (GP) Ib (Savage et al 1996). Adherent platelets become activated by an interaction between GPVI and collagen fibrils (Nieswandt et al 2001). Platelet aggregation, i.e. platelet– platelet adhesion, involves a short-lasting GPIb-mediated adhesion of platelets from the flowing blood to vWF at the surface of already immobilized platelets, and a stable adhesion, mediated by the binding of plasmatic fibrinogen (and possibly also plasmatic vWF), to activated GPIIb/IIIa receptors (Kulkarni et al 2000). Thus, although platelets initially adhere to already immobilized platelets in an activation-independent fashion, platelet activation, resulting in an increase of affinity of GPIIb/IIIa receptors for soluble fibrinogen, is required for stable aggregate formation (Figure 33.1). Platelets can be activated by a variety of stimuli, including: soluble mediators, e.g. adenosine diphosphate (ADP), TXA2/ prostaglandin (PG) endoperoxides, a-thrombin; matrix proteins, e.g. collagen; or mechanical factors (shear stress). Strong platelet stimuli are capable of activating all intracellular signalling pathways required for GPIIb/IIIa activation. In contrast, weak stimuli depend on additional amplification mechanisms in order to stimulate platelet aggregation. The two most important positive feedback mechanisms of platelet activation include the secretion of ADP from dense granules and the formation and release of TXA2. These mediators bind to specific receptors at the platelet surface, complete the initial platelet activation and recruit additional platelets into the activated fraction. Platelet Eicosanoids The most prevalent fatty acid in platelets is arachidonic acid (Marcus et al 1969). Most of platelet arachidonic acid can be The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
found in the sn-2 position of phosphatidylethanolamine, phosphatidylcholine and phosphatidylinositol. In resting platelets, the amount of free arachidonic acid is very low. This corresponds to a low rate of eicosanoid synthesis at basal conditions (Halushka et al 1997). However, platelet eicosanoid synthesis can be rapidly stimulated by the addition of free arachidonic acid. Upon platelet activation, endogenous arachidonic acid is rapidly cleaved from membrane phospholipids and metabolized to eicosanoids. The predominant enzyme involved in arachidonic acid liberation is the cytosolic (85 kDa) form of phospholipase A2 (Bartoli et al 1994). The major eicosanoids formed in platelets are TXA2 (via the cyclooxygenase pathway) and 12(S)-hydroxyeicosatetraenoic acid (12-S-HETE) (via the 12-lipoxygenase pathway) (Funk 2001). In the cyclooxygenase (COX) pathway, two oxygen molecules are first inserted into arachidonic acid by the COX activity of PGH synthase, resulting in the formation of PGG2. PGG2 is subsequently reduced to PGH2 by a peroxidase site in the PGH synthase. There are two isoforms of PGH synthase, referred to as COX-1 and COX-2 (Smith et al 1996; Vane et al 1998). Although COX-1 is the predominant isoform in platelets, COX-2 expression in platelets has also been reported (Matijevic-Aleksic et al 1995; Weber et al 1999; Hohlfeld et al 2000; Weber et al 2002a; Rocca et al 2002). However, the presence of COX-2 in platelets has been disputed (Patrignani et al 1999) and its possible functional significance still remains to be established (see below). PGH2 is metabolized to TXA2 by TX synthase. TXA2 is an unstable molecule (half-life approximately 30 s) and is non-enzymatically converted to the stable TXB2. In addition, PGH2 can be metabolized to prostacyclin (PGI2) by cells within the vessel wall (transcellular metabolism; see below) (Bunting et al 1976; Marcus et al 1980; Maclouf et al 1998). Arachidonic acid, released from a different ‘‘substrate pool’’, can also be metabolized by 12lipoxygenase to 12-S-hydroperoxyeicosatetraenoic acid (12-SHPETE) followed by peroxidation to 12-S-HETE (Marcus 1989).
Biological Effects of Thromboxane A2 TXA2 acts as a positive feedback mediator of platelet activation and also recruits additional platelets into the activated fraction. Agonist binding to thromboxane receptors (TP receptors) stimulates platelet shape change, aggregation and secretion (Johnson 1999). Despite pharmacological evidence for distinct classes of binding sites for TXA2 on platelets, the TP receptor is encoded by a single gene located at chromosome 19p13.3. Although the messenger RNA for two splice variants of TP receptors (TPa, TPb) is expressed in platelets, only TPa has been found at the protein level. TPa is coupled to Gq and its
374
CIRCULATORY SYSTEM
Figure 33.1 Mechanisms of platelet aggregation in vivo. Non-activated platelets transiently adhere via GPIb to vWF on the surface of already immobilized activated platelets. Adherent platelets become activated and firmly adhere via binding of plasmatic fibrinogen (and possibly also plasmatic vWF) to activated GPIIb/IIIa receptors. ADP and TX released from activated platelets complete the initial platelet activation and also recruit additional platelets into the activated fraction
stimulation leads to activation of phospholipase C, formation of inositol triphosphates/diacylglycerol, elevation of [Ca2+]i, and activation of protein kinase C (Halushka 2000). However, for TXA2-induced platelet aggregation, secretion of other agonists (e.g. adenosine diphosphate) with subsequent activation of Gicoupled receptors (e.g. P2Y12) might be necessary (Paul et al 1999). Platelet TP receptors can be activated by TXA2 but also by PGH2 (Mayeux et al 1988) or by the isoprostanes iPF2a-III and iPE2-III (Audoly et al 2000). In the vasculature, TXA2 causes vasoconstriction (Hamberg et al 1975; Bhagwat et al 1985), hypertrophy (Dorn 1997) and enhances mitogenesis of smooth muscle cells (Sachinidis et al 1995; Grosser et al 1997). PHARMACOLOGY OF ASPIRIN Mode of Action Although platelet-inhibitory effects of high doses of aspirin not related to COX inhibition have been reported (Buchanan et al 1982; Hanson et al 1985; Gaspari et al 1987; Ratnatunga et al 1992), the antiplatelet effects of low to medium doses of aspirin (75–325 mg/day) derive primarily from inhibition of platelet COX (Patrono et al 2001; Schro¨r 1997a). Aspirin acetylates a specific serine residue (Ser530) (al Mondhiry et al 1970; Roth and Majerus 1975; Roth et al 1975), resulting in a steric blockade of the COX channel and prevention of the access of arachidonic acid to the
catalytic site of the enzyme (Smith and Marnett 1991; Shimokawa and Smith 1992; Loll et al 1995; Figure 33.2). The acetyl group of aspirin is labile and susceptible to being transferred to biological substrates, e.g. at millimolar concentrations, aspirin is known to acetylate a variety of proteins, lipids, and nucleic acids (Pinckard et al 1969). In contrast, COX is rapidly (minutes) (Roth et al 1975) and specifically (Ser530) acetylated by aspirin at micromolar concentrations. This specific effect of aspirin is dependent on its initial binding (via the salicylate moiety) to Arg120. Even a weak binding of aspirin at this site is probably sufficient to create a high local concentration in the vinicity of Ser530, explaining the high selectivity for acetylation of this amino acid residue (Loll et al 1995). Pharmacokinetics Aspirin is rapidly absorbed with tmax values of 30–40 min after ingestion, resulting in a measurable inhibition of platelet function within 60 min (Jimenez et al 1992). With enteric-coated formulations, tmax is prolonged to 3–4 h. Oral bioavailability of aspirin is 40–50%. Presystemic deacetylation to salicylic acid can occur in the liver as well as in the portal blood. In addition, there is a relatively high esterase activity in the intestinal mucosa. Thus, a reduction of intestinal motility, resulting in a prolongation of aspirin exposition to intestinal esterases, may reduce the bioavailability of the compound. Accordingly, the bioavailability of
ASPIRIN AND ACTIVATED PLATELETS
375
Figure 33.2 Inhibition of platelet COX by aspirin. An oxidized haeme group is involved in COX activity. His309, His207 and His388 are essential for catalysis and haeme binding. Tyr385 is required for cyclooxygenase activity, while His386 is required for peroxidase activity. After an initial binding to Arg120, aspirin acetylates Ser530, resulting in a steric blockade of the COX channel and prevention of access of arachidonic acid to the catalytic site of the enzyme. Modified after Smith and Marnett (1991) and Shimikawa and Smith (1992)
enteric-coated formulations is markedly reduced to 12–25% (Bochner et al 1988) without affecting the bioavailability of salicylic acid. There is a large inter- and intra-subject variability in the pharmacokinetic parameters of aspirin at doses of 4100 mg/ day (Benedek et al 1995; Buchanan and Brister 1995). However, because platelet COX can be acetylated in the presystemic circulation (Pedersen and FitzGerald 1984; FitzGerald et al 1991), the antiplatelet effects of aspirin are unrelated to its systemic bioavailability. Aspirin concentration in plasma decays with a half-life (t1/2) of 20 min. Effects on Platelet Function When given orally to healthy volunteers, aspirin inhibits TX formation in a dose-dependent fashion (ID50=26 mg) (Patrono 1989). A log-linear inhibition of platelet COX activity is observed after single doses in the range 6–100 mg (Patrignani et al 1982; Figure 33.3). Because of irreversible enzyme inactivation (see below), the inhibitory effects of aspirin are cumulative on repeated dosing (Figure 33.4). When the dose–response effects at steady state after repeated dosing are compared with those after single doses, an eight-fold increase in potency is apparent (ID50=3 mg) (Patrono 1989). A rapid and complete inhibition of TX synthesis can be achieved with a loading dose of 300 mg in combination with 40 mg/day as a maintenance dose (Buerke et al 1995). Due to the limited mRNA content and protein synthesis in anucleate platelets, aspirin irreversibly inhibits platelet COX. Thus, formation of TX upon discontinuation of aspirin treatment depends on the release of new platelets, which are regenerated at a daily rate of approximately 10–15% (Di Minno et al 1983). However, the recovery of COX activity after a single dose of aspirin takes about 48 h (Catalano et al 1981; Patrono et al 1980). This has been interpreted as evidence for an inactivation of COX in megakaryocytes (Patrono 1989). In fact, megakaryocytes treated
Figure 33.3 Dose-dependent inhibition of platelet TX formation by single doses of aspirin. Data from Patrignani et al (1982)
with aspirin in vitro start to produce TX after a lag time of 12 h, with a recovery rate of 16% of control per day (Walenga et al 1984). Although in an early study no in vivo effects of aspirin on megakaryocytes could be detected (Huijgens et al 1986), a subsequent study demonstrated a transient (12 h) inhibition of human megakaryocyte COX by aspirin (van Pampus et al 1993). Since in most studies, platelet TX formation after low-dose aspirin was completely inhibited for at least 24 h (Preston et al 1981;
376
CIRCULATORY SYSTEM
Figure 33.4 Cumulative effects of low-dose aspirin (0.45 mg/kg) on platelet TX formation. Data from Patrignani et al (1982)
Hanley et al 1981; Mehta et al 1984), a once-daily application of aspirin is sufficient. When compared to competitive COX inhibitors, such as indomethacin or diclofenac, the inhibition of TX formation by aspirin in vitro has a steep concentration–response curve, reflecting a different (irreversible) type of enzyme inhibition (Figure 33.5). Although enzyme selectivity differs depending on the assay used (purified COX, broken cell preparations, isolated cells), aspirin is a relatively COX-1-selective inhibitor with COX-1:COX-2 selectivity ratios of 25–166 (Meade et al 1993; Mitchell et al 1994). COX-1 is the predominant COX isoform in platelets (see above). Accordingly, with the selective COX-2 inhibitor NS-398, a marked inhibition of platelet TX formation can only be observed at high inhibitor concentrations (510 mM). These effects probably reflect COX-1 inhibitory actions of the compound and demonstrate that COX-1 accounts for most of platelet COX activity. However, a small (410%) but significant inhibition of platelet TX formation is detectable at COX-2-selective concentrations (41 mM) of NS-398 (Panara et al 1995; Figure 33.5), indicating that at least some of the platelet TX may be formed via the COX2 pathway. With 120 mg aspirin, urinary TX metabolite excretion was inhibited only by 28+8%, despite a 94+1% inhibition of serum TX formation. Thus, it has been proposed that a 495% reduction in serum TX levels is required for a sufficient antithrombotic effect (Reilly and FitzGerald 1987). This conclusion was drawn from the observation that, at submaximal blockade of the capacity to generate TX ex vivo (serum TX formation), minor increments in the degree of inhibition result in a disproportionately greater inhibition of urinary TX metabolite excretion (Figure 33.6). However, the data were obtained from heterogenous experiments from two different studies involving three doses of aspirin, a TX synthase inhibitor, and a combination of aspirin with a TX synthase inhibitor (Figure 33.6). In fact, in a subsequent study, a much higher degree (480%) of inhibition of urinary TX metabolite excretion by aspirin (120 mg/day) has been reported (Knapp et al 1988). The discrepancy between marked TX formation in vivo after aspirin treatment despite almost maximal inhibition of serum TX formation may be partially explained by extraplatelet TX sources (Mehta and Roberts 1983). This hypothesis is supported by data demonstrating that, in patients with acute coronary syndrome, aspirin-resistant TX synthesis is
Figure 33.5 Concentration-dependent effects of aspirin, indomethacin, diclofenac and NS-398 on collagen (1 g/ml)-induced TX formation in citrated platelet-rich plasma. Data are means+SEM, n=20
ASPIRIN AND ACTIVATED PLATELETS
377
Figure 33.6 Relationship between serum TX formation and urinary TX metabolite excretion by different doses of aspirin and a TX synthase inhibitor (UK-38,485 200 mg, TSI). Data from Reilly and FitzGerald (1987)
dependent on COX activity of nucleated cells. This extraplatelet TX synthesis is either insensitive to antithrombotic doses of aspirin (COX-2) or rapidly recovers from COX inhibition by resynthesis of the enzyme (Cippolone et al 1997) and may be a factor in aspirin resistance (Halushka and Halushka 2002; see below). In addition, the lower efficacy of aspirin as compared to the combination of aspirin with a TX synthase inhibitor (Reilly and FitzGerald 1987) may be explained by TX formation resulting from the transfer of endothelial TX precursors (PG endoperoxides) to aspirinized platelets (Zou and Anges 1997; Camacho and Vila 2000). For example, thrombin-stimulated endothelial cells can restore the capacity of aspirinized platelets to produce TX (Karim et al 1996). Interestingly, this transfer of TX precursors correlates with the expression level of endothelial COX-2. This enzyme is rapidly (2–4 h) resynthesized in aspirintreated endothelial cells. Thus, induction of endothelial COX-2 would result in a rapid recovery of TX formation in aspirinized platelets via transcellular metabolism of PG endoperoxides. With competitive COX inhibitors (e.g. diclofenac), measurable inhibition of collagen (1 mg/ml)-induced platelet aggregation occurs when about 50% of TX formation is blocked (Figure 33.7). In fact, as little as 10% of normal platelets can partially restore platelet aggregation and TX formation (Di Minno et al 1983; Bradlow and Chetty 1982). Thus, for inhibition of TXdependent platelet activation in vivo, an almost complete inhibition of TX formation is desirable. When the effects of aspirin on platelet function are measured in vitro, several important issues need to be considered. First, aspirin can only affect TX-dependent platelet functions. For example, since arachidonic acid-induced platelet aggregation is dependent on endogenous TX formation, it can be completely inhibited by aspirin (Figure 33.8). Similarily, under normal circumstances, TX formation is required for platelet aggregation induced by low concentrations of collagen (41 mg/ml) (Figure 33.8). However, higher concentrations of collagen, as well as other ‘‘strong’’ platelet stimuli such as thrombin, can induce platelet aggregation independently of TX formation. Consequently, platelet aggregation induced by these stimuli is not inhibited by aspirin. In addition, under certain conditions, platelets can aggregate in response to low concentrations of collagen independently of TX
formation (aspirin pseudoresistance; see below). Thus, aspirinized platelets can display a full aggregatory response, depending upon the platelets studied and on the nature and concentration of the stimulus used. Second, under certain experimental conditions, marked inhibitory effects of aspirin can be observed in vitro which will not occur in vivo. For example, in citrated plasma, platelet stimulation with ADP results in a biphasic aggregation response (Packham et al 1989; Weber et al 1991). A first, reversible aggregation wave (primary aggregation) is followed by a second, irreversible response (secondary aggregation), which is accompanied by TX formation. Consequently, under these conditions, only the secondary aggregation can be blocked by aspirin (Figure 33.8). However, the occurrence of any aspirin-sensitive secondary aggregation upon stimulation with ADP is due to artificially low Ca2+ concentrations in citrated plasma (50–100 mM) as opposed to physiological values (1–2 mM). At physiological Ca2+
Figure 33.7 Relationship between the inhibition of collagen (1 g/ml)induced TX formation and platelet aggregation in citrated platelet rich plasma by diclofenac. Data are means+SEM, n=20
378
CIRCULATORY SYSTEM completely inhibit platelet TXA2 formation, with no significant inhibition of vascular prostacyclin synthesis (Clarke et al 1991). Similarily, a transdermal application of aspirin might also spare vascular prostacyclin synthesis and might limit the risk of gastrointestinal toxicity (Keimowitz et al 1993; McAdam et al 1996). Clinical Efficacy
Figure 33.8 Effects of aspirin (100 mM) in vitro on arachidonic acid-, ADP- and collagen-induced platelet aggregation in citrated plasma (original tracings)
concentrations, ADP does not stimulate TX formation and ADPinduced secondary aggregation does not occur. Accordingly, aspirin will not inhibit ADP-induced platelet aggregation in vivo. Thus, measurement of platelet aggregation in citrated plasma may lead to an overestimation of not only the platelet-inhibitory effects of aspirin but also of TP receptor antagonists or TX synthase inhibitors (Bretschneider et al 1994). Third, isolated platelets respond differently from platelets studied in whole blood, e.g. the presence of erythrocytes increases secretion of ADP and platelet TX formation. This modulation of platelet function by erythrocytes can be inhibited by aspirin (Valles et al 1991, 1998; Santos et al 1997). Taken together, in vitro platelet aggregation assays display only a limited sensitivity to detect the effects of aspirin. Thus, probably the most straightforward parameter to measure aspirin effects is the determination of serum TX formation. Measurements of serum TX evaluate the capacity of platelets to synthesize TX at virtually maximal stimulation (300–400 ng/ml TXB2) and do not reflect the actual TX formation in vivo, which is believed to be several orders of magnitude lower (1–2 pg/ml TXB2) (Patrono et al 1986). Thus, the efficacy of aspirin, as measured by the inhibition of serum TX formation, may underestimate the in vivo effects of the compound. On the other hand, measurement of serum TX formation avoids several methodological artifacts and most directly comprises the molecular mode of action of aspirin.
Effects on Endothelial Prostacyclin Synthesis Based on data demonstrating that selective inhibition of COX-2 results in a marked suppression of prostacyclin formation (Catella-Lawson et al 1999; McAdam et al 1999; Cullen et al 1998), it has been claimed that COX-2 is primarily responsible for the biosynthesis of prostacyclin in the endothelium. However, at antithrombotic doses, aspirin also inhibits endogenous prostacyclin formation (Hanley et al 1981; Preston et al 1981; FitzGerald et al 1983; Davi et al 1983; Mehta et al 1984). Because aspirin is a relatively selective COX-1 inhibitor, plasma concentrations obtained at antithrombotic doses of aspirin are unlikely to inhibit endothelial COX-2. The inhibition of prostacyclin generation by aspirin is especially pronounced in atherosclerotic patients (Weksler et al 1983; Knapp et al 1988), possibly reflecting the importance of platelet-derived PG endoperoxides for endothelial PG synthesis (Force et al 1991). Because platelet COX can be inhibited by aspirin in the presystemic (portal) circulation (Pedersen and FitzGerald 1984; FitzGerald et al 1991), some degree of platelet selectivity can be achieved by slow administration of low doses (e.g. using controlled release formulations) (Pedersen and FitzGerald 1984). For example, a formulation that releases aspirin at 10 mg/h has been shown to
The efficacy of aspirin is documented in randomized clinical trials including more than 100 000 patients (Antiplatelet Trialists’ Collaboration 1994; Antithrombotic Trialists’ Collaboration 2002). Overall, among high-risk patients (acute or previous vascular disease or other predisposing condition), aspirin therapy reduces the combined occurrence of major vascular events by 20– 25% (Patrono et al 2001). However, there appears to be surprisingly little evidence for an effectiveness of long-term aspirin treatment to reduce mortality (Cleland 2002a, 2002b). In addition, the efficacy of aspirin to prevent vascular complications is variable in different clinical settings. For example, there is a marked (about 50%) risk reduction in patients with coronary artery disease as compared to a low (if any) benefit in patients with peripheral arterial disease, acute stroke, or undergoing coronary artery bypass grafting or heart valve replacement (Antiplatelet Trialists’ Collaboration 1994; Antithrombotic Trialists’ Collaboration 2002). These differences probably reflect a variable contribution of TX-mediated platelet activation in different clinical settings. However, alternative explanations are possible. The efficacy of any antiplatelet therapy will of course critically depend on the involvement of platelets in the pathophysiology of the particular clinical condition. In addition, aspirin resistance may occur in some patient groups (see below). Aspirin doses below 75 mg/day have been suggested to be more effective and to cause fewer adverse effects, because low doses are reported to inhibit platelet TX synthesis without inhibiting PG synthesis in endothelial cells or the gastric mucosa (Patrono et al 2001). However, the efficacy of aspirin doses of 575 mg/day have been less widely studied than doses of 75–325 mg/day (Antithrombotic Trialists’ Collaboration 2002). In addition, 100 mg aspirin will compensate for one missed daily dose, as seen in noncompliant patients, while 40 mg will not (Weber et al 2000). Accordingly, available data suggest that doses of 575 mg/day may have a smaller effect as compared to medium doses (75– 325 mg/day) (Figure 33.9). Higher doses of aspirin are more gastrotoxic and are no more effective than medium doses. In fact, a recent study has demonstrated that high doses of aspirin (650 or 1300 mg/day) may be less effective than medium doses (81 or 325 mg/day) (Taylor et al 1999). Taken together, available evidence supports the use of aspirin at doses of 75–150 mg for the long-term prevention of vascular complications in high-risk patients (Antithrombotic Trialists’ Collaboration 2002). Aspirin has also been evaluated in primary prevention studies involving 450 000 persons at variable cardiovascular risk (Patrono et al 2001). In these studies, there was a significant (about 20%) reduction of vascular events by aspirin. However, the event rate on placebo is low in these persons. Thus, based on a risk–benefit stratification, aspirin treatment is not recommended for most healthy individuals. On the other hand, aspirin should be routinely considered for patients at high or intermediate risk (53%/year) of atherothrombotic complications (US Preventive Services Task Force 2002; Hayden et al 2002). Aspirin Resistance A variable portion (up to 57%) of patients with cerebrovascular, cardiovascular or peripheral vascular disease are ‘‘resistant’’ to
ASPIRIN AND ACTIVATED PLATELETS
Figure 33.9 Meta-analysis of the effects of different daily aspirin doses on the proportional reduction of vascular events in high-risk patients.The stratified odds ratios are plotted for each dose (black squares), along with their respective 99% confidence intervals (horizontal line). The combined results for all aspirin doses and the 95% confidence interval are plotted as an open diamond. Data from Antithrombotic Trialists’ Collaboration (2002)
aspirin treatment (Helgason et al 1993, 1994; Mueller et al 1997; Zimmermann et al 2001; Gum et al 2001; Eikelboom et al 2002). In some studies, however, an effective inhibition of platelet function by aspirin has been reported in these patient groups (Weksler et al 1985a, 1985b; Berglund and Wallentin 1991). One possible explanation for these divergent findings may be the fact that the term ‘‘aspirin resistance’’ describes a number of different phenomena, eventually resulting in the inability of aspirin to protect individuals from thrombotic complications, to cause a prolongation of bleeding time, to inhibit platelet aggregation ex vivo, or to inhibit platelet TX formation (Patrono et al 2001). A clinical approach to aspirin resistance, i.e. the failure of the compound to protect from an ischaemic event, despite regular intake of appropriate doses, is probably the most relevant one. However, in vitro or ex vivo tests are desirable in order to
379
individualize antiplatelet therapy in patients at risk of thrombotic complications. Using simple biochemical methods and functional in vitro studies, aspirin resistance can be classified into three distinct types (Weber et al 2002b): in healthy volunteers (‘‘aspirin responders’’), treatment with aspirin (100 mg/day for 5 days) resulted in a 495% inhibition of collagen (1 mg/ml)-induced TX formation and a complete inhibition of collagen (1 mg/ml)-induced platelet aggregation (Figure 33.10). In some patients with atherosclerosis, there was no effect of oral aspirin treatment (100 mg/day for at least 5 days) on collageninduced platelet aggregation or TX formation. However, addition of aspirin (100 mM) in vitro completely inhibited TX formation and platelet aggregation (Figure 33.10). Principally, patient noncompliance would mimic this type of aspirin resistance. Since the pharmacokinetics of low-dose aspirin is variable (see above), a reduced bioavailability of the compound may contribute to the reduced efficacy of aspirin in these patients. Thus, this type of aspirin resistance was designated ‘‘pharmacokinetic resistance— type I resistance’’. In some patients, even the addition of aspirin (100 mM) in vitro only partially inhibited platelet TX formation and collageninduced platelet aggregation. This type of aspirin resistance was designated ‘‘pharmacodynamic resistance—type II resistance’’. The mechanisms of type II resistance are unknown. One possible explanation for an impaired inhibition of thromboxane formation in vitro might be an increased platelet expression of platelet COX2, an isoform that is less sensitive to aspirin (Mitchell et al 1994). Interestingly, an incomplete suppression of the secretion of TX metabolites has been observed episodically in patients with acute coronary syndromes who were treated with intravenous aspirin (Vejar et al 1990). Extraplatelet sources of TX or transcellular metabolism of PG endoperoxides from COX-2-expressing vascular cells to aspirinized platelets (see above) has been proposed as a possible explanation for this phenomenon (Patrono et al 2001). However, this is unlikely to explain the failure of aspirin to inhibit platelet TX formation in platelet-rich plasma. In addition, polymorphisms in the COX-1 gene affecting Ser529 or Arg120 may represent the structural basis for aspirin resistance in these patients (Patrono et al 2001). In some patients, oral treatment with aspirin (100 mg/day for 5 days) completely inhibited TX formation. However, collagen
Figure 33.10 Typology of aspirin resistance as determined by the measurement of collagen (1 mg/ml)-induced TX formation and aggregation in citrated plasma (schematic tracings). Data from Weber et al (2002b) published by Taylor and Francis
380
CIRCULATORY SYSTEM
(1 mg/ml)-induced platelet aggregation was not inhibited (Figure 33.10). Because aspirin exerted the expected pharmacodynamic effect (i.e. inhibition of platelet TX formation) in these patients, this type of aspirin resistance was designated ‘‘pseudoresistance— type III resistance’’. Thus, in aspirin pseudoresistance, there is an alteration of platelet function, in that some in vitro tests of platelet function which detect aspirin effects in healthy volunteers are not affected (low concentrations of collagen are not dependent on the TX amplification pathway in these platelets to induce aggregation). Interestingly, an increased sensitivity to collagen in patients with ‘‘aspirin resistance’’ has been reported (Kawasaki et al 2000). These findings support the concept that the measurement of TX formation, rather than platelet aggregation assays, should be used to measure aspirin effects (see above). In fact, it is not clear at all if any in vitro parameter of platelet function, possibly reflecting ‘‘aspirin resistance’’ (including the measurement of TX formation) does correlate with the efficacy of the compound to reduce atherothrombotic complications in high-risk patients. Thus, there is an urgent need to develop reliable methods of assessing aspirin efficacy (Smout and Stansby 2002). Co-administration with Other Cyclooxygenase Inhibitors COX inhibitors (non-steroidal antiinflammatory drugs; NSAIDs) are widely used to treat inflammatory disease (Roberts and Morrow 2001). Many patients treated with NSAIDs also require aspirin for the secondary prevention of myocardial infarction or stroke. Because both the aspirin- and the NSAID-binding sites are located within a narrow hydrophobic channel within the COX, a potential competitive interaction between NSAIDs and aspirin is a matter of concern (Parks et al 1981; Livio et al 1982; Rao et al 1983). The irreversible inactivation of platelet COX by aspirin can be antagonized by ibuprofen and coxibs, albeit with different potencies (Ouellet et al 2001). A high level of COX-2 selectivity is associated with a low interference with COX inhibition by aspirin. Accordingly, administration of ibuprofen (400 mg/day) 2 h before aspirin (81 mg/day) to healthy volunteers significantly antagonized the irreversible COX inhibition by aspirin (Catella-Lawson et al 2001) (Figure 33.11). When the same medications are administered in the reverse order, aspirin was able to completely inhibit platelet COX. Administration of rofecoxib (25 mg/day) 2 h before aspirin (81 mg/day) did not antagonize the effects of aspirin. However, at a more clinically relevant ibuprofen dosing regimen (400 mg three times daily), the inhibitory effects of aspirin on platelet COX were inhibited, even when aspirin was administered before ibuprofen (Figure 33.11). Interestingly, administration of a clinically relevant delayed-release formulation of diclofenac (75 mg twice daily) did not interfere with COX inhibition by aspirin (Figure 33.12). This might reflect the lower potency and the shorter duration of action of diclofenac as compared to ibuprofen. Taken together, commonly used NSAIDs may limit the cardioprotective effects of aspirin. Such an interference is not likely to occur with diclofenac (75 mg twice daily) or with COX-2 selective inhibitors.
Figure 33.11 Effects of single doses of ibuprofen (400 mg/day) or rofecoxib (25 mg/day) on the inhibition of serum TX formation by aspirin (81 mg/day) in healthy volunteers (means+SD). The medications were administered for 6 days with an interval of 2 h in the indicated order. Measurements were performed at the indicated times after the administration of the first study drug. Data from Catella-Lawson et al (2001)
ANTIPLATELET EFFECTS OF OTHER THROMBOXANE-MODIFYING DRUGS Competitive Cyclooxygenase Inhibitors NSAIDs inhibit platelet TX formation through a competitive and reversible inhibition of platelet COX. Unlike aspirin, conventional NSAIDs inhibit both COX-1 and COX-2 with low selectivity (Mitchell et al 1994; Laneuville et al 1994; Patrignani et al 1994). The effectiveness of NSAIDs in the secondary prevention of
Figure 33.12 Effects of multiple doses of ibuprofen (400 mg t.i.d.) or delayed-release diclofenac (75 mg b.i.d.) on the inhibition of serum TX formation by aspirin (81 mg) in healthy volunteers (means+SD). The medications were administered for 6 days with an interval of 2 h in the indicated order. Measurements were performed at the indicated times after the administration of the first study drug. Data from Catella-Lawson et al (2001)
ASPIRIN AND ACTIVATED PLATELETS
381
Figure 33.13 Transcellular eicosanoid metabolisms between platelets and vascular cells. Vascular cells utilize platelet-derived PG endoperoxides to form prostacyclin (PGI2) which inhibits platelet fuction via IP receptors. Combined TP receptor antagonists/TX synthase inhibitors (TRASI) enhance the transfer of platelet-derived PG endoperoxides to vascular cells and block their TP receptor-dependent platelet stimulatory actions. Aspirin blocks platelet PG endoperoxide formation and, therefore, prevents the transcellular prostacylin formation in vascular cells. Platelets can also utilize endoperoxides derived from vascular cells to form TX (not shown)
atherothrombotic events is largely unknown. Although some small studies with indobufen, flurbiprofen or triflusal have demonstrated a reduction of vascular events (Patrono et al 2001), a large case-control study suggests that NSAIDs do not reduce the risk of a first myocardial infarction (Garcia-Rodriguez et al 2000). A possible explanation for the different efficacy of NSAIDs as compared to aspirin is the fact that, due to the reversible type of COX inhibition of NSAIDs, functionally significant inhibition of TX formation is achieved for only a portion of the dosing interval (Pedersen and FitzGerald 1985). Selective COX-2 inhibitors, such as rofecoxib, only marginally inhibit platelet TX formation and are not suitable as antiplatelet drugs. In fact, possible prothrombotic effects of COX-2 inhibitors, possibly resulting from a suppression of endothelial prostacyclin formation (Catella-Lawson et al 1999; McAdam et al 1999; Cullen et al 1998; Mukherjee et al 2001), are a matter of concern. Thromboxane Receptor Antagonists/Thromboxane Synthase Inhibitors In patients with atherosclerosis, increased prostacyclin generation can be observed which is paralleled by the degree of platelet activation (FitzGerald et al 1984). A possible explanation is the
finding that platelet PG endoperoxides can be utilized by cells of the vessel wall to form prostacyclin (Bunting et al 1976; Marcus et al 1980). In fact, in patients with advanced atherosclerosis, a significant part of vascular prostacyclin synthesis capacity depends on transcellular PG endoperoxide transfer (Force et al 1991). In this situation, any antiplatelet dose of aspirin will result in a reduction of prostacyclin production at the site of platelet– vessel wall interaction, e.g. in patients with atherosclerosis, aspirin (120 mg/day) inhibited prostacyclin metabolite excretion by 53% as compared to a 27% inhibition in young healthy controls (Knapp et al 1988). Inhibition of TX metabolite excretion was comparable in both groups (480%). In order to establish therapeutic alternatives to aspirin, TP receptor antagonists and TX synthase inhibitors have been developed (Catella-Lawson and FitzGerald 1997). In comparison to aspirin, TX synthase inhibitors have the advantage of inhibiting platelet TX synthesis and of increasing the formation of vascular prostacyclin synthesis via PG endoperoxide transfer (Figure 33.13). However, the use of TX synthase inhibitors is limited by platelet stimulatory actions of PG endoperoxides (Mayeux et al 1988). Thus, compounds that combine TX synthase inhibition with TP receptor antagonism, such as ridogrel, have been developed (Vermylen and Deckmyn 1992). However, in patients with myocardial infarction, ridogrel was not superior to aspirin (The RAPT Investigators 1994). A
382
CIRCULATORY SYSTEM
Figure 33.14 Effects of terbogrel (1 mM) on platelet-dependent 6-oxoPGF1a release from cultured vascular smooth muscle cells (means+SEM). Untreated and aspirinized smooth muscle cells were incubated with stimulated platelets. Terbogrel markedly stimulates 6-oxo-PGF1a release even from aspirinized smooth muscle cells, indicating that platelet-derived PG precursors (endoperoxides) are utilized to form prostacyclin. By inhibiting platelet PG endoperoxide production, aspirin blocks the transcellular prostacyclin formation in smooth muscle cells. Data from Muck et al (1998)
potential explanation might be the low potency of ridogrel as a TP receptor antagonist (Catella-Lawson and FitzGerald 1997). Terbogrel, a novel compound, selectively inhibits TX-dependent platelet activation and exhibits an equipotent (IC50 10 nM) potency as TX synthase inhibitor and TP receptor antagonist (Muck et al 1998). Terbogrel enhances prostacyclin synthesis in cultured vascular smooth muscle cells via transfer of plateletderived PG endoperoxides (Figure 33.14). Since prostacyclin has potent platelet inhibitory, vasodilatory and antiproliferative effects (Schro¨r 1997b; Weber et al 1998), combined TX synthase inhibitors/TP receptor antagonists, such as terbogrel, might be interesting for further in vivo investigations on platelet–vessel wall interactions in situations of endothelial injury.
REFERENCES Al-Mondhiry H, Marcus AJ and Spaet TH (1970) On the mechanism of platelet function inhibition by acetylsalicylic acid. Proc Soc Exp Biol Med, 133, 632–636. Antiplatelet Trialists’ Collaboration (1994) Collaborative overview of randomised trials of antiplatelet therapy: I. Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Br Med J, 308, 81–106. Antithrombotic Trialists’ Collaboration (2002) Collaborative metaanalysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. Br Med J, 324, 71–86. Audoly LP, Rocca B, Fabre J-E et al (2000) Cardiovascular responses to the isoprostanes iPF2a-III and iPE2-III are mediated via the TXA2 receptor in vivo. Circulation, 101, 2833–2840. Bartoli F, Lin H-K, Ghomashci F et al (1994) Tight binding inhibitors of 85 kDa phospholipase A2 but not 14 kDa phospholipase A2 inhibit release of free arachidonate in thrombin-stimulated human platelets. J Biol Chem, 269, 15625–15630. Benedek IH, Joshi AS, Pieniaszek HJ et al (1995) Variability in the pharmacokinetics and pharmacodynamics of low-dose aspirin in healthy male volunteers. J Clin Pharmacol, 35, 1181–1186.
Berglund U and Wallentin L (1991) Persistent inhibition of platelet function during long-term treatment with 75 mg acetylsalicylic acid daily in men with unstable coronary artery disease. Eur Heart J, 12, 428–433. Bhagwat SS, Hamann PR, Still WC et al (1985) Synthesis and structure of the platelet aggregation factor thromboxane A2. Nature, 315, 511–513. Bochner F, Williams DB, Morris DM et al (1988) Pharmacokinetics of low-dose oral modified release, soluble and intravenous aspirin in man, and effects on platelet functions. Eur J Clin Pharmacol, 35, 287–294. Bradlow BA and Chetty N (1982) Dosage frequency for suppression of platelet function by low dose aspirin therapy. Thrombosis Res, 27, 99– 110. Bretschneider E, Glusa E and Schro¨r K (1994) ADP-, PAF-, and adrenaline-induced platelet aggregation and thromboxane formation are not affected by a thromboxane receptor antagonist at physiological external Ca2+ concentrations. Thrombosis Res, 75, 233–242. Buchanan MR, Rischke JA and Hirsh J (1982) Aspirin inhibits platelet function independent of the acetylation of cyclooxygenase. Thrombosis Res, 25, 363–373. Buchanan MR and Brister SJ (1995) Individual variation in the effects of ASA on platelet function: implications for the use of ASA clinically. Can J Cardiol, 11, 221–227. Buerke M, Pittroff W, Meyer J and Darius H (1995) Aspirin therapy: optimized platelet inhibition with different loading and maintenance doses. Am Heart J, 130, 465–472. Bunting S, Gryglewski R, Moncada S and Vane JR (1976) Arterial walls generate from prostaglandin endoperoxides a substance (prostaglandin X) which relaxes strips of mesenteric and coeliac arteries and inhibits platelet aggregation. Prostaglandins, 12, 879–913. Camacho M and Vila L (2000) Transcellular formation of thromboxane A(2) in mixed incubations of endothelial cells and aspirin-treated platelets strongly depends on the prostaglandin I-synthase activity. Thrombosis Res, 99, 155–164. Catalano P, Smith JB and Murphy S (1981) Platelet recovery from aspirin inhibition in vivo; differing patterns under various assay conditions. Blood, 57, 99–105. Catella F, Healy D, Lawson JA and FitzGerald GA (1986) 11-Dehydrothromboxane B2: a quantitative index of thromboxane A2 formation in the human circulation. Proc Natl Acad Sci USA, 83, 5861–5865. Catella F and FitzGerald GA (1987) Paired analysis of urinary thromboxane B2 metabolites in humans. Thrombosis Res, 47, 647–656. Catella-Lawson F and FitzGerald GA (1997) Therapeutic aspects of platelet pharmacology. In von Bruchhausen F and Walter U (eds), Platelets and Their Factors: Handbook of Experimental Pharmacology, vol. 126. Berlin: Springer, 719–736. Catella-Lawson F, McAdam B, Morrison BW et al (1999) Effects of selective inhibition of cyclooxygenase II on sodium balance, hemodynamics, and vasoactive eicosanoids. J Pharmacol Exp Therapeut, 289, 735–741. Catella-Lawson F, Reilly M, Kapoor SC et al (2001) Cyclooxygenase inhibitors and the antiplatelet effects of aspirin. N Engl J Med, 345, 1809–1817. Cipollone F, Patrignani P, Greco A et al (1997) Differential suppression of thromboxane biosynthesis by indobufen and aspirin in patients with unstable angina. Circulation, 96, 1109–1116. Clarke R, Mayo G, Price P and FitzGerald GA (1991) Suppression of thromboxane A2 but not of systemic prostacyclin by controlled release aspirin. N Engl J Med, 235, 1137–1141. Cleland JGF (2002a) Is aspirin ‘‘the weakest link’’ in cardiovascular prophylaxis? The surprising lack of evidence supporting the use of aspirin for cardiovascular disease. Prog Cardiovasc Dis, 44, 275–292. Cleland JGF (2002b) Preventing atherosclerotic events with aspirin. Br Med J, 324, 103–105. Cullen L, Kelly L, Connor SO and Fitzgerald DJ (1998) Selective cyclooxygenase-2 inhibition by nimesulide in man. J Pharmacol Exp Therapeut, 287, 579–582. Davi G, Custro N, Novo S et al (1983) The effect of two low doses of aspirin on whole blood thromboxane and prostacyclin generation in healthy subjects. Thrombosis Haemostasis (Stuttgart), 50, 669–670. Di Minno G, Silver MJ and Murphy S (1983) Monitoring the entry of new platelets into the circulation after ingestion of aspirin. Blood, 61, 1081– 1085. Dorn GW II (1997) Role of thromboxane A2 in mitogenesis of vascular smooth muscle cells. Agents and Act Suppl, 48, 42–91.
ASPIRIN AND ACTIVATED PLATELETS Eikelboom JW, Hirsh J, Weitz JI et al (2002) Aspirin-resistant thromboxane biosynthesis and the risk of myocardial infarction, stroke, or cardiovascular death in patients at high risk for cardiovascular events. Circulation, 105, 1650–1655. Fitzgerald DJ, Roy L, Catella F and FitzGerald GA (1986) Platelet activation in unstable coronary disease. N Engl J Med, 315, 983–989. FitzGerald GA, Oates JA, Hawiger J et al (1983) Endogenous biosynthesis of prostacyclin and thromboxane and platelet function during chronic administration of aspirin in man. J Clin Invest, 71, 676–688. FitzGerald GA, Smith B, Pedersen AK and Brash AR (1984) Prostacyclin biosynthesis is increased in patients with severe atherosclerosis and platelet activation. N Engl J Med, 310, 1065–1068. FitzGerald GA, Lupinetti M, Charman SA and Charman WN (1991) Presystemic acetylation of platelets by aspirin: reduction in rate of drug delivery to improve biochemical selectivity for thromboxane A2. J Pharmacol Exp Therapeut, 259, 1043–1049. Force T, Milani R, Hibberd P et al (1991) Aspirin-induced decline in prostacyclin production in patients with coronary artery disease is due to decreased endoperoxide shift. Circulation, 84, 2286–2293. Funk CD (2001) Platelet eicosanoids. In Colman RW, Hirsh J, Marder VJ et al (eds), Hemostasis and Thrombosis. Philadelphia, PA: Lippincott, 533–539. Gracia-Rodriguez LA, Varas C and Patrono C (2000) Differential effects of aspirin and non-aspirin nonsteroidal anti-inflammatory drugs in the primary prevention of myocardial infarction in postmenopausal women. Epidemiology, 11, 382–387. Gaspari F, Vigano` G, Orisio S et al (1987) Aspirin prolongs bleeding time in uremia by a mechanism distinct from platelet cyclooxygenase inhibition. J Clin Invest, 79, 1788–1797. George NJ (2000) Platelets. Lancet, 355, 1531–1539. Grosser T, Zucker T-P, Weber A-A et al (1997) Thromboxane A2 induces cell signalling but requires platelet-derived growth factor to act as a mitogen. Eur J Pharmacol, 319, 327–332. Gum PA, Kottke-Marchant K, Poggio ED et al (2001) Profile and prevalence of aspirin resistance in patients with cardiovascular disease. Am J Cardiol, 88, 230–235. Halushka MK and Halushka PV (2002) Why there are some individuals resistant to the cardioprotective effects of aspirin? Could it be thromboxane A2? Circulation, 105, 1620–1622. Halushka PV, Pawate S and Martin ML (1997) Thromboxane A2 and other eicosanoids. In von Bruchhausen F and Walter U (eds), Platelets and Their Factors, Handbook of Experimental Pharmacology, vol 126. Berlin: Springer, 459–481. Halushka PV (2000) Thromboxane A2 receptors: where have you gone? Prostagland Lipid Mediat, 60, 175–189. Hamberg M, Svensson J and Samuelsson B (1975) Thromboxanes: a new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Natl Acad Sci USA, 72, 2294–2298. Hanley SP, Bevan J, Cockbill SR and Heptinstall S (1981) Differential inhibition by low-dose aspirin of human venous prostacyclin synthesis and platelet thromboxane synthesis. Lancet, 1, 696–671. Hanson SR, Harker LA and Bjomssorl TD (1985) Effects of plateletmodifying drugs on arterial thromboembolism in baboons: aspirin potentiates the antithrombotic actions of dipyridamole and sulfinpyrazone by mechanism(s) independent of platelet cyclooxygenase inhibition. J Clin Invest, 75, 1591–1599. Hayden M, Pignone M, Phillips C and Murlow C (2002) Aspirin for the primary prevention of cardiovascular events: a summary of the evidence for the US Preventive Services Task Force. Ann Intern Med, 136, 161–172. Helgason CM, Tortorice KL, Winkler SR et al (1993) Aspirin response and failure in cerebral infarction. Stroke, 24, 345–350. Helgason CM, Bolin KM, Hoff JA et al (1994) Development of aspirin resistance in persons with previous ischemic stroke. Stroke, 25, 2331– 2336. Hohlfeld T, Zimmermann N, Gams E and Schro¨r K (2000) Aspirin resistance coincides with platelet COX-2 expression but not with enhanced aspirin esterase activity. Circulation, 102(suppl), II-237 (abstr). Huijgens PC, van den Berg CA, Imandt LM et al (1986) Aspirin does not inhibit human megakaryocyte thromboxane synthesis in vivo. Scand J Haematol, 36, 92–95.
383
Jimenez AH, Stubbs ME, Tofler GH et al (1992) Rapidity and duration of platelet suppression by enteric-coated aspirin in healthy young men. Am J Cardiol, 69, 258–262. Johnson GJ (1999) Platelet thromboxane receptors: biology and function. In Rao GHR (ed.), Handbook of Platelet Physiology and Pharmacology. Norwell: Kluwer, 38–79. Karim S, Habib A, Levy-Toledano S and Maclouf J (1996) Cyclooxygenase-1 and -2 of endothelial cells utilize exogenous or endogenous arachidonic acid for transcellular production of thromboxane. J Biol Chem, 271, 12042–12048. Kawasaki T, Ozeki Y, Igawa T and Kambayashi J (2000) Increased platelet sensitivity to collagen in individuals resistant to low-dose aspirin. Stroke, 31, 591–595. Keimowitz RM, Pulvermacher G, Mayo G and Fitzgerald DJ (1993) Transdermal modification of platelet function. A dermal aspirin preparation selectively inhibits platelet cyclooxygenase and preserves prostacyclin biosynthesis. Circulation, 88, 556–561. Knapp HR, Healy C, Lawson J and FitzGerald GA (1988) Effects of lowdose aspirin on endogenous eicosanoid formation in normal and atherosclerotic men. Thrombosis Res, 50, 377–386. Kulkarni S, Dopheide SM, Yap CL et al (2000) A revised model of platelet aggregation. J Clin Invest, 105, 783–791. Laneuville O, Breuer DK, DeWitt DL et al (1994) Differential inhibition of human prostaglandin endoperoxide H synthases-1 and -2 by nonsteroidal antiinflammatory drugs. J Pharmacol Exp Therapeut, 271, 927–934. Livio M, Del Maschio A, Cerletti C and de Gaetano G (1982) Indomethacin prevents the long-lasting inhibitory effect of aspirin on human platelet cyclooxygenase activity. Prostaglandins, 23, 787–796. Loll PJ, Picot D and Garavito RM (1995) The structural basis of aspirin activity inferred from the crystal structure of inactivated prostaglandin H2 synthase. Nature Struct Biol, 2, 637–643. Maclouf F, Folco G and Patrono C (1998) Eicosanoids and isoeicosanoids: constitutive, inducible and transcellular biosynthesis in vascular disease. Thrombosis Haemostasis, 79, 691–705. Marcus AJ, Ullman HL and Safier LB (1969) Lipid composition of subcellular particles of human blood platelets. J Lipid Res, 10, 108– 114. Marcus AJ, Weksler BB, Jaffe EA and Broekman MJ (1980) Synthesis of prostacyclin from platelet-derived endoperoxides by cultured human endothelial cells. J Clin Invest, 66, 979–986. Marcus AJ (1989) Platelet eicosanoid metabolism. In Colman RW, Hirsh J, Marder VJ and Salzman EW (eds), Hemostasis and Thrombosis. Philadelphia, PA: Lippincott, 676–388. Matijevic-Aleksic N, Sanduja SK, Wang L-H and Wu KK (1995) Differential expression of thromboxane A synthase and prostaglandin H synthase in megakaryocytic cell line. Biochim Biophys Acta, 1269, 167–175. Mayeux PR, Morton HE, Gillard J et al (1988) The affinities of prostaglandin H2 and thromboxane A2 for their receptor are similar in washed human platelets. Biochem Biophys Res Commun, 157, 733– 739. McAdam B, Keimowitz RM, Maher M and Fitzgerald DJ (1996) Transdermal modification of platelet function: an aspirin patch system results in marked suppression of platelet cyclooxygenase. J Pharmacol Exp Therapeut, 277, 559–564. McAdam BF, Catella-Lawson F, Mardini I et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. Meade EA, Smith WL and DeWitt DL (1993) Differential inhibition of prostaglandin endoperoxide synthase (cyclooxygenase) isozymes by aspirin and other non-steroidal antiinflammatory drugs. J Biol Chem, 268, 6610–6614. Mehta J and Roberts A (1983) Human tissues produce thromboxane as well as prostacyclin. Am J Physiol, 244, R839–844. Mehta JL, Mehta P, Lopez L et al (1984) Platelet function and biosynthesis of prostacyclin and thromboxane A2 in whole blood after aspirin administration in human subjects. J Am Coll Cardiol, 4, 806–911. Mitchell JA, Akarasereenont P, Thiemermann C et al (1994) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697.
384
CIRCULATORY SYSTEM
Muck S, Weber A-A and Schro¨r K (1998) Effects of terbogrel on platelet function and prostaglandin endoperoxide transfer. Eur J Pharmacol, 344, 45–48. Mueller MR, Salat A, Stangl P et al (1997) Variable platelet response to low-dose ASA and the risk of limb deterioration in patients submitted to peripheral arterial angioplasty. Thrombosis Haemostasis (Stuttgart), 78, 1003–1007. Mukherjee D, Nissen SE and Topol EJ (2001) Risk of cardiovascular events associated with selective COX-2 inhibition. J Am Med Assoc, 286, 954–959. Nieswandt B, Brakebusch C, Bergmeier W et al (2001) Glycoprotein VI but not a2b1 integrin is essential for platelet interaction with collagen. EMBO J, 484, 211–216. Ouellet M, Riendeau D and Percival D (2001) A high level of cyclooxygenase-2 inhibitor selectivity is associated with a reduced interference of platelet cyclooxygenase-1 inactivation by aspirin. Proc Natl Acad Sci USA, 98, 14583–14588. Packham MA, Bryant NL, Guccione MA et al (1989) Effect of the concentration of Ca2+ in the suspending medium on the responses of human and rabbit platelet to aggregating agents. Thrombosis Haemostasis (Stuttgart), 3, 968–976. Panara MR, Greco A, Santini G et al (1995) Effects of the novel antiinflammatory compounds, N-[2-(cyclohexyloxy)-4-nitrophenyl] methanesulphonamide (NS-398) and 5-methanesulphonamido-6-(2,4difluorothiophenyl)-1-indanone (L-745,337), on the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. Br J Pharmacol, 116, 2429–2434. Parks WM, Hoak JC and Czervionke RL (1981) Comparative effect of ibuprofen on endothelial and platelet prostaglandin synthesis. J Pharmacol Exp Therapeut, 219, 415–419. Patrignani P, Filabozzi P and Patrono C (1982) Selective cumulative inhibition of platelet thromboxane production by low-dose aspirin in healthy subjects. J Clin Invest, 59, 1366–1372. Patrignani P, Panara MR, Greco A et al (1994) Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Therapeut, 271, 1706–1712. Patrignani P, Sciulli MG, Manarini S et al (1999) COX-2 is not involved in thromboxane biosynthesis by activated human platelets. J Physiol Pharmacol, 63, 661–667. Patrono C, Ciabattoni G, Pinca E et al (1980) Low-dose aspirin and inhibition of thromboxane B2 production in healthy subjects. Thrombosis Res, 17, 317–327. Patrono C, Ciabbatoni G, Pugliese F et al (1986) Estimated rate of thromboxane secretion into the circulation of normal humans. J Clin Invest, 77, 590–594. Patrono C (1989) Aspirin and human platelets: from clinical trials to acetylation of cylcooxygenase and back. Trends Pharmacol Sci, 10, 453–458. Patrono C, Coller B, Dalen JE et al (2001) Platelet-active drugs. The relationship among dose, effectiveness, and side effects. Chest, 119, 39S–63S. Paul BZS, Jin J and Kunapuli SP (1999) Molecular mechanism of thromboxane A2-induced platelet aggregation. J Biol Chem, 274, 29108–29114. Pedersen AK and FitzGerald GA (1984) Dose-related kinetics of aspirin. Presystemic acetylation of platelet cylcooxygenase. N Engl J Med, 311, 1206–1211. Pedersen AK and FitzGerald GA (1985) Cyclooxygenase inhibition, platelet function, and metabolite formation during chronic sulfinpyrazone dosing. Clin Pharmacol Therapeut, 37, 36–42. Pinckard RN, Hawkins D and Farr RS (1969) In vitro acetylation of plasma proteins, enzymes and DNA by aspirin. Nature, 219, 68–69. Preston FE, Whipps S, Jackson CA et al (1981) Inhibition of prostacyclin and platelet thromboxane A2 after low-dose aspirin. N Engl J Med, 304, 76–79. Rao GHR, Johnson GG, Reddy KR and White JG (1983) Ibuprofen protects platelet cyclooxygenase from irreversible inhibition by aspirin. Arteriosclerosis, 3, 383–388. Ratnatunga CP, Edmondson SF, Rees GM and Kovacs IB (1992) Highdose aspirin inhibits shear-induced platelet reaction involving thrombin generation. Circulation, 85, 1077–1082.
Reilly IA and FitzGerald GA (1987) Inhibition of thromboxane formation in vivo and ex vivo: implications for therapy with platelet inhibitory drugs. Blood, 69, 180–186. Roberts LJ and Morrow JD (2001) Analgesic-antipyretic and antiinflammatory agents and drugs employed in the treatment of gout. In Hardman JG, Limbird LE and Goodman Gilman A (eds), The Pharmacological Basis of Therapeutics. New York: McGraw-Hill, 687– 772. Rocca B, Secchiero P, Ciabattoni G et al (2002) Cyclooxygenase-2 expression is induced during human megakaryopoesis and characterizes newly formed platelets. Proc Natl Acad Sci USA, 99, 7634–7639. Roth GJ and Majerus PW (1975) The mechanism of the effect of aspirin on human platelets: I. Acetylation of a particulate fraction protein. J Clin Invest, 56, 624–632. Roth GJ, Stanford N and Majerus PW (1975) Acetylation of prostaglandin synthase by aspirin. Proc Natl Acad Sci USA, 72, 3073–3076. Sachinidis A, Flesch M, Ko Y et al (1995) Thromboxane A2 and vascular smooth muscle cell proliferation. Hypertension, 26, 771–780. Santos MT, Valles J, Aznar J et al (1997) Prothrombotic effects of erythrocytes on platelet reactivity. Reduction by aspirin. Circulation, 95, 63–68. Savage B, Saldivar E and Ruggeri ZM (1996) Initiation of platelet adhesion by arrest onto fibrinogen and translocation on von Willebrand factor. Cell, 84, 289–297. Schro¨r K (1995) Antiplatelet drugs. A comparative review. Drugs, 50, 7–28. Schro¨r K (1997a) Aspirin and platelets: the antiplatelet action of aspirin and its role in thrombosis treatment and prophylaxis. Semin Thrombosis Hemostasis, 23, 349–356. Schro¨r K (1997b) Prostacyclin (prostaglandin I2) and atherosclerosis. In Rubanyi GM and Dzau VJ (eds), The Endothelium in Clinical Practice. Source and Target of Novel Therapies. New York: Marcel Dekker, 1–44. Shimokawa T and Smith WL (1992) Prostaglandin endoperoxide synthase. The aspirin acetylation region. J Biol Chem, 267, 12387– 12392. Smith WL and Marnett LJ (1991) Prostaglandin endoperoxide synthase: structure and catalysis. Biochim Biophys Acta, 1083, 1–17. Smith WL, Garavito RM and DeWitt RM (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem, 271, 33157–33160. Smout J and Stansby G (2002) Aspirin resistance. Br J Surg, 89, 4–5. Taylor DW, Barnett HJM, Haynes RB et al for the ASA and Carotid Endarterectomy (ACE) Trial Collaborators (1999) Low-dose and highdose acetylsalicylic acid for patients undergoing carotid endarterectomy: a randomised controlled trial. Lancet, 353, 2179–2184. The RAPT Investigators (1994) Randomized trial of ridogrel, a combined thromboxane A2 synthase inhibitor and thromboxane A2/ prostaglandin endoperoxide receptor antagonist, vs. aspirin as adjunct to thrombolysis in patients with acute myocardial infarction. The ridogrel vs. aspirin patency trial (RAPT). Circulation, 89, 588–595. US Preventive Services Task Force (2002) Aspirin for the primary prevention of cardiovascular events: recommendation and rationale. Ann Intern Med, 136, 157–160. Valles J, Santos MT, Aznar J et al (1991) Erythrocytes metabolically enhance collagen-induced platelet responsiveness via increased thromboxane production, adenosine diphosphate release, and recruitment. Blood, 78, 154–162. Valles J, Santos MT, Aznar J et al (1998) Erythrocyte promotion of platelet reactivity decreases the effectiveness of aspirin as an antithrombotic therapeutic modality: the effect of low-dose aspirin is less than optimal in patients with vascular disease due to prothrombotic effects of erythrocytes on platelet reactivity. Circulation, 97, 350–355. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenases 1 and 2. Ann Rev Pharmacol Toxicol, 38, 97–120. Van Kooten F, Ciabattoni G, Patrono C et al (1997) Platelet activation and lipid peroxidation in patients with acute ischemic stroke. Stroke, 28, 1557–1563. Van Kooten F, Ciabattoni G, Koudstaal PJ et al (1999) Increased platelet activation in the chronic phase after cerebral ischemia and intracerebral hemorrhage. Stroke, 30, 546–549.
ASPIRIN AND ACTIVATED PLATELETS van Pampus EC, Huijgens PC, Zevenbergen A et al (1993) Influence of aspirin on human megakaryocyte prostaglandin synthesis. Eur J Haematol, 50, 264–268. Vejar M, Fragasso G, Hackett D et al (1990) Dissociation of platelet activation and spontaneous myocardial ischemia in unstable angina. Thrombosis Haemostasis (Stuttgart), 63, 163–168. Vermylen J and Deckmyn H (1992) Thromboxane synthase inhibitors and receptor antagonists. Cardiovasc Drugs Ther, 6, 29–33. Walenga RW, Miller JL and Stuart MJ (1984) Recovery of megakaryocyte thromboxane production in vitro after aspirin inhibition. Biochem Biophys Res Commun, 119, 982–988. Weber A-A, Hohlfeld T and Schro¨r K (1991) ADP-induced second wave aggregation in platelet-rich plasma from hypercholesterolemic rabbits. Thrombosis Res, 64, 703–712. Weber A-A, Zucker T-P, Hasse A et al (1998) Antimitogenic effects of vasodilatory prostaglandins in coronary artery smooth muscle cells. Basic Res Cardiol, 93(suppl 3), 54–57. Weber A-A, Zimmermann KC, Meyer-Kirchrath J and Schro¨r K (1999). Cyclooxygenase-2 in human platelets as a possible factor in aspirin resistance. Lancet, 353, 900. Weber A-A, Liesener S, Hohlfeld T and Schro¨r K (2000) 40 mg aspirin are not sufficient to inhibit platelet function under conditions of limited compliance. Thrombosis Res, 97, 365–367.
385
Weber A-A, Przytulski B, Schumacher M et al (2002a) Flow cytometry analysis of platelet cyclooxygenase-2 expression. Induction of platelet cyclooxygenase-2 in patients undergoing coronary artery bypass grafting. Br J Haematol, 116, 839–843. Weber A-A, Przytulski B, Schanz A et al (2002b) Towards a definition of aspirin resistance: a typological approach. Platelets, 13, 37–40. Weksler BB, Pett SB, Alonso D et al (1983) Differential inhibition by aspirin of vascular and platelet prostaglandin synthesis in atherosclerotic patients. N Engl J Med, 308, 800–805. Weksler BB, Tack-Goldman K, Subramanian VA and Gay WA (1985a) Cumulative inhibitory effect of low-dose aspirin on vascular prostacyclin and platelet thromboxane production in patients with atherosclerosis. Circulation, 71, 332–340. Weksler BB, Kent JL, Rudolph D et al (1985b) Effects of low-dose aspirin on platelet function in patients with recent cerebral ischemia. Stroke, 16, 5–9. Zimmermann N, Kienzle P, Weber A-A et al (2001) Aspirin resistance after coronary artery bypass grafting. J Thorac Cardiovasc Surg, 121, 982–984. Zou M and Anges C (1997) Cell–cell interaction between platelets and IL1b-stimulated vascular smooth muscle cells in synthesis of thromboxane A2. Prostagland Leukotrienes Essent Fatty Acids, 56, 85–91.
34 Generation of Vasoactive Prostanoids by the Cyclooxygenase-2 Pathway in the Cardiovascular System of the Rat P. J. Kadowitz, S. R. Baber, M. M. Mazim, M. Keebler, H. C. Champion, T. J. Bivalacqua, D. B. McNamara and A. L. Hyman Department of Pharmacology, Tulane University Health Sciences Center, New Orleans, LA, USA
Prostaglandins are involved in the regulation of physiological and pathophysiological processes (Vane et al 1987, 1998; Holtzmann 1991, 1992; Schonbeck et al 1999; Mukherjee et al 2001). The cyclooxygenases are the rate-limiting step in the formation of prostaglandins and, after the substrate arachidonic acid is released from cell membrane phospholipids, prostaglandins are rapidly formed (Vane 1971; Samuelsson et al 1978; Holtzmann 1991, 1992; Vane et al 1998). After arachidonic acid is converted into endoperoxide intermediates by the cyclooxygenase system, the unstable endoperoxide intermediates are converted into prostaglandins and thromboxane A2 by cell-specific terminal enzymes (Vane 1971; Samuelsson et al 1978; Holtzmann 1991, 1992; Vane et al 1998). Non-steroidal antiinflammatory drugs (NSAIDs) are used to treat fever, headache, menstrual pain and arthritis and are among the most widely used drugs (Vane et al 1998). Although NSAIDs are effective analgesic antiinflammatory and antipyretic agents, their use, particularly on a long-term basis, is limited by gastrointestinal (GI) side-effects, such as GI pain and dyspepsia, GI ulcers and bleeding (Vane 1971; Samuelsson et al 1978; Holtzmann 1991, 1992; Vane et al 1998). There have been attempts to limit GI and renal side-effects of NSAIDs, which until recently have not met with a great deal of success. The discovery that COX exists in two isoforms has stimulated a great deal of research on the biological role of the two COX isoforms and on the development of inhibitors that are selective for one isoform (De Witt 1971; Fu et al 1990; Hela and Neilson 1992; Feng et al 1993; Ford-Hutchinson 1993; Kennedy et al 1993). COX-1 is constitutively expressed in most cell types and believed to be involved in the regulation of physiological processes, whereas COX-2 is an inducible isoform readily upregulated by pathological conditions, such as inflammation (De Witt 1971; Fu et al 1990; Hela and Neilson 1992; Feng et al 1993; Ford-Hutchinson 1993; Kennedy et al 1993; Smith and De Witt 1995). Conventional NSAIDs, such as ibuprofen, indomethacin and aspirin, inhibit both COX-1 and COX-2, whereas a new class of antiinflammatory agents known as the coxibs are selective COX-2 inhibitors. The hypothesis that COX-2 is responsible for the generation of prostanoids in inflammation, whereas COX-1 is associated with the role of prostanoids in the regulation of physiological function, such as the maintenance of the integrity of the GI mucosa, led to the concept that COX-2 inhibitors would have less GI toxicity. The introduction of selective COX-2 inhibitors, such as celecoxib The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
and rofecoxib (etodolac and meloxicam are also considered to be selective COX-2 inhibitors) and large-scale clinical trials and comparisons with conventional NSAIDs have shown the efficacy of selective COX-2 inhibitors with a lower incidence of GI sideeffects (Fitzgerald and Patrono 2001). These early results, along with very effective marketing strategies, have led to excellent acceptance. Celecoxib and rofecoxib are among the most widely prescribed drugs for the treatment of arthritis, with sales on a worldwide basis in the billion dollar range. The catalytic activity and tertiary structures of COX-1 and COX-2 are similar, although COX-2 has a broader substrate affinity because the hydrophobic channel in the domain of the active site is less critical (Fitzgerald and Patrono 2001). The hypothesis that COX-1 is a constitutive ‘‘housekeeping’’ enzyme associated with physiological regulation of biological function, whereas COX-2 is an inducible COX isoform upregulated by pathophysiological processes, such as inflammation, seems simplistic and naı¨ ve (Fitzgerald and Patrono 2001). Although COX-2 is generally believed to be an inducible form of the enzyme, there is evidence that COX-2 is constitutively expressed in many organ systems, including the kidney, heart, brain and lung (Wilborn et al 1995; Asano et al 1996; Catrella-Lawson et al 1999; Watkins et al 1999; Bauer et al 2000; Camitta et al 2001). The expression of COX-2 is usually low under normal physiological conditions. However, the lung may be unusual in regard to the expression of COX-2. The cellular localization of COX-1 and COX-2 was compared in the lung, an organ with high COX activity, using immunogold–silver staining and RT–PCR (Fuert et al 1998). In this study, the expression of COX-1 and COX-2 was observed in the rat lung, with intense COX-2 staining in the wall of partially muscular arteries (Ermert et al 1998a). In very important studies from the same group of investigators, it was observed in the isolated buffer-perfused rat lung that arachidonic acid is converted into vasoactive prostanoids by the COX-2 pathway (Ermert et al 1998b). This study was among the first to show, on a functional basis, that COX-2 could play a physiological role in the regulation of vasomotor tone in the lung (Ermert et al 1998b). In an equally important study from the laboratory of Dr Benedict R. Lucchesi, evidence was provided that COX-2 in the coronary circulation generates PGI2, which is important in regulating vascular tone and platelet aggregation (Hennan et al 2001).
388
CIRCULATORY SYSTEM
Although it is known that arachidonic acid is converted into vasoactive prostanoids in the isolated perfused rat lung and in the coronary circulation of the dog, little is known about the role of COX-2 in the generation of vasoactive prostanoids in the pulmonary and systemic vascular beds in the intact-chest animal. The purpose of studies described in this chapter is to provide information about the role of COX-2 in the formation of vasoactive prostaglandins and thromboxane A2 (TXA2), hereafter referred to as prostanoids, in the intact-chest rat. METHODS Pulmonary arterial, pulmonary arterial wedge and systemic arterial pressures and cardiac output were measured in the intact-chest rat using a right-heart catheterization procedure (Hyman et al 1998). In these studies, Sprague–Dawley rats weighing 250–350 g were anaesthetized with Inactin (thiobutabarbital) (140 mg/kg i.p.), with supplemental doses given i.v. as needed to maintain a uniform level of anaesthesia. The rats were strapped to a fluoroscopic table. Body temperature was monitored with a rectal probe and maintained at 37 8C with a warming lamp. The trachea was cannulated with a 2.5 cm segment of PE 200 tubing in order to maintain airway patency, and the rats spontaneously breathed room air enriched with 100% O2 or were ventilated with a Harvard Model 683 rodent respirator. A femoral artery was catheterized with PE 90 tubing and systemic arterial pressure was measured using a Statham P23 transducer. Heart rate was determined using a Grass model 7P44 tachygraph.
The right external jugular vein was catheterized with PE 90 tubing, and the catheter tip was positioned at the confluence of the superior vena cava and the right atrium. This catheter was used for the i.v. injection of drugs and of the saline indicator for the measurement of cardiac output by the thermodilution technique. For the measurement of pulmonary arterial and wedge pressures, a specially designed 3F radiopaque catheter was passed from the left external jugular vein into the main pulmonary artery under fluoroscopic guidance. Pulmonary arterial pressure was measured from the catheter, and wedge pressure was measured when the catheter was advanced into the left or right pulmonary artery into the wedge position, with continuous monitoring of the pressure waveform. Pulmonary arterial and wedge pressure and systemic arterial pressure were recorded on a Grass Model 7 polygraph. Mean pressures were derived by electronic integration. Cardiac output was measured by the thermodilution technique. For the measurement of cardiac output, 200 ml 0.9% sodium chloride solution at 23 8C was injected into the right jugular vein catheter, with its tip at the confluence of the superior vena cava and the right atrium, using a Hamilton constant-rate syringe. Blood temperature changes were detected using a Columbus Instrument 3.5 F thermistor microprobe catheter, passed from the right carotid artery into the aortic arch, where blood temperature changes were monitored. Cardiac output was determined using a Columbus Instrument Cardiotherm Model 500 cardiac output computer, equipped with a small animal interface. Arterial blood gases and pH were measured using a Corning Model 178 analyser, with a 400 ml blood sample withdrawn from the femoral artery catheter.
Figure 34.1 Effect of the selective COX-2 inhibitor nimesulide on the increase in pulmonary arterial pressure and the decrease in systemic arterial pressure in response to arachidonic acid. Arachidonic acid in a dose of 3 mg/kg i.v. was injected at the beginning of the 2 min timing bar in the control period and after injection of nimesulide (3 mg/kg i.v.) in the rat
COX-2 AND CARDIOVASCULAR RESPONSES RESULTS Haemodynamic responses to i.v. injection of the sodium salt of arachidonic acid were measured in the anaesthetized rat. Injections of arachidonic acid in doses of 0.1–3 mg/kg i.v. caused dose-related increases in pulmonary arterial pressure and doserelated decreases in systemic arterial pressure. Cardiac output measured at the peak of the increase in pulmonary arterial pressure and pulmonary arterial wedge pressure was unchanged. Records from an experiment illustrating the response to the injection of arachidonic acid in a dose of 3 mg/kg i.v. are shown in Figure 34.1. The increase in pulmonary arterial pressure and the decrease in systemic arterial pressure were rapid, occurring within 10–15 s after the i.v. injection of the essential fatty acid substrate, and pressures slowly returned to baseline value over a 2–4 min period. In some experiments, pulmonary arterial pressure decreased to below baseline value after the pressor component of the response, and pulmonary arterial pressure slowly returned to baseline value. In order to ascertain the role of COX-2 in mediating haemodynamic responses to arachidonic acid, the effects of the selective COX-2 inhibitor nimesulide were investigated. Following the administration of nimesulide in a dose of 3 mg/kg i.v. the increases in pulmonary arterial pressure and the decreases in systemic arterial pressure in response to i.v. injections of arachidonic acid were reduced significantly. The inhibitory effect of nimesulide on the haemodynamic response to arachidonic acid is illustrated in the tracing from an experiment in Figure 34.1, where it can be seen that the increase in pulmonary arterial pressure and the decrease in systemic arterial pressure in response to the injection of arachidonic acid are markedly attenuated. The results of experiments in the intact-chest rat provide data in support of the hypothesis that prostanoid synthesis occurs by way of the COX-2 pathway in the rat. In addition, these results suggest that the predominant prostanoids formed by the COX-2 pathway have pressor activity in the pulmonary vascular bed and depressor activity in the systemic vascular bed in the rat. The hypothesis that the pressor activity in the lung is mediated by generation of TXA2 was investigated in experiments with the thromboxane receptor antagonist daltraban. The results of these experiments show that increases in pulmonary arterial pressure are markedly reduced by daltraban (5 mg/kg i.v.), a dose that significantly attenuated the pulmonary pressor response to i.v. injection of the TXA2 mimic U46619. These results suggest that the pulmonary pressor response to arachidonic acid is mediated in large part by the conversion of arachidonic acid to TXA2. These data suggest that TXA2 is a major product formed from arachidonic acid by the COX-2 pathway in the pulmonary vascular bed. It is unlikely that platelets play a significant role in the generation of TXA2 in the lung in these experiments, since the response is blocked by nimesulide, a selective COX-2 inhibitor, and platelets only possess the COX-1 isoform. However, additional experiments were carried out to determine whether nimesulide had an inhibitory effect on an arachidonic acidinduced platelet aggregation. In these experiments, rats receiving nimesulide (3 mg/kg i.v.) were bled and platelet-rich plasma was prepared. In aggregation experiments with platelet-rich plasma from nimesulide-treated rats, the aggregatory response to arachidonic acid was not impaired when compared with aggregatory responses to arachidonic acid in platelet-rich plasma prepared from rats treated with the nimesulide vehicle. The results of haemodynamic studies in the rat with nimesulide and daltraban and the platelet aggregation studies with nimesulide provide strong support for the hypothesis that TXA2 is generated by way of the COX-2 pathway in the lung and that platelets do not play a role in this response. The specific cell types in the lung that generate TXA2 by way of the COX-2 pathway have not yet
389
been identified, but are under study in the laboratory. The decrease in systemic arterial pressure in response to i.v. injection of arachidonic acid is attenuated by nimesulide but is not modified by daltraban. These data suggest that the decrease in systemic arterial pressure and systemic vascular resistance is mediated by the formation of vasodilator prostanoids by the COX-2 pathway, and that the systemic vasodilator response mediated by vasodilator prostanoids (PGI2 or PGE2) is modulated by the formation of TXA2. Therefore, under physiological conditions in the rat, TXA2 is formed and acts in the pulmonary circulation but not in the systemic vascular bed. It is possible, however, that in pathological conditions where platelets and inflammatory cells immigrate into systemic organs, such as the renal, intestinal, and hepatic vascular beds, TXA2 formation in these organs would increase and the vasoconstrictor effects of TXA2 would be expressed and could predominate. Pharmacological evidence supports the hypothesis that synthesis of vasoactive prostanoids occurs by way of the COX-2 pathway in the pulmonary and systemic vascular bed of the rat. These data suggest that the COX-2 pathway is constitutively active in the cardiovascular system of the rat under normal physiological conditions. This hypothesis goes against the widelyheld concept that COX-2 is an inducible isoform not normally present or present in low concentrations and is readily upregulated by cytokines at the site of inflammation. In order to measure the expression of COX-2 protein in the lung, Western blot analysis was performed on lung tissues from normal rats. Lung tissue from two rats was homogenized (Polytron, Brinkmann Instruments) in ice-cold buffer (HEPES 5 mM, pH 7.0, glycerol 26%, MgCl2 1.5 mM, EDTA 0.2 mM, DTT 0.5 mM, phenylmethysulphonyl fluoride 0.5 mM) with NaCl (300 mM final concentration) and was incubated on ice for 30 min. After centrifugation twice at 15 0006g and 4 8C for 20 min, the protein concentration was determined. For Western blot analysis, the supernatant was mixed with an equal volume of 2% SDS/1% bmercaptoethanol and fractionated using 8% SDS–PAGE (70 mg/ lane). Proteins were then transferred to a nitrocellulose membrane (Hybond-ECL, Amersham Life Sciences) by semi-dry electroblotting for 1 h. The membranes were blocked for 1 h at room temperature with blotto-Tween (5% non-fat dry milk, 0.1% Tween-20) and incubated with a primary polyclonal rabbit antiCOX-1 and antix-COX-2 (1:5000) (Santa Cruz Biotechnology). Bound antibody was detected with labelled rabbit anti-rabbit IgG secondary antibody (1:20 000) (Santa Cruz Biotechnology) and visualized using enhanced chemiluminescence. The results of these experiments are shown in Figure 34.2, where it can be seen that there is expression of COX-1 (70 kDa) or COX-2 (72 kDa) in the rat lung. The expression of a-actin (42 kDa) was used to control protein loading. Experiments to quantitate the expression of COX-1 and COX-2 in the lung, since different antibodies are used for COX-1 and COX-2, suggest that similar amounts of COX-1 protein are present in the lung under normal physiological conditions. The results of studies with the COX-2 inhibitor nimesulide suggest that synthesis of vasoactive prostanoids can proceed via the COX-2 pathway in the pulmonary and systemic vascular beds in the rat. However, the role of COX-2 and COX-1 in the mediation or modulation of responses to vasoactive hormones, such as angiotensin, was unknown until a recent report was published indicating that COX-1 and COX-2 have the ability to modulate pressor responses to angiotensin II in the mouse. Since it has been reported that COX-1 and COX-2 have opposite effects on the pressor response to angiotensin II, the effect of selective cyclooxygenase I and II inhibitors on responses to angiotensin II were investigated in the anaesthetized mouse. Following administration of the selective COX-1 inhibitor SC-560 in a dose of 5 mg/kg i.v., increases in systemic arterial pressure in response to
390
CIRCULATORY SYSTEM
Figure 34.2 Western blot analysis showing the expression of COX-1 and COX-2 protein in the rat lung in two animals. The expression of COX-1 (70 kDA protein) and COX-2 (72 kDa protein) and of a-actin as a control for protein loading is shown in the figure
i.v. injections of angiotensin II were not changed significantly. After administration of the selective COX-2 inhibitor nimesulide in a dose of 3 mg/kg i.v. pressor responses to angiotensin II were not altered. Following administration of SC-560 or nimesulide, decreases in systemic arterial pressure in response to i.v. injection of arachidonic acid were decreased in the anaesthetized mouse. These results suggest that, in anaesthetized CD-1 mice, responses to angiotensin II are not mediated or modulated by prostanoids in the COX-1 or the COX-2 pathway.
DISCUSSION The data presented in this chapter provide evidence in support of the hypothesis that COX-2 is expressed in the normal rat lung and that the formation of vasoactive prostanoids proceeds by way of this pathway. Although it is generally believed that COX-2 is an inducible isoform readily upregulated by inflammatory stimuli, it has been reported that both COX isoforms are present in normal rat lung (De Witt 1971; Fu et al 1990; Hela and Neilson 1992; Feng et al 1993; Kennedy et al 1993; O’Neill and FordHutchinson 1998; Ermert et al 1998a). The localization and mRNA expression have been studied using immunogold–silver staining and RT–PCR (Ermert et al 1998a). The expression of both COX-1 and COX-2 was demonstrated in rat lung and strong COX-2 expression was found in smooth muscle cells of partially muscular arteries and large veins (Ermert et al 1998a). In our studies, Western blot analysis showed the presence of COX-1 and COX-2 protein, and immunohistochemical studies revealed widespread immunostaining for both COX isoforms in the normal rat lung. The present results show that the selective COX-2 inhibitor nimesulide markedly attenuated the increase in pulmonary arterial pressure and the decrease in systemic arterial pressure in response
to i.v. injection of arachidonic acid in the rat. These data suggest that the pulmonary pressor response and systemic depressor response to arachidonic acid is mediated by the formation of prostanoids in the COX-2 pathway. The selectivity of nimesulide for COX-2 has been documented in the literature and was examined in the present study, using a platelet aggregation assay (Davis and Brogden 1994; Cullen et al 1998). Platelets are reported to only express COX-1, and COX-2 inhibitors do not alter arachidonic acid-induced platelet aggregation (Funk et al 1991; O’Neill and Ford-Hutchinson 1993; Cullen et al 1998). In our studies, arachidonic acid-induced platelet aggregation was not inhibited in platelet-rich plasma prepared from nimesulide-treated rats. In contrast, arachidonic acid-induced platelet aggregation was markedly inhibited in platelet-rich plasma from rats treated with the non-selective COX inhibitor sodium meclofenamate. These data on the selectivity of nimesulide provide support for the hypothesis that prostanoid formation can occur by way of the COX-2 pathway in the rat, since responses to arachidonic acid were attenuated by nimesulide. The mechanism by which prostanoids in the COX-2 pathway increase pulmonary arterial pressure and pulmonary vascular resistance was explored and, after treatment with the thromboxane receptor antagonist daltraban, the pulmonary pressor response to arachidonic acid was blocked. These data suggest that the pressor response is mediated in part by the generation of TXA2 within the lung, that TXA2 is formed in the lung via the COX-2 pathway. The observation that the pulmonary vasoconstrictor response was attenuated by a COX-2 inhibitor suggests that the TXA2 was generated by way of the COX-2 pathway in the lung and that platelets do not play an important role. The observation that the COX-2-mediated systemic vasodilator response to arachidonic acid was not altered by daltraban suggests that little if any TXA2 is formed from injected arachidonic acid in the systemic vascular bed of the rat. The data in the systemic vascular bed of the rat are in agreement with recent results in the coronary vascular bed of the dog (Hennan et al 2001). In the study by Hennan et al (2001) coronary vasodilator responses to arachidonic acid were attenuated by the COX-2 inhibitor celecoxib, whereas a thromboxane synthetase inhibitor had no significant effect on the response (Hennan et al 2001). Prostanoid formation can proceed by way of the COX-1 or COX-2 pathway. It is generally believed that COX-1 is constitutively expressed and is involved in ‘‘housekeeping’’ physiological processes, whereas COX-2 is an inducible isoform upregulated by inflammatory stimuli (De Witt 1971; Fu et al 1990; Hela and Neilson 1992; Feng et al 1993; Kennedy et al 1993; O’Neill and Ford-Hutchinson 1993; Smith and De Witt 1995). However, it has been reported that both COX-1 and COX-2 are constitutively expressed in the normal rat lung and a variety of other tissues, including human alveolar epithelial cells and mouse lung tumours, and that COX-2 may be implicated in physiological processes (Asano et al 1996; Willborn et al 1995; Catella-Lawson et al 1999; Watkins et al 1999; Bauer et al 2000; Camitta et al 2001). In our studies, Western blot analysis revealed the presence of COX-1 and COX-2 protein, and immunostaining for both COX isoforms was detected in the normal rat lung. In our experiments, injections of the prostanoid precursor arachidonic acid increased pulmonary vascular resistance, while at the same time decreased systemic vascular resistance, and these responses were blocked by nimesulide, a selective COX-2 inhibitor. These results suggest that vasoactive prostanoids are formed by the COX-2 pathway and extend the results of studies demonstrating the presence of COX-2 in the normal rat lung (Ermert et al 1998a). The selectivity of nimesulide for COX-2 was examined in platelet aggregation experiments in which the COX-2 inhibitor had no significant effect on arachidonic acid-induced platelet aggregation, an effect reported to be mediated by COX-1 in the platelet (Holtzmann
COX-2 AND CARDIOVASCULAR RESPONSES 1991; Ermert et al 1998a, 1998b, Hennan et al 2001). In order to exclude a non-specific inhibitory effect of nimesulide on vascular responses, the actions of the COX-2 inhibitor on vasodilator and vasoconstrictor responses were evaluated, and the results of these experiments show that nimesulide did not alter vascular responses to prostanoids and non-prostanoid agonists. The observation that the pulmonary vasoconstrictor response to arachidonic acid is inhibited by nimesulide or a thromboxane receptor antagonist suggests that TXA2 is a major product formed by way of the COX-2 pathway in the normal rat lung. The observation that the response is blocked by the COX-2 inhibitor in a dose that did not alter the platelet aggregatory response to arachidonic acid suggests a role for lung cells but not for platelets. The specific lung cells responsible for the generation of TXA2 in the rat lung have not been determined. The observation that daltraban blocked responses to the TXA2 mimic U46619 but did not alter responses to angiotensin II or norepinephrine suggests that the inhibitory effects of the antagonist are selective for thromboxane receptors. In summary, the results discussed in this chapter show that the COX-2 selective inhibitor nimesulide attenuates pulmonary vasoconstrictor and systemic vasodilator responses to arachidonic acid, whereas responses to exogenous prostanoids were not altered. Nimesulide did not inhibit the platelet aggregatory response to arachidonic acid. These data, along with immunohistological and Western blot analysis showing the presence of COX2 in the lung, suggest that prostanoid formation can occur by way of the COX-2 pathway. The observation that a thromboxane receptor antagonist blocked the pulmonary vasoconstrictor response to arachidonic acid suggests that TXA2 is a major product of the COX-2 pathway within the pulmonary vascular bed of the rat. These data suggest that the COX-2 enzyme may play a role in the physiological regulation of vascular function in the lung, which was previously thought to be mainly regulated by COX-1. Although it has been hypothesized that COX-2 may play a role in modulating responses to vasoactive hormones, such as angiotensin II, few if any data on this issue are currently available. It has been reported that responses to angiotensin II are modulated by the release of products in the cyclooxygenase pathway (Qi et al 2002). However, the effects of non-selective COX inhibitors on increases in systemic arterial pressure or regional vascular resistance in response to angiotensin II have been inconsistent (Qi et al 2002). Both COX-1 and COX-2 isoforms convert arachidonic acid to vasoactive prostanoids; however, the production of vasoactive prostanoids by COX-1 and COX-2 inhibitors may be regulated differently. It has recently been reported that selective COX-1 and COX-2 inhibitors have opposite effects on the pressor response to angiotensin II in the mouse (Qi et al 2002). Recent studies in our laboratory indicate that sodium meclofenamate, a non-selective COX inhibitor, has no significant effect on the pressor response to angiotensin II in the mouse. These data are consistent with the concept that vasoactive prostanoids from COX-1 and COX-2 may have differential effects on the response to angiotensin II in the mouse. In order to test the hypothesis that pressor responses to angiotensin II are modulated in a different manner by the two COX isoforms, the effects of selective COX-1 and COX-2 inhibitors on responses to angiotensin II and IV were investigated. The results of these studies show that SC-560, a selective COX1 inhibitor, and nimesulide, a selective COX-2 inhibitor, in doses that attenuate depressor responses to arachidonic acid, do not alter increases in systemic arterial pressure in response to i.v. injections of angiotensin II in the anaesthetized mouse. These data suggest that pressor responses to i.v. injections of angiotensin II are not modulated by prostanoids in the COX-1 and COX-2 pathways in the mouse. The reason for the difference in results in the anaesthetized mouse in our studies and in recently published
391
work is uncertain but may involve the method of administration of angiotensin II, bolus i.v. injection vs. i.v. infusion, the route of administration of the COX inhibitors, slow i.v. injection vs. oral administration, and the doses of the COX-1 and COX-2 inhibitors used in the two studies. In conclusion, the role of vasoactive prostanoids in the COX-1 and COX-2 pathways in the regulation or responses to vasoactive hormones is uncertain at the present time.
REFERENCES Asano K, Lilly CM and Drazen JM (1996) Prostaglandin G/H synthase-2 are the constitutive and dominant isoforms in cultured human lung epithelial cells. Am J Physiol, 271, L126–L181. Bauer AK, Dwyer-Nield LD and Malkinson AM (2000) High cyclooxygenase 1 (COX-1) and cyclooxygenase 2 (COX-2) contents in mouse lung tumors. Carcinogenesis, 21, 543–550. Camitta GW, Gabel SA, Chulada P et al (2001) Cyclooxygenase-1 and -2 knockout mice demonstrate increased cardiac ischemia/reperfusion injury but are protected by acute preconditioning. Circulation, 104, 2453–2458. Catella-Lawson F, McAdam B, Morrison BW et al (1999) Effects of specific inhibition of cyclooxygenase-2 on sodium balance hemodynamics and vasoactive eicosanoids. J Pharmacol Exp Ther, 289, 735–741. Cullen L, Kelly L, Connor SO et al (1998) Selective cyclooxygenase-2 inhibition by nimesulide in man. J Pharmacol Exp Ther, 287, 578–582. Davis R and Brogden RN (1994) Nimesulide an update of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy. Drugs, 48, 431–454. De Witt DL (1971) Prostaglandin endoperoxide synthase: regulation of enzyme expression. Biochem Biophys J, 1083, 121–134. Ermert L, Ermert M, Goppelt-Struebe M et al (1998a) Cyclooxygenase isoenzyme localization and mRNA expression in rat lungs. Am J Resp Cell Mol Biol, 18, 479–488. Ermert L, Ermert M, Althoff A et al (1998b) Vasoregulatory prostanoids generation proceeds via cyclooxygenase-2 in non-inflamed rat lungs. J Pharmacol Exp Ther, 286, 1309–1314. Feng L, Sun W, Xia W et al (1993) Cloning two isoforms of rat cyclooxygenase: differential regulation of their expression. Arch Biochem Biophys, 307, 361–368. Fitzgerald GA and Patrono C (2001) The coxibs, selective inhibitors of cyclooxygenase 2. N Engl J Med, 345, 433–442. Fu J-Y, Masferrer JL, Siebert K et al (1990) The induction and suppression of prostaglandin H2 synthase in human monocytes. J Biol Chem, 265, 16737–17640. Funk CD, Funk LB, Kennedy ME et al (1991) Human platelet/ erythroleukemia cell prostaglandin G/H synthase. FASEB J, 5, 2304–2312. Hela T and Neilson K (1992) Human cyclooxygenase-2 cDNA. Proc Natl Acad Sci USA, 889, 7384–7388. Hennan JK, Huang J, Barrett TD et al (2001) Effects of selective cyclooxygenase-2 inhibition on vascular responses and thrombosis in canine coronary arteries. Circulation, 104, 820–825. Holtzmann MJ (1991) Arachidonic acid metabolism: implications of biological chemistry for lung function and disease. Am Rev Respir Dis, 143, 188–203. Holtzmann MJ (1992) Arachidonic acid metabolism in airway epithelial cells. Ann Rev Physiol, 54, 303–329. Hyman AL, Hao Q, Tower A et al (1998) Novel catheterization technique for the in vivo measurement of pulmonary vascular responses in rats. Am J Physiol, 274, H1218–H1229. Kennedy BP, Chan CC, Culp SA et al (1993) Cloning and expression of rat prostaglandin endoperoxide synthase (cyclooxygenase)-2 cDNA. Biochem Biophys Res Commun, 197, 494–500. Mukherjee D, Nissen SE and Topol EJ (2001) Risk of cardiovascular events associated with selective COX-2 inhibitors. J Am Med Assoc, 286, 954–959. O’Neill G and Ford-Hutchinson AW (1993) Expression of mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human tissues. FEBS Lett, 330, 156–160.
392
CIRCULATORY SYSTEM
Qi Z, Hao CM, Langenbach RI et al (2002) Opposite effects of cyclooxygenase-1 and -2 activity on the pressor response to angiotensin II. J Clin Invest, 110, 61–69. Samuelsson B, Goldyne M, Granstrom E et al (1978) Prostaglandins and thromboxanes. Ann Rev Biochem, 47, 997–1029. Schonbeck U, Sukhov GK, Graber P et al (1999) Augmented expression of cyclooxygenase-2 in human atherosclerotic lesions. Am J Pathol, 155, 1281–1291. Smith WL and De Witt DL (1995) Biochemistry of prostaglandin H synthase-1 and synthase-2 and their differential susceptibility to nonsteroidal antiinflammatory drugs. Sem Nephrol, 15, 2179–2194.
Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature, 231, 232–235. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenase 1 and 2. Ann Rev Pharmacol Toxicol, 38, 97–120. Vane JR, Gryglewski RJ and Botting RM (1987) The endothelial cells as a metabolic and endocrine organ. Trends Pharmacol Sci, 8, 491–496. Watkins DN, Peroni DJ, Lenzo JC et al (1999) Expression and localization of COX-2 in human airways and cultured airway epithelial cells. Eur Respir J, 13, 999–1007. Wilborn J, De Witt DL and Peters-Golden M (1995) Expression and role of cyclooxygenase isoforms in alveolar and peritoneal macrophages. Am J Physiol, 268, L294–L301.
35 Eicosanoid Generation and Effects in Cardiac Muscle and Coronary Vessels Karsten Schro¨r Universita¨tsklinikum, Du¨sseldorf, Germany
Systematic research on the role of eicosanoids in the heart and coronary vessels started with the detection by Limas and Cohn (1973) of an enzymatic prostaglandin synthesizing and degrading activity in cardiac microsomes which, interestingly, was found to be insensitive to aspirin. In the same year, Kraemer and Folts (1973) demonstrated increased prostaglandin release from the heart during postocclusion reactive hyperaemia, and in 1977 it was found that prostacyclin (PGI2) is the major cyclooxygenase product in the hearts of rabbit and rat (Schro¨r et al 1977; De Deckere et al 1977). Later it was shown that prostaglandins might affect all aspects of cardiac and coronary function, specifically contractility, heart rate and coronary vessel tone, and that these actions are mediated by specific G-proteincoupled receptors. In addition, eicosanoids generated by the heart or coronary vessels also affect proliferation of cardiac and vascular myocytes and, therefore, became of interest as growthmodulating factors. The current interest in eicosanoids and cardiocoronary function is focused on prostaglandin receptors, their regulation, function and signal transduction. Other important topics are endothelium-derived hyperpolarizing factor(s), after identification as epoxyeicosatrienoic acids (EETs), i.e. monooxygenase products of arachidonic acid (Campbell et al 1996), as well as the interactions of eicosanoids with other vasoactive mediators, such as nitric oxide (NO) and peroxynitrite (ONOO7) or atrial natriuretic factor (ANF). Regarding the integrative function of the heart and its blood supply, it is also desirable to understand more completely the regulation and function of eicosanoids under pathophysiological conditions, i.e. under the influence of growth factors and proinflammatory cytokines and their contribution to cardiovascular diseases, specifically myocardial infarction, heart failure and restenosis of the coronary vasculature after endothelial injury, e.g. subsequent to percutaneous coronary interventions (PCI). Most recently, the possible cardiovascular side-effects of selective COX-2 inhibitors became an area of interest, in view of the exploding introduction of these compounds in the market of antiinflammatory drugs. This review is focused on COX products and epoxyeicosatrienoic acids (EETs), because these have been in the focus of scientific interest during the last decade and an improved knowledge of their generation and function will probably have a significant impact on cardiovascular pharmacology. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
GENERATION OF EICOSANOIDS BY THE HEART AND CORONARY VESSELS General Aspects Local levels of eicosanoids are entirely determined by biosynthesis. Two principally different mechanisms control biosynthesis: (a) substrate availability, i.e. free, unesterified arachidonic acid (AA) in the vicinity of synthesizing enzymes; and (b) the amount and activity of synthesizing enzymes. Under physiological conditions, i.e. in the absence of elevated levels of inflammatory cytokines or growth factors, the availability of free AA as the natural substrate is controlled by cytosolic phospholipase A2 (cPLA2) and the constitutively available cyclooxygenase (COX-1) (Smith 1992). Both become activated by acetylcholine, catecholamines, shear stress or other humoral and/or mechanical factors. Stimulation causes immediate (minutes) increase in eicosanoid production, eventually resulting in acute changes of myocardial contractile force, heart rate or coronary perfusion. There is also evidence for a functional coupling of COX-1 to a recently detected constitutive cytosolic PGE2 synthase (Tanioka et al 2000). Under pathological conditions, i.e. subsequent to the generation of proinflammatory cytokines or growth factors, e.g. in ischaemia/ reperfusion or heart failure, cPLA2 and COX-2 become induced (Smith 1992), possibly also other enzymes of downstream synthesizing pathways, most notably the inducible, membranebound PGE-synthase (mPGES) (Murakami et al 2000). This response requires longer periods of time (hours) because of transcriptional activation and protein synthesis, and leads to the generation of much larger amounts of eicosanoids. These, together with other mediators, such as NO (via iNOS), will modify cardiocoronary functions under noxious conditions. Cardiac Muscle Cells The heart can probably synthesize all kinds of prostaglandins, several lipoxygenase products and EETs (Bolton et al 1980; Schro¨r 1993; Campbell et al 1996). However, the basal prostaglandin production of cardiac myocytes is low under resting conditions and amounts to 0.1–2 ng/mg protein624 h in terms of PGE2. This is only about one-tenth of the amount synthesized by cardiac fibroblasts under comparable conditions
394
CIRCULATORY SYSTEMS
Table 35.1 Prostaglandin generation (pg/mg protein/10 min) in neonatal rat cardiomyocytes Prostaglandins
Non-treated
6-Keto-PGF1 PGE2 PGF2 TXB2
1872 679 258 108
(+201) (+80) (+31) (+33)
PMA-treated 16582 3998 248 142
(+2002) (+467) (+51) (+21)
PMA-treated+NS398 (1 mM) 1892 569 261 92
(+192) (+73) (+29) (+12)
PMA-treated+ASA (200 mM) 592 152 259 42
(+69) (+21) (+38) (+8)
From Adderley and Fitzgerald (1999), by permission of the American Society for Biochemistry and Molecular Biology.
(Mendez and LaPointe 2002). This high prostanoid-generating activity of cardiac fibroblasts might be related to their role in tissue repair and myocardial remodelling (see below). The major prostaglandin in neonatal rat cardiac myocytes under resting conditions is PGI2, followed by PGE2, PGF2a and TXA2 (Church et al 1994; Adderley and Fitzgerald 1999) (Table 35.1). Rapid prostaglandin release is caused by hormonal stimulation, e.g. by muscarinergic (acetylcholine) or adrenergic agonists (isoprenaline), ATP, angiotensin II or bradykinin. These compounds rapidly mobilize Ca2+ and release AA by activation of cPLA2 (Lin et al 1992). Acetylcholine acts via M2 and M3 in ventricular myocytes (Kan et al 1996), catecholamines via the badrenergic receptor (Schro¨r et al 1998). Proinflammatory cytokines such as interleukin-1b (IL-1b) induce COX-2 in cultured neonatal ventricular myocytes. This results in a preferential generation of PGE2 in addition to PGI2 and formation of minor amounts of PGF2a. This PGE2 production is associated with a significant increase in mPGES mRNA and protein levels. Growth factor (EGF) and PKC-induced ‘extracellular signal-related kinase’ (ERK1/2) activation are associated with enhanced PGI2 production in neonatal ventricular rat cardiomyocytes, while EGF-induced prostaglandin production takes place via a PKC-independent pathway (Quintaje et al 1998).
dependent induction of COX-2 in endothelial cells, might explain the dominant role of this isoform for PGI2 formation by the normal vasculature in vivo (McAdam et al 1999). In this context, it is interesting to note that vascular endothelial and smooth muscle cells contain comparable amounts of prostacyclin synthase but differ mainly in COX activity. The latter is much higher in endothelial cells than in smooth muscle cells (DeWitt et al 1983; Smith, 1986). In addition to cyclooxygenase and lipoxygenase products, cytochrome P450-derived monooxygenase metabolites of arachidonic acid are generated by coronary vascular endothelial cells. The coronary EDHF-synthase/cytochrome P450 2C8/9 gene in cultured porcine coronary endothelial cells has been found to be regulated by mechanical stress, eventually resulting in marked generation of several EETs, including 8,9-, 11,12- and 14,15-EET (Fisslthaler et al 2001). Importantly, this monooxygenase generates both EETs and O7 2 and, because coronary arteries are exposed to pronounced variations in vessel distension, this finding might have important consequences in endothelial dysfunction. In these conditions iNOS might become upregulated as well, eventually resulting in the generation of peroxynitrite, which inhibits prostacyclin synthase (Zou and Ullrich 1996).
Vascular Smooth Muscle Cells Coronary Vascular Cells Endothelial Cells In cultured bovine coronary artery endothelial cells under resting conditions as well as after stimulation with thrombin, arachidonic acid or Ca2+-inonophore, PGI2 is the major arachidonic acid metabolite. In addition, PGE2 is formed, as well as 11-, 12- and 15-HETE and 14,15-, 11,12-, 8,9- and 5,6-EET. The data indicate the presence of 12- and 15-lipoxygenases as well as P450monooxygenases in addition to cyclooxygenase in endothelial cells (Rosolowsky and Campbell 1996). Data in human saphenous vein endothelial cells suggest the absence of inducible mPGES (Soler et al 2000). Whether this is also true for endothelial cells from coronary arteries remains to be determined. Acetylcholine stimulates PGI2 synthesis in primary cultures of rabbit coronary endothelial cells by activation of cPLA2 in a PKC-independent manner. Stimulation by acetylcholine of PGI2 generation involves translocation of cPLA2 to the nuclear envelope and tyrosine phosphorylation of ERK via M3 muscarinergic receptors (Kan et al 1996). Acetylcholine possibly stimulates PGI2 by two mechanisms: via a PTX-insensitive Gprotein (Tyagi et al 1996) and by increasing the influx of extracellular Ca2+ via a G-protein-independent receptor-operated Ca2+-channel (Kan et al 1996, 1997). Shear stress stimulates the release of PGI2 in cultured porcine endocardial endothelial cells. This action appears to be related to intracellular Ca2+-mobilization (Hanada et al 2000) and also involves Gi (Berthiaume and Frangos 1992). Shear stress, i.e. flow-
In contrast to coronary endothelial cells, prostaglandin generation by vascular smooth muscle cells (SMCs) is very low under resting conditions (Eldor et al 1981), probably because of the low COX activity (Smith 1986). Accordingly, there is no stimulation of PGI2 by acetylcholine in cultured coronary vascular SMCs of the rabbit (Tyagi et al 1996). It is important to consider that SMCs in vivo express the contractile phenotype and, therefore, may behave differently from cultured cells in vitro, expressing the secretory phenotype. The latter is associated with the expression of receptors for growth factors at reduced contractile responses and corresponds to SMCs in the atherosclerotic plaque (Campbell and Campbell 1993). Prostaglandin generation by vascular SMCs can be markedly stimulated by the induction of COX-2 in response to serum and growth factors (Pritchard et al 1994; Rimarachin et al 1994; Bornfeldt et al 1997) or proinflammatory cytokines such as IL-1b or TNFa (Ohnaka et al 2000). This enhanced prostaglandin generation appears to be due to increased transcriptional activity of the COX-2 gene. In cultured SMCs from human popliteal arteries, i.e. mPGE-mRNA is expressed and upregulated by proinflammatory cytokines and phorbol ester (Soler et al 2000). This is interesting because of the tight coupling of mPGES and COX-2. Whether this also applies to coronary arteries remains to be shown. The half-life of COX-2 mRNA, amounting to 20– 30 min, is prolonged by thrombin (Rimarachin et al 1994) and angiotensin II (Ohnaka et al 2000), resulting in enhanced COX-2 levels. These might contribute to the thromboresistance of the neointima, partially via enhanced PGI2 production (Eldor et al 1981). The angiotensin II- and TNFa-stimulated COX-2
EICOSANOID GENERATION/EFFECTS IN HEART AND CORONARY VESSELS expression in vascular SMCs of the rat aorta is blocked by a MEK inhibitor, suggesting the involvement of the ERK pathway in this response (Young et al 2000). Similar results were obtained by Ohnaka and colleagues (2000), who reported that, in addition to ERK, the p38 MAP-kinase is also involved. More studies are necessary to clarify these signal transduction pathways in coronary SMCs. This is of particular interest for the pathophysiologically most important situation of acute COX-2 induction in vascular SMCs in vivo, i.e. vascular injury, subsequent to percutaneous coronary interventions (Pritchard et al 1994). PROSTAGLANDIN RECEPTORS Subtypes, Regulation and Signal Transduction After all human prostaglandin receptors had been cloned and sequenced, it was found that all major classes of these receptors are also present in the heart and coronary vessels, respectively, allowing for differential responses to the respective natural agonists (Narumiya and FitzGerald 2001). The intensity of this response depends on their number and activity state. Both can be regulated. Recently, interest has arisen in putative nuclear receptors for eicosanoids. It has been suggested that several COX products may act as agonists on peroxisome proliferation activating receptors (PPAR) (Bishop-Bailey 2000). However, it is not currently known whether or not eicosanoids, or metabolites thereof such as PGJ2, indeed function as physiological agonists on PPAR receptors or whether others of the numerous natural ligands are more important (Narumiya and FitzGerald 2001). Therefore, this issue will not be considered here in further detail. IP receptors in the heart and coronaries signal via Gs and Gq proteins and contribute to the regulation of coronary vessel tone. Because of the much higher affinity to Gs, stimulation of adenylate cyclase is the physiological response (Smyth et al 2000). Recently, IP receptor polymorphism has been described in man, eventually resulting in an impaired ligand binding and receptor activation (Stitham et al 2002). The clinical significance of this finding remains to be determined.
395
There is no evidence for transcriptional regulation of the IP receptor number in the heart or vasculature of adult cells, including the hypertrophied heart of spontaneously hypertensive rats (Nakagawa et al 1995) and isolated bovine coronary arteries subjected to 24 h of agonist exposure (Zucker et al 1998a). However, despite a stable transcription level, at least in heart and coronary vessels, the IP receptor is subject to both desensitization and downregulation, i.e. receptor internalization and degradation. The agonist-induced desensitization of the IP receptor occurs within hours in vitro, as seen from a reduced relaxation to the selective agonist cicaprost and a rightward parallel shift of the concentration–response curve, suggesting that the number of receptors was unchanged (Mu¨ller 1998). Alternatively, prevention of endogenous prostaglandin production by non-selective COX inhibitors may sensitize the smooth muscle against the agonist, suggesting some degree of downregulation by the endogenously formed PGI2 (Mu¨ller 1998) (Figure 35.1). In agreement with these findings, there is an inverse relationship between endogenous PGI2 formation and cAMP levels. The PGI2 mimetic iloprost did not stimulate cAMP formation in bovine aortic endothelial cells in the presence of significant endogenous PGI2 biosynthesis. However, there were increased cAMP levels after inhibition of COX by diclofenac or indomethacin, again pointing to an agonist-induced downregulation of the IP receptor (Schro¨der and Schro¨r 1992). Receptor downregulation is important with respect to the longterm actions of PGI2 and other prostanoids on cell growth and migration (Schro¨r 1997). Incubation of bovine coronary artery smooth muscle cells with iloprost for 24 h prevented the antimitogenic effects of cicaprost and iloprost, but not of PGE1 (Zucker et al 1998a). The cellular mechanisms of IP receptor regulation are not entirely clear and may differ between transfected cells and cells having the physiological ratio of receptor to G-protein and distal signalling cascades. In cells transfected with the IP receptor, desensitization was PKCdependent, whereas internalization was not (Smyth et al 2000). Another eicosanoid receptor of considerable interest in the heart and coronary vasculature is the TP receptor, activated by TXA2 and PG endoperoxides and signalling via different G-proteins, mainly
Figure 35.1 Sensitization and desensitization of the IP receptor by pretreatment for 4 h with cicaprost in isolated bovine coronary artery. No change of isoprenaline-induced relaxation. From Mu¨ller and Schro¨r unpublished
396
CIRCULATORY SYSTEMS
Gq (Halushka 2000). Thrombin treatment resulted in significant upregulation of the TP receptor mRNA in bovine coronary artery SMCs. This was associated with a potentiation of thrombininduced cell proliferation by a thromboxane mimetic (Zucker et al 1998b). EP receptors, coupled to the adenylate cyclase, were the first prostaglandin receptors that were identified in cardiac sarcolemmal membranes (Lopaschuk et al 1989). Further pharmacological and ligand-binding studies have identified this receptor as belonging to the EP3 subtype. These receptors inhibit adenylate cyclase activity and the increase in cAMP via a PTX-sensitive Gprotein, most likely Gi (Hohlfeld et al 1997). The receptor was upregulated by 50% in myocardial ischaemia, probably due to receptor externalization, because the upregulation was prevented by colchicine. The increase in receptor number was reversed after reperfusion, despite an intact G-protein coupling. It was suggested that this receptor mediates the cardioprotective actions of E-type prostaglandins and perhaps iloprost (see below). FP receptors for PGF2a, also coupled to Gq, mediate vasoconstrictor effects. However, in vascular SMCs and perhaps in cardiocytes, they appear to be much more important in mediating growth responses (see below) (Dorn et al 1992). EFFECTS OF PROSTAGLANDINS ON CARDIAC MUSCLE AND CORONARY VESSELS Cardiac Muscle Heart Rate and Contractility PGE2 and PGI2, synthesized in the heart, act as inhibitory modulators of b-adrenergic receptor-stimulated cardiac lipolysis (Ruan et al 1996). In addition, E-type prostaglandins are well known to inhibit catecholamine overflow after sympathetic nerve stimulation. There is little or no effect by prostaglandins or TXA2 on heart rate and contractility under physiological conditions (Schro¨r 1993). However, this picture changes significantly during myocellular injury, e.g. in ischaemia-reperfusion or heart failure. In these situations, endogenous prostaglandins, specifically PGI2, have repeatedly been shown to improve the reduced contractility of the heart in animal experiments, while TXA2 appears to have the opposite effect. This will be discussed in detail in the section on Myocardial Ischaemia and Infarction (see below). Myocyte Growth and Proliferation Growth and proliferation of cardiac myocytes are long-lasting responses that involve upregulation of COX-2 and possibly mPGES. Stimulation of growth by mitogenic factors (angiotensin II, PDGF and others) is a complex response and involves enhanced generation of eicosanoids as a modulating factor. Stimulation of neonatal rat myocardial myocyte growth was induced by exogenous PGE2, PGF2a, sulprostone or phenylephrine to a comparable extent. The effect of PGE2 was mimicked by the EP1/EP3 receptor agonist sulprostone and blocked by the EP1/EP2 receptor antagonist AH 6809. The prostacyclin mimetic beraprost was ineffective, although both beraprost and PGE2, increased cAMP levels. These data suggest that stimulation of cardiac myocyte growth by PGE2 is probably due to the activation of a different subset of EP receptors, while activation of IP receptors appears not to be involved (Mendez et al 2002). PGF2a induces myocardial hypertrophy in neonatal rat ventricular myocytes in vitro and cardiac growth in vivo (Lai et al 1996). This was associated with PLC activation, translocation
of PKC to the myocyte membrane and hydrolysis of inositol phosphates, typical for a Gq-coupled receptor. In addition, there was also activation of ERK and p38 MAP-kinases. However, pharmacological inhibitors and transfection experiments by others suggested that neither PKC nor ERK or p38 MAP-kinase were involved in PGF2a-induced growth signalling (Adams et al 1996, 1998). In this study, myocyte growth appears rather to be associated with protein tyrosine phosphorylation and JNK kinase activation, similar to phenylephrine and Ras-induced myocyte hypertrophy (Adams et al 1998). Fluprostenol, a stable analogue of PGF2a, also stimulates hypertrophy of rat cardiomyocytes (Lai et al 1996). In contrast, PGF2a does not cause cardiocyte hypertrophy or activation of Gq in mice (Deng et al 2000; HilalDandan et al 2000). This points to possible species differences that were also seen with smooth muscle cell proliferation (see below). Coronary Vessels Vasomotor Effects Prostaglandins are one group of autacoids that are involved in regulation of coronary perfusion. Coronary arteries of probably every species, including man, are relaxed by PGI2 at lower concentrations (51 mM). Higher concentrations produce contractions in the coronary arteries of pig (Dusting et al 1977) and man (Ginsburg et al 1980), possibly via Gq-coupling of the IP receptor, which only becomes relevant at high concentrations (see above). The vasodilator actions of PGI2 in isolated vessel preparations (Siegel et al 1989) and isolated hearts (Jackson et al 1993; Vesper and Schro¨r 1995) appear to be mediated by opening of KATP+ channels. This causes membrane hyperpolarization and secondary inhibition of activation of voltage-operated Ca2+ channels. It might also result in reduced generation of EDHF (Yajima et al 1999) (see below). Consequently, the vasorelaxing actions of prostacyclin are considerably reduced or even abolished in the presence of high extracellular K+ levels (Noll et al 1991), occurring, for example, locally in ischaemic areas of the myocardium (Figure 35.2). Studies in the isolated rat heart have confirmed that arachidonic acid-, PGD2-, PGE2- or iloprostinduced coronary vasodilation was inhibited by glibenclamide, suggesting that not only PGI2 but also other vasodilator prostaglandins act via opening of KATP+ channels (Bouchard et al 1994). PGE2, the other major arachidonic acid-derived COX product in the heart and coronary vasculature, has less predictable effects on coronary vessel tone. Isolated coronary arteries, such as bovine coronary arteries, are contracted. The situation in vivo is more complex; in most cases relaxation is seen. This variability is probably due to the different spectrum of EP receptors with antagonistic signalling. For PGE1 in the pig heart, a Gi-mediated inhibitory coupling to the adenylate cyclase has been demonstrated (Hohlfeld et al 1997). Similar results have been found in the rabbit and it has additionally been shown that PGE1 at submicromolar concentrations inhibits ICa2+ through L-type Ca2+ channels via a PTXsensitive mechanism after stimulation with isoprenaline (Yamamoto et al 1999), confirming earlier results on cAMP-dependent modulation of these channels in vascular smooth muscle cells (Sadoshima et al 1988). Both effects might contribute to the antiischaemic efficacy of E-type prostaglandins (see below). TXA2 has proved difficult to study because of the instability of the compound. Using thromboxane-generating systems containing considerable amounts of endoperoxides, which are equipotent ligands at the TP receptor (Halushka 2000), vasoconstriction was seen in coronary arteries (Ellis et al 1976). However, there was no high molar potency, e.g. threshold concentrations to contract pig
EICOSANOID GENERATION/EFFECTS IN HEART AND CORONARY VESSELS
397
Figure 35.2 Antagonism by high external K+ of the relaxing effects of PGI2 in isolated bovine coronary artery
coronaries were 4100 nM in vitro (Svensson and Hamberg 1976). In the presence of intact endothelium in vivo, there was only a short-lasting coronary vasoconstriction, requiring micromolar concentrations of TXA2 (Holzgrefe et al 1987). In contrast, a significant vasoconstriction is seen in areas of denuded endothelium (Folts et al 1976). In this case, when platelets become an important source of thromboxane generation, the vasoconstrictor response will be potentiated by other platelet-derived vasoconstrictors, such as serotonin (Schro¨r and Verheggen 1986). In human coronary artery segments from patients with marked endothelial dysfunction, TP receptor activation has been shown to act synergistically with serotonin via 5-HT1 receptors to cause vasoconstriction (Chester et al 1993). The contractile action of TXA2 in coronary artery smooth muscle cells is due to enhanced Ca2+ influx from the extracellular space through K+-activated Ca2+ channels (Scornik and Toro 1992), an effect opposite to PGI2 (Schro¨r 1993). Arachidonic acid-induced, endothelium-dependent relaxation of coronary arteries is mainly due to hyperpolarization and generation of transferable factor(s) (Feletou and Vanhoutte 1988). This endothelium-derived hyperpolarizing factor(s) (EDHF) was identified as EET(s) by Campbell’s group (Rosolowsky et al 1990a, 1990b; Rosolowsky and Campbell 1993, 1996; Campbell et al 1996). They showed in precontracted bovine coronary arteries that all four EETs (see above) relaxed the vessel at a comparable molar potency (IC50, 1–10 mM), being about 10-fold less potent than prostacyclin. EETs increase the open state probability of Ca2+-activated K+ channels, eventually resulting in hyperpolarization and inhibition of voltage-dependent Ca2+ influx. These effects are blocked by inhibitors of cytochrome P450 (SKF 525A, miconazole), high external K+ (20 mM) and inhibitors of Ca2+activated K+ channels (tetraethylammonium, charybdotoxin) but not by glibenclamide, a blocker of KATP+ channels, or arginine analogues that block NOS. There is no evidence for the involvement of second messengers, such as cAMP or cGMP, and no production of secondary metabolites in vitro (Campbell et al 1996). Similar effects have been described for dog (Rosolowsky et al 1990a) and pig coronary arteries (Ge et al 2000), further supporting the notion that EETs represent an EDHF. EET
(11,12) was found to partially mimic the EDHF function in coronary microarteries (Zou et al 2001), demonstrating a possible involvement in the regulation of coronary vascular resistance. The enhanced release of EETs in stenosed dog coronary arteries (Rosolowsky et al 1990b) would agree with this hypothesis.
Migration and Proliferation The contribution of different prostaglandin receptors to proliferation or hypertrophy of vascular smooth muscle cells has recently been studied in knockout mice. The data indicate that PDGFinduced proliferation is inhibited by PGE2 and PGI2 and probably mediated by the EP4 and IP receptor, respectively. Further downstream signalling for these receptors occurs via the cAMP– PKA pathway (Indolfi et al 1997). TXA2 stimulates PDGF-induced cell proliferation, while PGF2a has no effect (Fujino et al 2002). Dorn et al (1992) have previously suggested that PGF2a rather promotes hypertrophy (not proliferation) than vasoconstriction in rat vascular smooth muscle cells, while Sachinidis and co-workers (1995) showed a marked potentiation of growth responses by a thromboxane mimetic. The mitogenic signalling by prostanoids is still incompletely understood. In cultured coronary artery smooth muscle cells of guinea-pigs, thromboxane mimetics promote growth, which is preceded by activation of ERK and S6 kinase (Morinelli et al 1994). However, the ERK pathway can mediate both pro- and antiproliferative actions, depending on the availability of downstream targets, including the prostaglandin system (Bornfeldt et al 1997). Enhanced ERK-mediated cell proliferation and PGE2 production was seen in rat aortic smooth muscle cells, stimulated with angiotensin II. Both responses were blocked by COX-2-selective inhibitors. However, the possibility of enhanced thromboxane formation as a mediator was not excluded (Ohnaka et al 2000). Young et al (2000) also suggested that angiotensin II-induced proliferation of rat aortic smooth muscle cells involved COX-2dependent generation of a promitogenic factor, possibly TXA2, because a selective TP receptor antagonist blocked this response.
398
CIRCULATORY SYSTEMS
These somewhat surprising findings on the background of the probably also markedly stimulated generation of antimitogenic prostaglandins, such as PGE2 or PGI2, was explained by tolerance development (see above). However, thromboxane levels were not measured and the evidence is, therefore, indirect. Considering the numerous actions of angiotensin as a vasoconstrictor and growth factor in vascular cells, it would be interesting to know whether similar findings can also be obtained in coronary arteries. Most available evidence suggests that TXA2 is a promitogenic factor (Sachinidis et al 1995). Whether TXA2 is a mitogen by itself or primarily enhances the action of other mitogens, such as peptidergic growth factors or thrombin, appears to be speciesdependent (Fujino et al 2002). Studies in bovine coronary arteries have shown that thrombin upregulates TP receptor mRNA sixfold within 20 min and this resulted in a synergistic action of both factors on ERK activation and subsequent cell proliferation (Zucker et al 1998b). Expression of Adhesion Receptors Another area of considerable significance is the control of expression of adhesion receptors, such as integrins or immunoglobulin-type adhesion molecules (CAMs), by prostaglandins. Both adhesion receptors and their ligands appear to become upregulated by growth factors and inflammatory cytokines and trigger cell–cell interactions. Cicaprost has been found to inhibit TNFa- and IL-1b-induced expression of the adhesion molecules ICAM-1 and VCAM-1 in human coronary artery smooth muscle cells, possibly via influencing NF-kB-binding to the gene promotor region. Inhibition of this response appears to be specific for cAMP and might contribute to the antiatherosclerotic potential of prostacyclin (Braun et al 1997). Studies in human saphenous vein smooth muscle cells also indicated that COX-2derived products (PGE2) limit IL-1b-induced expression of adhesion molecules (Bishop-Bailey et al 1998). Interaction with Other Vasoactive Mediators PGI2, NO and EETs are vasodilators and will act synergistically on vascular smooth muscle cells. NO and prostaglandins also interact in control of myocardial performance in isolated cat papillary muscle (Mohan et al 1995). Vasodilatory prostaglandins, such as PGD2, release NO from coronary vascular endothelial cells and may relax coronary arteries by this mechanism (Braun and Schro¨r 1992). Experiments in dogs have shown that inhibition of NOS after 1 week of treatment with an arginine analogue resulted in enhanced prostacyclin production of coronary artery rings, which was explained by an upregulation of endothelial COX-1 (Beverelli et al 1997). In vitro studies demonstrated that shear stress-induced induction of iNOS is associated with enhanced NO and prostacyclin formation. Inhibition of NOS enhanced PGI2 production (Osanai et al 2001). These findings were confirmed in isolated hearts and, additionally, it was shown that NO, through cGMP-dependent kinase, phosphorylates a 104 kDa protein. This protein was associated with the inhibition of activity of COX-1 (Marcelin-Jimenez and Escalante 2001). Endothelial dysfunction of coronary resistance vessels in the ApoE knockout mouse mainly results in a more rapid inactivation of NO by superoxide anions. The function of the prostacyclin system in this model was not changed (Go¨decke et al 2002). The data collectively suggest a tight relationship between vasodilator prostaglandins and NO. The functional significance might lie in the adaption of coronary perfusion to the metabolic needs of the heart. NO might be more important under physiological conditions, which explains why inhibition of prostaglandin
formation by COX inhibition has only minor effects on coronary perfusion in vivo (Friedman et al 1981). Of particular interest is the interaction between peroxynitrite (ONOO7) and prostacyclin. Peroxynitrite irreversibly blocks PGI2 biosynthesis of bovine aortic microsomes within seconds, the IC50 being 50 nM (Zou and Ullrich 1996). Thus, ONOO7 is not only a potent vasoconstrictor on its own but also inhibits prostacyclin formation. This is due to tyrosine-nitrosylation of the PGI2 synthase which was shown in atherosclerotic bovine coronary artery-derived endothelial and vascular smooth muscle cells. This inhibits PGI2-dependent coronary vasodilation and promotes vasospasm (Zou et al 1999a). In early-state (human) atherosclerotic lesions there is reduced PGI2 formation at enhanced PGE2 release and unchanged COX activity. As seen with atherosclerotic bovine coronary arteries, this reduced PGI2 release is due to a selectively reduced prostacyclin synthase activity, allowing for the accumulation of vasoconstrictor PG endoperoxides (Zou et al 1999b). PGF2a, and to some extent PGE2 but not PGI2, increase the mRNA of ‘‘atrial natriuretic peptide’’ (ANP) as well as immunoreactive ANP secretion in neonatal rat atrial and ventricular cardiocytes. These effects are blocked by COX inhibition, suggesting an additional mechanism whereby eicosanoids can control cardiovascular homeostasis and which might be important for heart failure (Gardner and Schultz 1990). IMPLICATIONS OF PROSTAGLANDINS IN CARDIAC AND CORONARY DISEASES Since prostaglandins and perhaps other eicosanoids in pathological conditions are synthesized ‘‘on demand’’, they will have marked influences on the tissue pathology, because the amount at which these autacoids are available considerably exceeds their generation under physiological conditions. Interactions are possible at the level of release or action of other mediators, e.g. by influencing NO generation, release of catecholamines or generation of atrial natriuretic factor (ANP). Alternatively, eicosanoids might interfere directly with disturbed signalling of target cells via their specific receptors, e.g. via EP3, to inhibit the deleterious effect of catecholamines on cardiac metabolism and arrhythmias in myocardial ischaemia/reperfusion or by antagonizing the prothrombotic actions of enhanced, platelet-derived thromboxane on the coronary circulation via IP. Another pathophysiologically relevant issue is restenosis subsequent to PTCA-induced endothelial injury. Here, in addition to the antithrombotic events, control of smooth muscle cell migration and proliferation is involved. All these diseases have probably in common an upregulated COX-2 and possibly cPLA2, due to the promoting effects of proinflammatory cytokines and growth factors on gene transcription. Myocardial Ischaemia and Infarction Upregulation of COX-2 and Enhanced Prostaglandin Formation in Myocardial Ischaemia In healthy myocardium, COX-2 levels are low and probably restricted to the coronary endothelium. However, in induction of ischaemia in buffer-perfused hearts in vitro, i.e. in the absence of blood and corpuscular blood components such as platelets or white cells, there is a marked upregulation of COX-2 mRNA within 1 h, followed by COX-2 protein expression and PGI2 release (Schro¨r et al 1998) (Figure 35.3). Possibly ischaemia/ reperfusion-induced overflow of noradrenaline is involved, because both myocardial injury and COX-2 expression can be
EICOSANOID GENERATION/EFFECTS IN HEART AND CORONARY VESSELS
399
Figure 35.3 Expression of COX-2 mRNA and protein, 6-oxo-PGF1a release and noradrenaline overflow in ischaemic/reperfused hearts of rabbits. From Schro¨r et al (1998), published by Nature
reduced by the b-blockers pindolol and talinolol (Schro¨r et al 1998). COX-2 was also upregulated in isolated neonatal rat cardiomyocytes subjected to oxidative damage. Tissue injury could be prevented by iloprost and was aggravated by selective COX-2 inhibitors, while PGF2a was ineffective (Adderley and Fitzgerald 1999). COX-2 was also expressed in myocardial ischaemia in human patients suffering from dilatative cardiomyopathy (Wong et al 1998). These data suggest that myocellular injury in the heart is associated with the generation of protective prostanoids, mainly PGI2, which limit cell injury. An association of a single nucleotide polymorphism (C1117A) in exon 8 of the prostacyclin synthase gene with increased risk of myocardial infarction has been suggested (Nakayama et al 2002).
Actions of Prostanoids on Myocardial Ischaemia and Infarction Exogenous PGI2 protects the heart from ischaemic injury in animal experiments (Araki and Lefer 1980). As noted before, the vasodilator effects of iloprost under these controlled in vitro conditions can be separated from its cardioprotective actions because they are not modified by glibenclamide, a KATP+ channel blocker (Vesper and Schro¨r 1995) (Figure 35.4). Similar beneficial results were obtained with stimulation of endogenous prostacyclin generation by defibrotide (Hohlfeld et al 1993). This suggests that endogenous PGI2 has a cardioprotective potential, although its concentrations may not be sufficient for an optimum tissue protection. This might be due to peroxynitrite generation, eventually resulting in inhibition of prostacyclin synthase or downregulation of IP receptors at the cell membrane because of
exposure to high concentrations of PGI2. In line with these findings, doses of PGI2 or PGI2 mimetics in animal experiments were about 1000-fold higher on a weight basis than those that can be safely applied to man. However, the limited clinical experience with these compounds was rather negative, e.g. no protective effects of iloprost or PGE1 were seen in a clinical trial on heart transplant patients with severe cardiac failure (Stanek et al 1999). To clarify the relative contribution of TXA2 and PGI2, the dominating COX products released from the ischaemic heart, for tissue injury, experiments were carried out in IP and TP knockout mice. Infarct size and CK release were significantly larger in in vivo and in vitro buffer-perfused hearts in IP7/7 but not in the TP7/7 animals. It was also shown that these beneficial effects were independent of antiplatelet or antineutrophil actions of PGI2 (Xiao et al 2001). This confirms the cardioprotective actions of endogenous PGI2 in ischaemia/reperfusion but does not exclude positive effects from inhibition of the thromboxane pathway. Several studies in vivo suggest beneficial effects of thromboxane receptor antagonists and synthesis blockers. However, these effects may be largely associated with reduced platelet activation and platelet-primed activation of white cells, e.g. during reperfusion or thrombolysis (Thiemermann and Schro¨r 1984; Schro¨r and Thiemermann 1986; Vandeplassche et al 1993; Nichols et al 1993). In this context, it should be noted that neither PGI2 nor iloprost exhibits any significant spasmolytic activity against coronary artery contractions induced by platelet-derived thromboxane A2 and serotonin (Schro¨r and Verheggen 1986). There is a significant overflow of PGI2 and TXA2 from the ischaemic myocardium, indicating a possible relationship to arrhythmias (Coker et al 1981; Hirsh et al 1981). Prostacyclin has been considered an ‘‘endogenous antiarrhythmic substance’’ during ischaemia/reperfusion. Non-selective COX inhibitors did
400
CIRCULATORY SYSTEMS
Figure 35.4 Inhibition of coronary dilatory but not of cardioprotective actions of iloprost (ILO) by glibenclamide (GLI) in the Langendorff heart of rabbits subjected to ischaemia/reperfusion. Glibenclamide inhibits the ILO-induced coronary vasodilation (fall in coronary perfusion pressure, CPP) but not the cardioprotective effects as assessed from the increase in left ventricular end-diastolic pressure (LVEDP) and creatine kinase (CK). From Vesper and Schro¨r (1995). Reproduced by permission of [Birkha¨user Verlag Basel, Switzerland, from Mediators in the cardiovascular system: regional ischemia, ed. by K. Schro¨r and C.R. Pace-Asciak]
not protect from arrhythmias in vivo and abolished the antiarrhythmic action of nafazatrom and dazmegrel (Wainwright and Parratt 1991). These compounds also antagonized the antiarrhythmic effect of ischaemic preconditioning in man (Wang 2001). A clear antiarrhythmic effect has also been seen for the EP3 agonist M&B28.767 (Hohlfeld et al 2000). Another prostaglandin-mediated cardioprotective activity is the stimulation of EP3 receptors (Hohlfeld et al 1997), with subsequent inhibition of deleterious effects of ischaemia-induced catecholamine overflow. In the rabbit heart in vivo, both PGE1 and the EP1/EP3 agonist sulprostone reduced infarct size. In contrast to iloprost (Vesper and Schro¨r 1995), this effect was antagonized by KATP channel blockade (Hide et al 1995; Thiemermann and Zacharowski 2000). However, in the pig ischaemia model in vivo, glibenclamide did not antagonize the action potential shortening of PGE1, indicating that ventricular KATP+ channels were not involved (Hohlfeld et al 2000). In this respect, it is interesting to note that glibenclamide is a competitive antagonist of the TP receptor in dog coronary arteries (Cocks et al 1990). Both PGE1 and the selective EP3 agonist M&B28.767 reduced infarct size and ischaemia-related ventricular arrhythmias in the pig. In addition, these compounds inhibited the isoprenalineinduced inotropic responses, as well as the forskolin-induced increase in cAMP in EP3 receptor-transfected CHO cells. These data, as well as studies in the rat (Zacharowski et al 1999), suggest that E-type prostaglandins protect from ischaemia/reperfusion injury by a combined activation of repolarizing membrane currents and an inhibition of the deleterious effects of ischaemia-induced catecholamine overflow. Both actions appear to be mediated by EP3 receptors (Hohlfeld et al 2000). Possible modes of antiischaemic action via activation of inhibitory G-proteins involve: (a) antagonism of the detrimental effects of ischaemia-induced catecholamine overflow (Schro¨r et al 1981; Schro¨r and Funke 1985); (b) opening of ATP-sensitive K+channels (Kirsch et al 1990; Hide and Thiemermann 1996) that are also involved in myocardial protection by ischaemic preconditioning (Gross and Auchampach 1992; Hide and Thiemermann 1996); (c) inhibition of (excessive) Ca2+ entry during reperfusion via blockade of L-type Ca2+ channels (Yamamoto et al 1999).
Remodelling and Restenosis Cardiac fibroblasts, as the source of extracellular matrix for the left ventricle, subserve important functions for cardiac remodelling and fibrotic development following myocardial infarction or with pressure-overloaded cardiac hypertrophy (Yu et al 1997). In spontaneously hypertensive SHR rats, there was a greater arachidonic acid release and PGI2 formation than in normotensive animals. Beraprost, a PGI2 analogue, reduced growth both under resting conditions and after stimulation with angiotensin II. In addition, DNA synthesis and expression of collagens type I and III were also reduced. These data suggest that eicosanoids, particularly PGI2, are involved in collagen formation and, therefore, may control cardiac remodelling (Yu et al 1997). Later work by these authors demonstrated similar effects on PGI2 and collagen formation after stimulation of rat cardiac fibroblasts with bradykinin (Gallagher et al 1998). Bradykinininduced prostaglandin formation appears also to reduce fibrillar collagen formation after myocardial infarction (Carvalho Frimm et al 1996). These and other data (Brilla et al 1995) indicate that the neurohumoral response in hypertensive heart disease or myocardial infarction with activation of the cardiac or circulatory renin–angiotensin system is modulated by vasodilator prostaglandins, such as PGI2 or PGE2 (Brilla et al 1995). This hypothesis is further underlined by the finding that aspirin, at doses that do not affect cardiac hypertrophy in the rat, nevertheless affects interstitial fibrosis (Kalkman et al 1995). Hypoxia and experimental myocardial infarction are associated with elevated levels of PGF2a in the myocardium (Oudot et al 1995; Lai et al 1996). This might result in compensatory cardiac myocyte hypertrophy in animal models of myocardial ischaemia, resulting at least in part from the release of growth factors from damaged or stressed cells of the myocardium (Adams et al 1998). This will also stimulate cardiac growth in vivo (Lai et al 1996). Another aspect of growth-modulating actions of eicosanoids are their effects on restenosis, an issue of great relevance for coronary arteries after percutaneous coronary interventions. Prostacyclin synthase gene transfer, similar to the prostacyclin mimetic beraprost, was shown to inhibit neointimal formation in the rat after arterial injury by inhibiting proliferation of vascular smooth muscle cells (Harada et al 1999). A prostacyclin analogue
EICOSANOID GENERATION/EFFECTS IN HEART AND CORONARY VESSELS has been shown to inhibit intima hyperplasia of dog coronary arteries via inhibition of the cell cycle, specifically by preventing downregulation of p27kip1, a cyclin-dependent kinase (cdk) inhibitor (Ii et al 2001). On the other hand, the TP receptor may become up-regulated by thrombin (Zucker et al 1998b) or other growth factors and there is also evidence that TXA2 might ‘‘prime’’ other growth factors, e.g. PDGF, eventually resulting in a potentiation of mitogenesis (Grosser et al 1997). However, despite some positive data with beraprost, no clear improvement of restenosis by any eicosanoid has so far been convincingly demonstrated in man (Isogaya et al 1995), and this might be due at least partially to the complex pattern of eicosanoid formation under these conditions.
Pharmacological Implications—Selective COX-2 Inhibitors Enhanced COX-2-dependent PGI2 formation can be expected in patients suffering from chronic inflammatory diseases, such as atherosclerosis, possibly explaining the long-known enhanced prostacyclin generation in these patients (FitzGerald et al 1984). Specifically there is a marked upregulation of COX-2 in proliferating vascular smooth muscle cells and macrophages in the atherosclerotic plaques. Both COX-1 and COX-2 contribute to PGI2 production in these atherosclerotic patients, whereas thromboxane production depends only on COX-1 (Belton et al 2000). There is also an induction of COX-2 in the myocardium of patients with heart failure and in the atrial arterioles of patients undergoing cardiopulmonary bypass surgery (Wong et al 1998; Metais et al 2001), probably induced via the nuclear transcription factor NF-kB. In addition, non-selective COX inhibitors increased infarct size. A possible exception is ibuprofen, which was repeatedly shown to have a cardioprotective effect in animal experiments (Lefer and Polansky 1979; Judgutt et al 1980; Romson et al 1980). The reason for this is unclear but might be related to non-COXmediated actions of the compound, e.g. inhibition of certain transcription factors, such as NF-kB or AP-1. There were no such effects with indomethacin (Tegeder et al 2001). Studies in the dog suggest that COX-2 inhibitors might increase the risk of thrombotic coronary artery occlusion after endothelial injury and that this effect might be due to inhibition of COX-2dependent PGI2 formation (Hennan et al 2001). In these experiments, administration of the COX-2 inhibitor celexocib did not change the time to occlusive arterial thrombus formation or the bleeding time, but did prevent the beneficial effects of aspirin (4.6 mg/kg) on these parameters, if aspirin was given long enough (17 h) before celecoxib to allow for endothelial recovery of aspirin-inhibited COX-2 activity. These data suggest that vascular COX-2 is involved in the antithrombotic effects of aspirin in this model and that this response is sensitive to COX-2 inhibition. In conscious rabbits, the effect of COX-2 inhibition on the beneficial effects of ischaemic preconditioning and the severity of myocardial ischaemia was studied. Ischaemic preconditioning was associated with an upregulation of COX-2 mRNA and protein and with an enhanced PGI2 and PGE2 formation at unchanged thromboxane levels. COX-2 inhibition prevented the beneficial effects of late ischaemic preconditioning, suggesting that COX-2derived prostaglandins, probably PGI2, prevented the reduction in infarct size induced by preconditioning. Again, COX-2 inhibition per se did not change infarct size (Shinamura et al 2000). The possible interference of COX-2 with iNOS, another enzyme that mediates late ischaemic preconditioning (LaPointe and Sitkins 1998), was not studied. However, iNOS-derived products have previously been shown to stimulate COX-2 (Salvemini et al 1993).
401
Thus, an inhibition of COX-2 with subsequent reduction of endogenous PGI2 biosynthesis may be associated with an increase in adverse clinical outcome, especially in atherosclerotic individuals predisposed to vasculopathy and thrombosis (Hennan et al 2001; FitzGerald 2002).
ACKNOWLEDGEMENTS The work of the author’s group which is contained in this review was sponsored by several grants from the Deutsche Forschungsgemeinschaft, the SFB 242 and the Forschungsgruppe HerzKreislauf e.V. Du¨sseldorf. The author thanks Dr Jens Fischer, Dr Thomas Hohlfeld and Dr Artur-Aron Weber for critical reading of the manuscript and numerous valuable comments and suggestions. The author also thanks Erika Lohmann for competent secretarial assistance and Petra Kuger for preparation of the figures.
REFERENCES Adams JW, Migita DS, Yu MK et al (1996) Prostaglandin F2a stimulates hypertrophic growth of cultured neonatal rat ventricular myocytes. J Biol Chem, 271, 1179–1186. Adams JW, Sah VP, Henderson SA and Brown JH (1998) Tyrosine kinase and c-Jun NH2-terminal kinase mediate hypertrophic responses to prostaglandin F2a in cultured neonatal rat ventricular myocytes. Circ Res, 83, 167–178. Adderley SR and Fitzgerald DJ (1999) Oxidative damage of cardiomyocytes is limited by extracellular regulated kinases ½mediated induction of cyclooxygenase-2. J Biol Chem, 274, 5038–5046. Araki H and Lefer AM (1980) Role of prostacyclin in the preservation of ischemic myocardial tissue in the perfused cat heart. Circ Res, 47, 757– 763. Belton O, Byrne D, Kearney D et al (2000) Cyclooxygenase-1 and -2dependent prostacyclin formation in patients with atherosclerosis. Circulation, 102, 840–845. Berthiaume F and Frangos JA (1992) Flow-induced prostacyclin production is mediated by a pertussis toxin-sensitive G-protein. FEBS Lett, 308, 277–279. Beverelli F, Bea ML, Puybasset L et al (1997) Chronic inhibition of NO synthase enhances the production of prostacyclin in coronary arteries through upregulation of the cyclooxygenase type 1 isoform. Fund Clin Pharmacol, 11, 252–259. Bishop-Bailey D (2000) Peroxisome proliferator-activated receptors in the cardiovascular system. Br J Pharmacol, 129, 823–834. Bishop-Bailey D, Burke-Gaffney A, Hellewell PG et al (1998) Cyclooxygenase-2 regulates inducible ICAM-1 and VCAM-1 expression in human vascular smooth muscle cells. Biochem Biophys Res Commun, 249, 44–47. Bolton HS, Chanderbhan R, Bryant RW et al (1980) Prostaglandin synthesis by adult heart myocytes. J Mol Cell Cardiol, 12, 1287–1298. Bornfeldt KE, Campbell JS, Koyama H et al (1997) The mitogen-activated protein kinase pathway can mediate growth inhibition and proliferation in smooth muscle cells: dependence on the availability of downstream targets. J Clin Invest, 100, 875–885. Bouchard JF, Dumont E and Lamontagne D (1994) Evidence that prostaglandins I2, E2, and D2 may activate ATP sensitive potassium channels in the isolated rat heart. Cardiovasc Res, 28, 901–905. Braun M and Schro¨r K (1992) Prostaglandin D2 relaxes bovine coronary arteries by endothelium-dependent nitric oxide-mediated cGMP formation. Circ Res, 71, 1305–1313. Braun M, Pietsch P, Zepp A et al (1997) Regulation of tumor necrosis factor-a and interleukin-1b induced adhesion molecule expression in human vascular smooth muscle cells by cAMP. Arterioscler, Thromb, and Vasc Biol, 17, 2568–2575. Brilla CG, Zhou G, Rupp H et al (1995) Role of angiotensin II and prostaglandin E2 in regulation cardiac fibroblast collagen turnover. Am J Cardiol, 76, 8D–13D.
402
CIRCULATORY SYSTEMS
Campbell JH and Campbell GR (1993) Culture techniques and their applications to studies of vascular smooth muscle. Clin Sci, 85, 501– 513. Campbell WB, Gebremedhin D, Pratt PF and Harder DR (1996) Identification of epoxyeicosatrienoic acids as endothelium-derived hyperpolarizing factors. Circ Res, 78, 415–423. Carvalho Frimm de C, Sun Y and Weber KT (1996) Wound healing following myocardial infarction in the rat: role for bradykinin and prostaglandins. J Mol Cell Cardiol, 28, 1279–1285. Chester A-H, Allen SP, Tadjkarimi S and Yacoub M-H (1993) Interaction between thromboxane A2 and 5-hydroxytryptamine receptor subtypes in human coronary arteries. Circulation, 87, 874–880. Chierchia S, Patrono C, Crea F et al (1982) Effects of intravenous prostacyclin in variant angina. Circulation, 65, 470–476. Church DJ, Van den Bent V, Vallotton MB and Lang U (1994) Role of prostaglandin-mediated cyclic AMP formation in protein kinase Cdependent secretion of atrial natriuretic peptide in rat cardiomyocytes. Biochem J, 303, 217–225. Cocks TM, King SJ and Angus JA (1990) Glibenclamide is a competitive antagonist of the thromboxane A2 receptor in dog coronary artery in vitro. Br J Pharmacol, 100, 375–378. Coker SJ, Parratt JR, Ledingham IM and Zeitlin IJ (1981) Thromboxane and prostacyclin release from ischaemic myocardium in relation to arrhythmias. Nature, 291, 323–324. Coleman TA, Smith WL and Narumiya S (1994) VIIIth International Union of Pharmacology, classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. De Deckere EAM, Nugteren DH and Ten Hoor F (1977) Prostacyclin is the major prostaglandin released from the isolated perfused rabbit and rat heart. Nature, 268, 160–163. Deng X-F, Rokosh DG and Simpson PC (2000) Autonomous and growth factor-induced hypertrophy in cultured neonatal mouse cardiac myocytes. Circ Res, 87, 781–788. DeWitt DL, Day JS, Sonnenburg WK and Smith WL (1983) Concentrations of prostaglandin endoperoxide synthase and prostaglandin I2 synthase in the endothelium and smooth muscle of bovine aorta. J Clin Invest, 72, 1882–1888. Dorn GW II, Becker MW and Davis MG (1992) Dissociation of the contractile and hypertrophic effects of vasoconstrictor prostanoids in vascular smooth muscle. J Biol Chem, 267, 24897–24905. Dusting GJ, Moncada S and Vane JR (1977) Prostacyclin (PGI2) is a weak contractor of coronary arteries of the pig. Eur J Pharmacol, 45, 301– 304. Eldor A, Falcone DJ, Hajjar DP et al (1981) Recovery of prostacyclin production by the de-endothelialized rabbit aorta: critical role of neointimal smooth muscle cells. J Clin Invest, 67, 735–741. Ellis EF, Oelz O, Roberts LJ II et al (1976) Coronary arterial smooth muscle contraction by a substance released from platelets: evidence that it is thromboxane A2. Science, 193, 1135–1137. Feletou M and Vanhoutte PM (1988) Endothelium-dependent hyperpolarization of canine coronary smooth muscle. Br J Pharmacol, 93, 515–524. Fisslthaler B, Popp R, Michaelis UR et al (2001) Cyclic stretch enhances the expression and activity of coronary endothelium-derived hyperpolarizing factor synthase. Hypertension, 38, 1427–1432. FitzGerald GA (2002) Cardiovascular pharmacology of non-selective nonsteroidal antiinflammatory drugs and coxibs: clinical considerations. Am J Cardiol, 89(suppl), 26D–32D. FitzGerald GA, Smith B, Pedersen AK and Brash AR (1984) Prostacyclin biosynthesis is increased in patients with severe atherosclerosis and platelet activation. N Engl J Med, 310, 1065–1068. Folts JD, Crowell ED and Rowe GG (1976) Platelet aggregation in partially obstructed vessels and its elimination with aspirin. Circulation, 54, 365–370. Friedman PL, Brown EJ Jr, Gunther S et al (1981) Coronary vasoconstrictor effect of indomethacin in patients with coronary artery disease. N Engl J Med, 305, 1171–1175. Fujino T, Yuhki K-I, Yamada T et al (2002) Effects of the prostanoids on the proliferation or hypertrophy of cultured murine aortic smooth muscle cells. Br J Pharmacol, 136, 530–539. Gallagher AM, Yu H and Printz MP (1998) Bradykinin-induced reductions in collagen gene expression involve prostacyclin. Hypertension, 32, 84–88.
Gardner DG and Schultz HD (1990) Prostaglandins regulate the synthesis and secretion of the atrial natriuretic peptide. J Clin Invest, 86, 52–59. Ge ZD, Zhang XH, Fung PC and He GW (2000) Endothelium-dependent hyperpolarization and relaxation resistance to N(G)-nitro-L-arginine and indomethacin in coronary circulation. Cardiovasc Res, 46, 547– 556. Ginsburg R, Bristow MR, Harrison DC and Stinson EB (1980) Studies with isolated human coronary arteries. Some general observations, potential mediators of spasm, role of calcium antagonists. Chest, 78(suppl 1), 180–186. Go¨decke A, Ziegler M, Ding Z and Schrader J (2002) Endothelial dysfunction of coronary resistance vessels in apoE7/7 mice involves NO but not prostacyclin-dependent mechanisms. Cardiovasc Res, 53, 253–262. Gross GJ and Auchampach JA (1992) Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circ Res, 70, 223–232. Grosser T, Zucker T-P, Weber A-A et al (1997) Thromboxane A2 induces cell signaling but requires platelet-derived growth factor to act as a mitogen. Eur J Pharmacol, 319, 327–332. Halushka PV (2000) Thromboxane A2 receptors: where have you gone? Prostagland Lipid Mediat, 60, 175–189. Hanada T, Hashimoto M, Nosaka S et al (2000) Shear stress enhances prostacyclin release from endocardial endothelial cells. Life Sci, 66, 215–220. Harada M, Toki Y, Numaguchi Y et al (1999) Prostacyclin synthase gene transfer inhibits neointimal formation in rat balloon-injured arteries without bleeding complications. Cardiovasc Res, 43, 481–491. Hennan JK, Huang J, Barrett TD et al (2001) Effects of selective cyclooxygenase-2 inhibition on vascular responses and thrombosis in canine coronary arteries. Circulation, 104, 820–825. Hide EJ and Thiemermann C (1996) Sulprostone-induced reduction of myocardial infarct size in the rabbit by activation of ATP-sensitive potassium channels. Br J Pharmacol, 118, 1409–1414. Hide EJ, Ney P, Piper J et al (1995) Reduction by prostaglandin E1 or prostaglandin Eo of myocardial infarct size in the rabbit by activation of ATP-sensitive potassium channels. Br J Pharmacol, 116, 2435–2440. Hilal-Dandan R, Kanter JR and Brunton LL (2000) Characterization of G-protein signaling in ventricular myocytes from the adult mouse heart: differences from the rat. J Mol Cell Cardiol, 32, 1211–1221. Hirsh PD, Hillis LD, Campbell WB et al (1981) Release of prostaglandins and thromboxanes into the coronary circulation in patients with ischemic heart disease. N Engl J Med, 304, 686–691. Hohlfeld T, Strobach H and Schro¨r K (1993) Stimulation of endogenous prostacyclin protects the reperfused ischemic pig myocardium from ischemic injury. J Pharmacol Exp Therapeut, 264, 397–405. Hohlfeld T, Zucker T-P, Meyer J and Schro¨r K (1997) Expression, function, and regulation of E-type prostaglandin receptors (EP3) in the non-ischemic and ischemic pig heart. Circ Res, 81, 765–773. Hohlfeld T, Meyer-Kirchrath J, Vogel YC and Schro¨r K (2000) Reduction of infarct size by selective stimulation of prostaglandin EP(3) receptors in the reperfused ischemic pig heart. J Mol Cell Cardiol, 32, 285–296. Holzgrefe HH, Buchanan LV and Bunting S (1987) In vivo characterization of synthetic thromboxane A2 in canine myocardium. Circ Res, 60, 290–296. Ii M, Hoshiga M, Fukui R et al (2001) Beraprost sodium regulates cell cycle in vascular smooth muscle cells through cAMP signaling by preventing downregulation of p27(Kip1). Cardiovasc Res, 52, 500–508. Indolfi C, Avvedimento EV, Di Lorenzo E et al (1997) Activation of cAMP-PKA signaling in vivo inhibits smooth muscle cell proliferation induced by vascular injury. Nature Med, 3, 775–779. Isogaya M, Yamada N, Koike H et al (1995) Inhibition of restenosis by beraprost sodium (a prostaglandin I2 analogue) in the atherosclerotic rabbit artery after angioplasty. J Cardiovasc Pharmacol, 25, 947–952. Jackson WF, Ko¨nig A, Dambacher TH and Busse R (1993) Prostacyclininduced vasodilation in rabbit heart is mediated by ATP-sensitive potassium channels. Am J Physiol, 264, H238–H243. Judgutt BI, Hutchins GM, Bulkley BH and Becker LC (1980) Salvage of ischemic myocardium by ibuprofen during infarction in the conscious dog. Am J Cardiol, 46, 74–82. Kalkman EA, Sylen van RJ, Dijk van JP et al (1995) Chronic aspirin treatment affects collagen deposition in non-infarcted myocardium during remodeling after coronary artery ligation in the rat. J Mol Cell Cardiol, 27, 2483–2494.
EICOSANOID GENERATION/EFFECTS IN HEART AND CORONARY VESSELS Kan H, Ruan Y and Malik KU (1996) Involvement of mitogen-activated protein kinase and translocation of cytosolic phospholipase A2 to the nuclear envelope in acetylcholine-induced prostacyclin synthesis in rabbit coronary endothelial cells. Mol Pharmacol, 50, 1139–1147. Kan H, Ruan Y and Malik KU (1997) Signal transduction mechanism(s) involved in prostacyclin production elicited by acetylcholine in coronary endothelial cells of rabbit heart. J Pharmacol Exp Therapeut, 282, 113–122. Kirsch GE, Codina J, Birnbaumer L and Brown AM (1990) Coupling of ATP-sensitive K+-channels to A1-receptors by G-proteins in rat ventricular myocytes. Am J Physiol, 259, H820–H826. Kraemer RJ and Folts JD (1973) Release of prostaglandins following temporary occlusion of the coronary artery. Fed Proc, 32, 454. Lai J, Jin H, Yang R et al (1996) Prostaglandin F2 induces cardiac myocyte hypertrophy in vitro and cardiac growth in vivo. Am J Physiol, 271, H2197–H2208. LaPointe MC and Sitkins JR (1998) Phospholipase A2-metabolites regulate inducible nitric oxide synthase in myocytes. Hypertension, 31, 218–224. Lefer AM and Polansky WE (1979) Beneficial effects of ibuprofen in acute myocardial ischemia. Cardiology, 64, 265–279. Limas CJ and Cohn JN (1973) Isolation and properties of myocardial prostaglandin synthetase. Cardiovasc Res, 7, 623–628. Lin L-L, Lin AY and Knopf JL (1992) Cytosolic phospholipase A2 is coupled to hormonally regulated release of arachidonic acid. Proc Natl Acad Sci USA, 89, 6147–6151. Lopaschuk GD, Michalak M, Wandler EL et al (1989) Prostaglandin E receptors in cardiac sarcolemma—identification and coupling to adenylate cyclase. Circ Res, 65, 538–545. Marcelin-Jimenez G and Escalante B (2001) Functional and cellular interactions between nitric oxide and prostacyclin. Comp Biochem Physiol Toxicol Pharmacol CBP, 129, 349–359. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2. The human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. Mendez M and LaPointe MC (2002) Trophic effects of the cyclooxygenase-2 product prostaglandin E(2) in cardiac myocytes. Hypertension, 39, 382–388. Metais C, Bianchi C, Li J et al (2001) Serotonin-induced human coronary microvascular contraction during acute myocardial ischemia is blocked by COX-2 inhibition. Basic Res Cardiol, 96, 59–67. Mohan P, Brutsaert DL and Sys SU (1995) Myocardial performance is modulated by interaction of cardiac endothelium-derived nitric oxide and prostaglandins. Cardiovasc Res, 29, 637–640. Morinelli TA, Zhang LM, Newman WH and Meier KE (1994) Thromboxane A2/prostglandin H2-stimulated mitogenesis of coronary artery smooth muscle cells involves activation of mitogenactivated protein kinase and S6 kinase. J Biol Chem, 269, 5693–5698. Mu¨ller M (1998) Homologe Desensitisierung des IP-Rezeptors boviner Koronararterien. Inauguraldissertation, Medizinische Fakulta¨t der Heinrich-Heine-Universita¨t, Du¨sseldorf. Murakami M, Naraba H, Tanioka T et al (2000) Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem, 275, 32783–32792. Nakagawa O, Sasaki Y, Tanaka I et al (1995) Gene expression of prostacyclin receptor in the hypertrophied heart of spontaneously hypertensive rats. Clin Exp Pharmacol Physiol, 22(suppl 1), S270–272. Nakayama T, Soma M, Saito S et al (2002) Association of a novel single nucleotide polymorphism of the prostacyclin synthase gene with myocardial infarction. Am Heart J, 143, 797–801. Narumiya S and FitzGerald GA (2001) Genetic and pharmacological analysis of prostanoid receptor function. J Clin Invest, 108, 25–30. Nichols WW, Nicolini FA, Khan S et al (1993) Failure of prostacyclin analog iloprost to sustain coronary blood flow after recombinant tissue-type plasminogen-induced thrombolysis in dogs. Am Heart J, 126, 285–292. Noll G, Buehler FR, Yang Z and Lu¨scher TF (1991) Different potency of endothelium-derived relaxing factors against thromboxane, endothelin, and potassium chloride in intramyocardial porcine coronary arteries. J Cardiovasc Pharmacol, 18,120–126.
403
Ohnaka K, Numaguchi K, Yamakawa T and Inagami T (2000) Induction of cyclooxygenase-2 by angiotensin II in cultured rat vascular smooth muscle cells. Hypertension, 35, 68–75. Osanai T, Akutsu N, Fujita N et al (2001) Cross-talk between prostacyclin and nitric oxide under shear in smooth muscle cell: role in monocyte adhesion. Am J Physiol Heart Circ Physiol, 281, H177–182. Oudot F, Grynberg A and Sergiel KJ (1995) Eicosanoid synthesis in cardiomyocytes: influence of hypoxia, reoxygenation, and polyunsaturated fatty acids. Am J Physiol Heart Circ Physiol, 268, 308–315. Pritchard KA Jr, O’Banion MK, Miano JM et al (1994) Induction of cyclooxygenase-2 in rat vascular smooth muscle cells in vitro and in vivo. J Biol Chem, 269, 8504–8509. Quintaje SB, Rebsamen M, Church DJ et al (1998) MAP kinase mediates epidermal growth factor- and phorbol ester-induced prostacyclin formation in cardiomyocytes. J Mol Cell Cardiol, 30, 933–945. Rimarachin JA, Jacobson JA, Szabo P et al (1994) Regulation of cyclooxygenase-2 expression in aortic smooth muscle cells. Arterioscler Thromb, 14, 1021–1031. Romson JL, Bush LR, Hooch DW and Lucchesi BR (1980) The beneficial effects of oral ibuprofen on coronary artery thrombosis and myocardial ischemia in the conscious dog. J Pharmacol Exp Therapeut, 215, 271–278. Rosolowsky M and Campbell WB (1993) Role of PGI2 and EETs in the relaxation of bovine coronary arteries to arachidonic acid. Am J Physiol, 264, H327–H335. Rosolowsky M and Campbell WB (1996) Synthesis of hydroxyeicosatetraenoic (HETEs) and epoxyeicosatrienoic acids (EETs) by cultured bovine coronary artery endothelial cells. Biochim Biophys Acta, 1299, 267–277. Rosolowsky M, Falck JR and Campbell WB (1990a) Synthesis and biological activity of epoxyeicosatrienoic acid (EETs) by cultured bovine coronary artery endothelial cells. Adv Prostagland Thrombox Leukotriene Res, 21, 213–216. Rosolowsky M, Falck JR, Willerson JT and Campbell WB (1990b) Synthesis of lipoxygenase and epoxygenase products of arachidonic acid by normal and stenosed canine coronary arteries Circ Res, 66, 608–621. Ruan Y, Kan H, Cano C and Malik KU (1996) Modulation of badrenergic receptor-stimulated lipolysis in the heart by prostaglandins. Am J Physiol, 271, E556–562. Sachinidis A, Flesch M, Ko Y et al (1995) Thromboxane A2 and vascular smooth muscle cell proliferation. Hypertension, 26, 771–780. Sadoshima J-I, Akaike N, Kanaide H and Nakamura M (1988) Cyclic AMP modulates Ca-activated K channel in cultured smooth muscle cells of rat aortas. Am J Physiol, 255, H754–H759. Salvemini D, Misko TP, Masferrer JL et al (1993) Nitric oxide activates cyclooxygenase enzymes. Proc Natl Acad Sci USA, 90, 7240–7244. Schro¨der H and Schro¨r K (1992) Stimulation of cAMP by prostacyclin in endothelial cells. Naunyn-Schmiedeberg’s Arch Pharmacol, 347, 101– 104. Schro¨r K (1993) The effect of prostaglandins and thromboxane A2 on coronary vessel tone—mechanisms of action and therapeutic implications. Eur Heart J, 14(suppl 1), 34–41. Schro¨r K (1997) Prostacyclin (prostaglandin I2) and atherosclerosis. In Rubanyi GM and Dzau VJ (eds), The Endothelium in Clinical Practice. New York: Marcel Decker, 1–44. Schro¨r K and Funke K (1985) Prostaglandins and myocardial noradrenaline overflow after sympathetic nerve stimulation during ischemia and reperfusion. J Cardiovasc Pharmacol, 7(suppl 5), S50– S54. Schro¨r K and Thiemermann C (1986) Treatment of acute myocardial ischaemia with a selective antagonist of thromboxane receptors (BM 13.177). Br J Pharmacol, 87, 631–637. Schro¨r K and Verheggen R (1986) Prostacyclins are only weak antagonists of coronary vasospasm induced by authentic thromboxane A2 and serotonin. J Cardiovasc Pharmacol, 8, 607–613. Schro¨r K, Moncada S, Ubatuba FB and Vane JR (1977) Formation of prostacyclin (PGX) causes decrease in the coronary vascular resistance during application of arachidonic acid in the rabbit isolated heart. Naunyn-Schmiedeberg’s Arch Pharmacol, 297(suppl II), 124, R 31. Schro¨r K, Addicks K, Darius H et al (1981) PGI2 inhibits ischemiainduced platelet activation and prevents myocardial damage by inhibition of catecholamine release from adrenergic nerve terminals.
404
CIRCULATORY SYSTEMS
Evidence for cAMP as common denominator. Thromb Res, 21, 175– 180. Schro¨r K, Zimmermann KC and Tannha¨user R (1998) Augmented myocardial ischemia by nicotine—mechanisms and their possible significance. Br J Pharmacol, 125, 79–86. Scornik FS and Toro L (1992) U 46619, a thromboxane A2 agonist, inhibits KCa channel activity from pig coronary artery. Am J Physiol, 262, C708–C713. Shinamura K, Tang X-L, Wang Y et al (2000) Cyclooxygenase-2 mediates the cardioprotective effects of the late phase of ischemic preconditioning in conscious rabbits. Proc Natl Acad Sci USA, 97, 10197–10202. Siegel G, Carl A, Adler A and Stock G (1989) Effect of the prostacyclin analogue iloprost on K+ permeability in the smooth muscle cells of the canine carotid artery. Eicosanoids, 2, 213–222. Smith WL (1986) Prostaglandin biosynthesis and its compartmentation in vascular smooth muscle and endothelial cells. Ann Rev Physiol, 48, 251–262. Smith WL (1992) Prostanoid biosynthesis and mechanisms of action. Am J Physiol, 263, F181–F191. Smyth EM, Austin SC, Reilly MP and FitzGerald G (2000) Internalization and sequestration of the human prostacyclin receptor. J Biol Chem, 275, 32037–32045. Soler M, Camacho M, Escudero J-R et al (2000) Human vascular smooth muscle cells but not endothelial cells express prostaglandin E synthase. Circ Res, 87, 504–507. Stanek B, Sturm B, Frey B et al (1999) Bridging to heart transplantation: prostaglandin E1 vs prostacyclin versus dobutamine. J Heart Lung Transpl, 8, 358–366. Stitham J, Stojanovic A and Hwa J (2002) Impaired receptor binding and activation associated with a human prostacyclin receptor polymorphism. J Biol Chem, 277, 15439–15444. Svensson J and Hamberg M (1976) Thromboxane A2 and prostaglandin H2: potent stimulators of the swine coronary artery. Prostaglandins, 12, 943–950. Tanioka T, Nakatani Y, Semmyo N et al (2000) Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem, 275, 32775–32782. Tegeder I, Pfeilschifter J and Geisslinger G (2001) Cyclooxygenaseindependent actions of cyclooxygenase inhibitors. FASEB J, 15, 2057– 2072. Thiemermann C and Schro¨r K (1984) Comparison of the thromboxane synthetase inhibitor dazoxiben and the prostacyclin mimetic iloprost in an animal model of acute ischaemia and reperfusion. Biomed Biochim Acta, 43, S151–S154. Thiemermann C and Zacharowski K (2000) Selective activation of E-type prostanoid(3)-receptors reduces myocardial infarct size. A novel insight into the cardioprotective effects of prostaglandins. Pharmacol Therapeut, 87, 61–67. Tyagi MG, Kan H, Ruan Y and Malik KU (1996) Studies on the characterization of the subtype(s) of muscarinic receptor involved in prostacyclin synthesis in rabbit cardiomyocytes. J Recept Signal Transduct Res, 16, 273–296. Vandeplassche G, Hermans C, Somers Y et al (1993) Combined thromboxane A2 synthase inhibition and prostaglandin endoperoxide receptor antagonism limits myocardial infarct size after mechanical
coronary occlusion and reperfusion at doses enhancing coronary thrombolysis by streptokinase. J Am Coll Cardiol, 21, 1269–1279. Vesper A and Schro¨r K (1995) The cardioprotective actions of iloprost in myocardial ischemia of the rabbit can be separated from its vasodilatory effects mediated by KATP+-channel opening. In Schro¨r K and Pace-Asciak CR (eds), Mediators in the Cardiovascular System: Regional Ischemia. Agents Actions Suppl, 45, 93–99. Wainwright CL and Parratt JR (1991) Failure of cyclooxygenase inhibition to protect against arrhythmias induced by ischaemia and reperfusion: implications for the role of prostaglandins as endogenous myocardial protective substances. Cardiovasc Res, 25, 93–100. Wang L (2001) The role of prostaglandins in the antiarrhythmic effect of ischemic preconditioning. J Biomed Sci, 8, 406–410. Wong SCY, Fukuchi M, Melnyk P et al (1998) Induction of cyclooxygenase-2 and activation of nuclear factor-kB in myocardium of patients with congestive heart failure. Circulation, 98, 100–103. Xiao CA, Hara A, Yuhki K et al (2001) Roles of prostaglandin I(2) and thromboxane A(2) in cardiac ischemia-reperfusion injury: a study using mice lacking their respective receptors. Circulation, 104, 2210– 2215. Yajima K, Nishiyama M, Yamamoto Y and Suzuki H (1999) Inhibition of endothelium-dependent hyperpolarization by endothelial prostanoids in guinea-pig coronary artery. Br J Pharmacol, 126, 1–10. Yamamoto T, Habuchi Y, Tanaka H et al (1999) EP receptor-mediated inhibition by prostaglandin E(1) of cardiac L-type Ca(2+) current of rabbits. Am J Physiol, 277, H1369–H1374. Young W, Mahboubi K, Haider A et al (2000) Cyclooxygenase-2 is required for tumor necrosis factor-a and angiotensin II-mediated proliferation of vascular smooth muscle cells. Circ Res, 86, 906–914. Yu H, Gallagher AM, Garfin PM and Printz MP (1997) Prostacyclin release by rat cardiac fibroblasts: inhibition of collagen expression. Hypertension, 30, 1047–1053. Zacharowski K, Olbrich A, Piper J et al (1999) Selective activation of the prostanoid EP(3) receptor reduces myocardial infarct size in rodents. Arterioscler Thromb Vasc Biol, 19, 2141–2147. Zou M-H and Ullrich V (1996) Peroxynitrite formed by simultaneous generation of nitric oxide and superoxide selectively inhibits bovine aortic prostacyclin synthase. FEBS Lett, 382, 101–104. Zou M-H, Jendral M and Ullrich V (1999a) Prostaglandin endoperoxidedependent vasospasm in bovine coronary arteries after nitration of prostacyclin synthase. Br J Pharmacol, 126, 1283–1292. Zou M-H, Leist M and Ullrich V (1999b) Selective nitration of prostacyclin synthase and defective vasorelaxation in atherosclerotic bovine coronary arteries. Am J Pathol, 154, 1359–1365. Zou W, Yang Q, Yim AP and He GW (2001) Epoxyeicosatrienoic acids [EET(11,12)] may partially restore endothelium-derived hyperpolarizing factor-mediated function in coronary microarteries. Ann Thorac Surg, 72, 1970–1976. Zucker T-P, Grosser T, Morinelli T et al (1995) Potentiation of PDGFinduced growth responses in coronary artery smooth muscle cells by thromboxane. Agents Actions, 45(suppl), 53–58. Zucker T-P, Bo¨nisch D, Hasse A et al (1998a) Tolerance development to antimitogenic actions of prostacyclin but not of prostaglandin E1 in coronary artery smooth muscle cells. Eur J Pharmacol, 345, 213–220. Zucker T-P, Bo¨nisch D, Muck S et al (1998b) Thrombin-induced mitogenesis in coronary artery smooth muscle cells in potentiated by thromboxane A2 and involves upregulation of thromboxane receptor mRNA. Circulation, 97, 589–595.
Section Seven Digestive System
36 Perspectives and Clinical Significance of Eicosanoids in the Digestive System Chi Hin Cho1, Joshua Ka Shun Ko1,2 and Marcel Wing Leung Koo1 1University
of Hong Kong, and 2Hong Kong Baptist University, Hong Kong, China
Since the discovery that biologically active lipid factors derived from the prostate gland and seminal fluid could contract uterine smooth muscle, research on these factors, particularly eicosanoids, has been vigorously conducted. Formerly, much research in this area focused on the identification of products of the cyclooxygenase (COX) pathway, the respective enzymes and mechanisms of action, and the physiological or potential pathophysiological effects on various body cells and systems. Essentially, a moderate level of mRNA of both COX-1 and COX-2 was observed in the stomach, the small intestine and the liver (O’Neill and Ford-Hutchinson 1993). Lately, an additional pathway has been recognized for arachidonate metabolism, involving the lipoxygenase (LOX) enzyme system and its role in pathophysiology. It was not until recently that specific drug inhibitors for the COX and LOX pathways were derived to counteract the biological actions of products from arachidonate metabolism, leading to the recognition of the clinical implications of these bioactive compounds (Gerritsen 1996). The development of isozyme-selective COX inhibitors is of considerable clinical significance. While conventional non-steroidal antiinflammatory drugs (NSAIDs) are known to provoke substantial gastrointestinal (GI) side-effects, such as upper GI bleeding and peptic ulcerations due to inhibition of the COX-1 enzyme, highly selective COX-2 inhibitors, even at high doses, exhibited an antiinflammatory action devoid of those adverse effects (Famaey 1997; Kalgutkar et al 1998). Furthermore, these agents have potential applications in cancer chemoprevention, although their potential adverse effects on ulcer healing in the gastrointestinal (GI) tract remain controversial. Nevertheless, all these advances are in fact important contributions to the development of eicosanoid research in the GI tract. In addition to their pronounced physiological effects in circulation, motility and secretion, eicosanoids are considered to represent a large and important family of local mediators acting via receptors, which modulate different hormonal, immunological, neuronal and intracellular signals. Recently, a new frontier in eicosanoids research has been established. Some eicosanoids have been found to bind and activate a special class of intracellular target proteins, the peroxisome proliferator-activated receptors (PPAR). This receptor type belongs pharmacologically to the steroid/thyroid/ retinoid receptor superfamily, which, upon interaction with a proper ligand, will bind to specific DNA sequences and thereby regulate the activity of certain genes (Tsai and O’Malley 1994). Some subtypes of the PPAR are predominantly expressed in brown adipose tissue and the liver (Gelman et al 1999), and their actual implication in cellular and metabolic events in the biological system requires more investigation. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
Based on this information, the chapters in Section Seven will summarize the physiological and pathological actions of eicosanoids in the GI tract and the liver. In this chapter, we shall briefly discuss these actions, in particular the effects of eicosanoids in inflammation and carcinogenesis. This will provide some perspectives on clinical implications of prostaglandins (PGs)/leukotrienes (LTs) and also their antagonists in a variety of digestive disorders. LOCALIZATION OF EICOSANOID RECEPTORS IN THE DIGESTIVE SYSTEM Significant effects of eicosanoids are mediated through their interaction with receptors. Eicosanoid receptors have been found along the whole GI tract (Ding et al 1997). Prostaglandin (PG) receptors contain seven membrane-spanning domains that are characteristic of the G protein receptor family (Eberhart and DuBois 1995). A total of eight prostanoid receptors have been found (Narumiya et al 1991). In addition to the eight receptors, each encoded by a different gene, additional receptor diversity is generated through alternative splicing of several of the receptor subtypes (Kiriyama et al 1997). Thromboxane (TX) A2 was the first eicosanoid receptor (TP) cloned (Hirata et al 1991). These receptors are coupled with different signal transduction pathways, thus suggesting that PGs exert their effects via several downstream signalling pathways (Breyer et al 2001). PGE2 is unique among the PGs in that it mediates its effects through several pharmacologically distinct membrane receptors, EP1 to EP4, which are products of distinct genes (Ushikubi et al 1995). Each EP receptor is characterized by a unique pattern of coupling to intracellular signalling systems, reflecting distinct functions. EP1 is coupled to Ca2+ and thus is more closely related to the other constrictor prostanoid receptors, such as TP and FP receptors for PGF2a (Toh et al 1985). EP1 receptor mRNA is highly expressed in gastric muscularis mucosae (Abramovitz et al 1995). EP2 is coupled to cAMP. Activation of the EP2 receptor leads to an increase in cAMP levels, consistent with its ability to relax smooth muscle in vivo, and is closely related to other relaxant prostanoid receptors, such as the IP for prostacyclin and DP for PGD2 (Regan et al 1994). EP3 receptor was originally identified as a constrictor of smooth muscle expressed at relatively high levels in the stomach tissues (Schmid et al 1995). The EP4 receptor signals through increased cAMP. EP4 receptor mRNA is relatively highly expressed in ileum (Bastien et al 1994). In the stomach and small intestine, all four prostanoid receptors, EP1, EP2, EP3, and EP4 are localized in mucous cells and goblet cells, respectively (Northey
408
DIGESTIVE SYSTEM
et al 2000). EP1, EP3 and EP4 were detected in the parietal cells, while the chief cells expressed EP1 and EP3 (Northey et al 2000). The latter two types of EP receptors were also found in the muscularis mucosa, longitudinal muscle and enteric ganglia of the stomach and small intestine. The distribution pattern of EP receptors in the colon was similar to that in the small intestine, except that the former was accompanied by the presence of EP1, EP2 and EP3 receptors in the epithelial cells (Krause and Dubois 2000). TP receptor mRNA is expressed at low levels in most tissues; however, it is highly expressed in the mucus-secreting cells of the GI tract (Northey et al 2000). PROSTANOIDS AND GASTROINTESTINAL SECRETORY FUNCTIONS It has been observed that prostanoids have considerable actions on GI secretory activities. Prostaglandin E2 can induce a significant reduction of gastric acid secretion, while at the same time secretions of intestinal chloride ions, bicarbonate and mucus from the gut are all increased. Of the other eicosanoids, both TXA2 and leukotriene (LT) C4 increase intestinal chloride ion secretion and LTC4 and LTB4 both elevate GI mucus secretion (Davies and Rampton 1997). Additional epithelial fluid excretion is also induced by TX (Collins et al 1995). As a matter of fact, excess PG production is likely to contribute to the condition of diarrhoea. Nevertheless, this function could be part of the intestinal defence mechanism, as COX-2 is highly expressed in the epithelium of the GI tract in response to microbial infections. The resulting increase in PG production in turn stimulates chloride and fluid secretion from the mucosa, flushing the microorganisms away from the intestine. On the other hand, it has been reported that the presence of EP3 prostanoid receptors is essential for maintaining duodenal bicarbonate secretion and mucosal integrity against luminal acid (Takeuchi et al 1999), while PGE2 coupling to the EP4 receptor on colonic epithelial cells evokes mucin exocytosis, an important constituent of epithelial barrier function (Belly and Chadee 1999). Hence, it is apparent that the secretory activities induced by the prostanoids are related to intestinal epithelial integrity, and such actions could involve receptor binding. PROSTANOIDS AND GASTROINTESTINAL VASCULAR INTEGRITY Eicosanoids are important mediators of both physiological and pathophysiological responses in the microcirculation. The vascular endothelial cell in fact plays a crucial role in the biological actions of autocoids, growth factors, cytokines and haemostatic factors. It is believed that prostacyclin (PGI2) is the major PG produced by all endothelial cells. Besides endothelium-derived PGI2, microvessels might synthesize a large quantity of other eicosanoids, including the generation of PGE2 and PGD2. Alternatively, other locally produced eicosanoids, such as the LTs, lipoxins and cytochrome P450 metabolites of arachidonic acid, can have major influences on various microcirculatory functions in the GI tract, mostly in disease states and chronic inflammation. Thus, the potential functions of various eicosanoids in the regulation of GI mucosal blood flow, vascular permeability, leukocyte adhesion and transmigration and also angiogenesis are all aspects of interest that may be due to the primary effects of PGE2 and PGI2. Additionally, PGE2 also modulates vascular reactivity and permeability in concert with other mediators, such as histamine, bradykinin, neuropeptides and interleukins, primarily via vasodilation (Warren et al 1993). Since COX-1 is constitutively expressed in most body tissues, the
PGs formed along the COX-1 pathway are supposed to be involved in the regulation of tissue homeostasis. Apart from that, in the vasculature, COX-1 is important in the control of platelet aggregation through PGH2, a precursor for TXA2 in platelets, and for PGI2 in vascular endothelium. In addition, COX-1-derived PG is responsible for gastric cytoprotection, through the alteration of mucosal blood flow in the gastric circulation, which modifies the ischaemic conditions and mucosal damage or ulceration (Cho et al 1990). Other eicosanoids are also important contributors to increased vascular permeability, acting in concert with other vasoactive mediators. Actually, PG can modulate the release of vasoactive mediators from vascular endothelium. Findings have suggested that it interacts with some of the endothelium-derived vasodilator mediators, which in turn regulates gastric mucosal microcirculation and integrity (Whittle and Esplugues 1990). These mediators concomitantly possess additional properties, including the prevention of gastric mucosal barrier disruption as well as some cellular transport processes (Miller 1983). Capsaicin administration attenuates the beneficial properties of PGE2 and its analogue against acute gastric challenge, suggesting a permissive role for sensory neuropeptides in the mechanisms of action by this prostanoid in animals (Esplugues et al 1992). Invariably, the preservation of the gastric microvasculature, either by direct actions on endothelial cell integrity or by the prevention of the release of other vasoconstrictor or cytotoxic mediators (Boughton-Smith and Whittle 1988), may be an important mechanism in the protective actions of prostanoids in the GI tract. A balance between the release of the endogenous vasodilators and vasoconstrictors in the gastric microvasculature could be involved in the physiological control of the local microcirculation, providing a mechanism for rapid vascular responses to the functional needs of the GI mucosa. The interactions of PG with other diverse local vascular mediators, such as PG and nitric oxide, and neuropeptides like substance P, at the level of the endothelial barrier, could indeed preserve endothelial integrity (Ko and Cho 1999; Ko et al 1997). Thus, the GI tract is well prepared for the challenge by aggressors in attacking the mucosa. PROSTANOIDS AND GASTROINTESTINAL CYTOPROTECTION Gastric cytoprotection is classically defined as the property of certain prostanoids to protect the gastric mucosa from becoming inflamed and necrotic when exposed to noxious agents (Robert et al 1979). The protection occurs under the surface epithelium, where the pretreatment of PG has demonstrated protection of the deeper layers of the stomach against alcohol-induced haemorrhagic necrosis for up to 90% of the thickness of the gastric mucosa (Lacy and Ito 1982). The PG could also preserve the mucosal microvasculature and maintain rapid restitution and cell proliferation. Gastric lesions produced by non-steroidal antiinflammatory drugs (NSAIDs; e.g. aspirin, indomethacin) are linked to gastric mucosal PG deficiency. Correction of this deficiency and prevention of mucosal damage could be achieved by PG replenishment, namely ‘‘substitution cytoprotection’’ (Robert 1976). Products of the arachidonate metabolism include several TX and LT as damaging agents, and prostanoids, including the PGE and PGI types, as the protective mediators. The integrity of the gastric mucosa depends on the balance of these mediators. Under normal conditions, such balance is maintained, and the mucosa is well protected. However, in hostile circumstances, e.g. after intake of a necrotizing agent of very high concentration, the protective mediators may not be able to deal with the increasing levels of the ulcerogenic mediators and the uprising of the aggressive factors. It is apparent that COX-1 is
PERSPECTIVES AND CLINICAL SIGNIFICANCE involved in the regulation of normal gastric functions, whereas COX-2 deals with the crisis situations. However, it had been demonstrated that no spontaneous gastric lesions or ulcers developed in COX-1 knockout mice, where even fewer NSAIDinduced gastric ulcers were observed, despite the drastic drop in gastric PGE2 level (Langenbach et al 1995). Therefore a non-PG dependent pathway is suggested in the formation of NSAID ulcers in the GI tract. GASTROINTESTINAL AND HEPATIC INFLAMMATION The prevalence of gastroduodenal ulceration associated with the use of NSAIDs is in the approximate range 15–30%. NSAIDsinduced gastroduodenal injury leads to about 107 000 hospitalizations and 16 000 deaths each year (Lefkowith 1999). NSAIDs cause significant GI injury, including ulceration complicated by haemorrhage, perforation and intestinal stricture. The pathogenesis of NSAID-induced gastric damage remains controversial but includes reduction of local PG production and direct irritant effects. Altered mucin and bicarbonate secretion, along with reduction of mucosal blood flow, have also been the mechanisms responsible for direct adverse actions on the GI tract (Scheiman 1992). Inhibition of COX-2 not expressed in most healthy tissues would potentially inhibit most of the adverse effects associated with NSAIDs, which target at the constitutively expressed isoform, COX-1. The development of novel agents with high selectivity towards the inhibition of COX-2 showed that this hypothesis was well founded and that high levels of these drugs could be tolerated without serious adverse effects. More recently, however, concern has been expressed that COX-2 inhibition may in fact have a number of potential, previously hidden, pitfalls. These have arisen from the demonstration that COX-2 induction is not exclusively associated with the onset of inflammation, with expression limited to inflammatory sites. In fact, COX-2 is expressed more chronically, and is also seen during the resolution of inflammation and in areas of wound healing. Indeed, it has been reported that COX-2 may play a crucial role in gastric ulcer healing. Cyclooxygenase-2 protein is highly localized in the base of gastric ulcers in rats, and COX-2 mRNA expression might be regulated positively by IL-1b and TNFa and negatively by TGFb1 (Takahashi et al 1998). In humans, both COX-1 and COX-2 are upregulated in gastric ulcers (To et al 2001). Selective COX-2 inhibitors could delay gastric ulcer healing to the same
409
extent as traditional NSAIDs, which are non-selective COX inhibitors that impair the granulation tissue angiogenesis (Hull et al 1999). The application of COX-2-selective inhibitors during this period of time has been shown to be deleterious in that the resolution of inflammation is delayed and gastric ulcer healing is prolonged in some patients. Some ulcers have been shown to progress further to perforation (Colville-Nash and Gilroy 2001). This is not surprising when we consider inflammation as a component of the biological process of wound healing. Proinflammatory cytokines such as IL-1 and TNFa, which regulate the migration and activation of inflammatory cells, also regulate the epithelial component of wound healing (Fedyk et al 2001). It is also possible that delay in wound healing with COX-2 inhibitors may relate to the loss of vasodilation and enhanced vascular permeability induced by PGs. It is likely that drugs that inhibit inflammation may also retard wound healing if given in the long term. In fact, PGs may promote wound healing not only via protective effects but also via stimulatory effects on cell proliferation, angiogenesis or reconstruction of extracellular matrix in the ulcerated stomach (Figure 36.1). The involvement of eicosanoids in the formation of IBD is still a debatable issue in gastroenterology. Conflicting results have been reported. Previous studies have shown that the human colon is able to produce COX and LOX metabolites (Boughton-Smith et al 1983; Dreyling et al 1986). Levels of PGE2, PGF2a, PGD2, TXB2 and 12-hydroxyeicosatetraenoic acid (12-HETE), 15-HETE and LTB4 are increased in inflamed tissues compared with normal mucosa (Fretland et al 1990; Lauitsen et al 1988; Rampton and Collins 1993). These levels decline in IBD patients treated with either corticosteroids or sulphasalazine (Yang 1996). Also, some drugs currently being used for the treatment of IBD have been shown to have beneficial effects in experimental IBD, accompanied by the reduction of PGE2, PGI2 and TXA2 synthesis (Hiller et al 1991). Indeed, PG levels in patients with inactive IBD are not significantly different from those of normal control subjects (Rampton et al 1980; Sharon et al 1978). Although PG levels are increased in mucosal specimens of patients with active IBD, NSIAD treatment provides no clinical improvement in Crohn’s disease or ulcerative colitis (Gould et al 1981; Rampton et al 1980). In fact, evidence shows that NSAIDs exacerbate inflammation in IBD patients (Kaufmann and Taubin 1987). It is still uncertain whether PGs are involved in the inflammatory responses in the colon. It had been found that COX-2 protein and mRNA expression were induced in IBD patients, while COX-1 expression was not changed under a colitis condition (Hendel and Nielsen 1997;
Figure 36.1 Involvement of cyclooxygenase (COX) and lipoxygenase (LOX) in inflammation and ulcer formation and its healing in the gastrointestinal tract. PGs, postaglandins; LTs, leukotrienes
410
DIGESTIVE SYSTEM
Singer et al 1998). Moreover, there is a clear relationship between endoscopic activity of the colitis and the relative presence of mRNA for COX-2, indicating that COX-2 was involved in the acute inflammatory response of chronic IBD (Hendel and Nielsen 1997). However, it is still controversial whether COX-2 plays a crucial role in the pathogenesis of IBD. An in vitro study clearly demonstrated that pretreatment with indomethacin, which is a non-specific COX inhibitor, significantly prevented acute injury of human colonic cells (Stratton et al 1996). Other studies showed that pretreatment with COX inhibitors not only protected against the severity of IBD, but also attenuated the potentiating effect of cigarette-smoke exposure on colonic damage (Rachmileiwitz et al 1989; Guo et al 2001). This observation indicates that an acute injury of colonic cells is likely to be mediated by the COX enzyme. On the contrary, it was found that treatment with indomethacin or some other selective COX-2 inhibitors after the induction of experimental IBD resulted in exacerbation of inflammationassociated colonic injury in rats (Wallace et al 1992; Reuter et al 1996). A protective effect exerted by COX-2 expression in intestinal inflammation is suggested. In line with these findings, previous studies have demonstrated that selective COX-2 inhibitors prevent epithelial wound healing in the rat model of gastric ulcers and intestinal epithelial monolayers (Reuter et al 1996; Lesch et al 1999). These controversial findings suggested that COX-2 might possibly play different roles in the pathogenesis and healing process of IBD. Thus, inhibition of COX-2 activity may protect against inflammation but may lead to an impairment of wound healing in the inflamed intestinal tissue (Babak and Ma 2000). The role of leukotrienes in the inflammatory responses throughout the GI tract has also been studied. However, their physiological function has been less defined. LTs have been largely considered to be potent mediators of inflammation and allergic reactions in the GI tract. These have been well studied in the intestine, especially in IBD. The involvement of 5-LOX and its metabolites has been extensively explored (see Chapter 45, this volume, for details) but the therapeutic application is not yet confirmed to be clinically relevant. It seems that inhibition of LT biosynthesis is not effective as a single therapeutic modality in active IBD (Robert et al 1997). Lipoxygenase inhibitors show promise in animal models of hepatitis (Kennedy and Keeling 1987). The cytoprotective effects of different types of PGs and their synthetic analogues on liver disease has been demonstrated and reviewed (Quiroga and Prieto 1993). Prostaglandin products of the COX-2 pathway demonstrated a hepato-protective function against drug-induced liver toxicity (Reilly et al 2001). Also, the effect of cytoprotective drugs to prevent hepatic ischaemia and reperfusion injury was also found to involve endogenous PG stimulation (Aslan et al 2001). Liver hepatocytes synthesize small amount of PGE2, PGF2a, PGI2, TXA2 and LTs (Tran-Thi et al 1987), while Kupffer cells produce mainly PGD2, PGE2, TxA2, LTB4 and LTC4 (Brouwer et al 1988). All the PGs, with the exception of PGF2a, are protective (Sinclair et al 1990), while LTs such as LTB4 potentiate liver damage (Tiegs and Wendel 1988). The protective effect of PGs, such as PGE1 and PGE2, seems to be dependent on the stimulation of cAMP synthesis, which stabilizes the parenchymal cell membranes (Masaki et al 1992). PGI2 produced by the endothelial cells also exerts protection by preventing vasoconstriction, platelet aggregation and leukocyte adherence (Needleman et al 1986). The sustained release of PGE2 also exerts a negative feedback on further release of cytokines during acute inflammation (Quiroga and Prieto 1993). Chapter 37 (this volume) by Rudnick and Muglia elaborates the effects of eicosanoids on liver regeneration in both experimental models and human liver diseases. Indeed, the practical use of PGs in the treatment of hepatic disorders needs further basic and clinical investigations.
CANCERS IN THE DIGESTIVE SYSTEM Colorectal Neoplasm The involvement of PGs and other eicosanoids in the development of human cancer has been known for over two decades. COX plays an important role in cancer growth via angiogenesis, which offers a new strategy against cancer using COX inhibitor (Sawaoka et al 1999). Three independent lines of research indicate a possible beneficial association between NSAID use and colorectal cancer. First, epidemiological studies report a 40–50% reduction in the risk of colorectal cancer among individuals taking NSAIDs compared with those not taking these agents (Thun 1996). Second, familial adenomatous polyposis patients who take NSAIDs have a significant reduction in adenoma size and number (Giardiello et al 1993). Third, animal studies of colorectal carcinogenesis have demonstrated that NSAIDs are chemoprotective, causing a reduction in the frequency and number of premalignant and malignant lesions (Rao et al 1995; Reddy et al 1993). However, there was no association of changes in colon cancer risk with paracetamol and steroidal antiinflammatory drugs. It was noted that there was significant elevation of COX-2 mRNA but not COX-1 in colorectal adenocarcinomas when compared to normal colonic mucosa in the same patients (Eberhart et al 1994). COX-2 overexpression in intestinal epithelial cells leads to specific phenotypic changes, such as increased adhesion and inhibition of apoptosis, indicating that this enzyme may alter the tumorigenic potential of epithelial cells (DuBois et al 1996). Eicosanoids also mediate other cellular responses, including modulation of cellular adhesion, differentiation and mitogenesis. Ultimately, the elevated levels of PGs in human colorectal tumours as compared with other normal colonic mucosa (Rigas et al 1993) might accelerate the growth and invasion of cancer (Narisawa et al 1990). All these findings offer hope for the future development of improved chemopreventive agents. It has also been suggested that arachidonic acid metabolites derived from LOX play an important role in growth-related signal transduction, implying that intervention through this pathway should also be useful for cancer prevention. The LOX pathway also functions as a critical regulator of cell survival and apoptosis (Tang et al 1996). Specific 12-LOX expressed in a wide diversity of tumour cell lines and the metabolite 12(S)-HETE is a key modulatory molecule in metastasis, which provides the rationale for targeting these molecules in anticancer and antimetastasis therapeutic protocols (Honn et al 1994). In contrast, 15-LOX expression is reduced in human colorectal cancers. NSAIDs can increase 15-LOX enzymatic activity, with subsequent growth inhibition and apoptosis in colon cancer cells (Shureiqi et al 2000). It is likely that drug-mediated modulation through both COX and LOX could be an effective and therapeutic approach for cancer chemoprevention (Cuendet and Pezzuto 2000). Figure 36.2 summarizes the carcinogenic pathways of COX and LOX in the GI tract. Other Types of Cancer in the Digestive System The evidence that aspirin or NSAIDs prevent other human cancers in the digestive system is more limited than that pertaining to the colon and rectum. However, in an American Cancer Society study, aspirin use was inversely associated with fatal cancers of the oesophagus and stomach as well as the colon and rectum, but not generally with fatal cancers outside the digestive tract (Thun 1996). It has been suggested that NSAIDs would inhibit the development of tumours but did not reject or kill the established tumour cells.
PERSPECTIVES AND CLINICAL SIGNIFICANCE
411
Figure 36.2 Role of eicosanoids in gastrointestinal cancer development. COX-2, cyclooxygenase-2; LOX, lipoxygenase; PGEs, prostaglandin series; HETE, hydroxyeicosatetraenoic acid; LTB4, leukotriene B4; 13-S-HODE, 13-S-hydroxyoctadecadienoic acid. ), Enhancement; ), prohibition.
The involvement of COX in gastric tumorigenesis is less evidenced than in the colon. Recently, the stimulatory action on cell growth and the inhibition of apoptosis induced by 5-LOX and 12-LOX, respectively, have been indicated in human gastric cancer cells. Inhibitors of these LOX enzymes, but not the NSAIDs and antiinflammatory steroids, prevented such effects (Shimakura and Boland 1992; Wong et al 2001). Similarly, the contribution of COX-2 in human liver cancer is still unclear. However, patients with hepatocellular carcinoma exhibited more positive staining and more intense expression of COX-2 in cancer tissue (Shiota et al 1999). COX-2 expression in non-tumour tissue may play a positive role in the relapse of hepatocellular carcinomas (HCC) after surgery (Kondo et al 1999) The expression pattern of COX-2 protein has been well correlated with the differentiation grade, suggesting that abnormal COX-2 expression plays an important role in hepatocarcinogenesis. It was demonstrated that a high expression of COX-2 in well-differentiated HCC and a low expression in advanced HCC had been indicated, which was in contrast to its continuous expression during colorectal carcinogenesis. These findings suggest that COX-2 might play a role in the early stages of hepatocarcinogenesis but not in the advanced stages (Koga et al 1999). Indeed, inhibition of COX-2 can induce growth suppression of hepatoma cell lines, which is likely to be mediated via induction of apoptosis (Bae et al 2001). Recent studies have shown overexpression of COX-2 and 5LOX in exocrine pancreatic carcinomas, suggesting a potential role of the arachidonic acid cascade in the regulation of this type of cancer (Weddle et al 2001). It was shown that 5-LOX and 12LOX were unregulated in human pancreatic cancer cells. Metabolites of these enzymes could directly stimulate cancer cell proliferation, while blockade of the enzymes produced an opposite effect (Ding et al 1999; Ding and Adrian 2001). It is envisaged that perturbations of LOX activity by LOX inhibitors may be valuable for the treatment of human pancreatic cancer (Ding et al 1999). In a human pancreatic carcinoma cell line study, NSAIDs combined with a most effective chemotherapeutic drug, gemcitabine, inhibited cell growth to a greater degree than either compound alone. The effect results primarily from the inhibition of cell cycle progression rather than from direct induction of apoptotic cell death. The drug combination is promising as a treatment for pancreatic cancer. There are several lines of evidence, including experimental and epidemiological studies, that the use of NSAIDs may reduce the risk of oesophageal cancer. Immunohistochemical detection of
COX-2 revealed strong positive staining in the well-differentiated regions of oesophageal tumours. Poorly differentiated areas of the oesophageal tumours were negative (Ratnasinghe et al 1999a,b). Induction of apoptosis through the cytochrome c–caspase pathway by NSAIDs may be a mechanism by which NSAIDs can intervene in oesophageal carcinogenesis. This may be an indication of the potential of NSAIDs as chemopreventive agents in oesophageal cancer (Li et al 2000, 2001). Apart from COX, in human oesophageal cancers using paired normal and tumour surgical samples, 15-LOX was downregulated in human oesophageal carcinomas and NSAIDs induced 15-LOX expression during apoptosis in oesophageal cancer cells. This finding supports the concept that the loss of the proapoptotic role of 15-LOX in oesophageal cancers is not limited to human colorectal cancers (Shureiqi et al 2001). Modification of the level of 15-LOX by agents in the oesophagus may be an effective approach in treating oesophageal cancer. CONCLUSIONS The physiological functions and the pharmacological actions of PGs have been well defined in the upper GI tract. In particular, the role of PGs in mucosal defensive mechanisms is believed to have significant implications in the prevention of NSAID-induced mucosal injury. However, the observable side-effects on intestinal secretion and motility have been a problem for their general clinical applications. The involvement of COX-2 in ulcer formation and healing is still a debatable issue. Caution needs to be taken when COX-2 inhibitors are used in peptic ulcer patients. The physiological functions of PGs in the lower GI tract are different from those of the upper. The general stimulatory action on secretion and motility in the small and large intestines is not observed in the stomach. The inflammatory responses of LTs in the colon with different inflammatory disorders are widely studied. However, the complexity of the inflammatory reactions in the colon has limited the development of drugs derived from LTs as a sole agent for the treatment of IBD. It is an established fact that NSAIDs can reduce the incidence of colorectal cancer in humans. There is great potential for such drugs to be used in other cancer disorders in the digestive system. The involvement of 12LOX and 5-LOX in the carcinogenesis of GI cancer has been extensively studied. The metabolites of these enzymes, together with COX-2, seem to be the targets for drug development in chemoprevention. Combination of drug treatment with the
412
DIGESTIVE SYSTEM
inhibitors of both COX and LOX could be a new clinical perspective to cancer therapy.
REFERENCES Abramovitz M, Adam M, Boie Y et al (1995) Human prostanoid receptors: cloning and characterization. Adv Prostagland Thrombox Leukotriene Res, 23, 499–504. Aslan A, Karagu¨zel G, Elik M et al (2001) Pentoxifylline contributes to the hepatic cytoprotective process in rats undergoing hepatic ischaemia and reperfusion injury. Eur Surg Res, 33, 285–290. Babak M and Ma TY (2000) Eicosanoids and the small intestine. Prostagland Lipid Mediat, 61, 125–143. Bae SH, Jung ES, Park BS et al (2001) Expression of cyclooxygenase-2 (COX-2) in hepatocellular carcinoma and growth inhibition of hepatoma cell lines by a COX-2 inhibitor, NS-398. Clin Cancer Res, 7, 1410–1418. Bastien L, Sawyer N, Grygorczyk R et al (1994) Cloning, functional expression, and characterization of the human prostaglandin E2 receptor EP2 subtype. J Biol Chem, 269, 11873–11877. Belley A and Chadee K (1999) Prostaglandin E2 stimulates rat and human colonic mucin exocytosis via the EP4 receptor. Gastroenterology, 117, 1352–1362. Boughton-Smith NK and Whittle BJR (1988) Inhibition by 16,16dimethyl PGE2 of ethanol-induced gastric mucosal damage and leukotriene B4 and C4 formation. Prostaglandins, 35, 945–947. Boughton-Smith NK, Hawkey CJ and Whittle BJR (1983) Biosynthesis of lipoxygenase and cyclooxgenase products from (14C)-arachidonic acid by human colonic mucosa. Gut, 24, 1176–1182. Breyer RM, Bagdassarian CK, Myers SA and Breyer MD (2001) Prostanoid receptors: subtypes and signaling. Ann Rev Pharmacol Toxicol, 41, 661–690. Brouwer A, Barelds RJ, De Leeuw AM et al (1988) Isolation and culture of Kupffer cells from human liver. Ultrastructure, endocytosis and prostaglandin synthesis. J Hepatol, 6, 36–49. Cho CH, Chen BW, Hui WM et al (1990) Endogenous prostaglandins: its role in gastric mucosal blood flow and ethanol ulceration in rats. Prostaglandins, 40, 397–403. Collins RJ, Czerniak PM, Gorry SA and Trzasko JM (1995) Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature, 378, 406–409. Colville-Nash PR and Gilroy DW (2001) Potential adverse effects of cyclooxygenase-2 inhibition: evidence from animal models of inflammation. Biodrugs, 15, 1–9. Cuendet M and Pezzuto JM (2000) The role of cyclooxygenase and lipoxygenase in cancer chemoprevention. Drug Metab Drug Interact, 17, 109–157. Davies GR and Rampton DS (1997) Eicosanoids: role in gastrointestinal inflammation and cancer. Eur J Gastroenterol Hepatol, 9, 1033–1044. Ding M, Kinoshita Y, Kishi K et al (1997) Distribution of prostaglandin E receptors in the rat gastrointestinal tract. Prostaglandins, 53, 199– 216. Ding XZ and Adrian TE (2001) Role of lipoxygenase pathways in the regulation of pancreatic cancer cell proliferation and survival. Inflammopharmacology, 9, 157–164. Ding XZ, Kusznski CA, EL-Metwally TH and Adrian TE (1999) Lipoxygenase inhibition induced apoptosis, morphological changes and carbonic anhydrase expression inhuman pancreatic cancer cells. Biochem Biophys Res Commun, 266, 392–399. Dreyling KW, Hoppe U, Peska BA et al (1986) Leukotriene synthesis by human gastrointestinal tissues. Biochim Biophys Acta, 878, 184– 193. DuBois RN, Radhika A, Reddy BS and Entingh AJ (1996) Increased cyclooxygenase-2 levels in carcinogen-induced rat colonic tumors. Gastroenterology, 110, 1259–1262. Eberhart CE, Coffey RJ, Radhika A et al (1994) Upregulation of cyclooxygenase-2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology, 107, 1183–1188. Eberhart CE and Dubois RN (1995) Eicosanoids and the gastrointestinal tract. Gastroenterology, 109, 285–301.
Esplugues JV, Whittle BJR and Moncada S (1992) Modulation by opioids and by afferent sensory neurons of prostanoid protection of the rat gastric mucosa. Br J Pharmacol, 106, 846–852. Famaey JP (1997) In vitro and in vivo pharmacological evidence of selective cyclooxygenase-2 inhibition by nimesulide: an overview. Inflamm Res, 46, 437–446. Fedyk ER, Jones D, Critchley HO et al (2001) Expression of stromalderived factor-1 is decreased by IL-1 and TNF and dermal wound healing. J Immunol, 166, 5749–5754. Fretland DJ, Dijuric SW and Gaginella TS (1990) Eicosanoids and inflammatory bowel disease: regulation and prospects for therapy. Prostagland Leukotrienes Essent Fatty Acids, 41, 215–233. Gelman L, Fruchart JC and Auwerx J (1999) An update on the mechanisms of action of the peroxisome proliferator-activated receptors (PPARs) and their roles in inflammation and cancer. Cell Mol Life Sci, 55, 932–943. Gerritsen ME (1996) Physiological and pathophysiological roles of eicosanoids in the microcirculation. Cardiovasc Res, 32, 720–732. Giardello FM, Hamilton SR and Krush AJ (1993) Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med, 328, 1313–1316. Gould SR, Brash AR, Conolly ME and Lennard-Jones JE (1981) Studies of prostaglandins and sulphasalazine in ulcerative colitis. Prostagland Med, 6, 165–182. Guo X, Liu ESL, Ko JKS et al (2001) Protective role of cyclooxygenase inhibitors in the adverse action of passive cigarette smoking on the initiation of experimental colitis in rats. Eur J Pharmacol, 411, 193– 203. Hendel J and Nielsen OH (1997) Expression of cyclooxygenase-2 mRNA in active inflammatory bowel disease. Am J Gastroenterol, 92, 1170– 1173. Hiller K, Jewel R, Dorrel L and Smith CL (1991) Incorporation of fatty acids from fish oil and olive oil into colonic mucosal lipids and effects upon eicosanoid synthesis in inflammatory bowel disease. Gut, 32, 1151–1155. Hirata M, Hayashi Y, Ushikubi F et al (1991) Cloning and expression of cDNA for a human thromboxane A2 receptor. Nature, 349, 617–620. Honn KV, Tang DG, Gao X et al (1994) 12-Lipoxygenase and 12(S)HETE: role in cancer metastasis. Cancer Metast Rev, 13, 365–396. Hull MA, Thomson JL and Hawkey CJ (1999) Expression of cyclooxygenase 1 and 2 by human gastric endothelial cells. Gut, 45, 529–536. Kalgutkar AS, Crews BC, Rowlinson SW et al (1998) Aspirin-like molecules that covalently inactivate cyclooxygenase-2. Science, 280, 1268–1270. Kaufman HJ and Taubin HL (1987) Non-steroidal antiinflammatory drugs activate quiescent inflammatory bowel disease. Ann Intern Med, 107, 513–516. Kennedy NP and Keeling PW (1987) Review: lipoxygenase inhibitors and the gut. Aliment Pharmacol Therapeut, 1, 263–272. Kiriyama M, Ushikubi F, Kobayashi T et al (1997) Ligand binding specificities of the eight types and subtypes of the mouse prostanoid receptors expressed in Chinese hamster ovary cells. Br J Pharmacol, 122, 217–224. Ko JKS and Cho CH (1999). Co-regulation of mucosal nitric oxide and prostaglandins in gastric defensive mechanisms and their roles in adaptive cytoprotection. Inflamm Res, 48, 471–478. Ko JKS, Tang F and Cho CH (1997) Co-regulation of mucosal prostanoids and substance P by indomethacin in rat stomachs. Life Sci, 60, PL277–P281. Koga H, Sakisaka S, Ohishi M et al (1999) Expression of cyclooxygenase-2 in human hepatocelluar carcinoma: relevance to tumor dedifferentiation. Hepatology, 29, 6880–6896. Kondo M, Yamamoto H, Nagano H et al (1999) Increased expression of COX-2 in non-tumor liver tissue is associated with shorter disease-free survival in patients with hepatocellular carcinoma. Clin Cancer Res, 5, 4005–4012. Krause W and DuBois RN (2000) Eicosanoids and the large intestine. Prostagland Lipid Mediat, 61, 145–161. Lacy ER and Ito S (1982) Microscopic analysis of ethanol damage to rat gastric mucosa after treatment with a prostaglandin. Gastroenterology, 83, 619–625. Langenbach R, Morham SG, Tiano H et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced
PERSPECTIVES AND CLINICAL SIGNIFICANCE inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Lauitsen K, Laursen LS, Bukhave K and Rask-Madsen J (1988) In vivo profiles of eicosanoids in ulcerative colitis, Crohn’s colitis and Clostridium difficile colitis. Gastroenterology, 95, 11–17. Lefkowith JB (1999) Cyclooxygenase-2 specificity and its clinical implications. Am J Med, 106, 43S–50S. Lesch CA, Krause ER, Sanchez B et al (1999) Lack of beneficial effect of COX-2 inhibitors in an experimental model of colitis. Methods, findings. Exp Clin Pharmacol, 21, 99–104. Li M, Lotan R, Levin B et al (2000) Aspirin induction of apoptosis in esophageal cancer: a potential for chemoprevention. Cancer Epidemiol Biomarkers Prevent, 9, 545–549. Li M, Wu X and Xu XC (2001) Induction of apoptosis by cyclooxygenase2 inhibitor NS398 through a cytochrome c-dependent pathway in esophageal cancer cells. Int J Cancer, 93, 218–223. Masaki N, Ohta Y, Shirataki H et al (1992) Hepatocyte membrane stabilization by prostaglandins E1 and E2: favorable effects on liver injury. Gastroenterology, 102, 572–576. Miller TA (1983) Protective effects of prostaglandins against gastric mucosal damage: current knowledge and proposed mechanisms. Am J Physiol, 245, G601–G623. Narisawa T, Kusaka H, Yamazaki Y et al (1990) Relationship between blood plasma prostaglandin E2 and liver and lung metastases in colorectal cancer. Dis Colon Rectum, 33, 840–845. Narumiya S, Sugimoto Y and Ushikubi F (1991) Prostanoid receptors: structures, properties, and functions. Physiol Rev, 79, 1193–1226. Needleman P, Turk J, Jakschik BA et al (1986) Arachidonic acid metabolism. Ann Rev Biochem, 55, 69–102. Northey A, Denis D, Cirino M et al (2000) Cellular distribution of prostanoid EP receptors mRNA in the rat gastrointestinal tract. Prostagland Lipid Mediat, 62, 145–156. O’Neill GPO and Ford-Hutchinson AW (1993) Expression of mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human tissues. FEBS Lett, 330, 156–160. Quiroga J and Prieto J (1993) Liver cytoprotection by prostaglandins. Pharmacol Therapeut, 58, 67–91. Rachmileiwitz D, Simon PL, Schwartz LW et al (1989) Inflammatory mediators of experimental colitis in rats. Gastroenterology, 97, 326–337. Rampton DS and Collins CE (1993) Review article: thromboxanes in inflammatory bowel disease—pathogenic and therapeutic implications. Aliment Pharmacol Therapeut, 7, 357–367. Rampton DS, Sladen GE and Youlten LJ (1980) Rectal mucosal prostaglandin E2 release and its relation to disease activity, electrical potential difference, and treatment in ulcerative colitis. Gut, 21, 591– 596. Rao CV, Rivenson A, Simi B et al (1995) Chemoprevention of colon carcinogenesis by sulindac, a non-steroidal antiinflammatory agent. Cancer Res, 55, 1464–1472. Ratnasinghe D, Limburg P and Tangrea JA (1999a) Cyclooxygenase-2, non-steroidal antiinflammatory drugs and cancer of the esophagus. Gastroenterol Int, 12, 178–181. Ratnasinghe D, Tangrea J, Roth MJ et al (1999b) Expression of cyclooxygenase-2 in human squamous cell carcinomas of the esophagus: an immunohistochemical survey. Anticancer Res, 19, 171– 174. Reddy BS, Rao CV, Rivenson A and Kelloff G (1993) Inhibitory effect of aspirin on azoxymethane-induced colon carcinogenesis in F344 rats. Carcinogenesis, 14, 1493–1497. Regan JW, Bailey TJ, Pepperl DJ et al (1994) Cloning of a novel human prostanglandin receptor with characteristics of the pharmacologically defined EP2 subtype. Mol Pharmacol, 46, 213–220. Reilly TP, Brady JN, Marchick MR et al (2001) A protective role for cyclooxygenase-2 in drug-induced liver injury in mice. Chem Res Toxicol, 14, 1620–1628. Reuter BK, Asfaha, S, Buret A et al (1996) Exacerbation of inflammationassociated colonic injury in rat through inhibition of cyclooxygenase-2. J Clin Invest, 98, 2076–2085. Rigas B, Goldman IS and Levine L (1993) Altered eicosanoind levels in human colon cancer. J Lab Clin Med, 122, 518–523. Robert A (1976) Antisecretory, antiulcer, cytoprotective and diarrheogenic properties of prostaglandins. Adv Prostagland Thrombox Res, 1, 507–520.
413
Robert A, Nezamis JE, Lancaster C and Hanchar AJ (1979) Cytoprotection by prostaglandins in rats. Prevention of gastric necrosis produced by alcohol, HCl, NaOH, hypertonic NaCl, and thermal injury. Gastroenterology, 77, 433–443. Roberts WG, Simon TJ, Berlin RG et al (1997) Leukotrienes in ulcerative colitis: results of a multicenter trial of a leukotriene biosynthesis inhibitor, MK-591. Gastroenterology, 112, 725–732. Sawaoka H, Tsuiji S, Tsuiji M et al (1999). Cyclooxygenase inhibitors suppress angiogenesis and reduce tumor growth in vivo. Lab Invest, 79, 1469–1477. Scheiman JM (1992) Pathogenesis of gastroduodenal injury due to nonsteroidal antiinflammatory drugs: implications for prevention and therapy. Sem Arthrit Rheumat, 21, 201–210. Schmid A, Thierauch K, Schleuning W and Dinter H (1995) Splice variants of the human EP3 receptor for prostaglandin E2. Eur J Biochem, 15, 23–30. Sharon P, Ligumsky M, Rachmilewitz D and Zor U (1978) Role of prostaglandins in ulcerative colitis. Enhanced production during active disease and inhibition by sulfasalazine. Gastroenterol, 75, 638–640. Shimakura S and Boland CR (1992) Eicosanoind production by the human gastric cancer cell line AGS and its relation to cell growth. Cancer Res, 52, 1744–1749. Shiota G, Okubo M, Noumi T et al (1999) Cyclooxygenase-2 expression in hepatocellular carcinoma. Hepatogastroenterology, 46, 407–412. Shureiqi I, Chen D, Lee JJ et al (2000) 15-LOX-1: a novel molecular target of non-steroidal antiinflammatory drug-induced apoptosis in colorectal cancer cells. J Natl Cancer Inst, 92, 1136–1142. Shureiqi I, Xu X, Chen D et al (2001) Non-steroidal antiinflammatory drugs induce apoptosis in esophageal cancer cells by restoring 15lipoxygenase-1 expression. Cancer Res, 61, 4879–4884. Sinclair S, Abecassis M, Wong PY et al (1990) Mechanism of protective effect of prostaglandin E in murine hepatitis virus strain 3 infection: effects on macrophage production of tumor necrosis factor, procoagulant activity and leukotriene B4. Adv Exp Med Biol, 276, 533–542. Singer II, Kawkam DW, Schloemann S et al (1998) Cyclooxygenase-2 is induced in colonic epithelial cells in inflammatory bowel disease. Gastroenterology, 115, 297–306. Stratton MD, Sexe R, Peterson B et al (1996) The effects of trinitrobenzene sulfonic acid (TNB) on coloncyte arachindonic acid metabolism. J Surg Res, 60, 375–378. Takahashi S, Shigeta JI, Inoue H et al (1998) Localization of cyclooxygenase-2 and regulation of its mRNA expression in gastric ulcers in rats. Am J Physiol, 275, G1137–G1145. Takeuchi K, Ukawa H, Kato S et al (1999) Impaired duodenal bicarbonate secretion and mucosal integrity in mice lacking prostaglandin E-receptor subtype EP3. Gastroenterology, 117, 1128–1135. Tang DG, Chen YQ and Honn KV (1996) Arachidonate lipoxygenase as essential regulators of cell survival and apoptosis. Proc Natl Acad Sci USA, 93, 5241–5246. Thun MJ (1996) NSAID use and decreased risk of gastrointestinal cancers. Gastroenterol Clin N Am, 25, 333–348. Tiegs G and Wendel A (1988) Leukotriene-mediated liver injury. Biochem Pharmacol, 37, 2569–2573. To KF, Chan FKL, Cheng ASL et al (2001) Upregulation of cyclooxygenase-1 and -2 in human gastric ulcer. Aliment Pharmacol Therapeut, 15, 25–34. Toh H, Ichikawa A and Narumiya S (1995) Molecular evolution of receptors for eicosanoids. FEBS Lett, 361, 17–21. Tran-Thi TA, Gfyuko K, Henninger H et al (1987). Studies on synthesis and degradation of ecosanoids by rat hepatocytes in primary culture. J Hepatol, 5, 322–331. Tsai MJ and O’Malley BW (1994) Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Ann Rev Biochem, 63, 451–486. Ushikubi F, Hirata M and Narumiya S (1995) Molecular biology of prostanoid receptors; an overview. J Lipid Mediat Cell Signal, 12, 343–359. Wallace JL, Keenan CM, Gale D and Shoupe TS (1992) Exacerbation of experimental colitis by non-steroidal antiinflammatory drugs in not related to elevated leukotriene B4 synthesis. Gastroenterology, 102, 18–27. Warren JB, Wilson AJ, Loi RK and Coughlan ML (1993) Opposing roles of cyclic AMP in the vascular control of edema formation. FASAB J, 7, 1394–1400.
414
DIGESTIVE SYSTEM
Weddle DL, Tithoff P, Williams M and Schuller HM (2001) b-Adrenergic growth regulation of human cancer cell lines derived from pancreatic ductal carcinomas. Carcinogenesis, 22, 473–479. Whittle BJR and Esplugues JV (1990) Vascular interactions between endogenous prostanoids and vasoactive peptide mediators in gastric damage. Adv Prostagland Thrombox Leukotriene Res, 21, 761–765.
Wong BCY, Wang WP, Cho CH et al (2001) 12-Lipoyxgenase inhibition induced apoptosis in human gastric cancer cells. Carcinogenesis, 22, 1349–1354. Yang VW (1996) Eicosanoids and inflammatory bowel disease. Gastroenterol Clin N Am, 25, 317–332.
37 Eicosanoids and Liver Regeneration David A. Rudnick1 and Louis J. Muglia1 1Washington
University School of Medicine and 2St. Louis Children’s Hospital, St. Louis, MO, USA
The eicosanoids, which include prostaglandins, thromboxanes, leukotrienes and epoxyeicosatrienoic acids, are the bioactive endproducts of arachidonic acid metabolism (Decker 1985; Needleman et al 1986). Each of these classes of lipid mediators is synthesized in, act upon and degraded by the liver. They also mediate a number of physiological and pathophysiological actions on the liver. Their hepatic synthesis and metabolism has recently been extensively reviewed by Tolman (2000) as well as by Huber and Keppler (1990). The report by Tolman also summarizes the evidence for the hepatoprotective effects of prostaglandins and the hepatotoxic effects of leukotrienes. Eicosanoids, particularly prostaglandins, are known to be important mediators of normal and abnormal growth modulation in many tissues, including the liver. The purpose of this review is to summarize the evidence for the role of prostaglandins and other eicosanoids in hepatocellular growth regulation and in the regenerative response of the liver to injury in both experimental model systems and human liver disease.
EICOSANOID SYNTHESIS Eicosanoid production begins with the release of arachidonic acid from membrane phospholipids, usually by the action of phospholipase A2. This is followed by conversion to either prostaglandins and thromboxanes, leukotrienes, or epoxyeicosatrienoic acids by one of three different enzymatic pathways. The committed steps in these pathways are mediated by prostaglandin H synthase or cyclooxygenase (COX), lipoxygenase or cytochrome P450 epoxygenase, respectively (Huber and Keppler 1990; Tolman 2000). Each of these principal classes of eicosanoids has been detected in the liver. Prostaglandins and thromboxanes are produced by the COX-dependent conversion of arachidonic acid to prostaglandin H2 (PGH2), which is, in turn, converted to the various biologically active end-products by the actions of prostaglandin- and thromboxane-specific synthases. These endproducts, which include prostacyclin (PGI2), PGD2, PGE2, PGF2a, and thromboxane A2, then bind to specific receptors to mediate their effects on cells. COX, which catalyses the committed step in prostaglandin and thromboxane biosynthesis, exists as two distinct isozymes (Smith et al 1996). COX-1, the constitutive isoform, is present in most tissues, mediates the synthesis of prostaglandins required for normal or ‘‘housekeeping’’ functions, including maintenance of normal gastrointestinal, platelet and renal function, and appears to be inducible only in specific circumstances (e.g. gravid uterus; Gross et al 2000). In contrast, COX-2 is undetectable in most tissues, but is highly inducible, particularly by inflammatory mediators. Leukotrienes are The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
produced by the lipoxygenase-dependent conversion of arachidonic acid to leukotriene A4 (LTA4), which is converted, in turn, to the end-products LTB4, LTC4, LTD4 and LTE4 (Huber and Keppler 1990; Tolman 2000). A fourth product of arachidonic acid metabolism derives from the action of the cytochrome P450dependent epoxygenase on arachidonic acid, resulting in the production of the epoxyeicosatrienoic acids. Their activity as potent vasodilators and potential role in the pathogenesis of portal hypertension was discussed in detail by Tolman (2000) and they will not be discussed further in this review.
HEPATIC EICOSANOID SYNTHESIS The liver is composed of a number of cell types, including hepatocytes, biliary epithelial cells, endothelial cells, Kupffer cells (perisinusoidal reticuloendoethelial cells) and Ito cells (perisinusoidal lipocytes). Although each of these cell types is capable of synthesizing eicosanoids, the endothelial cells and Kupffer cells together appear to account for the majority of their synthesis (Kuiper et al 1988; Tolman 2000). Kupffer cell production of prostaglandins may be cytoprotective to hepatocytes and endothelial cells in the setting of endotoxin-stimulated inflammatory cascades (Tolman 2000). Hepatocellular production of prostaglandins and thromboxanes is more limited and is thought to mediate important cell-to-cell signalling events. Hepatocytes are also an important site of eicosanoid degradation (Tolman 2000). Endothelial cell production of vasodilatory prostaglandins is likely to protect the microcirculation of the liver during injury (Tolman 2000).
EICOSANOIDS AND LIVER REGENERATION The liver responds to multiple forms of injury with a highly regulated regenerative response (Diehl and Rai 1999; Michalopoulos and DeFrances 1997). Animal model systems have been important in the detailed analysis of the molecular signalling pathways that modulate this response. The best characterized of these models involves the analysis of hepatocellular proliferation after partial hepatectomy in rodents (Higgins and Anderson 1931). The important role of humoral factors in the modulation of liver regeneration was first demonstrated in parabiotic studies in which the circulatory systems of two rats were connected in series, partial hepatectomy was performed on one animal, and increased hepatocellular proliferation was detected in the other, non-operated animal (Fisher et al 1971; Moolten and Bucher 1967). A number of
416
DIGESTIVE SYSTEM
candidate humoral factors have subsequently been implicated in the modulation of this response, including cytokines, e.g. TNFa and IL-6, and growth factors, e.g. hepatocyte growth factor (HGF), epidermal growth factor (EGF) and transforming growth factor a (TGFa), as well as eicosanoids. As discussed below, a specific role for eicosanoid signalling during liver regeneration has been suggested by analyses of prostaglandin production and the effect of prostaglandin supplementation or inhibition on liver regeneration in vivo and on hepatocellular proliferation in vitro. The source of prostaglandin synthesis during the hepatic regenerative response, including the cell-type specificity and the COX-isozyme specificity, and also the mechanisms by which prostaglandins exert their influence on the hepatic regenerative response, have been the subject of additional investigations. The importance of eicosanoid signalling has also been demonstrated in other animal models of liver injury, and its clinical relevance in human liver disease has been investigated.
Eicosanoid Synthesis in the Regenerating Liver Because of the known relationship between prostaglandin production and cellular proliferation in lymphocytes and fibroblasts in vitro, the relationship between prostaglandin synthesis and liver regeneration in vivo has been examined using the rodent partial hepatectomy model system (MacManus and Braceland 1976). Early studies showed increased prostaglandin E content as well as increased prostaglandin synthetic capacity in the regenerating liver. The cellular source for PGE2 production in the regenerating liver was suggested by subsequent studies demonstrating LPS-stimulated PGE2 production in Kupffer cells harvested from regenerating but not quiescent rat liver (Callery et al 1991). In the original report by McManus and Braceland, increased serum prostaglandin levels were not detected after partial hepatectomy. Subsequent analyses, however, have shown that serum PGE2 levels are significantly and specifically increased following partial hepatectomy. In one study, this was shown to occur in a biphasic manner, peaking at 3 and 10 h after surgery (Tsujii et al 1993). Specificity for increased serum prostaglandin levels during liver regeneration was further demonstrated by a recent study, in which serum PGE2 levels were noted to be increased after partial hepatectomy but not after sham surgery (Casado et al 2001). The demonstration that PGE binding sites are specifically downregulated in hepatocytes harvested from regenerating but not quiescent rat liver further implicates the early synthesis of PGE2 during liver regeneration (Hashimoto et al 1991). The specific PGE receptor subtype was not identified in that analysis. Hepatic synthesis of other eicosanoid products during liver regeneration has also been demonstrated, e.g. studies have shown that a transition occurs from PGF2a production in the quiescent liver to thromboxane B2 production in the regenerating liver (Mahmud et al 1980; Miura and Fukui 1979). Studies in our laboratory have shown that, in addition to PGE2, the prostacyclin metabolite 6-keto-PGF1a can also be detected in regenerating mouse liver (Rudnick et al 2001). Increased serum levels of leukotrienes, specifically leukotriene B4, have also been detected during hepatic regeneration (Urade et al 1996). The differences in the specific eicosanoids detected in regenerating liver in these various reports may reflect temporal or cell-specific modulation of expression of the various synthases that act downstream of COX during prostaglandin biosynthesis. We have observed evidence for this in studies in our laboratory, which suggest that expression of prostaglandin D synthase is induced greater than five-fold in the regenerating liver after partial hepatectomy (Rudnick DA and Muglia LJ, unpublished observations).
Eicosanoids and Hepatocellular Proliferation In Vitro A number of studies have shown that prostaglandins exert mitogenic activity on hepatocytes in culture. Supplementation of primary neonatal and adult rat hepatocytes in culture with arachidonic acid or various prostaglandins, including PGA1, PGE1, PGE2, PGI2 and PGF2, increases cellular DNA synthesis and mitosis (Andreis et al 1981; Kimura et al 2000; Skouteris et al 1988; Tsujii et al 1993). The proliferative effect induced by arachidonic acid in these studies was generally more pronounced than that of the individual prostaglandins and was inhibited by co-treatment with relatively high concentrations of indomethacin, an inhibitor of both COX-1 and COX-2. Somewhat surprisingly, lower concentrations of indomethacin, and also of imidazole, a thromboxane synthesis inhibitor, stimulated hepatocellular proliferation in this in vitro model. Arachidonic acid but not PGF2a could also stimulate hepatocellular DNA synthesis in an in vitro model of isolated perfused liver, to a degree comparable to that induced after partial hepatectomy. Again this activity was inhibited by indomethacin, imidazole, and also by aspirin (Kanzaki et al 1979; Miura et al 1977). In addition to stimulating hepatocellular proliferation, prostaglandins, including prostacyclin, PGD2 and PGE1, have also been shown to inhibit TGFb-induced hepatocellular apoptosis in culture (Kroll et al 1998). Modulation of hepatocellular prostaglandin production by growth factors, including those implicated in the regulation of liver regeneration, has also been examined in cell culture. The proliferative activity of HGF and EGF on primary hepatocytes in culture is inhibited by indomethacin, suggesting that, at least in this cell culture model, these growth factors mediate their proliferative influence by inducing the synthesis of prostaglandins, which, in turn, direct subsequent downstream signalling events in an autocrine fashion (Adachi et al 1995). TGFa stimulates both PGE2 and PGF2a secretion from primary hepatocytes proliferating in cell culture (Skouteris et al 1988). In one study, specific inhibition of the EP1 prostaglandin receptor inhibited the stimulatory effects of PGE1 and PGE2 on hepatocellular DNA synthesis in culture, implicating this receptor subtype in mediating at least some of these downstream events (Kimura et al 2000). The specific intracellular signalling cascades activated by prostaglandin signalling and modulating increased cellular proliferation have also been examined in vitro. These appear to involve the activation of phospholipase C and a tyrosine kinase but not adenylate cyclase, based on inhibition of the proliferative activity of prostaglandins with specific pharmacological inhibitors of the former but not the latter (Kimura et al 2000; Refsnes et al 1995). The production of prostaglandins by and effects of prostaglandins on non-parenchymal cells of the liver has also been examined in cell culture model systems, e.g. Kupffer cells harvested from regenerating liver during the peak hepatocellular proliferative response exhibit increased PGE2 synthetic capacity (Callery et al 1991). The effect of prostaglandins on stellate cells in culture has also been examined (Mallat et al 1996). Stellate cells are activated to proliferate and synthesize components of the extracellular matrix that contribute to fibrosis in the setting of chronic liver disease. The expression of the endothelin receptor, which mediates the inhibitory action of endothelin on the proliferation of activated stellate cells in culture, is induced by prostaglandin supplementation. Furthermore, the COX antagonist ibuprofen blocks the growth-inhibitory activity of endothelin on stellate cells. Thus, prostaglandins appear to inhibit the proliferation of stellate cells while promoting the proliferation of hepatocytes.
LIVER REGENERATION Eicosanoids and Hepatocellular Proliferation In Vivo The importance of eicosanoid signalling, specifically that by prostaglandins in vivo during the hepatic regenerative response following injury, has been further emphasized in studies that show that pharmacological or genetic inhibition of prostaglandin synthesis impairs the hepatic regenerative response, while supplementation with prostaglandins or their precursors, or inhibition of leukotriene production, augments this response (Table 37.1). Several reports have demonstrated that administration of indomethacin impairs the induction of hepatocellular DNA synthesis after partial hepatectomy (MacManus and Braceland 1976; Miura and Fukui 1976, 1979; Rixon and Whitfield 1982; Rudnick et al 2001). Other pharmacological inhibitors of prostaglandin production have exhibited distinct effects on liver regeneration, e.g. mefenic acid, which is also reported to inhibit COX activity, had no effect on hepatocellular DNA synthesis but blocked cell cycle progression to mitosis. Similar results were observed with both hydrocortisone and dexamethasone, which were used in these studies to inhibit prostaglandin and thromboxane production by inhibiting the activity of phospholipase A2, thereby impairing eicosanoid synthesis by blocking arachadonic acid release from phospholipid membrane. Arachadonic acid supplementation specifically reversed the effects of hydrocortisone and dexamethasone but not of mefenic acid. These observations suggest that distinct events modulated by different eicosanoids differentially mediate hepatocellular DNA synthesis and later entry into mitosis (Rixon and Whitfield 1982). The possibility exists that the inhibitory activities of hydrocortisone and dexamethasone were independent of effects on arachadonic acid release from plasma membrane, although the reversal of such inhibition by supplementation with arachidonic acid further supports the important role of eicosanoid synthesis in hepato-
Table 37.1 Effect of inhibitors of eicosanoid synthesis or signalling on hepatocellular proliferation in rodent partial hepatectomy model in vivo Drug1 Indomethacin Indomethacin Mefenic acid Hydrocortisone Dexamethasone Indomethacin Dexamethasone Imidazole Daltroban Methylprednisolone AA-861 Indomethacin SC-236 SC-560 SC-236+SC-560 NS-398 1
Effect (%)
Reference
202,4 302,3,6 403,6 503,7 453,7 62,8 202 1502 1002 2202 2112 102,9 502 1302 102 245
(MacManus 1976) (Rixon and Whitfield 1982) (Rixon and Whitfield 1982) (Rixon and Whitfield 1982) (Rixon and Whitfield 1982) (McNeil et al 1985) (Besse et al 1991) (Besse et al 1991) (Besse et al 1991) (Azzarone et al 1992) (Urade et al 1996) (Rudnick et al 2001) (Rudnick et al 2001) (Rudnick et al 2001) (Rudnick et al 2001) (Casado et al 2001)
Target of inhibition: indomethacin, mefenic acid: COX-1 and COX-2; hydrocortisone, dexamethasone: PLA2; imidazole: thromboxane synthetase; daltroban: thromboxane receptor; AA-861: 5-lipoxygenase; SC-236, NS-398: COX2; SC-560: COX-1. For dosing details, see original reference. 2 Peak DNA synthesis as percentage of untreated control. 3 Peak mitoses as percentage of untreated control. 4 Not rescued by supplementation with PGE2, dimethyl PGE2 or PGF1a. 5 Peak PCNA expression (marker of DNA synthesis) as percentage of untreated control. 6 Not rescued by supplementation with arachidonic acid. 7 Rescued by supplementation with arachidonic acid. 8 Rescued by supplementation with dimethyl PGE2. 9 Not rescued by supplementation with PGE2, PGI2 or PGF1a alone or in combination.
417
cellular progression through the cell cycle. In contrast to these studies, a different study suggested that injection of methylprednisolone can stimulate hepatocellular DNA synthesis after partial hepatectomy in rats (Azzarone et al 1992). A mechanistic explanation for this apparent inconsistency remains unknown. Another study examining the effect of thromboxanes on liver regeneration showed no effect on hepatocellular DNA synthesis after partial hepatectomy with inhibition of thromboxane production by imidazole, or with inhibition of the thromboxane receptor with Daltroban (Besse et al 1991). Conflicting observations have been made regarding the influence of prostaglandin supplementation on the hepatic regenerative response to partial hepatectomy in the absence or presence of prostaglandin synthesis inhibitors (Tables 37.1 and 37.2). McNeil et al (1985) showed that supplementation with dimethyl PGE2, which has a longer in vivo half-life than PGE2, augmented DNA synthesis and progression through mitosis after partial hepatectomy in the absence or presence of indomethacin. They and others have also shown that the inhibitory effect of ethanol treatment on hepatocellular DNA synthesis and progression through mitosis during liver regeneration could be partially reversed by dimethyl PGE2 supplementation (Makowka et al 1982; McNeil et al 1985). In another study, PGE1 supplementation did not stimulate hepatocellular proliferation in rats subjected to partial hepatectomy, but did stimulate proliferation when continuously infused into the portal vein of non-operated dogs (Azzarone et al 1992). In the study by MacManus and Braceland, supplementation with PGE2, 15,15-dimethyl PGE2 or PGF1a did not rescue the inhibition of the hepatic regenerative response induced with indomethacin (MacManus and Braceland 1976). Similarly, Besse et al (1991) observed no augmentation of hepatic DNA synthesis after partial hepatectomy with PGE2 supplementation, but did note that a prostacyclin analogue, iloprost, augmented hepatocellular DNA synthesis in the absence and presence of dexamethasone after partial hepatectomy. We have examined the effect of intraperitoneal supplementation with a variety of prostaglandin mixtures on the hepatic regenerative response in mice treated with indomethacin, and have been unable to demonstrate any consistent enhancement of this response with PGE2, PGE2 in combination with PGI2 and PGF2a, or dimethyl PGE2 (Rudnick DA and Muglia LJ, unpublished observations). In contrast to prostaglandins and thromboxanes, leukotrienes, which are metabolites of 5-lipoxygenase, have been suggested to inhibit liver regeneration. Urade et al (1996) showed that inhibition of leukotriene synthesis using a specific 5-lipoxygenase inhibitor (AA-861) augmented the hepatocellular proliferative response after partial hepatectomy in rats. This effect was observed in the presence and absence of experimentally induced biliary obstruction. Because leukotrienes are excreted in the bile and can accumulate in the setting of obstructive jaundice, they could be important inhibitors of the regenerative response to liver injury in vivo in the setting of obstructive jaundice. Inhibition of 5lipoxygenase inhibition may have therapeutic value in this setting (Urade et al 1996). Table 37.2 Effect of eicosanoid supplementation on hepatocellular proliferation in rodent partial hepatectomy model in vivo Drug1 dmPGE2 PGE2 Iloprost PGE1 1
Effect (%) 2
3
140 ; 200 1202 150–2402 1062
Reference (McNeil et al 1985) (Besse et al 1991) (Besse et al 1991) (Azzarone et al 1992)
dmPGE2; 16,16-dimethyl PGE2; iloprost, PGI2 analogue. For dosing details, see original reference. Peak DNA synthesis as percentage of untreated control. 3 Peak mitoses as percentage of untreated control. 2
418
DIGESTIVE SYSTEM
Cyclooxygenase Specificity of Hepatic Prostaglandin Synthesis during Liver Regeneration The COX-isozyme specificity of prostaglandin synthesis during hepatic regeneration has been investigated in the partial hepatectomy model system, using potent and highly specific COX isozyme specific inhibitors (Gierse et al 1996; Penning et al 1997; Smith et al 1998) and COX null mice (Langenbach et al 1995; Morham et al 1995). In our own studies, liver regeneration was moderately impaired in the presence of a COX-2 specific inhibitor, SC-236, but delayed rather than reduced in the presence of a COX-1 specific inhibitor, SC-560. The specificity of these effects was further indicated by the demonstration that a mixture of the two isozyme-specific inhibitors recapitulated the degree of inhibition of liver regeneration seen with indomethacin. This observation also leaves open the possibility that distinct functional roles exist for COX-1- and COX-2-derived prostaglandins during liver regeneration. We also observed that the regenerative response was normal in COX-1 null mice, but significantly impaired in COX-1 null mice treated with the COX-2 specific inhibitor (Rudnick et al 2001). Consistent results have been reported by Casado et al (2001), who observed impaired liver regeneration in rats treated with another COX-2 specific inhibitor, NS-398, and also in the COX-2 null mouse. These investigators also reported that COX-2 expression was induced in the regenerating liver and primarily restricted to hepatocytes. Our own analyses of hepatic COX-2 and COX-1 gene expression during liver regeneration by cDNA microarray-based gene expression analysis showed slight induction of COX-2 mRNA early (2 h) and late (36 h) after partial hepatectomy, with no apparent induction of COX-1 over this same period of time (Rudnick DA and Muglia LJ, unpublished observation). Taken together, these data suggest that COX-2-derived prostaglandins play the predominant role during liver regeneration, and leave open the possibility that COX-1-derived prostaglandins are also involved. Prostaglandins and Signal Transduction during Liver Regeneration In Vivo One mechanism by which prostaglandins exert their effects on target cells is through interaction with various G-protein-coupled membrane-spanning receptors. Specific prostanoids interact with unique receptors, which in turn are coupled to distinct G-proteins. Different G-proteins can stimulate or inhibit the activity of adenylate cyclase, thereby modulating downstream signalling events in distinct ways. Additional specificity in this signalling system is introduced by the unique expression patterns of eight distinct receptors for each of different prostanoids (PGD2, PGE2, PGF2, PGI2 and thromboxane A2) in each of the principal liver cell types (hepatocytes, Kupffer cells, endothelial cells and stellate cells) (Negishi et al 1995). The relationship between prostaglandin signalling and the activation of cAMP-dependent secondary messenger signalling cascades during liver regeneration has been suggested by several observations. For example, the biphasic temporal pattern of peak serum PGE2 levels after partial hepatectomy observed by Tsujii et al (1993) correlates closely with the pattern of hepatic cAMP production in the regenerating liver (Fazia et al 1997). Consistent results were seen in an in vitro model in which perfusion of a hepatic remnant harvested after partial hepatectomy with PGE2 specifically induced elevated cAMP levels (Tsujii et al 1993). We examined the effect of COX inhibition on this signalling pathway by measuring hepatic activation of the cAMP-response element binding protein (CREB), a transcription factor whose activity is regulated by cAMP-dependent protein kinase phosphorylation. We observed
that CREB was specifically activated in the regenerating liver 1– 2 h and 12–24 h after partial hepatectomy in mice and that treatment of mice with indomethacin, using a dose sufficient to impair liver regeneration, entirely blocked this activation (Rudnick et al 2001). Although the specific genetic targets for prostaglandin-dependent transcriptional regulation by CREB during liver regeneration are not yet known, a number of candidate genes containing consensus CREB binding sites in their promoter regions have been described (Mayr and Montminy 2001). These candidates include growth factors and genes implicated in the modulation of cell cycling and cell survival. Microarray-based gene expression analysis is being applied to analyses of regulation of gene expression during liver regeneration (Li et al 2001; Kelley-Loughnane et al 2002) and will be a powerful tool for further elucidation of the role of prostaglandins and CREB activation in modulation of gene expression during liver regeneration. The cytokine signalling cascade has been implicated as important in the initiation of liver regeneration. In various in vivo studies, prostaglandins have been shown to enhance, attenuate or have no effect on cytokine signalling and the acute phase response in general. Prostaglandin-mediated attenuation of this signal transduction pathway has been illustrated by the observations that PGE2 produced from Kupffer cells in response to inflammatory stimuli can inhibit in an autocrine fashion both TNFa and IL-6 synthesis. Because prostaglandins have also been implicated as important in the initiation of the regenerative response, the interaction between prostaglandin signalling and cytokine signalling has been further examined using the partial hepatectomy model system (Yamada et al 1997). We have observed that inhibition of prostaglandin signalling, by treatment of mice with a dose of indomethacin sufficient to impair the hepatic regenerative response to partial hepatectomy, did not inhibit the hepatectomy-induced increase in serum IL-6, indicating that prostaglandins do not induce IL-6 during liver regeneration. In fact, peak serum IL-6 levels were significantly greater in indomethacin-treated mice subjected to partial hepatectomy than in their vehicle-treated counterparts, suggesting that, as has been observed in other model systems, prostaglandins downregulate IL-6 and the cytokine signalling cascade during liver regeneration (Rudnick et al 2001). Similarly, we observed that indomethacin did not impair IL-6-mediated and partial hepatectomy-induced activation of the transcription factor STAT3 after partial hepatectomy. Consistent with these observations, the addition of indomethacin to Kupffer cells harvested from regenerating liver and grown in culture increased LPS-induced IL1 and IL-6 levels (Goss et al 1992, 1993). Their studies and those of others (Williams and Shacter 1997) indicate that Kupffer cells are distinct from other macrophage populations in this response, e.g. indomethacin as well as the COX-2-specific inhibitor NS-398 can inhibit the induction of IL-6 gene expression from peritoneal macrophages harvested after an inflammatory stimulus in vivo and grown in culture (Hinson et al 1996). This effect was reversible by co-supplementation of the cultured cells with PGE2. The influence of prostaglandins on signalling pathways and gene expression as described above is summarized in Figure 37.1. One mechanism by which the coordinated interaction between these signalling pathways might be modulated during regeneration is suggested by the studies of Fennekohl et al (2000). These investigators showed that both in vitro and in vivo IL-6 treatment induces the hepatocellular expression of the Gs-coupled prostanoid receptors EP2-R and EP4-R, and that PGE2 stimulation of these newly induced receptors inhibits IL-6-induced a2-macroglobulin expression, thereby establishing a prostanoid-mediated feedback inhibition loop for attenuation of the cytokine signalling cascade. Such prostanoid signalling also appears to terminate IL-6 production.
LIVER REGENERATION
419
Figure 37.1 Prostaglandins and signal transduction during liver regeneration. Prostaglandins are required for activation of CREB and mediate inhibition of the cytokine signalling cascade to modulate gene expression during liver regeneration
A number of other observations have also suggested that prostaglandins interact with cytokines and other signalling molecules that have been implicated in the modulation of the hepatic regenerative response. TNFa can induce hepatocellular phospholipase A2 expression in the presence of HGF (Georgakopoulos et al 1995) and also can induce COX-2 expression (Jobin et al 1998). Prostaglandins themselves also induce the expression of HGF (Matsumoto et al 1995) and, conversely, HGF induces prostaglandin production in a variety of cell types (Hori et al 1993). The effects of prostaglandins on liver regeneration are also likely to involve the modulation of other signalling pathways not yet characterized. One likely possibility involves the peroxisomal proliferator activated receptors (PPARs). PPARs are transcriptional factors that are members of the nuclear steroid hormone receptor superfamily. They were first identified as cellular receptors responsible for mediating the biological effects of a diverse group of compounds, collectively termed ‘‘peroxisomal proliferators’’ (Bieri 1993; Latruffe and Vamecq 1997). The peroxisomal proliferators include hypolipidaemic drugs, industrial plasticizers, herbicides and other substances. When administered in vivo to rodents, these chemicals induce hepatomegaly, peroxisomal proliferation, fatty acid b-oxidation and, with chronic administration, hepatocellular carcinoma. Treatment of cultured liver cells in vitro stimulates DNA synthesis and suppresses apoptosis. The effects of peroxisomal proliferators on hepatocellular proliferation in vitro and in vivo suggest that the PPARs may play an important role in the modulation of growth control in the liver. Various prostaglandins and leukotrienes can bind to and activate specific PPARs (Yu et al 1995), e.g. prostaglandin J2, a metabolite of prostaglandin D, binds to and regulates the activity of the g isoform of PPAR. PPARs themselves have been shown to modulate the expression of COX-2 (Ledwith et al 1997). These observations suggest that eicosanoids may in part mediate their effects on hepatic regeneration by their interactions with PPARs and the resulting influence on PPAR-dependent transcriptional regulation. EICOSANOIDS AND LIVER DISEASE The cytoprotective effects of prostaglandins in experimental models of liver injury and the resulting unsuccessful attempts to use prostaglandins in the treatment of human liver disease, including acute liver failure, cirrhosis, liver transplant graft rejection and primary graft non-function, were reviewed by Tolman (2000). Prostaglandins have also been implicated in the promotion of unregulated cellular proliferation in many forms of
cancer (Subbaramaiah et al 1997). Several tumours have been shown to produce increased amounts of prostaglandins. Furthermore, inhibition of this prostaglandin production using COX inhibitors appears to be protective in animal models and in humans against the development of many cancers. The role of prostaglandin signalling in liver cancer has been suggested, although less well defined. COX-2, but not COX-1, appears to be overexpressed in well-differentiated hepatocellular carcinoma (HCC) compared to both normal liver tissue and poorly differentiated HCC (Koga et al 1999; Shiota et al 1999). HCC tissue also appears to contain decreased levels of eicosanoid precursors and increased levels of prostaglandin products compared to normal liver tissue (Hanai et al 1993). COX-2 overexpression has also been reported to occur during early stages and COX-1 overexpression in later stages of cholangiocarcinoma (Chariyalertsak et al 2001). These data suggest that prostaglandin synthesis is likely to be increased in HCC and cholangiocarcinoma compared to normal liver; however, the functional significance of this is unkown. Human HCC has also been reported to show decreased prostaglandin E1 binding capacity, suggesting alteration of hepatocellular prostaglandin receptor protein content during carcinogenesis (Virgolini et al 1989). One mechanism by which prostaglandin signalling may promote hepatic tumorigenesis is suggested by two recent reports demonstrating that COX-2 expression inhibits apoptosis in cholangiocarcinoma cells (Nzeako et al 2002) and that inhibition of COX-2 activity promotes apoptosis and inhibits proliferation in hepatoma cell lines (Bae et al 2001). SUMMARY AND CONCLUSIONS The role of eicosanoids, including prostaglandins, thromboxanes and leukotrienes, in the modulation of the hepatic growth response to injury has been and is being characterized by studies in vitro and in vivo. Our understanding of the signal transduction mechanisms by which eicosanoids exert these effects is also being advanced using these experimental model systems. In addition to their effects on hepatocellular proliferation after injury, the involvement of eicosanoids in mediating the uncontrolled cellular proliferation in the setting of liver cancer has also been suggested. Pharmacological modulation of prostaglandin synthesis and signalling has not yet been shown to have clinical benefit in the settings of human liver diseases. Increased understanding of the molecular basis through which these molecules mediate their biological effects provided by these kinds of analyses may lead us to the development of novel approaches to such modulation that will have therapeutic benefit.
420
DIGESTIVE SYSTEM
ACKNOWLEDGEMENTS This work was supported in part by grants from the National Institutes of Health, by an American Digestive Health Foundation/American Gastroenterological Association Research Scholar Award (to D.A.R.), and by a Burroughs Wellcome Fund Career Development Award (to L.J.M.). D.A.R. is a scholar of the Washington University School of Medicine Child Health Research Center of Excellence in Developmental Biology (HD33688).
REFERENCES Adachi T, Nakashima S, Saji S et al (1995) Roles of prostaglandin production and mitogen-activated protein kinase activation in hepatocyte growth factor-mediated rat hepatocyte proliferation. Hepatology, 21, 1668–1674. Andreis PG, Whitfield JF and Armato U (1981) Stimulation of DNA synthesis and mitosis of hepatocytes in primary cultures of neonatal rat liver by arachidonic acid and prostaglandins, Exp Cell Res, 134, 265– 272. Azzarone A, Francavilla A, Carrieri G et al (1992) Effects on in vivo and in vitro hepatocyte proliferation of methylprednisolone, azathioprine, mycophenolic acid, mizoribine and prostaglandin E1. Transpl Proc, 24, 2868–2871. Bae SH, Jung ES, Park YM et al (2001) Expression of cyclooxygenase-2 (COX-2) in hepatocellular carcinoma and growth inhibition of hepatoma cell lines by a COX-2 inhibitor, NS-398. Clin Cancer Res, 7, 1410–1418. Besse T, de Hemptinne B, Kabeya V and Lambotte L (1991) Stimulation of liver regeneration by prostacyclin. Transpl Proc, 23, 542–544. Bieri F (1993) Peroxisome prolferators and cellular signalling pathways. A review. Biol Cell, 77, 43–46. Callery MP, Mangino MJ and Flye MW (1991) Kupffer cell prostaglandin-E2 production is amplified during hepatic regeneration. Hepatology, 14, 368–372. Casado M, Callejas NA, Rodrigo J et al (2001) Contribution of cyclooxygenase 2 to liver regeneration after partial hepatectomy. FASEB J, 15, 2016–2018. Chariyalertsak S, Sirikulchayanonta V, Mayer D et al (2001) Aberrant cyclooxygenase isozyme expression in human intrahepatic cholangiocarcinoma. Gut, 48, 80–86. Decker K (1985) Eicosanoids, signal molecules of liver cells. Semin Liver Dis, 5, 175–190. Diehl AM and Rai R (1999) Liver regeneration. In Schiff R, Sorrell MF and Maddrey WC (eds), Schiff’s Diseases of the Liver, 8th edn. Philadelphia: Lippincott-Raven, 39–54. Fazia MAD, Servillo G and Sassone-Corsi P (1997) Cyclic AMP signalling and cellular proliferation: regulation of CREB and CREM. FEBS Lett, 410, 22–24. Fennekohl A, Lucas M and Puschel GP (2000) Induction by interleukin 6 of G(s)-coupled prostaglandin E(2) receptors in rat hepatocytes mediating a prostaglandin E(2)-dependent inhibition of the hepatocyte’s acute phase response. Hepatology, 31, 1128–1134. Fisher B, Szuch P, Levine M and Fisher ER (1971) A portal blood factor as the humoral agent in liver regeneration. Science, 171, 575–577. Georgakopoulos E, McMenamin M and Skouteris GG (1995) Phospholipase A2 is activated by tumor necrosis factor-a in primary hepatocytes stimulated by a deleted form of hepatocyte growth factor. Biochem Biophys Res Commun, 217, 1263–1270. Gierse JK, McDonald JJ, Hauser SD et al (1996) A single amino acid difference between cyclooxygenase-1 (COX-1) and -2 (COX-2) reverses the selectivity of COX-2 specific inhibitors. J Biol Chem, 271, 15810– 15814. Goss JA, Mangino MJ, Callery MP and Flye MW (1993) Prostaglandin E2 downregulates Kupffer cell production of IL-1 and IL-6 during hepatic regeneration. Am J Physiol, 264, G601–G608. Goss JA, Mangino MJ and Flye MW (1992) Prostaglandin E2 production during hepatic regeneration downregulates Kupffer cell IL-6 production. Ann Surg, 215, 553–559.
Gross G, Imamura T and Muglia LJ (2000) Gene knockout mice in the study of parturition. J Soc Gynecol Invest, 7, 88–95. Hanai T, Hashimoto T, Nishiwaki K et al (1993) Comparison of prostanoids and their precursor fatty acids in human hepatocellular carcinoma and non-cancerous reference tissues. J Surg Res, 54, 57–60. Hashimoto N, Watanabe T, Ikeda Y et al (1991) Downregulation of prostaglandin E2 receptors in regenerating rat liver and its physiological significance. Biochem Biophys Res Commun, 176, 226– 232. Higgins GM and Anderson RM (1931) Experimental pathology of the liver. 1. Restoration of the liver of the white rat following partial surgical removal. Arch Pathol, 12, 186–202. Hinson RM, Williams JA and Shacter E (1996) Elevated interleukin-6 is induced by prostaglandin E2 in a murine model of inflammation: possible role of cyclooxygenase-2. Proc Natl Acad Sci USA, 93, 4885– 4890. Hori T, Shibamoto S, Hayakawa M et al (1993) Stimulation of prostaglandin production by hepatocyte growth factor in human gastric carcinoma cells. FEBS Lett, 334, 331–334. Huber M and Keppler D (1990) Eicosanoids and the liver. Progr Liver Dis, 9, 117–141. Jobin C, Morteau O, Han DS and Balfour SR (1998) Specific NF-kB blockade selectively inhibits tumour necrosis factor-a induced COX-2 but not constitutive COX-1 gene expression in HT-29 cells. Immunology, 95, 537–543. Kanzaki Y, Mahmud I, Asanagi M et al (1979) Thromboxane as possible trigger of liver regeneration. Cell Mol Biol, 25, 147–152. Kelley-Loughnane N, Sabla GE, Ley-Ebert C et al (2002) Independent and overlapping transcriptional activation during liver development and regeneration in mice. Hepatology, 35, 525–534. Kimura M, Osumi S and Ogihara M (2000) Stimulation of DNA synthesis and proliferation by prostaglandins in primary cultures of adult rat hepatocytes. Eur J Pharmacol, 404, 259–271. Koga H, Sakisaka S, Ohishi M et al (1999) Expression of cyclooxygenase-2 in human hepatocellular carcinoma: relevance to tumor differentiation. Hepatology, 29, 688–696. Kroll B, Kunz S, Tu N and Schwarz LR (1998) Inhibition of transforming growth factor-b1 and UV light-induced apoptosis by prostanoids in primary cultures of rat hepatocytes. Toxicol Appl Pharmacol, 152, 240– 250. Kuiper J, Zijlstra FJ, Kamps JA and van Berkel TJ (1988) Identification of prostaglandin D2 as the major eicosanoid from liver endothelial and Kupffer cells. Biochim Biophys Acta, 959, 143–152. Langenbach R, Morham SG, Tiano HF et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Latruffe N and Vamecq J (1997) Peroxisome proliferators and peroxisomal proliferator activated receptors (PPARs) as regulators of lipid metabolism. Biochimie, 79, 81–94. Ledwith BJ, Pauley CJ, Wagner LK et al (1997) Inhibition of cyclooxygenase 2 expression by peroxisome proliferators and nontetradecanoylphorbol 12,13 myristate-type tumor promoters in immortalized mouse liver cells. J Biol Chem, 272, 3707–3714. Li W, Liang X, Leu JI et al (2001) Global changes in interleukin-6 dependent gene expression patterns in mouse liver after partial hepatectomy. Hepatology, 33, 1377–1386. MacManus JP and Braceland BM (1976) A connection between the production of prostaglandins during liver regeneration and the DNA synthetic response. Prostaglandins, 11, 609–620. Mahmud I, Fukui N and Miura Y (1980) Arachidonic acid metabolism in normal and regenerating liver and hepatoma. Adv Enzyme Reg, 18, 27– 37. Makowka L, Falk RE, Cohen MM et al (1982) Protective effect of 16,16 dimethyl prostaglandin E2 on acute ethanol induced inhibition of hepatic regeneration. Surg Forum, 33, 183–186. Mallat A, Preaux AM, Serradeil-Le Gal C et al (1996) Growth inhibitory properties of endothelin-1 in activated human hepatic stellate cells: a cyclic adenosine monophosphate-mediated pathway. Inhibition of both extracellular signal-regulated kinase and c-Jun kinase and upregulation of endothelin B receptors. J Clin Invest, 98, 2771–2778. Matsumoto K, Okazaki H and Nakamura T (1995) Novel function of prostaglandins as inducers of gene expression of HGF and putative mediators of tissue regeneration. J Biochem, 117, 458–464.
LIVER REGENERATION Mayr B and Montminy M (2001) Transcriptional regulation by the phosphorylation-dependent factor CREB. Nature Rev Mol Cell Biol, 2, 599–609. McNeil GE, Chen TS and Leevy CM (1985) Reversal of ethanol and indomethacin-induced suppression of hepatic DNA synthesis by 16,16dimethyl prostaglandin E2. Hepatology, 5, 43–46. Michalopoulos GK and DeFrances MC (1997) Liver regeneration. Science, 276, 60–66. Miura Y and Fukui N (1976) Pleiotypic responses of regenerating liver. Adv Enzyme Reg, 14, 393–405. Miura Y and Fukui N (1979) Prostaglandins as possible triggers for liver regeneration after partial hepatectomy. A review. Cell Mol Biol, 25, 179–184. Miura Y, Fukui N, Kanzaki Y and Mahmud I (1977) Induction of ornithine decarboxylase and DNA synthesis in isolated perfused liver after partial hepatectomy. Adv Enzyme Reg, 16, 195–203. Moolten FL and Bucher NLR (1967) Regeneration of rat liver: transfer of humoral agent by cross circulation. Science, 158, 272–274. Morham SG, Langenbach R, Loftin CD et al (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 83, 473–482. Needleman P, Turk J, Jakschik BA et al (1986) Arachidonic acid metabolism. Ann Rev Biochem, 55, 69–102. Negishi M, Sugimoto Y and Ichikawa A (1995) Molecular mechanisms of diverse actions of prostanoid receptors. Biochim Biophys Acta, 1259, 109–119. Nzeako UC, Guicciardi ME, Yoon J-H et al (2002) COX-2 inhibits fasmediated apoptosis in cholangiocarcinoma cells. Hepatology, 35, 552– 559. Penning TD, Talley JJ, Bertenshaw SR et al (1997) Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)1H-pyrazol-1-yl]benzenesulfonamide (SC-58635, celecoxib). J Med Chem, 40, 1347–1365. Refsnes M, Dajani OF, Sandnes D et al (1995) On the mechanisms of the growth-promoting effect of prostaglandins in hepatocytes: the relationship between stimulation of DNA synthesis and signalling mediated by adenylyl cyclase and phosphoinositide-specific phospholipase C. J Cell Physiol, 164, 465–473.
421
Rixon RH and Whitfield JF (1982) An early mitosis-determining event in regenerating rat liver and its possible mediation by prostaglandins or thromboxane. J Cell Physiol, 113, 281–288. Rudnick DA, Perlmutter DH and Muglia LJ (2001) Prostaglandins are required for CREB activation and cellular proliferation during liver regeneration. Proc Natl Acad Sci USA, 98, 8885–8890. Shiota G, Okubo M, Noumi T et al (1999) Cyclooxygenase-2 expression in hepatocellular carcinoma. Hepatogastroenterology, 46, 407–412. Skouteris GG, Ord MG and Stocken LA (1988) Regulation of the proliferation of primary rat hepatocytes by eicosanoids. J Cell Physiol, 135, 516–520. Smith CJ, Zhang Y, Koboldt CM et al (1998) Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc Natl Acad Sci USA, 95, 13313–13318. Smith WL, Garavito RM and DeWitt DL (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem, 271, 33157–33160. Subbaramaiah K, Zakim D, Weksler BB and Dannenberg AJ (1997) Inhibition of cyclooxygenase: a novel approach to cancer prevention. Proc Soc Exp Biol Med, 216, 201–210. Tolman KG (2000) Eicosanoids and the liver. Prostagland Lipid Mediat, 61, 163–174. Tsujii H, Okamoto Y, Kikuchi E et al (1993) Prostaglandin E2 and rat liver regeneration. Gastroenterology, 105, 495–499. Urade M, Izumi R and Kitagawa H (1996) Inhibition of 5-lipoxygenase promotes the regeneration of the liver after partial hepatectomy in normal and icteric rats. Hepatology, 23, 544–548. Virgolini I, Sinzinger H, Muller C and Hermann M (1989) Human hepatocellular carcinoma cancers show decreased prostaglandin E1 binding capacity. Br J Cancer, 59, 407–409. Williams JA and Shacter E (1997) Regulation of macrophage cytokine production by prostaglandin E2. Distinct roles of cyclooxygenase-1 and -2. J Biol Chem, 272, 25693–25699. Yamada Y, Kirillova I, Peschon JJ and Fausto N (1997) Initiation of liver growth by tumor necrosis factor: deficient liver regeneration in mice lacking type I tumor necrosis factor receptor. Proc Natl Acad Sci USA, 94, 1441–1446. Yu K, Bayona W, Kallen CB et al (1995) Differential activation of peroxisome proliferator-activated receptors by eicosanoids. J Biol Chem, 270, 23975–23983.
38 Eicosanoids and the Intestine Klaus Bukhave1 and Jørgen Rask-Madsen2 1The
Royal Veterinary and Agricultural University and 2University of Copenhagen, Denmark
Eicosanoids are abundant in the human gut. Their biological effects include regulation of intestinal fluid and ion transport, motor activity and epithelial and endothelial barrier function, in addition to modulation of the inflammatory response in a variety of disease states (Lauritsen et al 1989; Eberhart and DuBois 1995; Nielsen and Rask-Madsen 1996; Mohajer and Ma 2000; Krause and DuBois 2000). The term ‘‘eicosanoid’’, derived from the Greek word eicosa meaning 20, describes multiple products originating in the breakdown of membrane phospholipids, which leads to the release of free arachidonic acid (AA) and subsequently the formation of bioactive substances. Release of AA seems to be the rate-limiting step in the formation of most eicosanoids. This process may be catalysed not only by phospholipase A2, leading to AA and lysophopholipids, but also by the concerted action of phospholipase C and diglyceride lipase, which leads to the formation of diacylglycerol, inositoltriphosphate (IP3), monoacylglycerol and AA. Among the AA metabolites, prostaglandins (PGs), prostacycline (PGI2), thromboxanes (TXs) and leukotrienes (LTs) are the most important in the gut (Bukhave 1992). The named substances are all short-lived in the circulation. As any perturbation of cell membranes initiates de novo synthesis of eicosanoids through activation of phospholipase A2, measurements of their luminal release or the urinary excretion of their metabolites appear more reliable for studies in vivo (Lauritsen et al 1986; Bukhave 1992; Maclouf et al 1998). All of the named lipid mediators act through specific plasma membrane receptors and, for example, PGE2 interacts with four specific G protein-coupled receptors, among which the EP1 receptor increases intracellular Ca2+, EP2 and EP4 increase cyclic adenosine monophosphate (cAMP) and EP3 decreases cAMP (see Belley and Chadee 1999). Theoretically, the effect of a specific eicosanoid product may, therefore, differ in different cells, based on the type of receptor expressed and the signal transduction pathway used (Eberhart and DuBois 1995; Davies and Rampton, 1997; Krause and DuBois 2000). The purpose of this review is to provide a brief overview of the significance of eicosanoids in intestinal physiology and pathophysiology, primarily focusing on their relation to intestinal secretion, epithelial and endothelial barrier function, motor activity and the inflammatory response, in addition to the effects of inhibiting the biosynthesis of PGs and LTs in human disease.
ORIGIN AND PATHWAYS OF EICOSANOIDS The major eicosanoid products vary between cells and between the epithelial and subepithelial layers of the gut. The cyclooxygenase enzymes (COXs) catalyse the formation of prostanoids (PGs, TXs, PGI2) from AA, and consist of at least two enzymes The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
existing as distinctly regulated forms expressed by separate genes and differing in distribution, regulation and expression (Smith et al 2000). COX-1 is constitutively expressed in the intestine and regulates many of the physiological actions of PGs (O’Neill and Ford-Hutchinson 1993), while the expression of the COX-2 gene appears to be highly regulated by a number of transcription factors, in particular by NF-kB (Abate et al 1998), and is induced in inflammatory conditions, such as coeliac disease, Crohn’s disease and ulcerative or ischaemic colitis. Many physiological processes in the intestine are modulated by PG ligand–receptor interactions. The levels of PGs in cells are controlled by a regulatory protein, phospholipase A2 activating protein, and by activation of phospholipase C and D (Ribardo et al 2001). In the intestine, PGs are produced mainly by immune cells in the lamina propria and mesenchymal cells in the epithelium (Eberhart and DuBois 1995; Powell et al 1999) and to a lesser extent by the epithelial cells lining the luminal border of the mucosa. While most mammalian cells contain the COXs, relatively few cells contain the enzyme 5-lipoxygenase, which is responsible for the conversion of AA to LTA4 (Henderson 1994). The latter unstable intermediate can be further metabolized to LTB4 by an LTAhydrolase contained in neutrophils, macrophages and mast cells, but not in eosinophils (Rouzer et al 1986). Alternatively, LTA4 may be converted to LTC4 by a glutathione transferase contained in eosinophils but absent in neutrophils (Murphy et al 1979). The metabolism of LTC4 involves subsequent cleavage of glutamic acid and glycine, resulting in the formation of LTD4 and E4. BIOLOGICAL EFFECTS AND FUNCTIONAL ROLE OF EICOSANOIDS With regard to the functional role of eicosanoids, it may be speculated that they reinforce or synergize normal homeostatic mechanisms, which could occur in their absence, but proceed more efficiently in their presence. This may have a dramatic effect, especially in the diseased organism, where uncontrolled formation of eicosanoids may cause diarrhoea, in addition to being critical in chronic bowel inflammation for cell differentiation and proliferation (Rask-Madsen et al 1990). Table 38.1 lists diarrhoeal diseases in which eicosanoids play a major pathogenic role.
Epithelial Transport Intestinal absorption and secretion of fluid and electrolytes are tightly regulated by neural, paracrine, hormonal and immunological elements. As mentioned above, the effector cells, which are
424
DIGESTIVE SYSTEM
Table 38.1 Diarrhoeal diseases in which excess eicosanoid synthesis plays a major pathogenic role (Rask-Madsen 1986; Davies and Rampton 1997) Infectious diarrhoea (cholera, salmonellosis, shigellosis) Endocrine diarrhoea (carcinoid syndrome, medullary carcinoma of thyroid, vipoma) Rectal villous carcinoma Bile acid diarrhoea Laxative-induced diarrhoea Radiation colitis Irritable bowel syndrome Toddler’s diarrhoea Food intolerance Ulcerative colitis Crohn’s disease
likely to be involved in control of intestinal electrolyte transport, include immune cells in the lamina propria and mesenchymal cells in the epithelium (Eberhart and DuBois 1995; Powell et al 1999). Mast cells, phagocytes and neutrophils present in the lamina propria elaborate a host of soluble mediators, which are capable of stimulating chloride secretion (Eberhart and DuBois 1995), and the close proximity of subepithelial myofibroblasts to intestinal epithelial cells below the basement membrane suggests that these cells are paracrine regulators of intestinal electrolyte transport (Powell et al 1999). This means that the action of most secretagogues is indirect and requires paracellular and/or intracellular mediators to elicit the final secretory response (Armstrong 1987). Among these, eicosanoids have been shown to be important modulators of intestinal secretion and are thought to play roles in various types of watery diarrhoeas, as well as conditions of diarrhoeas associated with inflammation (RaskMadsen 1986, 1987; Rask-Madsen et al 1990) and intestinal anaphylaxis (Serafin and Austen 1987). Exogenous PGs stimulate secretion of fluid in both the small intestine and the large bowel. The relative potencies of various PGs in inducing fluid secretion is as follows: PGE24PGF2a4 PGA24PGD24PGI2 (Mohajer and Ma 2000). Markedly increased levels of AA metabolites have been demonstrated in animal models as well as human disease, in which eicosanoids are capable of stimulating intestinal fluid and chloride secretion, in addition to inhibiting intestinal NaCl and fluid absorption (RaskMadsen et al 1984). An early study demonstrated that exogenous + PGs stimulate ileal secretion of Cl7, HCO7 and fluid by 3 , Na increasing the serosa-to-mucosa unidirectional fluxes and decreasing the mucosa-to-serosa unidirectional fluxes of Cl7 and Na+ in rabbit tissues in vitro (Al-Awquati and Greenough 1972). Later, it was confirmed that the addition of PGE2 to the serosal side produces a reduction in net sodium transfer and an increase in the serosa-to-mucosa unidirectional flux of Cl7 in human jejunal tissues mounted in Ussing chambers in vitro (Bukhave and Rask-Madsen 1980) and in segments of rat jejunum in vivo (Brunsson et al 1987). These experiments further demonstrated that pretreatment with indomethacin increases the sensitivity of the tissue to exogenous PGE2 by inhibiting the formation of endogenous PGs in vitro, so that the dose–response curve for exogenous PGs was shifted to the left (Bukhave and Rask-Madsen 1980). During the past few decades, important advances have been made in the understanding of membrane transport processes as a result of studies on intestinal epithelia exposed to various secretagogues. Traditionally, cholera has served as a model for intestinal secretion in experimental work, and PGE2, in addition to cholera toxin (CT) and vasoactive intestinal polypeptide (VIP), have been considered the ‘‘classic’’ receptor-mediated cAMP stimuli. Thus, supraphysiological doses of PGE2 (1077–1074 M) increase
mucosal cAMP levels and their effects on ion transport in the small intestine parallel that of CT and VIP (Beubler et al 1986). The cAMP-dependent secretagogue, VIP, as well as the cAMPindependent secretagogue, 5-hydroxytryptamine (5-HT), reverse fluid absorption into secretion in tied-off loops of rat jejunum in vivo, but only 5-HT causes an increase in luminal release of PGE2 (Beubler et al 1986; Rask-Madsen et al 1990). Once generated, secondary messengers induce secretion by stimulating chloride and bicarbonate secretion and by inhibiting sodium and chloride absorption (Racusen and Binder 1980; Rask-Madsen et al 1984, 1990; Powell 1986; Diener et al 1988). Ca2+ seems to be the intracellular mediator of PG-induced intestinal secretion (Beubler et al 1986), and at doses which do not affect the intracellular levels of cAMP only the active enantiomer of the calcium channel blocker, L-verapamil, inhibits the secretory response to ‘‘lowdose’’ intra-arterial PGE2. Also in the rat small intestine, Ca2+ stimulates Cl7 secretion and inhibits Na+ and Cl7 absorption— an action that seems to be mediated by calmodulin, calmodulindependent kinases and PKC pathways (Powell et al 1985). PGs are considered to modulate secretion via central and peripheral neural transmission. Thus, PGs stimulate the release of acetylcholine from cholinergic enteric nerves (Keast 1987), which in turn activates muscarinic receptors on the epithelial cells, resulting in enhanced influx of Ca2+ through hydrolysis of phosphatidyl choline and, subsequently, stimulation of intestinal secretion. PGE2 administered orally to rats induces small intestinal secretion of fluid (Ruwart et al 1979), which is prevented by pretreatment with indomethacin. Also, intravenous administration of PGE2 and PGF2a stimulates a secretory response in the human jejunum (Milton-Thomson et al 1975). In the guinea-pig and rat colon, an electrogenic mechanism, similar to that found in other fluid secretory epithelia, is stimulated by PGE2 to secrete both Cl7 and K+. Recent data indicate that pathophysiological concentrations of PGE2 (4100 nM) stimulate electrogenic K+ secretion through activation of EP2 receptors, while electrogenic KCl secretion may occur in vivo through activation of DP receptors at physiological concentrations of PGD2 (5100 M) (Halm and Halm 2001). The traditional scheme of intracellular events leading to PGinduced secretion involves activation of the intracellular mediators, cAMP, cyclic guanidine phosphate (cGMP), IP3 (which releases ionized calcium) and diacylglycerol. Subsequently, the generated second messengers activate specific protein kinases (PKA, PKC and calmodulin-dependent kinases), which induce phosphorylation of specific cytoplasmatic and membrane-bound substrate proteins and thus activation of the substrate proteins leading to opening of anionic channels in the apical enterocyte membrane. The secretion of anions results in a lumen negative electrochemical gradient, leading to back-diffusion of positively charged Na+ and secretion of fluid. Cholera Substantial evidence has now accumulated that CT may cause diarrhoea by stimulating PG synthesis, in addition to involving 5hydroxytryptamine (5-HT) in the pathogenesis of CT-induced fluid and electrolyte secretion (Cassuto et al 1982; Nielsson et al 1983; Beubler et al 1986, 1989; Turvill et al 1998). This hypothesis is supported by the observations that acute cholera in humans is associated with increased mucosal PG synthesis (Bedwani and Okpako 1975; Speelman et al 1985) and that non-steroidal antiinflammatory drugs (NSAIDs) impair the secretory effect of CT (Wald et al 1997; Van Loon et al 1992). As indomethacin has been shown to inhibit CT-induced secretion in the absence of mucosal cAMP accumulation (Wald et al 1977; Beubler et al
THE INTESTINE 1986), PGs may play a primary role in the secretory mechanism. It is possible, therefore, that PGs stimulate cAMP production only at supraphysiological concentrations of PGs, through activation of membrane-associated adenylate cyclase (Bukhave and RaskMadsen 1980), and that the PG-induced stimulation of adenylate cyclase observed in homogenates of rodent intestinal cells (Kimberg et al 1971; Giganella et al 1978; Smith et al 1987) occurs due to the experimental conditions in vitro. PGE2 has also been shown to be an important intermediate in the transduction mechanism that leads to 5-HT-induced intestinal secretion (Beubler et al 1986). Therefore, CT may stimulate an apical receptor on the enterochromaffin cell by initiating release of 5-HT, which in turn causes formation of PGs to mediate the diarrhoeagenic action of CT. Recently, it was demonstrated in a rat model by use of selective COX-1 and COX-2 inhibitors, respectively, and by determining mRNA for COX-1 and COX-2, in addition to COX-2 protein levels in jejunal mucosa (Beubler et al 2001), that COX-1 is not involved and that COX-2 only is responsible for the profuse fluid secretion induced by CT. Other Secretory Diarrhoeas In a variety of clinical diarrhoeal conditions, besides cholera, PGs are considered to be responsible for inducing the diarrhoea (Table 38.1). The intestinal secretion observed in acute infectious diarrhoea may be considered an important protective host response, which causes dilution of toxins and other noxious agents, at the same time ensuring rapid emptying of the luminal contents due to the influence of PGs on intestinal motility. Invasive bacteria give rise to far greater PG production than toxin-producing, non-invasive bacteria (Eckmann et al 1997), in addition to a range of chemoattractants and proinflammatory cytokines, and COX inhibitors have no beneficial effects on the associated diarrhoea. In contrast, NSAIDs such as aspirin, indomethacin and piroxicam reduce secretion and promote intestinal absorption in toxin-mediated diarrhoea by inhibiting PG biosynthesis (Ludered et al 1980; Smith et al 1981; Van Loon et al 1992). Conversely, the addition of exogenous PGs prevents the antisecretory effects of COX inhibitors observed in vitro (Powell et al 1979; Smith et al 1981). Interestingly, loperamide apparently inhibits TXA2-induced secretion and reduces the diarrhoea observed in response to the antitumour drug, irinotecan, by blocking the calmodulin system in the colonic epithelium (Suzuki et al 2000). Nevertheless, both NSAIDs and loperamide are insufficient for treating high-purgent diarrhoea, which may be explained by an increase in prostanoid synthesis so extreme that even a 50–70% inhibition of the COX enzyme, obtained with clinically relevant doses, would permit local concentrations sufficiently high to cause a near-maximal secretory response (Bukhave and Rask-Madsen 1980; Rask-Madsen et al 1984; Rask-Madsen 1987). Epithelial and Endothelial Barrier Function PGs protect the intestinal mucosa from injury by noxious agents. This phenomenon is commonly referred to as ‘‘cytoprotection’’. Most studies have focused on the protective effects of exogenous PGs on the gastric mucosa, but small intestinal injury can also be prevented (Robert 1981). The term ‘‘adaptive cytoprotection’’ refers to the enhanced mucosal resistance of the mucosa observed following initial challenge with mild irritants (e.g. acid or ethanol). In the rat duodenum, the effect is associated with increased luminal release of PGE2 and bicarbonate, which is suppressed by indomethacin (Lugea et al 1992). In the proximal
425
human duodenum, COX inhibition by indomethacin decreases basal as well as stimulated secretion of bicarbonate (MertzNielsen et al 1995). Initially, the association between reduced duodenal mucosal bicarbonate secretion and ulceration was demonstrated in patients with duodenal ulcer, who also showed a ratio of bicarbonate secreted to the amount of PGE2 released into the lumen that was far less than in controls (Bukhave et al 1990). PGs also have a trophic effect on the intestinal mucosa and have been demonstrated to stimulate cellular proliferation (Uribe et al 1992). Thus, NSAIDs decrease small intestinal villus height, crypt depth and mitotic index in rats, while the proliferative response to intestinal resection is impaired in animals deficient in essential fatty acids. A major function of the epithelial layer is to act as a physical barrier between the outside environment and the organism. PGs appear to play an important role in maturation of this barrier (Mohajer and Ma 2000) and, in developing rats at least, maturation of the epithelial barrier function correlates with an age-related increase in PGE2 activity (Gerstle et al 1994). In contrast, intestinal ischaemia of the porcine ileum causes a drop in electrical resistance and increases epithelial permeability, which recovers within a few hours in the presence of PGs but not if PG synthesis is inhibited by indomethacin (Bilkslager et al 1997). Also, cAMP and the Ca2+ ionophore A23187 were capable of restoring epithelial resistance, suggesting that PG restoration of injured epithelial tight junction barrier function was mediated, at least in part, by upregulation of intracellular cAMP and Ca2+. In man, oral intake of aspirin and NSAIDs has consistently been shown to increase intestinal permeability characteristics (Bjarnason et al 1986, 1991), an effect which is prevented by coadministration of a PG analogue such as misoprostol (Bjarnason et al 1989). Similarly important to the functional integrity of the epithelium is the endothelial barrier, which is enhanced by PGs and disrupted by NSAIDs. In NSAID-associated enteropathy, which is postulated to be due to disruption of the combined epithelial and endothelial barrier, intestinal ulcerations and their complications, i.e. bleeding, perforation and strictures, are observed in a large proportion of patients on NSAID therapy (Allison et al 1992). Motor Activity Another mechanism of producing diarrhoea is through stimulation of small-intestinal motility. The effects of PGs on smooth muscle contraction in the colon, where both PGE2 and PGF2a stimulate contraction and motor activity (Burakoff et al 1990), are much greater than in the small intestine. The mechanism of stimulating small-intestinal motility is not entirely clear, but seems to depend on the specific type of PGs involved (Tollstro¨m et al 1988). Thus, PGE2 administered intraduodenally causes a delay in the initiation of the migrating motor complex and prevents its propagation, if already initiated, while the administration of PGF2a results in a burst of small-intestinal contractions with increased amplitude (Tollstro¨m et al 1988). Furthermore, PGE2 induces contraction of the longitudinal smooth muscle layer, while PGF2a initiates contraction of both longitudinal and circular smooth muscle layer (Bennett and Fleshler 1970; Waller 1973). Nevertheless, the diarrhoeagenic effects of PGs are mainly related to their action on intestinal secretion and less dependent on changes in motor activity. Inflammatory Response The role of LTs in the small intestine and the large bowel has recently been extensively reviewed (Mohajer and Ma 2000; Krause
426
DIGESTIVE SYSTEM
and DuBois 2000; Rask-Madsen 2001). Since their discovery in 1979 (Murphy et al 1979), LTs have been shown to increase neutrophil migration and degranulation, monocyte aggregation, adhesion of leukocytes to endothelial cells, release of lysosomal enzymes, superoxide production, capillary permeability, smooth muscle contraction and secretion of ions and mucus (FordHutchinson 1995; Nielsen and Rask-Madsen 1996; Krause and DuBois 2000; Mohajer and Ma 2000). LTs are also considered to be potent mediators of inflammatory and allergic reactions. They are released from leukocytes and other 5-lipoxygenase-expressing cells and exert their proinflammatory actions through binding to specific membrane receptors and, as suggested recently, the nuclear receptor peroxisome, proliferator-activated receptor-a (PPARa). For example, LTB4 is considered to be an endogenous ligand for PPARa, while the cyclopentenone 15D-12,14-PGJ2, which inhibits the activity of inhibitory-kB kinase, acts as an endogenous ligand for PPARg (Ricote et al 1998). Profiles of Eicosanoids in IBD The primary forms of idiopathic intestinal inflammation, ulcerative colitis and Crohn’s disease, are commonly denoted chronic inflammatory bowel disease (IBD), which is a chronic relapsing inflammation of the bowel causing significant morbidity. The aetiology remains contentious, but the pathophysiology is considered to be the consequence of an inappropriate overexpression of immune responses to antigens (bacterial products) normally present in the gut lumen. LTs play a significant role in IBD due to their amplification of the inflammatory response, which causes extensive tissue destruction in these patients (Nielsen and Rask-Madsen 1996; Krause and DuBois 2000). A link between eicosanoids and IBD was initially demonstrated by the discovery that colonic mucosa from patients with active ulcerative colitis contained high levels of PGs, whose synthesis was inhibited by the use of sulphasalazine (Sharon and Stenson 1978). Subsequently, TXs, PGI2 and LTB4 were found to be elevated in both ulcerative colitis and Crohn’s disease (Ligumsky et al 1981; Sharon and Stenson 1984). Profiles of eicosanoids appeared distinct in each type of bowel inflammation, although the levels were uniformly high in inflammation compared with those present in normal mucosa. Thus, concentrations of PGE2, TXB2 and LTB4 turned out to be markedly elevated in ulcerative colitis compared with those of Crohn’s colitis and Clostridium difficileinduced colitis (Lauritsen et al 1988b). While IBD is associated with excess eicosanoid formation in the target tissue of inflammation, COX inhibitors appear to worsen IBD, provoke a relapse, or even induce an ulcerative disease of the colon indistinguishable from ulcerative colitis (Lauritsen et al 1989; Eberhart and DuBois 1995; Nielsen and Rask-Madsen 1996; Krause and DuBois 2000; Mohajer and Ma 2000). The explanation for this was unclear until studies were carried out examining the inhibition of the 5-lipoxygenase pathway by 5aminosalicylic acid and, more importantly, the markedly increased production of LTB4 in IBD (Lauritsen et al 1984, 1986, 1987). In patients with active IBD, the large bowel was shown to produce LTs in vivo, as assessed by rectal equilibrium dialysis, and a positive correlation between concentrations of LTB4 and disease activity, as judged by clinical, endoscopical and histological gradings, was demonstrated (Lauritsen et al 1987). Furthermore, current therapies for ulcerative colitis, such as glucocorticoids and 5-aminosalicylic acid, were shown to produce significant inhibition of LTB4 synthesis (Lauritsen et al 1984, 1986, 1987, 1988a; Dreyling et al 1987). The extent to which increased levels were a primary event or the result of synthesis in neutrophils once recruited is still not clear, but functional studies suggest that LTB4 accounts for the majority of the chemoattractant stimulus
to continuing neutrophil infiltration observed in relapsing disease (Sharon et al 1984).
Targeting LTs Because of the proinflammatory profile of LTs, it was assumed that LT synthesis inhibitors and LT receptor antagonists might have a therapeutic potential in a variety of inflammatory diseases (Rask-Madsen et al 1992). The unique role of the enzyme 5lipoxygenase in the production of LTs made it a logical target for attempts to manipulate this pathway of AA metabolism in patients with IBD. The identification of LTs at the site of inflammation and the characterization of their potent proinflammatory actions resulted in the development of a number of antiLT drugs for the treatment of inflammatory conditions such as asthma, rheumatoid arthritis, multiple sclerosis, uveitis, gout, psoriasis and IBD. Since then, clinical studies have confirmed the therapeutic value of anti-LT therapy in asthma, but the results with LT biosynthesis inhibitors in psoriasis, arthritis and IBD were more or less disappointing, while specific LT-receptor antagonists have not yet been evaluated in IBD.
Pharmacodynamic Studies with Anti-LT Drugs The simple technique of rectal dialysis had shown that markedly increased LTB4 levels are reduced by selective 5-LO inhibitors, such as zileuton (Laursen et al 1990), FPL 64170XX (Kjeldsen et al 1995) or antagonists of 5-lipoxygenase-activating protein, such as MK-591 (Hillingsø et al 1995), glucocorticoids (Lauritsen et al 1986, 1987), 5-aminosalicylic acid (Lauritsen et al 1986, 1988a) or fish oil. Thus, a single oral dose of zileuton reduced the local release of LTB4 in rectal dialysates by approximately 70% (Laursen et al 1990), while oral MK-591 caused nearly complete inhibition of LTB4 in rectal dialysates from patients with distally located ulcerative colitis (Hillingsø et al 1995).
Clinical Experience with Anti-LT Drugs Reliable clinical data from randomised controlled trials on 5lipoxygenase inhibition are available only for zileuton and MK0591, both of which were withdrawn from clinical development due to lack of clinical efficacy in inducing and maintaining remission in patients with ulcerative colitis. The first randomized trial to test the efficacy of zileuton was carried out in 76 patients with relapsing UC (Laursen et al 1994). The study showed only a trend toward clinical benefit, although LTB4 concentrations in rectal dialysates fell significantly during the 4-week period. The disappointing clinical efficacy of zileuton was explained by the presence of endogenous LTs, still present in sufficient amounts to produce their effects. Another randomized trial of zileuton was carried out in 212 patients with active UC, but only published in abstract form (Peppercorn et al 1994). The clinical response and remission rates with zileuton 600 mg q.i.d. were statistically better than with placebo or zileuton 800 mg b.i.d. Later, a randomized clinical trial was performed comparing the ability of zileuton vs. mesalazine or placebo to prevent relapse in 305 patients with proven UC (Hawkey et al 1997). The relapse rates on mesalazine (37%) were significantly lower than those for placebo (57%) after 6 months, but a trend toward lower relapse rates in patients having received zileuton (46%) did not reach conventional levels of statistical significance. Consequently, the relevance of targeting LTB4 was questioned.
THE INTESTINE To evaluate whether more potent inhibition of LT biosynthesis would improve the marginal benefits obtained with zileuton, a randomized trial with MK-591 was carried out in 183 patients with active UC (Roberts et al 1997). Subsets of patients underwent rectal dialysis for determination of LTB4 concentrations. The median inhibition of baseline LTB4 levels varied from 88% to 98.6% for MK-591 and was 12% for placebo, but there was no correlation between reduction of LTB4 concentrations and induction of remission, and the outcome of active treatment did not differ significantly from that of placebo, in spite of marked inhibition of LT biosynthesis. The placebo response observed in this study was higher than those previously reported for ulcerative colitis, but even if remission rates in response to MK-591 were compared with placebo response rates observed in other studies, the therapeutic gain would be too small to be clinically relevant.
CONCLUSION PGs are important physiological modulators of intestinal epithelial functions (absorption/secretion) and intestinal motor activity, in addition to being crucial for the maintenance of epithelial and endothelial barrier integrity in health. PGs seem to play a primary role in the epithelial secretory mechanism, which is independent of cAMP accumulation, and inhibitors of COX have been shown to impair the secretory effect of CT in human cholera. The diarrhoeagenic effects of PGs seem to be related mainly to their stimulatory action on intestinal secretion and are less dependent on changes in motor activity. While PGs play no direct role in inducing bowel inflammation, IBD is associated with excess formation of LTs, which account for the majority of the chemoattractant stimulus to the continuing neutrophil infiltration observed in relapsing disease. Inhibitors of COX appear to worsen or even provoke a relapse of IBD. Despite the high potency of 5lipoxygenase inhibitors, LT-antagonists seem unable to induce or maintain remission.
REFERENCES Abate A, Oberle S and Schro¨der H (1998) Lipopolysaccharide-induced expression of cyclooxygenase-2 in mouse macrophages is inhibited by chloromethylketones and a direct inhibitor of NF-kB translocation. Prostagland Lipid Mediat, 56, 277–290. Al-Awquati Q and Greenough WB (1972) Prostaglandins inhibit intestinal sodium transport. Nature, 238, 26–27. Allison MC, Howatsaon AG, Torrance CJ et al (1992) Gastrointestinal damage associated with the use of non-steroidal antiinflammatory drugs. N Engl J Med, 327, 749–752. Armstrong WM (1987) Cellular mechanisms of ion transport in the small intestine. In Johnson LR (ed.), Physiology of the Gastrointestinal Tract. New York: Raven, 1251–1265. Bedwani JR and Okpako DT (1975) Effects of crude and pure cholera toxin on prostaglangin. Prostaglandins, 10, 117–127. Belley A and Chadee K (1999) Prostaglandin E2 stimulates rat and human colonic mucin exocytosis via the EP4 receptor. Gastroenterology, 117, 1352–1362. Bennett A and Fleshler B (1970) Prostaglandins and the gastrointestinal tract. Gastroenterology, 59, 790–800. Beubler E, Bukhave K and Rask-Madsen J (1986) Significance of calcium for the prostaglandin E2-mediated secretory response to 5hydroxytryptamine in the small intestine of the rat in vivo. Gastroenterology, 90, 1972–1977. Beubler E, Kollar G, Saria A et al (1989) Involvement of 5hydroxytryptamine, prostaglandin E2, and cyclic adenosine monophosphate in cholera toxin-induced fluid secretion in the small intestine of the rat in vivo. Gastroenterology, 96, 368–376.
427
Beubler E, Schuligoi R, Chopa AK et al (2001) Cholera toxin induces prostaglandin synthesis via post-transcriptional activation of cyclooxygenase-2 in the rat jejunum. J Pharmacol Exp Therapeut, 297, 940–945. Bjarnason I, Fehilly B, Smethurst P et al (1991) Importance of local versus systemic effects of non-steroidal antiinflammatory drugs in increasing small intestinal permeability in man. Gut, 32, 275–277. Bjarnason I, Smethurst, Fenn CG et al (1989) Misoprostol reduces indomethacin-induced changes in human intestinal permeability. Dig Dis Sci, 34, 407–411. Bjarnason I, Williams P, Smethurst TJ et al (1986) Effect of non-steroidal antiinflammatory drugs and prostaglandins on permeability of the human small intestine. Gut, 27, 1292–1297. Blikslager AT, Roberts MC, Rhoads JM and Argenzio RA (1997) Prostaglandin I2 and E2 have a synergistic role in rescuing epithelial barrier function in porcine ileum. J Clin Invest, 100, 1928–1933. Bukhave K (1992) Prostaglandin E2 in intestinal secretion with special reference to methodological problems in vivo and in vitro. Danish Med Bull, 39, 95–103. Bukhave K, Rask-Madsen J, Hogan DL et al (1990) Proximal duodenal prostaglandin E2 release and mucosal bicarbonate secretion are altered in duodenal ulcer patients. Gastroenterology, 99, 951–955. Bukhave K and Rask-Madsen J (1980) Saturation kinetics applied to in vitro effects of low prostaglandin E2 and F2a concentrations on ion transport across human jejunal mucosa. Gastroenterology, 78, 32–42. Burakoff R, Nastos E and Won S (1990) Effects of PGF2a and of indomethacin on rabbit small and large intestinal motility in vivo. Am J Physiol, 258, G231–G237. Brunsson I, Sjo¨qvist A, Jodal M and Lundgren O (1987) Mechanisms underlying the small intestinal fluid secretion caused by arachidonic acid, prostaglandin E1 and E2 in the rat in vivo. Acta Physiol Scand, 130, 633–642. Cassuto J, Jodal M, Tuttle R and Lundgren O (1982) 5Hydroxytryptamine and cholera secretion. Physiological and pharmacological studies in cats and rats. Scand J Gastroenterol, 17, 695–703. Davies GR and Rampton DS (1997) Eicosanoids: role in gastrointestinal inflammation and cancer. Eur J Gastroenterol Hepatol, 9, 1033–1044. Diener M, Bridges RJ, Knobloch SF and Rummel W (1988) Neuronally mediated and direct effects of prostaglandins on ion transport in rat colon descendens. Naunyn-Schmiedebergs Arch Pharmacol, 337, 69– 73. Dreyling KW, Kozuschek W et al (1987) Role of the endogenous leukotriene system in chronic inflammatory bowel disease. Progr Clin Biol Res, 242, 309–315. Eberhart CE and DuBois RN (1995) Eicosanoids and the gastrointestinal tract. Gastroenterology, 109, 285–301. Eckmann L, Stenson WF, Savidge TC et al (1997) Role of intestinal epithelial cells in the host secretory response to infection by invasive bacteria. Bacterial entry induces epithelial prostaglandin H synthase expression and prostaglandin E2 and prostaglandin F2a production. J Clin Invest, 100, 296–309. Ford-Hutchinson AW (1995) Leukotriene antagonists and inhibitors: clinical applications. Adv Prostagland Thromboxane Leukotriene Res, 23, 69–74. Gerstle JT, Seaton J, Kauffman GL and Colony PC (1994) The association between PGE2 activity and mucosal permeability in the proximal small bowel. J Surg Res, 57, 579–583. Giganella TS, Phillips SF, Dozois RR and Go VLW (1978) Stimulation of adenyl cyclase in homogenates of isolated intestinal epithelial cells from hamsters. Gastroenterology, 74, 11–15. Halm DR and Halm ST (2001) Prostanoids stimulate K secretion and Cl secretion in guinea-pig distal colon via distinct pathways. Am J Physiol Gastrointest Liver Physiol, 281, G984–G996. Hawkey CJ, Dube LM, Rountree LV et al and The European Zileuton Study Group for Ulcerative Colitis (1997) A trial of zileuton vs. mesalamine or placebo in the maintenance of remission of ulcerative colitis. Gastroenterology, 112, 718–724. Henderson WR (1994) The role of leukotrienes in inflammation. Ann Intern Med, 121, 684–697. Hillingsø J, Kjeldsen J, Laursen LS et al (1995) Blockade of leukotriene production by a single oral dose of MK-0591 in active ulcerative colitis. Clin Pharmacol Therapeut, 57, 335–341.
428
DIGESTIVE SYSTEM
Keast JR (1987) Mucosal innervation and control of water and ion transport in the intestine. Rev Physiol Biochem Pharmacol, 109, 2–37. Kimberg DV, Field M, Johnson J et al (1971) Stimulation of intestinal adenylcyclase by cholera enterotoxin and prostaglandins. J Clin Invest, 50, 1218–1230. Kjeldsen J, Laursen LS, Hillingsø J et al (1995) Selective blockade of leukotriene production by a single dose of the FPL 64170XX 0.5% enema in active ulcerative colitis. Pharmacol Toxicol, 77, 371–376. Krause W and DuBois RN (2000) Eicosanoids and the large intestine. Prostagland Lipid Mediat, 61, 145–161. Lauritsen K, Hansen J, Bytzer P et al (1984) Effects of sulphasalazine and disodium azodisalicylate on colonic PGE2 concentrations determined by equilibrium in vivo dialysis of faeces in patients with ulcerative colitis and healthy controls. Gut, 25, 1271–1278. Lauritsen K, Laursen LS, Bukhave K and Rask-Madsen J (1986) Effects of topical 5-aminosalicylic acid and prednisolone on prostaglandin E2 and leukotriene B4 levels determined by equilibrium in vivo dialysis of rectum in relapsing ulcerative colitis. Gastroenterology, 91, 837–844. Lauritsen K, Laursen LS, Bukhave K and Rask-Madsen J (1987) In vivo effects of orally administered prednisolone on prostaglandin and leucotriene production in ulcerative colitis. Gut, 28, 1095–1099. Lauritsen K, Laursen LS, Bukhave K and Rask-Madsen J (1988a) Use of colonic eicosanoid levels as predictors of relapse in ulcerative colitis: double-blind placebo-controlled study on sulphasalazine maintenance treatment. Gut, 29, 1316–1321. Lauritsen K, Laursen LS, Bukhave K and Rask-Madsen J (1988b) In vivo profiles of eicosanoids in ulcerative colitis, Crohn’s colitis, and Clostridium difficile colitis. Gastroenterology, 95, 11–17. Lauritsen K, Laursen LS, Bukhave K and Rask-Madsen J (1989) Inflammatory intermediaries in inflammatory bowel disease. Int J Colorect Dis, 4, 75–90. Laursen LS, Næsdal J, Bukhave K et al (1990) Selective 5-lipoxygenase inhibition in ulcerative colitis. Lancet, 335, 683–685. Laursen LS, Lauritsen K, Bukhave K et al (1994) Selective 5-lipoxygenase inhibition by zileuton in the treatment of relapsing ulcerative colitis: a randomized double-blind placebo-controlled multicentre trial. Eur J Gastroenterol Hepatol, 6, 209–215. Ligumsky M, Karmeli F, Sharon P et al (1981) Enhanced thromboxane A2 and prostacyclin production by cultured rectal mucosa in ulcerative colitis and its inhibition by steroids and sulfasalazine. Gastroenterology, 81, 444–449. Ludered JR, Demers LM, Nomides CT and Hayes AH (1980) Mechanism of action of castor oil: a biochemical link to the prostaglandins. Adv Prostagland Thromboxane Res, 8, 1633–1635. Lugea A, Salas A, Guarner F et al (1992) Duodenal mucosal resistance to intraluminal acid in the rat: role of adaptive cytoprotection. Gastroenterology, 102, 1129–1135. Maclouf J, Folco G and Patrono C (1998) Eicosanoids and isoeicosanoids: constitutive, inducible and transcellular biosynthesis in vascular disease. Thromb Haemostas, 79, 691–705. Mertz-Nielsen A, Hillingsø J, Bukhave K and Rask-Madsen J (1995) Indomethacin decreases gastroduodenal mucosal bicarbonate secretion in humans. Scand J Gastroenterol, 30, 1160–1165. Milton-Thomson GJ, Cummings JH, Newman A et al (1975) Colonic and small intestinal response to i.v. prostaglandins F2a and E2 in man. Gut, 16, 42–46. Mohajer B and Ma TY (2000) Eicosanoids and the small intestine. Prostagland Lipid Mediat, 61, 125–143. Murphy RC, Hammerstro¨m S and Samuelsson B (1979) Leukotriene C: a slow-reacting substance (SRS) from murine mastocytoma cells. Proc Natl Acad Sci USA, 76, 4274–4279. Nielsen OH and Rask-Madsen J (1996) Mediators of inflammation in chronic inflammatory bowel disease. Scand J Gastroenterol, 31(suppl 216), 149–159. Nilsson O, Cassuto J, Larsson PA et al (1983) 5-Hydroxytryptamine and cholera secretion: a histochemical and physiological study in cats. Gut, 24, 542–548. O’Neill GP and Ford-Hutchinson AW (1993) Expression of mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human tissues. FEBS Lett, 330, 156–160. Peppercorn M, Das K, Elson C et al (1994) Zileuton, a 5-lipoxygenase inhibitor, in the treatment of active ulcerative colitis: a double-blind, placebo controlled trial. Gastroenterology, 106, A751.
Powell DW (1986) Ion and water transport in the intestine. In Andreoli TE, Hoffman FG, Fanestil DD and Schultz SG (eds), Physiology of Membrane Disorders. New York: Plenum, 559–596. Powell DW, Berschneider HM, Lawson LD and Martens H (1985) Regulation of water and ion movements in intestine. CIBA Found Sympos, 112, 14–34. Powell DW, Tapper AJ and Morris SM (1979) Aspirin-stimulated intestinal electrolyte transport in rabbit ileum in vitro. Gastroenterology, 76, 1429–1437. Powell DW, Mifflin RC, Valentich JD et al (1999) Myofibroblasts. II. Intestinal subepithelial myofibroblasts. Am J Physiol, 277, C183–C201. Racusen LC and Binder HJ (1980) Effect of prostaglandins on ion transport across isolated colonic mucosa. Dig Dis Sci, 25, 900–904. Rask-Madsen J (1986) Eicosanoids and their role in the pathogenesis of diarrhoeal diseases. Clin Gastroenterol, 15, 545–566. Rask-Madsen J (1987) The role of eicosanoids in the gastrointestinal tract. Scand J Gastroenterol, 127(suppl), 7–19. Rask-Madsen J (2001) Leukotrienes as targets in the gut. Clin Exp Allergy Rev, 1, 309–312. Rask-Madsen J, Bukhave K and Beubler E (1990) Influence on intestinal secretion of eicosanoids. J Intern Med, 228(suppl 1), 137–144. Rask-Madsen J, Bukhave K, Bytzer P and Lauritsen K (1984) Prostaglandins in the gastrointestinal tract. Acta Med Scand, (suppl 685), 30–46. Rask-Madsen J, Bukhave K, Laursen LS and Lauritsen K (1992) 5Lipoxygenase inhibitors for the treatment of inflammatory bowel disease. Agents Actions, Special Conference Issue, C37–C46. Ribardo DA, Crowe SE, Kuhl KR et al (2001) Prostaglandin levels in stimulated macrophages are controlled by phospholipase A2-activating protein and by activation of phospholipase C and D. J Biol Chem, 276, 5467–5475. Ricote M, Li AC, Wilson TM et al (1998) The peroxisome proliferatoractivated receptor-g is a negative regulator of macrophage activation. Nature, 391, 79–82. Robert A (1981) Prostaglandins and the gastrointestinal tract. In Johnson LR (ed.), Physiology of the Gastrointestinal Tract. New York: Raven, 1407–1431. Roberts WG, Simon TJ, Berlin RG et al (1997) Leukotrienes in ulcerative colitis: results of a multicenter trial of a leukotriene biosynthesis inhibitor, MK-591. Gastroenterology, 112, 725–732. Rouzer CA, Matsumoto T and Samuelsson B (1986) Single protein from human leukocytes processes 5-lipoxygenase and leukotriene A4 synthase activities. Proc Natl Acad Sci USA, 83, 857–861. Ruwart MJ, Klepper MS and Rush BD (1979) Ionic composition of small intestinal secretion induced by PGE2. Prostagland Med, 2, 285–291. Serafin WE and Austen KF (1987) Mediators of immediate hypersensitivity reactions. N Engl J Med, 317, 30–34. Sharon P, Ligumsky M, Rachmilewitz D and Zor U (1978) Role of prostaglandins in ulcerative colitis, enhanced production during active disease and inhibition by sulfasalazine. Gastroenterology, 75, 638–640. Sharon P and Stenson WF (1984) Enhanced synthesis of leukotriene B4 by colonic mucosa in inflammatory bowel disease. Gastroenterology, 86, 453–460. Smith G, Warhurst G, Lees M and Turnberg L (1987) Evidence that PGE2 stimulates intestinal epithelial cell adenylate cyclase by a receptormediated mechanism. Dig Dis Sci, 32, 71–75. Smith PI, Blumberg JB, Stoff JL and Field M (1981) Antisecretory effects of indomethacin on rabbit ileal mucosa in vitro. Gastroenterology, 80, 356–365. Smith WL, Sewitt DL and Garavito RM (2000) Cyclooxygenases: structural, cellular, and molecular biology. Ann Rev Biochem, 69, 145–182. Speelman P, Rabbani GH, Bukhave K and Rask-Madsen J (1985) Increased jejunal PGE2 concentrations in patients with acute cholera. Gut, 26, 188–193. Suzuki T, Sakai H, Ikari A and Takeguchi N (2000) Inhibition of thromboxane A2-induced Cl7 secretion by antidiarrhoea drug loperamide in isolated rat colon. J Pharmacol Exp Therapeut, 295, 233–238. Tollstro¨m T, Hellstro¨m PM, Johansson C and Pernow B (1988) Effects of prostaglandin E2 and F2a on motility of small intestine in man. Dig Dis Sci, 33, 552–557.
THE INTESTINE Turvill JL, Mourad FH and Farthing MJ (1998) Crucial role for 5-HT in cholera toxin but not Escherichia coli heat-labile enterotoxin-intestinal secretion in rats. Gastroenterology, 115, 1009–1012. Uribe A, Alam M and Midtvedt T (1992) E2 prostaglandins modulate cell proliferation in the small intestinal epithelium of the rat. Digestion, 52, 157–164. Van Loon FPL, Rabbani GH, Bukhave K and Rask-Madsen J (1992) Indomethacin decreases jejunal fluid secretion, in addition to luminal
429
release of prostaglandin E2 in patients with acute cholera. Gut, 33, 643– 645. Wald A, Gotterer GS, Rajendra GR et al (1977) Effect of indomethacin on cholera-induced fluid movements, unidirectional sodium fluxes, and intestinal cAMP. Gastroenterology, 72, 106–110. Waller SL (1973) Prostaglandins and the gastrointestinal tract. Gut, 14, 402–417.
39 Eicosanoids and Stomach Physiology Brigitta M. Peskar Ruhr-University of Bochum, Bochum, Germany
BIOSYNTHESIS OF EICOSANOIDS IN THE STOMACH Formation of Prostanoids in Gastric Tissues Gastric mucosal and muscular tissues of various species, including rat, guinea-pig, rabbit, cat, dog, monkey and man, generate prostanoids from endogenous and exogenous substrate. The amount and pattern of prostanoid production varies between species. Fragments of human gastric mucosa release prostaglandin (PG) E2, PGD2, 6-keto-PGF1a and PGF2a during incubation at 378C, with PGE2 being the most abundant arachidonic acid derivative formed (Peskar et al 1984). In contrast, gastric mucosal tissues from rat, rabbit and cat synthesize more PGI2 than PGE2 (Gretzer et al 1998; Maricic et al 1999; Whittle and Vane 1987). Gastric tissues also generate relevant amounts of thromboxane (TX) B2 (Peskar et al 1984), which may, however, be derived from platelets or inflammatory cells trapped within the gastric vasculature. The prostaglandin-inactivating enzymes 15-hydroxy-PG-dehydrogenase and PG-D13-reductase have been demonstrated in the cytosolic fraction of human gastric mucosa (Peskar and Peskar 1976). Human gastric mucosal fragments release 15-keto-13,14dihydro-PGE2, 15-keto-13,14-dihydro-PGF2a and 6,15-diketo13,14-dihydro-PGF1a, the metabolites generated from PGE2, PGF2a and PGI2 by these enzymes, during incubation in vitro (Peskar et al 1984). Furthermore, high levels of 15-keto-13,14dihydro-PGE2 were generated by minced rat stomach after incubation with a hypertonic sucrose solution, as assessed by GC– MS techniques (Knapp et al 1978). Prostanoids in Gastric Juice Release of PGE2 into the gastric lumen was shown in rats (Shaw and Ramwell 1968), dogs (Dozois and Thompson 1974; RaskMadsen et al 1981), cats (Baker et al 1978) and humans (Baker et al 1979; Peskar et al 1980; Rask-Madsen et al 1981). In human gastric juice, concentrations of 6-keto-PGF1a were considerably lower than concentrations of PGE2 (Peskar et al 1980). In addition, 15-keto-13,14-dihydro-PGE2 was released into the lumen of the human stomach in amounts larger than those of unmetabolized PGE2 indicating rapid local metabolism of prostaglandins. Pentagastrin did not increase the concentrations of prostaglandins and prostaglandin metabolites in gastric juice but increased prostaglandin output parallel to the increase in juice volume (Peskar et al 1980).
noic acid (12-HETE) was shown to occur in human gastric tissue (Bennett et al 1981) and 12-HETE was generated from [14C] arachidonic acid in homogenates of canine (Ahlquist et al 1982) and guinea-pig (Ahlquist et al 1983) gastric mucosa. In the rat stomach, arachidonic acid was found to be metabolized via the 5lipoxygenase pathway to yield leukotriene (LT) C4 and LTB4. Under basal conditions, gastric mucosal formation of LTC4 was 5% that of 6-keto-PGF1a (Peskar et al 1986). Dispersed rat gastric mucosal cells separated by counterflow elutriation with respect to size synthesized LTC4 and smaller amounts of LTB4. The maximum release of leukotrienes was found in small, chromogranin A-positive, probably neuroendocrine cells (Huber et al 1993). Human gastric mucosal and muscular tissues released LTB4 and sulphidopeptide leukotrienes (consisting of a mixture of LTC4, LTD4 and LTE4) during incubation in vitro (Dreyling et al 1986). Immunohistochemical studies localized LTB4 in the cytoplasm of human parietal surface epithelial and vascular endothelial cells (Higuchi et al 1992). In children with Helicobacter pylori colonization, LTB4, LTC4 and LTE4 were determined in gastric juice and levels were significantly higher than in children without H. pylori infection (Kasirga et al 1999). In rats, intragastric instillation of ethanol elicited a dosedependent and marked increase in mucosal LTC4 formation without affecting release of cyclooxygenase products (Peskar et al, 1986). However, the selective inhibitor of 5-lipoxygenase MK-886 did not prevent ethanol-induced gastric injury, indicating that leukotrienes are not essential mediators of ethanol-induced damage (Peskar 1991). This is in line with the observation that various gastroprotective compounds, but also certain nonsteroidal antiinflammatory drugs (NSAIDs) without gastroprotective activity, inhibit the ethanol-stimulated release of LTC4 and 15-hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE) in rat gastric mucosa (Peskar 1991; Trautmann et al 1991). Cigarette smoke elevated rat gastric mucosal LTB4 concentrations parallel to aggravation of ethanol damage. This effect was suggested to contribute to neutrophil recruitment into the tissue. However, whether inhibition of LTB4 formation attenuates the injurious effect of cigarette smoke has not been studied (Chow et al 1998).
EFFECTS OF EICOSANOIDS ON GASTRIC SECRETION Prostaglandins and Acid Secretion
Formation of Lipoxygenase Products in Gastric Tissues Various lipoxygenase products have been identified in gastric tissues. Thus, using GC–MS techniques 12-hydroxyeicosatetraeThe Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
Robert et al (1967) were the first to demonstrate that prostaglandins of the E series inhibit acid secretion in the dog. The antisecretory effect of prostaglandins was confirmed in numerous
432
DIGESTIVE SYSTEM
studies. Gastric output of acid, fluid volume and pepsin induced by secretagogues such as histamine, pentagastrin or vagal stimulation as well as feeding was inhibited in dogs, rats, cats and monkeys by PGE2 and/or PGE1 administered intravenously, intraarterially and/or subcutaneously (Banerjee et al 1972; Dajani et al 1975, 1976; Gerkens et al 1978; Main and Whittle 1973; Nezamis et al 1971; Robert et al 1968). The antisecretory potency of PGE1 and PGE2 was significantly diminished after oral administration (Robert et al 1976). In humans, intravenous infusion of PGE1 and PGE2 reduced basal and pentagastrinstimulated acid output (Classen et al 1971; Newman et al 1975). Oral PGE2 inhibited acid secretion induced by liquid protein or following sham-feeding (Befrits and Johansson 1985; Reele and Bohan 1984), but not pentagastrin-stimulated gastric secretion in man (Horton et al 1968). Methylated PGE analogues such as 16,16-dimethyl-PGE2 administered topically showed potent and long-lasting antisecretory actions in vivo in several species including humans (Carter et al 1973; Robert et al 1976). Oral administration of methylated PGE derivatives inhibited gastric secretion also in duodenal ulcer patients during meal stimulation and pentagastrin or histamine infusion (Carter et al 1973; Peterson et al 1979). Although several methylated derivatives of E-type prostaglandins with potent antisecretory action have been developed, only misoprostol (a 16-methyl,16-hydroxy analogue of PGE1 methyl ester) was introduced into clinical therapy for the treatment of NSAID-induced gastric damage (Hawkey 2000). Whereas PGF2a had no or inconsistent antisecretory activity after parenteral administration in various species including humans (Newman et al 1975; Robert et al 1967), 15-methylPGF2a inhibited acid output in the monkey (Shea-Donohue et al 1982) and rat (Whittle 1976). PGD2 had no antisecretory effect on canine isolated parietal cells (Skoglund et al 1980). PGI2 administered parenterally inhibited acid secretion in the rat (Whittle et al 1978a), dog (Gerkens et al 1978; Kauffman et al 1979) and monkey (Shea-Donohue et al 1982), whereas the PGI2 breakdown product 6-keto-PGF1a had only weak antisecretory activity in the rat (Whittle et al 1978b) or dog (Kauffman et al 1979). Several PGI2 analogues were developed to increase potency and selectivity of action. Some of these analogues, e.g. (5a)-5,9epoxy,16-phenoxy-PGF1a, potently inhibited gastric acid in the rat and dog (Whittle and Boughton-Smith 1979). Leukotrienes and Acid Secretion Variable effects on acid secretion were found with the leukotrienes. Neither LTB4 nor LTC4 affected histamine-stimulated acid secretion in isolated human parietal cells (Jaramillo et al 1989). Using enriched rat parietal cells, it was found that LTB4 had no effect on basal or histamine, forskolin or dibutyryl cAMPstimulated acid secretion, whereas LTC4 and LTD4 potentiated prestimulated, but did not affect basal acid, secretion (Schepp et al 1989). On the other hand, LTC4 was found to be a potent inhibitor of gastric acid secretion in the stomach of dogs stimulated by histamine, pentagastrin and meat-feeding and to effectively reduce acid formation in canine isolated gastric glands stimulated by histamine or dibutyryl cAMP (Konturek et al 1987).
Exposure of guinea-pig gastric chief cells to indomethacin inhibited PGE2 formation but increased LTB4 release. Simultaneously, pepsinogen secretion and intracellular calcium concentrations were increased. Pretreating the cells with a 5-lipoxygenase inhibitor abolished the LTB4 generation induced by indomethacin and reduced indomethacin-induced pepsinogen secretion, suggesting that the lipoxygenase shunt may contribute to the stimulatory action of indomethacin on isolated chief cells (Fiorucci et al 1995b). MECHANISM OF ANTISECRETORY EFFECTS Effects on Gastric Circulation Vasodilating agents that lower systemic arterial blood pressure could decrease gastric mucosal blood flow and, as a consequence, interfere with acid secretory processes. In rat or dog, prostaglandins of the E and I series and their analogues had variable effects on gastric mucosal blood flow that did not correlate with their acid inhibitory potency. The ratio of mucosal blood flow to acid output either did not change or was elevated (Whittle and Vane 1987), indicating that prostaglandin-induced inhibition of acid secretion is not the result of reduced mucosal blood flow. Furthermore, prostaglandins reduced acid secretion in enriched parietal cells and isolated fundic glands of rabbit (Levine et al 1982) and isolated canine parietal cells (Skoglund et al 1982; Soll and Whittle 1981), suggesting that they interfere directly with the acid secretory process. Finally, PGE2, PGI2 and stable PGI2 analogues reduced acid output in the in vitro luminally perfused stomach of rats and mice (Boughton-Smith and Whittle 1981). Effects on Adenylate Cyclase In canine gastric mucosal cells enriched in their content of parietal cells, PGE2 inhibited histamine-stimulated aminopyrine uptake used as an index of parietal cell response and simultaneously inhibited histamine-stimulated cAMP production, in a dosedependent fashion, with maximal inhibition at 1 mM PGE2. PGI2 also inhibited both histamine-stimulated aminopyrine accumulation and cAMP production (Figure 39.1). At higher concentrations, however, PGE2 and PGI2 stimulated cAMP production (Soll 1980). In isolated canine fundic gastric mucosal cells and isolated rat gastric cells, PGE2 activation of adenylate cyclase was negatively correlated with the parietal cell content (Schepp et al 1983; Wollin et al 1979). From studies in canine parietal cells treated with pertussis toxin, it was concluded that prostanoids inhibit parietal cell function by receptor-mediated interaction with the inhibitory guanine nucleotide-binding protein of adenylate cyclase (Chen et al 1988). The stimulatory effect of PGE2 on adenylate cyclase activity in non-parietal cell fractions could possibly underlie the stimulatory action of prostaglandins on gastric mucus production, as cAMP can fully reproduce the effect of PGE2 on the formation and secretion of gastric mucus in rats (Bersimbaev et al 1985). GASTRIC MUCOSAL BLOOD FLOW
Leukotrienes and Pepsinogen Secretion
Effects of Prostaglandins
LTB4, LTC4, LTD4 and LTE4 stimulated release of pepsinogen from guinea-pig gastric chief cells in a dose-dependent manner and the effect was antagonized by LTB4 and LTD4 antagonists, respectively. Stimulation of pepsinogen secretion by leukotrienes was associated with increased nitric oxide (NO) generation and was attenuated by inhibition of NO synthase (Fiorucci et al 1995a).
Under resting conditions, intravenous infusions of PGE1 or PGE2 elevated gastric mucosal blood flow in the rat. This effect occurred although systemic arterial blood pressure was decreased, indicating a direct vasodilator action on the gastric mucosa (Main and Whittle 1975). Similarly, gastric intraarterial infusion of PGE2 increased total gastric blood flow and fundic mucosal blood flow in the dog
STOMACH PHYSIOLOGY
433
Figure 39.1 (A) Inhibitory effect of PGE2 on histamine (H)-stimulated acid-secretion as assessed by aminopyrine accumulation in parietal cells prepared and enriched from canine fundic mucosa. **p50.005 vs. uninhibited response, n=9 for each preparation. (B) The effect of PGE2 on histamine (10 mM)stimulated cyclic AMP production. *p50.05, **p50.005 vs. histamine, n=5 for each preparation. Data derived from Soll (1980)
(Gerber and Nies 1982; Kauffman and Whittle 1982). Synthetic prostaglandin analogues such as 16,16-dimethyl-PGE2, however, reduced gastric mucosal blood flow during pentagastrin or histamine stimulation of acid secretion as a consequence of their pronounced antisecretory effect, whereas the ratio of blood flow to acid output remained unchanged (Main and Whittle 1975; Miller et al 1980). The vasodilating effect of low doses of PGE2 was confirmed by microscopic visualization of rat gastric submucosal arterioles (Guth and Moler 1982). Although topical mucosal application of 16,16-dimethyl-PGE2 decreased gastric mucosal blood flow under resting conditions, the prostaglandin analogue prevented the reduction in gastric mucosal blood flow induced by luminal application of 50–100% ethanol (Pihan et al 1986). Similar to prostaglandins of the E series, PGI2 at low doses elevated and at high doses reduced gastric mucosal blood flow in rat and dog, presumably reflecting secondary effects after inhibition of acid secretion flow (Kauffman et al 1979; Whittle et al 1978a). Local intraarterial administration of PGF2a elevated vascular resistance in the dog and rabbit stomach (Kauffman and Whittle 1982; Salvati and Whittle 1981) and topical PGF2a induced vasoconstriction preferentially in the submucosal venules of the rat gastric microcirculation, as shown by in vivo microscopy (Whittle et al 1985). The finding that inhibition of prostaglandin biosynthesis by indomethacin reduces gastric mucosal blood flow in the rat suggests that endogenous vasodilating prostaglandins are involved in the maintenance of gastric mucosal blood flow under resting conditions (Main and Whittle 1973). Effects of Arachidonic Acid and Thromboxanes The vascular effects of arachidonic acid vary with the experimental conditions used. In the dog, close arterial injection to the stomach of low doses of arachidonic acid induced gastric vasodilation, which was inhibited by indomethacin, presumably resulting from the formation of a vasodilator prostaglandin (Kauffman and Whittle 1982). However, high concentrations of
arachidonic acid infused intraarterially elicited vasoconstriction (Kauffman and Whittle 1982; Whittle et al 1981). Under these experimental conditions, arachidonic acid is transformed to TXA2 by circulating platelets, as assessed by identification of the breakdown product TXB2 and inhibition of the vasoconstrictor action by indomethacin. Conversion to vasoconstricting TXA2 in the canine stomach concurrently challenged intraluminally with low concentrations of acidified bile salts obviously underlies the substantial mucosal injury induced by arachidonic acid (Whittle et al 1981). Similar to the vasoconstrictor action of TXA2 biosynthesized from arachidonic acid, stable thromboxane mimetics were found to be potent gastric vasoconstrictors in the dog, rabbit and rat (Kauffman and Whittle 1982; Salvati and Whittle 1981). In the rat, in vivo microscopy of the gastric submucosal microcirculation showed that the thromboxane mimetic evoked constriction of arterioles and intense focal venular constriction (Whittle et al 1985). Effects of Leukotrienes The 5-lipoxygenase product LTC4 was found to be a potent vasoconstrictor in rat gastric mucosal arterioles and, even more pronounced, venules (Whittle et al 1985). Similar to other gastric vasoconstricting compounds, LTC4 and LTD4, although not ulcerogenic when given alone, significantly aggravated damage induced by intraluminal noxious agents (Konturek et al 1988; Pihan et al 1988; Wallace et al 1990a). In the isolated rat stomach perfused in situ via the vasculature, injections of platelet-activating factor significantly increased the release of sulphidopeptide leukotrienes and simultaneously elicited long-lasting reductions of flow rates. The lipoxygenase inhibitors nordihydroguaiaretic acid and L-651,896 and the sulphidopeptide leukotriene receptor antagonist FPL 55712 did not modify the flow reduction, indicating that endogenous sulphidopeptide leukotrienes are of minor importance for the effect of platelet-activating factor on gastric vascular flow (Dembinska-Kiec et al 1989).
434
DIGESTIVE SYSTEM
GASTROPROTECTIVE ACTION OF PROSTAGLANDINS Gastroprotection in Experimental Animals Prostaglandins of the E and I series and their synthetic analogues prevent the induction of gastric mucosal damage induced by a great number of ulcerogens, including ethanol, strong acid, strong base, hypertonic solution, bile acids, boiling water, steroids, restraint or cold stress, reserpine, serotonin, non-steroidal antiinflammatory drugs (NSAIDs) and pylorus ligation. This phenomenon was initially described as ‘‘cytoprotection’’ by Robert et al (1979). The effect occurs 1 min after oral administration of the prostaglandin and is independent of inhibition of acid secretion, as it is found with prostaglandins that have little or no antisecretory activity and at doses below those necessary to reduce acid secretion, respectively (Robert et al 1979). Furthermore, prostaglandins protect against gastric injury in damage models that are acid-independent, such as ethanol-induced injury (Robert et al 1979) and after intragastric instillation of acid (Kauffman and Grossman 1978). The major targets of prostaglandin-induced protection are the deep necrotic lesions induced by necrotizing agents and antiinflammatory drugs resulting in macroscopically visible protection. However, histological studies have revealed that damage to the surface epithelial cells still occurs after prostaglandin treatment, despite prevention of deep necrotic lesions (Lacy and Ito 1982). From these findings it was suggested that the terms ‘‘gastroprotection’’ or ‘‘mucosaprotection’’ are preferable to the term ‘‘cytoprotection’’. Protection of surface epithelial cells in addition to protection of deeper necrosis induced by ethanol and aspirin was, however, demonstrated by scanning electron microscopy (Ohno et al 1985), suggesting that the extent of protection varies in different experimental models. Gastroprotection in Humans The gastroprotective action of prostaglandins was also demonstrated in human gastric mucosa. Thus, pretreatment with 16,16dimethyl-PGE2 greatly reduced alcohol (60%)-induced histological damage to gastric mucosal microvessels and haemorrhages (Tarnawski et al 1988a). In addition, oral administration of PGE2 significantly reduced faecal blood loss induced by aspirin and indomethacin in healthy volunteers (Cohen et al 1980). Similar observations were made using the PGE1 analogue misoprostol in non-antisecretory doses (Cohen et al 1985). Faecal blood loss was also significantly reduced by oral PGE2 in rheumatic patients taking indomethacin (Johansson et al 1980). As gastrointestinal lesions induced by NSAIDs are acid-sensitive, part of the protective action observed with the E-type prostaglandins could be related to inhibitory effects on acid secretion. MECHANISM OF PROSTAGLANDIN-INDUCED PROTECTION Vasodilatory Actions Despite the great efforts undertaken to elucidate the mechanisms underlying prostaglandin-induced gastroprotection, the phenomenon has not been unequivocally clarified. Prostaglandins influence various gastric functions that can strengthen mucosal defence reactions against injury. Maintenance of gastric mucosal blood flow was supposed to play a crucial role in the resistance of the gastric mucosa against damage. Sufficient blood flow ensures the delivery of nutrients and the removal of potentially injurious products such as hydrogen ions. 15-(S)-15-methyl-PGE2 protected against rat
gastric damage induced by acidified taurocholate and indomethacin co-administration and simultaneously increased gastric mucosal blood flow (Rush and Ruwart 1984). Vasodilation elicited by PGI2 was suggested to ensure bicarbonate delivery to the gastric mucosa necessary for an adequate buffering capacity, thus mediating protection against hypovolaemic shock-induced injury in the rat stomach (Starlinger et al 1981). In the dog, 16,16-dimethyl-PGE2 did not increase resting gastric blood flow and simultaneously did not protect against damage induced by acid-aspirin-hypovolaemic shock, whereas protection occurred after treatment with a PGE1 analogue which elevated resting gastric blood flow (Jensen et al 1981; Larsen et al 1981). Furthermore, the mucosal permeability changes and lesion production elicited by topical taurocholate or ethanol in the canine stomach were inhibited by vasodilating doses of PGE2 (Konturek and Robert 1982). In the rat, topical ethanol induced stasis of flow in the damage area, as assessed by in vivo fluorescent microscopy, and this was prevented by pretreatment with 16,16-dimethyl-PGE2 (Guth et al 1984; Lacy and Ito 1982). Whether these effects of ethanol-induced vascular changes represent an underlying mechanism or are the consequence of the protective action has not been clarified so far. Although it is tempting to speculate that the increase in gastric blood flow and/or the prevention of microcirculatory stasis mediate prostaglandin-induced protection, one should bear in mind that potent gastroprotective activity is found with agents that substantially reduce gastric mucosal blood flow, such as tachykinin NK2 analogues (Stroff et al 1996). Furthermore, 16,16-dimethyl-PGE2 prevented the reduction in viability, increase in LDH release and prominent ultrastructural damage induced by indomethacin and ethanol in isolated human gastric glands, indicating that prostaglandin protection does not necessarily depend on systemic actions, such as preservation of mucosal blood flow and maintenance of microvessel integrity (Tarnawski et al 1988b). Effects on Gastric Mucus, Bicarbonate and Fluid Secretion The role of gastric mucus secretion for the protective activity of prostaglandins is controversial. Intragastric administration of PGE2 and 16,16-dimethyl-PGE2 increased the release of viscous and soluble mucus in humans and/or rats (Johansson et al 1980; Bolton et al 1978). In another study, however, no increase in adherent mucus gel thickness was found with PGE2 or 16,16-dimethyl-PGE2, despite potent gastroprotection (Robert et al 1984). Gastric mucus could strengthen mucosal defence by acting as a physical barrier or by creating an unstirred layer of secreted bicarbonate, which helps to neutralize hydrogen ions diffusing back from the lumen into the mucosa. This effect may be enhanced by prostaglandin stimulation of gastric alkaline secretion observed in various species (Bolton et al 1978; Gascoigne and Hirst 1981; Kauffman et al 1980). Wallace and McKnight (1990) have shown that superficial damage of the rat gastric mucosa induced by hypertonic saline produces a mucoid cap that has a pH of 4–6. Inhibition of prostaglandin formation by NSAIDs resulted in complete dissipation of this high-pH microenvironment, with subsequent development of haemorrhagic erosions. Furthermore, certain prostanoids, including 16,16dimethyl-PGE2, have been found to stimulate net fluid output into the gastric lumen. This effect may contribute to mucosal protection by enhancing the disposal of acid or diluting topical irritants (Gascoigne and Hirst 1981). Effects on Ion Transport, Surface Hydrophobicity and Growth Factors Additional proposed mechanisms underlying prostaglandin protection include reduction of acid back-diffusion into the mucosal
STOMACH PHYSIOLOGY tissue elicited by so-called ‘‘barrier-breakers’’ (Bommelaer and 1979; Muller et al 1981), stimulation of gastric chloride transport, resulting in ionic exchange of cellular chloride for nutrient bicarbonate and elevated intracellular cAMP levels (Schiessel et al 1980), effects on surface hydrophobicity by modulating surfaceactive phospholipids (Lichtenberger et al 1983), and effects on non-protein sulphydryls (Szabo et al 1981). The gastroprotective effect of PGE2 was shown to be receptor subtype-specific. Thus, eicosanoid EP1 receptor agonists mimicked prostaglandin protection, whereas EP1 receptor antagonists blocked the protective effect of PGE2. Furthermore, the protective action of PGE2 totally disappeared in gene-disrupted mice lacking EP1 receptors (Araki et al 2000). Finally, it was demonstrated that PGE2, PGE1 and a stable PGI2 analogue strongly induced the expression of hepatocyte growth factor in human gastric fibroblasts by activation of a PGE-specific EP2 or EP4 receptor (Takahashi et al 1996a). Hepatocyte growth factor markedly stimulated proliferation and migration of rabbit gastric epithelial cells in an in vitro primary culture system (Takahashi et al 1995) and, in addition, protected against ethanol-induced gastric epithelial cell disruption through facilitating an actin–myosin contractile system (Takahashi et al 1996b). The acceleration of restitution evoked by PGE1 was abolished by anti-hepatocyte growth factor antibody, indicating that the action of the prostaglandin was mediated by hepatocyte growth factor (Takahashi et al 1996a). Taken together, these findings show that prostaglandins influence various gastric functions that may be important for the maintenance of gastric mucosal integrity. However, induction of such effects does not always correlate with the protective potency, suggesting that a complex system of interacting effects may underlie the protective action of prostaglandins. The relative contribution of specific effects may differ depending on the prostaglandin studied and the damage model used. CYCLOOXYGENASE-1 AND CYCLOOXYGENASE-2 Cyclooxygenase (COX) catalyses the initial step in the metabolism of arachidonic acid to PGH2, which is then converted to the various prostaglandins and thromboxanes by cell-specific enzymes. At least two isoforms of COX, COX-1 and COX-2, have been identified. Whereas COX-1 is expressed constitutively in most tissues, expression of COX-2 is usually low under basal conditions (Kargman et al 1996). Cytokines, endotoxin and mitogens rapidly upregulate COX-2 (Kujubu et al. 1991; Raz et al. 1989) and high expression of COX-2 is found at sites of inflammation, tissue damage and malignant transformation (Vane and Botting 1995). Initially, the concept was developed that COX1 functions as a housekeeping enzyme responsible for the maintenance of homeostasis reactions, particularly through formation of protective prostaglandins in the gastrointestinal tract, kidney and vasculature, whereas COX-2 yields the prostaglandins responsible for inflammatory reactions and malignant transformation (Vane and Botting 1995). Consequently, it has been proposed that inhibition of COX-2 mediates the therapeutic activity, and inhibition of COX-1 the toxicity of NSAIDs. Standard NSAIDs inhibit the COX pathway of prostaglandin biosynthesis, with little selectivity for the COX-1 and COX-2 isoform (Mitchell et al. 1993; Vane 1971). This underlies the antiinflammatory action but is also responsible for the development of side effects in various organ systems, including the gastrointestinal tract (Halter et al 2001; Hawkey 1990). Long-term NSAID use is commonly associated with gastric erosions (Hawkey 1990) and a two- to five-fold increase in relative risk and 30% attributable risk of ulcer perforation, upper gastrointestinal bleeding and death (Hawkey and Hudson 1994). Great effort has been undertaken to develop selective COX-2 inhibitors which do not affect the activity
435
of COX-1 when used in therapeutic doses (Masferrer et al 1994). COX-2 inhibitors were found to exert potent antiinflammatory and analgesic effects and to induce considerably less gastrointestinal injury than standard NSAIDs in experimental animals (Masferrer et al 1994; Schmassmann et al 1998) and man (Bombardier et al 2000; Jackson and Hawkey 2000; Laine et al 1999; Silverstein et al 2000; Whittle 2000). From these findings it has been suggested that only COX-1 is responsible for the maintenance of gastric mucosal integrity, with COX-2 having no role in mucosal resistance (Vane and Botting 1995). Later on, however, it was shown that both COX-1 and COX-2 contribute to gastric mucosal defence, with specific roles of the isoenzymes depending on the pathophysiological setting. Furthermore, in humans the selective COX-2 inhibitor celecoxib significantly reduced urinary excretion of the PGI2 metabolite 2,3-dinor 6keto-PGF1a, indicating that COX-2 is a major source of systemic PGI2 biosynthesis (McAdam et al 1999). Whether this finding is related to the observation that the incidence of myocardial infarction is higher among patients with rheumatoid arthritis treated with the COX-2 inhibitor rofecoxib than among those treated with the non-selective COX inhibitor naproxen (Bombardier et al 2000) remains to be finally clarified. Patients with rheumatoid arthritis or osteoarthritis co-treated with celecoxib and low-dose aspirin for cardioprophylaxis had no increased risk of cardiovascular events, but administration of aspirin blunted the gastroduodenal sparing effect of celecoxib (Silverstein et al 2000). ULCEROGENIC EFFECTS OF NON-SELECTIVE COX INHIBITORS In experimental animals, administration of NSAIDs causes gastric mucosal damage within a few hours (Masferrer et al 1994; Schmassmann et al 1998). Prostaglandin deficiency seems to be the primary mechanism of NSAID gastrotoxicity. This is supported by findings showing that, in rabbits, endogenous prostaglandin neutralization by active immunization with PGE2thyroglobulin conjugate or passive immunization with PGE2hyperimmune plasma leads to gastric ulcers (Redfern et al 1987). Prostaglandins activate a number of gastric functions capable of supporting gastric mucosal defence. NSAIDs can exert injurious actions on the gastric mucosa, as they interfere with such prostaglandin-dependent defence mechanisms (Halter et al 2001). Thus, inhibition of gastric prostaglandin biosynthesis resulted in a decrease in mucus and bicarbonate secretion, caused vasoconstriction and reduced gastric mucosal blood flow (Gana et al 1987; Whittle 1980). Furthermore, NSAIDs markedly increased the number of neutrophils adhering to the vascular endothelium in both gastric and mesenteric venules (Wallace et al 1993). Adherence is mediated by the expression of the b2 integrin (CD11/CD18) on neutrophils and of the intercellular adhesion molecule on the vascular endothelium (Wallace et al 1993). Indomethacin increased plasma levels of tumour necrosis factor-a, a proinflammatory cytokine that causes leukocyte margination by upregulating expression of adhesion molecules on both neutrophils and endothelial cells. Inhibition of the increase in tumour necrosis factor-a (TNFa) plasma levels by pentoxifylline diminished gastric damage and neutrophil margination, suggesting a mediator role of TNFa in indomethacin-induced leukocyte adhesion (Santucci et al 1994). In another study, however, pentoxifylline was found to prevent indomethacin-induced gastric damage but not leukocyte adhesion (Appleyard et al 1996). Neutrophil adherence results in microvascular stasis and mucosal injury due to ischaemia and release of oxygen-derived free radicals and proteases (Vaananen et al 1991). The severity of experimental NSAID gastrotoxicity was substantially diminished in rats with neutropenia following treatment with antineutrophil serum or
436
DIGESTIVE SYSTEM
methotrexate (Lee et al 1992; Wallace et al 1990b). Prostaglandinmediated maintenance of gastric mucosal blood flow and suppression of leukocyte adherence is COX isoenzyme-specific. Thus, in rats, the selective COX-1 inhibitor SC-560 reduced gastric mucosal blood flow without affecting leukocyte adherence, whereas the selective COX-2 inhibitor celecoxib promoted leukocyte adherence but did not reduce gastric mucosal blood flow (Wallace et al 2000). Further suggested mechanisms of gastrotoxicity include the ability of NSAIDs to stimulate gastric motor activity (Takeuchi et al 1986) and to decrease the hydrophobicity of the mucus gel layer of the gastric mucosa, thus converting the mucus gel from a nonwettable to a wettable state (Goddard et al 1987). Aspirin and most non-aspirin NSAIDs are weak organic acids. In the presence of acid they are deionized and penetrate into the gastric epithelial cells. There, at neutral pH, they are reionized and trapped within the cells, reaching high, injurious concentrations. NSAIDs uncouple mitochondrial oxidative phosphorylation (Somasundaram et al 2000), resulting in changes in mitochondrial morphology and a decrease in intracellular ATP (Halter et al 2001).
Leukotrienes in NSAID-induced Damage Reports on the role of leukotrienes in rat gastric mucosal damage induced by NSAIDs are controversial. Thus, in rats, ulcerogenic doses of indomethacin reduced gastric mucosal PGE2 formation without increasing release of LTC4 and at higher doses even decreased formation of the leukotriene (Peskar 1991). MK-886, a 5-lipoxygenase inhibitor, at a dose that abolished gastric LTC4 formation, did not inhibit indomethacin-induced gastric damage (Peskar 1991). Similarly, Lee and Feldmann (1992) did not find changes of rat gastric mucosal formation of LTC4 and LTB4 after administration of aspirin and neither MK-571, a LTD4 receptor antagonist, nor MK-886, a 5-lipoxygenase inhibitor, could reduce the mucosal lesions induced by aspirin. On the other hand, LTB4 has been suggested to contribute to the pathogenesis of NSAIDinduced gastric injury through its ability to promote leukocyte
adherence to the vascular endothelium. In rats, receptor antagonists for LTB4 and inhibitors of 5-lipoxygenase attenuated NSAID-induced leukocyte adherence (Asako et al 1992) and reduced the severity of NSAID-induced mucosal damage (Vaananen et al 1992). Kobayashi et al (1993) could not demonstrate effects of indomethacin on the gastric mucosal content of LTB4 and sulphidopeptide leukotrienes, but observed attenuation of mucosal damage and reduction of mucosal blood flow following indomethacin administration in rats treated with 5-lipoxygenase inhibitors and antagonists of sulphidopeptide leukotriene receptors. Enhanced gastric mucosal LTB4 synthesis was strongly associated with the presence of type C gastritis in patients taking NSAIDs, whereas synthesis of LTC4 was associated with H. pylori colonization but not NSAID use (Hudson et al 1993).
ULCEROGENIC EFFECTS OF COX-1 AND COX-2 INHIBITION Normal Gastric Mucosa Standard NSAIDs, such as indomethacin, induced dose-dependent gastric mucosal damage in rats and simultaneously nearmaximally inhibited gastric mucosal formation of 6-keto-PGF1a and platelet release of TXB2. Oral administration of the selective COX-1 inhibitor SC-560 suppressed formation of gastric mucosal 6-keto-PGF1a and platelet TXB2 to the same degree as indomethacin but did not damage the gastric mucosa (Figures 39.2 and 39.3: Gretzer et al 2001; Wallace et al 2000). Likewise, the selective COX-2 inhibitors rofecoxib or celecoxib did not injure the gastric mucosa when given alone. However, combined administration of SC-560 and rofecoxib or celecoxib induced severe gastric mucosal lesions (Figure 39.2; Gretzer et al 2001; Wallace et al 2000). Obviously, only simultaneous inhibition of both COX isoenzymes weakens the resistance of the gastric mucosa. This is possibly due to the specific interference of COX-1 and COX-2 inhibitors with defensive mechanisms such as gastric mucosal blood flow and suppression of leukocyte adhesion (Wallace et al 2000).
Figure 39.2 (A) Lesion formation in normal rat gastric mucosa 5 h after oral administration of the non-selective COX inhibitor indomethacin, the COX-1 inhibitor SC-560 and the COX-2 inhibitor rofecoxib. (B) Lesion formation after acid challenge. Rats were treated with indomethacin, SC-560 or rofecoxib 60 min before instillation of 1 ml of 200 mM HCl and gastric damage was assessed 60 min later. *p50.001 vs. vehicle-treated controls; &p50.05, .p50.001 vs. SC-560 alone. Data modified from Gretzer et al (2001)
STOMACH PHYSIOLOGY
437
Figure 39.3 Effect of indomethacin, SC-560 and rofecoxib on gastric mucosal 6-keto-PGF1a (A) and platelet TXB2 (B) formation in healthy rats. Gastric mucosal fragments were harvested 5 h after drug administration and incubated at 378C for 10 min. Blood was collected by cardiac puncture 5 h after drug administration and was incubated at 378C for 60 min. Release of eicosanoids was measured using radioimmunoassay. Values are the mean+SEM of 4–9 rats. *p50.001 vs. vehicle-treated controls. Data are modified from Gretzer et al (2001)
The lack of ulcerogenicity found with the isolated inhibition of COX-1 or COX-2 induced pharmacologically is in keeping with observations made in mice with disruption of COX genes. Thus, COX-1-deficient mice did not develop gastric lesions, although gastric PGE2 levels were 51% of those in wild-type animals (Langenbach et al 1995). Similarly, COX-2 deficient mice did not show gastric pathology, although severe defects occurred in other organ systems, such as the kidney (Morham et al 1995).
against damage (Whittle et al 1990). Whereas isolated inhibition of COX-2 did not induce gastric damage even after intragastric instillation of acid, the COX-2 inhibitors DFU or NS-398 induced severe damage when NO formation was simultaneously inhibited by L-NAME. Similarly, the COX-2 inhibitors were ulcerogenic in rats in which afferent neurons were defunctionalized by pretreatment with a neurotoxic dose of capsaicin, even without suppression of the NO system (Peskar 2001). Similar effects have been observed for indomethacin (Whittle et al 1990).
Acid-challenged Gastric Mucosa In contrast to normal gastric mucosa, in rats challenged with intraluminal acid at a concentration that did not induce gastric lesions given alone, selective inhibition of COX-1 induced by SC560 caused dose-dependent mucosal injury. Administration of the COX-2 inhibitors rofecoxib and DFU did not damage the acidchallenged gastric mucosa but significantly enhanced the injurious effect of the COX-1 inhibitor (Figure 39.2, Gretzer et al 2001). Intragastric instillation of acid induced the expression of COX-2 mRNA without effect on COX-1 mRNA levels. Pretreatment with dexamethasone prevented the acid-induced upregulation of COX2 expression and simultaneously aggravated the damage induced by SC-560 on the acid-challenged mucosa, similar to the effect of COX-2 inhibitors (Gretzer et al 2001). These findings show that the role of COX isoenzymes in gastric mucosal defence differs in normal mucosa and in mucosa exposed to a potentially noxious agent. Whereas in normal mucosa only combined inhibition of COX-1 and COX-2 is injurious, isolated inhibition of COX-1 alone is sufficient to interfere with mucosal defence when a potentially noxious agent is present in the gastric lumen. Pending injury induces the expression of COX-2, which then assists COX-1 in the maintenance of mucosal integrity. Interaction with Other Protective Mediators Various mediators such as prostaglandins, NO and afferent nerves have been found to act in concert to ascertain mucosal resistance
Ischaemia-reperfusion Damage In rats, ischaemia of the gastric artery followed by reperfusion increased mRNA levels of COX-2 but not COX-1 (Kishimoto et al 1998; Maricic et al 1999). Ischaemia-reperfusion of the gastric artery induced only minor mucosal damage which was markedly (up to four-fold) aggravated by pretreatment with the COX-2 inhibitors DFU and NS-398 and the non-selective COX inhibitor indomethacin. A similar augmentation of ischaemia-reperfusion injury occurred after pretreatment with dexamethasone, which inhibited the upregulation of COX-2 expression (Figure 39.4). The damage-aggravating effects of COX-2 inhibitors indomethacin and dexamethasone were fully reversed by co-administration of extremely low doses of 16,16-dimethyl-PGE2, which did not exert protection in the absence of COX inhibitors (Maricic et al 1999). The COX-1 inhibitor SC-560 also augmented ischaemiareperfusion gastric damage but considerably higher doses compared with the COX-2 inhibitors were necessary for the effect (Maricic et al 2001). These findings suggest that both COX-1 and COX-2 act to minimize damage induced by ischaemia-reperfusion, but COX-2 seems to play a more important role in this pathophysiological condition. This could be related to the stimulatory effect of COX-2 inhibition on leukocyte adherence (Wallace et al 2000), as formation of reactive oxygen metabolites by activated neutrophils plays a crucial role in the microvascular and parenchymal injury associated with ischaemia-reperfusion (Hernandez et al 1987).
438
DIGESTIVE SYSTEM
Figure 39.4 Effect of COX suppression on ischaemia-reperfusion-induced gastric mucosal damage. Oral pretreatment with the non-selective COX inhibitor indomethacin, the COX-2-selective inhibitor NS-398 or dexamethasone, which prevents upregulation of COX-2 expression, aggravated ischaemia-reperfusion-induced mucosal damage in a dose-dependent manner. Co-administration of 16,16-dimethyl-PGE2 (twice 4–15 ng/kg) counteracted the injurious effect of COX suppression. Each bar represents the mean+SEM of at least six rats. *p50.05, **p50.001 vs. ischaemia-reperfusion alone (Co); .p50.001 vs. COX inhibitors or dexamethasone without prostaglandin treatment. Data derived from Maricic et al (1999)
Adaptive Gastroprotection
EICOSANOIDS AND ULCER HEALING
Mild irritants, such as 20% ethanol instilled intragastrically, protect against damage induced by a subsequent instillation of necrotizing agents (Robert et al 1983). This ‘‘adaptive gastroprotection’’ depends on endogenous prostaglandins, as it is prevented by pretreatment with indomethacin (Gretzer et al 1998; Robert et al 1983). Similarly, the COX-2 inhibitors NS-398, L-745,337 and DFU abolished the protective effect of 20% ethanol (Figure 39.5). Pretreatment with dexamethasone at a dose that significantly reduced formation of inflammatory PGE2 had no effect on the protection induced by 20% ethanol, suggesting that the protective action of the mild irritant is mediated by a constitutive COX-2 (Gretzer et al 1998). Perfusion of the gastric lumen with 8% peptone protected against damage induced by subsequent perfusion with 50% ethanol. The protective effect of peptone was partially inhibited by pretreatment with indomethacin or COX-2 inhibitors but not dexamethasone, indicating that upregulation of COX-2 is not involved (Ehrlich et al 1998). Similarly, rebamipide, a 2-(1H)-quinolinone analogue with gastroprotective and ulcer-healing properties, was shown to selectively upregulate the expression of COX-2 in rat gastric mucosa without modulating COX-1 expression. Both the rebamipide-induced gastroprotection and increase in mucosal PGE2 formation was blocked by a COX-2 inhibitor (Sun et al 2000). In the rat stomach, repeated administration of endotoxin prevented the damaging effect of ethanol and simultaneously increased gastric mucosal mRNA levels for both COX-1 and COX-2. The endotoxin-induced gastric resistance to injury was abolished by pretreatment with indomethacin but not with the COX-2 inhibitor L-745,337 (Ferraz et al 1997). Hence, whereas the protective effects of 20% ethanol and intragastric peptone seem to be mediated by a constitutive COX-2, the protection induced by chronic administration of endotoxin seems to be mediated by an upregulated COX-1 without contribution of COX-2.
Role of COX-2 in Angiogenesis Angiogenesis plays a crucial role in gastric mucosal repair processes. Wound healing and angiogenesis is promoted by growth factors through modulation of COX-2 expression. Hepatocyte growth factor increased the expression of COX-2 in monolayers of rat gastric epithelial cells and the hepatocyte growth factor-facilitated restitution after artificial wounding was delayed by a COX-2 inhibitor (Horie-Sakata et al 1998). Using sponges implanted subcutaneously into the back of rats as an angiogenesis model, it was demonstrated that basic fibroblast growth factor markedly elevated levels of mRNA for COX-2 and vascular endothelium growth factor, parallel to an increase in angiogenesis. The stimulation of angiogenesis and upregulation of vascular endothelium growth factor induced by basic fibroblast growth factor was prevented by the COX-2 inhibitor NS-398 (Majima et al 2000). The non-selective COX inhibitors indomethacin and diclofenac, as well as the COX-2 inhibitors L745,337 and NS-398, suppressed angiogenesis in the ulcer base of chronic gastric cryo-ulcers and acetic acid-induced ulcers in rats (Schmassmann et al 1998; Shigeta et al 1998). Both non-selective NSAIDs and the selective COX-2 inhibitor NS-398 inhibited in vitro angiogenesis through direct effects on endothelial cells (Jones et al 1999). This action involved inhibition of mitogen-activated protein kinase (ERK2) activity, interference with ERK nuclear translocation, was independent of protein kinase C and had prostaglandin-dependent and prostaglandin-independent components (Jones et al 1999). COX-1 and COX-2 in Gastric Ulcers Whereas levels of COX-2 mRNA and protein are low in normal gastric mucosa, abundant expression of COX-2 is found in
STOMACH PHYSIOLOGY
439
Figure 39.5 Effect of COX inhibitors and dexamethasone on the protection conferred by 20% ethanol. Pretreatment with the non-selective COX inhibitor indomethacin or the COX-2-selective inhibitor NS-398 inhibited the protective effect of 20% ethanol against mucosal damage caused by 70% ethanol in a dose-dependent manner. Co-administration of 16,16-dimethyl-PGE2 (4 ng/kg) prevented the effect of NS-398. Pretreatment with dexamethasone did not diminish the protection conferred by 20% ethanol indicating that enzyme induction is not involved. Each bar represents the mean+SEM of six experiments. *p50.001 vs. damage induced by 70% ethanol (Co); ^p50.05, .p50.001 vs. the protective effect of 20% ethanol; ~ p50.001 vs. NS-398 alone. Data modified from Gretzer et al (1998)
ulcerated tissue. In mice, upregulation of COX-2 expression in gastric ulcers, elicited by subserosal injection of acetic acid, was associated with a three-fold higher prostaglandin formation compared with normal gastric mucosa, which was inhibited in vitro by the COX-2 inhibitor NS-398 (Mizuno et al 1997). Similarly, in rats with chronic cryo-ulcers, COX-2 immunoreactivity was negligible in the normal gastric wall, but after ulceration occurred in abundance in the cytoplasm of monocytes, macrophages, fibroblasts and endothelial cells in regions of maximal repair activity. The time course of COX-2 induction (Figure 39.6) closely paralleled the increase in epithelial cell proliferation, as assessed by the uptake of bromodeoxyuridine, with a maximum 5 days after ulcer induction. COX-1 immunoreactivity was located mainly in the mucous neck cells of the non-ulcerated mucosa, decreased after gastric ulceration in the mucosa adjacent to the ulcer crater and reappeared after day 5 in the apical cytoplasm of the regenerative epithelial cells (Schmassmann et al 1998). These findings demonstrate that in chronic gastric ulcers COX-1 and COX-2 differ in location and time sequence of expression. In acetic acid-induced ulcers in rats, increased expression of interleukin (IL)-1b, TNFa and transforming growth factor (TGF)-b1 mRNAs occurred in addition to elevated PGE2 production in the ulcerated tissue. Blockade of IL-1b and TNFa reduced the expression of COX-2 mRNA and prostaglandin production in a culture of isolated ulcer base. In contrast, COX-2 mRNA expression and PGE2 production were promoted by preventing the action of TGFb1 (Takahashi et al 1998). These findings confirm the upregulation of COX-2 in gastric ulcers in rats and the proposal that cytokines and growth factors are involved in the regulation of COX-2 expression.
Effect of COX Inhibition on Gastric Ulcer Healing Standard NSAIDs delay gastric ulcer healing in experimental animals and man (Halter et al 2001). In mice with acetic acid-
induced gastric ulcers, the COX-2 inhibitor NS-398 significantly delayed ulcer healing (Mizuno et al 1997). In rats with gastric cryo-ulcers, treatment with indomethacin, diclofenac or the COX2 inhibitor L-745,337 resulted in a dose-dependent impairment of ulcer healing, which was evident in the second week after ulcer induction (Figure 39.7; Schmassmann et al 1998). Epithelial cell proliferation in the ulcer margin and microvessel density in the ulcer bed were decreased, and the thickness of the granulation tissue below the ulcer crater and the gap between both edges of the muscularis mucosae were increased to the same extent by the nonselective COX inhibitors and the COX-2 selective inhibitor (Schmassmann et al 1998). A similar finding was reported by Shigeta et al (1998), who observed delayed healing of acetic acid-induced rat gastric ulcers associated with impaired regeneration of the mucosa, maturation of the ulcer base and angiogenesis in the base after treatment with indomethacin or NS-398. Local injection of hepatocyte growth factor or gastrin around the ulcers accelerated healing, raised mucosal blood flow at the ulcer margin and further upregulated COX-2 mRNA and protein in the ulcerated mucosa. Similarly, indomethacin or the COX-2 inhibitors NS-398 and rofecoxib inhibited generation of PGE2, reduced mucosal blood flow at the ulcer margin and attenuated the acceleration of healing by hepatocyte growth factor and gastrin (Brzozowski et al 2000).
EXPRESSION OF COX-2 IN H. PYLORI INFECTION AND GASTRIC CARCINOMA H. pylori colonization of the stomach is associated with chronic gastritis and peptic ulcer disease. H. pylori stimulated the formation of PGE2 in MKN 28 gastric mucosal cells in vitro (Romano et al 1998) and mucosal PGE2 formation in humans in vivo (Konturek et al 1998). In H. pylori gastritis, COX-2 and the inducible NO synthase are co-expressed in increased amounts (Fu et al 1999). Eradication of H. pylori reduced COX-2 protein
440
DIGESTIVE SYSTEM
Figure 39.6 Time-sequence of immunoreactivity for COX-1, COX-2 and bromodeoxyuridine (BrdU). Time-sequence of immunoreactivity of COX1 in the epithelial cells of the ulcer margin, COX-2 in non-epithelial cells in the connective tissue below the regenerative glands and BrdU in epithelial cells of the ulcer margin as index of epithelial cell proliferation. The maximal immunoreactivities for COX-2 and BrdU were on day 5. Data are derived from Schmassmann et al (1998)
expression parallel to the reduction of mucosal inflammation (McCarthy et al 1999). Expression of COX-2 may limit inflammation and injury in active gastritis but may also contribute to H. pylori-associated neoplastic transformation (Fu et al 1999). The prolonged use of NSAIDs has been associated with a reduced risk of cancer. The suggested target of these drugs is COX. Expression of COX-2 mRNA was elevated in human gastric biopsies containing intestinal metaplasia and frequency of COX-2 immunopositivity increased during progression from reactive epithelium to high-grade dysplasia (van Rees 2002). Expression of COX-2 further increased with advanced tumour stage and the presence of metastases (Rajnakova et al 2001) and correlated closely with the depth of invasion (Ohno et al 2001). COX-2 overexpression was associated with increased PGE2 biosynthesis and angiogenesis in gastric cancer (Uefuji et al 2000). Furthermore, treatment with indomethacin or the selective COX-2 inhibitor NS-398 significantly reduced the tumour volume of gastric cancer xenografts in nude mice by inducing apoptosis in cancer cells and inhibiting cancer cell replication (Sawaoka et al 1998).
CONCLUSIONS Exogenous prostaglandins, thromboxane, leukotrienes and their synthetic analogues exert pronounced effects on various gastric functions, such as secretory processes, gastric mucosal blood flow, ion transport processes, surface hydrophobicity and expression of growth factors. The most impressive actions of prostaglandins are the potent inhibition of acid secretion and the prevention of
Figure 39.7 Ulcer diameter–time curve. All COX inhibitors decreased the healing of gastric cryoulcers in rats predominantly in the second week after ulcer induction. The selective COX-2 inhibitor L-745,337 (265 mg/kg, p.o.) delayed gastric ulcer healing similarly to the two non-selective COX inhibitors diclofenac (262.5 mg/kg, p.o.) and indomethacin (260.5 mg/ kg, s.c.). Inhibition of gastric acid secretion by omeprazole accelerated gastric ulcer healing. *p50.002 vs. controls. Data derived from Schmassmann et al (1998)
gastric mucosal damage induced by a great variety of ulcerogens. Although numerous effects that strengthen gastric mucosal resistance have been described, the precise mechanisms underlying the protective activity of prostaglandins have not been unequivocally identified. Probably the phenomenon of gastroprotection is complex, with different factors contributing in different situations. Gastric mucosal and muscular tissues synthesize large amounts of eicosanoids from endogenous substrate via COX and smaller amounts via lipoxygenase pathways. Endogenous eicosanoids are involved in numerous physiological and pathophysiological gastric processes, such as maintenance of gastric mucosal blood flow, secretion of acid, bicarbonate and fluid, maintenance of gastric mucosal integrity, gastric mucosal restitution, angiogenesis, and gastric cancer. Inhibition of COX, the key enzyme in prostaglandin biosynthesis, interferes with the ability of the gastric mucosa to withstand injury. The intriguing concept that only prostaglandins generated via the COX-1 pathway are essential for the maintenance of a gastric mucosal integrity, whereas COX-2-derived prostaglandins are not involved in gastric mucosal defence, has recently been challenged. Thus, it is now clear that only simultaneous inhibition of COX-1 and COX-2 damages the gastric mucosa and that in certain pathophysiological situations, such as acid challenge of the gastric mucosa, ischaemia-reperfusion, suppression of NO biosynthesis and impairment of afferent nerve function, selective inhibition of COX-2 alone is highly ulcerogenic. Furthermore, it has been established that COX-2 is upregulated in gastric ulcers and is essential for rapid ulcer healing. Whether these effects and the finding that COX-2 is a major source of systemic, probably vascular, PGI2 biosynthesis will, at least partially, modulate the benefit of reduced gastrointestinal toxicity of the COX-2 inhibitors remains to be clarified.
STOMACH PHYSIOLOGY REFERENCES Ahlquist DA, Duenes JA, Madson TH et al (1982) Prostaglandin generation from gastroduodenal mucosa: regional and species differences. Prostaglandins, 24, 115–125. Ahlquist DA, Madson TH, Romero JC et al (1983) Factors influencing metabolism of arachidonic acid in guinea pig gastric mucosa. Am J Physiol, 244, G131–G137. Appleyard CB, McCafferty DM, Tigley AW et al (1996) Tumor necrosis factor mediation of NSAID-induced gastric damage: role of leukocyte adherence. Am J Physiol, 270, G42–G48. Araki H, Ukawa H, Sugawa Y et al (2000) The roles of prostaglandin E receptor subtypes in the cytoprotective action of prostaglandin E2 in rat stomach. Aliment Pharmacol Therapeut, 14 (suppl 1), 116–124. Asako H, Kubes P, Wallace J et al (1992) Indomethacin-induced leukocyte adhesion in mesenteric venules: role of lipoxygenase products. Am J Physiol, 262, G903–G908. Baker R, Jaffe BM, Reed JD et al (1978) Endogenous prostaglandins and gastric secretion in the cat. J Physiol, 278, 451–460. Baker R, Jaffe BM and Venables CW (1979) Endogenous prostaglandins in peptic ulcer disease. Gut, 20, 394–399. Banerjee AK, Phillips J and Winning WM (1972) E-type prostaglandins and gastric acid secretion in the rat. Nature (New Biol), 238, 177–179. Befrits R and Johansson C (1985) Oral PGE2 inhibits gastric acid secretion in man. Prostaglandins, 29, 143–152. Bennett A, Hensby CN, Sanger GJ and Stamford IF (1981) Metabolites of arachidonic acid formed by human gastrointestinal tissues and their actions on the muscle layers. Br J Pharmacol, 74, 435–444. Bersimbaev RI, Tairov MM and Salganik RI (1985) Biochemical mechanisms of regulation of mucus secretion by prostaglandin E2 in rat gastric mucosa. Eur J Pharmacol, 115, 259–266. Bolton JP, Palmer D and Cohen MM (1978) Stimulation of mucus and nonparietal cell secretion by the E2 prostaglandins. Am J Dig Dis, 23, 359–364. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med, 343, 1520– 1528. Bommelaer G and Guth PH (1979) Protection by histamine receptor antagonists and prostaglandin against gastric mucosal barrier disruption in the rat. Gastroenterology, 77, 303–308. Boughton-Smith NK and Whittle BJ (1981) The gastric antisecretory actions of prostaglandin E2 and stable prostacyclin analogues against different secretagogues in perfused whole-stomachs of rat or mouse in vitro. Br J Pharmacol, 72, 291–298. Brzozowski T, Konturek PC, Konturek SJ et al (2000) Involvement of cyclooxygenase (COX)-2 products in acceleration of ulcer healing by gastrin and hepatocyte growth factor. J Physiol and Pharmacol, 51, 751–773. Carter DC, Karim SM, Bhana D and Ganesan PA (1973) Inhibition of human gastric secretion by prostaglandin. Br J Surg, 60, 828–831. Chen MC, Amirian DA, Toomey M et al (1988) Prostanoid inhibition of canine parietal cells: mediation by the inhibitory guanosine triphosphate-binding protein of adenylate cyclase. Gastroenterology, 94, 1121–1129. Chow JY, Ma L and Cho CH (1998) Effect of cigarette smoke on ethanolinduced gastric mucosal lesions: the role of nitric oxide and neutrophils. Eur J Pharmacol, 342, 253–260. Classen M, Koch H, Bickhardt J et al (1971) The effect of prostaglandin E1 on the pentagastrin-stimulated gastric secretion in man. Digestion, 4, 333–344. Cohen MM, Cheung G and Lyster DM (1980) Prevention of aspirininduced faecal blood loss by prostaglandin E2. Gut, 21, 602–606. Cohen MM, Clark L, Armstrong L and D’Souza J (1985) Reduction of aspirin-induced fecal blood loss with low-dose misoprostol tablets in man. Dig Dis Sci, 30, 605–611. Dajani EZ, Driskill DR, Bianchi RG and Collins PW (1975) Comparative gastric antisecretory and antiulcer effects of prostaglandin E1 and its methyl ester in animals. Prostaglandins, 10, 205–215. Dajani EZ, Callison DA, Bianchi RG and Driskill DR (1976) Gastric antisecretory effects of E prostaglandins in Rhesus monkeys. Am J Dig Dis, 21, 1020–1028. Dembinska-Kiec A, Peskar BA, Muller MK and Peskar BM (1989) The effects of platelet-activating factor on flow rate and eicosanoid release
441
in the isolated perfused rat gastric vascular bed. Prostaglandins, 37, 69– 91. Dozois RR and Thompson CJ (1974) Presence of prostaglandins E2 and A2 in canine gastric secretions. Life Sci, 15, 975–986. Dreyling KW, Hoppe U, Peskar BA et al (1986) Leukotriene synthesis by human gastrointestinal tissues. Biochim Biophys Acta, 878, 184–193. Ehrlich K, Plate S, Stroff T et al (1998) Peptidergic and cholinergic neurons and mediators in peptone-induced gastroprotection: role of cyclooxygenase-2. Am J Physiol, 274, G955–G964. Ferraz JG, Sharkey KA, Reuter BK et al (1997) Induction of cyclooxygenase 1 and 2 in the rat stomach during endotoxemia: role in resistance to damage. Gastroenterology, 113, 195–204. Fiorucci S, Distrutti E, Santucci L and Morelli A (1995a) Leukotrienes stimulate pepsinogen secretion from guinea-pig gastric chief cells by a nitric oxide-dependent pathway. Gastroenterology, 108, 1709–1719. Fiorucci S, Santucci L, Gresele P et al (1995b) Effect of NSAIDs on pepsinogen secretion and calcium mobilization in isolated chief cells. Am J Physiol, 268, G968–G978. Fu S, Ramanujam KS, Wong A et al (1999) Increased expression and cellular localization of inducible nitric oxide synthase cyclooxygenase 2 in Helicobacter pylori gastritis. Gastroenterology, 116, 1319–1329. Gana TJ, Huhlewych R and Koo J (1987) Focal gastric mucosal blood flow in aspirin-induced ulceration. Annals of Surgery, 205, 399–403. Gascoigne AD and Hirst BH (1981) Prostaglandins alter the relationship between gastric hydrogen ion concentration and flow: evidence for stimulation of non-parietal secretion in the cat. J Physiol, 316, 427– 438. Gerber JG and Nies AS (1982) Canine gastric mucosal vasodilation with prostaglandins and histamine analogs. Dig Dis Sci, 27, 870–874. Gerkens JF, Gerber JC, Shand DG and Branch RA (1978) Effect of PGI2, PGE2 and 6-keto-PGF1a on canine gastric blood flow and acid secretion. Prostaglandins, 16, 815–823. Goddard PJ, Hills BA and Lichtenberger LM (1987) Does aspirin damage canine gastric mucosa by reducing its surface hydrophobicity? Am J Physiol, 252, G421–G430. Gretzer B, Ehrlich K, Maricic N et al (1998) Selective cyclooxygenase-2 inhibitors and their influence on the protective effect of a mild irritant in the rat stomach. Br J Pharmacol, 123, 927–935. Gretzer B, Maricic N, Respondek M et al (2001) Effects of specific inhibition of cyclooxygenase-1 and cyclooxygenase-2 in the rat stomach with normal mucosa and after acid challenge. Br J Pharmacol, 132, 1565–1573. Guth PH and Moler TL (1982) The role of endogenous prostanoids in the response of the rat gastric microcirculation to vasoactive agents. Microvasc Res, 23, 336–346. Guth PH, Paulsen G and Nagata H (1984) Histologic and microcirculatory changes in alcohol-induced gastric lesions in the rat: effect of prostaglandin cytoprotection. Gastroenterology, 87, 1083– 1090. Halter F, Tarnawski AS, Schmassmann A and Peskar BM (2001) Cyclooxygenase 2—implications on maintenance of gastric mucosal integrity and ulcer healing: controversial issues and perspectives. Gut, 49, 443–453. Hawkey CJ (1990) Non-steroidal antiinflammatory drugs and peptic ulcers. Br Med J, 300, 278–284. Hawkey CJ (2000) Non-steroidal antiinflammatory drug gastropathy. Gastroenterology, 119, 521–535. Hawkey CJ and Hudson N (1994) Mucosal injury caused by drugs, chemicals and stress. In Haubrich WS, Schaffner F and Berck JE (eds), Bockus Gastroenterology 5. Philadelphia, PA: WB Saunders, pp. 656– 699. Hernandez LA, Grisham MB, Twohig B et al (1987) Role of neutrophils in ischemia-reperfusion-induced microvascular injury. Am J Physiol, 253, H699–H703. Higuchi K, Arakawa T, Matsumoto T et al (1992) Immunohistochemical localization of cells that synthesize leukotriene B4 in human gastric mucosa. J Clin Gastroenterol, 14, S64–S67. Horie-Sakata K, Shimada T, Hiraishi H and Terano A (1998) Role of cyclooxygenase 2 in hepatocyte growth factor-mediated gastric epithelial restitution. J Clin Gastroenterol, 27, S40–S46. Horton EW, Main IH, Thompson CJ and Wright PM (1968) Effect of orally administered prostaglandin E1 on gastric secretion and gastrointestinal motility in man. Gut, 9, 655–658.
442
DIGESTIVE SYSTEM
Huber W, Trautmann M, Becker I et al (1993) Leukotriene B4 and C4 production in isolated rat gastric mucosal cells. Am J Physiol, 265, G1021–G1028. Hudson N, Balsitis M, Everitt S and Hawkey CJ (1993) Enhanced gastric mucosal leukotriene B4 synthesis in patients taking non-steroidal antiinflammatory drugs. Gut, 34, 742–747. Jackson LM and Hawkey CJ (2000) COX-2 selective non-steroidal antiinflammatory drugs: do they really offer any advantages? Drugs, 59, 1207–1216. Jaramillo E, Mardh S, Green K et al (1989) The effect of arachidonic acid and its metabolites on acid production in isolated human parietal cells. Scand J Gastroenterol, 24, 1231–1237. Jensen NF, Larsen KR, Davis EK et al (1981) Two prostaglandin analogs (16,16-dimethyl prostaglandin E2 and SC-29333) and the gastric mucosal barrier: a dose–response study. Surg Forum, 32, 124–126. Johansson C, Kollberg B, Nordemar R et al (1980) Protective effect of prostaglandin E2 in the gastrointestinal tract during indomethacin treatment of rheumatic diseases. Gastroenterology, 78, 479–483. Jones MK, Wang H, Peskar BM et al (1999) Inhibition of angiogenesis by non-steroidal antiinflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing. Nature Med, 12, 1418–1423. Kargman S, Charleson S, Cartwright M et al (1996) Characterization of prostaglandin G/H synthase 1 and 2 in rat, dog, monkey, and human gastrointestinal tracts. Gastroenterology, 111, 445–454. Kasirga E, Coker I, Aydogdu S et al (1999) Increased gastric juice leukotriene B4, C4 and E4 concentrations in children with Helicobacter pylori colonization. Turk J Pediatr, 41, 335–339. Kauffman GL and Grossman MI (1978) Prostaglandin and cimetidine inhibit the formation of ulcers produced by parenteral salicylates. Gastroenterology, 75, 1099–1102. Kauffman GL and Whittle BJ (1982) Gastric vascular actions of prostanoids and the dual effect of arachidonic acid. Am J Physiol, 242, G582–G587. Kauffman GL, Whittle BJ, Aures D et al (1979) Effects of prostacyclin and a stable analogue, 6-b-PGI1, on gastric acid secretion, mucosal blood flow, and blood pressure in conscious dogs. Gastroenterology, 77, 1301–1306. Kauffman GL, Reeve JJ and Grossman MI (1980) Gastric bicarbonate secretion: effect of topical and intravenous 16,16-dimethyl prostaglandin E2. Am J Physiol, 239, G44–G48. Kishimoto Y, Wada K, Nakamoto K et al (1998) Levels of cyclooxygenase-1 and -2 mRNA expression at various stages of acute gastric injury induced by ischemia-reperfusion in rats. Arch Biochem Biophys, 352, 153–157. Knapp HR, Oelz O, Sweetman BJ and Oates JA (1978) Synthesis and metabolism of prostaglandins E2, F2a and D2 by the rat gastrointestinal tract. Stimulation by a hypertonic environment in vitro. Prostaglandins, 15, 751–757. Kobayashi K, Fukuda T, Higuchi K et al (1993) Role of leukotrienes in indomethacin-induced mucosal damage in rats. J Clin Gastroenterol, 17, S11–S14. Konturek SJ and Robert A (1982) Cytoprotection of canine gastric mucosa by prostacyclin: possible mediation by increased mucosal blood flow. Digestion, 25, 155–163. Konturek SJ, Bilski J, Dembinski A et al (1987) Effects of leukotrienes on gastric acid and alkaline secretions. Gastroenterology, 92, 1209–1214. Konturek SJ, Brzozowski T, Drozdowicz D and Beck G (1988) Role of leukotrienes in acute gastric lesions induced by ethanol, taurocholate, aspirin, platelet-activating factor and stress in rats. Dig Dis Sci, 33, 806–813. Konturek JW, Dembinski A, Konturek SJ et al (1998) Infection of Helicobacter pylori in gastric adaptation to continued administration of aspirin in humans. Gastroenterology, 114, 245–255. Kujubu DA, Fletcher BS, Varnum BC et al (1991) TIS10, a phorbol ester tumor promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J Biol Chem, 266, 12866–12872. Lacy ER and Ito S (1982) Microscopic analysis of ethanol damage to rat gastric mucosa after treatment with a prostaglandin. Gastroenterology, 83, 619–625. Laine L, Harper S, Simon T et al (1999) A randomized trial comparing the effect of rofecoxib, a cyclooxygenase 2-specific inhibitor, with that of
ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology, 117, 776–783. Langenbach R, Morham SG, Tiano HF et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Larsen KR, Jensen NF, Davis EK et al (1981) The cytoprotective effects of (+/7)-15-deoxy-16-a, b-hydroxy-16-methyl PGE1 methyl ester (sc29333) vs. aspirin-shock gastric ulcerogenesis in the dog. Prostaglandins, 21 (suppl), 119–124. Lee M and Feldman M (1992) Non-essential role of leukotrienes as mediators of acute gastric mucosal injury induced by aspirin in rats. Dig Dis Sci, 37, 1282–1287. Lee M, Aldred K, Lee E and Feldman M (1992) Aspirin-induced acute gastric mucosal injury is a neutrophil-dependent process in rats. Am J Physiol, 263, G920–G926. Levine RA, Kohen KR, Schwartzel EH and Ramsay CE (1982) Prostaglandin E2-histamine in interactions on cAMP, cGMP and acid production in isolated fundic glands. Am J Physiol, 242, G21– G26. Lichtenberger LM, Graziani LA, Dial EJ et al (1983) Role of surfaceactive phospholipids in gastric cytoprotection. Science, 219, 1327– 1329. Main IH and Whittle BJ (1973) The effects of E and A prostaglandins on gastric mucosal blood flow and acid secretion in the rat. Br J Pharmacol, 49, 428–436. Main IH and Whittle BJ (1975) Potency and selectivity of methyl analogues of prostaglandin E2 on rat gastrointestinal function. Br J Pharmacol, 54, 309–317. Majima M, Hayashi I, Muramatsu M et al (2000) Cyclooxygenase-2 enhances basic fibroblast growth factor-induced angiogenesis through induction of vascular endothelial growth factor in rat sponge implants. Br J Pharmacol, 130, 641–649. Maricic N, Ehrlich K, Gretzer B et al (1999) Selective cyclooxygenase-2 inhibitors aggravate ischaemia-reperfusion injury in the rat stomach. Br J Pharmacol, 128, 1659–1666. Maricic N, Ehrlich KH and Peskar BM (2001) Effect of the cyclooxygenase-1 inhibitor SC-560 on mucosal damage during ischemia-reperfusion in the rat stomach. Gastroenterology, 120, A152. Masferrer JL, Zweifel BS, Manning PT et al (1994) Selective inhibition of inducible cyclooxygenase-2 in vivo is antiinflammatory and nonulcerogenic. Proc Natl Acad Sci USA, 91, 3228–3232. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. McCarthy CJ, Crofford LJ, Greenson J and Scheiman JM (1999) Cyclooxygenase-2 expression in gastric antral mucosa before and after eradication of Helicobacter pylori infection. Am J Gastroenterol, 94, 1218–1223. Miller TA, Henagan JM and Robert A (1980) Effect of 16, 16-dimethyl PGE2 on resting and histamine-stimulated gastric mucosal blood flow. Dig Dis Sci, 25, 561–567. Mitchell JA, Akarasereenont P, Thiemermann C et al (1993) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697. Mizuno H, Sakamoto C, Matsuda K et al (1997) Induction of cyclooxygenase-2 in gastric mucosal lesions and its inhibition by the specific antagonist delays healing in mice. Gastroenterology, 112, 387– 397. Morham SG, Langenbach R, Loftin CD et al (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell, 83, 473–482. Muller P, Fischer N, Dammann HG et al (1981) Simultaneous addition of 16, 16-dimethyl-prostaglandin E2 prevents aspirin and bile salt damage to human gastric mucosa. Zeitschr Gastroenterol, 19, 373–376. Newman A, de Moraes-Filho JP, Philippakos D and Misiewicz JJ (1975) The effect of intravenous infusions of prostaglandins E2 and F2a on human gastric function. Gut, 16, 272–276. Nezamis JE, Robert A and Stowe DF (1971) Inhibition by prostaglandin E1 of gastric secretion in the dog. J Physiol, 218, 369–383. Ohno T, Ohtsuki H and Okabe S (1985) Effects of 16,16-dimethyl prostaglandin E2 on ethanol-induced and aspirin-induced gastric
STOMACH PHYSIOLOGY damage in the rat. Scanning electron microscopic study. Gastroenterology, 88, 353–361. Ohno R, Yoshinaga K, Fujita T et al (2001) Depth of invasion parallels increased cyclooxygenase-2 levels in patients with gastric carcinoma. Cancer, 91, 1876–1881. Peskar BM (1991) Role of leukotriene C4 in mucosal damage caused by necrotizing agents and indomethacin in the rat stomach. Gastroenterology, 100, 619–626. Peskar BM (2001) Role of cyclooxygenase isoforms in gastric mucosal defence. J Physiol Paris, 95, 3–9. Peskar BM and Peskar BA (1976) On the metabolism of prostaglandins by human gastric fundus mucosa. Biochim Biophys Acta, 424, 430–438. Peskar BM, Gu¨nther B and Peskar BA (1980) Prostaglandins and prostaglandin metabolites in human gastric juice. Prostaglandins, 20, 419–427. Peskar BM, Weiler H and Meyer CH (1984) Inhibition of prostaglandin production in the gastrointestinal tract by antiinflammatory drugs. In Rainsford KD and Velo GP (eds), Advances in Inflammatory Research, vol 6. New York: Raven, pp. 39–50. Peskar BM, Lange K, Hoppe U and Peskar BA (1986) Ethanol stimulates formation of leukotriene C4 in rat gastric mucosa. Prostaglandins, 31, 283–293. Peterson W, Feldman M, Taylor I and Bremer M (1979) The effect of 15(R)-15-methyl prostaglandin E2 on meal-stimulated gastric acid secretion, serum gastrin, and pancreatic polypeptide in duodenal ulcer patients. Dig Dis Sci, 24, 381–384. Pihan G, Majzoubi D, Haudenschild C et al (1986) Early microcirculatory stasis in acute gastric mucosal injury in the rat and prevention by or sodium thiosulfate. 16,16-dimethyl prostaglandin E2 Gastroenterology, 91, 1415–1426. Pihan G, Rogers C and Szabo S (1988) Vascular injury in acute gastric mucosal damage. Mediatory role of leukotrienes. Dig Dis Sci, 33, 625– 632. Rajnakova A, Moochhala S, Goh PM and Ngoi S (2001) Expression of nitric oxide synthase, cyclooxygenase, and p53 in different stages of human gastric cancer. Cancer Lett, 172, 177–185. Rask-Madsen J, Bukhave K, Hovendal CP and Bech K (1981) Release of prostaglandin E2 into gastric juice during stimulation of muscarinic and gastrin receptors in dogs and in humans. Prostaglandins, 21 (suppl), 57–62. Raz A, Wyche A and Needleman P (1989) Temporal and pharmacological division of fibroblast cyclooxygenase expression into transcriptional and translational phases. Proc Natl Acad Sci USA, 86, 1657–1661. Redfern JS, Blair AJ, Lee E and Feldman M (1987) Gastrointestinal ulcer formation in rabbits immunized with prostaglandin E2. Gastroenterology, 93, 744–752. Reele SB and Bohan D (1984) Oral antisecretory activity of prostaglandin E2 in man. Dig Dis Sci, 29, 90–93. Robert A, Nezamis JE and Phillips JP (1967) Inhibition of gastric secretion by prostaglandins. Am J Dig Dis, 12, 1073–1076. Robert A, Nezamis JE and Phillips JP (1968) Effect of prostaglandin E1 on gastric secretion and ulcer formation in the rat. Gastroenterology, 55, 481–487. Robert A, Schultz JR, Nezamis JE and Lancaster C (1976) Gastric antisecretory and antiulcer properties of PGE2, 15-methyl PGE2 and 16, 16-dimethyl PGE2. Intravenous, oral and intrajejunal administration. Gastroenterology, 70, 359–370. Robert A, Nezamis JE, Lancaster C and Hanchar AJ (1979) Cytoprotection by prostaglandins in rats. Prevention of gastric necrosis produced by alcohol, HCl, NaOH, hypertonic NaCl and thermal injury. Gastroenterology, 77, 433–443. Robert A, Nezamis JE, Lancaster C et al (1983) Mild irritants prevent gastric necrosis through ‘‘adaptive cytoprotection’’ mediated by prostaglandins. Am J Physiol, 245, G113–G121. Robert A, Bottcher W, Golanska E and Kauffman GL (1984) Lack of correlation between mucus gel thickness and gastric cytoprotection in rats. Gastroenterology, 86, 670–674. Romano M, Ricci V, Memoli A et al (1998) Helicobacter pylori upregulates cyclooxygenase-2 mRNA expression and prostaglandin E2 synthesis in MKN 28 gastric mucosal cells in vitro. J Biol Chem, 273, 28560–28563. Rush BD and Ruwart MJ (1984) The role of accelerated colonic transit in prostaglandin-induced diarrhoea and its inhibition by prostacyclin. Br J Pharmacol, 83, 157–159.
443
Salvati P and Whittle BJ (1981) Investigation of the vascular actions of arachidonate lipoxygenase and cyclooxygenase products on the isolated perfused stomach of rat and rabbit. Prostaglandins, 22, 141– 156. Santucci L, Fiorucci S, Giansanti M et al (1994) Pentoxifylline prevents indomethacin induced acute gastric mucosal damage in rats: role of tumour necrosis factor a. Gut, 35, 909–915. Sawaoka H, Kawano S, Tsuji S et al (1998) Cyclooxygenase-2 inhibitors suppress the growth of gastric cancer xenografts via induction of apoptosis in nude mice. Am J Physiol, 274, G1061–G1067. Schepp W, Heim HK and Ruoff HJ (1983) Comparison of the effect of PGE2 and somatostatin on histamine stimulated 14C-aminopyrine uptake and cyclic AMP formation in isolated rat gastric mucosal cells. Agents Act, 13, 200–206. Schepp W, Kath D, Tatge C et al (1989) Leukotrienes C4 and D4 potentiate acid production by isolated rat parietal cells. Gastroenterology, 97, 1420–1429. Schiessel R, Matthews J, Barzilai A et al (1980) PGE2 stimulates gastric chloride transport: possible key to cytoprotection. Nature, 283, 671– 673. Schmassmann A, Peskar BM, Stettler C et al (1998) Effects of inhibition of prostaglandin endoperoxide synthase-2 in chronic gastrointestinal ulcer models in rats. Br J Pharmacol, 123, 795–804. Shaw JE and Ramwell PW (1968) Inhibition of gastric secretion in rats by prostaglandin E2. In Ramwell PW and Shaw JE (eds), Prostaglandin— Symposium of the Worcester Foundation for Experimental Biology. New York: Interscience, 55–56. Shea-Donohue PT, Nompleggi D, Myers L and Dubois A (1982) A comparison of the effects of prostacyclin and the 15(S),15-methyl analogs of PGE2 and PGF2a on gastric parietal and non-parietal secretion. Dig Dis Sci, 27, 17–22. Shigeta J, Takahashi S and Okabe S (1998) Role of cyclooxygenase-2 in the healing of gastric ulcers in rats. J Pharmacol Exp Therapeut, 286, 1383–1390. Silverstein FE, Faich G, Goldstein JL et al (2000) Gastrointestinal toxicity with celecoxib vs. non-steroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. J Am Med Assoc, 284, 1247–1255. Skoglund ML, Gerber JG, Murphy RC and Nies AS (1980) Prostaglandin production by intact isolated gastric parietal cells. Eur J Pharmacol, 66, 145–148. Skoglund ML, Nies AS and Gerber JG (1982) Inhibition of acid secretion in isolated canine parietal cells by prostaglandins. J Pharmacol Exp Therapeut, 220, 371–374. Soll AH (1980) Specific inhibition by prostaglandins E2 and I2 of histamine-stimulated [14C]aminopyrine accumulation and cyclic adenosine monophosphate generation by isolated canine parietal cells. J Clin Invest, 65, 1222–1229. Soll AH and Whittle BJ (1981) Prostacyclin analogues inhibit canine parietal cell activity and cyclic AMP formation. Prostaglandins, 21, 353–365. Somasundaram S, Sigthorsson G, Simpson RJ et al (2000) Uncoupling of intestinal mitochondrial oxidative phosphorylation and inhibition of cyclooxygenase are required for the development of NSAIDenteropathy in the rat. Aliment Pharmacol Therapeut, 14, 639–650. Starlinger M, Jakesz R, Matthews JB et al (1981) The relative importance of HCO37 and blood flow in the protection of rat gastric mucosa during shock. Gastroenterology, 81, 732–735. Stroff T, Plate S, Ebrahim JS, Ehrlich KH et al (1996) Tachykinin-induced increase in gastric mucosal resistance: role of primary afferent neurons, CGRP, and NO. Am J Physiol, 271, G1017–G1027. Sun WH, Tsuji S, Tsujii M et al (2000) Induction of cyclooxygenase-2 in rat gastric mucosa by rebamipide, a mucoprotective agent. J Pharmacol Exp Therapeut, 295, 447–452. Szabo S, Trier JS and Frankel PW (1981) Sulfhydryl compounds may mediate gastric cytoprotection. Science, 214, 200–202. Takahashi M, Ota S, Shimada T et al (1995) Hepatocyte growth factor is the most potent endogenous stimulant of rabbit gastric epithelial cell proliferation and migration in primary culture. J Clin Invest, 95, 1994– 2003. Takahashi M, Ota S, Hata Y et al (1996a) Hepatocyte growth factor as a key to modulate antiulcer action of prostaglandins in stomach. J Clin Invest, 98, 2604–2611.
444
DIGESTIVE SYSTEM
Takahashi M, Ota S, Hata Y et al (1996b) Constitutive expression of hepatocyte growth factor may maintain the sheet construction of gastric epithelial cells through facilitating actin–myosin contractile system. Biochem Biophys Res Commun, 219, 40–46. Takahashi S, Shigeta J, Inoue H et al (1998) Localization of cyclooxygenase-2 and regulation of its mRNA expression in gastric ulcers in rats. Am J Physiol, 275, G1137–G1145. Takeuchi K, Ueki S and Okabe S (1986) Importance of gastric motility in the pathogenesis of indomethacin-induced gastric lesions in rats. Dig Dis Sci, 31, 1114–1122. Tarnawski A, Stachura J, Hollander D et al (1988a) Cellular aspects of alcohol-induced injury and prostaglandin protection of the human gastric mucosa. Focus on the mucosal microvessels. J Clin Gastroenterol, 10, S35–S45. Tarnawski A, Brzozowski T, Sarfeh IJ et al (1988b) Prostaglandin protection of human isolated gastric glands against indomethacin and ethanol injury. Evidence for direct cellular action of prostaglandin. J Clin Invest, 81, 1081–1089. Trautmann M, Peskar BM and Peskar BA (1991) Aspirin-like drugs, ethanol-induced rat gastric injury and mucosal eicosanoid release. Eur J Pharmacol, 201, 53–58. Uefuji K, Ichikura T and Mochizuki H (2000) Cyclooxygenase-2 expression is related to prostaglandin biosynthesis and angiogenesis in human gastric cancer. Clin Cancer Res, 6, 135–138. Vaananen PM, Meddings JB and Wallace JL (1991) Role of oxygenderived free radicals in indomethacin-induced gastric injury. Am J Physiol, 261, G470–G475. Vaananen PM, Keenan CM, Grisham MB and Wallace JL (1992) Pharmacological investigation of the role of leukotrienes in the pathogenesis of experimental NSAID gastropathy. Inflammation, 16, 227–240. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature (New Biol), 231, 232–235. Vane JR and Botting RM (1995) New insights into the mode of action of antiinflammatory drugs. Inflamm Res, 44, 1–10. van Rees BP, Saukkonen K, Ristimaki A et al (2002) Cyclooxygenase-2 expression during carcinogenesis in the human stomach. J Pathol, 196, 171–179. Wallace JL and McKnight GW (1990) The mucoid cap over superficial gastric damage in the rat. A high-pH microenvironment dissipated by nonsteroidal antiinflammatory drugs and endothelin. Gastroenterology, 99, 295–304. Wallace JL, McKnight GW, Keenan CM et al (1990a) Effects of leukotrienes on susceptibility of the rat stomach to damage and investigation of the mechanism of action. Gastroenterology, 98, 1178– 1186.
Wallace JL, Keenan CM and Granger DN (1990b) Gastric ulceration induced by non-steroidal antiinflammatory drugs is a neutrophildependent process. Am J Physiol, 259, G462–G467. Wallace JL, McKnight W, Miyasaka M et al (1993) Role of endothelial adhesion molecules in NSAID-induced gastric mucosal injury. Am J Physiol, 265, G993–G998. Wallace JL, McKnight W, Reuter BK and Vergnolle N (2000) NSAIDinduced gastric damage in rats: requirement for inhibition of both cyclooxygenase 1 and 2. Gastroenterology, 119, 706–714. Whittle BJ (1976) Relationship between the prevention of rat gastric erosions and the inhibition of acid secretion by prostaglandins. Eur J Pharmacol, 40, 233–239. Whittle BJR (1980) Actions of prostaglandins on gastric mucosal blood flow. In Fielding LP (ed.), Gastrointestinal Mucosal Blood Flow. Edinburgh: Churchill Livingstone, 180–191. Whittle BJ (2000) COX-1 and COX-2 products in the gut: therapeutic impact of COX-2 inhibitors. Gut, 47, 320–325. Whittle BJR and Boughton-Smith NK (1979) 16-Phenoxy prostacyclin analogues—potent, selective antiulcer compounds. In Vane JR and Bergstrom S (eds), Prostacyclin. New York: Raven, 159–354. Whittle BJR and Vane JR (1987) Prostanoids as regulators of gastrointestinal function. In Johnson LR (ed), Physiology of the Gastrointestinal Tract, 2nd edn. New York: Raven, 143–180. Whittle BJ, Boughton-Smith NK, Moncada S and Vane JR (1978a) Actions of prostacyclin (PGI2) and its product, 6-oxo-PGF1a on the rat gastric mucosa in vivo and in vitro. Prostaglandins, 15, 955–967. Whittle BJR, Moncada S and Vane JR (1978b) Biological activities of some metabolites and analogues of prostacyclin. In De Las Heras FG and Vega S (eds), Medicinal Chemistry Advances. Oxford: Pergamon, 141–158. Whittle BJ, Kauffman GL and Moncada S (1981) Vasoconstriction with thromboxane A2 induces ulceration of the gastric mucosa. Nature, 292, 472–474. Whittle BJ, Oren-Wolman N and Guth PH (1985) Gastric vasoconstrictor actions of leukotriene C4, PGF2a, and thromboxane mimetic U-46619 on rat submucosal microcirculation in vivo. Am J Physiol, 248, G580– G586. Whittle BJ, Lopez-Belmonte J and Moncada S (1990) Regulation of gastric mucosal integrity by endogenous nitric oxide: interactions with prostanoids and sensory neuropeptides in the rat. Br J Pharmacol, 99, 607–611. Wollin A, Soll AH and Samloff IM (1979) Actions of histamine, secretin, and PGE2 on cyclic AMP production by isolated canine fundic mucosal cells. Am J Physiol, 237, E437–E443.
Section Eight Nervous System
40 Perspectives and Clinical Significance of Arachidonic Acid Release, Action and Metabolism in the Nervous System Christopher D. Breder Bristol Myers Squibb Co., Wallingford, CT, USA
Sixteen years have elapsed since the last edition of this series and L.S Wolfe’s accurate observation that to construct an allencompassing review of the role of arachidonic acid in the CNS was likely to be beyond the capability of any single reviewer (Wolfe 1988). The intervening time has been marked by a progression from chemical characterization of mediators in nervous tissue, to one where the scientist has the capacity to amplify genes involved in defined physiological processes or knockout genes in specific cellular types within an organ. These technological advances, coupled with the meteoric rise of therapeutic antagonists of the cyclooxygenase-2 (COX-2) enzyme from the laboratory bench to the marketplace, have propelled the study of eicosanoids in the CNS out of a peripheral niche of inquiry into a field of investigation in its own right. These investigations have led to a better understanding of the biochemical and anatomical origin of arachidonic acid and its metabolites in the CNS, as well as putative physiological and pathophysiological functions in nervous tissue. Indeed, one of the most exciting concepts that relates to all of the eicosanoids is that each of these lipid mediators and enzyme systems seems to be involved in both physiological and pathophysiological processes. This summary will provide a brief perspective on work that has focused on the cellular sources of arachidonic acid and its metabolites in nervous tissue, and some potential roles of these lipids in the normal and diseased nervous system. The reader is directed toward several excellent reviews of investigations performed in the invertebrate model, since this work will not be covered in detail (Piomelli 1991, 1994a; Schwartz 1991; Volterra et al 1992). CELLULAR SOURCES OF ARACHIDONIC ACID AND ITS ACTIVITY IN NERVOUS TISSUE Arachidonic Acid Formation in Nervous Tissues Arachidonic acid (AA) may be formed de novo from the desaturation of fatty acids or from arachidonate cleavage from membrane phospholipids. There is little literature on the relative contribution of each to the total CNS pool of bioactive AA. Desaturation is dependent on D-5 and -6 desaturase enzymes. Messenger RNA for both has been detected in nervous tissue; however, a detailed anatomical localization in normal tissue has not been described (Cho et al 1999a, 1999b). These enzymes are The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
prevalent in the pre- and early postnatal brain and almost undetectable in aged brain, suggesting an involvement for these systems in CNS development. Phospholipase A2 is the primary enzyme responsible for the direct cleavage of arachidonate from membrane phospholipids in the CNS, although phospholipases C or D, working in concert with phosphatases or other lipases, can lead to arachidonate production. The cellular source and physiological role of each system remains an active issue of ongoing inquiry. The phospholipase A2 enzyme cleaves the centre ester from phospholipids. Members of this enzyme family have been classified in over 11 families (Cho et al 1999a, 1999b; Six and Dennis 2000). The cytosolic form has been the most extensively studied with respect to stimulus-coupled production of AA. This isoform has been immunohistochemically localized to both neurons and glia throughout the brain (Ong et al 1999). The neuronal populations were in diverse regions but were most densely distributed in the olivary nuclei, cranial nerve nuclei and cerebellum. Ultrastructural localization has demonstrated that cPLA2-IR has predominantly postsynaptic profiles, primarily dendritic. This parallels the COX-2 cellular staining profile observed at the light microscopic level (Breder et al 1995). The phospholipase C (PLC) family demonstrates a similar multiplicity of members, each isoform with its own unique cellular and regional distribution. Early studies described the distribution of PLC-I (now termed PLCb) and PLC-II (now termed PLCg) as being primarily localized to neuronal populations with scattered astrocytic profiles (Gerfen et al 1988). PLC-III (now termed PLCd) was primarily observed in astrocytes (Choi et al 1989). The subsequent identification of at least eight isoforms in the CNS explains the discrepancy between reports and underscores the need for a systematic and complete characterization of the distribution of PLC in the CNS (Martelli et al 1996). PLA2 and PLC sometimes serve independent, complementary and even redundant roles in the generation of AA in the CNS. PLA2 is coupled to the 5-HT2 receptor in the cortical effects of serotonin (Felder et al 1990), whereas AA is generated through PLC in the cerebellar transductions of NMDA stimulation (Lazarewicz et al 1990). These systems may be redundant or sequential in the case of cerebral anoxia. A sequential relationship is noted in the accumulation of AA in LTP, which is marked by an initial rise in PLA2 activity, followed by PLC (Clements et al 1991). Both systems play distinct yet complementary roles in producing messengers involved in the elicitation of hyperalgesia
448
NERVOUS SYSTEM
by bradykinin. PLA2 generates AA used in PGE2 synthesis, while PLC is involved in the generation of prostacyclin (Taiwo et al 1990). Each paradigm will require careful pharmacological and temporal analysis to dissect the participation of these pathways. Phospholipase D (PLD) is involved in a complex AA synthetic pathway involving the subsequent action of a phosphatidic phosphohydrolase and diacylglycerol lipase (Ishimoto et al 1994). Two isoforms have been identified in brain with different cellular and, in the case of the neuronal populations, regional localization. An immunohistochemical study reported the distribution of PLD1-immunoreactive (-IR) neurons in discrete nuclei involved in diverse physiological roles (Lee et al 2000). Presumptive PLD-IR astrocytes were scattered over white matter in a pattern similar to that of GFAP-IR cells. A recent study of hybridization of the PLD1 and PLD2 mRNA suggested that the isoforms are developmentally regulated (Saito et al 2000). These authors concluded that the PLD1 was observed in ‘‘presumptive oligodendrocytes’’ and PLD2 in astrocytes. Their resolution of cellular identity does not allow for accurate comparison with the previous immunohistochemical study to resolve apparent conflicts in localization. Several diverse lines of research, including the demonstration of PLD coupling to metabotropic glutamate receptor signalling (Shinomura et al 2000) and the suppression of PLD activity under depolarizing conditions (Waring et al 1999), suggest that PLD generation of AA may play important physiological roles in the CNS. Arachidonic Acid Activity in Nervous Tissue AA has been implicated in numerous physiological roles within nervous tissue. Several excellent reviews present these studies in detail (Bazan 1989; Katsuki and Okuda 1995; Piomelli 1991, 1994b). These include AA activation of protein kinase C and phospholipase C and AA inhibition of Ca2+-calmodulin-dependent protein kinase II. Arachidonate has also been shown to directly depress the voltage-dependent, striatal Na+ current channel and the cortical GABAA receptor and directly enhance the cerebellar NMDA receptor-stimulated current and the inwardly rectifying K+ channel (Liu et al 2001). Receptorstimulated release of AA has been demonstrated for several muscarinic and serotonergic receptors, as well as mGluR1 and the SSTR4 somatostatin receptors; see Katsuki and Okuda (1995) for a comprehensive review. One of the original indications that AA might directly serve some physiological role in nervous tissue was through exploration of the mechanism of long-term potentiation (LTP), a synaptic phenomenon that has been proposed to be a model of learning. Evidence of AA’s role in LTP includes the suppression of LTP by inhibitors of PLA2 (Linden et al 1987) and the accumulation of AA following its initiation (Lynch et al 1989). Current evidence suggests that AA may actually act in concert with other stimuli to enhance LTP or that an AA metabolite mediates LTP. A detailed examination of this body of literature, with the benefit of our current understanding of AA metabolism, reveals many of the potential pitfalls in entertaining a hypothesis that AA might directly be involved in neurophysiological processes. These complicating issues have been most eloquently addressed in the report of Liu et al (2001), where the effects of AA on the human inwardly rectifying channel were studied. In this report, free radical generation, non-specific effects of fatty acids, changes in membrane fluidity and indirect effects from cyclooxygenase, lipoxygenase or cytochrome P450 metabolites were examined as factors that might confound the direct association of AA in the phenomenon of LTP. The authors astutely recognized possible effects from AA metabolites whose synthesis might not have been inhibited, such as from the ananamides. The lack of a specific
identified AA receptor or other unique binding site on certain channels, receptor or enzymes will continue to confound this issue.
PROSTAGLANDIN SYSTEMS IN NERVOUS TISSUE Cyclooxygenase Systems Prostaglandin endoperoxide H2 synthase or ‘‘cyclooxygenase’’ (COX) enzyme family converts AA to prostaglandin H2 (PGH2) (Smith et al 2000). Although early immunolocalization studies of ‘‘the cyclooxygenase’’ in the CNS were carried out without the knowledge of two distinct gene products, COX-1 and COX-2, there was speculation that multiple forms existed (Lysz and Needleman 1982). Another hypothesis that was energetically debated, even until recently, was that prostaglandins were utilized in distinct pathways within nervous tissue and were not likely to be simple ubiquitous autocoid or intracellular messengers. Some of the first data supporting this hypothesis emanated from anatomical studies of the PG synthesis machinery and binding sites in nervous tissue, e.g. receptor binding of PGE2 in the monkey diencephalon (see section on Receptor Systems in the CNS, below) revealed intense labelling within discrete nuclei (Watanabe et al 1988). COX-immunoreactivity in the sheep brain was observed in neurons of numerous distinct regions (Breder et al 1992), yet was clearly not observed in most of the cells within the CNS, as might have been expected. The principal cell type observed in these studies was neuronal, another unexpected observation, given the prevailing hypothesis that glia were the predominant source of prostaglandins in the CNS. The relative availability of good COX-2 specific antisera and the popular conception of COX-2 as being dynamically expressed in response to processes such as inflammation and nociception have contributed to a more extensive characterization of COX-2 in the CNS than COX-1. At the time of COX-2’s discovery, the expression of an ‘‘inducible form’’ of COX in the basal CNS was believed unlikely. Nevertheless, COX-2 expression has also been established in normal stomach and kidney tissue (O’Neill and Ford-Hutchinson 1993). In retrospect, it is of little wonder that the regulated COX-2 isozyme is expressed in these organs and in the CNS, as neither of these tissues is ever truly at rest, both are constantly integrating and responding to sensory input from their external and internal milieux. Studies describing the initial cloning of COX-2 gave evidence of mRNA or protein in brain (Xie et al 1991). Yamagata et al (1993) provided the first detailed investigation of the presence of COX-2 in the CNS. These investigators discovered that the ‘‘mitogen-inducible’’ isoform of COX was dramatically induced in a model of seizure. In this model, expression of the message was limited to telencephalic structures such as the cortex, hippocampus and amygdala. COX-2 protein in the basal state of the rat CNS was subsequently described in detail (Breder et al 1995). Only neurons were observed in brains from the basal state. The distribution of neurons in the telencephalon was identical to that described for the mRNA. The increased resolution and sensitivity of the immunohistochemical technique allowed for the identification of a few small cell groups in the diencephalon and brainstem. Improvements in the fixation technique have allowed the identification of COX-2 in astrocytes, where the expression seems to be linked to CNS inflammatory processes. Overall, the neuronal expression of COX-2 seems to be most robust in areas involved in complex integration of multiple sensory modalities, such as the hippocampus and associated cortical fields (Breder et al 1995). Complex neuropsychological testing will be needed to elucidate the true role of basally expressed COX-2 in the CNS.
ARACHIDONIC ACID IN THE NERVOUS SYSTEM Much of the literature regarding the role of COX-2 in the CNS has centred on the generation of fever and the mechanism of nociception, with more recent attention focusing on the neuropathology of ischaemic and excitotoxic injury. The central debate on the role of COX-2 in the generation of fever involves the cellular source of the enzyme in brain, which is the target of the antipyretics, such as aspirin. Compelling data based on cellspecific markers has suggested that this enzyme is induced in microglia (Elmquist et al 1997; Levi et al 1998; Minghetti and Levi 1995) or perivascular endothelial cells (Cao et al 1996) after immunogenic stimuli, such as infusion of lipopolysaccharide or cytokines. One study of COX-2 expression during a model of fever reported the induction of COX-2 mRNA in the paraventricular nucleus of the hypothalamus, where neuronal staining has been described under basal conditions (Lacroix and Rivest 1998). It is likely that the specific cellular COX-2 response is dependent on the chemical composition and the magnitude of the stimulus (Schiltz and Sawchenko 2002). Cell-type specific COX-2 knockout experiments will likely be needed to adequately address the role of COX-2 in these different cell types (Sauer 1998). Investigations into the role of COX-2 in nociception have contrasted the functions of the two isoforms. These systems are most readily differentiated by their distinct anatomical localization. COX-1 is the predominant form in DRG neurons, while COX-2 is observed in the dorsal horn of the spinal cord (Willingale et al 1997). Inhibition of COX-2 seems particularly effective in the reduction of allodynia and thermal hyperalgesia, but not mechanical hyperalgesia. Commensurate with the proposed synthesis of COX-2 in the dorsal horn, this pharmacological effect seems to be occurring at the level of the spinal cord. Systemic inhibitors, with relative selectivity for COX-1, reduce thermal hyperalgesia and visceral nociception (Goto et al 1998; Kusuhara et al 1998). Formulation of more specific inhibitors and standardization of nociceptive stimuli and assays will greatly reduce the variability in data and its interpretation in this line of inquiry. Another major path of inquiry has been into the neuropathology associated with COX-2 activity. This has been studied using mice transgenic for human COX-2 (Andreasson et al 2001) or by investigating the mechanisms of excitotoxicity of glutamate receptor agonists (Corasaniti et al 2000; Hewett et al 2000). Andreasson et al recently provided data linking overexpression of COX-2 with neuronal apoptosis and behavioural deficits of ageing. Apparently, although COX-2 may be involved in excitatory neurotransmission in parts of the brain crucial to cognition, too much activity is associated with cell death. A consistent portrayal of the basal distribution of COX-1-IR in the CNS remains elusive. Reports of the distribution of COX-1 have varied from discrete populations of neurons in the CA3 and CA4 regions of the hippocampus and layer 4 of the cortex (Yermakova et al 1999) to accounts of scattered neurons and ependymal cells. Descriptions of the distribution of COX-1 in the spinal cord have varied as well, possibly as a consequence of the specificity of the antisera used in these studies. This issue may ultimately be resolved through the combination of in situ hybridization and immunohistochemistry with cell-specific markers. A more consistent observation has been that COX-1 seems to be the predominant COX isoform observed in DRG neurons (Willingale et al 1997). Although the discovery of COX-2 seemed to relegate COX-1 into the role of ‘‘housekeeping gene’’, a role for COX-1 in dynamic processes is only now unfolding. One reported activity is the maintenance of cerebral blood flow and the increase in CBF in response to specific stimuli. This activity may be important in physiological responses to normal stimuli, such as hypercapnia, and may be vital in limiting infarct volume during ischaemic episodes (Iadecola et al 2001b). Many pathophysiological insults
449
to the CNS lead to an induction of expression of COX-1 in microglia (Deininger et al 2000; Deininger and Schluesener 1999; Schwab et al 2001). It is unclear whether prostaglandins synthesized in this scenario are mediators of cellular repair or are harbingers of tissue injury mediated through free radical mechanisms. An emerging role for COX-1 appears to be as a mediator in nociception. COX-1 is the primary isoform localized to neurons in the DRG (Chopra et al 2000; Willingale et al 1997). These are CGRP-IR cells that serve in nociception (Iadecola et al 2001b; Omana-Zapata and Bley 2001). COX-1 is also coupled to the generation of prostacyclin, which underlies the increase in ectopic activity of sensory neurons after neuropathic injury. Ultimately, it is likely that both COX-1 and COX-2 are constitutively expressed in the CNS, although differentially regulated in distinct cell types and in unique conditions. Prostaglandin/Thromboxane Synthase Systems Prostaglandin/thromboxane synthases metabolize PGH2, the product of COX, into active mediators, including PGD2, PGE2, PGF2, PGI2 (prostacyclin) or TXA2 (thromboxane). The most extensive work in CNS has focused on PGD synthase (PGDS), perhaps because of the relative quantity of its metabolite in tissues. Glutathione-dependent and independent (lipocalin type; b trace) isoforms of PGDS are known to exist in the nervous tissue, with the latter form predominating in the basal state (Beuckmann et al 2000; Hayaishi 1983; Vesin et al 1995). The glutathioneindependent form is thought to be developmentally regulated, in that PGDS-immunoreactivity is present in cortical layer 1 and 2 neurons in rat pups and oligodendroglia by 8 weeks. In the adult rat brain, glutathione-independent PGDS is primarily observed in various cells of the arachnoid, meningial and perivascular microglia and oligodendrocytes. The PGDS-immunoreactivity in each of these cells, with the exception of oligodendroglia, colocalizes with COX-2. PGDS-immunoreactivity is observed in cortical layer 1 and 2 neurons, however; mRNA is not detectable in this cell type. In the spinal cord, PGDS-immunoreactivity is observed in neurons of the superficial laminae and motor neurons. The glutathione-dependent and independent isoforms are differentially distributed to Schwann cells and small B1 neurons, respectively, within the dorsal root ganglia. Functionally, the PGD metabolite appears to be critical for regulation of the sleep– wake cycle (Hayaishi 2000) and may be involved in the spinal nociceptive process of tactile allodynia (Eguchi et al 1999). It seems likely that more isoforms of PGD synthase are present to subserve these expanding roles of PGD as a physiological mediator in the CNS. A synthase for PGE (PGES) has only recently been cloned and found to be differentially distributed in regions of the human brain (Jakobsson et al 1999). This molecule is highly inducible by immunogenic stimuli, such as IL-1b. A homologue was isolated from rat astrocytes that were stimulated with b-amyloid, providing a possible insight into the beneficial effects of COX inhibitors in Alzheimer’s disease (Satoh et al 2000) (see section on Pathophysiological Roles, below). This form of PGE synthase is also induced and co-expressed with COX-2 in the rodent brain after immunogenic stimulation. (Yamagata et al 2001). A novel, second cytosolic form of PGES has been cloned from rat brain that is functionally coupled to COX-1 (Tanioka et al 2000). LPS highly stimulates expression of the cPGES. Cellular distribution of this enzyme has not been described, although, given its regulation, it is likely to be found in microglia. It seems certain, given the variety of roles that PGE plays in the CNS, that more isoforms may be discovered or perhaps, like COX-2, PGES may be expressed in specific types of cells depending on the situation.
450
NERVOUS SYSTEM
Much less is known about the presence of thromboxane synthase in brain. TXB2, the stable metabolite of TXA2, can be detected in enriched cultures of microglia but not astrocytes. Furthermore, thromboxane synthase-IR has been detected in cultured microglia (Giulian et al 1996). A detailed examination of the distribution of this enzyme in intact tissue has not been reported. Eicosanoid Receptor Systems in the CNS; Comments on Their Physiology Considerable progress has been made on the identification of eicosanoid receptor systems in the CNS. Prostaglandin binding proteins have received specific focus because of concurrent advances in molecular biology and the interest in designing receptor-specific ligands. Prostaglandin E receptors, termed EPs, have been most extensively characterized in CNS tissue. In normal rodent brain, EP1 mRNA has been most densely observed in the paraventricular hypothalamic (PVH) and supraoptic nuclei and in scattered cells of the preoptic and tuberal regions of the hypothalamus (Batshake et al 1995; Oka et al 2000). EP2 mRNA has been primarily observed in the bed nucleus of the stria terminalis, lateral septal nucleus, subfornical organ, ventromedial hypothalamic nucleus, locus coeruleus and area postrema (Oka et al 2000; Zhang and Rivest 1999). Scattered cells hybridizing the EP2 mRNA are observed in the dorsal and ventral horns of the spinal cord (Kawamura et al 1997). EP3 exists as a family of splice variants, EP3a, EP3b and EP3g. The mRNA and protein-IR have been mapped in detail within the rodent CNS and show convincing overlap in their distribution (Ek et al 2000; Nakamura et al 2000; Oka et al 2000; Sugimoto et al 1994). Profuse labelling of neuropil in distinct nuclei and cortical regions is one of the most striking features of this staining pattern. This labelling is most prominent in layer 5 of the neocortex, extending to more superficial laminae in the periallo- and allocortical regions, medial preoptic area, dorsomedial hypothalamic nucleus, the midline, intralaminar and anterior thalamic nuclei, throughout the deep and anterior cortical nuclei of the amygdala, the parabrachial nucleus, nucleus of the solitary tract, and in several medullary catecholamine groups. In the spinal cord, lamina 2 is most heavily labelled. The cellular expression of the EP3 receptor is primarily in neurons, with additional dense labelling of ventricular ependymal cells. In many respects, particularly in the telencephalic structures, EP3 staining is complementary to the immunostaining of COX-2IR in normal rat brain. PGE released from these COX-2 neurons might activate these EP3 receptors in some sort of retrograde, or feedback loop of, neurotransmission. EP4 mRNA is most densely distributed in the BST, ventral septal/POA, magnocellular portion of the PVH, SON, PB, NTS and caudal ventral lateral medulla (cVLM). Numerous research strategies have been applied to elucidate the roles of EPs in neurophysiological systems. The primary foci have been on the mechanisms associated with nociception and in the generation of the febrile response (see Chapters 49 and 50). Each EP seems to have a unique role in the transmission of nociceptive signalling. While all of these receptor subtypes are found in dorsal root ganglia (DRG) and spinal cord, they have a unique regulation and profile of activities with respect to inflammatory and nociceptive stimuli. The development of increasingly specific ligands and the utilization of EP-specific knockout mice have been used in studies suggesting that EP1 plays a predominant role in allodynia, while EP2 and EP3 participate in hyperalgesia (Kawahara et al 2001; Minami et al 1994, 2001b). A recent study also suggests that EP1 may play an essential role in the hyperalgesia associated with pain in the postoperative scenario
(Omote et al 2001). There has been a growing consensus that PGE2 sensitizes sensory neurons to other signals of nociception. EP2 agonists are more similar to PGE2 than other subtypes in the sensitization of spinal lamina 3 and 4 neurons, while EP3 and EP4 seem to be involved in PGE-sensitization in the DRGs (Ahmadi et al 2002; Kumazawa et al 1996; Southall and Vasko 2001). The development of more specific ligands and functional assays will yield valuable insights into the eicosanoid-dependent mechanisms of nociception in the coming years. Considerable progress has been made with respect to the elucidation of central pathways mediating the acute phase response (APR) to infection, also known as the febrile response. This event is characterized by a resetting of the thermal set point resulting in the generation of fever, an elevation of plasma glucocorticoids, and onset of sleepiness or lethargy (Kawamura et al 1997; Saper et al 1994). The CNS orchestrates many components of this response through its detection of circulating cytokines, such as interleukin-1b (IL-1b). The APR has been modelled by infusions of various pyrogens by different routes, including intravenously, intramuscularly or intracerebroventricularly. There are few studies directly investigating the role of EP systems in the generation of fever during the APR. A pharmacological study suggested that EP1 receptors were involved in the early component of the increased temperature response to IL-1b injected into the cerebral ventricles (Oka et al 1998). A recent report examining the generation of fever in knockout mice that lack specific EPs suggests that the EP3 receptor is the critical PGsensitive protein (Ushikubi et al 1998), although the EP47/7 mice could not be adequately analysed because of low viability. Several studies have combined the localization of EPs with the identification of activated cells expressing the cFOS oncogene (Ek et al 2000; Oka et al 2000; Zhang and Rivest 1999). After immunogenic stimuli, activated (cFOS-IR) cells expressing EP1 were diffusely scattered throughout the preoptic and anterior periventricular area of the hypothalamus. EP2 message is induced in the BST, CEA, VMH, VLM and in the circumventricular organs, SFO and AP. After i.v. infusion of IL-1b, cFOS-IR staining appeared within cells in the MPO, VMPO, mNTS and rVLM. Interestingly, the level of EP3 hybridization did not appear to rise throughout the brain after this stimulus. Relatively few cFOS-IR contain EP3 mRNA after i.v. LPS. In contrast, relatively large numbers of cFOS-IR cells in the AV3V and POA express the EP4 mRNA. Of particular interest is the expression of EP4 mRNA in the parvocellular or neuroendocrine portion of the PVH and in the cVLM, where the A1 catecholaminergic neurons are found. These cell groups may carry PG-sensitive signals, underlying the central stimulation of plasma glucocorticoids following immunogenic stimulation (Zhang and Rivest 2000). EP4 mRNA is also induced in cells in the dorsal parvocellular PVH, which project to the spinal cord and may be involved in sympathetic regulation during the APR. The distribution of COX-2-IR neurons suggests that PGs may be involved in the regulation of several components of behaviour (Breder et al 1995). Functional analysis has confirmed this hypothesis, demonstrating EP-specific effects on the modulation of sensory integration in the telencephalon and control of the sleep–wake cycle. In cortex, norepinephrine and peptides, such as vasoactive intestinal peptide, cause a synergistic elevation of cAMP in a manner that is prostaglandin-dependent (Schaad et al 1987). Cortical application of PGE2 mimics this response. On the other hand, PGE2 inhibits norepinephrine release from cortex via the EP3 receptor and may dramatically potentiate the inhibitory properties of other putative transmitters, such as histamine (Exner and Schlicker 1995; Schlicker and Marr 1997). It seems feasible that PGE2, through unique EP systems, is involved in a feedback loop in cortical signalling. Further elucidation of EP types involved in the transduction of norepinephrine signals may
ARACHIDONIC ACID IN THE NERVOUS SYSTEM provide valuable information on the processing of sensory information in the CNS. EP-specific ligands have also been used to dissect the complex regulation of the sleep–wake cycle. Activation of EP4 receptors near the basal aspect of the third ventricle enhances sleep, while EP1 and EP2 receptors adjacent to the third ventricle increase wakefulness (Exner and Schlicker 1995; Schlicker and Marr 1997; Yoshida et al 2000). It would be of interest to determine whether these EP4-dependent effects are mediated by receptors in the ventral preoptic area of the hypothalamus, which has been hypothesized to be the site of a ‘‘sleep switch’’ (Saper et al 2001). CNS distribution of the receptor for PGD2, termed DP, parallels PGD synthase in that both are localized to the meninges and the dorsal and ventral horns of the spinal cord (Gerashchenko et al 1998; Mizoguchi et al 2001). A contemporary hypothesis on the role of PGD2 in sleep is that activation of DP leads to the production of some diffusable mediator that acts upon the ventral lateral preoptic nucleus, a region proposed to function as a switch in the onset of sleep (Saper et al 2001; Sherin et al 1996). The signal might then be relayed through the diffusely projecting histaminergic neurons of the tuberomammilary nucleus, to the entire cerebral cortex. The DP system in the spinal cord is presumably also involved in the role of PGD2 in producing tactile allodynia (Minami et al 1994b). The multitude of roles for PGD2 as a mediator in the CNS suggests that a more extensive system of synthases and receptors has yet to be described. Much less is known about the distribution and function of PGF2a in the CNS. Ligand binding studies suggest that this eicosanoid may act upon cerebral vasculature, where it probably acts as a vasoconstrictor (Li et al 1997). Resolution of its tissue distribution is limited to Northern blot analysis of whole brain and cultured astrocytes. Preliminary pharmacological evidence suggests that PGF2a may have similar roles to those of PGE2 in neuroendocrine regulation (Li et al 1997; Sutarmo et al 1998) and nociception (Minami et al 2001a). Ligand binding studies suggest the presence of two receptors for PGI2, also known as prostacyclin. The PGI agonist iloprost has binding sites in the caudal medulla, including the nucleus of the solitary tract and the spinal trigeminal nucleus (Matsumura et al 1995; Minami et al 2001a; Oida et al 1995). PGI2 is likely to be involved in autonomic regulation at these brainstem sites. The prostacyclin receptor (IP) cloned from the CNS has a similar distribution of expression to the pattern of binding described for iloprost but is also observed in cerebral arterial walls and DRG neurons. A second IP receptor has been suggested by the presence of binding activity of the prostacyclin agonist isocarbacyclin (Takechi et al 1996) in the telencephalon, thalamus and basal ganglia. Both iloprost and isocarbacyclin sites seem to be involved in the neuroprotective effects of prostacyclin following ischaemia. Prostacyclin also seems to be involved in specific components of neuropathic pain through its iloprost-sensitive binding sites (Matsumura et al 1995; Minami et al 2001a; Oida et al 1995; Omana-Zapata and Bley 2001; Takechi et al 1996). The development of more specific agonists or antagonists will likely lead to better understanding of the role of PGI in the CNS. Thromboxane receptors (TP) have been detected in cultured neurons and astrocytes and purified from oligodendrocytes and Schwann cells (Blackman et al 1998; Kitanaka et al 1995; Miggin and Kinsella 1998). More sensitive examination of the mRNA level indicates that cells of all glial lineages, including astrocytes, oligodendroglia and microglia, express TP. Immunocytochemical localization of TP suggests that at least one isoform is distributed throughout fibre tracts in the CNS. Certain neurophysiological systems utilize TX as a modulator of synaptic transmission. N-Methyl-D-aspartate (NMDA)-
451
mediated release of TX has been reported in both the hippocampus and the hypothalamus (Lazarewicz et al 1992; Yokotani et al 2001). In the hypothalamus, this has been associated with an increase in central sympathetic output, suggesting that the TXA2 system is acting upon the cell groups projecting to the spinal sympathetic preganglionic neurons. In the clinical arena, TX may be involved in the sequelae of cerebral ischaemia and vasospasm, possibly through the generation of free radicals or via a reduction of cerebral blood flow (Fisher 1993). Whether platelets or the damaged neural tissues generate the eicosanoid has not been resolved. A growing consensus on this matter suggests that the most beneficial scenario is one where the ratio of PGI:TX is maximized. LIPOXYGENASE SYSTEMS IN THE CNS The multitude of newly discovered lipoxygenase (LO) enzyme pathways and newfound roles for LO metabolites promise to make investigations within this aspect of eicosanoid neurobiology truly rewarding. The LO system comprises several families of enzymes (most notably the 5-, 12- and 15-LO) that stereo- and regiospecifically incorporate a hydroperoxy moiety into arachidonic acid, forming hydroxyperoxyeicosatetraenoic acids (HPETEs). This product may be reduced to a hydroxyeicosatetraenoic acid (HETE) or the HPETE may be metabolized to any of a variety of lipids that are LO family-specific. It is unclear whether the HPETE metabolites have inherent activity or, like PGH2 of the cyclooxygenase system, may serve as unstable intermediates for further transformation. Hwang et al (2000) have recently demonstrated that 12- and 15-HPETE were more potent than downstream 12- and 15-LO metabolites in activating the capsaicin receptor, VR1, suggesting that the HPETEs may have intrinsic bioactivity. The 12 LO family gives rise to a 12-HPETE that may be reduced to the corresponding HETE or metabolized into a ketoeicosatetraenoic acid (12-KETE) or a member of the hepoxilin subfamily. 12-KETE has been identified in the nervous tissue of Aplysia and may be a second messenger in histaminergic transmission (Piomelli et al 1989). A hepoxilin synthase metabolizes 12-HPETE into hypoxilin A3 and B3. These molecules have been detected in mammalian hippocampal tissue and have been shown to have inhibitory effects on both pre- and postsynaptic aspects of neurotransmission (Carlen et al 1989, 1994; PaceAsciak et al 1990b).The hepoxilins may be further metabolized into trioxilins or glutathionyl derivatives in neural tissue (PaceAsciak et al 1990a, 1990b, 1991). The roles for these metabolites have yet to be fully characterized. The apparent conflict regarding the cellular source of 12-LO metabolites in mammalian nervous tissue may stem from the multiplicity in this enzyme family. At least three mammalian 12LOs exist, the platelet, leukocyte and epidermal isoforms. An immunohistochemical survey in the canine brain detected 12-LOIR cells of neuronal and various glial cell types, while assays of several cerebral cell-type cultures have detected 12-LO activity (Nishiyama et al 1993). Kawajiri et al (2000) have investigated the distribution of leukocyte 12-LO-IR in rat tissues. This isoform has been observed in neurons of the spinal cord and superior cervical ganglion. Reagents that are isoform-specific must be used to characterize the specific 12-LO system in future experiments. Some of the most detailed characterizations of the LO metabolites have been with respect to their effects on vascular tone. The (S)-enantiomer of 12-HETE is produced by cerebral vessels and is a mediator of vasodilation, hyperpolarizing vascular muscle through its activation of a Ca2+ activated K+ channel (Faraci et al 2001). This activity contrasts with the vasoconstrictor 20-HETE, which is synthesized by the monooxygenation of AA
452
NERVOUS SYSTEM
by the cytochrome P450 4A subfamily (Medhora et al 2001) (see next section). From the 5-LO family, 5-HPETE may be metabolized further to one of the leukotrienes (LTA4-E4). Activation of 5-LO requires the activity of 5-lipoxygenase activating protein (FLAP; Lammers et al 1996). Both are localized throughout the brain in a distribution that suggests that they are co-expressed. The cellular profile of 5-LO-IR appears to change with ageing. In the young rat brain, neuronal 5-LO staining in amygdaloid neurons and cortical and hippocampal pyramidal cells is confined to the cell bodies, while in aged rats a luxuriant 5-LO-IR is observed in the apical dendrites (Manev et al 2000). This has been proposed as an anatomical substrate for the neurodegenerative process associated with ageing. The 5-LO leukotriene system was in fact one of the first active eicosanoid mediators studied in the brain because of its neuroendocrine activity. An LTC4 peptidergic system co-localizes with LHRH-IR in the anterior hypothalamus. LTC4, 5-HETE and 12-HETE can elicit LHRH release from the median eminence (Gerozissis et al 1985; Saadi et al 1990). A cholinoceptive 8-(S)-LO system has been described in nervous tissue of Aplysia (Tieman et al 1997). Interestingly, an 8-(R)-LO system has recently been cloned from mouse skin and its message appears to be relatively highly expressed in brain (Jisaka et al 1997). Further work differentiating these enzymes and the activities of their stereo- and regiospecific metabolites will need to be performed. Little is known about the 15-LO system in neural tissue. In addition to the formation of 15-HPETE, the 15-LO system acts in concert with 5-LO to form lipoxins, which have been proposed to act as neuronal second messengers in the activation of protein kinase Cg (Shearman et al 1989) at the molecular level, and in the regulation of the sleep–wake cycle (Sri et al 1994) at the physiological level. CYTOCHROME P450 SYSTEMS IN THE CNS One of the most significant areas of investigation in recent eicosanoid research has been in the role of the cytochrome P450 (CYP) mediated metabolism of AA. The CYPs are a superfamily of over 250 enzymes that are usually associated with the transformation of xenobiotics. A growing subpopulation of this group, viz the CYP1 and CYP2 families, are recognized for their capacity to monooxygenate AA, forming epoxyeicosatrienoic acids (EETS), or the CYP4A subfamily, for the ability to o- or o1-hydroxylate AA to form hydroxyeicosatetraenoic acids (HETES). Other fatty acids and their derivatives, including prostaglandins, may serve as substrates for these enzymes, possibly forming important physiological mediators in the CNS. The action of these CYPs upon AA is contingent on the activity of an NADPH CYP oxidoreductase co-enzyme (NCPO), which serves in electron transfer to the NADPH co-factor. This concept lends particular importance to the study of Norris et al (Norris et al 1994), who provided a detailed description of the distribution of NCPO in the rat CNS. The enzyme was primarily observed in neurons with a few glial populations. The regional neuronal distribution was a complex pattern of sensory, autonomic, cognititive and other areas involved in extrapyramidal motor function. This map may represent the basal distribution that is likely to change in terms of both cell type and regional localization, given a particular physiological or pathophysiological perturbation. The map gives a good collective approximation of the immunoreactivity described for the numerous CYP subfamily members in the rat CNS. Several studies have reported on CYP activity, mRNA and protein content and cellular localization. Few have addressed these issues with comprehensive examinations throughout the
entire intact CNS for any given physiological state or model of induction. Inhibition of the NCPO enzyme seems to have most effect on brain microsomal monooxygenase activity but little effect on that of mitochondrial monooxygenase activity, suggesting diverse functional pools of CYPs in the CNS (Bhagwat et al 1995). Selective regional immunohistochemical examinations of specific family members of the CYP2C, CYP2D and CYP2E subfamilies have demonstrated CYP-immunoreactive neuronal populations, particularly in the telencephalon, basal ganglia, substantia nigra and other regions involved in the motivational aspect of motor activity. One of the seminal studies localizing the CYP family in CNS tissue identified CYP2C11 message, protein and activity in cultured astrocytes (Alkayed et al 1996). This has been proposed to be a source of ‘‘endothelium-derived relaxing factor’’ for cerebral vasculature. A study of CYP2C11-immunoreactivity with respect to its distribution in the basal state or during increased cerebral perfusion (intracranial hypertension) would be of great value. The EETs were first identified functionally as neuroendocrine mediators in the rat CNS, most likely serving as second messengers in the dopaminergic stimulation of somatostatin from the hypothalamic median eminence. Probably the largest body of literature on the physiological (and pathophysiological) role of the CYP metabolites of AA has focused on the role in cerebrovascular regulation. The EETs are primarily involved in vasodilation through their action on a calcium-activated K+ channel, while 20-HETE has vasoconstrictor activity through inhibition of this same channel (see Medhora et al 2001 for an excellent review of this topic). Subdural perfusion of miconazole, an inhibitor of CYP epoxygenases, results in a decrease in blood flow suggesting that the EETs are tonically active in cerebrovascular autoregulation. A current hypothesis proposes that the HETEs and EETs are critical mediators in the coupling of cerebral metabolic activity to increased blood flow. This prospect holds great promise with respect to neuroprotection in the clinical arena, where surgical procedures that place the CNS at risk of ischaemia, such as cerebral aneurysm clipping and coronary artery bypass grafting, might benefit from the use of therapeutics specific to these systems. A small body of literature has also focused on the role of epoxygenase metabolites in the CNS control of the febrile or acute phase response to infection. AA metabolites of the CYP system seem to be functioning as downregulators of this response by reducing pyresis and acting as hypothalamic messengers in the cytokine induction of ACTH release. The recent demonstration of epoxygenase induction in response to starvation (Oleksiak et al 2000) and ethanol exposure (Tindberg and Ingelman-Sundberg 1996) (to name but a few studies) suggests that the metabolism of AA by CYPs is involved in a multitude of exciting roles in a variety of systems and cell types in the CNS. PATHOPHYSIOLOGICAL ROLES FOR EICOSANOIDS IN THE CNS Two important disease processes have drawn much attention recently because of their possible pathophysiological link to AA metabolism in nervous tissue. Alzheimer’s disease (AD) has been linked to the eicosanoid system through epidemiological studies of NSAID users and stroke has been associated using a more basic science approach. AD is a neurodegenerative process characterized by progression of neuropathological markers, such as neurofibrillary tangles and amyloid plaques, from areas of the brain involved in polymodal integration (e.g. the entorhinal cortex) to regions of more primary function. Early epidemiological evidence suggested that patients with rheumatoid arthritis, an inflammatory disease, had less risk
ARACHIDONIC ACID IN THE NERVOUS SYSTEM of AD (McGeer et al 1996). Since non-steroidal antiinflammatory drugs (NSAIDS) are the mainstays of therapy for this class of patient, a logical and fortuitous extension was that NSAIDs might be useful as therapeutics at some level of AD treatment. Primarily retrospective analyses of this hypothesis have demonstrated benefit of this class in the management of AD, particularly when the NSAID used was a non-specific inhibitor of both COX isoforms. Recent in vitro data has raised the prospect that certain NSAIDs may reduce production of b-amyloid through COXindependent mechanisms. A recent trial involving a therapeutic agent targeted to inhibit the COX-2 enzyme, was reported to have shown no increased benefit in the reduction of AD progression (McGeer 2000). Data from basic research has shown that levels of both COX isoforms may be affected in AD. COX-1 has been elevated in microglial populations (Hoozemans et al 2001; Yermakova et al 1999). Reports are conflicting as to whether COX-2 levels increase or decrease, although neurons seem to be the cell type most consistently described as being pathologically affected (Chang et al 1996; Ho et al 2001; Pasinetti and Aisen 1998; Tocco et al 1997). If the disease process does indeed begin in the more complex telencephalic structures involved in polymodal sensory integration, this would parallel the density of localization of COX-2immunoreactive neurons (Breder et al 1995). A possibility may exist that the redundancy of the COX systems ensures expression of the AD pathophysiology in the event that one isoform is inhibited. One might speculate that, as many transmitter systems previously implicated in AD, COX is localized in cells that are damaged (in the case of neurons) or reactive (in the case of glia). While the availability of therapeutics targeted at COX-2 have focused both clinical and basic research in this arena, other eicosanoid pathways may prove equally important. The clinical significance of all of these phenomena awaits prospective, randomized, controlled testing of receptor or enzyme-specific therapeutics in clinical trials. Knowledge of the role of eicosanoids in stroke has followed a more traditional route, with some of the first promising signs emanating from the laboratory. Induction of cerebral ischaemia by occlusion of the middle cerebral artery leads to induction of COX-2 expression in the ischaemic penumbra, a transition zone between the infarcted and ‘‘normal’’ brain. Inhibition of COX-2 leads to a reduction of infarct volume, suggesting a role for COX2 in the pathophysiology of this response (Iadecola et al 2001a). On the other hand, the induction of COX-2 may be an epiphenomenon, with the ‘‘COX-2-specific antagonists’’ affecting other biochemical pathways throughout the ischaemic brain. The induction of the COX-2 in the cortex may also cause release of diffusable mediators, capable of affecting the infarct process. It is of interest that accounts of COX-2 expression are not described in subcortical regions, such as the striatum, where these occluded vessels also feed. Recent work has also suggested that COX-1 plays a beneficial role in maintaining CBF in the penumbra of the ischaemic brain (Iadecola et al 2001b). Replication of the primary laboratory data should incite fruitful and significant clinical investigation. SUMMARY AND CONCLUSIONS The previous era has seen a transition in the study of eicosanoids in nervous tissue from descriptions of static systems to dynamic processes and single synthetic steps to participation in organism behaviours or adaptations. As advanced as we have become in these endeavours, in some respects we have only begun truly scientific inquiry, e.g. we have only recently shaken some unfounded ‘‘principles’’ on the ‘‘housekeeping’’ role of COX-1.
453
This is ironic, since prostaglandins as a family were once thought to be too ubiquitous to serve as neuronal transmitters in brain. The moral may be that eicosanoid metabolites and enzyme systems thought to be merely involved in ‘‘housekeeping functions’’ may play the most important roles, once we learn to ask the right basic research and clinical questions. The coming era will hopefully yield a robust a line of research into the lipoxygenases (and associated systems) and epoxygenases, as has been focused on cyclooxygenases. As these endeavours mature, a translation to clinical research must ensue. The fruits of well-designed and controlled studies will surely validate the scientific foundation that has been established.
REFERENCES Ahmadi S, Lippross S, Neuhuber WL and Zeilhofer HU (2002) PGE(2) selectively blocks inhibitory glycinergic neurotransmission onto rat superficial dorsal horn neurons. Nature Neurosci, 5(1), 34–40. Alkayed NJ, Narayanan J, Gebremedhin D et al (1996) Molecular characterization of an arachidonic acid epoxygenase in rat brain astrocytes. Stroke, 27(5), 971–979. Andreasson KI, Savonenko A, Vidensky S et al (2001) Age-dependent cognitive deficits and neuronal apoptosis in cyclooxygenase-2 transgenic mice. J Neurosci, 21(20), 8198–8209. Batshake B, Nilsson C and Sundelin J (1995) Molecular characterization of the mouse prostanoid EP1 receptor gene. Eur J Biochem, 231(3), 809–814. Bazan NG (1989) Arachidonic acid in the modulation of excitable membrane function and at the onset of brain damage. Ann NY Acad Sci, 559, 1–16. Beuckmann CT, Lazarus M, Gerashchenko D et al (2000) Cellular localization of lipocalin-type prostaglandin D synthase (b-trace) in the central nervous system of the adult rat. J Comp Neurol, 428(1), 62–78. Bhagwat SV, Boyd MR and Ravindranath V (1995) Brain mitochondrial cytochromes P450: xenobiotic metabolism, presence of multiple forms and their selective inducibility. Arch Biochem Biophys, 320(1), 73–83. Blackman SC, Dawson G, Antonakis K and Le Breton GC (1998) The identification and characterization of oligodendrocyte thromboxane A2 receptors. J Biol Chem, 273(1), 475–483. Breder CD et al (1992) Distribution and characterization of cyclooxygenase immunoreactivity in the ovine brain. J Comp Neurol, 322(3), 409–438. Breder CD et al (1995) Characterization of inducible cyclooxygenase in rat brain. J Comp Neurol, 355(2), 296–315. Cao C, Matsumura K, Yamagata K and Watanabe Y (1996) Endothelial cells of the rat brain vasculature express cyclooxygenase-2 mRNA in response to systemic interleukin-1b: a possible site of prostaglandin synthesis responsible for fever. Brain Res, 733(2), 263–272. Carlen PL, Gurevich N, Wu PH et al (1989) Actions of arachidonic acid and hepoxilin A3 on mammalian hippocampal CA1 neurons. Brain Res, 497(1), 171–176. Carlen PL, Gurevich N, Zhang L et al (1994) Formation and electrophysiological actions of the arachidonic acid metabolites, hepoxilins, at nanomolar concentrations in rat hippocampal slices. Neuroscience, 58(3), 493–502. Chang JW, Coleman PD and O’Banion MK (1996) Prostaglandin G/H synthase-2 (cyclooxygenase-2) mRNA expression is decreased in Alzheimer’s disease. Neurobiol Aging, 17(5), 801–808. Cho HP, Nakamura M and Clarke SD (1999a) Cloning, expression, and fatty acid regulation of the human D-5 desaturase. J Biol Chem, 274(52), 37335–37339. Cho HP, Nakamura MT and Clarke SD (1999b) Cloning, expression, and nutritional regulation of the mammalian D-6 desaturase. J Biol Chem, 274(1), 471–477. Choi WC, Gerfen CR, Suh PG and Rhee SG (1989) Immunohistochemical localization of a brain isozyme of phospholipase C (PLC III) in astroglia in rat brain. Brain Res, 499(1), 193–197. Chopra B, Giblett S, Little JG et al (2000) Cyclooxygenase-1 is a marker for a subpopulation of putative nociceptive neurons in rat dorsal root ganglia. Eur J Neurosci, 12(3), 911–920.
454
NERVOUS SYSTEM
Clements MP, Bliss TV and Lynch MA (1991) Increase in arachidonic acid concentration in a postsynaptic membrane fraction following the induction of long-term potentiation in the dentate gyrus. Neuroscience, 45(2), 379–389. Corasaniti MT, Strongoli MC, Piccirilli S et al (2000) Apoptosis induced by gp120 in the neocortex of rat involves enhanced expression of cyclooxygenase type 2 and is prevented by NMDA receptor antagonists and by the 21-aminosteroid U-74389G. Biochem Biophys Res Commun, 274(3), 664–669. Deininger MH, Meyermann R, Trautmann K et al (2000) Cyclooxygenase (COX)-1 expressing macrophages/microglial cells and COX-2 expressing astrocytes accumulate during oligodendroglioma progression. Brain Res, 885(1), 111–116. Deininger MH and Schluesener HJ (1999) Cyclooxygenase-1 and -2 are differentially localized to microglia and endothelium in rat EAE and glioma. J Neuroimmunol, 95(1–2), 202–208. Eguchi N, Minami T, Shirafuji N et al (1999) Lack of tactile pain (allodynia) in lipocalin-type prostaglandin D synthase-deficient mice. Proc Natl Acad Sci USA, 96(2), 726–730. Ek M, Arias C, Sawchenko P and Ericsson-Dahlstrand A (2000) Distribution of the EP3 prostaglandin E(2) receptor subtype in the rat brain: relationship to sites of interleukin-1-induced cellular responsiveness. J Comp Neurol, 428(1), 5–20. Elmquist JK et al (1997) Intravenous lipopolysaccharide induces cyclooxygenase 2-like immunoreactivity in rat brain perivascular microglia and meningeal macrophages. J Comp Neurol, 381(2), 119– 129. Exner HJ and Schlicker E (1995) Prostanoid receptors of the EP3 subtype mediate the inhibitory effect of prostaglandin E2 on noradrenaline release in the mouse brain cortex. Naunyn Schmiedebergs Arch Pharmacol, 351(1), 46–52. Faraci FM, Sobey CG, Chrissobolis S et al (2001) Arachidonate dilates basilar artery by lipoxygenase-dependent mechanism and activation of K(+) channels. Am J Physiol Regul Integr Comp Physiol, 281(1), R246–R253. Felder CC, Kanterman RY, Ma AL and Axelrod J (1990) Serotonin stimulates phospholipase A2 and the release of arachidonic acid in hippocampal neurons by a type 2 serotonin receptor that is independent of inositolphospholipid hydrolysis. Proc Natl Acad Sci USA, 87(6), 2187–2191. Fisher M (1993) Increased thromboxane biosynthesis in patients with acute cerebral ischemia. Stroke, 24(6), 912. Gerashchenko D, Beuckmann CT, Kanaoka Y et al (1998) Dominant expression of rat prostanoid DP receptor mRNA in leptomeninges, inner segments of photoreceptor cells, iris epithelium, and ciliary processes. J Neurochem, 71(3), 937–945. Gerfen CR, Choi WC, Suh PG and Rhee SG (1988) Phospholipase C I and II brain isozymes: immunohistochemical localization in neuronal systems in rat brain. Proc Natl Acad Sci USA, 85(9), 3208–3212. Gerozissis K, Vulliez B, Saavedra JM et al (1985) Lipoxygenase products of arachidonic acid stimulate LHRH release from rat median eminence. Neuroendocrinology, 40(3), 272–276. Giulian D, Corpuz M, Richmond B et al (1996) Activated microglia are the principal glial source of thromboxane in the central nervous system. Neurochem Int, 29(1), 65–76. Goto K, Ochi H, Yasunaga Y et al (1998) Analgesic effect of mofezolac, a non-steroidal antiinflammatory drug, against phenylquinone-induced acute pain in mice. Prostagland Lipid Mediat, 56(4), 245–254. Hayaishi O (1983) Prostaglandin D2: a neuromodulator. Adv Prostagland Thrombox Leukot Res, 12, 333–337. Hayaishi O (2000) Molecular mechanisms of sleep–wake regulation: a role of prostaglandin D2. Phil Trans R Soc Lond B Biol Sci, 355(1394), 275– 280. Hewett SJ, Uliasz TF, Vidwans AS and Hewett JA (2000) Cyclooxygenase-2 contributes to N-methyl-D-aspartate-mediated neuronal cell death in primary cortical cell culture. J Pharmacol Exp Ther, 293(2), 417–425. Ho L, Purohit D, Haroutunian V et al (2001) Neuronal cyclooxygenase-2 expression in the hippocampal formation as a function of the clinical progression of Alzheimer disease. Arch Neurol, 58(3), 487–492. Hoozemans JJ, Rozemuller AJ, Janssen I et al (2001) Cyclooxygenase expression in microglia and neurons in Alzheimer’s disease and control brain. Acta Neuropathol (Berl), 101(1), 2–8.
Hwang SW, Cho H, Kwak J et al (2000) Direct activation of capsaicin receptors by products of lipoxygenases endogenous capsaicin-like substances. Proc Natl Acad Sci USA, 97(11), 6155–6160. Iadecola C, Niwa K, Nogawa S et al (2001a) Reduced susceptibility to ischemic brain injury and N-methyl-D-aspartate-mediated neurotoxicity in cyclooxygenase-2-deficient mice. Proc Natl Acad Sci USA, 98(3), 1294–1299. Iadecola C, Sugimoto K, Niwa K et al (2001b) Increased susceptibility to ischemic brain injury in cyclooxygenase-1-deficient mice. J Cereb Blood Flow Metab, 21(12), 1436–1441. Ishimoto T, Akiba S, Sato T and Fujii T (1994) Contribution of phospholipases A2 and D to arachidonic acid liberation and prostaglandin D2 formation with increase in intracellular Ca2+ concentration in rat peritoneal mast cells. Eur J Biochem, 219(1–2), 401–406. Jakobsson PJ, Thoren S, Morgenstern R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96(13), 7220–7225. Jisaka M, Kim RB, Boeglin WE et al (1997) Molecular cloning and functional expression of a phorbol ester-inducible 8-S-lipoxygenase from mouse skin. J Biol Chem, 272(39), 24410–24416. Katsuki H and Okuda S (1995) Arachidonic acid as a neurotoxic and neurotrophic substance. Prog Neurobiol, 46(6), 607–636. Kawahara H, Sakamoto A, Takeda S et al (2001) A prostaglandin E2 receptor subtype EP1 receptor antagonist (ONO-8711) reduces hyperalgesia, allodynia, and c-fos gene expression in rats with chronic nerve constriction. Anesth Analg, 93(4), 1012–1017. Kawajiri H, Zhuang D, Qiao N et al (2000) Expression of arachidonate 12-lipoxygenase in rat tissues: a possible role in glucagon secretion. J Histochem Cytochem, 48(10), 1411–1419. Kawamura T, Yamauchi T, Koyama M et al (1997) Expression of prostaglandin EP2 receptor mRNA in the rat spinal cord. Life Sci, 61(21), 2111–2116. Kitanaka J, Hashimoto H, Sugimoto Y et al (1995) cDNA cloning of a thromboxane A2 receptor from rat astrocytes. Biochim Biophys Acta, 1265(2–3), 220–223. Kumazawa T, Mizumura K, Koda H and Fukusako H (1996) EP receptor subtypes implicated in the PGE2-induced sensitization of polymodal receptors in response to bradykinin and heat. J Neurophysiol, 75(6), 2361–2368. Kusuhara H, Matsuyuki H and Okumoto T (1998) Involvement of prostaglandins produced by cyclooxygenase-1 in murine visceronociception induced by phenylquinone. Prostagland Lipid Mediat, 55(1), 43–49. Lacroix S and Rivest S (1998) Effect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cyclooxygenase enzymes (COX-1 and COX-2) in the rat brain. J Neurochem, 70(2), 452–466. Lammers CH, Schweitzer P, Facchinetti P et al (1996) Arachidonate 5lipoxygenase and its activating protein: prominent hippocampal expression and role in somatostatin signaling. J Neurochem, 66(1), 147–152. Lazarewicz JW, Salinska E and Wroblewski JT (1992) NMDA receptormediated arachidonic acid release in neurons: role in signal transduction and pathological aspects. Adv Exp Med Biol, 318, 73–89. Lazarewicz JW, Wroblewski JT and Costa E (1990) N-methyl-D-aspartatesensitive glutamate receptors induce calcium-mediated arachidonic acid release in primary cultures of cerebellar granule cells. J Neurochem, 55(6), 1875–1881. Lee M, Jo Y, Chun M et al (2000) Immunohistochemical localization of phospholipase D1 in rat central nervous system. Brain Res, 864(1), 52– 59. Levi G, Minghetti L and Aloisi F (1998) Regulation of prostanoid synthesis in microglial cells and effects of prostaglandin E2 on microglial functions. Biochimie, 80(11), 899–904. Li DY, Hardy P, Abran D et al (1997) Key role for cyclooxygenase-2 in PGE2 and PGF2a receptor regulation and cerebral blood flow of the newborn. Am J Physiol, 273(4), R1283–R1290. Linden DJ, Sheu FS, Murakami K and Routtenberg A (1987) Enhancement of long-term potentiation by cis-unsaturated fatty acid: relation to protein kinase C and phospholipase A2. J Neurosci, 7(11), 3783–3792.
ARACHIDONIC ACID IN THE NERVOUS SYSTEM Liu Y, Liu D, Heath L et al (2001) Direct activation of an inwardly rectifying potassium channel by arachidonic acid. Mol Pharmacol, 59(5), 1061–1068. Lynch MA, Errington ML and Bliss TV (1989) Nordihydroguaiaretic acid blocks the synaptic component of long-term potentiation and the associated increases in release of glutamate and arachidonate: an in vivo study in the dentate gyrus of the rat. Neuroscience, 30(3), 693–701. Lysz TW and Needleman P (1982) Evidence for two distinct forms of fatty acid cyclooxygenase in brain. J Neurochem, 38(4), 1111–1117. Manev H, Uz T, Sugaya K and Qu T (2000) Putative role of neuronal 5lipoxygenase in an aging brain. FASEB J, 14(10), 1464–1469. Martelli AM, Lach S, Grill V et al (1996) Expression and immunohistochemical localization of eight phospholipase C isoforms in adult male mouse cerebellar cortex. Acta Histochem, 98(2), 131–141. Matsumura K, Watanabe Y, Onoe H and Watanabe Y (1995) Prostacyclin receptor in the brain and central terminals of the primary sensory neurons: an autoradiographic study using a stable prostacyclin analogue [3H]iloprost. Neuroscience, 65(2), 493–503. McGeer PL (2000) Cyclooxygenase-2 inhibitors: rationale and therapeutic potential for Alzheimer’s disease. Drugs Aging, 17(1), 1–11. McGeer PL, Schulzer M and McGeer EG (1996) Arthritis and antiinflammatory agents as possible protective factors for Alzheimer’s disease: a review of 17 epidemiologic studies. Neurology, 47(2), 425–432. Medhora M, Narayanan J and Harder D (2001) Dual regulation of the cerebral microvasculature by epoxyeicosatrienoic acids. Trends Cardiovasc Med, 11(1), 38–42. Miggin SM and Kinsella BT (1998) Expression and tissue distribution of the mRNAs encoding the human thromboxane A2 receptor (TP) a and b isoforms. Biochim Biophys Acta, 1425(3), 543–559. Minami T, Matsumura S, Okuda-Ashitaka E et al (2001a) Characterization of the glutamatergic system for induction and maintenance of allodynia. Brain Res, 895(1–2), 178–185. Minami T, Nakano H, Kobayashi T et al (2001b) Characterization of EP receptor subtypes responsible for prostaglandin E2-induced pain responses by use of EP1 and EP3 receptor knockout mice. Br J Pharmacol, 133(3), 438–444. Minami T, Nishihara I, Uda R et al (1994a) Characterization of EPreceptor subtypes involved in allodynia and hyperalgesia induced by intrathecal administration of prostaglandin E2 to mice. Br J Pharmacol, 112(3), 735–740. Minami T, Uda R, Horiguchi S et al (1994b) Allodynia evoked by intrathecal administration of prostaglandin E2 to conscious mice. Pain, 57(2), 217–223. Minghetti L and Levi G (1995) Induction of prostanoid biosynthesis by bacterial lipopolysaccharide and isoproterenol in rat microglial cultures. J Neurochem, 65(6), 2690–2698. Mizoguchi A, Eguchi N, Kimura K et al (2001) Dominant localization of prostaglandin D receptors on arachnoid trabecular cells in mouse basal forebrain and their involvement in the regulation of non-rapid eye movement sleep. Proc Natl Acad Sci USA, 98(20), 11674–11679. Nakamura K, Kaneko T, Yamashita Y et al (2000) Immunohistochemical localization of prostaglandin EP3 receptor in the rat nervous system. J Comp Neurol, 421(4), 543–569. Nishiyama M, Watanabe T, Ueda N et al (1993) Arachidonate 12lipoxygenase is localized in neurons, glial cells, and endothelial cells of the canine brain. J Histochem Cytochem, 41(1), 111–117. Norris PJ, Hardwick JP and Emson PC (1994) Localization of NADPH cytochrome P450 oxidoreductase in rat brain by immunohistochemistry and in situ hybridization and a comparison with the distribution of neuronal NADPH-diaphorase staining. Neuroscience, 61(2), 331–350. O’Neill GP and Ford-Hutchinson AW (1993) Expression of mRNA for cyclooxygenase-1 and cyclooxygenase-2 in human tissues. FEBS Lett, 330(2), 156–160. Oida H, Namba T, Sugimoto Y et al (1995) In situ hybridization studies of prostacyclin receptor mRNA expression in various mouse organs. Br J Pharmacol, 116(7), 2828–2837. Oka K, Oka T and Hori T (1998) PGE2 receptor subtype EP1 antagonist may inhibit central interleukin-1b-induced fever in rats. Am J Physiol, 275(6), R1762–R1765. Oka T, Oka K, Scammell TE et al (2000) Relationship of EP(1–4) prostaglandin receptors with rat hypothalamic cell groups involved in lipopolysaccharide fever responses. J Comp Neurol, 428(1), 20–32.
455
Oleksiak MF, Wu S, Parker C et al (2000) Identification, functional characterization, and regulation of a new cytochrome P450 subfamily, the CYP2Ns. J Biol Chem, 275(4), 2312–2321. Omana-Zapata I and Bley KR (2001) A stable prostacyclin analog enhances ectopic activity in rat sensory neurons following neuropathic injury. Brain Res, 904(1), 85–92. Omote K, Kawamata T, Nakayama Y et al (2001) The effects of peripheral administration of a novel selective antagonist for prostaglandin E receptor subtype EP1, ONO-8711, in a rat model of postoperative pain. Anesth Analg, 92(1), 233–238. Ong WY, Horrocks LA and Farooqui AA (1999) Immunocytochemical localization of cPLA2 in rat and monkey spinal cord. J Mol Neurosci, 12(2), 123–130. Pace-Asciak CR, Laneuville O, Gurevich N et al (1991) Formation of a glutathionyl-conjugate of hepoxilin A3 and its action in rat brain. Adv Prostagland Thrombox Leukot Res, 21B, 731–734. Pace-Asciak CR, Laneuville O, Su WG et al (1990a) A glutathione conjugate of hepoxilin A3: formation and action in the rat central nervous system. Proc Natl Acad Sci USA, 87(8), 3037–3041. Pace-Asciak CR, Wong L and Corey EJ (1990b) Hepoxilin A3 blocks the release of norepinephrine from rat hippocampal slices. Biochem Biophys Res Commun, 173(3), 949–953. Pasinetti GM and Aisen PS (1998) Cyclooxygenase-2 expression is increased in frontal cortex of Alzheimer’s disease brain. Neuroscience, 87(2), 319–324. Piomelli D (1991) Metabolism of arachidonic acid in nervous system of marine mollusk Aplysia californica. Am J Physiol, 260(5), R844–R848. Piomelli D (1994a) Eicosanoids in synaptic transmission. Crit Rev Neurobiol, 8(1–2), 65–83. Piomelli D, Feinmark SJ, Shapiro E and Schwartz JH (1989) 12-Ketoeicosatetraenoic acid. A biologically active eicosanoid in the nervous system of Aplysia. Ann NY Acad Sci, 559, 208–218. Saadi M, Gerozissis K, Rougeot C et al (1990) Leukotriene C4-induced release of LHRH into the hypophyseal portal blood and of LH into the peripheral blood. Life Sci, 46(25), 1857–1865. Saito S, Sakagami H and Kondo H (2000) Localization of mRNAs for phospholipase D (PLD) type 1 and 2 in the brain of developing and mature rat. Brain Res Dev, 120(1), 41–47. Saper CB, Chou TC and Scammell TE (2001) The sleep switch: hypothalamic control of sleep and wakefulness. Trends Neurosci, 24(12), 726–731. Saper CB et al (1994) The neurologic basis of fever, N Engl J Med, 330(26), 1880–1886. Satoh K, Nagano Y, Shimomura C et al (2000) Expression of prostaglandin E synthase mRNA is induced in b-amyloid treated rat astrocytes. Neurosci Lett, 283(3), 221–223. Sauer B (1998) Inducible gene targeting in mice using the Cre/lox system. Methods, 14(4), 381–392. Schaad NC, Schorderet M and Magistretti PJ (1987) Prostaglandins and the synergism between VIP and noradrenaline in the cerebral cortex. Nature, 328(6131), 637–640. Schiltz JC and Sawchenko PE (2002) Distinct brain vascular cell types manifest inducible cyclooxygenase expression as a function of the strength and nature of immune insults. J Neurosci, 22(13), 5606–5618. Schlicker E and Marr I (1997) Mutual interactions of the presynaptic histamine H3 and prostaglandin EP3 receptors on the noradrenergic terminals in the mouse brain. Neuroscience, 79(1), 247–254. Schwab JM, Seid K and Schluesener HJ (2001) Traumatic brain injury induces prolonged accumulation of cyclooxygenase-1 expressing microglia/brain macrophages in rats. J Neurotrauma, 18(9), 881–890. Schwartz JH (1991) Arachidonic acid metabolism: potential for diverse signalling within the same neuron. Biochem Soc Trans, 19(2), 387–390. Shearman MS, Naor Z, Sekiguchi K et al (1989) Selective activation of the g-subspecies of protein kinase C from bovine cerebellum by arachidonic acid and its lipoxygenase metabolites. FEBS Lett, 243(2), 177–182. Sherin JE, Shiromani PJ, McCarley RW and Saper CB (1996) Activation of ventrolateral preoptic neurons during sleep. Science, 271(5246), 216–219. Shinomura T, del Rio E, Breen KC et al (2000) Activation of phospholipase D by metabotropic glutamate receptor agonists in rat cerebrocortical synaptosomes. Br J Pharmacol, 131(5), 1011–1018.
456
NERVOUS SYSTEM
Six DA and Dennis EA (2000) The expanding superfamily of phospholipase A(2) enzymes: classification and characterization. Biochim Biophys Acta, 1488(1–2), 1–19. Smith WL, DeWitt DL and Garavito RM (2000) Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem, 69, 145– 182. Southall MD and Vasko MR (2001) Prostaglandin receptor subtypes, EP3C and EP4, mediate the prostaglandin E2-induced cAMP production and sensitization of sensory neurons. J Biol Chem, 276(19), 16083–16091. Sri KS, Matsumura H, Kubo E et al (1994) Effects of prostaglandin D2, lipoxins and leukotrienes on sleep and brain temperature of rats. Prostagland Leukotrienes Essent Fatty Acids, 51(2), 87–93. Sugimoto Y, Shigemoto R, Namba T et al (1994) Distribution of the messenger RNA for the prostaglandin E receptor subtype EP3 in the mouse nervous system. Neuroscience, 62(3), 919–928. Sutarmo SV, Shibuya I, Kabashima N et al (1998) Actions of prostaglandin E2 on rat supraoptic neurones. J Neuroendocrinol, 10(12), 927–936. Taiwo YO, Heller PH and Levine JD (1990) Characterization of distinct phospholipases mediating bradykinin and noradrenaline hyperalgesia. Neuroscience, 39(2,) 523–531. Takechi H, Matsumura K, Watanabe Y et al (1996) A novel subtype of the prostacyclin receptor expressed in the central nervous system. J Biol Chem, 271(10), 5901–5906. Tanioka T, Nakatani Y, Semmyo N et al (2000) Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem, 275(42), 32775–32782. Tieman TL, Steel DJ, Gor Y et al (1997) A pertussis toxin-sensitive 8lipoxygenase pathway is activated by a nicotinic acetylcholine receptor in Aplysia neurons. J Neurophysiol, 85(5), 2150–2158. Tindberg N and Ingelman-Sundberg M (1996) Expression, catalytic activity, and inducibility of cytochrome P450 2E1 (CYP2E1) in the rat central nervous system. J Neurochem, 67(5), 2066–2073. Tocco G, Freire-Moar J, Schreiber SS et al (1997) Maturational regulation and regional induction of cyclooxygenase-2 in rat brain: implications for Alzheimer’s disease. Exp Neurol, 144(2), 339–349. Ushikubi F, Segi E, Sugimoto Y et al (1998) Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature, 395(6699), 281–284. Vesin MF, Urade Y, Hayaishi O and Droz B (1995) Neuronal and glial prostaglandin D synthase isozymes in chick dorsal root ganglia: a light and electron microscopic immunocytochemical study. J Neurosci, 15(1), 470–476.
Volterra A, Trotti D, Cassutti P et al (1992) A role for the arachidonic acid cascade in fast synaptic modulation: ion channels and transmitter uptake systems as target proteins. Adv Exp Med Biol, 318, 147–158. Waring M, Drappatz J, Weichel O et al (1999) Modulation of neuronal phospholipase D activity under depolarizing conditions. FEBS Lett, 464(1–2), 21–24. Watanabe Y, Watanabe Y and Hayaishi O (1988) Quantitative autoradiographic localization of prostaglandin E2 binding sites in monkey diencephalon. J Neurosci, 8(6), 2003–2010. Willingale HL, Gardiner NJ, McLymont N et al (1997) Prostanoids synthesized by cyclooxygenase isoforms in rat spinal cord and their contribution to the development of neuronal hyperexcitability. Br J Pharmacol, 122(8), 1593–1604. Wolfe LS (1988) Arachidonic acid metabolites and the nervous system: an appraisal of the present situation. In Curtis-Prior P (ed.), Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids. Edinburgh: Churchhill Livingston, 411–421. Xie WL, Chipman JG, Robertson DL et al (1991) Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci USA, 88(7), 2692–2696. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11(2), 371–386. Yamagata K, Matsumura K, Inoue W et al (2001) Co-expression of microsomal-type prostaglandin E synthase with cyclooxygenase-2 in brain endothelial cells of rats during endotoxin-induced fever. J Neurosci, 21(8), 2669–2677. Yermakova AV, Rollins J, Callahan LM et al (1999) Cyclooxygenase-1 in human Alzheimer and control brain: quantitative analysis of expression by microglia and CA3 hippocampal neurons. J Neuropathol Exp Neurol, 58(11), 1135–1146. Yokotani K, Murakami Y, Okada S and Hirata M (2001) Role of brain arachidonic acid cascade on central CRF1 receptor-mediated activation of sympatho-adrenomedullary outflow in rats. Eur J Pharmacol, 419(2–3), 183–189. Yoshida Y, Matsumura H, Nakajima T et al (2000) Prostaglandin E (EP) receptor subtypes and sleep: promotion by EP4 and inhibition by EP1/ EP2. Neuroreport, 11(10), 2127–2131. Zhang J and Rivest S (1999) Distribution, regulation and colocalization of the genes encoding the EP2- and EP4-PGE2 receptors in the rat brain and neuronal responses to systemic inflammation. Eur J Neurosci, 11(8), 2651–2668. Zhang J and Rivest S (2000) A functional analysis of EP4 receptorexpressing neurons in mediating the action of prostaglandin E2 within specific nuclei of the brain in response to circulating interleukin-1b. J Neurochem, 74(5), 2134–2145.
41 Eicosanoid Pathways in the Ageing of the Central Nervous System Hari Manev and Tolga Uz University of Illinois, Chicago, IL, USA
Generally, ageing is associated with the diminishing of functional capabilities and with morphological, biochemical and molecular pathology-like alterations. These factors also apply to the ageing of the central nervous system (CNS). Thus, it has been noted that a variety of ageing-associated neuropathological conditions can be attributed to the increased vulnerability of the ageing brain to degeneration. Several clinical and experimental examples illustrate increased brain vulnerability in aged subjects. Hence, the degree of recovery from a stroke is far less in very old than in younger patients (Arboix et al 2000; Nakayama et al 1994). In experiments with animals, the severity of ischaemia-induced brain injury was greater in old than in young rats (Kharlamov et al 2000) and in old than in young mice (Nagayama et al 1999). It is also true that the remarkable morphological and functional plasticity of the brain, illustrated by its capability to produce new neurons even in adulthood (i.e. adult neurogenesis; van Praag et al 2002), may also be compromised during brain ageing. Thus, neurogenesis in the dentate gyrus of the hippocampus (a region of the brain that is important for learning and memory) is reduced in old rats compared with young rats (Kuhn et al 1996). This ageingreduced adult neurogenesis can be associated with age-related cognitive decline. Dietary restriction, a procedure known to increase the lifespan of rodents and to improve their learning and memory, was also capable of enhancing neurogenesis in the hippocampus of adult mice (Lee et al 2002). The mammalian eicosanoid pathways that encompass two major enzyme families, the lipoxygenases (LOX) that lead to synthesis of leukotrienes and the cyclooxygenases (COX) that lead to synthesis of prostaglandins and thromboxanes (collectively termed prostanoids; Funk 2001), are also present in the mammalian brain. The expression and the activity of LOX are tissue-specific; prominent CNS expression has been demonstrated for 5-LOX (Lammers et al 1996; Uz et al 1998). Neural cells also express 8-LOX (Jisaka et al 1997) and 12-LOX (Li et al 1997; Watanabe et al 1993). COX exists as two isoforms, referred to as COX-1 and COX-2. Whereas COX-1 is constitutively expressed, COX-2 is inducible and its expression in CNS neurons can be regulated (Breder et al 1995; Hoffmann 2000; O’Banion 1999). Moreover, biochemical studies have demonstrated that various leukotrienes and prostaglandins can be produced in the mammalian brain as well as in neurons (Adesuyi et al 1985; Bishai and Coceani 1992; Kaufmann et al 1997; Lindgren et al 1984; Shimizu et al 1987). Neuronal expression of 5-LOX and COX-2 is highly suggestive of a putative functional role of these enzymes in brain physiology and pathology. Moreover, recently it has been speculated that in the peripheral nervous system, i.e. sensory neurons, the product of The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
12-LOX activity, 12-HPETE (12-hydroperoxy-5,8,14-cis-10-transeicosatetraenoic acid), may act as an endogenous regulator of pain-related neuronal vanilloid (capsaicin) receptors (Piomelli 2001). Whereas their physiological role in the CNS still remains elusive (Piomelli 1994), a significant body of evidence indicates that 5-LOX and COX-2 are involved in neurodegenerative pathologies, particularly in brain injuries induced by ischaemia or stroke (Iadecola et al 2001; Rao et al 1999). These data, along with the findings of the aging-associated upregulation of the CNS 5-LOX pathway (Manev et al 2000; Qu et al 2000; Uz et al 1998) and COX pathway (Montine et al 2002), suggest that eicosanoid pathways may contribute to ageing-related increases in brain vulnerability. Alternatively, eicosanoid pathways might influence the plasticity of an ageing CNS through their putative involvement in the mechanisms of cell proliferation. Such mechanisms, which are ultimately responsible for adult neurogenesis, could also play a compensatory role during ageing and might determine how the brain copes with injury (Manev et al 2001a). EICOSANOID PATHWAYS AND CELL PROLIFERATION: RELEVANT FOR NEUROGENESIS? In a variety of cell types, metabolites of arachidonic acid function in growth signalling. For example, COX metabolite prostaglandins, such as prostaglandin E2, are capable of transactivating epidermal growth factor receptors and thereby of promoting colon cancer growth and gastrointestinal hypertrophy (Pai et al 2002). LOX metabolites, e.g. 5-HETE (5-hydroxy-6,8,11,14eicosatetraenoic acid), also increase the growth of tumour tissues, e.g. breast cancer cells (Avis et al 2001). In contrast, inhibitors of 5-LOX reduce cell proliferation. Hence, it was found that the 5LOX inhibitors MK-886 and AA-861 reduce proliferation and induce apoptosis of prostate cancer cell cultures (Ghosh and Myers 1997; Anderson et al 1998). In vitro studies indicate that 5LOX might be critical for the growth of brain tumours as well. Thus, proliferation of human glioma cells in culture was inhibited dose-dependently by application of 5-LOX inhibitors AA-863 and U-60,257 (Gati et al 1990). Moreover, cell-cycle progression of neuroblastoma cells in culture appears to involve the LOX pathway (possibly 5-LOX). Thus, inhibition of LOX in the early G1 phase of the cell cycle resulted in G1 phase arrest, whereas neuroblastoma cells stopped progressing through the S phase when LOX were inhibited in the early S phase of the cell cycle (van Rossum et al 2002).
458
NERVOUS SYSTEM
In the brain, the putative participation of the 5-LOX pathway in regulating the cell cycle might be conducive to the growth of tumours. Human brain tumours are capable of synthesizing leukotrienes and also express 5-LOX as a multitranscript family encompassing 2.7, 3.1, 4.8, 6.4 and 8.6 kilobases. Boado et al (1992) reported that both the abundance of 5-LOX mRNAs and the expression of the larger transcripts correlate significantly with tumour malignancy. In addition to the possible participation of the eicosanoid pathways in the growth of tumour cells, eicosanoids may also have non-pathological actions in cell proliferation. For example, in primary cultures of rat cerebellar granule neurons it was observed that the expression of 5-LOX is higher when neurons are immature and thus still proliferating (i.e. neuronal progenitors) than when the cells differentiate into mature neurons. Thus, the expression of 5-LOX decreases in differentiated post-mitotic neuronal cells (Uz et al 2001a). Treatment of neuronal progenitors with a 5-LOX antisense (directed against the endogenous 5-LOX mRNA to inhibit the translation of RNA into 5-LOX protein) decreased the cell content of 5-LOX protein and effectively reduced cell proliferation. In experiments with neuronal progenitors, [3H]thymidine incorporation was used as a marker for cell proliferation; the incorporation of labelled thymidine was significantly reduced by the 5-LOX inhibitors AA-861, MK-886, and L-655,238. This anti-proliferative effect of 5-LOX inhibitors was reversible, suggesting that the 5-LOX pathway might participate in neurogenesis (Uz et al 2001a). Similar to these findings that indicate a role for 5-LOX in neurogenesis are data suggesting that the COX pathway could be required for neurogenesis in the adult mammalian CNS (Kumihashi et al 2001). Transient global brain ischaemia causes a peculiar response in adult rodents; it stimulates neurogenesis in the dentate gyrus of the hippocampus and also induces COX-2. In these experimental conditions, proliferation of brain cells in response to ischaemia was studied in adult gerbils chronically treated with acetylsalicylic acid, a non-selective COX inhibitor. It was observed that acetylsalicylic acid significantly reduced the number of proliferating cells (Kumihashi et al 2001). These authors suggested that COX, probably COX-2, and prostaglandins play an important role in maintaining adult neurogenesis; e.g. during the stimulated (compensatory?) proliferation of neural cells after brain ischaemia. If an increase of adult neurogenesis is a compensatory response of the brain (e.g. to ischaemia), and if both 5-LOX and COX-2 pathways are needed for the maintenance of neurogenesis, one could assume that ageing-associated upregulation of the brain’s eicosanoid pathways (Montine et al 2002; Qu et al 2000; Uz et al 1998) might reflect an attempt by the ageing brain to mobilize such a compensatory response. The re-entry into a cell cycle of a dormant (quiescent) stem cell may be a beneficial response and it might lead to the production of new neurons that could improve the functioning of a brain region into which these new neurons are successfully incorporated (Malberg et al 2000; Manev et al 2001b; van Praag et al 2002). On the contrary, an unsuccessful re-entry into a cell cycle may be responsible for cell death and could become a cause of neurodegenerative pathology (Verdaguer et al 2002). In fact, it has been proposed that in an ageing-associated neurodegenerative disease, Alzheimer’s disease (AD), certain neurons may re-enter into the abortive cell cycle due to an imbalance between mitogenic signals and differentiation factors (Nagy 2000). It remains to be elucidated whether ageing-induced upregulation of genes of the eicosanoid pathway (e.g. 5-LOX) is a consequence of neuronal reentry into the cell cycle, i.e. other factors such as hormones (e.g. glucocorticoids) and epigenetic mechanisms (e.g. DNA methylation) may also regulate the neuronal expression of 5-LOX and/or COX-2.
GLUCOCORTICOIDS AND NEURONAL EICOSANOID PATHWAYS Whereas the tissue- and cell type-specific expression of LOX and COX genes may be determined by epigenetic factors, such as the degree of DNA methylation in the CpG islands of the promoter sequences of their respective genes (Song et al 2001; Uhl et al 2002), the extent of LOX and COX gene expression can be influenced by hormones. Hence, the sequence of the 5-LOX promoter indicates that this gene could be susceptible to hormonal regulation in tissues in which 5-LOX is expressed (Hoshiko et al 1990). Surprisingly, glucocorticoids, also known as adrenal ‘‘stress hormones’’, which are endowed with remarkable antiinflammatory activity, are capable of stimulating the expression of 5-LOX, a typical ‘‘inflammatory’’ enzyme. Thus, glucocorticoids (cortisol in humans and corticosterone in rodents, and dexamethasone, their synthetic analogue) stimulate the expression of 5-LOX (mRNA and protein) in the brain in vivo (Uz et al 1999) and in neural cultures in vitro (Uz et al 2001b). Studies with human cell cultures, i.e. monocytes (Riddick et al 1997) and mast cells (Colamorea et al 1999), confirm this stimulatory action of glucocorticoids, which appears to involve the glucocorticoid receptor (Uz et al 2001b). On the other hand, the expression of neuronal COX-2 appears to be suppressed by glucocorticoids (Yamagata et al 1993). More detailed studies on the regional expression of COX-2 in the rat brain revealed that the inhibitory action of glucocorticoids on neuronal COX-2 expression is region-specific. Thus, whereas the administration of a synthetic glucocorticoid dexamethasone to rats suppressed induced COX-2 expression in the cortex, it was ineffective in altering COX-2 expression in the hippocampus (Koistinaho et al 1999). The inhibitory action of dexamethasone on COX-2 expression might involve its effects on COX-2 mRNA stability. In an in vitro system using a b-globin–Cox-2 reporter construct, dexamethasone was found to destabilize reporter mRNAs by inhibiting mitogen-activated protein kinase p38; inhibition of p38 was antagonized by the antiglucocorticoid RU486 and was delayed by the inhibition of transcription, suggesting that ongoing glucocorticoid receptor-dependent transcription may be required for the inhibitory action of dexamethasone (Lasa et al 2001). It has been proposed that the glucocorticoid-triggered upregulation of the neuronal expression of 5-LOX or the failure of glucocorticoids to inhibit COX-2 (e.g. in the hippocampus) may contribute to the well-known neurotoxic actions of these stress hormones, e.g. glucocorticoids have been implicated in neurodegenerative changes in the brain, either in response to stress or during ageing, when the circulating levels of glucocorticoids are typically elevated. Moreover, newly evolving concepts of the neurobiology of a serious psychiatric illness, depression, also imply the involvement of glucocorticoids and glucocorticoidmediated hypotrophic and morphological changes in the CNS, e.g. in the hippocampus (Nestler et al 2002). Nevertheless, opposite views have also been expressed; i.e. it has been proposed, based on the findings of decrements in nuclear glucocorticoid receptor protein levels and DNA binding in the hippocampus of old rats, that glucocorticoids might exert beneficial effects on hippocampal functioning that are lost or impaired during ageing (Murphy et al 2002). EPIGENETIC MECHANISMS AND EICOSANOID PATHWAYS In spite of the same genetic code being shared by all the cells of an organism, not all genes are expressed all the time and there also is
EICOSANOIDS IN CENTRAL NERVOUS SYSTEM AGEING a significant tissue specificity in gene expression. Tissue-specific expression of genes is controlled by epigenetic mechanisms, such as the configuration of chromatin and the methylation of genomic DNA (Paulsen and Ferguson-Smith 2001). DNA methylation is a covalent modification by methylation at the 5-carbon position of cytosine residues at CpG dinucleotides catalysed by DNA methyltransferase (Dnmt1). Dnmt1 is expressed at high levels in the CNS and its activity has a profound influence on CNS neurons (Fan et al 2001). Generally, DNA methylation of promoter-containing CpG islands is associated with transcriptional inactivation, i.e. gene silencing; consequently, the same gene is not methylated in the tissue where it is expressed and is methylated in tissues or cells where it is not expressed. An additional mechanism that appears to act in concert with DNA methylation is histone acetylation; inhibitors of histone deacetylase (HDAC) can stimulate the expression of genes silenced by DNA methylation (Ng and Bird 1999). Epigenetic mechanisms, e.g. via abnormal DNA methylation, may have a profound effect on CNS gene expression and might contribute to the pathobiology of brain disorders (Costa et al 2002; Petronis et al 1999). Moreover, global and gene-specific hypomethylation occurs during ageing in various cells and tissues (Mays-Hoopes 1989), as does the hypermethylation of certain gene promoters (Issa et al 1994). The causes of these changes are not clear and their relevance for ageing-associated brain pathologies has not yet been elucidated. It has been speculated that epigenetic mechanisms may influence the incidence of ageingassociated neurodegenerative disorders, such as Alzheimer’s disease, e.g. via genetic imprinting that involves DNA methylation (Farrer et al 1991). Similar to the promoters of the so-called housekeeping genes, the 5-LOX promoter is characterized by the presence of repeated G+C-rich elements and the lack of TATAA or CCAAT boxes. In contrast to the promoters of housekeeping genes that are usually unmethylated, the methylation of the G+C-rich 5-LOX promoter appears to be cell-specific and associated with the suppression of 5-LOX mRNA expression. The first direct evidence that 5-LOX expression can be regulated by DNA methylation was provided by Uhl et al (2002). These authors analysed the methylation status of the 5-LOX promoter in human myeloid cell lines and found that the 5-LOX promoter is methylated in 5-LOX-negative cell lines and unmethylated in 5-LOX-positive cells. Pharmacological treatment of 5-LOX-negative cells with the demethylating agent 5-aza-20 -deoxycytidine (AdC) triggered the expression of 5-LOX primary transcripts and mature mRNA. Also in neurons, the expression of 5-LOX can be increased by the Dnmt1 inhibitor AdC or by a HDAC inhibitor, valproic acid (Manev and Uz 2002). Whereas AdC is effective only in proliferating cells (e.g. neuronal precursors), valproic acid was effective in both neuronal precursors and primary cultures of differentiated neurons. Although the role of DNA methylation in neuronal COX-2 expression has not yet been characterized, COX2 expression also appears to be regulated by DNA methylation, particularly in cancerous cells (Song et al 2001). Thus, in human gastric carcinoma cell lines without COX-2 expression, COX-2 CpG islands were hypermethylated and treatment with demethylating agents effectively reactivated the expression of COX-2. In addition, COX-2 promoter activity was completely blocked by in vitro methylation of all CpG sites in the COX-2 promoter region (Song et al 2001). Further research is needed to evaluate the contribution of the epigenetic regulation of LOX and COX to the pathobiology of CNS disorders, including ageing-associated alterations. Also, it should be explored whether epigenetic mechanisms could be targeted pharmacologically, and thus be used therapeutically or as a research tool to study the significance of the CNS expression of LOX and COX. Experimentally, genetic manipulations that have
459
produced transgenic animals have been used to explore the functioning of eicosanoid pathways. TRANSGENIC MICE TO STUDY BRAIN 5-LOX AND COX-2 Considering the well-established role of eicosanoid pathways in inflammation, transgenic mice deficient in 5-LOX or COX-2 were generated to study the role of these genes in inflammatory conditions. Recent and limited information is available on the use of transgenic animals to study the role of these genes in the brain. 5-LOX-deficient transgenic mice that do not synthesize leukotrienes (Chen et al 1994; Funk and Chen 2000) were used to probe CNS functioning behaviourally (Uz et al 2002). The following behavioural tests were employed: elevated plus-maze, marble burying, locomotor activity, rota-road, and the spontaneous alternations in the T-maze. In an elevated plus-maze, 5-LOXdeficient mice spent a shorter time on the closed arms, a longer time on the ‘‘anxiogenic’’ open arms, and entered the open arms more frequently than the control 5-LOX-expressing mice. As another marker of lowered expression of anxious behaviour, 5LOX-deficient mice covered fewer marbles in the marble-burying anxiety test. It was concluded that 5-LOX-deficient mice are less prone to anxiety, and it was suggested that 5-LOX might contribute to affective behaviours. However, it was stressed (Uz et al 2002) that further experiments with congenic 5-LOXdeficient mice backcrossed into inbred strains are needed. Transgenic mice were also produced to study the COX-2 gene. Behavioural studies in a mouse line expressing a constitutively elevated neuronal COX-2, which produced elevated levels of prostaglandins in the brain (Andreasson et al 2001), revealed that these animals developed an age-dependent deficit in spatial memory at 12 and 20 months but not at 7 months, and a deficit in aversive behaviour at 20 months of age. These behavioural changes were associated with a parallel age-dependent increase in neuronal cell death. On the other hand, in COX-2 deficient mice, the susceptibility to ischaemic brain injury and glutamate receptor-mediated neurotoxicity was decreased (Iadecola et al 2001). These studies with transgenic mice provide further evidence that neuronal 5-LOX and COX-2 may significantly influence CNS function, and indicate that this animal model is useful for studies of brain ageing. 5-LOX AND COX-2 IN THE PATHOBIOLOGY OF NEURODEGENERATION Ageing-associated neuropathology is often accompanied by pathological alterations of cerebral vasculature (Farkas et al 2001) and by the neurotoxic processes intrinsic to the CNS, such as glutamate receptor-mediated excitotoxicity (Hof et al 2002; Manev et al 1990; Schwarcz et al 1984). The eicosanoid pathways are important players in neurodegenerative processes, not only because of their neuronal effects but also because they influence the functioning of the vascular system and inflammatory processes. Another important mechanism mediated by eicosanoid pathways is oxidative stress, which ultimately contributes to the age-dependent vulnerability of the CNS to neurotoxins and age-dependent neuronal apoptosis (Kim and Chan 2001); thus, oxidative stress and lipid peroxidation in biomembranes are able to proceed enzymatically, e.g. via 5-LOX and COX-2 pathways. Numerous ageing-associated neurodegenerative disorders, including stroke, which is more debilitating in the elderly than in young subjects, include oxidative stress in their pathobiological mechanisms. Modern methodological approaches, such as the disease-linked microarray analysis of gene expression in the CNS, also point to
460
NERVOUS SYSTEM
the involvement of eicosanoid pathways in neurodegeneration; e.g. 5-LOX in multiple sclerosis (Whitney et al 2001). Nevertheless, the best evidence for the role of 5-LOX and COX-2 in neurodegeneration is exemplified by studies of brain ischaemia. Ohtsuki et al (1995) provided the first direct evidence showing that the CNS 5-LOX pathway is affected in stroke. They observed that neuronal 5-LOX is mobilized and activated during reperfusion of the ischaemic brain and that this activation is accompanied by an increased production of leukotrienes. Recently, Ciceri et al (2001) extended these early findings by demonstrating that the stroketriggered increased formation of leukotrienes (LTC4, LTD4 and LTE4) involves an activation of ionotropic glutamate receptors, which are known to participate in stroke-associated excitotoxic neuronal death. Protection against ischaemic neuronal injury was provided by LOX inhibitors (Arai et al 2001). COX-2 has also been linked to ischaemic stroke but its role in the mechanisms of ischaemic brain injury is not fully understood. Iadecola et al (2001) demonstrated that COX-2-deficient mice have a significant reduction in the severity of brain injury produced by the occlusion of the middle cerebral artery. These authors attributed the neuroprotective action of COX-2 deficiency to the attenuation of glutamate receptor-mediated excitotoxicity. They also concluded that COX-2 may be involved in pathogenic events occurring in both the early and late stages of cerebral ischaemia and that this eicosanoid pathway may be a valuable therapeutic target for treatment of human stroke. THE CNS EICOSANOID PATHWAYS AS A PUTATIVE ‘‘ANTI-AGEING’’ DRUG TARGET The presence and activity of 5-LOX and COX-2 in the brain indicate a functional role of eicosanoid pathways in the CNS. However, a full understanding of the physiological significance of these pathways in the brain has only started to emerge. An important element is the knowledge of eicosanoid-mediated cell signalling. The specific effects of 5-LOX metabolites, leukotrienes, are mediated by a family of metabotropic leukotriene receptors. Several of these receptors have been cloned and characterized. Thus, leukotriene LTB4 binds the receptors B-LT1 and B-LT2, whereas the cysteinyl leukotrienes (cysLTs) LTC4 and LTD4 bind the receptors CysLT1 and CysLT2. Of the known leukotriene receptors, CysLT2 is expressed in the brain (Heise et al 2000; Hui et al 2001) and the CNS synthesizes their ligands, the cysteinyl leukotrienes. Leukotrienes and prostaglandins also bind the peroxisome proliferator-activated receptors (PPAR), which act as transcription factors. Three types, PPAR-a, -b, and -g, are expressed in the mammalian CNS (Cullingford et al 1998) and activation of PPAR-g is effective in inhibiting glutamate excitotoxicity in primary cultures of cerebellar granule neurons (Uryu et al 2002). In the brain, a COX metabolite, prostaglandin H2, can be converted into prostacyclin by an action of prostacyclin synthase, which is localized not only in blood vessels but also in neurons and glia (Siegle et al 2000). A novel type of prostacyclin receptor termed IP2 has been found in the CNS; activation of this receptor was effective in attenuating ischaemic brain damage (Takamatsu et al 2002). Thus, drugs to be considered for targeting the CNS eicosanoid pathways might be directed to the CNS receptors for eicosanoids. Alternatively, they could be designed as selective 5-LOX or COX2 inhibitors or activators. In addition to the requirement for specificity, for such compounds to be used in the CNS they will also require blood–brain barrier permeability. Currently, drugs that interfere with histone deactylation are primarily considered for the treatment of cancer (Johnstone 2002). However, if further research confirms that pharmacology targeted to epigenetic
mechanisms, e.g. histone acetylation and DNA methylation, could alter the neuronal expression of 5-LOX or COX-2, one could envisage a novel line of drugs for the treatment of ageingassociated brain pathologies as well.
ACKNOWLEDGEMENT Support by the National Institute on Aging, Grant RO1AG15347, is kindly acknowledged.
REFERENCES Adesuyi SA, Cockrell CS, Gamache DA and Ellis EF (1985) Lipoxygenase metabolism of arachidonic acid in brain. J Neurochem, 45, 770–776. Anderson KM, Seed T, Vos M et al (1998) 5-Lipoxygenase inhibitors reduce PC-3 cell proliferation and initiate nonnecrotic cell death. Prostate, 37, 161–173. Andreasson KI, Savonenko A, Vidensky S et al (2001) Age-dependent cognitive deficits and neuronal apoptosis in cyclooxygenase-2 transgenic mice. J Neurosci, 21, 8198–8209. Arai K, Nishiyama N, Matsuki N and Ikegaya Y (2001) Neuroprotective effects of lipoxygenase inhibitors against ischemic injury in rat hippocampal slice cultures. Brain Res, 904, 167–172. Arboix A, Garcia-Eroles L, Massons J et al (2000) Acute stroke in very old people: clinical features and predictors of in-hospital mortality. J Am Geriat Soc, 48, 36–41. Avis I, Hong SH, Martinez A et al (2001) Five-lipoxygenase inhibitors can mediate apoptosis in human breast cancer cell lines through complex eicosanoid interactions. FASEB J, 15, 2007–2009. Bishai I and Coceani F (1992) Eicosanoid formation in the rat cerebral cortex. Contribution of neurons and glia. Mol Chem Neuropathol, 17, 219–238. Boado RJ, Pardridge WM, Vinters HV and Black KL (1992) Differential expression of arachidonate 5-lipoxygenase transcripts in human brain tumors: evidence for the expression of multitranscript family. Proc Natl Acad Sci USA, 89, 9044–9048. Breder CD, Dewitt D and Kraig RP (1995) Characterization of inducible cyclooxygenase in rat brain. J Comp Neurol, 355, 296–315. Chen XS, Sheller JR, Johnson EN and Funk CD (1994) Role of leukotrienes revealed by target disruption of the 5-lipoxygenase gene. Nature, 372, 179–182. Ciceri P, Rabuffetti M, Monopoli A and Nicosia S (2001) Production of leukotrienes in a model of focal cerebral ischaemia in the rat. Br J Pharmacol, 133, 1323–1329. Costa E, Davis J, Pesold C et al (2002) The heterozygote reeler mouse as a model for the development of a new generation of antipsychotics. Curr Opin Pharmacol, 2, 56–62. Colamorea T, DiPaola R, Macchia F et al (1999) 5-Lipoxygenase upregulation by dexamethasone in human mast cells. Biochem Biophys Res Commun, 265, 617–624. Cullingford TE, Bhakoo K, Peuchen S et al (1998) Distribution of mRNAs encoding the peroxisome proliferator-activated receptor a, b, and g, and the retinoid X receptor a, b, and g in rat central nervous system. J Neurochem, 70, 1366–1375. Fan G, Beard C, Chen RZ et al (2001) DNA hypomethylation perturbs the function and survival of CNS neurons in postnatal animals. J Neurosci, 21, 788–797. Farkas E, DeJong GI, Apro E et al (2001) Calcium antagonist decrease capillary wall damage in aging hypertensive rat. Neurobiol Aging, 22, 299–309. Farrer LA, Cupples LA, Connor L et al (1991) Association of decreased paternal age and late-onset Alzheimer’s disease. An example of genetic imprinting? Arch Neurol, 48, 599–604. Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science, 294, 1871–1875. Funk CD and Chen XS (2000) 5-Lipoxygenase and leukotrienes. Transgenic mouse and nuclear targeting studies. Am J Resp Crit Care Med, 161, S120–124.
EICOSANOIDS IN CENTRAL NERVOUS SYSTEM AGEING Gati I, Bergstrom M, Csoka K et al (1990) Effects of the 5-lipoxygenase inhibitors AA-863 and U-60,257 on human glioma cell lines. Prostagland Leukotrienes Essent Fatty Acids, 40, 117–124. Ghosh J and Myers CE (1997) Arachidonic acid stimulates prostate cancer cell growth: critical role of 5-lipoxygenase. Biochem Biophys Res Commun, 235, 418–423. Heise C, O’Dowd BF, Figueroa DJ et al (2000) Characterization of the human cysteinyl leukotriene 2 receptor. J Biol Chem, 275, 30531– 30536. Hof PR, Duan H, Page TL et al (2002) Age-related changes in GluR2 and NMDAR1 glutamate receptor subunit protein immunoreactivity in cortically projecting neurons in macaque and patas monkeys. Brain Res, 928, 175–186. Hoffmann C (2000) COX-2 in brain and spinal cord—implications for therapeutic use. Curr Med Chem, 7, 1113–1120. Hoshiko S, Radmark O and Samuelsson B (1990) Characterization of the human 5-lipoxygenase gene promoter. Proc Natl Acad Sci USA, 87, 9073–9077. Hui Y, Yang G, Galczenski H et al (2001) The murine cysteinyl leukotriene 2 (CysLT2) receptor. cDNA and genomic cloning, alternative splicing, and in vitro characterization. J Biol Chem, 276, 489–495. Iadecola C, Niwa K, Nogawa S et al (2001) Reduced susceptibility to ischemic brain injury and N-methyl-D-aspartate-mediated neurotoxicity in cyclooxygenase-2-deficient mice. Proc Natl Acad Sci USA, 98, 1294–1299. Issa J-P, Ottaviano YL, Celano P et al (1994) Methylation of the oestrogen receptor CpG islands links ageing and neoplasia in human colon. Nature Genet, 7, 536–540. Jisaka M, Kim RB, Boeglin WE et al (1997) Molecular cloning and functional expression of a phorbol ester-inducible 8S-lipoxygenase from mouse skin. J Biol Chem, 272, 24410–24416. Johnstone RW (2002) Histone-deacetylase inhibitors: novel drugs for the treatment of cancer. Nature Rev Drug Discov, 1, 287–299. Kaufman WE, Andreasson KI, Isakson PC and Worley PF (1997) Cyclooxygenases and the central nervous system. Prostaglandins, 54, 601–624. Kharlamov A, Kharlamov E and Armstrong DM (2000) Age-dependent increase in infarct volume following photochemically induced cerebral infarction: putative role of astroglia. J Gerontol Ser A Biol Sci Med Sci, 55, B135–141. Kim GW and Chan PH (2001) Oxidative stress and neuronal DNA fragmentation mediate age-dependent vulnerability to the mitochondrial toxin, 3-nitropropionic acid, in the mouse striatum. Neurobiol Dis, 8, 114–126. Koistinaho J, Koponen S and Chan PH (1999) Expression of cyclooxygenase-2 mRNA after global ischemia is regulated by AMPA receptors and glucocorticoids. Stroke, 30, 1900–1905. Kuhn HG, Dickinson-Anson H and Gage FH (1996) Neurogenesis in the dentate gyrus of the adult rat: age-related decrease of neuronal progenitor proliferation. J Neurosci, 16, 2027–2033. Kumihashi K, Uchida K, Miyazaki H et al (2001) Acetylsalycylic acid reduces ischemia-induced proliferation of dentate cells in gerbils. NeuroReport, 12, 915–917. Lammers CH, Schweitzer P, Facchinetti P et al (1996) Arachidonate 5lipoxygenase and its activating protein: prominent hippocampal expression and role in somatostatin signaling. J Neurochem, 66, 147– 152. Lasa M, Brook M, Saklatvala J and Clark AR (2001) Dexamethasone destabilizes cyclooxygenase 2 mRNA by inhibiting mitogen-activated protein kinase p38. Mol Cell Biol, 21, 771–780. Lee J, Seroogy KB and Mattson MP (2002) Dietary restriction enhances neurotrophin expression and neurogenesis in the hippocampus of adult mice. J Neurochem, 80, 539–547. Li Y, Maher P and Schubert D (1997) A role for 12-lipoxygenase in nerve cell death caused by glutathione depletion. Neuron, 19, 453–463. Lindgren JA, Hokfelt T, Dahlen SE et al (1984) Leukotrienes in the rat central nervous system. Proc Natl Acad Sci USA, 81, 6212–6216. Malberg JE, Eisch AJ, Nestler EJ and Duman RS (2000) Chronic antidepressant treatment increases neurogenesis in adult rat hippocampus. J Neurosci, 20, 9104–9110. Manev H and Uz T (2002) DNA hypomethylating agents 5-aza-20 deoxycytidine and valproate increase neuronal 5-lipoxygenase mRNA. Eur J Pharmacol, 445, 149–150.
461
Manev H, Costa E, Wroblewski JT and Guidotti A (1990) Abusive stimulation of excitatory amino acid receptors: a strategy to limit neurotoxicity. FASEB J, 4, 2789–2797. Manev H, Uz T, Sugaya K and Qu T (2000) Putative role of neuronal 5lipoxygenase in an aging brain. FASEB J, 14, 1464–1469. Manev H, Uz T, Manev R and Zhang Z (2001a) Neurogenesis and neuroprotection in the adult brain. A putative role for 5-lipoxygenase? Ann NY Acad Sci, 939, 45–51. Manev H, Uz T, Smalheiser NR and Manev R (2001b) Antidepressants alter cell proliferation in the adult brain in vivo and in neural cultures in vitro. Eur J Pharmacol, 411, 67–70. Mays-Hoopes LL (1989) DNA methylation in aging and cancer. J Gerontol, 44, 35–36. Montine KS, Montine TJ, Morrow JD et al (2002) Mouse cerebral prostaglandins, but not oxidative damage, change with age and are responsive to indomethacin treatment. Brain Res, 930, 75–82. Murphy EK, Spencer RL, Sipe KJ and Herman JP (2002) Decrements in nuclear glucocorticoid receptor (GR) protein levels and DNA binding in aged rat hippocampus. Endocrinology, 143, 1362–1370. Nagayama M, Aber T, Nagayama T et al (1999) Age-dependent increase in ischemic brain injury in wild-type mice and in mice lacking the inducible nitric oxide synthase gene. J Cerebr Blood Flow Metab, 19, 661–666. Nagy Z (2000) Cell cycle regulatory failure in neurons: causes and consequences. Neurobiol Aging, 21, 761–769. Nakayama H, Jorgensen HS, Raaschou HO and Olsen TS (1994) The influence of age on stroke outcome. The Copenhagen Stroke Study. Stroke, 25, 808–813. Nestler EJ, Barrot M, DiLeone RJ et al (2002) Neurobiology of depression. Neuron, 34, 13–25. Ng H-H and Bird A (1999) DNA methylation and chromatin modification. Curr Opin Genet Dev, 9, 158–163. O’Banion MK (1999) Cyclooxygenase-2: molecular biology, pharmacology, and neurobiology. Crit Rev Neurobiol, 13, 45–82. Ohtsuki T, Matsumoto M, Hayashi Y et al (1995) Reperfusion induces 5lipoxygenase translocation and leukotriene C4 production in ischemic brain. Am J Physiol, 268, H1249–1257. Pai R, Soreghan B, Szabo IL et al (2002) Prostaglandin E2 transactivates EGF receptor: a novel mechanism for promoting colon cancer growth and gastrointestinal hypertrophy. Nature Med, 8, 289–293. Paulsen M and Ferguson-Smith AC (2001) DNA methylation in genomic imprinting, development and disease. J Pathol, 195, 97–110. Petronis A, Paterson AD and Kenedy JL (1999) Schizophrenia: an epigenetic puzzle? Schizophren Bull, 25, 639–655. Piomelli D (1994) Eicosanoids in synaptic transmission. Crit Rev Neurobiol, 8, 65–83. Piomelli D (2001) The ligand that came from within. Trends Pharmacol Sci, 22, 17–19. Qu T, Uz T and Manev H (2000) Inflammatory 5-LOX mRNA and protein are increased in brain of aging rats. Neurobiol Aging, 21, 647– 652. Rao AM, Hatcher JF, Kindy MS and Dempsey RJ (1999) Arachidonic acid and leukotriene C4: role in transient cerebral ischemia of gerbils. Neurochem Res, 24, 1225–1232. Riddick CA, Ring WL, Baker JR et al (1997) Dexamethasone increases expression of 5-lipoxygenase and its activating protein in human monocytes and THP-1 cells. Eur J Biochem, 246, 112–118. Schwarcz R, Foster AD, French ED et al (1984) Excitotoxic models for neurodegenerative disorders. Life Sci, 35, 19–32. Shimizu T, Takusagawa Y, Izumi T et al (1987) Enzymic synthesis of leukotriene B4 in guinea-pig brain. J Neurochem, 48, 1541–1546. Siegle I, Klein T, Zou M-H et al (2000) Distribution and cellular localization of prostacyclin synthase in human brain. J Histochem Cytochem, 48, 631–641. Song SH, Jong HS, Choi HH et al (2001) Transcriptional silencing of cyclooxygenase-2 by hyper-methylation of the 50 CpG islands in human gastric carcinoma cells. Cancer Res, 61, 4628–4635. Takamatsu H, Tsukada H, Watanabe Y et al (2002) Specific ligand for a central type prostacyclin receptor attenuates neuronal damage in a rat model of focal cerebral ischemia. Brain Res, 925, 176–182. Uhl J, Klan N, Rose M et al (2002) The 5-lipoxygenase promoter is regulated by DNA methylation. J Biol Chem, 277, 4374–4379.
462
NERVOUS SYSTEM
Uryu S, Harada J, Hisamoto M and Oda T (2002) Troglitazone inhibits both post-glutamate neurotoxicity and low-potassium-induced apoptosis in cerebellar granule neurons. Brain Res, 924, 229–236. Uz T, Pesold C, Longone P and Manev H (1998) Aging-associated upregulation of neuronal 5-lipoxygenase expression: putative role in neuronal vulnerability. FASEB J, 12, 439–449. Uz T, Dwivedi Y, Savani PD et al (1999) Glucocorticoids stimulate inflammatory 5-lipoxygenase gene expression and protein translocation in the brain. J Neurochem, 73, 693–699. Uz T, Manev R and Manev H (2001a) 5-Lipoxygenase is required for proliferation of immature cerebellar granule neurons in vitro. Eur J Pharmacol, 418, 15–22. Uz T, Dwivedi Y, Qeli A et al (2001b) Glucocorticoid receptors are required for upregulation of neuronal 5-lipoxygenase (5-LOX) expression by dexamethasone. FASEB J, 15, 1792–1794. Uz T, Dimitrijevic N, Tueting P and Manev H (2002) 5-Lipoxygenase (5LOX)-deficient mice express reduced anxiety-like behavior. Restor Neurol Neurosci, 20, 15–20.
van Praag H, Schinder AF, Christie BR et al (2002) Functional neurogenesis in the adult hippocampus. Nature, 415, 1030–1034. van Rossum GS, Bijvelt JJ, van den Bosch H et al (2002) Cytosolic phospholipase A2 and lipoxygenase are involved in cell cycle progression in neuroblastoma cells. Cell Mol Life Sci, 59, 181–188. Verdaguer E, Garcia-Jorda E, Canudas AM et al (2002) Kainic acidinduced apoptosis in cerebellar granule neurons: an attempt at cell cycle re-entry. NeuroReport, 13, 413–416. Watanabe T, Medina JF, Haeggstrom JZ et al (1993) Molecular cloning of a 12-lipoxygenase cDNA from rat brain. Eur J Biochem, 212, 605–612. Whitney LW, Ludwin SK, McFarland HF and Biddison WEJ (2001) Microarray analysis of gene expression in multiple sclerosis and EAE identifies 5-lipoxygenase as a component of inflammatory lesions. Neuroimmunology, 121, 40–48. Yamagata K, Andreasson KI, Kaufmann WI et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386.
42 Arachidonate Metabolites in the Neurophysiological System: the Fever Pathway Ji Zhang and Serge Rivest Laboratory of Molecular Endocrinology, CHUL Research Centre and Department of Anatomy and Physiology, Laval University, Que´bec, Canada
When infectious microorganisms invade the body through its natural barriers, an array of systemic reactions promptly develops that serves to mitigate the deleterious effects of the invading pathogens and, ultimately, to restore health. Fever, altered neuroendocrine, cardiovascular and gastric functions, increased metabolism, behavioural changes, acute-phase protein production and activation of the immune system are characteristic host responses to systemic infection, injury and inflammation. The central nervous system (CNS) mediates a coordinated set of autonomic, endocrine and behavioural responses that constitute the cerebral component of the acute inflammation. Fever is an adaptive response to inflammatory stimulation mediated by the hypothalamus. The invading organisms and/or their degradation products raise the hypothalamic temperature ‘‘set-point’’ from the normal level to a higher level. A combination of heat conservation and heat production in the peripheral tissue results in an increased body temperature (viz. the fever) to match the elevated set-point. Experimental models of fever have shown that the febrile response depends strongly on the pyrogen dose and the route of its administration. When the bacterial endotoxin lipopolysaccharide (LPS) is administrated intravenously at a neutral (308C) ambient temperature, a low, near-threshold (1 mg/ kg) dose induces the first febrile phase only (monophasic fever), which is characterized by a relatively long latency and a temperature peak at *80 min. If the dose is somewhat higher (10–100 mg/kg), the first phase peaks earlier (50–70 min) and is followed by another temperature increase (second febrile phase), peaking at *140 min. The response to very low pyrogen doses (monophasic fever) can be considered a pure first (early) febrile phase, whereas at higher pyrogen doses the early phase becomes ‘‘contaminated’’ by a later, second phase (Romanovsky et al 1996a, 1996b). Each febrile phase is characterized by its own complex of associated ‘‘sickness symptoms’’: hyperalgesia, motor hyperexcitability, arterial hypertension and an increase in vigilance (the early phase) vs. hypoalgesia, low motor activity, norrmotension or hypotension, and sleepiness (the late febrile phase). As the vast majority of infections occur outside the brain, however, bacterial products or cytokines must somehow trigger the hypothalamic neurons inside the brain. The question was raised of how circulating pyrogens signal the brain. The purpose of the present chapter is to provide an integrative account of: 1. Neuronal and non-neuronal pathways involved in the receipt and transmission of inflammatory information from the periphery to higher CNS centres. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
2. The cellular and molecular bases of mechanisms underlying the induction of fever. 3. Essentially, we will discuss the involvement of PGE2 in this complex processes.
INTERACTIONS BETWEEN THE IMMUNE SYSTEM AND THE CENTRAL NERVOUS SYSTEM Systemic Inflammation and Exogenous Pyrogens Inflammation is a general name for reactions occurring after most kinds of tissue injuries or infections or immunological stimulation as a defence against foreign or altered endogenous substances. The local inflammatory reaction is characterized by an initial increase of blood flow to the site of injury, enhanced vascular permeability and the ordered, directional influx and selective accumulation of different effector cells from the peripheral blood at the site of injury. These cells mount a rapid, non-specific phagocytic response. The systemic inflammation is characterized by an elevation of body temperature, viz. fever, and an acute phase reaction that is initiated by a number of cytokines with inflammatory activities secreted by a variety of cell types. The cascades of inflammatory cytokines in different tissues represent amplification and regulatory pathways controlling the development of acute phase responses in vivo. Systemic injection with LPS, a component of the cell wall of Gram-negative bacteria, is frequently used as an exogenous pyrogen to mimic the endogenous effect during inflammation and sepsis. LPS is a powerful immune challenge associated with an increase in the circulating levels of different cytokines, such as tumour necrosis factor a (TNFa), interleukin (IL)-1b, and IL-6, in a pattern similar to that seen in natural infection (Andersson et al 1992). It is capable of inducing fever, septic shock and the acute-phase response, including the activation of the hypothalamic–pituitary–adrenal (HPA) axis, which was evidenced by the fact that intravenous administration of LPS to laboratory rodents produces marked elevations in ACTH and corticosterone secretion within 30– 60 min. This treatment also causes a wide expression of the immediate early genes (IEGs) c-fos and NGFI-B mRNA (indexes of cellular activation) throughout the rat brain, suggesting complex and redundant neuronal circuits to activate the endocrine hypothalamus and interfere with other systems (Elmquist et al 1996; Elmquist and Saper 1996; Laflamme et al 1999a; Rivest 1995; Rivest et al 2000). A complex of LPS and the serum protein LPS-binding protein (LBP) initiates signals through membrane-
464
NERVOUS SYSTEM
Figure 42.1 Hypothetical mechanisms of interaction between LBP, CD14 and TLR4 to function as the LPS signal transducer leading to activation of NF-kB and MAP kinases by Gram-negative bacteria. It is possible that CD14 acts as the principal LPS binding protein on the surface of monocytic cells and the newly formed complex reaches adjacent TLR4 receptors, which transduce the LPS signal via the general intracellular adaptor protein MyD88 (myeloid differentiation factor 88) that has the ability to interact with TLRs through its own carboxyl-terminal TIR domain. MyD88 recruits the serine kinase IRAK to engage the proinflammatory signal transduction pathways via its amino-terminal death domain. It has recently been shown that MD-2, a secreted protein with affinity for TLR4, can complement the LPS response if co-expressed with TLR4 in 293 cells. Hence, MD-2 may be required alongside TLR4 for reconstitution of the authentic LPS signal transduction pathway. Adapted from Beutler, 2000
bound CD14 in monocytes and myeloid cells. However, because of the presence, in the serum, of soluble CD14 that can substitute for membrane-bound CD14, CD14-deficient cells, such as endothelial and epithelial cells, also respond to LPS. Although LBP and CD14 were identified as factors that bind LPS, other evidence has suggested the presence of a co-receptor, i.e. Toll-like receptor 4 (TLR4), which transmits the LPS signal across the cell membrane (see Figure 42.1). The coexistence of both TLR4 and CD14 receptors in the circumventricular organs (CVOs) of the brain (Laflamme et al 2001) may be recognizing molecules from the endotoxin to trigger the proinflammatory signal transduction events in structures that can be reached from the systemic circulation. This causes activation of NF-kB signalling and transcription of proinflammatory molecules, first within the CVOs and thereafter across the brain parenchyma (Lacroix and Rivest 1998, 2001; Nadeau and Rivest 2000, 2001). Proinflammatory Cytokines and Endogenous Pyrogens Proinflammatory cytokines are produced by different cells of the myeloid lineage upon presentation of an antigen, and their secretion into the bloodstream is believed to be the key step in the neuronal activity and the subsequent neurophysiological responses that take place during immune stimuli. Cytokines influence many neuroendocrine systems, the most prominent of which is the activation of the HPA axis, resulting in release of adrenocorticotrophic hormone (ACTH) and glucocorticoids (Rivest et al 1995, 2000; Rivest 1995, 1999, 2001; Rivest and
Rivier 1995). Proinflammatory cytokine IL-1, especially the b form (IL-1b), is probably the most important molecule capable of modulating cerebral functions during systemic and localized inflammatory insults (Laflamme et al 1999b; Rivest 2001). Studies assessing the expression of IEG and neuropeptide mRNA within the rat PVN after administration of IL-1b have suggested a CNS site of action of IL-1 administrated via either peripheral or central routes. IL-1b administrated intravenously (Ericsson et al 1994) or intraperitoneally (Brady et al 1994) induces c-fos mRNA or Fos protein expression in the rat paraventricular nucleus of the hypothalamus (PVN). The Fos signal in the PVN co-localizes with CRF immunoreactivity or mRNA after either peripheral route or central route of IL-1b treatment (Ericsson et al 1994), indicating cellular activation of CRF-containing neurons. Despite the failure to detect IL-1 in the circulation, it is well established that IL-1 is involved in the development of the febrile response to exogenous pyrogens. Direct injection of low doses of IL-1 into the brain increases body temperature (Rothwell 1991; Rothwell and Hopkins 1995). The administration of neutralizing antibodies or an antagonist to IL-1 inhibits fever induced by various external infections or inflammatory stimuli in experimental animals. Recent data from studies using cytokine-deficient mice showed that IL-1b plays only a minor role in the induction of fever in response to injection of LPS (Zheng et al 1995). IL-1b null mice were unable to mount a normal acute-phase inflammatory response following turpentine injection, which causes localized leukocyte infiltration, oedema and tissue damage. The IL-1bdeficient mice did not develop fever following turpentine administration. In contrast, these mice responded similarly to
PGE2 AND NEUROINFLAMMATION wild-type controls in systemic inflammation, including fever induced by LPS. These data indicate that the molecular mechanism of LPS-induced fever is complex and employs multiple endogenous mediators, which may substitute for the lack of IL-1 (Zheng et al 1995). The recent characterization of receptor of the toll family has provided clear evidence that LPS is an exogenous ligand that has its own endogenous receptor and can trigger brain signalling by itself without the systemic release of proinflammatory cytokines (Laflamme and Rivest 2001). This explains the lack of significant effects of LPS in cytokine-deficient mice, at least on fever and some other autonomic functions (Laflamme et al 1999b; Laflamme and Rivest 2001). The binding of IL-1b to its cognate type I receptor leads to the formation of the IL-1 receptor-associated kinase (IRAK)–TNF receptor-associated factor 6 (TRAF6) complex, which recruits the general adaptor protein MyD88 and activates NIK/IKK kinases involved in the phosphorylation and degradation of IkBa (see Figure 42.2). NF-kB is then translocated into the nucleus and may bind to its kB consensus sequence on target genes (Baeuerle and Henkel 1994; Baeuerle and Baltimore 1996; Baeuerle 1998), such as cyclooxygenase-2 (COX-2), which is the best candidate gene involved in fever genesis. Other cytokines implicated in the febrile response include TNFa and IL-6. Evidence exists that TNFa acts in synergy with IL-1 during fever (Hopkins and Rothwell 1995). Like IL-1b, intravenous TNFa injection has a profound stimulating influence on the endothelium of the brain blood vessels. Indeed, a strong induction of IkBa and COX-2 mRNA was found within
465
endothelial cells, limiting the CNS vascular system in response to intravenous injection of TNFa (Lacroix and Rivest 1998; Laflamme et al 1999b). The binding of TNFa to its P55 receptor leads to the formation of the TNF-R1-associated death domain (TRADD)–TRAF2 complex, which activates the NF-kB signalling events (see Figure 42.2). The role of IL-6 in the febrile response is quite controversial. IL-6 has been defined as one of the principal endogenous pyrogens, from the observation that IL-6deficient mice are unable to develop normal fever in response to both LPS and IL-1b (Chai et al 1996). It has been also demonstrated that prostaglandins mediate IL-6-induced fever (Dinarello et al 1991) and HPA axis activation (Navarra et al 1992), but IL-6 is unable to stimulate prostaglandin formation in cerebral microvessels (Bishai and Coceani 1996; Vallie`res and Rivest 1999) or induce COX-2 mRNA synthesis in the rat brains (Lacroix and Rivest 1998). Possible Pathways by which IL-1b Signals the Brain The transport of soluble substances out of vascular compartments and into perivascular tissues (and vice versa) occurs via either paracellular or transcellular mechanisms. Within the cerebrovasculature, the paracellular route is particularly impermeable due to the presence of the blood–brain barrier (BBB). The BBB consists primarily of non-fenestrated endothelial cells that are connected by tight junctions. The paracellular ultrafiltration of solutes into and out of tissues that occurs in peripheral vascular beds does not,
Figure 42.2 The proinflammatory signal transduction pathways involving the nuclear factor kappa B (NF-kB). p50 and p65 are the two most common DNA-binding subunits of the NF-kB dimer that have the ability to trigger the transcription of numerous target genes. See text for details and abbreviations
466
NERVOUS SYSTEM
therefore, occur in most cerebrovascular beds, at least while the BBB remains intact. Furthermore, the large molecular sizes (13– 15 kDa) and hydrophilic nature of cytokines preclude their transcellular movement by simple diffusion to any appreciable extent. Indeed, early studies concluded that the BBB was impermeable to IL-1b (Blatteis 1990). It is nevertheless possible that endogenous cytokines diffuse across the BBB during extreme periods of fever or in the presence of high circulating levels during a long period of time (Turnbull and Rivier 1999). A number of regions of CNS relatively devoid of BBB permit cytokine interaction with the neuronal elements. The CVOs not only contain capillaries with rather greater permeability than the rest of the CNS, but the capillary density in these regions is extraordinarily high (Johnson and Gross 1993). Therefore, the CVOs have been proposed as potential sites of IL-1 action near the structures lining the AV3V region (in particular the OVLT), the SFO, the ME and in the brainstem (AP). The medullary group of cells may play an important role in the processing of visceral sensory information carried by sensory components of the vagus and glossopharyngeal nerves. The vagus may provide a rapid communication pathway for cytokine signalling between the periphery and the brain (Watkins et al 1995). This was evidenced by the fact that subdiaphragmatic vagotomy inverted the core temperature rises into falls and abrogated the concomitant increases in preoptic PGE2 levels induced in conscious guineapigs by intravenous LPS (Sehic et al 1996a, 1996b; Sehic and Blatteis, 1996). A strong body of evidence now supports the concept that cytokines, particularly IL-1, act on endothelium, brain–fluid interfaces with consequent release of local signalling molecules, such as prostaglandins (PGs). Indeed, brain microvessels exhibit a robust constitutive expression of the IL-1 type 1 receptor transcript (IL-1R1) (Ericsson et al 1995), whereas systemic intravenous injection of recombinant rat IL-1b causes a profound transcriptional activation of the gene encoding cyclooxygenase-2 (COX-2) (Lacroix and Rivest 1998) and IkBa (Laflamme and Rivest 1999) within the endothelial cells of the CNS blood vessels. The microsomal prostaglandin E synthase (mPGES) is also inducible by the proinflammatory cytokine IL-1b (Jakobsson et al 1999) and is co-localized with COX-2 in the perinuclear region of the endothelial cells (Ek et al 2001; Yamagata et al 2001). Complementary System, a Modulator of Inflammatory Response The complement system consists of more than 30 various glycoproteins present in the blood serum in an inactive state and these are activated by immune complexes (the classical pathway), by carbohydrates (the lectin pathway) or by other substances, mainly of bacterial origin (the alternative pathway). The complement system is a potent mechanism for initiating and amplifying inflammation. The intravenous administration of LPS triggers, within several minutes, the complement cascade via both the classical and alternative pathways, resulting in the production in blood of the anaphylatoxic C fragments, C4a, C3a and C5a. Proteins of the complement system produced by systemic immune cells (mainly Kupffer cells in the liver) may not reach the cerebral tissue, due to the presence of the BBB, but extrahepatic cellular sources of C3a and C5a are now known to be essential in the initiation and regulation of the inflammatory response. C5 mRNA is widely distributed across the brain parenchyma, remaining unchanged in response to circulating LPS. The C5aR mRNA levels increased in the cerebral endothelium at 1 h postLPS challenge, and the signal became gradually positive in microglial cells surrounding the capillaries and thereafter across the brain parenchyma (Nadeau and Rivest 2001). A single bolus
of endotoxin also caused a profound transcriptional activation of C3, C3aR and the factor B in numerous non-neuronal structures (Nadeau and Rivest 2001). The induction wave supports the concept of an integrated response of the complement system during endotoxaemia. C3a stimulates PGE2 production by Kupffer cells (KC) within 2 min (Puschel et al 1993). C5a and its desArg form are also potent stimuli of PGE2 synthesis (Sehic et al 1998). Pretreatment of rabbits and pigs with C-depleting agents (i.e. CVF) prevented the rise in plasma prostaglandins induced by LPS (Fink et al 1989), attenuated the first of the biphasic core temperature (Tc) rises after intravenous LPS, inverted the second into a Tc fall, and greatly reduced the usual fever-associated PGE2 increase in the preoptic area (Sehic et al 1998). NF-kB, an Important Transcription Factor in the Regulation of Inflammatory Response In recent years, the nuclear transcription factor NF-kB has been thought of as the ‘‘central switch’’ or ‘‘central mediator’’ of the immune response controlling the synthesis of different cytokines and chemokines, MHC molecules, proteins involved in antigen presentation, receptors required for neutrophil adhesion and migration across blood vessels. Normally, NF-kB is present in the cell cytoplasm, forming an inactive complex with an inhibitor known as IkBa (see Figure 42.2). After extracellular stimulation by growth factors, mitogens and cytokines that activate mitogenactivated protein (MAP) kinases, IkBa is phosphorylated by NFkB inducible kinase (NIK)/IkB kinase (IKK), ubiquitinated and degraded by cytoplasmic proteasomes (Baeuerle 1998). Free active NF-kB (the most common complex is the P50/P65 heterodimer) is then translocated into the nucleus, where it is able to regulate transcription of various genes in binding to its kB consensus sequence. After its degradation, IkBa is rapidly resynthesized to act as an endogenous inhibitory signal for NF-kB, and monitoring its de novo expression is a powerful tool to investigate the activity of the transcription factor within the CNS (Laflamme and Rivest 1999, 2001; Laflamme et al 1999b). NF-kB is activated by more than 150 stimuli and results in the induction of more than 150 genes that influence cell survival and maintenance of normal functional integrity. NF-kB can be activated by exposure of cells to LPS or inflammatory cytokines such as TNFa or IL-1, viral infection, UV irradiation, B or T cell activation, and by other physiological and non-physiological stimuli. Indeed, systemic administration of endotoxin, IL-1 or TNFa provoked a rapid expression of IkBa mRNA, first in the endothelium of the cerebral blood vessels and afterwards within parenchymal microglia across the rat brain (Laflamme and Rivest 1999). Genes regulated by NF-kB include COX-2, iNOS, prostaglandin E synthase-2, numerous cytokines, chemokines and vascular adhesion molecules. In response to systemic immunogenic stimuli, IkBa and COX-2 mRNA are strongly induced, and both of these two genes are co-localized within the brain endothelium (Laflamme et al 1999b). It is therefore likely that circulating proinflammatory molecules stimulate prostaglandin (PG) production via transcriptional activation of COX-2 through NF-kB signalling pathways within cells of the BBB. CONTRIBUTION OF PGE2 IN FEVER PRODUCTION PGE2 and Fever Over the last several years, accumulating evidence supports the critical importance of COX-2 in fever genesis. LPS, IL-1b and TNFa induce COX-2 mRNA in the brain microvasculature (Breder and Saper 1996; Cao et al 1999; Lacroix and Rivest 1998).
PGE2 AND NEUROINFLAMMATION NS-398, a selective COX-2 inhibitor, is able to completely inhibit fever generation in rats subjected to intraperitoneal LPS injection (Cao et al 1997a, 1997b). Oral administration of the selective COX-2 inhibitor DFU, 2 h after intravenous injection, reduced fever in squirrel monkeys (Chan et al 1997) and nimesulide, a partially selective COX-2 inhibitor, reduced fever induction in a dose-dependent fashion in rats injected subcutaneously with yeast (Taniguchi et al 1998). In this latter study, brain PGE2 levels were also found to be dose-dependent in mediating fever induction. A recent study by Li and his colleagues (1999) demonstrated that COX-2 gene-deficient mice are unable to develop a full fever in response to the intraperitoneal (i.p.) administration of a pyrogenic dose of LPS. The exact PG subtype(s) and the site(s) of action within the brain involved in the fever have recently been unravelled and compelling evidence points in the PG of E2 type. First, it is a potent hyperthermic agent acting directly or indirectly on thermoregulatory neurons in the preoptic anterior hypothalamus, the primary brain site in which body temperature is regulated. Microinjection of PGE1, PGE2 or PGE receptor agonists into the brain of rats has shown that PGE acts on the region at the anteroventral tip of the third ventricle, including the OVLT and the adjacent preoptic area, to produce fever (Oka et al 1997; Scammell et al 1996, 1998). Second, PGE2 levels increased in this region during the febrile response to lipopolysaccharide (Sehic et al 1996a). Third, COX inhibitors inhibit pyrogen fever, in parallel with reversal of PGE2 synthesis and the congenital absence of the PGE2 EP3 receptor impairs the febrile response to both exogenous pyrogen and endogenous pyrogen (Ushikubi et al 1998). This assumes that all cytokines and chemokines of peripheral origin, or any direct action of LPS independent of the induction of cytokines, always require signalling by PGE2 for fever induction. There is no doubt that PGE2 is an essential and diffusible signal for eliciting the fever response. Biosynthesis of PGE2 Prostanoids, a group of 20-carbon unsaturated fatty acid derivates, are produced via a complex enzyme cascade regulated by two principal enzymes, phospholipase A2 (PLA2) and cyclooxygenase. PLA2 mobilizes arachidonic acid from cellular phospholipids, which is then metabolized by COX to prostaglandin H2. PLA2 exists in both calcium-dependent and -independent isoforms. The extracellular or secretary (sPLA2) enzyme is activated continuously by the levels of calcium found in the extracellular environment, whereas the intracellular or cytosolic (cPLA2) enzyme is activated via increases in intracellular calcium elicited by inflammatory mediators. Cyclooxygenase also has two isoforms: COX-1 and COX-2 (see below). They catalyse two reactions. First, cyclooxygenase activity adds molecular oxygen to the unsaturated fatty acid arachidonic acid, generating prostaglandin G2 (PGG2). PGG2 is then converted to PGH2 by the peroxidase activity of the enzyme. Once generated, PGH2 is rapidly converted to prostaglandins (PGD2, PGE2, PGF2a), prostacylin (PGI2), and thromboxane A2 (TXA2) by tissue-specific synthases, whereas prostaglandin E synthase was recently identified and designated as microsomal PGES (mPGES) and cytosolic PGES (cPGES). These molecules or their derivates interact with specific receptors to modulate cell function. The diversity of the tissue-specific synthases and receptors gives rise to a wide range of potential biological functions for the prostanoids. Prostaglandin G, PGH, PGI and TXA are chemically unstable and are degraded into inactive products under physiological conditions, with a half-life of 30 s to a few minutes. Other PGs, although chemically stable, are metabolized quickly (Narumiya et al 1999). It is therefore believed that prostanoids work locally,
467
acting only in the vicinity of the site of production to serve as potent autocrine and paracrine mediators in a wide variety of physiological processes. Although PGE2 rises in blood promptly after the entry of microorganisms or after systemic exogenous pyrogen or endogenous pyrogen administration, it is now generally accepted that the PGE2 detected in the brain is not derived from the blood, but rather is produced directly in the brain. Cyclooxygenase and Prostaglandin E Synthase in the CNS: Cellular Source of PGE2 Cyclooxygenase, a rate-limiting enzyme in the formation of PGs, was isolated in 1976 (Hemler and Lands 1976) and cloned in 1988 (Merlie et al 1988). Two isoforms of COX have been characterized: the constitutive isoform COX-1 and the inducible isoform COX-2. COX-1 and COX-2 proteins show 60% homology in amino acid sequence, with the size of the mRNA for the inducible enzyme approximating 4.5 kb, and that of the constitutive enzyme being 2.8 kb. Both enzymes have a molecular weight of 71 kDa, and slightly different active sites for the natural substrate. COX-1 is expressed in most tissues under basal conditions and has clear physiological functions. COX-2 is induced as an immediate early gene in a wide variety of cell types and in response to a wide variety of stimuli and is implicated in the inflammatory responses. In the central nervous system, COX-1 is found in neurons throughout the brain, but it is most abundant in the forebrain, where prostaglandins may be involved in complex, integrative functions, such as control of the autonomic nervous system and in sensory processing. The site of COX-2 mRNA in the brain was first reported by Yamagata et al (1993). This group revealed, by using the in situ hybridization technique, that COX-2 mRNA was constitutively expressed in the neurons of some discrete telencephalic regions, including the cerebral cortex, hippocampus and amygdala. The level of COX-2 mRNA in the neurons was highly dependent on the neuronal activity; electrical convulsion or treatment with excitatory amino acid increased the amount of COX-2 mRNA, whereas an NMDA receptor antagonist decreased it. The distribution of COX-2-like immunoreactivity (COX-2-IR) in the rat brain was reported by Breder et al. (1995). It was mostly in line with that of COX-2 mRNA in the telencephalic region, although some differences were observed in the hypothalamic and lower brainstem regions. In order to clarify the brain sites where COX-2 is induced by pyrogenic stimuli, several groups conducted a series of experiments using in situ hybridization or/and immunohistochemistry techniques in animals challenged by inflammatory stimuli. COX-2 mRNA is induced in brain tissue by lipopolysaccharide (LPS), IL-1 and TNFa (Lacroix and Rivest 1998). In contrast, IL-6 is unable to activate prostaglandin formation in cerebral microvessels (Bishai and Coceani 1996). When injected systematically, it did not yield any COX-2 mRNA in these non-neuronal cells (Lacroix and Rivest 1998). There is compelling evidence, using in situ hybridization combined with immunohistochemistry, that COX2 is essentially expressed within the cerebral endothelium during systemic inflammatory challenges (Rivest 1999). A robust COX-2 mRNA signal is rapidly detected in the cerebral microvessels in response to i.v. IL-1b and TNFa injection, as well as after an intramuscular (i.m.) turpentine insult (Blais and Rivest 2001, Rivest 1999, 2001). These COX-2-expressing cells are determined as endothelium, because they are positive for a marker of the cerebral endothelium, i.e. von Willebrand Factor (vWF). Recent studies on microsomal-type prostaglandin E synthase, the terminal enzyme for PGE2 biosynthesis, provided conclusive evidence that brain endothelial cells are the sites of PGE2 production in response to a peripheral pyrogenic challenge
468
NERVOUS SYSTEM
(Yamagata et al 2001). By in situ hybridization, they observed that LPS-induced mPGES mRNA signals were mainly associated with brain blood vessels in the whole brain area. Immunohistochemical study demonstrated mPGES-like immunoreactivity in the perinuclear region of brain endothelial cells. Furthermore, mPGES was co-localized with COX-2, the enzyme essential for the production of the mPGES substrate PGH2. These two enzymes were functionally linked and this link is essential for inflammation and fever, since inhibition of COX-2 activity resulted in suppression of both PGE2 level in CSF and fever. PGE2 Receptors in the Brain: Distribution and Functional Studies Local production of PGE2 might therefore be a crucial step within the CNS to mediate the effects of circulating cytokines on the neuronal circuitry involved in the activation of the HPA axis and autonomic functions. As mentioned, the endothelium of the microvasculature is the source of PG formation into the brain during systemic inflammatory challenges. It is interesting to note that inflammation-induced COX-2 and mPGES are rather unspecific across the cerebral blood vessels and small capillaries, while the neuronal activity is limited to selective nuclei, including the endocrine hypothalamus. It is consequently possible that expression of specific PGE2 receptors within parenchymal cells adjacent to the site of production determines the action of the PG in the brain. Classic prostanoid receptors comprise a family of eight encoding transmembrane G-protein-coupled receptors. These receptors are classified on the basis of selective affinities for naturally occurring prostanoids. There are distinct receptors for TXA2, PGI2, PGF2a, PGD2 (viz. TP, IP, FP and DP, respectively) and four different receptors for PGE (EP1–4). Multiple alternatively spliced isoforms exist for the PGE receptors (EP3a, 3b, 3g) (Namba et al 1993). They share common extracellular and membrane-spanning regions, but differ in intracellular and carboxy-terminal domains. Each receptor is associated with a unique G-protein and consequently a unique second messenger system, namely elevation of intracellular Ca2+ (EP1) and stimulation (EP2, EP4) or inhibition of (EP3) of adynylate cyclase. Despite the presence of some conserved sequences, overall homology among the prostanoid receptors is quite limited, in the range 20–30%. On the other hand, the homology of a given type or subtype of receptor among various species is considerably higher (Narumiya et al 1999). Each of the eight types and subtypes of receptors shows selective ligand-binding specificity that distinguishes it from the others. In addition to transmembrane receptors, the peroxisome proliferator-activated receptor-g (PPAR-g) is a member of the nuclear receptor family of transcription factors that can be activated by binding to PGD derivates, such as 15-deoxy-D12,14 prostaglandin J2 (15d-PGJ2). In 1988, the distribution of [3H] PGE2 binding sites, presumably PGE2 receptors, was first demonstrated in monkey diencephalons (Watanabe and Hayaishi 1988), followed by more detailed analysis of [3H] PGE2 binding sites in the rat brain (Matsumura et al 1990, 1992). PGE2-binding sites were located in a number of discrete brain regions, including some thalamic and hypothalamic nuclei, ventral hippocampus, central grey, superior colliculus, parabrachial nucleus (PB), locus coeruleus (LC), raphe nuclei, spinal trigeminal nuclei, and the nucleus of solitary tract (NTS). In situ hybridization was widely used for the distinctive distribution of different subtypes of PGE2 receptors. EP1 mRNA was detected in the paraventricular (PVN) and supraoptic nuclei (SON) of the hypothalamus in the mouse brain (Batshake et al 1995). EP3 mRNA was widely distributed over the central nervous system (Sugimoto et al 1994), including the neurons of the cortex,
hippocampus, thalamus, hypothalamus, midbrain and lower brainstem. Recently, we reported a very distinct pattern of EP2 and EP4 receptors throughout the rat brain (Zhang and Rivest 1999). EP2 receptor mRNA was detected in the bed nucleus of the stria terminalis (BnST), lateral septum (LS), subfornical organ (SFO), ventromedial hypothalamic nucleus (VMH), central nucleus of the amygdala (CeA), LC and the area postrema (AP), whereas EP4 receptors were located in regions that are likely involved in the control of neuroendocrine and autonomic activities. These include the paraventricular nucleus of the hypothalamus (PVN), supraoptic nucleus (SON), parabrachial nucleus (PB), nucleus of the solitary tract (NTS) and the caudal ventrolateral medulla (cVLM). The most dramatic change was the profound transcriptional activation of the gene encoding the EP4 receptor over the parvocellular CRF neurons of the hypothalamic PVN, in response to different experimental models of systemic inflammation. Another interesting result is the expression of EP4 in the A2/C2 and A1 cell groups. Immune challenges (LPS or proinflammatory cytokines) also stimulated the EP4 biosynthesis within the A1 noradrenergic cells, whereas the A2/C2 group remained without significant effects (Zhang and Rivest 1999). Of interest is the fact that EP4-expressing neurons of the PVN, NTS and cVLM had immunoreactivity to Fos in animals that received systemic LPS and IL-1b, and it is possible that such activation depends on the local synthesis of PGE2 by cells of the microvasculature penetrating these regions. Indeed, we observed the same activation pattern in rats that were treated intracerebroventricularly with PGE2 (Zhang and Rivest 1999, 2000). Distinctive distributions of all four EP receptor mRNAs were found within the PGE2-sensitive anteromedial preoptic region (Oka et al 2000; Zhang and Rivest 1999, 2000). Among these, only the EP4 receptor mRNA was strongly expressed throughout the PGE2-sensitive-region, including the organum vasculosum of the lamina terminalis (OVLT), the ventromedial preoptic nucleus (VMPO) and the median preoptic nucleus (MnPO). Most Fosimmunoreactive neurons in this region after i.v. LPS also contained EP4 receptor mRNA. In contrast, little EP2 (Zhang and Rivest 1999) and EP3 receptor mRNA (Oka et al 2000) was found within the OVLT and VMPO and there is no clear evidence that these EP2 and EP3 receptor mRNA-expressing neurons were activated by immune challenges. Although, EP1 receptor mRNA was found throughout the PGE2-sensitive region, its expression was weak and about half of the Fos-immunoreactive cells contained EP1 receptor mRNA (Oka and Hori 1994). Therefore, taking the above findings into account, it led us to believe that EP4 was the key binding and functional receptor to mediate the neuronal activity including HPA axis activation and fever genesis, at least as revealed by IEG induction. Pharmacological evidence suggests that among the four subtypes of E-series prostaglandin (EP) receptors the EP1 receptor mediates fever responses. Oka and Hori administered a variety of drugs with agonist and antagonist properties for the four known EP receptors, to dissect their role in fever in rats (Oka and Hori 1994). They found that intracerebroventricular (i.c.v.) injection of 17-phenyl-o-trinor-PGE2 (an EP1 and EP3a receptor agonist), but not butaprost, M&B28767, or 11-deoxy-PGE1 (EP2, EP3a, and EP4 receptor agonists, respectively), mimicked PGE2 (i.c.v.)induced fever and that SC19220 (an EP1 receptor antagonist, i.c.v.) blocked PGE2 (i.c.v.) fever. Similar pharmacological study in pigs has suggested that the EP2 or EP3 receptor is necessary to produce fever (Parrott and Vellucci 1996). Ushikubi and colleagues examined mice bearing genetic deletions of the EP1–4 receptors (Ushikubi et al 1998). They found that only EP3 knockout animals failed to show an early phase of fever (up to 1 h) after injection of LPS (into the tail vein) or PGE2 (i.c.v.) (Ushikubi et al 1998).
PGE2 AND NEUROINFLAMMATION PROPOSED FEVER PATHWAYS Increasing evidence is accumulating from various studies that diverse mechanisms may underlie the febrile response. The apparent lack of LPS and IL-1 receptors within neuronal elements and the limited entry of LPS or cytokines into the CNS highlight the possibility that exogenous (LPS) or endogenous (IL-1b) pyrogen-mediated neuronal activation may result from LPS or cytokine–receptor interaction at vascular and/or other barrierrelated sites. Such interaction would trigger the release of secondary signalling molecules in a position to interact with components of circuitry controlling fever. Prostaglandin E2 is one of the most likely readily diffusible intermediates that are capable of interacting directly with hypothalamic neurosecretory processes. Brain endothelial cells constitute one of the routes for immune-CNS communication, in which the endothelial cells transform the blood-borne immune signal into a PGE2 signal, which in turn acts on the CNS neurons to evoke fever and other acute-phase responses (see Figure 42.3). Indeed, intravenous injection of LPS and recombinant rat IL-1b causes a profound transcriptional activation of the genes encoding the COX-2 and the mPGES, enzymes responsible for the production of PGE2 during inflammatory pathogenesis, within endothelial cells of the CNS blood vessels. A robust activation of NF-kB is also detected in the endothelium of the brain capillaries in response to different systemic inflammatory stimuli. This evidence strongly supports the hypothesis that circulating proinflammatory molecules stimulate prostaglandin production via transcriptional activation of
469
COX-2 through NF-kB signal pathways within the cells of the BBB (see Figure 42.3). Intravenous injection of LPS, IL-1b and localized intramuscular turpentine aggression, as well as central injection of its own ligand, PGE2, activate the transcription of the gene encoding EP4 mRNA and activate EP4-expressing neurons in the VMPO, PVN, NTS and cVLM (Zhang and Rivest 1999, 2000). IL-1b-induced EP4 transcription in the endocrine hypothalamus is completely abolished by the PG synthesis inhibitor ketorolac (Zhang and Rivest 2000). These results provide a substantial support for EP4 being the functional transmembrane receptor of PGE2 to mediate the effects of circulating immunogenic insults on neuroendocrine and autonomic functions.
SUMMARY AND CONCLUDING REMARKS From all the data discussed above, it is tempting to propose the following signalling cascade, leading to the fever in which brain endothelial cells play a pivotal role: 1. The endotoxin reaches the blood stream to bind with serum protein LPS-binding protein (LBP). The newly formed complex will bind to membrane CD14 receptor located on the mononuclear cell surface and therefore induces the release of cytokines (see Figure 42.4). IL-1b and TNFa, cytokines released by cells of myeloid origin during an inflammatory or infectious episode, will bind to their receptors, IL-1R1 and P55, expressed on the endothelium of the brain blood vessels.
Figure 42.3 Intracellular mechanisms mediating the influence of circulating interleukin-1ß (IL-1b) on the transcription of cyclooxygenase-2 (COX-2) within an endothelial cell of the blood–brain barrier. Although simplistic, both the MAP kinases and NF-kB pathways may be transduction/transcription signals in these processes. The production of the prostaglandin of E2 type is believed to be a key mediator to diffuse through the parenchymal brain and the neurons that control fever and the hypothalamic–pituitary–adrenal axis. The subsequent release of glucocorticoids is determinant for the immunosuppression of the systemic inflammation and downregulation COX-2 transcription. Glucocorticoids may increase IkBa transcription and/or interfere with NF-kB binding ability on COX-2 promoter in cerebral vascular cells
470
NERVOUS SYSTEM
2. Even though mCD14 is lacking on the endothelium, the soluble form of CD14 (sCD14) may allow a direct action of the bacterial endotoxin over barrier-related cells. 3. On the other hand, the intravascular complement cascade is activated in almost immediate reaction to the presence of LPS via both the classical and alternative pathways. The C fragment, C5, released by monocytes, may target its own receptor C5aR on cerebral endothelium upregulated by systemic LPS insult. 4. The fact that circulating endotoxin causes a rapid expression of CD14 mRNA within the circumventricular organs (CVOs), followed by a positive signal for CD14 transcript in microglia across the brain parenchyma, strongly suggested that LPS injected into the general circulation penetrates the OVLT, SFO, ME and AP tissues, which then allow the endotoxin to trigger locally the biosynthesis of its own receptor, CD14, within parenchymal structures surrounding the CVOs and then the entire brain of severely challenged animals. Activated microglia induced by LPS have the ability to produce a wide variety of cytokines, including IL-1b and TNFa. These cytokines may act on the nearby endothelial cells, in a paracrine manner, to bind their receptors respectively. By passing through these different pathways, simultaneously or respectively, bacterial endotoxin LPS triggers COX-2 and mPGES transcriptions through the NF-kB pathway. COX-2 and mPGES activation leads to PGE2 synthesis from the endothelium in the
preoptic area (POA). The PG may therefore diffuse through the parenchymal elements and bind to its EP4 receptor expressed at the surface of related neurons, which induces the cyclicAMP– cyclic AMP-responsive element binding protein transduction pathway and, in consequence, the induction of fever and activation of the HPA axis (Figure 42.3). Some reports have suggested that PVN may also have important roles in the fever pathway. Lesions of the parvocellular PVN block fever response to a systemic LPS challenge (Horn et al 1994; Lu et al 2001). The microinjection of PGE2 into the PVN is reported to increase core temperature in rats, although its effects are less remarkable than that of PGE2 injection into the anteromedial preoptic region. We propose here that systemic inflammatory insults activate the biosynthesis of PGE2 by endothelium in the PVN region. This newly formed PG has the ability to target its EP4 receptor expressed onto neurons of parvocellular PVN. These neurons, particularly those of the dorsal parvocellular division, may transmit signals directly, or indirectly via A5 noradrenergic cell groups or C1 adrenergic cell groups in the rostral VLM, to sympathetic preganglionic neurons at all levels of the thoracic spinal cord. Activation of these circuits may trigger sympathetic activity in brown adipose tissue, the heart, the adrenal gland and vasculature, which all together are involved in elevating body temperature (Zhang et al 2000). A neural route of pyrogen signal to the brain has also been proposed. The presence of LPS in the blood causes the immediate activation of complement and the consequent production of its
Figure 42.4 Schematic illustration of the possible circuits mediating the activation of PVN and the hypothalamic–pituitary–adrenal (HPA) axis during systemic innate immune response. The endotoxin LPS and cytokines use several pathways and sites of entry to communicate with the brain and neuroendocrine functions. It is suggested that circumventricular organs (organs devoid of blood–brain barrier) and the blood vessels (bv) are crucial target sites of LPS and proinflammatory cytokines of systemic origin, whereas activated regions involved in the autonomic control play a determinant role in the integration of information received from the periphery. Among these integrative structures, the PVN may be central to the appropriate control of homeostasis during immune challenge in directly controlling the autonomic outputs and the activity of the HPA axis. AP, area postrema; ARC, arcuate nucleus; BnST, bed nucleus of the stria terminalis; bv, blood vessels; chp, choroid plexus; CeA, central nucleus of the amygdala; DMH, dorsomedial nucleus of the hypothalamus; ME, median eminence; LC, locus coeruleus; LDT, laterodorsal tegmental nucleus; LPS, lipopolysaccharide; LRNm, lateral reticular nucleus medial; MPOA, medial preoptic area; NTS, nucleus of the solitary tract; OVLT, organum vasculosum of the lamina terminalis; PB, parabrachial nucleus; PP, posterior pituitary; PVN, paraventricular nucleus of the hypothalamus [parvocellular (pc) and magnocellular divisions (mc)]; SFO, subfornical organ; SON, supraoptic nucleus; VLM, ventrolateral medulla
PGE2 AND NEUROINFLAMMATION components C3a and C5a; these fragments will bind to Kupffer cells, which then are promptly stimulated to release mediators, such as cytokines and/or prostaglandins, potentially capable of activating nearby subdiaphragmatic vagal afferents, especially hepatic branch (Sehic et al 1998). The presence of type 1 IL-1 receptor and EP3 subtype PGE receptor in the bodies of the primary afferent vagal neurons located in the nodose ganglion strongly supports this hypothesis (Ek et al 1998). Vagal afferents then convey the pyrogenic message to the NTS, wherein it is passed to the A1/A2 noradrenergic cell groups, which transmit them to the POA/OVLT region via the ventral noradrenergic bundle, consequently developing the fever. It is important to note that the vagal route is essential for febrile pathogenesis only in response to minimal pyrogenic doses of LPS, i.e. for the development of the early febrile phase; when higher doses of LPS were administrated, vagotomy had no effect on the genesis of fever (Simons et al 1998). Thus the rapidity of neural communication between the immune system and the brain seems crucial for the onset of fever and less important for the later febrile phases.
ACKNOWLEDGEMENTS This research was supported by the Canadian Institutes of Health Research (CIHR), the former Medical Research Council of Canada (MRCC). While a PhD student in this laboratory, Ji Zhang held a Studentship from the MRCC. She is now a postdoctoral fellow at Astra-Zeneca (Montre´al, Canada). Serge Rivest is a MRCC Scientist and holds a Canadian Research Chair in Neuroimmunology.
REFERENCES Andersson J, Nagy S, Bjo¨rk L et al (1992) Bacterial toxin-induced cytokine production studied at the single-cell level. Immunol Rev, 127, 69–96. Baeuerle PA (1998) Proinflammatory signaling: last pieces in the NF-kB puzzle. Curr Biol, 8, R19–R22. Baeuerle PA and Henkel T (1994) Function and activations of NF-kB in the immune system. Ann Rev Immunol, 12, 141–179. Baeuerle PA and Baltimore D (1996) NF-kB: 10 years after. Cell, 87, 13– 20. Batshake B, Nilsson C and Sundelin J (1995) Molecular characterization of the mouse prostanoid EP1 receptor gene. Eur J Biochem, 231, 809– 814. Beutler B (2000) Tlr4: central component of the sole mammalian LPS sensor. Curr Opin Immunol, 12, 20–26. Bishai I and Coceani F (1996) Differential effects of endotoxin and cytokines on prostaglandin E2 formation in cerebral microvessels and brain parenchyma: implications for the pathogenesis of fever. Cytokine, 8, 371–376. Blais V and Rivest S (2001) Inhibitory action of nitric oxide on circulating tumor necrosis factor-induced NF-kB activity and COX-2 transcription in the endothelium of the brain capillaries. J Neuropathol Exp Neurol, 60, 893–905. Blatteis CM (1990) Neuromodulative actions of cytokines. Yale J Biol Med, 63, 133–146. Brady L, Lynn AB, Herkenham M and Gottesfeld Z (1994) Systemic interleukin-1 induces early and late patterns of c-fos mRNA expression in brain. J Neurosci, 14, 4951–4964. Breder CD and Saper CB (1996) Expression of inducible cyclooxygenase mRNA in the mouse brain after systemic administration of bacterial lipopolysaccharide. Brain Res, 713, 64–69. Breder CD, Dewitt D and Kraig RP (1995) Characterization of inducible cyclooxygenase in rat brain. J Comp Neurol, 355, 296–315. Cao C, Matsumura K and Watanabe Y (1997a) Induction of cyclooxygenase-2 in the brain by cytokines. Ann NY Acad Sci, 813, 307–309.
471
Cao C, Matsumura K, Yamagata K and Watanabe Y (1997b) Involvement of cyclooxygenase-2 in LPS-induced fever and regulation of its mRNA by LPS in the rat brain. Am J Physiol, 272, R1712–1725. Cao C, Matsumura K, Ozaki M and Watanabe Y (1999) Lipopolysaccharide injected into the cerebral ventricle evokes fever through induction of cyclooxygenase-2 in brain endothelial cells. J Neurosci, 19, 716–725. Chai Z, Gatti S, Toniatti C et al (1996) Interleukin (IL)-6 gene expression in the central nervous system is necessary for fever response to lipopolysaccharide or IL-1b: a study on IL-6-deficient mice. J Exp Med, 183, 311–316. Chan CC, Panneton M, Taylor AM et al (1997) A selective inhibitor of cyclooxygenase-2 reverses endotoxin-induced pyretic responses in nonhuman primates. Eur J Pharmacol, 327, 221–225. Dinarello CA, Cannon JG, Mancilla J et al (1991) Interleukin-6 as an endogenous pyrogen: induction of prostaglandin E2 in brain but not in peripheral blood mononuclear cells. Brain Res, 562, 199–206. Ek M, Kurosawa M, Lundeberg T and Ericsson A (1998) Activation of vagal afferents after intravenous injection of interleukin-1b: role of endogenous prostaglandins. J Neurosci, 18, 9471–9479. Ek M, Engblom D, Saha S et al (2001) Inflammatory response: pathway across the blood–brain barrier. Nature, 410, 430–431. Elmquist JK and Saper CB (1996) Activation of neurons projecting to the paraventricular hypothalamic nucleus by intravenous lipopolysaccharide. J Comp Neurol, 374, 315–331. Elmquist JK, Scammell TE, Jacobson CD and Saper CB (1996) Distribution of Fos-like immunoreactivity in the rat brain following intravenous lipopolysaccharide administration. J Comp Neurol, 371, 85–103. Ericsson A, Kova´cs KJ and Sawchenko PE (1994) A functional anatomical analysis of central pathways subserving the effects of interleukin-1 on stress-related neuroendocrine neurons. J Neurosci, 14, 897–913. Ericsson A, Liu C, Hart RP and Sawchenko PE (1995) Type-1 interleukin1 receptor in the rat brain: distribution, regulation, and relationship in sites of IL-1 induced cellular activation. J Comp Neurol, 361, 681–698. Fink MP, Rothschild HR, Deniz YF and Cohn SM (1989) Complement depletion with Naje haje cobra venom factor limits prostaglandin release and improves visceral perfusion in porcine endotoxic shock. J Trauma, 29, 1076–1084; discussion 1084–1085. Hemler M and Lands WE (1976) Purification of the cyclooxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyme. J Biol Chem, 251, 5575–5579. Hopkins SJ and Rothwell NJ (1995) Cytokines and the nervous system I: expression and recognition. Trends Neurosci, 18, 83–88. Horn T, Wilkinson MF, Landgraf R and Pittman QJ (1994) Reduced febrile responses to pyrogens after lesions of the hypothalamic paraventricular nucleus. Am J Physiol, 267, R323–328. Jakobsson PJ, Thoren S, Morgenstern R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96, 7220–7225. Johnson AK and Gross PM (1993) Sensory circumventricular organs and brain homeostatic pathways. FASEB J, 7, 678–686. Lacroix S and Rivest S (1998) Effect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cyclooxygenase enzymes (COX-1 and COX-2) in the rat brain. J Neurochem, 70, 452–466. Laflamme N and Rivest S (1999) Effects of systemic immunogenic insults and circulating proinflammatory cytokines on the transcription of the inhibitory factor kBa within specific cellular populations of the rat brain. J Neurochem, 73, 309–321. Laflamme N and Rivest S (2001) Toll-like receptor 4: the missing link of the cerebral innate immune response triggered by circulating Gramnegative bacterial cell wall components. FASEB J, 15, 155–163. Laflamme N, Lacroix S and Rivest S (1999a) An essential role of interleukin-1b in mediating NF-kB activity and COX-2 transcription in cells of the blood–brain barrier in response to systemic and localized inflammation, but not during endotoxemia. J Neurosci, 19, 10923– 10930. Laflamme N, Soucy G and Rivest S (2001) Circulating cell wall components derived from Gram-negative and not Gram-positive
472
NERVOUS SYSTEM
bacteria cause a profound transcriptional activation of the gene encoding toll-like receptor 2 in the CNS. J Neurochem, 79, 648–657. Laflamme N, Feuvrier E, Richard D and Rivest S (1999b) Involvement of serotonergic pathways in mediating the neuronal activity and genetic transcription of neuroendocrine corticotropin-releasing factor in the brain of systemically endotoxin-challenged rats. Neuroscience, 88, 223– 240. Li S, Wang Y, Matsumara K et al (1999) The febrile response to lipopolysaccharide is blocked in cyclooxygenase-2 (7/7), but not in cyclooxygenase-1 (7/7) mice. Brain Res, 825(1–2), 86–94. Lu J, Zhang YH, Chou TC et al (2001) Contrasting effects of ibotenate lesions of the paraventricular nucleus and subparaventricular zone on sleep–wake cycle and temperature regulation. J Neurosci, 21, 4864– 4874. Matsumura K, Watanabe Y, Onoe H et al (1990) High density of prostaglandin E2 binding sites in the anterior wall of the 3rd ventricle: a possible site of its hyperthermic action. Brain Res, 533, 147–151. Matsumura K, Watanabe Y, Imai-Matsumura K et al (1992) Mapping of prostaglandin E2 binding sites in rat brain using quantitative autoradiography. Brain Res, 581, 292–298. Merlie JP, Fagan D, Mudd J and Needleman P (1988) Isolation and characterization of the complementary DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase). J Biol Chem, 263, 3550–3553. Nadeau S and Rivest S (2000) Role of microglial-derived tumor necrosis factor in mediating CD14 transcription and NF-kB activity in the brain during endotoxemia. J Neurosci, 20, 3456–3468. Nadeau S and Rivest S (2001) The complement system is an integrated part of the natural innate immune response in the brain. FASEB J, 15, 1410–1412. Namba T, Sugimoto Y, Negishi M et al (1993) Alternative splicing of Cterminal tail of prostaglandin E receptor subtype EP3 determines Gprotein specificity. Nature, 365, 166–170. Narumiya S, Sugimoto Y and Ushikubi F (1999) Prostanoid receptors: structures, properties and functions. Physiol Rev, 79, 1193–1226. Navarra P, Pozzoli G, Brunetti L et al (1992) Interleukin-1b and interleukin-6 specifically increase the release of prostaglandin E2 from rat hypothalamic explants in vitro. Neuroendocrinology, 56, 61– 68. Oka T and Hori T (1994) EP1-receptor mediation of prostaglandin E2induced hyperthermia in rats. Am J Physiol, 267, R289–R294. Oka T, Hori T, Hosoi M et al (1997) Biphasic modulation in the trigeminal nociceptive neuronal responses by the intracerebroventricular prostaglandin E2 may be mediated through different EP receptors subtypes in rats. Brain Res, 771, 278–284. Oka T, Oka K, Scammell TE et al (2000) Relationship of EP(1–4) prostaglandin receptors with rat hypothalamic cell groups involved in lipopolysaccharide fever responses. J Comp Neurol, 428, 20–32. Parrott RF and Vellucci SV (1996) Effects of centrally administered prostaglandin EP receptor agonists on febrile and adrenocortical responses in the prepubertal pig. Brain Res Bull, 41, 97–103. Puschel GP, Hespeling U, Oppermann M and Dieter P (1993) Increase in prostanoid formation in rat liver macrophages (Kupffer cells) by human anaphylatoxin C3a. Hepatology, 18, 1516–1521. Rivest S (1995) Molecular mechanisms and neural pathways mediating the influence of interleukin-1 on the activity of neuroendocrine CRF motoneurons in the rat. Int J Dev Neurosci, 13, 135–146. Rivest S (1999) What is the cellular source of prostaglandins in the brain in response to systemic inflammation? Facts and controversies. Mol Psychiat, 4, 501–507. Rivest S (2001) How circuating cytokines trigger the neural circuits that control the hypothalamic–pituitary–adrenal axis. Psychoneuroendocrinology, 26, 761–788. Rivest S and Rivier C (1995) The role of CRF and interleukin-1 in the regulation of neurons controlling reproductive functions. Endocr Rev, 16, 177–199. Rivest S, Laflamme N and Nappi RE (1995) Immune challenge and immobilization stress induce transcription of the gene encoding the CRF receptor in selective nuclei of the rat hypothalamus. J Neurosci, 15, 2680–2695. Rivest S, Lacroix S, Vallie`res L et al (2000) How the blood talks to the brain parenchyma and the paraventricular nucleus of the hypothalamus during systemic inflammatory and infectious stimuli. Proc Soc Exp Biol Med, 223, 22–38.
Romanovsky AA, Shido O, Sakurada S et al (1996a) Endotoxin shock: thermoregulatory mechanisms. Am J Physiol, 270, R693–703. Romanovsky AA, Kulchitsky VA, Akulich NV et al (1996b) First and second phases of biphasic fever: two sequential stages of the sickness syndrome? Am J Physiol, 271, R244–253. Rothwell NJ (1991) Functions and mechanisms of interleukin-1 in the brain. Trends Pharmacol Sci, 12, 430–436. Rothwell NJ and Hopkins SJ (1995) Cytokines and the nervous system II: actions and mechanism of action. Trends Neurosci, 18, 130–136. Scammell TE, Elmquist JK, Griffin JD and Saper CB (1996) Ventromedial preoptic prostaglandin E2 activates fever-producing autonomic pathways. J Neurosci, 16, 6246–6254. Scammell TE, Griffin JD, Elmquist JK and Saper CB (1998) Microinjection of a cyclooxygenase inhibitor into the anteroventral preoptic region attenuates LPS fever. Am J Physiol, 274, R783–789. Sehic E and Blatteis CM (1996) Blockade of lipopolysaccharide-induced fever by subdiaphragmatic vagotomy in guinea-pigs. Brain Res, 726, 160–166. Sehic E, Ungar AL and Blatteis CM (1996a) Interaction between norepinephrine and prostaglandin E2 in the preoptic area of guineapigs. Am J Physiol, 271, R528–536. Sehic E, Li S, Ungar AL and Blatteis CM (1998) Complement reduction impairs the febrile response of guinea-pigs to endotoxin. Am J Physiol, 274, R1594–1603. Sehic E, Szekely M, Ungar AL et al (1996b) Hypothalamic prostaglandin E2 during lipopolysaccharide-induced fever in guinea-pigs. Brain Res Bull, 39, 391–399. Simons CT, Kulchitsky VA, Sugimoto N et al (1998) Signaling the brain in systemic inflammation: which vagal branch is involved in fever genesis? Am J Physiol, 275, R63–68. Sugimoto Y, Shigemoto R, Namba T et al (1994) Distribution of the messenger RNA for the prostaglandin E receptor subtype EP3 in the mouse nervous system. Neuroscience, 62, 919–928. Taniguchi Y, Yokoyama K, Ikesue A et al (1998) Inhibition by nimesulide of prostaglandin production in rat macrophages. Drugs Exp Clin Res, 24, 17–27. Turnbull AV and Rivier CL (1999) Regulation of the hypothalamic– pituitary–adrenal axis by cytokines: actions and mechanisms of action. Physiol Rev, 79, 1–71. Ushikubi F, Segi E, Sugimoto Y et al (1998) Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature, 395, 281–284. Vallie`res L and Rivest S (1999) Interleukin-6 is a needed proinflammatory cytokine in the prolonged neural activity and transcriptional activation of corticotropin-releasing factor during endotoxemia. Endocrinology, 140, 3890–3903. Watanabe Y and Hayaishi O (1988) Quantitative autoradiographic localization of prostaglandin E2 binding sites in monkey diencephalon. J Neurosci, 8, 2003–2010. Watkins LR, Maier SF and Goehler LE (1995) Cytokine-to-brain communication: a review and analysis of alternative mechanisms. Life Sci, 57, 1011–1026. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Yamagata K, Matsumura K, Inoue W et al (2001) Co-expression of microsomal-type prostaglandin E synthase with cyclooxygenase-2 in brain endothelial cells of rats during endotoxin-induced fever. J Neurosci, 21, 2669–2677. Zhang J and Rivest S (1999) Distribution, regulation and co-localization of the genes encoding the EP2- and EP4-PGE2 receptors in the rat brain and neuronal responses to systemic inflammation. Eur J Neurosci, 11, 2651–2668. Zhang J and Rivest S (2000) A functional analysis of EP4 receptorexpressing neurons in mediating the action of PGE2 within specific nuclei of the brain in response to circulating interleukin-1b. J Neurochem, 74, 2134–2145. Zhang YH, Lu J, Elmquist JK and Saper CB (2000) Lipopolysaccharide activates specific populations of hypothalamic and brainstem neurons that project to the spinal cord. J Neurosci, 20, 6578–6586. Zheng H, Fletcher D, Kozak W et al (1995) Resistance to fever induction and impaired acute-phase response in interleukin-1b-deficient mice. Immunity, 3, 9–19.
43 Prostanoids in Pain Tony L. Yaksh1, Patrick W. Mantyh2 and Camilla I. Svensson1 1University
of California at San Diego, La Jolla, CA, and 2University of Minnesota, Minneapolis, MN, USA
Several convergent lines of information emphasize the remarkable contribution made by the family of lipidic acids known as prostanoids to the cascade of events initiated by tissue injury and inflammation that lead to an attendant pain state. In the following sections, we will consider some of the key points that reflect the systems through which these actions are expressed.
OVERVIEW OF THE PROSTAGLANDIN CASCADE The formation of the prostanoids depends upon the initial action of a variety of phospholipase A2 isozymes that remove arachidonic acid from the cell membrane. The arachidonate is acted upon by the cyclooxygenase enzyme (COX) to form an intermediary, PGH2 (Marnett et al 1999). Cloning revealed the presence of two membrane-bound isozymes with a 60% amino acid sequence homology (Smith et al 2000b). Although structurally similar, they represent the expression of different gene and kinetic properties. Of greatest importance is that COX-1 is stably expressed, whereas COX-2 is subject to significant upregulation by a variety of nuclear regulatory factors and is subject to suicide inactivation (Vane et al 1998). In many cell systems, there is significant constitutive expression of COX-1, whereas COX-2 appears to be induced in a variety of cell types (including neurons) by a variety of membrane stimuli. Following PGH2 formation, a variety of prostaglandin synthetases convert the intermediary into a wide number of isoforms, including PGD2, PGE2 and PGF2 (Fitzpatrick and Soberman 2001), which then move extracellularly to interact through several families of G-protein-coupled receptors with several transmembrane spanning domains (DP, EP, FP, IP and TP-r) (Armstrong and Wilson 1995; Negishi et al 1995; Versteeg et al 1999). Activation of PG receptors by their respective ligands triggers intracellular signals that can be stimulatory [e.g. stimulation of adenylyl cyclase (Narumiya et al 1999; Negishi et al 1995) and activation of phospholipase C (Birnbaumer et al 1990; Yousufzai et al 1988)] or inhibitory [e.g. depressed cAMP production (Melien et al 1988; Negishi et al 1989)].
component of the pain states present after surgery, trauma and inflammation (Price 1997). In preclinical models, such hyperalgesia is typically defined in terms of an escape response induced by a stimulus of reduced intensity or by a decreased response latency/ increased response intensity to a given stimulus (Yaksh 1999). Considerable data, starting with the early studies on the action of the non-steroidal antiinflammatory drugs (NSAIDs), have implicated prostaglandins in this injury-evoked hyperalgesia (Yaksh and Svensson 2001). In the following sections, we will consider several aspects of the complex mechanisms involving prostaglandins in this behaviourally-defined phenomenon. The mechanisms underlying the pain states initiated by tissue injury and the role played by the prostanoids in the actions of NSAIDs may be broadly viewed in terms of two components: peripheral and spinal.
PERIPHERAL COMPONENTS OF HYPERALGESIA In the periphery, an intense stimulus gives rise to the activation of high-threshold, slowly conducting sensory afferents. The frequency of firing of these axons is proportional to the stimulus magnitude. If the stimulus is of a sufficient magnitude to initiate tissue injury, activity in the axon will continue after removal of the stimulus, and there will be the appearance of a sensitization of the terminals of the axon, such that for any given stimulus intensity there will be a greater discharge than would have been produced by that stimulus prior to the injury (Michaelis et al 1996; Treede et al 1992). This peripheral mechanism provides an important component to the exaggerated behavioural response associated with tissue injury. The origin of this spontaneous activity and sensitization is considered to be secondary to the local release of a wide variety of active factors from blood products released by extravasation, inflammatory cells, and local cells injured by the stimulus (Dantzer 2001).
Central PROPERTIES OF PAIN SECONDARY TO TISSUE INJURY AND INFLAMMATION From a behavioural perspective, two hallmarks of the behavioural state generated by tissue injury are ongoing pain and an exaggerated pain report evoked by a moderately strong stimulus applied to the injured tissue (Yaksh 1997). This latter condition, referred to as hyperalgesia, is an important psychophysical The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
In the face of an acute injury stimulus, input from small afferents will activate a variety of second-order neurons, with the frequency of the response being proportional to the input. If persistent, such input will, however, lead to a potent augmentation of the input/ output function of that neuron, such that the spinal neurons show an enhanced response to any given input and an increase in the size of the peripheral receptive field of the neuron (Willis 2001). This effect is considered to be an expression of a central
474
NERVOUS SYSTEM The NSAIDs have been shown to inhibit the essential cyclooxygenase-mediated synthesis of prostaglandins (Smith and Willis 1971; Vane 1971). Accordingly, to the degree that an augmented firing rate in response to a peripheral stimulus is dependent upon prostaglandins, cyclooxygenase inhibitors will return the activity of the system and the stimulation thresholds back towards the preinjury level. These properties are thus in accord with the common observation that NSAIDs acted in models of peripheral inflammation where the identifying characteristic was a ‘‘hyperalgesic state’’ (Ferreira 1972; Moncada et al 1975). Dissociation of Central vs. Peripheral Antihyperalgesic Actions of NSAIDs There is little doubt that prostaglandins, and hence COX isozymes, play a role at the peripheral terminal. Nevertheless, several lines of evidence support an equally important central (spinal role):
Figure 43.1 Acute primary afferent input initiated by an acute noninjurious stimulus (left) evokes a stimulus-dependent depolarization of dorsal horn neurons by the release of principally glutamate which acts through AMPA receptors leading to a transient depolarization and activation of downstream events such as CFOS. The pain behaviour (e.g. evoked escape latency) of the organism maps the input time course and is proportional to the magnitude of the physical stimulus. In the face of tissue injury, there is an ongoing afferent traffic leading to repetitive activity in dorsal horn neurons. This persistent depolarization initiates a cascade that includes the removal of the magnesium block from the NMDA receptor, and an increase in intracellular calcium secondary to activation of NK1 receptors and the NMDA ionophore. The increased calcium triggers activity in a variety of kinases leading to phosphorylation of membrane protein (such as PKC), phosphorylating the NMDA receptor and serving to enhance its activity and MAPKinases, which activates spinal PLA2, providing a source of arachidonic acid for constitutive COX-2 that leads to enhanced PGE2 release into the local extracellular milieu. PGE2 acts through EP receptors that are preterminal on sensory afferents, serving to increase terminal excitability and transmitter release. The net effect of this cascade is to enhance the preand postsynaptic responses initiated by a peripheral stimulus, leading to a condition in which the response to a given stimulus is markedly enhanced. The behavioural effects of intrathecal COX-2 inhibitors emphasize the importance of the COX-2 component of the post-tissue-injury facilitatory cascade and provides a paradigm explaining the absence of an effect of such isozyme inhibitors on acute, noxious and injurious stimuli
facilitation and possesses properties that serve to explain the hyperalgesia that is induced by tissue injury and inflammation.
Peripheral Actions of the PLA2/COX/Prostanoids in Hyperalgesia Local injury yields the release of active factors that can alter peripheral terminal sensitivity. It is appreciated that an important component of this peripheral sensitization is mediated by prostaglandins. These lipidic acids are locally synthesized by the enzyme cyclooxygenase and are released following focal tissue injury and inflammation (Ferreira 1972; Moncada et al 1975). They have been shown to sensitize peripheral nerve endings and to enhance pain behaviour in animals (Lim 1970). Sensitization of sensory neurons by prostaglandins is initiated in part by activation of adenylate cyclase and phospholipase C through prostanoid receptor binding, leading to enhanced protein kinase C (PKC) activity and increased levels of cAMP IP3 (Bley et al 1998).
1. Repetitive small primary afferent activation evokes facilitated activity in dorsal horn neurons that occurs without peripheral inflammation. This facilitated state is blocked by COX inhibitors (Attal et al 1988; Carlsson et al 1988; Groppetti et al 1988; Herrero et al 1997; Jurna and Brune 1990; Mazario et al 1999; Pitcher and Henry 1999; Willingale et al 1997). 2. This activity is observed in man. Thus, nociceptive reflexes evoked by electrical sural nerve stimulation is diminished by systemically delivered NSAIDs in humans (Piletta et al 1991; Willer et al 1989). 3. Systematic comparison of the clinical antihyperalgesic activity and the antiinflammatory activity of a wide variety of NSAIDs has indeed indicated that there is a significant dissociation between the antiinflammatory potency and the antihyperalgesic action of the several agents (McCormack and Brune 1991; Mehlisch 1983). 4. Finally, COX-2 inhibitors display an efficacy equal to that of mixed COX-1/2 inhibitors and show an immediate activity after an acute injury. This suggests that these agents must be acting where there is a constitutive expression of COX-2. As noted, such constitutive expression occurs within the central nervous system (CNS) and not at the site of injury. These observations suggest that COX inhibitors express their antihyperalgesic actions in human and animal models by an action that is at least partially independent of a peripheral antiinflammatory action.
SPINAL COMPONENTS OF HYPERALGESIA The biochemistry of spinal facilitation and the role played by prostaglandins reflect upon the pharmacology of the primary afferents and second-order neurons. This organization and the associated cascade are presented schematically in Figure 43.1. Small high-threshold primary afferents are known to contain and release peptides, such as substance P, and amino acids, such as glutamate. Substance P and excitatory amino acids evoke excitation in second-order neurons. Current electrophysiology suggests that acute excitation generated by a transient stimulus is mediated by the release of glutamate and a direct monosynaptic excitation of the second-order neuron by an AMPA class of glutamate receptors. In the face of a persistent excitation, a series of events are known to transpire that constitute a complex downstream cascade. First, in conjunction with the persistent activation of the AMPA receptor, there will be the release of sP and the activation of an excitatory neurokinin 1 (NK1) G-protein-
PROSTANOIDS IN PAIN coupled receptor present on dorsal horn neurons (Allen et al 1999; Todd 2002). This activation serves to produce an ongoing depolarization of the neuron. Additionally, the persistent depolarization serves to enable the NMDA receptor by removal of its constitutive Mg2+ block. At normal resting membrane potentials, the NMDA receptor is inactive because of the block of the ionophore by Mg2+. In this condition, occupancy by glutamate will not activate the ionophore (Parsons 2001; Yoneda and Ogita 1991). Given the modest depolarization of the membrane (as produced during repetitive stimulation secondary to the activation of AMPA and sP receptors), the Mg block is removed, permitting glutamate to now activate the NMDA receptor. The joint activation of the NK1 receptor and the NMDA receptor leads to increase in intracellular Ca2+ secondary to activation of IP3 and the opening of the NMDA ionophore, respectively. The importance of the role played by these several elements, outlined above, at the spinal level has been confirmed by the ability of spinally delivered antagonists of AMPA to prevent acute pain behaviour, while antagonists of the NK1 receptor and the NMDA ionophore have little effect upon acute nociception, but serve to reduce the hyperalgesia that arises secondary to tissue injury and inflammation (Hunt and Mantyh 2001; Yaksh et al 1999). Conversely, the intrathecal delivery of NMDA and NK1 agonists will initiate a hyperalgesic state (Malmberg and Yaksh 1992b). This increase in intracellular Ca2+, serves as a trigger for a subsequent series of intracellular events, including the activation of P38 Map kinase. Such activation of p38 MAPK (P-p38 MAP) has been demonstrated in astrocytoma cells after substance P (Fiebich et al 2000) and cerebellar neurons after NMDA (Kawasaki et al 1997). P-p38MAP in turn leads to activation of PLA2. Cytosolic (cPLA2) and secretory (PLA2) phospholipases release arachidonic acid (AA) from cellular membranes through hydrolysation of phospholipids (Bingham and Austen 1999; Farooqui et al 1997). In the presence of AA, cyclooxygenase (COX) products (prostaglandins; PG) are formed and released. Both COX isozymes have been identified in the normal spinal cord and, as will be discussed below, the spinal COX-2 isozyme appears to be primarily involved in the synthesis of the prostanoids responsible for the hyperalgesia (Svensson and Yaksh 2002). Consistent with these observations, in vitro and in vivo studies have shown peripherally a significant evoked release of spinal PGE2 (see below). As indicated schematically in Figure 43.2, these spinal prostanoids move extracellularly and facilitate transmitter release from primary and non-primary afferent terminals by an interaction with specific prostaglandins receptors. In DRGs, PGE2 enhances depolarization-evoked increases in intracellular calcium and a resulting release of sP from capsaicin-sensitive DRGs (Smith et al 2000a). Spinal release in vivo of glutamate evoked by IT capsaicin and NMDA is diminished by IT delivery of COX inhibitors (Malmberg and Yaksh 1995a; Sorkin 1992), indicating that spinal terminals from which glutamate is released are subject to facilitation by COX products. Aside from a presynaptic action, PGE2 has been shown to directly activate dorsal horn nociceptors (Baba et al 2001), a finding consistent with the presence of PGE2 receptor mRNA in dorsal horn neurons (Coleman et al 1994). In addition, it has been shown in dorsal spinal cord that PGE2 serves to antagonize the inhibitory effects of glycine receptor activation through a postsynaptic mechanism involving the EP2 receptors (Ahmadi et al 2002). In spinal dorsal horn this can lead to an enhanced responsiveness of dorsal horn neurons to large (Ab) afferent input (Sorkin and Puig 1996) and a remarkable tactile allodynia (Yaksh 1989). Importantly, these observations coincide with the observation that large afferent-evoked excitation in spinal dorsal horn is
475
augmented by peripheral injury (Baba et al 1999). The present discussions would suggest that this facilitation and its corresponding tactile allodynia may be mediated by a reduction in the glycine-mediated inhibition normally activated by such input. Spinal Actions of the PLA2/COX/Prostanoids in Hyperalgesia The importance of this spinal PLA2-prostanoid cascade, outlined above, to the behavioural sequelae of tissue injury is suggested by a number of convergent observations. Spinal Distribution of COX and Prostanoid Receptors COX-1 and COX-2 mRNA and protein are constitutively expressed in dorsal root ganglia (DRG) and in spinal dorsal horn, as shown by in situ hybridization (Chopra et al 2000; but see Ichitani et al 1997), Northern blotting (Beiche et al 1998a), immunohistochemistry (Willingale et al 1997) and immunoblot techniques (Beiche et al 1998a). Under control conditions COX-1
Figure 43.2 At a systems level, repetitive afferent input activating the cascade described in Figure 43.1, leading to the release of dorsal horn prostaglandins, will impact dorsal horn nociceptive processing in several ways. (1) PGE2 will enhance the opening of voltage-gated calcium channels and depress the opening of voltage-gated potassium channels in primary afferents. This serves to increase the depolarization-evoked release of primary afferent transmitters, leading to an enhanced excitatory drive. (2) PGE2 acting upon postsynaptic neurons has been shown to facilitate postsynaptic excitation, leading to an enhanced response to excitatory input. (3) PGE2 has been shown to inhibit the release of transmitter from inhibitory glycinergic and GABAergic interneurons. This ongoing release of inhibitory transmitter serves to modulate the evoked discharge of the second order dorsal horn neuron. Loss of that inhibition results in an exaggerated discharge of the second order neuron. See text for other details
476
NERVOUS SYSTEM
and COX-2 are in spinal neurons and non-neuronal cells (astrocytes). In DRGs, COX-1 immunoreactivity is observed in small- and medium-sized neuronal cell bodies (Chopra et al 2000). Non-neuronal cells expressing COX-1 and COX-2 include astrocytes (Hirst et al 1999), microglia (Bauer et al 1997), endothelial (Habib et al 1993) and leptomeningeal cells (Matsumura et al 1998), although by and large this non-neuronal expression is moderate. With regard to the prostanoid receptors, in situ hybridization and immunohistochemistry show EP1, EP2, EP3, EP4 (Beiche et al 1998b; Donaldson et al 2001; Kawamura et al 1997) and IP receptors (Matsumura et al 1995) in the superficial spinal cord, and DP, EP1, EP3 and IP receptors have been detected on DRG neurons (Oida et al 1995; Wright et al 1999; Svensson and Yaksh 2002).
Differential Role of Spinal COX-1 and COX-2 The role of COX-1, although it is constitutively expressed in the spinal cord, remains controversial. As noted above, several studies have failed to demonstrate a role for COX-1 (but see Lashbrook et al 1999; Ma and Eisenach 2002). This assertion is largely based on work with a single COX-1 inhibitor and from the observation that the maximum effect of COX-2 inhibitors is not distinguishable from COX-1/2 inhibitors. The absence of other well-defined COX-1 inhibitors, however, precludes dismissal of a COX-1 role. In any case, these points suggest that these isozymes do not play equivalent roles. Two possibilities may be that COX-1 requires higher AA concentrations than COX-2 (Kulmacz and Wang 1995), or that there may be differential coupling to cytosolic, secretory and non-calcium-dependent phospholipases (Leslie and Watkins 1985; Murakami et al 1999). Further study will doubtlessly reveal the role played by this constitutive isozyme.
Effects of Prostanoid Agonists The intrathecal delivery of a variety of prostanoids, including PGE1, PGE2, PGF2a, PGI2 and thromboxane B2, evokes thermal and mechanical hyperalgesia (Minami et al 1992, 1994; Saito et al 1995; Takano et al 1999; Uda et al 1990). Inhibition of Spinal PLA2 and COX Intrathecal injection of inhibitors of p38 MAPK or PLA2 results in a dose-dependent antihyperalgesia in the formalin flicking model and in the thermal hyperalgesia noted after intraplantar carrageenan (Svensson et al 2001, 2003), emphasizing a possible role of spinal PLA2 isozyme activation in facilitated processing. Intrathecal COX-1/2 inhibitors produce a dose-dependent/stereospecific reduction in pain states that involves central facilitation, such as the writhing test (Yaksh 1982), phase 2 of the formalin test (Malmberg and Yaksh 1992a; Malmberg and Yaksh 1993), thermal hyperalgesia induced by inflammation (Dirig et al 1998; Yamamoto and Nozaki-Taguchi 1997) or by IT-sP or IT-NMDA (Malmberg and Yaksh 1992b; Yamamoto and Sakashita 1998), but are without effect in acute pain. Examination of the respective role of the two isozymes indicates that the intrathecal injection of COX-2 but not COX-1 inhibitors dose-dependently and stereospecifically blocks the thermal hyperalgesia induced by intraplantar carrageenan, IT-sP and IT-NMDA, as well as the tactile allodynia resulting from a crush of the hind paw. Release of Spinal Prostanoids In vitro studies using spinal cord slices have shown that extracellular concentrations of PGs in the media are enhanced by local application of NK1, VR1 and glutamate receptor agonists (Dirig and Yaksh 1999), and this release is blocked by COX-2 inhibition (Dirig et al 1997). In vivo, PGE2 concentrations in spinal microdialysates or in cerebrospinal fluid are elevated following: (a) activation of small afferents (intraplantar formalin, heat; Coderre et al 1990; Malmberg and Yaksh 1995a, 1995b) that are capsaicin-sensitive (Hua et al 1997); (b) persistent inflammation (carrageenan in knee joint; intraplantar Freund’s adjuvant, intraplantar zymosan; Ebersberger et al 1999; Guhring et al 2000; Samad et al 2001; Yang et al 1996a); and (c) IT-sP, IT NMDA or IT-kainate (Hua et al 1999; Yang et al 1996b) or systemic cytokines (Samad et al 2001). With respect to the contributing enzyme, it is clear that, following carrageenan inflammation, IT NMDA and sP, this release is attenuated by the systemic delivery of COX-2 but not COX-1 inhibitors at doses that reverse the associated behavioural hyperalgesia (Yaksh et al 2001).
TIME-DEPENDENT CHANGES IN AFFERENT PROCESSING The above processes, outlined above, reflect events that occur over intervals of: (a) tens to hundreds of milliseconds (acute highintensity, non-injurious pain stimulus evoked by transient activation of the AMPA receptor); (b) seconds to minutes after tissue injury (activation of the NK1 and NMDA receptor to initiate the downstream cascade leading to central facilitation); and (c) events occurring over hours and perhaps days. In this last case, it is appreciated that the COX-2 isozyme is constitutively present in the spinal cord, but not in the periphery at the site of injury. Nevertheless, COX-2 protein is subject to significant upregulation that can be initiated by a number of nuclear transcription factors, including P38 MAP kinase. Such upregulation is known to occur in the periphery at the site of injury, as well as in the spinal cord after peripheral injury and inflammation. At the spinal level, the upregulation can be induced by the activation of spinal NMDA and NK1 receptors, indicating that input from the periphery can, by transmitter release, produce a stimulatory effect, perhaps mediated secondary to afferent transmitter release and p38 MAP kinase activation. The increase in COX-2 expression may thus provide additional synthetic capacity, enhancing the prostaglandin synthesis capacity of the systems, leading to a more profound hyperalgesia because of increased COX-2 expression peripherally at the site of injury and centrally at the spinal cord level. CENTRAL AND PERIPHERAL INTERACTION FOLLOWING PERIPHERAL INJURY In the above scenarios, it is appreciated that afferent input may lead to a central sensitization and to the activation of neural events that lead to an enhanced spinal expression of COX-2. There is also a growing appreciation that after tissue injury and inflammation central function can be altered not only by afferent input from the injury site, but also by the presence of circulating factors that can directly influence central excitability. Thus, circulating LPS, IL-1b or TNFa can induce the expression of COX-2 in perivascular microglia (Lacroix and Rivest 1998; Matsumura et al 1998; Samad et al 2001). The role of this upregulation is not certain, but it is hypothesized that it serves to provide additional sources of prostaglandins that can regulate the processing of small afferent input. These findings are thus consistent with a central sensitization by which afferent activation induced by peripheral injury and inflammation initiates a cascade of events that lead to a facilitated
PROSTANOIDS IN PAIN state of processing. With tissue injury and inflammation, circulating factors, such as TNFa and IL-1b, are elevated in the blood (Watkins et al 1999). Cells in proximity to the microvasculature, such as microglia or astrocytes, display increases in their COX-2 mRNA levels by 2 h, and this subsides by 24 h after i.v. cytokines (IL-1-b/TNFa) (Lacroix and Rivest 1998; Matsumura et al 1998; Quan et al 1998) or LPS (Matsumura et al 1998). The role played by both neural input generated by the injury and the circulating cytokines thus points to an important interaction underlying the regulation of an isozyme known to contribute to a hyperpathic state. It seems likely that the interaction will prove to be particularly complex, with the degree of upregulation of the spinal COX-2 isozyme reflecting the effects (a) produced by segmental inputs and (b) from an action upon the entire neuraxis secondary to the circulatory distribution of inflammatory products.
CENTRAL ACTIONS OF PROSTANOIDS IN HUMAN PAIN STATES While we are aware of no report showing COX-1 or -2 in human spinal tissues, prostaglandins have been observed in human lumbar CSF (Romero et al 1984). In any case, several lines of evidence provide strong support for the concept that prostaglandins within the neuraxis play as pervasive a role as that demonstrated in preclinical models: 1. The biceps femoris flexion reflex evoked by electrical stimulation of the sural nerve or the orbicularis oculi reflex is diminished by systemically delivered NSAIDs (Fabbri et al 1992; Guieu et al 1992; Piletta et al 1991; Willer et al 1989). These reflexes, resulting in the absence of tissue injury or inflammation from the direct activation of sensory afferents, support a non-peripheral (i.e. central) antinociceptive action. 2. The intrathecal delivery of lysine acetylsalicylate, a watersoluble agent, produced significant pain relief in cancer patients (Devoghel 1983, 1993; Pellerin et al 1987), a finding in parallel with the actions of IT NSAIDs in the preclinical models discussed above. 3. COX-2-selective inhibitors have significant analgesic activity in a variety of postsurgical pain states (Cannon and Breedveld 2001; Malmstrom et al 1999). In dental studies, pain occurs shortly after surgery. Here COX-2 inhibitors display a rapid onset of activity. This observation would be unexpected from a peripheral action, as COX-2 is not constitutively expressed in the periphery. The rapid onset of a COX-2-mediated effect argues that the effect occurs as the result of a mechanism in which COX-2 is constitutively expressed. In summary, prostaglandins play an important role in augmenting the sensitivity of afferent processing by an effect upon the peripheral terminals, but, in addition, they form an important link in the central cascade that is acutely initiated by tissue injury through the action of COX-2 isozyme. Over longer intervals of time, it is equally likely that the upregulation in the expression of the spinal COX-2 may contribute to developing hyperpathic states and perhaps account in part for the potent analgesic effects displayed by NSAIDs in a variety of chronic injury/inflammatory states. The information outlined in this chapter suggests that the hyperalgesic state induced by tissue injury is, at least in part, the consequence of a complex spinal cascade which serves to increase prostaglandin synthesis and release through a constituitive spinal COX-2 pathway. These prostanoids are part of the cascade that leads to a spinal state of facilitated processing. Current data strongly suggest that this phenomenon is at work in humans.
477
We would end this review by noting that we have come full circle (Yaksh and Svensson 2001). Flower et al (1980) noted that whereas ‘‘. . . the analgesic actions of morphine occur centrally, aspirin works peripherally . . . preventing the synthesis and release of prostaglandins in inflammation . . .’’. Later it was emphasized that the aspirin-like drugs are particularly effective in ‘‘settings . . . in which inflammation has caused sensitization of pain receptors to normally painless mechanical or chemical stimuli’’ (Insel 1990). Previously, Woodbury (1965) had noted that ‘‘. . . the salicylates are capable of alleviating certain types of pain by virtue of a selective depressant effect on the CNS, the mechanisms of which have not been elucidated. A subcortical site is suggested . . .’’
REFERENCES Ahmadi S, Lippross S, Neuhuber WL and Zeilhofer HU (2002) PGE(2) selectively blocks inhibitory glycinergic neurotransmission onto rat superficial dorsal horn neurons. Nature Neurosci, 5, 34–40. Allen BJ, Li J, Menning PM et al (1999) Primary afferent fibers that contribute to increased substance P receptor internalization in the spinal cord after injury. J Neurophysiol, 81, 1379–1390. Armstrong RA and Wilson NH (1995) Aspects of the thromboxane receptor system. Gen Pharmacol, 26, 463–472. Attal N, Kayser V, Eschalier A et al (1988) Behavioural and electrophysiological evidence for an analgesic effect of a nonsteroidal antiinflammatory agent, sodium diclofenac. Pain, 35, 341– 348. Baba H, Doubell TP and Woolf CJ (1999) Peripheral inflammation facilitates Ab fiber-mediated synaptic input to the substantia gelatinosa of the adult rat spinal cord. J Neurosci, 19, 859–867. Baba H, Kohno T, Moore KA and Woolf CJ (2001) Direct activation of rat spinal dorsal horn neurons by prostaglandin E2. J Neurosci, 21, 1750–1756. Bauer MK, Lieb K, Schulze-Osthoff K et al (1997) Expression and regulation of cyclooxygenase-2 in rat microglia. Eur J Biochem, 243, 726–731. Beiche F, Brune K, Geisslinger G and Goppelt-Struebe M (1998a) Expression of cyclooxygenase isoforms in the rat spinal cord and their regulation during adjuvant-induced arthritis. Inflamm Res, 47, 482– 487. Beiche F, Klein T, Nusing R et al (1998b) Localization of cyclooxygenase2 and prostaglandin E2 receptor EP3 in the rat lumbar spinal cord. J Neuroimmunol, 89, 26–34. Bingham CO III and Austen KF (1999) Phospholipase A2 enzymes in eicosanoid generation. Proc Assoc Am Physicians, 111, 516–524. Birnbaumer L, Abramowitz J and Brown AM (1990) Receptor–effector coupling by G proteins. Biochim Biophys Acta, 1031, 163–224. Bley KR, Hunter JC, Eglen RM and Smith JA (1998) The role of IP prostanoid receptors in inflammatory pain. Trends Pharmacol Sci, 19, 141–147. Cannon GW and Breedveld FC (2001) Efficacy of cyclooxygenase-2specific inhibitors. Am J Med, 110(suppl 3A), 6–12. Carlsson KH, Monzel W and Jurna I (1988) Depression by morphine and the non-opioid analgesic agents, metamizol (dipyrone), lysine acetylsalicylate, and paracetamol, of activity in rat thalamus neurones evoked by electrical stimulation of nociceptive afferents. Pain, 32, 313–326. Chopra B, Giblett S, Little JG et al (2000) Cyclooxygenase-1 is a marker for a subpopulation of putative nociceptive neurons in rat dorsal root ganglia. Eur J Neurosci, 12, 911–920. Coderre TJ, Gonzales R, Goldyne ME et al (1990) Noxious stimulusinduced increase in spinal prostaglandin E2 is noradrenergic terminaldependent. Neurosci Lett, 115, 253–258. Coleman RA, Smith WL and Narumiya S (1994) International Union of Pharmacology classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. Dantzer R (2001) Cytokine-induced sickness behavior: mechanisms and implications. Ann NY Acad Sci, 933, 222–234.
478
NERVOUS SYSTEM
Devoghel JC (1983) Small intrathecal doses of lysine-acetylsalicylate relieve intractable pain in man. J Int Med Res, 11, 90–91. Devoghel JC (1993) Intrathecal injection of lysine-acetylsalicylate in man with intractable cancer pain. In Jurna I and Yaksh TL (eds), Progress in Pharmacology and Clinical Pharmacology. Stuttgart: Gustav Fischers Verlag, 111–118. Dirig DM, Isakson PC and Yaksh TL (1998) Effect of COX-1 and COX-2 inhibition on induction and maintenance of carrageenan-evoked thermal hyperalgesia in rats. J Pharmacol Exp Therapeut, 285, 1031– 1038. Dirig DM, Konin GP, Isakson PC and Yaksh TL (1997) Effect of spinal cyclooxygenase inhibitors in rat using the formalin test and in vitro prostaglandin E2 release. Eur J Pharmacol, 331, 155–160. Dirig DM and Yaksh TL (1999) In vitro prostanoid release from spinal cord following peripheral inflammation: effects of substance P, NMDA and capsaicin. Br J Pharmacol, 126, 1333–1340. Donaldson LF, Humphrey PS, Oldfield S et al (2001) Expression and regulation of prostaglandin E receptor subtype mRNAs in rat sensory ganglia and spinal cord in response to peripheral inflammation. Prostagland Lipid Mediat, 63, 109–122. Ebersberger A, Grubb BD, Willingale HL et al (1999) The intraspinal release of prostaglandin E2 in a model of acute arthritis is accompanied by an upregulation of cyclooxygenase-2 in the spinal cord. Neuroscience, 93, 775–781. Fabbri A, Cruccu G, Sperti P et al (1992) Piroxicam-induced analgesia: evidence for a central component which is not opioid mediated. Experientia, 48, 1139–1142. Farooqui AA, Yang HC, Rosenberger TA and Horrocks LA (1997) Phospholipase A2 and its role in brain tissue. J Neurochem, 69, 889– 901. Ferreira SH (1972) Prostaglandins, aspirin-like drugs and analgesia. Nature New Biol, 240, 200–203. Fiebich BL, Schleicher S, Butcher RD et al (2000) The neuropeptide substance P activates p38 mitogen-activated protein kinase resulting in IL-6 expression independently from NF-kB. J Immunol, 165, 5606– 5611. Fitzpatrick FA and Soberman R (2001) Regulated formation of eicosanoids. J Clin Invest, 107, 1347–1351. Flower RJ, Moncada S and Vane JR (1980) Analgesic-antipyretics and antiinflammatory agents; drugs employed in the treatment of gout. In Gilman AG, Goodman LS and Gilman A (eds), The Pharmacological Basis of Therapeutics, 6th edn. New York: Macmillan, 682–728. Groppetti A, Braga PC, Biella G et al (1988) Effect of aspirin on serotonin and met-enkephalin in brain: correlation with the antinociceptive activity of the drug. Neuropharmacology, 27, 499–505. Guhring H, Gorig M, Ates M et al (2000) Suppressed injury-induced rise in spinal prostaglandin E2 production and reduced early thermal hyperalgesia in iNOS-deficient mice. J Neurosci, 20, 6714–6720. Guieu R, Blin O, Pouget J and Serratrice G (1992) Analgesic effect of indomethacin shown using the nociceptive flexion reflex in humans. Ann Rheum Dis, 51, 391–393. Habib A, Creminon C, Frobert Y et al (1993) Demonstration of an inducible cyclooxygenase in human endothelial cells using antibodies raised against the carboxyl-terminal region of the cyclooxygenase-2. J Biol Chem, 268, 23448–23454. Herrero JF, Parrado A and Cervero F (1997) Central and peripheral actions of the NSAID ketoprofen on spinal cord nociceptive reflexes. Neuropharmacology, 36, 1425–1431. Hirst WD, Young KA, Newton R et al (1999) Expression of COX-2 by normal and reactive astrocytes in the adult rat central nervous system. Mol Cell Neurosci, 13, 57–68. Hua XY, Calcutt NA and Malmberg AB (1997) Neonatal capsaicin treatment abolishes formalin-induced spinal PGE2 release. NeuroReport, 8, 2325–2329. Hua XY, Chen P, Marsala M and Yaksh TL (1999) Intrathecal substance P-induced thermal hyperalgesia and spinal release of prostaglandin E2 and amino acids. Neuroscience, 89, 525–534. Hunt SP and Mantyh PW (2001) The molecular dynamics of pain control. Nat Rev Neurosci, 2, 83–91. Ichitani Y, Shi T, Haeggstrom JZ et al (1997) Increased levels of cyclooxygenase-2 mRNA in the rat spinal cord after peripheral inflammation: an in situ hybridization study. NeuroReport, 8, 2949– 2952.
Insel PA (1990) Analgesic-antipyretics and antiinflammatory agents; drugs employed in the treatment of rheumatoid arthritis and gout. In Gilman AG, Rall TW, Nies AW and Taylor P (eds), Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 8th edn. New York: Macmillan, 640. Jurna I and Brune K (1990) Central effect of the non-steroid antiinflammatory agents, indomethacin, ibuprofen and diclofenac, determined in C fibre-evoked activity in single neurones of the rat thalamus. Pain, 41, 71–80. Kawamura T, Yamauchi T, Koyama M et al (1997) Expression of prostaglandin EP2 receptor mRNA in the rat spinal cord. Life Sci, 61, 2111–2116. Kawasaki H, Morooka T, Shimohama S et al (1997) Activation and involvement of p38 mitogen-activated protein kinase in glutamateinduced apoptosis in rat cerebellar granule cells. J Biol Chem, 272, 18518–18521. Kulmacz RJ and Wang LH (1995) Comparison of hydroperoxide initiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J Biol Chem, 270, 24019–24023. Lacroix S and Rivest S (1998) Effect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cyclooxygenase enzymes (COX-1 and COX-2) in the rat brain. J Neurochem, 70, 452–466. Lashbrook JM, Ossipov MH, Hunter JC et al (1999) Synergistic antiallodynic effects of spinal morphine with ketorolac and selective COX-1 and COX-2 inhibitors in nerve-injured rats. Pain, 82, 65–72. Leslie JB and Watkins WD (1985) Eicosanoids in the central nervous system. J Neurosurg, 63, 659–668. Lim RK (1970) Pain. Ann Rev Physiol, 32, 269–288. Ma W and Eisenach JC (2002) Morphological and pharmacological evidence for the role of peripheral prostaglandins in the pathogenesis of neuropathic pain. Eur J Neurosci, 15, 1037–1047. Malmberg AB and Yaksh TL (1992a) Antinociceptive actions of spinal non-steroidal antiinflammatory agents on the formalin test in the rat. J Pharmacol Exp Therapeut, 263, 136–146. Malmberg AB and Yaksh TL (1992b) Hyperalgesia mediated by spinal glutamate or substance P receptor blocked by spinal cyclooxygenase inhibition. Science, 257, 1276–1279. Malmberg AB and Yaksh TL (1993) Pharmacology of the spinal action of ketorolac, morphine, ST-91, U50488H, and L-PIA on the formalin test and an isobolographic analysis of the NSAID interaction [see comments]. Anesthesiology, 79, 270–281. Malmberg AB and Yaksh TL (1995a) Cyclooxygenase inhibition and the spinal release of prostaglandin E2 and amino acids evoked by paw formalin injection: a microdialysis study in unanesthetized rats. J Neurosci, 15, 2768–2776. Malmberg AB and Yaksh TL (1995b) The effect of morphine on formalinevoked behaviour and spinal release of excitatory amino acids and prostaglandin E2 using microdialysis in conscious rats. Br J Pharmacol, 114, 1069–1075. Malmstrom K, Daniels S, Kotey P et al (1999) Comparison of rofecoxib and celecoxib, two cyclooxygenase-2 inhibitors, in postoperative dental pain: a randomized, placebo- and active-comparator-controlled clinical trial. Clin Therapeut, 21, 1653–1663. Marnett LJ, Rowlinson SW, Goodwin DC et al (1999) Arachidonic acid oxygenation by COX-1 and COX-2. Mechanisms of catalysis and inhibition. J Biol Chem, 274, 22903–22906. Matsumura K, Cao C, Ozaki M et al (1998) Brain endothelial cells express cyclooxygenase-2 during lipopolysaccharide-induced fever: light and electron microscopic immunocytochemical studies. J Neurosci, 18, 6279–6289. Matsumura K, Watanabe Y and Onoe H (1995) Prostacyclin receptor in the brain and central terminals of the primary sensory neurons: an autoradiographic study using a stable prostacyclin analogue [3H]iloprost. Neuroscience, 65, 493–503. Mazario J, Roza C and Herrero JF (1999) The NSAID dexketoprofen trometamol is as potent as m-opioids in the depression of wind-up and spinal cord nociceptive reflexes in normal rats. Brain Res, 816, 512– 517. McCormack K and Brune K (1991) Dissociation between the antinociceptive and antiinflammatory effects of the non-steroidal antiinflammatory drugs. A survey of their analgesic efficacy. Drugs, 41, 533–547.
PROSTANOIDS IN PAIN Mehlisch DR (1983) Review of the comparative analgesic efficacy of salicylates, acetaminophen, and pyrazolones. Am J Med, 75, 47–52. Melien O, Winsnes R, Refsnes M et al (1988) Pertussis toxin abolishes the inhibitory effects of prostaglandins E1, E2, I2 and F2a on hormoneinduced cAMP accumulation in cultured hepatocytes. Eur J Biochem, 172, 293–297. Michaelis M, Habler HJ and Jaenig W (1996) Silent afferents: a separate class of primary afferents? Clin Exp Pharmacol Physiol, 23, 99–105. Minami T, Uda R, Horiguchi S et al (1992) Allodynia evoked by intrathecal administration of prostaglandin F2a to conscious mice. Pain, 50, 223–229. Minami T, Uda R, Horiguchi S et al (1994) Allodynia evoked by intrathecal administration of prostaglandin E2 to conscious mice. Pain, 57, 217–223. Moncada S, Ferreira SH and Vane JR (1975) Inhibition of prostaglandin biosynthesis as the mechanism of analgesia of aspirin-like drugs in the dog knee joint. Eur J Pharmacol, 31, 250–260. Murakami M, Kambe T, Shimbara S and Kudo I (1999) Functional coupling between various phospholipase A2s and cyclooxygenases in immediate and delayed prostanoid biosynthetic pathways. J Biol Chem, 274, 3103–3115. Narumiya S, Sugimoto Y and Ushikubi F (1999) Prostanoid receptors: structures, properties, and functions. Physiol Rev, 79, 1193–1226. Negishi M, Ito S and Hayaishi O (1989) Prostaglandin E receptors in bovine adrenal medulla are coupled to adenylate cyclase via Gi and to phosphoinositide metabolism in a pertussis toxin-insensitive manner. J Biol Chem, 264, 3916–3923. Negishi M, Sugimoto Y and Ichikawa A (1995) Molecular mechanisms of diverse actions of prostanoid receptors. Biochim Biophys Acta, 1259, 109–119. Oida H, Namba T, Sugimoto Y et al (1995) In situ hybridization studies of prostacyclin receptor mRNA expression in various mouse organs. Br J Pharmacol, 116, 2828–2837. Parsons CG (2001) NMDA receptors as targets for drug action in neuropathic pain. Eur J Pharmacol, 429, 71–78. Pellerin M, Hardy F, Abergel A et al (1987) [Chronic refractory pain in cancer patients. Value of the spinal injection of lysine acetylsalicylate. 60 cases]. Presse Me´d, 16, 1465–1468. Piletta P, Porchet HC and Dayer P (1991) Central analgesic effect of acetaminophen but not of aspirin. Clin Pharmacol Therapeut, 49, 350– 354. Pitcher GM and Henry JL (1999) NSAID-induced cyclooxygenase inhibition differentially depresses long-lasting vs. brief synapticallyelicited responses of rat spinal dorsal horn neurons in vivo. Pain, 82, 173–186. Price DD (1997) Psychophysical measurement of normal and abnormal pain states and analgesia. In Yaksh TL, Lynch CIII, Zapol WM et al (eds), Anesthesia: Biologic Foundations. Philadelphia: LippincottRaven, 719–733. Quan N, Whiteside M and Herkenham M (1998) Cyclooxygenase 2 mRNA expression in rat brain after peripheral injection of lipopolysaccharide. Brain Res, 802, 189–197. Romero SD, Chyatte D, Byer DE et al (1984) Measurement of prostaglandins in the cerebrospinal fluid in cat, dog, and man. J Neurochem, 43, 1642–1649. Saito Y, Kaneko M, Kirihara Y et al (1995) Intrathecal prostaglandin E1 produces a long-lasting allodynic state. Pain, 63, 303–311. Samad TA, Moore KA, Sapirstein A et al (2001) Interleukin-1b-mediated induction of COX-2 in the CNS contributes to inflammatory pain hypersensitivity. Nature, 410, 471–475. Smith JA, Davis CL and Burgess GM (2000a) Prostaglandin E2-induced sensitization of bradykinin-evoked responses in rat dorsal root ganglion neurons is mediated by cAMP-dependent protein kinase A. Eur J Neurosci, 12, 3250–3258. Smith JB and Willis AL (1971) Aspirin selectively inhibits prostaglandin production in human platelets. Nature New Biol, 231, 235–237. Smith WL, DeWitt DL and Garavito RM (2000b) Cyclooxygenases: structural, cellular, and molecular biology. Ann Rev Biochem, 69, 145– 182. Sorkin LS (1992) Release of amino acids and PGE2 into the spinal cord of lightly anesthetized rats during development of an experimental arthritis: enhancement of C-fiber evoked release. Soc Neurosci Abstr, 18, 42910.
479
Sorkin LS and Puig S (1996) Neuronal model of tactile allodynia produced by spinal strychnine: effects of excitatory amino acid receptor antagonists and a m-opiate receptor agonist. Pain, 68, 283–292. Svensson CI, Killremann KE, Dennis EA and Yaksh TY (2001) Effect of spinal administration of phospholipase A2 inhibitors on carrageenan and formalin induced inflammatory hyperalgesia. J Pain, 2(abstr), 34. Svensson CI, Yaksh TL (2002) The spinal phospholipase-cyclooxygenaseprostanoid cascade in nociceptive processing. Annu Rev Pharmacol Toxicol 42, 553–558. Svensson CI, Marsala M, Westerlund A et al (2003) Activation of p38 mitogen-activated protein kinase in spinal microglia is a critical link in inflammation-induced spinal pain processing. J Neurochem, 86, 1534–1544. Takano Y, Kuno Y, Sato E et al (1999) [The enhancement of formalin induced agitation behavior by intrathecal administration of prostaglandin E1]. Masui Jap J Anesthesiol, 48, 841–846. Todd AJ (2002) Anatomy of primary afferents and projection neurones in the rat spinal dorsal horn with particular emphasis on substance P and the neurokinin 1 receptor. Exp Physiol, 87, 245–249. Treede RD, Meyer RA, Raja SN and Campbell JN (1992) Peripheral and central mechanisms of cutaneous hyperalgesia. Progr Neurobiol, 38, 397–421. Uda R, Horiguchi S, Ito S et al (1990) Nociceptive effects induced by intrathecal administration of prostaglandin D2, E2, or F2a to conscious mice. Brain Res, 510, 26–32. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231, 232–235. Vane JR, Bakhle YS and Botting RM (1998) Cyclooxygenases 1 and 2. Ann Rev Pharmacol Toxicol, 38, 97–120. Versteeg HH, van Bergen en Henegouwen PM, van Deventer SJ and Peppelenbosch MP (1999) Cyclooxygenase-dependent signalling: molecular events and consequences. FEBS Lett, 445, 1–5. Watkins LR, Nguyen KT, Lee JE and Maier SF (1999) Dynamic regulation of proinflammatory cytokines. Adv Exp Med Biol, 461, 153– 178. Willer JC, De Broucker T, Bussel B et al (1989) Central analgesic effect of ketoprofen in humans: electrophysiological evidence for a supraspinal mechanism in a double-blind and cross-over study. Pain, 38, 1–7. Willingale HL, Gardiner NJ, McLymont N et al (1997) Prostanoids synthesized by cyclooxygenase isoforms in rat spinal cord and their contribution to the development of neuronal hyperexcitability. Br J Pharmacol, 122, 1593–1604. Willis WD (2001) Role of neurotransmitters in sensitization of pain responses. Ann NY Acad Sci, 933, 142–156. Woodbury DM (1965) Analgesics and antipyretics. In Goodman LS and Gilman A (eds), The Pharmacological Basis of Therapeutics, 3rd edn. London: Macmillan, 312–344. Wright DH, Nantel F, Metters KM and Ford-Hutchinson AW (1999) A novel biological role for prostaglandin D2 is suggested by distribution studies of the rat DP prostanoid receptor. Eur J Pharmacol, 377, 101– 115. Yaksh TL (1982) Central and peripheral mechanisms for the antianalgesic action of acetylsalicylic acid. In Barnet JM, Hirsh J and Mustard JF (eds), Acetylsalicylic Acid: New Uses for an Old Drug. New York: Raven Press, 137–152. Yaksh TL (1989) Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain, 37, 111–123. Yaksh TL (1997) Preclinical models of nociception. In Yaksh TL, Lynch CIII, Zapol WM et al (eds), Anesthesia: Biologic Foundations. Philadelphia: Lippincott-Raven, 685–718. Yaksh TL (1999) Spinal systems and pain processing: development of novel analgesic drugs with mechanistically defined models. Trends Pharmacol Sci, 20, 329–337. Yaksh TL, Dirig DM, Conway CM et al (2001) The acute antihyperalgesic action of NSAIDs and release of spinal PGE2 is mediated by the inhibition of constitutive spinal COX-2 but not COX-1. J Neurosci, 21, 5847–5853. Yaksh TL, Hua XY, Kalcheva et al (1999) The spinal biology in humans and animals of pain states generated by persistent small afferent input. Proc Natl Acad Sci USA, 96, 7680–7686.
480
NERVOUS SYSTEM
Yaksh TL and Svensson C (2001) Role of spinal cyclooxygenases in nociceptive processing. In Vane JR and Botting RM (eds), Therapeutic Roles of Selective COX-2 Inhibitors. London: William Harvey Press, 168–190. Yamamoto T and Nozaki-Taguchi N (1997) Role of spinal cyclooxygenase (COX)-2 on thermal hyperalgesia evoked by carageenan injection in the rat. NeuroReport, 8, 2179–2182. Yamamoto T and Sakashita Y (1998) COX-2 inhibitor prevents the development of hyperalgesia induced by intrathecal NMDA or AMPA. NeuroReport, 9, 3869–3873. Yang LC, Marsala M and Yaksh TL (1996a) Characterization of time course of spinal amino acids, citrulline and PGE2 release after
carrageenan/kaolin-induced knee joint inflammation: a chronic microdialysis study. Pain, 67, 345–354. Yang LC, Marsala M and Yaksh TL (1996b) Effect of spinal kainic acid receptor activation on spinal amino acid and prostaglandin E2 release in rat. Neuroscience, 75, 453–461. Yoneda Y and Ogita K (1991) Neurochemical aspects of the N-methyl-Daspartate receptor complex. Neurosci Res, 10, 1–33. Yousufzai SY, Chen AL and Abdel-Latif AA (1988) Species differences in the effects of prostaglandins on inositol trisphosphate accumulation, phosphatidic acid formation, myosin light chain phosphorylation and contraction in iris sphincter of the mammalian eye: interaction with the cyclic AMP system. J Pharmacol Exp Therapeut, 247, 1064–1072.
44 Eicosanoids: Roles in the Pathophysiology of Cerebral Ischaemia Robert W. Hickey1 and Steven H. Graham1,2 1University
of Pittsburgh School of Medicine and 2V.A. Pittsburgh Healthcare System Pittsburgh, PA, USA
Cerebral ischaemia occurs when there is an interruption in blood flow to the brain, resulting in a mismatch between the oxygen and nutrient supply to the brain and the metabolic demand. Eicosanoids may exacerbate the resulting cell death and dysfunction at several different sites and by several different mechanisms. These include effects on the vasculature, on the blood–brain barrier and on the neurons themselves. In this chapter, we will review the effects of eicosanoids on ischaemic injury at these sites, and review the evidence that pharmacological intervention targeted at eicosanoid production and receptor-mediated actions may ameliorate injury due to cerebral ischaemia.
CEREBROVASCULAR EFFECTS OF EICOSANOIDS A schematic representation of physiological and pathological effects of eicosanoids on vascular function is outlined in Figure 44.1. Vascular endothelial cells under physiological conditions express the inducible isoform of cyclooxygenase (COX-2). Endothelial COX-2 produces the potent vasodilator prostacyclin. Prostacyclin, together with the opposing actions of vasoconstrictors, helps to maintain blood flow within physiological levels across a range of blood pressures (autoregulation). COX-2 also functions to increase blood flow in response to neural activity (Niwa et al 2000). Diseased blood vessels containing cell-rich atherosclerotic plaques (macrophages, smooth muscle cells, endothelium) demonstrate augmented production of COX-2 (Schonbeck et al 1999; Stemme et al 2000). Under these pathological conditions, increased COX activity will produce markedly elevated concentrations of superoxide (O7 2 ), which is a metabolic by-product of COX activity. Superoxide can then react with nitric oxide (NO) to form the highly toxic free radical peroxynitrite (ONOO7). The formation of peroxynitrite contributes to free radical-mediated injury and also reduces the bioavailability of the potent vasodilator NO, thus favouring vasoconstriction. Vasoconstriction is also promoted by binding of the first-order prostaglandin, PGH2, to the thromboxane receptor. Vasoconstriction of an already narrowed vessel increases shear stress and attracts additional platelets to the thrombus. Platelets contain the constitutive isoform of COX (COX-1) and the prostaglandin synthase, thromboxane synthase. Thromboxane provides a positive feedback loop for additional platelet aggregation and vasoconstriction. The thrombus also traps and activates white blood cells containing high levels of lipoxygenases. Lipoxygenases metabolize The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
arachidonic acid into leukotrienes, which are potent vasoconstrictors and promote vascular permeability (causing oedema). Thus, under the pathological conditions found in atherosclerotic cerebrovascular disease, COX-dependent vasoconstriction may overwhelm COX-dependent vasodilation, thereby increasing the susceptibility to stroke (reviewed in Davidge 2001). EFFECTS OF EICOSANOIDS UPON THE BLOOD–BRAIN BARRIER AFTER ISCHAEMIA Another site of action of eicosanoids is the blood–brain barrier. TXA2, other prostaglandins and leukotrienes have potent chemotactic effects that promote platelet and neutrophil adhesion to the endothelium (reviewed in Chen et al 1986). Reactive oxygen species and cytokines produced by neutrophils can disrupt the blood–brain barrier. Prostaglandins and leukotrienes also directly promote opening of the blood–brain barrier (Black and Hoff 1985; Iannotti et al 1981). Increased permeability of the blood– brain barrier results in brain oedema, an important secondary injury factor after stroke. NEURONAL EFFECTS OF EICOSANOIDS IN CEREBRAL ISCHAEMIA Ischaemia causes a cellular energy crisis that overwhelms the ability to maintain membrane potentials. Ionic shifts create widespread depolarization and release of neurotransmitters into synaptic clefts. Release and accumulation of the excitatory neurotransmitter glutamate is a central component of ischaemic brain injury. Glutamate binds to several families of ligand-gated membrane channels that regulate movement of cations across the cell membrane. The NMDA receptor family regulates calcium influx and, when bound with the excitatory neurotransmitter glutamate, causes intracellular accumulation of calcium (Choi 1987). The excessive accumulation of glutamate associated with ischaemia causes NMDA activation sufficient to generate toxic levels of intracellular calcium. Elevated intracellular calcium content indiscriminately activates cytosolic enzymes, initiating multiple pathological cascades, including the translocation and activation of several phospholipases (Moskowitz et al 1985). Phospholipases then cleave phospholipids, releasing arachidonic acid, the parent compound of the eicosanoid pathways. Thus, the eicosanoid pathway is directly connected to the excitatory pathway of ischaemic brain injury.
482
NERVOUS SYSTEM
Figure 44.1 Physiological and pathological effects of eicosanoids upon vascular function. (A) Under physiological conditions, endothelial cells produce prostacyclin (PGI2) and nitric oxide (NO) via COX-2 and eNOS, respectively. PGI2 and NO are potent vasodilators. (B) Under pathological conditions of atherosclerosis, there is elevated expression of both COX-1 and COX-2 from the cellular constituents of the plaque (endothelium, smooth muscle and macrophages). COX activity produces superoxide ion (O7 2 ) as a by-product of prostaglandin production. The increased COX activity results in increased production of superoxide ion, which can react with NO to form the toxic molecule peroxynitrite (ONOO7). Peroxynitrite formation reduces the bioavailability of NO and thus increases vasoconstriction. Vasoconstriction, as well as mechanical obstruction and increased shearing forces, causes aggregation/activation of platelets and white blood cells. The platelets produce thromboxane (via COX-1), creating additional vasoconstriction and platelet aggregation/activation that contributes to thrombus formation. WBCs are attracted to the developing thrombus and produce COX-2, iNOS and lipoxygenases, each of which has products that can contribute to the injury cascade
ARACHIDONIC ACID METABOLISM AND EICOSANOID PRODUCTION IN CEREBRAL ISCHAEMIA Cerebral ischaemia results in activation of several enzymes that increase production of eicosanoids. In the following sections we will review the activation of enzymes and the production of various categories of eicosanoids after ischaemia. Furthermore, we will review evidence that the activity of these enzymes and their eicosanoid products exacerbates injury and that inhibition of enzyme activity or blockade of eicosanoid receptors may protect the brain from ischaemic injury.
PLA2 can contribute to cellular injury by several mechanisms, including disruption of lipid membranes, exacerbation of glutamate release, production of platelet-activating factor, uncoupling of oxidative phosphorylation, and production of lipid peroxides that can contribute to free radical lipid injury (reviewed in Farooqui et al 1997). Several lines of evidence suggest that PLA2 is an important mediator of ischaemic injury. PLA2 added directly to neuronal culture causes apoptotic cell death that can be attenuated by co-treatment with a COX-2 inhibitor (Yagami et al 2002). Treatment of rodents with the phospholipase inhibitors indoxam (Yagami et al 2002) and quinacrine (Phillis 1996) attenuates ischaemic injury. Finally, cPLA2-null mice have smaller infarcts, less oedema, and fewer neurological defects after transient middle cerebral artery occlusion (Bonventre 1997; Sapirstein and Bonventre 2000).
PHOSPHOLIPASE A2 IN CEREBRAL ISCHAEMIA Once ischaemia develops, there is an almost immediate (within 1 min) increase in phospholipase A2 (PLA2) activity (Davidge 2001). In vivo experiments suggest that the increased activity is mediated, in part, by glutamate (Bonventre 1997). In addition to the almost immediate increase in activity, ischaemia causes an increase in PLA2 protein levels (Saluja et al 1997, 1999). In global ischaemia (modelling cardiac arrest), increased protein is localized to astrocytes within the vulnerable CA1 hippocampus region at 72 h, suggesting an important role in delayed neuronal death (Clemens et al 1996; Walton et al 1997).
ARACHIDONIC ACID METABOLISM: CYCLOOXYGENASE AND LIPOXYGENASES The principal product of PLA2, arachidonic acid, is found in elevated concentrations within neurons almost immediately (30– 60 s) after the onset of ischaemia (Abe et al 1987; De Medio et al 1980; Westerberg et al 1987). Arachidonic acid is metabolized via the cyclooxygenase pathway to prostaglandins and thromboxanes. Alternatively, arachidonic acid can be metabolized by several distinct lipoxygenases into their associated products.
EICOSANOIDS IN CEREBRAL ISCHAEMIA Lipoxygenase metabolites linked to ischaemic brain injury include the leukotrienes and hydroxyeicosatetraenoic acids (HETEs). Reports of increased concentrations of the eicosanoids PGE2 (Carasso et al 1977), PGF2 (Egg et al 1980; Kostic et al 1984), and thromboxane (Fagan et al 1986) in human cerebrospinal fluid following stroke provide indirect evidence of ischaemia-induced increase in free arachidonic acid. Likewise, blood levels of leukotrienes (Katsura et al 1988) and urinary metabolites of thromboxane (Koudstaal et al 1993; McConnell et al 2001; van Kooten et al 1999) are elevated in patients with stroke. LEUKOTRIENES IN CEREBRAL ISCHAEMIA Leukotrienes can exacerbate ischaemic brain injury by enhancing vascular permeability, promoting tissue oedema, and attracting/ activating white blood cells. The leukotrienes LTC4, LTD4, and LTE4 are elevated in rodents for several hours after ischaemia (Ban et al 1989; Ciceri et al 2001; Moskowitz et al 1984; Namura et al 1994; Ohtsuki et al 1995). Treatment with lipoxygenase inhibitors results in less post-ischaemic oedema (Baskaya et al 1996; Dempsey et al 1986a) and neuronal cell death (Ciceri et al 2001; Rao et al 1999). An important consideration in experimental or clinical use of COX inhibitors is that COX inhibition will shunt arachidonic acid from the prostaglandin pathway into the lipoxygenase pathway (Dempsey et al 1986b). To the best of our knowledge, targeted inhibition of the lipoxygenase pathway has not been investigated in humans with stroke. HYDROXYEICOSATETRAENOIC ACIDS IN CEREBRAL ISCHAEMIA 12-HETE and its hepoxilin metabolites have been well characterized as important components of cardioprotective ischaemic preconditioning (Gabel et al 2001). Preconditioning is a phenomenon (described in both heart and brain) where brief sublethal injury increases tolerance to a subsequent potentially lethal injury. In brain, 12-HETE is an inhibitory modulator that can be produced by NMDA receptor-mediated activity. Recently, the hypothesis that 12-HETE is neuroprotective was tested in an in vitro model of excitotoxicity (Hampson and Grimaldi 2002). The investigators reported that 12-HETE protected cortical neurons from excitotoxic injury by limiting calcium influx, and they postulate the existence of a unique receptor or enzyme responsible for this effect. Additional experiments are required to clarify the role of 12-HETE and other hepoxilins in ischaemic brain injury. CYCLOOXYGENASE IN CEREBRAL ISCHAEMIA There are two isomers of cyclooxygenase, COX-1 and COX-2. The highest level of expression of COX-2 is found in brain. Worley et al found it to be identical to COX-2 found in other tissues (Yamagata et al 1993). Stimulation of the perforant path resulted in expression of COX-2 in hippocampus and physiological stimuli (cold stress)-induced COX-2 expression in cortex. This expression could be blocked by the NMDA antagonist MK801 and glucocorticoids (Yamagata et al 1993). Expression of COX-2 can be inhibited by a number of stimuli that inhibit synaptic activity, including NMDA antagonists and the usedependent Na+ channel antagonist, lamotrigine (Adams et al 1996). Basal expression of COX-2 is found in glutamatergic neurons, especially in dendritic processes, consistent with the role of COX-2 in producing neuronal responses to synaptic activity (Kaufmann et al 1996). Basal expression of COX-2 is prominent in hippocampus and limbic cortex, consistent with a high density
483
of excitatory amino acid receptors in these regions (Breder et al 1995). Spreading depression evoked by K+ injection in parietal cortex also induces expression of COX-2. COX-2 expression was also induced at sites several synapses distant from the spreading depression. This observation demonstrates that benign stimuli can increase COX-2 expression and thus may influence brain function for extended periods and at locations distant from the initial stimulus (Caggiano et al 1996). COX-2 activity may be exacerbated after cerebral ischaemic injury. COX-2 is upregulated acutely in human neurons, endothelial cells and glia after stroke (Tomimoto et al 2000). Others have also found prolonged (42 weeks) expression of COX-2 in macrophages and glia (reactive astrocytes) in human brain after stroke (Maslinska et al 1999). Nogawa et al (1997) found that COX-2 expression was induced by temporary focal ischaemia in rodents and that a COX-2 inhibitor decreased infarction volume when administered after middle cerebral artery occlusion. Nakayama et al (1996) found that COX-2 mRNA and protein were induced in CA1 hippocampal neurons after ischaemia. Treatment with a COX-2 inhibitor decreased concentration of PGE2 and increased CA1 neuronal survival after global ischaemia in rodents. Disruption of the COX-2 gene in mice also decreases the volume of infarction after middle cerebral artery occlusion (Iadecola et al 2001b). Furthermore, COX-2 gene disruption decreases the susceptibility of neurons to the effects of NMDA injected into brain (Iadecola et al 2001b). These data provide convincing evidence that COX-2 activity may exacerbate injury after cerebral ischaemia. COX-1 is the constitutive isoform of cyclooxygenase that is found in many cell types, including glia (O’Banion et al 1992). COX-1 may have an important role in maintaining inflammation in many cell types. Furthermore, COX-1 may mediate vascular responses that exacerbate ischaemia, including platelet aggregation and vasoconstriction. A recent report shows that COX-1 can also participate in vasodilation under some circumstances (Niwa et al 2001). COX-1 is found in microglia/macrophages in human brain following stroke, both early on and after several months (Schwab et al 2000). COX-1 expression is increased in neurons injured after brain trauma (Schwab et al 2001). Specific COX-1 inhibitors do not protect neurons against NMDA toxicity (Hewett et al 2000). Furthermore, COX-1-null mice have larger infarction volumes than wild-type controls (Iadecola et al 2001a). This effect is associated with an exacerbation of decrease in blood flow during middle cerebral artery occlusion. The exact mechanism and site by which COX-2 activity exacerbates injury is unclear. Cyclooxygenase has long been known to be an important mediator of inflammation. COX-2 inhibitors protect neurons against lipopolysaccaride-induced cell death in neuronal–glial co-cultures. Lipopolysaccharide does not induce cell death in pure neuronal culture, suggesting that glia are required in this COX-2-mediated mechanism (Araki et al 2001). Other studies suggest that COX-2 activity within the neuron itself produces injury. The COX-2-null mouse is resistant to the effects of intracerebral NMDA injections. Since the NMDA receptor is found exclusively on neurons, these data suggest that COX-2 inhibition is acting at a neuronal site. COX-2 inhibitors also protect neurons against NMDA and hypoxia in pure primary neuronal cultures (Hewett et al 2000; Li and Graham 2001). Ischaemic injury in vivo could potentially involve both neuronal and glial sites of action; further experiments are needed to determine the relative contributions of both mechanisms. COX-2 activity may have many effects in the vasculature and cerebral blood flow. Elegant experiments by Iadecola et al suggest that COX-2 may be one of the principal mechanisms by which increased cerebral metabolic activity is coupled to vasodilatation and increased regional blood flow. Both COX-2 inhibition and gene disruption blocked the increase in blood flow in mouse
484
NERVOUS SYSTEM
somatosensory cortex induced by whisker stimulation (Niwa et al 2000). COX-2 gene disruption had no effect on cerebral blood flow during middle cerebral artery occlusion (Iadecola et al 2001b). Thus, the exacerbation of injury after cerebral ischaemia by COX-2 is not explained by its effects on blood flow. Cyclooxygenase activity may result in oxidative stress. Oxidative stress has been implicated in the pathogenesis of neuronal death after cerebral ischaemia and in a variety of neurodegenerative diseases. Cyclooxygenase and nitric oxide synthase (NOS) may act synergistically via a variety of mechanisms to produce neuronal death (Dirnagl et al 1999). One such synergy is the activation of cyclooxygenase by peroxynitrate, one of the primary toxic products of nitric oxide (Landino et al 1996). Such a mechanism is suggested in cerebral ischaemia by the finding that COX-2 inhibitors are ineffective neuroprotectants in iNOS-null mice (Nagayama et al 1999). The discrepancy between inhibition or disruption of COX-1 and COX-2 is not well explained by oxidative stress, since both COX-1 and COX-2 have identical peroxidase functions (Lehmann 1994). Thus, another effect, such as the coupling of COX-1 and COX-2 to the production of different prostaglandins, may underlie the opposing actions of COX-1 and COX-2 inhibition on cell death after cerebral ischaemia. PROSTAGLANDINS IN CEREBRAL ISCHAEMIA Prostaglandins are produced in normal brain, and prostaglandin receptors are widely distributed throughout the CNS, suggesting that eicosanoids have important, widespread neuromodulatory effects. Indeed, prostaglandins have been linked to the physiological regulation of cerebral blood flow (already discussed), sleep–wake cycle, body temperature, and luteinizing hormonereleasing hormone. Recent investigations of prostaglandins in pathological states, such as inflammation and ischaemia, suggest that prostaglandins also have important pathological effects. Evidence is now emerging that prostaglandins have direct neuronal effects upon the ischaemic brain, independent of their vascular effects. Both neuroprotective and neurotoxic effects have been reported. Addition of the first-order PGH2 metabolites PGD2, PGE1, PGE2, PGF2a, and PGI2 protects against glutamate toxicity in vitro. A novel prostacyclin receptor subtype specific for brain has been identified, and administration of a ligand with high affinity for the receptor attenuates neuronal death in cultures exposed to oxygen stress or serum deprivation (Satoh et al 1999). Similarly, administration of a CNS-subtype PGI2 receptor ligand prior to and following forebrain ischaemia in the gerbil provided almost complete protection for CA1 neurons (Satoh et al 1999), and treatment in rats immediately after middle cerebral artery occlusion reduced infarct volume by 35% at 24 h (Takamatsu et al 2002). Prostaglandin concentrations are increased in spinal cord in response to peripheral inflammation and augment pain perception via PGE2 receptor-mediated suppression of inhibitory glycinergic neurotransmission, thus favouring excitatory over inhibitory input (Ahmadi et al 2002). Also, the prostaglandin E receptor agonist lipo-PGE1 protected spinal cord from apoptotic cell death induced by peripheral sciatic nerve constriction (Kawamura et al 1997). In contrast to these neuroprotective effects, addition of PGD2 or its cyclopentanone metabolite D12PGJ2 to neuronal culture is associated with mitochondrial dysfunction (Yagami et al 2002). The cyclopentanone 15-deoxyD12,14-PGJ2 is increased in spinal cord of patients with sporadic amyotrophic lateral sclerosis and causes p53-mediated activation of the Fas–FasL pathway associated with apoptosis when added to neurons in vitro (Kondo et al 2002). Thus, coupling of specific prostaglandin synthases with cyclooxygenase and the regional distribution of specific prostaglandin receptors may be important
determinants of neuronal fate. Additional experiments are required to characterize the distribution of both synthases and receptors in order to determine their relative contribution to ischaemic injury. Evidence is now emerging that prostaglandins are differentially expressed following an injury to mediate both inflammation and repair. In models of inflammation, two waves of COX-2 expression associated with distinct downstream prostaglandins and cyclopentanone metabolites have been described (reviewed in Gilroy 1999). Inhibition of the first wave of COX-2 expression attenuates inflammation, whereas inhibition of the second wave causes a prolongation of inflammation and attenuates tissue repair. Parallel studies have not been performed in injured brain. Discovery of a similar response in brain (bimodal COX-2/ prostaglandin expression associated first with inflammation and then with repair) would have important implications for clinical care. CLINICAL TRIALS OF EICOSANOIDS AND CYCLOOXYGENASE INHIBITORS IN CEREBRAL ISCHAEMIA Prostacyclin infusions have demonstrated favourable results in animal studies and in non-randomized human trials. However, randomized studies have failed to show a clear benefit (Bath and Bath 2000). A significant limitation of prostacyclin therapy is that it causes systemic hypotension. Systemic hypotension can decrease CBF and thus attenuate the beneficial effects of local vasodilation (Brown and Pickles 1982; Cook et al 1983). Administration of low-dose aspirin once a day for patients who present with transient ischaemic attacks (TIAs) reduces the risk of subsequent vascular events (Antiplatelet Trials Collaboration 1988). The mechanism by which aspirin prevents stroke is thought to be due to its effects on platelet aggregation. Platelets, which lack a nucleus and therefore the capability to synthesize new COX, are permanently inhibited by even intermittent, low doses of cyclooxygenase inhibitors. On the other hand, neurons (and other cells with nuclei) will synthesize new COX enzyme and recover function once all aspirin is either bound or cleared from the body. There are now two large prospective clinical trials investigating the use of 160–300 mg aspirin initiated after the onset of stroke (International Stroke Trial Collaborative Group 1997; Chinese Acute Stroke Trial Collaborative Group 1997). A total of approximately 40 000 adults were enrolled. Patients randomized to aspirin had reduced stroke recurrence and fewer deaths at 6 months follow-up. These trials do not differentiate between the beneficial effects secondary to platelet inhibition and direct neuronal effects. However, as discussed above, evidence from in vitro experiments (lacking vascular effects) and from in vivo experiments in which cerebral blood flow was unaffected demonstrate a benefit to COX-2 inhibition independent of vascular effects. A recent trial of adults treated with either a non-selective NSAID (naproxen) or a specific COX-2 inhibitor (rofecoxib) for rheumatoid arthritis found an increase in myocardial infarction among patients treated with the COX-2 inhibitor (Bombardier et al 2000). One potential explanation of this finding is that COX-2specific inhibition may limit endothelial prostacyclin production while leaving platelet thromboxane production unchecked, thereby increasing the risk of clot formation (Cheng et al 2002; Vane 2002). There have been no reports of increased risk of cerebrovascular events associated with use of COX-2 inhibitors. Nonetheless, this is an important clinical issue to resolve. It is likely that it is desirable to acutely inhibit both COX-1 and COX-2 in patients with stroke. COX-1 inhibition will inhibit
EICOSANOIDS IN CEREBRAL ISCHAEMIA platelet function, and COX-2 inhibition will attenuate the neuronal toxicity associated with COX-2 activity. Although non-specific cyclooxygenase inhibitors will achieve this goal, a combination of low-dose aspirin combined with a selective COX-2 inhibitor would be less likely to cause unwanted gastrointestinal bleeding and renal toxicity. Thus, low-dose aspirin combined with a selective COX-2 inhibitor may be an attractive therapy to test in future trials of patients with stroke.
SUMMARY Brain ischaemia causes an immediate increase in phospholipase activity and protein level accompanied by an increase in free arachidonic acid. Arachidonic acid is then metabolized by the cyclooxygenase or lipoxygenase pathways. Lipoxygenase generates leukotrienes, which can exacerbate tissue oedema and inflammation. COX is coupled to specific prostaglandin synthases that metabolize the first-order COX product PGH2 into their respective prostaglandins. COX-2, the predominant isoform in vascular endothelial cells, is coupled with production of the potent vasodilator prostacyclin. COX-1, the predominant isoform in platelets, is coupled with production of the potent vasoconstrictor thromboxane. Under pathological conditions, accumulating platelets and other pathological events favour the production of vasoconstricting prostanoids. Expression and activity of COX-2, the predominant isoform in neurons, is induced by cerebral ischaemia. COX-2 activity is associated with exacerbation of neuronal injury, independent of effects on cerebral blood flow. The availability of COX-2-specific inhibitors makes the COX-2 pathway an attractive target for stroke therapy. The direct neurotoxic or neuroprotective effects of specific prostaglandins and prostaglandin metabolites are poorly understood and represent an important direction for future research.
REFERENCES Abe K, Kogure K, Yamamoto H et al (1987) Mechanism of arachidonic acid liberation during ischemia in gerbil cerebral cortex. J Neurochem, 48, 503–509. Adams J, Collaco-Moraes Y and de Belleroche J (1996) Cyclooxygenase-2 induction in cerebral cortex: an intracellular response to synaptic excitation. J Neurochem, 66, 6–13. Antiplatelet Trialists’ Collaboration (1988) Secondary prevention of vascular disease by prolonged antiplatelet treatment. Br Med J, 296, 320–331. Ahmadi S, Lippross S, Neuhuber WL and Zeilhofer HU (2002) PGE(2) selectively blocks inhibitory glycinergic neurotransmission onto rat superficial dorsal horn neurons. Nature Neurosci, 5, 34–40. Araki E, Forster C, Dubinsky JM et al (2001) Cyclooxygenase-2 inhibitor ns-398 protects neuronal cultures from lipopolysaccharide-induced neurotoxicity. Stroke, 32, 2370–2375. Ban M, Tonai T, Kohno T et al (1989) A flavonoid inhibitor of 5lipoxygenase inhibits leukotriene production following ischemia in gerbil brain. Stroke, 20, 248–252. Baskaya MK, Hu Y, Donaldson D et al (1996) Protective effect of the 5lipoxygenase inhibitor AA-861 on cerebral edema after transient ischemia. J Neurosurg, 85, 112–116. Bath PM and Bath FJ (2000) Prostacyclin and analogues for acute ischaemic stroke. Cochrane Database Syst Rev, CD000177. Black KL and Hoff JT (1985) Leukotrienes increase blood–brain barrier permeability following intraparenchymal injections in rats. Ann Neurol, 18, 349–351. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med, 343, 1520– 1528.
485
Bonventre JV (1997) Roles of phospholipases A2 in brain cell and tissue injury associated with ischemia and excitotoxicity. J Lipid Mediat Cell Signal, 17, 71–79. Breder CD, Dewitt D and Kraig RP (1995) Characterization of inducible cyclooxygenase in rat brain. J Comp Neurol, 355, 296–315. Brown MM and Pickles H (1982) Effect of epoprostenol (prostacyclin, PGI2) on cerebral blood flow in man. J Neurol Neurosurg Psychiat, 45, 1033–1036. Caggiano AO, Breder CD and Kraig RP (1996) Long-term elevation of cyclooxygenase-2, but not lipoxygenase, in regions synaptically distant from spreading depression. J Comp Neurol, 376, 447–462. Carasso RL, Vardi J, Rabay JM et al (1977) Measurement of prostaglandin E2 in cerebrospinal fluid in patients suffering from stroke. J Neurol Neurosurg Psychiat, 40, 967–969. Chen ST, Hsu CY, Hogan EL et al (1986) Thromboxane, prostacyclin, and leukotrienes in cerebral ischemia. Neurology, 36, 466–470. Cheng Y, Austin SC, Rocca B et al (2002) Role of prostacyclin in the cardiovascular response to thromboxane A2. Science, 296, 539–541. Chinese Acute Stroke Trial (CAST) Group (1997) CAST: randomised placebo-controlled trial of early aspirin use in 20 000 patients with acute ischaemic stroke. Lancet, 349, 1641–1649. Choi DW (1987) Ionic dependence of glutamate neurotoxicity. J Neurosci, 7, 369–379. Ciceri P, Rabuffetti M, Monopoli A and Nicosia S (2001) Production of leukotrienes in a model of focal cerebral ischaemia in the rat. Br J Pharmacol, 133, 1323–1329. Clemens JA, Stephenson DT, Smalstig EB et al (1996) Reactive glia express cytosolic phospholipase A2 after transient global forebrain ischemia in the rat. Stroke, 27, 527–535. Collaboration AT (1988) Secondary prevention of vascular disease by prolonged antiplatelet treatment. Br Med J (Clin Res Ed), 296, 320– 331. Colville-Nash PR and Gilroy DW (2000) COX-2 and the cyclopentenone prostaglandins—a new chapter in the book of inflammation? Prostagland Lipid Mediat, 62, 33–43. Cook PJ, Maidment CG, Dandona P et al (1983) The effect of intravenous epoprostenol (prostacyclin, PGI2) on cerebral blood flow and cardiac output in man. Br J Clin Pharmacol, 16, 707–711. Davidge ST (2001) Prostaglandin H synthase and vascular function. Circ Res, 89, 650–660. De Medio GE, Goracci G, Horrocks LA et al (1980) The effect of transient ischemia on fatty acid and lipid metabolism in the gerbil brain. Ital J Biochem, 29, 412–432. Dempsey RJ, Roy MW, Cowen DE and Maley ME (1986a) Lipoxygenase metabolites of arachidonic acid and the development of ischaemic cerebral oedema. Neurol Res, 8, 53–56. Dempsey RJ, Roy MW, Meyer K et al (1986b) Development of cyclooxygenase and lipoxygenase metabolites of arachidonic acid after transient cerebral ischemia. J Neurosurg, 64, 118–124. Dirnagl U, Iadecola C and Moskowitz MA (1999) Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci, 22, 391–397. Egg D, Herold M, Rumpl E and Gunther R (1980) Prostaglandin F2a levels in human cerebrospinal fluid in normal and pathological conditions. J Neurol, 222, 239–248. Fagan SC, Castellani D and Gengo FM (1986) Prostanoid concentrations in human CSF following acute ischaemic brain infarction. Clin Exp Pharmacol Physiol, 13, 629–632. Farooqui AA, Yang HC, Rosenberger TA and Horrocks LA (1997) Phospholipase A2 and its role in brain tissue. J Neurochem, 69, 889– 901. Gabel SA, London RE, Funk CD et al (2001) Leukocyte-type 12lipoxygenase-deficient mice show impaired ischemic preconditioninginduced cardioprotection. Am J Physiol Heart Circ Physiol, 280, H1963–1969. Gilroy DW, Colville-Nash PR, Willis D et al (1999) Inducible cyclooxygenase may have anti-inflammatory properties. Nat Med, 5(6), 698–701. International Stroke Trial Group (1997) A randomized trial of aspirin, subcutaneous heparin, both, or neither among 19435 patients with acute ischaemic stroke. Lancet, 349, 1569–1581. Hampson AJ and Grimaldi M (2002) 12-Hydroxyeicosatetrenoate (12HETE) attenuates AMPA receptor-mediated neurotoxicity: evidence for a G-protein-coupled HETE receptor. J Neurosci, 22, 257–264.
486
NERVOUS SYSTEM
Hewett SJ, Uliasz TF, Vidwans AS and Hewett JA (2000) Cyclooxygenase-2 contributes to N-methyl-D-aspartate-mediated neuronal cell death in primary cortical cell culture. J Pharmacol Exp Ther, 293, 417–425. Iadecola C, Sugimoto K, Niwa K et al (2001a) Increased susceptibility to ischemic brain injury in cyclooxygenase-1-deficient mice. J Cereb Blood Flow Metab, 21, 1436–1441. Iadecola C, Niwa K, Nogawa S et al (2001b) Reduced susceptibility to ischemic brain injury and N-methyl-D-aspartate-mediated neurotoxicity in cyclooxygenase-2-deficient mice. Proc Natl Acad Sci USA, 98, 1294–1299. Iannotti F, Crockard A, Ladds G and Symon L (1981) Are prostaglandins involved in experimental ischemic edema in gerbils? Stroke, 12, 301– 306. Katsura K, Minamisawa H, Katayama Y et al (1988) Plasma levels of leukotriene C4, B4, slow reacting substance of anaphylaxis in chronological phases of cerebrovascular disease. Prostaglandins, 36, 655–665. Kaufmann WE, Worley PF, Pegg J et al (1996) COX-2, a synaptically induced enzyme, is expressed by excitatory neurons at postsynaptic sites in rat cerebral cortex. Proc Natl Acad Sci USA, 93, 2317–2321. Kawamura T, Akira T, Watanabe M and Kagitani Y (1997) Prostaglandin E1 prevents apoptotic cell death in superficial dorsal horn of rat spinal cord. Neuropharmacology, 36, 1023–1030. Kondo M, Shibata T, Kumagai T et al (2002) 15-Deoxy-D(12,14)prostaglandin J(2): the endogenous electrophile that induces neuronal apoptosis. Proc Natl Acad Sci USA, 99, 7367–7372. Kostic VS, Djuricic BM and Mrsulja BB (1984) Cerebrospinal fluid prostaglandin F2a in stroke patients: no relationship to the degree of neurological deficit. Eur Neurol, 23, 291–295. Koudstaal PJ, Ciabattoni G, van Gijn J et al (1993) Increased thromboxane biosynthesis in patients with acute cerebral ischemia. Stroke, 24, 219–223. Landino LM, Crews BC, Timmons MD et al (1996) Peroxynitrite, the coupling product of nitric oxide and superoxide, activates prostaglandin biosynthesis. Proc Natl Acad Sci USA, 93, 15069–15074. Lehmann WD (1994) Regio- and stereochemistry of the dioxygenation reaction catalyzed by (S)-type lipoxygenases or by the cyclooxygenase activity of prostaglandin H synthases. Free Radic Biol Med, 16, 241– 253. Li WJ and Graham SH (2001) Cyclooxygenase-2 inhibition but not cyclooxygenase-1 inhibition decreases death due to transient hypoxia in primary neuronal cultures. Soc Neurosci, 27(abstr). Maslinska D, Wozniak R, Kaliszek A and Modelska I (1999) Expression of cyclooxygenase-2 in astrocytes of human brain after global ischemia. Folia Neuropathol, 37, 75–79. McConnell JP, Cheryk LA, Durocher A et al (2001) Urinary 11-dehydrothromboxane B(2) and coagulation activation markers measured within 24 h of human acute ischemic stroke. Neurosci Lett, 313, 88–92. Moskowitz MA, Kiwak KJ, Hekimian K and Levine L (1984) Synthesis of compounds with properties of leukotrienes C4 and D4 in gerbil brains after ischemia and reperfusion. Science, 224, 886–889. Moskowitz N, Andres A, Silva W et al (1985) Calcium-dependent binding of calmodulin to phospholipase A2 subunits induces enzymatic activation. Arch Biochem Biophys, 241, 413–417. Nagayama M, Niwa K, Nagayama T et al (1999) The cyclooxygenase-2 inhibitor NS-398 ameliorates ischemic brain injury in wild-type mice but not in mice with deletion of the inducible nitric oxide synthase gene. J Cereb Blood Flow Metab, 19, 1213–1219. Nakayama M, Uchimura K, Zhu L et al (1996) Cyclooxygenase 2 promotes neuronal cell death after global ischemia in rat CA1 hippocampus. Soc Neurosci, 22, 1670(abstr). Namura Y, Shio H and Kimura J (1994) LTC4/LTB4 alterations in rat forebrain ischemia and reperfusion and effects of AA-861, CV-3988. Acta Neurochir Suppl (Wien), 60, 296–299. Niwa K, Araki E, Morham SG et al (2000) Cyclooxygenase-2 contributes to functional hyperemia in whisker-barrel cortex. J Neurosci, 20, 763– 770.
Niwa K, Haensel C, Ross ME and Iadecola C (2001) Cyclooxygenase-1 participates in selected vasodilator responses of the cerebral circulation. Circ Res, 88, 600–608. Nogawa S, Zhang F, Ross ME and Iadecola C (1997) Cyclooxygenase-2 gene expression in neurons contributes to ischemic brain damage. J Neurosci, 17, 2746–2755. O’Banion MK, Winn VD and Young DA (1992) cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc Natl Acad Sci USA, 89, 4888–4892. Ohtsuki T, Matsumoto M, Hayashi Y et al (1995) Reperfusion induces 5lipoxygenase translocation and leukotriene C4 production in ischemic brain. Am J Physiol, 268, H1249–1257. Phillis JW (1996) Cerebroprotective action of the phospholipase inhibitor quinacrine in the ischemia/reperfused gerbil hippocampus. Life Sci, 58, PL97–101. Rao AM, Hatcher JF, Kindy MS and Dempsey RJ (1999) Arachidonic acid and leukotriene C4: role in transient cerebral ischemia of gerbils. Neurochem Res, 24, 1225–1232. Saluja I, Song D, O’Regan MH and Phillis JW (1997) Role of phospholipase A2 in the release of free fatty acids during ischemiareperfusion in the rat cerebral cortex. Neurosci Lett, 233, 97–100. Saluja I, O’Regan MH, Song D and Phillis JW (1999) Activation of cPLA2, PKC and ERKs in the rat cerebral cortex during ischemia/ reperfusion. Neurochem Res, 24, 669–677. Sapirstein A and Bonventre JV (2000) Phospholipases A2 in ischemic and toxic brain injury. Neurochem Res, 25, 745–753. Satoh T, Ishikawa Y, Kataoka Y et al (1999) CNS-specific prostacyclin ligands as neuronal survival-promoting factors in the brain. Eur J Neurosci, 11, 3115–3124. Schonbeck U, Sukhova GK, Graber P et al (1999) Augmented expression of cyclooxygenase-2 in human atherosclerotic lesions. Am J Pathol, 155, 1281–1291. Schwab JM, Nguyen TD, Postler E et al (2000) Selective accumulation of cyclooxygenase-1-expressing microglial cells/macrophages in lesions of human focal cerebral ischemia. Acta Neuropathol (Berl), 99, 609–614. Schwab JM, Seid K and Schluesener HJ (2001) Traumatic brain injury induces prolonged accumulation of cyclooxygenase-1 expressing microglia/brain macrophages in rats. J Neurotrauma, 18, 881–890. Stemme V, Swedenborg J, Claesson H and Hansson GK (2000) Expression of cyclooxygenase-2 in human atherosclerotic carotid arteries. Eur J Vasc Endovasc Surg, 20, 146–152. Takamatsu H, Tsukada H, Watanabe Y et al (2002) Specific ligand for a central type prostacyclin receptor attenuates neuronal damage in a rat model of focal cerebral ischemia. Brain Res, 925, 176–182. Tomimoto H, Akiguchi I, Wakita H et al (2000) Cyclooxygenase-2 is induced in microglia during chronic cerebral ischemia in humans. Acta Neuropathol (Berl), 99, 26–30. van Kooten F, Ciabattoni G, Koudstaal PJ et al (1999) Increased platelet activation in the chronic phase after cerebral ischemia and intracerebral hemorrhage. Stroke, 30, 546–549. Vane JR (2002) Biomedicine. Back to an aspirin a day? Science, 296, 474– 475. Walton M, Sirimanne E, Williams C et al (1997) Prostaglandin H synthase-2 and cytosolic phospholipase A2 in the hypoxic–ischemic brain: role in neuronal death or survival? Brain Res Mol Brain Res, 50, 165–170. Westerberg E, Deshpande JK and Wieloch T (1987) Regional differences in arachidonic acid release in rat hippocampal CA1 and CA3 regions during cerebral ischemia. J Cereb Blood Flow Metab, 7, 189–192. Yagami T, Ueda K, Asakura K et al (2002) Human group IIA secretory phospholipase A2 induces neuronal cell death via apoptosis. Mol Pharmacol, 61, 114–126. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386.
45 NSAIDs in the Treatment of Alzheimer’s Disease Paul S. Aisen Georgetown University Medical Center, Washington, DC, USA
Alzheimer’s disease (AD) is among the most important health care problems (Aisen et al 1999). The prevalence exceeds 25% in individuals over the age of 85. As life-expectancy increases, the number of cases climbs: in the USA, it is estimated that there are over 4 million cases today and will be over 14 million by midcentury. At present, this is a uniformly fatal illness, typically progressing from mild impairment to death over a span of 7–10 years. The impact of AD on the families of patients is devastating, as they must witness the gradual disappearance of personality and function. The past decade has seen the development of effective drugs, the cholinesterase inhibitors, for the treatment of AD. Acetylcholine is the most important neurotransmitter for cognitive function, and cholinergic markers are depleted in the AD brain. Cognitive function can be improved by prolonging the activity of synaptic acetylcholine; this is accomplished by inhibiting acetylcholinesterase, the major enzyme involved in clearance. The first cholinesterase inhibitor approved by the FDA for the treatment of the cognitive symptoms of AD was tacrine (Davis et al 1992). Tacrine use is problematic, particularly because of hepatotoxicity, and it has now been supplanted by three newer cholinesterase inhibitors: donepezil (Rogers et al 1998), rivastigmine (Corey-Bloom et al 1998) and galantamine (Raskind et al 2000). These drugs carry minimal risk of serious toxicity, and have gained widespread acceptance in the treatment of mild to moderate stage AD. The benefit of cholinesterase inhibitors is presumably symptomatic, although some have suggested additional effects on the disease process that may alter the rate of progression (Giacobini 2001). These drugs improve cognitive performance, stabilize clinical status, preserve function, reduce the emergence of abnormal behaviours and relieve caregiver stress (Grutzendler and Morris 2001; Schneider 2001). But AD continues to progress despite cholinesterase inhibitor therapy. The major goal of pharmaceutical companies and academic investigators working on AD therapeutics is the development of disease-modifying therapy (Aisen and Davis 1997). In the past few years, this goal has seemed much closer, as investigators have reached a consensus regarding the inciting event in AD pathophysiology. A growing understanding of the genetics of the disease has played a major role in clarifying the aetiological mechanisms. While most cases of AD are sporadic and polygenic, a small number of cases are caused by a single mutation. Mutations of any one of three genes, coding for the amyloid precursor protein (APP), presenilin 1 and presenilin 2, can cause AD. The breakthrough in elucidating the pivotal mechanism of the disease was the revelation that each of the AD-causing mutations of these genes influences a single biochemical step: the cleavage of the The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
transmembrane APP by secretase enzymes to release the amyloidogenic fragment, amyloid b peptide (Ab). Along with the large number of studies demonstrating direct and indirect neurotoxicity of Ab, this is strong evidence that generation of Ab is the key event in the pathophysiology of AD. Several strategies for controlling this inciting pathway are evident, and each is now being actively pursued. The first strategy, blocking production of the Ab fragment, involves developing specific inhibitors of the b- or g-secretases that cleave APP at the two ends of the amyloidogenic peptide. While there has been some concern that g-secretase inhibition may interfere with vital pathways (Jack et al 2001), knockout studies in mice suggest that b-secretase function can be eliminated without adverse effects (Roberds et al 2001), encouraging the development of b-secretase inhibitors for AD therapy. The second strategy, increasing clearance of the Ab peptide, includes use of amyloid vaccines (Morgan et al 2000; Schenk et al 1999), and perhaps induction of degrading enzymes such as neprilysin (Hama et al 2001; Iwata et al 2000). Amyloid vaccine development involves the combination of all or part of the Ab fragment with an adjuvant, to induce a specific antibody response against Ab that results in opsonization of the peptide and phagocytosis by brain microglia. Several immunization strategies have yielded dramatic results in transgenic mice with AD-type amyloid deposition in brain, creating tremendous excitement in the field and in the media. However, the first phase II study of an amyloid vaccine was terminated early, because some subjects developed encephalitis. Finally, a variety of approaches to protecting neurons against amyloid toxicity have been pursued (Grundman et al 1998). Among these are antioxidant therapies, such as vitamin E (Sano et al 1997) and Ginkgo biloba (Le Bars et al 1997), hormonal treatments such as oestrogen (Mulnard et al 2000), and interventions aimed at delivery of neurotrophins to cholinergic neurons (Gozes and Brenneman 2000). Antiinflammatory drug therapy fits into this category. INFLAMMATION AND ALZHEIMER’S DISEASE Inflammatory mechanisms are active in the AD brain, and may contribute to neuronal damage (Aisen 1997; Akiyama et al 2000). Neuritic plaques are surrounded by activated microglial cells, which are mononuclear phagocytes with the capacity to release and respond to inflammatory mediators and to act as antigenpresenting cells (Giulian 1999). Among the inflammatory cytokines that are upregulated in AD brain and that may mediate both microglial activation and inflammatory neurotoxicity are
488
NERVOUS SYSTEM
interleukins 1 and 6 and tumour necrosis factor a. It has also been demonstrated in several systems that microglial cells stimulated by Ab or inflammatory mediators may release potent neurotoxins (Giulian et al 1994; London et al 1996). Complement proteins and regulatory proteins are upregulated in AD brain (McGeer and McGeer 1992). Interestingly, the complement cascade can be activated directly by interaction between C1q and the Ab peptide (Tacnet-Delorme et al 2001). Anaphylatoxin generation may contribute to microglial accumulation and activation, but also may augment amyloid neurotoxicity (Oda et al 1995). The membrane attack complex has been demonstrated along membranes, suggesting a role in neuronal cell lysis in AD (Webster et al 1997). While proof is still lacking, evidence continues to accumulate suggesting that inflammation in the AD brain is not merely a passive reaction to degenerating tissue, but rather contributes to loss of brain function. Inflammation may be a significant aspect of indirect amyloid neurotoxicity. ARE ANTIINFLAMMATORY DRUGS USEFUL IN THE THERAPY OF AD? As interest in the possible use of antiinflammatory drugs has grown during the past decade, a number of model systems have been used to test the efficacy of specific drugs in the suppression of mechanisms that may be relevant to AD. Most have been related to amyloid toxicity, in vitro or in vivo. As discussed above, it is generally believed that Ab plays a pivotal role in the pathophysiology of AD. Ab may be directly toxic to neuronal cells, but also contributes to destructive inflammatory activity in the brain (e.g. by stimulating cytokine release, generating free radicals, and activating the complement cascade) (Akiyama et al 2000). Can antiinflammatory treatment reduce amyloid toxicity and slow disease progression? ANTIINFLAMMATORY DRUGS FOR AD: EPIDEMIOLOGY Epidemiological support for the hypothesis that antiinflammatory drugs retard the development of AD has been accumulating for many years. Early studies suggested that the co-occurrence of rheumatoid arthritis and AD is unexpectedly rare, consistent with the idea that antiinflammatory drugs used to treat arthritis confer protection against AD (McGeer et al 1990). Studies in twin (Breitner et al 1994) and sibling (Breitner et al 1995) pairs indicate that use of steroid and non-steroidal antiinflammatory drugs (NSAIDs) reduces the risk of AD. A number of epidemiological surveys support the negative association between NSAID use and AD (McGeer et al 1996; Stewart et al 1997). A recently reported study provides particularly strong evidence of a protective effect of NSAIDs against AD (In’t Veld et al 2001). The investigators utilized the Rotterdam Study cohort, consisting of ageing individuals in a suburb of Rotterdam that have been followed longitudinally, with periodic cognitive assessments, for an average of 7 years. To study the impact of NSAID use, the authors analysed electronic pharmacy records, eliminating the important issue of recall bias that has clouded other studies of this type. The results indicate that chronic use (2 years or longer) of even low-dose NSAIDs markedly reduces the risk of AD. MECHANISMS OF NSAID ACTIVITY IN AD There is controversy regarding the mechanisms of this apparent beneficial effect of NSAIDs. Several considerations suggest that
the theory that NSAID use suppresses destructive inflammation in the AD brain may be simplistic. NSAIDs are not known to be effective in treating brain inflammation in other diseases; particularly since the epidemiological evidence supports the efficacy of low-dose NSAIDs, it seems unlikely that the benefit is mediated by suppression of inflammatory mechanisms such as complement activation. Further, the failure of glucocorticoids (Aisen et al 2000) and the antimalarial antiinflammatory drug hydroxychloroquine (Van Gool et al 2001) in clinical AD trials suggests that NSAIDs in particular, rather than antiinflammatory drugs in general, may be useful treatments. A number of studies indicate that NSAIDs may protect neurons against toxic mechanisms involved in AD. Indomethacin confers some protection to PC12 cells against amyloid peptide toxicity (Fagarasan and Aisen 1996). In a study that modelled microglial neurotoxicity, NSAIDs (but not steroids) were effective in blocking the toxicity of supernatant media from stimulated monocytic cells or postmortem human microglial cells against human neuroblastoma cell death (Klegeris et al 1999). Peripheral administration of NSAIDs also protects against inflammatory neurodegeneration in brain (Hauss-Wegrzyniak et al 1999; Scali et al 2000), and studies in humans suggest that NSAID use reduces brain microglial activity (Mackenzie and Munoz 1998). In vivo, cyclooxygenase (COX) inhibitors may reduce inflammatory neurotoxicity. A COX-2 inhibitor reduces inflammatory cholinergic damage in a rat model (Willard et al 2000). Indomethacin inhibits the toxicity of amyloid-stimulated mononuclear cells against organotypic brain cultures (Dzenko et al 1997). The degree to which NSAID neuroprotection is mediated by cyclooxygenase inhibition remains a matter of considerable debate. In favour of this mechanism, there is evidence that COX-2 upregulation contributes to neuronal stress. COX-2 elevation accompanies apoptotic cell death in vitro and in vivo (Ho et al 1998). Exposure to amyloid peptide upregulates COX-2 in neuroblastoma cell cultures (Pasinetti and Aisen 1998). Overexpression of COX-2 in mouse brain potentiates excitotoxicity (Kelley et al 1999) and causes neuronal apoptosis and cognitive deficits (Andreasson et al 2001), suggesting a harmful effect of the enzyme; however, it has also been suggested that COX-2 may serve a protective effect in stressed neurons (McGeer 2000). There is some evidence to support a role of COX-derived prostaglandins in AD-type brain inflammation (Fiebich et al 2001; Hoozemans et al 2001; Lee et al 1999; Montine et al 1999). Upregulation of COX-2 in AD brain (Ho et al 1999; Kitamura et al 1999; Pasinetti and Aisen 1998), and its correlation with clinical disease progression (Ho et al 2001) supports the theory that COX2 represents a therapeutic target. Peroxisome proliferator-activated receptors (PPARs) are a nuclear receptor superfamily involved in lipid metabolism and insulin sensitivity, but also in inflammatory responses (Combs et al 2000; Heneka et al 2000; Kitamura et al 1999). The finding that NSAIDs activate PPARs (Lehmann et al 1997) suggests the possibility that this interaction, rather than COX inhibition, mediates NSAID antiinflammatory effects. Further, it has been suggested that binding of NSAIDs to PPARs may be the mechanism of their putative beneficial effect in AD. Thus, NSAIDs may inhibit the inflammatory stimulation of microglia and monocytes by Ab by activating PPARg rather than by inhibiting COX, and a similar protective effect is seen with other PPARg agonists, such as the antidiabetic drug thiazolidinedione and the natural ligand prostaglandin J2 (Combs et al 2000). Interestingly, PPARg agonists also inhibit the expression of COX2 (Combs et al 2000). Most recently, a NSAID effect apart from COX inhibition or PPAR activation has been suggested. The non-specific NSAID ibuprofen reduces brain inflammatory activity in a transgenic
NSAIDS IN TREATMENT OF ALZHEIMER’S DISEASE mouse model of AD-type amyloid plaque deposition caused by expression of a mutant human APP gene (Lim et al 2000). Perhaps more interestingly, ibuprofen reduces amyloid deposition in this model (Lim et al 2000; Weggen et al 2001). While this effect on amyloid could be related to COX inhibition or PPAR activation (Lim et al 2000), an alternative mechanism has been proposed. Specific NSAIDs, including ibuprofen, preferentially decrease the production of the most highly amyloidogenic fragment (Ab42) in cultured cells (Weggen et al 2001), suggesting that the beneficial effect may be a result of modulation of secretase activity on APP. CLINICAL TRIALS OF ANTIINFLAMMATORY DRUGS IN AD Interest in the possibility that treatment with NSAIDs would slow the rate of cognitive decline in individuals with AD increased with the publication of the results of the first controlled study, a small pilot investigation of indomethacin (Rogers et al 1993). In this trial, analysis of participants able to complete the 6 month study revealed that subjects treated with indomethacin for 6 months showed stable cognitive performance, while subjects in the placebo group declined. Although this finding encouraged further investigation of this treatment strategy, the high drop-out rate (42% in the active drug group) also raised concerns about the tolerability of long-term NSAID therapy in this frail population. A second small study investigated the combination of the NSAID diclofenac with the prostaglandin misoprostol, with the aim of demonstrating efficacy without a high rate of gastrointestinal toxicity (Scharf et al 1999). In this trial, the planned analysis considered all randomized subjects (i.e. intent to treat). While there was a trend toward stability in the active treatment group, the primary analysis was negative. Also discouraging was the high drop-out rate, 50% of those in the active drug group. A new study using the preferential COX-2 inhibitor nimesulide is more encouraging with regard to tolerability in this population (Aisen et al 2003). In a 12 week randomized controlled pilot tolerability study of nimesulide in AD, there was no adverse effect on measures of cognition and behaviour. Tolerability was further assessed in a long-term open label continuation phase. Nimesulide was tolerated by 92% of subjects for at least 24 weeks, and a subset of subjects remained on the drug for as long as 2 years. Evaluation of the efficacy of NSAIDs in the treatment of AD requires a large trial. The Alzheimer’s Disease Cooperative Study (ADCS), a consortium of academic medical centres funded by the National Institute on Aging, conducted a multicenter randomized controlled trial of NSAID therapy to slow cognitive decline in subjects with mild to moderate AD (Aisen et al 2003). A total of 350 subjects were assigned to three treatment groups: naproxen 200 mg b.i.d., rofecoxib 25 mg q.i.d. and placebo; the duration of treatment was 1 year. The primary outcome measure in this study was the change in score on the cognitive subscale of the Alzheimer’s Disease Assessment Scale (Rosen et al 1984). Selection of treatment regimens for this trial was based on several of the considerations discussed above. While the epidemiological studies (Breitner et al 1994, 1995; In’t Veld et al 2001; McGeer et al 1996; Stewart et al 1997), as well as transgenic mouse studies (Lim et al 2000), support the use of a non-selective NSAID for AD, the two pilot studies (Rogers et al 1993; Scharf et al 1999) indicate that tolerability is a major problem with full-dose non-selective NSAID therapy in this population. Naproxen was selected because it is a widely used long-acting non-selective NSAID, but a low dose was chosen in consideration of long-term tolerability. The selective COX-2 inhibitor rofecoxib was selected based on animal (Ho et al 1998; Kelley et al 1999; Tocco et al 1997) and human studies (Ho et al 1999, 2001; Pasinetti and Aisen 1998) indicating the involvement of COX-2 in AD pathology and,
489
because this drug is well tolerated, with good penetration into the brain. The results of the ADCS NSAID trial were disappointing. There was no slowing of cognitive decline with either naproxen or rofecoxib. Secondary analysis of the clinical dementia stage, behaviour, function and quality of life showed no consistent benefit with either treatment. The ADCS NSAID trial demonstrated that neither of these NSAID regimens is effective in slowing cognitive decline in subjects with AD. The epidemiological literature provides strongest support for a preventive effect of these drugs. While it is much less costly to investigate NSAIDs as therapy for established disease, these drugs may in fact be useful only as preventive agents, and only with long-term (i.e. multiple year) exposure. Thus, the recent analysis of prescription drug use in the Rotterdam Study cohort indicates that maximal protection against AD requires at least 2 years of NSAID therapy (In’t Veld et al 2001); even if the drugs have a beneficial effect on mild to moderate AD, 1 year of treatment may be too short a time to demonstrate significant benefit. It is also possible that, by the clinical onset of disease, it is too late for a beneficial effect of NSAID treatment. Inflammatory activity varies with disease stage (Luterman et al 2000) and there is evidence that cyclooxygenase activity contributes to very early stage disease (Ho et al 2001). NSAID therapy may be most appropriate before clinical manifestations of the disease have appeared. Based on these considerations, Breitner and colleagues have initiated a long-term AD prevention trial of NSAID therapy. Similar to the ADCS NSAID trial, this study includes two active treatment arms: a low dose of a non-selective NSAID (naproxen) and a selective COX-2 inhibitor (celecoxib). Enrollment criteria include 72 years of age and older and family history of dementia; the estimated incidence of AD in these subjects is 3% per year. The planned duration of treatment is 7 years, and the primary outcome is diagnosis of dementia.
SUMMARY AND CONCLUSION Considering the scope of the problem, any intervention that reduces the incidence or slows the progression of AD may yield tremendous benefits. A large amount of evidence now supports the possibility that NSAIDs confer such benefits. Numerous epidemiological studies demonstrate a negative association between NSAID use and AD. Several plausible mechanisms that may mediate NSAID neuroprotection have been supported by cell culture and transgenic mouse studies. However, proof that NSAIDs are beneficial can only come from large, randomized, controlled, prospective clinical trials. The first major trial of NSAID therapy for established AD was recently completed with disappointing results. A larger trial of NSAIDs for the primary prevention of AD is also in progress, but it will take several more years to complete. Each of these trials is likely to yield definitive findings on the efficacy and tolerability of the treatment regimens. An obvious and important limitation of these studies is that the results may not be generalizable beyond their specific regimens and study populations. Issues of patient selection, drug selection, dose and duration of treatment may be crucial in determining the success of this strategy for the therapy of AD.
REFERENCES Aisen PS (1997) Inflammation and Alzheimer’s disease: mechanisms and therapeutic strategies. Gerontology, 43, 143–149.
490
NERVOUS SYSTEM
Aisen PS (2000) Antiinflammatory therapy for Alzheimer’s disease: implications of the prednisone trial. Acta Neurol Scand Suppl, 176, 85–89. Aisen PS and Davis KL (1997) The search for disease-modifying treatment for Alzheimer’s disease. Neurology, 48(suppl 6), S35–S41. Aisen PS, Marin DB and Davis KL (1999) Alzheimer’s Disease: Questions and Answers. Coral Springs: Merit Publishing International. Aisen PS, Davis KL, Berg JD et al (2000) A randomized controlled trial of prednisone in Alzheimer’s disease. Alzheimer’s Disease Cooperative Study. Neurology, 54, 588–593. Aisen PS, Schafer K, Grundman M et al (2003) Effects of rofecoxib or naproxen vs. placebo on Alzheimer disease progression: a randomized controlled trial. J Am Med Assoc 289(21), 2819–2826. Aisen PS, Schmeidler J and Pasinetti GM (2002) Randomized pilot study of nimesulide treatment in Alzheimer’s disease. Neurology, 54, 588–593. Akiyama H, Barger S, Barnum S et al (2000) Inflammation and Alzheimer’s disease. Neurobiol Aging, 21, 383–421. Andreasson KI, Savonenko A, Vidensky S et al (2001) Age-dependent cognitive deficits and neuronal apoptosis in cyclooxygenase-2 transgenic mice. J Neurosci, 21, 8198–8209. Breitner JC, Gau BA, Welsh KA et al (1994) Inverse association of antiinflammatory treatments and Alzheimer’s disease: initial results of a co-twin control study. Neurology, 44, 227–232. Breitner JCS, Welsh KA, Helms MJ et al (1995) Delayed onset of Alzheimer’s disease with non-steroidal antiinflammatory and histamine H2 blocking drugs. Neurobiol Aging, 16, 523–530. Combs CK, Johnson DE, Karlo JC et al (2000) Inflammatory mechanisms in Alzheimer’s disease: inhibition of b-amyloid-stimulated proinflammatory responses and neurotoxicity by PPARg agonists. J Neurosci, 20, 558–567. Corey-Bloom J, Anand R and Veach J (1998) A randomized trial evaluating the efficacy and safety of ENA 713 (rivastigmine tartrate), a new acetylcholinesterase inhibitor, in patients with mild to moderately severe Alzheimer’s disease for the ENA 713 B352 Study Group. Int J Geriatr Psychopharmacol, 1, 55–65. Davis KL, Thal LJ, Gamzu E et al (1992) Tacrine in patients with Alzheimer’s disease: a double-blind, placebo-controlled multicenter study (in preparation). Dzenko KA, Weltzien RB and Pachter JS (1997) Suppression of Abinduced monocyte neurotoxicity by antiinflammatory compounds. J Neuroimmunol, 80, 6–12. Fagarasan MO and Aisen PS (1996) IL-1 and antiinflammatory drugs modulate Ab cytotoxicity in PC12 cells. Brain Res, 723, 231–234. Fiebich BL, Schleicher S, Spleiss O et al (2001) Mechanisms of prostaglandin E2-induced interleukin-6 release in astrocytes: possible involvement of EP4-like receptors, p38 mitogen-activated protein kinase and protein kinase C. J Neurochem, 79, 950–958. Giacobini E (2001) Do cholinesterase inhibitors have disease-modifying effects in Alzheimer’s disease? CNS Drugs, 15, 85–91. Giulian D (1999) Microglia and the immune pathology of Alzheimer disease. Am J Hum Genet, 65, 13–18. Giulian D, Li J, Leara B and Keenen C (1994) Phagocytic microglia release cytokines and cytotoxins that regulate the survival of astrocytes and neurons in culture. Neurochem Int, 25, 227–233. Gozes I and Brenneman DE (2000) A new concept in the pharmacology of neuroprotection. J Mol Neurosci, 14, 61–68. Grundman M, Corey-Bloom J and Thal LJ (1998) Perspectives in clinical Alzheimer’s disease research and the development of antidementia drugs. J Neural Transm, (Suppl) 53, 255–275. Grutzendler J and Morris JC (2001) Cholinesterase inhibitors for Alzheimer’s disease. Drugs, 61, 41–52. Hama E, Shirotani K, Masumoto H et al (2001) Clearance of extracellular and cell-associated amyloid-b peptide through viral expression of neprilysin in primary neurons. J Biochem (Tokyo), 130, 721–726. Hauss-Wegrzyniak B, Willard LB, Del SP et al (1999) Peripheral administration of novel antiinflammatories can attenuate the effects of chronic inflammation within the CNS. Brain Res, 815, 36–43. Heneka MT, Klockgether T and Feinstein DL (2000) Peroxisome proliferator-activated receptor-g ligands reduce neuronal inducible nitric oxide synthase expression and cell death in vivo. J Neurosci, 20, 6862–6867. Ho L, Osaka H, Aisen PS and Pasinetti GM (1998) Induction of cyclooxygenase (COX)-2 but not COX-1 gene expression in apoptotic cell death. J Neuroimmunol, 89, 142–149.
Ho L, Pieroni C, Winger D et al (1999) Regional distribution of cyclooxygenase-2 in the hippocampal formation in Alzheimer’s disease. J Neurosci Res, 57, 295–303. Ho L, Purohit D, Haroutunian V et al (2001) Neuronal cyclooxygenase 2 expression in the hippocampal formation as a function of the clinical progression of Alzheimer disease. Arch Neurol, 58, 487–492. Hoozemans JJ, Veerhuis R, Janssen I et al (2001) Interleukin-1b induced cyclooxygenase-2 expression and prostaglandin E2 secretion by human neuroblastoma cells: implications for Alzheimer’s disease. Exp Gerontol, 36, 559–570. In’t Veld BA, Ruitenberg A, Hofman A et al (2001) Non-steroidal antiinflammatory drugs and the risk of Alzheimer’s disease. N Engl J Med, 345, 1515–1521. Iwata N, Tsubuki S, Takaki Y et al (2000) Identification of the major Ab142-degrading catabolic pathway in brain parenchyma: suppression leads to biochemical and pathological deposition. Nature Med, 6, 143– 150. Jack C, Berezovska O, Wolfe MS and Hyman BT (2001) Effect of PS1 deficiency and an APPg-secretase inhibitor on Notch1 signaling in primary mammalian neurons. Brain Res Mol Brain Res, 87, 166–174. Kelley KA, Ho L, Winger D et al (1999) Potentiation of excitotoxicity in transgenic mice overexpressing neuronal cyclooxygenase-2. Am J Pathol, 155, 995–1004. Kitamura Y, Shimohama S, Koike H et al (1999) Increased expression of cyclooxygenases and peroxisome proliferator-activated receptor-g in Alzheimer’s disease brains. Biochem Biophys Res Commun, 254, 582– 586. Klegeris A, Walker DG and McGeer PL (1999) Toxicity of human THP-1 monocytic cells towards neuron-like cells is reduced by non-steroidal antiinflammatory drugs (NSAIDs). Neuropharmacology, 38, 1017– 1025. Le Bars PL, Katz MM, Berman N et al (1997) A placebo-controlled, double-blind, randomized trial of an extract of Ginkgo biloba for dementia. North American EGb Study Group. J Am Med Assoc, 278, 1327–1332. Lee RK, Knapp S and Wurtman RJ (1999) Prostaglandin E2 stimulates amyloid precursor protein gene expression: inhibition by immunosuppressants. J Neurosci, 19, 940–947. Lehmann JM, Lenhard JM, Oliver BB et al (1997) Peroxisome proliferator-activated receptors a and g are activated by indomethacin and other non-steroidal antiinflammatory drugs. J Biol Chem, 272, 3406–3410. Lim GP, Yang F, Chu T et al (2000) Ibuprofen suppresses plaque pathology and inflammation in a mouse model for Alzheimer’s disease. J Neurosci, 20, 5709–5714. London JA, Biegel D and Pachter JS (1996) Neurocytopathic effects of bamyloid-stimulated monocytes: a potential mechanism for central nervous system damage in Alzheimer disease. Proc Natl Acad Sci USA, 93, 4147–4152. Luterman JD, Haroutunian V, Yemul S et al (2000) Cytokine gene expression as a function of the clinical progression of Alzheimer disease dementia. Arch Neurol, 57, 1153–1160. Mackenzie IRA and Munoz DG (1998) Nonsteroidal antiinflammatory drug use and Alzheimer-type pathology in aging. Neurology, 50, 986– 990. McGeer PL (2000) Cyclooxygenase-2 inhibitors: rationale and therapeutic potential for Alzheimer’s disease. Drugs Aging, 17, 1–11. McGeer PL and McGeer EG (1992) Complement proteins and complement inhibitors in Alzheimer’s disease. Res Immunol, 143, 621–624. McGeer PL, McGeer E, Rogers J and Sibley J (1990) Antiinflammatory drugs and Alzheimer disease. Lancet, 335, 1037. McGeer PL, Schulzer M and McGeer EG (1996) Arthritis and antiinflammatory agents as possible protective factors for Alzheimer’s disease: a review of 17 epidemiologic studies. Neurology, 47, 425–432. Montine TJ, Sidell KR, Crews BC et al (1999) Elevated CSF prostaglandin E2 levels in patients with probable AD. Neurology, 53, 1495–1498. Morgan D, Diamond DM, Gottschall PE et al (2000) A b peptide vaccination prevents memory loss in an animal model of Alzheimer’s disease. Nature, 408, 982–985. Mulnard RA, Cotman CW, Kawas C et al (2000) Estrogen replacement therapy for treatment of mild to moderate Alzheimer disease: a
NSAIDS IN TREATMENT OF ALZHEIMER’S DISEASE randomized controlled trial. Alzheimer’s Disease Cooperative Study. J Am Med Assoc, 283, 1007–1015. Oda T, Lehrer-Graiwer J, Finch CE and Pasinetti GM (1995) Complement and b-amyloid neurotoxicity in vitro: a model for Alzheimer disease. Alzheimer’s Res, 1, 29–34. Pasinetti GM and Aisen PS (1998) Cyclooxygenase-2 expression is increased in frontal cortex of Alzheimer’s disease brain. Neuroscience, 87, 319–324. Raskind MA, Peskind ER, Wessel T and Yuan W (2000) Galantamine in AD: a 6-month randomized, placebo-controlled trial with a 6-month extension. The Galantamine USA-1 Study Group. Neurology, 54, 2261–2268. Roberds SL, Anderson J, Basi G et al (2001) BACE knockout mice are healthy despite lacking the primary b-secretase activity in brain: implications for Alzheimer’s disease therapeutics. Hum Mol Genet, 10, 1317–1324. Rogers J, Kirby LC, Hempelman SR et al (1993) Clinical trial of indomethacin in Alzheimer’s disease. Neurology, 43, 1609–1611. Rogers SL, Farlow MR, Doody RS et al (1998) A 24-week, double-blind, placebo-controlled trial of donepezil in patients with Alzheimer’s disease. Neurology, 50, 136–145. Rosen WG, Mohs RC and Davis KL (1984) A new rating scale for Alzheimer’s disease. Am J Psychiat, 141, 1356–1364. Sano M, Ernesto C, Thomas RG et al (1997) A controlled trial of selegiline, a-tocopherol, or both as treatment for Alzheimer’s disease. N Engl J Med, 336, 1216–1222. Scali C, Prosperi C, Vannucchi MG et al (2000) Brain inflammatory reaction in an animal model of neuronal degeneration and its modulation by an antiinflammatory drug: implication in Alzheimer’s disease. Eur J Neurosci, 12, 1900–1912.
491
Scharf S, Mander A, Ugoni A et al (1999) A double-blind, placebocontrolled trial of diclofenac/misoprostol in Alzheimer’s disease. Neurology, 53, 197–201. Schenk D, Barbour R, Dunn W et al (1999) Immunization with amyloid-b attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature, 400, 173–177. Schneider LS (2001) Treatment of Alzheimer’s disease with cholinesterase inhibitors. Clin Geriatr Med, 17, 337–358. Stewart WF, Kawas C, Corrada M and Metter EJ (1997) Risk of Alzheimer’s disease and duration of NSAID use. Neurology, 48, 626– 632. Tacnet-Delorme P, Chevallier S and Arlaud GJ (2001) b-Amyloid fibrils activate the C1 complex of complement under physiological conditions: evidence for a binding site for Ab on the C1q globular regions. J Immunol, 167, 6374–6381. Tocco G, Freire MJ, Schreiber SS et al (1997) Maturational regulation and regional induction of cyclooxygenase-2 in rat brain: implications for Alzheimer’s disease. Exp Neurol, 144, 339–349. Van Gool WA, Weinstein HC, Scheltens PK and Walstra GJ (2001) Effect of hydroxychloroquine on progression of dementia in early Alzheimer’s disease: an 18-month randomised, double-blind, placebocontrolled study. Lancet, 358, 455–460. Webster S, Lue LF, Brachova L et al (1997) Molecular and cellular characterization of the membrane attack complex, C5b-9, in Alzheimer’s disease. Neurobiol Aging, 18, 415–421. Weggen S, Eriksen JL, Das P et al (2001) A subset of NSAIDs lower amyloidogenic Ab42 independently of cyclooxygenase activity. Nature, 414, 212–216. Willard LB, Hauss-Wegrzyniak B, Danysz W and Wenk GL (2000) The cytotoxicity of chronic neuroinflammation upon basal forebrain cholinergic neurons of rats can be attenuated by glutamatergic antagonism or cyclooxygenase-2 inhibition. Exp Brain Res, 134, 58–65.
46 Prostaglandins and Eicosanoids in Mental Illness A.I.M. Glen and B.M. Ross Ness Foundation, UHI Millennium Institute, Inverness, UK
The investigation of eicosanoids in mental illness could only happen with the development of new techniques, and some of these have already been described. Clinical observations which were subsequently linked to abnormalities in the eicosanoid system were, however, made much earlier and have been described by Horrobin (1977). Nissen and Spencer (1936) showed that patients with rheumatoid arthritis rarely seemed to suffer from schizophrenia. This observation has been made on a number of occasions over the years since then. Wagner-Jauregg, in studies which led to a Nobel Prize in 1927, showed that patients with schizophrenia improved when their psychosis was treated by giving them malaria. Horrobin subsequently hypothesized that the effect of malaria was to increase production of prostaglandins and proposed that these two observations, together with the fact that sufferers from schizophrenia do not experience pain as readily as normal individuals, suggested that there was an abnormality of the eicosanoid system in schizophrenia.
EARLY CLINICAL STUDIES In the absence of access to the brain for turnover studies of phospholipids until comparatively recently, the platelet has been studied extensively in schizophrenia. Platelet function has similarities to synaptic activity and may provide useful information on aspects of fatty acid metabolism linked to cell signalling. In an early study, Abdullah and Hamada (1975) found reduced formation of prostaglandin E1 following stimulation with ADP in subjects with schizophrenia. This reduction in platelet prostaglandin E1 may be in keeping with the finding of elevation of prostaglandin E2 metabolites in plasma of twins with schizophrenia (Mathe et al 1986; replicated by Idaka et al 1987), since there is a reciprocal relationship between the one series and the two series of prostaglandins. Although there were limited studies at this time of prostaglandins and their metabolites, a number of studies of the fatty acid precursors were being undertaken. Horrobin et al (1989) showed low levels of the highly unsaturated fatty acids in plasma of schizophrenic subjects and this was confirmed in the first study of fatty acids in post mortem brain in schizophrenia (Horrobin et al 1991). This provided only limited evidence of eicosanoid involvement in schizophrenia, but the parallel laboratory observations of fatty acid and eicosanoid interaction with receptor and post-synaptic function were now offering a hypothesis for the underlying problem of increased dopaminergic function in schizophrenia. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
MODULATION OF BEHAVIOUR AND NEUROTRANSMISSION BY EICOSANOIDS General Features The purpose of this section is not to provide an exhaustive list of the behavioural and neurochemical effects of eicosanoids, but to describe some illustrative examples of their interaction with several important neurotransmitter systems, with the aim of understanding what role they may play in mental and neurological illness. Polyunsaturated fatty acids and their oxygenated metabolites encompass a large number of chemicals with neuromodulatory effects. In other words, they do not function as classical neurotransmitters released in response to action potentials, but rather as modulators of neuronal activity, changing the likelihood of an action potential occurring. Operating through G-protein-linked cell-surface receptors, they act as regulators of second messengers such as cAMP, inositol phosphates and diacylglycerol, in turn affecting a variety of cellular processes, including calcium mobilization, protein kinase activity and protease activity (Wise 1997). Due to the widespread availability of cyclooxygenase inhibitors, much more is known regarding the central nervous system functions of cyclooxygenase products than of products of lipoxygenase. Several studies have reported a significant level of COX-1 and COX-2 expression within the brain (Kawasaki et al 1993; Tocco et al 1997), findings in keeping with in vitro experiments demonstrating the capacity of brain tissue to synthesize a variety of cyclooxygenase-derived compounds (Ellis et al 1981). The brain regional distribution of cyclooxygenase is, however, not uniform, suggesting that cyclooxygenase may be preferentially associated with specific neurotransmitters and/or neuronal pathways. Moreover, the large number of messenger molecules and receptors within the eicosanoid signalling system may allow a high degree of flexibility in their role as neuromodulators. Thus, although COX has a widespread brain distribution, the specific effect of cyclooxygenase inhibition is dependent upon which eicosanoid receptors are present in each neuronal locus. In this way, it has been suggested (Smith et al 1991) that cyclooxygenase inhibitors could have dissimilar effects upon the functioning of the same neurotransmitter system in different brain regions. There has been sparse study, however, of eicosanoids, their receptors and signalling systems in the brain, and no firm evidence exists to support such a hypothesis. However, it is notable that, although in situ autoradiography of radiolabelled PGE2 indicates a widespread distribution of PGE2 receptors in rodent brain (Matsumura et al 1992), the individual PGE2 receptor subtypes—EP1, EP2, EP3 and EP4—possess
494
NERVOUS SYSTEM
differing and highly specific neuroanatomical distributions (Zhang and Rivest 1999; Batshake et al 1995; Sugimoto et al 1994), indicating the possibility of regional heterogeneity in eicosanoid function. The behavioural effects of cyclooxygenase products have been mainly examined using cyclooxygenase inhibitors. Assessing the effects of the oxygenation products directly is extremely difficult, given both their very short biological half-lives and their failure to readily cross the blood–brain barrier. Although the continuous infusion of prostaglandins into the ventricles of the brain has been used as a means of changing their brain levels for long enough for a behavioural measure to be made (Saito et al 1987), use of inhibitors of the synthetic enzymes, cyclooxygenase and lipoxygenase, are more common, given their advantages of peripheral administration, stability and ability to produce long-lasting effects. For example, cyclooxygenase inhibition induces a small degree of hypothermia in rats, in keeping with their utility in control of fever, suggesting that a major site of action is in the temperature control systems of the brain (the hypothalamic– pituitary axis) (Linthorst and Reul 1998). Prostaglandins, specifically PGD2 and possibly PGE2, also play a key role in sleep behaviour (Mizoguchi et al 2001; Yoshida et al 2000). Cyclooxygenase inhibitors also reduce passive avoidance learning in chicks, indicating a role for cyclooxygenase in learning and memory (Holscher 1999). A study in elderly human subjects, however, suggested that cyclooxygenase inhibition may increase short-term memory performance (Bruce-Jones et al 1994). Modulation of Dopaminergic Neurotransmission Although the neurophysiological basis of such results is presently unclear, a picture is emerging of eicosanoids as important neuromodulators of most, and perhaps all, neurotransmitter systems. A large body of data suggests that eicosanoids play an important role in modulating the function of the neurotransmitter dopamine. Interest in such a possibility was encouraged by anecdotal clinical reports of COX inhibitors, particularly indomethacin, causing transient drug-induced psychosis in healthy subjects (Hoppmann et al 1991). Due to the putative involvement of the dopamine system in psychosis associated with both schizophrenia, a disorder thought to be associated with hyperdopaminergic activity, and the chronic use of amphetamine, a drug causing dopamine release, it was hypothesized that cyclooxygenase inhibitors act as dopaminergic agents, i.e. they stimulate pathways in the brain using dopamine as the neurotransmitter (Seeman and Niznik 1990). Dopamine-containing neurons can be approximately grouped into two major systems, the nigrastriatal that is involved in control of motor behaviour, and the mesocorticolimbic that is critical for the rewarding aspects of dopaminergics, such as amphetamine and cocaine. Within these systems, five dopamine receptors are utilized, characterized as D1like (D1, D3 and D5) and D2-like (D2 and D4). An effect of eicosanoids upon the nigrastriatal motor system is supported by experimental animal-based investigations. For example, the cyclooxygenase inhibitor indomethacin can: antagonize haloperidol, raclopride (both dopamine D2 receptor antagonists) and SCH23390 (a dopamine D1 receptor antagonist); induce catalepsy (reduced movement) (Ono et al 1992; Ross et al 2002); potentiate the ability of amphetamine to reduce the rate of bar-pressing in animals trained to press for food pellets (operant suppression) (Nielsen and Sparber 1984); and cause increased circling behaviour in rats lesioned in one hemisphere with 6-hydroxydopamine (Schwarz et al 1982). Similarly, injection of prostaglandins into the brain produces catalepsy (Ono et al 1986), as well as stimulating the activity of nigrastriatal dopamine neurons (Chu et al 2001).
For the mesocorticolimbic system, although cyclooxygenase inhibitors have no effect upon the firing rate of mesolimbic dopamine neurons when given alone, they do potentiate the ability of the opiate morphine to stimulate these neurons, suggesting an involvement of eicosanoids in the rewarding and addictive effects of opiates (Melis et al 2000; Fattore et al 2000). Furthermore, a role of eicosanoids in the mechanism in the direct dopaminergic class of drugs, which includes cocaine and amphetamine, is suggested by the ability of both cyclooxygenase and PLA2 inhibitors to reduce behavioural sensitization to these compounds (the phenomenon in which repeated administration of drug produces, rather counterintuitively, an increasing effect) (Reid et al 1996, 2002). Although recent behavioural evidence suggests that such observations may not be due entirely to stimulation of dopaminergic transmission (Ross et al 2002), in vitro studies are strongly supportive of an eicosanoid/dopamine link. Cell culture studies utilizing transfected dopamine receptors have suggested that dopamine D2 and D4 receptors can stimulate PLA2 activity (Chio et al 1994; Vial and Piomelli 1995; McCallister et al 1993), in particular the cPLA2 form of the enzyme. Modulation of arachidonic acid release may also underlie the important, but rather puzzling, ability of D1 receptors to potentiate the effect of dopamine D2 receptors at electrophysiological and behavioural levels, even though both receptors have opposing effects upon the activity of adenylyl cyclase and phospholipase C (Piomelli et al 1991). Futhermore, dopamine receptor agonists increase the incorporation of arachidonic acid into brain membranes, a parameter influenced by PLA2 action (Hayakawa et al 2001). Such findings suggest that enhanced dopamine receptor occupancy may lead to elevated PLA2 activity. Moreover, overstimulation of PLA2 activity by the dopaminergic agent cocaine appears to result in a compensatory downregulation via a negative-feedback mechanism in both experimental animals (Ross and Turenne 2002) and chronic human users of the drug (Ross et al 1996), while reduced dopaminergic input into the striatum in Parkinson’s disease leads to elevated PLA2 activity (Ross et al 2001). Although the experiments utilizing COX inhibitors outlined above suggest a role for oxygenated fatty acid products affecting dopaminergic transmission, intramembrane signalling by free fatty acid released by PLA2 is also indicated. For example, arachidonic acid has a direct effect upon the dopamine transporter, the membrane protein responsible for sequestering dopamine from the synapse and the target for the drug cocaine (Ingram and Amara 2000). Glutamate and GABA Glutamate and GABA represent the primary stimulatory and inhibitory neurotransmitters, respectively, in the cerebral cortex, both of which are central to the functioning of higher thought processes and memory. Glutamate receptors are involved in pathological conditions such as epilepsy, while overstimulation of the glutamate system, e.g. in stroke, can cause neuronal death. GABA receptors are the site of action of the two major classes of sedatives—benzodiazepine drugs such as valium, and barbiturates. Stimulation of one type of glutamate receptor, the NMDA receptor, leads to the release of free fatty acids and eicosanoids (Lazerewicz et al 1990; Mollace et al 1995), whereas brain c-fos expression and NMDA-induced allodynia are both prostaglandindependent processes (Lerea et al 1997; Dolan and Nolan 1999). Furthermore, administration of NMDA receptor antagonists to rats results in increased brain COX expression (Hashimoto et al 1997). An interaction with NMDA receptors may also underlie the psychotic effects of COX inhibitors described above, since drugs which inhibit this receptor, e.g. the hallucinogens
PROSTAGLANDINS AND EICOSANOIDS phenicyclidine (angel dust) and ketamine (Special-K), both exacerbate schizophrenia and produce schizophrenia-like symptoms in healthy individuals (Balster 1994). Arachidonic acid and PLA2 are also thought to play a key role in the ability of glutamate NMDA receptors to produce long-term potentiation, a phenomenon thought to be involved in learning and memory (Nishizaki et al 1999). The related phenomenon, long-term depression, which also requires NMDA receptor activation, also appears dependent upon activation of the lipoxygenase pathway (Normandin et al 1996). GABAergic transmission is also affected by eicosanoids. For example, GABA agonist-stimulated chloride uptake by rat synaptosomes is inhibited by prostaglandins, thromboxanes and arachidonic acid (Schwartz and Yu 1992), while arachidonic acid also enhances the release of GABA from neurons (Cheramy et al 1996). Furthermore, GABA receptor antagonist-induced tactile pain response is absent in PGD2-deficient mice, again suggesting a modulatory action of cyclooxygenase and PLA2 products upon GABAergic transmission (Eguchi et al 1999). Cyclooxygenase inhibitors also enhance the inhibition of GABAergic transmission by opiates in the brain’s nociceptive (pain) circuits (Vaughan et al 1997). Cyclooxygenase inhibitors achieve this, not via inhibition of prostaglandin synthesis per se, but by ‘shunting’ arachidonic acid into the lipoxygenase-dependent pathway. Thus, the site of action of the well-known antiinflammatory and pain-reducing uses of cyclooxygenase inhibitors are both at the injury itself and in the nociceptive circuits of the brain. These examples show the variety of ways in which eicosanoids modulate and coordinate brain function and behaviour. The variety of the molecules involved, as well as the receptors and signalling mechanisms mediating these effects, also make eicosanoids an attractive target for new drug design. As more specific inhibitors and stimulators of their function become available, the medical applications of eicosanoid-targeted drugs will likely increase, as will our knowledge of how eicosanoids function within the nervous system. In addition, it is likely that the abnormalities in eicosanoid function in psychiatric illness, reviewed in the following section, will soon be more amenable to treatment, opening up an array of new therapies. EICOSANOID FUNCTION IN PSYCHIATRIC DISEASE In Schizophrenia Much evidence has now accumulated for dysfunction of eicosanoid signalling in schizophrenia, a mental illness often characterized by misintepretation of reality resulting from hallucination and delusional ideas. Schizophrenia has a clear genetic component, although in identical twins there is only a 40% chance of both being affected, indicating there is also a major environmental component in the aetiology of the illness. Because drugs that reduce hallucination and delusional ideation have a prominent dopaminergic blocking effect, it was assumed for many years that the primary underlying abnormality was increased dopaminergic function. However, the evidence for this is scanty, since new scanning techniques (positron emission tomography) do not indicate any increase in dopaminergic receptors in untreated schizophrenic patients. An alternative hypothesis, now gaining ground, pinpoints the metabolism of key fatty acids, especially arachidonic acid, in phospholipids. PLA2 releases the arachidonic acid cell signaller from the sn-2 position in phospholipid complexes in cell membranes. As well as providing immediately available signaller, released arachidonic acid enters a cascade to provide prostaglandins and cytokines. The expression and availability of PLA2 is thus a key factor in the
495
investigation of mechanisms responsible for the reduced availability of arachidonic acid which has been described in neurodevelopmental disorders, especially in schizophrenia. Gattaz et al. (1995) first described increased activity of PLA2 in schizophrenia and later reviewed the evidence for this abnormality (Gattaz et al 1995). Ross has also reviewed the evidence and confirmed in his own studies increased levels of PLA2 in the plasma of schizophrenic patients (Ross 1999). These reports have presented measures of PLA2 activity. Since there have also been reports of the variable expression of cPLA2 in schizophrenia (Macdonald et al 2000), it would also be important to measure the amount of the enzyme expressed. Preliminary studies of cytosolic calcium-dependent PLA2 expression in schizophrenia using an ELISA technique support the view that there is increased expression of the enzyme associated with increased activity (Ramchand et al 1999; Wei and Hemmings 2000a, 2000b). MEASURING SIGNALLING FUNCTION IN THE CYCLOOXYGENASE PATHWAY—THE SKIN PATCH TEST The niacin patch test represents the first generation of functional tests of the cell signaller, arachidonic acid, with more sophisticated tests being likely to offer improvements in terms of accuracy and specificity. Researchers have concentrated on a simple model in the skin, whereby the effects of release of PGD2 activity derived from arachidonic acid is measured in terms of skin flushing. Horrobin first described reduced skin flushing in schizophrenia using oral medication. Since the original observation of significantly reduced skin flushing in schizophrenia using skin patch testing was reported (Ward et al 1998), there have been many replications. Studies carried out by Peet et al (2000), and by Berger et al (2002) in drug-free and drug-naive schizophrenic subjects show that the phenomenon is not caused by neuroleptics. Berger’s report of absence of skin flushing in unmedicated schizophrenic patients was part of a study of early onset schizophrenia in subjects suffering their first episode. About 80% of schizophrenic patients fail to flush with topical applications of methyl nicotinate (Ward and Glen 1999). Ross has shown that the phenomenon of reduced or absent flushing in schizophrenia, usually measured by clinical rating, can be confirmed using laser Doppler spectroscopy (Ross BM, personal communication). Two studies show familial linkage in skin flushing in schizophrenic patients (Waldo et al 1995; Waldo 1999). The niacin test suggests an abnormality in lipid signalling processes in the cyclooxygenase pathway in schizophrenia. Are these mirrored in the brain? Indeed, the blood–brain barrier appears to buffer the brain against changes in serum lipid levels more effectively than occurs in peripheral cells (Rappoport et al 2001). Recent developments in neuroimaging techniques are, however, allowing us to examine brain lipid metabolism in vivo for the first time. EICOSANOID-LINKED CELL SIGNALLING ABNORMALITIES IN PLATELETS Yao (1999) has reviewed the evidence that the platelet has similarities in its function to pre- and post-synaptic membranes and that it has similar calcium dynamics. His platelet studies of eicosanoid interaction in receptor-coupled intracellular signal generation in schizophrenia provided important new evidence in this field. Yao and his colleagues incorporated tritium-labelled arachidonic acid into resting platelets from schizophrenic patients on and off the dopamine-blocking drug haloperidol. The prelabelled platelets were then stimulated by thrombin to produce an
496
NERVOUS SYSTEM
eicosanoid cascade. The resting platelets incorporated over 85% of the labelled arachidonic acid into phospholipids, with only minute amounts found in free arachidonic acid, DAG and in the arachidonic acid metabolites in the cyclooxygenase and lipoxygenase pathways. When the platelets were activated with thrombin, increased tritium-labelled arachidonic acid was found in the eicosanoid pathways (Yao and Van Kammen 1966). Yao found that there was a decrease in the incorporation of the labelled arachidonic acid into phospholipids in the platelets of schizophrenic subjects in the drug-free condition compared to controls, and this held good for both haloperidol-treated and untreated subjects. The findings were similar to those of Demisch et al (1992), who also found abnormalities in major depressive disorder. Following treatment with haloperidol, Yao and his colleagues found an increase of 12-HETE in the activated platelets and suggested that one of the pharmacological effects of haloperidol might be to regulate the second messenger function via the arachidonic acid pathway. INVOLVEMENT OF THE IMMUNE SYSTEM IN MENTAL ILLNESS In the early studies of phenomenology in schizophrenia, there were indications of an abnormality in the immune system in that patients with schizophrenia seldom had rheumatoid arthritis. In a series of studies in Holland since 1996, Maes and his colleagues have provided convincing evidence of abnormalities in the immune system in both schizophrenia and depression (Maes et al 1996). Further discussion of this topic is not appropriate here, but the involvement of cytokines in immunity and mental illness is now well established. EVIDENCE FOR ABNORMALITIES OF EICOSANOID PRECURSORS IN BRAIN There is as yet no direct evidence for abnormalities of the eicosanoid system as far as the cyclooxygenase and lipoxygenase metabolites are concerned, although some studies have reported abnormalities in the systems in CSF (Horrobin et al 1991; Yao et al 2000). There is, however, substantial evidence for abnormalities in the eicosanoid precursors. This has been obtained using magnetic resonance spectroscopy and has been well reviewed by Williamson and Drost (1999) and Puri and Richardson (1999). EVIDENCE FROM EICOSANOID PRECURSOR TREATMENT RESPONSE IN OTHER PSYCHIATRIC ILLNESS Some 5% of the world population suffer from depression and of these 30% have depression resistant to treatment using current drugs. Of the world population, 1% suffer from schizophrenia. In both these illnesses mortality from suicide is increased. It is of great interest, therefore, that at the present time two double-blind trials of eicosopentaenoic acid (EPA) show large and statistically significant improvements in the treatment of depression (Nemets et al 2002; Peet et al 2002). The results of a further three trials still in progress are awaited with interest. A trial of EPA combined with docosahexaenoic acid showed a beneficial effect of high dose in bipolar depression (Stoll et al 1999). Five randomized placebo-controlled trials of EPA have been reported on schizophrenia. Four have shown significant improvements of 12–25% (Peet et al 2001; Peet and Horrobin 2002; Emsley et al 2002) but in one trial there was no significant difference between the placebo and the active drug (Fenton et al
2001). Trials of treatment with docosohexaenoic acid (DHA) in schizophrenia and depression have all proved negative, as have earlier studies with arachidonic acid-enriched or arachidonic precursor-enriched compounds. Treatment studies of neurodevelopmental disorders are currently being undertaken. Results so far show that attention deficit hyperactivity disorder (ADHD) is the most promising condition in terms of its response to EPA (Stevens and Burgess 1999; Richardson and Ross 2000).
MODE OF ACTION OF EPA The mode of action of EPA in depression and schizophrenia is not entirely clear. As has been described, both schizophrenia and depression are associated with abnormalities of fatty acids, especially arachidonic acid or its metabolites, in brain and peripheral cell membranes. It has been proposed that EPA acts by inhibiting PLA2, which is believed to be overactive in schizophrenia, and thus reducing loss of arachidonic acid by lipid peroxidation. It is of great interest in this context that lipid peroxidation in schizophrenia appears to be reduced during treatment with clozapine (Horrobin et al 1997) and that clozapine has been shown to increase the expression of apolipoprotein D in brain (Thomas et al 2001a, 2001b). Animal studies show that EPA has marked effects on neurotransmitter function in brain, although present at comparatively low levels. In depression, known from clinical studies to be associated with reduction in eicosanoids in peripheral membranes and abnormalities in fatty acid metabolites, EPA may act to increase eicosanoid production in the 3-series prostaglandins.
HUNTINGTON’S DISEASE AND EPA Huntington’s disease, a genetically determined illness causing movement disorder, has been found in preliminary studies to respond, both in the animal model and the human condition, to EPA. Following initial open studies with EPA (Vadaddi 1999), expression of the Huntington protein was shown to be reduced in the animal model (Lonergan et al 2002). A subsequent doubleblind trial showed significant reduction in symptoms over a 6 month period (Clifford et al 2002). In this brief review we have summarized the evidence from both laboratory studies and clinical trials for the involvement of eicosanoids in mental illness. Further double-blind clinical trials are now in progress, in both depression and schizophrenia, and the information we have on these is that the earlier results are being confirmed. This reinforces the view of the fundamental importance of this area of study.
REFERENCES Abdulla YH and Hamadah K (1975) Effect of ADP on PGE1 formation in blood platelets from patients with depression, mania and schizophrenia. Br J Psychiat, 127, 591–595. Balster RL (1994) The behavioural pharmacology of phencyclidine. In Psychopharmacology: The Third Generation of Progress. New York: Raven. Batshake B, Nilsson C and Sundelin J (1995) Molecular characterisation of the mouse prostanoid EP1 receptor gene. Eur J Biochem, 231, 809– 814. Berger GE, Yuen HP and McGorry PD (2002) The topical niacin flush test in early psychosis. Schizophr Res, 53, S38. Bruce-Jones PN, Crome P and Kalra L (1994) Indomethacin and cognitive function in healthy elderly volunteers. Br J Clin Pharmacol, 38, 45–51.
PROSTAGLANDINS AND EICOSANOIDS Cheramy A, Artaud F, Godeheu G et al (1996) Stimulatory effect of arachidonic acid on the release of GABA in matrix-enriched areas from the rat striatum. Brain Res, 742, 185–194. Chio CL, Drong RE, Riley DT et al (1994) D4 dopamine receptormediated signaling events determined in transfected Chinese hamster ovary cells. J Biol Chem, 269, 11813–11819. Chu YC, Tsou MY and Pan JT (2001) Prostaglandins play an important role in diurnal changes of tuberoinfundibular dopaminergic neuronal activity and prolactin secretion in ovariectomized, estrogen-treated rats. Brain Res Bull, 55, 87–93. Clifford JJ, Drago J, Natoli AL et al (2002) Essential fatty acids given from conception prevent topographies of motor deficit in a transgenic mouse model of Huntington’s disease. Neuroscience 109, 81–88. Demisch L, Gerbaldo H, Gebhart P et al (1987) Incorporation of 14Carachidonic acid into platelet phospholipids of untreated patients with schizophreniform or schizophrenic disorders. Psychiat Res, 22, 275– 282. Demisch L, Heinz K, Gerbaldo H and Kirsten R (1992) Increased concentrations of phosphatidylinositol and decreased esterification of arachidonic acid into phsopholipids in platelets from patients with schizoaffective disorders or atypic phasic psychoses. Prostagland Leukotrienes Essent Fatty Acids, 46, 47–52. Dolan S and Nolan AM (1999) N-methyl-D-aspartate induced mechanical allodynia is blocked by nitric oxide synthase and cyclooxygenade-2 inhibitors. NeuroReport, 10, 449–452. Eguchi N, Minami T, Shirafuji A et al (1999) Lack of tactile pain (allodynia) in lipocalin-type prostaglandin D synthase-deficient mice. Proc Natl Acad Sci USA, 1956, 726–730. Ellis EF, Wright KF, Wei EP and Kontos HA (1981) Cyclooxygenase products of arachidonic acid metabolism in cat cerebral cortex after experimental concusive brain injury. J Neurochem, 97, 892–896. Emsley RA, Myburgh CC, Osthuizen PP and van Rensburg SJ (2002) o-3 Fatty acids and schizophrenia: a randomized trial of ethyleicosapentaenoic acid (ethyl-EPA) vs. placebo as add-on treatment. Am J Psychiatry, 159, 1596–1598. Fattore L, Melis M, Diana M et al (2000) The cyclooxygenase inhibitor nimesulide induces conditioned place preference in rats. Eur J Pharmacol, 406, 75–77. Fenton WS, Dickerson F, Boronow J et al (2001) A placebo-controlled trial of o-3 fatty acid (ethyl eicosapentaenoic acid) supplementation for residual symptoms and cognitive impairment in schizophrenia. Am J Psychiatry, 158, 2071–2074. Gattaz WF, Schmitt A and Maras A (1995) Increased platelet phospholipase A2 activity in schizophrenia. Schizophr Res, 16, 1–6. Glen AI, Cooper SJ, Rybakowski J et al (1996) Membrane fatty acids, niacin flushing and clinical parameters. Prostagland Leukotriene Essent Fatty Acids, 55, 9–15. Hashimoto K, Minabe Y and Iyo M (1997) Expression of cyclooxygenase2 mRNA in gat retroplenial cortex following administration of phencylidine. Brain Res, 762, 259–263. Hayakawa T, Chang MC, Rapoport SI and Appel NM (2001) Selective dopamine receptor stimulation differentially affects [3H] arachidonic acid incorporation, a surrogate marker for phospholipase A2-mediated neurotransmitter signal transduction, in a rodent model of Parkinson’s disease. J Pharmacol Exp Ther, 296, 1074–1084. Holscher C (1999) Prostaglandins play a role in memory consolidation in the chick. Eur J Pharmacol, 294, 253–259. Hoppmann RA, Peden JG and Over SK (1991) Central nervous system side effects of non-steroidal antiinflammatory drugs. Arch Int Med, 151, 1309–1313. Horrobin DF (1977) Schizophrenia as a prostaglandin deficiency disease. Lancet, I, 936–937. Horrobin DF, Glen AIM and Cantrill RC (1997) Clozapine: elevation of membrane unsaturated lipid levels as a new mechanism of action. Schizophr Res, 24, 214. Horrobin DF, Manku MS, Hillman H and Glen AIM (1991) Fatty acid levels in the brains of schizophrenics and normal controls. Biol Psychiat, 30, 795–805. Horrobin DF, Manku MS, Morse-Fisher N et al (1989) Essential fatty acids in plasma phospholipids in schizophrenics. Biol Psychiat, 25, 562–568. Idaka E, Kaiya H and Horitsu H (1987) Determination of prostaglandin E2 in plasma of schizophrenicns and healthy controls by gas
497
chromatography–negative ion chemical ionization mass spectrometry and radioimminoassay. J Chromatogr, 4220, 373–378. Ingram SL and Amara SG (2000) Arachidonic acid stimulates a novel cocaine-sensitive cation conductance associated with the human dopamine transporter. J Neurosci, 20, 550–557. Kawasaki M, Yoshihara Y, Yamaji E and Watanabe (1993) Expression of prostaglandin endoperoxide synthase in rat brain. Mol Brain Res, 19, 39–46. Lazarewicz JW, Wrodleski JT and Costa E (1990) N-methyl-D-aspartatesensitive glutamate receptors influence calcium-mediated arachidonic acid release in primary cultures of cerebellar granule cells. J Neurochem, 55, 1875–1881. Lerea LS, Carlson NG, Simonato M et al (1997) Prostaglandin F2a is required for NMDA receptor-mediated induction of c-fos mRNA in dentate gyrus neurons. J Neurosci, 17, 117–124. Linthorst AC and Reul JM (1998) Brain neurotransmission during peripheral inflammation. Ann NY Acad Sci, 840, 139–152. Lonergan PE, Martin DS, Horrobin DF and Lynch MA (2002) Neuroprotective effect of eicosapentaenoic acid in hippocampus of rats exposed to g-irradiation. J Biol Chem, 277, 20804–20811. Macdonald DJ, Glen ACA, Boyle RM et al (2000) An ELISA method for Type IV cPLA2 in RBC: a potential marker for schizophrenia. Schizophr Res, 41, 259. Maes M, Smith R, Christophe A et al (1996) Fatty acid composition in major depression: decreased o-3 fractions in cholesteryl esters and increased C20:4-o-6/C20: 5-o-3 ratio on cholesteryl esters and phospholipids. J Affect Disord, 38, 35–46. Mathe AA, Eberhard G, Saaf J and Wetterberg L (1986) Plasma prostaglandin E2 metabolite in twins with schizophrenic disorder. Biol Psychiat 21, 1024–1030. Matsumura K, Watanabe Y, Imni-Matsumura K et al (1992) Mapping of prostaglandin E2 binding sites in rat brain using quantitative autoradiography. Brain Res, 581, 292–298. McAllister G, Knowles MR, Patel S et al (1993) Characterization of a chimeric D3/D2 dopamine receptor expressed in CHO cells. FEBS Lett, 324, 81–86. Melis M, Diana M and Gessa GL (2000) Cyclo-oxygenase-inhibitors increase morphine effects on mesolimbic dopamine neurons. Eur J Pharmacol, 387, R1–3. Mizoguchi A, Eguchi N, Kimura K et al (2001) Dominant localization of prostaglandin D receptors on arachnoid trabecular cells in mouse basal forebrain and their involvement in the regulation of non-rapid eye movement sleep. Proc Natl Acad Sci USA, 98, 11674–11679. Mollace V, Colasantin M, Rodino P et al (1995) NMDA-dependent prostaglandin E1 release by human cultured astroglial cells is driven by nitric oxide. Biochem Biophys Res Commun, 215, 743–749. Nemets B, Stahl Z and Belmaker RH (2002) o-3 Fatty acid treatment of depressive breakthrough during unipolar maintenance. Am J Psychiat, 159, 477–479. Nielsen JA and Sparber SB (1984) Indomethacin potentiates the operant behaviour suppresant and rectal temperature lowering effects of low doses of D-amphetamine in rats. Pharmacol Biochem Behav, 21, 219– 224. Nishizaki T, Nomura T, Matsuoka T and Tsujishita Y (1999) Arachidonic acid as a messenger for the expression of long-term potentiation. Biochem Biophys Res Commun, 254, 446–449. Normandin M, Gagne J, Bernard J et al (1996) Involvement of the 12lipoxygenase pathway of arachidonic acid metabolism in homosynaptic long-term depression of the rat hippocampus. Brain Res, 730, 40–46. Nissen AA and Spencer KA (1936) The psychogenic problem (endocrinal and metabolic) in chronic arthritis. N Engl J Med, 214, 576–581. Ono N, Abiru T, Shigiyama K and Kamiya H (1992) Influence of cycloxygenase inhibitors on the cataleptic behavior induced by haloperidol in mice. Prostagland Leukotriene Essent Fatty Acids, 46, 59–63. Ono N, Saito R, Abiru B et al (1986) Possible involvement of prostaglandins in cataleptic behavior in rats. Pharmacol Biochem Behav, 25, 463–471. Peet M, Brind J, Ramchand CN (2001) Two double-blind placebocontrolled pilot studies of eicosapentaenoic acid in the treatment of schizophrenia. Schizophr Res, 49, 243–251.
498
NERVOUS SYSTEM
Peet M and Horrobin DF (2002) EPA Multicentre Study Group. A doseranging study of the effects of ethyl eicosapentaenoate in patients with persistent schizophrenic symptoms. J Psychiat Res, 36, 7–18. Peet M, Ramchand CN, Shah SH et al (1998) Decreased membrane fluidity and elevated oxidative stress in unmedicated schizophrenics. Schizophr Res 36, 107. Piomelli D, Pilon C, Giros B et al (1991) Dopamine activation of the arachidonic acid cascade as a basis for D1/D2 receptor synergism. Nature, 353, 164–167. Puri BK and Richardson AJ (1999) Brain phospholipid metabolism in dyslexia assessed by magnetic resonance spectroscopy. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Ramchand CN, Wei J, Lee KH and Peet M (1999) Phospholipase A2 gene polymorphism and associated biochemical alterations in schizophrenia. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Rappoport SI, Chang NCJ and Spector AA (2001) Delivery and turnover of plasma-derived essential PUFAs in mammalian brain. J Lipid Res, 42, 678–685. Reid MS, Hsu K, Tolliver BK et al (1996) Evidence for the involvement of phospholipase A2 mechanisms in the development of stimulant sensitization. J Pharmacol Exp Ther, 276, 1244–1256. Reid MS, Ho LB, Hsu K et al (2000) Evidence for the involvement of cyclooxygenase activity in the development of cocaine sensitization. Pharmacol Biochem Behav, 71, 37–54. Richardson AJ and Ross MA (2000) Fatty acid metabolism in neurodevelopmental disorder: a new perspective on associations between attention-deficit/hyperactivity disorder, dyslexia, dyspraxia and the autistic spectrum. Prostagland Leukotriene Essent Fatty Acids, 63, 1–9. Ross BM (1999) Brain and blood phospholipase activity in psychiatric disorders. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Ross BM and Turenne SD (2002) Chronic cocaine adminstration reduces phospholipase A2 activity in rat brain striatum. Prostagland Leukotriene Essent Fatty Acids, 66, 479–483. Ross BM, Moszcynska A, Kalasinsky K and Kish SJ (1996) Phospholipase A2 activity is selectively decreased in the striatum of chronic cocaine users. J Neurochem, 67, 2620–2623. Ross BM, Mamalias N, Moszczynska A et al (2001) Elevated activity of phospholipid biosynthetic enzymes in substantia nigra of patients with Parkinson’s disease. Neuroscience, 39, 117–125. Ross BM, Brooks R, Lee M et al (2002) Effect of systemically administered cyclooxygenase inhibitors upon indices of dopaminergic function. Eur J Pharm, 450, 141–151. Saito R, Fujiwara M, Kamiya H and Ono N (1987) The effect of neurotransmitters on cataleptic behaviour induced by PGD2 in rats. Pharmacol Biochem Behav, 26, 543–546. Schwartz RD and Yu X (1992) Inhibition of GABA-gated chloride channel function by arachidonic acid. Brain Res 595, 405–410. Schwartz RD, Uretsky NJ and Qianchine JR (1982) Prostaglandin inhibition of amphetamine-induced circling in mice. Psychopharmacology, 78, 317–325. Seeman P and Niznik H (1990) Dopamine receptors and transporters in Parkinson’s disease and schizophrenia. FASEB J, 4, 2737–2744. Smith WL, Borgeat P and Fitzpatrick FA (1991) The eicosanoids: cyclooxygenase, lipoxygenase, and epoxygenase pathways. In Vance DE and Vance J (eds), Biochemistry of Lipids, Lipoproteins and Membranes. Amsterdam: Elsevier, 297–325. Stevens LJ and Burgess JR (1999) Essential fatty acids in children with attention-deficit/hyperactivity disorder. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press.
Stoll AL, Severus WE, Freeman MP et al (1999) o-3 Fatty acids in bipolar disorder: a preliminary double-blind, placebo-controlled trial. Arch Gen Psychiat, 56, 407–412. Sugimoto Y, Shigemoto R, Namba T et al (1994) Distribution of the prostaglandin E receptor EP3 in the mouse nervous system. Neuroscience, 62, 919–918. Tocco G, Freirc-Moar J, Schreiber SS et al (1997) Maturational regulation and regional induction of cyclooxygenase-2 in rat brain: implications for Alzheimer’s disease. Exp Neurol, 144, 339–349. Thomas EA, Danielson PE, Nelson PA et al (2001a) Clozapine increases apolipoprotein D expression in rodent brain: towards a mechanism for neuroleptic pharmacotherapy. J Neurochem 76, 789–796. Thomas EA, Dean B, Pavey G and Sutcliffe JG (2001b) Increased CNS levels of apolipoprotein D in schizophrenic and bipolar subjects; implications for the pathophysiology of psychiatric disorders. Proc Natl Acad Sci USA, 98, 4066–4071. Vaddadi KS (1999) Essential fatty acids and movement disorders. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Vaughan CW, Ingram Sl, Connor GA and Christie MJ (1997) How opioids inhibit GAHA-mediated neurotransmission. Nature, 290, 611– 614. Vial D and Piomelli D (1995) Dopamine D2 receptors potentiate arachidonate release via activation of cytosolic, arachidonic-specific phospholipase A2. J Neurochem, 64, 2765–2772. Waldo M (1999) Co-distribution of sensory gating and impaired niacin flush response in the parents of schizophrenics. Schizophr Res, 40, 49– 53. Waldo M, Myles-Worsley M, Madison A et al (1995) Sensory gating deficits in parents of schizophrenics. Am J Med Genetics Neuropsychiat Genet, 60, 506–511. Ward PE and Glen I (1999) Oral and topical niacin flush testing in schizophrenia. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Ward PE, Sutherland J, Glen EM and Glen AI (1998) Niacin skin flush in schizophrenia: a preliminary report. Schizophr Res, 29, 269–274. Wei J and Hemmings GP (2000a) Allelic association of two distinct cytosolic phosholipase A2 genes with schizophrenia. Schizophr Res, 41, 94. Wei J and Hemmings GP (2000b) The NOTCH4 locus is associated with susceptibility to schizophrenia. Nature Genet, 25, 376–377. Williamson PC and Drost DJ (1999) 31P magnetic resonance spectroscopy in the assessment of brain phospholipid metabolism in schizophrenia. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Wise H (1997) Neuronal prostacyclin receptors. Prog Drug Res, 49, 123– 154. Yao JK (1999) Red blood cell and platelet fatty acid metabolism in schizophrenia. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press. Yao JK and Van Kammen DP (1996) Incorporation of 3H-arachidonic acid into platelet phospholipids of patients with schizophrenia. Prostagland Leukotriene Essent Fatty Acids, 55, 21–26. Yao JK, Reddy R, McElhinny LG and Van Kammen DP (1998) Reduced status of plasma total antioxidant capacity in schizophrenia. Schizophr Res, 32, 1–8. Yao JK, Leonard S and Reddy R (2000) Membrane phospholipid abnormalities in postmortem brains from schizophrenic patients. Schizophr Res, 42, 7–17. Yoshida Y, Matsumura H, Nakajima T et al (2000) Prostaglandin E (EP) receptor subtypes and sleep: promotion by EP4 and inhibition by EP1/ EP2. NeuroReport, 11, 2127–2131. Zhang J and Rivest S (1999) Distribution, regulation and colocalization of the genes encoding the EP2- and EP4-PGE2 receptors in the rat brain and neuronal responses to systemic inflammation. Eur J Neurosci, 11, 2651–2668.
47 Essential Fatty Acids: Eicosanoid Precursors in the Treatment of Huntington’s Disease Krishna Vaddadi Monash Medical Centre (Monash University), Clayton, Victoria, Australia
The brain substance is unique in its high concentration of longchain polyunsaturated fatty acids of the o-3 and o-6 series (Clandinin et al 1980a, 1980b). Neuronal tissue on a dry weight basis contains over 50% of lipid and about one-third of this consists of polyunsaturated fatty acids (Sastry 1985). Normal dietary changes induce alterations to the composition of cell membranes and organelles in the brain (Tahin et al 1981; Hargreaves and Clandinin 1987). Polyunsaturated fatty acids and their metabolic products eicosanoids have been implicated in a number of neurological conditions, including Alzheimer’s disease, Huntington’s disease and epilepsy (Farooqui and Horrocks 1991).
ESSENTIAL FATTY ACIDS AND THEIR METABOLITES Neuronal membranes are largely composed of phospholipids, which are rich in essential fatty acids. Essential fatty acids must be taken in the diet as the body cannot synthesize them. There are two types of essential fatty acids, o-3 and o-6 (also known as n-3 and n-6) and these may exert biological actions in their own right or be metabolized to respective eicosanoids. The term ‘‘eicosanoids’’ refers collectively to a group of oxygenated 20-carbon compounds and includes prostaglandins (PGs), thromboxanes (TX) and leukotrienes and other types of hydroxy and hydroperoxy fatty acids (Funk 2001). The most important lipids in the neurons are the phospholipids, such as phosphatidyl-choline, Pethamolamine, P-inositol and P-serine. Each of these has a relatively hydrophilic, partially water-soluble head group with a ‘‘tail’’ consisting of two acyl groups. Different fatty acids may be present in the tail and their specific nature determines the properties of the phospholipid (PL) and the microenvironment within the cell membrane where that particular PL is found. Generally, the most important acyl groups are arachidonic acid (AA, 20:4 [n-6]) and docosahexaenoic acid (DHA, 22:6[n-3]) and these are most commonly found in the sn-2 position of the phospholipid (Horrobin 1999). This membrane PL structure offers a uniquely sensitive biochemical point at which genes and environment interact (Horrobin 1997). The two other fatty acids which give rise to agents active in the cell signalling process are dihomo-g-linolenic acid (DGLA,20:3[n-6]) and eicosapentaenoic acid (EPA, 20:5[n-3]): these are present in the brain in much smaller but still important amounts. The majority of the brain AA, DHA, DGLA and EPA are derived from their precursors in the diet, namely linoleic acid (18:2[n-6]) in the case of DGLA The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
and AA, and a-linolenic acid (18:3[n-3]) in the case of EPA and DHA; smaller but still important amounts may come directly from the diet, particularly seafood in the case of EPA and DHA. Their metabolism is summarized in Table 47.1. Linoleic and a-linolenic acid are converted to four key brain fatty acids by a series of alternating elongations in which two carbon atoms are added to the chain, and desaturations, in which a double bond is introduced into the chain. The rates of these chain elongations and desaturations vary and are influenced by a number of environmental factors. Diabetes, alcohol consumption, high saturated fat consumption, viral infections, catecholamines and steroid hormone production following stress are known to influence the supply of AA, DHA, DGLA and EPA to the PL-synthesising enzymes (Horrobin 1990). The neural membranes are highly interactive and dynamic and therefore the interaction of a ligand with its receptor may markedly affect neural membrane phopholipid metabolism (Farooqui and Horrocks 1991). The biochemical actions of membrane-associated proteins, such as receptors, ion channels and enzymes, can be modulated by the lipid composition of the membrane in which they are embedded (Horrobin and Manku 1990). The addition of two carbon atoms to the fatty acid (FA) chain, or the insertion of a double bond into FA, could substantially change the amount of benzodiazepine bound to a receptor (Witt and Nielsen 1994). Prostaglandins and leukotrienes are potent eicosanoid lipid mediators derived from phospholipase-released AA and are involved in many cellular processes. Prostaglandins are formed by most cells in the human body and are involved as autocrine or paracrine lipid mediators. They are not stored but synthesized de novo from membrane-released arachidonic acid when the cells are activated by specific stimuli. A number of enzymes regulate cellular levels of AA, which is kept esterified until mobilized by one or other member of the phospholipase A2 (PLA2) enzyme family. The PLA2 superfamily consists of a broad range of enzymes defined by their ability to catalyse the hydrolysis of the middle (sn-2) ester bond of phospholipids. The hydrolysis products of this reaction, a free fatty acid and a lysophospholipid, have many important downstream roles and are derived from a number of PLA2 enzymes (Six and Dennis 2000). Arachidonic acid itself can also function as a second messenger without being converted to eicosanoids.
EFAS AND NEURODEGENERATIVE DISORDERS Phospholipids are fundamental to neuronal structure, growth, remodelling and function. Studies in animals have shown that
500
NERVOUS SYSTEM
Table 47.1 Summary of the synthesis of essential fatty acids from the dietary precursors linoleic acid (n-6 series) and a-linolenic acid (n-3 series) n-6 Fatty acids Linoleic
; g-Linolenic ; Dihomo-g-linolenic ; Arachidonic ; Adrenic ; Tetracosatetraenoic ; Tetracosapentaenoic ; Docosapentaenoic
n-3 Fatty acids 18:2 18:3 20:3 20:4 22:4 24:4 24:5 22:5
a-Linolenic ; Octadecatetraenoic Elongase ; Eicosatetraenoic D-5-Desaturase ; Eicosapentaenoic Elongase ; Docosapentaenoic Elongase ; Tetracosapentaenoic D-6-Desaturase ; Tetrahexanoic b-Oxidation ; Docosahexaenoic D-6-Desaturase
18:3 18:4 20:4 20:5 22:5 24:5 24:6 22:6
essential fatty acids (EFAs) are critical in brain development (Simopoulos 1991). Therefore, any abnormalities in phospholipid metabolism arising out of genetic or environmental factors may have a role to play in the development of neurodegenerative disorders. Virtually all neurodegenerative disorders involve abnormal processing of neuronal proteins (Prusiner 2001). Most proteins in neurones are embedded in or are attached to phospholipid-rich membranes. The specific phospholipid structure in the vicinity of a protein will determine quaternary folding, and hence the function of that protein. The abnormal processing of neuronal protein can entail a misfolding of proteins, altered post-translational modification of newly synthesized proteins, abnormal proteolytic cleavage, improper expression or clearing of degraded protein in the neurones (Prusiner 2001). It is therefore conceivable that abnormalities in phospholipid structure and metabolism and their interactions with abnormal processing of proteins would contribute significantly to pathology and clinical manifestations in neurodegenerative disorders such as Huntington’s disease, Alzheimer’s disease and spinocerebellar ataxias. Additionally, because the neuronal tissue is high in its polyunsaturated fatty acid content, it is prone to oxidative stress and has only modest antioxidant defences. High oxygen consumption and modest antioxidant defences, coupled with a high concentration of iron and membrane polyunsaturated fatty acids in neurones, make them susceptible to lipid peroxidatation (Benkovic and Connor 1993). Reactive oxygen species (ROS), in particular the hydroxyl radical, can lead to functional alterations in lipids, proteins and nucleic acids. Accumulation of ROS is considered to be one factor which contributes to neurodegenerative changes in the brains of patients with Parkinson’s disease, Alzheimer’s disease and other dementias (Przedborski 2000; Mattson 2000). Therefore, essential fatty acids in optimal amounts, coupled with an effective antioxidant system, are essential for neuronal survival, function and the prevention of neurodegeneration.
ESSENTIAL FATTY ACIDS AND TARDIVE DYSKINESIA Eicosanoids, which are derived from metabolism of essential fatty acids, have a variety of actions in cell regulatory functions. They act as a source of second messengers in cellular signal transduction (Hudson et al 1993).
In relation to dopamine, a neurotransmitter which is involved in movement disorders such as Parkinson’s disease and neuroleptic-induced tardive dyskinesia, there exists an interaction between dopamine and prostaglandins (PGs). PGs of the E type and dopamine exert a physiological antagonism in the regulation of cyclic adenosine monophosphate (cAMP), PGE increasing it and dopamine lowering it (Myers et al 1978). Thus, a deficit of PGE1 will produce an apparent overactivity of the dopaminergic system, whilst an excess of PGE1 will result in reduced activity of the same system. As PGs are metabolic products of EFAs, the availability of EFAs is likely to influence the dopaminergic system and movements. Davidson et al (1988) demonstrated an interaction between EPA and dopamine in the cat caudate nucleus. They were able to show that both the n-3 and n-6 series of essential fatty acids were important for normal dopaminergic function. Application of dopamine causes release of large amounts of AA (n-6 series EFA) in Chinese hamster ovary cells transfected with both D1 and D2 receptors (Piomelli et al 1991). In presynaptic terminals, following depolarization there is intracellular calcium increase which triggers dopamine release and may at the same time enable D2 autoreceptors to release AA and its metabolites (Piomelli et al 1991). These in turn are involved in the autoinhibiting actions of dopamine, by regulating K+ and Ca2+ channel activities or protein phosphorylation. Therefore, EFAs and their metabolites, as they are released from neuronal membranes, are likely to have a significant impact on abnormal involuntary movements such as occur in tardive dyskinesia (TD). TD is a well-known complication of D2 receptor blocking neuroleptic therapy. Drugs like haloperidol, flupenthixol decanoate, etc. fall into this category. These drugs are commonly used to treat psychiatric disorders such as schizophrenia. TD is manifested mainly by a constellation of abnormal involuntary movements, usually of choreoathetoid type. Principally they affect muscles of the orofacial region, tongue, extremities and trunk (Jeste and Caligiuri 1993). Diminution of abnormal involuntary movements in neuroleptic-treated psychiatric patients has been reported when treated by g-linolenic acid (GLA), a PGE1 precursor (Vaddadi 1984), and also during PGE1 infusions to treat schizophrenia (Kaiya 1984). Reduction in abnormal movements have also been reported by Mellor et al (1995) using 10 g/ day MAX-EPATM, which contains n-3 series EFAs in a group of 20 chronic schizophrenics. In a double-blind placebo-controlled trial of EfamolTM capsules containing g-linolenic acid (45 mg/ capsule) in TD patients, the severity of abnormal movements correlated with levels of RBC membrane fatty acids of both the n-3 and the n-6 series (Vaddadi et al 1989). Dyskinetic movements were more severe in patients who had low levels of RBC membrane n-3 and n-6 series EFAs. There is evidence to suggest that RBC membrane fatty acid levels reflect those in the brain (Carlson et al 1986). Nilsson et al (1996) found, in their prevalence study of abnormal movements in the Swedish general population of men born in 1933, that plasma AA levels were highly significantly reduced in men who exhibited dyskinetic movements, even in the absence of antipsychotic drug exposure. Very recently, Emsley et al (2002), in a randomized controlled study, have shown that ethylEPA was dramatically effective in reducing TD in chronic schizophrenic patients. There are animal model studies of tetralin-induced dyskinesia that was reduced by administering either GLA or DGLA, and this action was blocked by aspirin, a cyclooxygenase inhibitor (Costall et al 1984; Nohria and Vaddadi 1982). It is likely that conversion of DGLA to PGs influenced the antidyskinetic action. This experience prompted Vaddadi to use EFAs in other neurodegenerative disorders, such as Huntington’s disease, to see whether this would be beneficial in either reducing abnormal movements or improving patients’ cognitive functioning or would halt the
ESSENTIAL FATTY ACIDS IN TREATMENT OF HUNTINGTON’S DISEASE progress of the disease. There has not been much published work in relation to EFAs and Huntington’s disease (HD). In a study by Sakai et al (1991), which included some HD patients, there were significantly lowered DHA levels compared to controls.
ESSENTIAL FATTY ACIDS IN THERAPY OF HUNTINGTON’S DISEASE Early Findings HD is a progressive neurological disorder that is inherited as an autosomal dominant disease. Symptoms typically start in mid-life, but can present in early childhood or even in the seventh decade or later. It is characterized by abnormal motor movements, choreoathetoid movements or rigidity, cognitive impairment and psychiatric symptoms such as depression and psychosis, sometimes leading to suicide. The disease causes neurodegenerative changes, mainly in the striatum, with the caudate nucleus being affected more severely than the putamen or globus pallidus. Diffuse gyral atrophy is also common and it progresses to other areas. The disease progresses to death within 15–20 years of symptom onset (Sharp and Ross 1996). The genetic defect is due to expansion of the trinucleotide CAG in the first exon of a gene on chromosome IV. The CAG repeats are translated to polyglutamine repeats in the expressed protein named huntingtin. HD occurs when the length of the poly Q domain exceeds 36Q (Huntington’s Disease Collaborative Research Group 1993).
The Role of Huntingtin Protein Huntingtin protein is a widely expressed protein which resides in the cell cytoplasm. Its exact role is not known, although several have been suggested. These include cellular vesicle trafficking and endocytosis (Kim et al 1999; Velier et al 1998). Perinuclear processing, essential for normal trafficking of secretory membranes and for mitochondrial assembly near the nucleus, has been purported to utilize huntingtin protein. There is evidence to suggest that huntingtin protein may have an antiapoptotic action (Rigamonti et al 2001).
Is Mutant Huntingtin Involved in the Process of Neuronal Death? There are several mechanisms suggesting why a polyglutamine expanded N-terminal end of huntingtin protein might lead to neuronal degeneration. Mutant huntingtin is proteolytically processed, and the resulting amino terminal fragments containing glutamine expansions form aggregates, which are deposited in the cytoplasmic and nuclear inclusions in the brains of HD patients and transgenic mouse models (Kaytor and Warren 1999). Intracellular aggregates are a common feature of polyglutamine disorders: they do not initiate the disorder, but are an indication of a neuron’s efforts to cope with mutant protein (Cummings et al 1999; Klement et al 1998). An expanded polyglutamine tract could make the protein resistant to degradation by altering its conformation (Cummings et al 1999). Insoluble aggregates containing huntingtin protein are found in the cytosol and nuclei of HD patients (DiFiglia et al 1997), transgenic animals (Davies et al 1997) and cellular models of HD (Saudou et al 1998). However, in HD, selective neuronal loss does not correlate well with the presence of aggregates and aggregates are seen in sites where there is no cell death (Orr 2001).
501
A number of neurodegenerative disorders are caused by mutations in genes that play a role in the ubiquitin–proteasome pathway (UPP) (Kaytor and Warren 1999). The UPP controls intracellular levels of a variety of short-lived proteins especially ones involved in maintaining cellular growth and metabolism. Redistribution of the proteasome complex to the polyglutamine aggregates over time causes a decrease in the level of proteasomal activity in the overall cellular environment and an increase in the aggregates. P53 and 70 kDa heat-shock protein (HSP 70) upregulation with expansion of polyglutamine protein accentuates the proteasomal malfunction. Such a reduction in proteasomal function is associated with disruption of mitochondrial membrane potential, release of cytochrome c into the cytosol and activation of caspase-9- and caspase-3-like proteases leading to cell death (Jana et al 2001). Mutant huntingtin protein may be involved in transcriptional events through interaction with transcription factors. It may disrupt the formation of protein complexes that regulate transcription and RNA processing: breakdown of mutant huntingtin protein and its N-terminal fragments in the nucleus can repress transcription aberrantly in HD and cause neuronal dysfunction (Cha 2000; Kegel et al 2002). It has been suggested that expanded polyglutamine alters protein conformation, resulting in aberrant protein interactions (Perutz 1999), including interactions of the expanded polyglutamine with cellular proteins containing short polyglutamine stretches. CREB binding protein (CBP) is a co-activator for CREB-mediated transcription and contains an 18-(human) glutamine stretch. CREB-mediated gene transcription promotes cell survival, and CBP is a major mediator of survival signal in mature neurons: CBP has been found in polyglutamine aggregates in vitro and in vivo (Kazantsev et al 1999; Preisinger et al 1999; McCampbell et al 2000; Steffan et al 2000). Nucifora et al (2001) showed that the expanded polyglutamine may exert toxic effects within cells by sequestering CBP or other proteins containing short polyglutamine stretches away from their critical sites of action. It has been suggested that the short polyglutamine repeats and the expanded polyglutamine repeats in mutant huntingtin protein directly interact, and there may well be interactions with other proteins. These types of interactions may provide explanations for ‘‘toxic gain of function’’ (Paulson and Fischbeck 1996; Housman 1995). The loss of normal functions of huntingtin protein, coupled with the acquisition of ‘‘toxic’’ characteristics of mutant huntingtin, such as mutant huntingtin binding of CREB binding protein (CBP), might lead to destruction of striatal neurones in Huntington’s disease.
Essential Fatty Acids and Transgenic Mouse Model of Huntington’s Disease In order to understand the pathology and the processes involved in Huntington’s disease, several animal models have been developed. These include intrastriatal injection of glutamate analogues, such as tainic acid (Coyle and Schwarcz 1976), quinolinic acid (Beal et al 1986) or a mitochondrial inhibitor, 3nitropropionic acid (3-NP) (Brouillet et al 1993). These models show acute toxic effects and are dissimilar to the slow and variable course of the neurodegenerative process in human HD. There are, however, now well-established transgenic mouse models of HD which provide better models. Transgenic mice have an extra gene added. The new added gene may be an aberrant form of a human gene linked to a known disease such as HD. The mutated form of the human HD gene is added to the mouse genome to generate a mouse model of HD.
502
NERVOUS SYSTEM
Various forms of transgenic mice with different abnormalities, ranging from targeted disruption of the HD gene to a cascade of transgenics, generated using either a full-length or a shorter part of the human HD gene carrying either expanded CAG repeats (up to 150 triplets) or short CAG repeat segments (<20 CAG), have been created. These have been well reviewed by Sipione and Cattaneo (2001) and Brouillet (2000). Those transgenic mice which express either a part of or the fulllength huntingtin with an expanded polyglutamine segment show emergence and progression of an HD-like phenotype with ageing. These mice suffer from progressive weight loss and a generalized lack of muscle bulk, have reduced life-span, motor impairment characterized by abnormal gait, resting tremor and hind limb clasping, early cognitive deficits and histological and neurochemical abnormalities. Mangiarini et al (1996) developed the extensively studied R6 transgenic mouse line to understand the molecular pathology of HD. The R6 line constitutes a series of mice transgenic for a human genomic fragment containing promoter elements, exon 1 and a portion of intron 2 of the huntingtin gene, which carry the following approximate CAG repeat expansions: R6/1, 115; R6/2, 150; R6/5, 128–156 (Mangiarini et al 1996; Davies et al 1997; Carter et al 1999). The R6/2 line has received the most extensive study, due to early emergence and rapid progression of the phenotype. However, the R6/1 line shows a later age at onset and more insidious progression of the phenotype, with subsequent seizures, bradykinesia, weight loss and death, thus showing more homology with the typical course of illness in HD (Carter et al 1999).
EFA Supplementation Improves Movement Disorder in a Transgenic Mouse (R6/1) Model of Huntington’s Disease Essential fatty acids are involved in a number of cellular functions. Due to our previous experience in using EFAs in tardive dyskinesia and successful open human single case studies, we investigated the systemic effects of treatment with EFAs or placebo, given throughout life, on the emergence and progression of phenotype in the R6/1 transgenic mouse model of HD, using assessment techniques which include a novel, ethologically-based approach to dissect neurological impairment into topographical domains of function at a naturalistic level (Clifford et al 2002). The R6/1 line was chosen, since for the reasons noted above it is considered more homologous to the human HD course. Male breeder R6/1 HD transgenic mice were crossed with hybrid (CBA6C57 BL/6) F1 females and transgenic progeny were identified by PCR (Clifford et al 2002). Female mice were randomized to receive oral EFA on alternate days or placebo (coconut oil) in addition to their laboratory feed. The EFA mix included linoleic acid, g-linolenic acid (GLA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Dietary supplementation was continued throughout pregnancy and to the time of weaning, at which point transgenic males and normal male littermates were identified from tail DNA PCR. From weaning, transgenic mice derived from mothers randomized to EFAs continued to receive EFAs on alternate days and those derived from mothers who received placebo continued to receive placebo (for details of the methodology, see Clifford et al 2002). Assessment of mouse behaviour was undertaken at regular intervals by a technique previously described (Clifford et al 2000, 2002; Ross et al 2000).
Summary of the Main Findings The gross neurological phenotype of R6/1 transgenic mice given placebo broadly resembled that previously reported for their R6/2 counterparts but developed more gradually (Mangiarini et al 1996). These R6/1 mice exhibited a slow onset of neurological impairment and an insidious course of dysfunction. Loss of body weight was evident from approximately 12 weeks of age, with emergence of the ‘‘feet-clasping’’ posture on tail suspension from approximately 30 weeks. These gross aspects of the phenotype progressed over a period of 1 year (Clifford et al 2002). Impairments in gait emerged over weeks 38–43, with progressive stride length reduction. Locomotion and overall rearing remained normal up to week 27, followed by progressive diminution over weeks 33–43; both rearing free and rearing to wall showed a profile of progressive decline, while a less clear pattern was evident for rearing seated. Sniffing was slightly diminished at week 27 and this progressed further over weeks 33–43; grooming was elevated as compared to controls at week 27 and this effect was sustained over weeks 38–43. Among other topographies of behaviour, sifting was markedly diminished at week 27, with only limited progression thereafter. Chewing remained normal up to week 38, but by week 43 this behaviour was diminished. Quantitative receptor autography performed at week 43 indicated R6/1 transgenic mice to be characterized in each of the striatum nucleus accumbens and olfactory tubercle by >70% loss of D1-like receptors, and in the striatum by 60% loss of D2-like receptors (Clifford et al 2002). Following treatment with EFAs (48% linoleic acid, 6% glinolenic acid, 3% eicosapentaenoic acid and 2% docosahexaenoic acid) from conception to adulthood, R6/1 transgenic mice showed marked amelioration of the phenotype as compared to placebo-treated transgenic mice. Relative to placebo, treatment with EFAs exerted no effect on normal mice but materially reduced by approximately 50% the incidence of the ‘‘feetclasping’’ posture on tail suspension in HD transgenics, while there was no effect on loss of body weight. With regard to gait, treatment with EFAs exerted no consistent effect in normal mice, but protected HD transgenics against progressive reduction of stride length. Essential fatty acids exerted little or no effect on behavioural topography in normal mice but protected completely against progressive reductions in locomotion, overall rearing and sniffing in HD transgenics, with some ‘‘overshoot’’ of locomotion above the level of EFA-treated normal mice. EFA-treated transgenics showed reduction in grooming evident in HD transgenics. No effect of EFA was observed in autoradiographic studies of D1-like or D2-like receptors in normal mice. EFA treatment also failed to influence the marked loss of D1-like receptors from each of the striatum, nucleus accumbens and olfactory tubercle, or of D2-like receptors from the striatum, in HD transgenics. In Australia, where this experiment was undertaken, for animal welfare reasons it is not possible to obtain death as a specific experimental end point. We noted, therefore, that within the 43week time frame five of 34 (15%) EFA-treated HD transgenic male mice died; the corresponding value for placebo-treated mice was 11 of 26 (42%). This study clearly demonstrated that when EFAs were given from conception on alternate days throughout adulthood, treatment protected substantively against the emergence of the HD phenotype. As both n-3 and n-6 fatty acids were administered, it was not possible to specify to what extent each of the series of EFAs contributed towards the effect and at what specific period these EFAs need be given. EFAs exerted no effect on loss of body weight or of D1- and D2-like receptors.
ESSENTIAL FATTY ACIDS IN TREATMENT OF HUNTINGTON’S DISEASE EFA Supplementation Studies in Huntington’s Disease My initial interest in treating HD patients began when I observed that, at least in some patients with neuroleptic-induced tardive dyskinesia, there was some improvement noticed in abnormal involuntary movements on supplementation with a mixture of essential fatty acids, mainly linoleic and g-linolenic acid (Vaddadi 1984). These original observations were replicated using o-3 fatty acids in schizophrenic patients with neuroleptic-induced TD (Mellor et al 1995; Emsley et al 2002). In a study conducted by Williamson et al (1997), using magnetic resonance spectroscopy with HD patients, decreased levels of phosphomonoesters and phophodiesters in the prefrontal region were found as compared to normal individuals and patients with schizophrenia. My initial open label studies using EFAs, a combination of linoleic and g-linolenic acid in an 82 year-old woman with a diagnosis of HD, confirmed by DNA test, were extremely encouraging. This elderly woman with severe dyskinesia and cognitive impairment improved significantly for 172 weeks. Her quality of life was significantly better due to reduced movements and some improvement in memory. This improvement was maintained until her death due to a chest infection (for a full description, see Vaddadi 1999). The second case previously described by Vaddadi (1999) is that of a man diagnosed as HD gene positive, who responded well intially to a mixture of highly unsaturated fatty acids. He was subsequently switched to pure ethyl eicosapentaenoate. This improvement in his movement disorder scores and cognitive functioning, as assessed by neuropsychological testing, was maintained for 5.5 years, subsequent to which he developed mood disorder and showed some decline. This response in movement disorder scores and cognitive functioning is not a normally observed phenomenon in HD patients; they normally show progressive neurological decline over a 5 year period. The initial positive response led us to do a double-blind placebo-controlled study of HD patients using a mixture of highly unsaturated fatty acids containing both o-6 and o-3 fatty acids. Each capsule (1000 mg) of essential fatty acids contained glinolenic acid 70 mg, eicosapentaenoic acid 35 mg, docasahexaenoic acid 20 mg and a-lipoic acid (antioxidant) 50 mg, with linoleic acid as a carrier. The identical looking placebo capsule contained hydrogenated coconut oil (containing no unsaturated fatty acids) and antioxidants similar to the ones in the essential fatty acid capsules. Each patient received eight capsules daily. All patients had HD with clinical signs such as chorea, and the HD diagnosis was confirmed by DNA testing. The details of this study are described by Vaddadi et al (2002). The trial design was for a period of 2 years, with assessment at baseline and at 6 Table 47.2 Changes in the Rockland–Simpson Dyskinesia Rating Scale (RSDRS) and the Unified Huntington’s Disease Motor Rating Scale (UHDRS-motor) RSDRS
Direction of change Improved Unchanged Deteriorated Mean baseline score Mean end of study score Mean change from baseline
UHDRS-motor
Placebo
Active
Placebo
Active
8 1 1 6 64.4+3.0 71.4+5.1
9 7 0 2 60.0+4.5 55.7+3.8
8 1 0 7 34.8+5.7 45.0+7.6
9 5 0 4 33.2+5.4 29.8+8.5
+7.0+3.7 74.3+2.1
+10.3+4.7 73.4+4.0
Change from baseline to the last assessment is shown. Higher scores are worse and a plus sign indicates deterioration while a minus sign indicates improvement. Reproduced from Vaddadi et al (2002), by permission of NeuroReport
503
monthly intervals. The main assessments were the Rockland– Simpson Dyskinesia Rating Scale (which has been used extensively in dyskinesia research; Simpson et al 1979) and the Unified Huntington’s Disease Rating Scale (UHDRS; 1996). Table 47.2 shows the results of the above study. This trial was stopped early as the results of another smaller trial became available, conducted at the Hammersmith Hospital. Puri et al (2002) used pure ethyl eicosapentaenoate, one of the components of the mixture used in our study. The Hammersmith trial showed clear evidence of benefit and hence, on ethical grounds, it was felt appropriate to stop the trial and evaluate the results and offer existing patients active treatment. A new international multicentre trial was then established and is currently being conducted at Johns Hopkins, Emory and Harvard Medical Schools in the USA, the University of British Columbia in Canada, Monash University in Melbourne and the Hammersmith Hospital in London. Ethyl-EPA appeared to show modest beneficial effects in motor scores as measured by (TMS4) sub-scale (Leavittt et al 2003). The details of this trial will be published during 2004. In our trial the average duration of treatment was 19 months. Table 47.2 shows that in the placebo group, six patients became worse, one was unchanged and one was better. In contrast in the active group, seven patients improved and two deteriorated. Analysis of variance showed a difference between active and placebo groups at p=0.01. This difference in the motor score was measured by the Rockland–Simpson Dyskinesia Rating Scale. When Unified Huntington Disease Rating Scale (1996) scores were analysed in the placebo group, seven patients deteriorated and one improved, whereas in the active group five improved and four deteriorated, and the difference between the groups was of borderline significance at p=0.08. The cognitive and behavioural assessment scale results showed non-significant changes. There were trends to a lower rate of deterioration on active therapy than on placebo, as shown by timed Stroop tests. The results of this study are significant and clinically important. These must be assessed against the background of a dominant genetic condition, with death within 15–20 years after onset, which causes a great deal of physical and emotional distress and for which there is no treatment. In this condition, invasive procedures such as neurotransplantation are seriously considered (Bachoud-Levi et al 2000). In the Hammersmith Hospital pilot trial (Puri et al 2002) patients were randomized to placebo or purified ethyl-EPA, which was supplied by Laxdale Research, UK. The dose of ethyl-EPA in this study was 2 g daily in four capsules each containing 500 mg lipid. This was a 6 month randomized, double-blind, placebocontrolled study which included seven inpatients with advanced stage III HD. There were three patients on ethyl-EPA and four on placebo. They were rated using the UHDRS. Patients on ethylEPA improved on the orofacial component of the UHDRS, while those on placebo deteriorated on this scale ( p<0.03). In this study 3-D MRI brain scans were done at baseline and then at follow-up at 6 months (Puri et al 2002). The volume scans for each patient were accurately registered and subtraction images were obtained to demonstrate changes between baseline and follow-up images. Four patients successfully underwent MRI scanning at baseline and at 6 months, two were on placebo and two were on ethyl-EPA. The subtraction images showed that an increase in ventricular size had taken place in both patients on the placebo; in contrast, the subtraction images from the two patients treated with ethyl-EPA showed evidence of an overall decrease in ventricular size. This was a small but carefully conducted study that showed an improvement in both cerebral structure and movement disorder scores in response to ethyl-EPA. The clinical results in HD patients in open-label studies, two double-blind placebo-controlled studies, and in the transgenic
504
NERVOUS SYSTEM
mouse model using essential fatty acids are all consistent. This is very encouraging as it might offer a novel therapy with no toxicity in HD.
Unsaturated Fatty Acids and Protein Interactions
In treating HD and other neurodegenerative disorders, if one could arrest or slow down the neurodegenerative process in specific parts of the brain, it would be a milestone in therapeutic advance. Cells die either by necrosis or apoptosis. Inhibiting programmed cell death (apoptosis) in neurones after ischaemic brain injury promotes functional recovery in animal models of stroke (Martinou et al 1994; Hara et al 1997). It is therefore plausible to defer cell death in neurodegenerative disorders and maintain cellular function by inhibiting apoptosis.
The phospholipid environment determines the final tertiary and quarternary structure of membrane proteins and may modify their functioning. It has been proposed by many investigators that HD is caused by toxic ‘‘gain of function’’, which is in turn caused by abnormal protein–protein interactions. Huntingtin protein has been shown to interact with a variety of proteins (Walling et al 1998). This alteration of normal (or the conferral of novel) protein–protein interactions may be at the crux of the altered activity of huntingtin, its nuclear localization, inclusion formation and cell death (Walling et al 1998). It is therefore possible that essential fatty acids, by creating a highly flexible and unsaturated cellular environment, prevent these abnormal protein–protein interactions and also prevent the cascade of events leading to neuronal death.
Inhibition of Caspases
Replacement of Lost Membrane Fatty Acids
The caspases are a family of cysteine proteases that are important for regulating apoptosis (Alnemri et al 1996). There is evidence of caspase-1 activation in the brains of mice and humans with HD. It has been shown that in the transgenic mouse model of HD expression of a dominant-negative caspase-1 mutant extends survival and delays appearance of neuronal inclusions, neuroreceptor alterations and onset of symptoms, indicating that caspase-1 has a major role in the pathogenesis of neurodegenerative diseases such as HD (Ona et al 1999). Elevation of mature IL-1b levels has been used as sensitive and specific evidence for caspase-1 activation, as caspase-1 is the main enzyme to activate pro-IL-1b in vivo and convert it to mature 1L1b (Li et al 1995). Levels of mature IL-b in R6/2 mice were elevated to 268% of those of wild-type control mice ( p=0.0007) (Ona et al 1999). Il-1b levels in the brains of HD patients (HD grade 3–4) were significantly increased to 213% that of normal controls. Intracerebroventricular administration of a caspase inhibitor delayed disease progression and mortality in the mouse model of HD (Ona et al 1999). McGahon et al (1999) have shown that in aged rats there is a decrease in the brain antioxidant defences and there is an increase in IL-1b concentrations. There is a decrease in brain AA in aged rats, and there is an impairment in the ability to sustain long-term potentiation in the dentate gyrus. These changes, including elevated IL-1b concentrations, are reversed by supplementing the diet with AA, glinolenic acid (n-6 series EFAs) and also eicosapentaenoic acid. Therefore, there is evidence that inhibition of caspase activity as measured by IL-1b concentrations can be reversed by essential fatty acids, which might help to halt neuronal degeneration.
It is also conceivable that in neurodegenerative disorders the membrane fatty acids are just lost through an oxidative process, and that oral supplementation with essential fatty acids simply corrects this loss.
How Do Essential Fatty Acids Work in Huntington’s Disease?
Inhibition of Phospholipase A2 The intracellular enzyme phospholipase A2 (PLA2) catalyses the hydrolysis of membrane phospholipids to release fatty acids. Arachidonic acid (AA) and docosahexaenoic acid are predominantly found at the sn-2 position of membrane phospholipid. PLA2 specifically removes fatty acids from the sn-2 position (Horrobin 1997). Increased PLA2 activity accelerates the breakdown of neuronal membrane phospholipids. One of the ways of counteracting membrane breakdown would be to at least partially block phospholipase A2 (Finnen and Lovell 1991). EPA has also been shown to inhibit irradiation-induced apoptosis (Lonergan et al 2002) Therefore, supplementation with EPA may prevent loss of membrane fatty acids and stop or minimize the cascade of cell apoptosis signalling events and thus neuronal death in HD.
CONCLUSION Whatever the mechanism, the use of eicosapentaenoic acid, and possibly other essential fatty acids, is emerging as an important new development in the management of tardive dyskinesia and Huntington’s disease.
ACKNOWLEDGEMENTS Thanks to Dr David Horrobin of Laxdale Research, Sterling, Scotland, for reviewing this manuscript.
REFERENCES Alnemri ES, Livingston DJ, Nicholson DW et al (1996) Human ICE/ CED-3 protease nomenclature. Cell, 87, 171. Bachoud-Levi AC, Remy P, Nguyen JP et al (2000) Motor and cognitive improvements in patients with Huntington’s disease after neural transplantation. Lancet, 356, 1975–1979. Beal MF, Kowall NW, Ellison DW et al (1986) Replication of the neurochemical characteristics of Huntington’s disease by quinolinic acid. Nature, 321, 168–171. Benkovic SA and Connor JR (1993) Ferritin, transferrin, and iron in selected regions of the adult and aged rat brain. J Comp Neurol, 338, 97–113. Brouillet E (2000) Animal models of Huntington’s disease: from basic research on neuronal death to assessment of new therapeutic strategies. Funct Neurol, 15, 239–251. Brouillet E, Jenkins BG, Hyman BT et al (1993) Age-dependent vulnerability of the striatum to the mitochondrial toxin 3nitropropionic acid. J Neurochem, 60, 356–359. Carlson SE, Carver JD and House SG (1986) High fat diets varying in ratios of polyunsaturated to saturated fatty acid and linoleic to linolenic acid: a comparison of rat neural and red cell membrane phospholipids. J Nutr, 116, 718–725. Carter RJ, Lione LA, Humby T et al (1999) Characterization of progressive motor deficits in mice transgenic for the human Huntington’s disease mutation. J Neurosci, 19, 3248–3257. Cha JH (2000) Transcriptional dysregulation in Huntington’s disease. Trends Neurosci, 23, 387–392.
ESSENTIAL FATTY ACIDS IN TREATMENT OF HUNTINGTON’S DISEASE Clandinin MT, Chappell JE, Leong S et al (1980a) Extrauterine fatty acid accretion in infant brain: implications for fatty acid requirements. Early Hum Dev, 4, 131–138. Clandinin MT, Chappell JE, Leong S et al (1980b) Intrauterine fatty acid accretion rates in human brain: implications for fatty acid requirements. Early Hum Dev, 4, 121–129. Clifford JJ, Drago J, Natoli AL et al (2002) Essential fatty acids given from conception prevent topographies of motor deficit in a transgenic model of Huntington’s disease. Neuroscience, 109, 81–88. Clifford JJ, Usiello A, Vallone D et al (2000) Topographical evaluation of behavioural phenotype in a line of mice with targeted gene deletion of the D2 dopamine receptor. Neuropharmacology, 39, 382–390. Costall B, Kelly ME and Naylor RJ (1984) The antidyskinetic action of dihomo-g-linolenic acid in the rodent. Br J Pharmacol, 83, 733–740. Coyle JT and Schwarcz R (1976) Lesion of striatal neurones with kainic acid provides a model for Huntington’s chorea. Nature, 263, 244–246. Cummings CJ, Reinstein E, Sun Y et al (1999) Mutation of the E6-AP ubiquitin ligase reduces nuclear inclusion frequency while accelerating polyglutamine-induced pathology in SCA1 mice. Neuron, 24, 879–892. Davidson B, Kurstjens NP, Patton J and Cantrill RC (1988) Essential fatty acids modulate apomorphine activity at dopamine receptors in cat caudate slices. Eur J Pharmacol, 149, 317–322. Davies SW, Turmaine M, Cozens BA et al (1997) Formation of neuronal intranuclear inclusions underlies the neurological dysfunction in mice transgenic for the HD mutation. Cell, 90, 537–548. DiFiglia M, Sapp E, Chase KO et al (1997) Aggregation of huntingtin in neuronal intranuclear inclusions and dystrophic neurites in brain. Science, 277, 1990–1993. Emsley RA, Myburgh CC, Oosthuizen PP and van Rensberg SJ (2002) o-3 Fatty acids and schizophrenia: a randomised trial of ethyleicospentaenoate (EPA) vs. placebo as add-on treatment (presented at winter Schizophrenia Workshop, Davos, Switzerland, 24–29 Feb 2002). Schizophren Res, 53. Farooqui AA and Horrocks LA (1991) Excitatory amino acid receptors, neural membrane phospholipid metabolism and neurological disorders. Brain Res Brain Res Rev, 16, 171–191. Finnen MJ and Lovell CR (1991) Purification and characterisation of phospholipase A2 from human epidermis. Biochem Soc Trans, 19, 91S. Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science, 294, 1871–1875. Hara H, Fink K, Endres M et al (1997) Attenuation of transient focal cerebral ischemic injury in transgenic mice expressing a mutant ICE inhibitory protein. J Cereb Blood Flow Metab, 17, 370–375. Hargreaves KM and Clandinin MT (1987) Phosphatidylethanolamine methyltransferase: evidence for influence of diet fat on selectivity of substrate for methylation in rat brain synaptic plasma membranes. Biochim Biophys Acta, 918, 97–105. Horrobin DF (1990) g-Linolenic Acid. Rev Contemp Pharmacother, 1, 1–41. Horrobin DF (1997) Fatty acids, phospholipids and schizophrenia. In Handbook of Essential Fatty Acid Biology: Biochemistry, Physiology and Behavioural Neurobiology. Humana Press, Totona, NJ, 245–256. Horrobin DF (1999) The phospholipid concept of psychiatric disorders and its relationship to the neurodevelopmental concept of schizophrenia. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press, 285–296. Horrobin DF and Manku MS (1990) Clinical biochemistry of essential fatty acids. In Horrobin DF (ed.), Omega-6 Essential Fatty Acids: Pathophysiology and Roles in Clinical Medicine. New York: Wiley-Liss, 21–53. Housman D (1995) Gain of glutamines, gain of function? Nature Genet, 10, 3–4. Hudson CJ, Young LT, Li PP and Warsh JJ (1993) CNS signal transduction in the pathophysiology and pharmacotherapy of affective disorders and schizophrenia. Synapse, 13, 278–293. Huntington Study Group (1996) Unified Huntington’s Disease Rating Scale: reliability and consistency. Huntington Study Group. Mov Disord, 11, 136–142. Huntington’s Disease Collaborative Research Group (1993) A novel gene containing trinucleotide repeat that is unstable on Huntington’s disease chromosomes. Cell, 72, 971–983. Jana NR, Zemskov EA, Wang G and Nukina N (2001) Altered proteasomal function due to the expression of polyglutamineexpanded truncated N-terminal huntingtin induces apoptosis by
505
caspase activation through mitochondrial cytochrome c release. Hum Mol Genet, 10, 1049–1059. Jeste DV and Caligiuri MP (1993) Tardive dyskinesia. Schizophr Bull, 19, 303–315. Kaiya H (1984) Prostaglandin E1 treatment of schizophrenia. Biol Psychiat, 19, 457–462. Kaytor MD and Warren ST (1999) Aberrant protein deposition and neurological disease. J Biol Chem, 274, 37507–37510. Kazantsev A, Preisinger E, Dranovsky A et al (1999) Insoluble detergentresistant aggregates form between pathological and non-pathological lengths of polyglutamine in mammalian cells. Proc Natl Acad Sci USA, 96, 11404–11409. Kegel KB, Meloni AR, Yi Y et al (2002) Huntingtin is present in the nucleus, interacts with the transcriptional corepressor C-terminal binding protein, and represses transcription. J Biol Chem, 277, 7466– 7476. Kim M, Velier J, Chase K et al (1999) Forskolin and dopamine D1 receptor activation increase huntingtin’s association with endosomes in immortalized neuronal cells of striatal origin. Neuroscience, 89, 1159– 1167. Klement IA, Skinner PJ, Kaytor MD et al (1998) Ataxin-1 nuclear localization and aggregation: role in polyglutamine-induced disease in SCA1 transgenic mice. Cell, 95, 41–53. Leavitt BR, Hayden MR, Ross CA et al (2003) Ethyl-Eicosapentaenoate (Ethyl-EPA) in Huntington’s disease: A randomized placebocontrolled trial. Poster presentation at the American College of Neurology Conference in Hawaii, USA, March 2003. Li P, Allen H, Banerjee S et al (1995) Mice deficient in IL-1b-converting enzyme are defective in production of mature IL-1b and resistant to endotoxic shock. Cell, 80, 401–411. Lonergan PE, Martin DS, Horrobin DF and Lynch MA (2002) Neuroprotective effect of eicosapentaenoic acid in hippocampus of rats exposed to g-irradiation. J Biol Chem, 277, 20804–20811. Mangiarini L, Sathasivam K, Seller M et al (1996) Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell, 87, 493–506. Martinou JC, Dubois-Dauphin M, Staple JK et al (1994) Overexpression of BCL-2 in transgenic mice protects neurons from naturally occurring cell death and experimental ischemia. Neuron, 13, 1017–1030. Mattson MP (2000) Free radical-mediated disruption of cellular ion homeostasis, mitochondrial dysfunction, and neuronal degeneration in sporadic and inherited Alzheimer’s disease. In Poli G, Kadenas E and Packer L (eds), Free Radicals in Brain Pathophysiology. New York: Marcel Dekker, 323–357. McCampbell A, Taylor JP, Taye AA et al (2000) CREB-binding protein sequestration by expanded polyglutamine. Hum Mol Genet, 9, 2197– 2202. McGahon BM, Murray CA, Horrobin DF and Lynch M (1999) Agerelated changes in oxidative mechanisms and LTP are reversed by dietary manipulation. Neurobiol Aging, 20, 643–653. Mellor JE, Laugharne JD and Peet M (1995) Schizophrenic symptoms and dietary intake of n-3 fatty acids. Schizophr Res, 18, 85–86. Myers PR, Blosser J and Shain W (1978) Neurotransmitter modulation of prostaglandin E1-stimulated increases in cyclic AMP. II. Characterization of a cultured neuronal cell line treated with dibutyryl cyclic AMP. Biochem Pharmacol, 27, 1173–1177. Nilsson A, Horrobin DF, Rosengren A et al (1996) Essential fatty acids and abnormal involuntary movements in the general male population: a study of men born in 1933. Prostagland Leukotriene Essent Fatty Acids, 55, 83–87. Nohria V and Vaddadi KS (1982) Tardive dyskinesia and essential fatty acids: an animal model study. In Horrobin DF (ed.), Clinical Uses of Essential Fatty Acids. Montreal: Eden Press. Nucifora FC Jr, Sasaki M, Peters MF et al (2001) Interference by huntingtin and atrophin-1 with cbp-mediated transcription leading to cellular toxicity. Science, 291, 2423–2428. Ona VO, Li M, Vonsattel JP et al (1999) Inhibition of caspase-1 slows disease progression in a mouse model of Huntington’s disease. Nature, 399, 263–267. Orr HT (2001) Beyond the Qs in the polyglutamine diseases. Genes Dev, 15, 925–932. Paulson HL and Fischbeck KH (1996) Trinucleotide repeats in neurogenetic disorders. Ann Rev Neurosci, 19, 79–107.
506
NERVOUS SYSTEM
Perutz MF (1999) Glutamine repeats and neurodegenerative diseases: molecular aspects. Trends Biochem Sci, 24, 58–63. Piomelli D, Pilon C, Giros B et al (1991) Dopamine activation of the arachidonic acid cascade as a basis for D1/D2 receptor synergism. Nature, 353, 164–167. Preisinger E, Jordan BM, Kazantsev A and Housman D (1999) Evidence for a recruitment and sequestration mechanism in Huntington’s disease. Philos Trans R Soc Lond B Biol Sci, 354, 1029–1034. Prusiner SB (2001) Shattuck lecture—neurodegenerative diseases and prions. N Engl J Med, 344, 1516–1526. Przedborski S and Jackson-Lewis VR (2000) ROS and Parkinson’s disease: a view to a kill. In Poli G, Kadenas E and Packer L (eds), Free Radicals in Brain Pathophysiology. New York: Marcel Dekker, 273–290. Puri BK, Bydder GM, Counsell SJ et al (2002) MRI and neuropsychological improvement in Huntington disease following ethyl-EPA treatment. NeuroReport, 13, 123–126. Rigamonti D, Sipione S, Goffredo D et al (2001) Huntingtin’s neuroprotective activity occurs via inhibition of procaspase-9 processing. J Biol Chem, 276, 14545–14548. Ross SA, Wong JY, Clifford JJ et al (2000) Phenotypic characterization of an a-4 neuronal nicotinic acetylcholine receptor subunit knock-out mouse. J Neurosci, 20, 6431–6441. Sakai T, Antoku Y, Iwashita H et al (1991) Chorea-acanthocytosis: abnormal composition of covalently bound fatty acids of erythrocyte membrane proteins. Ann Neurol, 29, 664–669. Sastry PS (1985) Lipids of nervous tissue: composition and metabolism. Prog Lipid Res, 24, 69–176. Saudou F, Finkbeiner S, Devys D and Greenberg ME (1998) Huntingtin acts in the nucleus to induce apoptosis but death does not correlate with the formation of intranuclear inclusions. Cell, 95, 55–66. Sharp AH and Ross CA (1996) Neurobiology of Huntington’s disease. Neurobiol Dis, 3, 3–15. Simopoulos AP (1991) Omega-3 fatty acids in health and disease and in growth and development. Am J Clin Nutr, 54, 438–463.
Simpson GM, Lee JH, Zoubok B and Gardos G (1979) A rating scale for tardive dyskinesia. Psychopharmacology (Berl), 64, 171–179. Sipione S and Cattaneo E (2001) Modeling Huntington’s disease in cells, flies, and mice. Mol Neurobiol, 23, 21–51. Six DA and Dennis EA (2000) The expanding superfamily of phospholipase A(2) enzymes: classification and characterization. Biochim Biophys Acta, 1488, 1–19. Steffan JS, Kazantsev A, Spasic-Boskovic O et al (2000) The Huntington’s disease protein interacts with p53 and CREB-binding protein and represses transcription. Proc Natl Acad Sci USA, 97, 6763–6768. Tahin QS, Blum M and Carafoli E (1981) The fatty acid composition of subcellular membranes of rat liver, heart, and brain: diet-induced modifications. Eur J Biochem, 121, 5–13. Vaddadi KS (1984) Essential fatty acids and neuroleptic drug induced tardive dyskinesia—preliminary clinical observations. IRCS Med Sci, 12, 678–679. Vaddadi KS (1999) Essential fatty acids and movement disorders. In Peet M, Glen I and Horrobin DF (eds), Phospholipid Spectrum Disorder in Psychiatry. Carnforth: Marius Press, 285–296. Vaddadi KS, Courtney P, Gilleard CJ et al (1989) A double-blind trial of essential fatty acid supplementation in patients with tardive dyskinesia. Psychiatry Res, 27, 313–323. Vaddadi KS, Soosai E, Chiu E and Dingjan P (2002) A randomised, placebo-controlled, double blind study of treatment of Huntington’s disease with unsaturated fatty acids. NeuroReport, 13, 29–33. Velier J, Kim M, Schwarz C et al (1998) Wild-type and mutant huntingtins function in vesicle trafficking in the secretory and endocytic pathways. Exp Neurol, 152, 34–40. Walling HW, Baldassare JJ and Westfall TC (1998) Molecular aspects of Huntington’s disease. J Neurosci Res, 54, 301–308. Williamson PC, Malla A, Cortese L et al (1997) Phosphorus 31 magnetic resonance spectroscopy in patients with Huntington disease. Arch Gen Psychiatry, 54, 186–188. Witt MR and Nielsen M (1994) Characterization of the influence of unsaturated free fatty acids on brain GABA/benzodiazepine receptor binding in vitro. J Neurochem, 62, 1432–1439.
Section Nine Reproductive System
48 Perspectives and Clinical Significance of Eicosanoids in Obstetric and Gynaecological Practice I. Z. MacKenzie John Radcliffe Hospital, Oxford, UK
During the past 50 years there have been considerable advances in obstetric and gynaecological practice. Developments in anaesthesia, with the widespread use of intrapartum epidural analgesia, have radically improved management for many labouring woman, while the adoption of regional rather than general anaesthesia for Caesarean section has reduced the morbidity of the operation. The availability of cross-matched blood, with the recognition of the significance of the increasing number of red cell antigens and their potential impact for the recipient of a transfusion, as well as of the complications for the foetus in utero, represent major advances in obstetrics. The exploitation of endoscopy for diagnostic and therapeutic purposes has radically changed many aspects of gynaecological treatment, reducing surgical morbidity and promoting more rapid postoperative recovery. Alongside these advances must be placed the expanding knowledge of the physiological roles and the exploitation of the pharmacological functions of the eicosanoids for a large variety of therapeutic strategies in both obstetric and gynaecological practice, which have provided significant improvements in clinical results. The following chapters highlight areas that relate to male and female reproduction, detailing current understanding of the functions of the eicosanoids. A lot of that knowledge has already been transposed, where applicable, into routine or research clinical practice. It is probable that further roles will be identified with the basic and applied research that continues, resulting in further advances in treatment protocols. FERTILITY Male Fertility At present there are few recognized therapeutic uses of the prostaglandins in male reproduction and fertility. The physiological functions of prostaglandins in semen or seminal plasma, both replete in prostaglandins, remain speculative. As yet they do not appear to have a role in enhancing or reducing male fertility. However, the concept of the prostaglandins influencing the immune response in the male is possibly relevant, since such modulation is potentially lost when extracorporeal conceptions by in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) are used. It is reassuring, therefore, that there have been no suggestions to date of any immunological compromise to the partners receiving embryos conceived by such techniques, nor to The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
the resultant offspring. It is possible that this disturbance to the immune response could be responsible for the failure of cases of extracorporeal conceptions and unsuccessful implantation following embryo transfer. This possibility is supported by the observation that women who do not abstain from coitus following embryo transfer during IVF treatment have a higher successful implantation rate than those who do abstain (Tremellen et al 2000). It is tempting to suggest that a deficiency of prostanoids might have a role in unexplained male subfertility, but the evidence presented by Kelly (Chapter 49, this volume) strongly indicates that the prostaglandins have no function in either male ejaculation of sperm or sperm transport in the female. Despite this, it is common practice to reduce the seminal fluid content of sperm preparations introduced transcervically into the uterus during intrauterine insemination (IUI) to minimize the possibility of uterine contractility being stimulated and thereby reduce the chances of fertilization. Prostaglandin E1 is, however, used as a therapy for treating erectile dysfunction in the male. This problem has been estimated to affect 52% of men 40–70 years of age with a rate of around 65% by the age of 65 years (Feldman et al 1994). It has been shown that it relaxes the corporeal smooth muscle, thereby facilitating penile erection (Porst 1997; Engel and McVary 1998). The intrapenile injection of prostaglandin E1 (alprostadil) systemically into the dorsum of the corpus carvernosum, although very effective in some men, is associated with a high attrition rate, perhaps because of complications with priapism (Fallon 1995). Oral sildenafil (Viagra) has a 55% success rate (Leungwattanakij et al 2001) and intracavernous alpostadil has been found to be useful and safe in those men who do not respond satisfactorily to sildenafil (Shabsigh et al 2000). Intraurethral instillation of PGE1 by pessary or gel (MUSE) was introduced in 1997 (PadmaNathan et al 1997) as a more acceptable administration approach for prostaglandins and has been much studied since. This results in the vasoactive agent passing into the corpus spongiosum, with subsequent transfer into the corpora cavernosa through vascular channels. As illustrated in Figure 48.1, this treatment modality is effective for approximately 30–50% of subjects but the long-term consistency rates appear to be only around 30% (Mulhall et al 2001). At present both intracavernosal and intraurethral prostaglandin E1 are recommended in the UK guidelines for the treatment of erectile dysfunction (Ralph and McNicholas 2000).
510
REPRODUCTIVE SYSTEM
Figure 48.1 Successes with at-home intraurethral alprostadil 1000 mg suppository (MUSE) to treat erectile dysfunction in 61 subjects according to patient’s age. Data from Mulhall et al (2001)
Female Fertility
MANAGEMENT OF ABORTION
The function of the prostaglandins in female fertility, particularly that surrounding implantation, remains very uncertain. It seems highly likely that any small disturbance in homeostasis could result in discouraging implantation and lead to female infertility. A greater understanding of the process might allow further elucidation of the mechanisms for some cases of unexplained infertility. It is known that PGE2 is responsible for increased local vascular permeability at the implantation site and that the blastocyst contains PGE2 and its enzyme precursors. The fact that increased local vascular permeability is one of the earliest signs of implantation indicates that prostaglandin E2 is involved with implantation. Since non-steroidals, including indomethacin and diclofenac, as well as dexamethasone, all discourage implantation in small mammals, and this effect can be mitigated by a variety of prostaglandins, it is possible that the addition of a prostaglandin around the time of extracorporeal fertilization and at the time of embryo transfer might enhance the success of IVF pregnancy rates; the work by Tremellen and colleagues (2000) already referred to supports this concept. However, as cautioned by Poyser (Chapter 51, this volume), the prostaglandins may have a role in spacing, hatching and development of the blastocyst. The contribution by prostaglandins and the leukotrienes to implantation in the human clearly needs much further study. A better understanding might improve techniques of assisted fertility. It might also provide an additional approach to fertility control in the female. The role of the prostaglandins in luteolysis in the human similarly remains obscure; the available evidence clearly indicates that the process of luteolysis differs among animal species. The observation that prostaglandin F2a is luteolytic in cattle led to attempts to prove the same pharmacological effect in humans and produce a method of early ‘medical’ abortion (Mocsary and Csapo 1975; Karim 1973; Ylikorkala et al 1974). This potential clinical application of the prostaglandins was unfortunately not realized, but there remains the possibility of developing approaches that discourage or dislodge the implanting blastocyst or preserve or curtail the functioning corpus luteum. There are currently no proven therapies used that depend upon a known influence on prostaglandin production or modulation that affects implantation or luteal survival.
Therapeutic Abortion The introduction of the prostaglandins to terminate pregnancy during the first and second trimester of pregnancy has had a dramatic impact on the opportunities for reducing the morbidity of therapeutic abortion, especially during the second trimester. Figure 48.2 illustrates the use of prostaglandins in England and Wales during the past 20 years. The introduction and licensing of the antiprogestin mifepristone, as discussed by Cowan and Calder (Chapter 52, this volume), has resulted in a much greater acceptance of the concept of medical abortion during both the first and second trimesters of pregnancy. Apart from the advantage of avoiding any surgical procedure in a high proportion of cases, and thus the risk of genital tract trauma and anaesthetic risks when the operation is performed using a general anaesthetic, the addition of the antiprogestin to the prostaglandin protocol has allowed midtrimester medical abortion to be provided as a day case procedure. Figure 48.3 illustrates the impact the introduction of mifepristone has had during the past 30 years on bed occupancy for midtrimester abortion in one gynaecological unit. With the change in the abortion law in England and Wales in 1991, the opportunity to provide a reliable method of pregnancy termination for cases of severe foetal abnormality at around 20– 22 weeks’ gestation has been particularly welcome. There have been reports of foetal anomalies in pregnancies of women exposed to both mifepristone (Pons et al 1991) and misoprostol (Fonseca et al 1991) when the pregnancy has continued unintentionally. The anomalies noted, sirenomelia (mermaid syndrome) following mifepristone and Mobius syndrome following misoprostol, are unlikely to have been provoked by either drug; the need for continued vigilance is evident. Spontaneous Abortion When women are referred to a gynaecological unit and a diagnosis of early missed abortion (blighted ovum) or incomplete abortion is made, they are usually advised to undergo uterine curettage under anaesthetic to remove any retained products of conception and to control haemorrhage. While this management seems appropriate for those who are bleeding heavily, it may not
CLINICAL SIGNIFICANCE OF EICOSANOIDS IN OBSTETRICS/GYNAECOLOGY
511
Figure 48.2 Change in use of prostaglandins for therapeutic abortion in England and Wales, 1983–2000. Data from the annual reports of abortion statistics for England and Wales, Office of Population Censuses and Surveys (2002)
be essential for many and a more conservative ‘‘wait-and-see’’ approach may be better. As an example of this philosophy, the administration of a prostaglandin with or without prior treatment with mifepristone has been tried with some success for incomplete abortion (Henshaw et al 1993; Chung et al 1999) and early missed abortion (Creinin et al 1997; Ayres-de-Campos et al 2000; Demetroulis et al 2001). Although there are still relatively few studies reported, a review of these studies suggests that it is an appropriate approach to management (Blanchard et al 2002). It appears that patient selection is important, avoiding those with an obvious risk of excess bleeding. Further objective study is still needed to decide whether medical management should be adopted as part of routine gynaecological practice in preference to a more ready resort to surgical evacuation.
Ectopic Pregnancy During the 1970s studies were conducted into the effects of PGF2a and PGE2 on the Fallopian tubes in rabbits and humans. These studies concluded, with relatively little supporting evidence, that PGF2a caused tubal contraction, while PGE2 caused tubal relaxation (Hahlin et al 1987). This suggested a possible role of PGF2a in the management of tubal ectopic pregnancy. Initial trials with PGF2a 0.5–1.5 mg injected into an unruptured tubal ectopic pregnancy, as well as into the ovary containing the corpus luteum, were highly successful (Lindblom et al 1987). However, a subsequent report of 18 cases using PGF2a 2.0 mg in combination with PGE2 given systemically identified two episodes of cardiac arrhythmia and one case of hypertension with marked pulmonary oedema (Egarter and Husslein 1988). Experience using 15-methylPGF2a in 26 patients 3 years later reported success in 92% without complication (Lindblom et al 1991).
Figure 48.3 Impact of mifepristone on day-case admissions for midtrimester prostaglandin induction of abortion (unpublished data)
Work in this area has continued but has concentrated on examining the local injection of cytotoxic agents or the use of minimal access surgery. To date there has been no clear evidence of which, if any, medical management of the unruptured ectopic will surpass the surgical removal of the gestation sac.
MANAGEMENT OF LABOUR Labour Induction and Augmentation By the end of the 1980s, the prostaglandins had become a part of labour induction in virtually all obstetric units in the UK and
512
REPRODUCTIVE SYSTEM
elsewhere in the developed world. A survey conducted in 1989 produced the results depicted in Figure 48.4. The demands of women regarding their care at the time of childbirth have increased, especially during the past decade. While induction has been seen by some as an intervention to be avoided if possible, the need for planned delivery will always be present. The introduction of the prostaglandins during the 1970s into the schedules of methods available provided an opportunity to reduce the degree of invasion experienced by women when delivery is considered essential. As described by Cowan and Calder (Chapter 52, this volume), the prostaglandins now have an established place in obstetric practice. They have not only improved patient satisfaction with a less invasive approach to labour induction, but have also enhanced the outcome of labour, especially in the more difficult situations, with greater prospects of vaginal delivery rather than Caesarean section. Most of the available evidence suggests that the prostaglandins offer no advantage over the alternative of intravenous oxytocin titration for stimulating labour in patients in whom the membranes have ruptured but contractions have not started or when labour is rather dilatory (Hannah 1996). During the past decade the introduction of misoprostol, the PGE1 analogue which is stable at room temperature and is effective if taken orally (unlike PGE2), has been the major focus of attention for labour induction. At present there are few studies that have examined the efficacy of this newer cheaper PGE1 analogue in nulliparae and multiparae separately; studies involving women of mixed parity can potentially obscure important differences in clinical efficacy. With the suggested possible increase in the incidence of hyperstimulation, it seems that it will be some time yet before this preparation can be recommended over the alternative licensed preparations available for clinical use.
Missed Abortion and Intrauterine Foetal Death Unlike early missed abortion, the prostaglandins have revolutionized the management of patients in whom foetal death has occurred but spontaneous abortion or labour does not start. The place for uterine priming with an antiprogestin has not yet been firmly established. Progesterone levels are generally low in cases of foetal death, and an antiprogestin would not therefore seem to be indicated, although the administration of epostane, the
3b-hydroxysteroid dehydrogenase inhibitor, was found to further reduce the circulating concentrations of progesterone and appears to enhance treatment results (Selinger 1988). Cabrol et al (1985) first reported successful induction of labour for proven foetal death using mifepristone 200 mg 12 hourly for 2 days, and subsequently a prospective randomized trial showed that mifepristone 200 mg three times daily for 2 days resulted in 63% of 46 patients delivering their missed abortion or stillbirth within 72 h of the start of treatment, compared with only 17% of patients randomized to placebo (Lelaidier et al 1993). Further investigation in this area is needed before it will become clear whether the inevitable delay associated with the combination approach with an antiprogestin and prostaglandin is worthwhile. The recent finding that 25% women with an intrauterine foetal death are delivered by Caesarean section because the attempt at labour induction was unsuccessful illustrates the need for improved induction methods (MacKenzie and Gould 1997).
Management During the Third Stage of Labour The prostaglandins play a fundamental part in the physiological processes of uterine contractility and myometrial retraction, vasoconstriction and platelet activation immediately following delivery. Prostaglandin metabolites rapidly increase in the peripheral circulation at the time of placental separation, pointing to a dramatic increase of myometrial prostaglandin synthesis, which is an important part of the mechanism responsible for the prevention and control of primary post-partum haemorrhage (Husslein and Sinzinger 1984). The administration of prostaglandins for routine management of the third stage of labour was initially proposed by Kerekes and Domokos (1979) in a prospective randomized trial using PGF2a 1– 2 mg intramyometrially compared with intramuscular ergometrine or no oxytocin injection; the PGF2a resulted in lower blood loss. Routine intramyometrial injections would not be acceptable as a routine and, interestingly, intramyometrial injection of 15-methylPGF2a at Caesarean section offers no advantage over oxytocin (Catanzarite 1990; Chou and MacKenzie 1994). Studies using intramuscular injections of prostaglandins have consistently suggested that blood loss at delivery can be reduced compared with ergometrine, but at the expense of higher rates of vomiting and diarrhoea.
Figure 48.4 Use of prostaglandins for labour induction in the UK by 1989. Data derived from MacKenzie and Boland (1993)
CLINICAL SIGNIFICANCE OF EICOSANOIDS IN OBSTETRICS/GYNAECOLOGY Misoprostol is active orally and is extensively metabolized during and/or prior to its gastrointestinal absorption; several metabolites are produced, one of which, misoprostol acid, is biologically active. Absorption is extremely rapid and the acid is detected in the systemic circulation within 2 min of its oral ingestion, with peak plasma levels reached in 30 min (Zieman et al 1997; Danielson et al 1999). Figure 48.5 illustrates the kinetics of oral and vaginally administered misoprostol. Meta-analyses of a number of studies comparing oral misoprostol with ergometrine or oxytocin suggested that there might be parity as far as efficacy is concerned. Because of this and the stability of misoprostol, together with the ease of administration, the possibility of using this as a routine for third stage management, especially in the underdeveloped world, prompted a large multinational multicentre study to be conducted; post-partum haemorrhage is a significant cause of maternal death in underdeveloped countries (Fawcus et al 1997; Li et al 1996). Unfortunately, this study found that oral misoprostol 600 mg was not as effective at parenteral oxytocin 10 IU at preventing excessive blood loss, and had a high rate of causing maternal shivering and pyrexia (Gulmezoglu et al 2001). The prostaglandins have been widely used to control atonic post-partum haemorrhage threatening the life of the patient. A variety of treatments have been used, including intramyometrial PGF2a 250–1000 mg (Tagaki et al 1976), intramuscular 15-methylPGF2a 500 mg (Corson and Bolognese 1977), PGE2 20 mg vaginal pessaries (Hertz et al 1980) or intravenous infusion at 10 mg/min (Henson et al 1983), gemeprost 1 mg pessaries vaginally (Barrington and Roberts 1993) and rectally (Raiman and Tai 1991) and, more recently, misoprostol tablets rectally (O’Brien et al 1998). Acute uterine inversion has also been managed with 15-methylPGF2a 500 mg 250 mg (Heyl et al 1980) and PGE2 0.1–0.2 mg intramyometrially (Thiery and Delbeke 1985). By 1991, intramuscular 15-methyl-PGF2a (carboprost) had become commercially available. However, concern followed that this prostaglandin treatment could have potentially serious sideeffects. Five cases of maternal arterial desaturation were described following the intramuscular or intramyometrial injection of 15methyl-PGF2a to control atonic post-partum haemorrhage (Hankins et al 1988), although this serious hazard was not identified during a study examining intramyometrial 15-methylPGF2a for routine use at elective Caesarean section (Chou and MacKenzie 1994). Severe hypotension and collapse have also been reported with intramuscular and intramyometrial injections of
513
prostaglandins (Popat et al 1991; Mahmood and Smith 1989); this appears to be the result of considerable overdosing with prostaglandins. There is little doubt that the prostaglandins can be very valuable in this very worrying situation, but care must be maintained to avoid excessive dosing and inadvertent intravenous injection with planned intramuscular or intramyometrial therapy. Whether there would be any advantage in combining oxytocin with prostaglandins in massive post-partum haemorrhage is worth considering, since prostaglandins induce gap junction formation, which enhances myometrial contractility. The two drugs in combination could provoke an even more rapid myometrial contraction by stimulating the inositol phospholipid pathway, increasing intracellular calcium transfer (Lopez Bernal and Watson 1992). PROSTAGLANDINS AND THE NEONATE As described by Momma (Chapter 53, this volume), there has been a much greater understanding of ductus arteriosus physiology during the past 30 years with the progress in prostaglandin research. The impact of the prostaglandins on maintaining ductal patency and the antiprostaglandins promoting ductal closure has been developed to great therapeutic benefit. The possibility of premature ductal closure occurring in the foetus of women who were given indomethacin for tocolysis to control or prevent preterm labour was recognized many years ago and led some clinicians to caution against using indomethacin for this purpose. This and other non-steroidal antiinflammatory drugs (NSAIDs) are known to cross the placenta and reach the foetal circulation in measurable amounts (MacKenzie et al 1985). Although not a foetal risk before 24 weeks’ gestation, the risk of provoking premature ductal closure in utero, however, increases with advancing gestation. The observation that Doppler studies can assess foetal ductus arteriosus blood flow suggests that this should provide a reliable guide to ductal function and be used for monitoring the impact on the foetus of women treated with NSAIDs for tocolysis (Reiss 2001; Sherer and Divon 1996). The observation highlighted by Momma, that glucocorticoids are potent ductus constrictors in preterm rats (Momma et al 1981), is important, since there is now widespread acceptance that the maternal administration of corticosteroids promotes the maturation of a number of foetal systems and reduces the
Figure 48.5 Mean+SD plasma concentrations of misoprostol acid following the oral and vaginal administration of 400 mg misoprostol tablets by 20 women during 7–13 weeks’ gestation. Data derived from Zieman et al (1997)
514
REPRODUCTIVE SYSTEM
morbidity and mortality of the very preterm foetus when delivery is necessary or inevitable (Crowley 1989). Since this prophylaxis has been practised for more than 10 years, it seems likely that any detrimental effect on ductus physiology would have been identified by now. Whether the augmenting effect that vitamin A has on promoting ductal closure is an added risk for those women who take vitamin A during the antenatal period needs to be assessed. Similarly, misoprostol possibly crosses the placenta and reaches the foetal circulation when given for labour induction; this might influence ductal closure and should be investigated. There is evidence that PGE2 given for labour induction reaches the foetal circulation (Sellers and MacKenzie 1985; MacKenzie et al 1989) but no evidence that it compromises ductal function following delivery, but any difference in the kinetics of the misoprostol might be significant. It seems likely that any significant ill-effects of PGE2 on the foetus or neonate would have been recognized by now, since it has been widely used for more than 25 years for labour induction in the UK, with more than 1 million mothers and babies exposed to this prostaglandin. Caution should be shown until more is known about the placental kinetics of misoprostol and longitudinal studies on neonatal and infant outcome have been reported.
OTHER AREAS OF CLINICAL RELEVANCE Menstrual Symptomatology The realization that prostaglandin concentrations in menstrual fluid were elevated in women troubled by spasmodic dysmenorrhoea led to the rational treatment of the condition with prostaglandin synthase inhibitors (Zhang and Li Wan Po 1998). With the same observation made for women complaining of heavy menstrual losses, when an organic explanation was not apparent, the same approach was introduced with similar symptomatic benefit (Rees 1999). The further elucidation of the enzymatic pathways identifying the role of COX-1 and COX-2 has led to the prescribing of more specific inhibitors (Morrison et al 1999). The use of these synthase inhibitors has also been recommended to relieve the discomfort that patients experience with minor outpatient gynaecological procedures that involve manipulating or dilating the cervix, as is required with the insertion of intrauterine contraceptive devices, diagnostic aspiration curettage or hysteroscopy (Buttram et al 1979). In contrast, the administration of local prostaglandins prior to those procedures that involve cervical distension or dilatation has been shown to have some therapeutic benefit compared with placebo (Lauersen et al 1982; Elder and Lewis 1991; Fung et al 2002); there has been little further investigation to confirm the limited information provided by this study.
Modulating Detrusor Contractility The role of the various prostanoids in the physiology and pathophysiology and pharmacology of urinary bladder function has received some attention over the years but few conclusions have been reached. After some initial interest in exploring the role of prostaglandins in managing bladder muscle dysfunction in the 1980s (Mikhailidis et al 1987), with some renewed enthusiasm in the 1990s (Maggi 1993), there remains no established role for the prostaglandins in urogynaecology, despite the acknowledgement that they probably can modify reflex micturition.
SUMMARY The uses to which the prostaglandins and ‘‘antiprostaglandins’’ have been put in obstetric and gynaecological practice have continued to expand during the past 25 years. As a consequence, there has been a dramatic improvement in many aspects of the care that is provided. As further research into fertility techniques proceed, it seems probable that advances will occur for both partners. The continuing elucidation of the enzymatic pathways of the eicosanoids should herald more specific treatment of nonpregnancy-associated gynaecological disorders.
REFERENCES Ayres-de-Campos D, Teixeira-da-Silva J, Campos I and Patricio B (2000) Vaginal misoprostol in the management of first-trimester missed abortions. Int J Gynaecol Obstet, 71, 53–57. Barrington JW and Roberts A (1993) The use of gemeprost pessaries to arrest postpartum haemorrhage. Br J Obstet Gynaecol, 100, 691–692. Blanchard K, Clark S, Winikoff B et al (2002) Misoprostol for women’s health: a review. Obstet Gynecol, 99, 316–332. Buttram V, Izu A and Henzl MR (1979) Naproxen sodium in uterine pain following intrauterine contraceptive device insertion. Am J Obstet Gynecol, 134, 575–578. Cabrol D, Bouvier D-M, Mermet E et al (1985) Induction of labour with mifepristone after intrauterine foetal death. Lancet, 2, 1019. Catanzarite VA (1990) Prophylactic intramyometrial carboprost tromethamine does not substantially reduce blood loss relative to intramyometrial oxytocin at routine cesarean section. Am J Perinatol, 7, 39–42. Chou MM and MacKenzie IZ (1994) A prospective, double-blind, randomized comparison of prophylactic intramyometrial 15-methyl prostaglandin F2a, 125 mg, and intravenous oxytocin, 20 U, for the control of blood loss at elective Cesarean section. Am J Obstet Gynecol, 171, 1356–1360. Chung TKH, Lee DTS, Cheung LP et al (1999) Spontaneous abortion: a randomized, controlled trial comparing surgical evacuation with conservative management using misoprostol. Fertil Steril, 71, 1054–1059. Corson SL and Bolognese RJ (1977) Postpartum uterine atony treated with prostaglandins. Am J Obstet Gynecol, 129, 918–919. Creinin MD, Moyer R and Guido R (1997) Misoprostol for medical evacuation of early pregnancy failure. Obstet Gynecol, 89, 768–772. Crowley P (1989) Promoting pulmonary maturity. In Chalmers I, Enkin M and Keirse MJNC (eds), Effective Care in Pregnancy and Childbirth. Oxford: Oxford University Press, 746–764. Danielson KG, Marions L, Rodriguez A et al (1999) Comparison between oral and vaginal administration of misoprostol on uterine contractility. Obstet Gynecol, 93, 275–280. Demetroulis C, Saridogan E, Kunde D and Naftalin AA (2001) A prospective randomized controlled trial comparing medical and surgical treatment for early pregnancy failure. Hum Reprod, 16, 365–369. Egarter C and Husslein P (1988) Prostaglandins in the treatment of tubal pregnancy. Eicosanoids Fatty Acids, 5, 44–50. Elder MG and Lewis G (1991) Randomised double blind placebo controlled trial of gemeprost pessaries for cervical dilatation in nonpregnant women. Acta Obstet Gynecol Scand, 70, 149–152. Engel JD and McVary KT (1998) Transurethral alprostadil as therapy for patients who withdrew from or failed prior intracavernous injection therapy. Urology, 51, 687–692. Fallon B (1995) Intracavernous injection therapy for male erectile dysfunction. Urol Clin N Am, 22, 833–845. Fawcus S, Mbizvo M, Lindmark G and Nystrom L (1997) A communitybased investigation of maternal mortality from obstetric haemorrhage in rural Zimbabwe. Trop Doc, 27, 159–163. Feldman HAGI, Hatzichristou DG, Krane RJ et al (1994) Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol, 151, 54–57. Fonseca W, Alencar AJC, Mota FSB and Coelho HLL (1991) Misoprostol and congenital malformations. Lancet, 338, 56. Fung TM, Lam MH, Wong SF and Ho LC (2002) A randomised placebocontrolled trial of vaginal misoprostol for cervical priming before
CLINICAL SIGNIFICANCE OF EICOSANOIDS IN OBSTETRICS/GYNAECOLOGY hysteroscopy in postmenopausal women. Br J Obstet Gynaecol, 109, 561–565. Gulmezoglu AM, Villar J, Ngoc TN et al (2001) WHO multicentre randomised trial of misoprostol in the management of the third stage of labour. Lancet, 358, 689–695. Hahlin M, Bokstrom H and Lindblom B (1987) Ectopic pregnancy: in vitro effects of prostaglandins on the oviduct and corpus luteum. Fertil Steril, 47, 935–941. Hankins GD, Berrman GK, Scott RT and Hood D (1988) Maternal arterial desaturation with 15-methyl prostaglandin F2a for uterine atony. Obstet Gynecol, 72, 367–370. Hannah ME (1996) Prostaglandins vs. oxytocin for prelabour rupture of membranes (PROM) at term (37 weeks and onwards). In Keirse MJNC, Renfrew MJ, Neilson JP and Crowther CA (eds), Pregnancy and Childbirth Module of The Cochrane Database of Systematic Reviews, 1996. London: British Medical Journal Publishing Group. Henshaw RC, Cooper K, El-Refaey H et al (1993) Medical management of miscarriage: non-surgical uterine evacuation of incomplete and inevitable spontaneous abortion. Br Med J, 306, 894–895. Henson G, Gough JD and Gillmer MD (1983) Control of persistent primary postpartum haemorrhage due to uterine atony with intravenous prostaglandin E2: case report. Br J Obstet Gynaecol, 90, 280–282. Hertz RH, Sokol RJ and Dierker LJ (1980) Treatment of postpartum uterine atony with prostaglandin E2 vaginal suppositories. Obstet Gynecol, 56, 129–139. Heyl W, Stubblefield PG and Phillippe M (1980) Recurrent inversion of the puerperal uterus managed with 14(S)-15-methyl prostaglandin F2a and uterine packing. Obstet Gynecol, 63, 263–264. Husslein P and Sinzinger H (1984) Concentration of 13,14-dihydro-15keto-PGF2a in the maternal peripheral plasma during labor of spontaneous onset. Br J Obstet Gynaecol, 91, 228–231. Karim SMM (1973) Intrauterine prostaglandins for outpatient termination of very early pregnancy. Lancet, ii, 894. Kerekes L and Domokos D (1979) The effect of prostaglandin F2a on third stage of labour. Prostaglandins, 18, 161–166. Lauersen NH, Kurkulos M, Graves ZR and Leeds L (1982) A new IUD insertion technique utilizing cervical priming with prostaglandin. Contraception, 26, 59–63. Lelaidier C, Benifla JL, Fernandez H et al (1993) The value of RU-486 (mifepristone) in medical indications of the induction of labor at term. Results of a double-blind randomized prospective study (RU-486 vs. placebo). J Gynecol, Obstet Biol Reprod, 22, 91–100. Leungwattanakij S, Flynn V Jr and Hellstrom WJ (2001) Intracavernosal injection and intrurethral therapy for erectile dysfunction. Urol Clin N Am, 28, 343–354. Li XF, Forney JA, Kotelchuck M and Glover LH (1996) The post-partum period: the key to maternal mortality. Int J Gynaecol Obstet, 54, 1–10. Lindblom B, Kallfelt B, Hahlin M and Hamberger L (1987) Local prostaglandin F2a injection for termination of ectopic pregnancy. Lancet, i, 776–777. Lindblom B, Hahlin M, Lundorff P and Thorburn J (1991) Treatment of tubal pregnancy by laparoscope-guided injection of prostaglandin F2a. Fertil Steril, 54, 404–409. Lopez Bernal A and Watson SP (1992) Cellular endocrinology. In Drife JO and Calder AA (eds), Prostaglandins and the Uterus. London: Springer Verlag, 213–235. MacKenzie IZ and Boland J (1993) Current therapeutic uses of prostaglandins in obstetrics in the United Kingdom. Contemp Rev Obstet Gynaecol, 5, 9–14. MacKenzie IZ and Gould S (1997) Stillbirth and Caesarean section. J Obstet Gynaecol, 17, 431–434. MacKenzie IZ, Graf AK and Mitchell MD (1985) Prostaglandins in the foetal circulation following maternal ingestion of a prostaglandin synthetase inhibitor during mid-pregnancy. Int J Gynaecol Obstet, 23, 455–458. MacKenzie IZ, Castle B and Bellinger J (1989) Transfer of prostaglandins to the foetus after prostaglandin E2 vaginal pessary administration. Am J Obstet Gynecol, 161, 920–923. Maggi CA (1993) Prostanoids as local modulators of reflex micturition. Pharmacol Res, 25, 13–20.
515
Mahmood TA and Smith NC (1989) Severe hypotension following intramyometrial injection of prostaglandins in a case of primary atonic postpartum haemorrhage. J Obstet Gynaecol, 9, 217–218. Mikhailidis DP, Jeremy JY and Dandona P (1987) Urinary bladder prostanoids—their synthesis, function and possible role in the pathogenesis and treatment of disease. J Urol, 137, 577–582. Mocsary P and Csapo AI (1975) Work in progress—menstrual induction with PGF2a and PGE2. Prostaglandins, 10, 545–551. Momma K, Nishihara S and Ota Y (1981) Constriction of the foetal ductus arteriosus by glucocorticoid hormones. Paediat Res, 15, 19–21. Morrison BW, Daniels SE, Kotey P et al (1999) Rofecoxib, a specific cyclooxygenase-2 inhibitor, in primary dysmenorrhea. Obstet Gynecol, 94, 504–508. Mulhall JP, Jahoda AE, Ahmed A and Parker M (2001) Analysis of the consistency of intraurethral prostaglandin E1 (MUSE) during at-home use. Urology, 58, 262–266. O’Brien P, El-Refaey H, Gordon A and Rodeck CH (1998) Rectally administered misoprostol for the treatment of postpartum haemorrhage unresponsive to oxytocin and ergometrine: a descriptive study. Obstet Gynecol, 92, 212–214. Office of Population Censuses and Surveys (2002) Abortion Statistics 2000. England & Wales. London: HMSO. Padma-Nathan H, Hellstrom WJ, Kaiser FE et al (1997) Treatment of men with erectile dysfunction with transurethral alprostadil. N Engl J Med, 336, 1–7. Pons JC, Imbert MC, Elefant E et al (1991) Development after exposure to mifepristone in early pregnancy. Lancet, 338, 763. Popat MT, Suppiah N and White JB (1991) Cardiac arrest following intramyometrial injection of prostaglandin E2. Anaesthesia, 46, 236. Porst H (1997) Transurethral alprostadil with MUSE (medicated urethral system for erection) vs. intracavernous alprostadil—a comparative study in 103 patients with erectile dysfunction. Int J Impotency Res, 9, 187–192. Raiman S and Tai C (1991) Intrarectal gemeprost in the treatment of intractable postpartum haemorrhage due to uterine atony. J Obstet Gynaecol, 11, 264–265. Ralph D and McNicholas T (2000) UK management guidelines for erectile dysfunction. Br Med J, 321, 499–503. Rees MCP (1999) Refractory menorrhagia. In Sheth S and Sutton C (eds), Menorrhagia. Oxford: Isis Medical Media, 127–132. Reiss RE (2001) Maternal diseases and therapies affecting the foetal cardiovascular system. In Allen HD, Gutgesell HP, Clark EB and Driscoll DJ (eds), Moss Adams Heart Disease in Infants, Children and Adolescents. Balitmore: Williams & Wilkins, 551–568. Selinger M (1988) Progesterone inhibition in mid-pregnancy with Epostane. Contemp Rev Obstet Gynaecol, 1, 67–79. Sellers S and MacKenzie IZ (1985) Prostaglandin release following vaginal prostaglandin treatment for labour induction. London: Royal Society of Medicine Symposium, 77–84. Shabsigh R, Padma-Nathan H, Gittleman M et al (2000) Intracavernous alprostadil alfadex (Edex/Viridal) is effective and safe in patients with erectile dysfunction after failing sildenafil (Viagra). Urology, 55, 477– 480. Sherer DM and Divon MY (1996) Prenatal ultrasonographic assessment of the ductus arteriosus: a review. Obstet and Gynecol, 87, 630–637. Tagaki S, Yoshida T, Togo Y et al (1976) The effects of intramyometrial injection of prostaglandin F2a on severe post-partum hemorrhage. Prostaglandins, 12, 565–579. Thiery M and Delbeke L (1985) Acute puerperal uterine inversion: twostep management with a b-mimetic and a prostaglandin. Am J Obstet Gynecol, 153, 891–892. Tremellen KP, Valbuena D, Landeras J et al (2000) The effect of intercourse on pregnancy rates during assisted human reproduction. Human Reprod, 15, 2653–2658. Ylikorkala O, Jouppila P, Ylostalo P and Jarvinen PA (1974) Intrauterine injection of prostaglandin F2a for termination of early pregnancy in outpatients. Prostaglandins, 7, 57–70. Zhang WY and Li Wan Po A (1998) Efficacy of minor analgesia in primary dysmenorrhoea: a systematic review. Br J Obstet Gynaecol, 105, 780–789. Zieman M, Fong SK, Benowitz NL et al (1997) Absorption kinetics of misoprostol with oral or vaginal administration. Obstet Gynecol, 90, 88–92.
49 Prostaglandins and Male Reproductive Physiology Rodney W. Kelly University of Edinburgh Centre for Reproductive Biology, Edinburgh, UK
Prostaglandins in the male tract have been an enigma since these vasoactive and smooth-muscle stimulants were first characterized as ‘‘prostaglandin’’ and ‘‘vesiglandin’’ (Von Euler 1936). There is good reason to suppose that human seminal plasma is a potent modulator of the immune system and several reviews have covered this area (Alexander and Anderson 1987; James and Hargreave 1984; Kelly and Critchley 1997) but explanations are now possible that implicate seminal plasma components as inducers of tolerance for the spermatozoal antigens. Maintenance of tolerance in the presence of infection of the reproductive tract may have been a problem that had to be overcome relatively late in evolution as infections of the reproductive tracts spread, and accordingly the solutions to the problem may have varied widely between species. This chapter describes the prostaglandins in semen and in the testis and in both locations they apparently play a role in maintaining immune tolerance. PROSTAGLANDINS IN HUMAN SEMINAL PLASMA An agent that contracted the uterine smooth muscle was identified in human seminal plasma some 70 years ago (Kurzrock and Liebb 1930) and this agent was later characterized as ‘‘prostaglandin’’ by von Euler (1936) who assumed that these lipids were synthesized in the prostate. In these early studies a second lipid was identified in rhesus monkey seminal vesicles and called ‘‘vesiglandin’’. The prostatic origin of prostaglandin has since proved to be incorrect and it is now accepted that the prostaglandins in human semen originate largely from the seminal vesicle (Bendvold et al 1985; Gerozissis et al 1982; Hamberg, 1976). Moreover, since rhesus monkey semen contains but one major prostaglandin, 19hydroxy-PGE-1 (Kelly et al 1976), it is likely that the original ‘‘vesiglandin’’ was this 19-hydroxylated prostaglandin. Von Euler also reported that ‘‘prostaglandin could not be traced in ejaculates from the horse or the bull, nor in secretions from the pig’’ and reflected that such a distribution threw into doubt any suggestion that the seminal PGs were involved in the process of ejaculation. We might also add that, for the same reason, they are unlikely to be involved in the process of sperm transport in the female. Human semen has four major components, PGE1, PGE2 (Samuelsson 1963), 19-hydroxy-PGE1 and 19-hydroxy-PGE2 (Jonsson et al 1975; Taylor and Kelly 1974). The prostaglandins of the A and B series originally reported in semen (Hamberg and Samuelsson 1966) are dehydration artefacts of the corresponding E series prostaglandins (Middleditch 1975). Human seminal plasma also contains smaller amounts of 8-isoprostaglandins (Taylor and Kelly 1975), which we now know can be formed by The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
autooxidation of fatty acids. However, because these products are abundant and their likely precursor by a non-enzymic action must be 19-hydroxy-arachidonic acid (which has not been identified, particularly in semen), the 8-iso-compounds in semen may be derived from cyclooxygenase products. The cytochrome p450 responsible for 19-hydroxylation has now been revealed as CYP4F8 (Bylund et al 2000). Surprisingly, the substrate for this enzyme is not PGE but the endoperoxide PGH. There is a very limited distribution of Cyp4F8, with some expression in liver but none in leukocytes (Bylund et al 2000). This is the first p450 enzyme with a specific affinity for prostaglandin endoperoxides and this route of synthesis of 19-hydroxy-PGE (Figure 49.1) removes all possibility that 19-hydroxlation is a catabolic route for inactivation of PGE. Although these early preliminary studies suggest a very limited distribution of the 19hydroxylase (and consequently of 19-hydroxy-PGE), the elucidation of the sequence allows further searches in other physiological situations. The mean levels of PGE and 19-hydroxy-PGE in human semen are around 70 and 270 mg/ml, respectively (Bendvold et al 1987; Templeton et al 1978) (Table 49.1) with a surprisingly wide range of concentrations found in normal fertile semen. The relationship between semen prostaglandin concentration and fertility is not straightforward. Many of the earlier studies were carried out before the instability of the E series PGs in semen was recognized and the disparity between samples collected from the clinic over an undefined period of time and freshly collected control samples may have biased the results. Thus, there are no reliable data showing a deficiency of prostaglandin in semen from infertile men. There is, however an association between prostaglandin content and semen quality, with 19-hydroxy-PGE being higher and 19hydroxy-PGF being lower in ejaculates with normal sperm motility (Bendvold et al 1984). This may be due to suppression of leukocytes in semen. Since prostaglandins of the E series regulate the phagocytic actions and activation of macrophages and neutrophils (Kunkel et al 1986a, 1986b; Wahl et al 1979; Weissman et al 1973), the seminal prostaglandin will limit secretion of cytokines and activated oxygen species from phagocytic and cytotoxic leukocytes and these activated moieties are detrimental to spermatozoal function (Aitken et al 1994; Wolff et al 1990; Wongkham et al 1991). Since most sperm preparation techniques involve centrifugation, both spermatozoa and white blood cells will be brought into close contact and, under these in vitro conditions, damage may be amplified. There may also be a direct effect of seminal prostaglandin on spermatozoal function, since 19-hydroxy-PGE stimulates ATP levels in these cells (Gottlieb et al 1988), PGE significantly raises intracellular cAMP in spermatozoa (Aitken and Kelly 1985) and PGE (but
518
REPRODUCTIVE SYSTEM A more likely action of seminal prostaglandin is on the immune system of the recipient female, and in the mouse there is direct evidence since an increase in weight of the draining lymph nodes follows exposure to spermatozoa, and this almost certainly represents a ‘‘suppressive’’ immune response (Alexander and Anderson 1987). The exact nature of any such suppresser response is being re-evaluated in the light of the discovery of a specific population of regulatory T cells that are associated with tolerance of food antigen in the gut. The role of prostaglandins in regulatory cell generation is still unclear, but one important feature of the E series prostaglandins is that both PGE and 19-hydroxy-PGE (and consequently human seminal plasma) have an effect on cytokine production, stimulating IL-10 and inhibiting IL-12 (Kelly et al 1997). This cytokine switch has far-reaching consequences in favouring Th-2 cells (Malefyt et al 1992; Trinchieri 1993), anergy and regulatory cell (Tr1) generation (Groux et al 1996, 1997). Regulatory T cells are involved in oral tolerance (Groux et al 1997; Weiner et al 1994) and, as detailed below, they are likely to be involved in modulating immune responses to spermatozoal antigen in both the testis and the female tract.
PROSTAGLANDINS IN MALE REPRODUCTIVE TRACT—ANDROGEN DEPENDENCE
Figure 49.1 The pathway of biosynthesis of 19-hydroxy-PGE, the major prostaglandin of human seminal plasma
not 19-hydroxy-PGE) has an effect on semen quality (Aitken et al 1986). These relatively small effects on semen quality, taken with the very limited species distribution of the seminal prostaglandin, suggest a target for the prostaglandin other than the spermatozoa. However, at this point it should be said that there is some evidence for a prostaglandin receptor on spermatozoa which is involved in transient increases in intracellular Ca2+, and recently it was proposed (Schaefer et al 1998) that such receptors did not match the characteristics of the known classified receptors (Coleman et al 1990).
Prostaglandins are found in most areas of the male reproductive tract and they are largely under the control of androgen. PGE predominates in the extratubular spaces of the testis, whereas PGF is the major prostanoid found in Sertoli cells (Carpenter 1974). The role of PGF in the testis is not clear, but it has been suggested that this stimulator of contraction of peritubular myoid cells mediates pulsatile tubular constriction by an IP3 mechanism and thus stimulates spermatozoal transport within the testis (Tripiciano et al 1998). PGE and PGF are the predominant prostaglandins in the mature rat vas deferens (Gerozissis and Dray 1983). The prepubertal rat vas produces mainly PGI (measured as 6-oxo-PGF), whereas mature animals produce PGE and PGF as major products. Both PGE and PGF are reduced by castration and reappear with testosterone treatment. PGE is more androgen-dependent than PGF, whereas 6-oxo-F is not affected by androgen withdrawal (Gerozissis and Dray 1983). More recently, evidence has been presented that the major cyclooxygenase isoform in the vas deferens is PGHS-2 (McKanna et al 1998), although the agents causing activation of this inducible enzyme at this location are unclear. We do know, however, that the vas enzyme is androgen-dependent from castration experiments. This report begs the question whether all the androgen-dependent (and therefore inducible) PGHS is actually PGHS-2, although in the epididymis it appears that COX-1 and COX-2 are both androgen-dependent (Cheuk et al 2000). The seminal vesicle production of both PGE and 19hydroxy-PGE in man is certainly dependent on testosterone
Table 49.1 Mean levels of PGE and 19-hydroxy-PGE in human semen
PGE 19-Hydroxy-PGE PGF 19-Hydroxy-PGF
Mean (mg/ml)
Standard error of mean
73 67 267 246 2.1 3.2 18 13.3
15.3 49 51 108 0.38 3.0 7.4
Range
(n)
Reference
2–272 8.4–160 53–1094 71–523 0.1–7.0 0.1–13.1 3–62 1–31
22 30 22 31 22 27 22 29
(Templeton et al 1978) (Bendvold et al 1987) (Templeton et al 1978) (Bendvold et al 1987) (Templeton et al 1978) (Bendvold et al 1987) (Templeton et al 1978) (Bendvold et al 1987)
PROSTAGLANDINS AND MALE REPRODUCTIVE PHYSIOLOGY
519
(Skakkebaek et al 1976). The original extraction of PGHS-1 from sheep seminal vesicles has always suggested that this is the isoform responsible for seminal prostaglandin production, but seminal vesicles obtained at slaughter are unlikely to be stimulated and there is now direct evidence for a marked expression of COX-2 in the mature human seminal vesicle (Kirschenbaum et al 2000). IMMUNOSUPPRESSIVE ACTIVITY OF HUMAN SEMINAL PLASMA Two simple but limited tests of immunosuppressive activity have been used over the last two decades: the lymphoproliferation assay, testing effects on T cell replication, and the natural killer (NK)-lysis activity test, which examines effects on the cytotoxic action of NK cells. PGE has been shown to be immunosuppressive in both the lymphoproliferation (Chouaib and Fradelizi 1982; Rappaport and Dodge 1982) and the NK cell assay (Bankhurst 1982; Brunda et al 1980; Droller et al 1978; Jondal et al 1981) and is responsible for the normal suppressive action of monocytes (Fischer et al 1981). Although human semen was shown to be highly immunosuppressive in several studies (James et al 1983; Lord et al 1977; Prakash et al 1976; Shivaji et al 1990; Stites and Erickson 1975), the complexity of the possible contributory agents (James and Hargreave 1984) led to uncertainty on the role of the prostaglandin component. Several immunosuppressive agents have now been identified; semen contains high levels of inhibitors of complement (the cascade of enzymes which is often activated by antibody action and results in cell death) (reviewed in Kelly 1995) and appreciable amounts (160 ng/ml) of the proform of TGFb, but the major activity is due to the prostaglandins, variously ascribed to PGE (Vallely et al 1988), 19-hydroxy-PGE (Tarter et al 1986) and recognized as arising from both fractions (Quayle et al 1989). One problem about the use of both the lymphoproliferation assay and the NK cell assay is that in general they employ peripheral blood mononuclear cell preparations that include monocytes, which probably function as accessory cells, secreting cytokines such as IL-12 (previously known as natural killer cell stimulatory factor), which stimulates NK cell killing, and IL-10, which inhibits T cell proliferation (Taga and Tosato 1992). EFFECTS OF PGE ON T CELL FUNCTION PGE has a major effect in determining the type of T cell response to a presented antigen. All these effects are apparently mediated by an increase in intracellular cAMP (either EP2 or EP4 receptors) and, although the effects are different, they have the same overall endpoint, stimulating a Th2 response and attenuating a Th1 response. T cell responses vary in a continuum from a fully cytotoxic response, characterized by secretion of cytokines such as IL-2 and interferon-g, to a humoral (or antibody) response, characterized by secretion of cytokines such as IL-4, IL-5 and IL-10 (FernandezBotran et al 1988; Mosmann 1988). PGE affects T helper cell function in three ways: first, PGE stimulates IL-10 and inhibits IL-12 (Kraan et al 1995; Strassmann et al 1994) and this will affect the cytokine environment by an effect of monocytes; second, PGE affects differentiation of naı¨ ve T cells, favouring a Th2 outcome (Katamura et al 1995); and third, Th1 and Th2 cells respond differently to an elevated intracellular cAMP level induced by PGE (Betz and Fox 1991; Hilkens et al 1995; Snijdewint et al 1993). EFFECTS OF PGE ON CO-STIMULATORY SIGNALS AND ANERGY One of the major co-stimulatory molecules, B7.2 (CD86), can be induced in microglial cells by a combination of interferon-g and
Figure 49.2 The effect of dilutions of human semen on cytokine release from whole blood preparations. Redrawn from Kelly et al (1997)
lipopolysaccharide. This critical modulator of anergy is inhibited by PGE (Iglesias et al 1997) and in this way the very high levels of prostaglandins in human seminal plasma might be ensuring an anergic outcome of T cell presentation. This effect is shared with IL-10 and other agents that elevate cAMP but is not seen with TGFb. Although this appears to be a direct effect, PGE will have an important indirect effect by cytokine modulation. Expression of both B7.1 and B7.2 is stimulated by interferon-g (Li et al 1996; Orlikowsky et al 1996) and inhibited by IL-10 (Li et al 1996; Orlikowsky et al 1996; Ding and Shevach 1992; Ding et al 1993) and thus the interferon-g:IL-10 ratio is critical to B7 expression. This ratio is very sensitive to the action of PGE, since PGE both inhibits interferon-g production by T cells and stimulates IL-10 production by both T cells and monocytes. There is thus a major effect of PGE to maintain low levels of B7 and consequently avoid priming the adaptive immune system. Antigen presented in the absence of co-stimulatory signals leads to the generation of anergic T cells. Anergic cells are antigenspecific unresponding cells, but in some circumstances such cells act as suppressor or regulatory cells. If antigen is presented to the T cell in the presence of IL-10, antigen-specific anergic cells are formed (Groux et al 1996), but such cells can be stimulated to proliferate clonally. Such cells are termed regulatory cells (Tr1) and after clonal proliferation they can be shown to secrete a large amount of IL-10 with some TGFb and IL-5, but not IL-4 (Groux et al 1997). Although the generation of Tr1 cells is not directly linked to PGE, the powerful effects of PGE in stimulating IL-10 (Figure 49.2) provides a mechanism by which such cells could be generated. Moreover, regulatory cells in the gut are known to release PGE as well as TGFb (Khoury et al 1992). A further role of PGE is to inhibit IL-12 formation (Kraan et al 1995) and the absence of IL-12 is a prerequisite for anergic cell formation (Parijs et al 1997). This property of PGE on IL-12 is seen also with 19hydroxy-PGE and human seminal plasma (Figure 49.3) (Kelly et al 1997).
520
REPRODUCTIVE SYSTEM
Figure 49.3 The main prostaglandin components of semen are active in both stimulating IL-10 and inhibiting IL-12 (Kelly et al 1997). The huge concentrations of prostaglandin in human semen will most probably reach lymph nodes through lymphatic drainage and impose the above cytokine changes which will ensure tolerance for the presented antigen
TOLERANCE FOR SPERMATOZOAL ANTIGEN— THE EFFECT OF SEMINAL PLASMA The question of how the immune system distinguishes between self and non-self is currently the subject of debate. The original hypothesis was that T cells sensitized to antigens present at the time of birth were deleted or that perinatal events gave rise to a population of tolerant T cells that ensured no hostile response to these self-antigens. This idea took little account of problems presented by a maturing reproductive system that would generate new antigen from spermatozoon and conceptus. One answer was that neither spermatozoa nor the villous trophoblast displayed either class II or class I MHC antigens, and the extravillous trophoblast only had non-classical MHC class I antigens (HLA-G and HLA-E), which were mostly monomorphic within the species (Loke and King 1996). However, it is clear that the absence of MHC antigens is not a full answer, since anti-sperm antibodies can be induced and there is new evidence of T cell suppression by the trophoblast (Munn et al 1998). Recent explanations of tolerance suggest that the immune system is able to distinguish between self and non-self, but the outcome of an antigen–T cell encounter is determined by the conditions under which the antigen is presented. Thus, the interaction of the ab T cell with a professional antigen-presenting cell involves interaction with the T cell receptor and other costimulatory signals on the surface of the presenting cell. These costimulatory signals include CD40 and CD54, but the best established are CD80 and CD86 (B7.1 and B7.2) (Medzhitov and Janeway 1997). When antigen is presented with simultaneous expression of costimulatory molecules, then a primed T cell results. In the absence of co-stimulatory signal, the usual outcome is an anergic T cell, although in the presence of highly antigenic peptide sequences, e.g. those derived from virus, the need for co-stimulation is relaxed (Bachmann et al 1998). The ‘‘danger’’ model proposes that the immune system distinguishes between those entities which are either potentially harmful or innocuous (Matzinger 1994) and that this decision is based on a second signal induced by products (such as lipopolysaccharide) that are recognizable as originating from a potentially harmful source (in the case of LPS, bacteria). An alternative but similar hypothesis proposes that the presence
of pathogen-associated signals (pathogen-associated molecular patterns, or PAMPs) is recognized by pattern recognition receptors that are germline-encoded (Medzhitov and Janeway, 1997) and the best example of these are the Toll-like receptors (see Chapter 20, this volume). Although these theories of recognition of ‘‘danger’’ by the immune system are still controversial, it is clear that an autoimmune response can be generated by events such as viral infection. There are also obvious changes to cytokine production in the presence of infection, which would raise the likelihood of the adaptive immune response system mounting a fulminating T cell-mediated attack on subsequent exposure to antigen. In the case of spermatozoa, the absence of MHC antigens may once have been sufficient protection (and may still be so in some other species), but with the advent of infection in semen and/or the female tract the following scenario may have arisen. A serious threat to survival of the species would have been presented as the adaptive immune system generated primed T cells after intercourse, and this was followed by anaphylaxis after a subsequent encounter with sperm antigen. This course of events would have occurred to a different extent in different species, leading to markedly different methods of countering this threat. In many primates, where mating is not controlled by seasonal or cyclical considerations, the sexual spread of genital tract infection may have reached a crisis point earlier than in species where mating was seasonally and/or cyclically restricted and a single partner was the norm. If such a crisis arose, there would be a powerful selection in favour of a component of semen that dampened the cell-mediated response, and there is good reason to suppose that, in man, prostaglandins of the E series fulfil that role. Optimal induction of tolerance involves adequate exposure of the immune system to the antigen, and the extensive leukocytosis of the human cervix following intercourse (Pandya and Cohen 1985) might be a method of ensuring this encounter. In this way it is certain that the spermatozoal antigen will be presented in the presence of sufficient tolerizing agent (in man, prostaglandin of the E series) to ensure a beneficial result. A more important function of exposing spermatozoa antigen to the female’s immune system may, however, be a preparation for pregnancy. Since the trophoblast and spermatozoa share many antigens, it is possible that exposure to semen will induce an ‘‘active tolerance’’ for the
PROSTAGLANDINS AND MALE REPRODUCTIVE PHYSIOLOGY
521
implanting embryo (Kelly 1997; Robertson et al 1997). Thus, it is envisaged that regulatory cells induced by exposure to specific antigens, in the presence of seminal plasma components (PGs in man and TGFb in the mouse), will be created and subsequent exposure to trophoblast antigen will initiate replication of these cells, with accompanying beneficial cytokine release. Although this role for seminal plasma is still hypothetical, regulatory cells in decidua are recognized (Daya et al 1987).
PROSTAGLANDINS AND TESTICULAR FUNCTION An immune response that threatens survival of the individual is clearly counterproductive and therefore key organs such as the brain and the eye have restricted immune responses. The testis is similarly protected and, given that the testis has an efficient lymph drainage system (Setchell et al 1990), such immune privilege is now thought to be due to alterations in both the efferent and the afferent arms of the immune system (Ksander and Streilein 1994). In the case of the testis, immune privilege has been demonstrated by noting prolonged survival of grafts of allogeneic (i.e. from a different individual of the same species) tissue (Head et al 1983). Testicular tolerance can be demonstrated by injecting antigen into the testis, which results in systemic tolerance (Ren et al 1996). This has been termed ‘‘orchidic tolerance’’ and has similarities with oral tolerance. Tolerance is transferable to other animals by injection of splenic T cells from tolerized animals (Ren et al 1996) and thus is likely to depend on antigen-specific active suppression mediated by cytokines. Oral tolerance is also transferable and may involve a population of regulatory cells (Tr1) that secrete IL-10 (Groux et al 1997). The regulatory cells that are responsible for the tolerance are both CD4+ cells, which secrete IL-4 and IL-10, and CD8+ cells, which secrete TGFb (Yotsukura et al 1997). As in the study of several regulatory cell systems, the secretion of PGE as a contributing factor has not been addressed, although there is evidence that regulatory cells involved in oral tolerance secrete PGE (Khoury et al 1992). Moreover, there is good evidence that PGE plays a major role in determining the suppressive microenvironment that will ensure a regulatory cell outcome from antigen presentation to naive T cells. Macrophages and Leydig cells are the major interstitial cells in the testis (Figure 49.4) and the macrophages are a major source of PGE (Kern and Maddocks 1995). The Leydig cell and the interstitial macrophage have an intimate relationship and not only may the macrophage influence steroidogenesis, but also the Leydig cell may influence the secretion of tolerizing signals from the macrophage. In animals treated with ethane dimethyl sulphonate to selectively destroy the Leydig cells, tolerance is reduced, whereas silica, which inhibits macrophage function, has little effect on tolerance (Huhtaniemi et al 1986). These experiments suggest that the effect of the macrophage is to modulate testosterone production by the Leydig cell and that this affects tolerance. However, the Leydig cell may be responsible for the unique prostaglandin and cytokine secretion from the testicular macrophage. Testicular macrophages produce large amounts of PGE (and PGF) without stimulation (Kern and Maddocks 1995) and this accounts for the limited response of these cells to lipopolysaccharide (Kern et al 1995). When indomethacin is introduced in in vitro preparation of testicular macrophages, LPS is effective at stimulating both IL-1 and TNFa. Little is known about IL-10 production by rat interstitial testicular cells but, since PGE is known to stimulate IL10 (Kraan et al 1995; Strassmann et al 1994) and antigen presented in the presence of IL-10 leads to regulatory (Tr1) cells (Groux et al 1997), there is good reason for implicating testicular PGE production in the tolerance that is displayed in this organ. In addition, an absence of IL-12 is necessary for the establishment of
Figure 49.4 A view of the interstitial cells of the rat testis. Several blood vessels are shown with associated pericytes. Leydig cells are indicated (L) and the prostaglandin secreting macrophages are denoted with arrows. Portions of the seminiferous tubules can be seen on three sides of the photograph. Photograph kindly supplied by Drs R.M. Sharpe and J.B. Kerr
anergy (Parijs et al 1997) and PGE is very effective at ablating IL12 release (Kraan et al 1995). The role of PGE in tolerance is not easily assessed by depletion. Indomethacin breaks oral tolerance (Louis et al 1996) but in the experiments where selective knockout of PGHS-1 and PGHS-2 was achieved there was suggestion of a compensation of one isoform for the other (Lagenbach et al 1995). In addition, NSAIDs may induce gastric ulceration by an effect on immune tolerance. This has always been likely to occur via the PGHS-2 enzyme isoform and consequently has largely been ignored since PGHS-1 was thought to be the beneficial regulatory enzyme. A recent reappraisal of this status (Wallace 1999) suggests a possible inducible role of PGHS-2 in oral tolerance. THE RELEVANCE OF PROSTAGLANDINS TO SEXUALLY TRANSMITTED DISEASE The powerful immunosuppressive properties of human seminal plasma will clearly affect the ability of the recipient female
522
REPRODUCTIVE SYSTEM
immune system to respond to infection in semen with a cellmediated response. In the case of viral infection, the intracellular organism is usually eliminated by killing the infected cell, although in the case of infection of vital organs such as the liver, intracellular virus can be eliminated by a cytokine-mediated mechanism that does not involve cell death (Guidotti and Chisari 1996). Even if cell death is not involved, virus can still only be eradicated by a cytokine profile favouring cell-mediated responses (Th1). The effect of human seminal plasma in restricting cellmediated responses will therefore seriously restrict the host immune system’s ability to cope with semen-borne infection. Clearly the female reproductive tract cannot have responses totally ablated by semen but it is likely that certain organisms have taken advantage of the protection provided by the immunosuppressive activity in human seminal plasma. One defence against the transmission of virus in semen is the provision of a low pH due to a natural colonization of the vagina with lactobacilli. Another protection is the natural antibiotics, such as the defensins, which are remarkably abundant in the lower female tract (Quayle et al 1998; Valore et al 1998) and these will protect against infection ascending through the cervix. There is, however, also evidence for antimicrobial activity in semen, since secretory leukocyte protease inhibitor (SLPI), which is abundant in human semen (Ohlsson et al 1995), has been shown to have antifungal, antibacterial and antiviral activity (McNeely et al 1995; Tomee et al 1998). Human seminal plasma may therefore partly restrict the immune system in the female reproductive tract but at the same time it compensates by the provision of antimicrobial activity, and the ratio of these two actions may prove important in the dissemination of sexually transmitted disease. CONCLUSIONS The volume of evidence that supports an immunomodulatory role for human seminal plasma is now substantial. However, there are many questions to be answered and these may have to await a better understanding of the role of prostaglandin in controlling the immune system. First, why is it that monocytes produce both PGE and PGF? Several roles for PGE have been established but why is PGF there also? The identified roles of PGF as a smooth muscle contracting agent and a luteolytic in certain species do not sit easily with the production by the macrophage in the testis or elsewhere. Second, it is of interest to know whether 19-hydroxyPGE appears elsewhere than in human semen; it would be curious in the extreme if nature had derived a new prostaglandin solely for its action in semen. Third, what is the interaction between the suppressive molecules secreted by monocytes and other antigenpresenting cells? The current list includes IL-10, TGFb, IL-1 receptor antagonist and PGE. It would be particularly relevant to know what the relationship is between PGE and TGFb, since both have such a similar spectrum of activity. Last, what are the agents controlling prostaglandin secretion from the seminal vesicle? Is the seminal vesicle cyclooxygenase enzyme inducible and how is it regulated by androgen? An added unknown is whether the p450 enzyme that 19-hydroxylates in the primate seminal vesicle (Bylund et al 1999, 2000; Oliw et al 1988) is inducible with steroid? There are thus many questions left unanswered at this stage and with these answers relevant to areas as diverse as transplant immunology, sexually transmitted disease and conditions of poor implantation, such as preeclampsia, solutions should be urgently sought. REFERENCES Aitken RJ and Kelly RW (1985) Analysis of the direct effects of prostaglandins on human sperm function. J Reprod Fert, 73, 139–146.
Aitken RJ, Irvine S and Kelly RW (1986) Significance of the intracellular calcium and cyclic adenosine 3,5-monophosphate in the mechanism by which prostaglandins influence human sperm function. J Reprod Fert, 77, 452–462. Aitken RJ, West K and Buckingham D (1994) Leukocyte infiltration into the human ejaculate and its association with semen quality, oxidative stress and sperm function. J Androl, 15, 343–352. Alexander NJ and Anderson DJ (1987) Immunology of semen. Fertil Steril, 47, 192–205. Bachmann MF, Zinkernagel RM and Oxenius A (1998) Immune responses in the absence of costimulation: viruses know the trick. J Immunol, 161, 5791–5794. Bankhurst AD (1982) The modulation of human natural killer cell activity by prostaglandins. J Clin Lab Immunol, 7, 85–91. Bendvold E, Svanborg K, Eneroth P et al (1984) The natural variations in prostaglandin concentration in human seminal fluid and its relation to sperm quality. Fert Steril, 41, 743–747. Bendvold E, Svanborg K, Bygdeman M and Noren S (1985) On the origin of prostaglandins in seminal fluid. Int J Fertil, 8, 37–43. Bendvold E, Gottlieb C, Svanborg K et al (1987) Concentration of prostaglandins in seminal fluid of fertile men. Int J Androl, 10, 463– 469. Betz M and Fox BS (1991) Prostaglandin E2 inhibits production of Th1 lymphokines but not of Th2 lymphokines. J Immunol, 146, 108–113. Brunda MJ, Herberman RB and Holden HT (1980) Inhibition of natural killer cell activity by prostaglandins. J Immunol, 124, 2682–2687. Bylund J, Finnstrom N and Oliw EH (1999) Gene expression of a novel cytochrome P450 of the CYP4F subfamily in human seminal vesicles. Biochem Biophys Res Commun, 261, 169–174. Bylund J, Hidestrand M, Ingelman-Sundberg M and Oliw EH (2000) Identification of CYP4F8 in human seminal vesicles as a prominent 19hydroxylase of prostaglandin endoperoxides. J Biol Chem, 275, 21844– 21849. Carpenter MP (1974) Prostaglandins in the rat testis. Lipids, 9, 397–406. Cheuk BL, Leung PS, Lo AC and Wong PY (2000) Androgen control of cyclooxygenase expression in the rat epididymis. Biol Reprod, 63, 775– 780. Chouaib S and Fradelizi D (1982) The mechanism of inhibition of human IL-2 production. J Immunol, 129, 2463–2468. Coleman RA, Kennedy I, Humphrey PP et al (1990) Prostanoids and their receptors. In Hansch C, Sammes PG and Taylor JB (eds) Comprehensive Medicinal Chemistry. Oxford: Pergamon, 643–714. Daya S, Rosenthal KL and Clark DA (1987) Immunosuppressor factor(s) produced by decidua-associated suppressor cells: a proposed mechanism for fetal allograft survival. Am J Obstet Gynecol, 156, 344–350. Ding L and Shevach EM (1992) IL-10 inhibits mitogen-induced T cell proliferation by selectively inhibiting macrophage co-stimulatory function. J Immunol, 148, 3133–3139. Ding L, Linsley PS, Huang LY et al (1993) IL-10 inhibits macrophage costimulatory activity by selectively inhibiting the upregulation of B7 expression. J Immunol, 151, 1224–1234. Droller MJ, Schneider MU and Perlmann P (1978) A possible role of prostaglandins in the inhibition of natural and antibody dependent lymphocyte cytotoxicity against tumor cells. Cell Immunol, 39, 165– 177. Fernandez-Botran R, Sanders VM, Mosmann TR and Vitetta ES (1988) Lymphokine-mediated regulation of the proliferative response of clones of T helper 1 and T helper 2 cells. J Exp Med, 168, 543–558. Fischer A, Durandy A and Griscelli C (1981) Role of prostaglandin E2 in the induction of non-specific T lymphocyte suppressor activity. J Immunol, 126, 1452–1455. Gerozissis K and Dray F (1983) In vitro prostanoid production by rat vas deferens. J Reprod Fertil, 67, 389–394. Gerozissis K, Jouannet P, Soufir JC and Dray F (1982) Origin of prostaglandins in human semen. J Reprod Fertil, 65, 401. Gottlieb C, Svanborg K, Eneroth P and Bygdeman M (1988) Effect of prostaglandins on human sperm function in vitro and seminal adenosine triphosphate concentration. Fertil Steril, 49, 322–327. Groux H, Bigler M, de Vries JE and Roncarolo MG (1996) Interleukin-10 induces a long-term antigen-specific anergic state in human CD4+ T cells. J Exp Med, 184, 19–29.
PROSTAGLANDINS AND MALE REPRODUCTIVE PHYSIOLOGY Groux H, O’Garra A, Bigler M et al (1997) A CD4(+) T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature, 389, 737–742. Guidotti LG and Chisari FV (1996) To kill or to cure: options of host defense against viral infection. Curr Opin Immunol, 8, 478–483. Hamberg M (1976) Biosynthesis of prostaglandin E1 by human seminal vesicles. Lipids, 11, 249–250. Hamberg M and Samuelsson B (1966) Prostaglandins in human seminal plasma. J Biol Chem, 241, 257–263. Head JR, Neaves WB and Billingham RE (1983) Immune privilege in the testis. Transplantation, 36, 423–431. Hilkens CMU, Vermeulen H, van Neerven RJJ et al (1995) Differential modulation of T helper type 1 (Th1) and T helper type 2 (Th2) cytokine secretion by prostaglandin E2 critically depends on interleukin-2. Eur J Immunol, 25, 59–63. Huhtaniemi I, Nikula H, Lehtonen E and Hovatta O (1986) Testicular morphology and endocrine function after silica-induced damage of interstitial macrophages. In Stefanini M, Conti M, Geremia R and Ziparo E (eds), Molecular and Cellular Endocrinology of the Testis. Amsterdam: Elsevier, 57–62. Iglesias BM, Cerase J, Ceracchini C et al (1997) Analysis of B7-1 and B7-2 costimulatory ligands in cultured mouse microglia: Upregulation by interferon-g and lipopolysaccharide and downregulation by interleukin-10, prostaglandin E(2) and cyclic AMP-elevating agents. J Neuroimmunol, 72, 83–93. James K and Hargreave TB (1984) Immunosuppression by seminal plasma and its possible clinical significance. Immunol Today, 5, 357–363. James K, Harvey J, Bradbury AW et al (1983) The effect of seminal plasma on macrophage function—a possible contributory factor in sexually transmitted disease. AIDS Res, 1, 45–57. Jondal M, Merrill J and Ullberg M (1981) Monocyte induced human natural killer cell suppression followed by increased cytotoxic activity during short term in vitro culture with autologous serum. Scand J Immunol, 14, 555–563. Jonsson HT, Middleditch BS and Desiderio DM (1975) Prostaglandin in human seminal fluid: two novel compounds. Science, 187, 1093–1094. Katamura K, Shintaku N, Yamauchi Y et al (1995) Prostaglandin E2 at priming of naive CD4+ T cells inhibits acquisition of ability to produce IFNg and IL-2, but not IL-4 and IL-5. J Immunol, 155, 4604–4612. Kelly RW (1995) Immunosuppressive mechanisms in semen: implications for contraception. Hum Reprod, 10, 1686–1693. Kelly RW (1997) Prostaglandins in primate semen: biasing the immune system to benefit spermatozoa and virus. Prostagland Leukotrienes Essent Fatty Acids, 57, 113–118. Kelly RW and Critchley HOD (1997) Immunomodulation by human seminal plasma: a benefit for spermatozoa and pathogen? Hum Reprod, 12, 2200–2207. Kelly RW, Taylor PL, Hearn JP et al (1976) 19-Hydroxy prostaglandin E1 as a major component of the semen of primates. Nature, 260, 544–555. Kelly RW, Carr GC and Critchley HOD (1997) A cytokine switch induced by human seminal plasma: an immune modulation with implications for sexually transmitted disease. Hum Reprod, 12, 677–681. Kern S and Maddocks S (1995) Indomethacin blocks the immunosuppressive activity of rat testicular macrophages cultured in vitro. J Reprod Immunol, 28, 189–201. Kern S, Robertson SA, Mau VJ and Maddocks S (1995) Cytokine secretion by macrophages in the rat testis. Biol Reprod, 53, 1407–1416. Khoury SJ, Hancock WW and Weiner HL (1992) Oral tolerance to myelin basic-protein and natural recovery from experimental autoimmune encephalomyelitis are associated with downregulation of inflammatory cytokines and differential upregulation of transforming growth factorb, interleukin-4, and prostaglandin-E expression in the brain. J Exp Med, 176, 1355–1364. Kirschenbaum A, Liotta DR, Yao S et al (2000) Immunohistochemical localization of cyclooxygenase-1 and cyclooxygenase-2 in the human fetal and adult male reproductive tracts. J Clin Endocrinol Metab, 85, 3436–3441. Kraan TCTMV, Boeije LCM, Smeenk RJT et al (1995) Prostaglandin E2 is a potent inhibitor of human interleukin-12 production. J Exp Med, 181, 775–779. Ksander BR and Streilein JW (1994) Regulation of the immune response within privileged sites. In Granstein RD (ed.), Mechanisms of Immune Regulation, vol 5. Basel: Karger, 117–145.
523
Kunkel SL, Chensue SW and Phan SH (1986a) Prostaglandins as endogenous mediators of interleukin 1 production. J Immunol, 136, 186–192. Kunkel SL, Wiggins RC, Chensue SW and Larrick J (1986b) Regulation of macrophage tumor necrosis factor production by prostaglandin E2. Biochem Biophys Res Commun, 137, 404–410. Kurzrock R and Liebb CC (1930) Biochemical studies of human semen: II. The action of semen on the human uterus. Proc Soc Exp Biol Med, 28, 268–272. Lagenbach R, Morham SG, Tiano HF et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Li J, Yang Y, Inoue H et al (1996) The expression of co-stimulatory molecules CD80 and CD86 in human carcinoma cell lines: its regulation by interferon-g and interleukin-10. Cancer Immunol Immunother, 43, 213–219. Loke YW and King A (1996) Immunology of human implantation: an evolutionary perspective. Hum Reprod, 11, 283–286. Lord EM, Sensabaugh GF and Stites DP (1977) Immunosuppressive activity of human seminal plasma 1, Inhibition of in vitro lymphocyte activation. J Immunol, 118, 1704–1711. Louis E, Franchimont D, Deprez M et al (1996) Decrease in systemic tolerance to fed ovalbumin in indomethacin-treated mice. Int Arch Allerg Immunol, 109, 21–26. Malefyt Rd, Yssel H, Roncarolo M et al (1992) Interleukin-10. Curr Opin Immunol, 4, 314–320. Matzinger P (1994) Tolerance, danger, and the extended family. Ann Rev Immunol, 12, 991–1045. McKanna JA, Zhang MZ and Wang JL et al (1998) Constitutive expression of cyclooxygenase-2 in rat vas deferens. Am J Physiol, 275, R227–233. McNeely TB, Dealy M, Dripps DJ et al (1995) Secretory leukocyte protease inhibitor—a human saliva protein exhibiting anti-human immunodeficiency-virus-1 activity in vitro. J Clin Invest, 96, 456–464. Medzhitov R and Janeway CA (1997) Innate immunity: impact on the adaptive immune response. Curr Opin Immunol, 9, 4–9. Middleditch BS (1975) PGA: fact or artifact? Prostaglandins, 9, 409–411. Mosmann TR (1988) Directional release of lymphokines from T cells. Immunol Today, 9, 306–307. Munn DH, Zhou M, Attwod JT et al (1998) Prevention of allogeneic fetal rejection by tryptophan catabolism. Science, 281, 1191–1193. Ohlsson K, Bjartell A and Lilja H (1995) Secretory leukocyte protease inhibitor in the male genital tract—PSA-induced proteolytic processing in human semen and tissue localization. J Androl, 16, 64–74. Oliw EH, Kinn A and Kvist U (1988) Biochemical characterization of prostaglandin 19-hydroxylase of seminal vesicles. J Biol Chem, 263, 7222–7227. Orlikowsky T, Wang ZQ, Dudhane A et al (1996) The cell-surface marker phenotype of macrophages from HIV-1-infected subjects reflects an IL-10-enriched and IFN-g-deprived donor environment. J Interferon Cytokine Res, 16, 957–962. Pandya IJ and Cohen J (1985) The leukocytic reaction of the human uterine cervix to spermatozoa. Fertil Steril, 43, 417–421. Parijs LV, Perez VL, Biuckians A et al (1997) Role of interleukin-12 and co-stimulators in T cell anergy in vivo. J Exp Med, 186, 1119–1128. Prakash C, Coutinho A and Moller G (1976) Inhibition of in vitro immune responses by a fraction from seminal plasma. Scand J Immunol, 5, 77– 88. Quayle AJ, Kelly RW, Hargreave TB and James K (1989) Immunosuppression by seminal prostaglandins. Clin Exp Immunol, 75, 387–391. Quayle AJ, Porter EM, Nussbaum AA et al (1998) Gene expression, immunolocalization, and secretion of human defensin-5 in human female reproductive tract. Am J Pathol, 152, 1247–1258. Rappaport RS and Dodge GR (1982) Prostaglandin E inhibits the production of interleukin-2. J Exp Med, 155, 943–948. Ren J, Singh AK, Gregerson DS and Shichi H (1996) Induction of immunotolerance in rats by intratesticular administration of an eicosapeptide of bovine S-antigen. Autoimmunity, 25, 19–31. Robertson SA, Mau VJ, Hudson SN and Tremellen KP (1997) Cytokine– leukocyte networks and the establishment of pregnancy. Am J Reprod Immunol, 37, 438–442.
524
REPRODUCTIVE SYSTEM
Samuelsson B (1963) Isolation and identification of prostaglandins from human seminal plasma. J Biol Chem, 238, 3229–3234. Schaefer M, Hofmann T, Schultz G and Gudermann T (1998) A new prostaglandin E receptor mediates calcium influx and acrosome reaction in human spermatozoa. Proc Natl Acad Sci USA, 95, 3008– 3013. Setchell BP, Uksila J, Maddocks S and Po¨lla¨nen P (1990) Testis physiology relevant to immuneregulation. J Reprod Immunol, 18, 19– 32. Shivaji S, Scheit K-H and Bhargava PM (1990) Immunosuppressive factors in seminal plasma. In Shivaji S, Scheit K-H and Bhargava PM (eds), Proteins of Seminal Plasma. New York: Wiley, 375–389. Skakkebaek NE, Kelly RW and Corker CS (1976) Prostaglandin concentrations in the semen of hypogonadal men during treatment with testosterone. J Reprod Fert, 47, 119–121. Snijdewint FG, Kalinski P, Wierenga EA et al (1993) Prostaglandin E2 differentially modulates cytokine secretion profiles in human T helper lymphocytes. J Immunol, 150, 5321–5329. Stites DP and Erickson RP (1975) Suppressive effect of seminal plasma on lymphocyte activation. Nature, 253, 727–729. Strassmann G, Patil Koota V, Finkelman F et al (1994) Evidence for the involvement of interleukin-10 in the differential deactivation of murine peritoneal macrophages by prostaglandin E2. J Exp Med, 180, 2365– 2370. Taga K and Tosato G (1992) IL-10 inhibits human T cell proliferation and IL-2 production. J Immunol, 148, 1143–1148. Tarter TH, Cunningham-Rundles S and Koide SS (1986) Suppression of natural killer cell activity by human seminal plasma in vitro: identification of 19-OH-PGE as the suppressor factor. J Immunol, 136, 2862–2867. Taylor PL and Kelly RW (1974) 19-hydroxylated E prostaglandins as the major prostaglandins of human semen. Nature, 250, 665–667. Taylor PL and Kelly RW (1975) The occurrence of 19-hydroxy F prostaglandins in human semen. FEBS Lett, 57, 22–25. Templeton AA, Cooper I and Kelly RW (1978) Prostaglandin concentrations in the semen of fertile men. J Reprod Fertil, 52, 147– 150.
Tomee JFC, Koeter GH, Hiemstra PS and Kauffman HF (1998) Secretory leukoprotease inhibitor: a native antimicrobial protein presenting a new therapeutic option? Thorax, 53, 114–116. Trinchieri G (1993) IL-12 and its role in the generation of Th1 cells. Immunol Today, 14, 335–338. Tripiciano A, Filippini A, Ballarini F and Palombi F (1998) Contractile response of peritubular myoid cells to prostaglandin F2a. Mol Cell Endocrinol, 138, 143–150. Vallely PJ, Sharrard RM and Rees RC (1988) The identification of factors in seminal plasma responsible for suppression of natural killer cell activity. Immunology, 63, 451–456. Valore EV, Park CH, Quayle AJ et al (1998) Human b-defensin-1: an antimicrobial peptide of urogenital tissues. J Clin Invest, 101, 1633– 1642. Von Euler US (1936) On the specific vasodilating and plain muscle stimulating substances from accessory genital glands in man and certain animals (prostaglandin and vesiglandin). J Physiol (Lond), 88, 213–234. Wahl LM, Olsen CE, Wahl SM et al (1979) Prostaglandin and cAMP regulation of macrophage involvement in connective tissue destruction. Ann NY Acad Sci, 332, 271–278. Wallace JL (1999) Selective COX-2 inhibitors: is the water becoming muddy? Trends Pharmacol Sci, 20, 4–6. Weiner HL, Friedman A, Miller A et al (1994) Oral tolerance: immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral administration of autoantigens. Ann Rev Immunol, 12, 809–837. Weissman G, Zurier RB and Hofstein S (1973) Leukocytes as secretory organs of inflammation. Agents Actions, 3, 370–379. Wolff H, Politch JA, Martinez A et al (1990) Leukocytospermia is associated with poor semen quality. Fertil Steril, 53, 528–536. Wongkham S, Tanphaichitr N and Anderson DJ (1991) The effect of polymorphonuclear leukocytes on human sperm. J Sci Soc Thailand, 17, 71–76. Yotsukura J, Huang H, Singh AK and Shichi H (1997) Regulatory cells generated by testicular tolerization to retinal S-antigen: possible involvement of IL-4, IL-10, and TGF-b in the suppression of experimental autoimmune uveoretinitis. Cell Immunol, 182, 89–98.
50 Prostaglandin F2a: the Luteolytic Hormone John A. McCracken University of Connecticut, Storrs, CT, USA
Luteolysis (or corpus luteum regression, demise or involution) is the process by which the corpus luteum (in monoovulatory species) or corpora lutea (in polyovulatory species) undergo regression. The corpus luteum (Figure 50.1) is formed in the ovary from the cells surrounding the ovulatory follicle(s) after ovulation has occurred under the influence of luteinizing hormone (LH). The corpus luteum is a unique tissue because of its transient existence as an endocrine gland. The only other endocrine structures that function in such a temporary manner are preovulatory follicles and the placenta, although the latter usually has a longer functional lifespan. The principal secretory product of the corpus luteum is the steroid hormone progesterone, which maintains pregnancy in mammals by inducing uterine quiescence and by suppressing maternal immune responses to foetal antigens. In addition, progesterone influences tubal transport of fertilized ova and creates a uterine milieu which supports blastocyst development and implantation. Progesterone also reduces cyclic ovarian activity in most mammals during pregnancy and also contributes to mammary development. If conception does not occur in species with a functional cyclic luteal phase (which includes primates, domestic animals and the guinea-pig), the corpus luteum undergoes luteolysis (involution or regression), so that a new ovarian cycle can be initiated and, hence, another opportunity for conception is achieved. The term ‘‘luteolysis’’ is something of a misnomer, since lysis of luteal cells does not actually occur. Instead, as luteolysis progresses, cells of the corpus luteum are gradually removed by the process of programmed cell death known as apoptosis (see Mechanisms of Luteolytic Action, below). However, the term ‘‘luteolysis’’ is now firmly entrenched in the literature and its use will no doubt continue. The first most obvious indication of the onset of luteolysis is a marked decline in peripheral plasma progesterone due to a decrease in its secretion from the corpus luteum. Such a decline in progesterone production is generally termed ‘‘functional luteolysis’’ as distinct from ‘‘structural luteolysis’’ which, as the name suggests, signifies changes in the cellular composition of the gland and its involution to form the corpus albicans, a small residual structure in the ovary composed of connective tissue and collagen (Figure 50.1). However, the distinction between functional and structural luteolysis is not clear and the two events may be part of a continuum. Indeed, emerging evidence suggests that biochemical events controlling the extracellular matrix, and hence the cellular integrity of the gland, may be one of the earliest events of luteolysis (see below). The relative occurrence of luteolysis in different types of mammalian reproductive cycles is shown in Figure 50.2. Each type of cycle is described briefly below. The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
Primate Menstrual Cycles In primates, ovulation and corpus luteum formation occurs under gonadotrophic control. The corpus luteum in primates produces both progesterone and oestrogen. Towards the end of the luteal phase of the cycle, the corpus luteum undergoes luteolysis, thus terminating the cycle. Unlike most mammals, luteolysis, in primates, is not dependent on the presence of the uterus (see later) and primate luteolysis is currently considered to involve asyet unidentified intraovarian factors (see later). In Old World primates, including humans, the decline in progesterone due to luteolysis causes shedding of the endometrium, which results in menstruation, an external sign of the end of the reproductive cycle. Menstruation occurs only in Old World primates including humans since, unlike New World primates, a spiral arteriolar complex exists in the endometrium, the contraction of which induces menstruation. In non-human primates, the mid-cycle increase in oestrogen from the preovulatory follicle induces an increase in sexual receptivity (oestrus), whereas in the human female there is no overt increase in sexual receptivity at this time. Thus, non-human primates exhibit a transitional form of both oestrous and menstrual cycles. When conception occurs in primates, the lifespan of the corpus luteum is extended by chorionic gonadotrophin secreted by the implanting blastocyst. Pregnancy is maintained initially by progesterone from the corpus luteum until the placenta assumes this function (the luteo–placental shift). Domestic Animal Cycles As in primates, ovulation and corpus luteum formation are initiated by pituitary gonadotrophins. During a relatively short follicular phase, increasing oestrogen from the preovulatory follicle induces a brief period of sexual receptivity (oestrus) as well as inducing the preovulatory surge of LH. Thus, the time of mating is synchronized with the time of ovulation. If the cycle is infertile, the corpus luteum undergoes luteolysis, which terminates the luteal phase and a new ovarian cycle is initiated. If conception occurs, the luteolytic signal emanating from the uterus is subverted by the blastocyst and the corpus luteum continues to secrete progesterone for pregnancy maintenance. In some species, e.g. mouse and goat, the corpora lutea remain the sole source of progesterone throughout gestation and the initiation of labour (parturition) in these species may involve a luteolytic mechanism. In other species, such as the sheep and horse, the placenta later becomes the principal source of progesterone for pregnancy
526
REPRODUCTIVE SYSTEM
maintenance (the luteo–placental shift). The guinea-pig is included in this category because its reproductive cycle is very similar to domestic animals.
Since ovulation and conception are achieved only by a mating stimulus, induced ovulators are reproductively quite efficient. Canine Cycle
Laboratory Rodent Cycles In rodents, e.g. rat, mouse and hamster, corpora lutea form after ovulation but a mating stimulus is required to cause the corpora lutea to become functional under the influence of pituitary prolactin. If conception occurs the corpora lutea persist throughout pregnancy (22 days), but if mating is infertile, the corpora lutea become temporarily functional, i.e. pseudopregnancy. However, most rodents develop a luteolytic mechanism which curtails the length of pseudopregnancy. Since rodents do not spontaneously develop a functional luteal phase after ovulation unless they are mated, rodents are reproductively efficient, i.e. they have an opportunity to conceive every 4–5 days.
Induced Ovulators The induced or reflex ovulators do not have repetitive cycles with distinct follicular and luteal phases. Instead, species such as the rabbit, cat, mink, ferret and camel have waves of follicular development during which there is a relatively long period of sexual receptivity. A mating stimulus is required for the neurogenic release of LH and thus the induction of ovulation and corpus luteum formation. Corpora lutea are maintained in both fertile (pregnancy) and in infertile matings (pseudopregnancy). The latter may last as long as a normal pregnancy (e.g. ferret) or be terminated by a luteolytic mechanism (e.g. rabbit).
Canines (e.g. dog and wolf) usually exhibit only two ovulatory cycles per year and ovulation and corpora lutea are formed spontaneously under gonadotrophic control. Corpora lutea secrete progesterone in both fertile and infertile matings and in the latter case (pseudopregnancy) progesterone production persists for a period approximating that of pregnancy. It appears that a luteolytic mechanism has not developed in these animals to shorten the lifespan of corpora lutea when pregnancy is not established, and no signal is required from the embryo to extend the lifespan of corpora lutea during gestation. Thus, canines are not very efficient reproductively, since they do not have repetitive cycles and only have a chance to conceive about twice per year. Comparative aspects of corpus luteum development and function have been reviewed previously (McCracken and Schramm 1988; Keyes and Wiltbank 1988; Auletta and Flint 1988; Behrman et al 1993; Poyser 1995; McCracken et al 1999; Niswender et al 2000). IDENTIFICATION OF PGF2a AS THE MAMMALIAN LUTEOLYTIC HORMONE The quest for an explanation of luteolysis was initiated by the pioneering studies of Leo Loeb in the 1920s. Working with his favorite model, the guinea-pig, Loeb observed that hysterectomy during the luteal phase prevented regression of corpora lutea at the end of the cycle (Loeb 1923, 1927). This discovery by Loeb opened up a new dimension concerning the regulation of corpus
Figure 50.1 Diagram of ovary showing the development of a ruptured ovulatory follicle into a corpus luteum and its transformation during luteolysis into a corpus albicans. From Patten in Patten’s Human Embryology (1976), with permission of Blakiston Division of McGraw-Hill Book Co.
LUTEOLYSIS luteum function. Especially perplexing at that time was the fact that it was known that hysterectomy did not have any effect on ovarian cyclicity in women. Indeed, the role of the uterus in regulating the lifespan of the corpus luteum in non-primate species proved to be elusive. Some 12 years after Loeb’s discovery, Marx (1935) expressed the then-current view that ‘‘Few problems in endocrinology have been as resistant to clarification’’. It should be remembered that at the time of Loeb’s discovery progesterone had not yet been identified and even the existence of prostaglandins was not known. Progesterone was not identified and synthesized until 1934 (Slotta et al 1934; Butenandt and Westphal 1934) and the prostaglandins not until 1957 (Bergstro¨m and Sjo¨vall 1957), shortly before Loeb’s death. Leo Loeb (1869–1959) was a medical pathologist from Germany who had trained in several universities throughout Europe before migrating to North America in 1897. After spells in Chicago, Montreal and Pennsylvania, in 1915 he settled in Washington University in St. Louis, where he worked for the rest of his career. He died at the
527
age of 91 after having published over 400 papers. He is credited with being one of the founders of experimental pathology and cancer research in North America (Corner 1960). His major interests were in wound healing and the experimental transplantation of tumours. Remarkably, in a paper on the corpus luteum of the guinea-pig (Loeb 1906) he likened the corpus luteum to a kind of temporary tumour by commenting that ‘‘the corpus luteum can be considered to be a transitory tumour, the study of which is, therefore, of general pathological interest’’. This was quite prophetic, because recent work on the regulation of the extracellular matrix of the corpus luteum during luteolysis (Towle et al 2002) indicates some common ground with the regulation of tumorigenesis and its prevention (see Mechanisms of Luteolytic Action, below). A similar persistence of corpora lutea after hysterectomy was reported subsequently in most of the domestic animals and in a variety of pseudopregnant rodents and induced ovulators (Anderson et al 1969). These species, unlike primates, all have a
Figure 50.2 The role of luteolysis in controlling the lifespan of the corpus luteum during the reproductive cycle or during pseudopregnancy in different species (see text for details). From McCracken et al (1999), by permission of the American Physiological Society
528
REPRODUCTIVE SYSTEM
bicornate uterus. In most of them, the absence of one uterine horn (either congenitally or surgically) results in the persistence of the corpora lutea only in the ovary adjacent to the absent uterine horn. However, in primates, congenital absence of the uterus, or its surgical removal, has no effect on the lifespan of corpora lutea (see McCracken et al 1999). The failure of corpora lutea to regress normally at the end of the luteal phase following hysterectomy in non-primates indicated that the uterus produced a factor which caused luteolysis at the end of the cycle. Selective ligation of ovarian and uterine blood vessels of the ovary and uterus indicated that, in many of these species, vascular connections were essential for the local transport of the putative luteolysin from the uterus to the ovary (see McCracken et al 1973). Moreover, in the sheep, transplantation of one ovary with vascular anastomoses to a jugulo-carotid skin loop, with the uterus remaining in situ in the abdomen, prevented luteolysis (Goding and McCracken 1966; Goding et al 1967). Transplantation of one ovary and its contiguous uterine horn to a jugulocarotid skin loop resulted in regular cyclical regression of the corpus luteum (McCracken et al 1969a, 1970a, 1971). These findings proved unequivocally that the luteolytic factor from the uterus in the sheep acted locally on the ovary and could not be transmitted via the systemic circulation. Possible candidates for such a luteolytic factor were the prostaglandins (PGs), acidic lipids originally isolated from seminal plasma of sheep and several other species (Bergstro¨m and Sjovall 1957). Two of these new substances, PGF2a and PGE2, were soon identified in human menstrual fluid (Eglinton et al 1963) and were shown to be
produced in large quantities by human endometrium (Pickles 1967). Based on their abundance in the uterus and their known vasoactive properties, prostaglandins were suggested as potential candidates for the uterine luteolytic factor. High levels of PGF2a administered subcutaneously to female rats caused a shortening of pseudopregnancy and a reduction in the progesterone content of the corpora lutea, although the corpora lutea did not actually regress (Pharriss and Wyngarden 1969). In the rat study, it was unclear whether the administered PGF2a acted directly on the corpora lutea or, as the authors suggested, indirectly via the pituitary. However, the infusion of very small amounts of PGF2a directly into the arterial supply of the transplanted ovary of the sheep caused a sustained decline in ovarian progesterone secretion, followed by the onset of oestrus and an LH peak, while the same amount of PGF2a infused systemically had no effect on ovarian progesterone secretion (McCracken et al 1970b; McCracken 1971). Because PGF2a was known to be rapidly metabolized in the peripheral circulation, particularly by the lungs (Ferreira and Vane 1967), these infusion experiments in the sheep strengthened the proposal that the locally acting uterine luteolytic factor in the sheep might be PGF2a. Because of its known rapid clearance rate in the systemic circulation, PGF2a would only be effective if delivered locally to the ovary from the uterus. In the sheep, the ovarian artery is very convoluted and closely apposed to the surface of the utero-ovarian vein before entering the hilus of the ovary (Figure 50.3). This curious vascular arrangement is also conspicuous in the cow and had been suggested as having some physiological function (Vollmerhaus 1964). Subsequent
Figure 50.3 Anatomical relationship in situ of ovarian artery and utero-ovarian vein of the sheep and the experimental design used to demonstrate the countercurrent transfer of 3H-PGF2a from a utero-ovarian vein to its adherent ovarian artery. About 1% of PGF2a secreted by the uterus reaches the ovary directly without passing through the pulmonary circulation. From McCracken et al (1972), reproduced by permission from Nature New Biol, 238, 129– 134, # 1972 Macmillan Magazines Ltd
LUTEOLYSIS
Figure 50.4 Time course of countercurrent transfer of 3H-PGF2a from a utero-ovarian vein to its adherent ovarian artery. From McCracken et al (1972), reproduced by permission from Nature New Biol, 238, 129–134, # 1972 Macmillan Magazines Ltd
experiments indicated that a small fraction (about 1%) of radioactively labelled PGF2a infused into a uterine vein of the sheep passed directly via counter-current transfer into the closely adherent ovarian artery and, thus, bypassed the systemic route (McCracken et al 1971, 1972; see Figure 50.4). Such a mechanism
529
provided an explanation for the local effect on the ovary of the putative uterine luteolysin PGF2a. Recently, high concentrations of the PG transporter have been identified in the bovine uterine vein/ovarian artery juxtaposition, as well as in other reproductive tissues (Banu et al 2003). The concentration of the PG transporter, which is selectively expressed in vascular smooth muscle cells, was higher in the utero-ovarian vascular complex on the side where the ovary contained the corpus luteum, especially in the late luteal phase. Such a finding may explain the mechanism underlying the counter current transfer of PGF2a in the utero-ovarian vascular pedicle of ruminants and other species. However, it remained to be determined whether PGF2a was actually secreted by the uterus at the time of luteolysis in sheep. Using a combination of gas chromatography and mass spectrometry, PGF2a was identified and measured in uterine venous blood collected over the time of luteolysis in sheep. PGF2a concentration increased some 20-fold in uterine vein blood of the sheep coincident with the onset of luteolysis (McCracken et al 1971, 1972). Moreover, when the calculated amount of PGF2a produced by the uterus during luteolysis was infused into a uterine vein adjacent to the ovary containing the corpus luteum, premature luteolysis was consistently observed, but not when the same amount was infused into the peripheral circulation (McCracken 1971; McCracken et al 1972). These uterine vein infusions of physiological levels of PGF2a not only confirmed the local counter-current transfer of PGF2a in the utero-ovarian vasculature, but also proved the identity of PGF2a as the local uterine luteolytic hormone in the sheep. Further studies in the sheep, with high frequency sampling of uterine vein blood, revealed that PGF2a was secreted by the uterus as a series of five to eight pulses (see Figure 50.5) each lasting about 1 h and occurring at intervals of 6–9 h over the course of luteolysis (McCracken et al 1973; Thorburn et al 1973; Barcikowski et al 1974). The infusion of low
Figure 50.5 Utero-ovarian vein plasma concentration of PGF2a, oestradiol-17b, progesterone and LH in blood sampled every 2 h from sheep bearing a utero-ovarian autotransplant over the course of luteolysis. From Barcikowski et al (1974). Copyright 1974. The Endocrine Society
530
REPRODUCTIVE SYSTEM
levels of PGF2a, in the form of 1 h pulses into the arterial supply of the sheep ovary, indicated that the pulsatile infusion of PGF2a caused luteolysis with only 1/40th of the minimal amount of PGF2a required to cause luteolysis by continuous infusion (Schramm et al 1983). Thus, a pulsatile pattern of PGF2a appears to be advantageous for luteolysis. Later, pulses of the primary metabolite of PGF2a, 15-keto-13,14-dihydro-PGF2a (PGFM), were observed in the peripheral blood of sheep over the course of luteolysis, which occurred at the same frequency as pulses of PGF2a in uterine vein blood (Kindahl et al 1976a; Peterson et al 1976). The role of PGF2a as a uterine luteolytic hormone was further supported by the finding that administration of indomethacin to several different species, either systemically or via the intrauterine route, delayed or prevented luteolysis (Lewis and Warren 1977). Also, immunization against PGF2a, either passively (Fairclough et al 1976, 1981) or actively (Scaramuzzi and Baird 1976; Ronayne et al 1990), was found to delay luteolysis in the sheep. The identity of PGF2a as the uterine luteolytic hormone has been confirmed in several other species, including the cow, goat, pig, mare, reindeer and guinea-pig. For data on the pulsatile release of PGF2a in these species, see McCracken et al (1999). Also, a luteolytic role for PGF2a as a means of curtailing the length of pseudopregnancy has been established in the rat, rabbit, hamster and mouse. The identification of PGF2a as the luteolytic ormone in these smaller species was based on one or more of the following: (a) a luteolytic effect of administered PGF2a; (b) an elevation of PGF2a in uterine vein blood or PGFM in peripheral blood during luteolysis; (c) prolonged cycles in animals treated with indomethacin or immunized against PGF2a; and increased uterine secretion of PGF2a following treatment of ovariectomized animals with oestrogen or progesterone. In the mouse there is evidence that a luteolytic mechanism is required for the induction of parturition. Mice rendered deficient in either the gene for the PGF2a receptor (Sugimoto et al 1997, 1998) or the PLA2 enzyme (Bonventre et al 1997; Uozumi et al 1997) do not show the normal prepartum drop in progesterone and do not undergo parturition. Ovariectomy or treatment with a progesterone antagonist removed the progesterone blockade of parturition. There are species differences in the degree of local control of the corpus luteum by the adjacent uterine horn. In the guinea-pig, as in sheep, there appears to be an absolute requirement for a local transfer of uterine PGF2a to the adjacent ovary. Although direct measurement of a local transfer of PGF2a has not been demonstrated in the guinea-pig, other indirect evidence supports this conclusion, e.g. the unilateral effect of hysterectomy, the correct anatomical vasculature in the utero-ovarian pedicle, the transfer of xenon gas from uterus to ovary (Einer-Jensen 1974) and, importantly, the extremely rapid clearance of PGF2a in the peripheral circulation of the guinea-pig (Granstro¨m and Kindahl 1982). However, in some species, such as the horse and the rabbit, it appears that PGF2a or one of its metabolites reaches the ovary via the systemic circulation, because the same amount of PGF2a is equally effective when given by the systemic or intrauterine routes. Moreover, in these two species, there is no evidence of a vascular arrangement in the utero-ovarian vessels, which would be consistent with a counter-current transfer mechanism (Del Campo and Ginther 1972, 1973). In other species, such as the pig and the cow, there is evidence for both a local and a systemic route of transfer. Such a conclusion is supported by the finding that in these two species the ovarian vasculature is consistent with a counter-current transfer mechanism. In addition, a relatively large proportion of radioactively labelled PGF2a (about 40%) passes through the pulmonary circulation unchanged in the pig and the cow, whereas in the sheep <1% passes unchanged (Davis et al 1980, 1985). In support of a systemic contribution in the cow,
normal cycles were observed in beef cows following hemihysterectomy, regardless of which ovary contained the corpus luteum (Ward et al 1976). Moreover, when one ovary of the cow was transplanted to a jugulo-carotid loop, leaving the intact uterus in the abdomen, cycle lengths were normal (Barcikowski et al 1976). In this instance, PGF2a secreted by the whole in situ uterus would produce a peripheral blood level of PGF2a twice that of the hemi-hysterectomized cow. In the alpaca, a member of the camel family indigenous to South America, it appears that the luteolytic effect of the uterus is mediated locally in the right uterine horn, but is mediated both systemically and locally in the left uterine horn (Fernandez-Baca et al 1979). In the sheep, progesterone and several other ovarian steroids are transferred from the ovarian vein to the adjacent ovarian artery, indicating the presence of an ovarian–ovarian counter-current system, i.e. in addition to a utero-ovarian transfer system for PGF2a in this species (McCracken et al 1984a). An ovarian–ovarian countercurrent transfer system has also been reported to exist in the human ovarian vascular pedicle, with no evidence for a uteroovarian transfer mechanism (Bendz 1977; Bendz et al 1979), which is consistent with the observed lack of a uterine effect on primate ovarian function. HORMONAL REGULATION OF UTERINE PGF2a SYNTHESIS It was established early on that the endometrium was the site of PGF2a synthesis, since high concentrations had been observed in human endometrium (Pickles et al 1965) and were found later in the endometrium of sheep (Wilson et al 1972a, 1972b) and guineapigs (Poyser 1972). Endometrial prostaglandin synthesis appeared to be controlled by both oestrogen and progesterone. Oestrogen alone had a modest stimulatory effect but a prior period of progesterone treatment enhanced oestrogen-stimulated production of PGF2a in several species. However, in the sheep, it subsequently became apparent that oestrogen and progesterone indirectly controlled PGF2a synthesis by regulating receptors for oxytocin in the endometrium (McCracken 1980). This conclusion was based on the finding that mechanical stimulation of the uterus evoked the secretion of PGF2a only early and especially late in the luteal phase of the cycle, while no effect was seen during the mid-luteal phase. Since mechanical stimulation of the female reproductive tract causes the release of neurohypophyseal oxytocin via the Ferguson reflex (Ferguson 1941), and since exogenous oxytocin had been shown to cause a shortening of the cycle in the cow, it seemed likely that oxytocin released reflexly by mechanical stimulation of the uterus mediated the observed stimulatory effect of mechanical stimulation on PGF2a secretion. Indeed, it was found that physiological levels of oxytocin, infused into a uterine artery in the sheep, mimicked the cyclical pattern of uterine PGF2a secretion caused by mechanical stimulation (Wilson et al 1974; Roberts and McCracken 1976; McCracken 1980). Therefore, it appeared that the cyclical variation in the ability of oxytocin to evoke the uterine secretion of PGF2a was due to a cyclical variation in the concentration of oxytocin receptors in the endometrium. Such a proposal was strengthened by reports that other target sites for oxytocin action, such as mammary gland, oviduct and uterus, contained high-affinity oxytocin-binding sites that appeared to be increased by oestrogen (Soloff 1975; Soloff et al 1975). Subsequently, it was shown that the stimulation of PGF2a by oxytocin from the ovine uterus, both in vitro and in vivo, was positively correlated with the relative abundance of oxytocin receptors in this tissue (Roberts et al 1976; McCracken 1980). A similar correlation was subsequently demonstrated in the cow (Mirando et al 1993) and the pig (Mirando et al 1996). A model for hormonal regulation of PGF2a
LUTEOLYSIS
531
Figure 50.6 A model for the endocrine control of PGF2a synthesis in the endometrial cell of the sheep during luteolysis (see text for explanation). From McCracken et al (1995), with permission from Kluwer Academic/Plenum Publisher
in the ovine uterus is depicted in Figure 50.6. It is proposed that oestrogen enhances the formation of endometrial oxytocin receptors, while, during the luteal phase, progesterone reduces the concentration of oxytocin receptors by blocking the action of oestrogen. However, since progesterone catalyses the destruction of its own receptor (Millgrom et al 1973; Vu Hai et al 1977) and decreases mRNA for its receptor (Clarke 1990), progesterone action will eventually decline towards the end of the luteal phase, thus allowing the return of oestrogen action and the synthesis of oxytocin receptors. Later, direct evidence was found for an increase in oxytocin receptors in sheep following the withdrawal of 5 days of progesterone treatment in the presence of oestradiol (McCracken et al 1984b; Leavitt et al 1985). Uterine receptors for oxytocin were upregulated in parallel 6–12 h after withdrawal of progesterone to simulate the loss of progesterone action (Figure 50.7). Also, only endometrial luminal epithelial cells were found to express oestrogen and oxytocin receptors at the time of luteolysis, while progesterone receptors were not detectable in luminal or glandular epithelia during this time (Wathes et al 1996). The greatly enhanced production of PGF2a by oxytocin at the end of the luteal phase most likely results from the priming action of progesterone on lipid precursors in the endometrium during the luteal phase (Brinsfield and Hawk 1973; Boshier et al 1981). There is now evidence that similar hormonal mechanisms regulate endometrial synthesis of PGF2a in other species, such as the cow, the sow and the mare (McCracken et al 1999). However, evidence points to a direct control of endometrial PGF2a by oestrogen and progesterone in primates (Eldering et al 1993) and in guinea-pigs (Poyser 1984).
REGULATION OF PULSATILE UTERINE PGF2a SECRETION In most species, PGF2a is released from the uterus in a pulsatile manner (see McCracken et al 1999), which appears to be advantageous for luteolysis. The model shown in Figure 50.5 explains the cellular regulation of PGF2a synthesis via the hormonal induction of endometrial oxytocin receptors. However, it does not account for the role of circulating levels of oxytocin. Episodes of oxytocin neurophysin (Fairclough et al 1980) and oxytocin itself (Flint and Sheldrick 1983) were found to occur synchronously with peaks of PGFM (the main metabolite of PGF2a) in the peripheral plasma of sheep during luteolysis. These pulses of oxytocin from the posterior pituitary appear to be regulated by an interplay between oestrogen and progesterone via their respective receptors in the hypothalamus, in a fashion rather similar to the regulation of oxytocin receptors in the uterus (McCracken et al 1995, 1996). In the sheep and other ruminants, the regulation of blood levels of oxytocin is more complex because, in addition to the neurohypophysis, the corpus luteum is an ectopic site of oxytocin synthesis and secretion (Wathes et al 1982; Fields et al 1983). The gene for oxytocin or its neurophysin is expressed in the corpus luteum of both the sheep (Jones and Flint 1988; Ivell et al 1989, 1990b) and the cow (Ivell and Richter 1984), thus confirming de novo synthesis of oxytocin in ruminant corpora lutea. Moreover, low levels of PGF2a were found to selectively release luteal oxytocin in the sheep without having any effect on progesterone secretion (Lamsa et al 1989; Custer et al 1995, 1996). Recent work indicates that the finite store of oxytocin
532
REPRODUCTIVE SYSTEM
2. 3. 4. 5.
Figure 50.7 Time course of response of uterine receptors for oxytocin, oestradiol, and progesterone in three ovariectomized sheep after the withdrawal of a 5 day infusion of progesterone (500 mg/h) in the presence of a continuous infusion of oestradiol (0.5 mg/h). Endometrial and myometrial tissues were biopsied at 71 h and at +3 h, +6 h, +12 h, following the withdrawal of progesterone infusion to simulate the loss of progesterone action. A parallel time-dependent increase in oestradiol and oxytocin receptors occurred in both endometrial and myometrial tissue. nRe, nuclear oestrogen receptor; cRe, cytosolic oestrogen receptor; ROT, oxytocin receptor; cRp, cytosolic progesterone receptor. From McCracken et al (1999), by permission of the American Physiological Society
in the corpus luteum, released by low levels of PGF2a, acts to supplement periodic discharges of oxytocin from the neurohypophysis, thus amplifying the magnitude of each luteolytic pulse of PGF2a from the uterus. However, it is not certain that a supplemental contribution of oxytocin from the corpus luteum is an absolute requirement for luteolysis in ruminants. For example, episodic pulses of PGFM of reduced magnitude occur in peripheral blood at a normal frequency in hormone-treated ovariectomized (Silvia and Raw 1993) or luteectomized (Mann 1999) sheep. Since a corpus luteum was not present in these animals, it was not established whether these reduced pulses of PGF2a (PGFM) were sufficient to induce luteolysis. However, experimental depletion of luteal oxytocin in cows (Kotwica and Skarzynski 1993) and sheep (McCracken et al 2000) did not affect the cycle length in these animals, suggesting that a luteal contribution of oxytocin may not be essential for luteolysis in ruminants. A model for the endocrine control of the pulsatile secretion of PGF2a from the sheep uterus during luteolysis is shown in Figure 50.8 (see numbers in model). 1. At the end of the luteal phase, loss of progesterone action, in both the hypothalamus and the uterus due to catalysis of its
own receptor, results in the return of oestrogen action in these tissues. Oestrogen action increases endometrial oxytocin receptors and also causes changes in the frequency of the central oxytocin pulse generator. Low levels of PGF2a will be generated in the uterus. These initial low levels of PGF2a, acting on a high-sensitivity state of the luteal PGF2a receptor (HFPR), appear to initiate a supplemental release of luteal oxytocin. Luteal oxytocin will serve to enhance uterine PGF2a secretion, which will now inhibit progesterone synthesis via a lowsensitivity state of the PGF2a receptor (LFPR) and, at the same time, release more oxytocin in a closed-loop fashion.
Additional luteolytic pulses of PGF2a will follow when luteal PGF2a receptors and/or second messenger systems have recovered, and/or when endometrial oxytocin receptors are further upregulated in time to respond to the next discharge of oxytocin via the central oxytocin pulse generator. In species such as the sow and the mare that do not synthesize oxytocin in the ovary, it is likely that oxytocin secreted by the neurohypophysis alone may be sufficient to generate the observed pulses of uterine PGF2a. Thus, in the sheep and other species, the uterus can be regarded as a transducer which, under appropriate conditions, converts neural signals (oxytocin) into local hormonal signals (PGF2a) to cause luteolysis. In addition to the well-established interplay between ovarian steroids and the anterior pituitary–hypothalamic axis for the initiation of the ovarian cycle (via the gonadotrophins), there is now good evidence for an interplay between ovarian steroids and the posterior pituitary–hypothalamic system for the termination of the reproductive cycle (via oxytocin). Although the primate uterus is an abundant source of prostaglandins (Pickles 1967), the presence of the uterus is not required for luteolysis, since ovarian cyclicity is normal in hysterectomized subjects (Beling et al 1970). However, many of the components in the sheep utero-ovarian system are present within the primate ovary itself, viz. small amounts of oxytocin (Ivell et al 1990a; Einspanier et al 1994), oxytocin receptors (Fuchs et al 1990; Copland et al 2002), PGF2a (Challis et al 1976) and receptors for PGF2a (Powell et al 1974), as well as different luteal cell types (Ohara et al 1987). Moreover, oxytocin infused into the corpus luteum of the rhesus monkey (Auletta et al 1984) and the human (Bennegard et al 1987, 1991; Bennegard-Eden et al 1995) caused a shortening of the luteal phase. In the human study, the simultaneous infusion of a PG synthesis inhibitor with oxytocin into the corpus luteum prevented luteal regression. Thus, the local production of a metabolite of arachidonic acid in the primate ovary could account for an intra-ovarian process of luteolysis in the absence of the uterus. Potential differences in the control of luteolysis in different classes of regularly cycling mammals are depicted in Figure 50.9 and are described below.
Ruminants The uterus can be regarded as a transducer which converts neurohypophyseal oxytocin signals, generated by the central oxytocin pulse generator, into pulses of PGF2a which then causes luteolysis. However, in ruminants a finite store of oxytocin in the corpus luteum may also act to supplement oxytocin signals from the central oxytocin pulse generator, and hence amplify PGF2a pulses from the uterus. It is not certain whether such an amplification of PGF2a by luteal oxytocin is a necessary requirement for luteolysis in ruminants.
LUTEOLYSIS
533
Figure 50.8 A model for the neuroendocrine control of pulsatile uterine PGF2a secretion during luteolysis in the sheep (see text for explanation of numbers). Reproduced from McCracken et al (1995)
Non-ruminants Since the corpora lutea of most non-ruminants contain very little oxytocin, it is likely that these species rely mainly on the neurohypophysis as a source of oxytocin. Thus, the central pulse generator signals of oxytocin are being transduced by the uterus into luteolytic PGF2a pulses without amplification by supplemental releases of luteal oxytocin. However, a possible contributory role to luteolysis by oxytocin produced by the endometrium of the sow (Trout et al 1995) and the mare (Behrendt-Adam et al 1999) remains to be determined. Primates Since hysterectomy is without effect on ovarian cyclicity, and since oxytocin and oxytocin receptors have been identified in the primate corpus luteum and/or ovary, oxytocin from the ovary and/or a central oxytocin pulse generator could act on ovarian oxytocin receptors to generate low levels of intra-ovarian arachidonic acid or PGF2a, which may then cause luteolysis. The transducing function of the uterus would thus be bypassed in this paradigm. However, when the uterus is present, as in the intact subject, the withdrawal of progesterone action in the uterus due to intra-ovarian luteolysis acts as a major stimulus for endometrial synthesis of PGs, which are considered to contribute to the vascular and contractile events associated with menstruation. MECHANISM OF LUTEOLYTIC ACTION OF PGF2a Receptors for PGF2a in the Corpus Luteum Evidence has accumulated that the luteolytic action of PGF2a is mediated via specific receptors for PGF2a in the corpus luteum. The first direct evidence for the existence of receptors for PGs in any tissue was provided by Fried et al (1969), who showed that biologically inert 7-oxa-acetylenic analogues of PGF1a blocked
the uterotonic action of natural PGs. Subsequently, using radiolabelled PGF2a, specific receptor binding sites for PGF2a were identified in the corpora lutea of sheep, human, cow, rat and pig as well as several other species. In addition, the PGF2a receptor has been cloned in the mouse, rat, cow, human and sheep (see McCracken et al 1999 for references). In the sheep corpus luteum, mRNA for the PGF2a receptor was localized to the large granulosa-derived steroidogenic luteal cells (Rueda et al 1995; Anderson et al 2001), whereas in the cow corpus luteum it was found in both large and small steroidogenic cells as well as in a multiple-passaged cell line of bovine luteal endothelial cells (Mamluk et al 1998). However, a subsequent study found that heterogenous luteal endothelial cells isolated from the bovine corpus luteum of pregnancy, which had undergone only one passage, did not express the PGF2a receptor (Cavicchio et al 2002). A large number of mechanisms have been proposed to explain the luteolytic action of PGF2a in the corpus luteum. Broadly, these can be divided into factors mediating functional luteolysis (the reduction of progesterone secretion or antisteroidogenic action), and structural luteolysis (the physical involution of the corpus luteum). However, it is currently unclear whether functional and structural luteolysis are truly separate events. Functional luteolysis is considered to be due to a blockade of LH action (the major luteotrophic hormone in most species), potentially mediated by a variety of substances such as adenosine, calcium, hydrogen peroxide, leukotrienes, heat shock proteins, microtubules, reactive oxygen and, more recently, endothelin-1, sterol carrier protein-2 and steroidogenic acute regulatory protein (StAR). Agents mediating structural luteolysis include apoptosis, immune cells, tumour necrosis factor-a (TNFa), monocyte chemoattractant protein-1 (MCP-1) and gap junctions (see McCracken et al 1999 for references to these potential mediators of functional and structural luteolysis). Receptors for PGF2a are located on the plasma membrane of luteal cells, where the luteolytic effect of PGF2a is thought to be initiated. Some studies report the presence of both high- and lowaffinity sites for PGF2a in luteal tissue, while others report only
534
REPRODUCTIVE SYSTEM
Figure 50.9 Comparative hypothetical models for the regulation of luteolysis in major groups of mammals that exhibit regular cyclical regression of the corpus luteum—see text for explanation. OTR, oxytocin receptor; OPG, oxytocin pulse generator. From McCracken et al (1999), by permission of the American Physiological Society
high-affinity sites. There is evidence in the sheep of a differential sensitivity of the corpus luteum to PGF2a in vivo. Subluteolytic levels of PGF2a, infused intra-arterially, can selectively evoke the release of luteal oxytocin without any effect on progesterone secretion, whereas much higher levels of PGF2a are required to depress progesterone secretion (Lamsa et al 1989; Custer et al 1995). As previously discussed, this differential sensitivity appears to be important in relation to the supplemental release of luteal oxytocin during luteolysis in ruminants. It has been proposed that the luteolytic action of PGF2a in sheep may be enhanced by the induction of PGG/PGH synthase-2 (COX2) within the corpus luteum itself (Tsai and Wiltbank 1997). Such an induction of COX-2 by PGF2a may be relevant to the eventual irreversible decline in luteal function after a finite number of endogenous pulses of PGF2a in sheep (Barcikowski et al 1974; see
Figure 50.5). Moreover, it was found that at least four 1 h long pulses of ultra-low levels of PGF2a, given into the arterial supply of the sheep ovary at intervals of 6 h, was required to cause permanent suppression of progesterone secretion (Schramm et al 1983). Similarly, a series of 1 h long pulses of higher levels of PGF2a, given into the systemic circulation at physiological intervals, was required to cause permanent luteolysis in sheep (Custer et al 1997). In this connection, six systemic infusions of PGF2a (20 mg/min), each lasting 30 min, given at mid-cycle in sheep, caused a depletion of luteal oxytocin but had no effect on plasma progesterone levels (McCracken et al 2000). However, our previous work had shown that three or four systemic infusions of PGF2a (20 mg/min), each of 1 h duration, caused a premature fall in progesterone levels and the early onset of oestrus. An important conclusion from these results is that the 1 h duration of each endogenous pulse of PGF2a
LUTEOLYSIS (i.e. in addition to their number, frequency and amplitude) appears to be critical for the induction of luteolysis in sheep and, most likely, other species as well. When a single 1 h pulse of PGF2a was given at mid-cycle, either into the ovarian artery (Schramm et al 1983) or into the systemic circulation (Custer et al 1997), the decline in progesterone was temporary, followed by recovery by 16–24 h. This suggests that with each 1 h long sequential endogenous pulse of PGF2a, or with each sequential 1 h long pulsed infusion of physiological levels of PGF2a, there may be a cumulative increase in COX-2 within the corpus luteum. The net effect of this build-up of luteal COX-2 would most likely be irreversible PGF2a-induced changes in the cellular and structural integrity of the corpus luteum. In support of this view, a recent study in sheep indicated that indomethacin (a potent inhibitor of COX action), implanted into the corpus luteum at mid-cycle, did not prevent the expected decline in plasma progesterone during luteolysis. However, the drug appeared to block structural regression of the corpus luteum (Griffeth et al 2002). This result implies that endogenous pulses of PGF2a from the uterus are sufficient to mediate the antisteroidogenic effect of PGF2a, but that PGF2a generated within the corpus luteum (via COX-2 induced by a series of endogenous uterine pulses of PGF2a) may be necessary to complete structural luteolysis. Such a notion would be consistent with the finding that a single 1 h long pulse of PGF2a given at mid-cycle causes only a temporary drop in progesterone levels followed by normal luteal function (Schramm et al 1983; Custer et al 1997; see also Prostaglandins and the Extracellular Matrix, below). The results of the above experiment with indomethacin are consistent with an earlier finding in sheep that co-administration of indomethacin with Cloprostenol (a synthetic analogue of PGF2a) prevented an increase in luteal PGF2a production but peripheral levels of progesterone fell at the expected time of luteolysis (Guthrie 1979). However, in the latter study, an effect of indomethacin on structural luteolysis was not reported. The results of the indomethacin blockade experiments also raise the possibility that the antisteroidogenic action of endogenous pulses of uterine PGF2a may be mediated by the extracellular action of PGF2a on the plasma membrane (e.g. blockade of LH–receptor interaction), whereas structural luteolysis may depend on intracellular PGF2a generated via PGF2a-induced COX-2 activity. The apparent separation of functional and structural luteolysis with indomethacin may also be relevant to the separation of functional and structural luteolysis which occurs in the rat. In this species, PGF2a can cause only functional luteolysis, whereas prolactin, secreted later, is required to cause structural luteolysis. Overall, these results with indomethacin support the early studies of Patek and Watson (1976), who proposed that luteal synthesis of PGF2a might contribute to uterine-induced luteolysis in the pig. Moreover, they provide support for the unifying hypothesis of Rothchild (1981), that PGF2a, whether of uterine or ovarian origin, might stimulate the synthesis of PGF2a within the corpus luteum itself and hence induce luteolysis. Several recent studies have suggested that endothelin-1 (ET1), a 21 amino acid peptide synthesized by endothelial cells, may play a role in the antisteroidogenic action of PGF2a (Girsh et al 1995), e.g. co-culture of bovine luteal cells with endothelial cells increased the sensitivity of the cultures to the antisteroidogenic effect of PGF2a. Additional studies indicated that ET1 reduced progesterone synthesis by bovine steroidogenic luteal cells, an effect which was blocked by an ET-1 receptor antagonist (Girsh et al 1996a, 1996b). These authors proposed that the antisteroidogenic effect of PGF2a is caused by the release of ET-1 from luteal endothelial cells. The ET-1 so released would then act on steroidogenic luteal cells to diminish progesterone production. Such a proposal would require the presence of receptors for PGF2a in luteal endothelial cells. However, as
535
mentioned above, it is not clear whether PGF2a receptors are actually expressed in bovine luteal endothelial cells. One study, using a multiple passaged endothelial cell line from bovine corpora lutea, reported the expression of mRNA for the PGF2a receptor in both endothelial and steroidogenic luteal cells (Mamluk et al 1998). On the other hand, a more recent study found that heterogenous luteal endothelial cells isolated from the bovine corpus luteum of pregnancy, which had undergone only one passage, did not express the PGF2a receptor (Cavicchio et al 2002). Moreover, these endothelial cells were unresponsive to exogenous PGF2a and progesterone but released monocyte chemoattractant protein-1 (MCP-1) in response to tumour necrosis factor-a (TNFa) and interferon. However, it is possible that PGF2a may interact indirectly with endothelial cells. For example, matrix metalloproteinases, whose activity is increased in luteal tissue by PGF2a (see next section), is known to generate ET-1 in vascular tissues (Fernandez-Patron et al 1999, 2001). It seems that further studies will be required to determine the exact role and mechanism of action of ET-1 in the process of PGF2ainduced luteolysis. A key step in the process of steroidogenesis is the cleavage of cholesterol to pregnenolone by the P-450 side-chain cleavage enzyme (P-450scc) located on the inner mitochondrial membrane of steroidogenic cells. Several studies have indicated an effect of PGF2a on cholesterol transport to and into the mitochondria of luteal cells. In rat luteal cells, PGF2a treatment decreases the mRNA and the level of sterol carrier protein-2, a protein considered to control intracellular transport of cholesterol to mitochondria (McLean et al 1995). Also of particular interest is the potential action of PGF2a on StAR, a protein considered to transport cholesterol across the outer mitochondrial membrane (Clark et al 1994; Bose et al 2002). A reduction in StAR message has been demonstrated after high doses of PGF2a in several species, an effect which may be regulated by heat shock proteins (Liu and Stocco 1997). Moreover, the message for StAR is reduced via the removal of LH by hypophysectomy and restored by the administration of exogenous LH (Juengal et al 1995). Thus, the abrogation of LH action at the cellular level, resulting in a reduction in StAR function, and hence reduced mitochondrial entry of cholesterol, could provide an attractive explanation for the antisteroidogenic action of PGF2a. PGF2a and the Extracellular Matrix of the Corpus Luteum Regressive changes in a variety of structures, especially those in the reproductive system, involve degradation of the extracellular matrix (ECM) which surrounds the cellular elements of these tissues (Woessner 1991; Birkedal-Hansen et al 1993; Smith et al 1999; Curry and Osteen 2001). Degradation of the ECM is controlled by matrix metalloproteinases (MMPs), whose activity is regulated by a number of factors, including tissue inhibitors of matrix metalloproteinases (TIMPs) which form tight noncovalent bonds with MMPs, thereby inhibiting their activity (Salamonsen 1996). An in vivo model in sheep was used to investigate the relationship between functional luteolysis (as manifested by peripheral plasma progesterone concentration) and structural luteolysis (as manifested by protein levels of regulators of the ECM within the corpus luteum). To detect the earliest biochemical changes within the corpus luteum at the onset of luteolysis, PGF2a (20 mg/min) was infused for 1 h into the systemic circulation of sheep during mid-cycle (Custer et al 1997). About 1% of the infused PGF2a escapes metabolism by the lungs and causes a 40% reduction in peripheral plasma progesterone in 6–9 h, identical to that seen after the first endogenous pulse of PGF2a during natural luteolysis in this species (Zarco et al 1988). Because only one pulse of PGF2a was given at mid-cycle, the
536
REPRODUCTIVE SYSTEM
decline in peripheral plasma progesterone was temporary and levels returned to control values by 16–24 h, whereas during natural luteolysis, peripheral plasma progesterone continues to decline under the influence of additional pulses of PGF2a (Barcikowski et al 1974; Schramm et al 1983). To measure the dynamic changes in mediators of the ECM within luteal tissue in vivo (Towle et al 2002), corpora lutea were harvested at 0 h (controls) and at 1, 8, 16 and 24 h after beginning a single 1 h-long systemic infusion of PGF2a (n=three sheep at each time point). Since TIMPs and MMPs are implicated in regulating the integrity of the ECM, temporal changes in luteal tissue were determined for protein levels of TIMP-1 and -2 which were correlated with temporal changes in the activity of MMP-2 and -9, using Western blot analysis and gelatin zymography, respectively, according to methods previously described by Goldberg et al (1996). To monitor the functional state of the corpus luteum, plasma progesterone concentration was measured in peripheral blood collected hourly from time zero until the time of corpus luteum removal from each individual animal. As shown in Figure 50.10, there was a precipitous fall in protein levels of TIMP-1 (60%, p<0.05) and TIMP-2 (90%, p<0.05) in luteal tissue 1 h after beginning the infusion of PGF2a. In contrast to the sharp fall in TIMP-1 and -2 at 1 h, the level of active MMP-2 increased to 165% of controls ( p<0.05) at 8 h, which coincided with the nadir in peripheral plasma progesterone concentration (Towle et al 2002). These results, using a physiological model in vivo, indicate that marked changes in known regulators of the ECM, signifying the onset of structural regression, actually precede the decline in plasma progesterone, signifying functional regression of the corpus luteum. The fall in peripheral plasma progesterone concentration by 8 h may reflect the influence of alterations in the ECM on the steroidogenic capacity of luteal cells, e.g. by
inducing changes in LH receptor conformation/localization in the plasma membranes of steroidogenic luteal cells. In this connection, LH receptors are reported to exhibit fast lateral diffusion in the plasma membrane of luteal cells and show receptor selfaggregation during desensitization (Horvat and Roess 2000). In support of an effect on LH receptors, there is an unexplained rapid block of gonadotrophin uptake in vivo by rat corpora lutea following PGF2a administration (Behrman and Hitchen 1976). Moreover, PGF2a-induced luteolysis in the marmoset monkey is associated with a rapid loss of LH mRNA and LH binding to steroidogenic luteal cells (Duncan et al 1998). In the latter study, the same effect on LH mRNA and LH binding was obtained by administering a GnRH antagonist, which in this case results in the loss of the ligand (LH). Also, luteolysis was induced in sheep by depleting LH from the blood circulation with a constant infusion of antiserum against bovine LH (McCracken et al 1999). Taken together, these results indicate that during luteolysis the deprivation of LH action on luteal cells, potentially via alterations in the ECM, is the primary cause of the antisteroidogenic action of PGF2a. The observed reduction in TIMP-2 may activate MMP-2 by interaction with membrane type 1-metalloproteinase which may be linked to integrins in the plasma membrane (Strongin et al 1995; Giancotti and Ruoslahti 1999). The disruption of the ECM and its associated integrin signalling system could eventually also initiate apoptosis of luteal cells by the recently described sensor Bmf, which appears to couple integrin-linked alterations in the cytoskeleton to the activation of apoptosis (Puthalakath et al 2001). Indeed, rapid alterations in the cytoskeleton of ovine steroidogenic luteal cells has been reported following pharmacological doses of PGF2a (Murdoch 1996). Recently, the intraluteal administration of L-NAME (Nonitro-L-arginine methyl ester), an inhibitor of nitric oxide (NO)
Figure 50.10 Percentage change from controls in plasma progesterone concentration (circles), TIMP-1 protein levels (triangles), TIMP-2 protein levels (diamonds), and MMP-2 activity (squares) following a 1 h systemic infusion of PGF2a (20 mg/min) at mid-cycle in sheep. Based on data from Towle et al (2002), by permission
LUTEOLYSIS synthase (NOS), was found to prolong the lifespan of the corpus luteum in cattle (Jaroszewski and Hansel 2000). In addition, the infusion of the same NOS inhibitor into the aorta, cranial to the origin of the ovarian arteries, blocked PGF2a-induced luteolysis in cattle (Skarzynski et al 2001). It has also been suggested that NO may be luteolytic in the human corpus luteum (Friden et al 2000). Thus, there is evidence that NO may play a mediatory role in luteolysis in ruminants and other species. In this connection, the putative generation of NO during luteolysis may be relevant to our finding of an acute fall in TIMP-1 and -2 after PGF2a. Frears et al (1996) reported that peroxynitrite (ONOO7) was capable of inactivating TIMP-1, thereby potentiating MMP-mediated degradation of the ECM. NO is synthesized by NOS, which also has the ability to catalyse O7 2 production. The subsequent generation of the potent radical ONOO7 may catabolize TIMPs as well as other biological molecules (Ortega and de Artinano 2000). The sequence of changes in TIMP-1 and -2 and MMP-2, and perhaps MMP-9 as well, and their relationship with declining plasma progesterone levels following additional pulses of PGF2a, given at appropriate physiological intervals, remain to be determined. The rapid and dramatic effects observed with a physiological level of PGF2a on known regulators of the ECM (TIMPs and MMPs) in luteal tissue in vivo suggest that prostaglandins may have hitherto unsuspected effects on the ECM of other tissues which are sensitive to prostaglandins. For example, changes in the ECM are known to occur at the site of rupture of Graafian follicles and in the endometrium during menstruation (Salamonsen 1996) and prostaglandins have also been implicated in both these processes (Sirois and Dore 1997; Eldering et al 1993). Moreover, the unexplained beneficial effects of non-steroidal anti-inflammatory drugs (NSAIDs) on certain cancers (Marx 2001) may be attributed to their attenuation of prostaglandin action on the ECM, especially since the ECM has been implicated in the progression of tumorigenesis in a number of organs and tissues (Liotta and Kohn 2001). This view is strengthened by the recent report that the overexpression of the inducible form of cyclooxygenase (COX-2) is sufficient to induce tumorigenesis in transgenic mice (Liu et al 2001). It may seem paradoxical that prostaglandins (specifically PGF2a) promote regression of luteal tissue, whereas there is evidence that prostaglandins promote tumour growth in susceptible tissues and that NSAIDs prevent it. However, recent work has shown that, in some tissues, TIMP-2 suppresses TKR-growth factor independently from metalloproteinase inhibition (Hoegy et al 2001). Thus, suppression of TIMP2 by a prostaglandin, such as occurs in the corpus luteum, may increase growth factor activation in tissues susceptible to cancer, and hence promote tumorigenesis. It seems that Leo Loeb’s insight in comparing the corpus luteum to a transitory tumour (Loeb 1906) may have some validity after all. Structural Luteolysis As luteolysis progresses, major structural changes occur in the corpus luteum in most species, mediated in part by the above changes in regulators of the ECM as well as the invasion of immune cells and macrophages. In the marmoset corpus luteum, TIMP-1 levels decline significantly 24 h after inducing luteolysis with either a GnRH antagonist or PGF2a (Duncan et al 1996). The reduction in TIMPs in luteal tissue would be consistent with an increase in MMP activity. It is also possible that TIMPs may have an effect on steroidogenesis, since within the sequences expressed in the bovine corpus luteum there is a 124 base pair homology between StAR and TIMP (Hartung et al 1995). It is also likely that degradation of the ECM by the activation of MMPs will initiate a variety of processes which contribute to the involution of the structure of the corpus luteum. As alluded to
537
above, it is likely that activation of membrane type MMPs will affect the function of integrins and hence the coupling of integrinlinked alterations in the cytoskeleton to the activation of apoptosis of luteal cells (Puthalakath et al 2001). ‘‘Apoptosis’’ is a term which was introduced to designate a series of morphological changes during programmed cell death (Kerr et al 1972). Cell death has been categorized into two types, necrosis and apotosis (Wyllie et al 1980). Cell necrosis is characterized by swelling of the cell and rupture of the plasma and organelle membranes resulting in lysis of the cell. Necrosis of cells is caused by toxins or prolonged ischaemia. Apoptosis, in contrast, denotes a selective or programmed death of cells which is marked by chromatin condensation and nuclear shrinkage in the presence of membrane integrity. Protuberances later appear in the plasma membrane in the form of globules (apoptotic bodies) which are phagocytosed by macrophages or adjacent tissue cells. Immune cells and macrophages also participate in the structural decline of the corpus luteum via the production of tumour necrosis factor-a (TNFa) and other cytokines (Benyo and Pate 1992). More recently, the mRNA levels and activity of caspase-3, a protein thought to be involved in cell death, was found to increase in the ovine corpus luteum after PGF2a treatment in vivo (Rueda et al 1999). Moreover, treatment of bovine luteal cells with H2O2 caused activation of caspase-3 as well as increased expression of p53 and Bax mRNA during induced apoptosis (Nakamura and Sakamoto 2001). Monocyte chemoattractant protein-1 (MCP-1) is a potent chemotactic agent for monocytes/macrophages that is expressed in regressing rat corpora lutea (Townson et al 1996). Moreover, in hypophysectomized rats, MCP-1 expression increases in response to exogenous prolactin (Bowen et al 1996), a known luteolyic agent in this species. The expression of MCP-1 and the distribution of immune cells in the corpus luteum have been correlated during the bovine oestrous cycle (Townson et al 2002). Moreover, treatment with exogenous PGF2a causes a marked increase in MCP-1 mRNA expression within corpora lutea of sheep and cows (Tsai et al 1997). MCP-1 is also expressed in blood vessel walls of the human corpus luteum and may serve to attract macrophages during regression (Senturk et al 1999). MCP-1 was not found in steroidogenic luteal cells, but other cell types such as endothelial cells, fibroblasts and macrophages themselves are known to express MCP-1 (Leonard and Yoshimura 1990). Indeed, a recent study found that endothelial cells from bovine corpora lutea of pregnancy showed increased expression of MCP-1 mRNA and protein following treatment with TNFa and interferon-g but not with PGF2a or progesterone (Cavicchio et al 2002). Thus, endothelial MCP-1, stimulated by PGF2a-induced cytokines, is a likely mediator of the invasion of macrophages which may serve to phagocytose apoptotic cells during structural regression. In this connection, a recent study has identified a protein (MGF-E8) that is secreted by activated macrophages which binds to apoptotic cells and transports them to phagocytes for engulfment (Hanayama et al 2002). Blood Flow During Luteal Regression Early studies had suggested that PGF2a might induce luteal regression via a constricting effect on blood flow through the utero-ovarian vein of the rat (Pharriss and Wyndgarden 1969; Pharriss et al 1970, 1972). Some initial studies, using pharmacological doses of PGF2a, implied that PGF2a acted by reducing either luteal or ovarian blood flow, as determined by microspheres or Doppler measurements (Kirton et al 1972; Novy and Cook 1973; Niswender et al 1976; Ford et al 1979). However interest in blood flow as a mediator of luteolysis waned with the finding that PGF2a depressed progesterone production in vitro (O’Grady et al
538
REPRODUCTIVE SYSTEM
Figure 50.11 Miniaturized Geiger Mu¨ller tube (diameter 1.5 mm) inserted through the centre of the corpus luteum of the in situ ovary of the sheep on day 10 of the cycle. A fine catheter was inserted into the ramus uterinus branch of the ovarian artery to allow the continuous infusion of PGF2a into the ovary and the periodic injection of 85krypton into the corpus luteum. A wide-bore catheter was inserted into the utero-ovarian vein to monitor progesterone secretion rate. From Einer-Jensen N and McCracken JA (1981) Acta Vet Scand, (Suppl) 77, 89–101 with permission
Figure 50.12 Capillary blood flow in luteal tissue (solid circles) and ovarian progesterone secretion rate (open circles) during the infusion of PGF2a into the arterial supply of the in situ ovary of the sheep on day 10 of the cycle. Progesterone secretion rate fell precipitously, while there was no change in capillary blood flow through the corpus luteum. From EinerJensen and McCracken (1981), by permission
1972; Henderson and McNatty 1975). In addition, other studies, using much lower doses of PGF2a, reported no effect on ovarian blood flow (Bruce and Hillier 1974; Janson et al 1975). Consistent with the latter findings, in the female rabbit, the withdrawal of oestrogen, the major luteotrophin in this species, caused a precipitous fall in progesterone levels within 24 h, whereas blood flow remained unchanged (Keyes and Wiltbank 1988). Because it was considered that a regional change in luteal blood flow might occur without a change in total ovarian blood flow, capillary blood flow was measured in the corpus luteum of the sheep after the intra-arterial administration of PGF2a (Einer-Jensen and McCracken 1976, 1981). A miniaturized Geiger–Mu¨ller tube was inserted through the corpus luteum (Figure 50.11) and 85krypton dissolved in saline was infused periodically for 1 min into the arterial supply of the ovary. The clearance rate of 85krypton from 0.5 mm of luteal tissue surrounding the probe was determined by recording b-emission. Capillary blood flow in luteal tissue was determined by the half-life, according to the method of Lassen and Ingvar (1961). As shown in Figure 50.12, capillary blood flow in luteal tissue remained unchanged after infusing PGF2a into the ovarian artery, while progesterone levels had declined by *50% after 2 h. These results are consistent with the finding that total ovarian blood flow in sheep does not decrease until after the onset of functional luteolysis (Mattner and Thorburn 1969). Degenerative changes in the extracellular matrix, as well as apoptotic changes of the cellular components of the corpus luteum, including endothelial cells, may explain the reduction in luteal blood flow as structural luteolysis progresses. It is also possible
LUTEOLYSIS
Figure 50.13 Episodic pulses of PGF2a in utero-ovarian venous plasma in cyclic non-pregnant and early pregnant sheep over the normal time of luteolysis. From Barcikowski et al (1974), published by The Endocrine Society reproduced with permission
that the eventual decline in luteal blood flow may be caused in part by the abrogation of LH action in the corpus luteum by PGF2a. For example, the infusion of purified ovine LH into the arterial supply of the ovary of conscious sheep causes a dramatic and sustained increase in ovarian blood flow (McCracken et al 1969b). Thus, a blockade of LH action in the corpus luteum by PGF2a would be expected to have the opposite effect on blood flow. ABROGATION OF LUTEOLYSIS IN EARLY PREGNANCY In animals which normally undergo luteolysis towards the end of the reproductive cycle, mechanisms have evolved to prevent luteolysis when the animal becomes pregnant. Such an abrogation of luteolysis allows the continued secretion of luteal progesterone which is necessary for pregnancy maintenance, either throughout gestation (e.g. goat and pig) or until the placenta assumes this function (e.g. sheep and primate). In the event of pregnancy, several strategies have emerged to prevent luteolysis in different species. Early studies in sheep demonstrated that the transfer of an embryo, or its secreted proteins, to a synchronized donor prevented luteolysis even late in the cycle (Moor and Rowson 1966; Moor 1968). Since PGF2a was identified as the proximate cause of luteolysis in sheep (McCracken et al 1972), it seemed likely that the capacity of an embryo to prevent luteolysis depended on either its ability to inhibit the production of PGF2a (antiprostaglandin secreting effect) or by the production of a substance which protected the corpus luteum against the luteolytic action of PGF2a (luteoprotective effect). There is, indeed, evidence that one or both mechanisms may be operative in a number of species.
539
Figure 50.14 The concentration of PGFM in jugular venous plasma of cyclic non-pregnant (A) and early pregnant sheep (B) over the normal time of luteolysis. The large pulses of PGFM normally seen during luteolysis were diminished or absent during early pregnancy. Reproduced from Peterson et al (1976). Copyright # 1976, with permission from Elsevier
Antiprostaglandin Secreting Effect of Pregnancy In species such as the domestic animals and the guinea-pig, where pulses of uterine PGF2a have been identified as the proximate cause of luteolysis, there is a marked diminution or absence of peaks of PGF2a in uterine venous plasma in the early pregnant animal (Barcikowski et al 1974; see Figure 50.13). This finding was subsequently confirmed by the report that pulses of PGFM in peripheral blood were also diminished or absent during early pregnancy in the sheep (Peterson et al 1976; see Figure 50.14). However, in some species, such as the goat, basal plasma levels of PGF2a metabolites are higher in early pregnancy in the presence of diminished pulsatile PGF2a secretion (Kindahl et al 1984). Because endometrial oxytocin receptors had been correlated with the production of uterine PGF2a production in cycling sheep (Roberts et al 1976; McCracken 1980), it seemed likely that the reduction of PGF2a pulses in early pregnancy might be related to a reduction in endometrial receptors for oxytocin. This idea was supported by the finding that oxytocin infused into a uterine artery of the sheep on day 16 of pregnancy did not evoke the secretion of PGF2a, as was found in the cyclic animal on day 16 (McCracken 1980; McCracken et al 1984b). Also, the concentration of oxytocin receptors in the endometrium of pregnant sheep was found to be 33 fmol/mg protein on day 16 compared to a concentration >250 fmol/mg protein in the non-pregnant animal at this time (McCracken 1980; McCracken et al 1984b). Such a difference in oxytocin receptor levels between pregnant and nonpregnant sheep was confirmed later by Sheldrick and Flint (1985). Subsequent results indicated that a protein secreted by the sheep embryo (Martal et al 1979), identified as interferon-t (Imakwa
540
REPRODUCTIVE SYSTEM
et al 1987; Roberts et al 1992), blocked the formation of endometrial oxytocin receptors by suppressing the transcription of the oestrogen receptor-a gene (Fleming et al 2001). In the early pregnant cow, the blastocyst also produces interferon-t which blocks the formation of endometrial receptors for oxytocin and hence supresses oxytocin-induced luteolytic pulses of PGF2a (Mann et al 1999). However, the cow may differ from the sheep in that the embryo appears to suppress oxytocin receptors without causing a change in endometrial receptors for oestrogen (Mann et al 1999; Robinson et al 2001) Also, there is evidence that the endometrium of the cow may produce an inhibitor of prostaglandin synthesis, which may help to reduce basal levels of endometrial PGF2a production during the establishment of early pregnancy (Basu et al 1988; Thatcher et al 1995). This is consistent with the fact that basal levels of PGF2a in early pregnancy are lower in the cow compared to the goat and the sheep. However, there is some evidence that an inhibitor of prostaglandin synthesis may also exist in sheep endometrium (Basu 1988). The ability of the bovine blastocyst to produce enough interferon-t to abrogate luteolysis appears to depend on an adequate luteal phase production of progesterone (Mann et al 1999; Mann and Lamming 2001). This conclusion is supported by an earlier finding that higher maternal progesterone levels during early pregnancy in cows are correlated with greater lengths of the bovine blastocyst (Boyd et al 1969). In the pig, oestrogen secreted by the implanting blastocyst is thought to change endometrial PGF2a secretion from an endocrine mode to an exocrine mode, thus preventing the secretion of PGF2a into the uterine venous system (Gross et al 1988). The equine blastocyst does not elongate rapidly as it does in ruminants and pigs, but rather maintains contact with the endometrium by traversing the uterus every 2 h during the establishment of pregnancy (Leith and Ginther 1984). PGF2a and/or PGE2 produced by the blastocyst or the endometrium has been proposed as a mediator of the intrauterine migration of the blastocyst in the mare (Stout and Allen 2001). In the guinea-pig, PGF2a production by the endometrium is suppressed during early pregnancy. However, unlike domestic animals in which the suppression of PGF2a secretion is a local phenomenon, there is evidence that an unidentified systemic factor may suppress PGF2a production in the early pregnant guinea-pig (Poyser 1995). In the rat, cervical stimulation reflexly releases pituitary secretion of prolactin, which causes the corpus luteum to secrete progesterone (Smith et al 1975). In the rabbit, a placental luteotrophin in combination with follicular oestrogen is thought to maintain the corpus luteum during pregnancy (Gadsby and Keyes 1984). However, the increase in uterine PGF2a production seen during pseudopregnancy in the rabbit is suppressed during pregnancy by an unknown mechanism. Endometrial production of PGs is suppressed during early ectopic and intrauterine pregnancies in women (Abel et al 1980). Such a supression of endometrial PGs most likely results from the continued secretion of progesterone by the corpus luteum, which is rescued by hCG secreted by the implanting blastocyst. Luteoprotective Effects of Pregnancy In addition to the suppression of endometrial PGF2a synthesis, there is evidence that in some species the corpus luteum is resistant to the luteolytic action of PGF2a during early pregnancy, e.g. in early pregnant sheep, PGF2a is less efficient in causing luteolysis when injected into the ovarian artery (Mapletoft et al 1976) or into ovarian follicles adjacent to the corpus luteum, compared to similar injections in non-pregnant sheep (Inskeep et al 1975). Also, the luteolytic effect of exogenous oestrogen is only partially effective in early pregnant sheep compared to non-pregnant sheep
(Kittok and Britt 1977). The ovine corpus luteum was found to be more resistant to a PGF2a analogue when multiple blastocysts were present compared to the presence of a single blastocyst (Nancarrow et al 1982). It has been suggested that PGE2, which is produced by the blastocyst or endometrium in the sheep and pig, may counteract the luteolytic effect of PGF2a, since PGE2 delays luteolysis in non-pregnant sheep when infused into the uterus or into an ovarian artery (Pratt et al 1977). Based on its ability to stimulate cyclic AMP production and its known vasodilator effect, PGE2 may be one of the factors which protects the corpus luteum against the potential luteolytic effect of PGF2a in early pregnancy. It has also been suggested that luteal activity of 15hydroxyprostaglandin dehydrogenase, the main enzyme responsible for catabolizing PGF2a, may be greater during early pregnancy in sheep (Silva et al 2000). In primates, progesterone from the corpus luteum is necessary to maintain pregnancy in its early stages until placental production of progesterone begins. The extension of the lifespan of the corpus luteum during early pregnancy in primates is caused by the secretion of chorionic gonadotrophin by the blastocyst around the time of implantation (Knobil 1973). Thus, during early pregnancy in primates, the rescue of the corpus luteum is primarily a luteotrophic event which overrules luteolysis. In domestic animals, however, the rescue of the corpus luteum is primarily due to the prevention of uterine PGF2a secretion and possibly, in part, to a luteoprotective effect. The corpus luteum of early pregnant monkeys is relatively resistant to the luteolytic effect of administered prostaglandin analogues compared to the non-pregnant monkey (McCracken et al 1980), suggesting that one function of chorionic gonadotrophin in primates is to override the potential luteolytic effect of endogenous prostaglandins. Moreover, the administration of human chorionic gonadotrophin to rhesus monkeys during the mid-luteal phase of the cycle prevents corpus luteum regression by the intraluteal infusion of PGF2a (Auletta and Kelm 1994). These findings suggest that, in primates, chorionic gonadotrophin may rescue the corpus luteum, not by suppressing the synthesis of potential luteolysins such as prostaglandins, but rather by overriding their action by a luteotrophic process.
ACKNOWLEDGEMENTS The work in the author’s laboratory was most recently supported by National Institutes of Health Grants HD-08129 and HD-20290 and by United States Department of Agriculture Grant 98-35203-6635. The author is most grateful to Juanita McIntosh for typing the manuscript and to Judy M. Nordberg for reference tracing.
REFERENCES Abel MH, Smith SK and Baird DT (1980) Suppression of concentration of endometrial prostaglandin in early intra-uterine and ectopic pregnancy in women. J Endocrinol, 85, 379–386. Anderson LL, Bland KP and Melampy RM (1969) Comparative aspects of uterine–luteal relationships. Rec Prog Horm Res, 25, 57–104. Anderson LE, Wu YL, Tsai SJ and Wiltbank MC (2001) Prostaglandin F2a receptor in the corpus luteum: recent information on the gene, messenger ribonucleic acid, and protein. Biol Reprod, 64, 1041–1047. Auletta FJ and Flint APF (1988) Mechanisms controlling corpus luteum function in sheep, cows, non-human primates, and women especially in relation to the time of luteolysis. Endocr Rev, 9, 88–105. Auletta FJ and Kelm LB (1994) Mechanisms controlling luteum function in the rhesus monkey (Macaca mulatta): Inhibitory action of hCG on luteolysis induced by PGF2a. J Reprod Fertil, 102, 215–220.
LUTEOLYSIS Auletta FJ, Paradis DK, Wesley M and Duby RT (1984) Oxytocin is luteolytic in the rhesus monkey (Macaca mulatta). J Reprod Fertil, 72, 401–406. Banu S, Arosh JA, Chapeldaine P and Fortier MA (2003) Molecular cloning and spatio-temporal expression of the prostaglandin transporter: a basis for the action of prostaglandins in the bovine reproductive system. Proc Natl Acad Sci, 100, 11747–11752. Barcikowski B, Carlson JC, Wilson L and McCracken JA (1974) The effect of endogenous and exogenous estradiol-178 on the release of prostaglandin F2a from the ovine uterus. Endocrinology, 95, 1340– 1349. Barcikowski B, Pichova D, Skarda J and Jerlicka L (1976) The oestrous cycle in cattle after transplantation of the ovary. Proc Sym RIA in Anim Physiol Slovak Acad Sci Kosice & Czech Acad Sci. Prague, 89–100. Basu S (1988) Endogenous inhibition of arachidonic acid metabolism in the endometrium of the sheep. Prostagland Leukotriene Essent Fatty Acids, 35, 147–152. Basu S, Albihn A and Kindahl H (1988) Studies of endogenous inhibitors of prostaglandin biosynthesis in bovine endometrial tissue collected by biopsy. J Vet Med A, 35, 141–151. Behrendt-Adam CY, Adams MH, Simpson KS and McDowell KJ (1999) Oxytocin-neurophysin I mRNA abundance in equine uterine endometrium. Domest Anim Endocrinol, 16, 183–192. Behrman HR and Hichens M (1976) Rapid block of gonadotropin uptake by corpora lutea in vivo induced by prostaglandin F2a. Prostaglandins, 12, 83–95. Behrman HR, Endo T, Aten R and Musicki B (1993) Corpus luteum function and regression. Reprod Med Rev, 2, 153–160. Beling CG, Marcus SL and Markham SM (1970) Functional activity of the corpus luteum following hysterectomy. J Clin Endocrinol Metab, 30, 30–39. Bendz A (1977) The anatomical basis for a possible counter-current exchange mechanism in the human adnex. Prostaglandins, 13, 355– 362. Bendz A, Einer-Jensen N, Lundgren O and Janson PO (1979) 85Krypton exchange between blood vessels of the human uterine adnexa. J Reprod Fertil, 57, 137–142, Bennegard B, Hahlin M and Dennefors B (1987) Antigonadotropic effect of oxytocin on the isolated human corpus luteum. Fertil Steril, 47, 431– 435. Bennegard B, Hahlin M, Wanneberg E and Noren H (1991) Local luteolytic effect of prostaglandin F2a in the human corpus luteum. Fertil Steril, 56, 1070–1076. Bennegard-Eden B, Hahlin M and Kindahl H (1995) Interaction between oxytocin and prostaglandin F2a in human corpus luteum? Hum Reprod, 10, 2320–2324. Benyo DF and Pate JL (1992) Tumour necrosis factor-a alters bovine luteal cell synthetic capacity and viability. Endocrinology, 130, 854– 960. Bergstro¨m S and Sjo¨vall J (1957) The isolation of prostaglandin. Acta Chem Scand, 11, 1086. Birkedal-Hansen H, Moore WGI, Bodden MK et al (1993) Matrix metalloproteinases: a review. Crit Rev Oral Biol Med, 4, 197–250. Bonventre JV, Huang Z, Taheri MR et al (1997) Reduced fertility and postischaemic brain injury in mice deficient in cytosolic phospholipase A2. Nature 390, 622–625. Bose HS, Lingappa VR and Miller WL (2002) Rapid regulation of steroidogenesis by mitochondrial protein import. Nature 417, 87–91. Boshier DP, Holloway H and Millener NM (1981) Triacylglycerols in the rat uterine epithelium during the oestrous cycle and early pregnancy. J Reprod Fertil, 62, 441–446. Bowen JM, Keyes PL, Warren JS and Townson DH (1996) Prolactininduced regression of the rat corpus luteum: expression of monocyte chemoattractant protein-1 and invasion of macrophages. Biol Reprod, 54, 1120–1127. Boyd H, Bacsich P, Young A and McCracken JA (1969) Fertilization and embryonic survival in the dairy cow. Br Vet J, 125, 87–97. Brinsfield TH and Hawk HW (1973) Control by progesterone of the concentration of lipid droplets in epithelial cells of the sheep endometrium. J Anim Sci, 36, 919–922. Bruce NW and Hillier K (1974) The effect of prostaglandin F2a on ovarian blood flow and corpora lutea regression in the rabbit. Nature, 249, 176–177.
541
Butenandt A and Westphal U (1934) Zur Isolierung und Charakterisierung des Corpus-Luteum Hormons. Berichte Deutsch Chem Geseilsch, 67, 1440–1442. Cavicchio VA, Pru JK, Davis BS et al (2002) Secretion of monocyte chemoattractant protein-1 (MCP-1) by endothelial cells of the bovine corpus luteum: regulation by cytokines but not prostaglandin F2a. Endocrinology, 143, 3582–3589. Challis JR, Calder AA, Dillway S et al (1976) Production of prostaglandins E and F2a by corpora lutea, corpora albicantes and stroma from the human ovary. J Endocrinol, 68, 401–408. Clark BJ, Wells J, King SR and Stocco DM (1994) The purification, cloning, and expression of a novel luteinizing hormone-induced mitochondrial protein in MA-10 mouse Leydig tumour cells. Characterization of the steroidogenic acute regulatory protein (StAR). J Biol Chem, 269, 28314–28322. Clarke CE (1990) Cell-specific regulation of progesterone receptors in the female reproductive system. Mol Cell Endocrinol, C29–C33. Copland JA, Zlatnik MG, Ives KL and Soloff MS (2002) Oxytocin receptor regulation and action in a human granulosa-lutein cell line. Biol Reprod, 66, 1230–1236. Corner GN (1960) Leo Loeb (1869–1959). In Dictionary of American Biography, suppl 6, 1956–1960. Publr.: American Council of Learned Societies, 1980. Curry TE Jr and Osteen KG (2001) Cyclic changes in the matrix metalloproteinase system in the ovary and uterus. Biol Reprod, 65, 1285–1296. Custer EE, Lamsa JC, Eldering JA and McCracken JA (1995) Identification of functional high and low affinity states of the prostaglandin F2a receptor in the ovine corpus luteum in vivo and their role in hormone pulsatility. Endocrinology, 3, 761–764. Custer EE, Eldering JA, Lamsa JC and McCracken JA (1996) Differential actions of PGF2a on the ovine corpus luteum. Reprod Dom Anim, 31, 449–454. Custer EE, Gagnon GC, Nay MG and McCracken JA (1997) Regression of the ovine corpus luteum by pulsatile systemic infusions of PGF2a. Biol Reprod, 56(suppl), A460, 197. Davis AJ, Fleet IR, Harrison FA and Maule Walker FM (1980) Pulmonary metabolism of prostaglandin F2a in the conscious nonpregnant ewe and sow. J Physiol, 301, 86P. Davis AJ, Fleet IR, Hansford PA et al (1985) Pulmonary metabolism of prostaglandin F2a in the conscious non-pregnant cow. J Physiol (Lond), 358, 107P. Del Campo CH and Ginther OJ (1972) Vascular anatomy of the uterus and the ovaries and the unilateral luteolytic effect of the uterus: guineapigs, rats, hamsters and rabbits. Am J Vet Res, 33, 2561–2578. Del Campo CH and Ginther OJ (1973) Vascular anatomy of the uterus and the ovaries and the unilateral luteolytic effect of the uterus: horses, sheep and swine. Am J Vet Res, 34, 305–316. Duncan WC, Illingworth PJ and Fraser HM (1996) Expression of tissue inhibitor of metalloproteinases-1 in the primate ovary during induced luteal regression. J Endocrinol, 151, 203–213. Duncan WC, Illingworth PJ, Young FM and Fraser HM (1998) Induced luteolysis in the primate: rapid loss of luteinizing hormone receptors. Hum Reprod, 13, 2532–2540. Eglinton G, Raphael RA, Smith GN et al (1963) Isolation and identification of two smooth muscle stimulants from menstrual fluid. Nature, 200, 960–993. Einer-Jensen N (1974) Local transfer of 133Xenon from the uterine horn to the ipsilateral ovary in the mouse, hamster and guinea-pig. J Reprod Fertil, 40, 479–482. Einer-Jensen N and McCracken JA (1976) 85Krypton measurement of capillary blood flow in the ovine corpus luteum during PGF2a-induced luteolysis. Adv Prostagland Thrombox Res, 2, 901–902. Einer-Jensen N and McCracken JA (1981) Physiological aspects of corpus luteum blood flow and of the counter-current system in the ovarian pedicle of the sheep. Acta Vet Scand 77(suppl), 89–101. Einspanier A, Ivell R, Rune G and Hodges JK (1994) Oxytocin gene expression and oxytocin immunoactivity in the ovary of the common marmoset monkey (Callithrix jacchus). Biol Reprod, 50, 1216–1222. Eldering JA, Nay MG, Hoberg LM et al (1993) Hormonal regulation of endometrial prostaglandin F2a production during the luteal phase of the rhesus monkey. Biol Reprod, 49, 809–815.
542
REPRODUCTIVE SYSTEM
Fairclough RJ, Small JF, Peterson AJ and McGowan LT (1976) The effect of estradiol-17b, progesterone and PGF2a antiplasma on luteal function in the ewe. J Reprod Fertil, 46, 523–524. Fairclough RJ, Moore LG, McGowan LT et al (1980) Temporal relationship between plasma concentrations of 13,14-dihydro-15keto-prostaglandin F and neurophysin I/II around luteolysis in sheep. Prostaglandins, 20, 199–208. Fairclough RJ, Smith JF and McGowan LT (1981) Prolongation of the oestrous cycle in cows and ewes after passive immunization with PGF antibodies. J Reprod Fertil, 62, 213–219. Ferguson JKW (1941) A study of the motility of intact uterus at term. Surg Gynecol Obst, 73, 359–366. Fernandez-Baca S, Hansel W, Saatman R et al (1979) Differential luteolytic effects of right and left uterine horns in the alpaca. Biol Reprod, 20, 586–595. Fernandez-Patron C, Radomski MW and Davidge ST (1999) Vascular matrix metalloproteinase-2 cleaves big enothelin-1 yielding a novel vasoconstrictor. Circ Res, 85, 906–911. Fernandez-Patron C, Zouki C, Whittal R et al (2001) Matrix metalloproteinases regulate neutrophil endothelial cell adhesion through generation of endothelin-1 (1-32). FASEB J, 15, 2230–2240. Ferreira SH and Vane JR (1967) Prostaglandins: their disappearance from and release into the circulation. Nature, 216, 868–873. Fields PA, Eldridge RK, Fuchs AR et al (1983) Human placental and bovine corpora luteal oxytocin. Endocrinology, 112, 1544–1546. Fleming JGW, Youngsok C, Johnson GA et al (2001) Cloning of the ovine estrogen receptor-a promoter and functional regulation by ovine interferon-t. Endocrinology, 142, 2879–2887. Flint APF and Sheldrick EL (1983) Evidence for a systemic role for ovarian oxytocin in luteal regression in sheep. J Reprod Fertil, 47, 215– 225. Ford SP, Christenson RK and Chenault JR (1979) Pattern of blood flow to the uterus and ovaries of ewes during the period of luteal regression. J Anim Sci, 49, 1510–1516. Frears ER, Zhang Z, Blade DR et al (1996) Inactivation of tissue inhibitor of metalloproteinase-1 by peroxynitrite. FEBS Lett, 381, 21–24. Friden BE, Runesson E, Hahlin M and Brannstrom M (2000) Evidence for nitric oxide acting as a luteolytic factor in the human corpus luteum. Mol Hum Reprod, 6, 397–403. Fried J, Santhanakrishnan TS, Himizu J et al (1969) Prostaglandin antagonists: synthesis and smooth muscle activity. Nature, 223, 208– 210. Fuchs A-R, Behrens O, Helmer H et al (1990) Oxytocin and vasopressin binding sites in human and bovine ovaries. Am J Obstet Gynecol, 163, 1961–1967. Gadsby JE and Keyes PL (1984) Control of corpus luteum function in the pregnant rabbit: role of the placenta (‘‘placental luteotropin’’) in regulating responsiveness of corpora lutea to estrogen. Biol Reprod, 31, 16–24. Giancotti FG and Ruoslahti E (1999) Integrin signalling. Science, 285, 1208–1032. Girsh E, Greber Y and Meidan R (1995) Luteotrophic and luteolytic interactions between bovine small and large luteal cells and endothelial cells. Biol Reprod, 52, 954–962. Girsh E, Milvae RA, Wang W and Meidan R (1996a) Effect of endothelin1 on bovine luteal cell function: role in prostaglandin F2a-induced antisteroidogenic action. Endocrinology, 137, 1306–1312. Girsh E, Wang W, Mamluk R et al (1996b) Regulation of endothelin-1 expression in the bovine corpus luteum: Elevation by prostaglandin F2a. Endocrinology, 137, 5191–5196. Goding JR and McCracken JA (1966) Preparation of a jugulo-carotid ovarian autotransplant in the sheep. Proc 2nd Internat Congr Horm Ster Milan, A 473, 259. Goding JR, McCracken JA and Baird DT (1967) The study of ovarian function in the ewe by means of a vascular autotransplantation technique. J Endocrinol, 39, 37–52. Goldberg MJ, Moses MA and Tsang PCW (1996) Identification of matrix metalloproteinases and metalloproteinase inhibitors in bovine corpora lutea and their variation during the bovine estrous cycle. J Anim Sci, 74, 849–857. Granstro¨m E and Kindahl H (1982) Species differences in circulating prostaglandin metabolites: relevance for the assay of prostaglindin release. Biochim Biophys Acta, 713, 555–569.
Griffeth RJ, Nett TM, Burns PD et al (2002) Is luteal production of PGF2a required for luteolysis? Biol Reprod, 66(suppl 1), abstr 465. Gross TS, Lacroix MC, Bazer FW et al (1988) Prostaglandin secretion by perfused porcine endometrium: further evidence for an endocrine vs. exocrine secretion of prostaglandins. Prostaglandins, 35, 327–341. Guthrie HD (1979) In discussion of luteolysis. Adv Exp Med Biol, 112, 621. Hanayama R, Tanaka M, Miwa K et al (2002) Identification of a factor that links apoptotic cells to phagocytes. Nature, 417, 182–187. Hartung S, Rust W, Balvers M and Ivell R (1995) Molecular cloning and in vivo expression of the bovine steroidogenic acute regulatory protein. Biochem Biophys Res Commun, 215, 646–653. Henderson KM and McNatty KP (1975) A biochemical hypothesis to explain the mechanism of luteal regression. Prostaglandins, 9, 779– 797. Hoegy S, Oh HR, Corcoran M and Stetler-Steveson W (2001) Tissue inhibitor of metalloproteinase-2 (TIMP-2) suppresses TKR-growth factor signaling. J Biol Chem, 276, 3203–3214. Horvat RD and Roess DA (2000) Desensitized LH receptors are selfassociated in large, slowly diffusing complexes. Biol Reprod, 62(suppl 1), abstr 2, 102. Imakawa K, Anthony RV, Kazemi M et al (1987) Interferon-like sequence of ovine trophoblast protein secreted by embryonic trophoderm. Nature, 330, 337–379. Inskeep EK, Smutny WJ, Butcher RL and Pexton JE (1975) Effects of intrafollicular injections of prostaglandins in non-pregnant and pregnant ewes. J Anim Sci, 41, 1098–1104. Ivell R and Richter D (1984) The gene for the hypothalamic peptide hormone oxytocin is highly expressed in the ovine corpus luteum: biosynthesis, structure, and sequence analysis. EMBO J, 3, 2351– 2354. Ivell R, Hunt N, Lamsa JC and McCracken JA (1989) Oxytocin gene expression in the ovine corpus luteum. Endocrinology, 124, A1544. Ivell R, Furuya K, Brackman B et al (1990a) Expression of the oxytocin and vasopressin genes in human and baboon tissues. Endocrinology, 127, 2990–2996. Ivell R, Hunt N, Abend N et al (1990b) Structure and ovarian expression of the oxytocin gene in sheep. Reprod Fertil Dev, 2, 703–711. Janson PO, Albrecht I and Ahren K (1975) Effects of prostaglandin F2a on ovarian blood flow and vascular resistance in the pseudo-pregnant rabbit. Acta Endrocrinol, 79, 337–350. Jaroszewski JJ and Hansel W (2000) Intraluteal administration of a nitric oxide synthase blocker stimulates progesterone and oxytocin secretion and prolongs the life span of the bovine corpus luteum. Proc Soc Exp Biol Med, 224, 50–55. Jones DSC and Flint APF (1988) Concentrations of oxytocin-neurophysin prohormone mRNA in corpora lutea of sheep during the oestrous cycle and in early pregnancy. J Endocrinol, 117, 409–414. Juengel JL, Meberg BM, Turzillo AM et al (1995) Hormonal regulation of messenger ribonucleic acid encoding steroidogenic acute regulatory protein in ovine corpora lutea. Endocrinology, 136, 5423–5429. Kerr JRF, Wyllie AH and Currie AR (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer, 26, 239–257. Keyes PL and Wiltbank MC (1988) Endocrine regulation of the corpus luteum. Ann Rev Physiol, 50, 465–482. Kindahl H, Edqvist L-E, Bane A and Granstro¨m E (1976) Blood levels of progesterone and 15-keto-13,14-dihydroprostaglandin F2a during the normal oestrous cycle and early pregnancy in heifers. Acta Endocrinol, 82, 134–149. Kindahl H, Basu S, Fredriksson G et al (1984) Levels of prostaglandin F2a metabolites in blood and urine during early pregnancy. Anim Reprod Sci, 7, 133–148. Kirton KT, Gutknecht GD, Bergstro¨m KK et al (1972) Prostaglandins and reproduction. J Reprod Med, 9, 266–270. Kittok RJ and Britt JH (1977) Corpus luteum function in ewes given estradiol during the estrous cycle of early pregnancy. J Anim Sci, 45, 336–341. Knobil E (1973) On the regulation of the primate corpus luteum. Biol Reprod, 8, 246–258. Kotwica J and Skarzynski D (1993) Influence of oxytocin removal from the corpus luteum on secretory function and duration of the oestrous cycle in cattle. J Reprod Fertil, 97, 411–417.
LUTEOLYSIS Lamsa JC, Kot SJ, Nay MG and McCracken JA (1989) Prostaglandin F2a stimulated release of ovarian oxytocin in the sheep in vivo: threshold and dose dependency. Biol Reprod, 40, 1215–1223. Lassen NA and Ingvar DH (1961) The blood flow of the cerebral cortex determined by radioactive krypton-85. Experientia, 17, 42–43. Leavitt WW, Okulicz WC, McCracken JA et al (1985) Rapid recovery of nuclear estrogen receptor and oxytocin receptor in the ovine uterus following progesterone withdrawal. J Steroid Biochem, 22, 687–691. Leith GS and Ginther OJ (1984) Characterization of intrauterine mobility of the early equine conceptus. Theriogenology, 22, 401–408. Leonard EJ and Yoshimura T (1990) Human monocyte chemoattractant protein-1 (MCP-1). Immunol Today, 11, 97–101. Lewis PE and Warren JE (1977) Effect of indomethacin on luteal function in ewes and heifers. J Anim Sci, 45, 763–767. Liotta LA and Kohn EC (2001) The microenvironment of the tumor–host interface. Nature, 411, 375–379. Liu Z and Stocco DM (1997) Heat shock-induced inhibition of acute steroidogenesis in MA-10 cells is associated with inhitition of the synthesis of the steroidogenic acute regulatory protein. Endocrinology, 138, 2722–2728. Liu CH, Chang S-H, Narko K et al (2001) Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem, 276, 18563–18569. Loeb L (1906) The formation of the corpus luteum in the guinea pig. J Am Med Assoc, 46, 416–423. Loeb L (1923) The effect of extirpation of the uterus on the life and function of the corpus luteum in the guinea-pig. Proc Soc Exp Biol Med (New York), 20, 441–443. Loeb L (1927) Effects of hysterectomy on system of sex organs and on periodicity of sexual cycle in guinea-pigs. Am J Physiol, 83, 202– 224. Mamluk R, Chen D-B, Greber Y et al (1998) Characterization of messenger ribonucleic acid expression for prostaglandin F2a and luteinizing hormone receptors in various bovine cell types. Biol Reprod, 58, 849–856. Mann GE (1999) The role of luteal oxytocin in episodic secretion of PGF2a at luteolysis in the ewe. Anim Reprod Sci, 57, 167–175. Mann GE and Lamming GE (2001) Relationship between maternal endocrine environment, early embryo development and inhibition of the luteolytic mechanism in cows. Reproduction, 121, 175–180. Mann GE, Lamming GE, Robinson RS and Wathes DC (1999) The regulation of interferon-t production and uterine hormone receptors during early pregnancy. J Reprod Fert Suppl, 54, 317–328. Mapletoft RJ, Lapin DR and Ginther OJ (1976) The ovarian artery as the final component of the local luteotropic pathway between a gravid uterine horn and ovary in ewes. Biol Reprod, 15, 414–421. Martal J, Lacroix M-C, Loudes C et al (1979) Trophoblastin, an antiluteolytic protein present in early pregnancy in sheep. J Reprod Fertil, 56, 63–73. Marx R (1935) Influence of hysterectomy on uterine balance. Am J Surg, 28, 117–121 Marx J (2001) Anti-inflammatories inhibit cancer growth—but how? Science, 291, 581–582. Mattner PE and Thorburn GD (1969) Ovarian blood flow in sheep during the oestrous cycle. J Reprod Fertil, 19, 547–549. McCracken JA (1971) Prostaglandin F2a and corpus luteum regression. Ann NY Acad Sci, 180, 456–472. McCracken JA (1980) Hormone receptor control of PGF2a secretion by the ovine uterus. Adv Prostagland Thrombox Res, 8, 1329–1344. McCracken JA and Schramm W (1988) Prostaglandins and corpus luteum regression. In Curtis-Prior PB (ed.), Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids. Edinburgh: Churchill-Livingstone, 425–462. McCracken JA, Caldwell BV, Tillson SA et al (1969a) Ovarian steroid secretion and LH levels in sheep after autotransplantation of the ovary, uterus, cervix and vagina to the neck. 2nd Annual Meeting of the Society for the Study of Reproduction, Davis, CA. Biol Reprod, A3(suppl), 11. McCracken JA, Uno A, Goding JR et al (1969b) The in vivo effects of sheep pituitary gonadotrophins on the secretion of steriods by the autotransplanted ovary of the ewe. J Endocrinol, 45, 425–440.
543
McCracken JA, Glew ME and Levy LK (1970a) The regulation of corpus luteum function by gonadotrophins and related compounds. Adv Biosci, 4, 377–397. McCracken JA, Glew ME and Scaramuzzi RJ (1970b) Corpus luteum regression induced by prostaglandin F2a. J Clin Endocrinol Metab, 30, 544–546. McCracken JA, Baird DT and Goding JR (1971) Factors affecting the secretion of steroids from the transplanted ovary in the sheep. Rec Prog Horm Res, 27, 537–582. McCracken JA, Carlson JC, Glew ME et al (1972) Prostaglandin F2a identified as a luteolytic hormone in the sheep. Nature New Biol, 238, 129–134. McCracken JA, Barcikowski B, Carlson JC et al (1973) The physiological role of prostaglandin F2a in corpus luteum regression. Adv Biosci, 9, 599–624. McCracken JA, Einer-Jensen N and Fried J (1979) Prostaglandin F2a and its 13-dehydro analogs: comparative luteolytic effects in vivo. Adv Exp Med Biol, 112, 577–601. McCracken JA, Glew ME, Hull SS Jr et al (1980) Comparative luteolytic effects of prostaglandin F2a and its 13-dehydro analogs in vivo. Adv Prostagland Thrombox Res, 6, 365–380. McCracken JA, Schramm W and Einer-Jensen N (1984a) The structure of steroids and their diffusion through blood vessel walls in a countercurrent system. Steroids, 43, 291–303. McCracken JA, Schramm W and Okulicz WC (1984b) Hormone receptor control of pulsatile secretion of PGF2a from ovine uterus during luteolysis and its abrogation in early pregnancy. Anim Reprod Sci, 7, 31–55. McCracken JA, Custer EE, Lamsa JC and Robinson AG (1995) The central oxytocin pulse generator: a pacemaker for luteolysis. Adv Exp Med Biol, 395, 133–154. McCracken JA, Custer EE, Eldering JA and Robinson AG (1996) The central oxytocin pulse generator: a pacemaker for the ovarian cycle. Acta Neurobiol Exp, 56, 819–832. McCracken JA, Custer EE and Lamsa JC (1999) Luteolysis: a neuroendrocrine-mediated event. Physiol Rev, 79, 263–323. McCracken JA, Amico JA, Newbury MK and Milvae RA (2000) Depletion of oxytocin from ovine corpus luteum does not extend cycle length. Biol Reprod, 62(suppl 1), abstr 92. McLean MP, Billheimer JT, Warden KJ and Irby RB (1995) Prostaglandin F2a mediates ovarian sterol carrier protein-2 expression during luteolysis. Endocrinology, 136, 4963–4972. Millgrom E, Thi L, Atger M and Baulieu EE (1973) Mechanisms regulating the concentration and conformation of progesterone receptor(s) in the uterus. J Biol Chem, 248, 6366–6374. Mirando MA, Willard CB and Whiteaker SS (1993) Relationships among endometrial oxytocin receptors, oxytocin-stimulated phosphinosotide hydrolysis and prostaglandin F2a secretion in vitro, and plasma concentrations of ovarian steroids before and during corpus luteum regression in cyclic heifers. Biol Reprod, 48, 874–882. Mirando MA, Uzumeu M, Carnahan KG and Ludwig TE (1996) A role for oxytocin during luteolysis and early pregnancy in swine. Reprod Dom Anim, 31, 455–462. Moor RM (1968) The effect of the embryo on corpus luteum function. J Anim Sci, 7(suppl 1), 97–118. Moor RM and Rowson LE (1966) Local uterine mechanisms affecting luteal function in sheep. J Reprod Fertil, 11, 307–310. Murdoch WJ (1996) Microtubular dynamics in granulosa cells of periovulatory follicles and granulosa-derived (large) lutein cells of sheep: relationships to the steroidogenic folliculo-luteal shift and functional luteolysis. Biol Reprod, 54, 1135–1140. Nakamura T and Sakamoto K (2001) Reactive oxygen species upregulates cyclooxygenase-2, p 53, and Bax mRNA expression in bovine luteal cells. Biochem Biophys Res Commun, 284, 203–210. Nancarrow CD, Evison BM and Connell PJ (1982) The effect of embryos on luteolysis and termination of early pregnancy in sheep with cloprostenol. Biol Reprod, 26, 263–269. Niswender GD, Reimers TJ, Diekman MA and Nett TM (1976) Blood flow: a mediator of ovarian function. Biol Reprod, 14, 64–81. Niswender GD, Juengel JL, Silva PJ et al (2000) Mechanisms controlling the function and lifespan of the corpus leuteum. Physiol Rev, 80, 1–29. Novy MJ and Cook MJ (1973) Redistribution of blood flow by prostaglandin F2a in the rabbit ovary. Am J Obstet Gynecol, 117, 381–385.
544
REPRODUCTIVE SYSTEM
O’Grady JP, Kohorn EI, Glass RH et al (1972) Inhibition of progesterone synthesis in vitro by prostaglandin F2a. J Reprod Fertil, 30, 153–156. Ohara A, Mori T, Tall S et al (1987) Functional differentiation in steroidogenesis of two types of luteal cells isolated from mature human corpora lutea of menstrual cycle. J Clin Endocrinol Metab, 65, 1192– 1200. Ortega MA and de Artinano AA (2000) Nitric oxide reactivity and mechanisms involved with its biological effects. Pharmacol Res, 42, 421–427. Patek CE and Watson J (1976) Prostaglandin F and progesterone secretion by porcine endometrium and corpus luteum in vitro. Prostaglandins, 12, 97–111. Peterson AJ, Tervit HR, Fairclough RJ et al (1976) Jugular levels of 13,14dihydro-15-keto-prostaglandin F and progesterone around luteolysis and early pregnancy in the ewe. Prostaglandins, 12, 551–558. Pharriss BB and Wyngarden LJ (1969) The effect of prostaglandin F2a on the progestogen content of ovaries from pseudopregnant rats. Proc Soc Exp Biol Med, 130, 92–94. Pharriss BB, Cornette JC and Gutknecht GD (1970) Vascular control of luteal steroidogenesis. J Reprod Fertil, 10(suppl), 97–103. Pharriss BB, Tilson SA and Erickson RR (1972) Prostaglandin in luteal function. Rec Prog Horm Res, 28, 51–89. Pickles VR (1967) Prostaglandins in the human endometrium. Int J Fertil, 12, 335–338. Pickles VR, Hall WJ, Best FA and Smith GN (1965) Prostaglandins in endometrium and menstrual fluid from normal and dysmenorrhoeic subjects. J Obstet Gynaecol Br Commonw, 72, 185–192. Powell WS, Hammerstro¨m S, Samuelsson B and Sjo¨berg B (1974) Prostaglandin F2a receptor in human corpora lutea. Lancet, 1, 1120. Poyser NL (1972) Production of prostaglandins by the guinea-pig uterus. J Endocrinol, 54, 147–159. Poyser NL (1984) Prostaglandin production by the uterus of the nonpregnant and early pregnant guinea-pig. Anim Repro Sci, 7, 1–30. Poyser NL (1995) The control of prostaglandin production by the endometrium in relation to luteolysis and menstruation. Prostagland Leukotriene Essent Fatty Acids, 53, 147–195. Pratt BR, Butcher RL and Inskeep EK (1977) Antiluteolytic effect of the conceptus and of PGE2 in ewes. J Anim Sci, 46, 784–791. Puthalakath H, Villunger A, O’Reilly LA et al (2001) Bmf: a proapoptotic BH3-only protein regulated by interaction with the myosin V actin motor complex, activated by anoikis. Science, 293, 1829–1832. Roberts JS and McCracken JA (1976) Does prostaglandin F2a released from the uterus by oxytocin mediate the oxytocic action of oxytocin? Biol Reprod, 15, 457–463. Roberts JS, McCracken JA, Gavagan JE and Soloff MS (1976) Oxytocinstimulated release of prostaglandin F2a from ovine endometrium: correlation with estrous cycle and oxytocin receptor binding. Endocrinology, 99, 1107–1114. Roberts JS, Cross JC and Leaman DW (1992) Interferons as hormones of pregnancy. Endocr Rev, 13, 432–452. Robinson RS, Mann GE, Lamming GE and Wathes DC (2001) Expression of oxytocin, oestrogen and progesterone receptors in uterine biopsy samples throughout the oestrous cycle and early pregnancy in cows. Reproduction, 122, 965–979. Ronayne E, Quirke JF, Enright WJ and Roche JF (1990) Effect on immunization of ewes against prostaglandin F2a on the life-span of corpora lutea and oestrous behavior during two breeding seasons. J Reprod Fertil, 90, 175–183. Rothchild I (1981) The regulation of the mammalian corpus luteum. Rec Prog Horm Res, 37, 183–298. Rueda BR, Botros IW, Pierce KI et al (1995) Comparison of the mRNA levels for the PGF2a receptor (FP) during luteolysis and early pregnancy in the ovine corpus luteum. Endocrine, 3, 781–787. Rueda BR, Hendry IR, Tilly JL and Hamernik DL (1999) Accumulation of caspase-3 messenger ribonucleic acid and induction of caspase activity in the ovine corpus luteum followin prostaglandin F2a treatment in vivo. Biol Reprod, 60, 1087–1092. Salamonsen LA (1996) Matrix metalloproteinases and their tissue inhibitors in endocrinology. Trends Endocrinol Metab, 7, 28–43. Scaramuzzi RJ and Baird DT (1976) The oestrous cycle of the ewe after active immunization against prostaglandin F2a. J Reprod Fertil, 46, 39–47.
Schramm W, Bovaird L, Glew ME et al (1983) Corpus luteum regression induced by ultra-low pulses of prostaglandin F2a. Prostaglandins, 26, 347–364. Senturk LM, Seli E, Gutierrez LS et al (1999) Monocyte chemotactic protein-1 expression in human corpus luteum. Mol Hum Reprod, 8, 697–702. Sheldrick EL and Flint APF (1985) Endocrine control of uterine oxytocin receptors in the ewe. J Endocrinol, 106, 249–258. Silvia WJ and Raw RE (1993) Regulation of pulsatile secretion of prostaglandin F2a from the ovine uterus by ovarian steroids. J Reprod Fertil, 98, 341–347. Silva PJ, Juengel JL, Rollyson MK and Niswender GD (2000) Prostaglandin metabolism in the ovine corpus luteum: catabolism of prostaglandin F2a coincides with resistance of the corpus luteum to PGF2a. Biol Reprod, 63, 1229–1236. Sirois J and Dore M (1997) The late induction of prostaglandin G/H synthase-2 in equine preovulatory follicles supports its role as a determinant of the ovulatory process. Endrocrinology, 138, 4427– 4434. Skarzynski DJ, Jaroszewski JJ, Bah MM et al (2001) An inhibitor of nitric oxide synthase (L-NAME) blocks prostaglandin F2a-induced luteolysis in cattle. Abstr 4th International Conference on Farm Animal Endocrinology, Parma, Italy, 7–10 October 2001. Slotta KH, Ruschig H and Fels E (1934) Reindarstellung der Hormone aus dem Corpus Luteum. Ber Dtsch Chem Geselschaft, 67, 1270. Smith MS, Freeman ME and Neill JD (1975) The control of progesterone secretion during the estrous cycle and early pseudopregnancy in the rat: prolactin, gonadotropin and steroid levels associated with rescue of the corpus luteum of pseudopregnancy. Endocrinology, 96, 219– 226. Smith MF, McIntush EW, Ricke WA et al (1999) Regulation of ovarian extracellular matrix remodeling by metalloproteinases and their tissue inhibitors: effects of follicular development, ovulation and luteal function. J Reprod Fertil, 54(suppl.), 367–381. Soloff MS (1975) Uterine receptors for oxytocin: Effect of estrogens. Biochem Biophys Res Commun, 65, 205–212. Soloff MS, Rees HD, Sar M and Stumpf WE (1975) Autoradiographic localization of radioactivity from [3H] oxytocin in the mammary gland and oviduct of the rat. Endocrinology, 96, 1475–1477. Stout TA and Allen WR (2001) Role of prostaglandins in intrauterine migration of the equine conceptus. Reproduction, 121, 771–775. Strongin A, Collier I, Bannikov G et al (1995) Mechanism of cell surface activation of 72 kDa type IV collagenase. J Biol Chem, 270, 5331–5338. Sugimoto Y, Yamasaki A, Segi E et al (1997) Failure of parturition in mice lacking the prostaglandin F receptor. Science, 277, 681–683. Sugimoto Y, Segi E, Tsuboi K et al (1998) Female reproduction in mice lacking the prostaglandin F receptor. Roles of prostaglandin and oxytocin receptors in parturition. Adv Exp Med Biol, 449, 317– 321. Thatcher WW, Meyer MD and Danet-Desnoyers G (1995) Maternal recognition of pregnancy. J Reprod Fertil, 49(suppl), 15–28. Thorburn GD, Cox RI, Currie WB et al (1973) Prostaglandin F and progesterone concentrations in the utero-ovarian venous plasma of the ewe during the oestrous cycle and early pregnancy. J Reprod Fertil, 18(suppl.), 151–158. Towle TA, Tsang PCW, Milvae RA et al (2002) Dynamic in vivo changes in tissue inhibitors of metalloproteinases -1 and -2, and matrix metalloproteinases -2 and -9, during prostaglandin F2a-induced luteolysis in sheep. Biol Reprod, 66, 1515–1521. Townson DH, Warren JS, Flory CM et al (1996) Expression of monocyte chemoattractant protein-1 in the corpus luteum of the rat. Biol Reprod, 54, 513–520. Townson DH, O’Connor CL and Pru JK (2002) Expression of monocyte chemoattractant protein-1 and the distribution of immune cell populations in the bovine corpus luteum throughout the estrous cycle. Biol Reprod, 66, 361–366. Trout WE, Smith GW, Gentry PC et al (1995) Oxytocin secretion by the endometrium of the pig during maternal recognition of pregnancy. Biol Reprod, 52(suppl), A529, 189. Tsai S-J and Wiltbank MC (1997) Prostaglandin F2a induces expression of prostaglandin G/H synthase-2 in the ovine corpus luteum: a potential positive feedback loop during luteolysis. Biol Reprod, 57, 1016–1022.
LUTEOLYSIS Tsai S-J, Juengel JL and Wiltbank MC (1997) Hormonal regulation of monocyte chemoattractant protein-1 messenger ribonucleic acid expression in corpora lutea. Endocrinology, 138, 4517–4520. Uozumi N, Kume K, Nagase T et al (1997) Role of cytosolic phospholipase A2 in allergic response and parturition. Nature, 390, 618–622. Vollmerhaus B (1964) Investigation of the vascular architecture of the bovine female genital tract. Zentralbl Veterinaermed, 11, 597–646. Vu Hai MT, Logeat F, Warembourg M and Milgrom E (1977) Hormonal control of progesterone receptors. Ann NY Acad Sci, 286, 199–209. Ward KE, Longwell LC, Kreider JL and Godke RA (1976) Effect of unilateral hysterectomy on cycling beef heifers. J Anim Sci, 43, A376, 309. Wathes DC, Swann RW, Pickering T et al (1982) Neurohypophysial hormones in the human ovary. Lancet, ii, 410–412. Wathes DC, Mann GE, Payne JH et al (1996) Regulation of oxytocin, oestradiol and progesterone receptor concentrations in different
545
uterine regions by oestradiol, progesterone and oxytocin in ovariectomized ewes. J Endocrinol, 151, 375–393. Wilson L, Cenedella RJ and Butcher RL (1972a) Prostaglandin F2a in the uterus of ewes during early pregnancy. Prostaglandins, 1, 479–482. Wilson L, Cenedella RJ, Butcher RL and Inskeep EK (1972b) Levels of prostaglandins in the uterine endometrium during the ovine estrous cycle. J Anim Sci, 34, 93–99. Wilson L, Roberts JS and McCracken JA (1974) The release of prostaglandin F (PG) and oxytocin (OT) during uterine massage. Proceedings of the 7th Meeting, Society for the Study of Reproduction, Ottawa, Canada, abstr 164. Woessner JF (1991) Matrix metalloproteinases and their inhibitors in connective tissue remodeling. FASEB J, 5, 2145–2154. Wyllie AH, Kerr JFR and Currie AR (1980) Cell death: the significance of apoptosis. Intern Rev Cytol, 68, 251–306. Zarco L, Stabenfeldt GH, Quirke JK et al (1988) Release of prostaglandin F2a and the timing of events associated with luteolysis in ewes with oestrous cycles of different length. J Reprod Fertil, 83, 517–526.
51 Prostaglandins in Implantation Norman L. Poyser University of Edinburgh, Edinburgh, UK
The attachment of the blastocyst to the uterine endometrium by the process of implantation is crucial for the establishment of pregnancy and the development of the blastocyst into a foetus. The blastocyst consists of an outer layer of cells (the trophoblast) and an inner layer of cells (which develop into the foetus). The inner layer of cells alone will not implant without the trophoblast being present, whereas the trophoblast alone will implant but fails to form a foetus. Implantation is therefore dependent upon an interaction between the trophoblast and the endometrium. In some species, including the mouse, hamster, rat, rabbit, guinea-pig and human, implantation takes place within 4–7 days of the egg being fertilized. In other species, such as the cow, sheep, pig and horse, implantation occurs at least 2 weeks after fertilization. Implantation is delayed in some species if, for example, the female is lactating. The majority of the studies that have examined the processes involved in implantation have been carried out in small rodents and rabbits. The findings obtained in these species point to the possible processes involved in implantation in humans which, for obvious reasons, is quite difficult to study. It is clear that, during early pregnancy, there is a short ‘‘implantation window’’ during which the blastocyst will implant. This ‘‘window’’ lasts for about 30 h in mice (Milligan et al 1995), with day 4 of pregnancy being regarded as the main day of implantation in this species. Outside this window, implantation will not occur. However, the implantation window can be brought forward in mice by treatment with splenocytes or thymocytes (which are mostly T lymphocytes) on day 2 of pregnancy, as they affect endometrial differentiation (Takabatake et al 1997a, 1997b; Fujita et al 1998). In species in which many blastocysts are present, the blastocysts are evenly spaced in each uterine horn down the antimesometrial side. This is achieved initially by uterine contractions randomly distributing the blastocysts along the horn. The lumen of the uterine horn then closes round the blastocyst and holds it in place. Any further spacing of the blastocysts takes place by differential elongation of the uterus. At the site of blastocyst attachment, one of the earliest signs of implantation is a local increase in vascular permeability, resulting in a local oedema. This site can be visualized by injecting a blue dye into the blood stream. The dye passes out of the blood stream into the surrounding tissue at the implantation site. The stromal cells at the implantation site become transformed into decidual cells. This decidual cell reaction (decidualization) takes place to a greater or lesser extent, depending on species. Also the extent of penetration by trophoblast cells into the endometrium is species-specific (Finn and Porter 1975; Hogarth 1978).
The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
HORMONAL CONTROL In all species, the uterus has to be primed with progesterone for several days in order for implantation to occur. The corpus luteum in the ovary is the source of this progesterone. In pregnant, ovariectomized rats maintained on progesterone, implantation does not occur. Oestradiol is also required for blastocyst attachment to the endometrium. In intact, pregnant rats a surge of oestradiol from the ovary occurs on day 4 and implantation takes place on day 5 of pregnancy (Watson et al 1975). Tamoxifen (an oestrogen receptor antagonist) prevents implantation in the rat (Watson et al 1975). An identical oestradiol surge occurs in pseudopregnant rats (Shaikh and Abraham 1969), which indicates that the presence of the blastocyst is not necessary for increased oestradiol output from the ovary. Oestradiol is also necessary for implantation in the mouse and, as in the rat, there is a preimplantation surge of oestradiol from the ovary (Gerber et al 1979). Implantation occurs in pregnant, ovariectomized rabbits maintained solely on progesterone, which suggests that maternal oestradiol is not involved in this species. However, the intrauterine installation of the antioestrogenic drug, CI-628, during early pregnancy in rabbits prevents implantation (Bhatt and Bullock 1974; Dey et al 1976), indicating that oestradiol is indeed necessary for implantation in this species. The source of the oestrogen is the rabbit blastocyst (Dickmann et al 1975; Wu and Lin 1982). Consequently, oestradiol of maternal and/or blastocyst origin acting on a progesterone-primed uterus is the hormonal stimulus for implantation. There appears to be a stronger, localized oestrogenic effect in uterine tissues in close proximity to the blastocyst (De Hertogh et al 1986). The actions of oestradiol may be mediated initially by the ‘‘oestrogen-response finger protein’’ (Efp), which is a putative transcription regulator that possibly amplifies oestradiol action in target cells. In the endometrium of pregnant mice, Efp mRNA is expressed in luminal and glandular epithelial cells on days 1 and 2, and in both epithelial and stromal cells on days 3 and 4. On day 5 after implantation, Efp mRNA expression is higher in stromal cells surrounding the implantation chamber (Das et al 1997). Another transcription factor, Hoxa-10, is necessary for implantation in mice, as mice lacking the Hoxa-10 gene are infertile due to pregnancy failure at the uterine level (Benson et al 1996; Satokata et al 1995). Hoxa-10 shows increased expression also in the human uterus during the mid-luteal phase at the time of implantation (Taylor et al 1998). Cyclin D3 is one of the genes upregulated in the mouse uterus at the implantation site. The upregulation of this
548
REPRODUCTIVE SYSTEM
gene is dependent on Hoxa-10 activation, since upregulation does not occur in Hoxa-10-deficient mice (Das et al 1999). The ‘‘switching on’’ of protein synthesis leads to the production in the uterus of many substances (including local hormones, cytokines, growth factors, and enzymes) around the time of the implantation process. Of the multitude of locally produced factors that are implicated in the implantation process, clear evidence needs to be provided as to which of these factors are essential. Some of this evidence is available. LOCAL FACTORS INVOLVED IN IMPLANTATION Implantation fails to take place in mice lacking the gene for leukaemia inhibitory factor (Stewart et al 1992) or for the interleukin-11 receptor (Robb et al 1998). This is compelling evidence that leukaemia inhibitory factor and endogenous ligands for the interleukin receptor are essential for implantation. Splenocytes and thymocytes from day 4 mice, which are able to bring forward the implantation window, induce the expression of LIF mRNA in the mouse uterus (Takabatake et al 1997a; Fujita et al 1998). Suppression of calcitonin gene expression in rats also prevents implantation (Zhu et al 1998a). There is a peak of calcitonin release from the pregnant rat uterus on day 4 just prior to implantation. Progesterone stimulates calcitonin release from the rat uterus, whereas oestradiol inhibits it (Zhu et al 1998b). Implantation is blocked by receptor antagonists for histamine in rats and rabbits (Brandon and Wallis 1977; Hoos and Hoffman 1983) and for interleukin-1 in mice (Simon et al 1994). Antagonists of histamine synthesis and of histamine release from mast cells also prevent implantation in the rabbit and mouse (Dey et al 1978; Dey and Johnson 1980; Dey 1981). Concentrations of mRNA for histidine decarboxylase are increased on days 3 and 4 of pregnancy in the mouse. This increase occurs in the uterine epithelial cells and is due to an action of progesterone but not oestradiol (Paria et al 1998). In rats, decidualization is prevented by an inhibitor of the angiotensin converting enzyme, inferring that angiotensin II is involved in this process (Squires and Kennedy 1992). An inhibitor of cathepsins B and L prevents implantation in mice (Afonso et al 1997). Monoclonal antibodies directed to the Ley oligosaccharide or to an early pregnancy factor-induced suppressor factor (EPF-S-1) disrupt implantation in mice (Athanasas et al 1995; Zhu et al 1995; Wang et al 1998). Inhibitors of nitric oxide synthase reduce the implantation rates in rats, suggesting that nitric oxide is involved in the implantation process (Biswas et al 1998; Chwalisz et al 1999; Duran-Reyes et al 1999). This is some of the evidence indicating which local factors appear essential for implantation to take place. What is the evidence, therefore, that prostaglandins have an essential role in the implantation process?
monkey (Nayak et al 1997). Dexamethasone, which can inhibit prostaglandin synthesis by reducing the activity of phospholipase (PL) A2 and by preventing the synthesis of prostaglandin H synthase-2 (PGHS-2), also inhibits implantation in the rat (Johnson and Dey 1980). The antiimplantation action of indomethacin is largely overcome by the simultaneous administration of prostaglandin (PG) E2 or PGF2a (Lau et al 1973; Saksena et al 1976; Hoffman 1978; El-Banna 1980; Gupta et al 1989; Chida and Mettler 1989). The outputs of PGE2, 6-ketoPGF1a and, to a lesser extent, PGF2a from the pregnant rat uterus superfused in vitro are significantly increased just prior to implantation (Figure 51.1; Poyser 1985). Indomethacin decreases prostaglandin concentrations in the rat uterus at the time of implantation (Garg and Chaudhuri 1983). PGE2 and PGF2a will induce blastocysts to implant in mice (Holmes and Gordashko 1980). These studies indicate that prostaglandins have a role in implantation, so what possible roles may they have in this process?
Spacing of Blastocysts In indomethacin-treated, pregnant rats in which implantation has been partially inhibited (i.e. there is a reduced number of implanted blastocysts), the implantation sites are not evenly spaced along the two uterine horns (Wellstead et al 1989). This finding indicates that, due to their spasmogenic activity on smooth muscle, prostaglandins are involved in the even spacing of blastocysts.
Blastocyst Hatching Before the blastocyst can implant, the zona pellucida surrounding it has to be shed. Various inhibitors of PGE2 synthesis and action prevent this ‘‘hatching process’’ in mice and hamsters (Biggers et al 1978; Baskar et al 1981; Hurst and MacFarlane 1981; Terranova and Dey 1982; van der Weiden et al 1993). Mouse blastocysts contain small quantities of PGE2 and have the necessary enzymes to synthesize prostaglandins (Racowsky and Biggers 1983; Niimura and Ishida 1987; Marshburn et al 1990; van der Weiden et al 1996). These findings are consistent with PGE2 being involved in the shedding of the zona pellucida. It has been suggested that, as PGE2 promotes water passage across epithelia, the PGE2 produced by the
EFFECTS OF NON-STEROIDAL ANTIINFLAMMATORY DRUGS (NSAIDs) AND THE INVOLVEMENT OF PROSTAGLANDINS Indomethacin delays, reduces or totally inhibits implantation, depending on the dose used, in rats, mice, hamsters, rabbits and ferrets when administered just prior to or during the ‘‘implantation window’’ (O’Grady et al 1972; Lau et al 1973; El-Banna et al 1976; Saksena et al 1976; Kennedy 1977; Evans and Kennedy 1978; Hoffman 1978; Hoffman et al 1978; El-Banna 1980; Holmes and Gordashko 1980; Phillips and Poyser 1981; Terranova and Dey 1982; Garg and Chaudhury 1983; Jones et al 1986; Mead et al 1988; Gupta et al 1989; Chida and Mettler 1989; Wellstead et al 1989; Paria et al 1991; Kanayama et al 1996). Diclofenac sodium (another NSAID) inhibits or delays implantation in the rhesus
Figure 51.1 Mean (+SEM, n=6) outputs of PGF2a, PGE2 and 6-ketoPGF1a from the early pregnant rat uterus superfused in vitro. {Significantly higher than corresponding value on day 2. From Poyser (1985), by permission
PROSTAGLANDINS IN IMPLANTATION blastocyst may induce water uptake into the blastocyst, causing the zona pellucida to burst (Biggers et al 1978). Blastocyst Development After hatching, the formation of trophoblast outgrowths takes place. The production of these outgrowths by mice blastocysts in vitro is inhibited by indomethacin, and this inhibition is largely overcome by PGE2 and/or PGF2a treatment. These findings indicate that prostaglandins are involved in blastocyst development (Chida and Mettler 1989). Vascular Permeability Indomethacin prevents the increase in vascular permeability which occurs at the implantation site in rats, mice, hamsters and rabbits, but not ferrets (Kennedy 1977; Evans and Kennedy 1978; Hoffman et al 1978; Phillips and Poyser 1981; Mead et al 1988; Gupta et al 1989; Paria et al 1991). The concentrations of PGE2, PGI2 and PGF2a are higher at the sites of increased vascular permeability than in the surrounding areas in the rat uterus at the time of implantation (Kennedy 1977; Kennedy and Zamecnik 1978). PGE2 concentrations are also increased at these sites in the pregnant hamster uterus (Evans and Kennedy 1978). PGE2, but not PGI2 or PGF2a, acting on the rat uterus produces an increase in vascular permeability, suggesting that PGE2 is the prostaglandin responsible for increasing the permeability of blood vessels at the site of implantation (Kennedy 1979a, 1979b, 1980). Decidual Cell Reaction (Decidualization) Decidualization is induced in the rat and mouse uterus, which has been adequately primed with progesterone and oestradiol, by applying an artificial stimulus (e.g. trauma or vegetable oil) to the endometrium. Indomethacin prevents this decidual cell reaction in response to an artificial stimulus (Castracane et al 1974; Tobert 1976; Sananes et al 1976, 1981; Rankin et al 1979; Jonsson et al 1979; Miller and Morchoe 1982a; Buxton and Murdoch 1982; Peleg and Lindner 1982). PGE2 or PGF2a induces the decidual cell reaction when placed in the suitably primed uterus of the rabbit or rat (Hoffman et al 1977; Sananes et al 1981; Kennedy and Lukash 1982; Miller and Morchoe 1982b; Kennedy 1986; Tawfik et al 1987a; Kennedy and Doktorick 1988; Kennedy and Ross 1993). PGE2 is generally more effective than PGF2a (Hoffman et al 1977; Kennedy 1986). PGE2 or PGF2a overcomes inhibition of decidualization produced by indomethacin (Kennedy 1985). Arachidonic acid also induces decidualization in the rat uterus and indomethacin inhibits this action of arachidonic acid, which suggest that endogenously produced prostaglandins are able to initiate the decidual cell reaction (Sananes et al 1981). Decidualization induced by artificial stimuli in the rat or mouse uterus is followed by a rapid rise (within 1–5 min) of PGE2 and PGF2a concentrations in the uterus. These increases are prevented by indomethacin treatment (Anteby et al 1975; Rankin et al 1979; Jonsson et al 1979; Kennedy 1983; Kennedy and Lukash 1982; Milligan and Lytton 1983). It is apparent that the decidual cell reaction is dependent on an increase in the production of prostaglandins by the uterus. CONTROL OF PROSTAGLANDIN PRODUCTION BY THE UTERUS Uterine Prostaglandin Synthesis During Early Pregnancy As prostaglandins are not stored in tissues, they have to be synthesized when they are required. The main pathway of
549
prostaglandin synthesis involves phospholipase (PL) A2 releasing arachidonic acid from phospholipids, and then prostaglandin H synthase (PGHS) converts the free arachidonic acid into PGH2. The PGH2 is then acted upon by any of five enzymes to produce the appropriate prostaglandin or thromboxane. Activation of PLA2 is the rate-limiting step, whereas the concentration of PGHS determines the amounts of prostaglandin(s) formed. The prostaglandin-synthesizing capacity of the uterus (which is indicative of the PGHS concentration) generally exhibits peaks on days 4 and 5 of pregnancy in the rat (Phillips and Poyser 1981; Garg and Pandhi 1981; Malathy et al 1986; Kogo et al 1991). Since similar changes occur in the pseudopregnant rat uterus, the general increase in the prostaglandin-synthesizing capacity of the uterus is not dependent upon the presence of the trophoblast (Fenwick et al 1977). Notwithstanding, there is evidence that the trophoblast to some extent directs where in the uterus these changes occur (Parr et al 1988). Immunohistochemical labelling shows there to be a low concentration of PGHS in the pregnant rat uterus on day 1. On day 4, there is increased labelling of PGHS in the luminal and glandular epithelium, in stromal cells adjacent to the luminal epithelium, and in blood vessels. On day 5, there is intense labelling of PGHS in the early differentiating stromal cells adjacent to the luminal epithelium at the implantation sites, whereas peripheral stromal cells exhibit only weak labelling of PGHS. This agrees with increased concentrations of prostaglandins being present at the implantation site compared with surrounding areas, as noted previously. By day 7, the differentiated cells of the primary decidual zone show only weak labelling of PGHS, whereas stromal cells in the secondary decidual zone are intensely labelled as the trophoblast penetrates into the endometrium (Parr et al 1988). The gross changes in prostaglandin-synthesizing capacity and prostaglandin output are generally controlled by oestradiol acting on a progesteroneprimed uterus, although progesterone does not necessarily have to be present for oestradiol to act (Pakrasi et al 1983, 1984; Brown et al 1984; Tawfik et al 1987b; Kogo et al 1991; Nakayama et al 1991). Tamoxifen treatment reduces the increase in uterine prostaglandin-synthesizing capacity seen during early pseudopregnancy (Fenwick et al 1977). In parallel with the increases in uterine prostaglandin-synthesizing capacity, PLA2 activity increases during early pregnancy around the time of implantation. These increases in PLA2 are also controlled by oestradiol acting on a uterus treated with progesterone, although progesterone may reduce the stimulant activity of oestradiol to some extent (Cox et al 1982; Pakrasi et al 1983, 1984; Novaro et al 1996). PLA2 activity in the early pregnant rabbit uterus is also highest at the time of implantation (Hoffman et al 1984). Quinacrine, a putative PLA2 inhibitor, reduces the number of blastocysts that implant in mice (Holmes et al 1988). Fertility is greatly impaired in female mice lacking the gene for cytosolic PLA2. Pregnancy failure in these mice often occurred around the time of implantation (Bonventre et al 1997), suggesting that the presence of cytosolic PLA2 is necessary for the normal production of prostaglandins at the time of implantation. Involvement of Growth Factors and Cytokines Oestradiol induces implantation in hypophysectomized rats maintained on progesterone. It has been shown that epidermal growth factor (EGF) can replace oestradiol in this response. The action of EGF in inducing implantation is inhibited by indomethacin, which suggests that prostaglandins mediate the action of EGF (Johnson and Chatterjee 1993a, 1993b, 1995). The concentration of EGF binding sites in the pregnant rat uterus increases after day 2 and reaches a peak on day 5 (Chakraborty et al 1989).
550
REPRODUCTIVE SYSTEM
In the mouse, oestradiol stimulates an increase in the concentration of EGF in luminal and glandular epithelial cells, but not in stromal cells (Huet-Hudson et al 1990). On the other hand, oestradiol increases the concentration of EGF receptor mRNA and protein in stromal cells, but not in luminal and glandular epithelial cells (Das et al 1994). EGF stimulates prostaglandin production by the stromal cells and not the epithelial cells of mouse uterus (Paria et al 1991). EGF also stimulates prostaglandin output from stromal cells of the rat uterus sensitized for the decidual cell reaction, partly by increasing cytosolic PLA2 levels (Bany and Kennedy 1995a; Bany et al 1999). These observations therefore lead to a ‘‘two-cell’’ hypothesis for PGE2 synthesis (Figure 51.2). The increase in oestradiol output prior to implantation stimulates EGF synthesis in epithelial cells and EGF receptor synthesis in stromal cells. The EGF released from epithelial cells combines with the EGF receptors on the stromal cells, which subsequently causes prostaglandin synthesis (particularly of PGE2) by the stromal cells to increase. The PGE2 released from the stromal cells may then be involved in the spacing of blastocysts and implantation. There are also EGF receptors in the myometrium (Das et al 1994), so EGF released from epithelial cells may act on these receptors to stimulate PGE2 production by myometrial cells. The PGE2 released by these cells may act on the circular smooth muscle to help in the spacing of the blastocysts. EGF also stimulates PGF2a output from stromal cells but in smaller quantities than PGE2 (Paria et al 1991; Bany and Kennedy 1995a). Although PGF2a will induce decidualization, it may have a less important role than other prostaglandins. Implantation is blocked in mice by an interleukin-1 (IL-1) receptor antagonist (Simon et al 1994). The concentration of IL-1 in the mouse uterus rapidly increases during early pregnancy and reaches a maximum on day 4 (De et al 1993). This IL-1 is produced by the epithelial cells. IL-1a acts on the stromal cells of both the mouse and rat uterus to stimulate PGE2 and PGF2a production although, as with EGF stimulation, PGE2 is synthesized in greater quantities than PGF2a (Jacobs and Carson 1993; Jacobs et al 1994; Bany and Kennedy 1995b). Thus, it appears that both EGF and IL-1a may mediate the stimulatory effect of oestradiol on prostaglandin production at the time of implantation in rats and mice by a process involving both epithelial and stromal cells. The effects of EGF and IL-1a on prostaglandin production by rat stromal cells are additive (Bany and Kennedy 1995b), which indicates that different intracellular pathways are involved.
Figure 51.2 The two-cell hypothesis by which oestradiol stimulates PGE2 synthesis by endometrial cells during early pregnancy in rats and mice. Oestradiol stimulates the output of epidermal growth factor (EGF) from epithelial cells and an increase in EGF receptor (EGF-R) concentration in stromal cells. The EGF produced by the epithelial cells stimulates PGE2 production by the stromal cells via combining with EGF receptors
Prostaglandin H Synthase Isoforms Arachidonic acid is converted into PGH2 by the action of prostaglandin H synthase (PGHS), of which there are two forms, PGHS-1 and PGHS-2. The question is raised as to whether the synthesis of prostaglandins by the uterus for implantation involves one or both forms of the enzyme. PGHS-1 mRNA in the pregnant mouse uterus is highly expressed in the epithelium on day 4 until the initiation of attachment reaction in the evening, after which expression falls. This PGHS-1 gene expression coincides with the generalized uterine oedema required for luminal closure, suggesting that PGHS-1 activity is involved (Chakraborty et al 1996). In contrast, PGHS-2 mRNA is expressed in the luminal epithelium and subepithelial stromal cells on the anti-mesometrial side of a uterine horn exclusively surrounding the blastocyst at the time of attachment on day 4. This increased expression persists into day 5. PGHS-2 mRNA is not expressed at the sites of blastocyst apposition in delayed implantation, during which the mice only receive progesterone. However, PGHS-2 mRNA is readily expressed at these sites after the delay has been terminated by treatment with oestradiol (Chakraborty et al 1996). Surprisingly, progesterone and oestradiol fail to increase PGHS-2 mRNA expression in the uterus of ovariectomized mice, whereas PGHS-1 mRNA is increased in the uterine epithelium following treatment with this combination of steroid hormones (Chakraborty et al 1996). EGF and IL-1a increase the concentration of prostaglandin H synthase (PGHS) in stromal cells obtained from the uterus of ovariectomized rats sensitized for the decidual cell reaction (Bany and Kennedy 1995a, 1995b). In these cells, EGF increases the concentration of both PGHS-1 and -2 mRNA and protein (Bany and Kennedy 1995c, 1997). IL-1a increases the concentration of mRNA for PGHS-2 but not for PGHS-1 (Bany and Kennedy 1995c, 1999). This observation is in agreement with the previous finding that IL-1a increases the concentration of PGHS-2 protein, but not PGHS-1 protein, in stromal cells of the mouse uterus (Jacobs et al 1994). In the early pregnant mouse uterus, PGHS-2 is localized to the implantation sites in the newly differentiated stromal cells at the time of blastocyst attachment. PGHS-1 is localized to the epithelial cells at this time (Jacobs et al 1994). In the uterus of mice lacking the gene for leukaemia inhibitory factor, PGHS-1 expression at the time of implantation is normal. PGHS-2 expression in these mice is present in the luminal epithelium but is lacking in the underlying stromal cells surrounding a blastocyst, indicating that activity of leukaemia inhibitory factor is necessary for the increase in PGHS-2 in stromal cells (Song et al 2000). The increase in PGHS-2 in stromal cells in response to a decidualization stimulus is apparently dependent upon activation of the p38 MAPkinase pathway (Scherle et al 2000). PGHS-2 concentrations are also increased in the endometrium at the implantation site in mink (Song et al 1998). All these studies suggest that increased production of prostaglandins by stromal cells at the implantation site is dependent upon the increased synthesis of PGHS-2 by these cells. PGHS-1 is largely confined to epithelial cells and may be responsible for basal uterine prostaglandin production. If so, this basal production is not essential for implantation, as experiments on gene knockout mice have shown. Female mice lacking the gene for PGHS-1 are fertile, and implantation is not impaired (Langenbach et al 1995). Thus, the presence of PGHS-1 in the mouse uterus is not essential for implantation. On the other hand, female mice lacking PGHS-2 are infertile (Dinchuk et al 1995). This is primarily due to lack of ovulation, since increased prostaglandin production by ovarian follicles, which is necessary for the ovulatory process, is dependent on the increased synthesis of PGHS-2 in the follicles (Sirois and Richards 1992; Sirois et al 1992). It is not possible, therefore, to
PROSTAGLANDINS IN IMPLANTATION study the importance of PGHS-2 in the uterus for implantation in normally mated, PGHS-2-deficient mice, since these mice do not become pregnant. This difficulty was overcome by examining the implantation rates of blastocysts transferred from normal mice to PGHS-2-deficient mice in which the uterus has been suitably primed with steroid hormones for implantation to take place. Implantation fails to occur in a very large majority of these mice and, where implantation does take place, these mice contain only one implanted blastocyst each (Lim et al 1997). These and other studies indicate that implantation, as well as ovulation, is impaired in PGHS-2-deficient mice (Matsumoto et al 2001). In addition, the treatment of mice containing the PGHS-2 gene with a selective inhibitor of PGHS-2 (DuP697) during early pregnancy considerably reduces the number of blastocysts that implant (Lim et al 1997). However, the treatment of rats with a NS-398 (another selective inhibitor of PGHS-2), in doses that reduce pain and inflammation, fails to reduce the implantation rate (Poyser 1999). Either higher doses of NS-398 are required to affect implantation or lack of PGHS-2 in the rat uterus may be compensated for by PGHS-1. Studies on mice have shown that simultaneous inhibition of PGHS-1 and PGHS-2 produces more severe effects on early pregnancy than inhibition of either isoform alone, and there is evidence that the PGHS-1 isoform compensates for the lack of the PGHS-2 isoform and vice versa (Reese et al 1999, 2001). It is clear that PGHS-2 is involved in implantation, which is consistent with the observations described earlier that the concentrations of PGE2 and PGHS-2 are higher at the site of the implanting blastocyst than in surrounding areas. These findings also imply that the blastocyst has some control over the expression of the PGHS-2 gene in the uterus (Chakraborty et al 1996). How the blastocyst exerts this control is far from clear. Rabbit and rat blastocysts are able to produce oestrogens (Dey and Dickmann 1974; Dickmann and Dey 1974; Dickmann et al 1975, 1977; Wu and Lin 1982), so a high local concentration of oestradiol may be responsible for this effect. There is evidence that histamine produced by the blastocyst has a role in implantation (Ferrando and Nalbandov 1968; Brandon and Wallis 1977; Dey et al 1978, 1979a, 1979b; Brandon and Raval 1979; Dey and Johnson 1980; Johnson and Dey 1980; Dey 1981; Hoos and Hoffman 1983; Ohta 1985). Whether histamine affects PGHS-2 expression at the site of implantation is not known. PROSAGLANDIN PRODUCTION BY THE BLASTOCYST As outlined earlier, prostaglandins produced by the blastocyst may be involved in the shedding of the zona pellucida, as indicated by studies in mice (Racowsky and Biggers 1983; Niimura and Ishida 1987; Marshburn et al 1990; van der Weiden et al 1996). It is possible that prostaglandins produced locally by the blastocyst have other roles in the implantation process. Rabbit blastocysts contain increased amounts of PGE2 and PGF2a around the time of implantation. They are also capable of synthesizing increased quantities of PGE2 and PGF2a on days 6 and 7 of pregnancy (Dickmann and Spilman 1975; Dey et al 1980; Sharma 1980; Pakrasi and Dey 1982; Racowsky and Biggers 1983; Harper et al 1983, 1989; Pakrasi et al 1985; Kasamo et al 1986). Prostaglandins are also synthesized by the blastocyst of sheep, cow, pig and monkey prior to implantation (Shemesh et al 1979; Hyland et al 1982; Marcus 1981; Lewis et al 1982; Lewis and Waterman 1983, 1985; Davis et al 1983; Neulen 1991). The increase in prostaglandin by sheep blastocysts is associated with an increase in the concentration of PGHS-2, but not of PGHS-1 (which is undetectable), in the trophoblast cells between days 8
551
and 17 of pregnancy. Maximum levels of PGHS-2 are present between days 14 and 16, which is just prior to the attachment of the blastocyst to the uterus on days 17 and 18 (Charpigney et al 1997). The factors controlling these changes in PGHS-2 expression in the trophoblast are not known, but this change is developmentally regulated. In this context, receptors for EGF are present on the cell surface of mouse blastocysts, suggesting that growth factors may have a role in blastocyst development and, therefore, prostaglandin synthesis (Paria and Dey 1990; Wiley et al 1992). The expression of these EGF receptors is controlled by the ovarian steroid hormones, with oestradiol having a stimulatory action (Paria et al 1992).
INVOLVEMENT OF CYCLIC ADENOSINE-3,5-MONOPHOSPHATE (cAMP) Oestradiol increases cAMP concentrations in the uterus, and cAMP mimics the early action of oestradiol in stimulating RNA synthesis in the blastocyst and uterus of the rat and mouse (Mahla and Prasad 1970; Holmes and Bergstrom 1976). Implantation is induced in the progesterone-primed uterus of the mouse by cAMP (or its dibutyryl derivative), although the pregnancies are not maintained to term (Holmes and Bergstrom 1975; Webb 1975). Nevertheless, these findings suggest that cAMP is involved in the implantation process. PGE2 and PGI2 increase cAMP concentrations in many tissues, so this raises the question as to whether the actions of prostaglandins in implantation are mediated by cAMP. The concentration of cAMP is raised at the site of increased vascular permeability in the rat uterus (Vilar-Rajus et al 1982). PGE2 increases the concentration of cAMP in the rat uterus suitably primed for an implantation response. However, cAMP or its more stable analogues fail to cause an increase in vascular permeability in the rat uterus (Kennedy 1983; Johnston and Kennedy 1984). Therefore it appears that, although PGE2 increases cAMP formation, the action of PGE2 in increasing vascular permeability at the implantation site is not mediated through cAMP. The application of deciduogenic stimulus to the suitably primed rat or mouse uterus results in increases in both adenyl cyclase activity and cAMP concentration (Rankin et al 1977; Sanders et al 1986). However, cAMP applied to the mouse uterus only results in a mild decidual reaction in some animals (Leroy et al 1974). This would suggest that cAMP does not mediate decidualization of the uterus produced by prostaglandins. However, this does not rule out the possibility that cAMP mediates some of the actions of PGE2. An early event in the decidual cell reaction is an increase in the activity of alkaline phosphatase in the endometrial stromal cells. This increase in enzyme activity is prevented by indomethacin; treatment with PGE2 overcomes this inhibition, indicating that prostaglandins have a physiological role in stimulating uterine alkaline phosphatase activity (Yee and Kennedy 1988). This stimulatory action of PGE2 is mediated by cAMP (Yee and Kennedy 1991, 1993). Therefore, PGE2-induced cAMP formation appears to have a role in implantation in rodents. In rabbits, adenyl cyclase activity is higher in endometrial stromal cells than in epithelial cells, although activity of the enzyme in both cell types is lower on day 6.5 than on day 1. However, there is increased activity of adenyl cyclase in the endometrium at the implantation site on day 9. PGE2 stimulates adenyl cyclase activity in stromal and epithelial cells, although it has a greater effect on stromal cells (Fortier et al 1987, 1989, 1990). Thus, PGE2-induced cAMP formation in the endometrium of rabbits may have a role in the processes immediately following implantation.
552
REPRODUCTIVE SYSTEM
PROSTAGLANDIN RECEPTORS INVOLVED IN IMPLANTATION Studies performed on pseudopregnant rats showed that specific binding of PGE2 to endometrial membranes is not detectable on day 2, is low on days 3 and 4, and then increases greatly on day 5 to reach a maximum on day 6 (Kennedy et al 1983a). There are no high-affinity PGE2 binding sites in the endometrium during the oestrous cycle (Kennedy et al 1983b). The treatment of ovariectomized rats with oestradiol and progesterone, alone or in combination, showed that the appearance of high-affinity PGE2 binding sites in the endometrium is solely progesterone-dependent (Kennedy et al 1983b). These binding sites first appeared after 48 h of progesterone treatment (Kennedy et al 1983b), which is the same time course for them to appear during pseudopregnancy in response to the increase in plasma concentrations of endogenous progesterone. The increase in the PGE2 binding sites occurs in the stromal cells and not the epithelial cells of ovariectomized, progesterone-treated rats (Kennedy et al 1983b). These binding studies were performed before receptors for PGE2 were isolated and characterized, so it is not clear which PGE2 receptors were involved in these binding studies. It is now known that there are four receptors for PGE2 (EP1–EP4). Studies on PGE2 receptor gene expression in the uterus of pregnant and pseudopregnant mice around the expected time of implantation have been performed (Yang et al 1997; Katsayama et al 1997; Lim and Dey 1997). There is increased expression of the mRNA for the EP2 receptor in the luminal epithelium on days 4 and 5 of pregnancy and pseudopregnancy. In ovariectomized mice, progesterone treatment upregulates EP2 mRNA levels. Oestradiol augments the action of progesterone although, on its own, oestradiol downregulates the mRNA levels for EP2. The level of EP3 mRNA increases in the circular smooth muscle of pregnant mice up to day 5. There is also an increase in EP3 mRNA levels in a subpopulation of cells in the stromal bed at the mesometrial side of the uterus. There are increased levels of mRNA for the EP4 receptor in luminal and, to a lesser extent, glandular epithelial cells, and in the stromal cells of mice on days 3 and 4 of pregnancy. By day 5 of pregnancy, EP4 receptor gene expression is localized in the luminal epithelium and decidualizing stroma around the implanting blastocyst (Yang et al 1997). Similar changes in EP receptor gene expression in pseudopregnant mice and in ovariectomized mice treated with progesterone and oestradiol show that these changes are controlled by the ovarian steroid hormones. Progesterone exerts the main control (Yang et al 1997; Katsayama et al 1997). It has been suggested from these studies in mice that expression of the EP3 receptor in the circular smooth is involved in myometrial activity and the spacing of the blastocysts. Expression of the EP2 receptor in the luminal epithelium may be involved in blastocyst implantation signalling, and EP4 expression in the endometrial stroma may be involved in the decidualization of the stromal cells (Katsayama et al 1997). In the pseudopregnant rat uterus, mRNA for EP2, EP3 and EP4 but not EP1 receptors is detectable in the endometrium around the time of implantation. The levels of mRNA for EP2 and EP3, but not for EP4, increase from day 1 to day 5 of pseudopregnancy. On day 5, the mRNA for EP2 receptors is localized to the luminal epithelium and the antimesometrial pole, whereas the mRNA for EP3 receptors is localized in the glandular epithelium and stroma. The mRNA for EP4 receptors is concentrated in the antimesometrial stroma. The cell-specific expression of EP2, EP3 and EP4 receptors on day 5 in these rats is dependent upon oestradiol given on day 4. The application of a decidugenic stimulus on day 5 results in a decrease in the levels of the mRNA for these three EP receptors (Papav and Kennedy 2000).
Many studies indicate the PGE2 produced in the pregnant uterus, particularly by stromal cells, of several species may act on EP2, EP3 and EP4 receptors to cause many of the localized changes associated with the implantation process. However, genetically modified mice lacking EP2, EP3 or EP4 receptors (or even EP1 receptors) do not exhibit any failure of implantation (Fleming et al 1998; Segi et al 1998; Ushikubi et al 1998; Audaly et al 1999; Kennedy et al 1999; Tilley et al 1999). Mice lacking the EP2 receptor have reduced fertility, but this is due to an inhibition of ovulation and fertilization rather than an inhibition of the subsequent reproductive processes during pregnancy (Kennedy et al 1999; Tilley et al 1999; Matsumoto et al 2001). These studies suggest that lack of only one of the receptors for EP2, EP3 or EP4 does not have a detrimental effect on implantation and that this single loss may be compensated for by other means. It would be interesting to study the multiple loss of two or more EP receptor subtypes on the implantation process to see whether it is disrupted. Most of the evidence suggests that PGE2 is the main prostaglandin involved in implantation. However, the lack of failure of the implantation process in mice deficient in any one of the EP receptors may indicate that a prostaglandin other than PGE2 is the major one involved. Implantation proceeds normally in mice lacking the receptor for PGF2a (i.e. the FP receptor), suggesting that PGF2a is not necessary for implantation. Parturition fails to take place in these mice, since the lack of the FP receptor in the ovary prevents PGF2a, produced by the pregnant uterus at term, from causing luteolysis and the subsequent fall in ovarian progesterone output necessary for parturition to take place (Sugimoto et al 1997). Therefore, if earlier in pregnancy PGF2a is involved in the implantation process, it cannot be acting via the FP receptor. PGI2 is the major prostaglandin synthesized by homogenates of the uterus removed from rats on day 5 of pseudopregnancy and pregnancy (Phillips and Poyser 1981). PGI2 is released in increased quantities from the pregnant, superfused rat uterus around the time of implantation (Figure 50.1; Poyser 1985), and is present in increased concentrations at the implantation site in the rat uterus (Kennedy 1977; Kennedy and Zamecnik 1978). PGI2 is the most abundant prostaglandin at the implantation site in the mouse uterus. The administration of carbaprostacyclin (a PGI2 analogue), but not PGE2, to suitably impregnated PGHS-2-deficient mice increases the implantation rate and the incidence of a decidual response to a level seen in wild-type mice (Lim et al 1999). These observations suggest that PGI2 may be the main prostaglandin involved in the implantation process. However, genetically modified mice lacking the IP receptor for PGI2 are fertile and therefore show no failure of implantation (Murata et al 1997). Consequently, either PGI2 is not involved in the implantation process or another receptor for PGI2 is involved. Cicaprost (another PGI2 analogue) does not increase the implantation rate in PGHS-2-deficient mice, although it is an agonist at the IP receptor (Lim et al 1999). Cicaprost, unlike PGI2 and carbaprostacyclin, does not activate the nuclear receptor, peroxisome proliferator-activated receptor d (PPARd) (see Lim and Dey 2000). The administration of L-165041, a selective agonist for PPARd, increases the implantation rate in PGHS-2-deficient mice to a similar extent as treatment with carbaprostacyclin (Lim et al 1999). It has been proposed, therefore, that PGI2 is the main uterine prostaglandin involved in implantation and decidualization, with its actions being mediated by PPARd (Lim et al 1999; Lim and Dey 2000; Matsumoto et al 2001). PGE2 may only have an ancillary role since, when administered with carbaprostacyclin to PGHS-2deficient mice, it increases embryonic and decidual growth (Lim et al 1999). The receptor(s) involved in this action of PGE2 are not apparent.
PROSTAGLANDINS IN IMPLANTATION INVOLVEMENT OF OTHER EICOSANOIDS The production of leukotriene (LT) B4 and LTC4 by the early pregnant rat uterus shows a similar profile to that for PG production, with peaks on days 4 and 5 (Malathy et al 1986). Leukotriene production, like prostaglandin production, appears to be stimulated by oestradiol acting on a progesterone-primed uterus (Tawfik et al 1987b). Decidualization of the rat uterus is induced by LTC4, and the effects of LTC4 and PGE2 administered together on the decidual cell reaction are greater than when the two eicosanoids are added separately (Tawfik et al 1987a). The intraluminal infusion of the leukotriene receptor inhibitor, FPL 55712, prevents decidualization in the rat, and the simultaneous infusion of LTC4 with FPL 55712 overcomes the inhibitory effect of the antagonist (Tawfik and Dey, 1988). These studies indicate that the leukotrienes, as well as prostaglandins, have a role in the implantation process. However, it has been reported that the 5lipoxygenase inhibitor, AA861, does not prevent implantation in mice (Pakrasi 1997), which casts some doubt on a role for leukotrienes in the implantation process. Finally, anandamide, a derivative of arachidonic acid, is an endogenous ligand for both the brain-type (CB1-R) and spleen type (CB2-R) cannabinoid receptors. The periimplantation mouse uterus contains the highest concentrations of anandamide discovered in a mammalian tissue. The anandamide concentration changes during pregnancy, with a decrease in concentration being associated with an increase in receptivity of the uterus for implantation and an increase in concentration being associated with a decrease in receptivity. Anandamide concentrations in the uterus are highest at the interimplantation sites and lowest at the implantation sites. CP 55940, a cannabinoid receptor agonist, reduces the rate of hatching of mouse blastocysts from the zona pellucida in vitro and, when administered systemically, prevents implantation in mice. These inhibitory effects of CP 55940 are prevented by GR141716, a CB1-R antagonist (Schmid et al 1997). These studies indicate that activation of cannabinoid receptors in the uterus prevents implantation, and that a reduction in this activation is necessary for implantation to proceed. SUMMARY There is much evidence that prostaglandins are necessary for implantation to occur, especially in rodents and rabbits. Whether prostaglandins are necessary for implantation in other species, including humans, needs further study. Prior to implantation, prostaglandins may be involved in the spacing of the blastocysts along the uterine horns, the hatching of the blastocyst from the zona pellucida, and the subsequent development of the blastocyst. During the implantation process, prostaglandins are involved in the increase in vascular permeability at the implantation site and in the transformation of stromal cells into decidual cells (the ‘‘decidual cell reaction’’). Increased prostaglandin production occurs in the stromal cells of the uterus at the time of implantation in rodents and rabbits. This increase is due to oestradiol acting on a progesterone-primed uterus. The action of oestradiol may be mediated by growth factors (e.g. EGF) and cytokines (e.g. IL-1) produced by the epithelial cells. However, as prostaglandin production is higher at the implantation site than in surrounding areas, the close proximity of the blastocyst has some control over endometrial prostaglandin synthesis. In fact, the blastocysts may also produce prostaglandins for the process of implantation. Much evidence has accumulated over the years that PGE2 is the main prostaglandin involved in implantation. However, the finding that implantation is not detrimentally affected in mice lacking EP1, EP2, EP3 or EP4 receptors raises serious doubts about a preeminent role for PGE2 in this process. There is now some
553
evidence that PGI2 may be the major prostaglandin involved in implantation, with its actions being mediated by the nuclear PPARd and not the more classical IP receptor.
REFERENCES Afonso A, Romagnano L and Babiarz B (1997) The expression and function of cystatin C and cathepsin B and cathepsin L during mouse embryo implantation and placentation. Development, 124, 3415–3426. Anteby SO, Bauminger S and Zor U (1975) Prostaglandin synthesis in decidual tissue of the rat uterus. Prostaglandins, 10, 991–999. Athanasas Platsis S, Hoskin MJ, Rolfe BE et al (1995) A monoclonal antibody to an early pregnancy factor-induced suppressor factor (EPFS-1) disrupts implantation in mice. Am J Reprod Immunol, 33, 243–252. Audaly LP, Nguyen M, Koller BH and Caffman TM (1999) The prostaglandin EP4 receptor subtype is an essential component of blood pressure regulation in males but not females. FASEB J, 13, A1549. Bany BM and Kennedy TG (1995a) Interleukin-1a regulates prostaglandin production and cyclooxygenase activity in sensitized rat endometrial stromal cells in vitro. Biol Reprod, 53, 126–132. Bany BM and Kennedy TG (1995b) Regulation by epidermal growth factor of prostaglandin and cyclooxygenase activity in sensitized rat endometrial stromal cells. J Reprod Fertil, 104, 57–62. Bany BM and Kennedy TG (1995c) Regulation of cyclooxygenase activity by epidermal growth factor and interleukin-1a in rat endometrial stromal cells. J Reprod Fertil, abstr ser 15, 57. Bany BM and Kennedy TG (1997) Regulation of cyclooxygenase gene expression in rat endometrial stromal cells: the role of epidermal growth factor. Dev Genet, 21, 109–115. Bany BM and Kennedy TG (1999) Role of interleukin 1 in the regulation of cyclooxygenase gene expression in rat endometrial stromal cells. J Reprod Fertil, 115, 125–131. Bany BM, Schultz GA and Kennedy TG (1999) Regulation of cytosolic phospholipase A2 in rat endometrial stromal cells; the role of epidermal growth factor. Mol Reprod Dev, 52, 335–340. Baskar JF, Torchiana DF, Biggers JD et al (1981) Inhibition of hatching of mouse blastocysts in vitro by various prostaglandin antagonists. J Reprod Fertil, 63, 359–363. Benson GV, Lim H, Paria BC et al (1996) Mechanisms of reduced fertility in Hoxa-10 mutant mice: uterine homeosis and loss of maternal Hoxa10 expression. Development, 122, 2687–2696. Bhatt BM and Bullock DW (1974) Binding of oestradiol to rabbit blastocysts and its possible role in implantation. J Reprod Fertil, 39, 65–70. Biggers JD, Leonov BV, Baskar JF and Fried J (1978) Inhibition of hatching of mouse blastocysts in vitro by prostaglandin antagonists. Biol Reprod, 19, 519–533. Biswas S, Kabir SN and Pal AK (1998) The role of nitric oxide in the process of implantation in rats. J Reprod Fertil, 114, 157–161. Bonventre JV, Huang Z, Taheri MR et al (1997) Reduced fertility and postischaemic brain injury in mice deficient in cytosolic phospholipase A2. Nature, 390, 622–625. Brandon JM and Raval PJ (1979) Interaction of oestrogen and histamine during ovum implantation in the rat. Eur J Pharmacol, 57, 171–177. Brandon JM and Wallis RM (1977) Effect of mepyramine, a histamine H1-, and burimamide, a histamine H2-receptor antagonist, on ovum implantation in the rat. J Reprod Fertil, 50, 251–254. Brown C, Gosden RG and Poyser NL (1984) Effects of age and steroid treatment on the rat uterus in relation to implantation. J Reprod Fertil, 70, 649–656. Buxton LE and Murdoch RN (1982) Lectins, calcium ionophore A23187 and peanut oil as deciduogenic agents in the uterus of pseudopregnant mice: effects of tranylcypromine, indomethacin, iproniazid and propranolol. Austral J Biol Sci, 35, 63–72. Castracane VD, Saksena SK and Shaikh AA (1974) Effect of IUDs, prostaglandins and indomethacin on decidual cell reaction in the rat. Prostaglandins, 6, 397–403. Chakraborty C, Tawfik OW and Dey SK (1989) Epidermal growth factor binding in the rat uterus during the peri-implantation period. Biochem Biophys Res Commun, 153, 564–569.
554
REPRODUCTIVE SYSTEM
Chakraborty C, Das SK, Wong J and Dey SK (1996) Developmental expression of the cyclooxygenase-1 and cyclooxygenase-2 genes in the peri-implantation mouse uterus and their differential regulation by the blastocyst and ovarian steroids. J Mol Endocrinol, 16, 107–122. Charpigny G, Reinaud P, Tomby JP et al (1997) Cyclooxygenase-2 unlike cyclooxygenase-1 is highly expressed in ovine embryos during the implantation period. Biol Reprod, 57, 1032–1040. Chida S and Mettler L (1989) Effects of indomethacin, prostaglandin E2 and prostaglandin F2a on mouse blastocyst attachment and trophoblastic outgrowths in vitro. Prostaglandins, 37, 411–416. Chwalisz K, Winterhager E, Thienel T and Garfield RE (1999) Synergistic role of nitric oxide and progesterone during the establishment of pregnancy in the rat. Hum Reprod, 14, 542–552. Cox C, Cheng HC and Dey SK (1982) Phospholipase A2 activity in the rat uterus during early pregnancy. Prostagland Leukotrienes Med, 8, 375– 382. Das SK, Tsukamura H, Paria BC et al (1994) Differential expression of epidermal growth factor receptor (EGF-R) gene and regulation of EGF-R bioactivity by progesterone and oestrogen in the adult mouse uterus. Endocrinology, 134, 971–981. Das N, Wang J and Dey SK (1997) Uterine preparation for implantation in the mouse is associated with coordinate expression of estrogenresponse finger protein and estrogen receptor. Mol Reprod Dev, 46, 499–506. Das SK, Lim H, Paria BC and Dey SK (1999) Cyclin D3 in the mouse uterus is associated with decidualization during early pregnancy. Mol Endocrinol, 22, 91–101. Davis DL, Pakrasi PL and Dey SK (1983) Prostaglandins in swine blastocysts. Biol Reprod, 128, 1114–1118. De M, Sanford TR and Wood GW (1993) Expression of interleukin 1, interleukin 6 and tumour necrosis factor a in mouse uterus during the peri-implantation period of pregnancy. J Reprod Fertil, 97, 83–89. De Hertough R, Ekka E, Vanderheyden I and Glorieux B (1986) Estrogen and progesterone receptors in the implantation sites and interembryonic segments of rat uterus endometrium and myometrium. Endocrinology, 119, 680–684. Dey SK (1981) Role of histamine in implantation: inhibition of histidine decarboxylase induces delayed implantation in the rabbit. Biol Reprod, 24, 867–869. Dey SK and Dickmann Z (1974) D5-3b-Hydroxysteroid dehydrogenase activity in rat embryos on days 1 through 7 of pregnancy. Endocrinology, 95, 321–322. Dey SK and Johnson DC (1980) Histamine formation by mouse preimplantation embryos. J Reprod Fertil, 60, 457–460. Dey SK, Dickmann Z and Sen Gupta JS (1976) Evidence that the maintenance of early pregnancy in the rabbit requires ‘blastocyst estrogen’. Steroids, 28, 481–485. Dey SK, Villanueva C, Chien SM and Crist RD (1978) The role of histamine in implantation in the rabbit. J Reprod Fertil, 53, 23–26. Dey SK, Johnson DC and Santos JG (1979a) Is histamine production by the blastocyst required for implantation in the rabbit? Biol Reprod, 21, 1169–1173. Dey SK, Villanueva C and Abdou NI (1979b) Histamine receptors on rabbit and endometrial cell membranes. Nature, 278, 648–649. Dey SK, Chien SM, Cox CL and Crist RD (1980) Prostaglandin in the rabbit blastocyst. Prostaglandins, 19, 449–453. Dickmann Z and Dey SK (1974) Steroidogenesis in the preimplantation rat embryo and its possible influence on morula-blastocyst transformation and implantation. J Reprod Fertil, 37, 91–93. Dickmann Z and Spilman CH (1975) Prostaglandins in rabbit blastocysts. Science, 190, 997. Dickmann Z, Dey SK and Sen Gupta JS (1975) Steroidogenesis in rabbit preimplantation embryos. Proc Natl Acad Sci USA, 72, 298–300. Dickmann Z, Sen Gupta JS and Dey SK (1977) Does ‘‘blastocyst estrogen’’ initiate implantation? Science, 195, 687–688. Dinchuk JE, Car BD, Focht RJ et al (1995) Renal abnormalities and altered inflammatory response in mice lacking cyclooxygenase 2. Nature, 378, 406–409. Duran-Reyes G, Gomez-Melendez MR, Morali-de la Brena G et al (1999) Nitric oxide synthesis inhibition suppresses implantation and decreases cGMP concentrations and protein peroxidation. Life Sci, 65, 2259– 2268. El-Banna AA (1980) The degenerative effect on rabbit implantation sites by indomethacin. 1. Timing of indomethacin action, possible effects of
uterine proteins and the effect of replacement doses of PGE2. Prostaglandins, 20, 587–600. El-Banna AA, Sacher B and Schilling E (1976) Effect of indomethacin on egg transport and pregnancy in the rabbit. J Reprod Fertil, 46, 375– 378. Evans CA and Kennedy TG (1978) The importance of prostaglandin synthesis for the initiation of blastocyst implantation in the hamster. J Reprod Fertil, 54, 255–261. Fenwick L, Jones RL, Naylor B et al (1977) Production of prostaglandins by the pseudopregnant rat uterus, in vitro, and the effect of tamoxifen with the identification of 6-keto-prostaglandin F1a as a major product formed. Br J Pharmacol, 59, 191–199. Ferrando G and Nalbandov AV (1968) Relative importance of histamine and estrogen on implantation in the rat. Endocrinology, 83, 933–937. Finn CA and Porter DG (1975) The Uterus. London: Paul Elik (Scientific Books), 57–93. Fleming EF, Athirakil K, Oliverio MI et al (1998) Urinary concentrating function in mice lacking EP3 receptors for prostaglandin E2. Am J Physiol, 275, F955–F961. Fortier MA, Boulanger M, Boulet AP and Lambert RD (1987) Cellspecific localisation of prostaglandin E2-sensitive adenylate cyclase in rabbit endometrium. Biol Reprod, 36, 1025–1034. Fortier MA, Boulet AP and Lambert RD (1989) Regulation of adenyl cyclase activity and stimulation response in relation to endometrial receptivity in the rabbit. J Reprod Fertil, 85, 443–450. Fortier MA, Boulet AP, Dugre FJ and Lambert RD (1990) Local alteration in adenylate cyclase activity and stimulation response at implantation site in rabbit endometrium during early pregnancy. Biol Reprod, 42, 106–113. Fujita K, Nakayama T, Takabatake K et al (1998) Administration of thymocytes derived from non-pregnant mice induces an endometrial receptive stage and leukemia inhibitory factor expression in the uterus. Hum Reprod, 13, 2888–2894. Garg SK and Chaudhury RR (1983) Evidence for a possible role of prostaglandins in implantation in rats. Arch Int Pharmacodyn Therap, 262, 299–307. Garg SK and Pandhi P (1981) Uterine prostaglandin levels in pregnant ovariectomized rats. Ind J Med Res, 73, 782–786. Gerber GB, Jacquet P, Leonard A and Maes J (1979) Evolution des taux d’oestradiol, de progeste´rone et de prostaglandines E et F2a durant le gestation chez la souris. Compt Rend Soc Biol, 173, 644–659. Gupta A, Huet YM and Dey SK (1989) Evidence for prostaglandins and leukotrienes as mediators of phase I of oestrogen action in implantation in the mouse. Endocrinology, 124, 546–548. Harper MJK, Norris CJ and Rajkumar K (1983) Prostaglandin release by zygotes and endometria of pregnant rabbits. Biol Reprod, 28, 350–362. Harper MJK, Jones MA, Norris CJ and Woodard DS (1989) Prostaglandin synthesis by day-6 rabbit blastocysts in vitro. J Reprod Fertil, 86, 315–325. Hoffman LH (1978) Antifertility effects of indomethacin during early pregnancy in the rabbit. Biol Reprod, 18, 148–153. Hoffman LH, Strong GB, Davenport GR and Frolich JC (1977) Deciduogenic effects of prostaglandins in the pseudopregnant rabbit. J Reprod Fertil, 50, 231–237. Hoffman LH, DiPetro DL and McKenna TJ (1978) Effects of indomethacin on uterine capillary permeability and blastocyst development in rabbits. Prostaglandins, 15, 823–828. Hoffman LH, Davenport GR and Brash AR (1984) Endometrial prostaglandins and phospholipase activity related to implantation in rabbits: effects of dexamethasone. Biol Reprod, 30, 544–555. Hogarth P (1978) Biology of Reproduction. Glasgow: Blackie, 77–89. Holmes PV and Bergstrom S (1975) Induction of blastocyst implantation in mice by cyclic AMP. J Reprod Fertil, 43, 329–332. Holmes PV and Bergstrom S (1976) Cyclic adenosine monophosphateinduced changes in the surface morphology of diapausing blastocysts and the effects on implantation. Am J Obstet Gynecol, 124, 301–306. Holmes PV and Gordashko BJ (1980) Evidence of prostaglandin involvement in blastocyst implantation. J Embryol Exp Morphol, 55, 109–122. Holmes PV, Hellberg P and Sjoblom R (1988) Evidence that establishment of pregnancy requires activation of lipoxygenase and phospholipase A2. Experentia, 44, 772–773.
PROSTAGLANDINS IN IMPLANTATION Hoos PC and Hoffman LH (1983) Effect of histamine receptor antagonists and indomethacin on implantation in the rabbit. Biol Reprod, 29, 833– 840. Huet-Hudson YM, Chakraborty C, De SK et al (1990) Estrogen regulates the synthesis of epidermal growth factor in mouse uterine epithelial cells. Mol Endocrinol, 4, 510–523. Hurst PR and MacFarlane DW (1981) Further effects of non-steroidal antiinflammatory compounds on blastocyst hatching in vitro and implantation rates in the mouse. Biol Reprod, 25, 777–784. Hyland JH, Manns JG and Humphrey WD (1982) Prostaglandin production by ovine embryos and endometrium in vitro. J Reprod Fertil, 65, 299–304. Jacobs AL and Carson DD (1993) Uterine epithelial cell secretion of interleukin-1a induces prostaglandin E2 (PGE2) and F2a secretion by uterine stromal cells. Endocrinology, 132, 300–308. Jacobs AL, Hwang D, Julian J and Carson DD (1994) Regulated expression of prostaglandin endoperoxide synthase-2 by uterine stroma. Endocrinology, 135, 1807–1815. Johnson DC and Chatterjee S (1993a) Embryo implantation in the rat uterus induced by epidermal growth factor. J Reprod Fertil, 99, 557– 559. Johnson DC and Chatterjee S (1993b) Epidermal growth factor (EGF) replaces estradiol for the initiation of embryo implantation in the hypophysectomized rat. Placenta, 14, 429–438. Johnson DC and Chatterjee S (1995) The role of arachidonic acid and/or its metabolites in embryo implantation initiated by epidermal growth factor. Prostagland Leukotrienes Essent Fatty Acids, 52, 29–34. Johnson DC and Dey SK (1980) Role of histamine in implantation: Dexamethasone inhibits estradiol-induced implantation in the rat. Biol Reprod, 22, 1136–1141. Johnston MEA and Kennedy TG (1984) Estrogen and uterine sensitization for the decidual cell reaction in the rat: role of prostaglandin E2 and adenosine 3’,5’-cyclic monophosphate. Biol Reprod, 31, 959–966. Jones M, Cao Z-D, Anderson W et al (1986) Capillary permeability changes in the uteri of recipient rabbits after transfer of blastocysts from indomethacin-treated donors. J Reprod Fertil, 78, 261–273. Jonsson HT, Rankin JC, Ledford BE and Baggett B (1979) Uterine prostaglandin levels following stimulation of the decidual cell reaction: effects of indomethacin and tranylcypromine. Prostaglandins, 18, 847– 857. Kanayama K, Osada H, Nariari K and Endo T (1996) Inhibitory effects of indomethacin on implantation and its related phenomena. J Int Med Res, 24, 358–362. Kasamo M, Ishikawa M, Yamashita K et al (1986) Possible role of prostaglandin F in blastocyst implantation. Prostaglandins, 31, 321– 336. Katsayama M, Sugimoto Y, Morimoto K et al (1997) Distinct cellular localization of the messenger ribonucleic acid for prostaglandin E receptor subtypes in the mouse uterus during pseudopregnancy. Endocrinology, 138, 344–350. Kennedy CR, Zhang Y, Brandon S et al (1999) Salt-sensitive hypertension and reduced fertility in mice lacking the EP2 receptor. Nature Medicine, 5, 217–220. Kennedy TG (1977) Evidence for a role of prostaglandins in the initiation of blastocyst implantation in the rat. Biol Reprod, 16, 286–291. Kennedy TG (1979a) Prostaglandins and increased endometrial vascular permeability resulting from the application of an artificial stimulus to the uterus of the rat sensitized for the decidual cell reaction. Biol Reprod, 20, 560–566. Kennedy TG (1979b) Does prostaglandin I2 (PGI2) mediate the increased endometrial vascular permeability which results from the application of a deciduogenic stimulus to the sensitized rat uterus? Biol Reprod, 20(suppl 1), 99A. Kennedy TG (1980) Estrogen and uterine sensitization for the decidual cell reaction: role of prostaglandins. Biol Reprod, 23, 955–962. Kennedy TG (1983) Prostaglandin E2, adenosine 3’,5’-cyclic monophosphate and changes in endometrial vascular permeability in rat uteri sensitized for the decidual cell reaction. Biol Reprod, 29, 1069– 1076. Kennedy TG (1985) Evidence for the involvement of prostaglandins throughout the decidual cell reaction in the rat. Biol Reprod, 33, 140– 146.
555
Kennedy TG (1986) Intrauterine infusion of prostaglandins and decidualization in rats with uteri differentially sensitized for the decidual cell reaction. Biol Reprod, 34, 327–335. Kennedy TG and Doktorick PE (1988) Effects of analogues of prostaglandin E2 and prostaglandin F2a on the decidual cell reaction in the rat. Prostaglandins, 35, 207–220. Kennedy TG and Lukash LA (1982) Induction of decidualization in rats by the intrauterine infusion of prostaglandins. Biol Reprod, 27, 253– 260. Kennedy TG and Ross HE (1993) Effect of prostaglandin E2 on rate of decidualization in rats. Prostagland, 46, 243–250. Kennedy TG and Zamecnik J (1978) Concentration of 6-ketoprostaglandin F1a is markedly elevated at site of blastocyst implantation in rat. Prostaglandins, 16, 599–605. Kennedy TG, Martel D and Psychoyos A (1983a) Endometrial prostaglandin E2 binding: characterization in rats sensitized for the decidual cell reaction and changes during pseudopregnancy. Biol Reprod, 29, 556–564. Kennedy TG, Martel D and Psychoyos A (1983b) Endometrial prostaglandin E2 binding during the estrous cycle and its hormonal control in ovariectomized rats. Biol Reprod, 29, 565–571. Kogo H, Ogawa K and Takeo S (1991) The effects of ovarian steroids in controlling rat uterine prostaglandin F2a during the peri-implantation period. Prostagland Leukotrienes Essent Fatty Acids, 43, 29–34. Langenbach R, Morham SG, Tiano HF et al (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell, 83, 483–492. Lau IF, Saksena SK and Chang MC (1973) Pregnancy blockade by indomethacin, an inhibitor of prostaglandin synthesis: its reversal by prostaglandin and progesterone in mice. Prostaglandins, 4, 795–803. Leroy F, Vasande J, Shetgen G and Brasseur D (1974) Cyclic AMP and the triggering of the decidual reaction. J Reprod Fertil, 39, 207–211. Lewis GS and Waterman RA (1983) Effects of endometrium on metabolism of arachidonic acid by bovine blastocysts in vitro. Prostaglandins, 25, 881–889. Lewis GS and Waterman RA (1985) Metabolism of arachidonic acid in vitro by bovine conceptuses recovered during early pregnancy. Prostaglandins, 30, 263–284. Lewis GS, Thatcher WW and Bazer FW (1982) Metabolism of arachidonic acid in vitro by bovine blastocysts and endometrium. Biol Reprod, 27, 431–439. Lim H and Dey SK (1997) Prostaglandin E2 receptor subtype EP2 gene expression in the mouse uterus coincides with the differentiation of the luminal epithelium for implantation. Endocrinology, 138, 4599–4606. Lim H and Dey SK (2000) PPARd as a prostacyclin receptor for blastocyst implantation. Trends Endocrinol Metab, 11, 137–142. Lim H, Paria BC, Das SK et al (1997) Multiple female reproductive failures in cyclooxygenase 2-deficient mice. Cell, 91, 197–209. Lim H, Gupta RA, Ma W-G, Moller DE et al (1999) Cyclooxygenase-2derived prostacyclin mediates embryo implantation in the mouse via PPARd. Genes Dev, 13, 1561–1574. Mahla S and Prasad MRN (1970) Stimulation of RNA synthesis in the blastocyst and the uterus of the rat by adenosine 3’,5’-monophosphate (cyclic AMP). J Reprod Fertil, 23, 327–329. Malathy PV, Cheng HC and Dey SK (1986) Production of leukotrienes and prostaglandins in the rat uterus during the peri-implantation period. Prostaglandins, 32, 605–614. Marcus GJ (1981) Prostaglandin formation by the sheep embryo and endometrium as an indication of maternal recognition of pregnancy. Biol Reprod, 25, 56–64. Marshburn PB, Shabonowitz RB and Clark MR (1990) Immunohistochemical localization of prostaglandin H synthase in the embryo and uterus of the mouse from ovulation through implantation. Mol Reprod Dev, 25, 309–316. Matsumoto H, Ma W, Smalley W et al (2001) Diversification of cyclooxygenase-2-derived prostaglandins in ovulation and implantation. Biol Reprod, 64, 1557–1565. Mead RA, Bremner S and Murphy BD (1988) Changes in vascular permeability during the periimplantation period in the ferret (Mustela putorius). J Reprod Fertil, 82, 293–298. Miller MM and Morchoe CCC (1982a) Inhibition of artificially induced decidual cell reaction by indomethacin in the mature oophorectomized rat. Anat Rec, 204, 223–230.
556
REPRODUCTIVE SYSTEM
Miller MM and Morchoe CCC (1982b) Decidual cell reaction induced by prostaglandin F2a in the mature oophorectomized rat. Cell Tissue Res, 225, 189–199. Milligan SR and Lytton FDC (1983) Changes in prostaglandin levels in the sensitized and non-sensitized uterus of the mouse after the intrauterine instillation of oil or saline. J Reprod Fertil, 67, 373–377. Milligan SR, Cohen PE and Finn CA (1995) The minimum requirements for oestradiol to induce uterine sensitivity for implantation and decidualization in mice. Hum Reprod, 10, 1502–1506. Murata T, Ushikubi F, Matsuoka T et al (1997) Altered pain response in mice lacking prostacyclin receptor. Nature, 388, 678–681. Nakayama R, Yasuda K, Okumara T and Saito K (1991) Effect of 17bestradiol on PAF and prostaglandin levels in oophorectomized rat uterus. Biochim Biophys Acta, 1085, 235–240. Nayak NR, Ghosh D, Lasley BL and Sengupta J (1997) Anti-implantation activity of luteal phase mifepristone administration is not mimicked by prostaglandin synthesis inhibitor or prostaglandin analogue in the rhesus monkey. Contraception, 55, 103–114. Niimura S and Ishida K (1987) Immunohistochemical demonstration of prostaglandin E-2 in preimplantation mouse embryo. J Reprod, 80, 505–508. Neulen J, Williams RF, Schaefer W et al (1991) Arachidonic acid metabolism and prostaglandin production by primate preimplantation blastocysts. Prostaglandins, 42, 121–126. Novaro V, Rettori V, Gonzalez ET et al (1996) Interaction between uterine PGE and PGF2a production and the nitridergic system during embryonic implantation in the rat. Prostaglandins, 51, 363–374. O’Grady JP, Caldwell BV, Auletta FJ and Speroff L (1972) The effects of an inhibitor of prostaglandin synthesis (indomethacin) on ovulation, pregnancy and pseudopregnancy in the rabbit. Prostaglandins, 1, 97– 106. Ohta Y (1985) Histochemical localization of prostaglandin synthetase in the rat endometrium with reference to decidual cell reaction. Proc Japan Acad, 61(series B), 467–470. Pakrasi PL (1997) Evaluation of the role of leukotrienes on ovum implantation in mice. J Biosci, 22, 267–271. Pakrasi PL and Dey SK (1982) Blastocyst is the source of prostaglandins in the implantation site in the rabbit. Prostaglandins, 24, 73–77. Pakrasi PL, Cheng HC and Dey SK (1983) Prostaglandins in the uterus: modulation by steroid hormones. Prostaglandins, 26, 991–1010. Pakrasi PL, Dey SK and Johnson DC (1984) Studies on the temporal pattern of prostaglandin synthesis in the uterus of the delayed implanting rat with or without implantation inducing stimuli. Prostagland Leukotrienes Med, 14, 365–382. Pakrasi PL, Becka R and Dey SK (1985) Cyclooxygenase and lipoxygenase pathways in the preimplantation rabbit uterus and blastocyst. Prostaglandins, 29, 481–496. Papav KD and Kennedy TG (2000) Characterisation of temporal and cellspecific changes in transcripts for prostaglandin E2 receptors in pseudopregnant rat endometrium. Biol Reprod, 62, 1515–1525. Paria BC and Dey SK (1990) Preimplantation embryo development in vitro: cooperative interactions among embryos and role of growth factors. Proc Natl Acad Sci USA, 87, 4756–4760. Paria BC, Das SK, Gupta A and Dey SK (1991) Reversal of indomethacin-induced inhibition of implantation in the mouse by epidermal growth factor. Prostaglandins, 42, 191–199. Paria BC, Das SK, Andress GK and Dey SK (1992) Altered expression of the blastocyst epidermal growth factor receptor gene during delayed implantation in the mouse. Biol Reprod, 46(suppl 1), 70. Paria BC, Das N, Das SK et al (1998) Histidine decarboxylase gene in the mouse uterus is regulated by progesterone and correlates with uterine differentiation for blastocyst implantation. Endocrinology, 139, 3958– 3966. Parr MB, Parr EL, Munaretto K et al (1988) Immunohistochemical localization of prostaglandin synthase in the rat uterus and embryo during the peri-implantation period. Biol Reprod, 38, 333–344. Peleg S and Lindner HR (1982) The effect of prostaglandins on progestin receptor translocation and on decidual cell reaction in vivo and in vitro. Endocrinology, 110, 1647–1652. Phillips CA and Poyser NL (1981) Studies on the involvement of prostaglandins in implantation in the rat. J Reprod Fertil, 62, 73–81. Poyser NL (1985) Prostaglandin output from the early pregnant rat uterus superfused in vitro, and the effect of A23187 and trifluoroperazine. J Reprod Fertil, 74, 271–277.
Poyser NL (1999) A comparison of the effects of indomethacin and NS398 (a selective prostaglandin H synthase-2 inhibitor) on implantation in the rat. Prostagland Leukotrienes Essent Fatty Acids, 61, 297–301. Racowsky C and Biggers JD (1983) Are blastocyst prostaglandins produced endogenously? Biol Reprod, 29, 379–388. Rankin JC, Ledford BE and Baggett B (1977) Early involvement of cyclic nucleotides in the artificially stimulated decidual cell reaction in the mouse uterus. Biol Reprod, 17, 549–554. Rankin JC, Ledford BE, Jonsson HT and Baggett B (1979) Prostaglandin, indomethacin and decidual cell reaction in the mouse uterus. Biol Reprod, 20, 399–404. Reese J, Brown N, Paria BC et al (1999) COX-2 compensates in the uterus of COX-1 deficient mice during the preimplantation period. Endocrinology, 150, 23–31. Reese J, Zhao X, Ma W-G et al (2001) Comparative analysis of pharmacologic disruption of cyclooxygenase-1 and cyclooxygenase-2 function in female mice. Endocrinology, 142, 3198–3206. Robb L, Li R, Hartley L et al (1998) Infertility in female mice lacking the receptor for interleukin 11 is due to a defective uterine response to implantation. Nature Med, 4, 303–308. Saksena SK, Lau IF and Chang MC (1976) Relationship between oestrogen, prostaglandin F2a and histamine in delayed implantation in the mouse. Acta Endocrinol, 81, 801–807. Sananes N, Baulieu E-E and Le Goascogne C (1976) Prostaglandin(s) as inductive factor of decidualization in the rat uterus. Mol Cell Endocrinol, 6, 153–158. Sananes N, Baulieu E-E and Le Goascogne C (1981) A role for prostaglandins in decidualization of the rat uterus. J Endocrinol, 89, 25–33. Sanders RB, Bekairi AM, Asbulaban FS and Yochim JM (1986) Uterine adenylate cyclase in the rat: responses to a decidual-inducing stimulus. Biol Reprod, 35, 100–105. Satokata I, Benson G and Maas R (1995) Sexually dimorphic sterility phenotypes in Hoxa 10-deficient mice. Nature, 374, 460–463. Scherle PA, Ma W, Lim H et al (2000) Regulation of cyclooxygenase-2 induction in the mouse uterus during decidualization. An event of early pregnancy. J Biol Chem, 275, 37086–37092. Schmid PC, Paria BC, Krebsbach RJ et al (1997) Changes in anandamide levels in mouse uterus are associated with uterine receptivity for embryo implantation. Proc Natl Acad Sci USA, 94, 4188–4192. Segi E, Sugimoto Y, Yamasaki A et al (1998) Patent ductus arteriosus and neonatal death in prostaglandin EP4-deficient mice. Biochem Biophys Res Commun, 246, 7–12. Shaikh AA and Abraham GE (1969) Measurement of estrogen surge during pseudopregnancy in rats by radioimmunoassay. Biol Reprod, 1, 378–380. Sharma SC (1980) Temporal changes in PGE, PGFa oestradiol 17b and progesterone in uterine venous plasma and endometrium during early pregnancy. In Crastes de Paulet A, Thaler-Dao H and Dray F (eds), Prostaglandins and Reproductive Physiology. Paris: Inserm, 243–263. Shemesh M, Milaguir F, Ayalon N and Hansel W (1979) Steroidogenesis and prostaglandin synthesis by cultured bovine blastocysts. J Reprod Fertil, 56, 181–185. Simon C, Frances A, Piquette GN et al (1994) Embryonic implantation in mice is blocked by interleukin-1 receptor antagonist. Endocrinology, 134, 521–528. Sirois J and Richards JS (1992) Purification of a novel, distinct isoform of prostaglandin endoperoxide synthase induced by human chorionic gonadotropin in granulosa cells of rat pre-ovulatory follicles. J Biol Chem, 267, 6382–6388. Sirois J, Simmons DL and Richards JS (1992) Hormonal regulation of messenger ribonucleic acid encoding a novel isoform of prostaglandinH synthase in rat preovulatory follicles—induction in vivo and in vitro. J Biol Chem, 267, 11586–11592. Song JH, Sirois J, Houde A and Murphy BD (1998) Cloning, development and immunohistochemistry of cyclooxygenase 2 in the endometrium during embryo implantation and gestation in the mink (Mustela vison). Endocrinology, 139, 3629–3636. Song H, Lim H, Das SK et al (2000) Dysregulation of EGF family of growth factors and COX-2 in the uterus during the preattachment and attachment reactions of the blastocyst with the luminal epithelium correlates with implantation failure in LIF-deficient mice. Mol Endocrinol, 14, 1147–1161.
PROSTAGLANDINS IN IMPLANTATION Squires PM and Kennedy TG (1992) Evidence for a role of a uterine reninangiotensin system in decidualization in the rat. J Reprod Fertil, 95, 791–802. Stewart CL, Kaspor P, Brunet LJ et al (1992) Blastocyst implantation depends on maternal expression of leukemia inhibitory factor. Nature, 359, 76–79. Sugimoto Y, Yamasaki A, Segi E et al (1997) Failure of parturition in mice lacking the prostaglandin F receptor. Science, 277, 681–683. Takabatake K, Fujiwara H, Goto Y et al (1997a) Splenocytes in early pregnancy promote embryo implantation by regulating endometrial differentiation in mice. Hum Reprod, 12, 2102–2107. Takabatake K, Fujiwara H, Goto Y et al (1997b) Intravenous administration of splenocytes in early pregnancy changes the implantation window in mice. Hum Reprod, 12, 583–585. Tawfik OW and Dey SK (1988) Further evidence for the role of leukotrienes as mediators of decidualization in the rat. Prostaglandins, 35, 379–380. Tawfik OW, Sagrillo C, Johnson DC and Dey SK (1987a) Decidualization in the rat: Role of leukotrienes and prostaglandins. Prostagland Leukotrienes Med, 29, 221–228. Tawfik OW, Huet YM, Malathy PV et al (1987b) Release of prostaglandins and leukotrienes from the rat uterus is an early estrogenic response. Prostaglandins, 34, 805–816. Taylor HS, Arici A, Olive D and Igarashi P (1998) HOXA10 is expressed in response to sex steroids at the time of implantation in the human endometrium. J Clin Invest, 101, 1379–1384. Terranova PF and Dey SK (1982) Indomethacin delays zona-shedding and implantation in the ovariectomised progesterone-treated hamster. Prostaglandins, 24, 165–172. Tilley SC, Audaly LP, Hick SHE et al (1999) Reproductive failure and reduced blood pressure in mice lacking the EP2 prostaglandin E2 receptor. J Clin Invest, 103, 1539–1543. Tobert JA (1976) A study of the possible role of prostaglandins in decidualization using a non-surgical method for the instillation of fluids into the rat uterine lumen. J Reprod Fertil, 47, 391–397. Ushkubi F, Segi E, Sugimoto Y et al (1998) Impaired febrile response in mice lacking the prostaglandin E receptor subtype EP3. Nature, 395, 281–284. van der Weiden RMF, Verdjik RM, Poelmann RE et al (1993) The influence of indomethacin on the hatching of mouse embryos. Prostagland Leukotrienes Essent Fatty Acids, 49, 683–686. van der Weiden RMF, Wisse LJ, Helmerhorst FM et al (1996) Immunohistochemical and ultrastructural localization of prostaglandin
557
H synthase in the preimplantation mouse embryo. J Reprod Fertil, 107, 161–166. Vilar-Rajus C, Castro-Osuna G and Hicks JJ (1982) Cyclic AMP and cyclic GMP in the implantation site of the rat. Int J Fertil, 27, 56–59. Wang XQ, Zhu ZM, Fenderson BA et al (1998) Effects of monoclonal antibody directed to the Ley on implantation in the mouse. Mol Hum Reprod, 4, 295–300. Watson J, Anderson FB, Alam M et al (1975) Plasma hormones and pituitary luteinizing hormone in the rat during the early stages of pregnancy and after post-coital treatment with tamoxifen (ICI 46474). J Endocrinol, 65, 7–17. Webb FTG (1975) Implantation of ovariectomised mice treated with dibutyryl adenosine 3’,5’-monophosphate (dibutyryl cyclic AMP). J Reprod Fertil, 42, 511–517. Wellstead JR, Bruce NW and Rahima A (1989) Effects of indomethacin on spacing of conceptuses within the uterine horn and on fetal and placental growth in the rat. Anat Rec, 225, 101–105. Wiley LM, Wu J-X, Harari I and Adamson ED (1992) Epidermal growth factor receptor mRNA and protein increase after the four-cell preimplantation stage in murine development. Dev Biol, 149, 247–260. Wu JT and Lin G-M (1982) Effect of aromatase inhibitor on oestrogen production in rabbit blastocysts. J Reprod Fertil, 66, 655–662. Yang ZM, Das SK, Wang J et al (1997) Potential sites of prostaglandin action in the preimplantation mouse uterus: differential expression and regulation of prostaglandin receptor genes. Biol Reprod, 56, 368–379. Yee GM and Kennedy TG (1988) Stimulatory effects of prostaglandins upon endometrial alkaline phosphatase activity during the decidual cell reaction in the rat. Biol Reprod, 38, 1129–1136. Yee GM and Kennedy TG (1991) Role of cyclic adenosine 3’,5’monophosphate in mediating the effect of prostaglandin E2 on decidualization in vitro. Biol Reprod, 45,163–171. Yee GM and Kennedy TG (1993) Prostaglandin E2, cAMP and cAMPdependent protein kinase isozymes during decidualization of rat endometrial cells in vitro. Prostaglandins, 46, 117–138. Zhu ZM, Kojima N, Straud MR et al (1995) Monoclonal antibody directed to Ley oligosaccharide inhibits implantation in the mouse. Biol Reprod, 52, 903–912. Zhu L-J, Baychi MK and Baychi JC (1998a) Attenuation of calcitonin gene expression in pregnant rat uterus leads to a block in embryonic implantation. Endocrinology, 139, 330–339. Zhu L-J, Cullinan-Bove C, Polihronis M et al (1998b) Calcitonin is a progesterone-regulated marker that forecasts the receptive state of endometrium during implantation. Endocrinology, 139, 3923–3934.
52 Parturition and the Clinical Interruption of Pregnancy S. Cowan and A. A. Calder University of Edinburgh, Edinburgh, UK
The physiological process of parturition is extremely complex and, although the majority of cases proceed without complication, we do not yet fully understand the underlying biological mechanisms and so our ability to intervene effectively where necessary is limited. The process of labour has been well characterized in terms of regular uterine contractions resulting in effacement and dilatation of the cervix, allowing passage of the mature foetus to the outside world. Prior to this the myometrium gradually becomes more responsive to stimuli over months, the cervix ripens over several weeks and eventually effectors initiate the process of labour. It is this synchronous preparatory process and its regulators that must be unravelled. Prostaglandins certainly play a major role at the cellular level, but what are the mechanisms by which they are activated and which messengers and triggers are responsible for initiation of the cascade of biological changes which culminate in established labour (Thiery 1979; Wiqvist 1979)? Prostaglandins have been used for decades in labour induction and termination of pregnancy but are not entirely without risk. Further therapeutic progress can only come from a clearer understanding of the mechanisms involved. PROSTAGLANDINS AND PARTURITION Prostaglandins in Labour Prostaglandins are eicosanoids, of which those believed to be most important in human labour are of the 2 series, PGE2 and PGF2a. They are derived from the cyclooxygenase (COX) metabolism pathway of arachidonic acid, obtained from deesterified membrane phospholipids. They are locally acting paracrine regulators, rapidly catabolized from the circulation in the lung, and also locally metabolized by degradative enzymes, so requiring prompt tissue response for any effect (Kelly et al 1992; Moore 1985). Prostaglandins are produced by many tissues, but notably for this chapter by amnion, chorion, decidua, placenta, cervix, leukocytes and platelets. There is no storage mechanism, so they are produced and released as required, and act through cell-surface binding (Geirsson and Greer 1990). Prostaglandins induce many effects throughout all systems, including pro/anticoagulant properties, immunomodulation, vasoconstriction/dilatation and bronchoconstriction/dilatation, hence their potency and potential for unwanted side-effects as well as desired attributes. There are five groups of prostaglandins. Those involved in pregnancy are PGE2 and PGF2a, prostacyclin (PGI2) and thromboxane, but in labour only the first two have identified roles. PGE2 levels are known to increase in late pregnancy and The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
early labour, whereas PGF2a rises in established labour (Moore 1985). Many sources of prostaglandins have been identified in the cervix (Ellwood et al 1980), the uterus, placenta and foetal membranes. The amnion is particularly capable of PGE2 synthesis (Keirse 1979; Mitchell 1987a, 1987b) and it has been shown that amniotic fluid prostaglandin levels gradually increase towards term, with a sharper rise in labour and as labour progresses (Reece et al 1996; Steinborn et al 1995; Gibb 1998). This is now known to be as a result of increased synthesis, rather than reduced catabolism. As prostaglandins and their controlling enzymes (phospholipases, cyclooxygenases and prostaglandin dehydrogenase) are fairly ubiquitous, it has proved difficult to localize production sites, although high prostaglandin levels have been reported in the amnion, decidua (PGF2a) and myometrium. There are two cyclooxygenase enzymes, COX-1, a constitutive enzyme, and COX-2, an inducible enzyme that increases in labour (Whittle et al 2001; Bennett et al 1992). Particularly high levels of the prostaglandin-inactivating prostaglandin dehydrogenase (PGDH) have been found in the chorion and, although placental levels do not fall in labour, chorionic PGDH expression was seen to be lower in both term and preterm labour compared to nonlabour (Whittle et al 2001). Another theory is that with varying membranous PGDH distribution there may be regional variation, most pertinently covering the internal cervical os, where the prostaglandin can escape catabolism and traverse the membranes in order to effect cervical ripening and myometrial activity (Challis et al 1999; Van Meir et al 1997). Glucocorticoids increase amniotic COX-2 expression and decrease PGDH in the chorion, thereby promoting prostaglandin activity (Gibb 1998). It is possible that PGDH expression may be varied in some areas by inflammatory cytokines as a result of infection, and in others by progesterone and cortisol competitively binding to their steroid receptors. There are several regulators of prostaglandin synthesis: steroid hormones, inflammatory mediators, growth factors and cytokines, and they seem to act in a multifactorial manner. Prostaglandin receptors are seven-transmembrane G-coupled receptors divided into five groups, specific to their prostaglandin. The EP group are specific to PGE and are further subdivided into four isoforms; EP1 and EP3 stimulate the adenyl cyclase system causing muscle relaxation, whereas EP2 and EP4 are inhibitory and cause contraction (Arias 2000). Both PGE and PGF2a cause myometrial contractility, whilst PGE, in addition, has specific effects on cervical ripening. Cervical PGE production increases at term and in labour and administration of PGE results in cervical changes akin to those seen
560
REPRODUCTIVE SYSTEM
Figure 52.1 Initiation of labour
physiologically. PGs not only increase the total number of myometrial gap junctions, but also release stored intracellular calcium and facilitate its transfer across these gap junctions, a process necessary for synchronized myometrial activity. Initiation of Labour The trigger for initiation of labour has been well documented for many mammals but the same hypothesis cannot be advanced for all species. In sheep, labour is driven by foetal hypothalamo– pituitary axis (HPA) maturation, corticotrophin-releasing hormone (CRH) and ACTH release, causing raised foetal adrenal cortisol levels, which in turn cause an increased oestrogen production by the placenta, via increased 17a-hydroxylase activity, with a concurrent reduction in progesterone levels. This results in prostaglandin synthesis and so labour ensues (LundinSchiller and Mitchell 1990). Yet the same has not until recently seemed plausible for humans, as anencephalic foetuses can be spontaneously delivered around term and the same alteration of placental enzymes is not possible. Nor in humans is there the prelabour circulatory progesterone withdrawal, although there may be subtle local intrauterine changes. In the human foetus the effect of the cortisol surge may have a different emphasis, to promote key enzymes responsible for foetal lung maturation and preparing other organs to meet the demands of extrauterine function (see Figure 52.1). Recent publications would support the postulation of foetal HPA maturation being the primary trigger. Whittle et al (2001) in their
review of the glucocorticoid regulation of parturition discuss the central role of cortisol, both its direct prostaglandin effects and its indirect intrauterine CRH-mediated effects, stimulating COX-2 expression and initiating human labour. The human mechanism is different to that of sheep in that it is adrenal androgen production that drives the HPA axis as opposed to a placental source. Dehydroepiandrosterone (DHEAS) is then converted by placental sulphatase and aromatase to oestradiol. So, a similar mechanism may operate in sheep and humans after all, with the exception of the androgen precursor source, where humans rely on foetal and maternal adrenal supplies to provide the term rise in intrauterine oestrogen, without a drop in progesterone.
Cervical Ripening Before labour begins satisfactory preparation of the cervix is required. Both physiologically and therapeutically, this is now referred to as ‘‘cervical ripening’’ and applies to the process whereby the cervical structure undergoes significant connective tissue remodelling, resulting in greater compliance than is found in the firm fibrous non-pregnant cervix. Cervical ripening takes place over a number of weeks in the latter stages of pregnancy, before the dramatic cervical changes seen during labour. Ledger et al (1985) confirmed that this cervical change was an active process when it was seen in cervices surgically separated from the uterus, where the effects could not be as a passive result of uterine contractions.
PARTURITION AND THE CLINICAL INTERRUPTION OF PREGNANCY Structure The nature of cervical tissue was well described by Danforth (1947). Distinct from the muscular uterine corpus, the cervix is a fibrous organ with less than 15% of its content contributed by smooth muscle, mostly found peripherally and increasingly abundant nearer to the corpus. The remaining extracellular matrix consists of dense collagen (66% type I, 33% type III) fibril bundles and elastic fibres with intervening ground substance. Ground substance is composed of proteoglycan complexes, consisting of various glycosaminoglycan (GAGs) side chains on core proteins linked to a hyaluronic acid chain. These complexes bind tightly, investing the collagen fibrils and providing rigidity. In general, the longer the GAG chain the greater the binding affinity to collagen. The dominant GAGs in the cervix are chondroitin sulphate and dermatan sulphate. They contain iduronic acid, which binds very strongly, helping to confer strength. Certain other glucuronic acid-containing GAGs, such as heparan sulphate, bind less strongly, leading to destabilization of the organized fibril bundles, loss of mechanical strength and, thus, enhanced compliance (Uldbjerg et al 1983a). This is represented by smaller widely scattered dissociated fibrils seen at term, compared to the non-pregnant state. There are many biochemical changes in the cervix pre-labour, the end result being a reduction in overall collagen content (50– 70%), an increase in water content (80–86%), and a change in the proportions of different GAGs with a relative increase in heparan sulphate (Uldbjerg et al 1983b; Ekman et al 1986). Recently it has been suggested that a large keratan sulphate proteoglycan may be involved in the matrix reorganization at term, but this has not yet been fully characterized (Fischer et al 2001). Hyaluronic acid content also increases in labour and, as a hydrophilic compound, may be partially responsible for the higher water content, but there are other contributory factors such as increased vascular permeability (Cabrol et al 1990). The fibroblast is undoubtedly central to this process as a source of collagen, GAGs and lytic enzymes, but there is more to consider.
561
The antiprogestin mifepristone (Radestad et al 1990), oestrogen, cytokines and other inflammatory mediators, such as nitric oxide donors (Thomson et al 1998), are all implicated in the cervical ripening process but the exact mechanisms of interaction remain unclear. Cytokines Cytokines offer many avenues of investigation, some of which provide clear evidence of their role in parturition. Cervical and uterine isthmic biopsies were found to produce far greater quantities of IL-1, TNFa, IL-6 and IL-8 (Winkler et al 1998; Sennstrom et al 2000) in labour and postpartum. Amniotic fluid interleukin-8 (IL-8) levels increase in labour, but a much greater increase is seen in preterm labour with positive amniotic fluid bacterial cultures (Romero et al 1991). IL-8 is produced by cervical fibroblasts (Barclay et al 1993) and has a dual function, to induce neutrophil activation and migration and cause degranulation to release collagenase. IL-8 is also produced by the myometrium, decidua and membranes in rising quantities throughout the stages of labour (Osmers et al 1995), whereas progesterone suppresses its synthesis. Other proinflammatory cytokines (TNFa and IL-1) stimulate IL-8 and also have a similar effect on PGE2 (Sato et al 2001; Barclay et al 1993). A synergistic action between IL-8 and PGE has been proposed, where PGE provides the vasodilation and vascular permeability for IL-8 neutrophil recruitment. IL-8 is a potent neutrophil chemoattractant, while IL-6 activates neutrophils. MMP release (Ledingham et al 1999) from the cervix is known to be associated with IL-8, IL-1 and PGE (El Maradny et al 1996). Cytokines and proteoglycans may interact to promote the remodelling process (Belayet et al 1999; Sennstrom et al 2000). Platelet activating factor (PAF) (Sugano et al 2001) and monocyte chemotactic protein-1 (MCP-1) (Yamamoto et al 2000; Denison et al 2000) further increase the number of potential candidates for cervical ripening agents and it seems probable that many more will follow.
Inflammatory Response
Nitric Oxide
Leukocytes also appear to have a major role. Cervical ripening has been compared to an inflammatory response (Liggins 1981), perhaps orchestrated by the fibroblast with increased local prostaglandins, a marked inflammatory neutrophil infiltration (Junqueira et al 1980), increased vascularity and stromal oedema. The reduced collagen content seen in late pregnancy is a result of both proteoglycan constituent variation and collagen breakdown from enzymatic degradation. Extracellular matrix degradation is achieved via collagenases/ matrix metalloproteinases. They are divided into interstitial collagenases (MMP-1, -8, -13), gelatinases (MMP-2,-9), stromelysins (MMP-3,-7,-10,-11) and membrane-type MMPs (-14,-15,16,-17). Leukocyte collagenase (MMP-8), fibroblast collagenase (MMP-1) and leukocyte elastase are shown to increase throughout gestation and even more so at the time of parturition, as the cervical collagen content decreases (Uldbjerg et al 1983a; Rajabi et al 1988). Tissue inhibitors of metalloproteinases (TIMP-1, -2 and -4) modulate MMP activity and are also present in the cervix (Ledingham et al 1999; Denison et al 2000). Immediately prelabour, the ripening process seems to accelerate, with even more proteolytic activity associated with this inflammatory cell infiltrate (Knudsen et al 1997). Prostaglandins, which are known to increase physiologically at this stage and are also effective therapeutically, are only one of several factors involved in initiation of this process.
Synthesized by macrophages, myometrium, cervix, foetal membranes and other cell types, such as cardiac muscle, nitric oxide (NO) release is stimulated by IL-1a, TNFa and oestrogen and inhibited by dexamethasone. However, PGE2 can have either influence, as can progesterone and several cytokines. In rats, uterine and placental NO production is seen to increase in pregnancy and then downregulate before term, whereas cervical NO production is the opposite (Chwalisz and Garfield 1998). The same has not yet been demonstrated in humans. Of course, the cervical source of NO is likely to be the leukocytes, differing from the myometrial source. The mechanism of action involves cGMP stimulation and effects include smooth muscle relaxation, felt to be important in the maintenance of uterine quiescence, and increased vascular permeability and immunomodulation via IL-8 and MMPs. NO also induces COX-2 and PGE. There is conflicting evidence about the role of NO in labour, but several studies have shown some cervical ripening effect (Thomson et al 1998; Chwalisz et al 1997). Uterine Contractility Actin and myosin filaments within smooth muscle fibres require phosphorylation by the enzyme myosin light chain kinase (MLCK) to allow contraction of the myosin light chains.
562
REPRODUCTIVE SYSTEM
MLCK activity is inhibited by cAMP and cGMP. MLCK is also calmodulin/calcium-dependent, which in turn is dependent on inositol triphosphate opening the sarcoplasmic calcium channels, increasing intracellular calcium levels (Husslein and Egarter 1992; Arias 2000). Calcium can also enter cells from the extracellular compartment via voltage-gated channels. Gap junctions facilitate the depolarization required to open these. Intermediate filaments provide a cross-link between muscle fibres to coordinate uterine contractility throughout the myometrium. Gap junctions increase in number throughout gestation and particularly in labour, corresponding to uterine activity and cervical dilatation (Garfield et al 1998). High intracellular calcium levels and the presence of gap junctions are the two essential components for the stimulation of myometrial activity, and PGE2, PGF2a and oxytocin have been shown to promote both of these (Garfield et al 1998). It is thought that amniotic PGE2 stimulates decidual PGF2a production, which is responsible for myometrial contractility. Expectedly, progesterone inhibits gap junction formation and antiprogestogens induce the same. INDUCTION OF LABOUR Background Induction of labour is an intervention appropriate when it is no longer considered desirable for the pregnancy to continue and vaginal delivery is the preferred option. As an option it should not be taken lightly, as initial intervention encourages and often necessitates later intervention, so clinicians must assess the balance of risk in every case. This may be for foetal or maternal reasons, where the risks of the foetus remaining in utero outweigh the benefits of continued gestation. Induction is commonly performed nowadays, in excess of 20% of pregnancies in some regions. The single most common indication is prolonged pregnancy, where it has been shown from perinatal mortality statistics that the risk of unexplained intrauterine death increases significantly beyond 41 weeks (Yudkin et al 1987; Hilder et al 1998; Cotzias et al 1999). There has been hesitation to introduce routine elective induction prior to 42 weeks, due to the reported increased rate of caesarean delivery without perinatal outcome improvement. This has now been refuted, as evidence grows that elective induction after 41weeks does not increase caesarean deliveries (Rand et al 2000; RCOG Clinical Effectiveness Support Group 2001). Other reasons to intervene may be pregnancyinduced hypertension, intrauterine growth restriction, obstetric cholestasis and several other situations where foetal or maternal well-being is threatened. In 2001 the Royal College of Obstetricians and Gynaecologists prepared guidelines detailing recommendations for induction of labour from indications and service provision to drug regimes. It is a detailed account of many meta-analyses of available evidence in induction issues and merits attention from all clinicians involved in this field (RCOG Clinical Effectiveness Support Group 2001). Historically, many techniques of labour induction have been passed over for new improved methods. For many years, herbal to mechanical methods have in some way stimulated endogenous prostaglandins, before any awareness of the existence of such compounds now known to have such a major role in labour. Amniotomy has served women well for centuries, but requires a ripe cervix. Oxytocin was the main therapeutic option from the early twentieth century until the 1960s, when prostaglandins were developed. It was relatively recently in the history of induction that the importance of cervical ripening began to be appreciated; it is currently well established that unripe cervices lead to prolonged labour, with associated complications of maternal pyrexia, sepsis and postpartum haemorrhage, failed induction, an
increased risk of caesarean delivery and foetal effects, such as poor Apgar scores, birth asphyxia and neonatal sepsis (Calder 1979; Edwards and Richards 2000). Prostaglandins have provided us with a new tool with which to minimize these sequelae by increasing the efficacy of the induction process. Cervical Assessment Many authors have commented on the importance of cervical assessment before induction of labour and several methods, from examination to biochemical markers or ultrasound assessment, have been explored. The most cost-effective and widely used system is that of the Bishop score (Edwards and Richards 2000), or a modified version, where five cervical parameters are scored and the total reflects ripeness. A low score is unfavourable, but equips the clinician with pertinent information allowing decisions regarding suitability and method of induction. Prostaglandins and Labour Induction Systemic prostaglandins were quickly superseded by local applications in order to avoid the side-effects experienced. After Foley catheter placement in the extraamniotic space, prostaglandin administration soon became much more convenient as tablets and gels were developed, and later a hydrogel polymer (Arias 2000; Calder and Elder 1993; Keirse 1992). Dinoprostone (PGE2) is currently the agent of choice, either as a vaginal/intracervical gel or as a slow-release polymer insert, but a requirement for special storage and the relatively high cost of these preparations can be prohibitive for some countries. PGE2 effects cervical ripening as well as relaxing the cervical smooth muscle and contracting the myometrium. Misoprostol, a PGE1 analogue, is an inexpensive alternative and simplifies storage requirements, but there are licensing and dosage issues, despite many studies supporting its efficacy and safety. Prostaglandin Formulations The currently favoured formulations of dinoprostone are intravaginal gels (also intracervical gel in the USA), and hydrogel polymer inserts. These are found to be of similar efficacy but intracervical gels are possibly more difficult to administer and vaginal gels perform slightly better. The controlled-release polymer insert is equally effective but more expensive. Vaginal tablets are also used and are cheaper but require increased use of oxytocin (RCOG Clinical Effectiveness Support Group 2001; Witter 2000). Certainly, PGE is considered safe for induction without affecting the caesarean section rate. A recent trial compares vaginal gel to an insert, which is left in situ for 12 h and so may reduce the number of examinations required. With regard to patient acceptability, a slight preference for the latter was found when scoring for satisfaction with labour, but not for satisfaction with the induction process itself (Tomlinson et al 2001). Misoprostol, a synthetic PGE1 analogue developed for the prevention of NSAID-induced peptic ulcers, was always known to have abortifacient properties, and as such was contraindicated for use in pregnancy. After development for use in termination, the first attempts to evaluate misoprostol as a labour induction agent were in 1992. Misoprostol performed well against placebo and oxytocin alone, by increasing the 24 h delivery rate to 79% (Sanchez-Ramos and Kaunitz 2000; Wing 1999) A comparison with dinoprostone confirmed the efficacy whilst raising a few questions about safety. With a 50 mg vaginal dose every 3 h, a high
PARTURITION AND THE CLINICAL INTERRUPTION OF PREGNANCY
563
Table 52.1 Prostaglandin comparisons—induction of labour Study Chyu 1997 n=73 Kolderup et al 1999 n=159 Sanchez-Ramos et al 1998 n=223 Kwon 2001 n=160 Rowlands 2001 n=126 How 2001 n=330 Rozenberg 2001 n=370
Dinoprostone
Misoprostol
Ic gel Insert Ic gel 0.5 mg/6 h
– –
Insert 10 mg – Vaginal gel – Vaginal gel
50 mg/4 h v 50 mg/3 h v 50 mg/6 h v 50 mg/6 h o 100 mg/6 h v 25 mg/4 h o+v 25 mg/4 h v 25 mg/4 h o 50 mg/6 h
Mean time to delivery (h) 26.4 20.6 28.9 19.8 17.4 11.6 19.3 27.3 26.3 15.4 14.2 13.5 23.9 18.2 14
CS rate (%)
23 17 25.3 20.6 30 17 32 16.2 17.9
TS (%)
7 21 7.3 9 1.6 9.6 34 32 10 3.2 6.5
HS (%)
0 0 0 16.1 22 25 4 2.7 1.1
Ic, intracervical; Insert, intravaginal polymer insert; CS, caesarean section; TS, tachysystole; HS, hyperstimulation; v, vaginal administration; o, oral administration.
rate of meconium (30%) and tachysystole (37%) occurred, despite no other adverse neonatal outcomes (Wing 1999). With half the dose, 66% of women still delivered within 24 h with a 50% reduction in tachysystole. Misoprostol has been repeatedly shown to effectively induce labour, resulting in improved cervical score, shortened delivery interval and reduced requirement for other oxytocics (Danielian et al 1999; Hofmeyr et al 1999; SanchezRamos and Kaunitz 2000; Wing 1999), yet the ideal regime has not been established due to concerns about its uterotonic effect and association with hyperstimulation and higher rates of meconium-stained liquor (Goldberg et al 2001). Vaginally administered misoprostol has different pharmacokinetics to oral, with a slower and lower peak plasma concentration, which may be sustained for up to 4 h (Sanchez-Ramos and Kaunitz 2000; Hofmeyr et al 2001). This results in a greater exposure overall, while avoiding the surge seen with oral administration. The result is greater efficacy when considering delivery outcomes against time (Kwon et al 2001; How et al 2001) but concern remains with foetal safety. Currently lower dose regimes are being assessed. Misoprostol has repeatedly been shown to be as efficacious, if not more, as dinoprostone but sufficient safety data have yet to be established, particularly in those women with a uterine scar (Rozenberg et al 2001; Rowlands et al 2001; Kolderup et al 1999). Clearly, until the safety issues are clarified this therapy will not be universally welcomed. Mifepristone 200 mg, given 24, 48 or 72 h before induction of labour, results in a slightly shorter delivery interval and a reduced requirement for either prostaglandins or oxytocin, but not the more dramatic effect we see in early pregnancy (Elliott et al 1998; Wing et al 2000).
Uterine Rupture Uterine rupture is a rare but catastrophic complication of labour following a previous caesarean delivery, and can result in foetal death with serious maternal morbidity. Induction of labour is always performed with caution in these cases, but restricted prostaglandin use has been successful. The incidence of uterine rupture is 0.2–1.5%, depending on the definition used (SanchezRamos et al 2000). There is an associated uterine dehiscence, where scar separation is noted incidentally at caesarean section or post-delivery where there have been no clinical signs or complications. Prostaglandin E2 use has not been shown to significantly increase the risk of rupture.
Uterine disruption has been reported in 5.6% women with previous caesarean deliveries treated with misoprostol, compared to only 0.2% in those not exposed (Plaut et al 1999). However, uterine rupture has also been reported in parous women without prior uterine surgery (Bennett 1997). Currently there is not enough data to support safety of misoprostol use in those women with a scarred uterus until enough properly conducted controlled studies have been completed. TERMINATION OF PREGNANCY Background Many prostaglandins and prostaglandin analogues have been employed for termination of pregnancy for over 30 years. Prior to this, clinical procedure relied upon surgical aspiration, hysterotomy or amnioinfusion of various hypertonic solutions (e.g. glucose or urea). Around 192 000 terminations of pregnancy procedures are performed in the UK each year at present (RCOG Guideline Development Group 2001). There are two main methods of termination of pregnancy, surgical and medical. Surgical, largely vacuum aspiration, and medical induced abortion techniques have both benefited from the use of prostaglandins. Indeed, prostaglandin use has dramatically changed medical methods to the extent that where it was once a rare alternative to vacuum aspiration it is now commonplace and in many cases preferred. Prostaglandins and Termination of Pregnancy Prostaglandins have been particularly valuable in pregnancy termination, as it is known that myometrial stimulation occurs in all trimesters with exogenous prostaglandin administration (Mitchell 1987b). It has also been shown that prostaglandins are synthesized in the cervix itself and prostaglandin receptor presence in the cervix has been demonstrated (Uldbjerg et al 1987). Therefore, termination of pregnancy either medically or surgically in the first trimester, and medically at any gestation, can be effected by prostaglandin use. Administration Initially prostaglandins were administered by intravenous injection but, with a particularly short half-life of 1 min or less, it was
564
REPRODUCTIVE SYSTEM
soon discovered that duration of action could be significantly extended to 13–20 h by extra/intraamniotic administration (Bygdeman 1993). In comparing PGE2 and PGF2a, it was found that PGE2 had a 5–10-fold increase in potency, a much greater effect on cervical ripening and a better side-effect profile for intra/ extraamniotic administration; this then became the treatment of choice (Calder 1987). Analogues were then found to possess an even greater potency and longer half-life, making treatments more effective, whilst intraamniotic/intramuscular administration dispensed with intravenous infusions. Prostaglandin analogues have been developed that are not substrates for the initial stage of degradation by 15-dehydrogenase. This increases potency over their parent compounds, with a prolonged duration of action and greater specificity to uterine smooth muscle (Bygdeman, 1992). Gemeprost (16,16-dimethyltrans-D2-PGE1 methyl ester), carboprost [15(S)-15-methyl PGF2a] and sulprostone (16-phenoxy-17,18,19,20-tetranor PGE2 methylsulphonamide) were those in use initially, administered vaginally, intramuscularly/intraamniotically and intramuscularly, respectively, with success rates of 94–100% (Norman 1992). Later pharmacological developments transformed the method of administration from invasive procedures to local vaginal application, with much greater patient acceptability, reduced incidence of side-effects and reduced requirement for skilled medical staff. Vaginal prostaglandins were first used in 1972 for second trimester termination of pregnancy after Sandberg had shown absorption into the systemic circulation in 1968 (Karim et al 1979). The 1980s saw virtually universal use of vaginal prostaglandin analogues, mostly gemeprost, for both medical abortion and cervical ripening pre-surgical vacuum aspiration. Extraamniotic prostaglandin analogue infusion was still popular for second-trimester-induced abortion at this stage (Keirse 1992; Bygdeman 1992; Calder and Greer, 1992). More recently, the discovery of two drugs, mifepristone and misoprostol, has further improved the efficacy of medically induced abortion. Antiprogestins Mifepristone is an antiprogestin that competes with progesterone at the receptor level, thus producing a reduced progesterone effect and sensitizing the uterus to prostaglandins (Kelly and Bukman 1990). Csapo’s theory, that progesterone withdrawal results in uterine activity, is borne out here, despite many studies being unable to show a physiological reduction in progesterone in normal term pregnancy. Not only is uterine activity stimulated after approximately 36 h, but increased myometrial sensitivity to prostaglandins also results. The exact mechanism of action is not known but increased leukocyte and monocyte numbers, with elastase and MMP stimulation, is seen (Denison et al 2000). Mifepristone alone is not a particularly effective abortifacient (success rates of 60–70%) (Cameron et al 1986; Bygdeman, 1993) but, in combination with prostaglandin analogues, results in successful termination of pregnancy in 92–99% of cases (Peyron 1993; McKinley et al 1993; Spitz et al 1998; Schaff et al 1999; Norman et al 1991; Schaff et al 2000; Ashok et al 1998b; Baird et al 1995; Silvestre et al 1990). Mifepristone is also helpful in reducing side-effects, as the prostaglandin dose required can be significantly reduced. Misoprostol Misoprostol has been exploited for the precise purposes of therapeutic termination of pregnancy with great success (Norman et al 1991, Peyron 1993; Jain and Mishell 1994; el-Refaey and
Templeton 1995). Although there has been recent controversy surrounding its use in pregnancy in terms of licensing (Hale and Zinberg 2001; Friedman 2001; Gebhardt 2001; Mackenzie 2001; Wagner 2001), misoprostol has been extensively evaluated (Goldberg et al 2001) and introduced into routine clinical practice for termination of pregnancy. First-trimester Medical Termination of Pregnancy Early use of PGE2 and PGF2a intravenously/intraamniotically was superseded by the development of prostaglandin analogues, with a reduction in both side-effects and necessity for invasive procedures/administration. The success rate for complete termination of pregnancy at this time was 90–100%, but a rate of 50% of gastrointestinal upset and 30% of significant uterine pain revealed room for improvement. Many different regimes of prostaglandins have been evaluated with regard to success rates, side-effects and complications, but here we will concentrate on current regimes using a combination of mifepristone and either vaginal gemeprost or oral/vaginal misoprostol, as these have been shown to be the most effective and best tolerated to date. Readers are referred to summaries of prior regimes (Norman 1992; Keirse 1992; Bygdeman 1992, 1993; Calder and Greer 1992; Calder 1987). Much of the following information relates to pregnancies of less than 9 weeks’ gestation, as failure rates beyond this time were higher and associated with excess bleeding. As a result, surgical termination has been recommended by many for gestations between 9 and 12 weeks, (RCOG Guideline Development Group 2001; Calder 1987) and up to 15 weeks, although it is standard practice to offer medical management beyond 12 weeks. Mifepristone alone is relatively free of side-effects, but did not approach the efficacy of prostaglandins (Cameron et al 1986). A combined approach allowed a reduction in prostaglandin dosage and attendant side-effects, with equivalent efficacy of 94–100% (Rodger et al 1989; Silvestre et al 1990; Cameron et al 1986; UK Multicentre Trial 1990). Most regimes used a similar pattern, in which oral mifepristone (600 mg) is followed 48 h later by the prostaglandin analogue. Mifepristone maintains efficacy at a reduced dose of 200 mg in combination with either gemeprost or misoprostol (Norman et al 1991; Baird et al 1995; McKinley et al 1993; Ashok et al 1998b; Schaff et al 1999, 2000; WHO Task Force on Post-ovulatory Methods of Fertility Regulation 2000). Misoprostol has been explored as a prostaglandin alternative since 1991 and has been shown to be extremely effective (Peyron 1993; McKinley et al 1993; Spitz et al 1998), with the advantage of oral administration. Paradoxically many investigators are now using misoprostol vaginally as efficacy improves in the 7–9 week group (Ashok et al 1998b; Schaff et al 1999, 2000). Complete abortion rates of 97.5% have been achieved in women up to 63 days pregnant (Ashok et al 1998b). Vaginal misoprostol alone has been recently evaluated with an 85% success rate and low patient acceptability (Ngai et al 2000). The timing of misoprostol dosage was looked at by Schaff et al (2000), who found that efficacy was not compromised if 800 mg of self-administered vaginal misoprostol was given either 1, 2 or 3 days after 200 mg mifepristone (complete abortion rates 98%, 98% and 96%, respectively). This allows more flexibility in planning termination of pregnancy services. UK studies have not yet evaluated self-administration but it seems to be an acceptable method for the majority of women (94%), with 85% considering the side-effects to be acceptable (Schaff et al 1999). Pregnancy termination at 9–13 weeks has traditionally been by surgical methods, due to increased failure rates and excessive blood loss with medical methods at these gestations. However,
PARTURITION AND THE CLINICAL INTERRUPTION OF PREGNANCY
565
Table 52.2 First-trimester pregnancy termination studies Study
Gestation (days)
Mifepristone dose (mg)
Oral misoprostol dose
Success* rate (%)
556
– 200 600 600 200 200 600 200
200–600 mg 200–1000 mg 400 mg 400+200 mg 0.5 mg gemeprost, v 600 mg 600 mg 600 mg
600
400 mg
200
800 mg, v
5 86 96.9 98.7 96.7 94.6 93.6 93.6 92 83 77 97.5
556
200
800 mg self-admin, v
97
563
–
800 mg day 1, 3, 5, v
85
556
200
549
200 600
800 g day 1, v day 2 day 3 400 mg 400 mg
Norman et al 1991 n=33 Peyron 1993 n=895 Baird et al 1995 n=800 McKinley et al 1993 n=220 Spitz et al 1998 n=2121 Ashok et al 1998b n=2000 Schaff et al 1999 n=933 Ngai et al 2000 n=80 Schaff et al 2000 n=2255 WHO Task force 2000 n=1589
550 563 563 549 50–56 57–63 563
Note
549 days 97.5% 450 days 89%
Procedure acceptable 94%
98 98 96 89.3 88.1
*Complete abortion; v, vaginal administration.
Table 52.3 Second-trimester drug regimes—vaginal administration unless otherwise stated Abortion 524 h (%)
Study
Prostaglandin
Jain and Mishell 1994 n=55 el-Refaey and Templeton 1995* n=70 Wong 1998 n=140 Jain 1999 n=84 Webster 1996 n=70 Nuutila 1997 n=81
PGE2 20 mg/3 h Misoprostol 200 mg/12 h vMiso 600 mg, 400 mg/3 h vMiso 600 mg, 400 mg oral/3 h Gemeprost 1 mg/3 h Misoprostol 400 mg/3 h Misoprostol 200 mg/6 h Misoprostol 200 mg/12 h Misoprostol 800 mg vag, then 400 mg oral/3 h Misoprostol 100 mg/6 h Misoprostol 200 mg/12 h Gemeprost 1 mg/3 h
81 89 58.6 80 80.6 86.5
Mean interval (h)
Success rate (%)
12 10.6 6.0 6.7 19.5 14.1 13.8 14 6.9
100 81 97 97 91 87.2 89.2 94–97
23.1 27.8 14.5
74 92 89
*With mifepristone pretreatment.
Ashok et al (1998a) achieved success rates of 95% at this gestation, using a combination of mifepristone and vaginal misoprostol, with 93% of women treated as day cases. With further evaluation this regime may extend the use of a medical alternative throughout the first trimester.
Second-trimester Medical Termination of Pregnancy Extraamniotic/intraamniotic and intramuscular prostaglandin administration were the mainstay of treatment options from the early 1970s and have been previously summarized (Bygdeman 1993; MacKenzie 1992). Ten years on, vaginal preparations were introduced. Gemeprost and later misoprostol became well established for this use and their evaluation has followed that for the first trimester, with improved efficacy, reduced side-effects and shortened treatment times when used in combination with mifepristone. Common regimes are mifepristone followed by repeated gemeprost pessaries or vaginal/oral misoprostol (RCOG Guideline Development Group 2001). Abortion within 24 h can
be achieved in 90–97% of cases, depending on the misoprostol regime (Goldberg et al 2001).
Surgical Termination of Pregnancy Surgical vacuum aspirations are performed throughout the first trimester, although medical methods are better at very early gestations (56 weeks), and up to a recommended gestation of 12– 15 weeks, depending on unit policy (RCOG Guideline Development Group 2001). Cervical priming before surgical termination of pregnancy reduces the operative morbidity associated with aspiration procedures, by reducing the force required to achieve adequate cervical dilatation. Prostaglandins administered 3 h beforehand have shown a reduced incidence in uterine perforation, cervical damage, false passages and incomplete evacuation. The only problem lies with the usual side-effects seen with prostaglandin administration, those of abdominal pain, nausea, vomiting and diarrhoea (Ledingham et al 2001). Gemeprost is UK licensed for
566
REPRODUCTIVE SYSTEM
this purpose and is effective, producing a lower cervical resistance, but other agents have been looked at to attempt a reduction in these preoperative symptoms. Vaginal misoprostol has been shown to be at least as effective, with the added benefits of a reduced side-effect profile and much reduced expense (el-Refaey et al 1994). Clinical trials have found 400 mg misoprostol vaginally 3 h prior to surgery to be the optimal regime (Fong et al 1998; Henry and Haukkamaa 1999). Miscarriage Similarly, mifepristone and misoprostol use is now extending to managing silent or incomplete miscarriage, giving women an option to avoid surgical intervention or prolonged conservative management. SUMMARY Clearly, prostaglandins are a major factor in parturition and they have already contributed greatly to our understanding of this process and provided useful clinical tools. The challenge is to further define the role of specific prostaglandins and their receptors, in combination with the many inflammatory mediators known to be involved. There are clearly very complex interactions at play and continued scientific research will hopefully gradually unravel these pathways. An ideal induction agent would cause cervical ripening without uterine activity, allowing this process to complete uninterrupted prior to the initiation of active labour, and without any unwanted side-effects. As there are so many mechanisms involved and their mediators often have wide-ranging effects, the difficulty is with the specificity of effect. Many of the previously mentioned mediators, such as cytokines and nitric oxide, are under investigation as the search continues. It may well be that a combined approach may be the way forward in optimizing our clinical procedure. REFERENCES Arias F (2000) Pharmacology of oxytocin and prostaglandins. Clin Obstet Gynaecol, 43, 455–468. Ashok PW, Flett GM and Templeton A (1998a) Termination of pregnancy at 9–13 weeks’ amenorrhoea with mifepristone and misoprostol. Lancet, 352. Ashok PW, Penney GC, Flett GM and Templeton A (1998b) An effective regimen for early medical abortion: a report of 2000 consecutive cases. Hum Reprod, 13, 2962–2965. Baird DT, Sukcharoen N and Thong KJ (1995) Randomized trial of misoprostol and cervagem in combination with a reduced dose of mifepristone for induction of abortion. Hum Reprod, 10, 1521–1527. Barclay CG, Brennand JE, Kelly RW and Calder AA (1993) Interleukin-8 production by the human cervix. Am J Obstet Gynecol, 169, 625–632. Belayet HM, Kanamaya N, Khatun S et al (1999) Binding of interleukin-8 to heparan sulphate enhances cervical maturation in rabbits. Mol Hum Reprod, 5, 261–269. Bennett BB (1997) Uterine rupture during induction of labour at term with intravaginal misoprostol. Obstet Gynecol, 89, 832–833. Bennett PR, Henderson DJ and Moore GE (1992) Changes in expression of the cyclooxygenase gene in human fetal membranes and placenta with labour. Am J Obstet Gynecol, 167, 213–216. Bygdeman M (1992) In Elder MG (ed.), Clinical Obstetrics and Gynaecology—Prostaglandins, vol 6, no 4. London: Ballie`re-Tindall, 893–903. Bygdeman M (1993) In Vane J and O’Grady J (eds), Therapeutic Applications of Prostaglandins. London: Edward Arnold, 85–91. Cabrol D, Dallot E, Bienkiewicz A et al (1990) Cyclooxygenase and lipoxygenase inhibitors—induced changes in the distribution of
glycosaminoglycans in the pregnant rat uterine cervix, Prostaglandins, 39, 515–523. Calder AA (1979) In Keirse MJNC, Anderson ABM and Gravenhorst JB (eds), Human Parturition, vol 15. Leiden: Leiden University Press, 201– 217. Calder AA (1987) In Hillier K (ed.) Eicosanoids and Reproduction. London: MTP Press Limited, 184–194. Calder AA and Elder MG (1993) In Vane J and O’Grady J (eds), Therapeutic Applications of Prostaglandins. London: Edward Arnold, 92–104. Calder AA and Greer IA (1992) In Elder MG (ed.), Clinical Obstetrics and Gynaecology, vol 6, no 4. London: Ballie`re-Tindall, 771–786. Cameron IT, Michie AF and Baird DT (1986) Therapeutic abortion in early pregnancy with antiprogestogen RU 486 alone or in combination with prostaglandin analogue (Gemeprost). Contraception, 34, 459–468. Challis JR, Patel FA and Pomini F (1999) Prostaglandin dehydrogenase and the initiation of labour. J Perinat Med, 27, 26–34. Chwalisz K and Garfield RE (1998) New molecular challenges in the induction of cervical ripening; nitric oxide as the final metabolic mediator of cervical ripening. Hum Reprod, 13, 245–248. Chwalisz K, Shao-Qing S, Garfield RE and Beier HM (1997) Cervical ripening in guinea pigs after a local application of nitric oxide. Hum Reprod, 12, 2093–2101. Chyu JK and Strassner H (1997) Prostaglandin E2 for cervical ripening: a randomized comparison of Cervidil vs. Prepidil. Am J Obstet Gynecol, 177(3), 606–611. Cotzias CS, Paterson-Brown S and Fisk N (1999) Prospective risk of unexplained stillbirth in singleton pregnancies at term: population based analysis. Br Med J, 319, 287–288. Danforth, DN (1947) The fibrous nature of the human cervix, and its relation to the isthmic segment in the gravid and nongravid uterus. Am J Obstet Gynecol, 53, 541–560. Danielian P, Porter B, Ferri N, Summers J and Templeton A (1999) Misoprostol for induction of labour at term: a more effective agent than dinoprostone vaginal gel. Br J Obstet Gynaecol, 106, 793–797. Denison FC, Riley SC, Elliott CL et al (2000) The effect of mifepristone administration on leukocyte populations, matrix metalloproteinases and inflammatory mediators in the first trimester cervix. Mol Hum Reprod, 6, 541–548. Edwards RK and Richards DS (2000) Preinduction cervical assessment. Clin Obstet Gynecol, 43, 440–446. Ekman G, Malmstrom A, Uldbjerg N and Ulmsten U (1986) Cervical collagen: an important regulator of cervical function in term labour. Obstet Gynecol, 67, 633–636. El Maradny E, Kanamaya N, Halim A et al (1996) Biochemical changes in the cervical tissue of rabbit induced by interleukin-8, interleukin-1b, dehydroepiandrosterone sulphate and prostaglandin E2: a comparative study. Hum Reprod, 11, 1099–1104. el-Refaey H, Calder L, Wheatley DN and Templeton A (1994) Cervical priming with prostaglandin E1 analogues, misoprostol and gemeprost. Lancet, 343, 1207–1209. el-Refaey H and Templeton A (1995) Induction of abortion in the second trimester by a combination of misoprostol and mifepristone: a randomized comparison between two misoprostol regimens. Hum Reprod, 10, 475–478. Elliott CL, Brennand JE and Calder AA (1998) The effects of mifepristone on cervical ripening and labour induction in primigravidae. Obstet Gynecol, 92, 804–809. Ellwood DA, Mitchell MD et al (1980) The in vitro production of prostanoids by the human cervix during pregnancy: preliminary observations. Br J Obstet Gynaecol, 87, 210–214. Fischer D, Kuth A and Winkler M et al (2001) A large keratan sulfate proteoglycan present in human cervical mucus appears to be involved in the reorganization of the cervical extracellular matrix at term. J Soc Gynecol Invest, 8, 277–284. Fong YF, Singh K and Prasad RN (1998) A comparative study using two dose regimens (200 mg or 400 mg) of vaginal misoprostol for preoperative cervical dilatation in first trimester nulliparae. Br J Obstet Gynaecol, 105, 413–417. Friedman MA (2001) Manufacturer’s warning regarding unapproved uses of misoprostol. N Engl J Med, 344, 61. Garfield RE, Saade G, Buhimschi I et al (1998) Control and assessment of the uterus and cervix during pregnancy and labour. Hum Reprod Update, 4, 673–695.
PARTURITION AND THE CLINICAL INTERRUPTION OF PREGNANCY Gebhardt DOE (2001) Misoprostol in a topsy-turvy world. J Med Ethics, 27, 205. Geirsson RT and Greer IA (1990) Prostaglandins: a key factor in human labor. Acta Obstet Gynecol Scand, 69, 371–373. Gibb W (1998) The role of prostaglandins in human parturition. Ann Med, 30, 235–241. Goldberg AB, Greenberg MB and Darney PD (2001) Misoprostol and pregnancy. N Engl J Med, 344, 38–47. Hale RW and Zinberg S. (2001) Use of misoprostol in pregnancy. N Engl J Med, 344, 59–60. Henry AM and Haukkamaa M (1999) Comparison of vaginal misoprostol and gemeprost as pretreatment in first trimester pregnancy interruption. Br J Obstet Gynaecol, 106, 540–543. Hilder L, Costeloe K and Thilaganathan B (1998) Prolonged pregnancy: evaluating gestation-specific risks of fetal and infant mortality. Br J Obstet Gynaecol, 105, 169–173. Hofmeyr GJ, Alfirevic Z, Matonhodze B et al (2001) Titrated oral misoprostol solution for induction of labour: a multi-centre, randomised trial. Br J Obstet Gynaecol, 108, 952–959. Hofmeyr GJ, Gulmezoglu AM and Alfirevic Z (1999) Misoprostol for induction of labour: a systematic review. Br J Obstet Gynaecol, 106, 798–803. How HY, Leaseburg L, Khoury JC et al (2001) A comparison of various routes and dosages of misoprostol for cervical ripening and the induction of labour. Am J Obstet Gynecol, 185, 911–915. Husslein P and Egarter CH (1992) In Elder MG (ed.), Clinical Obstetrics and Gynaecology—Prostaglandins, vol 6, no 4. London: Ballie`re Tindall, 755–770. Jain JK and Mishell DR (1994) A comparison of intravaginal misoprostol with prostaglandin E2 for termination of second trimester pregnancy. N Engl J Med, 331, 290–293. Jain JK, Kuo J and Mishell DR Jr (1999) A comparison of two dosing regimens of intravaginal misoprostol for second-trimester pregnancy termination. Obstet Gynecol, 93(4), 571–575. Junqueira LCU, Zugaib M, Montes GS et al (1980) Morphological and histological evidence for the occurrence of collagenolysis, and for the role of neutrophilic polymorphonuclear leukocytes during cervical ripening. Am J Obstet Gynecol, 138, 273–281. Karim SMM, Ng SC and Ratnam SS (1979) In Karim SMM (ed.), Practical Applications of Prostaglandins and their Synthesis Inhibitors. Lancaster, UK: MTP Press Limited. Keirse MJNC (1979) In Keirse MJNC, Anderson ABM and Gravenhorst JB (eds), Human Parturition, vol 15. Leiden: Leiden University Press, 101–142. Keirse MJNC (1992) In Elder MG (ed.), Clinical Obstetrics and Gynaecology—Prostaglandins, vol 6, no 4. London: Ballie`re-Tindall, 787–808. Kelly RW and Bukman A (1990) Antiprogestogenic inhibition of uterine prostaglandin inactivation: a permissive mechanism for uterine stimulation. J Steroid Biochem Mol Biol, 37, 97–101. Kelly RW, Cheng L, Thong J et al (1992) In Drife JO and Calder AA (eds), Prostaglandins and the Uterus. Berlin: Springer-Verlag, 3–13. Knudsen UB, Uldbjerg N, Rechberger T and Fredens K (1997) Eosinophils in human cervical ripening. Obstet Gynecol, 72, 165–168. Kolderup L, McLean L, Grullon K et al (1999) Misoprostol is more efficacious for labour induction than prostaglandin E2, but is it associated with more risk? Am J Obstet Gynecol, 180, 1543–1550. Kwon JS, Davies GAL and MacKenzie VP (2001) A comparison of oral and vaginal misoprostol for induction of labour at term: a randomised trial. Br J Obstet Gynaecol, 108, 23–26. Ledger WL, Webster M, Harrison LP et al (1985) Increase in cervical extensibility during labor induced after isolation of the cervix from the uterus in pregnant sheep. Am J Obstet Gynecol, 151, 397–402. Ledingham M, Denison FC, Riley SC and Norman JE (1999) Matrix metalloproteinases-2 and -9 and their inhibitors are produced by the human uterine cervix but their secretion is not refulated by nitric oxide donors. Hum Reprod, 14, 2089–2096. Ledingham M, Thomson AJ, Lunan CB et al (2001) A comparison of isosorbide mononitrate, misoprostol and combination therapy for first trimester pre-operative cervical ripening: a randomised control trial, Br J Obstet Gynaecol, 108, 276–280. Liggins GC (1981) In Ellwood DA and Anderson ABM (eds), The Cervix in Pregnancy and Labour, Clinical and Biological Investigations. Edinburgh: Churchill Livingstone, 1–9.
567
Lundin-Schiller S and Mitchell MD (1990) The role of prostaglandins in human parturition. Prostagland, Leukotrienes Essent Fatty Acids, 39, 1–10. MacKenzie IZ (1992) In Drife JO and Calder AA (eds), Prostaglandins and the Uterus. Berlin: Springer-Verlag, 119–133. Mackenzie WE (2001) Misoprostol and the politics of fear. Lancet, 357, 1296. McKinley C, Thong KJ and Baird DT (1993) The effect of dose of mifepristone and gestation on the efficacy of medical abortion with mifepristone and misoprostol. Hum Reprod, 8, 1502–1505. Mitchell MD (1987a) In Hillier K (ed.), Eicosanoids and Reproduction. Lancaster, UK: MTP Press, 89–107. Mitchell MD (1987b) In Hillier K (ed.), Eicosanoids and Reproduction. Lancaster, UK: MTP Press, 108–127. Moore PK (1985) Prostanoids: Pharmacological, Physiological and Clinical Relevance. London: Cambridge University Press. Ngai C SW, Tang LCH, Chan YM and Ho PC (2000) Vaginal misoprostol alone for medical abortion up to 9 weeks of gestation: efficacy and acceptability, Hum Reprod, 15, 1159–1162. Norman JE (1992) In Drife JO and Calder AA (eds), Prostaglandins and the Uterus. Berlin: Springer-Verlag, 101–118. Norman JE, Thong KJ and Baird DT (1991) Uterine contractility and induction of abortion in early pregnancy by misoprostol and mifepristone. Lancet, 338, 1233–1236. Nuutila M, Toivonen J, Ylikorkala O and Halmesmaki E (1997) A comparison between two doses of intravaginal Misoprostol and Gemeprost for induction of second-trimester abortion. Obstet Gynecol, 90(6), 896–900. Osmers RGW, Blaser J, Kuhn W and Tschesche H (1995) Interleukin-8 synthesis and the onset of labour. Obstet Gynecol, 86, 223–229. Peyron R (1993) Early termination of pregnancy with mifepristone (RU 486) and the orally active prostaglandin misoprostol. N Engl J Med, 328, 1509–1513. Plaut MM, Schwartz ML and Lubarsky SL (1999) Uterine rupture associated with the use of misoprostol in the gravid patient with a previous caesarean section. Am J Obstet Gynecol, 180, 1535–1542. Radestad A, Bygdeman M and Green K (1990) Induced cervical ripening with mifepristone (RU486) and bioconversion of arachidonic acid in human pregnant uterine cervix in the first trimester. Contraception, 41, 283–292. Rajabi MR, Dean DD, Beydoun SN and Woessner JF (1988) Elevated tissue levels of collagenase during dilation of uterine cervix in human parturition. Am J Obstet Gynecol, 159, 971–976. Rand L, Robinson JN, Economy KE and Norwitz ER (2000) Post-term induction of labour revisited. Obstet Gynecol, 96, 779–783. RCOG Clinical Effectiveness Support Group (2001) Induction of labour— evidence-based clinical guideline, No. 9. London: RCOG. RCOG Guideline Development Group (2001) The care of women requesting induced abortion. London: RCOG. Reece MS, McGregor JA, Allen KGD et al (1996) Prostaglandins in selected reproductive tissues in preterm and full term gestations. Prostagland, Leukotrienes Essent Fatty Acids, 55, 303–307. Rodger MW, Logan AF and Baird DT (1989) Induction of early abortion with mifepristone (RU 486) and two different doses of prostaglandin pessary (gemeprost). Contraception, 39, 497–502. Romero R, Ceska M, Avila C et al (1991) Neutrophilattractant/activating peptide-1/interleukin-8 in term and preterm parturition. Am J Obstet Gynecol, 165, 813–820. Rowlands S, Bell R, Donath S et al (2001) Misoprostol versus dinoprostone for cervical priming prior to induction of labour in term pregnancy: a randomised controlled trial. Austral NZ J Obstet Gynaecol, 41, 145–152. Rozenberg P, Chevret S, Goffinet F et al (2001) Induction of labour with a viable infant: a randomised clinical trial comparing intravaginal misoprostol and intravaginal dinoprostone. Br J Obstet Gynaecol, 108, 1255–1262. Sanchez-Ramos L, Peterson DE, Delke I et al (1988) Labor induction with prostaglandin E1 misoprostol compared with dinoprostone vaginal insert: a randomized trial. Obstet Gynecol, 91(3), 401–405. Sanchez-Ramos L, Gaudier FL and Kaunitz AM (2000a) Cervical ripening and labor induction after previous cesarean delivery. Clin Obstet Gynecol, 43, 513–523.
568
REPRODUCTIVE SYSTEM
Sanchez-Ramos L and Kaunitz AM (2000b) Misoprostol for cervical ripening and labor induction: a systematic review of the literature. Clin Obstet Gynecol, 43, 475–488. Sato T, Michizu H, Hashizume K and Ito A (2001) Hormonal regulation of PGE2 and COX-2 production in rabbit uterine cervical fibroblasts. J Appl Physiol, 90, 1227–1231. Schaff EA, Eisinger SH, Stadalius LS et al (1999) Low-dose mifepristone 200mg and vaginal misoprostol for abortion. Contraception, 59, 1–6. Schaff EA et al (2000) Vaginal misoprostol administered 1, 2 or 3 days after mifepristone for early medical abortion. JAMA, 284, 1948–1953. Sennstrom MB, Ekman G, Westergen-Thorsson G et al (2000) Human cervical ripening, an inflammatory process mediated by cytokines. Mol Hum Reprod, 6, 375–381. Silvestre L, Dubois C et al (1990) Voluntary interruption of pregnancy with mifepristone (RU 486) and a prostaglandin analogue. A large scale French experience. N Engl J Med, 322, 645–648. Spitz IM, Bardin CW, Benton L and Robbins A (1998) Early pregnancy termination with mifepristone and misoprostol in the United States. N Engl J Med, 338, 1241–1247. Steinborn A, Gunes H and Halberstadt E (1995) Signal for term parturition is of trophoblast and therefore of fetal origin. Prostaglandins, 50, 237–252. Sugano T, Narahara H, Nasu K et al (2001) Effects of platelet-activating factor on cytokine production by human uterine cervical fibroblasts. Mol Hum Reprod, 7, 475–481. Thiery M (1979) In Keirse MJNC, Anderson ABM and Gravenhorst JB (eds), Human Parturition, vol 15. Leiden: Leiden University Press, 155– 164. Thomson AJ, Lunan CB, Ledingham M et al (1998) Randomised trial of nitric oxide donor versus prostaglandin for cervical ripening before first-trimester termination of pregnancy. Lancet, 352, 1093–1096. Tomlinson AJ, Archer PA and Hobson S (2001) Induction of labour: A comparison of two methods with particular concern to patient acceptability. J Obstet Gynaecol, 21, 239–241. UK Multicentre Trial (1990) The efficacy and tolerance of mifepristone and prostaglandin in first trimester termination of pregnancy. Br J Obstet Gynaecol, 97, 480–486. Uldbjerg, N., Ekman, G., Malmstrom, A et al (1983a) Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. Am J Obstet Gynecol, 147, 662–666. Uldbjerg N, Ulmsten U and Ekman G (1983b) The Ripening of the human uterine cervix in terms of connective tissue biochemistry. Clin Obstet Gynecol, 26, 14–26.
Uldbjerg N, Ulmsten U and Ekman G (1987) In Hillier K (ed.), Eicosanoids and Reproduction. Lancaster, UK: MTP Press Limited, 163–183. Van Meir CA, Ramirez MM, Matthews SG et al (1997) Chorionic prostaglandin catabolism is decreased in the lower uterine segment with term labour. Placenta, 18, 109–114. Wagner M (2001) Misoprostol and the politics of convenience. Lancet, 357. Webster D, Penney GC and Templeton A (1996) A comparison of 600 and 200 mg mifepristone prior to second trimester abortion with the prostaglandin misoprostol. Br J Obstet Gynaecol, 103(7), 706–709. Whittle WL, Patel FA, Alfaidy N et al (2001) Glucocorticoid regulation of human and ovine parturition: the relationship between fetal hypothalamic-pituitary-adrenal-axis activation and intrauterine prostaglandin production. Biology of Reproduction, 64, 1019–1032. WHO Task Force on post-ovulatory methods of fertility regulation (2000) Comparison of two doses of mifepristone in combination with misoprostol for early medical abortion: a randomised trial. Br J Obstet Gynaecol, 107, 524–530. Wing DA (1999) Labour induction with misoprostol. Am J Obstet Gynaecol, 181, 339–345. Wing DA, Fassett MJ and Mishell DR (2000) Mifepristone for preinduction cervical ripening beyond 41 weeks’ gestation: a randomized controlled trial. Obstet Gynecol, 96, 543–548. Winkler M, Fischer D, Hlubek M et al (1998) Interleukin-1beta and interleukin-8 concentrations in the lower uterine segment during parturition at term. Obstet Gynecol, 91, 945–949. Wiqvist N (1979) In Keirse MJNC, Anderson ABM and Gravenhorst JB (eds), Human Parturition, vol 15. Leiden: Leiden University Press, 189– 200. Witter FR (2000) Prostaglandin E2 preparations for preinduction cervical ripening, Clin Obstet Gynecol, 43, 469–474. Wong KS, Ngai CSW, Wong AYK (1998) Vaginal Misoprostol compared with vaginal Gemeprost in termination of second trimester pregnancy: a randomized trial. Contraception, 58(4), 207–210. Yamamoto T, Eckes B, Mauch C et al (2000) Monocyte chemoattractant protein-1 enhances gene expression and synthesis of matrix metalloproteinase-1 in human fibroblasts by an autocrine IL-1a loop. J Immunol, 164, 6174–6179. Yudkin PL, Wood L and Redman CWG (1987) Risk of unexplained stillbirths at different gestational ages. Lancet, 1, 1192– 1194.
53 Foetal and Neonatal Ductus Arteriosus Kazuo Momma Tokyo Women’s Medical University, Tokyo, Japan
FOETAL DUCTUS ARTERIOSUS Anatomy and Physiology In mammalian foetuses, the ductus arteriosus is widely patent between the main pulmonary artery and the descending aorta (Walsh et al 1974; Rudolph 2001). As shown in Figure 53.1, the foetal ductus arteriosus has the same diameter as the main pulmonary artery and descending aorta, forming a smooth arch together with the main pulmonary artery and descending aorta as viewed in sagittal section. The ductus arteriosus is a typical muscular artery, and the medial layer consists of abundant smooth muscle cells and a small amount of elastic fibres (Walsh et al 1974). The aorta and pulmonary arteries are elastic arteries, and the medial layer contains a small amount of smooth muscle cells and abundant elastic fibres, which can be demonstrated with Elastica–Van Gieson stain (Figure 53.2). The abundant smooth muscle cells in the ductus arteriosus form the basis of its strong neonatal constriction and rapid closure after birth. In the foetus, the lung is filled with fluid and does not function in gas exchange. The foetal pulmonary circulation has very high resistance, and 90% of the right ventricular output passes through the ductus to the descending aorta (Rudolph 2001), to return to the placenta via the umbilical arteries. In the foetus, right ventricular output is slightly more than that of the left ventricle (Rudolph 2001). Since 55% of total cardiac output is ejected by the right ventricle, half of the systemic output flows through the foetal ductus arteriosus. The ductus arteriosus is an essential passway in the foetus.
Mechanisms of Foetal Patency Oxygen Hypoxia and low environmental pO2 form the basis of foetal ductal patency (Dawes 1968; Rudolph 2001). In the foetal circulation, right ventricular blood is more desaturated than left ventricular blood because of preferential flow of superior vena caval blood to the right and inferior vena caval and ductus venosus blood to the left ventricle. Because of this preferential flow, the pO2 of blood in the foetal ductus is about 20 mmHg (Rudolph 2001). pO2-dependent ductal constriction, studied with foetal lamb ductal strips, occurs at 20–150 mmHg of pO2 (Dawes 1968; Rudolph 2001). The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
Prostaglandins Foetal plasma PG and ductal PG receptors. In the early 1970s, studies with prostaglandin and its inhibitors revealed that prostaglandins are essential for ductal dilatation of the nearterm foetus (Heymann 1983; Coceani et al 1986; Clyman 1990). Shown first in rats and rabbits, this was confirmed in other species, including humans. Non-steroidal antiinflammatory drugs (NSAIDs), such as aspirin and indomethacin, are inhibitors of the prostaglandin synthase cyclooxygenase. It was established that the antipyretic and analgesic effects of NSAIDs act through their inhibition of cyclooxygenase and reduction of prostaglandin synthesis. In vivo and in vitro studies have shown that all acidic NSAIDs constrict the foetal ductus arteriosus (Momma 1988; Momma and Takeuchi 1983; Momma and Takao 1990; Momma et al 1984). Both prostaglandin type E and type I are synthesized in the ductal wall (Clyman 1990). In the foetal ductus, the medial smooth muscle is far more sensitive to PGE1 and PGE2 than to PGI2 in in vitro studies (Clyman 1990). In vivo study in rat neonates shows that the neonatal ductus is 1000 times more sensitive to dilatation by PGE1 than to PGI2 analogues. Circulating prostaglandins in foetal plasma have additional effects in dilating the foetal ductus arteriosus (Clyman 1990; Corbet 1998). The plasma concentration of foetal PGE1 and PGE2 is about five times higher than maternal levels (Lewis et al 1978). The source of foetal PGE is the placenta. COX-2 is induced in cotyledonary tissues during late gestation (Wimsatt et al 1993). Foetal pulmonary blood flow is small, and catabolism of plasma prostaglandins in the pulmonary circulation is minimal. Neonatal plasma concentrations of PGE1 and PGE2 fall rapidly after birth (Challis et al 1978) because of termination of placental supply and increased catabolism by the pulmonary circulation after birth. In the early 1990s, two cyclooxygenases, type 1 (COX-1) and type 2 (COX-2), were identified (Vane and Botting 1998; Feldman et al 2000; Crofford et al 2000). As discussed in other chapters in this book, COX-1 is constitutive, and works in homeostasis, protecting the gastric mucosa, activating platelets and maintaining renal function. COX-2 is inducible and pathological. It is expressed locally in inflammation, fever and pain. Interestingly, COX-2 is induced physiologically in reproduction. A focal increase of COX-2 expression is found at the site of blastocyst attachment in the uterus (Crofford et al 2000). COX-2 is 10–100 times more abundant than COX-1 in full-term foetal membranes (Slater et al 1995). COX-2 is upregulated during labour in the foetal membranes (Slater et al 1995). The selective COX-2
570
REPRODUCTIVE SYSTEM
Figure 53.1 Right sagittal section of the foetal rat. The right ventricle continues to the main pulmonary artery, the ductus arteriosus and the descending aorta, forming a smooth arch. The ductus is widely open: A, anterior; DA, ductus arteriosus; dAo, descending aorta; I, inferior; LA, left atrium; lPA, left pulmonary artery; lSVC, left superior vena cava; mPA, main pulmonary artery; MV, mitral valve; P, posterior; PV, pulmonary vein; RV, right ventricle; S, superior; TV, trisuspid valve
Figure 53.2 Histology of the ductus arteriosus (DA) and aortic arch (Ao) in the foetal rat. The ductus arteriosus is constricted with maternally administered indomethacin (10 mg/kg). Elastica–van Gieson stain. The aorta contains abundant elastic fibres which are darkly stained. The ductus contains smaller amount of elastic fibres
FOETAL AND NEONATAL DUCTUS ARTERIOSUS inhibitor effectively prevents preterm delivery (Sawdy et al 1997). PGE2 in foetal plasma, in which the concentration is several times higher than in the maternal plasma, is probably produced by abundant COX-2 in the foetal membranes. Developmental changes in PGE2 receptor subtypes in the ductus arteriosus has been studied in foetal and neonatal pigs (Bhattacharya et al 1999) and lambs (Bouayad et al 2001). In the foetus of 75% and 90% gestation, EP2, EP3 and EP4 receptors are present in the ductus. In immediate neonates, only the EP2 receptor is present in the ductus, suggesting decreased sensitivity to PGE2 in the neonatal ductus (Smith et al 1994; Bouayad et al 2001). The developmentally essential role of EP4 in closure of the neonatal ductus, was shown in the study of mice lacking the EP4 receptor (Nguyen et al 1997; Segi et al 1998). In these mice, the ductus arteriosus failed to close after birth and they died of heart failure shortly after. COX in ductus tissues. COX-1 and COX-2 in foetal ductal tissues have been studied in animal experiments. Only COX-1 is present at mid-gestation and COX-2 increases markedly at term in foetal ductal tissue in pigs (Guerguerian et al 1998). In the lamb foetus, both COX-1 and COX-2 contribute to the production of PGE2 by ductal tissue of the term foetus, but COX-2 is more abundant than COX-1 in term ductus tissue (Guerguerian et al 1998). COX-2 is expressed in the endothelium in the foetal lamb
571
ductus arteriosus, while COX-1 is expressed in endothelium and smooth muscle (Clyman et al 1999). COX inhibitors: non-steroidal antiinflammatory drugs (NSAIDs)— experimental findings. Many antiinflammatory drugs available before 1990 block both COX-1 and COX-2. These are mixed COX blockers. Momma and Takeuchi (1983) and Momma et al (1984a) studied foetal ductal constriction by more than 60 NSAIDs in rats. These drugs were administered through an orogastric tube in near-term rats and foetal ductal diameters were studied 4 h later, following atlas dislocation, foetal rapid wholebody freezing, cutting on a freezing microtome, and measurement of the ductal and pulmonary trunk diameters with a micrometer (Momma et al 1985; Figure 53.3). Transplacental transfer of indomethacin to the foetus was studied in rats. After orogastric administration, plasma concentration of indomethacin peaked in 4 h in the mothers and in 8 h in the foetuses (Momma and Takao 1987). Foetal plasma concentrations were 30–50% of the maternal concentrations. Foetal ductal constriction was maximal 4 and 8 h after maternal administration (Figure 53.4). The time courses of foetal ductal constriction were similar following administration of other NSAIDs (Momma and Takao 1987). Every acidic NSAID was transferred through the placenta and constricted the foetal ductus arteriosus dose-dependently (Figure 53.5). Ductal constriction varied according to the clinical dose and differed from drug to drug. These NSAIDs are classified into four groups
Figure 53.3 Morphologies of the constricted foetal ductus arteriosus following maternal orogastric administration of indomethacin (10 mg/kg). Ductal constriction is most severe at 8 h. LPA, left pulmonary artery; MV, mitral valve; other abbreviations as in previous figures. Reproduced from Momma et al (1985), published by the International Pediatric Research Foundation
572
REPRODUCTIVE SYSTEM
Figure 53.4 Time courses of foetal ductal constriction by maternally administered aspirin (100 mg/kg, orogastric), indomethacin (10 mg/kg, orogastric) and NS 398 (selective COX-2 inhibitor, 10 mg/kg, s.c. injection) in near-term rats. Maximal constriction occurs at 4 and 8 h
Figure 53.5 Foetal ductal constriction by six NSAIDs. Dose–response curves. Inner diameter ratio of the ductus to the main pulmonary artery (DA/PA) is plotted on the y axis. In the control rat, DA/PA is 1.0. Every NSAID induces foetal ductal constriction dose-dependently. Moderate foetal ductal constriction is induced by the clinical doses of indomethacin (mixed inhibitor), celecoxib and rofecoxib (two selective COX-2 inhibitors). The full dose of SC 560 (selective COX-1 inhibitor) causes only mild ductal constriction
FOETAL AND NEONATAL DUCTUS ARTERIOSUS
573
Table 53.1 Foetal ductal constriction following administration of clinical doses of antiinflammatory drugs in rats. Drugs are grouped according to potency. Measurements were made 4 h after orogastric administration, except for six drugs administered by intramuscular (i.m.) or intraperitoneal (i.p.) injection. Selective COX-2 inhibitors* are included, in addition to the mixed COX inhibitors (Momma et al 1984a) Grade
DA:PA ratio
NSAID
Severe
0.4–0.7
Celecoxib*, clidanac, diclofenac sodium, etodolac*, fenbufen, fenoprofen calcium, flufenamic acid, flurbiprofen, glafenin, ibuprofen, indomethacin, ketoprofen, mefenamic acid, naproxen, pranoprofen, protizinic acid, rofecoxib*, suprofen, tolmetin
Moderate
0.7–0.9
Aluminium flufenamate, azapropazone, betamethasone, choline salicylate, clofezone, dexamethasone, fentiazac, floctafenine, ketophenylbutazone, meloxicum*, metiazinic acid, oxyphenbutazone, phenylbutazone, piroxicam, sulindac, suxibutazone, tiaprofenic acid
Mild
0.9–1.0
Acetaminophen, alclofenac, aminopyrine, antipyrine, aspirin, aspirin aluminium, aspirin-DL-lysine, benoxaprofen, bucetine, bucolome, hydrocortisone (i.m.) isopropylantipyrine, nicotinoylaminoantipyrine, phenacetin, prednisolone (i.m.), sasapyrine, sodium salicylate (i.m.), sulpyrine, tiaramide HCl
None
1.0–1.1 (control: 1.02+0.03)
Benzydamine HCl, ethoxybenzamide, mepirizole, perisoxal citrate, phenylacetylglycine dimethylamide, salicylamide, sodium salicylamide-o-acetate (i.p.), tinolidine HCl
in Table 53.1, according to foetal ductal constriction and clinical dose (Momma 1988). It is noteworthy that indomethacin, ibuprofen and selective COX-2 inhibitors, including celecoxib and rofecoxib, are strong constrictors of the foetal ductus, while aspirin and acetaminophen are rather mild constrictors at the clinically used doses (Momma and Takao 1990; and unpublished data). The cardiac effects of foetal ductal constriction were further studied in rats. Indomethacin-induced foetal ductal constriction for 24 h caused right ventricular concentric hypertrophy and left ventricular dilatation (Momma and Takao 1989a). At the same time, foetal ductal constriction increased the blood flow to the
ascending aorta and aortic isthmus, resulting in increased foetal isthmic diameter (Momma and Ando 1994). Foetal ductal constriction by indomethacin varies according to foetal stage. In rats, the foetal stage is short. The embryonic stage ends at the 14th day of gestation and the foetal stage ranges from day 15 to day 21.5. Foetal ductal constriction by indomethacin is strong on day 21, but very weak on days 19 and 20 (Figure 53.6) (Momma and Takao 1987). With the full dose of indomethacin, the foetal ductal inner diameter diminishes to 20% of the control in the near-term foetus, but only to 75% in the preterm foetus. In experimental lambs (gestation period 150 days), foetal ductus sensitivity to indomethacin also changes in the course of gestation.
Figure 53.6 Dose–response curves of foetal ductal constriction by indomethacin, rofecoxib and SC 560 in pre-term foetal rats. The maximum effects of these three drugs were the same and the foetal ductus constricted to about 75% of the original diameter
574
REPRODUCTIVE SYSTEM
The foetal lamb ductus does not constrict following administration of indomethacin before 100 days of gestation, but consistently constricts after 103 days of gestation (Coceani et al 1992). The mechanism of this change is related to nitric oxide, as stated later in this chapter. Foetal ductal constriction by selective COX-2 inhibitors (rofecoxib, celecoxib, NS 398) was studied in vivo and in vitro in the lamb (Clyman et al 1999; Takahashi et al 2000; Coceani et al 2001). These COX-2 inhibitors induced ductal constriction in the near-term foetus to the same or a slightly smaller degree compared to the constriction induced by indomethacin. As shown in Figure 53.5, selective COX-2 inhibitors constricted the foetal ductus arteriosus of near-term rats dose-dependently, and maximum effects were the same as with indomethacin, a mixed COX inhibitor (Toyoshima et al 2000). A selective CO-1 inhibitor, SC 560, constricted the near-term foetal ductus only weakly, suggesting a major role of COX-2 and a minor role of COX-1 in the near-term foetal ductal patency. In the preterm foetal ductus, both COX-1 inhibitor and COX-2 inhibitor constricted the foetal ductus to the same degree as the mixed blocker, indomethacin, suggesting an equal role of COX-1 and COX-2 in foetal ductal patency (Momma et al unpublished data). These are compatible with the presence and dominance of COX-1 and COX-2 in the foetal ductus and foetal membrane. Non-steroidal antiinflammatory drugs (NSAIDs) in pregnant women. Neonates may sustain persistent pulmonary hypertension (PPHN) following maternal ingestion of NSAIDs, such as aspirin and indomethacin (Clyman 1990). The first case of PPHN was reported in a mature newborn baby following maternal ingestion of aspirin for arthritis. Indomethacin is widely used for the treatment of tocolysis. In the late 1970s, it was clearly shown, clinically and experimentally, that maternally-administered aspirin and indomethacin constrict the foetal ductus arteriosus, increase foetal pulmonary vascular smooth muscle, and induce postnatal persistent pulmonary hypertension (Heymann 1983; Clyman 1990). Since the late 1980s, foetal ultrasound studies have demonstrated foetal ductal constriction by accelerated blood flow in the foetal ductus following maternal treatment with indomethacin. In the normal foetus, flow velocity in the ductus arteriosus is less than 1 m/s, and foetal ductal constriction may be demonstrated by increased ductal flow velocity (Huhta et al 1987; Sherer and Divon 1996). Administration of indomethacin should be stopped if such accelerated flow is noticed. Ductal flow velocity returns to a normal level within 3 days after stopping indomethacin, and subsequent foetal and neonatal clinical course is usually uneventful (Reiss 2001). Indomethacin can induce PPHN in both term and pre-term infants (Long 1990). However, foetal ductal sensitivity to indomethacin differs, depending on foetal maturity (Moise 1993; Vermillion et al 1997). Indomethacin does not constrict the human foetal ductus before 24 weeks’ gestation. After 24 weeks, the incidence of ductal constriction by indomethacin increases and almost all foetal ducti constrict with indomethacin after 30 weeks of gestation (Moise 1993; Vermillion et al 1997). These observations suggest that some mechanism other than prostaglandins maintains dilatation of the foetal ductus in the pre-term foetus. The constrictive effect of selective COX-2 inhibitors on the foetal ductus arteriosus is interesting because of possible harm to the foetus following administration to pregnant women. In nearterm pregnant rats following gastric administration, selective COX-2 inhibitors, including celecoxib, rofecoxib, meloxicam, nimesulide and NS-398, induced dose-dependent constriction of the foetal ductus arteriosus (Figure 53.5, Table 53.1). These effects are comparable to that of indomethacin. These selective COX-2 inhibitors can thus damage the foetus during pregnancy.
Nitric Oxide (NO) Since the discovery of nitric oxide (NO) as an endotheliumderived relaxing factor (EDRF) (Ignarro 1990), the presence of the NO synthase and NO receptor in the foetal ductus arteriosus has been demonstrated in the lamb ductus arteriosus (Coceani et al 1992; Fox et al 1996; Clyman et al 1998). However, in the nearterm foetal lamb, in vivo and in vitro studies failed to show the dilating role of NO (Fox et al 1996; Clyman et al 1998). Momma and Toyono (1999) have shown that NO is the major dilator of the foetal ductus in the pre-term foetal rat at day 19 of gestation (Figure 53.7). In contrast, prostaglandins are the major dilator of the ductus and NO is an additional dilator in the near-term foetal rat at day 21 of gestation (Figure 53.8) (Momma and Toyono 1999). The mechanism and physiological meaning of the shift from NO to prostaglandins is speculative (Yallampalli et al 1994). NO works to inhibit migration, proliferation and apoptosis of cells, which are important mechanisms of the organic closure of the neonatal ductus arteriosus following functional closure. It is speculated that the shift of a dilating role from NO to prostaglandins occurs as a preparation for postnatal organic ductus closure. Endothelin 1 (ET1) ET1 is a potent vasoconstrictive substance produced in the vascular endothelial cells (Yanagisawa et al 1988). Coceani et al (1992) have reported that it is a potent constrictor of the neonatal ductus arteriosus working in response to oxygen. Foetal ductal constriction by indomethacin and L-NAME in rats is inhibited dose-dependently by bosentan, a potent inhibitor of endothelin synthase (Clozel et al 1994; Momma et al 2003). Therefore, the foetal ductus arteriosus remains patent, with a balance of a constrictor such as endothelin and dilators, such as NO and prostaglandins, with the dilators dominant. Neonatal ductal constriction is delayed following administration of bosentan in rats (Momma et al, unpublished data), showing the physiological role of endothelin in neonatal ductal constriction. Glucocorticoid Hormones Momma reported foetal ductal constriction by glucocorticoid hormones, including betamethasone, dexamethasone, prednisolone and hydrocortisone (Momma et al 1981). Foetal ductal constriction by glucocorticoid is dose-dependent, and a massive dose of maternally-administered glucocorticoid hormones can constrict the foetal ductus arteriosus almost completely (Momma et al 1981). Combined administration of glucocorticoid hormone and indomethacin causes a remarkably increased constriction of the foetal ductus arteriosus (Momma and Takao 1989b). Foetal ductal constriction by glucocorticoid hormones is more prominent in pre-term than in near-term foetal rats (Momma et al 1981). A clinical case of foetal ductal constriction by maternally administered dexamethasone has been reported (Azancot et al 1995). Molecular mechanisms of foetal ductal constriction by glucocorticoid hormones are probably multiple. It is speculated that mechanisms include inhibition of NO synthase (Nasjletti et al 1984; Radomski et al 1990; Bustamante et al 1996), inhibition of PG synthesis, and inducing endothelin discharge. On-going study shows synergistic ductus-constricting effects of glucocorticoid hormones and L-NAME (Figure 53.8) (Momma, unpublished data) and this is compatible with these multiple molecular mechanisms. Retinoic Acid and Vitamin A The constrictive response of the foetal ductus arteriosus to indomethacin can be increased by pretreatment with retinoic acid
FOETAL AND NEONATAL DUCTUS ARTERIOSUS
575
Figure 53.7 Sagittal section of the pre-term foetal rat on the 19th day. Control (A) and 4 h after i.m. injection of L-NAME (10 mg/kg), a potent inhibitor of NO synthase (B). The ductus arteriosus is constricted severely in (B). A, anterior; I, inferior; LV, left ventricle; P, posterior; RA, right atrium; S, superior; other abbreviations as in previous figures
Figure 53.8 Foetal ductal constriction by maternally administered indomethacin (Ind; 1 mg/kg, orogastric), L-NAME (10 mg/kg, i.m.) and dexamethasone (Dex, 1 mg/kg, i.m.) 4 h later in near-term rats. Each drug constricted the foetal ductus mildly. Combined administration constricted the ductus severely
576
REPRODUCTIVE SYSTEM
and vitamin A and attenuated by pretreatment with indomethacin in rats. The retinoic acid receptor is expressed dominantly in the media of the foetal ductus arteriosus (Colbert et al 1996), suggesting its specific role in its maturation. Retinoic acid, administered 1 and 2 days before term, increases foetal ductal constriction by indomethacin in near-term foetal rats (Momma 2000). Vitamin A has the same effect, and its clinical dose (1 mg/kg), administered 1 and 2 days before term, increases foetal ductal constriction by indomethacin in near-term foetal rats (Figure 53.9) (Momma et al 1998). The oxygen sensitivity of the foetal ductus arteriosus is increased by maternally administered vitamin A (Wu et al 2001). Neonatal ductal constriction is accelerated by prenatal vitamin A in rats (Sasaki et al 2000). Retinoic acid and vitamin A themselves do not constrict the foetal ductus arteriosus. It remains to be checked clinically whether aspirin, ibuprofen and indomethacin cause more foetal ductal constriction following intake of increased vitamin A in pregnant women. Clinically, indomethacin treatment of the patent ductus arteriosus in premature infants may not be effective in some cases (Moore et al 2001). Clinical studies are necessary to determine whether pretreatment with vitamin A increases ductal constriction in the premature baby treated with indomethacin.
infant (Norton et al 1993; Souter et al 1998), neonatal periventricular haemorrhage and impaired renal function (Souter et al 1998). Clinically and experimentally, early exposure of the foetal ductus to indomethacin is associated with attenuated ductal constriction (Clyman et al 2001). Clinically, the incidence of patent ductus arteriosus in premature babies was increased in those exposed to indomethacin in foetal life secondary to its administration to mothers to treat tocolysis (Norton et al 1993). Premature babies with foetal ductal exposure to indomethacin were more resistant to postnatal treatment (Norton et al 1993). The mechanisms of these changes include increased ductal NO production and remodelling (Clyman et al 2001). The same phenomenon was observed in rats experimentally. Exposure of pre-term foetal rats to indomethacin on foetal days 19 and 20 induced a lack of constrictive response to indomethacin at foetal day 21 (Konishi et al 1986). L-NAME, a potent NO synthase inhibitor, constricts the foetal ductus arteriosus on foetal day 21 following early exposure to indomethacin in rats (Momma, unpublished data). NEONATAL DUCTUS ARTERIOSUS Anatomy and Physiology
Early Exposure of the Foetal Ductus to Indomethacin Prenatal administration of indomethacin to the pregnant woman may cause several side-effects in the neonate, including increased incidence and severity of patent ductus arteriosus in the premature
Neonatal ductal closure (Olley 1987; Neal 1990; Rudolph 2000) Clinically, the morphology and haemodynamics of the neonatal ductus arteriosus have been studied by two-dimensional and
Figure 53.9 Frontal cut of the dilated DA in (A) (no vitamin A, no indomethacin) and (B) (vitamin A on 19th and 20th day), and constricted DA in (C) (indomethacin on 21st day) and in (D) (vitamin A on the 19th and 20th day and indomethacin on the 21st day) in near-term foetal rats. AoA, aortic arch; RPA, right pulmonary artery; other abbreviations as in previous figures. Reproduced from Momma et al (1998), published by the International Pediatric Research Foundation
FOETAL AND NEONATAL DUCTUS ARTERIOSUS Doppler echocardiography (Moore et al 2001). In the initial hours after birth, the ductus is widely patent and bidirectional flow is observed: 4 h later, pulmonary arterial pressure falls and the ductus arteriosus becomes narrower. Flow through the ductus is from aorta to pulmonary artery at this stage. About 15 h later, the ductus arteriosus closes (Rudolph 2001). Ductal closure may be delayed in some infants, especially those that are premature. However, in most mature infants, the ductus arteriosus closes completely by 72 h after birth (Moore et al 2001). Closure of the neonatal ductus is functional at first. It may reopen if exposed to hypoxia or prostaglandins. After several days, the closed ductus undergoes organic change. Apoptosis and fibrosis occur and it gradually becomes a fibrotic cord. In experimental animals, neonatal ductal constriction, its time course, its morphology, and its response to dilating prostaglandins were studied. Neonatal ductal constriction in rats and rabbits is very rapid, and the ductus closes by 1 h after birth in these small animals. Neonatal ductal constriction occurs uniformly (Figure 53.10). PGE can dilate the closed ductus initially, but ductal dilatation by PGE is progressively attenuated, and no dilatation occurs by 24 h after complete closure (Figure 53.11) (Momma et al 1980). In the postnatal constricted ductus, ductal tissue becomes severely hypoxic, and severe hypoxia inhibits production of prostaglandins and nitric oxide (Kajino et al 2000). Apoptosis of the ductal medial layer occurs (Imamura et al 2000) and the ductal smooth muscle cells are replaced by connective tissue. The role of oxygen is clear as an initiating mechanism of neonatal ductal closure. The ductal tissue is very sensitive to oxygen. Constriction of ductal tissue is minimal at 20 mmHg pO2, and becomes maximal at 150 mmHg pO2 (Rudolph 2001). Foetal pO2 in the ductus is about 20 mmHg, and increases to 60 and 100 mmHg after birth (Rudolph 2001). Foetal ductal constriction by oxygen is less prominent in the premature foetus (Rudolph 2001).
577
Decrease of prostaglandins is another factor in the promotion of neonatal ductal constriction (Rudolph 2001). The placenta produces prostaglandins with foetal plasma concentration of PGE about five times higher than the maternal concentration. After birth, neonatal plasma concentration of PGE falls rapidly, promoting ductal constriction. Neonatal mice without COX-1 and COX-2 die with patent ductus arteriosus (Loftin et al 2001). This shows that COX-1 and COX-2 are important for postnatal ductal remodelling. The molecular mechanism of oxygen in constriction of the ductus arteriosus is under study. The constrictive response of the ductus arteriosus is very sensitive to oxygen, and this reaction is rather specific in the ductus. Oxygen dilates the pulmonary artery and the different reaction of these arteries to oxygen suggests different molecular mechanisms intervening between oxygen and the contracting elements of the smooth muscle cells of these arteries. In vitro studies have shown production of endothelin (Coceani et al 1992) and inhibition of the potassium channels (Nakanishi et al 1993; Tristani-Firouzi et al 1996; Reeve and Weir 2000) as the mechanism of neonatal ductus constriction by oxygen. Patent Ductus Arteriosus in the Premature Baby Patent ductus arteriosus (PDA) is frequent in the premature baby, with increased incidence in those premature with lower body weights, younger gestational age and respiratory distress syndrome (Clyman 1990; Corbet 1998). The plasma concentration of PGE is increased in these infants compared with premature infants without PDA. Indomethacin and other NSAIDs are used to close the PDA in the premature infant (Clyman 1996; Corbet 1998). Indomethacin
Figure 53.10 Right sagittal sections of the neonatal ductus of rats at 0, 30, 60 and 90 min after birth. The ductus arteriosus constricts uniformly and closes by 90 min in neonatal rats. RAA, right atrial appendage; other abbreviations as in previous figures. Reproduced from Momma et al (1985), published by the International Pediatric Research Foundation
578
REPRODUCTIVE SYSTEM
Figure 53.11 Age-related ductus-dilating effect of PGE1 in newborn rabbits. PGE1, 0.5 mg/g, was injected subcutaneously to mature newborn rabbits, and 30 min later the ductal size was measured. The ductus-dilating effect was most prominent 1 h after birth, but was lost 24 h later. Reproduced from Momma et al (1980), published by the International Pediatric Research Foundation
treatment is indicated for those infants with symptomatic PDA, such as those with respiratory distress necessitating respirator control, or with congestive heart failure (Clyman 1996). Indomethacin treatment is contraindicated in infants with renal failure with elevated serum creatinine, jaundice, necrotizing enterocolitis, and haemorrhagic tendency (Corbet 1998). Orogastric or intravenous indomethacin, 0.2 mg/kg, have been used (Clyman 1996), but inconsistent gastroenteric absorption of indomethacin in the premature infant favours intravenous administration. It can be repeated two or three times, at 12 h intervals, if PDA persists on echocardiography. Administration of indomethacin may be associated with transient oliguria or anuria, but urine output usually recovers in a few days if indomethacin is stopped. Other possible side-effects include intracranial haemorrhage, necrotizing enterocolitis, and haemorrhagic tendency. Closure of the ductus is associated with improvement of respiratory distress and congestive heart failure. In an effort to find a more effective and safer therapy, ibuprofen has been used to close the patent ductus arteriosus in the premature infant. Compared with indomethacin, ibuprofen had the same effect of closing the ductus, and with fewer side-effects, including oliguria (Patel et al 1995; Van Overmeier et al 1997, 2000). Another line of new pharmacological treatment was studied, based on the importance of NO in preterm ductal dilatation. Combined inhibition of prostaglandins and nitric oxide in the preterm baboon induced remarkably accelerated ductal closure (Seidner et al 2001), suggesting a new clinical strategy. Therapeutic Dilatation of the Neonatal Ductus Prostaglandin E1 (PGE1) is essential in the medical treatment of many severe neonatal congenital cardiac defects (Corbet 1998; Rudolph 2001). In these diseases, foetal circulation is maintained without difficulty because of placental gas exchange and widely
patent ductus arteriosus. After birth, gas exchange shifts to the lung and the ductus arteriosus begins to close. Prostaglandin treatment of the ductus arteriosus is needed in the following three categories of diseases (Momma et al 1984b; Momma 1996): pulmonary atresia; ductus-dependent systemic circulation; and complete transposition of the great arteries. Pulmonary Atresia The first category includes those congenital heart diseases in which the pulmonary circulation depends on the patent ductus arteriosus. This category includes several entities associated with pulmonary atresia and severe pulmonary stenosis. Pulmonary atresia may be associated with intact ventricular septum, tetralogy of Fallot, double outlet right ventricle, single ventricle, tricuspid atresia, transposition of the great arteries, and Ebstein’s anomaly (Momma and Takeuchi 1983; Corbet 1998). Infants with pulmonary atresia develop cyanosis in the first few days as the ductus constricts. They develop severe hypoxaemia and metabolic acidosis without proper treatment. Inhalation of pure oxygen is not effective because of inadequate pulmonary blood flow. Intravenous infusion of PGE1 is life-saving. Because of rapid catabolism and short half-life, PGE1 is infused continuously intravenously. The initial starting dose is 30–100 ng/ kg/min. The effects of ductal dilatation usually appear in 30 min following the start of infusion. Cyanosis lessens and arterial pO2 increases (Figure 53.12) (Momma 1996; Corbet 1998). Respiratory distress becomes less. Metabolic acidosis disappears. A continuous murmur of PDA appears. Ductal flow becomes prominent on the echocardiogram. After 5–10 h, the dose of PGE1 can be decreased to a minimal maintenance dose, about 10 ng/kg/min. PGE1 infusion is not without side-effects, which include respiratory depression and apnoea, diarrhoea, elevation of temperature, twitching, and bleeding tendency (Corbet 1998).
FOETAL AND NEONATAL DUCTUS ARTERIOSUS
579
Figure 53.12 The effects of PGE1 in dilating the ductus arteriosus in those neonates with pulmonary atresia. The x axis is time after start of infusion; y axis is arterial pO2. The conventional PGE1 (PGE1 CD) dilates the ductus and pO2 increases in 30 min. Lipo-PGE1 is another preparation of PGE1, in which PGE1 is included in a small lipid particle. Smaller dose of lipo-PGE1 is effective compared with conventional PGE1. Reproduced from Momma (1996). Copyright # 1996, by permission of Elsevier
Apnoea may be treated by stimulation of the skin, but may require intubation and a respirator. Prolonged administration of PGE1 over 1 or 2 months induces transient thickening of the periosteum of the long bones including tibia and fibula (Momma and Takeuchi 1983). It disappears in several months after stopping PGE1 infusion. Oral PGE2 is available (Momma 1988; Corbet 1998) but the effect is less reliable and intravenous PGE1 is generally recommended for initial treatment. Oral PGE2 can be used in cases where surgical treatment is not urgently needed. In the vast majority of neonates with pulmonary atresia, shunt operation is indicated following PGE1 treatment. PGE1 treatment serves to stabilize their general condition, allowing safer diagnostic procedures and surgical intervention. Historically, surgical results in these neonates improved dramatically following clinical application of PGE1.
Ductus-dependent Systemic Circulation Some neonatal congenital heart diseases depend for their systemic blood flow partially or totally on a patent ductus arteriosus. These diseases include hypoplastic left heart syndrome, interruption of the aortic arch, and coarctation of the aorta (Corbet 1998). In hypoplastic left heart syndrome, systemic blood flow depends totally on the ductus. In interruption or coarctation of the aortic arch, blood flow to the lower half of the body is supplied through the ductus arteriosus. As the ductus constricts after birth in these diseases, these patients may develop severe circulatory shock, anuria, and metabolic acidosis a few days after birth. Infusion of PGE1 dilates the ductus arteriosus, increases systemic and renal blood supply, and
stabilizes the general condition of the patient, providing enough time to establish a diagnosis and to prepare for surgical intervention.
Complete Transposition of the Great Arteries In this disease, the pulmonary and systemic circulations are parallel, and the patient survives by shunts between these two circulations. Shunts take place through the foramen ovale and ductus arteriosus immediately after birth. When the ductus remains open, blood shunts from the aorta to the pulmonary artery through the ductus, and between the atria via the foramen ovale. The aortic oxygen content depends on the oxygenated blood flow from the left atrium to the right atrium. As the ductus constrict a few days after birth, the patient develops severe cyanosis and hypoxaemia. Intravenous PGE1 infusion dilates the ductus arteriosus, increases pulmonary blood flow and elevates aortic oxygen saturation (Corbet 1998). Neonatal arterial switch operation for transposition can be achieved with a high success rate at the present time (Castaneda et al 1994; Kirklin et al 1993).
Effects The ductus-dilating effect of intravenously-infused PGE1 appears within 1 min and lasts for 15–20 min after the infusion is stopped (Momma 1988). Orogastrically-administered PGE2 dilates the ductus for 3 h (Momma 1988). Prolonged PGE1 infusion for more than 12 h usually causes a greater increase of arterial pO2. In those cases in which PGE1 infusion is initiated in the early neonatal
580
REPRODUCTIVE SYSTEM
period, the ductus remains responsive to PGE1 infusion for several months. PGE1 infusion fails to dilate the ductus even in the early neonatal period if the ductus is already completely closed. The effect of PGE1 infusion is small or absent if it is infused in infants initially after 1 month of age (Momma 1988). REFERENCES Azancot-Benisty A, Benifla JL, Matias A et al (1995) Constriction of the fetal ductus arteriosus during prenatal betamethasone therapy. Obstet Gynecol, 85, 874–876. Bhattacharya M, Asslin P, Hardy P et al (1999) Developmental changes in prostaglandin E2 receptor subtypes in porcine ductus arteriosus. Circulation, 100, 1751–1756. Bouayad A, Kajino H, Waleh N et al (2001) Characterization of PGE2 receptors in fetal and newborn lamb ductus arteriosus. Am J Physiol Heart Circ Physiol, 280, 2342–2349. Bustamante SA, Pang Y, Romero S et al (1996) Inducible nitric oxide synthese and the regulation of central vessel caliber in the fetal rat. Circulation, 94, 1948–1954. Castaneda AR, Jonas RA, Mayer JE and Hanley FL (1994) Cardiac Surgery of the Neonate and Infant. Philadelphia: Saunders, 65–87. Challis JRG, Hart I, Lois TM et al (1978) Prostaglandins in the sheep fetus: importance for fetal function. In Coceani F and Olley PM (eds), Advances in Prostaglandin and Thromboxane Research, vol 4. New York: Raven, 115–132. Clozel M, Breu V, Gray GA (1994) Pharmacological characterization of bosentan, a new potent orally active non-peptide endothelin receptor antagonist. J Pharmacol Exp Therapeut, 270, 228–235. Clyman RI (1990) Development physiology of the ductus arteriosus. In Long WA (ed.), Fetal and Neonatal Cardiology. Philadelphia: W.B. Saunders, 64–75. Clyman RI (1996) Recommendations for the postnatal use of indomethacin: an analysis of four separate treatment strategies. J Pediatr, 126, 601–607. Clyman RI, Waleh N, Black SM et al (1998) Regulation of ductus arteriosus patency by nitric oxide in foetal lamb: the role of gestation, oxygen tension, and vasa vasorum. Pediatr Res, 43, 633–644. Clyman RI, Hardy P, Waleh N et al (1999) Cyclo-oxygenase-2 plays a significant role in regulating the tone of the fetal ductus arteriosus. Am J Physiol, 45, R913–R921. Clyman RI, Chen YQ, Chemtob S et al (2001) In utero remodeling of the foetal lamb ductus arteriosus. The role of antenatal indomethacin and avascular zone thickness on vasa vasorum proliferation, neointimal formation, and cell death. Circulation, 103, 1806–1812. Clozel M, Breu V, Gray GA et al (1994) Pharmacological characterization of bosentan, a new potent orally active non-peptide endothelin antagonist. J Pharmacol Exp Therapeut, 270, 228–235. Coceani F, Huhtanen D, Hamilton NC et al (1986) Involvement of intramural prostaglandin E2 in prenatal patency of the lamb ductus arteriosus. Can J Physiol Pharmacol, 64, 737–744. Coceani F, Kelsey L and Seidlitz E (1992) Evidence for an effector role of endothelin in closure of the ductus arteriosus at birth. Can J Physiol Pharmacol, 70, 1061–1064. Coceani F, Kelsey L and Seidlitz E (1993) Occurrence of endotheliumderived relaxing factor-nitric oxide in the lamb ductus arteriosus. Can J Physiol Pharmacol, 72, 82–88. Coceani F, Ackerley C, Seidlitz E and Kelsey L (2001) Function of cyclooxygenase-1 and cyclo-oxygenase-2 in the ductus arteriosus from foetal lamb: differential development and change by oxygen and endotoxin. Br J Pharmacol, 132, 241–251. Colbert MC, Kirby ML and Robbins J (1996) Endogenous retinoic acid signaling colocalizes with advanced expression of the adult smooth muscle myosin heavy chain isoform during development of the ductus arteriosus. Circ Res, 78, 790–798. Corbet AJ (1998) Medical manipulation of the ductus arteriosus. In Garson A Jr, Bricker TJ, Fisher DJ and Neish SR (eds), The Science and Practice of Pediatric Cardiology, 2nd edn. Baltimore: Williams & Wilkins, 2489–2513. Crofford LJ, Lipsky PE, Brooks P et al (2000) Basic biology and clinical application of specific cyclooxygenase-2 inhibitors. Arthrit Rheumat, 43, 4–13.
Dawes GS (1968) Fetal and Neonatal Physiology. Chicago: Year Book, 66– 90, 91–103, 106–114, 125–133, 160–172. Feldman M and McMahon AT (2000) Do cyclooxygenase-2 inhibitors provide benefits similar to those of traditional non-steroidal antiinflammatory drugs, with less gastrointestinal toxicity? Ann Intern Med, 132, 134–143. Fox JJ, Ziegler JW, Ivy DD et al (1996) Role of nitric oxide and cGMP system in regulation of ductus arteriosus tone in ovine fetus. Am J Physiol, 271 (Heart Circ Physiol, 40), H2638–H2645. Guerguerian AM, Hardy P, Bhattacharya M et al (1998) Expression of cyclooxygenase in ductus arteriosus of foetal and newborn pigs. Am J Obstet Gynecol, 179, 1618–1626. Hawkey CJ (1999) COX-2 inhibitors. Lancet, 353, 307–314. Heymann MA (1983) Patent ductus arteriosus. In Adams FH and Emmanouillides GC (eds), Moss’ Heart Disease in Infants, Children and Adolescents, 3rd edn. Baltimore: Williams & Wilkins, 158–171. Huhta JC, Moise KJ, Fisher DJ et al (1987) Detection and quantification of constriction of the fetal ductus arteriosus by Doppler echocardiography. Circulation, 75, 406–412. Ignarro LJ (1990) Biosynthesis and metabolism of endothelium-derived nitric oxide. Ann Rev Pharmacol Toxicol, 30, 535–560. Imamura S, Nishikawa T, Hiratsuka E, Takao A and Matsuoka R (2000) Behavior of smooth muscle cells during arterial ductal closure at birth. J Histochem Cytochem, 48, 35–44. Kajino H, Chen Y-Q, Chemtob S et al (2000) Tissue hypoxia inhibits prostaglandin and nitric oxide production and prevents ductus arteriosus reopening. Am J Physiol, 279, R278–R286. Kirklin JW and Baratt-Boyes BG (1993) Cardiac Surgery, 2nd edn. New York: Churchill Livingstone, 1383–1468. Konishi T, Momma K and Takao A (1986) Inhibition of rat ductus arteriosus maturation by indomethacin. Japan J Neonatol, 22, 430–436 (in Japanese with English summary). Lewis PE, Cefalo RC, Naulty JS et al (1978) Maternal and fetal prostaglandin levels during acute maternal hypoxia in the pregnant ewe. In Coceani F and Olley PM (eds), Advances in Prostaglandin and Thromboxane Research, vol 4. New York: Raven, 133–135. Loftin CD, Trivedi DB, Tiano H et al (2001) Failure of ductus arteriosus closure and remodeling in neonatal mice deficient in cyclooxygenase-1 and cyclooxygenase-2. Proc Natl Acad Sci USA, 98, 1059–1064. Long WA (1990) Persistent pulmonary hypertension of the newborn syndrome (PPHN). In Long WA (ed.), Fetal and Neonatal Cardiology. Philadelphia: W.B. Saunders, 627–655. Moise KJ (1993) Effect of advancing gestational age on the frequency of fetal ductal constriction in association with maternal indomethacin use. Am J Obstet Gynecol, 168, 1350–1353. Momma K, Uemura S, Nishihara S and Ota Y (1980) Dilatation of the ductus arteriosus by prostaglandins and prostaglandin precursors. Pediatr Res, 14, 1074–1077. Momma K, Nishihara S and Ota Y (1981) Constriction of the fetal ductus arteriosus by glucocorticoid hormones. Pediatr Res, 15, 19–21. Momma K and Takeuchi H (1983) Constriction of fetal ductus arteriosus by non-steroidal antiinflammatory drugs. Prostaglandins, 26, 631–643. Momma K, Hagiwara H and Konishi T (1984a) Constriction of fetal ductus arteriosus by non-steroidal anti-inflammatory drugs: study of additional 34 drugs. Prostaglandins, 28, 527–536. Momma K, Takao A, Sone K and Tashiro M (1984b) Prostaglandin E1 treatment of ductus-dependent congenital heart disease. Int Angiol, 3(suppl), 33–38. Momma K, Konishi T and Hagiwara H (1985) Characteristic morphology of the constricted fetal ductus arteriosus following maternal administration of indomethacin. Pediatr Res, 19(5), 493–500. Momma K and Takao A (1987) In vivo constriction of the ductus arteriosus by non-steroidal antiinflammatory drugs in near-term and pre-term foetal rats. Pediatr Res, 22, 567–572. Momma K (1988) Ductus arteriosus and cardiovascular system in the neonate. In Curtis-Prior PB (ed.), Prostaglandins: Biology and Chemistry of Prostaglandins and Related Eicosanoids. Edinburgh: Churchill Livingstone, 476–489. Momma K and Takao A (1989a) Right ventricular concentric hypertrophy and left ventricular dilatation by ductal constriction in fetal rats. Circulat Res, 64, 1137–1146. Momma K and Takao A (1989b) Increased constriction of the ductus arteriosus with combined administration of indomethacin and betamethasone in fetal rats. Pediatr Res, 25, 69–75.
FOETAL AND NEONATAL DUCTUS ARTERIOSUS Momma K and Takao A (1990) Transplacental cardiovascular effects of four popular analgesics in rats. Am J Obstet Gynecol, 162, 1304–1310. Momma K and Ando M (1994) In situ morphology of fetal aortic isthmus following ductal constriction in rats. Fetal Diagn Therapy, 9, 53–61. Momma K (1996) Lipo-PGE1 treatment of the neonate with critical congenital heart disease and ductus arteriosus-dependent circulation. Adv Drug Delivery Rev, 20, 177–180. Momma K, Toyono M and Miyagawa-Tomita S (1998) Accelerated maturation of fetal ductus arteriosus by maternally administered vitamin A in rats. Pediatr Res, 43, 629–632. Momma K and Toyono M (1999) The role of nitric oxide (NO) in dilating the foetal ductus arteriosus in rats. Pediatr Res, 46, 311–315. Momma K (2000) Accelerated maturation of the ductus arteriosus with retinoic acid in the fetal rat. In Clark E and Takao A (eds), Developmental Mechanisms of Heart Disease. New York: Futura, 261–264. Momma K, Nakanishi T and Imamura S (2003) Inhibition of in vivo constriction of fetal ductus arteriosus by endothelial receptor blockade in rats. Pediatr Res, 53, 479–485. Moore P, Brook MM and Heymann MA (2001) Patent ductus arteriosus. In Allen HD, Gutgesell HP, Clark EB and Driscoll DJ (eds), Moss Adams’ Heart Disease in Infants, Children and Adolescents. Baltimore: Willaims & Wilkins, 652–669. Nakanishi T, Gu H, Hagiwara N and Momma K (1993) Mechanisms of oxygen-induced contraction of ductus arteriosus isolated from the fetal rabbit. Circulat Res, 72, 1218–1228. Nasjletti A, Erman A, Cagen LM and Baer PG (1984) Plasma concentrations, renal excretion, and tissue release of prostaglandins in the rat with dexamethasone-induced hypertension. Endocrinology, 114, 1033–1040. Neal WA (1990) Patent ductus arteriosus. In Moller JH and Neal WA (eds), Fetal, Neonatal and Infant Cardiac Disease. Norwalk: Appleton & Lange, 391–400. Nguyen MT, Camenisch T, Snouwaert JN et al (1997) The prostaglandin receptor EP4 triggers remodelling of the cardiovascular system at birth. Nature, 390, 78–81. Norton ME, Merrill J, Cooper BAB et al (1993) Neonatal complications after the administration of indomethacin for premature labor. N Engl J Med, 329, 1602–1607. Olley PM (1987) The ductus arteriosus, its persistence and its patency. In Anderson RH, Macartney FJ, Shinebourne EA and Tynana M (eds), Paediatric Cardiology. Edinburgh: Churchill Livingstone, 931–958. Patel J, Marks KA, Roberts I et al (1995) Ibuprofen treatment of patent ductus arteriosus. Lancet, 346, 255. Radomski MW, Palmer RMJ and Moncada S (1990) Glucocorticoids inhibits the expression of an inducible, but not the constitutive, nitric oxide synthase in vascular endothelial cells. Proc Natl Acad Sci USA, 87, 10043–10047. Reeve HL and Weir EK (2000) Regulation of ion channels in the ductus arteriosus. In Weir EK, Archer SL and Reeves JT (eds), The Fetal and Neonatal Pulmonary Circulations. Armonk, NY: Futura, 319–330. Reiss RE (2001) Maternal diseases and therapies affecting the fetal cardiovascular system. In Allen HD, Gutgesell HP, Clark EB and Driscoll DJ (eds), Moss Adams’ Heart Disease in Infants, Children and Adolescents. Baltimore: Williams & Wilkins, 551–568. Rudolph AM (2000) The fetal circulation and its adjustments after birth. In Moller JH, Hoffman JIE (eds), Pediatric Cardiovascular Medicine. New York: Churchill Livingstone, 60–64.
581
Rudolph AM (2001) Congenital Diseases of the Heart, Clinical– Pathological Considerations, fully revised and updated 2nd edn. New York: Futura, 155–196. Sasaki Y and Momma K (2000) Accelerated maturation of neonatal ductus arteriosus by prenatally administered vitamin A in rats. Japan J Prematurity Neonatol, 12, 97–101 (in Japanese with English summary). Sawdy R, Slater D, Fisk N et al (1997) Use of a cyclo-oxygenase type-2selective non-steroidal anti-inflammatory agent to prevent preterm delivery. Lancet, 350, 265–266. Segi E, Sugimoto Y, Yamasaki A et al (1998) Patent ductus arteriosus and neonatal death in prostaglandin receptor EP4-deficient mice. Biochem Biophys Res Comm, 246, 7–12. Seidner SR, Chen YQ, Opysko PR et al (2001) Combined prostaglandin and nitric oxide inhibition produces anatomic remodeling and closure of the ductus arteriosus in the premature baboon. Pediatr Res, 50, 365– 373. Sherer DM and Divon MY (1996) Prenatal ultrasonographic assessment of the ductus arteriosus: a review. Obstet Gynecol, 87, 630–637. Slater D, Berger L, Moore GE and Bennett PR (1995) Changes in the expression of types 1 and 2 cyclo-oxygenase in the human fetal membranes at term. Am J Obstet Gynecol, 172, 77–82. Smith GCS, Coleman RA and McGrath JC (1994) Characterization of dilator prostanoid receptors in the fetal rabbit ductus arteriosus. J Pharmacol Exp Therapeut, 271, 390–396. Souter D, Harding J, McCowan L et al (1998) N Z J Obstet Gynecol, 38, 11–16. Takahashi Y, Roman C, Chemtob S et al (2000) Cyclooxygenase-2 inhibitors constrict the fetal lamb ductus arteriosus both in vitro and in vivo. Am J Physiol, 278, R1496–R1505. Toyoshima K, Takeda S and Momma K (2000) Fetal ductal constriction by selective cyclooxygenase-2 inhibitor (NS-398) in rats. Japan J Neonatol, 36, 593–597 (in Japanese with English summary). Tristani-Firouzi M, Reeve HL and Tolarova SL (1996) Oxygen-induced constriction of rabbit ductus arteriosus occurs via inhibition of a 4aminopyrodine voltage-sensitive potassium channel. J Clin Invest, 98, 1959–1965. van Overmeier B, Follens I, Hartmann S et al (1997) Treatment of patent ductus arteriosus with ibuprofen. Arch Dis Childhood, 76, F179–F184. van Overmeier B, Smets K, Lecoutere D et al (2000) A comparison of ibuprofen and indomethacin for closure of patent ductus arteriosus. N Engl J Med, 343, 674–681. Vane JR and Botting RM (1998) Mechanism of action of nonsteroidal anti-inflammatory drugs. Am J Med, 104(3A), 2S–8S. Vermillion ST, Scardo JA, Lashus AG and Wiles HB (1997) The effect of indomethacin tocolysis on fetal ductus arteriosus constriction with advancing gestational age. Am J Obstet Gynecol, 177, 256–261. Walsh SZ, Meyer WW and Lind JL (1974) The Human Fetal and Neonatal Circulation. Springfield, IL: Charles C. Thomas, 59–86. Wimsatt J, Nathanielsz PW and Sirois J (1993) Induction of prostaglandin endoperoxide synthase isoform-2 in ovine cotyledonary tissues during late gestation. Endocrinology, 133, 1068–1073. Wu G-R, Jing S, Momma K and Nakanishi T (2001) The effect of vitamin A on contraction of the ductus arteriosus in fetal rat. Pediatr Res, 49, 747–754. Yallampalli C, Byam-Smith M, Nelson SO and Garfield RE (1994) Steroid hormones modulate the production of nitric oxide and cGMP in the rat uterus. Endocrinology, 134, 1971–1974. Yanagisawa M, Kurihara H, Kimura S et al (1988) A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature, 332, 411–415.
Section Ten Conclusions and Correlations
54 Biochemical Interactions of Platelet-activating Factor with Eicosanoids Joseph T. O’Flaherty and Robert L. Wykle Wake Forest University School of Medicine, Winston-Salem, NC, USA
Platelet-activating (PAF) and eicosanoids have intertwining biochemistries. In part, this reflects their common molecular origin. The major glycerophospholipid classes, i.e. those containing phosphocholine (PC), phosphoethanolamine (PE), phosphoserine or phosphoinositol head groups, comprise 1-Oacyl, 1-O-alkyl and 1-O-alk-1’-enyl subclasses. They typically contain a saturated long chain residue at the sn-1 position and arachidonic acid (AA) or other unsaturated fatty acid esterified to the sn-2 position. Hydrolysis of key phospholipids at sn-2 thus yields the AA used for eicosanoid synthesis as well as the 1-Oalkyl-2-lyso-GPC (GPC is sn-glycero-3-phosphocholine) that is acetylated to 1-O-alkyl-2-acetyl-GPC, i.e. PAF. In consequence, production of the two types of mediators can be coordinated. Beyond this, PAF and the eicosanoids stimulate each other’s production and have overlapping or even synergistic actions: these agents not only form but also act cooperatively. This chapter reviews recent work that has delved into these issues. It places some emphasis on human polymorphonuclear neutrophils (PMN), a cell type favoured for study because of its responsiveness to, and high content of precursors for, PAF. PMN and the lipid mediators they produce are implicated in diverse host defence and self-injury responses. BIOLOGICAL PROFILES OF PAF AND EICOSANOIDS PAF acts through a receptor belonging to the rhodopsin superfamily (Honda et al 1991). The gene for this receptor exists as a single copy that has two 5’-non-coding exons with separate promoters and start sites. Hence, two transcripts with identical open reading frames can and frequently do form by alternative splicing in various tissues (Mutoh et al 1993). The PAF receptor often shows less sensitivity to pertussis toxin than do the rhodopsin superfamily receptors for other chemotactic factors. This may indicate that PAF receptors link with the Gaq or Ga12 to a far greater extent than the Gai class of G proteins (Amatruda et al 1993; Ali et al 1994, 1998). More typical chemotactic factor receptors appear to link with Gai as well as, or in preference to, Gaq and Ga12 G protein classes. Such variations may underlie differences in the activity of PAF compared to other chemotactic factors (see below). PAF, like most chemotactic factors, stimulates its target cells to make leukotrienes (LT), hydoxyicosatetraenoates (HETE), prostaglandins and other eicosanoids. These AA metabolites resemble PAF in that they can activate cells by binding to known LTB4, The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
prostanoid, thromboxane A2 (Smith 1997; Yokomizo et al 1997) or presumed 5-HETE (Smith 1997; O’Flaherty et al 1998) rhodopsin superfamily receptors. Based on their sensitivities to bacterial toxins, the receptors for excitatory eicosanoids, LTB4 and 5-HETE being prominent examples, link mainly to the Gai class of G proteins (O’Flaherty et al 1988); those for inhibitory eicosanoids, e.g. prostaglandins E2 and D2, link to Gaa in addition to other G protein classes (Smith 1997). There is, then, a fundamental analogy between PAF and eicosanoids: they are synthesized in stimulated cells and bind to receptors which, depending on G protein linkages, upregulate or downregulate cell functions. In vivo, PAF causes: platelet thrombosis; PMN, monocyte, and eosinophil migration; smooth muscle contraction; enhanced vascular permeability; and lipid mediator synthesis. PAF may thereby contribute to allergy (Tamura et al 1996), inflammation (Tjoelker et al 1995; Crespo et al 1996) and syndromes involving shock (Lopes-Martins et al 1996; Ishii et al 1997), ischaemiareperfusion (Silver et al 1996), coagulation (Morooka and Natsume 1996) and multiple organ failure (Muguruma et al 1996). In the latter capacities, PAF apparently acts in concert with eicosanoids and both mediator types seem inextricably involved in defence and toxic reactions, e.g. 5-lipoxygenaseknockout mice are protected from PAF-induced anaphylaxis (Chen et al 1994). PAF may also orchestrate slower evolving events, e.g. transgenic mice that are made to overexpress PAF receptors not only have increased bronchial reactivity and endotoxin sensitivity but also develop chronic epidermal hyperplasia and melanotic tumours (Ishii et al 1997). PAF is likewise implicated in the progressive neuronal apoptosis seen in HIV-infected monocyte-neuron cultures and, by extrapolation, HIV-related dementia (Perry et al 1998). Finally, mutations in LIS-1 are associated with Miller–Dieker and isolated lissencephaly syndromes. LIS-1 codes for a subunit of acetylhydrolase, an enzyme that inactivates PAF (Ho et al 1997; Pe´terfy et al 1998). The abnormally high PAF levels attending the loss of acetylhydrolase activity, it is speculated, disrupt neuron migration during foetal development to cause brain malformation (Manya et al 1998). As implied by these examples, persistent dysfunction may result from prolonged PAF receptor excitation without involving eicosanoids. In the short term, then, PAF commonly works in conjunction with eicosanoids. In its prolonged actions, however, PAF may have no such connection. Perhaps related to this, the stimulus-induced short-term but not the constitutional and long-term synthesis of PAF involves metabolic pathways that release AA.
586
CONCLUSIONS AND CORRELATIONS
PAF SYNTHESIS IN STIMULATED CELLS The Deacylation–Acetylation Pathway As reviewed previously (Rossi and O’Flaherty 1991), early studies indicated that the sequential actions of phospholipase A2 (PLA2) and acetyltransferase deacylate and acetylate resident 1-O-alkyl-2acyl-GPC to mediate the stimulus-induced generation of PAF. Figure 54.1 shows that this route necessitates AA release. Various types of evidence support this scheme. Cells incorporate [3H]AA at the sn-2 carbon of 1-O-alkyl-2-lyso-GPC and on challenge release this label and synthesize [3H]-eicosanoids as well as PAF (Rossi and O’Flaherty 1991). Moreover, PMN from rats deprived of polyunsaturated fatty acids have little capacity to make LTB4 or PAF in response to stimulation. Addition of AA to these cells reverses both defects (Ramesha and Pickett 1986). Similarly, HL60 cells depleted of AA in vitro are poor PAF producers, and AA repletion of the cells returns PAF synthetic responses to normal (Suga et al 1990). Finally, macrophages from cPLA2 knockout mice fail to make either PAF or LTB4 when stimulated (Uozumi et al 1997) (cPLA2 is defined below). These results imply that PLA2-induced AA release is essential for PAF synthesis during cell stimulation. This conclusion supports the deacylation–acetylation pathway of Figure 54.1. However, it is equally compatible with a newly defined route wherein coenzyme A-independent transacylase (CoA-IT) contributes to PAF production. CoA-IT and Phospholipid Remodelling In processing exogenous AA, cells first activate the fatty acid with CoA using ATP-dependent acyl-CoA ligases (MacDonald and Sprecher 1991), a cloned enzyme which acts selectively on polyenoates (Cao et al 1998). An acyl-CoA:lyso-phospholipid acyltransferase then esterifies this AA to 1-O-acyl phospholipids, after which any one of three phospholipid remodelling reactions may shift the AA to 1-O-alkyl and 1-O-alk-1’-enyl subclasses of PC and PE (Blank et al 1992; Choy et al 1997; Yamashita et al 1997). In the Lands cycle, PLA2 hydrolyses 1-O-acyl-2-arachidonoyl phospholipids to release AA. Released AA is in turn esterified to 1-O-alkyl and 1-alk-l’-enyl-GPC and -GPE by the sequential actions of ligase and acyltransferase. This cycle requires
CoA and ATP. A CoA-dependent but ATP-independent pathway may also contribute to AA movement. Here, acyltransferase may reverse its usual direction to accept AA from 1-acyl phospholipids and pass it to 1-O-alkyl and 1-O-alk-l’-enyl-GPC and -GPE. This route prefers 1-acyl phospholipids as fatty acid acceptor, is less active on PE and is not specific for AA (Sugiura et al 1995; Yamashita et al 1995). In a third pathway, remodelling proceeds in the absence of CoA and ATP. The CoA-IT conducting this reaction is selective for AA, prefers PE and PC to other phospholipids, and effects a direct exchange (i.e. no free fatty acid is detectable) from a donor phospholipid to an acceptor lysophospholipid (Kramer and Deykin 1985; Sugiura et al 1987; Blank et al 1995; Nixon et al 1996). Since most tissues express robust CoA-IT activity, this enzyme, rather than the Lands cycle or reversal of acyltransferase activity, may make an important contribution to stimulus-induced AA release and PAF synthesis. Several findings support this notion. First, CoA-IT has the phospholipid specificity required to form PAF precursor. Second, on the basis of mass, the AA released during cell stimulation, at least that found in PMN, originates mainly from the 1-O-alk-l’-enyl subclass of PE (Chilton and Connell, 1988). The 1-O-alk-1’-enyl-2-lyso-GPE (GPE is snglycero-3-phosphoethanolamine) product of this reaction is a good acceptor for CoA-IT and therefore well suited to sponsor AA transfers that generate lyso-PAF. Third, the addition of lysoPE and lyso-PC, regardless of subclass, to isolated cell membranes triggers the formation of lyso-PAF and PAF by causing AA to move out of endogenous 1-O-alkyl-2-arachidonoyl-GPC. The effect occurs independently of CoA and ATP in the membranes of PMN (Nieto et al 1991; Venable et al 1991, 1993), platelets (Kramer and Deykin 1983), macrophages (Sugiura et al 1987), monocytes (Winkler et al 1991), HL-60 cells (Uemura et al 1991), amnion WISH cells (Toyoshima et al 1994), mast cells (Colard et al 1993) and various rat organs (Blank et al 1995). Fourth, PMN treated with tumour necrosis factor-a doubles the Vmax of CoA-IT and concurrently shifts the AA in diverse phospholipids to 1-Oalk-l’-enyl- and 1-O-alkyl-2-lyso-GPE. These cells also exhibit a heightened PAF synthetic capacity (Winkler et al 1994). (However, we were unable to find any change in CoA-IT activity in PMN challenged with tumour necrosis factor-a or various other stimuli, under conditions virtually identical to Wimbler et al (Baker et al 2002).) Lipopolysaccharide and protein kinase C activators stimulate similar responses in other cell types (Breton
Figure 54.1 The deacylation–acetylation pathway. PLA2 acts on 1-O-alkyl-2-arachidonoyl-GPC to form 1-O-alkyl-2-lyso-GPC. Acetyltransferase then acetylates the 1-O-alkyl-2-lyso-GPC to PAF while the released AA is oxygenated to eicosanoids, e.g. PMN use 5-lipoxygenase (Lox) to convert AA to LTB4, 5-HETE, and, if a specific NADP+-dependent dehydrogenase is active, 5-oxo-ETE
INTERACTIONS OF PLATELET-ACTIVATING FACTOR WITH EICOSANOIDS and Colard 1991; Si et al 1997; Svetlov et al 1997). Finally, CoAIT inhibitors block not only the transfer of AA from 1-O-acyl-2arachidonoyl phospholipids to 1-O-alk-l’-enyl- and 1-alkyl-2lyso-GPC and -GPE in resting PMN (Chilton et al 1995) but also dampen AA release, eicosanoid formation and PAF synthetic responses in monocytes, PMN and mast cells (Winkler et al 1995, 1998). Taken together, these findings support the scheme depicted in Figure 54.2. In resting cells, acyltransferase esterifies exogenous AA primarily to 1-O-acyl-2-lyso-GPC. CoA-IT then moves this AA to 1-O-alk-1’-enyl- and 1-O-alkyl-2-lyso-GPE and -GPC. During cell stimulation, PLA2 attacks principally the alkyl and alk-l’-enyl subclasses of PE to yield products that have defined fates: AA enters eicosanoid synthetic pathways, while 1-O-alk-l’enyl- and 1-O-alkyl-2-lyso-GPE act as acceptors for CoA-IT, sponsoring AA transfers that generate the 1-O-alkyl-2-lyso-GPC used for PAF synthesis.
587
envision that CoA-IT works coordinately with deacylation– acetylation pathway reactions. PLA2 releases AA primarily from 1-O-alk-l’-enyl-2-arachidonoyl-GPE, 1-O-alkyl-2-arachidonoylGPE, 1-O-acyl-2-arachidonoyl-GPC, 1-O-alkyl-2-arachidonoylGPC and possibly other phospholipids. CoA-IT transfers AA to the lyso products of the latter reaction, perhaps with some illdefined aspect of the system favouring PLA2 attack on the alk-l’enyl and alkyl subclasses of PE and/or CoA-IT-mediated transfers of AA away from 1-O-alkyl-2-arachidonoyl-GPC. By moving AA into preferred targets of cPLA2 (e.g. PE) and out of PAF precursors, CoA-IT supplements the AA release and PAF synthetic responses implemented by PLA2. Although the extent of this supplementation awaits further analysis, PLA2 remains in all events critical for regulating PAF and eicosanoid synthesis.
PLA2 CoA-IT vs. Deacylation–Acetylation in PAF and Eicosanoid Synthesis CoA-IT is a constitutionally active enzyme whose regulation is not understood. In vitro, it transfers AA between PE and PC with little apparent directional or subclass preferences. The PLA2 involved in AA release (see next section) similarly shows only minor preference for PE over PC or for one phospholipid subclass over another, at least as determined in cell-free systems. The intrinsic activities of these enzymes, then, do not explain the net flow of AA from PC to PE demanded by Figure 54.2. Indeed, they allow for PLA2 to deacylate 1-O-alkyl-2-arachidonoyl-GPC and for CoA-IT to transfer AA into 1-O-alkyl-2-lyso-GPC. Working in this direction, CoA-IT would paradoxically reduce the availability of substrate for PAF synthesis. We accordingly
The PLA2 class of enzymes fall into four categories: Ca2+independent PLA2 (iPLA2), acetylhydrolases, cytosolic PLA2 (cPLA2) and secretory PLA2 (sPLA2). The iPLA2 class deacylates sn-2 fatty acids from phospholipids without selectivity for AA. This class may serve housekeeping functions in remodelling phospholipids for the Lands cycle (Balsinde et al 1997; Balsinde and Dennis 1997). Acetylhydrolases release acetate from PAF or short-chain fatty acids from oxidatively fragmented PC (Van Lenten et al 1995; Stafforini et al 1997). The latter species form during stress and have the property of stimulating cells via PAF receptors (Heery et al 1995). Acetylhydrolases thus detoxify certain phospholipids to block inflammatory, allergic and cellular apoptotic responses (Matsuzawa et al 1997; Stafforini et al 1997). The remaining two PLA2 classes mediate AA release and PAF synthesis. Stimulated cells commonly activate mitogen-activated
Figure 54.2 CoA-IT-dependent reactions. In resting cells (left side), free AA is esterified to CoA by ligase and then to 1-O-acyl-2-lyso-GPC by acyltransferase. CoA-IT transfers the AA to 1-O-alk-1’-enyl- (or 1-O-alkyl-)-2-lyso-GPE. In stimulated cells (right side), PLA2 releases AA from 1-O-alk1’-enyl- (or 1-alkyl-)-2-arachidonoyl-GPE. The AA is oxygenated to eicosanoids while CoA-IT transfers AA from 1-O-alkyl-2-arachidonoyl-GPC to 1-Oalk-1’-enyl (or 1-O-alkyl)-2-lyso-GPE. The latter product is acetylated to PAF by acetyltransferase
588
CONCLUSIONS AND CORRELATIONS
protein kinases (MAPK) and raise cytosolic Ca2+. MAPK activate cPLA2; Ca2+ transients cause cPLA2 to translocate to membranes and release mainly sn-2-AA from PE and PC (Roberts 1996; Leslie 1997). sPLA2 are secreted by stimulated cells and, in the presence of millimolar extracellular CA2+, release sn-2 fatty acids from cell surface phospholipids (Diez et al 1994; Cupillard et al 1997; Tischfield 1997). Given their distinct dispositions, cPLA2 and sPLA2 seem positioned to perform very differently. This appears to be the case. Monocytes depleted of cPLA2 by antisense RNA or treated with cPLA2-inhibitors have reduced prostaglandin but normal leukotriene and PAF synthetic responses. Treated with sPLA2 inhibitors, however, the cells show a complete reversal in this pattern, making normal levels of prostaglandins while failing to produce leukotrienes or PAF (Marshall et al 1997; Winkler et al 1997). Monocytes, then, may use cPLA2 to generate products available to the prostaglandin metabolic pathway and sPLA2 to form products accessing leukotriene and PAF synthetic enzymes. In a distinctly different example of this division of labour, cPLA2 and sPLA2 mediate early and late phases of AA release, respectively, in macrophages. However, inhibition of cPLA2 blocks both of these phases. Furthermore, AA, when added to the cPLA2-blocked cells, restores late-phase AA release. These findings imply that the two enzymes intercommunicate, with the AA released by cPLA2 signalling for, or at least permitting, sPLA2 secretion (Balsinde and Dennis 1996; Marshall et al 1997; Murakami et al 1998). The dialogue may also proceed in the opposite sense, i.e. secreted sPLA2 forms 1-O-acyl-2-lyso-GPC and 1-O-alkyl-2-lyso-GPC from cell surface PC. These and other lyso phospholipids that are produced by sPLA2 may penetrate into cells and sponsor CoA-IT transfers that move fatty acids out of intracellular 1-O-acyl-2-acyl-GPC and 1-O-alkyl-2-acyl-GPC. Both the first and second generations (i.e. the products formed by sPLA2 and CoA-IT, respectively) of 1-O-alkyl-2-lyso-GPC may be acetylated to PAF (O’Flaherty et al 1996a). Finally, the AA released by sPLA2 may also enter cells and be converted to eicosanoids such as LTB4 and 5-HETE (Wijkander et al 1995). PAF, LTB4 and 5-HETE (O’Flaherty et al 1996) bind cognate receptors to mobilize MAPK, Ca2+ and cPLA2, while 1-O-acyl-2lyso-GPC may accomplish this by directly activating G proteins (Yuan et al 1996). sPLA2 also bind its cell receptors to activate the MAPK–cPLA2 axis (Lambeau et al 1996; Huwiler et al 1997; Mosior et al 1998). While the occurrence of these events in any given cell may vary with the availability of key enzymes, products and receptors, mounting evidence supports the following: cPLA2 and sPLA2 are conversant components in a network of cell signals. They attack respective internal and cell surface phospholipids to generate products that may selectively enter one or another metabolic pathway. Activation of cPLA2 may regulate the secretion of sPLA2. Then the eicosanoids derived from the AA released by sPLA2, the PAF and 1-O-acyl-2-lyso-PC that are formed by sPLA2 and the sPLA2 itself, may bind to receptors and directly activate G proteins or some other signals to mobilize cPLA2 and thereby the deacylation–acetylation and CoA-IT pathways. Various combinations of these direct and positive feedback reactions appear responsible for regulating the synthesis of PAF and eicosanoids in many cell types. MAPK AND ACETYLTRANSFERASE IN PAF SYNTHESIS The ERK class of MAPK phosphorylates and thereby activates cPLA2 in vitro and in vivo. p38 Kinase and c-Jun N-terminal kinase classes of MAPK also phosphorylate cPLA2 in vitro, but their in vivo effects are uncertain (Leslie 1997). We recently observed that the inhibition of ERK activation with PD 98059 or
the inhibition of p38 kinase by SB 203580 resulted in a modest decline in the ability of PMN to activate cPLA2, release AA and synthesize PAF in response to diverse stimuli. Blockade of both MAPK classes, however, reduced responses by nearly 100%. Thus, ERK and p38 Mpk appear to have overlapping and perhaps redundant abilities to regulate cPLA2 and therefore to mount AA release and PAF synthetic responses in PMN (Nixon et al 1999a). Studies in other cells have yielded different results. Cultured mast cells require ERK to activate cPLA2 and release AA. The activation of p38 kinase, however, antagonizes these responses (Zhang et al 1997). A third pattern emerges in human platelets. Here, cPLA2 activation and AA release proceed normally when either ERK or p38 kinase is blocked (Kramer et al 1995, 1996). The operation of other routes may explain this apparently anomalous result: platelets may use c-Jun N-terminal kinases rather than ERK or p38 kinase to activate cPLA2; they may rely on Ca2+-induced cPLA2 translocations that allow unphosphorylated enzyme to digest membrane phospholipids (Leslie 1997); phosphoinositol-4,5-bisphosphate, which forms during platelet stimulation, may promote cPLA2 activity (Mosior et al 1998); or another PLA2 besides cPLA2 may be responsible for AA release in platelets. Studies to date thus indicate that there are crucial, cell-specific differences in the signals that modulate cPLA2. Among these signals, ERK, sometimes in association with p38 Mpk, may have a dominant role in at least some cell types. In addition to PLA2, acetyltransferase is required for the synthesis of PAF (Wykle et al 1980). The acetyltransferase that acts on lyso-PAF differs from the one that acts upon lysophosphatidic acid in the de novo pathway of PAF synthesis (see later). Both enzymes, however, localize with nuclear membranes (Lee et al 1986; Baker and Chang 1997). Since the nuclear membranes of monocytes contain 1-O-alkyl-2-arachidonoylGPC, 1-O-alkyl-2-arachidonoyl-GPE, 1-O-alk-l’-enyl-1-arachidonoyl-GPE and CoA-IT activity (Surette and Chilton 1998), PAF synthesis, like that of eicosanoids, may take place in or near to the cell nucleus. In any event, lyso-PAF-specific acetyltransferase activity rises 2–4-fold in stimulated cells (Ninio and Bessou 1991; Garcia et al 1993; Rodriguez et al 1993). This response may be mediated by a phosphorylation reaction. cAMP-dependent protein kinase, protein kinase C, and a calmodulin-dependent kinase have been implicated in activating acetyltransferase (reviewed in Nixon et al 1999b). A recent study, however, points to p38 kinase as a key regulator of the enzyme. A broad series of stimuli induced PMN to increase acetyltransferase activity. This increase correlated with the ability of the stimuli to activate p38 kinase but not activate ERK. Furthermore, a p38 kinase inhibitor blocked these responses by 80–100%, while an inhibitor of ERK activation had no such effect. In agreement with these findings, recombinant activated p38 kinase tripled the activity of acetyltransferase in membrane fractions of PMN. Under the same condition, recombinant activated ERK1 and ERK2 were inactive. Finally, protein phosphatase treatment of membrane fractions from stimulated PMN returned the elevated acetyltransferase activity to control values (Nixon et al 1996). It therefore appears that ERK and p38 kinase regulate the initial as well as last steps in the PAF synthetic responses of PMN; they share responsibility for activating cPLA2, while p38 kinase, apparently operating independently of ERK, activates acetyltransferase. The applicability of the latter result to other cells has not yet been investigated. ALTERNATIVE PATHWAYS FOR PAF AND EICOSANOID SYNTHESIS A de novo pathway (Figure 54.3) synthesizes PAF starting with 1acyl-dihydroxyacetone phosphate (Snyder 1995). Critical steps in
INTERACTIONS OF PLATELET-ACTIVATING FACTOR WITH EICOSANOIDS
589
Figure 54.3 The de novo pathway. 1-O-Alkyl-dihydroxyacetone phosphate (DHAP) synthase exchanges fatty alcohol (R-OH) for the acyl residue in 1-Oacyl-DHAP. Product 1-O-alkyl-DHAP is reduced to 1-O-alkyl-2-lyso-sn-3-glycerophosphate (GPA) by an NADPH-dependent oxidoreductase. Acetyltransferase acetylates this lyso PA, and a phosphohydrolase converts 1-O-alkyl-2-acetyl-GPA product to 1-O-alkyl-2-acetyl-glycerol. A cholinephosphotransferase converts the latter intermediate to PAF by adding the phosphocholine supplied by cytosine diphosphate (CDP) choline in the illustrated side reaction
the pathway include acetylation of 1-O-alkyl-2-lyso-phosphatidic acid by a specific acetyltransferase, hydrolysis of 1-O-alkyl-2acetyl-phosphatidate at the sn-3 carbon to yield 1-O-alkyl-2acetyl-glycerol, and transfer of choline phosphate from CDPcholine to 1-O-alkyl-2-acetylglycerol. Rate-limiting factors in this pathway include cholinephosphotransferase activity, CDP-choline availability and acetyltransferase activity. Since the synthesis of PAF from a fatty alcohol (i.e. the first catalytic step in Figure 54.3) has never been demonstrated to occur in whole cells, initiation of the de novo pathway could possibly involve the generation of 1-Oalkyl-2-lyso-phosphatidate by lysophospholipase D acting on 1O-alkyl-2-lyso-GPC (Wykle and Schremmer 1974). In any event, it is postulated that the de novo pathway participates in the constitutive and physiological, rather than stimulus-induced, production of PAF. De novo PAF synthesis from fatty alcohol would not involve the sn-2 deacylation of phospholipids and therefore would proceed without the release of AA. Analogous routes exist for mobilizing AA, e.g. a phospholipase C may remove the polar head group from phosphatidylinositols or, alternatively, a phospholipase D and phosphohydrolase may sequentially cleave PC. The product of either route, diacylglycerol, is then acted on by lipases to release AA. Lipase-dependent AA release does not form lyso phospholipids and in consequence does not form PAF precursors. One or another of these pathways reportedly operates in pancreatic islet cells, B lymphocytes, neural tissue, platelets and PMN (Balsinde et al 1991; Garcia et al 1993; Konrad et al 1994). INTERACTIONS OF EICOSANOIDS AND PAF IN STIMULATING CELLS Although there are many examples of eicosanoids interacting with PAF to stimulate cells (Rossi and O’Flaherty 1991), those between 5-HETE and PAF are exceptional. 5-HETE and structural analogues (i.e. its keto metabolite, 5-oxo-ETE and a series of 5-HETEs and 5-oxo-ETEs containing either a hydroxy or keto
residue at carbon 15 or 20) only weakly stimulate PMN but nonetheless increase the PMN-stimulating potency of PAF up to 1000-fold. The compounds have no such impact on C5a, IL-8, LTB4 or formylated oligopeptide chemotactic factors. 5-HETE cooperates with PAF to stimulate responses far downstream from (e.g. degranulation) as well as close to (e.g. inositol trisphosphate formation) the receptor-G protein linkage. Indeed, in preliminary studies, we find that PMN responses to PAF are resistant to pertussis toxin under conditions where responses to 5-HETE are totally blocked and those to other chemotactic factors are partially blocked by the toxin. These sensitivity patterns suggest that PAF receptors act primarily through Gaq or Ga12 G proteins, 5-HETE receptors act almost exclusively through Gai G proteins, and the receptors for other chemotactic factors operate more or less equally through both sets of G proteins. This division in receptor–G protein linkages, which is a common feature of rhodopsin superfamily receptors (Ji et al 1998), immediately suggests three possibilities. First, chemotactic factor receptor may enlist separate sets of G proteins that individually govern quite different and limited aspects of cell excitation but together produce optimal cell activation. Second, receptors for PAF and 5-HETE may represent opposite extremes in this governance. Third, tissues may vary the nature or number of chemotactic factors they express to control not only the recruitment but also the functionality of PMN. The model may also have other important applications. In eosinophils, for example, the strongest 5-HETE analogue, 5-oxo-ETE, is an exceedingly potent chemotactic factor but a weak degranulating agent, while PAF has the opposite pattern of activity. Acting together, however, the two agents exhibit synergy in eliciting especially prominent responses (Powell et al 1995; O’Flaherty et al 1996a). Moreover, both agents are highly active in vivo, acting individually to recruit large numbers of eosinophils to rat lung (Stamatiou et al 1998). 5-OxoETE and PAF, by mobilizing discretely different signal pathways, may cooperate to stimulate eosinophils and thereby interact to modulate allergic or other eosinophil-based toxic reactions. Relevant to this notion, 5-lipoxygenase-deficient mice are
590
CONCLUSIONS AND CORRELATIONS
protected from the lethal effect of PAF (Chen et al 1994) and, furthermore, drugs that block this enzyme attenuate not only human bronchoconstrictive responses to PAF in mild asthma but also reduce the chronic pulmonary dysfunction in mild to moderate asthma (Israel et al 1993; Go`mez et al 1998). The latter effect appears due at least in part to the dampening of eosinophil recruitment, a notion supported by animal studies which find a clear relation between 5-lipoxygenase, eosinophils and allergic reactivity (Irvin et al 1997; Ngase et al 1997). In all of these instances, then, the toxic actions of PAF may require the endogenous formation of one or another agonist of 5-HETE receptors.
CONCLUSIONS The stimulus-induced synthesis of PAF begins with the release of fatty acids, particularly AA, from resident 1-O-alkyl-2-arachidonoyl-GPC. The response is initiated by mobilizing cell CA2+ and MAPK cascades to activate cPLA2. Activated cPLA2 releases AA from internal phospholipid stores in a reaction that, at least in some cells, leads to the secretion of sPLA2. Secreted sPLA2 attacks cell surface phospholipids and may also feed back to activate cPLA2, either by binding to its plasmalemma receptor or by causing formation of G protein-linked receptor agonists (e.g. PAF, LTB4, 5-HETE) or products that have a more direct effect on G proteins (e.g. 1-O-acyl-2-lyso-GPC). The activated G proteins then proceed to mobilize the MAPK–cPLA2 axis. The products of PLA2 cleavage enter metabolic pathways, with AA being converted to eicosanoids and 1-O-alkyl-2-lyso-GPC being acetylated to PAF by acetyltransferase. One MAPK class, p38 kinase, directly regulates the latter enzyme, at least in PMN. CoAIT contributes to eicosanoid and PAF synthesis by transferring AA to PE from other phospholipids. CoA-IT may thereby reload PE fractions with releasable AA while also forming the PAF precursor, 1-O-alkyl-2-lyso-GPC. These relations, i.e. a shared molecular origin, regulation of synthesis by the same enzymes, cross-talk between different synthetic enzymes, and capacity to stimulate each other’s synthesis, ensure that stimulated cells produce eicosanoids and PAF in concert. This may be of particular importance in situations where the products exhibit synergism, e.g. in PMN or eosinophils where PAF and 5-oxo-ETE utilize different signalling cascades and thereby achieve optimal levels of cell stimulation when acting together.
REFERENCES Ali H, Richardson RM, Tomhave ED et al (1994) Regulation of stably transfected platelet activating factor receptor in RBL-2H3 cells. Role of multiple G proteins and receptor phosphorylation. J Biol Chem, 269, 24557–24563. Ali H, Sozzani S, Fisher I et al (1998) Differential regulation of formyl peptide and platelet-activating factor receptors. Role of phospholipase Db3 phosphorylation by protein kinase A. J Biol Chem, 273, 11012– 11016. Amatruda TT III, Gerard NP, Gerard C and Simoff MI (1993) Specific interactions of chemoattractant factor receptors with G-proteins. J Biol Chem, 268, 10139–10144. Baker RR and Chang H-Y (1997) Neuronal nuclear acetyltransferases involved in the synthesis of platelet-activating factor are located in the nuclear envelope and show differential losses in activity. Biochim Biophys Acta, 1345, 197–206. Baker PR, Owen JS, Nixon AB et al (2002) Regulation of plateletactivating factor synthesis in human neutrophils by MAP kinases. Biochim Biophys Acta, 1592, 175–184. Balsinde J, Balboa MA and Dennis EA (1997) Antisense inhibition of group VI Ca2+-independent phospholipase A2 blocks phospholipid
fatty acid remodeling in murine P388D1 macrophages. J Biol Chem, 272, 29317–29321. Balsinde J and Dennis DA (1997) Function and inhibition of intracellular calcium-independent phospholipase A2. J Biol Chem, 272, 16069– 16072. Balsinde J and Dennis EA (1996) Distinct roles in signal transduction for each of the phospholipase A2 enzymes present in P388D1 macrophages. J Biol Chem, 271, 6758–6765. Balsinde J, Diez E and Mollinedo F (1991) Arachidonic acid release from diacylglycerol in human neutrophils. Translocation of diacylglycerol– deacylating enzyme activities from an intracellular pool to plasma membrane upon cell activation. J Biol Chem, 266, 15638–15643. Blank ML, Smith ZL, Fitzgerald V and Snyder F (1995) The CoAindependent transacylase in PAF biosynthesis: tissue distribution and molecular species selectivity. Biochim Biophys Acta, 1254, 295–301. Blank ML, Smith ZL and Snyder F (1992) Contributing factors in the trafficking of [3H]arachidonate between phospholipids. Biochim Biophys Acta, 1124, 262–272. Breton M and Colard O (1991) Protein kinase C promotes arachidonate mobilization through enhancement of CoA-independent transacylase activity in platelets. Biochem J, 280, 93–98. Cao Y, Traer E, Zimmerman GA et al (1998) Cloning, expression, and chromosomal localization of human long-chain fatty acid-CoA ligase 4 (FACL4). Genomics, 49, 327–330. Chen K-S, Sheller JR, Johnson EN and Funk CD (1994) Role of leukotrienes revealed by targeted disruption of the 5-lipoxygenase gene. Nature, 372, 179–182. Chilton FH and Connell TR (1988) 1-ether-linked phosphoglycerides. Major endogenous sources of arachidonate in the human neutrophil. J Biol Chem, 263, 5260–5265. Chilton FH, Fonteh AN, Sung C-M et al (1995) Inhibitors of CoAindependent transacylase block the movement of arachidonate into 1ether-linked phospholipids of human neutrophils. Biochemistry, 34, 5403–5410. Choy PC, Skrzypczak M, Lee D and Jay FT (1997) Acyl-GPC and alkenyl/alkyl-GPC:acyl-CoA acyltransferases. Biochim Biophys Acta, 1348, 124–133. Colard O, Bidault J, Breton M and Ninio E (1993) Biosynthesis of plateletactivating factor in cultured mast cells. Involvement of the CoAindependent transacylase demonstrated by analysis of the molecular species of platelet-activating factor. Eur J Biochem, 216, 835–840. Crespo MS, Alonso A, Bayon Y and Garcia Rodriguez MC (1996) Platelet-activating factor is an effector of rapid reactions and an inductor of late responses in immune-mediated injury. In Nigam S (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 195–199. Cupillard L, Koumanov K, Matte´i M-G et al (1997) Cloning, chromosomal mapping, and expression of a novel human secretory phospholipase A2. J Biol Chem, 272, 15745–15752. Diez E, Chilton FH, Stroup G et al (1994) Fatty acid and phospholipid selectivity of different phospholipase A2 enzymes studied by using a mammalian membrane as substrate. Biochem J, 301, 721–726. Garcia C, Montero M, Alvarez J and Sanchez Crespo M (1993) Biosynthesis of platelet-activating factor (PAF) induced by chemotactic peptide is modulated at the lyso-PAF:acetyl-CoA acetyltransferase level by calcium transient and phosphatidic acid. J Biol Chem, 268, 4001–4008. Go`mez FP, Iglesia R, Roca J et al (1998) The effects of 5-lipoxygenase inhibition by zileuton on platelet-activating-factor-induced pulmonary abnormalities in mild asthma. Am J Resp Crit Care Med, 157, 1559– 1564. Heery JM, Kozak M, Stafforini DM et al (1995) Oxidatively modified LDL contains phospholipids with platelet-activating factor-like activity and stimulates the growth of smooth muscle cells. J Clin Invest, 96, 2322–2330. Ho YS, Swenson L, Derewenda U et al (1997) Brain acetylhydrolase that inactivates platelet-activating factor is a G-protein-like trimer. Nature, 385, 89–93. Honda Z-I, Nakamura M, Miki I et al (1991) Cloning by functional expression of platelet-activating factor receptor from guinea-pig lung. Nature, 349, 342–346. Huwiler A, Staudt G, Kramer RM and Pfeilschifter J (1997) Cross-talk between secretory phospholipase A2 and cytosolic phospholipase A2 in rat renal mesangial cells. Biochim Biophys Acta, 1348, 257–272.
INTERACTIONS OF PLATELET-ACTIVATING FACTOR WITH EICOSANOIDS Irvin CG, Tu Y-P, Sheller JR and Funk CD (1997) 5-lipoxygenase products are necessary for ovalbumin-induced airway responsiveness in mice. Am J Physiol, 272, L1053–L1058. Ishii S, Nagase T, Tashiro F et al (1997) Bronchial hyperreactivity, increased endotoxin lethality and melanocytic tumorigenesis in transgenic mice overexpressing platelet-activating factor receptor. EMBO J, 16, 133–142. Israel E, Rubin P, Kemp JP et al (1993) The effect of inhibition of 5lipoxygenase by zileuton in mild-to-moderate asthma. Ann Intern Med, 119, 1059–1066. Ji TH, Grossmann M and Ji I (1998) G Protein-coupled receptors. I. Diversity of receptor–ligand interactions. J Biol Chem, 273, 17299– 17302. Konrad RJ, Major CD and Wolf BA (1994) Diacylglycerol hydrolysis to arachidonic acid is necessary for insulin secretion from isolated pancreatic islets: sequential actions of diacylglycerol and monoacylglycerol lipases. Biochemistry, 33, 13284–13294. Kramer RM and Deykin D (1983) Arachidonoyl transacylase in human platelets. Co-enzyme A-independent transfer of arachidonate from phosphatidylcholine to lysoplasmenylethanolamine. J Biol Chem, 258, 13806–13811. Kramer RM, Roberts EF, Hyslop PA, et al (1995) Differential activation of cytosolic phospholipase A2 (cPLA2) by thrombin and thrombin receptor agonist peptide in human platelets. Evidence for activation of cPLA2 independent of the mitogen-activated protein kinases ERK 1/2. J Biol Chem, 270, 14816–14823. Kramer RM, Roberts EF, Um SL et al (1996) p38 mitogen-activated protein kinase phosphorylates cytosolic phospholipase A2 (cPLA2) in thrombin-stimulated platelets. Evidence that proline-directed phosphorylation is not required for mobilization of arachidonic acid by cPLA2. J Biol Chem, 271, 27723–27729. Lambeau G, Ancian P, Nicolas JP et al (1996) A family of receptors for secretory phospholipases A2. Compt Rend Se´anc Soc Biol Fil, 190, 425– 435. Lee T-C, Malone B and Snyder F (1986) A new de novo pathway for the formation of 1-alkyl-2-acetyl-sn-glycerols, precursors of platelet activating factor. Biochemical characterization of 1-alkyl-2-lyso-snglycero-3-P:acetyl-CoA acetyltransferase in rat spleen. J Biol Chem, 261, 5373–5377. Leslie CC (1997) Properties and regulation of cytosolic phospholipase A2. J Biol Chem, 272, 16709–16712. Lopes-Martins RAB, Arau´jo CV, Estato V et al (1996) Platelet-activating factor. A possible role in the modulation of the vasomotor tone and blood pressure. In Nigam S (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 223–230. MacDonald JIS and Sprecher H (1991) Phospholipid fatty acid remodeling in mammalian cells. Biochem Biophys, 1084, 105–121. Manya H, Aoki J, Watanabe M, et al (1998) Switching of plateletactivating factor acetylhydrolase catalytic subunits in developing rat brain. J Biol Chem, 273, 18567–18572. Marshall LA, Bolognese B, Winkler JD and Roshak A (1997) Depletion of human monocyte 85 kDa phospholipase A2 does not alter leukotriene formation. J Biol Chem, 272, 759–765. Matsuzawa A, Hattori K, Aoki J et al (1997) Protection against oxidative stress-induced cell death by intracellular platelet-activating factoracetylhydrolase II. J Biol Chem, 272, 32315–32320. Morooka S and Natsume Y (1996) Effect of SM-12502 on disseminated intravascular coagulation (DIC) in tumor-bearing rats. In Nigam S (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 189–193. Mosior M, Six DA and Dennis EA (1998) Group IV cytosolic phospholipase A2 binds with high affinity and specificity to phosphatidylinositol 4,5-bisphosphate resulting in dramatic increases in activity. J Biol Chem, 273, 2184–2191. Muguruma K, Furukawa M, Tjoelker LW et al (1996) The prevention of necrotizing enterocolitis. The role of platelet-activating factor acetylhydrolase. In Nigam S (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 129–133. Murakami M, Shimbara S, Kambe T et al (1998) The functions of five distinct mammalian phospholipase A2s in regulating arachidonic acid release. Type IIA and type V secretory phospholipase A2s are functionally redundant and act in concert with cytosolic phospholipase A2. J Biol Chem, 273, 14411–14423.
591
Mutoh H, Bito H, Minami M et al (1993) Two different promoters direct expression of two distinct forms of mRNAs of human plateletactivating factor receptor. FEBS Lett, 322, 129–134. Ngase T, Ishii S, Katayama H et al (1997) Airway responsiveness in transgenic mice overexpressing platelet-activating factor receptor. Roles of thromboxanes and leukotrienes. Am J Resp Crit Care Med, 156, 1621–1627. Nieto ML, Venable ME, Bauldry SA et al (1991) Evidence that hydrolysis of ethanolamine plasmalogens triggers synthesis of platelet-activating factor via a transacylation reaction. J Biol Chem, 266, 18699–18706. Ninio E and Bessou G (1991) Biosynthesis of PAF-acether. XVI. Acetyltransferase specificity determines composition of PAF-acether molecular species in human polymorphonuclear neutrophils. Biochim Biophys Acta, 1085, 136–139. Nixon AB, Greene DG and Wykle RL (1996) Comparison of acceptor and donor substrates in the CoA-independent transacylase reaction in human neutrophils. Biochem Biophys Res Commun, 1300, 187–196. Nixon AB, O’Flaherty JT, Salyer JK and Wykle RL (1999b) AcetylCoA:1-O-alkyl-2-lyso-sn-glycero-3-phosphocholine acetyltransferase is directly activated by p38 kinase. J Biol Chem. O’Flaherty JT, Jacobson D and Redman J (1988) Mechanism involved in the mobilization of neutrophil calcium by 5-hydroxyeicosatetraenoate. J Immunol, 140, 4323–4328. O’Flaherty JT, Kuroki M, Nixon AB et al (1996a) 5-oxo-Eicosanoids and hematopoietic cytokines cooperate in stimulating neutrophil function and the mitogen-activated protein kinase pathway. J Biol Chem, 271, 17821–17828. O’Flaherty JT, Kuroki M, Nixon AB et al (1996b) 5-oxo-Eicosatetraenoate is a broadly active, eosinophil-selective stimulus for human granulocytes. J Immunol, 157, 336–342. O’Flaherty JT, Taylor JS and Thomas MJ (1998) Receptors for the 5-oxo class of eicosanoids in neutrophils. J Biol Chem, 273, 32535–32541. Perry SW, Hamilton JA, Tjoelker LW et al (1998) Platelet-activating factor receptor activation. An initiator step in HIV-1 neuropathogenesis. J Biol Chem, 273, 17660–17664. Peterfy M, Gyuris T, Grosshans D, et al (1998) Cloning and characterization of cDNAs and the gene encoding the mouse platelet-activating factor acetylhydrolase 1b a subunit/lissencephaly-1 protein. Genomics, 47, 200–206. Powell WS, Chung D and Gravel S (1995) 5-oxo-6,8,11,14Eicosatetraenoic acid is a potent stimulator of human eosinophil migration. J Immunol, 154, 4123–4132. Ramesha CS and Pickett WC (1986) Platelet-activating factor and leukotriene biosynthesis is inhibited in polymorphonuclear leukocytes depleted of arachidonic acid. J Biol Chem, 261, 7592–7595. Roberts MF (1996) Phospholipases: structural and functional motifs for working at an interface. FASEB J, 10, 1159–1172. Rodriguez CG, Montero M, Alvarez J et al (1993) Dissociation of plateletactivating factor production and arachidonate release by the endomembrane CA2+-ATPase inhibitor thapsigargin. Evidence for the involvement of a CA2+-dependent route of priming in the production of lipid mediators by human polymorphonuclear leukocytes. J Biol Chem, 268, 24751–24757. Rossi AG and O’Flaherty JT (1991) Bioactions of 5-hydroxyicosatetraenoate and its interaction with platelet-activating factor. Lipids, 26, 1184–1188. Si S, Liu H, Chao W and Olson MS (1997) Regulation of plateletactivating factor (PAF) biosynthesis via co-enzyme A-independent transacylase in the macrophage cell line ICI-21 stimulated with lipopolysaccharide. Biochim Biophys Acta, 1346, 120–130. Silver D, Dhar A, Slocum M et al (1996) Role of platelet-activating factor in skeletal muscle ischemia-reperfusion injury. In S Nigam (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 217–221. Smith WL (1997) Molecular biology of prostanoid biosynthetic enzymes and receptors. In Honn KV, Nigam S and Marnett LJ (eds), Eicosanoids and Other Bioactive Lipids in Cancer, Inflammation, and Radiation Injury 2. New York: Plenum, 989–1011. Snyder F (1995) Platelet-activating factor: the biosynthetic and catabolic enzymes. Biochem J, 305, 689–705. Stafforini DM, McIntyre TM, Zimmerman GA and Prescott SM (1997) Platelet-activating factor acetylhydrolases. J Biol Chem, 272, 17895– 17898.
592
CONCLUSIONS AND CORRELATIONS
Stamatiou P, Hamid Q, Taha R et al (1998) 5-oxo-ETE induces pulmonary eosinophilia in an integrin-dependent manner in brown Norway rats. J Clin Invest, 102, 2165–2172. Suga K, Kawasaki T, Blank ML and Snyder F (1990) An arachidonoyl (polyenoic)-specific phospholipase A2 activity regulates the synthesis of platelet-activating factor in granulocytic HL-60 cells. J Biol Chem, 265, 12363–12371. Sugiura T, Kudo N, Ojima T et al (1995) Co-enzyme A-dependent cleavage of membrane phospholipids in several rat tissues: ATPindependent acyl-CoA synthesis and the generation of lysophospholipids. Biochim Biophys Acta, 1255, 167–176. Sugiura T, Masuzawa Y, Nakagawa Y and Waku K (1987) Transacylation of lyso platelet-activating factor and other lysophospholipids by macrophage microsomes. Distinct donor and acceptor selectivities. J Biol Chem, 262, 1199–1205. Surette ME and Chilton FH (1998) The distribution and metabolism of arachidonate-containing phospholipids in cellular nuclei. Biochem J, 330, 915–921. Svetlov SI, Liu H, Chao W and Olson MS (1997) Regulation of plateletactivating factor (PAF) biosynthesis via co-enzyme A-independent transacylase in the macrophage cell line IC-21 stimulated with lipopolysaccharide. Biochim Biophys Acta, 1346, 120–130. Tamura G, Takishima T, Mue S et al (1996) Effect of a potent plateletactivating factor antagonist, WEB-2086, on asthma. A multicenter, double-blind placebo-controlled study in Japan. In Nigam S (ed.), Platelet-activating Factor and Related Lipid Mediators 2. New York: Plenum, 371–380. Tischfield JA (1997) A reassessment of the low molecular weight phospholipase A2 gene family in mammals. J Biol Chem, 272, 17247– 17250. Tjoelker LW, Wilder C, Eberhardt C, et al (1995) Anti-inflammatory properties of a platelet-activating factor acetylhydrolase. Nature, 374, 549–502. Toyoshima K, Narahara H, Frenkel RA and Johnston JM (1994) Coenzyme A-independent transacylation in amnion-derived (WISH) cells. Arch Biochem Biophys, 314, 224–228. Uemura Y, Lee T-C and Snyder F (1991) A co-enzyme A-independent transacylase is linked to the formation of platelet-activating factor (PAF) by generating the lyso-PAF intermediate in the remodeling pathway. J Biol Chem, 266, 8268–8272. Uozumi N, Kume K, Nagase T et al (1997) Role of cytosolic phospholipase A2 in allergic response and parturition. Nature, 390, 618–622. Van Lenten BJ, Hama SY, de Beer FC et al (1995) Antiinflammatory HDL becomes proinflammatory during the acute phase response. J Clin Invest, 96, 2758–2767. Venable ME, Nieto ML, Schmitt JD and Wykle RL (1991) Conversion of 1-O-[3H]alkyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine to lyso platelet-activating factor by the CoA-independent transacylase in membrane fractions of human neutrophils. J Biol Chem, 266, 18691– 18698.
Venable ME, Olson SC, Nieto ML and Wykle RL (1993) Enzymatic studies of lyso platelet-activating factor acylation in human neutrophils and changes upon stimulation. J Biol Chem, 268, 7965– 7975. Wijkander J, O’Flaherty JT, Nixon AB and Wykle RL (1995) 5Lipoxygenase products modulate the activity of the 85 kDa phospholipase A2 in human neutrophils. J Biol Chem, 270, 26543– 26549. Winkler JD, Bolognese BJ, Roshak AK et al (1997) Evidence that 85 kDa phospholipase A2 is not linked to CoA-independent transacylasemediated production of platelet-activating factor in human monocytes. Biochim Biophys Acta, 1346, 173–184. Winkler JD, Fonteh AN, Sung C-M et al (1995) Effects of CoAindependent transacylase inhibitors on the production of lipid inflammatory mediators. J Pharmacol Exp Therapeut, 274, 1338–1347. Winkler JD, Sung C-M, Bennett CF and Chilton FH (1991) Characterization of CoA-independent transacylase activity in U937 cells. Biochim Biophys Acta, 1081, 339–346. Winkler JD, Sung C-M, Chabot-Fletcher M et al (1998) b-Lactams SB 212047 and SB 216754 are irreversible, time-dependent inhibitors of co-enzyme A-independent transacylase. Mol Pharmacol, 53, 322–329. Winkler JD, Sung C-M, Huang L and Chilton FH (1994) CoAindependent transacylase activity is increased in human neutrophils after treatment with tumor necrosis factor a. Biochim Biophys Acta, 1215, 133–140. Wykle RL, Malone B and Snyder F (1980) Enzymatic synthesis of 1-alkyl2-acetyl-sn-glycero-3-phosphocholine, a hypotensive and plateletaggregating lipid. J Biol Chem, 255, 10256–10260. Wykle RL and Schremmer JM (1974) A lysophospholipase D pathway in the metabolism of ether-linked lipids in brain microsomes. J Biol Chem, 249, 1742–1746. Yamashita A, Sugiura T and Waku K (1997) Acyltransferases and transacylases involved in fatty acid remodeling of phospholipids and metabolism of bioactive lipids in mammalian cells. J Biochem, 122, 1– 16. Yokomizo T, Izumi T, Chang K et al (1997) A G-protein-coupled receptor for leukotriene B4 that mediates chemotaxis. Nature, 387, 620–624. Yuan Y, Schoenwaelder SM, Salem HH and Jackson SP (1996) The bioactive phospholipid, lysophosphatidylcholine, induces cellular effects via G-protein-dependent activation of adenylyl cyclase. J Biol Chem, 271, 27090–27098. Zhang C, Baumgartner RA, Yamada K and Beaven MA (1997) Mitogenactivated protein (MAP) kinase regulates production of tumor necrosis factor-a and release of arachidonic acid in mast cells. Indications of communication between p38 and p42 MAP kinases. J Biol Chem, 272, 13397–13402. Zhang Y, Styhler A and Powell WS (1996) Synthesis of 5-oxo-6,8,11,14eicosatetraenoic acid by human monocytes and lymphocytes. J Leukocyte Biol, 59, 847–854.
55 Eicosanoid Precursors as Pharmaceuticals The Late David F. Horrobin
INTRODUCTION The idea that nutrients, hormones and intermediates in human biochemical pathways might have therapeutic actions had its heyday in the first half of the twentieth century. Deficits of thyroid, adrenal, gonadal and pituitary function which had earlier all been recognized as diseases with dramatic and characteristic groups of symptoms were broadly understood and became treatable by replacement therapy. Similarly, diseases like beri beri, pellagra and rickets were recognized as caused by deficits of specific nutrients which were co-factors in biochemical reactions and became easily and quickly treatable. By 1950 it was thought that diseases relating to essential nutrients were all identified and understood. At about that time the University of Oxford rejected a large donation offered to establish a human nutrition research centre on the grounds that the science of human nutrition was largely complete (Ewin 2001). What was reasonably well understood was the impact of essential nutrients whose body stores were modest and deficits of which would lead to unequivocal symptoms within a matter of weeks or months. The possibility that there might be essential nutrient disorders or deficits which would produce diseases only over many decades, or diseases which might be marked by a slow disintegration of the biological system rather than a clear syndrome of acute onset, had not been developed. This was in part, of course, because only relatively small numbers of people lived beyond their 60s or 70s and understanding of degenerative disease processes was in its infancy. But Robert McCarrison had developed his ideas, imperfect in retrospect but remarkably prescient at the time, to the effect that nutrition in childhood and over a lifetime was the key to health and freedom from degenerative disorders. One of the people inspired by McCarrison was Hugh Sinclair, who was one of the first to begin to understand the implications of essential fatty acids for human health (Ewin 2001; Sinclair 1956). He recognized that a nutrient like linoleic acid, which was stored in the body in large amounts and which might have to be metabolized to perform many of its functions, might lead to a pattern of disorder quite different from that seen with most nutritional and metabolic diseases. In a much derided letter to the Lancet in 1956 (Sinclair 1956), he outlined his idea that defects of essential fatty acid biochemistry might be responsible for many of the chronic disorders of humankind, from cardiovascular disease, to arthritis, to respiratory disease, to neurological disease and to cancer. He happened to be about 50 years ahead of his time! Only now is it being recognized that many of the diseases he speculated might be related to EFA metabolism are indeed due, or at least partly due, to the sorts of problems Sinclair envisaged. Twenty years ago it The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
began to be appreciated that Sinclair might indeed have been on to something (Horrobin 1982, 1983, 1990, 1992). Only now is that ‘‘something’’ being realised in specific treatments for specific diseases.
EFA METABOLISM AND PHARMACEUTICAL CANDIDATES There are two series of essential fatty acids, n-3 (o-3) and n-6 (o-6). The pathways of metabolism of the main dietary precursors are summarized in Figure 55.1. Both linoleic acid (LA), the parent compound of the n-6 series, and a-linolenic acid (ALA), the parent compound of the n-3 series, are converted to their ultimate metabolites by an alternating series of desaturations, in which a new cis double bond is inserted into the carbon chain, and elongations in which two further carbon atoms are added to the chain. Many investigators have realized that, whereas the elongation steps tend to be rapid and non-rate-limiting, the desaturation steps are slow and regulated by many different factors (Brenner 1981; Horrobin 1992). As examples, insulin, zinc and magnesium promote desaturation, while the stress-related hormones adrenaline and cortisol inhibit it (Brenner 1981; Horrobin 1992). The fatty acids compete with each other, especially for desaturation. In particular, administration of the n-3 fatty acids is able to inhibit both the first D-6-desaturation (D6D) and the second D-5desaturation (D5D) steps (Nassar et al 1986a, 1986b, 1987; Horrobin 1983, 1992). As a result there are many interesting therapeutic possibilities raised by the idea that particular members of the pathway may be usefully administered, often with other members of the pathway to control their metabolism. For example, if it is desired to raise the concentration of dihomog-linolenic acid (DGLA), which seems to have many desirable actions, this can be achieved by administering DGLA or its immediate precursor, g-linolenic acid (GLA), together with the o-3 acid (EPA) to block the D5D conversion of DGLA to the potentially harmful arachidonic acid (AA) (Nassar et al 1986a, 1986b, 1987).
Fatty Acids as Drug Candidates When reviewing the EFAs chain in Figure 55.1, there are various criteria which may be used to identify potential drug candidates. Some of them may appear mundane but are crucial to practical use as a pharmaceutical.
594
CONCLUSIONS AND CORRELATIONS
Figure 55.1 An outline of the pathways of metabolism of o-6 and o-3 fatty acids showing the key molecules likely to be of therapeutic interest
1. Is it present in substantial (more than 1 g/day) amounts in most diets? If so, it is unlikely to be useful as a therapeutic agent because only very large doses are likely to have much impact on pathophysiology. Linoleic acid (LA) and a-linolenic acid (ALA) can be rejected as systemic pharmaceuticals for this reason. They have low biological activity at the sorts of doses which are realistic for pharmaceuticals, primarily because to exert their biological effects they must be metabolized to other substances. 2. Can they be synthesized or otherwise manufactured in the correct isomeric form? All EFAs must have all their double bonds in the cis configuration if they are to be physiologically active. The interesting EFAs have three to six double bonds and hence three to six chiral centres. Complete chemical syntheses are known but they all lead to variable amounts of trans isomers at each double bond. Since the trans isomers may be competitive inhibitors of the all-cis compounds, the synthesized products can have wildly ranging biological activity, so making them unsuitable as pharmaceuticals. Without a great improvement in the ability to synthesize pure chiral compounds with all of the double bonds in the cis configuration, it is unlikely that full chemical synthesis will be a practical proposition. This therefore means that only those fatty acids which are available in large amounts from living organisms—which unerringly make the all-cis compounds— are likely to be useful drugs. The main biological sources which are viable are plants for g-linolenic acid (GLA) and stearidonic acid (SA), and algae and fungi for arachidonic acid (AA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). EPA and DHA may also be prepared from oils from marine mammals, fish and shellfish, which accumulate these fatty acids primarily via the food chain rather than by making them themselves. All of these fatty acids may also be made by genetic engineering of oilseed plants, such as soy or rape. For the foreseeable future, therefore, the main drug candidates are GLA and AA of the n-6 series and SA, EPA or DHA of the n-3 series. They may, of course, be converted into many different derivatives which may have different pharmacokinetic and pharmacological properties, such as salts, ethyl esters, amides, diols, triglycerides or phospholipids.
3. What fatty acids have useful and important biological actions at reasonable dose levels? This, of course, is the key question assuming that the first two criteria have been fulfilled. A drug must have a useful therapeutic effect, it must achieve that therapeutic effect without too many or too serious adverse effects, so leading to a favourable risk:benefit ratio, and preferably it should achieve that effect at a dose of under 2 g/ day and certainly under 4 g/day. Higher doses may be acceptable for life-threatening conditions but for most illnesses most patients are unlikely to be willing to take high doses for any length of time. The next sections consider the possible mechanisms of action of the five candidate EFAs and the possible advantages and disadvantages of using them. POTENTIAL MECHANISMS OF ACTIONS Many potential mechanisms of action are discussed in other chapters of this book and only the most important ones are summarized here. Actions of EFAs Themselves Some of the potential effects are dependent upon the fatty acids and some on their metabolites. Until very recently, the former action has been regarded as of little importance, an attitude which is beginning to change.
Structure as Influencing Function The fatty acids can be incorporated into many types of complex lipid, including phospholipids, ether lipids, cholesteryl esters and triglyceride. Many and perhaps most proteins are now known to be embedded in, attached to or associated with such complex lipids. Even quite small variations in the lipid environment, such as the insertion of a single additional double bond into a fatty
EICOSANOID PRECURSORS AS PHARMACEUTICALS acid, can produce large changes in the quaternary structure of lipid-associated proteins. Such changes in quaternary structure can have important functional consequences, such as changes in enzyme activity or in receptor function (Witt and Nielsen 1994). Because of the very large numbers of proteins whose function can be modulated in this way, it is clear that it may be very difficult to predict the overall effects of an intervention with any fatty acid. The fatty acids which are most abundant in structural lipids are AA and DHA, and so it is they that are most likely to influence function by modifying structure. GLA and DGLA may be converted to AA, and SA and EPA may be converted to DHA, even though these reactions may be relatively rate-limited in humans. One indirect consequence of this phenomenon, which seems to be understood by only very few scientists, relates to cell culture. For normal integrated function, cells require structured lipids of a particular composition. However, cultured cells have structural lipid compositions which are very different from those of the same cell in a normal animal or human (Reynolds et al 2001). These differences may lead to large numbers of subtle but important differences between protein quaternary structures in cultured cells and in their in vivo counterparts. This may lead to defective understanding of the in vivo situation. It may also mean that cultured cells are limited in their value as screening systems in assays for proteins which must be in a precisely accurate quaternary configuration to function normally. Fatty Acids Themselves as Biologically Active Agents It used to be thought that fatty acids were largely inert molecules which had biological activity largely as a result of their conversion to other molecules, such as prostaglandins or leukotrienes. This can now be seen to be untrue. Fatty acids, and particularly the EFAs, have a wide range of biological actions in themselves. There are binding sites for fatty acids on many enzymes and proteins which allow the fatty acid to change the activity of that enzyme or protein. For example, many ion channels are now known to have specific fatty acid binding sites (Kang and Leaf 1996), while the synaptic regulating protein GAP-43 is switched from an inactive to an active configuration by binding to arachidonic acid (Benowitz and Routtenberg 1997). Enzymes known to be regulated by fatty acids include phospholipases, protein kinases, fatty acid desaturases and lipoxygenases, and cyclooxygenases. For example, EPA inhibits phospholipase A2 (PLA2) and D-5-desaturases (D5D), and also competitively reduces the metabolism of arachidonic acid by cyclooxygenases and lipoxygenases, and inhibits caspases such as caspase 1 or caspase 3 (Martin et al 2002). Fatty acids can also be bound to many other proteins, such as fatty acid binding and transport proteins, some of which, like CD36, have other functions, or apolipoproteins like ApoD and ApoE or albumin. In some cases the binding is simply related to transport, but in others it has an influence on function also. Over the past decade, two important related classes of receptors have also been found that have unsaturated fatty acid as their natural ligands. One class is the peroxisome proliferator activated receptors (PPARs), of which there are three main classes, a, g and d (Xu et al 1999; Forman et al 1997). The other is the class of receptors activated by retinoids, which only function as heterodimers with PPAR when one component is bound to a retinoid and the other to an unsaturated fatty acid (Schoonjans et al 1996). Both the PPARs and the retinoid receptors have an everexpanding range of roles in regulating genes and many different aspects of metabolism. Finally, fatty acids themselves may have many different effects on genome and proteome function by regulating gene expression, transcription and translation (Clarke 2000, 2001; Kliewer et al
595
1997). For example, they can regulate the production of a range of cytokines and also of many cell cycle-regulating factors, such as cyclin D and cyclin E (Palakurthi et al 2001; Chiu et al 2001). At least some of these effects seem not to be related to the other effects of the fatty acids noted above, e.g. the effect on cyclin D seems to involve the modulation of calcium ion movements (Palakurthi et al 2001). Fatty Acids as Precursors of Other Biologically Active Molecules Much of the rest of this book is devoted to describing the almost infinite number of highly active molecules which can be formed from EFAs and in particular from DGLA, AA and EPA, as a result of the activity of an ever-expanding range of cyclooxygenases, lipoxygenases, P450 and other enzyme systems. The numbers of biologically active compounds which are formed are bewildering but leave little doubt that EFAs are molecules which influence many different biological systems. ADVANTAGES AND DISADVANTAGES OF USING EFAS AS DRUGS The main advantage of using EFAs as drugs is that they are all normal nutrients or intermediates in human metabolism. This means that they cannot be inherently toxic at any reasonable dose level. It does not mean that there can be no adverse effects, because a particularly high dose of a single fatty acid could theoretically lead to an unbalanced excess of one of the activities described in the past section. In practice this does not seem to occur, at least as far as acute and serious toxicity is concerned. Even AA, which on theoretical grounds, because of its potential conversion to proinflammatory cyclooxygenase and lipoxygenase metabolites, ought to be toxic but has in practice, when given to normal individuals at high dose, shown no evidence of any important adverse effects (Nelson et al 1997). This lack of any serious toxicity is, of course, a major benefit when considering the possible use of an EFA as a drug. It means that the human body has highly efficient mechanisms for dealing with any excess, either as a result of storage in adipose tissue or in phospholipid membranes, or as a result of the efficient oxidation or other metabolic routes of disposal of a compound with which the body is evolutionarily adapted to deal. On the other hand, this lack of apparent toxicity has led biomedical scientists and physicians to ignore almost completely another aspect of drug action, namely the dose–response relationship. Pharmacology texts are dominated by simplistic discussion of dose–response as though this were almost always some form of relatively simple linear, log-linear, sigmoid or other relationship, which might be summed up as ‘‘more drug=more effect’’ until a plateau is reached. Such dose–response relationships depend on situations where a drug acts on a single binding site and in which there are no feedback mechanisms. Simplistic notions like these have been fed by reductionist efforts to get more and more simple systems, ideally ones in which the target protein is isolated in pure form: simple dose–response relationships are then seen because no others are possible. But most biological systems and most drug actions are not like that. Even simple drugs in simple systems often have complex dose–response relationships which usually depend on the presence of more than one receptor or other binding site, and with different affinities for the different binding locations (Horrobin 1977; Calabrese 2001). Thus, the effect of binding to an ultra-high-affinity site may kick in first, leading to a particular effect. At higher concentrations, as binding sites with lower affinities are activated, so there may be a second peak in the
596
CONCLUSIONS AND CORRELATIONS
first type of response, or a quite different response may kick in, or the original response may be reversed. This last effect is exceedingly common, possibly due to the wide occurrence of negative feedback control systems. As a result there are frequently bell-shaped dose–response curves which are particularly common in the eicosanoid field (Horrobin 1977; Calabrese 2001). The most biologically active fatty acids, such as GLA, DGLA, AA, SA, EPA and DHA, have multiple actions, often at least 10 main ones and sometimes many more. These multiple actions do not all occur at the same concentrations but may kick in at different points on a dose–response curve which may range from as low as about 10713 molar to as high as 1073 molar. As a result, it is likely that in whole biological systems the dose–response relationships for the biological actions of fatty acids will be exceedingly complex. They are highly unlikely to be simple monotonic affairs. They may well be bell-shaped or even more complex, showing several peaks and troughs as concentrations or doses are progressively raised. Most investigators looking at fatty acid effects seem to be completely unaware of this concept. Even if theoretically aware, they do not incorporate it into their studies. Almost everyone seems to operate on the principle that more is better and therefore that fatty acid effects can be understood by looking at what happens when a single very high dose is administered. As a result, much of the literature is worthless as far as understanding the biological effects and therapeutic potential of fatty acids is concerned. Another major problem of most studies is that they use complex fatty acid mixtures, such as vegetable oils, fish oils or relatively crude lipid extracts, and then report the results as though the effects were all consequences of a single fatty acid component which is the compound the investigator happens to be most interested in. There seems to be almost no understanding of the fact that most fatty acids are biologically active to at least some degree, that each fatty acid component of a mixture will have a range of actions which kick in at different concentrations and doses, and that different fatty acids may synergize, inhibit, add to or subtract from the effects of other fatty acids. The potential complications and problems of interpretation are almost infinite: just because, say, one DHA-containing oil mix has an effect at one particular high dose, it does not mean that a different DHAcontaining oil mix given at the same or a different dose will have similar actions. Yet the literature is full of such extrapolations and misinterpretations and, as a result, many papers, even though produced with much effort, are close to valueless. Progress will only be made when single pure fatty acids are studied at a carefully chosen range of doses and when these single fatty acid studies are followed by carefully designed interaction studies. So long as most people in the field pursue the simplistic ‘‘more-is-better, detailed fatty acid composition of the mix does not matter’’ approach, progress will be slow or non-existent. This is a tragedy, because there is emerging evidence that specific fatty acids can be safe and highly effective treatments for important illnesses. Studies of Fish Oils at Different Doses The possibility of different effects at different doses and with different impurified or partially purified preparations is well illustrated by the cardiovascular studies on fish oils. Studies using relatively low doses have generated strongly beneficial outcomes on various end-points (Burr et al 1989; Gruppo Italiano per lo Studio della Streptochinasi nell’Infarto Miocardico (GISSI) 1999). In contrast, randomized controlled studies with large doses of o-3 fatty acids have shown no differences between active and placebo, but sometimes with the active group tending to do slightly worse (Leaf et al 1994; Johansen et al 1999). It is uncertain whether these confusing outcomes arise because the precise fatty
acid compositions of the products used in the low dose and high dose studies were not quite the same, or whether the higher doses were ineffective because of a dose–response relationship or a mixture of both. One important issue here relates to the impact of one group of fatty acids on the other. What is clear is that both groups of EFAs are important for the normal function of most organs, and that the long chain n-6 EFAs play central roles in a very wide range of signal transduction processes (Horrobin 1983, 1995; Horrobin et al 2002). If one had to choose, one would have to conclude that the n-6 EFAs are actually even more important than the n-3 EFAs. It is very easy to produce a clinically apparent EFA deficiency state by depriving an animal of n-6 EFAs, but deficiencies of n-3 EFAs produce clear results only after several generations. Because recently n-3 EFAs have tended to be at relatively low levels in the Western diet, most recent concerns have related to possible adverse effects of n-6 EFAs, or of a high n-6:n-3 ratio (Simopoulos et al 2000; Okuyama et al 1997). Most of these concerns are theoretical and are contradicted by the actual data, which show that long chain EFAs of both groups are absolutely required for human health. One important problem that may arise is that high doses of n-3 EFAs may actually deplete cells of AA. While this has often been portrayed as desirable, the actual clinical or experimental evidence from humans that this is so is zero. In contrast, it may well be possible that the repeated failures of clinical trials that have administered very high doses of n-3 EFAs have been caused because these high doses, but not the successful lower doses, actually cause AA depletion. Another issue is the differentiation between different fatty acids of the same class, notably EPA and DHA. Because they are at least slowly interconvertible, there has been a tendency to treat them as equivalent. This has had two consequences: first, in clinical studies little attention has been paid to the precise composition of the n-3 oils, even when partially purified: the possibility that different EPA:DHA ratios might have different effects has rarely been considered. Second, it has been assumed that effects produced by one will be the same as or closely similar to effects produced by the other. These are dangerous assumptions, as shown in important studies by Peet, by Edwards, by Finnen and by Tisdale. Peet, in a randomized study compared the effects in schizophrenia of an EPA-enriched oil, a DHA-enriched oil and a placebo oil. The effects of DHA were not different from placebo, whereas EPA produced a strong beneficial effect (Peet et al 2001). Edwards obtained a similar result in depression (R Edwards, personal communication). Finnen and Lovell (1991) showed that EPA but not DHA could inhibit a phospholipase A2. Tisdale and colleagues (Price and Tisdale 1998) showed that EPA could block cancer cachexia induced by over-production of a zinc a-2 macroglobulin. DHA had a much weaker effect than EPA, possibly dependent upon retroconversion to EPA. These experiences indicate that there will be no short cuts to the development of EFA-based drugs. Dose–response studies will be extremely important in order to define the optimum dose. Such dose–response studies should be conducted on a pilot scale prior to any commitment to large-scale trials. Single pure compounds will be required in order to define biological activity clearly, to guarantee that the effects reported can be consistently achieved in other trials, and to allow the proper study of interactions between fatty acids. INDIVIDUAL FATTY ACIDS AS DRUGS Of the theoretically interesting drug candidates, GLA, DGLA and AA of the n-6 series, and SA, EPA and DHA of the n-3 series, only two have achieved the status of prescription drugs. These are
EICOSANOID PRECURSORS AS PHARMACEUTICALS (a) GLA, which, in the form of evening primrose oil, is approved in a number of Western European countries for eczema and for breast pain, and (b) EPA, which, in the form of the pure ethyl ester, is approved in Japan for the treatment of peripheral vascular disease and elevated triglycerides. EPA/DHA and other fatty acids in a mix have also been approved in Western Europe for lowering of triglycerides and prevention of death after myocardial infarction. AA and DHA from various natural sources have been approved by most developed countries, including the USA, as additives to infant formula to make it more like human breast milk in composition. However, these are additives to food as distinct from pharmaceuticals and they are applied in the form of natural plant oils, natural phospolipids or microbial cells.
Ethyl-EPA This is the only pure, chemically defined EFA that has been approved anywhere for any pharmaceutical purpose. In Japan it is licensed at a dose of 1.8 g/day (66300 mg gelatin capsules) for the treatment of peripheral vascular disease and elevated triglycerides (Kawamura et al 1983). It was developed by Mochida. Laxdale is developing similarly pure ethyl-EPA for the management of neurodegenerative disorders, schizophrenia and depression. In the latter two conditions, dose–response studies have compared placebo with 1 g, 2 g or 4 g/day. The results were surprising. In schizophrenia 1 g/day had a modest effect, 2 g/day had a substantial effect, whereas 4 g/day was slightly worse than placebo (Peet et al 2002b). The beneficial effect was strongly correlated with the effect of treatment on red cell AA, rather than with EPA or DHA levels. 1 g and 2 g/day elevated AA, possibly by inhibiting PLA2-induced loss of AA from phospholipids. 4 g/day, in contrast, reduced AA levels, probably both by direct competition for incorporation and by reduced synthesis of AA by D5D (Peet et al 2002b; Horrobin et al 2002). This result, obtained in a single experiment with a single fully defined pure compound, parallels the results obtained in cardiovascular studies. Low doses, which elevate the phospholipid concentration composition of AA, have positive effects, whereas high doses, which deplete AA, have neutral or negative effects. A similar dose-ranging study was performed in depression (Peet et al 2002a). The results were similar, except that this time the optimum close was only 1 g/day with 2 g and 4 g/day being less effective. Presumably, different mechanisms are involved in depression and schizophrenia so it is not surprising that optimal doses will be different. The principle of performing careful dose– response studies of pure compounds as the only way to define how patients may benefit from EFA treatment is therefore strongly supported. It should be noted that in psychiatric disorders single dose levels of DHA have proved to be ineffective in both depression and attention deficit hyperactivity disorder (Marangell et al 2000; Voigt et al 2001). It is unknown whether this is because DHA is completely ineffective or whether there might be efficacy at different doses.
GLA as Evening Primrose Oil Triglycerides GLA, in the form of the triglyceride extracted from evening primrose seeds, has been approved as a treatment for atopic eczema and for breast pain in the UK and other European countries (Morse et al 1989; Mansel 1994). The dose is around 300–1000 mg/day of the GLA. Higher doses might possibly be more effective but would be difficult to administer in this format, as the GLA makes up only around 9% of the oil. The availability
597
of a pure single compound could be a substantial advance which would allow proper dose–response studies. The mechanism of action has been only partially defined, but in both illnesses may include antiinflammatory, immunomodulating and receptor-related effects. There is also likely to be a PPARrelated effect, since GLA binds to PPAR receptors with a higher affinity than any other fatty acid (Xu et al 1999). In atopic eczema there is now solid evidence of impaired D6D conversion of dietary LA to GLA and further metabolites, resulting in low concentrations of DGLA and AA. This offers a clear rationale for using GLA.
FUTURE PROSPECTS AND CONCLUSIONS The literature dealing with EFAs as pharmaceuticals for the management of specific diseases is tantalizing, both in its promise and its inadequacy. Much of the problem relates to a failure to think clearly about multiple mechanisms of action, dose– responses and the value of pure compounds. Simplistic and sometimes downright silly beliefs and statements abound. However, what is being achieved with pure ethyl-EPA indicates that the promise, far from being an illusion, is real. When other pure fatty acids are studied in careful dose–response studies we can expect the situation to be considerably clarified. With good fortune clear, safe therapeutic effects will result.
REFERENCES Benowitz LI and Routtenberg A (1997) GAP-43: an intrinsic determinant of neuronal development and plasticity. Trends Neurosci, 20, 84–91. Brenner RR (1981) Nutritional and hormonal factors influencing desaturation of essential fatty acids. Prog Lipid Res, 20, 41–48. Burr ML, Fehily AM, Gilbert JF et al (1989) Effects of changes in fat, fish, and fibre intakes on death and myocardial reinfarction: diet and reinfarction trial (DART). Lancet, 2, 757–761. Calabrese EJ (2001) Prostaglandins: biphasic dose responses. Crit Rev Toxicol, 31, 475–487. Chiu LC, Ooi VE and Wan JM (2001) Eicosapentaenoic acid modulates cyclin expression and arrests cell cycle progression in human leukemic K-562 cells. Int J Oncol, 19, 845–849. Clarke SD (2000) Polyunsaturated fatty acid regulation of gene transcription: a mechanism to improve energy balance and insulin resistance. Br J Nutr, 83, S59–66. Clarke SD (2001) Polyunsaturated fatty acid regulation of gene transcription: a molecular mechanism to improve the metabolic syndrome. J Nutr, 131, 1129–1132. Ewin JH (2001) Fine Wine and Fish Oil. New York: Oxford University Press. Finnen MJ and Lovell CR (1991) Purification and characterisation of phospholipase A2 from human epidermis. Biochem Soc Trans, 19, 91S. Forman BM, Chen J and Evans RM (1997) Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors alpha and delta. Proc Natl Acad Sci USA, 94, 4312–4317. Gruppo Italiano per lo Studio della Streptochinasi nell’Infarto Miocardico (GISSI) (1999) Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI–Prevenzione trial. Lancet, 354, 447–455. Horrobin DF (1977) Interactions between prostaglandins and calcium: the importance of bell-shaped dose–response curves. Prostaglandins, 14, 667–677. Horrobin DF (1982) Clinical Uses of Essential Fatty Acids. Montreal: Eden Press. Horrobin DF (1983) The regulation of prostaglandin biosynthesis by the manipulation of essential fatty acid metabolism. Rev Pure Appl Pharmacol Sci, 4, 339–383. Horrobin DF (1990) o-6 Essential Fatty Acids. Pathophysiology and Roles in Clinical Medicine. New York: Wiley-Liss.
598
CONCLUSIONS AND CORRELATIONS
Horrobin DF (1992) Nutritional and medical importance of g-linolenic acid. Prog Lipid Res, 31, 163–194. Horrobin DF (1995) Abnormal membrane concentrations of 20- and 22carbon essential fatty acids: a common link between risk factors and coronary and peripheral vascular disease? Prostagland Leukotriene Essent Fatty Acids, 53, 385–396. Horrobin DF, Jenkins K, Bennett CN and Christie WW (2002) Eicosapentaenoic acid and arachidonic acid: collaboration and not antagonism is the key to biological understanding. Prostagland Leukotriene Essent Fatty Acids (in press). Johansen O, Seljeflot I, Hostmark AT and Arensen H (1999) The effect of supplementation with o-3 fatty acids on soluble markers of endothelial function in patients with coronary heart disease. Arterioscler Throm Vasc Biol, 19, 1681–1686. Kang JX and Leaf A (1996) Evidence that free polyunsaturated fatty acids modify Na+ channels by directly binding to the channel proteins. Proc Natl Acad Sci USA, 93, 3542–3546. Kawamura M, Naito C, Hayashi H et al (1983) [Effects of 4 weeks’ intake of polyunsaturated fatty acid ethylester rich in eicosapentaenoic acid (ethylester) on plasma lipids, plasma and platelet phospholipid fatty acid composition and platelet aggregation; a double blind study]. Nippon Naika Gakkai Zasshi, 72, 18–24. Kliewer SA, Sundseth SS, Jones SA et al (1997) Fatty acids and eicosanoids regulate gene expression through direct interactions with peroxisome proliferator-activated receptors a and g. Proc Natl Acad Sci USA, 94, 4318–4323. Leaf A, Jorgensen MB, Jacobs AK et al (1994) Do fish oils prevent restenosis after coronary angioplasty? Circulation, 90, 2248–2257. Mansel RE (1994) Breast Pain. Br Med J, 309, 866–868. Marangell LB, Zboyan HA, Cress KK et al (2000) A double-blind, placebo-controlled study of docosahexaenoic acid in the treatment of depression. Inform, 11, S78. Martin DSD, Lonergan P, Boland B et al (2002) Apoptotic changes in the aged brain are triggered by IL-1-induced activation of p38 and reversed by dietary supplementation with eicosapentaenoic acid. Prostagland Leukotriene Essent Fatty Acids (in press). Morse PF, Horrobin DF, Manku MS et al (1989) Meta-analysis of placebo-controlled studies of the efficacy of Epogam in the treatment of atopic eczema. Relationship between plasma essential fatty acid changes and clinical response. Br J Dermatol, 121, 75–90. Nassar BA, Huang Y-S and Horrobin DF (1986a) Response of tissue phospholipid fatty acid composition to dietary (n-6) and replacement with marine (n-3) and saturated fatty acids in the rat. Nutr Res, 6, 1397–1406. Nassar BA, Huang YS, Manku MS et al (1986b) The influence of dietary manipulation with n-3 and n-6 fatty acids on liver and plasma phospholipid fatty acids in rats. Lipids, 21, 652–656. Nassar BA, Manku MS, Huang YS et al (1987) The influence of dietary marine oil (Polepa) and evening primrose oil (Efamol) on
prostaglandin production by the rat mesenteric vasculature. Prostagland Leukotriene Med, 26, 253–263. Nelson GJ, Schmidt PC, Bartolini G et al (1997) The effect of dietary arachidonic acid on platelet function, platelet fatty acid composition, and blood coagulation in humans. Lipids, 32, 421–425. Okuyama H, Kobayashi T and Watanabe S (1997) Dietary fatty acids— the n-6/n-3 balance and chronic elderly diseases. Excess linoleic acid and relative n-3 deficiency syndrome seen in Japan. Prog Lipid Res, 35, 409–457. Palakurthi SS, Aktas H, Grubissich LM et al (2001) Anticancer effects of thiazolidinediones are independent of peroxisome proliferatoractivated receptor g and mediated by inhibition of translation initiation. Cancer Res, 61, 6213–6218. Peet M, Brind J, Ramchand CN et al (2001) Two double-blind placebocontrolled pilot studies of eicosapentaenoic acid in the treatment of schizophrenia. Schizophr Res, 49, 243–251. Peet M, Horrobin DF and E-E Multicentre Study Group (2002a) A doseranging study of the effects of ethyl eicosapentaenoate in patients with ongoing depression in spite of apparently adequate treatment with standard drugs. Arch Gen Psychiatry, 59, 913–919. Peet M, Horrobin DF and EPA Multicentre Study Group (2002b) A doseranging exploratory study of the effects of ethyl-eicosapentaenoate in patients with persistent schizophrenic symptoms. J Psychiatric Res, 36, 7–18. Price SA and Tisdale MJ (1998) Mechanism of inhibition of a tumor lipidmobilizing factor by eicosapentaenoic acid. Cancer Res, 58, 4827–4831. Reynolds LM, Dalton CF and Reynolds GP (2001) Phospholipid fatty acids and neurotoxicity in human neuroblastoma SH-SY5Y cells. Neurosci Lett, 309, 193–196. Schoonjans K, Staels B and Auwerx J (1996) Role of the peroxisome proliferator-activated receptor (PPAR) in mediating the effects of fibrates and fatty acids on gene expression. J Lipid Res, 37, 907–925. Simopoulos AP, Leaf A and Salem N (2000) Workshop statement on the essentiality of and recommended dietary intakes for o-6 and o-3 fatty acids. Prostagland Leukotriene Essent Fatty Acids, 63, 119–121. Sinclair HM (1956) Deficiency of essential fatty acids and atherosclerosis, etcetera. Lancet, 1, 381–383. Voigt RG, Llorente AM, Jensen CL et al (2001) A randomized, doubleblind, placebo-controlled trial of docosahexaenoic acid supplementation in children with attention-deficit/hyperactivity disorder. J Pediatr, 139, 189–196. Witt MR and Nielsen M (1994) Characterisation of the influence of unsaturated free fatty acids on brain GABA/benzodiazepine receptor binding in vitro. J Neurochem, 62, 1432–1439. Xu HE, Lambert MH, Montana VG et al (1999) Molecular recognition of fatty acids by peroxisome proliferator-activated receptors. Mol Cell, 3, 397–403.
56 Pharmaceutical Exploitation: Cyclooxygenase and Lipoxygenase Inhibitors Paola Patrignani and Maria G. Sciulli ‘‘G. D’Annunzio’’ University of Chieti School of Medicine, Chieti, Italy
The eicosanoids are a family of biologically active compounds that participate in cell–cell communication (Fitzpatrick and Soberman 2001), as a consequence of the activation of membraneassociated receptors that belong to the G-protein-coupled rhodopsin-type family (Narumiya and FitzGerald 2001). They are formed from polyunsaturated fatty acids, principally, arachidonic acid (AA). AA is a 20-carbon essential fatty acid that contains four double bonds (5,8,11,14-eicosatetraenoic acid), esterified to the phospholipids of cell membranes. Since the concentration of free AA in the cell is very low, the biosynthesis of eicosanoids depends primarily upon the capacity of the cell to release AA from the membranes in response to different physical, chemical and hormonal stimuli. Once released, AA is metabolized enzymatically into three series of biologically active compounds, collectively termed eicosanoids (Figure 56.1A): (a) the prostanoids, i.e. prostaglandin (PG)E2, PGF2a, PGD2, prostacyclin (PGI2) and thromboxane (TX)A2; (b) the leukotrienes (LTs), i.e. LTB4, LTC4, LTD4 and LTE4, and the hydroperoxyeicosatetraenoic acids (HPETE)s; and (c) the epoxyeicosatrienoic acids (EETs) and hydroxyeicosatetraenoic acids (HETEs). Recently, a non-enzymatic pathway of AA conversion has been discovered, giving rise to a novel series of biologically active compounds that are isomers of eicosanoids, termed isoeicosanoids (Morrow et al 1990; Lawson et al 1999; Roberts et al 1999). In this chapter, the therapeutic strategies developed to affect biosynthesis and/or action of prostanoids and leukotrienes in humans, e.g. non-steroidal antiinflammatory drugs (NSAIDs), selective COX-2 inhibitors (coxibs) and antileukotriene drugs (antagonists of LT receptors and inhibitors of LT biosynthesis), will be reviewed. BIOSYNTHESIS AND PHARMACOLOGICAL MODULATION OF PROSTANOIDS Inducible and Constitutive Pathways of Prostanoid Biosynthesis The biosynthesis of prostanoids is triggered by different types of stimuli that may be classified primarily according to the time employed to induce detectable levels of them in the surrounding milieu of the cell. Thus, a rapid and delayed prostanoid biosynthesis can occur (Fitzpatrick and Soberman 2001). The coordinated activity of three consecutive enzymatic steps is involved in both rapid and delayed prostanoid biosynthesis (Figure 56.1B): (a) the release of AA from membrane phosphoThe Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
lipids carried out by phospholipases (PLs), primarily PLA2 (Dennis 2000; Fitzpatrick and Soberman 2001); (b) the transformation of AA to the unstable endoperoxide PGH2 by prostaglandin H synthases (PGHSs) (Marnett et al 1999; Smith and Langenbach 2001); and (c) its metabolization to the different prostanoids by isomerases, which have different structures and exhibit a cell- and tissue-specific distribution (Jakobsson et al 1999; Tanioka et al 2000; Ueno et al 2001). What differentiates these two pathways of prostanoid biosynthesis is the involvement of an enzymatic machinery, constitutively expressed (for rapid biosynthesis) or induced in response to inflammatory and mitogenic stimuli (for delayed biosynthesis) (Figure 56.2). Moreover, inducible but not constitutive prostanoid biosynthesis is sensitive to glucocorticoids (Masferrer et al 1994; Santini et al 2001). Two isoforms of PGHS have been identified, also referred to as COX-1 and COX-2, that catalyse the same reactions, i.e. the conversion of AA to PGG2 by their cyclooxygenase (COX) activity and, then, the reduction of PGG2 to PGH2 by their peroxidase (HOX) activity (Marnett et al 1999; Smith and Langenbach 2001) (Figure 56.1B). The reactions catalysed by PGHS isozymes are controlled by a complex balance between hydroperoxide generation and removal at the level of the cell. Thus, the initiation of the PGHS activity requires the generation of a hydroperoxide tone (higher for COX-1 than for COX-2) (Kulmacz and Wang 1995). In contrast, high levels of hydroperoxides cause an irreversible loss of PGHS activity (Wu et al 1999). Cellular glutathione (GSH) may modulate prostanoid biosynthesis. GSH is involved in the inhibition of cyclooxygenase catalysis of PGHS isozymes because it is a necessary co-factor in the peroxide-removing reaction catalysed by GSH-peroxidase (Hemler and Lands 1980; Kulmacz and Wang 1995). The expression of the two isozymes of PGHS is differently regulated (Table 56.1) (reviewed by Smith and Langenbach 2001). COX-1 displays the characteristics of a ‘‘housekeeping gene’’ and is constitutively expressed in virtually all tissues. It is mainly utilized in the immediate biosynthesis of prostanoids, which occurs within several minutes after stimulation with Ca2+ mobilizers. In contrast, the inducible COX-2 is an absolute requirement for delayed prostanoid biosynthesis, which lasts for several hours following proinflammatory stimuli (Fitzpatrick and Soberman 2001) (Figure 56.2). However, this simplified paradigm of constitutive COX-1 and inducible COX-2 has many exceptions: COX-1 can be regulated during development (Rocca et al 1999; Smith and Langenbach 2001), whereas COX-2 is constitutively expressed in the brain (Yamagata et al 1993), reproductive tissues
600
CONCLUSIONS AND CORRELATIONS
[Image not available in this electronic edition.]
Figure 56.1 (A) Biosynthesis of eicosanoids. Arachidonc acid, a 20-carbon fatty acid containing four double bonds, is released from the sn2 position in membrane phospholipids by phospholipases and is metabolized enzymatically into three series of biologically active compounds, collectively termed eicosanoids: (a) the prostanoids, i.e. prostaglandin (PG)E2, PGF2a, PGD2, prostacyclin (PGI2) and thromboxane (TX) A2; (b) the leukotrienes (LTs), i.e. LTB4, LTC4, LTD4 and LTE4, and the hydroperoxyeicosatetraenoic acids (HPETEs); and (c) the epoxyeicosatrienoic acids (EETs) and hydroxyeicosatetraenoic acids (HETEs). (B) Production of prostanoids. The coordinated activity of three consecutive enzymatic steps is involved in prostanoid biosynthesis: (a) the release of AA from membrane phospholipids carried out by phospholipase A2; (b) the transformation of AA to the unstable endoperoxide PGH2 by PGH synthases (COX-1 and COX-2); and (c) its metabolization to the different prostanoids by isomerases which have different structures and exhibit a cell- and tissue-specific distribution. The different prostanoids activate specific cell-membrane receptors that belong to the G-protein-coupled rhodopsin-type family. COX, cyclooxygenase; HOX, peroxidase. Adapted by permission from FitzGerald and Patrono (2001). Copyright &2001 Massachusetts Medical Society. All rights reserved
(Kniss 1999) and kidney (Harris et al 1994; Vio et al 1997; Ferreri et al 1999). At least two GSH-dependent forms of terminal synthases, involved in the production of PGE2, have recently been identified (Jakobsson et al 1999; Forsberg et al 2000; Tanioka et al 2000; Mancini et al 2001). One isoform, present in the cytosol (cPGES), constitutively expressed and unresponsive to proinflammatory stimuli, is identical to the heat shock protein 90-associated protein p23 and promotes COX-1-mediated rapid PGE2 biosynthesis
(Tanioka et al 2000). The other isoform (mPGES), localized to the microsomal compartment, is coordinately induced with COX-2 in response to proinflammatory stimuli (Murakami et al 2000; Stichtenoth et al 2001) (Figure 56.2). The formation of biologically important prostanoids, such as PGE2 and TXA2, is involved in a variety of pathophysiological processes. These include modulation of the inflammatory reaction, gastrointestinal (GI) cytoprotection and ulceration, angiogenesis and cancer, haemostasis and thrombosis, renal
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS
Figure 56.2 Time-course of COX-1, COX-2 and mPGES expression and prostanoid biosynthesis in monocytes stimulated with bacterial endotoxin (LPS). Isolated human monocytes (2–3106/ml) were incubated for 0, 4, 24, 48 and 72 h at 378C with LPS (10 mg/ml) and PGE2 levels were measured in the medium by radioimmunoassay, while COX-1, COX-2 and mPGES levels were evaluated in monocyte lysates by Western blot
601
in the gastric mucosa, where prostanoids play an important role in protecting the stomach from erosion and ulceration (FitzGerald and Patrono 2001; Patrono et al 2001b). COX-1 and COX-2, being widely expressed in the kidney (Harris et al 1994; Vio et al 1997; Ferreri et al 1999), appear to produce vasodilating prostanoids that maintain renal plasma flow and glomerular filtration rate, especially during conditions of systemic vasoconstriction (Dunn 2000; Brater et al 2001). Moreover, COX-2-dependent prostanoid biosynthesis contributes to renal sodium balance (Catella-Lawson et al 1999). Thus, NSAID therapy is often associated with adverse changes in renal electrolyte excretion and the impairment of renal perfusion (Dunn 2000; Brater et al 2001). In addition, NSAID therapy may also cause blood pressure destabilization (Whelton 2001), attributable, at least in part, to the ability of NSAIDs to attenuate the antihypertensive effects of several agents, including diuretics and angiotensin-converting enzyme (ACE) inhibitors (Mene` et al 1995). Finally, NSAIDs may occasionally produce acute renal failure (Dunn and Zambraski 1980; Fored et al 2001). The development of coxibs, a new family of antiinflammatory drugs, represents a response to the unsatisfactory safety profile of NSAIDs. Therefore, it was hoped that coxibs would be better tolerated than non-selective NSAIDs but equally efficacious. Development of Coxibs
haemodynamics, and progression of kidney disease (Smith and Langenbach 2001). Thus, it is not surprising that drugs inhibiting the activity of PGHS isozymes, e.g. NSAIDs and coxibs, may have desirable as well as untoward effects on a variety of human diseases (FitzGerald and Patrono 2001; Patrono et al 2001b).
Mechanism of Therapeutic and Toxic Effects of NSAIDs Aspirin and related NSAIDs originally were found to prevent the synthesis of prostanoids from AA in tissue homogenates (Vane 1971). It is now well established that both the therapeutic effects and side-effects of these drugs are largely dependent on the inhibition of the cyclooxygenase activity of PGHS isozymes (Vane 1994; FitzGerald and Patrono 2001; Patrono et al 2001b). The inhibition of COX-2 is thought to mediate the antipyretic, analgesic and antiinflammatory actions of NSAIDs, while the inhibition of COX-1 results in unwanted side-effects, particularly in the GI tract. In fact, COX-1 is the major PGHS isoform expressed in platelets and gastric mucosa of normal humans (Kargman et al 1996; Patrignani et al 1999; Rocca et al 2002). NSAID toxicity in the GI mucosa, leading to ulceration, bleeding, perforation and obstruction, is the result of inhibition of COX-1 activity in platelets, which increases the tendency of bleeding, and Table 56.1 Expression of COX-1 and COX-2 COX-1
COX-2
Chromosome 9 Constitutive, found in all tissues; highly expressed in stomach, kidney and platelets Can increase from 2- to 4-fold
Chromosome 1 Predominantly inducible in response to inflammatory, mitogenic stimuli in macrophages/monocytes, synoviocytes, fibroblasts, vascular cells Constitutive in kidney, brain, uterus, vascular cells and malignant cells Can increase from 10- to 20-fold
X-ray crystallography of the 3-D structure of COX-1 and COX-2 has evidenced that these two enzymes are very similar, in fact they contain a cyclooxygenase active site characterized by a hydrophobic, long and narrow channel (Picot et al 1994; Kurumbail et al 1996; Luong et al 1996). However, the cyclooxygenase channel of COX-2 is wider than that of COX-1, due to a single amino acid difference, i.e. at position 523 an isoleucine molecule in COX-1 is substituted with a valine (smaller by a single methyl group) in COX-2. The smaller valine molecule in COX-2 leaves a gap in the wall of the channel, giving access to a side-pocket (Figure 56.3A). Most non-selective NSAIDs are acidic and inhibit prostanoid biosynthesis by competing with AA for binding to arginine (Arg) at position 120, which is present in both COX isozymes (Figure 56.3B) (Mancini et al 1995). Non-selective NSAIDs first bind both PGHS isozymes near the solvent-accessible opening of the hydrophobic channel; then they translocate along the length of this channel and associate with the cyclooxygenase active site by a reversible interaction (Figure 56.3B). An additional third step is involved in selective COX-2 inhibition by celecoxib and rofecoxib, two coxibs that are diarylheterocycles containing a sulphonamide and a phenylsulphone moiety, respectively (Figure 56.4): the formation of a tightly bound enzyme–inhibitor complex, which involves the optimization of inhibitor and protein conformational changes in the active site and the side-pocket (Figure 56.3C) (Walker et al 2001). Recently, a new wave of coxibs, valdecoxib and etoricoxib (Figure 56.4), that are more selective than celecoxib and rofecoxib, respectively, have been developed (Talley et al 2000; Riendeau et al 2001). Clinical Effects of Non-selective and Selective Inhibition of PGHS isozymes Selectivity is a measure of relative concentrations of a drug required to inhibit each PGHS isozyme (often expressed as COX1/COX-2 IC50 ratio; IC50 is the concentration of the drug required to inhibit the activity of COX-1 and COX-2 by 50%), and applies to in vitro studies performed during drug screening. Many biological assays have been developed to assess the selectivity
602
CONCLUSIONS AND CORRELATIONS
Figure 56.4 Coxibs: chemical structures
Figure 56.3 (A) The cyclooxygenase active site of COX-1 and COX-2 is characterized by a hydrophobic, long and narrow channel. However, the cyclooxygenase channel of COX-2 is wider than that of COX-1 due to a single amino acid difference, i.e. at position 523 an isoleucine molecule in COX-1 is substituted with a valine (smaller by a single methyl group). (B) Most non-selective NSAIDs compete with arachidonic acid for the binding to the arginine (Arg) at position 120 that is present in both COX-1 and COX-2. (C) Selective COX-2 inhibitors, such as celecoxib and rofecoxib, which are diarylheterocycles containing a sulphonamide or a phenylsulphone moiety, respectively, interact with COX-2 channel forming a tight complex that involves their adjustment in COX-2 channel and side-pocket and conformational changes. In contrast, selective COX-2 inhibitors bind COX-1 with a reversible and lowaffinity interaction
towards the cyclooxygenase activity of COX-2 vs. COX-1 by NSAIDs in vitro that utilize purified enzyme systems derived from microsomal membranes of cells transfected with the murine (Meade et al 1993) or human isozymes (Laneuville et al 1994), or cultured cells, which selectively express COX-1 and COX-2
(Mitchell et al 1993) (Figure 56.5A). The relative degree of isoform selectivity of cyclooxygenase inhibitors varies, depending on the experimental conditions of the assays used (Brooks et al 1999). To provide a more physiological assessment of isoform selectivity by cyclooxygenase inhibitors, a model of COX-2 expression in human whole blood in response to bacterial endotoxin (lipopolysaccharide, LPS), a known stimulus for COX-2 induction, has been developed (Patrignani et al 1994) (Figure 56.5B). The measurement of PGE2 production in response to LPS added to heparinized blood samples reflects the timedependent induction of COX-2 in circulating monocytes (Masferrer et al 1994; Patrignani et al 1994). The parallel measurement of TXB2 production during whole blood clotting is used as an index of platelet COX-1 activity (Patrono et al 1980). The whole blood assays allow the assessment of COX-1/COX-2 selectivity both in vitro and ex vivo (Patrignani et al 1994; Panara et al 1998, 1999; Ehrich et al 1999; McAdam et al 1999). Based on this methodological approach, a large number of cyclooxygenase inhibitors have been screened in vitro and they can be grouped in three categories (Figure 56.6): (a) selective COX-1 inhibitors, e.g. aspirin and S-indobufen; (b) non-selective COX inhibitors—the vast majority of the NSAIDs examined exhibit COX-1/COX-2 ratios of 0.5–3.0; and (c) selective COX-2 inhibitors, with COX-1/COX-2 IC50 ratios of 18–300. However, the assessment of selectivity in vitro is not informative of what may actually happen in vivo at therapeutic plasma concentrations of the drugs. Using whole blood assays, it is possible to evaluate whether the administration of clinical doses of the drug causes the inhibition ex vivo of monocyte COX-2 without affecting platelet COX-1. Biochemical COX-2 selectivity can be defined as a ‘‘COX-1-sparing’’ effect of the drug at and above clinical doses, i.e. throughout the whole range of therapeutic drug levels. Since the administration of therapeutic doses of nimesulide and meloxicam to healthy subjects causes a profound suppression of monocyte COX-2 activity associated with a detectable inhibition of platelet COX-1 (Panara et al 1998, 1999), these two drugs cannot be considered selective COX-2 inhibitors. In contrast, rofecoxib inhibits monocyte COX-2 activity without affecting COX-1 activity at doses in excess of those recommended
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS
603
Figure 56.5 (A) The effects of cyclooxygenase inhibitors on COX-1 and COX-2 can be evaluated in vitro, in purified or recombinant isozymes or in cultured cells and circulating blood cells, which selectively express the two isozymes. The whole blood assays allow the assessment of PGHS isozyme inhibition ex vivo, i.e. the effects of circulating concentrations of therapeutic doses of the drug. The inhibition of prostanoid can be assessed in vivo through the measurement of prostanoids in biological fluids, such as synovial fluid, or their major enzymatic metabolites in urine. However, these measurements do not give information about the isozyme and the cell type inhibited by the drug. (B) Whole blood assays to evaluate the effects of cyclooxygenase inhibitors on platelet COX-1 (a) and monocyte COX-2 (b) activities. The measurement of PGE2 production in response to bacterial endotoxin (LPS) added to heparinized blood samples reflects the time-dependent induction of COX-2 in circulating monocytes (Patrignani et al 1994) while the parallel measurement of TXB2 production during whole blood clotting is used as an index of platelet COX-1 activity. Reprinted from Patrono et al (1980). Copyright &1980 by permission of Elsevier
for clinical use, both in healthy subjects (Ehrich et al 1999) and in patients with rheumatoid arthritis (RA) (Patrignani et al in preparation; Figure 56.7A). Moreover, therapeutic doses of the drug did not affect gastric PGE2 biosynthesis ex vivo (Wight et al 2001). This drug represents an appropriate tool to test the hypothesis that the antiinflammatory and analgesic effects of NSAIDs are dependent on the inhibition of COX-2, while their typical side-effects are due to the inhibition of COX-1. A large randomized, double-blind GI outcomes study has been performed to assess the risk of clinically important upper GI (UGI) events associated with rofecoxib vs. the non-selective NSAID naproxen: the Vioxx Gastrointestinal Outcomes Research (VIGOR) study
(Bombardier et al 2000). The incidence of GI perforation, GI haemorrhage or symptomatic peptic ulcer was significantly (p50.001) lower in patients with RA treated with rofecoxib vs. naproxen (Bombardier et al 2000; Figure 56.7B) consistently with the COX-2 selectivity of the former even at high doses (Ehrich et al 1999 and Figure 56.7A). In whole blood assays in vitro, celecoxib showed a COX-1/ COX-2 IC50 ratio of 30 (Figure 56.6). Not surprisingly, such a limited COX-2 selectivity was associated with large interindividual variability in the inhibition of platelet COX-1 activity following single oral doses of celecoxib, 100–800 mg, in healthy subjects (McAdam et al 1999) (Figure 56.8). Inadequate biochemical
604
CONCLUSIONS AND CORRELATIONS
Figure 56.6 The biochemical selectivity of cyclooxygenase inhibitors in whole blood assays of COX isozyme activity. The relative degree of isoform selectivity is expressed as COX-1/COX-2 IC50 ratios. IC50 is the concentration of the drug required to inhibit the activity of COX-1 and COX-2 by 50%
selectivity of celecoxib at 800 mg daily may have contributed, at least in part, to the failure of the Celecoxib Long-term Arthritis Safety Study (CLASS; Silverstein et al 2000) in reaching a statistically significant difference in its primary end-point between celecoxib and comparator NSAIDs, as well as to the apparent heterogeneity of the results obtained in separate comparisons of celecoxib to ibuprofen vs. celecoxib to diclofenac (FitzGerald and Patrono 2001). Due to a wide distribution of COX-2 in the kidney (Harris et al 1994; Vio et al 1997; Ferreri et al 1999), the administration of coxibs produces qualitative changes in urinary prostanoid excretion, glomerular filtration rate, sodium retention and development of peripheral oedema that are similar to, but no worse than, those caused by non-selective NSAIDs (CatellaLawson et al 1999; McAdam et al 1999; Brater et al 2001; Whelton 2001). Rare cases of acute renal failure and acute tubulointerstitial nephritis have been described in patients taking selective COX-2 inhibitors (Dunn 2000; Perazella and Eras 2000; Rocha and Fernandez-Alonzo 2001). Therapeutic doses of rofecoxib and celecoxib inhibit monocyte COX-2 activity to a similar extent as conventional NSAIDs (Van Hecken et al 2000) and this is associated with comparable clinical efficacy to traditional NSAIDs in patients with osteoarthritis (OA) and RA (Simon et al 1998; Bensen et al 1999; Simon et al 1999; Cannon et al 2000; Clemett and Goa 2000; Day et al 2000; Hawkey et al 2000). Whether rofecoxib and celecoxib are equally efficacious is not known. Recently, Geba et al (2002) have reported that rofecoxib 25 mg/day is superior to celecoxib 200 mg/ day and acetaminophen 4000 mg/day in treating OA of the knee. However, some important limitations of this study (i.e. the absence of the placebo group and the dose of celecoxib, presumably not equivalent to rofecoxib and acetaminophen doses) prevent a final conclusion from being reached. A new wave of highly selective COX-2 inhibitors (etoricoxib and valdecoxib) has been developed (Talley et al 2000; Riendeau et al 2001) (Figures 56.4 and 56.6). The rationale of developing COX-2 inhibitors with improved biochemical selectivity over those of commercially available coxibs may have two distinct potential advantages. In principle, it should lead to a clear-cut separation of COX-2- from COX-1-dependent effects, by virtue of maximizing the likelihood of exposed patients being in the 80– 100% range of COX-2 inhibition and in the 0–20% range of
COX-1 inhibition throughout the dosing interval. Whether this will allow using higher coxib doses for improved efficacy remains to be established, given the dose-dependence of mechanism-based renal effects associated with these agents (Brater et al 2001). Second, it could diminish the probability of the COX-2 inhibitor interfering with permanent inactivation of platelet COX-1 by lowdose aspirin in patients requiring antiinflammatory and antiplatelet therapy (Catella-Lawson et al 2001). Recent studies suggest that the likelihood of selective COX-2 inhibitors sharing this pharmacodynamic interaction of conventional NSAIDs is inversely related to their COX-2 selectivity (Ouellet et al 2001). Etoricoxib and valdecoxib are commercially available in some countries for the treatment of arthritis and acute pain (Daniels et al 2002; Leung et al 2002, Makarowski et al 2002; Schumacher et al 2002). Whether their administration will be associated with a detectably improved clinical efficacy and/or safety vis-a`-vis existing coxibs remains to be established. The commercially available coxibs are non-acidic and highly lipophilic, with long half-lives (411 h) and characterized by a homogeneous distribution in the body (Table 56.2). In contrast, acidic non-selective NSAIDs (e.g. diclofenac) achieve high concentrations in inflamed tissues and this might be associated with improved clinical efficacy. However, acidic NSAIDs also achieve high concentrations in the stomach wall that may be associated with the gastric toxicity of non-selective NSAIDs. Thus, it would be desirable to combine the tissue-targeted distribution of the acidic non-selective NSAIDs with the COX1-sparing effects of highly selective COX-2 inhibitors. Lumiracoxib, the first representative of this new series of coxibs, is in advanced clinical development for the treatment of arthritis and pain. Aspirin The best-characterized mechanism of action of aspirin is related to its capacity to inactivate permanently the cyclooxygenase activity of COX-1 and COX-2 (Majerus 1983; Roth and Majerus 1975; Roth et al 1975; Burch et al 1979; Smith et al 1996). The molecular mechanism of permanent inactivation of cyclooxygenase activity by aspirin is related to the blockade of the cyclooxygenase
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS channel, as a consequence of the acetylation of a strategically located serine residue (i.e. Ser529 in the human COX-1 and Ser516 in the human COX-2), which prevents access of the substrate to the catalytic site of the enzyme (Loll et al 1995). Because aspirin has a short half-life (15–20 min) in the human circulation and is approximately 100-fold more potent in inhibiting platelet COX-1 than monocyte COX-2 (Cipollone et al 1997; Figure 56.6), it is ideally suited to act on anucleate platelets, inducing a permanent defect in TXA2-dependent platelet function. Moreover, since aspirin probably also inactivates COX1 in relatively mature megakaryocytes, and since only 10% of the platelet pool is replenished each day, once-a-day dosing of aspirin (75–100 mg) is able to maintain virtually complete inhibition of platelet TXA2 production. In contrast, the inhibition of COX-2dependent pathophysiological processes (e.g. hyperalgesia and inflammation) requires larger doses of aspirin (because of the decreased sensitivity of COX-2 to aspirin) and a much shorter dosing interval (because nucleated cells rapidly resynthesize the enzyme). Thus, there is an approximately 100-fold variation in daily doses of aspirin when it is used as an antiinflammatory rather than as an antiplatelet agent. Furthermore, the benefit–risk profiles of the drug depend on the dose and indication, since its GI toxicity is dose-dependent. Permanent inactivation of platelet COX-1 by aspirin may lead to the prevention of thrombosis as well as to excess bleeding (Patrono et al 2001a). Inhibition of platelet function is largely responsible for the two-fold increase in the risk of upper GI bleeding associated with daily doses of aspirin in the range 75–100 mg, inasmuch as a similar relative risk is associated with other antiplatelet agents that do not act on PGHS isozymes (Heeschen and Hamm 2000). At higher doses of aspirin, inhibition of GI COX-1-dependent cytoprotection amplifies bleeding risk by causing new mucosal lesions. Acetaminophen (Paracetamol) Acetaminophen is the first-line therapy recommended by the American College of Rheumatology for the treatment of pain and stiffness associated with OA (ACR Subcommittee on Osteoarthritis Guidelines 2000), mainly because of its ‘‘assumed’’ lower toxicity on GI tract and kidney as compared with NSAIDs. The drug is considered an analgesic and antipyretic agent with weak antiinflammatory effects (Robak et al 1978). Acetaminophen has been shown to affect prostanoid biosynthesis in brain microsomes (Flower and Vane 1972) and in spinal cord (Muth-Selbach et al 1999). In contrast, it is a weak PGHS inhibitor in spleen homogenates (Flower et al 1972). Acetaminophen is only a modulator of PGHS isozymes (Ouellett and Percival 2001; Boutaud et al 2002) and its inhibitory effects may be influenced by the presence of different components of the intracellular and extracellular milieu (e.g. reduced inhibition by the presence of hydroperoxides and enhanced inhibition by the presence of GSH). This may lead to intersubject variability in the clinical efficacy of the drug that is presumably involved in the preference of NSAIDs over acetaminophen by rheumatic patients (Pincus et al 2000; Wolfe et al 2000). Moreover, acetaminophen (4 g/day) has been recently reported to have lower clinical efficacy than the selective COX-2 inhibitor rofecoxib (25 mg) during a 6-week controlled trial in OA patients (Geba et al 2002). The improved renal and GI safety profile of acetaminophen over NSAIDs is not confirmed by population-based observational studies. In fact, Garcia Rodriguez and Hernadez-Diaz (2001) have shown that, similar to oral steroids and aspirin, acetaminophen use is associated with a two-fold increased risk of upper GI complications when taken at daily doses 52 g. This effect may be due, at least in part, to the capacity of acetaminophen to affect gastric prostanoid biosynthesis (Konturek et al 1981). Recently, it
605
Figure 56.7 (A) Relation between steady-state plasma concentrations of rofecoxib and inhibition of platelet COX-1 and monocyte COX-2, as measured ex vivo using the whole blood assays. Data were obtained from nine patients with rheumatoid arthritis (RA) who were given 50 mg rofecoxib once daily for 7 days. Blood was drawn 4 h after the last dose of rofecoxib. Rofecoxib inhibits monocyte COX-2 activity without affecting COX-1 activity at doses two-fold higher than those recommended for the treatment of RA patients. (B) The incidence of gastrointestinal (GI) perforation, GI haemorrhage, or symptomatic peptic ulcer was significantly lower in patients with osteoarthritis (OA) (*Langman et al 1999) and RA ({Bombardier et al 2000) treated with rofecoxib vs. NSAIDs. In a combined analysis of eight trials of patients with OA, treatment with rofecoxib was associated with a significantly (p50.05) lower incidence of peptic ulcer than treatment with NSAIDs. Similarly, in a large randomized, double-blind GI outcomes study [Vioxx Gastrointestinal Outcomes Research (VIGOR) study] the risk of peptic ulcer was significantly (p50.001) halved in the rofecoxib group vs. the naproxen group
has been demonstrated that therapeutic concentrations (100– 300 mM) of acetaminophen may inhibit platelet TXA2 biosynthesis in vitro by 50–80% (Sciulli et al in preparation). However, the administration of 1 g of the drug to healthy subjects caused only 44% inhibition of platelet COX-1 activity ex vivo that was not associated with any significant inhibitory effect on AA-induced platelet aggregation (Catella-Lawson et al 2001). These results may suggest that the toxicity of the drug is only marginally dependent on the inhibition of prostanoid biosynthesis, but may involve its capacity to induce drug–protein adducts and oxidative stress (Hinson et al 1995; Delanty et al 1996). Thus, acetaminophen use was associated with increased risk of chronic renal failure (Fored et al 2001), despite its inability to reduce renal prostanoid biosynthesis (Bippi and Frolich 1990). Cardiovascular Effects of COX-2 Inhibition Prostacyclin is thought to be part of a homeostatic defence mechanism that limits the consequences of platelet activation in vivo (FitzGerald et al 1984; Cheng et al 2002). Coxibs have been
606
CONCLUSIONS AND CORRELATIONS
Table 56.2 Pharmacokinetics and metabolism of coxibs
Chemistry COX-1/COX-2 IC50 ratio in vitro Pharmacokinetics Oral bioavailability (%) Time to maximal plasma concentration (h) Maximal plasma concentration (ng/ml) Half-life (h) Vol dist. (litres) Bound in plasma (%) Metabolism Main pathway of liver metabolism Urinary excretion (%) Approved daily doses (mg) For osteoarthritis For rheumatoid arthritis For acute pain and primary dysmenorrhea Familial adenomatous polyposis
Celecoxib
Rofecoxib
Valdecoxib
Etoricoxib
Sulphonamide derivative 30
Sulphonyl derivative 276
Sulphonamide derivative 61
Sulphonyl derivative 344
22–40 2–4
92–93 2–3
83 2.3
100 1
705* 11 455+166 97
320** 10–17 86–91 87
161*** 8–11 86 98
788**** 22 120 Not known
Oxidation by cytochrome P-450 (2C9, 3A4) 29
Cytosolic reduction
200 200–400 Up to 400
12.5–25 25 Up to 50
10 10 Up to 40
60 90 Up to 120
800
Not approved
Not approved
Not approved
72
Oxidation by cytochrome Oxidation by cytochrome P-450 (2C9, 3A4) P-450 (3A4) 70 60
After the administration of 200*, 25**, 10*** and 40****mg, respectively.
shown to decrease urinary excretion of prostacyclin metabolites in normal subjects, indicating that the production of prostacyclin is also decreased (Catella-Lawson et al 1999; McAdam et al 1999). In contrast, they do not affect platelet TXA2 biosynthesis (Ehrich et al 1999; McAdam et al 1999). The cardiovascular implications of these effects are currently debated, based on the results of the VIGOR trial showing a statistically significant difference in acute myocardial infarction (MI) rates between rofecoxib and naproxen (0.4 vs. 0.1%, respectively). However, at least three possible explanations can be entertained: (a) a cardioprotective effect of naproxen: however, there is contradictory evidence that naproxen reduces the risk of MI (Garcı` a Rodriguez et al 2000; Rahme et al 2002; Ray et al 2002; Solomon et al 2002; Watson et al 2002); (b) a thrombogenic effect of coxibs: however, the size of the effect is not biologically plausible if due to incomplete inhibition of a single mediator of ‘‘thromboresistance’’, i.e. prostacyclin (McAdam et al 1999); moreover, such an explanation is not substantiated by CLASS results (Silverstein et al 2000), although a smaller coxib effect cannot be excluded; (c) the play of chance: the apparent difference in VIGOR might represent uneven distribution of a small number of events occurring over a short time-frame in a low-risk population; a meta-analysis of all coxib trials might address this possibility. Thus, more information is needed on the cardiovascular effects of selective inhibitors of COX-2. In the meantime, patients with arthritis who have had a major cardiovascular event should be treated with low-dose aspirin in combination with a highly selective COX-2 inhibitor. In contrast, the co-administration of a non-selective NSAID should be avoided. In fact, the concomitant administration of the nonselective NSAID ibuprofen, but not rofecoxib, has been shown to antagonize the irreversible inhibition of platelet COX-1 activity by low-dose aspirin (Catella-Lawson et al 2001). Moreover, it should be mentioned that selective COX-2 inhibition may have potential beneficial effects in addition to low-dose aspirin in the setting of acute coronary syndromes. Thus, COX-2 expressed in inflammatory cells (Masferrer et al 1994; Patrignani et al 1994; Cipollone et al 2001) may account for aspirin-resistant TXA2 biosynthesis in acute coronary syndromes (Cipollone et al 1997, 2000). Baseline urinary excretion of 11dehydro-TXB2 (a major enzymatic metabolite of TXA2) predicted the future risk of MI or cardiovascular death in a subgroup of high-risk aspirin-treated patients participating in the HOPE trial
(Eikelboom et al 2002). However, whether the co-administration of low-dose aspirin with a highly selective COX-2 inhibitor will maintain the advantage of a coxib over a non-selective NSAID with respect to GI side-effects remains to be tested. BIOSYNTHESIS AND PHARMACOLOGICAL MODULATION OF LEUKOTRIENES Biosynthesis of Leukotrienes Leukotrienes (LTs) are potent lipid mediators that have an important role in asthma and other disease processes characterized by inflammation, cellular proliferation and altered vasoconstriction, as well as in homeostasis (Samuelsson 1983; Henderson 1994; Lewis et al 1990). LT biosynthesis occurs predominantly in myeloid cells and is triggered by different stimuli, including antigens, microbes, cytokines and immune complexes. Upon activation, cytosolic phospholipase A2 (cPLA2), a calcium-dependent enzyme, releases AA from membrane phospholipids (Murakami et al 1997; Dennis 2000), then free AA is converted by 5-lipoxygenase (5-LO) to 5HPETE, which is unstable and is transformed both enzymatically, via the HPETE peroxidase, and non-enzymatically into 5-HETE (Figure 56.1A). 5-LO has also another enzymatic activity that catalyses the conversion of 5-HPETE to the epoxide LTA4. Maximal activity of 5-LO requires Ca2+, ATP and hydroperoxides (Rouzer and Samuelsson 1985; Rouzer and Samuelsson 1986) (Figure 56.9). Moreover, its efficient utilization of endogenously released AA in intact cells needs a 18 kDa helper protein, termed 5-LO activating protein (FLAP) (Dixon et al 1990). LTA4 is transformed into two classes of biological compounds, LTB4 and cysteinyl LTs (LTC4, LTD4 and LTE4), by the activity of LTA4 hydrolase (Wetterholm et al 1991; Haeggstrom 2000) and LTC4 synthase (Lam et al 1994), respectively (Figure 56.1A). 5-LO, a 78 kDa protein localized in a soluble intracellular compartment of resting cells, translocates to the nuclear envelope, upon activation, and co-localizes with LTC4 synthase and FLAP, which are integral nuclear envelope proteins (Peters-Golden and Brock 2001) (Figure 56.9). Neutrophils are the main source of LTB4 (Sala et al 1996). However, neutrophil-derived LTA4 may be processed by red
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS
607
blood cells to form LTB4 (Fitzpatrick et al 1984) (Figure 56.10). Eosinophils, basophils/mast cells and monocytes/macrophages have been demonstrated to express both 5-LO/FLAP and LTC4 synthase, and are thus capable of producing LTC4 from endogenous AA. In contrast, platelets and endothelial cells lack 5-LO activity but contain LTC4 synthase (Tornhamre et al 1998) or microsomal glutathione S-transferase-II (GST-II) (Scoggan et al 1997) and can produce LTC4 when supplied with LTA4 via transcellular mechanisms (Feinmark and Cannon 1986; Maclouf and Murphy 1988) (Figure 56.10). Leukotriene Receptors
Figure 56.8 Relation between plasma concentrations of celecoxib and inhibition of platelet COX-1 and monocyte COX-2 activities, as measured ex vivo using the whole blood assays. Data were obtained from healthy subjects treated with single doses (100–800 mg) of celecoxib (McAdam et al 1999). In the whole blood assays in vitro, celecoxib showed a COX-1/ COX-2 IC50 ratio of 30 (see Figure 56.6). Such a limited COX-2 selectivity was associated with large interindividual variability in the inhibition of platelet COX-1 activity ex vivo. Thus, at the higher doses of celecoxib, detectable inhibition of platelet COX-1 was demonstrated
Pharmacological studies have shown that cysteinyl LTs activate at least two receptors, designated CysLT1 and CysLT2 (Coleman et al 1995; Metters 1995). CysLT1 receptor activation by LTD4 causes contraction and proliferation of smooth muscle, oedema, eosinophil migration and damage of the mucus layer of the lung (Dahle´n et al 1980; Piper 1984; Lewis et al 1990). CysLT1 receptor, a glycosylated G-protein-coupled receptor (GPCR) (Metters and Zamboni 1993), is known to signal through elevation of intracellular Ca2+ (Chan et al 1994; Lynch et al 1999). The human CysLT1 receptor has been cloned (Lynch et al 1999; Sarau et al 1999). Expression of CysLT1 mRNA is mainly in the spleen, peripheral blood leukocytes and lung (Lynch et al 1999; Sarau et al 1999). Recently, the molecular cloning and characterization of human CysLT2 has been reported (Heise et al 2000). CysLT2 is a GPCR (Coleman et al 1995; Metters 1995). CysLT2 mRNA has been detected in lung macrophages and airway smooth muscle cells, cardiac Purkinje cells, adrenal medulla cells, peripheral blood leukocytes and brain (Heise et al 2000). The prevalence of CysLT2 receptor on Purkinje cells in the heart and its activation in vascular tissues suggest that, in physiological conditions, cysteinyl LTs may regulate cardiac contractility and maintain vascular tone via the CysLT2 receptor (Folco et al 2000). Experimental data have established that LTB4 acts at specific receptors that are designed BLT1 and BLT2 receptors (Coleman et al 1995; Yokomizo et al 1997, 2000). In leukocytes, LTB4 induces chemokinetic and chemotactic response in vitro (Ford-Hutchinson et al 1980). In vivo, LTB4 increases leukocyte rolling and adhesion to the venular endothelium, followed by their migration into the extravascular space (Dahle´n et al 1981). The activation of the high-affinity BLT1 receptor on leukocytes elicits a pertussis toxinsensitive G-linked chemotactic response (Yokomizo et al 1997). BLT2 receptor is expressed ubiquitously, in contrast to BLT1 which is expressed predominantly in leukocytes. BLT2 receptor binds LTB4 with much lower affinity than BLT1 but its function is presently unknown. LTB4 can also bind and activate the intranuclear transcription factor peroxisome proliferator-activated receptor a (PPARa), resulting in the activation of genes that terminate inflammatory processes (Devchand et al 1996). Drugs Affecting LT Formation and Action: Antileukotrienes
Figure 56.9 Activation of 5-lipoxygenase (5-LO). Upon activation, cytosolic phospholipase A2 (cPLA2) and 5-LO, localized in a soluble intracellular compartment of resting cells, translocate to the nuclear envelope and co-localize with LTC4 synthase and FLAP, which are integral nuclear envelope proteins. Then, free AA is converted by 5-LO to 5-HPETE, which is unstable and is transformed to the epoxide LTA4
Cysteinyl LTs have potent bronchoconstrictor properties (Dahle´n et al 1980) and induce pathophysiological responses similar to those associated with asthma (McFadden and Gilbert 1992; Drazen et al 1999). Specifically, 5-LO products can cause tissue oedema and migration of eosinophils and can stimulate airway secretion. Moreover, they stimulate proliferation of both smooth muscle and various haematopoietic cells. Leukotrienes have been detected from sputum, bronchoalveolar lavage fluid, plasma, nasal secretion, and urine during spontaneous exacerbations of asthma and after antigen challenge (Knapp 1990; Knapp et al
608
CONCLUSIONS AND CORRELATIONS
Figure 56.10 Production of eicosanoids by direct synthesis or transcellular metabolism. Neutrophils express the enzymatic machinery (5-LO, FLAP and LTA4 hydrolase) to produce LTB4 by direct synthesis. Moreover, neutrophil-derived LTA4 may be processed by red blood cells (containing LTA4 hydrolase) to LTB4, and by platelets and endothelial cells (containing LTC4 synthase and microsomal glutathione S-transferaseII) to LTC4 when supplied with LTA4, via transcellular mechanisms. Adapted from Maclouf et al (1988)
LTs and LTB4, whereas the receptor antagonists only block the action of cysteinyl LTs. However, clinical results do not support the hypothesis that zileuton is significantly more efficacious than the CysLT1 antagonists in the treatment of asthma, but it has to be pointed out that no head-to-head comparative studies have been performed. Nevertheless, both pharmacokinetic and safety issues suggest the use of CysLT1 receptor antagonists as the antileukotriene drugs mostly commonly prescribed. In fact, zileuton is recommended to be taken four times a day and liver enzymes should be monitored, because in about 4–5% of patients treated with zileuton an elevation of liver enzymes has been documented. In the majority of patients this occurs during the first 2 months of therapy. The presence of non-responder patients to antileukotriene drugs has been reported and may suggest the occurrence of nonleukotriene-dependent asthma mechanisms or pharmacogenetic factors (Funk et al 1989; Drazen et al 1999). Although antileukotrienes are active anti-asthma drugs, their position in the guidelines for asthma therapy has not yet been clearly stated. The National Heart, Lung, and Blood Institute guidelines position antileukotriene drugs as an alternative to lowdose inhaled steroids, but indicate that inhaled corticosteroids are the preferred option. Thus, to define the advantages of adding an antileukotriene drug to traditional therapy in chronic persistent asthma, further clinical studies have to be performed.
CONCLUSIONS 1992; Wenzel et al 1990; Smith et al 1991). These findings suggest the contribution of 5-LO products in asthma. Thus, antileukotriene drugs have been developed. These drugs consist of synthesis inhibitors that are either direct inhibitors of the 5-LO enzyme or antagonists of FLAP (Drazen et al 1999a) and receptor antagonists of CysLT1 receptor (Drazen et al 1999b; Barnes 2000). Although many antileukotriene drugs have been studied in humans, only few have entered clinical practice: the 5-LO inhibitor zileuton (Carter et al 1991), licensed in the USA, the CysLT1 receptor antagonists montelukast (Jones et al 1995), licensed worldwide, pranlukast (Dubb and Rebuck 1998) licensed in Japan, and zafirlukast (Krell et al 1990), licensed worldwide. Clinical trials have shown that antileukotriene drugs provide bronchodilatory effects in patients with symptomatic asthma (Israel et al 1993; Reiss et al 1997). Furthermore, these studies demonstrate an additive improvement on lung function with b2agonists. Studies in patients with symptomatic asthma, not taking inhaled corticosteroids, have shown that the administration of antileukotriene drugs causes an improvement of lung function, accompanied by a decrease in symptoms and b2-agonist use. Moreover, reduction in asthma exacerbations has been reported (Spector et al 1994; Reiss et al 1998). The effects of asthma exacerbations have been confirmed in a meta-analysis of studies with zafirlukast that showed a 50% decrease in exacerbation rates among patients taking the drug (Barnes et al 1996). The administration of LT antagonists to patients taking lowand high-dose steroids is associated with an increase in pulmonary function and a reduction in asthma exacerbations (Virchow et al 1997a, 1997b; Nayak et al 1998; Laviolette et al 1999). Moreover, in patients taking inhaled steroids, montelukast was able to decrease the dose of steroids (Lofdahl et al 1999). The efficacy of leukotriene antagonists in patients taking steroids is presumably due to the fact that corticosteroids cannot prevent the production of LTs in vivo (O’Shaughnessy et al 1993). On the whole, these findings clearly demonstrate that antileukotriene drugs are active in the treatment of asthma. Zileuton, theoretically, should have broader action than the CysLT1 receptor antagonists, in that it inhibits the formation of cysteinyl
In the last decade, the medical scientific community has moved from the discovery of COX-2 to clinical trials demonstrating that rofecoxib, a highly selective COX-2 inhibitor, causes similar therapeutic effects but significantly fewer GI complications than conventional NSAIDs. Etoricoxib and valdecoxib represent a second wave of COX-2 inhibitors with a slightly higher COX-1/ COX-2 biochemical selectivity than rofecoxib and celecoxib, respectively. Whether their administration will be associated with a detectably improved clinical efficacy and/or safety vis-a`-vis existing coxibs remains to be established. Lumiracoxib, an acidic coxib in advanced clinical development, can theoretically achieve higher concentrations in inflamed tissues than the non-acidic available coxibs, and consequently it could have superior antiinflammatory effects. This hypothesis should be verified in head-to-head comparative clinical trials. However, it is imperative to clarify, in the near future, the cardiovascular and renal effects of selective COX-2 inhibitors. The development of antileukotriene drugs has contributed to elucidate some details of the pathophysiological features of asthma. However, a clearer role for these agents in asthma management guidelines remains to be defined. Thus, it is urgent to design appropriate clinical trials to clarify whether antileukotriene drugs may be used for first-line preventive monotherapy or second-line additive controller therapy. Finally, it is extremely important to characterize the determinants that dictate the response to antileukotriene drugs in this setting.
REFERENCES ACR Subcommittee on Osteoarthritis Guidelines (2000) Recommendations for the medical managment of osteoarthritis of the hip and knee. Arthritis Rheum, 43, 1905–1915. Barnes NC (2000) Effects of antileukotrienes in the treatment of asthma. Am J Respir Crit Care Med, 161, S73–S76. Barnes NC, Black B, Syrett N et al (1996) Reduction of exacerbations of asthma in multinational trials with zafirlukast (Accolate). Allergy, 51, 84.
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS Bensen WG, Fiechtner JJ, McMillen JI et al (1999) Treatment of osteoarthritis with celecoxib, a cyclooxygenase-2 inhibitor: a randomized controlled trial. Mayo Clin Proc, 74, 1095–1105. Bippi H and Frolich JC (1990) Effects of acetylsalicylic acid and paracetamol alone and in combination on prostanoid synthesis in man. Br J Clin Pharmacol, 29, 305–310. Bombardier C, Laine L, Reicin A et al (2000) Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med, 343, 1520– 1528. Boutaud O, Aronoff DM, Richardson JH et al (2002) Determinants of the cellular specificity of acetaminophen as an inhibitor of prostaglandin H(2) synthases. Proc Natl Acad Sci USA, 99, 7130–7135. Brater DG, Harris C, Redfern JF and Gertz BJ (2001) Renal effects of COX-2 selective inhibitors. Am J Nephrol, 21, 1–15. Brooks P, Emery P, Evans JF et al (1999) Interpreting the clinical significance of the differential inhibition of cyclooxygenase-1 and cyclooxygenase-2. Rheumatology, 38, 779–788. Burch JW, Stanford PW and Majerus PW (1979) Inhibition of platelet prostaglandin synthetase by oral aspirin. J Clin Invest, 61, 314–319. Cannon GW, Caldwell JR, Holt P et al (2000) Rofecoxib, a specific inhibitor of cyclooxygenase 2, with clinical efficacy comparable with that of diclofenac sodium: results of a one-year, randomized, clinical trial in patients with osteoarthritis of the knee and hip. Arthritis Rheum, 43, 978–987. Carter GW, Young PR, Albert DH et al (1991) Five-lipoxygenase inhibitory activity of zileuton. J Pharmacol Exp Ther, 256, 929–937. Catella-Lawson F, McAdam B, Morrison BW et al (1999) Effects of specific inhibition of cyclooxygenase-2 on sodium balance, hemodynamics, and vasoactive eicosanoids. J Pharmacol Exp Ther, 289, 735–741. Catella-Lawson F, Reilly MP, Kapoor SC et al (2001) Cyclooxygenase inhibitors and the antiplatelet effects of aspirin. N Engl J Med, 345, 1809–1817. Chan CC, Ecclestone P, Nicholson DW et al (1994) Leukotriene D4induced increases in cytosolic calcium in THP-1 cells: dependence on extracellular calcium and inhibition with selective leukotriene D4 receptor antagonist. J Pharmacol Exp Ther, 269, 891–896. Cheng Y, Austin SC, Rocca B et al (2002) Role of prostacyclin in the cardiovascular response to thromboxane A2. Science, 296, 539–541. Cipollone F, Ciabattoni G, Patrignani P et al (2000) Oxidant stress and aspirin-insensitive thromboxane biosynthesis in severe unstable angina. Circulation, 102, 1007–1013. Cipollone F, Patrignani P, Greco A et al (1997) Differential suppression of thromboxane biosynthesis by indobufen and aspirin in patients with unstable angina. Circulation, 96, 1109–1116. Cipollone F, Prontera C, Pini B et al (2001) Overexpression of functionally coupled cyclooxygenase-2 and prostaglandin E synthase in symptomatic atherosclerotic plaques as a basis of prostaglandin E2dependent plaque instability. Circulation, 104, 921–927. Clemett D and Goa KL (2000) Celecoxib: a review of its use in osteoarthritis, rheumatoid arthritis and acute pain. Drugs, 59, 957–980. Coleman RA, Eglen RM, Jones RL et al (1995) Prostanoid and leukotriene receptors: a progress report from the IUPHAR working parties on classification and nomenclature. Adv Prostagland Thrombox Leukotriene Res, 23, 283–285. Dahle´n SE, Bjork J, Hedqvist P et al (1981) Leukotrienes promote plasma leakage and leukocyte adhesion in postcapillary venules: in vivo effects with relevance to the acute inflammatory response. Proc Natl Acad Sci USA, 78, 3887–3891. Dahle´n SE, Hedqvist P, Hammarstrom S and Samuelsson B (1980) Leukotrienes are potent constrictors of human bronchi. Nature, 288, 484–486. Daniels SE, Desjardins PJ, Talwalker S et al (2002) The analgesic efficacy of valdecoxib vs. oxycodone/acetaminophen after oral surgery. J Am Dent Assoc, 133, 611–621. Day R, Morrison B, Luza A et al (2000) A randomized trial of the efficacy and tolerability of the COX-2 inhibitor rofecoxib vs ibuprofen in patients with osteoarthritis. Rofecoxib/Ibuprofen Comparatory Study Group. Arch Intern Med, 160, 1781–1787. Delanty N, Reilly M, Pratico` D et al (1996) 8-Epi PGF2a: specific analysis of an isoeicosanoid as an index of oxidant stress in vivo. Br J Clin Pharmacol, 42, 15–19.
609
Dennis EA (2000) Phospholipase A2 in eicosanoid generation. Am J Respir Crit Care Med, 161, S32–S35. Devchand PR, Keller H, Peters JM et al (1996) The PPARa-leukotriene B4 pathway to inflammation control. Nature, 384, 39–43. Dixon RA, Diehl RE, Opas E et al (1990) Requirement of a 5lipoxygenase-activating protein for leukotriene synthesis. Nature, 343, 282–284. Drazen JM, Israel E and O’Byrne PM (1999a) Treatment of asthma with drugs modifying the leukotriene pathway. N Engl J Med, 340, 197–206. Drazen JM, Yandava CN, Dube L et al (1999b) Pharmacogenetic association between ALOX5 promoter genotype and the response to anti-asthma treatment. Nature Genet, 22, 168–170. Dubb JW and Rebuck AS (1998) Pranlukast—the first orally active cysteinyl leukotriene receptor antagonist marketed for the treatment of asthma. In Drazen JM, Dahle´n E and Lee TH (eds), The lung in Health and Disease, vol 120. New York: Marcel Dekker, 347–363. Dunn MJ (2000) Are COX-2 selective inhibitors nephrotoxic? Am J Kidney Dis, 35, 976–977. Dunn MJ and Zambraski EJ (1980) Renal effects of drugs that inhibit prostaglandin synthesis. Kidney Int, 18, 609–622. Eikelboom JW, Hirsh J, Weitz JI et al (2002) Aspirin-resistant thromboxane biosynthesis and the risk of myocardial infarction, stroke, or cardiovascular death in patients at high risk for cardiovascular events. Circulation, 105, 1650–1655. Ehrich EW, Dallob A, De Lepeleire I et al (1999) Characterization of rofecoxib as a cyclooxygenase-2 isoform inhibitor and demonstration of analgesia in the dental pain model. Clin Pharmacol Ther, 65, 336– 347. Feinmark SJ and Cannon PJ (1986) Endothelial cell leukotriene C4 synthesis results from intercellular transfer of leukotriene A4 synthesized by polymorphonuclear leukocytes. J Biol Chem, 261, 16466–16472. Folco G, Rossoni G, Buccellati C et al (2000) Leukotrienes in cardiovascular diseases. Am J Respir Crit Care Med, 161, S112–S116. Ford-Hutchinson AW, Bray MA, Doig MV et al (1980) Leukotriene B, a potent chemokinetic and aggregating substance released from polymorphonuclear leukocytes. Nature, 286, 264–265. Ferreri NR, An SJ and McGiff JC (1999) Cyclooxygenase-2 expression and function in the medullary thick ascending limb. Am J Physiol, 277, F360–F368. FitzGerald GA and Patrono C (2001) The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med, 345, 433–442. FitzGerald GA, Smith B, Pedersen AK and Brash AR (1984) Increased prostacyclin biosynthesis in patients with severe atherosclerosis and platelet activation. N Engl J Med, 310, 1065–1068. Fitzpatrick FA, Liggett W, McGee J et al (1984) Metabolism of leukotriene A4 by human erythrocytes: a novel cellular source of leukotriene B4. J Biol Chem, 259, 11403–11407. Fitzpatrick FA and Soberman R (2001) Regulated formation of eicosanoids. J Clin Invest, 107, 1347–1351. Flower R, Gryglewski R, Herbaczynska-Cedro K and Vane JR (1972) Effects of antiinflammatory drugs on prostaglandin biosynthesis. Nature New Biol, 238, 104–106. Flower RJ and Vane JR (1972) Inhibition of prostaglandin synthetase in brain explains the anti-pyretic activity of paracetamol (4acetamidophenol). Nature, 240, 410–411. Fored CM, Ejerblad E, Lindblad P et al (2001) Acetaminophen, aspirin, and chronic renal failure. N Engl J Med, 345, 1801–1808. Forsberg L, Leeb L, Thoren S et al (2000) Human glutathione dependent prostaglandin E synthase: gene structure and regulation. FEBS Lett, 471, 78–82. Funk CD, Hoshiko S, Matsumoto T et al (1989) Characterization of the human 5-lipoxygenase gene. Proc Natl Acad Sci USA, 86, 2587–2591. Garcia Rodriguez LA and Hernandez-Diaz S (2001) The risk of upper gastrointestinal complications associated with non-steroidal antiinflammatory drugs, glucocorticoids, acetaminophen, and combinations of these agents. Arthritis Res, 3, 98–101. Garcia Rodriguez LA, Varas C and Patrono C (2000) Differential effects of aspirin and non-aspirin non-steroidal antiinflammatory drugs in the primary prevention of myocardial infarction in postmenopausal women. Epidemiology, 11, 382–387. Geba GP, Weaver AL, Polis AB et al (2002) Efficacy of rofecoxib, celecoxib, and acetaminophen in osteoarthritis of the knee: a
610
CONCLUSIONS AND CORRELATIONS
randomized trial. Vioxx, Acetaminophen, Celecoxib Trial (VACT) Group. J Am Med Assoc, 287, 64–71. Haeggstrom JZ (2000) Structure, function, and regulation of leukotriene A4 hydrolase. Am J Respir Crit Care Med, 161, S25–S31. Harris RC, McKanna JA, Akai Y et al (1994) Cyclooxygenase-2 is associated with the macula densa of rat kidney and increases with salt restriction. J Clin Invest, 94, 2504–2510. Hawkey C, Laine L, Simon T et al (2000) Comparison of the effect of rofecoxib (a cyclooxygenase 2 inhibitor), ibuprofen, and placebo on the gastroduodenal mucosa of patients with osteoarthritis: a randomized, double-blind, placebo-controlled trial. The Rofecoxib Osteoarthritis Endoscopy Multinational Study Group. Arthritis Rheum, 43, 370–377. Heeschen C and Hamm CW (2000) Difficulties with oral platelet glycoprotein IIb/IIIa receptor antagonists. Lancet, 355, 330–331. Heise CE, O’Dowd BF, Figueroa DJ et al (2000) Characterization of the human cysteinyl leukotriene 2 receptor. J Biol Chem, 275, 30531– 30536. Hemler ME and Lands WE (1980) Evidence for a peroxide-initiated free radical mechanism of prostaglandin biosynthesis. J Biol Chem, 255, 6253–6261. Henderson WR Jr (1994) The role of leukotrienes in inflammation. Ann Intern Med, 121, 684–697. Hinson JA, Pumford NR and Roberts DW (1995) Mechanisms of acetaminophen toxicity: immunochemical detection of drug–protein adducts. Drug Metab Rev, 27, 73–92. Israel E, Rubin P, Kemp JP et al (1993) The effect of inhibition of 5lipoxygenase by zileuton in mild-to-moderate asthma. Ann Intern Med, 119, 1059–1066. Jakobsson PJ, Thoren S, Morgenstern R and Samuelsson B (1999) Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA, 96, 7220–7225. Jones TR, Labelle M, Belley M et al (1995) Pharmacology of montelukast sodium (Singulair), a potent and selective leukotriene D4 receptor antagonist. Can J Physiol Pharmacol, 73, 191–201. Kargman S, Charleson S, Cartwright M et al (1996) Characterization of prostaglandin G/H synthase 1 and 2 in rat, dog, monkey, and human gastrointestinal tracts. Gastroenterology, 111, 445–454. Knapp HR (1990) Reduced allergen-induced nasal congestion and leukotriene synthesis with an orally active 5-lipoxygenase inhibitor. N Engl J Med, 323, 1745–1748. Knapp HR, Sladek K and FitzGerald GA (1992) Increased excretion of leukotriene E4 during aspirin-induced asthma. J Lab Clin Med, 119, 48–51. Kniss DA (1999) Cyclooxygenases in reproductive medicine and biology. J Soc Gynecol Investig, 6, 285–292. Konturek SJ, Obtulowicz W, Sito E et al (1981) Distribution of prostaglandins in gastric and duodenal mucosa of healthy subjects and duodenal ulcer patients: effects of aspirin and paracetamol. Gut, 22, 283–289. Krell RD, Aharony D, Buckner CK et al (1990) The preclinical pharmacology of ICI-204,219. A peptide leukotriene antagonist. Am Rev Respir Dis, 141, 978–987. Kulmacz RJ and Wang LH (1995) Comparison of hydroperoxide initiator requirements for the cyclooxygenase activities of prostaglandin H synthase-1 and -2. J Biol Chem, 270, 24019–24023. Kurumbail RG, Stevens AM, Gierse JK et al (1996) Structural basis for selective inhibition of cyclooxygenase-2 by antiinflammatory agents. Nature, 384, 644–648. Lam BK, Penrose JF, Freeman GJ and Austen KF (1994) Expression cloning of a cDNA for human leukotriene C4 synthase, an integral membrane protein conjugating reduced glutathione to leukotriene A4. Proc Natl Acad Sci USA, 91, 7663–7667. Laneuville O, Breuer DK, Dewitt DL et al (1994) Differential inhibition of human prostaglandin endoperoxide H synthase-1 and -2 by nonsteroidal antiinflammatory drugs. J Pharmacol Exp Ther, 271, 927– 934. Laviolette M, Malmstrom K, Lu S et al (1999) Montelukast added to inhaled beclomethasone in treatment of asthma. Montelukast/ Beclomethasone Additivity Group. Am J Respir Crit Care Med, 160, 1862–1868.
Lawson JA, Rokach J and FitzGerald GA (1999) Isoprostanes: formation, analysis and use as indices of lipid peroxidation in vivo. J Biol Chem, 274, 24441–24444. Leung AT, Malmstrom K, Gallacher AE et al (2002) Efficacy and tolerability profile of etoricoxib in patients with osteoarthritis: a randomized, double-blind, placebo and active-comparator controlled 12-week efficacy trial. Curr Med Res Opin, 18, 49–58. Lewis RA, Austen KF and Soberman RJ (1990) Leukotrienes and other products of the 5-lipoxygenase pathway. Biochemistry and relation to pathobiology in human diseases. N Engl J Med, 323, 645–655. Lofda¨hl CG, Reiss TF, Leff JA et al (1999) Randomised, placebocontrolled trial of effect of a leukotriene receptor antagonist, montelukast, on tapering inhaled corticosteroids in asthmatic patients. Br Med J, 319, 87–90. Loll PJ, Picot D and Garavito RM (1995) The structural basis of aspirin activity inferred from the crystal structure of inactivated prostaglandin H2 synthase. Nature Struct Biol, 2, 637–643. Luong C, Miller A, Barnett J et al (1996) Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nature Struct Biol, 3, 927–933. Lynch KR, O’Neill GP, Liu Q et al (1999) Characterization of the human cysteinyl leukotriene cys-LT1 receptor. Nature, 399(6738), 789–793. Maclouf JA, Foleo G and Patrono C (1998) Eicosanoids and isoeicosanoids: constitutive, inducible and transcellular biosynthesis in vascular disease. Thromb Haemost, 79, 691–705. Maclouf JA and Murphy RC (1988) Transcellular metabolism of neutrophil-derived leukotriene A4 by human platelets. A potential cellular source of leukotriene C4. J Biol Chem, 263, 174–181. Majerus PW (1983) Arachidonate metabolism in vascular disorders. J Clin Invest, 72, 1521–1525. Makarowski W, Zhao WW, Bevirt T and Recker DP (2002) Efficacy and safety of the COX-2 specific inhibitor valdecoxib in the management of osteoarthritis of the hip: a randomized, double-blind, placebocontrolled comparison with naproxen. Osteoarthr Cartilage, 10, 290– 296. Mancini JA, Blood K, Guay J et al (2001) Cloning, expression, and upregulation of inducible rat prostaglandin e synthase during lipopolysaccharide-induced pyresis and adjuvant-induced arthritis. J Biol Chem, 276, 4469–4475. Mancini JA, Riendeau D, Falgueyret JP et al (1995) Arginine 120 of prostaglandin G/H synthase-1 is required for the inhibition by nonsteroidal antiinflammatory drugs containing a carboxylic acid moiety. J Biol Chem, 270, 29372–29377. Marnett LJ, Rowlinson SW, Goodwin DC et al (1999) Arachidonic acid oxygenation by COX-1 and COX-2. Mechanisms of catalysis and inhibition. J Biol Chem, 274, 22903–22906. Masferrer JL, Reddy ST, Zweifel BS et al (1994) In vivo glucocorticoids regulate cyclooxygenase-2 but not cyclooxygenase-1 in peritoneal macrophages. J Pharmacol Exp Ther, 270, 1340–1344. McAdam BF, Catella-Lawson F, Mardini IA et al (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA, 96, 272–277. McFadden ER and Gilbert IA (1992) Asthma. N Engl J Med, 327, 1928– 1937. Meade EA, Smith WL and DeWitt DL (1993) Differential inhibition of prostaglandin endoperoxide synthase (cyclooxygenase) isozymes by aspirin and other non-steroidal antiinflammatory drugs. J Biol Chem, 268, 6610–6614. Mene` P, Pugliese F and Patrono C (1995) The effects of non-steroidal antiinflammatory drugs on human hypertensive vascular disease. Semin Nephrol, 15, 244–252. Metters KM (1995) Leukotriene receptors. J Lipid Mediat Cell Signal, 12, 413–427. Metters KM and Zamboni RJ (1993) Photoaffinity labeling of the leukotriene D4 receptor in guinea pig lung. J Biol Chem, 268, 6487– 6495. Mitchell JA, Akarasereenont P, Thiemermann C et al (1993) Selectivity of non-steroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA, 90, 11693–11697. Morrow JD, Hill KE, Burk RF et al (1990) A series of prostaglandin F2like compounds are produced in vivo in humans by a noncyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci USA, 87, 9383–9387.
CYCLOOXYGENASE AND LIPOXYGENASE INHIBITORS Murakami M, Nakatani Y, Atsumi G et al (1997) Regulatory functions of phospholipase A2. Crit Rev Immunol, 17, 225–283. Murakami M, Naraba H, Tanioka T et al (2000) Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem, 275, 32783–32792. Muth-Selbac US, Tegeder I, Brune K and Geisslinger G (1999) Acetaminophen inhibits spinal prostaglandin E2 release after peripheral noxious stimulation. Anesthesiology, 91, 231–239. Nayak AS, Anderson P, Kharous BL et al (1998) Equivalence of adding zafirlukast vs. double-dose inhaled corticosteroids in asthmatic patients symptomatic on low dose inhaled corticosteroids (abstr). J Allergy Clin Immunol, 101, S333. Narumiya S and FitzGerald GA (2001) Genetic and pharmacological analysis of prostanoid receptor function. J Clin Invest, 108, 25–30. O’Shaughnessy KM, Wellings R, Gillies B and Fuller RW (1993) Differential effects of fluticasone propionate on allergen-evoked bronchoconstriction and increased urinary leukotriene E4 excretion. Am Rev Respir Dis, 147, 1472–1476. Ouellet M and Percival MD (2001) Mechanism of acetaminophen inhibition of cyclooxygenase isoforms. Arch Biochem Biophys, 387, 273–280. Ouellet M, Riendeau D and Percival MD (2001) A high level of cyclooxygenase-2 inhibitor selectivity is associated with a reduced interference of platelet cyclooxygenase-1 inactivation by aspirin. Proc Natl Acad Sci USA, 98, 14583–14588. Panara MR, Padovano R, Sciulli MG et al (1998) Effects of nimesulide on costitutive and inducile prostanoid biosynthesis in human beings. Clin Pharmacol Ther, 63, 672–681. Panara MR, Renda G, Sciulli MG et al (1999) Dose-dependent inhibition of platelet cyclooxygenase-1 and monocyte cyclooxygenase-2 by meloxicam in healthy subjects. J Pharmacol Exp Ther, 290, 276–280. Patrignani P, Panara MR, Greco A et al (1994) Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther, 271, 1705–1712. Patrignani P, Sciulli MG, Manarini S et al (1999) COX-2 is not involved in thromboxane biosynthesis by activated human platelets. J Physiol Pharmacol, 50, 661–667. Patrono C, Ciabattoni G, Pinca E et al (1980) Low-dose aspirin and inhibition of thromboxane B2 production in healthy subjects. Thromb Res, 17, 317–327. Patrono C, Coller B, Dalen JE et al (2001a) Platelet-active drugs: the relationships among dose, effectiveness, and side effects. Chest, 119, 39S–63S. Patrono C, Patrignani P and Garcia Rodriguez LA (2001b) Cyclooxygenase-selective inhibition of prostanoid formation: transducing biochemical selectivity into clinical read-outs. J Clin Invest, 108, 7–13. Perazella MA and Eras J (2000) Are selective COX-2 inhibitors nephrotoxic? Am J Kidney Dis, 35, 937–940. Peters-Golden M and Brock TG (2001) Intracellular compartmentalization of leukotriene synthesis: unexpected nuclear secrets. FEBS Lett, 487, 323–326. Picot D, Loll PJ and Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature, 367, 243– 249. Pincus T, Swearingen C, Cummins P and Callahan LF (2000) Preference for non-steroidal antiinflammatory drugs vs. acetaminophen and concomitant use of both types of drugs in patients with osteoarthritis. J Rheumatol, 27, 1020–1027. Piper PJ (1984) Formation and actions of leukotrienes. Physiol Rev, 64, 744–761. Rahme E, Pilote L and LeLorier J (2002) Association between naproxen use and protection against acute myocardial infarction. Arch Intern Med, 162, 1111–1115. Ray WA, Stein CM, Hall K et al (2002) Non-steroidal antiinflammatory drugs and risk of serious coronary heart disease: an observational cohort study. Lancet, 359, 118–123. Reiss TF, Chervinsky P, Dockhorn RJ et al (1998) Montelukast, a oncedaily leukotriene receptor antagonist, in the treatment of chronic asthma: a multicenter, randomized, double-blind trial. Montelukast Clinical Research Study Group. Arch Intern Med, 158, 1213–1220.
611
Reiss TF, Sorkness CA, Stricker W et al (1997) Effects of montelukast (MK-0476); a potent cysteinyl leukotriene receptor antagonist, on bronchodilation in asthmatic subjects treated with and without inhaled corticosteroids. Thorax, 52, 45–48. Riendeau D, Percival MD, Brideau C et al (2001) Etoricoxib (MK-0663): preclinical profile and comparison with other agents that selectively inhibit cyclooxygenase-2. J Pharmacol Exp Ther, 296(2), 558–566. Robak J, Wieckowski A and Gryglewski R (1978) The effect of 4acetamidophenol on prostaglandin synthetase activity in bovine and ram seminal vesicle microsomes. Biochem Pharmacol, 27, 393–396. Roberts LJ II, Brame CJ, Chen Y and Morrow JD (1999) Novel eicosanoids. Isoprostanes and related compounds. Methods Mol Biol, 120, 257–285. Rocca B, Spain LM, Pure E et al (1999) Distinct roles of prostaglandin H synthases 1 and 2 in T cell development. J Clin Invest, 103, 1469–1477. Rocca B, Secchiero P, Ciabattoni G et al (2002) Cyclooxygenase-2 expression is induced during human megakaryopoiesis and characterizes newly formed platelets. Proc Natl Acad Sci USA, 99, 7634–7639. Rocha JL and Fernandez-Alonso J (2001) Acute tubulointerstitial nephritis associated with the selective COX-2 enzyme inhibitor, rofecoxib. Lancet, 357, 1946–1947. Roth GJ and Majerus PW (1975) The mechanism of the effect of aspirin on human platelets. I. Acetylation of a particulate fraction protein. J Clin Invest, 56, 624–632. Roth GJ, Stanford N and Majerus PW (1975) Acetylation of prostaglandin synthase by aspirin. Proc Natl Acad Sci USA, 72, 3073–3076. Rouzer CA and Samuelsson B (1985) On the nature of the 5-lipoxygenase reaction in human leukocytes: enzyme purification and requirement for multiple stimulatory factors. Proc Natl Acad Sci USA, 82, 6040–6044. Rouzer CA and Samuelsson B (1986) The importance of hydroperoxide activation for the detection and assay of mammalian 5-lipoxygenase. FEBS Lett, 204, 293–296. Sala A, Bolla M, Zarini S et al (1996) Release of leukotriene A4 vs. leukotriene B4 from human polymorphonuclear leukocytes. J Biol Chem, 271, 17944–17948. Samuelsson B (1983) Leukotrienes: mediators of immediate hypersensitivity reactions and inflammation. Science, 220, 568–575. Santini G, Patrignani P, Sciulli MG et al (2001) The human pharmacology of monocyte cyclooxygenase 2 inhibition by cortisol and synthetic glucocorticoids. Clin Pharmacol Ther, 70, 475–483. Sarau HM, Ames RS, Chambers J et al (1999) Identification, molecular cloning, expression, and characterization of a cysteinyl leukotriene receptor. Mol Pharmacol, 56, 657–663. Schumacher HR Jr, Boice JA, Daikh DI et al (2002) Randomised double blind trial of etoricoxib and indometacin in treatment of acute gouty arthritis. Br Med J, 324, 1488–1492. Scoggan KA, Jakobsson PJ and Ford-Hutchinson AW (1997) Production of leukotriene C4 in different human tissues is attributable to distinct membrane bound biosynthetic enzymes. J Biol Chem, 272, 10182– 10187. Silverstein FE, Faich G, Goldstein JL et al (2000) Gastrointestinal toxicity with celecoxib vs non steroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS Study: a randomized controlled trial. Celecoxib Long-term Arthritis Study. J Am Med Assoc, 284, 1247–1255. Simon LS, Lanza FL, Lipsky PE et al (1998) Preliminary study of the safety and efficacy of SC-58635, a novel cyclooxygenase 2 inhibitor: efficacy and safety in two placebo-controlled trials in osteoarthritis and rheumatoid arthritis, and studies of gastrointestinal and platelet effects. Arthritis Rheum, 41, 1591–1602. Simon LS, Weaver AL, Graham DY et al (1999) Antiinflammatory and upper gastrointestinal effects of celecoxib in rheumatoid arthritis: a randomized controlled trial. J Am Med Assoc, 282, 1921–1928. Smith CM, Christie PE, Hawksworth RJ et al (1991) Urinary leukotriene E4 levels after allergen and exercise challenge in bronchial asthma. Am Rev Respir Dis, 144, 1411–1413. Smith WL, Garavito RM and DeWitt DL (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem, 271, 33157–33160. Solomon DH, Glynn RJ, Levin R and Avorn J (2002) Non-steroidal antiinflammatory drug use and acute myocardial infarction. Arch Intern Med, 162, 1099–1104.
612
CONCLUSIONS AND CORRELATIONS
Smith WL and Langenbach R (2001) Why there are two cyclooxygenase isozymes. J Clin Invest, 107, 1491–1495. Spector SL, Smith LJ and Glass M (1994) Effects of 6 weeks of therapy with oral doses of ICI 204,219, a leukotriene D4 receptor antagonist, in subjects with bronchial asthma. ACCOLATE Asthma Trialists Group. Am J Respir Crit Care Med, 150, 618–623. Stichtenoth DO, Thoren S, Bian H et al (2001) Microsomal prostaglandin E synthase is regulated by proinflammatory cytokines and glucocorticoids in primary rheumatoid synovial cells. J Immunol, 167, 469–474. Tanioka T, Nakatani Y, Semmyo N et al (2000) Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem, 275, 32775–32782. Talley JJ, Brown DL, Carter JS et al (2000) 4-[5-Methyl-3-phenylisoxazol4-yl]-benzenesulfonamide, valdecoxib: a potent and selective inhibitor of COX-2. J Med Chem, 43, 775–777. Tornhamre S, Sjolinder M, Lindberg A et al (1998) Demonstration of leukotriene-C4 synthase in platelets and species distribution of the enzyme activity. Eur J Biochem, 251, 227–235. Ueno N, Murakami M, Tanioka T et al (2001) Coupling between cyclooxygenase, terminal prostanoid synthase, and phospholipase A2. J Biol Chem, 276, 34918–34927. Vane JR (1971) Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature New Biol, 231, 232–235. Vane JR (1994) Towards a better aspirin. Nature, 367, 215–216. Van Hecken A, Schwartz JI, Depre M et al (2000) Comparative inhibitory activity of rofecoxib, meloxicam, diclofenac, ibuprofen, and naproxen on COX-2 vs. COX-1 in healthy volunteers. J Clin Pharmacol, 40, 1109–1120. Vio CP, Cespedes C, Gallardo P and Masferrer JL (1997) Renal identification of cyclooxygenase-2 in a subset of thick ascending limb cells. Hypertension, 30, 687–692. Virchow JC, Hassall SM and Harris A (1997a) Improved asthma control over 6 weeks with zafirlukast in patients on high-dose inhaled corticosteroids (abstr). Eur Respir J, 10, 437S.
Virchow JC, Hassall SM, Summerton L et al (1997b) Reduction of asthma exacerbations with zafirlukast in patients on inhaled corticosteroids (abstr). Eur Respir J, 10, 420S. Walker MC, Kurumbail RG, Kiefer JR et al (2001) A three-step kinetic mechanism for selective inhibition of cyclo-oxygenase-2 by diarylheterocyclic inhibitors. Biochem J, 357, 709–718. Watson DJ, Rhodes T, Cai B and Guess HA (2002) Lower risk of thromboembolic cardiovascular events with naproxen among patients with rheumatoid arthritis. Arch Intern Med, 162, 1105–1110. Wenzel SE, Larsen GL, Johnston K et al (1990) Elevated levels of leukotriene C4 in bronchoalveolar lavage fluid from atopic asthmatics after endobronchial allergen challenge. Am Rev Respir Dis, 142, 112– 119. Wetterholm A, Medina JF, Radmark O et al (1991) Recombinant mouse leukotriene A4 hydrolase: a zinc metalloenzyme with dual enzymatic activities. Biochim Biophys Acta, 1080, 96–102. Whelton A (2001) Renal aspects of treatment with conventional nonsteroidal antiinflammatory drugs vs. cyclooxygenase-2-specific inhibitors. Am J Med, 110, 33S–42S. Wight NJ, Gottesdiener K, Garlick NM et al (2001) Rofecoxib, a COX-2 inhibitor, does not inhibit human gastric mucosal prostaglandin production. Gastroenterology, 120, 867–873. Wolfe F, Zhao S and Lane N (2000) Preference for non-steroidal antiinflammatory drugs over acetaminophen by rheumatic disease patients: a survey of 1799 patients with osteoarthritis, rheumatoid arthritis, and fibromyalgia. Arthritis Rheum, 43, 378–385. Wu G, Wei C, Kulmacz RJ et al (1999) A mechanistic study of selfinactivation of the peroxidase activity in prostaglandin H synthase-1. J Biol Chem, 274, 9231–9237. Yamagata K, Andreasson KI, Kaufmann WE et al (1993) Expression of a mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron, 11, 371–386. Yokomizo T, Izumi T, Chang K et al (1997) A G-protein-coupled receptor for leukotriene B4 that mediates chemotaxis. Nature, 387, 620–624. Yokomizo T, Kato K, Terawaki K et al (2000) A second leukotriene B4 receptor, BLT2. A new therapeutic target in inflammation and immunological disorders. J Exp Med, 192, 421–432.
57 Pharmaceutical Exploitation: Eicosanoids and their Analogues David F. Woodward and June Chen Allergan Inc., Irvine, CA, USA
The discovery of the prostanoids (Kurzrok and Lieb 1930; Goldblatt 1933; von Euler 1934) and their structural elucidation occurred many years ago (Bergstro¨m and Sjo¨vall 1957; Hamberg and Samuelsson 1973; Hamberg et al 1975; Moncada et al 1976; Moncada and Vane 1980) but their therapeutic potential has been incompletely realized. Prostanoids remain attractive drug targets as they often produce profound biological effects. Since the prostanoids are local hormones with a wide distribution, the potential for side-effects associated with systemic administration is high. Thus, most prostanoid-based therapies to date rely on local drug administration. The development of a prostanoid receptor classification (Kennedy et al 1982) and the subsequent structural elucidation of the receptors (Coleman et al 1994) has facilitated a more informed approach to prostanoid drug design. This will likely lead to the emergence of further successful therapeutic modalities utilizing the prostanoid technology base.
MALE ERECTILE DYSFUNCTION Prostaglandins (PGs) were originally discovered in human seminal fluid (Kurzrok and Lieb 1930; Goldblatt 1933; von Euler 1934) and the natural prostaglandins found early and continued use for indications generally related to reproduction. Prostaglandin E1 PGE1) is a naturally occurring prostaglandin that is biosynthesized from eicosatrienoic acid (Rhyage and Samuelsson 1965; Nugteren and von Dorp 1965) and interacts with both EP and IP receptor subtypes (Funk et al 1993; Regan et al 1994a, 1994b; Bastien et al 1994; Miller and Gorman 1979). It has been suggested that prostacyclin (PGI2) plays a role in the penile erection process, thereby providing rational therapy for male impotence (Jeremy et al 1986). PGE1 is available under the trade name alprostadil in various forms. Alprostadil is given locally for the treatment of male impotence due to erectile dysfunction. This results in a dose-dependent erectile response (Linet et al 1996) and titration studies have been recommended before patients undertake self-injection therapy (Krane et al 1989). Alprostadil has been used as an alternative to, and in combination with, the vasodilator papaverine and the non-selective a-adrenoceptor antagonist phentolamine (Waldhauser and Schramek 1988; Lee et al 1989). Fibrosis may occur as a result of repeated intracavernosal injection. Side-effects specifically related to alprostadil include penile pain and burning (Waldhauser and Schramek 1988). Systemic side-effects may also occur (Krane et al 1989) but nevertheless alprostadil appears to be a popular choice for selfinjection therapy (Meghdadi et al 1999). Intrauretheral supposiThe Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
tories have been explored as an alternative to intracavernosal injection. The required dose of alprostadil is higher and the overall success rate for suppositories is about 50% (PadmaNathan et al 1997; Guay et al 2000). In addition to therapy, alprostadil is also employed as a diagnostic tool for male impotence. Alprostadil may assist in identifying haemodynamic impairment as the origin of male impotence. The initiation of penile erection involves reflexogenic mechanisms elicited by local stimuli, which are modulated by psychogenic stimuli (Krane et al 1989; Andersson 2001). Erection of the penis involves neurogenic and endothelium-mediated control of penile smooth muscle (Krane et al 1989; Andersson 2001). Dilatation of the cavernosal and helicine arteries leads to increased blood flow to the lacunar spaces, which are dilated as a result of concomitant relaxation of the trabecular smooth muscle. The increased blood pressure transmitted via the dilated helicine arteries expands the trabecular walls. In turn, the subtunical venules are compressed, which reduces venous outflow in the lacunar space. Thus, intralacunar pressure is elevated and the penis becomes rigid. The development of an erection in response to alprostadil would imply a normal vascular status and impotence of neurological or psychological origin would require consideration. The future of locally administered agents to achieve penile erection may be uncertain. Early reports on oral sildenafil show clear superiority in terms of convenience and naturalness of the erection (Montorsi et al 2001).
TERMINATION OF PREGNANCY Prostaglandins appear to play a prominent role in the uterine contraction and cervical ripening associated with the induction of labour. All receptor subtypes originally described by the original prostanoid receptor classification (Coleman et al 1994) are represented in the human myometrium (Senior et al 1991, 1992, 1993; Fernandes and Crankshaw 1995; Brodt-Eppley and Myatt 1999; Milne et al 2001). Receptors mediating contraction (EP3, FP and TP) are present in the upper uterine segment and these would generate expulsive forces when stimulated (Senior et al 1993). The cervix allows passage of the foetus by becoming softened and dilated. Cervical ripening involves a series of biochemical changes that result in the disintegration and dispersion of collagen fibres and an increase in ground substance. The softened and dilated cervix may be readily further dilated by uterine contraction. Cervical ripening and dilatation can be induced independent of
614
CONCLUSIONS AND CORRELATIONS
uterine motility and this provides the basis for the use of prostaglandins in pregnancy termination. Prostaglandins were introduced in 1972 (Wiqvsit et al 1972) as pretreatment to dilate the cervix before vacuum aspiration. This procedure is commonly employed for the termination of first trimester and early second trimester pregnancies. Prostaglandin pretreatment facilitates vacuum aspiration and helps avoid the risk of damage to the cervix or uterus that may occur in later pregnancies with mechanical cervical dilatation methods. Such damage could lead to cervical incompetence and spontaneous abortion in future pregnancies. Prostaglandins may be used in conjunction with a progestogen antagonist, which serves to sensitize the uterus to prostaglandins. Initial studies involved the natural prostaglandins PGF2a and PGE2. Prostaglandins have been given by the vaginal, intramuscular, intravenous and, in cases of foetal death in utero, by the intraamnial route. Vaginal administration is now the preferred route. PGF2a is commercially available as dinoprost. Although it is effective at any stage of the gestation period, dinoprost is not used routinely for the induction of labour because of a higher incidence of side-effects compared to PGE2. Pharmaceutical formulations of PGE2 are named dinoprostone. The local administration of dinoprostone, typically as a gel, has proved to be an effective and tolerable method for preoperative cervical ripening (Uldbjerg et al 1983; Wingerup et al 1981). Side-effects with natural prostaglandins include diarrhoea, nausea, vomiting and uterine pain. Such adverse events may be minimized by the use of structural PG analogues. These include the PGE analogues gemeprost, misoprostol, enisoprost sulprostone and carbaprost, which is 15-methyl PGF2a. Gemeprost (Christensen et al 1983; Christensen and Bygdeman 1984) and misoprostol (Goldberg et al 2001) provide effective pretreatment for vacuum aspiration, with minimal side-effects. Gemeprost is thermolabile and requires refrigerated storage (Christin-Maitre et al 2000). Misoprostol has been reported to be more effective than both dinoprostone (Varaklis et al 1995) and gemeprost (Bartley et al 2001). Sulprostone is no longer used as an abortifacient due to a link with increased cardiovascular complications (Ulmann et al 1992). Enisoprost, administered orally, has also been reported to induce uterine activity and soften the cervix during first trimester pregnancy (Pukkinen et al 1989). Oral administration is not, however, favoured for prostaglandins because of increased sideeffects. The presence and potent effects of prostaglandins on human female reproductive tissue suggest possible future uses for receptor-selective PG analogues. Several substances that increase myometrial responsiveness converge upon a final common pathway of prostaglandin production (Gillin 1994). Thus, prostaglandin-based therapies may be employed for prevention of pre-term labour (Senior et al 1993), dysmenorrhoea and endometriosis (Benedetto 1989). CARDIOVASCULAR IMPLICATIONS Alprostadil and dinoprostone are used to treat neonates with congenital heart disease by maintaining the patency of the ductus arteriosus before surgery. The ductus arteriosus connects the main pulmonary artery to the descending aorta and serves as a shunt to the aorta in the foetal circulation. The ductus arteriosus closes just after birth as an adaptation to respiration. Several side-effects may occur in neonates treated with alprostadil for congenital heart defects (Lewis et al 1981). Dinaprostone may be dosed orally for 4 weeks and is beneficial in allowing more time for infant growth before surgery (Silove et al 1985; Thanopoulus et al 1987). Alprostadil and dinoprostone may also be administered intravenously.
Although natural prostaglandins are currently employed for maintaining the ductus arteriosus, selective agonists may represent safer future therapy. In mice where the gene encoding the EP4 receptor was disrupted, the ductus arteriosus remained patent (Segi et al 1998). Thus, selective EP4 agonists may provide more targeted therapy. Prostaglandins have also found utility in treating Raynaud’s syndrome. Primary Raynaud’s phenomenon involves vasospasm of the digits and other parts of the body, which may even include visceral organs (Belch and Ho 1996). Raynaud’s phenomenon also occurs secondary to other disorders, notably connective tissue disease. Raynaud’s phenomenon often results from systemic atherosclerotic obstructive arteriolar disease (Belch and Ho 1996). Microcirculatory flow may also be impeded by vasospasm and is frequently triggered by cold. Platelet aggregation and clumping may play a further role and, consequently, treatments involving vasodilatation and inhibition of platelet aggregation have been clinically tested but a statistically significant benefit was not obtained (Lau et al 1993; Belch et al 1995). Thus, iloprost (Belch et al 1995) and cicaprost (Lau et al 1993), which are stable prostacyclin mimetics, did not entirely meet therapeutic expectations when administered orally. A synthetic PGE1 analogue, limaprost, has also been found efficacious in the treatment of Raynaud’s syndrome (Murai et al 1989). PGE1 has also been used to treat peripheral arterial occlusive disease. Clear improvement was apparent following intraarterial PGE1 (Tru¨bestein et al 1987). Prostacyclin (PGI2) has also been tried for peripheral vascular disease. The prostacyclin mimetic iloprost has also been employed for treating severe Raynaud’s phenomenon associated with scleroderma (Stratton et al 2001). Prostacyclin is supplied as the sodium salt and is commercialized under the name epoprostenol. It has also been used to prevent platelet aggregation in blood from patients undergoing kidney dialysis. Continuous prostacyclin infusion, in low doses, has also been reported to preserve renal function in high-risk coronary bypass patients (Morgera et al 2002). Epoprostenol has been reported to produce some encouraging clinical improvement in pulmonary hypertension (Barst et al 1996; Higenbottam et al 1998). Due to its instability, epoprostenol must be administered by intravenous infusion. Inhalation has been successfully employed as an alternative route of administration in adults with primary and secondary pulmonary hypertension (Olschewski et al 1996). GASTROINTESTINAL IMPLICATIONS Stimulation of the PGE2-sensitive EP3 receptor results in inhibition of gastric acid secretion, with resultant protection of the mucosal barrier (Whittle 1976). Misoprostol has been shown to be effective in treating both gastric and duodenal ulcers, although it is certainly not more efficacious than histamine H2receptor antagonists (Rachmilewitz et al 1986). Misoprostol is commercially available as cytotec and arthotec. Misoprostol may be of value in ulcer prevention in patients receiving chronic nonsteroidal antiinflammatory (NSAID) therapy. For this reason, arthotec comprises misoprostol and a NSAID. The utility of misoprostol is limited in treating gastroduodenal ulceration by its side-effects, the most common side-effects being diarrhoea and abdominal pain. Misoprostol and MB 28767 are EP3 agonists that also exhibit activity at EP2 (Lawrence et al 1992; Lawrence and Jones 1992) and TP receptors (Banerjee et al 1985; Lawrence and Jones 1992; Woodward et al 1995). GR 63799 is a potent and selective EP3 agonist that appears to have been developed as a gastric cytoprotective agent (Bunce et al 1990). GR 63799 was never commercialized, presumably because of side-effects.
EICOSANOIDS AND THEIR ANALOGUES FUTURE OPPORTUNITIES The advent of the prostanoid receptor classification and the availability of human recombinant receptors have stimulated, and continue to stimulate, interest in prostanoids as drugs. Future opportunities are not entirely within the intended scope of this review and are confined to two examples, osteoporosis and PPAR receptors as future targets. A promising future use of prostaglandin agonists is in the treatment of osteoporosis. Systemic administration of natural prostaglandins has long been known to increase bone mass and this has been reported in neonates treated with PGE1 to maintain patency of the ductus arteriosus (Ueda et al 1980; Nadroo et al 2000). In the context of the current classification for prostanoid receptors, EP2, EP4 and FP receptors have attracted attention for designing drugs to treat osteoporosis. EP2 and FP receptors offer agonist drug targets (Hartke and Lundy 2001). The FP receptor also appears to represent a promising target for the design of anabolic agents for treating osteoporosis. The proposal that 15-deoxy-D12,14 PGJ2 is an endogenous ligand for the isoform of the peroxisome proliferator-activated receptor (PPARg) opened up the nuclear receptor superfamily as a new direction for prostanoid-based drug design. Fibroblast differentiation into adipocytes is induced by PPARg expression and, consequently, this receptor is a potential target for antiobesity drugs. The insulin-sensitizing antidiabetic drugs, the thiazolidinediones, also selectively activate PPARg, suggesting that 15-deoxyD12,14 PGJ2 may serve as an endogenous regulator of insulin activity (Forman et al 1995). It has further been suggested that different prostaglandins may promote or attenuate adipogenesis through opposing effects on PPARg. Thus, 15-deoxy-D12,14 PGJ2 functions as a PPARg activator, whereas PGF2a inhibits adipogenesis by a MAP kinase-mediated phosphorylation mechanism (Serrero and Lepak 1997; Reginato et al 1998). This again indicates further potential therapeutic utility for FP receptor agonists and identifies PGJ2 and its serum-catalysed metabolic products as a new template for antiobesity and antidiabetic drug design.
REFERENCES Andersson K-E (2001) Pharmacology of penile erection. Pharmacol Rev, 53, 417–450. Banerjee AK, Tuffin DP and Walker JL (1985) Pharmacological effects of (+)-11-deoxy-16-phenoxy prostaglandin E1 derivatives in the cardiovascular system. Br J Pharmacol, 84, 71–80. Barst RJ, Rubin LJ, Long WA et al (1996) A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med, 334, 296–301. Bartley J, Brown A, Elton R and Baird DT (2001) Double-blind randomized trial of mifepristone in combination with vaginal gemeprost or misoprostol for induction of ‘‘abortion’’ up to 63 days gestation. Hum Reprod, 16, 2098–2102. Bastien L, Sawyer N, Grygorczyk R et al (1994) Cloning, functional expression, and characterization of the human prostaglandin E2 receptor EP2 subtype. J Biol Chem 269, 11873–11877. Belch JJ, Capell HA, Cooke ED et al (1995) Oral iloprost as a treatment for Raynaud’s syndrome: a double-blind multicenter placebo controlled study. Ann Rheum Dis, 54, 197–200. Belch JJ and Ho M (1996) Pharmacotherapy of Raynaud’s phenomenon. Drugs, 52, 682–695. Benedetto C (1989) Eicasanoids in primary dysmenorrhea, endometriosis, and menstrual migraine. Gynecol Endocrinol, 3, 71–94. Bergsto¨m S and Sjo¨vall J (1957) The isolation of prostaglandin. Acta Chem Scand, 11, 1086. Brodt-Eppley J and Myatt L (1999) Prostaglandin receptors in lower segment myometrium during gestation and labor. Obstet Gynecol, 93, 89–93.
615
Bunce KT, Clayton NM, Coleman RA et al (1990) GR 63799—a novel prostamoid with selectivity for EP3 receptors. Adv Prostagland Thrombox Leukotriene Res, 21, 379–382. Christensen NJ, Bygdeman M and Green K (1983) Comparison of different prostaglandin analogues and laminaria for preoperative dilatation of the cervix in late first trimester abortion. Contraception, 27, 51–61. Christensen NJ and Bygdeman M (1984) Cervical dilatation with 16, 16dimethyl-trans-D2 PGE1 methyl ester (Cervagem) prior to vacuum aspiration. A double-blind, placebo-controlled randomized study. Contraception, 29, 457–464. Christin-Maitre S, Bouchard P and Spitz IM (2000) Medical termination of pregnancy. N Engl J Med, 342, 946–956. Coleman RA, Smith WL and Narumiya S (1994) VIII International Union of Pharmacology classification of prostanoid receptors: properties, distribution and structure of the receptors and their subtypes. Pharmacol Rev, 46, 205–229. Fernandes B and Crankshaw D (1995) Functional characterization of the prostanoid DP receptor in human myometrium. Eur J Pharmacol, 283, 73–81. Forman BM, Tontonoz P, Chen J et al (1995) 15-deoxy-D12,14prostaglandin J2 is a ligand for the adipocyte determination factor PPARg. Cell, 83, 803–812. Funk CD, Furci L, Fitzgerald GA et al (1993) Cloning and expression of a cDNA for the human prostaglandin E receptor EP1 subtype. J Biol Chem, 268, 26767–26772. Gillin AG (1994) Maintenance of high risk pregnancies: role of prostaglandins and other mediators. Aust NZ Obstet Gynecol, 34, 351–356. Goldberg AB, Greenberg MB and Darney PD (2001) Misoprostol and pregnancy. N Engl J Med, 344, 38–47. Goldblatt MW (1933) A depressor substance in seminal fluid. J Soc Chem Indust (Lond), 52, 1056–1057. Guay AT, Perez JB, Velasquez E et al (2000) Clinical experience with intraurethral alprostadil (MUSE1) in the treatment of men with erectile dysfunction. Eur Urol, 38, 671–676. Hamberg M and Samuelsson B (1973) Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Natl Acad Sci USA, 70, 899–903. Hamberg M, Hedquist P, Strandberg K et al (1975) Prostaglandin endoperoxides iv. Effects on smooth muscle. Life Sci, 16, 451–462. Hartke JR and Lundy MW (2001) Bone anabolic therapy with selective prostaglandin analogs. J Musculoskel Neuron Interact, 2, 25–31. Higenbottam T, Butt AY, McMahon A et al (1998) Long-term intravenous prostaglandin (epoprostenol or iloprost) for treatment of severe pulmonary hypertension. Heart, 80, 151–155. Jeremy JY, Morgan RJ, Mikhailidis et al (1986) Prostacyclin, synthesis by the corpora cavernosa of the human penis: evidence for muscarinic control and pathological implications. Prostaglandins Leukot Med, 23, 211–216. Kennedy I, Coleman RA, Humphrey PPA et al (1982) Studies on the characterization of prostanoid receptors; a proposed classification. Prostaglandins, 24, 667–689. Kennedy CRJ, Zhang Y, Brandon S et al (1999) Salt-sensitive hypertension and reduced fertility in mice lacking the prostaglandin EP2 receptor. Nature Med, 5, 217–220. Krane RJ, Goldstein I and Saenz de Tejada (1989) Impotence. N Engl J Med, 321, 1648–1659. Kurzrok R and Lieb CC (1930) Biochemical studies on human semen II. The action of semen on the human uterus. Proc Soc Exp Biol Med, 28, 268–272. Lau CS, Belch JJ, Cappell H et al (1993) A randomized, double-blind study of cicaprost, an oral prostacyclin analogue, in the treatment of Raynaud’s phenomenon secondary to systemic sclerosis. Clin Exp Rheumatol, 11, 35–40. Lawrence RA and Jones RL (1992) Investigation of the prostaglandin E (EP-) receptor subtype mediating relaxation of the rabbit jugular vein. Br J Pharmacol, 63, 1339–1346. Lawrence RA, Jones RL and Wilson NH (1992) Characterization of receptors involved in the direct and indirect actions of prostaglandin E and I on the guinea-pig ileum. Br J Pharmacol, 105, 271–278. Lee LM, Stevenson RWD and Szasz G (1989) Prostaglandin E1 vs. phentolamine/papaverine for the treatment of erectile impotence: a double-blind comparison. J Urol, 141, 549–550.
616
CONCLUSIONS AND CORRELATIONS
Lewis AB, Greed MD, Heymann MA et al (1981) Side-effects of therapy with prostaglandin E1 in infants with critical congenital heart disease. Circulation, 64, 238–246. Linet OI, Ogrinc FG and the Alprostadil Study Group (1996) Efficacy and safety of intracavernosal journal of alprostadil in men with erectile dysfunction. N Engl J Med, 334, 873–877. Meghdadi S, Porst H, Stackl W et al (1999) Presence of PGE1 binding determines the erectile response to PGE1. Prostagland Leukotrienes Essent Fatty Acids, 60, 111–113. Miller OV and Gorman RR (1979) Evidence for distinct prostaglandin I2 and D2 receptors in human platelets. J Pharmacol Exp Therapeut, 210, 134–140. Milne SA, Perchick GB, Boddy SC and Jabbour HN (2001) Expression, localization and signaling of PGE2 and EP2/EP4 receptors in human non-pregnant endometrium across the menstrual cycle. J Clin Endocrinol Metab, 86, 4453–4459. Moncada S, Gryglewski RJ, Bunting S and Vane JR (1976) An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature (Lond), 263, 663–665. Moncada S and Vane JR (1980) Biological significance and therapeutic potential of prostacyclin. J Med Chem, 23, 591–593. Montorsi F, Althof S, Orazem J and Giulliano F (2001) Comparison of satisfaction between prostaglandin E1 injections and oral viagra (sildenafil citrate) in patients with erectile dysfunction. Eur Urol, 39, 3. Morgera S, Woydt R, Kern H et al (2002) Low-dose prostacyclin preserves renal function in high-risk patients after coronary bypass surgery. Crit Care Med, 30, 107–112. Murai C, Sasaki T, Osaki H et al (1989) Limaprost for Raynaud’s phenomenon. Lancet, Nov, 1218. Nadroo AM, Shringari S, Garg M and Sowailen AM (2000) Prostaglandin induced cortical hyperostosis in neonates with cyanotic heart disease. J Perinat Med, 28, 447–452. Nugteren DH and van Dorp DA (1965) The participation of molecular oxygen in the biosynthesis of prostaglandins. Biochim Biophys Acta, 98, 654–656. Olschewski H, Walmrath D, Schermuly R et al (1996) Aerosolized prostacyclin and iloprost in severe pulmonary hypertension. Ann Intern Med, 124, 820. Padman-Nathan H, Hellstro¨m WJ, Kaiser FE et al and the Medicated Urethral System for Erection (MUSE1) Study Group (1997) Treatment of men with erectile dysfunction with transurethral alprostadil. N Engl J Med, 336, 1–7. Pukkinen MO, Cohen H and Vuento M (1989) Oral administration of the prostaglandin enisoprost induces uterine activity and softens the cervix during first trimester pregnancy. Acta Obstet Gynecol Scand, 68, 149– 152. Rachimlewitz D, Chapman JW, Nicholson PA and the International Study Group (1986) A multi-center international controlled comparison of two dosage regimens of misoprostol with cimetidine in treatment of gastric ulcer in outpatients. Dig Dis Sci, 31, 755–805. Regan JW, Bailey TJ, Donello JE et al (1994a) Molecular cloning and expression of human EP3 receptors: evidence of three variants with different carboxyl termini. Br J Pharmacol, 112, 377–385. Regan JW, Bailey TJ, Pepperl DJ et al (1994b) Cloning of a novel human prostaglandin receptor with characteristics of the pharmacologically defined EP2 subtype. Mol Pharmacol, 46, 213–220. Reginato MJ, Krakow SL, Bailey ST and Lazar MA (1998) Prostaglandins promote and block adipogenesis through opposing effects on peroxisome proliferator-activated receptor-g. J Biol Chem, 273, 1855–1858.
Rhyage R and Samuelsson B (1965) Origin of oxygen incorporated during biosynthesis of prostaglandin E1. Biochem Biophys Res Commun, 19, 279–282. Segi E, Sugimoto Y, Yamasaki A et al (1998) Patent ductus anteriosus and neonatal death in prostaglandin receptor EP4-deficient mice. Biochem Biophys Res Commun, 246, 7–12. Senior J, Marshall K, Sangha R et al (1991) In vitro characterization of prostanoid EP receptors in the non-pregnant human myometrium. Br J Pharmacol, 102, 747–753. Senior J, Marshall K, Baxter GS and Clayton JK (1992) In vitro characterization of prostanoid FP-, DP-, IP-, and TP-receptors on the non-pregnant human myometrium. Br J Pharmacol, 107, 215–221. Senior J, Marshall K, Sangha R and Clayton JK (1993) In vitro characterization of prostanoid receptors on human myometrium at term pregnancy. Br J Pharmacol, 108, 501–506. Serrero G and Lepak NM (1997) Prostaglandin F2a receptor (FP receptor) agonists are potent adipose differentiation inhibitors for primary culture of adipocyte precursors in defined medium. Biochem Biophys Res Commun, 233, 200–202. Silove ED, Roberts DGV and De Giovanni JV (1985) Evaluation of oral and low-dose intravenous prostaglandin E2 in management of ductus dependent congenital heart disease. Arch Dis Child, 60, 1025–1030. Stratton R, Shiwen X, Martini G et al (2001) Iloprost suppresses connective tissue growth factor production in fibroblasts and in the skin of scleroderma patients. J Clin Invest, 108, 241–250. Thanopoulos, AA and Frimas C (1987) Prostaglandin E2 administration in infants with ductus-dependent cyanotic congenital heart disease. Eur J Pediat, 146, 279–282. Tru¨bestein G, Ludwig M, Diehm C et al (1987) Prostaglandin E1 bei arterieller Verschlusskrankeit in Stadium III and IV. Deutsch Med Wochenschr, 112, 955–959. Uldbjerg N, Ekman G, Herltoft P et al (1983) Human cervical connective tissue and its reaction to prostaglandin E2. Acta Obstet Gynecol Scand Suppl, 113, 163. Ueda K, Saito A, Nakamo H et al (1980) Cortical hyperostosis following long-term administration of prostaglandin E1 in infants with cyanotic heart disease. J Pediat, 97, 834–836. Ulmann A, Silvestre L, Chemma L et al (1992) Medical termination of early pregnancy with mifepristone (RU 486) followed by a prostaglandin analogue. Acta Obstet Gynecol Scand, 71, 278–283. Varaklis K, Gumina R and Stubblefield PG (1995) Randomized controlled trial of vaginal misoprostol and intracervical prostaglandin E2 gel for induction of labor at term. Obstet Gynecol, 86, 541–544. Von Euler US (1934) Zur Kenntnis der pharmakilogischen accesorischen geschlechtsdru¨sen. Arch Exp Pathol Pharmakol, 175, 78. Waldhauser M and Schramek (1988) Efficiency and side effects of prostaglandin E1 in the treatment of erectile dysfunction. J Urol, 140, 525–527. Whittle BJ (1976) Relationship between the prevention of rat gastric erosions and the inhibition of acid secretion by prostaglandins. Eur J Pharmacol, 40, 233–239. Wingerup L, Ulmsten M, Bygdeman G et al (1981) A new prostaglandin E2-gel for pretreatment of the cervix in nulliporous patients having a late first trimester termination of pregnancy. Arch Gynecol, 231, 1. Wiquist N, Beguin F, Bygdeman M et al (1972) Induction of abortion by extra-amniotic prostaglandin administration. Prostaglandins, 1, 37–53. Woodward DF, Fairbairn CE, Krauss AH-P et al (1995) Radioligand binding analysis of receptor subtypes in two FP receptor preparations that exhibit different functional rank orders of potency in response to prostaglandins. J Pharmacol Exp Therapeut, 273, 285–291.
Epilogue
Foremost among the various advances in scientific knowledge mentioned in the Preface was the identification of an alternative cyclooxygenase designated COX-2 (e.g. Masferrer et al 1990). Current thinking is that the activity of this inducible cyclooxygenase is amplified under inflammatory conditions and that it coexists with the now-classical COX-1 enzyme, which is normally involved with important ‘‘housekeeping’’ or constitutive functions in the gastrointestinal tract and renal system, where its inhibition risks adverse reactions. However, it has also been suggested that this thesis may be ingenuous, ‘‘. . . an oversimplification of the biologic reality . . .’’ (Fitzgerald and Patrono 2001) since there are a number of organs and tissues in which it is believed that COX-2 serves in a constituitive role (e.g. see this volume Chapter 34, Kadowitz et al). Nevertheless, substantial effort had been made by the pharmaceutical industry towards the development of new inhibitor products with COX-2 selectivity. The first two such compounds were Celebrex (celecoxib) from Pharmacia and Vioxx (rofecoxib) from Merck Sharp & Dohme and these have been followed by a second wave, with improved COX-2 selectivity, comprising etoricoxib (Merck) and valdecoxib (Pharmacia) and most recently a further new but different wave in the form of lumiracoxib (Novartis) is in the latter stages of development. These fascinating compounds and their evolution are dealt with in considerable detail in this volume in two complementary contributions: Chapter 16 by Rainsford, Inhibitors of Eicosanoids and Chapter 56 by Patrignani & Sciulli, Pharmaceutical Exploitation: Cyclooxygenase and Lipoxygenase Inhibitors. Also considered are certain new COX-2-selective NSAIDs (such as etodolac and numesulide) which have shown improved gastrointestinal safety, as well as novel compounds with dual cyclooxygenase-5-lipoxygenase (COX-LOX) inhibitory activity, nitric oxide (NO) donating cyclooxygenase (CIONOD) inhibitors (Rainsford 2001) and prodrugs such as nabumetone (Roth 2000), which prevents gastric damage through reducing neutrophil infiltration and activation (Ishiwata et al 2003), which may well prove themselves to be superior to both non-selective NSAIDs and the newer COX-2 selective Celebrex-type molecules, when they have been more widely evaluated. Currently there is no obvious choice for chronic antiinflammatory drug treatment in arthritis, but the ideal compound would definitely spare COX-1 activity in its important physiological roles. Meanwhile, it should be mentioned that contrary to the varied beneficial aspects of NSAIDs on human and animal health, a very recent report concludes that aspirin may be associated with an increased risk of pancreatic cancer among women (Schernhammer 2004). A third variety of cyclooxgenase enzyme (COX-3) has been hypothesized (Willoughby et al 2000) which could explain the antipyretic and analgesic actions of paracetamol (Botting 2000) and has recently been described (Chandrasekharan et al 2002; Simmons 2003). The further investigation of this enzyme and its The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
relationship with COX-1 and COX-2 will no doubt prove of great interest over the next few years. As we have witnessed, there have been several significant advances in our eicosanoid knowledge over the last fifteen years, which have led to new understandings and new medicaments. One of these concerns the leap from ‘‘slow reacting substance’’ to leukotrienes described by Holgate and Dahlen (1996). In their contribution ‘‘Leukotrienes in Aspirin-intolerant Asthma’’, Sampson and Holgate (Chapter 21) explain that in this condition there is an overexpression of leukotriene synthase, which may be exacerbated when NSAIDs inhibit COX-1-derived PGE2 synthesis. The most important proinflammatory mediators in asthma are the peptido- or cysteinyl leukotrienes, of which LTD4 plays a key role because of its effects on inflammation, bronchoconstriction and mucus production. However, Zeneca’s (Accolate) zafirlukast is the first of a new class of drug modifier which blocks the effects of LTD4 and the second is Merck Pharmaceuticals’ (Sinulair) monleukast. There now also exists a 5-LO inhibitor (Zyflo) zileuton from Abbott. There have been developments reported in this book in many areas of eicosanoid activity, such as those mentioned earlier (isoprostanes, cannabinoids, receptor knockout mice) plus new viewpoints in biosynthesis and metabolism, refined systems of analysis, new therapeutic antagonists (see above and Chapter 14 ‘‘Eicosanoid Antagonists’’) as well as in the myriad of roles for cytokines. Here (‘‘Cytokines and Eicosanoids in Arthritis’’ Chapter 31) we read of the provocative relationships between antiinflammatory (IL-13) and pro-inflammatory (TNEa and IL-17) cytokines with LOX, COX and prostaglandins and are immediately caught up in a desire to know more about these current mysteries. Finally, there are new insights into the roles of eicosanoids in immunology, endocrinology and metabolism, especially ageing, in inflammation and the circulatory and digestive systems and in the substantial and comprehensive contributions in aspects of the central nervous and reproductive systems. Although this book ends with Section Ten, entitled ‘‘Conclusions and Correlations’’, this is no finale, but an inspiration to seek yet further into the control mechanisms of eicosanoid metabolic pathways, their interdependence and their varied actions on organ systems, with a view to improving understanding and knowledge in order to develop new therapies and greater wisdom about dietary polyunsaturated fatty acids for the benefit of human well-being. REFERENCES Botting RM (2000) Mechanism of action of acetaminophen: is there a cyclooxygenase 3? Clin Infect Dis, 31 (Suppl 5), S202–S210.
618
EPILOGUE
Chadrasekharan NV, Dai H, Roos KL et al (2002) COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure and expression. Proc Natl Acad Sci USA, 99, 13926–13931. Fitzgerald GA and Patrono C (2001) The Coxibs, selective inhibitors of cyclooxygenase-2. New Eng J Med, 345, 433–442. Holgate ST and Dahlen S-E (1996) From slow reacting substance to leukotrienes. A testimony of scientific endeavour and achievements. In Holgate S and Dahlen SE (eds), SRS-A to Leukotrienes: The Dawning of a New Treatment. Oxford: Blackwell Science, 1–23. Ishiwata Y, Okamoto M, Yokochi S et al (2003) Non-steroidal antiinflammatory drug, nabumetone, prevents indomethacin-induced gastric damage via inhibition of neutrophil functions. J Pharm Pharmacol, 55, 229–237.
Masferrer JL, Zweifel BS, Selbert K et al (1990) Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J Clin Invest, 86, 1375–1379. Rainsford KD (2001) The ever-emerging antiinflammatories. Have there been any real advances? J Physiol Paris, 95, 11–29. Roth SH (2000) Rationale for using nabumetone and clinical experience. Drugs, 59, 35–41. Schernhammer ES, Kang JH, Chan AT et al (2004) A prospective study of aspirin use and the risk of pancreatic cancer in women. J Natl Cancer Inst, 96, 22–28. Simmons DL (2003) Variants of cyclooxygenase-1 and their roles in medicine. Thromb Res, 110, 265–268. Willoughby DA, Moore AR, Colville-Nash PR (2000) COX-1, COX-2 and COX-3 and the future treatment of chronic inflammatory disease. Lancet, 255, 646–648.
Index Note: page numbers in italics refer to figures and tables. A-64077 see zileuton A-85761 see ABT-761 AA-2414 (seratrodast; BRONICA) 166, 170 abortion missed 512 spontaneous 510–11 therapeutic 510, 511, 563–6, 613–14 ABT-761 (A-85761) 139, 142–3 Accolate see zafirlukast acetaminophen 605 cyclooxygenase inhibition 604, 605 inhibitory potency 197 mode of action 322–3 acetylcholinesterase (AChE) 99, 103, 104 acute phase proteins (APPs) 330, 331 N-acylethanolamines 179–80, 181–2 see also anandamide N-acyltransferase 180 adhesion receptors, control by PGs 398 adrenergic receptors, cPLA2 activation 22 adrenocorticotropic hormone (ACTH) 327 adult respiratory distress syndrome (ARDS) 136 ageing of central nervous system 311, 449, 457–60 cyclooxygenase systems 303–4, 305, 311, 449 effects on PG function 299–314 cardiovascular system 304–7, 311, 312, 313 central nervous system 311 clinical applications 312–13 diagnostic possibilities 312 dietetic treatment 312 epidermis 311–12 haemato-immunological system 310–11 hepato-gastrointestinal system 309–10, 312 influence on biosynthesis 299–304 influence on metabolism 304 pharmacological treatment 312–13 preventive possibilities 313 pulmonary system 307 renal system 307–9, 312, 313 AH6809 165, 169 AH22921 169 AH23848 165, 169 AL-3138 165, 169 AL-8810 165, 169 alcohol-induced liver disease (AILD) 152–3, 154 allodynia 322, 333, 335, 336, 475 alprostadil 313 ductus arteriosus 614 erectile dysfunction 509, 510, 613 Alzheimer’s disease (AD) 235, 452–3, 487 amyloid vaccines 487 cholinesterase inhibitors 487
The Eicosanoids. Edited by Peter Curtis-Prior. &2004 John Wiley & Sons, Ltd: ISBN 0 471 48984 0
COX-2 inhibitors 311, 323, 453 cyclooxygenase in 153, 311, 323, 453, 458 effects of ageing 311, 458, 459 epigenetic mechanisms 459 isoprostanes 153, 216 NSAIDs 235, 311, 323, 453, 487–9 AMPA receptors, pain 474–5, 476 analgesia, mechanism of 338, 477, 603 analytical methods 97 bioassay 98, 111–15 and eicosanoid hydrophobicity 97 and enzymatic catabolism of eicosanoids 97 enzyme immunoassay 99–100, 103–9, 118–19, 123, 196–8 mass spectrometry 98–9, 117–20 and non-enzymatic hydrolysis 97 and non-enzymatic oxidation 97 radioimmunoassay 99, 118–19, 123 recovery experiments 97 sample extraction 117–18 sample purification 100, 117, 118, 124 time-resolved fluoroimmunoassay 100–1, 123–6 anandamide 179–84 biosynthesis 70, 179–81 blastocyst implantation 553 chemical synthesis 91, 92 degradation 181–4 oxygenation 184 angiogenesis gastric mucosal repair 438 tumours 343, 410 angiotensin and cyclooxygenase 391 renin–angiotensin system 366–9 angiotensin-converting enzyme inhibitors (ACE-I) 272 annexins glucocorticoid action 205 sPLA2 activity 24 antigen-presenting cells (APCs) 237 PGE2 interaction 231, 237–9, 240–1 see also macrophages; monocytes antileukotrienes see leukotriene antagonists; lipoxygenase inhibitors antioxidants COX-1 regulation 36, 37–8 isoprostanes as clinical markers of 154 neurodegenerative disorders 500 AP-1, glucocorticoid receptor effects on 329–30 apoptosis lipoxygenases 411 luteolysis 537 lymphocytes 241 role of COX-2 342, 419
arachidonic acid (AA) Alzheimer’s disease 235 biochemical/molecular pharmacology cyclooxygenase pathway 131, 143–9 lipoxygenase pathway 131–43 non-enzymatic pathway 131, 149–54 biosynthesis of eicosanoids 3–9, 70, 229, 599– 601, 606–7 and ageing 299, 300–3 cerebral ischaemia 482–3 cyclooxygenase 3–5, 268 cyclooxygenase isoforms 45–6, 268 in gut 423 by heart and coronary vessels 393, 394 inhibitors see cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors; lipoxygenase inhibitors; non-steroidal antiinflammatory drugs lipoxygenases 5–7, 23, 53, 56, 57, 86, 87, 606–7 in liver 415 and platelet activating factor 585–90 platelets 373 role of PLA2 3, 22–4, 46, 268 vasoactive in rat cardiovascular system 387–91 see also arachidonic acid (AA), biochemical/molecular pharmacology central nervous system activity in 448 cerebral ischaemia 481, 482–3 formation in 447–8 mental illness 495, 597 metabolism by cytochrome P450s 452 and cyclooxygenase 131, 143–9 cancer 35 cerebral ischaemia 482–3 co-oxidation 33–4, 35 coupling with PLA2 46 cyclooxygenasic function 29, 30–1, 36, 43, 599 in gut 423 natural regulation of 35–7, 46 peroxidasic function 3–4, 29, 31, 32–3, 43, 599 PG biosynthesis 3–5, 268, 387–91, 393, 423, 599–601 inhibition see cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors; non-steroidal antiinflammatory drugs reason for isoforms 45–6, 268 dietary sources 257, 299, 594, 596 gastric mucosal blood flow 433 and inositol phosphoglycerides 261
620 arachidonic acid (AA) (cont.) isoprostanes from 83, 85, 149–54, 211–16 leukotriene biosynthesis 5–7, 23, 53, 56, 57, 86, 87, 606–7 inhibitors see leukotriene antagonists; lipoxygenase inhibitors linoleic acid precursor 229, 230, 257–60, 261, 499 lipoxygenases biochemical/molecular pharmacology 131– 43 cerebral ischaemia 482–3 leukotriene biosynthesis 5–7, 23, 53, 56, 57, 86, 87, 606–7 inhibitors see leukotriene antagonists; lipoxygenase inhibitors synthetic products 87, 88, 90–1 neonates 230, 259–60 niacin patch test 495 nutritional importance 230, 259–60, 262 platelet activating factor 585, 586–9, 590 pregnancy 230, 259 regulation of gene expression 261–2 therapeutic possibilities 594–7 N-arachidonoylethanolamine see anandamide 2-arachidonyl glycerol 4, 91, 92, 179 arginine, cyclooxygenase active sites 30 arthritis 347–56 acetaminophen 605 and cardiovascular disease 606 COX-2 inhibitors 323–4, 604, 606 NSAIDs 604, 606 aspartic acid, sPLA2 activity 19, 20 aspirin xiii arthritis 323, 606 asthma intolerant to 142, 231–2, 247–52 bone 294 cancer 35, 410 cardioprotective action 232–4, 271, 378–80, 401, 606 Celecoxib Long-term Arthritis Safety Study (CLASS) 323 cerebral ischaemia 484, 485 chemical structure 145, 231 clinical efficacy 378 co-administration of 380, 401, 485, 606 and cyclooxygenase 145–6, 189, 601 cancer 35 COX-1/COX-2 selectivity 197, 604 lipoxin biosynthesis 145–6, 189, 192 mechanism of inhibition 193 mode of action 374, 604–5 platelet function effects 375–7, 378, 379–80 regulation of 38, 283 ductus arteriosus 571, 572 endothelial prostacyclin synthesis effects 378 inhibitory potency 197 mode of action 189, 374, 604–6 pharmacokinetics 374–5 platelet function effects 375–80, 605, 606 in pregnancy 574 resistance to 378–80, 606 Reye’s syndrome 232 stroke 484, 485 aspirin-intolerant asthma (AIA) 142, 231–2, 247–52 asthma 136, 607–8 allergic 222 antileukotriene drugs 141, 142–3, 154, 172, 232, 608 aspirin-intolerant 142, 231–2, 247–52 biomarkers for 5-LO in 141–2 atherosclerosis 232 effects of ageing 305, 306
INDEX 15-lipoxygenase 57 oxidized LDL (oxLDL) 270 pharmacology of isoprostanes in 152, 154 platelets 267, 269–70, 271, 306, 381 prostacyclin endothelial 267, 268, 269–72, 305 selective COX-2 inhibitors 401 Raynaud’s syndrome 614 statins 267, 272 atherothrombosis 232–4, 267 aspirin 232–4, 271, 378 COX-2 inhibitors 232, 233–4, 324 and endothelial secretory function 267–73 haemostasis 223, 232–3, 269, 369 B cell stimulating factor see interleukin 6 B lymphocytes apoptosis 241 effects of ageing on PG function 310, 311 PGE2 interaction with 231, 241, 242 Bartter syndrome 231 BAY u3405 (ramatroban; BAYNAS) 166, 170–1 BAY u9773 168, 173 Bayer x1005 140 BAYNAS see BAY u3405 beraprost, elderly patients 313 bioassay 98, 111–15 biological samples bioassay 111, 112, 113–15 enzyme immunoassay 100, 119, 196–8 extraction 117–18 mass spectrometry 119, 120 purification 117, 118, 124 biotin-labelled eicosanoids 101 bladder muscle dysfunction 514 blood pressure, renal function 222, 361–9, 601 blood samples enzyme immunoassay 100, 196–8 mass spectrometry 120 BLT receptors 135, 163, 171, 607 antagonists 141, 155, 164, 168, 171–2 BM-531 166, 171 BM 13,505 (daltroban) 166, 170 BMS 180,291 (ifetroban) 166, 170 bone 289–94 leukotrienes 294 metastatic disease 344 NSAIDs 235, 290, 294 osteoporosis 234, 289, 294, 615 PG production in 290–1 remodelling 222, 234–5, 289–90, 291–4 bradykinin, pain 336, 337, 352 brain see central nervous system; neurological disorders BRONICA (AA-2414; seratrodast) 166, 170 BW A4C 139 BW A868C 163, 165 C/EBPs COX-2 gene 48, 238 sPLA2 25, 26 calcium, cPLA2 activation 22, 587–8 calcium antagonists, elderly patients 313 calcium-binding domain (CalB) cPLA2 activity 17, 18 eicosanoid synthesis 22 calcium binding loop, sPLA2 19, 20 calcium-independent PLA2 (iPLA2) 17, 18 phospholipid metabolism 21–2, 587 prostacyclin synthesis 268 regulation of COX-1 35–6 thromboxane synthesis 268
cancer cyclooxygenase in 35, 47, 154, 221, 323, 341–4 angiogenesis 343, 410 apoptosis 342, 419 colorectal 221, 323, 341, 343, 344, 410 co-oxidation 35 COX-2 inhibitors 47, 154, 323, 341, 342, 343, 344, 440 extracellular matrix 537 free radicals 35 gastric 439–40 hepatic 411, 419 immune suppression 342–3 invasiveness 344 metastasis 344 lipoxygenases in 410, 411–12, 457 NSAIDs 35, 323, 341, 342, 410–11, 440, 537 5-oxoeicosatetraenoic acid 65 cannabinoid receptors, endogenous ligands of see endocannabinoids captopril 140, 313 carbacyclin analogues 82, 84 carboprost, postpartum haemorrhage 513 carcinogenesis see cancer cardiovascular system and diseases xiv, 144, 232–4 ageing effects on PG function 304–7, 312, 313 aspirin 232–4, 271, 378–80 co-administered drugs 380, 401, 606 atherosclerosis see atherosclerosis atherothrombosis see atherothrombosis COX-2 inhibitors 198, 232, 233–4, 324, 380– 1, 605–6 ductus arteriosus 220, 513–14, 569–80, 614 eicosanoid generation 393 cardiac muscle cells 393–4 coronary vascular cells 394–5 and COX-2 inhibitors 401 endothelial cells 394 in rat 387–91 vascular smooth muscle cells 394–5 endothelial secretory function 267–73, 311 essential fatty acids 259 glucocorticoids 328 isoprostanes 152, 154, 306 myocardial infarction see myocardial infarction pathophysiology of 5-LO-derived products 137 physical activity 313 prostanoids 144, 396–401 adhesion receptors 398 and ageing 304–7, 311, 312, 313 antiplatelet drug mechanisms 381, 382 atherothrombosis 232–4, 267, 269–72, 305 cardiac remodelling 400–1 COX-2 inhibitors 232, 233–4, 324, 401, 484, 605–6 ductus arteriosus 220, 513–14, 569, 571, 578–80, 614 generation in 387–91, 393–5, 401 haemostasis 223, 232–4, 269, 311, 369 heart contractility 396 heart rate 396 implications in disease 398–401 myocardial infarction 396, 398–400, 401, 484, 606 myocyte growth 396 and peroxynitrite 398 pharmacological implications 401 Raynaud’s syndrome 614 receptors 395–6, 400 restenosis 398, 400–1
INDEX vascular smooth muscle cells 394–5, 397–8 vasomotor effects 396–7, 398 smoking 313 cascade superfusion bioassay 98, 111, 112, 113– 15 caspases Huntington’s disease 504 luteolysis 537 caveolae, plasma membrane 261 caveolins 261 Celebrex see celecoxib celecoxib xi, 149, 154, 323, 601, 617 Alzheimer’s disease 453, 489 cancer 341, 343, 344 cardiovascular outcomes 324 Celecoxib Long-term Arthritis Safety Study (CLASS) 323–4, 604 chemical structure 149, 602 COX-1/COX-2 selectivity 197, 199, 603–4 development 199 ductus arteriosus 571, 572, 573–4 gastrointestinal outcomes 198, 323–4, 387, 436 inhibitory potency 197, 199 metabolism 606 osteoarthritis 604 pharmacokinetics 606 in pregnancy 574 and renal function 308–9, 324 cell factors, COX-1 regulation 36–7 cell signalling, essential fatty acids 261–2 cellular phospholipase A2 (PLA2) 17–19, 25–6 calcium-independent (iPLA2) 17, 18, 21–2, 35–6, 268, 587 cytosolic (cPLA2) 17, 18 cerebral ischaemia 482 coupling with COX 46 eicosanoid biosynthesis 3, 22, 46, 268, 300, 393, 549 female reproductive system 223, 549 pain 475 platelet activating factor 586, 587–9, 590 regulation of COX-1 35–6 regulation of synthesis of 24 schizophrenia 495 central nervous system 447 ageing of 311, 449, 457–60 Alzheimer’s disease see Alzheimer’s disease arachidonic acid activity in 448 cerebral ischaemia 482–3 formation in 447–8 mental illness 495 metabolism by cytochrome P450s 452 blood–brain barrier 465–6, 481 brain development 230, 259–60, 262 cerebral ischaemia 453, 457, 460, 481–5 cyclooxygenase systems 448–9, 457 ageing 311, 449, 457, 458, 459–60 cerebral ischaemia 481, 482–5 epigenetic regulation 459 fever 147, 221, 449, 465, 466–8, 469 mental illness 493–4 neurodegenerative disease 311, 323, 449, 453, 458, 459–60, 488 neurogenesis 458 pain 321–3, 350–1, 449, 453, 474, 475–7 cytochrome P-450 systems 452 essential fatty acids 230, 259–60, 262, 499– 503 fever 147, 221, 449, 450, 463–71 Huntington’s disease 496, 499–504 lipoxygenase systems 451–2, 457–8, 459–60, 482–3
mental illness 493–6, 596, 597 pain 321–3, 333, 334, 335–6, 350–1, 449, 450, 473–7 prostaglandins 449–51 ageing effects 311 cerebral ischaemia 481, 482–3, 484 fever 147, 221, 450, 466–8, 469, 470, 471 mental illness 493–6 pain 322, 333, 335–6, 450, 474, 475, 476, 477 sleep 222, 451 tardive dyskinesia 500–1 sleep 222, 451 stroke 457, 460, 481, 483, 484–5 thromboxane receptors 451 thromboxanes 450, 451 tumours 457–8 cerebral ischaemia 453, 457, 460, 481–5 chemotactic activity bioassay 115 childbirth see labour cholera 424–5 cholesterol, PGF2a action on 535 cicaprost 82, 84, 313 cicletanine 313 cigarette smoking 313, 431 CIONODS 191, 617 circulatory system eicosanoid generation in 387–91, 393–5, 401 heart and coronary vessels effects of PGs on 396–401, 484 eicosanoid generation in 393–5, 401 PG receptors 395–6 platelets see platelets renal function 222, 361–70, 601 see also cardiovascular system and diseases coenzyme A-independent transacylase (CoAIT) 586–7, 590 colorectal cancer 221, 323, 341, 343, 344, 410 complementary system 466, 471, 488 conjugate additions, PG synthesis 74–9, 78–80 conjugation, eicosanoid metabolism 8, 10–11, 64 co-oxidation, COX–arachidonic acid system 33–4, 35 Corey synthesis, prostaglandins 72–4, 74–6 cornea 61, 62 corpus luteum regression see luteolysis corticosteroids Alzheimer’s disease 488 antiinflammatory 205–6, 327–31, 488 asthma 608 ductus arteriosus 574 see also glucocorticoids corticotrophin releasing factor (CRF) 327 cortisol 327 COX see cyclooxygenase (COX) COX-1 see cyclooxygenase-1 (COX-1) COX-2 see cyclooxygenase-2 (COX-2) COX-189 (lumiracoxib) 148, 149, 154, 201, 604 coxibs, development 191 see also specific coxibs CP-105696 172 CP-195543 172 Crohn’s disease 137, 138, 426 CRTH2 antagonist 164, 171, 222 cyclic AMP (cAMP) arthritis 351 blastocyst implantation 551 gastric secretion 432 intestinal secretion 424–5 PG immunomodulation 238, 240, 241, 242, 519 cyclic AMP binding protein (CREB) COX-2 gene 48, 49
621 Huntington’s disease 501 liver regeneration 418 cyclic AMP response element (CRE) 47, 48, 238, 321 cyclooxygenase (COX; PGH synthase) xix, 29, 43 and ageing 303–4, 305 central nervous system 311, 449, 457, 458, 459–60 angiogenesis cancer 343, 410 gastric mucosal repair 438 angiotensin, responses to 391 bioassay 113–14 biochemical/molecular pharmacology 131, 143–9 biosynthesis of 38, 43–4, 45–6 cancer 35, 47, 154, 221, 323, 341–4, 411 angiogenesis 343, 410 apoptosis 342, 419 colorectal 221, 323, 341, 343, 344, 410 co-oxidation 35 COX-2 inhibitors 47, 154, 323, 341, 342, 343, 344, 440 extracellular matrix 537 free radicals 35 gastric 439–40 hepatic 411, 419 immune suppression 342–3 invasiveness 344 metastasis 344 oesophageal 411 pancreatic 411 cardiovascular system haemostasis 232, 233–4, 369 myocardial infarction 398–9, 401, 484 PG biosynthesis 387–91, 393, 394–5, 401 central nervous system 448–9, 457 and ageing 311, 449, 457, 458, 459–60 cerebral ischaemia 481, 482–5 epigenetic regulation 459 fever 147, 221, 449, 465, 466–8, 469, 470 mental illness 493–4 neurodegenerative disorders 311, 323, 449, 453, 458, 459–60, 488 neurogenesis 458 pain 321–3, 350–1, 449, 453, 474, 475–7 cloning the second gene 44 co-oxidation 33–4, 35 diabetes 234, 283–5 enzyme immunoassay 104–6, 196–8 evidence for second gene 44 fever 147, 221, 449, 465, 466–8, 469, 470 free radicals co-oxidation 34, 35 pharmacological regulation 39 scavenging properties 37–8 suicide inactivation 37–8 gastrointestinal system 407, 423, 601 cytoprotection 408–9, 425, 440 diarrhoeas 425 gastric carcinoma 439–40 H. pylori infection 439–40 inflammation 409–10 NSAID COX-2 selectivity for safety 198 secretory function 408, 425 ulcer healing 438–9 ulcerogenesis 436–8, 440 hepatic system 411, 415, 418, 419 and immune response 238, 343 see also fever above inflammation 146–8, 268, 321, 336 arthritis 349–52, 356 central nervous system 459
622 cyclooxygenase (COX; PGH synthase) (cont.) inflammation (cont.) gastrointestinal system 409–10 isoforms, reason for 45–6, 268 see also cyclooxygenase-1; cyclooxygenase2; cyclooxygenase-3 mechanisms of activity 30–2 cyclooxygenasic 30–1, 32, 33, 38 peroxidasic 30, 31–2, 34, 35 sites for 30 triggering agents 32–3 natural regulation of 35–8, 46 antioxidants 36, 37–8 by cell factors 36–7 ligand stimulation 43–4, 45–6 by plasma factors 36–7 precursor availability 35–6 suicide inactivation 5, 37–8 and NSAIDs 30, 45, 145, 147–8, 189, 321, 601 aspirin see aspirin, and cyclooxygenase aspirin-intolerant asthma 248, 249, 250, 251–2 cancer 35, 323, 341, 342, 410, 440 COX-1/COX-2 inhibitor selectivity 191, 196–8, 601–4 COX-2 selectivity for safety 198 gastrointestinal tract cytoprotection 409, 425, 435 inhibition mechanism 192–8 pain 474, 477 regulation of 38, 283 see also cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors pain 146–7, 321–3, 335, 336, 338, 350–1, 449, 453, 474, 475–7 peroxidation-triggering agents 32–3 availability 36 hydrogen peroxide 33, 36 hydroperoxide ROOH 32–3, 34, 36 nitric oxide 33 peroxynitrite 33, 36, 37 PG biosynthesis 3–5, 13, 23, 45–7, 229, 599– 601 AA cyclooxygenation 29, 30–1, 36, 43, 46, 599 AA peroxidation 3–4, 29, 31, 32–3, 43, 46, 599 and ageing 303–4, 305 in brain 467–8 cancer 323, 341–2 in cardiovascular system of rat 387–91 coupling with PLA2 46 coupling with prostanoid synthases 46–7, 268 and diabetes 283–5 endothelium 268, 270 in gut 423 in heart and coronary vessels 393, 394–5, 401 inhibitors see cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors; non-steroidal antiinflammatory drugs ligand induced 43–4, 45–6 in liver 415, 418 PGH2 reactivity 4–5, 143, 144 platelets 373, 374, 375–7, 378, 380 prostacyclin 4, 46, 47, 232, 268, 270 suicide inactivation of COX 5 see also biochemical/molecular pharmacology above pharmacological regulation of 38, 39, 45, 601
INDEX see also cyclooxygenase (COX; PGH synthase), and NSAIDs reproductive system blastocyst implantation 550–1 ductus arteriosus 569–74, 577 labour 559 luteolysis 533–5 male 518, 521 singlet oxygen generation 34 structure of COX-1 29–30, 32 suicide inactivation 5, 37–8 thromboxane synthesis 4, 29, 46, 47, 144, 268, 605 cyclooxygenase-1 (COX-1; PGH synthase-1) xi, 29, 321 active site of 30 angiotensin, responses to 391 aspirin 145, 189, 604–5 aspirin-intolerant asthma 251–2 and cardiovascular disease 233 inhibition mechanism 193 pharmacology of 376, 379 resistance to 379 biosynthesis of 38 cancer 47, 323, 410, 419 central nervous system 448, 449 Alzheimer’s disease 453 cerebral ischaemia 482, 483, 484–5 COX-2 gene compared 44–5, 599, 601 COX-2 protein compared 45 distribution in cell 30 efficiency of AA use 46 enzyme immunoassay 104–6, 196–8 gastrointestinal system 407, 408–10, 423, 425, 435, 440, 601 ulcer healing 438–9 ulcerogenesis 436–8, 440 and immune response 238 inflammation 268 pain perception 335, 336, 338 inhibition of see cyclooxygenase-1 (COX-1) inhibitors; non-steroidal antiinflammatory drugs liver disease 419 liver regeneration 418 natural regulation of 35–8, 46 by cell factors 36–7 by plasma factors 36–7 precursor availability 35–6 suicide inactivation 37–8 triggering agent availability 36 pain 335, 336, 338, 449, 475–6, 477 PG biosynthesis 3–4, 13, 45–6, 229, 268, 599– 601 and ageing 303–4 in cardiovascular system of rat 387–91 coupling with PLA2 46 coupling with prostanoid synthases 46–7, 268 and diabetes 283, 285 in gut 423 in heart and coronary vessels 393 inhibition see cyclooxygenase-1 (COX-1) inhibitors; non-steroidal antiinflammatory drugs pharmacological regulation of 38, 39, 601 see also cyclooxygenase-1 (COX-1) inhibitors; non-steroidal antiinflammatory drugs purification 30 reason for COX-2 45–6, 268 reproductive system blastocyst implantation 550–1 ductus arteriosus 569, 571, 577
labour 559 male 518, 521 structure 29–30, 32, 387 cyclooxygenase-1 (COX-1; PGH synthase-1) inhibitors xi, 35–9, 45, 145, 189–91, 601 aspirin see aspirin bone remodelling 235 cardiovascular disease 233, 380–1 cerebral ischaemia 484–5 COX-1/COX-2 selectivity 148, 191, 196–8, 601–4 ductus arteriosus 574 NSAID mechanism 192–3, 196–8 pain 476 ulcerogenic effects 436–8, 440 cyclooxygenase-2 (COX-2; PGH synthase-2) xi, 29, 321 and ageing 303–4, 305 central nervous system 311, 449, 457, 458, 459–60 Alzheimer’s disease 311, 323, 453, 458, 488 angiogenesis 343, 410, 438 angiotensin, responses to 391 apoptosis 342, 419 aspirin aspirin-intolerant asthma 251–2 inhibition mechanism 193 pharmacology 145–6, 189, 376, 377, 378, 379, 380, 604–5 resistance 379, 606 biosynthesis of 38 bone PG production in 290–1 remodelling 291–2, 293, 294 cancer 47, 154, 221, 323, 341–4, 411 colorectal 323, 341, 343, 344, 410 gastric 439–40 hepatic 411, 419 oesophageal 411 pancreatic 411 cardiovascular system myocardial infarction 398–9, 401, 484 PG biosynthesis 387–91, 393, 394–5, 401 central nervous system 448–9 ageing 311, 449, 457, 458, 459–60 cerebral ischaemia 481, 482, 483–5 epigenetic regulation 459 fever 449, 465, 466–8, 469, 470 neurodegenerative disorders 311, 323, 449, 453, 458, 459–60, 488 neurogenesis 458 pain 321–3, 350–1, 474, 475–7 cloning cDNA for 44 COX-1 gene compared 44–5, 599, 601 COX-1 protein compared 45 diabetes 234, 283–5 efficiency of AA use 46 enzyme immunoassay 104–6, 196–8 fever 449, 465, 466–8, 469, 470 gastrointestinal system 407, 423, 601 colorectal cancer 323, 341, 343, 344, 410 cytoprotection 409, 435 gastric carcinoma 439–40 H. pylori infection 439–40 inflammation 409–10 inflammatory bowel disease 410 oesophageal cancer 411 pancreatic cancer 411 secretory function 408, 425 ulcer formation 409, 436–8, 440 ulcer healing 438–9 wound healing 409, 410, 438 gene regulation 47–9, 601 cis-acting sites 47–8
INDEX during inflammation 321, 459 signalling pathways 48–9, 394–5 transcription factors 48 glucocorticoid response element 205 and immune response 238 cancer 343 see also fever above inflammation 147–8 arthritis 349–51, 352, 356 cancer 323, 342–3 central nervous system 323, 459 gastrointestinal system 409–10 pain 321–3, 335, 336, 338, 350–1, 474 regulation of expression during 321 therapeutic use of inhibitors 323–4 inhibitors see cyclooxygenase-2 (COX-2; PGH synthase-2) inhibitors liver regeneration 418 message stability 47–9 natural regulation of 38, 46 neuronal 311 pain 321–3, 335, 336, 338, 350–1, 449, 474, 475–7 PG biosynthesis 3–4, 13, 45–6, 268, 599–601 and ageing 303–4, 305 2-arachidonyl glycerol 4 in bone 290–1 in cardiovascular system of rat 387–91 in central nervous system 467–8 coupling with PLA2 46 coupling with prostanoid synthases 46–7, 268 and diabetes 283–5 in gut 423 in heart and coronary vessels 393, 394–5, 401 inhibition see cyclooxygenase-2 (COX-2; PGH synthase-2) inhibitors PGI2 in endothelium 268, 270 and PGE2 synthase 155 pharmacological regulation of 38, 39, 601 see also cyclooxygenase-2 (COX-2; PGH synthase-2) inhibitors reason for COX-1 45–6, 268 reproductive system blastocyst implantation 550–1 ductus arteriosus 569–74, 577 labour 559 luteolysis 533–5 male 518, 521 selectivity 148, 196–8, 601–4 side pocket 148, 601 structure 148, 387 cyclooxygenase-2 (COX-2; PGH synthase-2) inhibitors xi, xix, 38, 39, 45, 189–92, 321– 4, 601, 617 analgesia mechanism 338, 477 anti-ageing 460 arthritis 323–4, 604, 606 biochemical/molecular pharmacology 145–6, 147–9, 154 bone PG production in 290 remodelling 235, 294 cancer 47, 154, 323, 341, 342, 343, 344, 440 cardiovascular system 198, 324, 380–1, 401, 605–6 atherothrombosis 232, 233–4 co-administered drugs 380, 401, 606 myocardial infarction 401, 484, 606 vasoactive prostanoid generation in 388, 389–91 central nervous system 149 ageing effects 311, 460
Alzheimer’s disease 311, 323, 488, 489 cerebral ischaemia 453, 483, 484–5 pain 474, 476 clinical profile 148 COX-1/COX-2 selectivity 148, 191, 196–8, 601–4 COX-2 hypotheses 148, 149, 322 COX-2 side pocket 148, 601 COX-2 structural profile 148, 601 coxib metabolism 604, 606 coxib pharmacokinetics 604, 606 development of specific xi, 148–9, 198–201, 601, 602 ductus arteriosus 569–74 elderly patients Alzheimer’s disease 311, 323 gastrointestinal tract 310, 324 indications 324 renal function 308–9 rheumatoid diseases 312–13, 324 gastrointestinal tract 310, 435 inhibitor development 148, 149, 191 ulcerogenic effects 221, 323–4, 409, 436–8, 440, 603 inhibitory potency 197 NSAID mechanism 192–8 NSAID modification into selective 148, 601– 2 pain 322, 338, 474, 476 pharmacological profile 148 in pregnancy 574 and prostacyclin 232, 233–4, 324, 401, 484, 605–6 renal system 148–9, 198, 308–9, 324, 601, 604 cyclooxygenase-3 (COX-3) 192, 322–3 cyclooxygenase/lipoxygenase (COX/LOX) inhibitors 191, 202–5, 352, 356, 617 cyclopentenone prostaglandins cancer 343 cerebral ischaemia 484 isoprostanes 213 cysteinyl leukotriene antagonists (Cys-LTRAs) 140–1, 154, 155, 164, 172–3, 608 asthma 141, 143, 172, 232, 248, 608 cardiovascular disease 137 chemical structures 168 classification 163 clinical studies 142, 143 gastric mucosal blood flow 433 lung disease 136 see also asthma above cysteinyl leukotriene receptors (Cys-LT receptors) 135–6, 163, 172, 607 antagonists see cysteinyl leukotriene antagonists anti-ageing drug targets 460 cardiovascular diseases 137 inflammatory diseases 137 lung diseases 136 cysteinyl leukotrienes (Cys-LTs) xix action on melanocytes 133, 137 antagonists see cysteinyl leukotriene antagonists aspirin-intolerant asthma 142, 247, 248–52, 250, 252 bioassay 112, 113 biosynthesis 6, 7, 86, 131, 431, 606, 607 in platelets 134, 137, 607 blastocyst implantation 553 bone 294 cardiovascular diseases 137 central nervous system 460, 483 cerebral ischaemia 483 enzyme immunoassay 100, 106–9, 119
623 eye inflammation 138 gastrointestinal system 138 biosynthesis in stomach 431 and COX/LOX inhibitors 202, 203, 204 gastric mucosal blood flow 433 NSAID-induced damage 436 secretory functions 408, 432 immunoaffinity purification 118 inflammatory diseases 137 lung diseases 136, 607 mass spectrometry 120 metabolism 9, 131 pathophysiology 136–7, 607–8 receptors see cysteinyl leukotriene receptors renal diseases 137 skin 137, 138 solid phase extraction 117 stroke 460 cytochrome P-450 systems (CYPs) central nervous system 452 eicosanoid biosynthesis 8–9, 394 eicosanoid metabolism 9–10 luteolysis 535 oxidized derivatives 89–92 seminal plasma 517 cytokines xiv, 347 and bone PG production in 290 remodelling 293 and cyclooxygenase arthritis 349–52, 353 COX-2 gene 47, 48, 49, 321 diabetes 284, 285 eicosanoid synthesis 46, 394 fever 466–7 gastric ulcers 439 inflammation 321, 322, 336, 349–52, 356, 409 message stability 49 pain 322, 336, 350–1, 476–7 promoter sequence 238 regulation 35–6, 43–4, 46 diabetes 277, 284, 285 fever 147, 221, 450, 463–70, 471 functions 347, 348–50 and glucocorticoids 205, 328, 329–30 Huntington’s disease 504 inflammation 347, 348 Alzheimer’s disease 487–8 arthritis 347–56 cyclooxygenase 321, 322, 336, 349–52, 356, 409 hepatic 410 major signal transduction pathways 354–5 pain 322, 336, 337, 350–1, 476–7 liver regeneration 418–19 PG immunomodulation 237–9, 240–1, 242 effects of ageing on 310–11 seminal plasma 518, 519, 520–1, 522 see also cytokines, inflammation, arthritis and PLA2 arthritis 349–51 COX-1 regulation 35–6, 46 in eicosanoid synthesis 22, 24 regulation of synthesis of 24–5, 35–6 secreted 22, 24–5 pleotropic actions 348 properties 347, 348–50 reproductive system blastocyst implantation 549–50 cervical ripening 561 luteolysis 537 cytosolic phospholipase A2 (cPLA2) 17, 18 cerebral ischaemia 482
624 cytosolic phospholipase A2 (cPLA2) (cont.) coupling with COX 46 eicosanoid biosynthesis 3, 22, 46, 268, 300, 393, 549 female reproductive system 223, 549 pain 475 platelet activating factor 586, 587–9, 590 regulation of COX-1 35–6 regulation of synthesis of 24 schizophrenia 495 daltroban (BM 13,505) 166, 170 dementia, Alzheimer’s see Alzheimer’s disease (AD) dendritic cells 237, 238–9, 240 depression 496, 596, 597 dermatitis, atopic 242 developmental disorders Bartter syndrome 231 ductus arteriosus 220, 514, 569–80, 614 essential fatty acids 230, 260 diabetes mellitus 277–85 antidiabetic drug design 615 and COX-2 234, 283–5 cytokines 277, 284, 285 essential fatty acids 258, 259 isoprostanes 153 pathogenesis 234, 277, 278 prostaglandins 277–85 diarrhoeal diseases 423, 424–6 diclofenac Alzheimer’s disease 489 with aspirin 380 chemical structure 145 COX-1/COX-2 selectivity 197, 604 gastrointestinal outcomes 198 inhibitory potency 197 diet elderly people 299–300, 312, 313 essential fatty acids xiii–xiv, 230, 257, 259–60, 262 and ageing 299–300, 312, 313 Huntington’s disease 502–3 therapeutic possibilities 594, 596 food antigen tolerance 242 neonates 230, 259–60 digestive system see gastrointestinal system and diseases; hepatic system and diseases dinoprostone 562, 563, 614 docosahexaenoic acid (DHA) 499 biosynthesis 170, 500 dietary sources 257, 594, 596 isoprostane-like compounds from 215, 216 linolenic acid as precursor of 229, 230, 257– 60 mental illness 496, 596 neonates 230, 259–60 nutritional importance 230, 259–60, 262 pregnancy 230, 259 regulation of gene expression 261–2 therapeutic possibilities 594–7 Zellweger’s syndrome 230, 260 domitroban (S-145) 166, 170 dopaminergic neurotransmission 494, 500 ductus arteriosus foetal 220, 514, 569–76 neonatal 513–14, 576–80, 614 DuP-697 148, 149 eicosapentaenoic acid (EPA) 499 dietary sources 257, 594, 596 Huntington’s disease 496, 503 isoprostane-like compounds from 215 mental illness 496, 596, 597
INDEX therapeutic possibilities 594–7 electrospray ionization 99, 120 enalapril 313 endocannabinoids 179–84 chemical synthesis 92 structures 91 see also anandamide; 2-arachidonyl glycerol endogenous cannabinoids see endocannabinoids endothelial barrier function, gastrointestinal 425 endothelial prostacyclin 268, 269–72, 305 endothelial secretory function and atherothrombosis 267–73 effects of ageing on 311 endothelin-1 (ET-1) ductus arteriosus 574 luteolysis 535 endothelium-derived hyperpolarizing factor (EDHF) 397 enisoprost 614 enprostil 313 enzyme immunoassay (EIA) 99–100, 103–9, 118–19, 123, 196–8 enzyme-linked immunosorbent assay (ELISA) 99–100 eosinophils aspirin-intolerant asthma 249, 250, 251, 252 leukotrienes 65, 133, 134, 607 5-oxoeicosatetraenoic acid 64, 65, 66 epidermal growth factor (EGF) 549–50, 551 epidermis see skin epithelial barrier function, gastrointestinal 408– 9, 425 epithelial cells, 5-oxoeicosatetraenoic acid 65 epithelial transport, gastrointestinal 423–4 epoprostenol 313, 614 epoxide hydrolases 12–13 epoxy-eicosatrienoic acids (EETs) central nervous system 452 chemical synthesis 89–90 eicosanoid biosynthesis 8–9, 10, 394, 599 mass spectrometry 120 as substrate for epoxide hydrolases 12–13 vasomotor effects 397, 398 Epstein–Barr virus 242 erectile dysfunction 509, 613 essential fatty acids 257 see also linoleic acid; linolenic acid estrogen see oestrogen etodolac chemical structure 145 COX-1/COX-2 selectivity 197 inhibitory potency 197 etoricoxib 148, 149, 154, 601 chemical structure 149, 602 COX-1/COX-2 selectivity 604 development 201 metabolism 606 pharmacokinetics 606 safety 201 europium, time-resolved fluoroimmunoassay 123–4 evening primrose oil xiii–xiv, 299–300, 597 extracellular matrix (ECM) cervical ripening 561 luteolysis 535–7 eye development 230, 259, 260 inflammation 138 12-oxoeicosatetraenoic acid 61, 62 fatty acid amide hydrolase (FAAH) anandamide biosynthesis 179–80, 181
anandamide degradation 181–4 catalytic properties 182–3 inhibitors 182–3 molecular properties 183–4 female reproductive system see reproductive system fertility female see reproductive system male see male reproductive system fetal development disorders see foetal development, disorders of fever 147, 221, 449, 450, 463–71 fish oils xiv five lipoxygenase activating protein (FLAP) aspirin-intolerant asthma 249 central nervous system 452 leukotriene biosynthesis 7, 53, 606, 607 inhibitors 139, 154, 608 5-lipoxygenase action mechanism 133 fluoroimmunoassay, time-resolved 100–1, 123– 6 flurbiprofen chemical structure 145 COX-1/COX-2 selectivity 197, 604 inhibitory potency 197 foetal death, missed intrauterine 512 foetal development disorders of Bartter syndrome 231 ductus arteriosus 220, 513–14, 569–80 essential fatty acids 230, 260 essential fatty acids 230, 259, 262 5-oxo-7-glutathionyl-eicosatrienoic acid (FOG7) 8, 11, 64, 120 food antigen tolerance 242 FPL-55712 140 free radicals cardiovascular disease 152 cyclooxygenase co-oxidation 34, 35 pharmacological regulation 39 scavenging properties 37–8 suicide inactivation 37–8 and isoprostanes 152, 211 lung disease 152 see also nitric oxide furosemide 313 G-protein-coupled receptors (GPCRs) see endocannabinoids; leukotriene receptors; prostaglandin receptors; thromboxane A2 receptors G-proteins diabetes 280–1 5-oxoeicosatetraenoic acid coupling to 66 GABAergic transmission 494, 495 gas chromatography–mass spectrometry (GCMS) 98–9, 117, 119–20 gastric carcinoma 439 gastrointestinal system and diseases 407, 423–7 acetaminophen 605 ageing effects on PG function 309–10, 312 cancer 221, 323, 341, 343, 344, 410–12, 439– 40 cholera 424–5 COX-2 inhibitors 310, 435 inhibitor development 148, 149, 191 ulcerogenic effects 221, 323–4, 409, 436–8, 440, 603 COX/LOX inhibitors 202, 203–4 cytoprotection 408–9, 425, 434–5, 440, 614– 15 diarrhoeal diseases 423, 424–6 endothelial barrier function 425
INDEX epithelial barrier function 408–9, 425 epithelial transport 423–4 food antigen tolerance 242 homeostasis 221 inflammation 137, 138, 409–10, 411, 425–7 inflammatory diseases Crohn’s 137, 138, 426 IBD 137, 138, 409–10, 426 ulcerative colitis 426–7 leukotrienes biosynthesis in stomach 431 COX/LOX inhibitors 202, 203–4 gastric mucosal blood flow 433 inflammation 137, 138, 410, 425–7 NSAID-induced damage 436 pepsinogen secretion 432 secretory functions 408, 432 localization of eicosanoid receptors in 407–8 motor activity 425 and NSAIDs see non-steroidal antiinflammatory drugs (NSAIDs), gastrointestinal system secretory functions 408, 424–5, 431–2, 614 stomach 431–40 adenylate cyclase 432 antisecretory mechanisms 432 cancer 439–40 cyclooxygenase 435, 436–40 eicosanoid biosynthesis 431, 440 eicosanoid receptors in 407–8 gastric circulation 432 gastric mucosal blood flow 432–3, 434 gastric secretion 431–2, 614 H. pylori infection 431, 439–40 PG-induced protection 434–5, 614–15 ulcer healing 438–9 ulcerogenic effects of inhibitors 435–8 vascular integrity 408 gemeprost 564, 565, 614 glomerulonephritis 137 glucocorticoids (GCs) 191, 205–6, 327 Alzheimer’s disease 488 antiinflammatory effects 205–6, 327–8, 488 cyclooxygenase 205 COX-1 regulation 36, 38 COX-2 message stability 49 neuronal pathways 458 ductus arteriosus 574 lipoxygenase expression 458 neuronal eicosanoid pathways 458 PG production in bone 290 physiological function 327 and PLA2 36, 38 receptors (GRs) 328 cytoplasmic signalling 331 histone acetylation 331 homodimerization 328–9 and inflammatory cytokines 329–30 transactivation 328, 329 transrepression 329 glucoronidation, leukotriene metabolism 11 glutamate receptors 494–5 NMDA see NMDA receptors glutathione (GSH) and cyclooxygenase activity 33, 36, 599 leukotriene biosynthesis 7 leukotriene conjugation 8, 11, 64 glutathione-S-transferase (GST) see leukotriene C4 (LTC4) synthase glycerophospholipids 3 phospholipase A2 activity see phospholipase A2 glyceryl-prostaglandins 4, 5 gout 201, 350
GR32191 (vapiprost) 166, 170 gynaecology see reproductive system haem, COX activity 30, 31–2, 33, 34 isozyme efficiency 46 suicide inactivation 38 haemostatic system 223, 232–4, 269, 311, 369, 373 heart and heart disease see cardiovascular system and diseases Helicobacter pylori infection 431, 439–40 heparin, COX-1 regulation 36, 37 hepatic system and diseases 415–19 cancer 411, 419 inflammation 410 isoprostanes 152–3, 154, 213 leukotrienes 410, 417 prostaglandins ageing effects on 309, 310 inflammation 410 synthesis of 415, 416, 417, 418 regeneration 415–19 thromboxane synthesis 415 and zileuton 608 hepatocyte stimulating factor see interleukin 6 hepoxilins (Hx) 86, 89, 91 high-performance liquid chromatography (HPLC) 117, 118 histone acetylation 331 HIV infection 242 Huntington’s disease (HD) 496, 499–504 hydrogen peroxide, and COX activity 33, 36 hydroperoxide ROOH, triggering action 32–3, 34, 36 hydroperoxyeicosatetraenoic acids (HPETEs) 86 biosynthesis 6, 7, 8, 9, 53, 54, 55, 56, 131, 606 chemical synthesis 87–8, 90 5-hydroperoxyeicosatetraenoic acid (5HPETE) biosynthesis 6, 7, 53, 54, 606 central nervous system 452 chemical synthesis 90 leukotriene biosynthesis 6, 7, 53, 54, 131 5-lipoxygenase mechanism 133–4, 606 oxoeicosanoid formation 64 synthetic leukotrienes from 88 8-hydroperoxyeicosatetraenoic acid (8-HPETE) 57 12-hydroperoxyeicosatetraenoic acid (12-HPETE) biosynthesis 8, 9, 54, 55, 56, 131 central nervous system 451, 452 diabetes 234 oxoeicosanoid formation 61, 62 15-hydroperoxyeicosatetraenoic acid (15-HPETE) action 57 biosynthesis 9, 54, 55, 56, 131 chemical synthesis 90 hydroxyeicosanoid dehydrogenases 12, 61, 62– 4, 66 hydroxyeicosatetraenoic acids (HETEs) 61, 86 chemical synthesis 87–8, 90 formation of 15-oxoeicosanoids 61 5-hydroxyeicosatetraenoic acid (5-HETE) 7, 61, 131 bone 294 chemical synthesis 90 neutrophils 134 oxoeicosanoids 61, 63, 64, 66, 131 and platelet activating factor 589–90 8-hydroxyeicosatetraenoic acid (8-HETE) 57
625 12-hydroxyeicosatetraenoic acid (12-HETE) 10, 56, 131 activity 62, 66 biosynthetic pathway 134 bone 294 central nervous system 451, 483 cerebral ischaemia 483 colorectal cancer 410, 411 in gastric tissue 431 high-performance liquid chromatography 118 oxoeicosanoid formation 61–2, 131 oxoeicosanoid metabolism 62 platelets 134, 137 psoriasis 137 15-hydroxyeicosatetraenoic acid (15-HETE) 61, 131 action 57 aspirin-intolerant asthma 252 cancer 411 high-performance liquid chromatography 118 lipoxin synthesized from 88–9 20-hydroxyeicosatetraenoic acid (20-HETE) 9, 10, 118 mass spectrometry 120 13-hydroxy-octadecadienoic acid (13-HODE) 61, 62 15-hydroxyprostaglandin dehydrogenases (15hPG dh) 12, 14, 61, 238, 431 hyperalgesia 322, 333, 335, 336, 450, 473–6 hyperprostaglandinaemia 230–1 hypertension pulmonary 574, 614 renal function 222, 361–9, 601 ibuprofen Alzheimer’s disease 488–9 with aspirin 380, 606 cardioprotective action 380, 401, 606 chemical structure 145 COX-1/COX-2 selectivity 197, 604 ductus arteriosus 571, 578 gastrointestinal outcomes 198 inhibitory potency 197 ICI 192,605 166, 171 ICI-204,219 see zafirlukast ifetroban (BMS 180,291) 166, 170 IkB fever 465, 466 and glucocorticoid receptor 329, 330 and PLA2 synthesis 24, 26 iloprost 82, 84 elderly patients 313 Raynaud’s syndrome 614 immune system 231, 237–42 cancer 342–3 effects of ageing on 310–11 fever 463–71 immunomodulation by seminal plasma 517– 22 infection 242 5-LO-derived product pathophysiology 137 mental illness 496 tolerance 242 see also inflammation immunoaffinity chromatography 118 immunoaffinity purification 117, 118 immunoassay see enzyme immunoassay; radioimmunoassay; time-resolved fluoroimmunoassay immunoglobulin production aspirin-intolerant asthma 247, 248 and prostaglandins 241, 242
626 indomethacin Alzheimer’s disease 489 blastocyst implantation 548, 549 chemical structure 145 COX-1/COX-2 selectivity 604 diabetes 279, 283 dopaminergic neurotransmission 494 ductus arteriosus 513, 571, 572, 573, 574, 576, 577–8 gastrointestinal outcomes 204 inhibitory potency 197 mechanism of COX inhibition 193 in pregnancy 574, 576 inflammation xiii, xiv, 321–4, 333 antiinflammatories, non-steroidal see nonsteroidal antiinflammatory drugs antiinflammatory steroids 205–6, 327–31, 488 brain 459 in Alzheimer’s disease 235, 323, 487–8 fever 463–71 cardiovascular disease 234, 267, 270–1 cervical ripening 561 complementary system 466, 471, 488 cyclooxygenase 146–8 cyclooxygenase-1 268, 335 cyclooxygenase-2 147–8 cancer 323, 342–3 neurodegenerative disorders 323, 459 and PGs 147–8, 321, 322, 336 regulation of expression during 321 therapeutic use of inhibitors 323–4 fever 147, 449, 450, 463–71 glucorticoids 327–31 lipoxygenases 459 nociception 322, 333–8 pain secondary to see pain role of PGs 146–8, 223, 231, 335–8 atherosclerosis 267, 270–1, 401 and COX-2 147–8, 321, 322, 336 see also immune system; inflammatory diseases inflammatory bowel disease (IBD) 137, 138, 409–10, 426 inflammatory diseases arthritis 323–4, 347–56, 604, 605, 606 asthma as see asthma atherosclerosis as see atherosclerosis bone loss in 293 COX-2 inhibitors xi, 149, 401, 410 Crohn’s 137, 138, 426 IBD 137, 138, 409–10, 426 leukotrienes in 137, 138, 410, 425–7 neurodegenerative see Alzheimer’s disease psoriasis 137 ulcerative colitis 426–7 inositol phosphoglycerides (IPGs) 261 insulin secretion 277–9, 280–1, 282, 285 insulin-like growth factor-1 (IGF-1) 24–5 interferons (IFNs) arthritis 347 inflammation 348, 349 luteolysis 537 PG immunomodulation 240, 519 pleotropic actions 348 role 348 interleukin 1 (IL-1) xiv arthritis 347, 349, 350–2, 356 blastocyst implantation 550 bone remodelling 293 cyclooxygenase arthritis 349, 350, 351–2 biosynthesis 43–4 COX-2 gene 47, 48 diabetes 284, 285
INDEX eicosanoid synthesis 46 fever 450, 464–7, 469, 470, 471 5-LO inhibitors 356 pleotropic actions 348 properties 349 role 348, 349 see also interleukin 1b (IL-1b) interleukin 1b (IL-1b) arthritis 349, 350–1 fever 450, 464–7, 469, 470 gastric ulcers 439 heart and coronary vessels 394 Huntington’s disease 504 pain 322, 336, 350–1, 477 and sPLA2 24–5 interleukin 2 (IL-2) 240, 348, 349, 356, 519 interleukin 6 (IL-6) arthritis 347, 349, 350, 352, 356 cervical ripening 561 fever 465 glucocorticoid receptor effects 330 liver regeneration 418 pleotropic actions 348 properties 350 role 348, 350 interleukin 8 (IL-8) 348, 356, 561 interleukin 10 (IL-10) 348 arthritis 350, 356 PG immunomodulation 240, 241, 242 seminal plasma 519, 521, 522 interleukin 12 (IL-12) 348 PG immunomodulation 237, 238–9, 241, 242 seminal plasma 519, 521 isoketals (IsoKs) 213–15 isoleukotrienes 90, 92 isoprostanes (iPGs; iPs; IsoPs) 149–55, 211–16 ageing effects 306 alcohol-induced liver disease 152–3, 154 biological properties 150–1, 211–16 biosynthesis 70, 150, 151, 211, 212 cardiovascular dysfunction 152, 154, 306 chemical synthesis 83–4, 85–6 classes of 85 cyclopentenone group 213 diabetes 153 in evaluation of antioxidants 154 iPF2a-III (8-epi-PGF2a; 8-iso-PGF2a; 15-F2t-IsoP) 150 alcohol-induced liver disease 152–3, 154 biological properties 150–1, 211–13 biomarkers of oxidative stress 153 cardiovascular disease 152, 154 diabetes 153 enzyme immunoassay 103–4 in evaluation of antioxidants 154 in lungs 151, 152 nomenclature 150, 151 receptors 153 renal diseases 153, 154 solid-phase extraction 117–18 vascular effects 212–13 iPF2a-VI alcohol-induced liver disease 152–3 cardiovascular disease 154 as clinical marker 153 isoketals 213–15 liver disease 152–3, 154, 213 in lungs 151–2 mass spectrometry 119, 120 neurological disorders 153, 216 nomenclature 150, 152 oxidant injury 211–16 oxidative stress markers 153–4, 211, 269 pain 335
and prostanoid receptors 151–2, 153, 213 receptors 153, 155 renal diseases 153, 154 vascular effects 212–13 JTE-522 149 ketorolac 197, 201 kidney see renal system and diseases kinases AA activation of 261 arthritis 350 COX-2 gene cancer 342 signalling pathways 49, 395 fever 464, 465, 466, 469 in gastrointestinal system 424, 438 in heart and coronary vessels 394, 395, 396, 397–8 labour 561–2 and oxoeicosanoids 63, 64, 65 pain 334, 337, 474, 475, 476 phospholipases in eicosanoid biosynthesis 22, 24, 25 platelet activating factor 587–9, 590 labour 223, 559–63 cervical assessment 562 cervical ripening 560–1, 562 induction 511–12, 514, 562–3 initiation 560 postpartum haemorrhage 513 third stage 512–13 uterine contractility 561–2 uterine rupture 563 leukocytes atherosclerosis 267 cerebral ischaemia 482 cervical ripening 561 see also eosinophils; lymphocytes; mast cells; monocytes; neutrophils leukotriene A4 (LTA4) biosynthesis 6, 7, 53, 54, 57, 86, 87, 131, 606 5-lipoxygenase mechanism 133–4 platelets 134 leukotriene A4 (LTA4) hydrolase 7, 606 inhibitors 140, 313 5-lipoxygenase mechanism 133–4 leukotriene antagonists 140–1, 154, 155, 164, 171–3, 608 asthma 141, 143, 172, 232, 248, 608 cardiovascular disease 137 chemical structures 168 classification 163 clinical studies 142–3 gastrointestinal disease 426, 433 lung disease 136 see also leukotriene antagonists, asthma leukotriene B4 (LTB4) antagonists 141, 155, 164, 168, 171–2 arthritis 351 as biomarker for 5-LO in disease 141, 142 biosynthesis 6, 7, 23, 86, 87, 131, 133, 431, 606–7 blastocyst implantation 553 bone 294 cardiovascular disease 137 central nervous system 460 chemotactic activity bioassay 115 eosinophils 65 eye inflammation 138 gastrointestinal system 138 biosynthesis in stomach 431
INDEX inflammatory bowel disease 426 NSAID-induced damage 436 secretory functions 408, 432 ulcerative colitis 426, 427 high-performance liquid chromatography 118 inflammatory disease 137, 351, 426 metabolism 9–10, 12 12-hydroxyeicosanoid dehydrogenase 12, 62 in keratinocytes 135 oxoeicosanoids 62–3, 131–3 neutrophils 64–5, 134, 135 and oxoeicosanoids 62–3, 64–5, 66, 131–3 pain 335 receptors (BLT) 135, 163, 171, 607 antagonists 141, 155, 164, 168, 171–2 renal disease 137 skin 138 smoking 431 leukotriene C4 (LTC4) action on melanocytes 133, 137 aspirin-intolerant asthma 248–9, 250, 252 bioassay 112, 113 biosynthesis 6, 7, 86, 131, 431, 606, 607 blastocyst implantation 553 bone 294 central nervous system 460, 483 cerebral ischaemia 483 enzyme immunoassay 106–9, 119 gastrointestinal system 138 biosynthesis in stomach 431 gastric mucosal blood flow 433 NSAID-induced damage 436 secretory functions 408, 432 mass spectrometry 120 metabolism 9, 131 pathophysiology in disease 136, 137 platelets 134, 137, 607 receptors 135, 136 antagonists 172–3 skin 138 leukotriene C4 (LTC4) synthase aspirin-intolerant asthma 249–50 leukotriene biosynthesis 7, 131, 134, 607 leukotriene conjugation 8, 11, 64 platelets 134, 607 leukotriene D4 (LTD4) aspirin-intolerant asthma 248 bioassay 112, 113 biosynthesis 6, 7, 131, 606 central nervous system 460, 483 cerebral ischaemia 483 eye inflammation 138 gastric mucosal blood flow 433 gastric secretion 432 mass spectrometry 120 pathophysiology in disease 136, 137 receptors 135, 136 antagonists 140–1, 142, 172–3 skin 138 leukotriene E4 (LTE4) aspirin-intolerant asthma 142, 248, 249, 250, 252 biosynthesis 6, 7, 131, 431, 606 cerebral ischaemia 483 enzyme immunoassay 100 in gastric tissue 431 immunoaffinity purification 118 mass spectrometry 120 pathophysiology in disease 136 receptors 135, 136 antagonists 172–3 solid phase extraction 117
leukotriene receptors 135–6, 137, 163, 171, 172, 607 antagonists see leukotriene antagonists anti-ageing drug targets 460 bone 294 leukotrienes (LTs) xix, 131–43 action mechanisms of 5-LO 133–4 analytical methods 97 bioassay 98, 111, 112, 113, 114–15 enzyme immunoassay 99, 100, 103, 104, 106–9, 119 mass spectrometry 98, 99, 120 sample extraction 117–18 antagonists see leukotriene antagonists asthma 607–8 aspirin-intolerant 142, 247–52 biomarkers for 5-LO in disease 141–2 biosynthesis 3, 5–7, 13, 70, 131, 229, 606–7 inhibitors of see lipoxygenase inhibitors lipoxygenases 5–7, 23, 53–4, 56, 57, 86, 87, 133–4 and platelet activating factor 586 in stomach 431 blastocyst implantation 553 bone 294 cardiovascular diseases 137 central nervous system 460, 481, 483 cerebral ischaemia 481, 483 eosinophils 65, 133, 134, 607 eye 138 gastrointestinal system biosynthesis in stomach 431 COX/LOX inhibitors 202, 203–4 gastric mucosal blood flow 433 inflammation 137, 138, 410, 425–7 NSAID-induced damage 436 pepsinogen secretion 432 secretory functions 408, 432 hepatic system 410 regeneration 417 inhibitors see lipoxygenase inhibitors lung diseases 136, 607 see also leukotrienes (LTs), asthma metabolism 9 conjugation 10–11 by 12-hydroxyeicosanoid dehydrogenase 12, 62 b-oxidation 10, 12 o-oxidation 9–10 oxoeicosanoids 61, 62–3 neutrophils 64–5, 133, 134, 606–7 and oxoeicosanoids 61, 62–3, 64–5, 66 pain 335 pathophysiology in disease 136–7, 607–8 receptors see leukotriene receptors renal diseases 137 skin 137–8 source of 5-LO 133 stroke 460 licofelone 204–5, 352 limaprost 313, 614 linoleic acid (LA) xiii–xiv, 229, 257 and ageing 299, 312 biological functions 260–1 chemical structure 229, 230 dietary sources 257, 312, 594, 596 Huntington’s disease 499, 500–4 metabolism 229–30, 257–8, 261, 499 neonates 230, 259–60 neurodegenerative disorders 499–504 nutritional importance 230, 259–60, 262, 502–3, 593, 596 pregnancy 230, 259 and prostaglandins 260–1
627 therapeutic possibilities 593–7 linolenic acid (ALA) xiii–xiv, 70, 229, 257 and ageing 299–300, 312, 313 biological functions 260–2 chemical structure 229 dietary sources 257, 299–300, 312, 594, 596 isoprostane-like compounds from 215 lipid rafts 261 metabolism 229–30, 257–8, 499 neonates 230, 259–60 neurodegenerative disorders 499–500 Huntington’s disease 499, 500–4 tardive dyskinesia 500, 502–3 nutritional importance 230, 259–60, 262, 502–3, 593, 596 pregnancy 230, 259 and prostaglandins 260–1 regulation of gene expression 261–2 therapeutic possibilities 593–7 g-linolenic acid (GLA) xiii–xiv and ageing 299–300 dietary sources 257, 299–300, 594 dihomo (DGLA) 299–300, 499, 594–7 Huntington’s disease 502–3 tardive dyskinesia 500 therapeutic possibilities 594–7 lipid mediator production, PLA2 role 22–4, 26 lipid rafts 261 lipocortin, COX-1 regulation 36 lipopolysaccharide (LPS) cyclooxygenase inhibitor selectivity 602 immune response to 238 fever 463–4, 465, 466–7, 468, 469, 470–1 and testicular function 521 lipopolysaccharide-binding protein (LBP), fever 463–4, 469 lipoprotein repair, PLA2 22 lipoxins (Lxs) xiv, 86 biosynthesis 56, 70, 133, 194–5 aspirin-activated 145–6, 189, 192, 193 chemical synthesis 88–9 15-epi-lipoxins 145–6, 189, 192, 193 aspirin-intolerant asthma 252 lipoxygenase inhibitors 138–41, 608 arthritis 352, 356 COX/LOX inhibitors 191, 202–5, 352, 356, 617 in gastrointestinal system 138, 426–7 hepatitis 410 ischaemic neuronal injury 460 5-lipoxygenase (5-LO) 139–41, 142–3, 154, 608 anti-ageing 460 arthritis 352, 356 aspirin-intolerant asthma 248 central nervous system 457, 460 COX/LOX inhibitors 202–5, 352, 356 eye inflammation 138 gastrointestinal disease 426–7 hydroxamic acid derivatives 139–40 lung diseases 136 lipoxygenases 53–8, 84, 131–43 anandamide 184 cancer 410, 411–12, 457–8 central nervous system 451–2, 457–8, 459–60, 482–3 cerebral ischaemia 482–3 eye 138 gastrointestinal system 138 inhibitors see lipoxygenase inhibitors leukotriene biosynthesis 5–7, 23, 53–4, 56, 57, 86, 87, 133–4, 606–7 leukotriene receptors 135–6 5-lipoxygenase (5-LO) 53–4 anandamide 184
628 lipoxygenases (cont.) 5-lipoxygenase (5-LO) (cont.) arthritis 351 aspirin-intolerant asthma 248–9, 252 biomarkers for in disease 141–2 bone 294 cancer 411–12, 457–8 cardiovascular diseases 137 central nervous system 452, 457–8, 459–60 epigenetic regulation 459 immune system 137 inflammatory diseases 137, 351, 459 inhibitors see lipoxygenase inhibitors, 5-lipoxygenase (5-LO) leukotriene biosynthesis 5–7, 13, 23, 53–4, 86, 87, 131, 133–4, 606–7 lung diseases 136 mechanism of action 133–4, 606–7 molecular evolution 57 neutrophils 134 pathophysiology in disease 136–7 renal diseases 137 source of 133 8-lipoxygenase (8-LO) 57, 452 12-lipoxygenase (12-LO) 8, 23, 54–6, 86, 131 anandamide 184 biosynthetic pathway 86, 88 cancer 410, 411–12 central nervous system 451, 457 functions 55–6 isozymes 55 leukotriene synthesis in platelets 134 molecular evolution 57–8 oxoeicosanoid formation 61, 62 15-lipoxygenase (15-LO) 8, 23, 56–7, 86, 131 anandamide 184 atherosclerosis 57 biosynthetic pathway 86, 89 cancer 410, 411 central nervous system 452 lipoxin biosynthesis 56, 133 molecular evolution 57–8 oxoeicosanoid formation 61 molecular evolution 57–8 reaction mechanism of soybean 53, 54 skin 137–8 synthetic products 84–9 liquid chromatography (LC) 117 with mass spectrometry (LC-MS) 117, 120 liver see hepatic system and diseases lumiracoxib (COX-189) 148, 149, 154, 201, 604 lungs and lung diseases see pulmonary system luteolysis domestic animal cycles 525–6 PGF2a 223, 510, 525–40 early pregnancy 538–40 identification of role 526–30 mechanism of action 533–8 regulation of secretion of 531–3 uterine synthesis 530–1 primate cycles 525 LY255283 171 LY293111 172 lymphocytes 237 apoptosis 241 autoimmune disease 347, 349 effects of ageing 310, 311 immunosuppression by seminal plasma 519, 520, 521 and PGE 231, 237, 238–41, 242, 519 transendothelial migration 242 macrophages calcium-independent PLA2 17
INDEX immune response 237–8 NSAID inhibitory effects 196 major histocompatibility complex (MHC) 237, 238, 520 male reproductive system androgen dependence 518–19 erectile dysfunction 509, 613 fertility 509, 517–21 and immune system 517–22 seminal plasma 517–22 sexually transmitted disease 521–2 testicular function 521 see also prostate cancer mass spectrometry 98–9, 117–20 electrospray ionization 99, 120 mast cells aspirin-intolerant asthma 250, 251 leukotriene synthesis 607 PG synthesis in inflammation 336 platelet activating factor 588–9 matrix metalloproteinases (MMPs) 241–2 cervical ripening 561 luteolysis 535–7 melanocytes, leukotriene C4 133, 137 meloxicam 308 COX-1/COX-2 selectivity 197, 602, 604 gastrointestinal outcomes 198 inhibitory potency 197 membrane repair, PLA2 22 mental illness 493–6, 596, 597 mifeprostine 510, 511, 512, 563, 564–5 misoprostol 313 Alzheimer’s disease 489 gastric ulcers 614–15 labour induction 512, 514, 562–3 postpartum haemorrhage 513 therapeutic abortion 510, 564–6, 614 third stage labour 513 MK-467 see montelukast MK-0591, ulcerative colitis 426, 427 ML-3000 139 monocyte chemoattractant protein-1 (MCP-1) cervical ripening 561 luteolysis 537 monocytes COX-1/COX-2 inhibitor selectivity 602–3, 604 immune response 237–9, 240, 241–2 immunosuppression by seminal plasma 519 5-oxoeicosatetraenoic acid 63–4, 65 transendothelial migration 242 monooxygenase pathway (cytochrome P450) 89–92 montelukast (MK-467; Singulair) 136, 141, 142, 143, 168, 173, 248, 608 muscarinic signalling, cPLA2 22 myocardial infarction cardiac remodelling 400 COX-2 inhibitors 401, 484, 606 pharmacology of isoprostanes in 152 prostaglandins in 398–400 protective agents 232–3, 234 myosin light chain kinase (MLCK) 561–2 nabumetone 191, 198 NADPH CYP oxidoreductase co-enzyme (NCPO) 452 naproxen Alzheimer’s disease 489 cardioprotection 606 chemical structure 145 COX-1/COX-2 selectivity 197, 603, 604 gastrointestinal outcomes 198, 323, 324, 603, 606
inhibitory potency 197 and renal function 308–9 natural killer (NK) cells 239 immunosuppression by seminal plasma 519 NCX-1015 205 negative-ion chemical ionization (NICI) 119 neonates ductus arteriosus 513–14, 576–80, 614 nutrition 230, 259–60 persistent pulmonary hypertension 574 nervous system see central nervous system neuroleptics, tardive dyskinesia 500 neurological development, essential fatty acids 230, 259–60, 262 neurological disorders 235 ageing effects 311, 457–60 Alzheimer’s disease see Alzheimer’s disease cell proliferation 457–8 COX-2 inhibitors 149, 311, 323, 453 cyclooxygenase in 153, 311, 323, 453, 458, 459–60 epigenetic mechanisms 459 glucocorticoids 458 Huntington’s disease 496, 499–504 ischaemia-induced 453, 457, 460, 481–5 isoprostanes 153, 216 lipoxygenases in 457–8, 459–60 pathobiology of neurodegeneration 459–60 neuroprostanes 70, 83 neutrophils cerebral ischaemia 481 leukotrienes 133, 134 LTB4 activity 64–5, 134, 135 synthesis in 134, 606–7 5-oxoeicosatetraenoic acid 63, 64–5, 66, 134 12-oxoeicosatetraenoic acid 62 platelet activating factor 585, 586–7, 589 transendothelial migration 242 NF-IL-6 arthritis 352 COX-2 gene 47, 48, 238 pancreatic islet 284, 285 NFkB arthritis 349, 350, 352 bone remodelling 291, 292–3 COX-2 gene 48, 238, 283–4 arthritis 350 cancer 342 fever 466, 469 pancreatic islet 284, 285 glucocorticoid receptor effects on 329–30 and immune response 238 fever 464, 465, 466, 469 mechanism of analgesia 338 PLA2 24, 25 receptor activator of (RANKL) 291, 292–3 niacin patch test, schizophrenia 495 nicotine 313, 431 nimesulide 308 Alzheimer’s disease 489 chemical structure 199 COX-1/COX-2 selectivity 197, 602, 604 development 199 gastrointestinal outcomes 198 inhibitory potency 197 vasoactive prostanoid generation 388, 389, 390–1 nitric oxide (NO) atherosclerosis 267, 271–2 cervical ripening 561 and cyclooxygenase activity 33 in ageing 303–4 cerebral ischaemia 481, 482 inflammation 336
INDEX ductus arteriosus 574, 578 and glucocorticoid 328 luteolysis 536 NCX-1015 205 nociception 334, 336 ulcerogenic effects of COX inhibitors 437 vasomotor effects 398 nitric oxide (NO) donating cyclooxygenase (CIONOD) inhibitors 191, 617 nitric oxide (NO) donating NSAIDS 201–2 nitric oxide (NO) donating prednisolone analogue, NCX-1015 205 nitrogylcerin, elderly patients 313 NMDA receptors cerebral ischaemia 481, 483 mental illness 494–5 pain 474, 475, 476 non-steroidal antiinflammatory drugs (NSAIDs) xiii, 145 Alzheimer’s disease 235, 311, 323, 453, 487–9 analgesia mechanism 338, 477, 603 arthritis 323–4, 604, 606 aspirin see aspirin asthma, aspirin-intolerant 247, 248–9, 250, 251, 252 blastocyst implantation 548–9 bone 235, 290, 294 cancer 35, 323, 341, 342, 410–11, 440, 537 cardioprotective action 232–4, 271, 378–81, 401, 606 coronary vessels, PG production in 395 COX/LOX inhibitors 191, 202–5, 352, 356, 617 and cyclooxygenase 30, 45, 145, 147–8, 189, 321, 601 aspirin-intolerant asthma 248, 249, 250 cancer 35, 323, 341, 342, 410, 440 COX-1/COX-2 inhibitor selectivity 191, 196–8, 601–4 gastrointestinal inflammation 409, 425, 435 inhibition mechanism 192–8 pain 474, 477 regulation of 38, 283 see also cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors development of coxibs 148, 191, 601 diabetes 234, 277, 278–9, 280, 284 ductus arteriosus 513, 571–4, 576, 577–8 in elderly patients 308–9, 310, 312, 324 extracellular matrix 537 food antigen tolerance 242 gastrointestinal system 191, 310, 312, 323, 409, 435, 601 barrier functions 425 COX/LOX inhibitors 202–5 gastric carcinoma 440 safety in 198, 323–4, 603, 604 secretory diarrhoeas 425 ulcerogenic effects 435–6 and the immune system 242 inhibitory potency 197 and lipoxygenase, colorectal cancer 410 mental illness 494–5 and neurotransmission 494–5 nitric oxide-donating 201–2 osteoarthritis 604 pain 474, 477 in pregnancy 574, 576 and renal function 308–9, 324, 601, 604 NS-398 148, 149, 437, 438, 439, 440, 572 ductus arteriosus 572 inhibitory potency 197 in pregnancy 574
NSAIDs see non-steroidal antiinflammatory drugs nuclear factor-kb see NFkB nutrition essential fatty acids xiii–xiv, 230, 257, 259–60, 262 and ageing 299–300, 312, 313 Huntington’s disease 502–3 therapeutic possibilities 594, 596 neonates 230, 259–60 obesity, antiobesity drug design 615 obstetrics see reproductive system oesophageal cancer 411 oesophagitis 138 oestradiol, blastocyst implantation 547, 549–50, 551, 553 oestrogen and bone PG production in 290 remodelling 234, 235 and luteolysis 525 abrogation in early pregnancy 540 regulation of uterine PGF2a 530–2 oncogenesis see cancer ONO-1078 see pranlukast ONO-4057 171 ONO-8711 165, 167, 169 ONO-8713 165, 167, 169 ONO-AE2-227 165, 169 ornoprostol 313 osteoarthritis (OA) 349, 350, 604, 605 osteoporosis 234, 289, 294, 615 see also bone, remodelling ovulation 220–1 corpus luteum regression see luteolysis b-oxidation, eicosanoid metabolism 10, 11, 12 o-oxidation, eicosanoid metabolism 9–10, 11 oxidative stress 22 cerebral ischaemia 484 isoprostane markers 153–4, 211, 269 oxidized LDL markers 270 renal function 307–8 oxoeicosanoids 7, 61–6 5-oxoeicosatetraenoic acid (5-oxo-ETE) 62–6 biological activity 64–5 conjugation 8, 11, 64 formation 7, 8, 62–4, 131 inflammatory disease 137 leukotriene synthesis in neutrophils 134 leukotriene synthesis in platelets 134 lung diseases 136 metabolism 64 platelet activating factor 589–90 receptor 65–6 12-oxoeicosanoids (12-oxo-ETE) 61–2, 131, 137 15-oxoeicosatetraenoic acid (15-oxo-ETE) 61 5-oxo-7-glutathionyl-eicosatrienoic acid (FOG7) 8, 11, 64, 120 13-oxo-octadecadienoic acid (13-oxo-ODE) 61 15-oxo-PGE2 12, 14 oxyradical production, NSAID mechanisms 193–4 oxytocin 513 and luteolysis 530–3, 539 pain 333, 473–7 allodynia 322, 333, 335, 336, 475 analgesia mechanism 338, 477, 603 chemical mediators 334–5 cyclooxygenases 146–7, 321–3, 335, 336, 338, 350–1, 449, 453, 474, 475–7 hyperalgesia 322, 333, 335, 336, 450, 473–6
629 isoprostanes 335 nerve sensitization 334, 336–7 nociceptors 334 physiological process 333 prostaglandins 146–7, 223, 321–3, 335–8, 450, 474, 475, 476, 477 stimulus transmission 333–4 types of 333 pancreatic cancer 411 pancreatic islet, PG synthesis 277, 278–9, 280–1, 284–5 paracetamol see acetaminophen parecoxib 201 parturition see labour PD 138387 (COX-2/5-LO) 139, 140 penile erection dysfunction 509, 613 pepsinogen secretion 432 peptido-leukotrienes (p-LTs) see cysteinyl leukotrienes peroxisome proliferator activated receptors (PPARs) 407 Alzheimer’s disease 488 anti-ageing drug targets 460 blastocyst implantation 552 essential fatty acids 262, 597 8-lipoxygenase 57 liver regeneration 419 LTB4 activation of 607 and secreted PLA2 24 therapeutic possibilities 460, 615 peroxisomes, essential fatty acid metabolism 229–30, 260 peroxynitrite cerebral ischaemia 482, 484 luteolysis 536 and prostacyclin 398 triggering action 33, 36, 37 persistent pulmonary hypertension (PPHN) 574 PGH synthase see cyclooxygenase (COX) phosphodiesterase, biosynthesis of anandamide 180–1 phospholipase A2 (PLA2) xiv Alzheimer’s disease 235 arthritis 349–51, 352 calcium-independent see calcium-independent PLA2 (iPLA2) and cell physiology 20–4 eicosanoid synthesis 22–4 lipid mediator production 22–4 lipoprotein repair 22 membrane repair 22 phospholipid metabolism 20–2 cellular see cellular phospholipase A2 (PLA2) cerebral ischaemia 482 classes 17–20 and cyclooxygenase eicosanoid biosynthesis 46 regulation of 35–6 cytosolic see cytosolic phospholipase A2 (cPLA2) dopaminergic neurotransmission 494 eicosanoid biosynthesis 3, 22–4, 46, 268, 600 and ageing 300, 301–2 in heart and coronary vessels 393 in uterus 549 Huntington’s disease 504 inhibitors 26, 191–2, 205–6 in nervous tissue 447–8 pain 474, 475–6 platelet activating factor 586, 587–9, 590 acetylhydrolases 19–20, 587–8 receptors 24 regulation of synthesis of 24–5, 35–6 schizophrenia 495
630 phospholipase A2 (PLA2) (cont.) secreted (sPLA2) see secreted phospholipase A2 (sPLA2) phospholipase C (PLC) 300, 302, 447–8 phospholipase D (PLD) 300, 302, 448 phospholipases cerebral ischaemia 481, 482 eicosanoid biosynthesis 3, 300–3 see also phospholipase A2; phospholipase C (PLC); phospholipase D (PLD) phospholipids 499 eicosanoid biosynthesis 3–4, 70, 600 isoprostane formation from 150, 211 phospholipase A2 see phospholipase A2 phosphorylation cPLA2 22 sPLA2 synthesis 24, 25, 26 physical activity 313 plasma factors, COX-1 regulation 36–7 plasmalogen-selective phospholipase A2 17, 19 platelet activating factor (PAF) xiv, 585–90 acetylhydrolases 19–20, 587–8 platelets 373–82 aggregation mechanisms 373, 374, 377–8 aspirin 374–80, 605, 606 aspirin-intolerant asthma 248 atherosclerosis 267, 269–70, 271, 306, 381 cerebral ischaemia 481, 482, 484 COX-1/COX-2 inhibitor selectivity 602–3, 604 effects of ageing on 306–7 haemostasis 223, 232–3, 269, 369, 373 leukotriene synthesis in 134, 137, 607 non-aspirin cyclooxygenase inhibitors 380–2 5-oxoeicosatetraenoic acid 63, 64 12-oxoeicosatetraenoic acid 62 psoriasis 137 Raynaud’s syndrome 614 schizophrenia 495–6 polymorphonuclear neutrophils (PMNs) leukotrienes 133, 134, 202, 203 platelet activating factor 585, 586–7, 589 pranlukast (ONO-1078) 136, 141, 142, 143, 168, 172, 248, 608 pregnancy abrogation of luteolysis in early 538–40 antiprostaglandin secreting effect of 539–40 blastocyst implantation 547–53 cervical ripening 560–1, 562, 565–6, 613–14 domestic animal cycles 525–6 ectopic 511 essential fatty acids in 230, 259 foetal development disorders see foetal development, disorders of luteoprotective effects 540 missed abortion 512 missed intrauterine foetal death 512 NSAIDs in 574, 576 parturition see labour primate menstrual cycles 525 silent miscarriage 566 spontaneous abortion 510–11 therapeutic abortion 510, 511, 563–6, 613–14 primrose oil xiii–xiv, 299–300, 597 progesterone blastocyst implantation 547, 548 and luteolysis 525–6 abrogation in early pregnancy 538–9, 540 identification of PGF2a 528, 530 PGF2a action mechanism 535–7 PGF2a receptors 533–5 regulation of uterine PGF2a 530–2 propranolol, elderly patients 313
INDEX prostacyclin (prostaglandin I2; PGI2) 69 and ageing 303 cardiovascular system 304, 305, 306, 307 gastrointestinal system 309, 310 pharmacological treatment 312, 313 pulmonary system 307 renal system 307, 309, 312 skin 312 analytical methods 97 bioassay 113, 114 time-resolved fluoroimmunoassay 125, 126 antagonists 164, 169 biosynthesis 4, 29, 46, 47, 144 and ageing 303 aspirin 378 and atherosclerosis 232, 268–9, 270, 305, 401 in heart and coronary vessels 393, 394, 395, 401 in stomach 431 blastocyst implantation 552 cardiovascular system 144 antiplatelet drug mechanisms 381, 382 atherothrombosis 232, 233–4, 267, 269–72, 305 cardiac remodelling 400 COX-2 inhibitors 232, 233–4, 324, 401, 484, 605–6 effects of ageing on 304, 305, 306, 307 generation in 393, 394, 395, 401 haemostasis 223, 269, 369 myocardial ischaemia/infarction 398–400, 401, 484, 606 and peroxynitrite 398 restenosis 400–1 vasomotor effects 396, 397, 398 central nervous system 451 cerebral ischaemia 481, 482, 484 pain 476 chemical synthesis 80–2 derivation 71 endothelial 268, 269–72, 305 gastrointestinal system and ageing 309, 310 biosynthesis in stomach 431 gastric mucosal blood flow 432, 433, 434 vascular integrity 408 hepatic system 410 hydrolytic inactivation 72 inflammation 146, 223 instability 269 pain 146, 223, 335, 337, 476 peripheral vascular disease 614 receptors (IPs) 143–4, 146, 164, 169, 219 blastocyst implantation 552 central nervous system 451, 476, 484 cerebral ischaemia 484 in coronary vessels 395 haemostasis 269 in heart 395 pain 223, 335, 337, 476 renal actions 144 dialysis 614 effects of ageing on 307, 309, 312 stable analogues 81–2, 83, 84, 313, 614 prostacyclin (prostaglandin I2; PGI2) synthase 47 and ageing 303 and atherosclerosis 268, 270 restenosis 400–1 prostaglandin As (PGAs) 70–1 derivation 71, 73 formation from PGE 364–5 hypertension 365–9 prostaglandin B (PGB) 70–1, 73
prostaglandin biosynthesis 3, 23, 69, 70, 229, 599–601 by blastocysts 551 in cardiovascular system 393 cardiac muscle cells 393–4 coronary vascular cells 394–5 and COX-2 inhibitors 401 endothelial cells 394 of rat 387–91 vascular smooth muscle cells 394–5 in central nervous system 467–8 cyclooxygenase 3–5, 13, 23, 45–7, 229, 599– 601 AA cyclooxygenation 29, 30–1, 36, 43, 46, 599 AA peroxidation 3–4, 29, 31, 32–3, 43, 46, 599 and ageing 303 cancer 323, 341–2, 439–40, 599 in cardiovascular system 387–91, 393, 394– 5, 401 in central nervous system 467–8 coupling with PLA2 46 coupling with prostanoid synthases 46–7, 268 diabetes 283–5 endothelium 268, 270 in gut 423 inhibitors see cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors; non-steroidal antiinflammatory drugs ligand stimulation 43–4, 45–6 in liver 415, 418 PGH2 reactivity 4–5, 143, 144 platelets 373 prostacyclin 4, 46, 47, 232, 268, 270 reason for isoforms 45–6, 268 suicide inactivation 5 see also biochemical/molecular pharmacology above in gut 423, 424–5 influence of age on 299–314 in liver 415, 416, 417, 418 pancreatic islet 277, 278–9, 280–1, 284–5 in stomach 431, 440 in uterus 530–1, 549–51 prostaglandin C (PGC) 70–1, 73 prostaglandin D2 (PGD2) 69–70 and ageing 309, 311, 312 allergic asthma 222 antagonists 163–4, 165 biosynthesis, COX 46, 144 cancer 343 cardiovascular actions 144 central nervous system 222, 451 CRTH2 agonism 171, 222 derivation 71 and iso-PGD2 84 pain 335 receptors (DP) 143–4, 163–4, 220 allergic asthma 222 central nervous system 222, 451 chemical structures 165 pain 335, 476 sleep 222, 451 schizophrenia 495 secondary PGs derived from 73 sleep 222, 451 prostaglandin D (PGD) synthase 449, 451 prostaglandin E1 (PGE1) xix and ageing 299, 300, 309 analogues 313, 614 see also alprostadil; misoprostol
INDEX bioassay 111, 112–13 diabetes 278, 279 ductus arteriosus 569, 578–80 and essential fatty acids 261 gastric secretion 431–2, 614 high-performance liquid chromatography 118 myocardial infarction 400 pain 476 Raynaud’s syndrome 614 reproductive system ductus arteriosus 569, 578–80 erectile dysfunction 509, 510, 613 female fertility 510 labour induction 512, 514, 562–3 postpartum haemorrhage 513 seminal plasma 517–18 therapeutic abortion 510 third stage labour 513 tardive dyskinesia 500 vasomotor effects 396 prostaglandin E2 (PGE2) xix, 69–70 and ageing 303 cardiovascular system 304, 305, 306, 307 central nervous system 311 gastrointestinal system 309–10 hepatic system 309 immune system 310–11 low salt diets 313 pharmacological treatment 312, 313 renal system 307, 308 skin 312 analogues 313, 614 antagonists 164–9 aspirin-intolerant asthma 232, 248–9, 251–2 bioassay 111, 112, 113–14 biosynthesis 4, 43–4, 45, 46, 144, 600–1 and ageing 303 in bone 290 in brain 467–8 in central nervous system 467–8 in heart and coronary vessels 393–4 in hepatic system 416 NSAID inhibition of 196 in stomach 431 bone production in 290 remodelling 222, 234, 291–4 cancer 221, 342, 343, 344, 439–40 cardiovascular actions 144, 220 adhesion receptors 398 effects of ageing 304, 305, 306, 307 myocyte growth 396 vasomotor 396 central nervous system 450–1 biosynthesis in 467–8 cerebral ischaemia 483 fever 466–8, 469, 470 mental illness 493–4 pain 450, 475, 476 tardive dyskinesia 500 chemical synthesis 72–4, 74, 75, 82 and COX-2 155, 220, 221 in central nervous system 467–8 effects of ageing 305 inhibitor selectivity 602–3 regulation of expression of 321 derivation 71 developmental disorders Bartter syndrome 231 ductus arteriosus 220, 514, 569, 571, 579 diabetes 278, 279, 280–1, 282, 285 and essential fatty acids 261 fever 466–8, 469, 470
formation of PGA2 from 364–5 gastrointestinal system 221 and ageing 309–10 biosynthesis in stomach 431 cholera 425 COX inhibitors 438 cytoprotection 425, 434, 435 gastric carcinoma 439–40 gastric juice 431 gastric mucosal blood flow 432–3, 434 H. pylori infection 439–40 motor activity 425 secretory functions 408, 424, 425, 431–2, 614 ulcer healing 439 vascular integrity 408 haemostasis 223 hepatic system 410 and ageing 309 regeneration 416, 417, 418 synthesis in 416 high-performance liquid chromatography 118 and immune response 231, 237–42, 310–11, 519 see also prostaglandin E2 (PGE2), fever inflammation 146, 223 arthritis 350, 351–2 regulation of COX-2 expression 321 see also prostaglandin E2 (PGE2), pain and iso-PGE2 84 b-oxidation 10, 11 o-oxidation 9, 11 pain 146–7, 223, 322, 335, 336, 337–8, 450, 475 receptors see prostaglandin E receptors renal system and ageing 307, 308 vascular homeostasis 222, 365, 366–9 reproductive system blastocyst implantation 548–9, 550, 552, 553 cervical ripening 561 ductus arteriosus 220, 514, 569, 571 ectopic pregnancy 511 female fertility 510 labour 559–60, 561, 562 labour induction 512, 514, 562, 563 luteoprotective effects of pregnancy 540 postpartum haemorrhage 513 seminal plasma 517–19, 521, 522 termination of pregnancy 564, 614 uterine rupture 563 secondary PGs derived from 73 time-resolved fluoroimmunoassay 124–5, 126 prostaglandin E2 (PGE2) monoglyceride 5 prostaglandin E receptors (EPs) 143–4, 146–7, 163, 164–9, 220 aspirin-intolerant asthma 249 blastocyst implantation 552, 553 bone remodelling 222, 292–3 central nervous system 450–1 cerebral ischaemia 484 fever 466–8, 469, 470 mental illness 494 pain 450, 476 colorectal cancer 221 diabetes 280–1, 283 ductus arteriosus 220, 571, 614 fertilization 220–1 fever 147, 221, 450, 468, 469, 470, 471 gastrointestinal system 221, 407–8, 424, 614–15 in heart and coronary vessels 396 liver regeneration 418 and lymphocytes 240 myocardial infarction 400
631 osteoporosis 615 ovulation 220–1 pain 223, 335, 336, 450, 476 vascular homeostasis 222 prostaglandin E synthase (PGES) 46, 47, 600 central nervous system 449, 466, 467–8 coupling of COX-1 to 393 and COX-2 155, 467–8 prostaglandin E3 xix prostaglandin endoperoxide G (PGG) 29 prostaglandin endoperoxide G1 (PGG1) 70 prostaglandin endoperoxide G2 (PGG2) 70, 71 and arachidonic acid 29, 31, 32, 43, 46 chemical synthesis 75–6 COX inhibition mechanism 193 prostaglandin endoperoxide G3 (PGG3) 70 prostaglandin endoperoxide H (PGH) 29 prostaglandin endoperoxide H synthase see cyclooxygenase prostaglandin endoperoxide H2 (PGH2) 71 bioassay 113–14 chemical synthesis 75–6, 79 COX 4–5, 43, 45, 46, 143, 144, 193 and PGI2 synthesis 268, 269 reactivity 4–5, 143, 144 stable mimetic analogues of 79–80 prostaglandin F, bone remodelling 234 prostaglandin F1a (PGF1a) and ageing 304, 307, 309, 313 time-resolved fluoroimmunoassay 125, 126 prostaglandin F2 (PGF2) 69–70, 119 prostaglandin F2a (PGF2a) and ageing 304 cardiovascular system 304, 306 central nervous system 311 gastrointestinal system 309–10 hepatic system 309 physical activity 313 renal system 307, 308 skin 312 antagonists 164, 169 bioassay 111, 112, 113 biosynthesis 144, 394, 530–1 cancer 342, 537 cardiovascular system and ageing 304, 306 cell migration and proliferation 397 myocyte growth 396 vasoactive mediation 398 central nervous system 451 and ageing 311 pain 476 chemical synthesis 72–4, 74, 75 derivation 71 diabetes 279 gastrointestinal motor activity 425 pain 335, 476 receptors (FP) 143, 144, 163, 164, 169, 219, 222–3 in corpus luteum 533–5 in heart and coronary vessels 396 osteoporosis 615 therapeutic possibilities 615 reproductive system blastocyst implantation 549, 552 ectopic pregnancy 511 immunomodulation by seminal plasma 518, 522 labour 223, 559–60, 562 luteolysis 223, 510, 525–40 postpartum haemorrhage 513 termination of pregnancy 564, 614 third stage labour 512 time-resolved fluoroimmunoassay 124–5, 126
632 8-epi-prostaglandin F2a (8-epi-PGF2a) see isoprostanes, iPF2a-III 8-iso-prostaglandin F2a (8-iso-PGF2a) see isoprostanes, iPF2a-III prostaglandin H (PGH) synthase see cyclooxygenase (COX) prostaglandin I2 see prostacyclin prostaglandin J (PGJ) 70–1 cerebral ischaemia 484 derivation 71, 73 therapeutic possibilities 615 15-d-prostaglandin J2 (15-d-PGJ2) 231, 615 prostaglandin receptors 82–3 activities 144, 395 allergic asthma 222 antagonists 145–6, 147, 163–71 aspirin-intolerant asthma 249 blastocyst implantation 552, 553 bone remodelling 222, 292–3 central nervous system 222, 450–1 cerebral ischaemia 484 mental illness 494 see also below fever; pain chemical structures 165 classification 143–4, 163, 395 in corpus luteum 533–5 diabetes 280–1, 283, 284, 615 ductus arteriosus 220, 571, 614 expression 144 fertilization 220–1 fever 147, 221, 450, 466–8, 469, 470, 471 future opportunities 615 gastrointestinal system 221, 407–8, 424, 614– 15 haemostasis 269 heart and coronary vessels 395–6, 400 inflammation 146–7, 223 and isoprostane receptors 151–2, 153, 213 liver regeneration 418 and lymphocytes 240 major phenotypes 220 myocardial infarction 400 osteoporosis 615 ovulation 220–1 pain 146–7, 223, 335, 336, 337, 351, 450, 476 properties of mouse 220 receptor-knockout mice 219–23 sleep 222, 451 therapeutic possibilities 615 vascular homeostasis 222 prostaglandins (PGs) xi, xiii–xiv, xix, 69–71, 143–9 1-series xiii–xiv, 69 2-series 69 3-series xiv, 69 and ageing 299–314 clinical applications 312–13 diagnostic possibilities 312 dietetic treatment 312 pharmacological treatment 312–13 preventive possibilities 313 allergic asthma 222 analytical methods 97 bioassay 98, 111–15 enzyme immunoassay 99–100, 103–4, 119 high-performance liquid chromatography 118 mass spectrometry 98, 120 sample extraction 117–18 time-resolved fluoroimmunoassay 100–1, 123–6 antagonists see prostaglandin receptors, antagonists aspirin-intolerant asthma 232, 248–9, 251–2
INDEX biosynthesis see prostaglandin biosynthesis bladder muscle dysfunction 514 and bone production in 290–1 remodelling 222, 234–5, 291–4, 615 cancer 221, 323, 341–4, 410–11, 419, 439–40, 537 cardiovascular system 144, 396–401 adhesion receptors 398 and ageing 304–7, 311, 312, 313 atherothrombosis 232, 233–4, 267, 269–72, 305 cardiac remodelling 400–1 COX-2 inhibitors 401, 484 ductus arteriosus 220, 513–14, 569, 571, 578–80, 614 generation in 387–91, 393–5, 401 haemostasis 223, 232–4, 269, 311, 369 heart contractility 396 heart rate 396 myocardial infarction 398–400, 401, 484, 606 myocardial ischaemia 398–400, 401 myocyte growth 396 and prostaglandin receptors 395–6, 400 Raynaud’s syndrome 614 restenosis 398 vascular smooth muscle cells 394–5, 397–8 vasomotor effects 396–7, 398 central nervous system 448–51 and ageing 311 biosynthesis in 467–8 cerebral ischaemia 481, 482–3, 484 dopaminergic neurotransmission 494 fever 147, 221, 450, 466–8, 469, 470, 471 mental illness 493–6 pain 322, 333, 335–6, 450, 474, 475 tardive dyskinesia 500–1 chemical synthesis 69, 71, 72–83 conjugate addition approach 74–9, 78–80 Corey synthesis 72–4, 74–6 prostacyclin 80–2 stable mimetic analogues 79–80, 81–2 developmental disorders Bartter syndrome 231 ductus arteriosus 220, 513–14, 569, 571, 578–80, 614 diabetes 277–85, 615 and essential fatty acids 260–1 fever 147, 221, 450, 466–8, 469, 470, 471 gastrointestinal system 221 and ageing 309–10, 312 biosynthesis in stomach 431, 440 cancers 410–11, 439–40 cholera 424–5 cytoprotection 408–9, 425, 434–5, 440, 614–15 diarrhoeal diseases 424–5 epithelial barrier function 408–9, 425 gastric carcinoma 439–40 gastric juice 431 gastric mucosal blood flow 432–3, 434 H. pylori infection 439–40 inflammation 409–10, 411, 426 motor activity 425 receptors 407–8 secretory functions 408, 411, 424–5, 431–2, 614 ulcer healing 439 vascular integrity 408 hepatic system and ageing 309, 310 cancers 411, 419 inflammation 410
regeneration 416, 417, 418–19 signal transduction 418–19 synthesis in 415, 416, 417, 418 hyperprostaglandinaemia 230–1 and immune response 231, 237–42 cancer 343 effects of ageing on 310–11 seminal plasma 517–22 see also prostaglandins (PGs), fever inflammation 146–8, 223, 231, 335–8 atherosclerosis 267, 270–1, 401 cancer 343 and COX-2 147–8, 321, 322, 336 gastrointestinal system 409–10, 411, 426 hepatic system 410 inhibitors see cyclooxygenase-1 (COX-1) inhibitors; cyclooxygenase-2 (COX-2) inhibitors; non-steroidal antiinflammatory drugs and isoprostane receptors 151–2, 153, 213 mental illness 493–6 metabolism 9, 302 and ageing 304–14 conjugation 10–11 by 15-hydroxyprostaglandin dehydrogenase 12, 14, 61 b-oxidation 10, 11 o-oxidation 9, 11 and neonates 513–14 pain 146–7, 223, 321–3, 335–8, 450, 474, 475, 476, 477 primary see prostaglandin D2; prostaglandin E2; prostaglandin F2 pulmonary system 222, 307 receptors see prostaglandin receptors renal actions 144 effects of ageing on 307–9, 312, 313 vascular homeostasis 222, 361–9 reproductive system see reproductive system; specific prostaglandins secondary see prostaglandin A; prostaglandin B; prostaglandin C; prostaglandin J skin 311–12 sleep 222, 451 vascular homeostasis 222, 361–9 prostanoids 219 see also prostaglandins; thromboxanes prostate cancer 65, 341, 344 psoriasis 137 psychiatric illness 493–6, 596, 597 pulmonary atresia 578–9 pulmonary hypertension 574, 614 pulmonary system and diseases ageing effects on PG function 307 asthma see asthma isoprostanes in 151–2 pathophysiology of leukotrienes in 136, 607– 8 role of PGD2 222 radioimmunoassay (RIA) 99, 118–19, 123 ramatroban (BAY u3405; BAYNAS) 166, 170–1 Raynaud’s syndrome 614 reactive oxygen species (ROS) 152, 500 renal system and diseases acetaminophen 605 antihypertensive function 222, 361–9, 601 Bartter syndrome 231 COX-2 inhibitors 148–9, 198, 324, 601, 604 dialysis 614 isoprostanes 153, 154 NSAIDs 308–9, 324, 601, 604 pathophysiology of leukotrienes in 137
INDEX prostaglandins 144 effects of ageing on 307–9, 312, 313 vascular homeostasis 222, 361–9 renin–angiotensin system 366–9 reproductive system 509–14 and bone remodelling 234, 235 female fertility 220–1, 510 blastocyst implantation 547–53 luteolysis 223, 510, 525–40 labour 222–3, 559–63 augmentation 511–12 induction 511–12, 514, 562–3 postpartum haemorrhage 513 third stage 512–13 luteolysis 223, 510, 525–40 male see male reproductive system menstrual cycle 525 menstrual symptomatology 514 ovulation 220–1 corpus luteum regression see luteolysis pregnancy abrogation of luteolysis in early 538–40 antiprostaglandin secreting effect of 539–40 blastocyst implantation 547–53 cervical ripening 560–1, 562, 565–6, 613–14 domestic animal cycles 525–6 ectopic 511 essential fatty acids in 230, 259 foetal development disorders see foetal development, disorders of luteoprotective effects 540 missed abortion 512 missed intrauterine foetal death 512 NSAIDs in 574, 576 parturition see reproductive system, labour primate menstrual cycles 525 silent miscarriage 566 spontaneous abortion 510–11 therapeutic abortion 510, 511, 563–6, 613– 14 retinoic acid, ductus arteriosus 574, 576 retinoid X receptors (RTR) 262 Reye’s syndrome 232 rheumatoid diseases COX-2 inhibitors 312–13, 323–4, 604 COX-2 protein 349–50 cytokine production 349, 352 immune response 347, 348 ridogrel 381 rofecoxib 149, 154, 323, 601, 617 Alzheimer’s disease 489 with aspirin 380, 606 cardiovascular disease 232, 324, 380, 606 chemical structure 149, 602 COX-1/COX-2 selectivity 197, 199, 602–3, 604 development 199 ductus arteriosus 571, 572, 573–4 gastrointestinal outcomes 198, 323, 324, 387, 436, 437, 439, 603 inhibitory potency 197, 199 metabolism 606 osteoarthritis 604 pharmacokinetics 606 in pregnancy 574 and renal function 198, 308, 309, 324 Vioxx Gastrointestinal Outcomes Research (VIGOR) 323, 324, 603, 606 S-145 (domitroban) 166, 170 S-5751 163–4, 165 salt, dietary 313 samples bioassay 111, 112, 113–15
enzyme immunoassay 100, 119 extraction 117–18 mass spectrometry 119, 120 purification 100, 117, 118, 124 sandwich enzyme-linked immunosorbent assay (ELISA) 99–100 SC-560 foetal ductus arteriosus 574 ulcerogenic effects 436, 437 vasoactive prostanoid generation 389–90, 391 SC-19220164 165, 166 SC-51089 165, 166 SC-51322 165, 166 schizophrenia 493, 495–6, 596, 597 secreted phospholipase A2 (sPLA2) 19, 20, 25–6 eicosanoid synthesis 22, 24, 46, 268 pain 475 platelet activating factor 587–8 regulation of COX-1 35–6 regulation of synthesis of 24–5, 26 senescence see ageing septic shock 238, 242 seratrodast (AA-2414; BRONICA) 166, 170 serine COX activity 30, 45 aspirin 145, 189, 193 PLA2 activity 17, 18, 22 serum samples, enzyme immunoassay 100 sexually transmitted diseases 521–2 sildenafil 613 Singulair see montelukast skin ageing 311–12 atopic dermatitis 242 pharmacology of leukotrienes in 137–8 blood flow 137–8 epidermal proliferation 138 inflammation 137 vascular permeability 138 psoriasis 137 skin patch test, schizophrenia 495 sleep role of EP receptors 451 role of PGD2 222, 451 slow-reacting substance of anaphylaxis (SRS-A) 111, 112, 114, 131 slow-reacting substances (SRS) 131 smoking 313, 431 elderly people 313 gastric tissue 431 sodium chloride, dietary 313 solid-phase extraction (SPE) 117–18 solid-phase immobilized epitope immunoassay (SPIE-IA) 107–9, 119 spinal cord see central nervous system SQ 29,548 166, 169–70 StAR, luteolysis 535 statins, endothelial action 267, 272 stearidonic acid (SA) 594–7 steroids Alzheimer’s disease 488 antiinflammatory 205–6, 327–31, 488 asthma 608 ductus arteriosus 574 see also glucocorticoids stomach 431–40 adenylate cyclase 432 antisecretory mechanisms 432 cancer 439–40 cyclooxygenase 435, 436–40 eicosanoid biosynthesis 431, 440 eicosanoid receptors in 407–8 gastric circulation 432 gastric mucosal blood flow 432–3, 434
633 gastric secretion 431–2, 614 H. pylori infection 431, 439–40 PG-induced protection 434–5, 614–15 ulcer healing 438–9 ulcerogenic effects of inhibitors 435–8 stroke 457, 460 aspirin 484, 485 blood–brain barrier 481 COX inhibitors 484–5 COX-2 activity 483 leukotrienes after 483 prostanoids after 483 substance P 334–5, 336, 337, 474–5 sulindac chemical structure 145 COX-1/COX-2 selectivity 197 inhibitory potency 197 sulprostone 614 superfusion bioassay 111 superfusion cascade bioassay 98, 111, 112, 113–15 synthetic eicosanoids 69–92 anandamide 92 isoprostanes 83–4 lipoxygenase products 84–9 monooxygenase pathway (cytochrome P450) 89–92 prostaglandins 69, 71, 72–83 thromboxane A2 76, 79–80 T lymphocytes 237 apoptosis 241 autoimmune disease 347, 349 effects of ageing on PG function 310, 311 immunosuppression by seminal plasma 519, 520, 521 and PGE 231, 237, 238–41, 242, 519 transendothelial migration 242 tandem mass spectrometry (MS–MS) 119–20 tardive dyskinesia (TD) 500–1, 502–3 Terbogrel 382 tetrahydrocannabinol 91, 92 thin-layer chromatography (TLC) 117, 118 thrombosis, role of prostanoids 223, 232–4 see also atherothrombosis thromboxane A2 (TXA2) 69 and ageing 304, 305–7, 308, 309, 312, 313 antagonists 146, 147, 164, 166, 169–71, 312, 381–2 bioassay 113–14 biosynthesis 4, 144, 600–1 and aspirin 605, 606 and atherosclerosis 232, 268–9 in cardiovascular system of rat 389, 390, 391 COX-1 activity 29, 46, 47, 605 COX-2 activity 46, 47, 606 platelets 373, 605 cardiovascular actions 144 atherothrombosis 232–4, 267, 269–71 effects of ageing on 304, 305–7, 312 haemostasis 223, 232–3, 269, 369 mitogenic 397–8, 401 myocardial ischaemia and infarction 399–400, 606 restenosis 401 vasomotor effects 396–7 central nervous system 450, 451, 481 cerebral ischaemia 481 chemical synthesis 76, 79, 80 derivation 71 hydrolytic inactivation 72 and immune response 237–8 instability 269
634 thromboxane A2 (TXA2) (cont.) physical activity 313 platelet function 223, 232–3, 269, 369, 373–4 pulmonary system 307 receptor-knockout mice 219, 223 receptors see thromboxane A2 receptors renal system 307, 308, 309 secretory diarrhoeas 425 smoking 313 stable mimetic analogues of 79–80, 81 thromboxane A2 monoglyceride 5 thromboxane A2 receptors (TPs) 82–3, 143, 144, 163 antagonists 146, 147, 164, 166, 169–71, 312 and atherosclerosis 269, 270–1 central nervous system 451 gastrointestinal system 407, 408 gastric mucosal blood flow 433 secretory functions 408 in heart and coronary vessels 395–6 and isoprostanes 151–2, 153, 213, 269 platelets 373–4 receptor-knockout mice 219, 223 thromboxane B2 (TXB2) enzyme immunoassay 119 myocardial infarction 606 pain 476 sample extraction 117, 118 sample purification 118 thromboxane synthase 47, 268 in brain 450 inhibitors 381–2 thromboxanes cerebral ischaemia 481 enzyme immunoassay 103 hepatic synthesis 415 platelet function 373–4 aspirin 375–80, 605 competitive cyclooxygenase inhibitors 380–1 non-aspirin inhibitors 380–2 in stomach 431 thromboxane A2 see thromboxane A2 thromboxane B2 see thromboxane B2
INDEX time-resolved fluoroimmunoassay (TR-FIA) 100–1, 123–6 timnodonic acid 69, 70, 84, 86 tissue inhibitors of matrix metalloproteinases (TIMPs) 535–7, 561 tissue samples bioassay 111, 112, 113–15 enzyme immunoassay 100 a-tocopherol COX-1 regulation 36 isoprostanes as clinical markers 154 triglycerides, COX-1 regulation 36 tryptase, aspirin-intolerant asthma 251 tumorigenesis see cancer tumour necrosis factor a (TNFa) arthritis 347, 349, 350, 351–2, 356 cyclooxygenase arthritis 349, 350, 351–2 cis-acting sites of COX-2 gene 47 fever 467 gastric ulcers 439 pain 477 regulation 35–6 fever 465, 466, 467, 469, 470 inflammatory pain 336, 477 5-LO inhibitors 356 liver regeneration 419 luteolysis 537 and PLA2 24, 35–6, 352 pleotropic actions 348 properties 350 role 348, 350 tyrosine COX active sites 30, 45 COX cyclooxygenasic function 31, 46 COX peroxidasic function 32, 34, 35, 46 sPLA2 activity 19, 20 U44069 80 U46619 79–80, 81 U-57302 171 ubiquitin-proteasome pathway (UPP) 501 UCB 35440 139, 140 ulcerative colitis 426–7
urinary bladder function 514 urine samples enzyme immunoassay 100, 119 mass spectrometry 119, 120 purification 100, 118 solid-phase extraction 117–18 valdecoxib 149, 154, 601 chemical structure 149, 602 COX-1/COX-2 selectivity 604 development 201 metabolism 606 pharmacokinetics 606 safety 201 vapiprost (GR32191) 166, 170 vascular development, essential fatty acids 259 vascular disease, peripheral 614 see also cardiovascular system and diseases vascular homeostasis, renal function 222, 361–9, 601 vascular injury, NSAID-induced 203–4 vascularization gastric mucosal repair 438 tumours 343, 410 Vioxx see rofecoxib viral infections 242 vitamin A 574, 576 vitamin E 36, 154 YM158 168, 173 Z-335 166, 171 zafirlukast (ICI-204,219; Accolate) 136, 608 asthma 141, 143, 172, 608 chemical structure 142, 168 Zellweger’s syndrome 230, 260 zileuton (A-64077) 136, 139 asthma 142, 154, 248, 608 ulcerative colitis 426 ZK138357 164 ZM325802 169